stereoelectronic effects in nucleosides and nucleotides and their
TRANSCRIPT
Stereoelectronic Effects in
Nucleosides and Nucleotides and
their Structural Implications
(including Appendix on literature up to 2005)
Christophe Thibaudeau, Parag Acharya and Jyoti Chattopadhyaya*
*To whom correspondence should be addressed.
E-mail: [email protected]
F +4618554495
T +46184714577
www.boc.uu.se
Department of Bioorganic Chemistry
Biomedical Center, Uppsala University, Sweden
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
2
Christophe Thibaudeau, Parag Acharya and Jyoti Chattopadhyaya*
Dept of Bioorganic Chemistry
Uppsala University
Biomedical Center, Box 581
S-751 23 Uppsala
Dr C. Thibaudeau has completed his Ph.D
in Feb, 1999 at the Dept of Bioorganic Chemistry
under the supervision of Prof J. Chattopadhyaya
Dr P. Acharya has completed his Ph.D
in Dec, 2003 at the Dept of Bioorganic Chemistry under
the supervision of Prof J. Chattopadhyaya
Dr J. Chattopadhyaya is professor of Bioorganic Chemistry
at the Uppsala University
Use the following for citation in your reference:
Uppsala University Press
First Edition: 1999; Second Edition, 2005
Copyright © May, 1999 by J. Chattopadhyaya, C. Thibaudeau and P. Acharaya
ISBN 91-506-1351-0
pp 166
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
3
PREFACE
The three essential components of DNA and RNA are
the aglycone, the pentofuranose sugar in β-D stereochemistry
and the phosphodiester. The phosphates at the backbone
makes DNA/RNA to behave as polyelectrolyte, the pentose
sugar gives the intrinsic flexibility with relatively low energy
barrier for conformational interconversions (according to the
demand of the environment)compared to the hexopyranoses,
and the aglycones help in the self-assembly process throgh the
stacking and hydrogen bonding. Thus, the stability of the
folded structure of nucleic acids is usually attributed to the
interplay of various forces such as hydrogen bonding,
stacking, electrostatics, hydrophobics and hydration.
Are the physico-chemical roles of the aglycone, sugar
and phosphate correlated and interdependent? If so, what are
their thermodynamics? Energetically, how do they respond to
the local change of the environment, and affect the
recognition process that culminate into function?
It has long been qualitatively known that they are not
isolated structural elements. The electronic nature of the
aglycone dictates certain preferred sugar torsions, which are
again correlated with certain phosphate torsions – they are
interdependent. X-ray crystallographic and NMR studies have
demonstrated how the sugar and phosphate moieties can
adopt different conformations in DNA and RNA. It has been
shown that the rotations about C-O and P-O ester bonds are
restricted, and certain sugar-phosphate torsions are preferred
over the others.
Prior to the work summarized here from the Uppsala
group, very little was known about the dynamic
interdependency of the conformational changes between the
three essential components of DNA and RNA or the energetics
involved in this process. Through our solution NMR studies,
we have attempted to show that the electronic nature of the
aglycone as well as those of all other substituents of the sugar
moiety dictate the intrinsic character of the pentose-sugar
conformation, which in turn dictate the phosphate backbone
torsions. The important aspect of this dynamic
interdependency of the aglycone-sugar-phosphate orientation
is that it can be modulated by the change of the environment
with a relatively much smaller energy penalty compared to the
hexopyranose counterparts. This intrinsic flexibile character of
the pentose ring is perhaps the evolutionary basis for their
adoption in nucleic acids for storage of genetic information,
almost error-free transcription, as well as for selective gene
expression in the translation machinery. The present evidences
suggest that the mechanism of this modulation of the pentose
conformation in DNA and RNA is stereolectronic in character,
and the concerted conformational change is unidirectional,
originating from the aglycone to the sugar and further to the
phosphate. In this monograph, we have explored the nature of
the stereoelectronic forces arising from the gauche and
anomeric interactions, that are partly responsible for the self-
organization of nucleosides and nucleotides. Most importantly,
we have experimentally measured the strength and the
interplay of these interactions, for the first time, and shown the
significance of their effects in dictating the overall dynamics
and the structure of nucleos(t)ides and their analogs. This
process has enabled engineering of specific conformations in a
predictable manner in nucleos(t)ides by having appropriate
substituent(s) in the sugar moiety.
The intrinsic flexibility of pentoses in natural
nucleosides and nucleotides, owing to their lower
energy barrier for interconversions compared to the
hexopyranoses, is dictated by the energetics of
stereoelectronic effects, which simply can be tuned
and modulated by choice of substituents and their
ionization state as well as by their complexation with
potential ligands present in the medium.
Stereoelectronic effects operate by appropriate
orbital overlap between the donor and acceptor
through-bond and through-space interactions. The
strength of the stereoelectronic effects induced
stabilization is proportional to the square of the
overlap of the donor and acceptor orbitals, and is
inversely proportional to their energy difference.
Finally, a new section has been added
covering the latest in the field (Appendix: Chapter
10), which makes this monograph current till the end
of 2002.
We believe this monograph is of considerable
value to those medicinal and pharmaceutical
chemists in the academia or in the industry, who
wish to understand fundamental mechanism involved
in the design of structure of modified nucleos(t)ides,
and use this knowledge to rationally develop
antisense, RNAi or triplexing agents or specific
enzyme inhibitors.
We thank Swedish Natural Science Research
Council (Vetenskapsrådet), Swedish Organization
for Strategic Research (SSF) and Uppsala University
for generous financial support in our research
projects
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
4
Table of Contents
page
1. Introduction. Stereoelectronic effects in hexopyranoses 8
1.1 Equatorial monosubstituted cyclohexane 9
1.2 The Edward-Lemieux effect 9
1.3 ∆∆G° versus ∆∆H° estimates for the anomeric effect 10
1.4 The generalized anomeric effect 11
1.5 Influence of the nature and configuration of substituents on the anomeric effect 11
1.6 Effect of the polarity of the solvent on the anomeric effect 13
1.7 The reverse anomeric effect 13
1.8 Hyperconjugation as the origin of the anomeric effect 43
1.9 The nature of the electron lonepairs 16
1.10 Dipole-dipole (electrostatic) interactions as origin of the anomeric effect 17
1.11 Alternative explanations for the origin of the anomeric effect 18
1.12 The gauche effect 19
1.13 The energetics of the gauche effect 20
1.14 Possible origins of the gauche effect 20
2. Stereoelectronic effects in nucleosides and nucleotides 22
2.1 Structure of nucleic acids 22
2.2 The pseudorotation concept 22
2.3 The two-state N � S equilibrium in β-nucleosides 25
2.4 The two-state N � S equilibrium in α-nucleosides 27
2.5 The two-state N S equilibrium in carbocyclic nucleosides 27
2.6 Energy barriers of the pseudorotation cycle of β-D-nucleosides 34
2.7 Steric effect of the nucleobase on the sugar conformation 35
2.8 O4'-C1'-N1/9 stereoelectronic effect (the anomeric effect) 36
2.9 Effect of the aglycone on the conformation of nucleos(t)ides and oligos 38
2.9.1 Configuration-dependent sugar conformation in furanosides 39
2.9.2 Effect of electron-withdrawing nucleobases on pseudorotamer populations 39
2.9.3 Effect of the protonated nucleobase on pseudorotamer populations 40
2.9.4 Effect of base-modifications on the stability of nucleic acids 42
2.10 The gauche effects in α- and β-nucleosides 43
2.10.1 2'-Deoxynucleosides 43
2.10.2 Ribonucleosides and nucleotides 44
2.11 The gauche effects of sugar substituents and the self-organization of DNA/RNA 44
2.11.1 Studies on nucleosides 44
2.11.2 Studies on oligonucleotides 46
3. Methods to quantitate stereoelectronic effects in nucleos(t)ides 46
3.1 Thermodynamics of the two-state N � S equilibria 46
3.2 1H-NMR spectra [temperature, pD, ligand-dependent spectra of nucleos(t)ides] 47
3.3 Pseudorotational analyses of 3JHH with PSEUROT and some practical hints 48
3.3.1 Incorporation of coupling constant errors in PSEUROT calculations 48
3.3.2 Parameters to be fixed or optimized during PSEUROT calculations 48
3.3.3 Electronegativity of the substituents on each HCCH fragment 49
3.3.4 Priority rule to number the substituents on the HAC1C2HB fragment 49
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
5
3.3.5 Translation of HCCH torsion angles into endocyclic torsion angles 50
3.3.6 General operational conditions for PSEUROT 50
3.4 Generalised Karplus-type equation 53
3.4.1 EOS Karplus-Altona equation for 3JHH 53
3.4.2 Reparametrized EOS equation for carbocyclic nucleosides 54
3.4.3 Karplus equation for interpretation of 3JHF 55
3.4.4 Refined Karplus equation for 3JHH based on Fourier formalism 55
3.5 Translation of experimental 3JHH and 3JHF into pseudorotational parameters 56
3.6 Principle of iterations with PSEUROT 66
3.7 Estimation of ∆Hº, ∆Sº and ∆Gº of the N� S equilibrium 61
3.7.1 Methodology 61
3.7.2 Accuracy of thermodynamics 62
3.7.3 Influence of λN1/9 on the thermodynamics 62
3.7.4 Influence of the nature of the aglycone dictates the thermodynamics 66
3.8 New Karplus equation to interpret 3JHF coupling constants 66
3.8.1 Dataset of (3JHF,ΦHF) pairs for monofluoronucleosides 67
3.8.2 Parametrization of the Karplus equation 68
3.8.3 Pseudorotational analyses of 3JHF in fluoronucleosides to validate the Karplus equation 69
4. The quantitation of the stereoelectronic effects in nucleos(t)ides 71
4.1 Quantitation of the anomeric and gauche effects by regression analyses 71
4.1.1 Stereoelectronic effects in neutral β-D-Ns 71
4.1.2 Effect of the 5'CH2OH versus 5'CH2OMe 72
4.1.3 Effect of the nucleobase 73
4.1.4 Gauche effects 73
4.1.5 Energetic equivalence of mirror-image β-D-dNs and β-L-dNs 74
4.1.6 3'-phosphate has stronger gauche effect than 3'-hydroxy 74
4.1.7 Stereoelectronic effects in neutral α-D-N 75
4.1.8 Weakening of the effects of 5'CH2OH and 5'CH2OMe in α-nucleosides 75
4.1.9 Weakening of the effect of the nucleobase in α-nucleosides 75
4.1.10 Weaker 3'-hydroxy gauche effect in α- compared with β-D-Ns 76
4.1.11 Limitations of regression analysis to quantitate stereoelectronic effects 76
4.2 Quantitation of the anomeric and gauche effects by pairwise comparisons 78
4.3 Strengths of ∆H° and -T∆S° to ∆G° of pseudorotation of neutral nucleosides 78
4.4 Nucleobase-dependent anomeric effects in neutral nucleosides 78
4.4.1 Effect of the 5'CH2OH versus 5'CH2OMe 82
4.4.2 The effect of the nucleobase is sugar-dependent 82
4.5 The gauche effect of the 3'-substituent in neutral dNs 85
4.5.1 The influence of the C1'-substituent 85
4.5.2 3'-substituent electronegativity dictates 3'-gauche effect 86
4.5.3 Stronger gauche effect in nucleosides than in 1,2-difluoroethane 90
4.6 The 2'-OH effect in ribonucleos(t)ides is nucleobase-dependent 90
4.7 3'-gauche effect modulation by 2'-OH in ribonucleos(t)ides 92
4.8 Drive of pseudorotation in β-nucleosides by the nature of the nucleobase 93
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
6
4.8.1 The two-state N � S equilibrium is evidenced by pKa values from ∆G° 93
4.8.2 Identical pKas of the nucleobases in 2',3'-dideoxy, 2'-deoxy and ribo series 93
4.8.3 Anomeric effect in β-D-ddNs is modulated by the nature of the nucleobase 95
4.8.4 The orbital mixing as the origin of the O4'-C1'-N1/9 anomeric effect 95
4.8.5 No reverse anomeric effect in pentofuranosyl nucleosides 97
4.8.6 Variable tunablity of anomeric effect in β-D-ddNs, β-D-dNs and β-D-rNs 97
4.8.7 Correlation of the electronic nature of aglycone with pseudorotational state 99
5. Comparison of stereoelectronic effects in α- and β-D-nucleosides 102
5.1 The relative magnitude of the anomeric and gauche effects in Neutral state 102
5.1.1 Anti orientation of the nucleobase in α/β-D-ddN and α/β-D-dN 103
5.1.2 The balance of ∆H° and -T∆S° in neutral α- and β-D-N 104
5.1.3 Weakening of 5'-substituent effect in α- compared with β-nucleosides 104
5.1.4 3'-gauche effect weakens anomeric effect in α-D-dN compared to α-D-ddN 105
5.1.5 Weakening of the 3'-gauche effect in α-D/L-dN compared with β-D/L-dN 105
5.2 The relative magnitude of the anomeric and gauche effects in the ionic states 105
5.2.1 Virtually identical pKa values of the nucleobase in α- and β-nucleosides 105
5.2.2 Predominant enthalpy over entropy in the ionic states of α-D-ddN and -dN 105
5.2.3 Weaker anomeric effect in α-D-ddN gives poorer flexibility 105
5.2.4 The interplay of pD-independent ∆H° and ∆S° in α-D-dN 106
5.2.5 Poor correlation of the nature of aglycone with pseudorotation in α-D-N 106
6. Quantitation of the anomeric effects in C- and N-nucleosides 107
6.1 The anomeric effect in C-nucleosides 107
6.1.1 Effect of the C1'-pyrimidine aglycone on the conformation of the sugar 108
6.1.2 Effect of the C1'-purine aglycone on the drive of the sugar conformation 110
6.1.3 pD-tunable anomeric effect in C-nucleosides 110
6.1.4 Transmission of the nature of the C-aglycone drives the N� S equilibrium in C-nucleosides 111
6.1.5 Estimates for the thermodynamics of the N � S equilibrium 111
6.1.6 Enhanced anomeric effect upon protonation of the aglycone 111
6.1.7 Weaker anomeric effect upon deprotonation of the aglycone 112
6.1.8 Quantitation of the anomeric effect in C-nucleosides 112
6.1.9 Comparison of pD-induced flexibility in C- and N-nucleosides 112
6.1.10 Correlation of the effect of the aglycone and its electronic nature 113
6.2 Quantitation of anomeric effect in N-nucleosides using C-nucleoside as reference 114
7. The interdependency of the sugar and phosphate conformation 118
7.1 Methods to assess the preferred conformation across the phosphate backbone 119
7.2 No correlation of sugar and phosphate conformation in 2'-dN-3'-ethylphosphates 120
7.3 Interaction of 2'-OH with vicinal 3'-phosphate in ribonucleotides 121
8. Application of stereoelectronic effects in oligonucleotides 124
8.1 Design of antisense oligonucleotides via the gauche engineering 124
8.2 Fused nucleosides to engineer preorganized DNA structures 127
8.3 Enzyme recognition by fused carbocyclic nucleosides of fixed conformation 127
8.4 The conformational transmission in the self-cleaving Lariat-RNA 127
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
7
8.5 The conformational transmission by dihydrouridine in RNA 129
8.6 Preferential recognition of 3'-anthraniloyladenosine by Elongation Factor Tu 131
8.7 Conformational changes in nucleotides induced by interaction with metal ions 133
8.8 RNA as molecular wire 136
8.9 The importance of O4' in the self-organization of oligo-DNA 146
9. References
148
10. Appendix (New Additions of references during 2003-4) 160
10.1 The nature of Gauche Effect between 2'-OH and Glycosyl-Nitrogen.
10.2 The pseudorotational barrier of the pentose moiety of nucleosides and nucleotides.
10.3 The hydrogen bonding, hydration and pKa of 2'-OH in adenosine and adenosine 3'-
ethylphosphate.
10.4 Latest articles (1998 – 2002) on the aspects of Anomeric (AE) and Gauche (GE) Effects and
discussions/comments on these works
10.5 The sensitivity of the RNase H discriminates the local structure changes owing to
conformational transmission induced by 3'-endo sugar constrained nucleotides in the
antisense strand of the antisense-RNA hybrid duplex.
10.6 The influence of flouro substitution at the sugar moiety in modulating the furanose
onformation of flourinated nucleosides.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
8
1. Introduction. Stereoelectronic effects in hexopyranoses
Stereoelectronic effects1-7 can be defined as phenomenological effects, which allow to explain
"seemingly abnormal" conformational preferences. Their influence upon the structure of transition
states and the course of chemical reactions is known. The extensive application by Deslongchamps2
and Kirby1 of the principles underlying anomeric hyperconjugation to rationalize and predict reaction
pathways has given rise to the theories of stereoelectronic control, kinetic anomeric effect and to the
antiperiplanar lone-pair hypothesis. The preponderant role of steroelectronic effects in the course of
biochemical reactions has been pointed out for many enzymes, such as ribozymes8-10, serine
proteases11, lysozyme12, or ribonuclease9,10,13. Most of the studies in the field of stereoelectronic
effects have been devoted to six-membered rings, and the state-of-the-art for such systems has been
the subject of many review articles1-7. On the other hand, to the best of our knowledge, no unifying
model regarding the influence of stereoelectronic effects on the conformation of heterocyclic five-
membered rings at the monomeric or oligomeric level has yet come out in the literature. This is
possibly due to the fact that, in contrast with rather unflexible and consequently conformationally
biased six-membered rings, saturated five-membered rings, such as the pentofuranose moiety in
nucleosides and nucleotides, are generally involved in complex conformational equilibria.
Some early experimental studies have shown in a qualitative manner how the anomeric and
gauche effects control the bias of conformational equilibria of the pentofuranose moiety in
nucleos(t)ides and of other saturated five-membered rings. A significant progress has however only
been accomplished during the last decade with the development of new methodologies in our
laboratory which give accurate estimates of the magnitude of energetics of stereoelectronic effects
driving the two-state North �South pseudorotational equilibrium in nucleos(t)ides14-45.
At the oligomeric level, recent investigations46-54 on modified oligonucleotides have been
conducted as a part of the pursuit for developing antisense compounds with increased nuclease-
resistance compared with the parent DNA and RNA, and with improved hybridization with the target
RNA. The analysis of the thermodynamic properties and of the three-dimensional structure55-61 of
these oligonucleotides has shown that a particular modification of one of the three components of the
constituent nucleotides, i.e. the nucleobase, the sugar moiety and the phosphate backbone alters both
their stability and overall structure. The actual participation of stereoelectronic effects to the observed
structural changes has been adressed only qualitatively46,49,62-64.
The potential power of the application of the gauche / anomeric engineering strategy to design
new oligonucleotides endowed with therapeutic properties is evident. A detailed appraisal of the
qualitative and quantitative results from the early as well as from state-of-the-art studies on the
manifestation, magnitude and origin of stereoelectronic effects in the pentofuranose ring in
nucleosides and nucleotides is therefore required in order to understand its role in the self
organization of DNA, RNA and their analogues. This is the purpose of the present monograph, which
is organized as follows: (i) We first introduce the concept of stereoelectronic effects by summarizing
the most important findings of conformational studies on pyranose derivatives (Section 1). (ii) The
main results of qualitative conformational studies on nucleosides and nucleotides showing the
influence of the nature of the substituents at C1'-C4' on the conformation of the constituent
pentofuranose sugar are subsequently presented and analyzed in detail (Section 2). (iii) Sections 3 - 7
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
9
are devoted to the quantitation of the thermodynamics of stereoelectronic effects driving
pseudorotational equilibria in nucleos(t)ides using the methodology (described in Section 3)
developed by us in Uppsala. In Sections 3 - 7, we show how specific conformational preferences of
nucleos(t)ides can be engineered by altering either the nature or the configuration of one of their
constituents, i.e. heterocyclic nucleobase, pentofuranose moiety or C2' and C3' substituents, or the pH
of the medium. Evidences regarding the transmission of information between the basic components
within the nucleotidyl unit are presented in Sections 4 - 7. (iv) We finally present recent works, both
from our laboratory and others, on the impact of the anomeric and gauche effects on the
threedimensional structure and biological function of oligonucleotides (Section 8).
Stereoelectronic effects can be classified into a few categories: (i) The original Edward-
Lemieux effect65-67 which designates the preference of the electronegative methoxy or acetoxy group
at C2 in 2-substituted tetrahydropyran for the axial over the equatorial orientation. (ii) The generalized
anomeric effect (GAE), which is an extension of the original Edward-Lemieux effect in the case of
any acyclic or cyclic R-X-A-Y system. (iii) The reverse anomeric effect allows to rationalize a
conformational preference opposite to that expected on the basis of the (generalized) anomeric effect,
i.e. an increased tendency of the anomeric substituent in R-X-A-Y fragment to assume an equatorial
orientation in comparison with its steric effect. (iv) The gauche effect refers to a stereoelectronic
preference for conformations in which the best donor lonepair or bond is antiperiplanar to the best
acceptor bond.
1.1 Equatorial monosubstituted cyclohexane
Except in rare cases (such as bulky V-shaped alkyl groups68), the most stable conformation of
monosubstituted cyclohexane (1) is the chair form E1 in which the substituent X takes up an
equatorial orientation (Fig 1A). This is owing to the fact that in the axial chair conformation A1, X
exerts energetically unfavourable steric repulsions with the axially oriented H3 and H5. The tendency
of X in cyclohexane to adopt an equatorial orientation is dictated by its effective size and can be
estimated from the Gibbs free-energy (i.e. ∆G°steric(cylcohex.)) of the A1 � E1 equilibrium (Eq 1):
∆G°steric(cyclohex.) = -RT ln [E1]/[A1] ..... Eq 1.
In a study69 based on temperature-dependent 13C-NMR spectra of CFCl3-CDCl3 solutions, it
has been shown that ∆G°steric(cyclohex.) at 300 K is ≈ -7.3 kJmol-1 (X = Me), -7.5 kJmol-1 (X = Et)
and -9.3 kJmol-1 (X = iPr).
1.2 The Edward-Lemieux effect
The acid hydrolysis of methyl β-pyranosides of glucose, mannose and galactose, in which the
anomeric methoxy group is equatorial, is faster than that of the α-counterparts, in which it is axial,
which has led Edward65 to propose that the axial orientation of the anomeric substituent in
pyranosides is more stable than the equatorial one. This is opposite to what one would predict on the
basis of simple steric effect rule (vide supra). Edward has attributed this unusual conformational
behaviour to the destabilization of the equatorial conformer by electrostatic repulsions between the
anomeric substituent and the ring dipole induced by the presence of the endocyclic oxygen atom. In
his studies on various acetylated α- versus β-aldohexopyranoses, Lemieux66,67,70 also found that the
acetoxy substituent at C2 adopts preferentially an axial orientation. He introduced the term "anomeric
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
10
effect" to designate this preference. An extension of the original65-67,70 Edward-Lemieux effect to the
analysis of the conformational equilibrium of 2-methoxytetrahydropyran (2) allows to predict
correctly71 that the axial chair A2 is more stable than its equatorial counterpart E2 (Fig 1B). The Gibbs
free-energy (∆G°heterocycle) of the A2 � E2 conformational equilibrium in 2 is calculated according to
Eq 2: ∆G°heterocycle = -RT ln [E2]/[A2] ..... Eq 2
Figure 1. Axial � Equatorial anomeric conformational
equilibrium in six-membered rings. (A) The drive of the A1 �
E1 equibrium in monosubstituted cyclohexane (1) toward the
equatorial conformer E1 minimizes steric repulsions of the
anomeric group with H3 and H5. (B) The O-C-Ome anomeric
effect in 2-methoxytetrahydropyran (2) pushes its conformational
equilibrium toward the A2 form in spite of unfavourable steric
repulsions, which are reduced in the E2 chair.
The original definition of the anomeric effect
therefore implies that the stereoelectronic component of the O-C-O effet prevails over the
counteracting steric effect. The Gibbs free-energy of the anomeric effect (∆∆G°AE, kJmol-1) in 2 is
estimated using Eq 3:
∆∆G°AE = ∆G°heterocycle - F * ∆G°steric(cyclohex.) - 0.08 ..... Eq 3
where F (= 1.53) stands for Franck's correction factor72 and accounts for the fact that in 2 steric
repulsions generated by the axial orientations of X, H3 and H5 are greater than in the parent
monosubstituted cyclohexane (1), owing to the shorter C-O bond in 2 compared with C-C bond in (1).
1.3 ∆∆G° versus ∆∆H° estimates for the anomeric effect
The question of which thermodynamic quantity [i.e. either enthalpy (∆∆H°), or Gibbs free-
energy (∆∆G°), Table 1] is adequate to estimate the magnitude of the anomeric effect that drives the
conformational equilibrium of 2-substituted tetrahydropyran has been adressed in many studies71,75,91,
92. An initial work by Booth et al 91 based on temperature-dependent 13C-NMR spectra showed that -
T∆S° overrides the negligible ∆H° in the drive of the conformational equilibrium of 2-
methoxytetrahydropyran (2) in CDCl3-CFCl3. It was therefore suggested that the stabilization of A2
over E2 form of 2 was the result of the entropy term only. Lemieux71 however pointed out that such a
negligible ∆H° value is only obtained when 2-methoxytetrahydropyran is solvated by hydrogen-
bonding or polar solvents, and he showed that in CCl4 the origin of the anomeric effect is mainly
enthalpic. The relative preference69 for equatorial over axial chair forms of monosubstituted
cyclohexane (1) carrying either Me, Et or iPr group is strongly temperature-dependent, as the result of
entropy effects. It is therefore preferable to estimate the magnitude of the anomeric effect in terms of
∆∆H°ΑΕ (Eq 4) rather than ∆∆G°AE (Eq 3) in order to eliminate any misleading entropy-based
contribution92:m∆∆H°AE = ∆H°heterocycle - ∆H°steric(cyclohex.) ..... Eq 4
where ∆H°steric(cyclohex.) and ∆H°heterocycle represent the ∆H° contributions to ∆G° of the
conformational equilibria for substituted cyclohexane (Fig 1A) and tetrahydropyran (Fig 1B),
respectively. Juaristi et al 86 have recently estimated the strength of the enthalpic S-C-P(O) anomeric
effect in 2-(diphenylphosphinoyl)-1,3-dithiane by replacing ∆H°steric(cyclohex.) in Eq 4 by
F*∆H°steric(cyclohex.) in order to take into consideration Franck's argument72 regarding different
H X
H
OMe
O
H
H X
O
OMe
(A)
(B)
1: A1 chair 1: E1 chair
2: A2 chair 2: E2 chair
ΔGo
steric (cyclohex.) = -AX
ΔGo
heterocycle
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
11
effective steric sizes of the anomeric substituents in substituted cyclohexane and tetrahydropyran
(Table 1).
Enthalpic anomeric stabilizations have been found experimentally for substituted heterocyclic
six-membered rings, such as 2-substituted tetrahydropyrans, 2-carbomethoxy- and 2-cyano-
piperidines87 and they have been extensively reviewed4,6 (Table 1).
1.4 The generalized anomeric effect
The concept of generalized anomeric effect93-96 results from an extension of the original
Edward-Lemieux effect to describe the preference for gauche over trans orientation of R-X with
respect to A-Y in R-X-A-Y fragments. R-X-A-Y may be either acyclic or constitute a part of a ring
other than the original substituted tetrahydropyran. X designates an element with at least one pair of
non-bonding electrons, R and A have intermediate electronegativities and Y is more electronegative
than A. Typically1-6, R stands for C or H, A is a tetrahedral (anomeric) center such as C or Si, X is
either O, N or S whereas Y designates F, Cl, Br, O, N or P. Further manifestations and extensions of
the generalized anomeric effect include the “benzylic anomeric effect” 97 which allows to explain the
preferred perpendicular orientation of C-X bond with respect to the plane of the benzene ring (X =
S(O)Me, SO2Me, Cl, etc) in ArCH2X compounds. The preference of chlorine and methoxy in α-
chloro and α-methoxycyclohexanone oximes for an axial orientation has been attributed to the
“vinylogous anomeric effect” 98, whereas the “syn anomeric effect” is responsible for the relatively
stable synperiplanar orientation of the nitrogen lonepair with respect to the C-F bond in
fluoromethylamine, as shown by ab initio calculations99. When R-X-A-Y is part of a heterocycle and
both X and Y possess lonepairs of electrons, the conformation of the heterocycle is driven by the fine
balance between the endo and exo anomeric effects3, which operate in opposite directions: The endo
anomeric effect involves one of the lonepairs of the endocyclic atom X, which tends to be
antiperiplanar with respect to the exocyclic A-Y bond, whereas for the exo anomeric effect it is one of
the electron pairs of the exocyclic Y atom which adopts an anti orientation with respect to the
endocyclic R-X bond. Depending on the relative magnitudes of endo and exo anomeric effects, the
overall estimate (i.e. endo + exo ) for the magnitude of the anomeric effect may be either positive or
negative3,4,6.
1.5 Influence of the nature and configuration of substituents on the anomeric effect
The magnitude of the anomeric effect depends on the nature of the anomeric group, or other
substituents and on their relative configurations. It is also modulated by the polarity and nature of the
solvent.
Conformational studies on acetylated and benzoylated β-D-ribo and xylopyranose derivatives
have shown that the magnitude of the anomeric effect is regularly increased as the electron-
withdrawing character of the anomeric substituent increases, H100,101 < OMe102 < OAc103 < OBz104 <
Halogen101. Similarly, other experimental studies have shown that the preference of the anomeric
substituent X for axial orientation in 2-substituted tetrahydropyran decreases steadily when X is
changed: Br73,74, Cl73,74 > OMe76 > OH80 > (CH2)2N81 > MeHN81. The relative
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
12
Table 1. The magnitude of the anomeric effect in heterocyclic 6-membered ringsa
Axial Equatorial
Z
Y
Y
R
R'
XZ
R'
R
X
ΔGo
heterocycle
R R' Y Z X ∆∆G°AE1 (ref.) b,c ∆∆G°AE2 b,d ∆∆H°AE (ref.) e
H H O CH2 Br f ≈ 9.6 (73) / 13.4 (74) 10.8 l -
H H O CH2 Cl f ≈ 9.6 (73) / 11.3 (74) 10.8 l 8.9 m (75)
H H O CH2 MeO 8.0 (76) 10.3 3.1 m (75)
H H O CH2 EtO 7.5 (76) 9.8 -
H H O CH2 Me3CO 5.8 (76) 8.1 -
H H O CH2 PhO 6.7 (77) 8.3 -
H H O CH2 AcO 5.8 (78) 7.6 -
H H O CH2 MeS 6.2 (79) 8.5 -
H H O CH2 HO 3.9 (80) 6.2 2.5 m (75)
H H O CH2 (CH2)2N 4.1 (81) 7.6 -
H H O CH2 MeHN 1.5 (81) 4.4 ≈ 0 m (75)
H H O CH2 MeO2C -0.5 (82) 2.4 -
Me H O CH2 HO 5.0 (78) 7.3 -
Me H O CH2 MeO 7.7 (83) 10.0 -
Me H O CH2 EtO 7.2 (83) 9.5 -
Me H O CH2 AcO g 5.7 (84) 7.6 -
Me H O CH2 MeO2C h 0.1 (85) 3.0 -
Me H O CH2 Cl f 11.3 (74) 12.3 -
Me H O CH2 Br f 13.4 (74) 14.6 -
Me H O CH2 I 13.0 (74) 14.0 -
H Me O CH2 MeO 7.2 (83) 9.5 -
H Me O CH2 EtO 7.0 (83) 9.2 -
H Me O CH2 AcO g 6.0 (84) 7.8 -
H Me O CH2 MeS 5.6 (83) 7.9 -
H Me O CH2 Me3CS 5.8 (83) 8.1 -
H Me O CH2 MeO2C h 0.1 (85) 3.0 -
H H S S POPh2 i 14.2 n (86)
H H O CH2 COOMe j 0.5 n (6,87)
H H NH CH2 COOMe j 4.5 n (6,87)
H H O CH2 CN k 2.3 n (6,87)
H H NH CH2 CN k 10.1 n (6,87)
a In kJ mol-1. Other data on the anomeric effect can also be found in ref.3. NMR spectra have been recorded in CCl4
unless stated otherwise. ∆G°
steric(cyclohex.) values (Eq 1, Fig. 1A) are taken from refs. 88-90. b ∆∆G°
AE1 and
∆∆G°
AE2 values represent the free-energy of the Y-C-X anomeric effect in each compound. c ∆∆G°
AE1 values have been
estimated using the equation: ∆∆G°
AE1 = ∆G°
heterocycle - ∆G°
steric(cyclohex.) where ∆G°
heterocycle is the free-energy of
the Axial � Equatorial equilibrium of the heterocyclic six-membered ring (Eq 2, Fig 1B). ∆G°
steric(cyclohex.) is the
estimate for the steric effect of X in the substituted cyclohexane counterpart. For (CH2)2N, ∆G°
steric(cyclohex.) has been
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
13
assumed equal to that of (CH3)2N. See ref. 3 and the refs. indicated in parentheses in col. 6 for the original ∆G°
heterocycle
values. d ∆∆G°
AE2 has been calculated according to Eq 3. See ref. 3 and the refs. indicated in parentheses in col. 6 for
the original ∆G°
heterocycle values. e ∆∆H°
AE values designate the enthalpy of the Y-C-X anomeric effect in each
compound. f Neat. g In acetic acid. h In methanol. i In CDCl3. j In ether/toluene-d8. k In CFCl3/CDCl3.
l
Using ∆G°
heterocycle values from ref. 73. m In CFCl3/CDCl3. ∆∆H°
AE has been estimated from the simple relation:
∆∆H°
AE = ∆H°
heterocycle - ∆H°
steric(cyclohex.) (Eq 4) where ∆H°
heterocycle is the enthalpy of the Axial � Equatorial
equilbrium of the heterocyclic six-membered ring. ∆H°
steric(cyclohex.) is the enthalpy of the steric effect of X in the
substituted cyclohexane counterpart. n ∆∆H°
AE has been derived from the equation: ∆∆H°
AE = ∆H°
heterocycle - F x
∆G°
steric(cyclohex.), by analogy with the method described in footnote d for free-energy estimates of the anomeric effect
(see ref. 6 and 86 for the values of F).
populations of chair conformations with axially or equatorially oriented anomeric substituent in
pyranose derivatives are also controlled by the electronegativity of the substituents at C4 and C53.
1.6 Effect of the polarity of the solvent on the anomeric effect
The influence of the polarity of the solvent upon the magnitude of the anomeric effect has
been experimentally evidenced in numerous NMR studies on substituted heterocyclic 6-membered
rings such as 2-hydroxy-80,84, 2-alkoxy-71,83,105 and 2-alkylthio-tetrahydropyran79,83, 2-alkoxy-1,3-
dioxane83, 5-halogen-, 5-methoxy- and 5-ethoxy-2-isopropyl-1,3-dioxane95. Polar solvents stabilize
the relatively more polar conformation, i.e. the equatorial chair, more efficiently than the apolar ones.
Therefore the anomeric effect is weakened as the dielectric constant of the solvent increases. In the
case of other substituents at C2 in tetrahydropyran, such as N(CH2)281, the modulation of the bias of
the conformational equilibrium is much reduced. The influence of the solvent on the magnitude of the
anomeric effect should however be considered with caution, especially when there are no accurate
estimates for the actual change of the size (or bulkiness, i.e. A-value) of the anomeric group as a
function of the polarity of the solvent.
1.7 The reverse anomeric effect
An electronegative anomeric group in substituted tetrahydropyrans preferentially adopts an axial
orientation, despite unfavorable steric interactions, owing to the anomeric effect. However, it has been
claimed that if this electronegative substituent is positively charged, the bias of the conformational
equilibrium is just opposite to that predicted in terms of anomeric effect, i.e. the anomeric group
prefers the equatorial orientation more than expected on the basis of steric effect alone 7,106-110. This
tendency has been attributed to the reverse anomeric effect (see Section 4.8 for a comparison with the
pentofuranose system with the protonated aglycone30). The concept of reverse anomeric effect has
been introduced by Lemieux106 to explain the preference of pyridinium in N-(tetra-O-acetyl-α-D-
glucopyranosyl)-pyridinium or 4-methylpyridinium bromides and N-(tri-O-acetyl-α-D-2-deoxy-2-
iodo-mannopyranosyl)-pyridinium perchlorate in aqueous solution for an equatorial orientation, as
shown by his conformational analysis based on the interpretation of vicinal proton-proton coupling
constants. A reverse anomeric effect has also been noted in the case of other pyranosides107,108 with a
pyridinium group at C2. However, whether the preference of pyridinium for the equatorial orientation
is the result of (i) a specific stereoelectronic interaction, or (ii) simply the necessity of the anomeric
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
14
substituent to avoid steric repulsions disfavouring an axial position as a result of a possible increase of
steric bulk of the anomeric group in the protonated (P) compared to neutral (N) state, or (iii) a specific
solvation of the cation, is still a matter of debate7, especially since no reverse anomeric effect has
been observed for the corresponding glucopyranosylammonium ions, in which the bulkiness of the
anomeric group is comparatively significantly reduced111.
The increase (experimentally evidenced by the change of vicinal 3JHH by 2-3 Hz) of the
population of chair conformers with equatorial anomeric substituents in N-(tetra-O-acetyl-α-D-
glucopyranosyl)imidazolium94, N-(tetra-O-acetyl-α-D-mannopyranosyl) imidazolium94, and N-(tri-O-
acetyl-α-D-xylopyranosyl)imidazolium107 in comparison with their neutral imidazole counterparts in
lipophilic media (i.e. in CDCl3 and Me2CO-d6) has also been attributed to the reverse anomeric
effect. In this context, it is important to point out that the positive charge on the remote nitrogen in
imidazolium does not increase dramatically the effective size of imidazolium in comparison with
imidazole, as shown recently by Perrin et al. 112 with help of a 1H-NMR titration method at low
temperature. However, the situation in the lipophilic medium changes completely when the coupling
constants of the nonacetylated parent compounds are measured in aqueous solution 113: No significant
change (well within the experimental accuracy) in 3JHH is found between their P and N states.
Another investigation by Perrin's laboratory114 has shown that the tendency of imidazolium in N-(D-
glucopyranosyl)imidazolium and its tetraacetate to adopt an axial orientation in DMSO-d6, CD3OD
and aqueous solution is slightly greater than that of imidazole in the neutral counterparts. This result,
which is in agreement with what one would expect basing on the simple anomeric effect rule, violates
the concept of the reverse anomeric effect. It also contradicts the initial results of Lemieux94. On the
other hand, Thatcher et al110 have shown that in the case of N-(tri-O-acetyl-α-D-
xylopyranosyl)imidazole in CDCl3 there is a clear enhancement (by ≈ 35 %) of the population of
equatorial chair 1C4 upon addition of trifluoroacetic acid (TFA) with respect to the neutral
counterpart, and this enhancement is not the result of the change of the ionic strength of the solution
induced by addition of TFA. This work confirms the initial findings of Paulsen et al107 and supports
the existence of a reverse anomeric effect for the xylopyranose derivative.
1.8 Hyperconjugation as the origin of the anomeric effect
As yet, no consensus has been reached regarding the origin of the anomeric effect in the
hexose system, but recent evidences from our lab on the pentose systems in nucleosides and
nucleotides point to the orbital mixing as the most probable reason for the anomeric effect (see
Section 4.8). Two models have received outstanding recognition among all those proposed for the
origin of the anomeric effect:
(i) The anomeric effect has often been interpreted as the result of a stabilizing
hyperconjugative interaction115,116 between the filled orbital of one of the electron lonepairs (nX) of
the endocyclic heteroatom X with the antibonding orbital (σ∗A-Y
) of the A-Y bond.
(ii) Alternatively, the anomeric effect can also be attributed to the attempt of the system to
minimize electrostatic repulsions65,117,118 between the dipole formed by the A-Y bond and the dipole
induced by the presence of the heteroatom X. (iii) A few other explanations have also been proposed,
such as Eliel's119 "rabbit-ear effect", which also involves electrostatic interactions, however this time
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
15
directly between lonepairs of electrons. Finally, 4e- orbital mixing destabilizing effects120 have also
been advocated.
Lucken originally explained116 the unusually low nuclear quadrupole resonance frequencies of
35Cl in X-C-Cl fragments of α-halogenoethers by an overlap between a p-type orbital of the
heteroatom X and the antibonding orbital (σ*C-Cl
) of the C-Cl bond. His original interpretation has
given rise to the very popular molecular orbital overlap model which is currently used to explain the
anomeric effect [Fig 2(A-C2)]. In α-halogenoethers, the preference of the halogen atom for an axial
orientation is concomittant with a characteristic shortening of the C-O bond and a lengthening of the
C-Cl bond115 compared with tetrahedral values (Fig 2E).
Figure 2. The rationalization of the anomeric effect in
6-membered rings with help of the delocalization115 and
molecular orbital overlap116 models. (A) Overlap of an
electron lonepair orbital of the endocyclic oxygen with
the σ*C-Cl antibonding orbital of C-Cl bond in 2-
chlorotetrahydropyran (using sp3 hybridized oxygen
lonepairs). (B) The magnitude of the anomeric effect is
proportional to S2 (S is the overlap between the oxygen
lonepair orbital and the σ*C-Cl orbital) and inversely
proportional to the energy difference between both
orbitals, i.e. ∆E(σ*C-Cl - nO)4,5. The nO σ*C-Cl
interaction results in the formation of two new orbitals.
(C1) and (C2) Orientation of the electron lonepairs of the
endocyclic oxygen [either sp3 hybridized, i.e. 1nsp3 and
2nsp3 in (C1) or sp2 hybridized, i.e. 1nsp2 (p-type) and 2nsp2
(s-type) in (C2)] with respect to the exocyclic C-Cl bond.
The nO σ*C-Cl overlap is maximal in the axial conformer
[A(C1) in (C1) or A(C2) in (C2)] in which the torsion
angle β [defined as 1nsp3-O1-C2-Cl in (C1) or 1nsp2 (p-
type)-O1-C2-Cl in (C2)] is closest to 180°, whereas in the
other conformer [E(C1) in (C1) or E(C2) in (C2)] the
oxygen lonepairs are either gauche (for both nsp3 orbitals
in E(C1) and for the 2nsp2 (p-type) orbital in E(C2)) or cis
(in the case of the 2nsp2 (s-type) orbital in E(C2)) to the
exocyclic C2-Cl bond. The anomeric effect therefore
favours the axial over the equatorial conformer. When the
oxygen lonepairs orbitals are sp2 hybridized as in (C2),
the magnitude of the anomeric effect is expected to be reduced in comparison with a situation in which lonepairs are sp3
hybridized as in (C1), owing to the fact that β is closer to 180° in A(C1) than in A(C2). (D) The anomeric effect has been
alternatively explained by the hyperconjugation of one of the endocyclic oxygen lonepairs to the exocyclic C-Cl bond,
which corresponds to the double bond � no-bond resonance in terms of valence bond theory. (E) Shortening of the
endocyclic bond C2-O and concomitant lengthening of the exocyclic bond C2-F in the axial conformer A3 of tri-O-
benzoyl-2-fluorotetrahydropyran compared with the equatorial counterpart E4 of tri-O-acetyl-2-fluorotetrahydropyran,
supporting the hyperconjugative origin of the anomeric effect101.
The hyperconjugation model (Fig 2D), initially proposed by Romers et al115 allows to explain the O-
C-O anomeric effect in 2-substituted-tetrahydropyran (as well as the generalized anomeric effect in
other systems) by the delocalization of one of the lonepairs of the endocyclic oxygen into the
O
O
Cl
H1
Cl
H1
Cl
H1
Cl
O
O
OAc
AcO
H1
AcO
BzO
BzO
OBz
F
F
Cl
Cl
(A)
O
1nsp3
2nsp3
β
α
1nsp3
O
2nsp3
β
α
β
1nsp2 (p-type)
OO2nsp2 (s-type)
β
2nsp2 (s-type)
1nsp2 (s-type)
1nsp3
2nsp3
2nsp3
1nsp3
nOHyperconjugative
stabilization
ΔE (σ*C-Cl - nO)
AE α S2 /ΔE (σ*C-Cl - nO)
σ*C-Cl
σ*C-Cl
O
Cl
O
Cl−
(B)
(C1)
(D)
(C2)
(E)
A2 E2
A(C2) E(C2)
1.339 Å
1.367 Å
1.398 Å
1.406 Å
A(C1) E(C1)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
16
antibonding orbital of the exocyclic C-O bond. This hyperconjugation, which corresponds to a
double-bond � no-bond resonance in terms of valence-bond theory (Fig 2D), is most efficient when
the lonepair orbital and the antibonding orbital of the exocyclic C-O bond are in the same plane, i.e.
when the lonepair adopts an antiperiplanar orientation with respect to the exocyclic C-O bond [Fig 2,
(C1) & (C2)]. This is only achieved in the axial conformer A2, which is therefore more stable than the
equatorial counterpart E2 (Fig 2A). The energy stabilization resulting from the n → σ* interaction is
inversely proportional to the energy difference [∆E(σ*C-Cl
- nO)] between both orbitals and proportional
to square of their overlap (Fig 2B). The electron-donating abilities of orbitals decrease in the order:
nC- > nN > nO > σC-S > σC-H > σC-C > σC-O > σC-F, whereas the electron accepting capability of
orbitals is reduced in the order: σ∗C-Cl > σ∗C-S > σ∗C-F > σ∗C-O > σ∗C-C > σ∗C-H 4,5.
Delocalization is expected to result in the lengthening of the exocyclic C-O bond, whereas the
concomittant shortening of the endocyclic C-O bond and opening of the O-C-O bond angle with
respect to "standard" (tetrahedral) values are observed due to the reduced and enhanced p-character of
the endocyclic and exocyclic C-O bonds, respectively.
The hyperconjugation model is therefore strongly supported by the various examples of
structural data found in the literature, such as in the case of pyranose derivatives, which show
significant changes of the lengths of the exocyclic C-O (or C-Y in the general case) and endocyclic C-
O bonds as well as of the value of O-C-O (or O-C-Y) bond angle (Fig 2E)1,121,122 in chair conformers
with axially versus equatorially oriented anomeric groups.
1.9 The nature of the electron lonepairs
The question of the adequate hybridization state (i.e. sp2 versus sp3) of the electron lonepairs
orbitals has given rise to some controversy. The use of sp3 hybridized lonepairs allows to predict
correctly conformational preferences induced by the generalized anomeric effect1. However,
experimental and theoretical studies123-126 have led to the conclusion that the nonbonding oxygen
lonepairs are nonequivalent: One of them occupies a higher energy orbital with high p-type character,
whereas the other occupies an orbital of lower energy with predominant s-type character. In recent
studies on the anomeric effect117,127,128, the sp2 hybridization model of the electron lonepairs orbitals
has therefore gained widespread recognition.
Basing on the results of their statistical analysis of X-ray crystal structures of COCOC
molecular fragments, Cossé-Barbi and Dubois have shown129 that the five-membered furanose ring
orients polar anomeric groups in axial or pseudoaxial position (95%) much more efficiently than the
six-membered pyranose ring (56%). They argued that this stronger anomeric effect in furanose
compared with pyranose is the result of the better ability of one of the lonepairs of the endocyclic
oxygen to participate in stabilizing hyperconjugation, which can be easily understood with help of the
sp2 (not sp3) hybridization model of lonepairs. Indeed, sp2 hybridization makes the nO → σ
*
C-X
interaction more efficient in furanose than in pyranose130, since it results in a perfect antiperiplanar
orientation of the p-type lonepair orbital of the endocyclic oxygen atom with respect to the exocyclic
C-X bond in the former but not in the latter [compare panel (C2) in Fig 2 with panel (C) in Fig 9]. If
the lonepairs of the endocyclic oxygen were sp3 hybridized, the situation would be just reverse, i.e.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
17
the overlap between nO and σ
*
C-X orbital would be optimal in the pyranose system but not for the
furanose counterpart [compare panel (C1) in Fig 2 with panel (B) in Fig 9].
1.10 Dipole-dipole (electrostatic) interactions as origin of the anomeric effect
Already in 1955, Edward65 proposed that the destabilization of the E2 chair of 2-
methoxytetrahydropyran (2) (with equatorially oriented 2-OCH3 group) with respect to the axial
counterpart A2 is the result of electrostatic repulsions between 2-OCH3 and the ring dipole. The ring
dipole is itself the resultant of individual dipole moments from endocyclic C-O bonds and the
lonepairs of the endocyclic oxygen. More recently, the electrostatic model has been refined and the
anomeric effect has been attributed to electrostatic repulsions117,118 between the exocyclic C-X dipole
and the ring dipole in the E2 chair (Fig 3A). The repulsion between two dipoles is function of their
magnitudes as well as the distance and the angle between them74,131. In the equatorial conformer E2,
both dipoles are nearly parallel and repel each other much more than in the axial form A2 where the
angle between them is nearly 90°. As a result, the E2 conformer is destabilized by electrostatic
repulsions compared to the A2 counterpart and the anomeric group 2-OCH3 preferentially adopts an
axial orientation.
The experimentally found (Section 1.6) dependence of the magnitude of the anomeric effect
upon the polarity of the solvent has been considered as the best evidence for the electrostatic origin of
the anomeric effect: Polar solvents stabilize the more polar conformation (i.e. the equatorial chair E2)
more efficiently than solvents with lower dielectric constants, therefore the strength of the anomeric
effect will be reduced in polar compared to apolar solvents. In a recent calorimetric study of the heats
of hydrolysis of the anomeric forms of 4,6-dimethyl-2-methoxytetrahydropyran, Wiberg and
Marquez118 have also shown that the nature of the solvent strongly influences ∆H° and ∆G° of the
isomerization equilibrium, which led them to conclude that the anomeric effect is mainly induced by
the electrostatic repulsion between the C-O dipoles, the magnitude of which decreases when the
dielectric constant of the solvent is increased.
However, there are also known examples of an increase in the strength of the anomeric effect
induced by an increased polarity of the solvent, such as for trans-2,5-bis(trimethylsiloxy)-1,4-
dioxane132. This particular result cannot be explained by the electrostatic model. Instead, it has been
rationalized in terms of hyperconjugative interactions (Section 1.8). More importantly, in contrast
with the hyperconjugation model, the electrostatic model does not allow to explain the changes of
bond lengths and bond angles that are commonly associated with the existence of the anomeric effect
in heterocyclic 6-membered rings (Section 1.8).
The question of the relative importance of hyperconjugative interactions versus dipole-dipole
repulsions as the origin of anomeric effect in non-polar solvents has been raised in a recent
comparative conformational study of 2-methoxy-1,3-dimethylhexahydropyrimidine (3) and 2-
methoxy-1,3-dioxane (4)117. It was argued that since the dipole moment of CH3-NH2 is smaller than
that of CH3-OH, substitution of nitrogen in 3 for oxygen in 4 should lead to increased electrostatic
repulsions in 4 compared to 3. Conversely, as evident from the lower ionization potential of CH3NH2
in comparison with that of CH3OH, replacing oxygen in 4 by nitrogen in 3 should raise the energy of
the filled nN lonepair orbital in the later with respect to that of nO orbital in the former and therefore
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
18
strengthen the n→σ* hyperconjugative interactions in 3 compared to 4. As a result, if the anomeric
effect operates mainly through dipole-dipole (electrostatic) interactions, then replacing oxygen by
nitrogen should reduce the strength of the anomeric effect in 3 compared to 4. Conversely, if
hyperconjugative interactions predominate, then the strength of the anomeric effect should be
increased in 3 compared to 4. Perrin's conformational analysis based on temperature-dependent 1H
and 13C NMR spectra showed that the proportion of axial conformer is nearly the same in 3 and 4.
However, after correction of the results for the steric effects induced by the presence of methyl groups
in 3 (supported by molecular mechanics and AM1 calculations), it turned out that the anomeric effect
is weaker in 3, and it was therefore concluded that the anomeric effect is mainly the result of
electrostatic interactions. Additionally, it was also suggested that the bond length changes that are
often invoked as a key evidence for the existence of hyperconjugative interactions (Section 1.8) can
also be explained in terms of dipole-dipole interactions. A subsequent reinvestigation of the origin of
the anomeric effect by Salzner128 based on ab initio geometry optimizations (at HF/6-31G* level) of
various conformers of hexahydropyrimidine (5), 2-hydroxypiperidine (6) and 2-
hydroxyhexahydropyrimidine (7) and on the NBO analysis of the resulting wave functions showed
that indeed the magnitude of the anomeric effect in 6 and 7 is much reduced in comparison with the
oxygen analogs. However, it was also shown in that work that the total energies of the conformers do
not correlate with total dipole moments, which discredited the predominance of electrostatic
repulsions. Instead, the weaker anomeric effects in 6 and 7 in comparison with the oxygen
counterparts were interpreted as the results of the competition between OH/NH bond repulsions and
hyperconjugative stabilizations.
1.11 Alternative explanations for the origin of the anomeric effect
According to Eliel's original rabbit-ear effect model119
, the anomeric effect in R-X-A-Y
fragment arises from repulsions between electron lonepairs of X and Y heteroatoms (Fig 3B). This
model has been used in particular to explain the drive of the conformation of 2-
alkoxytetrahydropyran76,95 and tripepiperideine133 by the O-C-O and N-C-N anomeric effects,
respectively. Alternatively, it has also been suggested by Box120 that four electrons orbital mixing
destabilizing interactions allow to account for X-C-Y anomeric effect controlling the conformation of
glycosides: according to this model, the occupied (2+2) electron lonepair orbitals of X and Y
heteroatoms interact to form two new orbitals, occupied by the four electrons (Fig 3C). However, the
energy destabilization of the system induced by the newly formed high-energy orbital is greater than
the stabilization gained by the formation of the new low-energy orbital, therefore such lonepair-
lonepair interactions, on the overall, are destabilizing, and the system attempts to adopt a
conformation in which there are fewer lonepair orbital interactions. As a result, the anomeric
substituent adopts preferentially an axial orientation.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
19
1.12 The gauche effect
The "gauche effect"1,134-143 designates the tendency of R-X-Y-R' and X-C-C-Y fragments
(where X and Y represent electronegative atoms or groups) to adopt preferentially gauche over
antiperiplanar orientations in spite of unfavourable steric repulsions and electrostatic interactions (i.e.
the relative energy difference between trans and gauche conformers, ∆Et-g, is negative, as depicted in
Fig 4). For 1,2-disubstituted ethane in the gas phase, there is a linear relationship between ∆Et-g and
the sum of Huggins electronegativities of both substituents135. 1,2-dimethoxyethane exists
preferentially in a gauche conformation in solution, but the stabilization of the gauche conformer is
much reduced in the gas phase144. On the other hand, ab initio calculations predict almost identical
energies for the gauche and trans rotamers142. A study of the energy relationship of the isomers of
1,2-difluorethylene, 1,2-difluorodiazene and several other halogen- and oxygen-substrituted
ethylenes145 show that the cis isomer has the lower energy, which is generally referred to as the cis
effect.
Figure 3. The anomeric effect in 2-substituted
tetrahydropyran as the result of electrostatic interactions.
Lonepairs on the endocyclic and exocyclic oxygen atoms are
represented assuming sp3 hybridization. (A) Destabilization
of the equatorial chair E2 with respect to the axial counterpart
A2 by dipole-dipole repulsions65,117,118
. (B) Prediction of the
relative stabilities of B1 - B6 conformers of 2-
methoxytetrahydropyran (2) using Eliel's rabbit-ear model95
of interaction between the electron lonepairs of the
endocyclic and exocyclic oxygens. ("black" lonepair: in the
front, "empty": at the back, "grey": in the plane of the
drawing). The equatorial conformers B1 - B3 as well as the
axial forms B4 - B5 are destabilized owing to one or two (for
the B2 form only) repulsion(s) between lonepairs of the
endocyclic and exocyclic oxygens. The axial conformer B6,
in which the anomeric group OMe adopts an axial
orientation, is the most stable, since only in this situation
there is no destabilizing lonepair-lonepair interaction. (C)
The four electrons interaction model120
. The interaction
between two filled lonepairs orbitals of the endocyclic [nO
(endo)] and exocyclic [nO (exo)] oxygens produces two new orbitals. This interaction is destabilizing, since the stabilization
(∆E1) of the system gained from the formation of [nO(endo) + nO(exo)] orbital is weaker than the corresponding
destabilization (∆E2) owing to the newly formed high-energy [nO(endo) - nO(exo)] orbital .
“Attractive” and “repulsive” gauche effects define a greater preference of a X-C-C-Y fragment for
gauche and anti orientations, respectively, than that expected on the basis of steric effects and
electrostatic interactions alone. It has been shown138-140 that the extent of stabilization of the 1,2-
diaxial versus 1,2-diequatorial conformers of 1,2-trans-disubstituted cyclohexane is dictated by the
fine balance between steric, electrostatic and gauche interactions between vicinal X and Y
substituents: For strongly electronegative X and Y substituents (i.e. O/O, F/Cl, F/Br, F/I or F/O pairs),
a strong attractive gauche effect operates, whereas when the electronegativity of X and Y is
intermediate (such as O/Cl, O/I or Cl/I pairs), the observed conformational preferences can be
explained by steric and electrostatic repulsions alone. In contrast, for S/S, S/Cl, S/Br or S/O
O
O
XX
A2 chair E2 chair
O OO
O O O
O O O
O O O
Me
Me
Me
MeMe
Me
(A)
(B) B1 B2 B3
B4 B5 B6
(C)
ΔE2
nO (endo) - nO (exo)
ΔE1
nO (endo) + nO (exo)
ΔE2 > ΔE1
nO (endo)nO (exo)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
20
disubstitution, the population of gauche rotamers is less than expected (owing to the repulsive gauche
effect). In 5-substituted-1,3-dioxan, conformers with axially oriented C5 substituent are preferred as
the result of the attractive gauche effect146, whereas in 5-methoxy- and 5-methylthio-1,3-dithianes
repulsive S/O and S/S gauche effects are operating147.
1.13 The energetics of the gauche effect
As stated above, the gauche effect implies the electronic energy preference of a gauche
rotamer over the corresponding anti rotamer1. The experimental energetic relationship of various
rotamers for 1,2-difluoroethane (as derived from Raman148,149, infrared149, NMR spectroscopy150 and
electron diffraction151-153 studies) or 1,2-dimethoxyethane (by NMR spectroscopy144) in the gas
phase has been well worked out. These experimental works have shown that the gauche rotamer has a
lower energy than the anti counterpart by 2.4 - 3.4 kJmol-1 (in the case of difluoro substitution),
which compares well with results obtained from theoretical studies141,154 (i.e. ab initio calculations at
the MP2/6-311++G** and TZ+2D+P levels).
Figure 4. Plot of the relative energy of 1,2-
disubstituted ethane conformers as a function of the
ΦXCCY torsion angle, showing the stabilization [∆E(t-
g) < 0] of gauche+ and gauche-
rotamers over the trans
counterpart, as the result of the attractive X-C-C-Y
gauche effect.
1.14 Possible origins of the gauche effect
The origin of the gauche effect is still a matter of controversy: (i) It has been shown that the
preferred gauche orientation in ethylene glycol142,155 results from hydrogen bonding. (ii) The
repulsive gauche effects for some 1,2-disubstituted cyclohexane derivatives (e.g. for S/S
disubstitution pattern) have been attributed to the through-space interaction (or "hockey sticks effect")
between lonepair orbitals on the X and Y heteroatoms in the X-C-C-Y fragment138,156,157. Such an
overlap between two doubly-occupied lonepairs is destabilizing: It results in the formation of a
bonding orbital and its antibonding counterpart, however the destabilizing influence of the latter is
greater than the stabilizing effect of the former (Fig 5, Panel A1). On the other hand, Hoffmann158
suggested that the lonepair interaction does not only take place through-space but it has also a
significant through bond component (Fig 5, Panels A2 and A3). The actual repulsive or attractive159
character of the gauche effect is the result of the fine balance between through-space and through-
bond contributions. Hoffmann's analysis is supported by the results of a photoelectron spectroscopic
study of 3,7,9-trihetero derivatives of bicyclo[3.3.1]nonane, showing the preponderant influence of
the through-bond component160.
(iii) σ →σ∗ orbital interactions between the best donor σ orbital and the best acceptor σ*
orbital have also been suggested136,161. The efficiency of these interactions is proportional to the
0 60 120 180 240 300 360
0
1
2
3
4
Φ X-C-C-Y
( ° )
ΔE
(t-g)
Y
X
Y X
Y
X
Y
X
X Y
Y
X
X
Y
gauche −
gauche +
eclipsed
cis
trans
eclipsed
cis
Relative Energy
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
21
square of the overlap between the donor and acceptor orbitals and inversely proportional to the
difference between their energies. An example of σC-H →σ∗C-F orbital overlap in the gauche
conformer of 1,2-difluoroethane is shown in Fig 5B-C.
Figure 5. (A1)-(A3) Through-space (A1) and through-bond (A2 and A3) interactions138,156-160
between lonepair orbitals
(Φ1 and Φ2) on the X and Y heteroatoms in the X-C-C-Y
fragment, as the origin of attractive or repulsive gauche
effects. Through-space (A1) four electron interactions
result in the formation of a bonding orbital (Φ1 + Φ2),
which is stabilizing (∆E1 < 0) whereas the antibonding
counterpart (Φ1 - Φ2) is destabilizing (∆E2 > 0). ∆E2 >
∆E1 and therefore the system is destabilized on the
overall. (Φ1 + Φ2) can in turn interact with σC-C [through-
bond component of the origin of the gauche effect, (A3)].
This four-electron interaction is again destabilizing since
∆E4 > ∆E3. On the other hand, the concomittant two-
electron interaction between (Φ1 - Φ2) and σ*C-C is
stabilizing, as shown in Panel (A2). (B) and (C): σC-H
→σ*C-F orbital overlap to explain the gauche effect in
1,2-difluoroethane136,161
. The σ →σ* overlap is most
efficient when it involves the best donor and best acceptor
σ and σ* orbitals, respectively. In the case of the trans
conformer, the σ orbital is σC-F, which is much poorer
donor than σC-H in the corresponding gauche conformer
[Section 1.8]. Therefore, the σC-H →σ*C-F interaction
stabilizes the gauche conformer with respect to the trans
counterpart, and by bond-no bond resonance the olefin (3)
is envisioned as a canonical form. (D) The gauche effect
is shown in relation with bent bonds143,162
. In the trans
form, the C-C bond paths are bent in the opposite
direction (as shown by the dashed lines), which leads to a
reduced overlap and a weaker C-C bond. On the contrary,
in the gauche rotamer, bond paths are bent essentially in
the same direction. Therefore the trans rotamer is
destabilized with respect to the gauche counterpart.
Wolfe originally defined the gauche effect134 as "a tendency to adopt that structure which has
the maximum number of gauche interactions between the adjacent electron pairs and/or polar bonds",
which he rationalized in terms of a nuclear-electron attraction prevailing over nuclear-nuclear and
electron-electron repulsions. This definition enables to predict correctly the preferred gauche
orientation of two vicinal electron lonepairs R and R' in a R-X-Y-R' system or of X-C and C-Y polar
bonds in X-C-C-Y fragment, but not the antiperiplanar orientation136 of a lonepair with respect to a
vicinal polar bond, i.e. owing to the anomeric effect (Sections 1.2 - 1.6). It is therefore preferable to
define1 gauche effect as "a stereoelectronic preference for conformations in which the best donor
lonepair or bond is antiperiplanar to the best acceptor bond". (iii) The preferential gauche arrangement
in 1,2-difluoroethane has also been explained by the fact that in this rotamer, the C-C bond paths (i.e.
the regions with maximum electron density connecting both carbon atoms163) are bent in the same
direction, whereas in the relatively less stable trans form bending occurs in opposite directions, which
results into a reduced overlap and a weaker C-C bond143 (Fig 5D). This theory is supported by the
C C
F
F
C C
FF
(D)
H
H
F
H
H
F
H
H
F
H
F
H
F
H
H
H
F-
trans (1) gauche (2) (3)
(B)
trans (1) gauche (2)
H
F
H H
H
F
H
H
H F
H
F
180ο
60ο
(C)
YX
YX
X
Y
YX YX
YX YX
ΔE1
ΔE2
Φ1
Φ2
Φ1 +Φ2
Φ1 - Φ2
Φ1 +Φ2
Φ1 - Φ2
σ
σ*
(A1)
(A2)
(A3)
ΔE3
ΔE4
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
22
comparison of geometrical parameters for the gauche and anti rotamers of 1,2-difluoroethane, as
obtained from an experimental study based upon high-resolution infrared spectroscopy162 or from ab
initio calculations (using hybrid-DFT164 or RHF/6-311++G** level149) .
2. Stereoelectronic effects in nucleosides and nucleotides
2.1 Structure of nucleic acids
Endowed with unique capabilities, such as the storage of the genetic information, induction of
cellular differentiation, propagation of tumor viruses as well as splicing and autocatalysis, deoxyribo
(DNA) and ribonucleic acids (RNA) are polymers built up of monomers, the nucleosides, which are
covalently linked through 3'→5' phophodiester linkages165. Nucleosides consist of a pentofuranose
sugar and a heterocyclic nucleobase (adenin-9-yl, guanin-9-yl, cytosin-1-yl, thymin-1-yl, uracil-1-yl or
modified C-166 or N-heterocycle165) connected by the covalent C1'-N1/9 or C1'-C (glycosyl) bond. In
natural DNA and RNA, the D-sugar enantiomer and β-glycosyl linkage are preferred over the L- and
α-counterparts. The conformation along the phosphate backbone in nucleotides can be fully
defined167,168 by the torsion angles α, β, γ, δ, ε and ζ (Section 7.1).
The stability of the three-dimensional structure of nucleic acids is usually attributed to only a
few types of forces: (i) Strong intermolecular H-bonds between complementary nucleobases; (ii)
Intramolecular base-base stacking interactions; (iii) Electrostatic interactions and hydrophobic forces
between adjacent base pairs and across the nucleotide chain165, and hydration. The purpose of the
next sections is to present recent progress in our understanding of the interplay of stereoelectronic
gauche and anomeric effects in the drive of the conformation of simple nucleosides and nucleotides
and their analogs as well as the self-organization of oligonucleotides.
2.2 The pseudorotation concept
Saturated heterocyclic six-membered rings are much less flexible than their five-membered
counterparts169. The activation energy barrier for the conversion of one chair form of cyclohexane
into the other is170,171 about 42 kJmol-1. 1H- and 13C NMR spectra of 2-methoxy-1,3-
dimethylhexahydropyrimidine (3)117 measured in various solvents have shown that the ∆G* value for
the ring inversion between two chair forms is ≈ 37 ± 2 kJmol-1 (the two interconverting chair
conformers could be isolated at low temperature). On the other hand, the activation energy barrier for
the interconversion between the two preferential puckering modes, i.e. N- and S-type pseudorotamers,
of the pentofuranose moiety in purine nucleosides is much reduced in comparison, i.e. below 20 - 25
kJmol-1 (vide infra).
The pseudorotation concept has been introduced by Kilpatick et al. 172 to describe the
continuous interconversions between an infinite number of indefinite puckered forms of the
cyclopentane ring. Pseudorotation173 allows cyclopentane to relieve strains which would be induced
by 120° bond angles and eclipsed methylene groups if it would adopt a planar geometry.
A barrier to planarity of cyclopentane of 22 kJmol-1 has been reported176. The concept of
pseudorotation has been applied for the first time to furanose by Hall et al. 177 in their conformational
analyses of pentofuranosyl fluorides.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
23
Kilpatrick originally described172 the puckered states of cyclopentane by sequential out-of-
plane displacements (of the carbon atoms) from the least square plane of the unpuckered ring. The
displacement zi of the ith ring carbon in a particular puckered conformer was calculated using Eq 5:
zi = (2/5)1/2 q cos 2 (Φ + 2πi/5) ..... Eq 5
where, q and Φ represent the maximum puckering amplitude and the phase angle of pseudorotation of
the pseudorotamer, respectively. Eq 5 holds exactly in the case of the infinitesimal diplacement in an
equilateral pentagon. Cremer and Pople have later on devised a generalized set of puckering
coordinates valid for equilateral and non-equilateral five-membered rings178. This formalism is
however quite cumbersome since it requires the
Figure 6. The pseudorotation wheel (E =
envelope; T = twist) for pentofuranosyl D-
nucleosides. The hyperspace of geometries
accessible to N- and S-type pseudorotamers is
within the shaded areas for β-D-dNs, β-D-rNs
and β-D-arabinonucleosides (-1° < PN < 34°,
137° < PS < 194°, 30° < Ψm(N), Ψm(S) <
46°)174 and within the empty oval shaped
areas for α-D-dNs, α-D-rNs, α-D-xylo- and α-
D-arabinofuranosyl nucleosides175 (-18° < PN
< 19°, 168° < PS < 224°, 28° < Ψm(N), Ψm(S)
< 49°).
knowledge of 15 cartesian
coordinates. Altona et al. have solved
this problem by proposing an alternative description179-181 of a puckered geometry of the
pentofuranose ring in nucleic acid derivatives by two parameters P and Ψm, related in turn to the five
endocyclic torsion angles νi (i = 0...4) (Eq 6a): νi = Ψm cos (P+4π(i-2)/5) ..... Eq 6a
P, the phase angle of pseudorotation (0° < P < 360°), shows which part of the ring is mostly
puckered, whereas Ψm, the maximum puckering amplitude, provides information about the largest
deviation of the endocyclic torsions from zero. The ensemble of puckered forms that may be adopted
by the pentofuranose moiety in nucleos(t)ides is represented in the form of the pseudorotation cycle
(Fig 6). In nucleosides, the endocyclic torsions νi (i = 0...4) are defined as follows: ν0 [C4'-O4'-C1'-
C2'], ν1 [O4'-C1'-C2'-C3'], ν2 [C1'-C2'-C3'-C4'], ν3 [C2'-C3'-C4'-O4'] and ν4 [C3'-C4'-O4'-C1']. P
and Ψm are derived from Eqs 6b and 6c:
tan P = [(ν4 + ν1) - (ν3 + ν0)] / [2ν2 (sin 36° + sin 72°)] .....Eq 6b
Ψm = ν2 / cos P .....Eq 6c
R
O
R'
O
R
N
H
H R'
HOH2C
OHOH2C
R
N
H
H
R'
N
HOH2CH
H
H
H
H
H
H
H
H
O
H
H
HR'
R
CH2OHN 0E
144º
3E
108º 0
1T
0º
T
306º
234º
198º
T
4
3T
0
4T
3
4T
3
2T
1
2T
T1
0
0E
3E
1E
4E
2E
1E
4E
2E
324º
288º
252º
216º
180º
72º
36º40
ο
90º
40ο
342º
162º
54º
18º
126º
20ο
0ο
2
3
270º
2
1
4
020
ο
T
West (W)
South (S)
East (E)
North (N)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
24
For practical purposes, Eq 6a may be regarded as nearly exact for equilateral five-membered
rings, since it reproduces the values of the endocyclic torsion angles of a cyclopentane pseudorotamer
with an accuracy better than 0.1°182. In a study of 178 β-D-furanosides, the endocyclic torsion angles
could be calculated using Eq 6a with an r.m.s. error of 0.4 - 0.9° 183,184. For non-equilateral rings, a
more accurate expression (r.m.s. = 0.2 - 0.4°) for the dependence of the endocyclic torsions on P and
Ψm is given by Eq 7183-186:
νi = ai * Ψm cos (P+εi+4π(i-2)/5) ..... Eq 7
where, ai and εi represent correction factors that are included to overcome differences in endocyclic
bond lengths.
Altona's model has not only been used to describe the conformation of the pentofuranose
moiety in nucleosides and nucleotides187-226 but also the puckered states of cyclopentane179, the
pyrrolidine ring in L-proline and in its derivatives182,227, the ring D in steroids180, and other five-
membered rings115,228-231.
Ab initio calculations232 have shown that pseudorotation of cyclopentane occurs without any
substantial change in potential energy whereas in unsymmetrically substituted saturated five-
membered rings potential energy thresholds induced by the presence of the exocyclic substituents
limit the flexibility of the system and lead to prefered puckering modes233.
O
OR'B
O
OR'
B
RO
OR
O
B
HO
HO
O
OR'
B
O
OR'
BRO
OR
North (N) β-D-sugar (C3'-endo-C2'-exo)
South (S) β-L-sugar (C2'-endo-C3'-exo)
(iv)
O
B
O
B
HO
OH
(iii)
HO
North (N) β-L-sugar(C2'-exo-C3'-endo)
South (S) β-D-sugar (C2'-endo-C3'-exo)
HO
North (N) α-L-sugar(C2'-exo-C3'-endo)
South (S) α-L-sugar (C2'-endo-C3'-exo)
South (S) α-D-sugar (C2'-endo-C3'-exo)
North (N) α-D-sugar (C3'-endo-C2'-exo)
(i) (ii)
B = adenin-9-yl, guanin-9-yl, cytosin-1-yl, thymin-1-yl or uracil-1-yl (in RNA)
D-series L-series
O
B
OH
HO
Scheme 1: The D- and L- mirror image relationship for the two-state dynamic N � S sugar equilibrium in α-D-
dNs (i) , α-L-dNs (ii), β-D-dNs (iii) and β-L-dNs (iv). In L-nucleosides234, the N sugar is redefined as being the
form with maximal negative value for the endocyclic torsion [C1'-C2'-C3'-C4']. Note that in α-D-dNs (i) and α-L-
dNs (ii), the aglycone B becomes more pseudoaxial as the anomeric effect becomes stronger in the S-type
conformation, whereas in β-D-dNs (iii) and β-L-dNs (iv), this is achieved in the N-type conformation. Hence, the
sign for the energetic drive of the anomeric effect in α-nucleosides is opposite to that of β-counterparts (Sections 4
and 5).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
25
2.3 The two-state N � S equilibrium in β-D-nucleosides
In D-nucleosides, in the typical N (P = 0°) and S (P = 180°) conformations, the [C1'-C2'-C3'-
C4'] torsion angle has a maximal positive and negative value, respectively, whereas in the L-
enantiomers the signs, as defined by Hoffmann and Altona234, are opposite as a result of the mirror-
image relationship of D- and L-enantiomers (Scheme 1).
A perusal174 of 178 crystal structures of β-D-ribo-, 2'-deoxyribo- and arabinonucleosides
shows that the phase angle values of their puckered pentofuranose moieties are not evenly distributed
over the whole 0 - 360° range but are instead clustered in the North (N, -1° < P < 34°, centered around
C3'-endo) and South (S, 137° < P < 194°, centered around C2'-endo) regions, while Ψm is within the
30° - 46° range with an average value of 38.6 ± 3° (Fig 6). The N and S regions are nearly equally
populated for ribonucleosides whereas for 2'-deoxyribonucleosides there is a clear preference for S-
type geometries (≈ 3:1 for the ratio of S versus N pseudorotamer populations). Only a few E-type
pseudorotamers and no W-type conformers were found among these crystal structures, which is
consistent with an analysis based upon simple steric effects: W-type conformations are energetically
highly disfavoured owing to the pseudoaxial orientation of both the nucleobase and the 5'CH2OH
group and to the fact that the C2' and C3' substituents are eclipsed. Analogously, the energy
destabilization of E-type conformations can be attributed to the eclipsed orientation of the C2' and C3'
substituents. 1H-NMR studies in aqueous solution193-204,206-208,212,216,219,220,222,224-226,235-240, have
shown that the puckered geometries of the furanose moiety in nucleos(t)ides interconvert rapidly (in
the NMR timescale) since only time-average chemical shifts and coupling constants can be extracted
from the spectra. This means that the activation energy barrier between the interconverting
pseudorotamers is significantly reduced in comparison with heterocyclic six-membered rings, which
often adopt a single chair conformation (vide infra).
However, two distinctly identifiable and dynamically interconverting N and S conformations have
been observed in some B � Z DNA241,242, A � Z RNA243,244 or A-form � B-form lariat
RNA245,246 transitions.
The NMR results studies as well as de Leeuw et al's statistical analysis of the distribution of P
values of the crystal structures of nucleos(t)ides (vide supra) suggest that the conformation of the
pentofuranose moiety of nucleos(t)ides in aqueous solution is adequately described by a two-state N
� S equilibrium model, since no third state is found yet. This is also supported by the following
observations, based upon the results of our own conformational studies on nucleosides and
nucleotides14-44: (i) PN, PS, Ψm(N) and Ψm(S), as obtained from pseudorotational analyses of
experimental vicinal proton-proton coupling constants (3JHH) for β-D-2',3'-dideoxynucleosides19 are
virtually the same in the analyses which have been performed using the 3JHH data at any single
temperature or in other analyses based upon the whole set of temperature-dependent 3JHH coupling
constants (i.e. at seven temperatures within the 283 K - 353 K range). (ii) Van't Hoff plots of the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
26
ratios of mole fractions of S- and N-type sugars (xS/xN) as a function of the inverse of temperature
give straight lines for all nucleosides with correlation coefficients ≥ 0.95. (iii) We have recently
shown that ∆H˚, ∆S° and ∆G° of the N � S equilibrium in β-30,32,36, and α-nucleosides36,37 are pD-
dependent (Sections 4 - 6). The shift of the N � S equilibrium in β- (or α-) nucleosides toward more
N- (or S-type) sugars at acidic pDs upon protonation of the nucleobase, and toward S-type (or N-type)
conformers at alkaline pDs upon deprotonation, results from the transmission of pD-tunable electronic
character of the nucleobase to steer the sugar conformation through modulable anomeric and gauche
effects. The pDs at the inflection points of the plots of pD-dependent ∆G° values gave the pKas of the
nucleobases, and these values were virtually the same as those published247 or independently
determined through monitoring of pD-dependent chemical shifts of non-exchangeable aromatic and
anomeric protons30,32,36,37. Raman spectroscopy study on A and T containing DNAs248 also support
that the two-state N � S equilibrium model also holds for the pentofuranose sugar moiety.
Throughout this whole book and in all papers cited from this laboratory, the signs of the
thermodynamic quantities have arbitrarily chosen in such a way that the positive values indicate the
drive of the N � S equilibrium to N, whereas the negative values describe the drive to S.
Interconversions between N- and S-type puckered furanose conformers most likely occur
along the pseudorotational cycle (i.e. via puckered geometries) rather than through a planar
intermediate, which is disfavoured as the result of high strain energies181,249,250. The fact that only a
few E-type puckered geometries and no W-type pseudorotamer were identified among the crystal
strutures of 178 β-D nucleos(t)ides suggests that N- to S- interconversion proceeds via an E-type
puckered geometry intermediate, not through the W region.
The intrinsic flexibility of the pentofuranose moiety is absent in cyclic nucleosides such as
2',3'-O-, 8,2'-O-, 8,3'-O-anhydronucleosides or in cyclic nucleotides165. This results in a reduced
puckered amplitude, for instance in 2',3'-cyclic nucleotides251-253 or to unusual puckering modes of
3',5'-cyclic nucleotides, both in solution254 and in the solid state255-257.
The effects of packing forces and hydrogen bonds upon the observed puckering modes of
nucleosides in the solid state are well known165. Solution and solid state conformations of
pentofuranose differ dramatically for some nucleosides: Adenosine258 crystallizes as the N-type
conformer, whereas its hydrochloride salt crystallizes as the S-type form259 and our NMR studies30
show a clear preference (≈ 67% at room temperature) for S-type conformers at neutral pD, and the
pseudorotational equilibrium is shifted to more N-type sugars at acidic pD (≈ 55% South conformer at
298 K). 3'-azidothymidine17 in D2O is involved in an equipopulated two-state N � S equilibrium,
despite the fact that it crystallizes in the S-type puckered conformers: P = 175° with Ψm = 32° and P =
215° with Ψm = 36°260-265. 3'-fluorothymidine crystallizes266,267 in the form of S-type conformers,
however the average phase angle value for these conformers (P ≈ 171°, Ψm ≈ 34°) deviates from that
found for the preferred S-type sugar geometry (≥ 90%, P ≈ 160°, Ψm ≈ 34°) by 1H NMR studies17,26.
For all β-D-2',3'-dideoxynucleosides (ddNs), the pentofuranose sugar is mostly puckered as the N-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
27
type conformer in D2O solution19 (≥ 75% at 298K), whereas in the solid state, it adopts S-type
geometries (P = 194°, Ψm = 37° for ddA268, P = 208° and Ψm = 34° for ddC269).
An integrated conformational study on 4'-thiothymidine and (E)-5-(2-bromovinyl)-2'-deoxy-4'-
thiouridine in solution (based on the interpretation of 3JHH with help of the pseudorotation concept
and on the analysis of nOe enhancements) and in the solid state (from X-ray crystallography) has been
recently reported by our lab270. It was found that the crystal structures of both thymidine derivatives
are grossly similar with a 4'-thiofuranose moiety in S-type conformation (P = 180° and Ψm = 48° for
both compounds). Owing to the nonequilateral nature of 4'-thionucleosides, the pseudorotation
equation (Eq 7) was reparametrized basing on ab initio calculations. The solution structure of the
major conformer of 4'-thiothymidine and its derivative were similar to the crystal structure, showing
that the presence of S4' vis-a-vis O4' does not result in any additional flexibility of the thiofuranose
ring. The similarity of the bias of the N � S equilibrium in thionucleosides and thymidine show that
the magnitude of interplay of gauche and anomeric effects is comparable in both, thereby showing the
insignificant effect of S vis-a-vis O. It is also noteworthy that the bond lengths of C1'-S and C4'-S are
essentially the same as found in many other natural C-nucleosides271-282 and N-nucleosides174,283,
suggesting that the identification of the anomeric effect on the basis of unequal bond lengths of C4'-
X4' versus C1'-X4' is an oversimplification5.
2.4 The two-state N � S equilibrium in α-nucleosides
A recent perusal175 of the few crystal structures available in the literature for α-nucleosides
has shown that their constituent pentofuranose moieties also tend to adopt predominantly
conformations belonging to the C2'-endo and C3'-endo domains (-18° < PN < 19°, 168° < PS < 224°,
28° < Ψm(N), Ψm(S) < 49°) with a few exceptions found in the C4'-endo region. Therefore, the
hypothesis of the two-state N � S equilibrium has also been retained during our conformational
studies on α-nucleosides (Section 5). In that work175, basing on the results of potential energy
calculations on both α-purine and α-pyrimidine nucleosides, it was suggested that interconversion
between N- and S-type α-sugars proceeds through the O4'-exo rather than through the higher O4'-endo
activation energy barrier, in contrast with what has been claimed for β-counterparts (vide infra).
2.5 The two-state N � S equilibrium in carbocyclic nucleosides
The chemical and enzymatic vulnerability of the glycosyl bond of natural nucleosides has
inspired the synthesis of a wide gamut of carbocyclic nucleosides284-291 in which a cyclopentane ring
replaces the ribose moiety. However, these more stable carbocyclic nucleosides are generally less
potent284 than their natural counterparts. This is attributed to the more flexible nature of the
cyclopentane ring where both anomeric and gauche effects inherent to the natural nucleosides17,19-
30,32,33,35-38,40-44 are lacking. Efforts to reduce the flexibility of the cyclopentane ring have resulted in
the design of conformationally constrained carbocyclic nucleosides292,293 (see Sections 8.2 and 8.3).
The X-ray crystal structures of four conformationally unconstrained carbocyclic
nucleosides294-297 are so far known, i.e. for (-)-aristeromycin294 (P = 89.0° and Ψm = 40.8), (+)-
carbathymidine295 (P = 118.6° and Ψm = 41.2°), 1-(c-2-fluoro-t-4-hydroxy-c-3-hydroxymethyl-r-1-
cyclopentyl)uracil296 (P = 95.7° and Ψm = 43.1°) and the carbocyclic analogue of 2'-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
28
deoxyguanosine297 (P = 293.3° and Ψm = 35.8°). The crystal structures59 of 6'-α-methylcarbocyclic
thymidine in four different forms (P = 147.4° and Ψm = 47.6°; P = 141.9° and Ψm = 49.1°; P = 153.6°
and Ψm = 44.5°; P = 111.2° and Ψm = 37.6°) and of 6'-α-hydroxycarbocyclic thymidine (P = 127.9°
and Ψm = 44.2°; P = 141.1° and Ψm = 45.6°; P = 140.0° and Ψm = 48.2°; P = 123.4° and Ψm = 49.0°)
as constituents in modified oligo-DNA duplexes have also been determined using X-ray
crystallography.
These data suggest that the cyclopentyl rings in carbocyclic nucleosides are prone to adopt a
greater variety of conformations than the pentofuranosyl counterpart in the corresponding natural N-
nucleosides. The greater flexibility of carbocyclic nucleosides compared with natural pentofuranosyl
nucleosides may be attributed to the absence of O4' oxygen in the former, whereas in the latter O4' is
involved in stereoelectronic interactions with the glycosyl nitrogen (i.e. the O4'-C1'-N1/9 anomeric
effect) and with O2' and/or O3' (i.e. through gauche effects). However, any possible contribution of
different solvation and electrostatic potential in the cyclopentyl compared with pentofuranosyl
analogues cannot be excluded. The conformational variations of cyclopentane moieties in carbocyclic
nucleosides in the solid state clearly suggested that a knowledge of their dependable solution
conformations is highly desirable in order to compare the solution and the solid state structures.
Hence, a modified Karplus equation correlating the 3JHH values to the corresponding proton-proton
torsion angle specifically in carbocylic nucleosides is now available (Eq 8b in Section 3.4.2).
∆G° of the protonation � deprotonation equilibrium drives the two-state N � S equilibrium
of the furanose moiety in N-30,36,37 and C-nucleosides32 through the modulation of the interplay of
gauche and anomeric effects (Sections 4 - 6). It has been found that the electronic nature of the
aglycone is reflected, through the anomeric effect, in ∆G° of the N � S equilibrium of the constituent
pentofuranose moiety. In fact, the pKa value of the constituent nucleobase can be independently
measured from a pD-dependent change of ∆G° of the N � S equilibrium, which is closely similar to
the pKa value obtained from the pD-dependent chemical shift plots. We have examined41 3JHH
of aristeromycin and its 2' and 3'-deoxy derivatives at pD ≈2 and ≈7 in the 278 K - 358 K range to
examine whether the protonation � deprotonation equilibrium of the adenin-9-yl base also drives the
pseudorotational equilibrium of their constituent cyclopentane moieties. As expected, temperature-
dependent 3JHH values are virtually unchanged from acidic to neutral pDs (± 0.15 Hz, Table 1 in ref.
41) suggesting that the change of the electronic nature of the aglycone in aristeromycin has no effect
on the drive of the constituent cyclopentane conformation, thereby establishing the role of the
anomeric effect in the drive of the sugar conformation in N- and C-nucleosides17,19-30,32,33,35-38,40-42.
The comparison of the Energy barrier of the pseudoroation in natural pentofuranosyl
nucleoside versus the carbocyclic counterpart.
Our ab initio calculations (Fig 7) at HF/3-21G* level with the GAUSSIAN 94 program298 on
various pseudorotamers (with defined P values in the 0 - 360° range at a constant Ψm for all
conformers of a particular compound) of aristeromycin (8), 2'-deoxyaristeromycin (9), 3'-
deoxyaristeromycin (10), 2',3'-dideoxyaristeromycin (11) and their natural pentofuranosyl
counterparts have allowed us to further shed light on the validity of the two-state N � S (or W)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
29
equilibrium model for the cyclopentane ring in carbocylic nucleosides (as suggested by the presence
of two major energy wells, vide infra) and to assess the effect of the substitution of CH2 for O4' upon
the energy barrier of pseudorotation in the modified versus natural nucleosides. This comparison is
intended to improve our understanding of the energy penalty of stereoelectronic gauche and anomeric
effects in natural nucleosides and nucleotides.
(i) 2',3'-dideoxynucleosides constitute the system of choice to analyze and compare the relative
flexibility of sugar conformation in carbocyclic versus natural pentofuranosyl nucleosides since 2'-
and 3'-OH groups are absent, which prevents any H-bonding interaction between themselves or with
the nucleobase [see subsections (iii) and (iv) below]. Both for 2',3'-dideoxyaristeromycin (11) and β-
D-ddA (30), N-type pseudorotamers (P = 0°) are slightly preferred over the S-type counterparts (P =
180°) [by ≈ 0.7 kcalmol-1 for 11 and ≈ 1.2 kcalmol-1 for 30 (Compare Panels (A) and (B) in Fig 7].
However, whereas for β-D-ddA, the activation energy barrier (∆∆G°‡) in the E region for the N to S
interconversion is ≈ 7.6 kcalmol-1 (see Section 2.6 for an upper limit of energy of activation of
pseudorotation found experimentally) the same transition in 2',3'-dideoxyaristeromycin is more facile
(∆∆G°‡ ≈ 3.7 kcalmol-1) which uniquevocally shows that the O4'-C1'-N9 anomeric effect efficiently
hampers free pseudorotation in the former. If we now for the sake of convenience consider that 2',3'-
dideoxyaristeromycin and β-D-ddA (30) are isosteric, then a simple subtraction of their above ∆∆G°‡ values gives us an idea of the influence of the anomeric effect on the energy barrier of the
pseudorotation in β-D-ddA (30), which is 3.9 kcalmol-1. For nearly all optimized pseudorotamers of
11 and 30, the nucleobase is in anti orientation and β is found in the anti range while γ+ rotamers are
preferred [Panels (I) and (J) in Fig 7].
(ii) For β-D-dA (37), among all possible pseudorotamers, N and S-type sugar conformations
(both having the same energy) are energetically preferred as shown by the energy profile [Panel (D) in
Fig 7]. The activation energy barrier from N to S through E is reduced (∆∆G°‡ ≈ 4.7 kcalmol-1) in
comparison with β-D-ddA (7.6 kcalmol-1, see panel (B) in Fig 7), showing the effect of 3'-OH. The
constituent nucleobase remains in the anti orientation and γt, ε3t and βt rotamers (see the legend of
Fig 7 for definition of the ε3 torsion angle) are preferred throughout the whole pseudorotational cycle
[Panel (L) in Fig 7]. For 2'-deoxyaristeromycin (9), among all pseudorotamers, the S- or W-type
conformers (210° < P < 240°) are energetically preferred by ≈ 3 kcalmol-1 over the N-type
counterparts (330° < P < 60°) [Panel (C) in Fig 7]. However, in contrast with the case of β-D-dA, the
interconversion between the energetically preferred N- and S-/W-type conformations takes place
through a very small barrier in the E-region (∆∆G°‡ ≈ 1.1 kcalmol-1) which is consistent with the
absence of the O4'-C1'-N9 anomeric and [O3'-C3'-C4'-O4'] gauche effects, as discussed for 2',3'-
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
30
Bas
eO
OR
O
OH
OH
Bas
e
Bas
eO
OH O
HO
H
O
OH
Bas
e
OB
ase
OH
OH
O
OR
'
Bas
eO
R
N
NNN
NH2
NN
N
NNH2
O
NH
NO
O
NH
NO
O
H3C
NH
NO
O
F
Bas
eO
OH
O
P
OH
−O O
Bas
eO
OH
O
P
O
−O O
H2
CC
H3
Bas
eO
OH
O
P
OH
−O O
OH
Bas
eO
OH
O
P
O
−O O
H2
C
CH3
OH
H
NN
N
NH
O
OO
H OH
OH
H
NN
N
N
NH2
OO
H OH
OH
H
N
N
N
NH2
O
OH O
HO
H
NN
H
NH2
OO
OH OH
OH
RN
NR
'
O
OO
OH O
HO
H
AO
OH
OH
NH
NNN
O
NH2
16:
R =
Me;
R' =
OM
e; R
" =
H
15:
R =
R"
= H
; R
' = O
Me
1
3:
R =
R"
= H
; R
' = O
H
14:
R =
H;
R' =
R"
= O
H
28: α
-L-d
C
cyto
sin
-1-y
l (C
)
Bas
e =
aden
yl-
9-y
l (A
)
25: α
-D-d
C (
R =
R' =
H)
9:
2'-d
eoxy-a
rist
ero
myci
n
(R
= O
H;
R' =
H)
10:
3'-d
eoxy-a
rist
ero
myci
n
(
R =
H;
R' =
OH
)
hyp
oxan
thin
-9-y
l (I
)
17: α
-D-d
dA
23:
3',5
'-d
iOM
e-α
-D-d
A (
R =
R' =
Me)
18: α
-D-d
dG
21
: α
-D-d
A (
R =
R' =
H)
22:
3'-O
Me-α
-D-d
A (
R =
Me;
R' =
H)
19: α
-D-d
dC
20: α
-D-d
dT
24
: α
-D-d
G (
R =
R' =
H)
27: α
-L-d
A
26
: α
-D-T
(R
= R
' = H
)
gu
anin
-9-y
l (G
)
52: β
-D-C
54: β-D
-U
34: β-D
-dd
T (
R =
H)
31: β-D
-dd
G (
R =
H)
29: α
-L-T
33: β-D
-dd
C (
R =
H)
50: β
-D-A
32:
5'-O
Me-β
-D-d
dG
(R
= M
e)
51: β
-D-G
47: β
-L-d
G
46: β
-L-d
A
53: β
-D-r
T
30: β-D
-dd
A (
R =
H)
γ
β
55:
5-F
-β-D
-U
35: β-D
-dd
U (
R =
H)
49: β
-L-T
imid
azo
l-
1-y
l (I
m)
thym
in-1
-yl
(T)
ura
cil-
1-y
l (U
)5
-F-u
raci
l-1
-yl
[5-F
-dU
(4
5)
& 5
-F-U
(5
5)]
Na+
64
: β
-D-d
AM
P
65: β-D
-dG
MP
66: β-D
-dC
MP
67: β-D
-TM
P
68: β-D
-dU
MP
ε
δ
β
74: β-D
-AM
P
48: β
-L-d
C
72: β-D
-TM
PE
t
γ
ζ
69: β-D
-dA
MP
Et
70: β-D
-dG
MP
Et
71: β-D
-dC
MP
Et
β
76: β-D
-CM
P
78: β-D
-rU
MP
α
77: β-D
-rT
MP
75: β-D
-GM
P
Na+
Na+
α
6
β
9
Na+
δ ζ
79: β-D
-rA
MP
Et
80: β-D
-rG
MP
Et
81: β-D
-rC
MP
Et
82: β-D
-rT
MP
Et
83: β-D
-rU
MP
Et
73: β-D
-dU
MP
Et
1 28
8
ε
75
3
7
4
5
9
1
57:
Fo
rmyci
n A
56:
Fo
rmyci
n B
58:
9-d
eaza
-A
6 3
1
42
5
3
721
9
6
8
2
4
5
6
5
3
60: ψ
-U:
R =
R' =
H12
6
59: ψ
-iso
C
44
61:
1-M
e-ψ
-U:
R
= M
e, R
' = H
3
63: β-D
-3'-d
A6
2:
1,3
-diM
e-ψ
-U:
R
= R
' = M
e
8:
Ari
ster
om
yci
n (
R =
R' =
OH
)
HO
OR
H
R'
R"
Bas
eO
OR
'
OR
AO
H RR
'
12:
R =
R' =
R"
= H
H6'
H6"
39
: 3
',5'-d
iOM
e-β-D
-dA
(
R =
R' =
Me)
37: β
-D-d
A (
R =
R' =
H)
38
: 3
'-O
Me-β
-D-d
A
(
R =
Me;
R' =
H)
41: β-D
-dG
(R
= R
' = H
)
42: β-D
-dC
(R
= R
' = H
)
NH
NNN
O
45
: 5
-F-β
-D-d
U (
R =
R' =
H)
44
: β-D
-dU
(R
= R
' = H
)
43: β-D
-T (
R =
R' =
H)
40: β-D
-dIm
(R
= R
' = H
)
36: β-D
-dd
I (R
= H
)1
1:
2',3
'-d
ideo
xy-a
rist
ero
myci
n
(
R =
R' =
H)
59
b: ψ
-iso
C, H
Cl
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
31
O
R F3"
H2"
B
H2'
H3'
H4'
H5"
H5'
H1'
O
OR1
H3"
OR
H2'
F3'
H4'
H5"
H5'
H1'
BO
OM
MT
r
F3"
H2"
B
HO
H3'
H4'
H5"
H5'
H1'
O
OH H
3"
H2"
B
F2'
H3'
H4'
H5"
H5'
H1'
O
OH
H3"
H2'
B
F2"
H3'
H4'
H5"
H5'
H1'
O
OR
F3'
H2"
B
H2'
H3"
H4'
H5"
H5'
H1'
CO
OC
H3
CO
OC
H3
HA
F
HB
O
CO
OH
CO
OH
HA
F
HB
R1
R3
R2
R4
Cl
Cl
Cl
Cl
Cl
Cl
F
F
F
F2
H4
FF
F
F
F
F
F
O
OH O
H
F2'
B
F2"
H3'
H4'
H5"
H5'
H1'
N
H2ax
H2eq
H4ax
H4eq
F
R
H
HN
H6ax
H6eq
H4ax
H4eq
F
H
H
CO
OH
N
H2ax
H2eq
H4ax
H4eq
F
CO
O
H
Me
Me
Me
FH
O
Hax
Me
FM
e
Me
HB
HA
OH
O
OH O
H
H2'
B
F2"
H3'
H4'
H5"
H5'
H1'
O
OR
F3'
H2"
B
H2'
H3"
H4'
H5"
H5'
H1'
G =
guan
in-9
-yl
96
: B
= U
(F
2'd
dU
) 88
: R
= O
H, B
= T
(F
LT
)
115
: X
= N
O2
92
: R
= R
1 =
H, B
= A
(F
XA
)
90
: R
= N
H2, B
= T
(A
FL
T) 99
91
: R
= O
H, B
= U
(F
3"d
dU
)
98
:R =
H, B
= U
(F
3'd
dU
)
93
: R
= H
; R1 =
MM
Tr,
B
= A
(F
XA
5)
97
: B
= U
(F
2"d
dU
)
89
: R
= O
Tr,
B =
T (
FL
T5)
104
C =
cyto
sin-1
-yl
A =
aden
in-9
-yl
T =
thym
in-1
-yl
103
: R1 =
FA
; R2 =
FB
;
R3 =
HA
; R4 =
FB
; 100
94
: R
= C
S-O
Ph;
R1 =
MM
Tr,
B =
A (
FX
A25)
101
: R1 =
FA
; R2 =
HB
;
R3 =
HA
; R4 =
FB
;
95
: B
= T
(F
3A
T)
102
: R1 =
HA
; R2 =
HB
;
R3 =
FA
; R4 =
FB
;
U =
ura
cil-
1-y
l
85
: B
= G
(diF
G)
86
: B
= T
(diF
T)
110
109
112
: B
= C
(F
2"C
)
87
: B
= C
(diF
C)
105
: R
= H
106
: R
= C
OO
H
108
84
: B
= A
(diF
A)
107
O
OH X
TH5"
H5'
114
: X
= O
Me
116
: X
= O
CF3
111
:R =
H, B
= A
(F
3'd
dA
)
113
: X
= N
H2
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
32
(A)
2',3'-did
eo
xya
riste
rom
ycin
(11
)
P (o
)
-60
060
120
180
240
E (kcal mol-1
)
048
12
16
(C)
2'-d
eoxyari
ste
rom
ycin
(9
)
P (o
)
-60
060
120
180
240
E (kcal mol-1
)
048
12
16
(E)
3'-de
oxya
riste
rom
ycin
(10
)
P (o
)
-60
060
120
180
240
E (kcal mol-1
)
048
12
16
(G)
ari
ste
rom
ycin
(8)
P (o
)
-60
060
120
180
240
E (kcal mol-1
)
048
12
16
(B) β
-D-d
dA
(30
) P (o
)
-60
060
120
180
240
E (kcal mol-1
)
048
12
16
(D) β-D
-dA
(37)
P (o
)
-60
060
120
180
240
E (kcal mol-1
)
048
12
16
(F) β
-D-3
'-d
A (63
) P (o
)
-60
060
120
180
240
E (kcal mol-1
)
048
12
16
(H) β-D
-A (50
)
P (o
)
-60
060
120
180
240
E (kcal mol-1
)
048
12
16
(I)
2',3
'-d
ide
oxya
riste
rom
ycin
(11
)
P (o
)-60
060
120
180
240
Torsion angle (o)
-240
-180
-120
-600
60
χ
γ
β
(J) β-D
-dd
A (30
) P (o
)-60
060
120
180
240
Torsion angle (o)
-240
-180
-120
-600
60
χ
γ
β
(K)
2'-d
eoxya
riste
rom
ycin
(9
)
P (o
)-60
060
120
180
240
Torsion angles (o)
-240
-210
-180
-150
-120
-90
-60
χ
γβε 3
(L) β
-D-d
A (37
)
P (o
)-60
060
120
180
240
Torsion angle (o)
160
170
180
190
200
χγ
βε 3
(N) β
-D-3
'-d
A (63) P (o
)
-60
060
120
180
240
Torsion angles (o)
-240
-180
-120
-600
60
d(H-bond) (Å)
2345
d(2'-OH-N
3)
d(5'-OH-O
4')
χ
γβε 2
(M)
3'-d
eo
xya
riste
rom
ycin
(10
)
P (o
)
-60
060
120
180
240
Torsion angles (o)
-240
-180
-120
-600
603456
d(2'-OH-N
3)
χ γβ
ε 2
d(H-bond) (Å)
(P) β
-D-A
(50)
P (o
)-60
060
120
180
240
Torsion angles (o)
-240
-180
-120
-60
d(H-bond) (Å)
246
d(3'-OH-O
2')
χγ
β
ε 2
ε 3
d(2'-OH-O
3')
d(5'-OH-O
4')
d(2'-OH-N
3)
(O)
ari
ste
rom
ycin
(8
)
P (o
)-60
060
120
180
240
Torsion angles (o)
-240
-180
-120
-600
60
d(H-bond) (Å)
246
χ γβ
ε 2
d(2'-OH-N
3)
d(2'-OH-O
3')
d(3'-OH-O
2')
ε 3
Fig
ure
7.
The
plo
ts o
f th
e en
ergy (
E),
of
var
ious
tors
ion a
ngle
s an
d o
f in
tera
tom
ic d
ista
nce
s fo
r a
b i
nit
io o
pti
miz
ed p
seud
oro
tam
ers
of
2',3
'-d
ideo
xyar
iste
rom
yci
n (11
) [P
anel
s (A
) an
d
(I)]
, 2
'-d
eoxyar
iste
rom
yci
n (9
) [P
anel
s (C
) an
d (K
)],
3'-d
eoxyar
iste
rom
yci
n (10
) [P
anel
s (E
) an
d (M
)],
aris
tero
myci
n (8
) [P
anel
s (G
) an
d (O
)] a
nd
of
thei
r p
ento
fura
no
syl
counte
rpar
ts
β-D
-dd
A (30
) [P
anel
s (B
) an
d (J
)], β
-D-d
A (37
) [P
anel
s (D
) an
d (L
)], β
-D-3
'-d
A (63
) [P
anel
s (F
) an
d (N
)] a
nd
β-D
-A (50
) [P
anel
s (H
) an
d (P
)] a
s a
funct
ion o
f th
e p
has
e an
gle
of
pse
ud
oro
tati
on (
P).
The
calc
ula
tio
ns
hav
e b
een p
erfo
rmed
by c
onst
rain
ing o
nly
tw
o e
nd
ocy
clic
to
rsio
ns
(ν0 a
nd
ν4)
to t
he
app
rop
riat
e val
ues
in o
rder
to
sw
eep
P f
rom
0 t
o 3
60°
in 3
0°
(fo
r 8
- 10
, 37
, 50
and
63
or
20°
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
33 (for 11 and 30) intervals at a common Ψm value [41° for 8 - 11 (as in the crystal structure292 of 8), 40° for 30 and 36° (as
in the crystal structure of adenosine259) for 37, 50 and 63] whereas the rest of the molecule has been freely optimized. For
all conformers, the starting geometries used as input for the optimization with GAUSSIAN program298 were as follows (ε2
and ε3 designate the [2'-OH-O2'-C2'-C1'] and [3'-OH-O3'-C3'-C4'] torsion angles, respectively): χ = 120°, β = γ = 180° for
2',3'-dideoxyaristeromycin (11); χ = -170°, β = γ = 180° for β-D-ddA (30); χ = 120°, β = γ = ε3 = 180° for 2'-
deoxyaristeromycin (9); χ = -170°, β = γ = ε3 = 180° for β-D-dA (37); χ = 120°, β = γ = ε2 = 180° for 3'-
deoxyaristeromycin (10); χ = -170°, β = γ = ε2 = 180° for β-D-3'-dA (63); χ = 120°, β = γ = ε2 = 180° and ε3 = -60° for
aristeromycin (8) and χ = -170°, β = γ = ε2 = 180° and ε3 = -60° for β-D-A (50).
dideoxyaristeromycin under the subsection (i). It is also noteworthy that whereas the values of the γ, ε3
and β torsion angles (in the anti domain) do not change drastically from one pseudorotamer of 2'-
deoxyaristeromycin to the other, the nucleobase adopts a syn orientation in the high energy W-type
pseudorotamers while for all other sugar conformations it is anti [Panel (K) in Fig 7]. (iii) For 3'-
deoxyaristeromycin (10), the W-type (P = 240°) pseudorotamers are preferred among all possible
geometries and they are more stable by ≈ 6 kcalmol-1 than the N-type counterparts giving rise to a
local energy minimum at P ≈ 330° [Panel (E) in Fig 7]. The interconversion between N- and S-type
pseudorotamers is possible, provided that the energy barrier in the E-region (∆∆G°‡ ≈ 4.3 kcalmol-1) is
overcome. It is noteworthy that the preference for W-type (P = 240°) geometries of 3'-deoxyaristeromycin results from the gauche+ orientation of 2'-OH hydroxy group (ε2 ≈ 60°) which
enables its hydrogen-bonding interaction [d(2'-OH-N3) ≈ 2.5 Å] with the nucleobase in antiperiplanar
orientation [Panel (M) in Fig 7]. On the other hand, the relatively higher energy of other
pseudorotamers in the W-region (P = 270° and P = 300°) is associated with a syn orientation of the
nucleobase. The energy profile presented in Panel (F) for the natural counterpart β-D-3'-dA (63) is far
more complex than that of 3'-deoxyaristeromycin and suggests a strong preference for E-type (P = 90°)
or S-type (P = 180° or 210°) pseudorotamers (by ≈ 5 - 5.5 kcalmol-1) over the local energy minimum in the N-region (P = 0°). The preference for E-type (ε2
+ with β-) and S-type (ε2+ with β+) pseudorotamers
is owing to a stabilizing H-bonding interaction between 2'-OH proton and N3 [d(2'-OH-N3) ≈ 2.0 Å] and
between 5'-OH and O4' [d(5'-OH-O4') ≈ 2.0 Å] (for E-type conformations) [Panel (N) in Fig 7].
(iv) For aristeromycin (8), the S/W type conformers (120° < P < 240°) are preferred over the N-
type (330° < P < 30°) geometries by ≈ 10 kcalmol-1 [Panel (G) in Fig 7]. The N- to S/W-type
geometries interconversions preferably take place through the activation energy barrier in the E (∆∆G°‡
≈ 4.8 kcalmol-1 from N to S; ≈ 13.0 kcalmol-1 from S to N) rather than the W (∆∆G°‡ ≈ 5.8 kcalmol-1
from N to S; ≈ 14.1 kcalmol-1 from S to N) region of the pseudorotation cycle. The increased stability
of S/W-type pseudorotamers with respect to other geometries is due to 2'-OH....N3 and 3'-OH....O2' hydrogen-bonding interactions which are permitted by ε2
+ and ε3t orientations [Panel (O) in Fig 7]. On
the other hand for P values in the range 270° < P < 90°, a single H-bonding interaction is possible
between 2'-OH and O3'. As for 2'-deoxy and 3'-deoxyaristeromycin (see (ii) and (iii)), in the relatively
unstable conformers in the W-region (P = 270° and P = 300°), a syn orientation of the nucleobase is
observed, whereas for the energy minima the nucleobase is anti. In fact, the energy profile [Panel (G)
in Fig 7] and the dependence of the glycosyl torsion angle χ on the value of P appear to be very similar.
The plot presented Panel (H) in Fig 7 shows that the energy of various pseudorotamers of β-D-A (50)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
34
is much less sensitive to the phase angle value than for the aristeromycin counterpart [Panel (G) in Fig
7], possibly because of the fact that all pseudorotamers are stabilized by the H-bonding interaction
between 2'-OH and O3' [Panel (P) in Fig 7]. The absolute energy minimum is found for N-type
conformations of β-D-A (P = 0°) which are preferred over secondary minima in the E/S-regions at P =
120° (by 0.9 kcalmol-1) and P = 210° (by 2.3 kcalmol-1). The activation energy barriers between N-
and E- (P = 120°) (through P = 60°: 5.5 kcalmol-1 and 4.6 kcalmol-1) and from E- to S- (P = 210°)
(through P = 150°: 3.5 kcalmol-1 and 2.2 kcalmol-1) are reduced in comparison with what was found
for aristeromycin. For all pseudorotamers, γt rotamers are preferred and the nucleobase is in anti orientation with ε2
t for 270° < P < 90° and ε2- for 90° < P < 270°. The preference for P = 0° among all
pseudorotamers in the N region correlates nicely with the fact that in this situation ε2 is in the maximal
trans orientation. However, this trans orientation does not result into an H-bond between 2'-OH and N3 as shown by the fact that the d(2'-OH-N3) distance is greater than 4 Å at any P value. On the other hand,
the presence of the secondary energy minimum at P = 120° may be attributed to a hydrogen-bond
between 5'-OH and O4' permitted by the β- orientation.
In summary, the energy plots in Fig 7 show the energy penalties induced by some specific
conformational properties of the cyclopentane ring owing to the absence of O4' in carbocyclic
nucleosides vis-a-vis natural pentofuranosyl nucleosides, in which O4' is the cause of both gauche and
anomeric effects. Work is now in progress in our laboratory to perform the above calculations at a
higher basis set in order to delineate the influence of hydration on the pseudorotational energy and
conformational profile in the carbocylic as well as in pentofuranosyl nucleosides.
2.6 Energy barriers of the pseudorotation cycle of β-D-nucleosides
Early quantum chemical, CNDO or classical potential energy calculations250,299-302 and more
recent calculations using consistent force field method303 (permitting bond stretching and bond angle
bending) on unsubstituted ribo and deoxyribofuranose have shown that the N-type and S-type sugars
are preferred among all pseudorotamers, and that they have almost the same energy. Calculations on 1-
amino-ribose, -2-deoxyribose or -3-deoxyribose304,305 and on nucleosides themselves306 have shown
that the activation energy barrier for N- to S- interconversions is clearly greater in the W (≈ 24 kJ/mol
for 2-deoxyribofuranose, ≈ 31 kJ/mol for ribofuranose) than in the E region (≈ 7.5 kJ/mol for 2-
deoxyribofuranose, ≈ 16 kJ/mol for ribofuranose).
The measurement307 of 13C longitudinal relaxation times of single tertiary carbons showed that
the energy of activation of the N- to S-type sugar interconversion in purine nucleosides is ≈ 20 ± 2
kJ/mol, whereas for pyrimidines the internal motions are slow compared with the rotational diffusion of the whole molecule, as a result of a possible H-bonding between the 5'CH2OH group and the base,
and the authors suggested that this might raise the apparent activation energy of pseudorotation for
uridine and cytosine above 25 kJ/mol.
A recent temperature-dependent 2H and 13C relaxation study from our laboratory308 on
selectively deuterated thymidines and allofuranoses as well as their comparisons with the
conformationally constrained analogues and abasic sugars did not allow us to determine the activation
energy barrier of pseudorotation because the internal motions are heavily coupled with the overall
molecular reorientations, which prevents dissection of the observed activation energy barrier of 20-23
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
35
(±0.9) kJmol-1 into contribution from pseudorotational interconversions, rotation around the glycosyl
and C4'-C5' torsions and overall tumbling. Nevertheless, this experimental estimate308 gives an upper
limit for the pseudorotational barrier (see Appendix for detailed discussion on the estimates of the
barrier).
2.7 Steric effect of the nucleobase on the sugar conformation
The inspection of the three molecular fragments in nucleos(t)ides (i.e. the aglycone, the sugar
and the phosphate) allows us to estimate the bias of the two-state N � S pseudorotational equilibrium
of the constituent pentofuranose moiety on the basis of various steric and stereoelectronic effects.
The N-nucleobase influences the pentofuranose conformation through its inherent steric effect
and counteracting stereoelectronic interactions within O4'-C1'-N1/9 fragment. In β-D-nucleosides,
among all pseudorotamers, steric repulsions penalize O4'-exo (W-type) conformation304-306,309 most
owing to the 1,3-diaxial orientation of the nucleobase and 5'CH2OH group and to the eclipsed
arrangement of C2' and C3' substituents (Fig 6). In O4'-endo (E-type) pseudorotamers, the steric
repulsions between the nucleobase and the 5'CH2OH group are minimized, however the substituents at
C2' and C3', just as in the case of W-type conformers, are in the unfavorable eclipsed orientation (Fig.
6). Conformational analyses of nucleos(t)ides in aqueous solution are performed on the basis of a two-
state N �S equilibrium model (vide supra). In terms of steric effect alone, for β-D-nucleosides, S-type
pseudorotamers are energetically favoured in comparison with N-type counterparts, since the
pseudoequatorially oriented nucleobase in the former exerts less steric repulsions with the other
substituents on the pentofuranose moiety than when it is pseudoaxial in the later (Figs 6 and 8A).
In α-D-nucleosides, the nucleobase and the 5'CH2OH group are on opposite faces of the
pentofuranose sugar, therefore their steric interactions will be minimal in comparison with the situation
in β-nucleosides. In N-type conformers, the nucleobase and 3'-OH are pseudoequatorial while 2'-OH is
pseudoaxial, whereas the opposite is true for S-type pseudorotamers. Both in W- and E-type
conformers, 2'-OH and 3'-OH are pseudoaxial, whereas the nucleobase is pseudoequatorial in W-type
geometry and pseudoaxial in the E-type counterpart. In α-D-2'-deoxynucleosides, the nucleobase exerts
stronger steric repulsions with 3'-OH and H4' in N-type than in S-type pseudorotamers. Therefore, in
solution, taking into consideration of a two-state N � S equilibrium model, S-type pseudorotamers
would be energetically favoured over the N-type counterparts if the steric effect alone were considered.
Potential energy calculations on α- and β-2'-deoxy (or ribo) adenosine and thymidine have
shown that syn conformations of the nucleobase in the α-anomers are energetically less favourable over
the entire range of phase angle values, and that the energy barrier of anti � syn interconversions is
higher in the α-ribo than in the α-2'-deoxy series. Highly hypothetical syn orientations of a nucleobase
in α-2'-deoxynucleosides may be stabilized by hydrogen-bonding interactions between N3 (or O2) and
3'-OH.
From the previous observations, it can be derived that as the bulk of the nucleobase is
increased, one expects the bias of the two-state N � S pseudorotational equilibrium to be shifted
toward more S-type and N-type conformations in β- and α-nucleosides, respectively.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
36
2.8 O4'-C1'-N1/9 stereoelectronic effect (the anomeric effect)
The anomeric effect in heterocyclic six-membered rings is commonly explained in terms of
stabilizing hyperconjugative interactions (molecular orbital overlap model)115,116 or, alternatively, by
destabilizing dipole-dipole repulsions65,117,118 (Section 1.3 and Table 1 for the energetics of the
anomeric effect in 2-substituted hexopyranoses and derivatives).
Similarly, the O4'-C1'-N1/9 anomeric effect in nucleosides and nucleotides may be explained
either (i) by stabilizing nO4'
→σ∗C1'-N1/N9 orbital interactions between the orbital of one of the
endocyclic O4' electron lonepairs (nO4'
) and the antibonding orbital of the C1'-N1/9 glycosyl bond
(σ∗C1'-N1/N9
) (Fig 9), or (ii) by destabilizing electrostatic repulsions between two dipoles [i.e. the dipole
of the furanose ring, on one hand (which is itself the resultant of C4'-O4', O4'-C1' individual dipole
moments and of the dipole induced by O4' lonepairs) and the dipole oriented from C1' to N1/9, on the
other (Fig 8B-C)].
In β-D-nucleosides, O4'-C1'-N1/9 stereoelectronic interactions are most efficient in W-type
sugar geometries, where one of the O4' lonepairs is in optimal antiperiplanar orientation with respect to
the C1'-N1/9 bond, and at the same time dipole-dipole repulsions are minimal since the angle between
both dipoles is nearly 90°. However, for the steric reasons stated above, West-type conformers have
neither been found in crystal structures of β-D-nucleos(t)ides nor in the conformational equilibria in
solution. Conversely, in E-type pseudorotamers, both O4' lonepairs are gauche with respect to the C1'-
N1/9 bond, which makes hyperconjugation (or molecular orbital overlap) least efficient;
Simultaneously, dipole-dipole repulsions are maximal, owing to the fact that both dipoles are nearly
parallel.
Figure 8: Steric effect of the base and
rationalization of the anomeric effect in
nucleos(t)ides in terms of electrostatic repulsions,
as examplified for β-D-dA (37). (A) The drive of
the two-state N �S equilibrium toward S- over
N-type pseudorotamers by the steric effect of the
nucleobase. (B) & (C) Stronger electrostatic
repulsions between the pentofuranose ring dipole
(black arrow) and the C1'-N9 dipole (white
arrow) in S-than in N-type sugars. Both dipoles
are nearly parallel in the S-type pseudorotamers
whereas they are nearly perpendicular in N-type
sugars (β << α, Panel C).
The anomeric effect is therefore often
invoked as one of the factors responsible
for the activation energy barrier
encountered in the East region of the pseudorotational cycle for β-D-nucleosides, which, besides the
stronger barrier in the West region, also hampers free pseudorotation of the constituent pentofuranose.
The comparison of the geometries of N- and S-type pseudorotamers of β-D-nucleosides shows
that dipole-dipole repulsions (Fig 8B-C) and O4'-C1'-N1/9 hyperconjugative interactions (Fig 9) are
OHOH2C
O
HOH2C
OH
HO
NN
N
N
N
N
N
N
NH2
NH2
H1'
N
C2'
C4'
H1'
N
C2'
C4'
O4'C
O
OH
HO
NN
N
N9
N
N
N
N9
NH2
NH2
H
HO
HH
H
HH
H
H
CHO
HH
(B)
(C)
(A)
αβ
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
37
reduced and enhanced, respectively, in the former compared to the latter. Therefore, O4'-C1'-N1/9
stereoelectronic interactions drive the two-state N � S pseudorotational equilibrium in β-D-
nucleosides, in aqueous solution, toward N-type conformations.
Figure 9. Rationalization of the O4'-C1'-
N1/9 anomeric effect in nucleos(t)ides in
terms of molecular orbital overlap and
hyperconjugation, as examplified for β-D-
dA (37). The O4' lonepairs orbitals are
represented using either the sp2 [i.e. higher
energy 1nsp2(p-type)
lonepair with
predominant p-type character and lower
energy 2nsp2(s-type)
lonepair with
predominant s-type character] or sp3 [i.e. 1n
sp3 and 2nsp3
lonepairs with the same
energy] hybridization models123-126,129,130
(Section 1.9). (A) Molecular orbital overlap
model116: The overlap (i.e. nO4' σ*C1'-N9)
of a lonepair orbital of O4' [1nsp2 (p-type)]
with the σ*C1'-N9 antibonding orbital of the
glycosyl bond results in the stabilization of
N- over S-type pseudorotamers and shifts
the pseudorotational equilibrium toward N.
(B) & (C) Influence of sp3 (i.e. 1nsp3
and
2nsp3
, Panel (B)) versus sp2 (1nsp2
(p-type)
and 2nsp2
(s-type), Panel (C)) hybridization of
the O4' lonepairs orbitals on their relative
orientation with respect to the glycosyl bond
and on the efficiency of the O4'-C1'-N9 stereoelectronic interactions. For a typical N-type pseudorotamer (the angles in the
projections are calculated for P = 0° and Ψm = 40°, assuming perfect trigonal symmetry), the nO4' σ*C1'-N9 interaction is
possible due to the near antiperiplanar orientation of either 1nsp3
(β1 ≈ 132°) or 1n
sp2 (p-type) (β2 ≈ 162°) with respect to
σ*C1'-N9 orbital, whereas in the S-type sugar counterpart (the angles are calculated for P = 160°, Ψm = 40°) the efficiency of
the interaction is much reduced as the result of the more accute β1 and β
2 angles compared to those in the N-type sugar. (D)
Double-bond � No-bond resonance resulting from hyperconjugation of one of the O4' lonepairs to the glycosyl C1'-N9
bond115. This model is consistent with the shortening of O4'-C1' bond compared to O4'-C4' bond observed in the crystal
structures174,310,311 of β-nucleos(t)ides. (E) This overlap, which results in the formation of a stabilizing (occupied) and a
destabilizing (unoccupied) orbitals, becomes more efficient as the square of the overlap (S) between both orbitals increases
and as the difference (∆E) between their respective energies decreases4,5. A maximal interaction requires an antiperiplanar
orientation of nO4' relative to the C1'-N9 bond.
The significant change (Fig 2E)1,121,122 of the endocyclic and exocyclic C-O bond lengths in
combination with concommitant closing or opening of the corresponding bond angles in crystal
structures of chair conformers of various 2-substituted-tetrahydropyrans with axially versus
equatorially oriented anomeric substituents is considered as a strong evidence for the hyperconjugative
origin of the anomeric effect in such systems. A perusal174,310,311 of the crystal structures of β-D-N-
nucleosides also shows that the C1'-O4' bond is 0.03 - 0.04 Å shorter than the C4'-O4' bond, which
OHOH2C
O
HOH2C
OH
HO
N
N
N
NH2
C2'
N9
H1'O4'
C4'
C2'
N9
H1'
O4'
C4'
C2'
N9
H1'O4'
C4'
C2'
N9
H1'
O4'
C4'
N9
OHOH2C
HOO
HOH2C
HO
(A)
(B)
(C)
(D)
β1 = 132ο
β1 = 96ο
β2 = 162ο
β2 = 126ο
σ*C1'-N9
δ+
δ−
2nsp3
1nsp2 (p-type)
2nsp2 (s-type)
1nsp2 (p-type)
1nsp3
1nsp3
2nsp3
2nsp2 (s-type)
1nsp2 (p-type)
1nsp2 (p-type)
(E)nO4'
σ*C1'-N9
ΔE(σ*C1'-N9 - nO4')
Stabilization due to anomeric effect (AE)
N
N
N
NH2
N9
N
N
N
NH2
N9
NN
N
NH2
N9
N-type sugar S-type sugar
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
38
goes hand in hand with a possible hyperconjugative origin for the anomeric effect in nucleosides (Fig
9).
2.9 Effect of the aglycone on the conformation of nucleos(t)ides and oligos
History of the discovery of stereoelectronic effects in nucleos(t)ides: The first set of eight
papers qualitatively indicating the influence of the electronic character of the C1'-aglycone on the
conformation of the constituent pentofuranose sugar in nucleos(t)ides were the result of almost
simultaneous independent observations: (i) That the UV spectrum of 1-methylcytosine and cytosine
nucleosides varies with the nature of the carbohydrate component is the first milestone paper that
established the cross-talk between the aglycone and the constituent sugar312. (ii) Guschlbauer et al
showed313 that the confomation of the ribose moiety changes with the protonation of the guanin-9-yl
base at N7 in 2'-GMP. (iii) As shown314 by Sarma et al., the sugar conformation is indeed different in
oxidized and reduced β-nicotinamide mononucleotides. (iv) Remin and Shugar showed315 that the
protonation of cytidine or arabinocytidine changes the 3J1'2' value by about 0.2 Hz. (v) Altona and
Sundaralingam did an important qualitative correlation on the nature of sugar pucker depending on the
nature of the aglycone basing on their X-ray crystal structures data181.
(vi) Subsequently, Altona and Sundaralingam extended their X-ray correlation data to the
coupling constant correlation, showing again the qualitative dependence of 3J1'2' with the nature of the
aglycone316. (vii) The preference of S versus N conformer is pH-dependent, as demonstrated for the
first time by Guschlbauer et al on guanosine phosphates317. (viii) A similar observation was also made
by Hruska et al on 2'-O-methyladenosine318.
Further qualitative evidences:
In 1987, we first showed14 basing on thermodynamic estimates on a set of four isomeric 2'/3'-
deoxynucleosides that the net result of the gauche effect and the anomeric effect is of major importance
in determining the overall furanose conformation. Beside the subsequent quantitative works from our
labs (Sections 4 - 6), many qualitative studies have been conducted to understand how the electronic
nature,309,319-323 protonation state198, bulkiness or substitution pattern207,239,324-327 and configuration
of the nucleobase175,328-346 or of the C1 substituent234 modulate the sugar conformation in nucleosides
and nucleotides as well as in furanosides347-351. The conformational analysis46 in solution of a dimer
containing a 4'-oxofuran derivative352, based on pseudorotational analysis of vicinal 3JHH, has shown
that the modified nucleoside adopts almost exclusively (89 %) S-type puckered geometries, as a result
of the cooperative drive by the O5'-C4'-O4' anomeric effect and the [O3'-C3'-C4'-O4'] gauche effect
(vide infra).
2.9.1 Configuration-dependent sugar conformation in furanosides
At the sugar level, spectroscopic studies based upon the interpretation of 3JHH, 3JCH and 3JCC
coupling constants347,349,353 have shown that the α- and β-anomers of D-ribofuranose (and of their 1-
methyl derivatives) adopt preferentially S- and N-type puckered geometries, respectively, owing to the
O1-C1-O4 anomeric effect, which places the anomeric group in pseudoaxial or nearly pseudoaxial
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
39
orientation. The contribution of the exo-anomeric effect to the relative stabilities and preferred
conformations of α- and β-D-erythrofuranose and -threofuranose, as obtained from ab initio
calculations, has also been adressed348.
2.9.2 Effect of electron-withdrawing nucleobases on pseudorotamer populations
1H-NMR studies have shown that in 2-substituted tetrahydropyran, the magnitude of the O-C-O
anomeric effect increases as the anomeric group becomes more electronegative (Section 1.5). In
nucleosides and nucleosides, similar trends have also been found:
(i) Egert et al., in their integrated approach321 consisting of an analysis of crystallographic data
and 1H-NMR spectra as well as quantum-chemical calculations on 5-substituted uridines, have shown
that electron-withdrawing substituents at C5 induce a lengthening and shortening of N1-C1' and C1'-
O4' bonds, respectively, in comparison with the reference compound, uridine. For electron-donating
substituents, the opposite situationwas observed. They invoked a possible charge transfer from one of
the O4' lonepairs to the C5 substituent to rationalize these tendencies. We, on the other hand, argue that
electron-withdrawing (or electron-donating) substituents at C5 are expected to strengthen (or weaken)
the nO4' →σ∗C1'-N9 interactions (i.e. the anomeric effect), which would in turn also lead to the
experimentally observed lengthening (shortening) and shortening (lengthening) of C1'-N1 and O4'-C1'
bond lengths, respectively. In that work, the concomitant opening (closing) of the glycosyl torsion
angle χ in 5-substituted uridines with electron-withdrawing (electron-donating) substituents at C5,
respectively, with respect to uridine, was explained by the fact that small χ values favour nO4' →πC5-C6
and nO4' →π*C5-C6 interactions, which are destabilizing and stabilizing, respectively. For electron-
withdrawing at C5, the stabilizing nO4' →π*C5-C6 interactions are strengthened while the destabilizing
nO4' →πC5-C6 interactions is weakened (in comparison with the reference compound uridine), leading
to the overall stabilization of the system, therefore such substituents favour small values of χ. In
contrast, electron-donating groups at C5 will induce a strengthening of the destabilizing nO4' →πC5-C6
interactions and a weakening of the nO4' →π*C5-C6 interactions (i.e. an overall destabilization of the
system) and in turn greater χ values will be found in comparison with uridine. A qualitative correlation
between the preference of the sugar moiety in 5-substituted uridine derivatives for N-type versus S-type
geometries and the value of χ was also proposed.
(ii) In their conformational study on 5-substituted uridine derivatives in aqueous solution by
1H-NMR spectroscopy, Uhl et al.322 have shown that the population of N-type pseudorotamers (% N)
increases from 44% to ≈ 90% going from NH2 to NO2 substituent at C5, and a plot of % N as a
function of the Hammett constant, σp, of the substituent at C5, gave a straightline. This is also
consistent with our proposition that electron-withdrawing (electron-donating) groups at C5 are
expected to strengthen (weaken) nO4' →σ∗C1'-N9 interactions (i.e. the anomeric effect) and therefore
stabilize (destabilize) the N- over S-type pseudorotamers. This is clearly evident from our quantitative
analysis showing that the protonation of the nucleobase in adenosine, guanosine and cytidine
nucleosides and in their deoxy derivativies strengthen the anomeric effect (i.e. increase of N-type
conformation) , whereas it is weakened by the deprotonation (i.e. decrease of N-type conformation),
compared to the neutral state (see Section 4 for detailed study).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
40
(iii) We have quantitatively shown that the sugar conformation can be steered by altering the
electronic character of the nucleobase as well as by the change of the electronegativity of the sugar
substituent (see section 4). Subsequently, Rosemeyer and Seela have confirmed this by showing225 that
the extent of the preference for N-type pseudorotamers of 7-substituted 7-deaza-2'-deoxyadenosines
increases as the Hammett contant σm of the 7-substituent increases. This result is consistent with the
expected increase in the strength of the O4'-C1'-N9 anomeric effect as the glycosyl nitrogen becomes a
better electron-acceptor.
2.9.3 Effect of the protonated nucleobase on pseudorotamer populations
The protonation of a nucleobase in a nucleotide directly controls its hydrogen-bonding
capabilities, therefore the overall three-dimensional structure of oligonucleotides is dictated by the pH
of the medium. For instance, the local DNA triple helix formed354,355 by the binding of a natural
homopyrimidine oligonucleotide to a target DNA duplex is much more stable in the acidic solution
than at neutral pH, owing to the fact that Hoogsteen base-pairing with the third pyrimidine strand
requires that the cytosin-1-yl nucleobases be protonated. However, substitution of cytidine in the
Hoogsteen strand for 5-methylcytosine or 5-bromouridine allows to increase its affinity for the DNA
duplex under physiological pH356,357. pH-dependent conformational transitions and stabilities of C.A
and G.A mismatches in DNA358-362 have been extensively studied. It has also been suggested363,364
that although the secondary structure of the oligoDNA d(A+-G)10 is presumably helical, it is stabilized
not by stacking bases or hydrogen-bonding base pairs but instead by ionic bonds between positively
charged 2'-deoxyadenosine residues and distal negatively charged phosphates. The formation of
unusual parralel double-stranded DNA duplex365-367 or four-strand tetrads (the i-motif)368,369 with two
parallel stranded base-paired duplexes at low pH has also been experimentally evidenced.
The acid-base character of nucleobases in nucleosides and nucleotides varies widely165,247,370-
381: Whereas adenosine165 (pKa ≈ 3.5) and cytidine165 (pKa ≈ 4.2) can be easily protonated in the
acidic solution at N1 and N3, respectively, uridine165 (pKa ≈ 9.4), 5-fluoro-2'-deoxyuridine382-384 (pKa
≈ 7.8) and thymidine165 (pKa ≈ 9.9) are deprotonated at N3 in alkaline solution. Guanosine is either
protonated379 at N7 (pKa ≈ 1.9) in the acidic medium or deprotonated165 at N1 in the alkaline solution
(pKa ≈ 9.4). pKa values for C-nucleosides are known for formycin A385,386 (pKa = 4.4 (protonation at
N3) and 9.6 (deprotonation at N7)), formycin B387,388 (pKa = 8.8 (deprotonation at N1) and 10.4
(deprotonation at N7)), pseudoisocytidine389 (pKa = 3.7 (protonation at N1) and 9.0 (deprotonation at
N3)) and pseudouridine390 (pKa = 9.0 (mixed deprotonation at N1 and N3)). We have recently reported
for the first time34 the pKa values corresponding to protonation at N3 in 9-deazaadenosine (pKa = 6.0)
and formycin B (pKa = 1.3). Most pKa values of nucleobases in nucleos(t)ides are rather different from
the physiological pH, therefore one might conclude that pH-induced conformational transitions are not
likely to occur in DNA and RNA near physiological pH (≈ 6.8 - 7.3), unless the pKa of a particular
nucleobase changes drastically as a result of change of the microenvironment359,360,363,369,396-398. This
can be examplified by the pKa values of adenin-9-yl and cytosin-1-yl moieties, which are significantly
different at some oligonucleotides than those found in their monomers: For example, the pKa of
adenin-9-yl in the A25 residue (located close to the cleavage site in a lead-dependent ribozyme) is
6.5397,398 which is unusually high compared with the pKa of adenin-9-yl in adenosine (3.5). Clearly,
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
41
any slight deviation of the medium from the physiological pH may impose an ionic character on
nucleobases with standard pKa values. Additionally, transition or soft metal ions, by binding to the
nucleobase391-395 may also impose a change in the ionic character of the nucleobase, equivalent to the
effect of its protonation or deprotonation (depending upon which of the sites in a nucleobase a metal
ion prefers to bind!).
The affinity of a pyrimidine oligonucleotide399 cytidine residues for a target DNA duplex
increases by 10-fold when the pH is changed from 7.6 to 5.8, whereas the equilibrium binding constant
for another pyrimidine oligonucleotide in which 5-methylcytosin-1-yl nucleotides have been
incorporated increases by a factor 20 upon the same change in pH. An analysis of these data in terms of
two-state model revealed that the above pyrimidine oligonucleotides with cytidine and its 5-methyl
counterpart form triple helical structures with apparent pKas of 5.5 [(C+GC) triplets] and 5.7
[(m5C+GC) triplets], respectively. These pKa values are respectively 1.2 and 1.3 unit higher than those
known for the nucleoside counterparts, 2'-dC (pKa = 4.3) and 5-methyl-2'-dC (pKa = 4.4), respectively.
From the above studies397-399, it is clear that some specific adenin-9-yl and cytosin-1-yl nucleotides,
under certain folded structural states would form the protonated species at a pH close to the
physiological pH. Since the magnitude of the O4'-C1'-N1/9 anomeric effect is enhanced in the P
compared to the N state of the nucleobase in nucleosides (see the sections below), those residues are
more prone to take a N-type conformation, which in turn may dictate the local phosphate backbone to
adopt a specific conformation, which together may constitute a recognition element for a ligand
binding.
Owing to the key role played by protonation of the constituent nucleobases in the biological
function of DNA and RNA, it appeared necessary to obtain reliable estimates of the actual magnitude
of the anomeric effect in the N, P and D states of the nucleobases in nucleosides and nucleotides
(Sections 4 - 6).
As stated in Section 2.9, protonation of the nucleobase in various purine and pyrimidine
nucleosides and nucleotides, such as adenin-9-yl at N1 in 2'-O-methyladenosine318 or arabinoadenosine
derivatives198, guanin-9-yl at N7 in 2'-, 3'- and 5'-phosphates of guanosine313,317, cytosin-1-yl at N3 in
arabinocytidine or its methyl derivatives315 results in the shift of the N � S pseudorotational
equilibrium of the constituent pentofuranose moiety toward more N-type pseudorotamers (as
experimentally evidenced by the change in 3JHH coupling constants), in which the nucleobase adopts a
pseudoaxial orientation.
In six-membered rings, on the other hand, it has been shown that the preference of imidazolium
or pyridinium as anomeric groups in 2-substituted pyranose derivatives for equatorial positions is
greater than that of their neutral counterparts94,106-108. This has been attributed to the reverse anomeric
effect (Section 1.7). If the reverse anomeric effect were to play any role in the drive of the sugar
conformation in β-D-nucleosides at acidic pH, one should observe an increase in the population of S-
type pseudorotamers, in which the nucleobase adopts a pseudoequatorial orientation, in the acidic
compared to the neutral pH. Since the above qualitative studies198,313,315,317,318,400 as well as our
recent quantitative works30,32,36,37,44 (described in detail in Sections 4 - 6 and 8.8) show opposite
trends, i.e. a greater preference for N-type conformations in the P compared with the N state, one can
therefore conclude that no reverse anomeric effect operates in the protonated pentofuranosyl β-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
42
nucleosides. Instead, the shift of the two-state N � S pseudorotational equilibrium in nucleosides
toward N-type conformers at acidic pH can be easily explained in terms of pD-dependent magnitude of
the anomeric effect318: As the nucleobase becomes protonated, a partial positive charge is created at
the glycosyl nitrogen. As a result, nO4' →σ∗C1'-N9 interactions become more efficient. Owing to the
strengthening of the anomeric effect, the tendency of the nucleobase to adopt pseudoaxial orientations
in N-type pseudorotamers is greater in protonated nucleosides than in their neutral counterparts.
Protonation of guanosine at N7 promotes delocalization of the lonepair of the glycosyl nitrogen
N9, making it slightly positively charged, which has a certain azine-type character, and therefore its
15N chemical shift resonates downfield by 6.7 ppm compared to N9 in neutral guanine moiety401.
Similarly, upon protonation of N3 in 1-methylimidazole, the delocalization of the N1 lonepair results
in its partial positive character, which promtes its 7.4 ppm downfield shift compared to the counterpart
in the neutral form402. The delocalization of the N9 lonepair in N1-protonated adenosine to stabilize
N1H+ species also results in partly positively charged N9, which resonates downfield by 6.6 ppm in
comparison with N9 in neutral adenine403. On the other hand, in cytidine the chemical shift of N1 is
hardly affected (the downfield shift is only 1.1 ppm404) upon protonation of cytosin-1-yl at N3.
Upon its deprotonation at N3 in 3'-O-methylarabinouridine405, uracil-1-yl tends to prefer more
pseudoequatorial orientations than in the N state, as reflected in the increase of the population of S-type
pseudorotamers. This can be attributed to the fact that deprotonation of uracil-1-yl reduces the ability
of the glycosyl nitrogen to be involved in nO4' →σ∗C1'-N9 interactions, i.e. a weakening of the
anomeric effect, simply because of the electron-rich character of the conjugate base of uracil moiety.
2.9.4 Effect of base-modifications on the stability of nucleic acids
The modification of the chemical nature of nucleobases affects the overall stability of
DNA/RNA heteroduplexes has been reviewed47,49. However, the influence of the O4'-C1'-N1/9
anomeric effect (Section 4) as a result of base modification, or its modulation by the change of the
aromatic character by the change of pH of the medium (i.e. protonation-deprotonation equilibrium) or
by its binding to any specific ligand (i.e. association-dissociation equilibrium) may affect the stability
of the duplexes, which has not been addressed hithertofore in the literature.
The stabilization of DNA/RNA duplexes (with respect to parent unmodified duplexes), as a
result of modification of some constituent nucleobase(s) in the oligodeoxyribonucleotide chain, may be
attributed to either of the following events: (i) Increased stacking interactions in the case of 5-propynyl
dU406, 5-(amino-ethyl-3-acrylimido) dU49,407, 7-halo-7-deaza408,409 and 7-propyne-7-deaza410 purines
modifications, or (ii) to the shielding of the negative phosphate charges by 5-amino-hexyl-substituted
pyrimidines411, or (iii) to the possibility to form an additional hydrogen-bond in the case of 2-
aminoadenosine412,413. On the other hand, the destabilization of the heteroduplex DNA/RNA upon
modification of the nucleobase has been explained either by (i) loss of hydrophobic interactions (upon
substitution of thymin-1-yl by uracil-1-yl)414, or (ii) the inability of the thymin-1-yl nucleobase to adopt
anti orientation around the glycosyl torsion when it is substituted at C6414, or (iii) to the loss of
hydrogen-bonding sites in O2 or O4 substituted thymin-1-yl415.
2.10 The gauche effects in α- and β-nucleosides
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
43
In addition to O4'-C1'-N1/9 stereoelectronic interactions and the associated steric effect of the
nucleobase, the gauche effect also plays an important role to explain the preferred conformation of the
pentofuranose moiety in all nucleosides and nucleotides. The origin and the energetics of the gauche
effect in simple 1,2-disubstituted ethanes have been discussed in Sections 1.12 - 1.14. In this section,
we only present the relevant 1,4-gauche interactions which are involved in the drive of the sugar
conformation in nucleosides and nucleotides.
2.10.1 2'-Deoxynucleosides
In α- and β-2'-deoxynucleosides, the gauche effect of [O4'-C4'-C3'-O3'] fragment stabilizes
pseudorotamers in which the O4'-C4' and O3'-C3' bonds are in a gauche orientation (i.e. W- and S-
type), over those where they are in a trans arrangement (i.e. E- and N-type)304,305. However, as
discussed in Section 2.3-2.5, W-type pseudorotamers are neither observed in the solid state nor in
solution, owing to the high energy penalty resulting from the eclipsed orientation of C2' and C3'
substituents, therefore the [O4'-C4'-C3'-O3'] gauche effect alone drives the sugar conformation of
nucleos(t)ides preferentially toward S-type pseudorotamers. In α-2'-deoxynucleosides, since O4'-C1'-
N1/9 stereoelectronic interactions cooperate with the [O4'-C4'-C3'-O3'] gauche effect, one expects that
the conformation of the pentofuranose sugar will be biased toward S-type conformations. The results of
our recent study37 on the thermodynamics of the N � S equilibrium in α-D/L-2'-deoxynucleosides are
in agreement with this simple reasoning (Section 5). However, we have also shown that the magnitude
of stereoelectronic interactions in α-series is in general considerably reduced than in the β-
counterparts. In β-dNs, the anomeric effect drives the two-state N � S pseudorotational equilibrium in
solution toward N-type forms, whereas the [O4'-C4'-C3'-O3'] gauche effect favours S-type geometries
(vide supra). The experimentally observed preference of β-dNs and β-2'-deoxy mono- and
oligonucleotides for S-type conformations165 therefore suggests that the [O4'-C4'-C3'-O3'] gauche
effect is the predominant factor controlling the conformation of the constituent sugar moieties,
prevailing over the counteracting anomeric effect. This is consistent with our recent estimates of the
magnitudes of anomeric and gauche effects in such systems20,30.
2.10.2 Ribonucleosides and nucleotides
In ribonucleosides and nucleotides, three additional gauche effects determine the preferred
orientations within [O2'-C2'-C1'-O4'], [O2'-C2'-C1'-N1/9] and [O2'-C2'-C3'-O3'] fragments. Both in α-
and β-ribonucleos(t)ides, among all possible pseudorotamers, W-type geometries are the most favoured
by the [O4'-C4'-C3'-O3'] and [O4'-C1'-C2'-O2'] gauche effects, since only in that situation both
fragments are in a gauche orientation. Conversely, E-type pseudorotamers are the most disfavoured,
owing to the trans orientation of O4'-C4' and O4'-C1' bonds with respect to C3'-O3' and C2'-O2',
respectively. However, W-type pseudorotamers are energetically penalized (vide supra ), and in
aqueous solution, the sugar moiety is involved in a two-state N � S equilibrium. The [O4'-C4'-C3'-
O3'] and [O4'-C1'-C2'-O2'] fragments are in a trans and a gauche orientation in N-type pseudorotamers,
respectively, whereas in the S-type counterparts, this opposite is true, therefore the [O4'-C4'-C3'-O3']
and [O4'-C1'-C2'-O2'] gauche effects cancel each other. The [O2'-C2'-C1'-N1/9] gauche effect drives
the pseudorotational equilibrium of the sugar moiety in β-ribonucleosides toward S-type
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
44
pseudorotamers, whereas in α-ribonucleosides C1'-N1/9 and C2'-O2' bonds are in gauche orientation
both in N- and S-type conformations. The [O4'-C4'-C5'-O5'] gauche effect is one of the factors (besides
the steric or hydrogen-bonding interactions between 4'-CH2OH group and the nucleobase) that controls
the preferred conformation across the γ torsion angle.
Several models have been proposed to account for the gauche effects within X-C-C-Y
fragments, X and Y being electronegative elements or groups (Section 1.14). A simple explaination is
based on σ → σ* interactions between best donor σ and best acceptor σ* orbitals as shown for 2'-
deoxynucleosides in Fig 10 (A & B). Possible contributions from bond-bending, or a combination of
through-space and through bonds interactions to the mechanism of the gauche effect can however not
be excluded (Section 1.14).
In all studies on the effect of 1,4-gauche interactions upon the sugar conformation in
nucleosides and nucleotides, the estimates obtained for the magnitudes of these gauche effects actually
correspond to overall strength, consisting of steric, stereoelectronic and electrostatic components.
2.11 The gauche effects of sugar substituents and the self-organization of DNA/RNA
2.11.1 Studies on nucleosides
1H- and 13C-NMR studies on N-nucleosides17,46,193,195-199,201-204,206,208,
213,216,219,222,226,235,416-423, and on C-nucleosides224,237 as well as theoretical works304,305 have
qualitatively shown that the preferred conformation of the pentofuranose moiety in nucleosides,
nucleotides and their derivatives is strongly affected by the electronic nature and relative configuration
of the substituents at C2' and C3' positions.
We, on the other hand, have shown in our recent quantitative studies, that as the nucleobase in
β-D-N- and C-nucleosides becomes electron-deficient in the protonated state, the strength of O4'-C1'-
N1/9 stereoelectronic interactions increases, whereas in the deprotonated state, the N9/1 becomes more
electron-rich resulting in the weakening of O4'-C1'-N1/9 stereoelectronic interactions. This has been
discussed in details in Sections 4-6.
Systematic analyses of the conformational preferences of 2'-deoxy-2'-substituted uridine199,235
and adenosine206,417,418 derivatives have allowed us to derive linear relationships between the
population of the N-type pseudorotamer and the electronegativity of the 2'-substituent: As the 2'-
substituent becomes more electronegative, the population of N-type pseudorotamers linearly increases
as a result of the increased strength of the 2'-substituent gauche effects. 2'-thionucleosides416 prefer
more S-type geometries than in 2'-deoxynucleosides. Similarly, the two-state N � S equilibrium in 2'-
methylthionucleosides421 is strongly (> 70 %) biased toward S-type conformations in CD3OD, and the
effect of 2'-SMe has been attributed both to its reduced electronegativity (i.e. resulting in weaker [S2'-
C2'-C1'-O4'] and [S2'-C2'-C1'-N1/9] gauche effects) and increased steric bulk (resulting in the
destabilization of N-type pseudorotamers).
The population of S-type pseudorotamers in 3'-deoxy-3'-substituted arabinofuranosyladenine201
is also dictacted by the tuning of the gauche effect of [X3'-C3'-C4'-O4'], which is in turn controlled by
the electronegativity of the 3'-substituent (X). Similarly, substitution of 3'-OH in β-D-2'-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
45
deoxynucleosides by a more electronegative 3'-OPO2213, 3'-NO2
422 or 3'-F17,219,424 results in the
increased preference of the sugar moiety for S-type conformations. From a comparison of the
conformational properties of natural and 3'-modified thymidine dimers it has been found423 that S-type
pseudorotamers are more and more stabilized as the 3'-substituent is changed in the following order: 3'-
NH2 < 3'-N3 < 3'-O-H < 3'-O-P < 3'-O-S. The two-state N � S equilibrium in 2',3'-cis-fused furano-
and pyrrolidino-β-D-nucleosides is driven to S- (in case of C3'-O or C3'-N substitution) or N-type
sugars (for C2'-O or C2'-N).
Figure 10. Rationalization of the [X3'-C3'-
C4'-O4'] gauche effect in a 2',3'-dideoxy-3'-
substituted-β-D-nucleoside in terms σ σ* interactions between best donor (σC3'-H3') and
best acceptor (σ*C4'-O4') orbitals. The σ σ*
interactions stabilize the S- over N-type
pseudorotamers [Panel (A)], owing to the fact
that in the former the C3'H3' and C4'O4'
bonds are almost antiperiplanar, and therefore
the torsion angle between σC3'-H3' and σ*C4'-
O4' orbitals is much reduced in the S-type
sugar geometries (β ≈ -37°) compared to the
N-type counterparts (β ≈ -93°) [Panel (B)]
(Torsions angles have been calculated from
the values of the endocyclic torsion angle ν3
from ab initio optimized (at HF/6-31G*)
geometries of typical N- (PN = 21°; Ψm(N) =
35°) and S-type (PS = 143°; Ψm(S) = 34°)
pseudorotamers of 3'-fluorothymidine,
assuming simple trigonal symmetry. (C) The
extent of the energy stabilization of S-type
pseudorotamers through [X3'-C3'-C4'-O4']
gauche effect is proportional to the square of
the overlap between σC3'-H3' and σ*C4'-O4'
orbitals (i.e. S2) and inversely proportional
between the difference between their energies.
The sugar moiety prefers S- (or N) type conformers more in the furano than in the pyrrolidino
analog, owing to the stronger gauche effect of [O4'-C4'-C3'-O3'] (or [O4'-C1'-C2'-O2'] and [O2'-C2'-
C1'-N]) in the former compared with [O4'-C4'-C3'-N3'] (or [O4'-C1'-C2'-N2'] and [N2'-C2'-C1'-N]) in
the latter419,420. The gauche effect of the fluorine substituent, due to its high electronegativity, has a
profound stereoelectronic effect on the stereochemical orientation of the neighbouring groups, thereby
fluorine substituent governs the overall conformation of the sugar ring199,206,208,222.The sugar moieties
in 2'-α-fluoro-2',3'-β-D-dideoxyuridine and 3'-β-fluoro-2',3'-β-D-dideoxyuridine adopt exclusively N-
type conformations, owing to the cooperative drive of the [F2"(α)-C2'-C1'-O4'] and [F3'(β)-C3'-C4'-
O4'] gauche effects, respectively with the O4'-C1'-N1/9 anomeric effect. In contrast, as a result of
configuration-dependent gauche effect, the two-state N � S equilibrium in 2'-β-fluoro-2',3'-β-D-
dideoxyuridine and 3'-α-fluoro-2',3'-β-D-dideoxyuridine are strongly biased to the S-type conformers
because of the predominance of the [F2'(β)-C2'-C1'-O4'] and [F3"(α)-C3'-C4'-O4'] gauche effects,
respectively, over the anomeric effect226. 3'-methyl-thymidine in a TT dimer adopts preferentially N-
HOH2CO
HOH2C
Base
Base
H
H4'
C5'O4'
C2'
C2'
H4'
C5' H3'
X
C1'
C1'
σC3'H3'
α S2 / ΔE (σ*C4'O4' - σC3'H3')
σ*C4'O4'
Stabilization due to gauche effect (GE)
ΔE (σ*C4'O4' - σC3'H3')
O
O4'
X
X
σ*C4'O4'
σC3'H3'
σ*C4'O4'
σC3'H3'
H3'
XσC4'O4'
σ*C4'O4'
σ C3'H3'σC3'H3' β1 = -93ο
H3'
β1 = -37ο
σC4'O4'
σC4'O4'
σ*C4'O4'
where S = overlap between σ*C4'O4' and σC3'H3'GE
North, trans [O4'-C4'-C3'-X]
HF/6-31G* geometry; X = F HF/6-31G* geometry; X = F
South, gauche [O4'-C4'-C3'-X]
(A)
(B)
(C)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
46
type conformations425, owing to the lack of [O3'-C3'-C4'-O4'] gauche effect, whereas 3'-OH in the
thymidine counterpart drives the sugar conformation toward S-type pseudorotamers.
2.11.2 Studies on oligonucleotides
The application of the gauche effect concept in the design of oligonucleotides with specific
conformational properties will be discussed in detail in Section 8.1.
3. Methods to quantitate stereoelectronic effects in nucleos(t)ides
In our laboratory, a new strategy has been recently elaborated to obtain reliable accurate
estimates for the magnitudes of the O4'-C1'-N1/9 anomeric effect and the gauche effects of [O4'-C4'-
C3'-O3'], [O4'-C1'-C2'-O2'] and [O2'-C2'-C1'-N1/9] fragments controlling the sugar conformation in
nucleosides and nucleotides14-45. Our method is based on two essential steps: (i) The initial calculation
of the thermodynamics (i.e. ∆H°, ∆S° and ∆G°) of the two-state N � S equilibrium in 12 - 83 (Section
3.1-3.7). (ii) We have subsequently correlated the experimental ∆H° values with the structural features
of 12 - 83 either through semi-quantitative regression analysis or pairwise comparisons (Section 3.10
and following).
3.1 Thermodynamics of the two-state N � S equilibrium
∆H°, ∆S° and ∆G° of the two-state N � S equilibria of abasic sugars 12 - 1620,37, α-D-2',3'-
dideoxynucleosides (α-D-ddNs) 17 - 2036, α-D-2'-deoxynucleosides (α-D-dNs) 21 - 2637, α-L-2'-
deoxynucleosides (α-L-dNs) 27 - 2937, β-D-2',3'-dideoxynucleosides (β-D-ddNs) 30 - 3620,36, β-D-2'-
deoxynucleosides (β-D-dNs) 37 - 4520,30, β-L-2'-deoxynucleosides (β-L-dNs) 46 - 4937, β-D-
ribonucleosides (β-D-rNs) 50 - 5520,30, β-D-ribo-C-nucleosides (β-D-C-rNs) 56 - 6224,25,32, β-D-3'-dA
6327,426, 3'-monophosphates of β-D-2'-deoxynucleosides (β-D-dNMPs) 64 - 6823, 3'-ethylphosphates
of β-D-2'-deoxynucleosides (β-D-dNMPEts) 69 - 7323, 3'-monophosphates of β-D-ribonucleosides (β-
D-rNMPs) 74 - 7828 and 3'-ethylphosphates of β-D-ribonucleosides (β-D-rNMPEts) 79 - 8328 have
been derived in two steps: (i) Vicinal proton-proton coupling constants (3JHH) extracted from their 1H-
NMR spectra have been initially translated into the parameters defining the geometry of the N- [PN and
Ψm(N)] and S-type [PS and Ψm(S)] sugar pseudorotamers as well as their mole fraction at the
equilibrium using the PSEUROT203,209,427 program. (ii) Van't Hoff analysis of the temperature-
dependent mole fractions of the N- and S-type conformations has subsequently allowed to derive ∆H°,
∆S° and ∆G° values of their N � S equilibria at each pD, and in the case of nucleosides exhibiting pD-
dependent conformational preferences, we have subsequently estimated the values of ∆H°, ∆S° and
∆G° in each of their P, N and deprotonated (D) states from a nonlinear fitting procedure.
3.2 1H-NMR spectra [temperature, pD, ligand-dependent spectra of nucleos(t)ides]
One-dimensional 1H-NMR spectra of 5 - 20 mM D2O solutions of 12 - 83 have been recorded
between 278 K and 358 K (in 5 K or 10 K steps) at one or several pDs within the 0.5 - 12.0 range at
600 MHz or 500 MHz using Bruker DRX 600, DRX 500 and AMX 500 spectrometers.
For abasic sugars 12 - 16, β-D-ddU (35), β-D-ddI (36) and mononucleotides 64 - 73 and 75 -
83, the spectra have been recorded at one (neutral) pD only. At neutral pD, 50% of all molecules of 3'-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
47
monophosphates β-D-dNMPs 64 - 68 and β-D-rNMPs 74 - 78 are expected to have a net charge of -1
(pKa ≈ 1.5 for the first dissociation of the monophosphate moiety165) whereas the remaining 50% will
carry a charge of -2 (pKa ≈ 6.7 for its second dissociation165). However, the corresponding 3'-
ethylphosphates β-D-dNMPEts 69 - 73 and β-D-rNMPEts 79 - 83 carry a single net negative charge at
the neutral pD. In order to assess any possible influence of the ionization state of the phosphate moiety
upon the bias of the two-state N � S equilibrium in β-D-rNMPs compared with their β-D-rNMPEts
counterparts, we have also recorded 1H-NMR spectra of β-D-AMP (74) at several pDs in the range
from 6.5 to 8.4 (0.5 pD unit resolution). Under this pD range, only the ionization state of the phosphate
moiety is expected to change, not that of the constituent adenin-9-yl, since its pKa is ≈ 3.5 (section
2.8). We found that 3JHH and 3JHP remain constant throughout the whole 6.5 - 8.4 pD range, showing
that the ionization state of the 3'-monophosphate moiety does not affect the preferred conformation of
the pentofuranose moiety. For L-nucleosides 27 - 29 and 46 - 49 and for D-nucleosides 22, 23, 38 and
39, the 1H-NMR spectra have been recorded at two or three pDs (one in each of their P, N and/or D
states).
For all 1H-NMR experiments, 8 - 32 scans were typically recorded using a spectral width of ≈
10 ppm. The FIDs were processed using a slight gaussian apodization in order to enhance resolution,
and the final spectra consisted of 64 K datapoints. The pD values correspond actually to pH* values,
since they have been obtained simply by reading the values displayed on a pH meter (equipped with a
calomel electrode calibrated with pH 4 and 7 standard buffers in H2O) without any further correction
for the deuterium isotope effect. The pD of each sample has been adjusted by the simple addition of
microliter volumes of D2SO4 or NaOD solutions (typically 0.1 - 0.5 N). The small (< 0.1 ppm) and
irregular change in the chemical shifts of all resonances from 278 K to 358 K suggested that
aggregation was negligible. The 1H resonances were assigned through decoupling experiments and
one-dimensional nOe difference experiments. 1H-1H coupling constants (i.e. geminal 2JHH and vicinal
3JHH) have been verified with help of the DAISY428 simulation and iteration program package.
Identical values have been found at 1mM and 20 mM concentrations.
The errors on experimental 3JHH coupling constants (± 0.1 Hz in most cases, except for some
α-D-ddNs and β-D-ddNs at acidic pDs owing to the near isochronocity of some of H2', H2", H3' and
H3" multiplets36) have been estimated from the fit of simulated (or iterated) spectra to the
corresponding experimental 1H-NMR spectra. The influence of these errors upon the geometries and
relative populations of N- and S-type pseudorotamers engaged in the two-state equilibrium in aqueous
solution has been investigated during the pseudorotational analyses with a slightly modified version of
PSEUROT program (vide infra).
3.3 Pseudorotational analyses of 3JHH with PSEUROT and some practical hints
Some practical hints to perform dependable PSEUROT calculations on natural and unnatural
systems for quantitative estimation of the thermodynamics of intramolecular stereoelectronic effects
are perhaps important for new users. Before discussing in details the steps involved in performing
dependable PSEUROT calculations, we would like to urge a potential user of PSEUROT to consider
the following guidelines in addition to those described in the manual on PSEUROT 6.0 (Prof C.
Altona, Gorlaeus Laboratories, University of Leiden, 2300 Leiden, The Netherlands).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
48
3.3.1 Incorporation of coupling constant errors in PSEUROT calculations
Measurements of experimental coupling constants are prone to errors of at least 0.1 Hz. In the
original PSEUROT program this is not taken into consideration. We have therefore modified version
5.4 of PSEUROT to accomodate such error by generating datasets of coupling constants (subsequently
used to perform individual calculations with PSEUROT) with a gaussian distribution around the
experimental values (see our website:
http://bioorgchem.boc.uu.se/WWW_secret/manual_pages/local/ps54ran.html). Readers are directed to
Section 3.6 for details.
3.3.2 Parameters to be fixed or optimized during PSEUROT calculations
PSEUROT allows to perform a multilinear fit of the experimental coupling constants to P and
Ψm of the N and S conformers and their mole fractions in either of the following ways:
(i) All parameters can be freely optimized during the calculation. This can be done only when
the number of observables (i.e. temperature-dependent coupling constants) is greater than the number
of unknowns (P and Ψm of the N and S conformers and mole fraction xN or xS of a pseudorotamer).
This is a very important consideration! It is also noteworthy that this is also the reason why
pseudorotational calculations on 2'- or 3'-dNs are more relibale than on the ribo counterparts.
(ii) In the case of two-state N �S equilibrium strongly biased (i.e. ≥ 70%) to either of the N- or
S-type pseudorotamers, it is advised to obtain as much information as possible on the geometry of the
major conformer, and also the evidence of why two-state equilibrium is a valid concept on the
unnatural compound in question. This can be achieved by performing a set of PSEUROT runs in which
P(minor) and Ψm(minor) of the minor pseudorotamer are constrained to different values in each
calculation. P(minor) and Ψm(minor) are chosen in such a way that over the whole set of calculations the
ensemble of P(minor) and Ψm(minor) values that one can reasonably expect to be accessible to the minor
pseudorotamer is surveyed. The hyperspace of "reasonable" geometries that can be adopted by the
minor pseudorotamer may be assessed basing on either a perusal of crystal structures of the compound
of interest or some of its derivatives, or on the results of ab initio calculations at a higher basis set such
HF/6-31G* (even semiemperical or molecular mechanics based geometry can show some guidelines
when it is not possible to perform ab intio calculations). Further hints can also be found in the manual
of PSEUROT version 6.0, p. 24, which can be obtained from Prof. Altona's group.
(iii) In contrast, when there is no clear preference for either N- or S-type pseudorotamers, a
PSEUROT calculation is performed in which Ψm(N) and Ψm(S) are fixed to an identical value during
each calculation, and this value is incremented in the following runs with PSEUROT in such a way that
all possible "reasonable" values are surveyed throughout the whole set of calculations.
(iv) In the case of conformationally constrained compounds, it is required that the coupling
constants show some variation as a function of temperature. If not, it is impossible to derive
thermodynamics of stereoelectronic effects for these compounds.
3.3.3 Electronegativity of the substituents on each HCCH fragment
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
49
In the earlier versions of PSEUROT (e.g. 3B), the generalized equation (referred to as the EOS
equation, Eq 8 in Section 3.4) was based on Huggins group electronegativities. In more recent versions,
the Karplus equations incorporate the values of substituent parameters λ429,430 to take into account the
effect of the electronegativity of the substituents on the HCCH fragment of the coupling protons, which
present several advantages over the Huggins electronegativities (see Section 3.4 for details). In a
situation where λ is not known for a particular substituent (which is very often the case in unnatural
nucleosides used in the antisense work, for instance!), it is advised to use as a starting point the value
for the chemical group (among the 50 presented in the article429) whose chemical nature is closest of
that of the group of interest in the unnatural nucleoside. An alternative would consist in performing a
series calculations in which hypothetical λ values (between the limiting values in the scale 0 to 1.4) are
successively used in order to assess the effect of the unknown λ value on the thermodynamics of the N
� S equilibrium. For compounds with a known pKa value, one can determine the pH-dependent
thermodynamics of the N � S equilibrium using a series of λ values, and figure out which one gives
the correct pKa value of the aglycone or of any other ionizeable substituent. In our recent studies on the
pD-dependent modulation of the thermodynamics of the N � S equilibrium in N- and C-nucleosides
(Sections 4 - 6) through tunable gauche and anomeric effects, it was necessary to assess the influence
of the change of the electron-density of the glycosyl nitrogen upon its λ value. Our unpublished results
(see Section 3.7(c) for details) show that the effect of a particular λ(N1/9) value chosen within the 0.0 -
1.4 range on ∆H° and ∆S° of the N � S pseudorotational equilibrium is not significant.
3.3.4 Priority rule to number the substituents on the HAC1C2HB fragment
The original convention431 proposed by Altona's group to differentiate the substituents on the
HAC1C2HB fragment according to their relative orientations with respect to the coupling protons HA
and HB is better shown under the form of Scheme 2 (see also p. 14 in the manual of PSEUROT version
6.0).
The substituent S1 on C1 is said to be "positive" (ζ = +1 in the Karplus equation) since the
projected valency angle between HA and S1, counting clockwise from HA, amounts to ≈ +120°.
Conversely, S2 is a "negative" substituent, since the projected valency angle between HA and S2,
counting anticlockwise from HA, amounts to ≈ -120° (ζ = +1 in the Karplus equation). Analogously, S3
and S4 are "positive" and "negative" substituents, respectively, owing to the fact that the projected
valency angles between HB and S3, on one hand, and between HB and S4, on the other (counting
clockwise and anticlockwise from HB, respectively) are ≈ +120° and ≈ -120°.
3.3.5 Translation of HCCH
torsion angles into endocyclic
torsion angles
In the second step of
PSEUROT (Step 4a in Scheme 3
and step 2 in Scheme 4), proton-
proton (HCCH) torsion angles are
translated into the corresponding
HA
(-) S2 S1 (+)
(+) S3 HB
S4 (-)
HB
(-) S4 S3 (+)
(+) S1 HA
S2 (-)
S2
S3
C2
S4
HB
C1
S1
HA
Φ
α1 = +120°α2 = -120°
α3 = +120°α4 = -120°
Φ
Scheme 2. Definition of "positive" and "negative" substituents on the HAC1C2HB fragment
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
50
endocyclic torsion angles (ν0 .... ν4) via linear relationships such as: Φ = Aνi + Bi (Eq 11 in Section
3.5). For natural pentofuranosyl nucleosides, A and B have been determined for the β-D-ribose, β-D-
2'-deoxyribose, β-D-arabinose, β-D-lyxose, β-D-xylose, α-D-lyxose and α-L-xylose configurations
from linear regressions based on torsion angles extracted from crystal structures174,182,205 (see also the
discussion in the manual of PSEUROT version 6.0). For other modified nucleosides, A and B pairs can
be derived from correlation plots of HCCH versus νi torsion angles as extracted from ab initio
optimized geometries (e.g. at HF/3-21G* or higher basis sets) of various pseudorotamers of the
nucleoside of interest with selected P values (for instance, 12 structures with 0° < P < 330° at
30° resolution and a common puckering amplitude). Care is to be taken so as to ensure that each
pseudorotamer optimized ab initio will have the desired phase angle value, i.e. by constraining the
values of two endocyclic torsions (e.g. ν0 and ν4) during the calculation.
3.3.6 General operational conditions for PSEUROT
In our conformational studies on nucleosides and nucleotides, 3JHH coupling constants have
been interpreted in terms of a two-state N � S equilibrium with help of the program
PSEUROT203,209,427. The validity of the two-state equilibrium has been experimentally evidenced by
the analysis of the distribution of sugar conformations in the X-ray crystal structures of nucleosides, as
well as by the results of NMR studies in solution both in our own lab and elsewhere32,36,37,174, 230,241-
244,245,248 (Section 2.3 and the following).
PSEUROT Program
Experimental 3JHH
Coupling Constants for
a nucleoside X
Karplus-Altona Equation
AH
, BH parameters
from ab initio
calculations or
crystal structure
Endocyclic Torsion Angle (ν0 - ν4)
Pseudorotational Concept
Phase Angle of Pseudorotation (P) and Maximum Puckering Amplitude (Ψm) and Equilibrium Populations for X
(5a)
(4a)
(3a)
X-ray Crystal Structure of X
Phase Angle of Pseudorotation (P) and Maximum Puckering Amplitude (Ψm)
(6)(7)
Φ(HCCH) torsion angles
Φ(HCCH) torsion angles
(2)
Experimental 3JHH in
constrained nucleosides
Corresponding Φ(HCCH) torsions angles from X-ray
structures
A New 7 Parameter Karplus
type equation
(9)(10)
Φ(HCCH) torsion angles
Phase Angle of Pseudorotation (P) and Maximum Puckering Amplitude (Ψm) and Equilibrium Populations
(3b)
(4b)
(5b)
Endocyclic Torsion Angle (ν0 - ν4)
Experimental 3JHH
Coupling Constants
(8)
(11) (12)
Ab Initio Calculations
on X
The two-state N S equilibrium is a valid model for
pseudorotational analyses of 3JHH
of the nucleoside X
(1)
Scheme 3: Iterative structure elucidation of nucleosides using NMR-PSEUROT203,209,427 analyses of 3JHH coupling
constants, ab initio calculations and comparisons. Ab initio calculations on aristeromycin (8), 2'-deoxyaristeromycin (9) and
3'-deoxyaristeromycin (10)41 allowed us to validate the two-state N � S equilibrium model (Steps 1 & 2) in such
compounds and to derive AH and BH values required for step 4a (see below). Three translation steps are involved in
PSEUROT: (i) Experimental 3JHH are first used to calculate ΦHCCH torsion angles (Step 3a) using Eqs 8a-9a; (ii) ΦHCCH
are used to calculate the endocyclic torsion angles (νi) (Step 4a). AH and BH are published174 for β-D-dNs/rNs. For α-D-
dNs, 120° was subtracted from AH and BH known for Φ1'2' (Φ1'2") in parent β-D-dNs. For α-/β-L-dNs, AH and BH are the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
51
same (but with opposite signs) as those of α-/β-D counterparts234. For β-D-dNMPs/dNMPEts, we used the same values as
for β-D-dNs. For β-D-rNMPs/rNMPEts and β-D-C-rNs we assumed AH and BH identical to those of β-D-rNs. For β-D-
ddNs, AH and BH were derived from regression analysis of HCCH torsion angles versus the corresponding νi (taken from
crystal structures268,269,432-434). For α-D-ddNs, we have subtracted 120° from the values found for Φ1'2' (Φ1'2") for the
parent β-D-ddNs. (iii) PN, PS, Ψm(N), Ψm(S) and xS are finally calculated (Step 5a) from all νi values using Eq 6. For
aristeromycin, we found41 that the geometry of the cyclopentyl ring in the solid state (steps 6 and 7) is different (step 8)
from the solution state geometry derived from PSEUROT analyses (based on the original431 Karplus-Altona Eqs 8a-9a). To
verify whether this was due to a poor parametrization of Eqs 8a-9a for H2CCH2 fragments, we specifically reparametrized
them (Eq 8b-9b, steps 9 & 10) for carbocyclic nucleosides. PSEUROT analyses (steps 3b - 5b) based on Eqs 8b-9b
produced the same geometries (step 11) for the cyclopentyl ring in aristeromycin, its 2' and 3'-deoxy derivatives as the initial
analyses (steps 3a-5a), however the r.m.s. errors were reduced when Eqs 8b-9b were used.
Five parameters are necessary to define the position of the equilibrium at a certain temperature or pD:
The phase angles of the N- (PN) and S-type (PS) pseudorotamers and their respective puckering
amplitudes [Ψm(N) and Ψm(S)] as well as the mole fraction of one of the conformers (xN or xS). In the
case of 12 (10 experimental 3JHHs), of α-D-ddNs 17 - 20 and β-D-ddNs 30 - 36 (8 3JHHs) and of 13,
15 and 16 (7 3JHHs), the system is overdetermined and PN, PS, Ψm(S), Ψm(S) and xS can in principle
be estimated simultaneously from a single set of coupling constants. For α-D-dNs 21 - 26, α-L-dNs 27
- 29, β-D-dNs 37 - 45, β-L-dNs 46 - 49, β-D-dNMPs 64 - 68, β-D-dNMPEts 69 - 73 and β-D-3'-dA
(63), five 3JHHs are available at each temperature. Since there are as many unknowns as knowns, if the
calculation is performed without constraining one of the unknowns to some value estimated by another
method (for instance from X-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
52
PSEUROT Program
Experimental 3JHH
Coupling Constants
Karplus-Altona Equation
Proton-Proton Torsion Angle
AH , BH parameters from ab initio calculations or crystal structure
Endocyclic Torsion Angle (ν0 - ν4)
Pseudorotational Concept
Phase Angle of Pseudorotation (P) and
Maximum Puckering Amplitude (Ψm)
and Equilibrium Populations
Experimental 3JHF
Coupling Constants
Proton-Fluorine Torsion Angle
Endocyclic Torsion Angle (ν0 - ν4)
3JHF and Corresponding Proton-Fluorine
Torsion Angles from Conformationally
Fixed Compounds
AF , BF parameters from ab initio calculation or crystal structure
Proton-Fluorine Torsion Angle
Limiting 3JHF
Coupling Constants
Extrapolation of experimental 3JHF to pure major conformer
New Karplus-Type Equation
for 3JHF Coupling Constants
(9)
(8)(6)
(7)
(5)
(10)
(3)
(2)
(1)
(4)
Scheme 4: The "PSEUROT+JHF" program39, as an extension of the original PSEUROT program for pseudorotational
analysis of 3JHF in combination with 3JHH coupling constants. We first performed pseudorotational analyses of 3JHH
coupling constants of monofluoronucleosides to find out P, Ψm and xS (at various temperatures) of the N and S forms
(Steps 1 - 3). From the plot of temperature-dependent xS as a function of temperature-dependent experimental 3JHF, we
derived limiting 3JHF values for the major N or S pseudorotamer of each nucleoside (step 5). The corresponding HCCF
torsion angles for all 3JHF coupling constants were calculated from the corresponding endocyclic torsion angles (step 4),
basing upon AF and BF values (Eq 11 applied to ΦHCCF torsion angles) derived from ab initio optimized geometries of our
monofluoronucleosides. A dataset of 57 pairs of the above limiting (3JHF , ΦHCCF) from our monofluoronucleosides and
from conformationally constrained cyclic fluorinated organic compounds (step 6), which were added in order to better
define the values of 3JHF for ΦHCCF around ± 90° and 180°, was subsequently used to parametrize a new Karplus-type
equation specifically for 3JHF (Step 7), which includes a term accounting for HCC and FCC bond angle changes. The
validity of Eq 8c was proven by the fact that pseudorotational analyses based on 3JHF alone (steps 8 - 10) of a set of other
monofluoro nucleosides yielded nearly the same values for P, Ψm of the N and S pseudorotamers as our initial analyses
which relied exclusively upon 3JHH data (Steps 1 - 3). Eq 8c was finally used in combination with the Karplus-Altona
equation430 to derive for the first time accurate estimates for P, Ψm of the N and S pseudorotamers and xS (at various
temperatures) for difluoronucleosides 84 - 87 (steps 1 - 3 and 8 - 10).
ray crystal structures or ab initio calculations), the accuracy of the results will strongly depend upon the
accuracy of the experimental 3JHH coupling constants. For all other compounds, only 3 (or 4 in the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
53
case of abasic sugar 14) 3JHHs are available at each temperature and the iteration with PSEUROT is
only possible if at least two (or one) parameter(s) are (is) fixed.
Calculations with PSEUROT are based on three principal steps (Schemes 3 & 4): (i)
Translation of experimental 3JHH coupling constants into the corresponding proton-proton torsion
angles (Φ) [Step 3a in Scheme 3, step 1 in Scheme 4] with help of the generalised Karplus-Altona
equation. In some modified nucleosides it might be necessary to reparametrize this Karplus-Altona
equation to take account of the influence of a specific substituent on the bond length, bond angle,
through-space transmission effect (i.e. the Barfield effect) etc; (ii) ΦHCCH are themselves used to
calculate the corresponding endocyclic torsion angles of the pentofuranose moiety (ν0 ... ν4) (Step 4a
in Scheme 3, step 2 in Scheme 4); (iii) PN, PS, Ψm(N) and Ψm(S) of the N-type and S-type
pseudorotamers as well as xS are ultimately derived from all endocyclic torsion angles obtained in step
(ii) using the law of pseudorotation (Eq 6, Step 5a in Scheme 3, step 3 in Scheme 4).
3.4 Generalised Karplus-type equation
Karplus-type equations (Step 3a in Scheme 3, Step 1 in Scheme 4) allow to translate vicinal
coupling constants into the torsion angle between the coupling nuclei, and their use in conformational
analysis is well known215,217,435-437.
3.4.1 EOS Karplus-Altona equation for 3JHH
The generalised EOS equation (Eq 8a) developed by Altona's group431 describes the
dependence of a vicinal 3JHH coupling constant upon the corresponding proton-proton torsion angle,
the Electronegativity and relative Orientations of the Substituents in the H-C-C-H fragment (EOS).
3J
HH = P1 cos2Φ+ P2 cosΦ + P3 + ∑
−=
Δ
m1i
χi(g) [P4 + P5 cos2 (ζiΦ + P6 |∆χi(g)|)] ... Eq 8a
Φ designates the H-C-C-H torsion angle. The first three-terms in Eq 8a show the dependence of 3JHH
upon the H-C-C-H torsion angle alone, using the same formalism as that originally proposed by
Karplus et al438. The introduction of a phase-shifted cosine square function in the remaining terms
allows to take into consideration the influence of the electronegativity of the non-hydrogen substituents
on the H-C-C-H fragment and of their relative orientations with respect to the coupling protons. It is
assumed that the effects of m substituents are additive. The value of ζi (± 1) reflects the orientation of
the substituent i with respect to the coupling protons. Δχi(g) represents the difference between the
group electronegativities of the substituent i and of hydrogen, which is used as a reference, in the
Huggins scale439. Δχi(g) is calculated according to Eq 9a: Δχi (g) = Δχi(α) - P7 ∑−=
Δ
n1j
χj (β-substituent)
..... Eq 9a. Δχi(g) takes into account both the electronegativity of the α-substituent [Δχi(α)] and
the electron-withdrawing or donating effect of n β-substituents [Δχj (β−substituent)]. No term in Eq 8a
takes into account a possible effect of H-C-C bond angles and C-C bond distances440,441, or the
possible influence of through space orbital interactions 442-444. P1 - P7 have been optimized separately
for HCCH, H2CCH and H2CCH2 fragments using a set of 315 experimental 3JHH coupling constants
and Φ values (as derived from molecular mechanics calculations), of conformationally constrained six-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
54
membered rings, which implies that the torsion angles are clustered into the gauche and trans regions.
Eq 8a allows to predict the experimental 3JHH with an accuracy (r.m.s.) of 0.48 Hz.
Recently, Eq 8a has been reparametrized430 using a set of 299 pairs of (3JHH, Φ). In that work,
Huggins Δχi(g) values (Eq 9a) were replaced by empirical substituent parameters (λ) and the overall
r.m.s. error between experimental and theoretical 3JHH values dropped down from 0.48 Hz to 0.36 Hz.
Additionally, it was found that a separate parametrization for HCCH, H2CCH and H2CCH2 fragments
was no longer necessary.
3.4.2 Reparametrized EOS equation for carbocyclic nucleosides
We have recently determined41 the solution conformations of aristeromycin (8), 2'-
deoxyaristeromycin (9) and 3'-deoxyaristeromycin (10) from pseudorotational analyses of 3JHH
coupling constants using PSEUROT version 3B203-203 program, which is based on the Karplus-Altona
EOS equation (Eq 8a). We found serious discrepancies between the X-ray crystal structure (P = 89°,
Ψm = 41°) of aristeromycin (8) and its structure calculated by NMR-PSEUROT conformational
analysis (35° < P [3/4
T - 0/4
T] < 65°, 35° < Ψm < 45°) (128° < P [1E] < 131°, 34° < Ψm < 36°), as
well as relatively high errors in the NMR-PSEUROT analyses for aristeromycin and its 2'-deoxy and
3'-deoxy derivatives [∆Jmax ≤ 1.6 Hz (i.e. maximal difference between experimental and PSEUROT-
calculated 3JHH) and r.m.s. error ≤ 0.7 Hz]. These observations have prompted us to reparametrize (Eq
8b) the Karplus equation implemented in the PSEUROT program by using torsion angles derived from
solid state geometries of conformationally constrained nucleosides and their corresponding
experimental 3JHH.
3JHH
= 13.41 cos2Φ -0.98 cosΦ + 1.37
+ ∑−=
Δ
m1i
χi(g) [0.20 - 2.26 cos2 (ζiΦ + 0.39 |∆χi(g)|)] ......Eq 8b
where, Δχi (g) = Δχi(α) + 0.071 ∑−=
Δ
n1j
χj (β-substituent) ..... Eq 9b
The χ2 value for the fitting of the experimental ΦHH vs 3JHH
data using Eq. 8b is 7.2 Hz2,
which corresponds to an r.m.s. error of 0.40 Hz. For comparison, the r.m.s. error of the original
Haasnoot-Altona's equation (Eq 8a) was in the range 0.36 - 0.51 Hz, depending upon the substitution
patterns of H-C-C-H fragments (0.48 Hz in the case where all H-C-C-H fragments are considered
together for a common set of parameters). With the help of our new Karplus-type equation [Eq. 8b],
the difference between the experimental and calculated 3JHH couplings, ∆Jmax, was below 0.5 Hz for
40 data points, and in the range from 0.8 -1.3 Hz for five data points. It should also be noted that in our
approach we have a common equation for all H-C-C-H fragments. Moreover, the small P7 value
indicates the reduced influence of β-substituents. Therefore, for practical reasons, β-substituents may
even be excluded when the determination of the electronegativities of the substituents is performed.
The results of the PSEUROT analyses performed with the standard Haasnoot-Altona Karplus
equation (EOS equation: Eq 8a) are also very comparable in terms of geometry with those based on our
reparametrized equation (Eq. 8b). Both series of PSEUROT analyses suggest that the predominant
conformation of the cyclopentane ring in carbocyclic nucleosides is defined by 128° < P < 140° for
aristeromycin (8), 105° < P < 116° for 2'-deoxyaristeromycin (9) and 118° < P < 127° for 3'-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
55
deoxyaristeromycin (10), with the puckering amplitude in the range from 34° to 40°. However,
PSEUROT analyses based on our Karplus equation produced a smaller r.m.s. error by ≤ 0.14 Hz and
∆Jmax error by ≤ 0.5 Hz than those performed with the standard Haasnoot-Altona's equation.
3.4.3 Karplus equation for interpretation of 3JHF
In the case of difluoronucleosides 84 - 87, the conformation of the constituent sugar moiety
cannot be determined on the basis of the unique experimental 3JHH available, i.e. 3J3'4' . However, four
additional vicinal coupling constants, i.e. proton-fluorine couplings 3JHF, can also be extracted from
their 1H-NMR spectra. It was clear to us that these 3JHF could also be used to find out the preferred
conformation of the sugar moiety in 84 - 87 in solution through pseudorotational analyses, provided
that we have at our disposal an accurate Karplus-type equation to translate them into the corresponding
HCCF torsion angles. It is this observation that has prompted us to parametrize39 a new seven-term
Karplus equation specifically for 3JHF coupling constants using a dataset consisting of
monofluoronucleosides and conformationally constrained fluorinated organic compounds (Eq 8c) [see
Section 3.8 for details]. Eq 8c has been constructed basing upon Eq 8a, however two significant
improvements have been made in comparison with the original formalism: (i) We have used
λ substituent parameters to account for the electronegativity of the substituents on the H-C-C-F
fragment of interest, in view of Altona's latest work430 and (ii) we have incorporated a cosine squared
term that allows to reproduce the experimentally observed variation in 3JHF coupling constants as a
function of the HCC (aHCC) and HCF (aFCC) bond angle values.
3JHF = 40.61 cos2Φ - 4.22 cosΦ + 5.88 + Σ λi [-1.27 - 6.20 cos2(ξi Φ + 0.20 λi)]
- 3.72 [(aFCC + aHCC)/2 - 110]
....Eq. (8c)
Using Eq 8c, we have been able to elucidate the conformation of a series of mono and
difluorinated nucleosides using a combination 3JHH and 3JHF coupling constants (Section 3.8). Our
study has shown that the geometries of the N- and S-type pseudorotamers derived from calcualtions
with PSEUROT based on either type of coupling constants were nearly identical.
3.4.4 Refined Karplus equation for 3JHH based on Fourier formalism
The formalism proposed in Eq 8a suffers itself from the following limitations. (i) It implies a
strict additivity of the effects of the substituents, in spite of the fact that linear correlations between the
value of 3JHH and the sum of the substituent electronegativities for monosubstituted ethanes break
down for 1,1-disubstituted ethanes carrying highly electronegative substituents. (ii) It poorly
reproduces the small experimental 3JHH values (< 1.2 Hz) for torsion angles in the ≈ ± 90° range. As a
result, Eq 8a has only been incorporated in the earlier versions of the PSEUROT program (version 3B).
In more recent versions (including version 5.4427, which has been used to perform the pseudorotational
analyses for 12 - 83), the dependency of 3JHH upon the HCCH torsion angle Φ and the nature of the
substituents is formulated as a truncated Fourier series429,445,446 in Φ (up to 3Φ) with coefficients
expanded as a Taylor series in the empirical substituent parameter λ, as shown in Eq 10a-b. 3J
HH = C0+ C1 cosΦ + C2 cos2Φ + C3 cos3Φ + S2 sin2Φ .... Eq 10a
where C0 - C3 and S2 are calculated according to Eq 10b, as follows:
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
56
C0 = 7.01 - 0.58 Σi λi - 0.24 (λ1λ2 + λ3λ4); C1 = -1.08; C2 = 6.54 - 0.82 Σi λi + 0.20 (λ1λ4 + λ2λ3);
C3 = -0.49; S3 = 0.68 Σi (ζiλ2i) ... Eq 10b
The new formalism is clearly advantageous: Terms reflecting the individual influence of each
substituent are still included, as in Eq 8, however new λ cross-terms are also introduced to take into
account possible interdependent substituent effects. Additionally, the influence of β-substituents upon 3JHH is incorporated in the value of empirically optimized λ parameters. The λ scale has been
developed429 through a least-squares fitting procedure, basing upon the change in the value of vicinal
3JHH as a function of the substitution pattern in mono- and 1,1-disubstituted ethanes, using two
references, i.e. λ(H) = 0.0 and λ(OR) = 1.4.
During the pseudorotational analyses performed on experimental 3JHH for 12 - 83, the
following λ values429,446 for the substituents at C1' - C4' have been used: λ(H) = 0.0; λ(C1') = 0.62 in
nucleosides or 0.67 in abasic sugars; λ(C2' or C3' deoxy) = 0.67; λ(C2' or C3' ribo) = 0.62; λ(C4') =
0.62; λ(C5') = 0.68; λ(O4') = 1.27; λ(OH) = 1.26; λ(OMe) = 1.27; λ(OPO3H-/OPO3Et-) = 1.27; λ(C-
aglycone at C1') = 0.45. For the glycosyl nitrogen of the nucleobase in nucleosides, we have used λ =
0.58 at all pD values.
3.5 Translation of experimental 3JHH and 3JHF into pseudorotational parameters
Proton-proton torsion angles (Φ) derived from step 3a in Scheme 3 and Step 1 in Scheme 4 can
in turn be used to calculate endocyclic torsion angles (ν0 .... ν4) of the pentofuranose moiety in 12 - 83,
using simple linear relationships (Eq 11, Step 4a in Scheme 3 and Step 2 in Scheme 4): Φ = Aνi + Bi
.... Eq 11
In the case of β-D-dNs and β-D-rNs, the values of A and B parameters, taken from the
literature174, have been determined from linear regressions, basing on torsion angles extracted from the
crystal structures of nucleosides. For α-D-dNs, we have subtracted 120° from the values published for
Φ1'2' (and Φ1'2") in the parent β-D-dNs, whereas for α- and β-L-dNs, A and B have been obtained by
reversing the signs of both A and B values used for their α- and β-D counterparts, respectively, as
suggested by Altona et al234. For β-D-dNMPs and β-D-dNMPEts, we have used the same A and B
values as for β-D-dNs, whereas A and B for β-D-rNMPs, β-D-C-rNs and β-D-rNMPEts have been
assumed identical to those known for β-D-rNs. For β-D-ddNs, A and B values have been determined
from regression analysis of some proton-proton torsion angles Φ as a function of the corresponding
νi values. Φ and νi values have been extracted from the crystal structures of β-D-ddA268, β-D-ddC269,
β-D-ddT432, β-D-ddU433 and β-D-2',3'-dideoxyribavirin434. Finally, for α-D-ddNs, we have subtracted
120° from the values found for Φ1'2' (Φ1'2") for the parent β-D-ddNs. Step 3 in PSEUROT translate the
values of the endocyclic torsion angles νi found in Step 2 into PN, Ψm(N) of the N-type pseudorotamer
and PS, Ψm(S) of the S-type sugar via the law of pseudorotation (Eq 6a).
3.6 Principle of iterations with PSEUROT
PSEUROT iterates the values of PN, Ψm(N), PS, Ψm(S), and of the mole fractions of the
conformers (xS) using a Netwon-Raphson minimization procedure, in such a way that a best fit is
obtained between experimental 3JHH and those back-calculated using the best fit PN,S, Ψm(N,S) and xS
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
57
values. Initially, PSEUROT calculates theoretical 3JHH values using the values of P and Ψm of both N-
and S-type pseudorotamers and their respective populations defined in the user's input. The
experimental and calculated 3JHH are compared. In the following iteration steps, random changes are
made in P and Ψm values, and/or in the populations, depending upon the user's input, which specifies
the parameters to be optimized or constrained during the calculation. The discrepancy between
experimental and calculated 3JHH is monitored and optimized during the iteration procedure. When the
best fit is found, the optimal PN, Ψm(N), PS and Ψm(S) of the N- and S-type conformers and their
relative populations are printed out, together with the error analysis, which shows both individual
differences between experimental and calculated 3JHH as well as an overall root mean square error
(r.m.s.).
Incorporation of the error in experimental 3JHH during the PSEUROT calculation: We have
modified the original version 5.4 of the PSEUROT program 203,209,427 to assess the propagation of
errors in experimental 3JHH throughout the calculations and during the subsequent treatment of the
results. Our modified program32 retains all features of the original version 5.4, all changes are
additions. The estimated error, expressed as standard deviation (σ), of each 3JHH and the desired
number of sets of randomly varied 3JHH to be generated and subsequently analyzed by pseudorotational
analyses, are included in the input file. Typically, for each set of contrained P and Ψm values (vide
infra), 1000 data sets are generated and individually analyzed. Each generated dataset contains
randomly varied 3JHHs but over all data sets, each 3JHH is normally distributed around its experimental
value with the given σ. The output from our modified program consists of statistical data [average, σ
and skew of the calculated geometrical parameters, of the mole fractions and of the "randomized" 3JHHs] and of the results from all the individual pseudorotational analyses (the calculated P and Ψm
values and mole fractions). We have discarded results which fall outside given ranges for: (i) The
individual errors between experimental and calculated coupling constants, i.e. Jcalc-Jexp, (ii) the root
mean square error (r.m.s.) in 3JHH, (iii) PN, PS, Ψm(N) and Ψm(S). PN, PS, Ψm(N) and Ψm(S) values
are considered as "reasonable" when they are within the ranges found for the sugar moieties in the
crystal structures of nucleosides and nucleotides, i.e. typically174: -40° < PN < 40°, 120° < PS < 200°,
30° (or 25° for α-nucleosides175) < Ψm(N) < 45° and 30° (or 25° for α-nucleosides175) < Ψm(S) < 45°.
Typically, Jcalc-Jexp and r.m.s. in 3JHH are considered as acceptable when they do not exceed Jcalc-Jexp
and r.m.s., respectively, of the best fit analyses by more than 0.1 Hz.
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
58
Tab
le 2.
∆H° a
nd -
T∆
S° c
ontr
ibuti
on
s (k
Jmol-
1)a
-c t
o ∆
G2
98 o
f th
e tw
o-s
tate
N �
S p
seudoro
tati
onal
equil
ibri
um
in a
bas
ic s
ugar
s 12 -
16 a
nd
nucl
eosi
des
17 -
63 w
ith f
ull
y p
roto
nat
ed, neu
tral
and f
ull
y d
epro
tonat
ed n
ucl
eobas
es a
nd t
he
corr
espondin
g p
Ka
val
ues
d,e
Com
pound
Full
y p
roto
nat
ed n
ucl
eobas
e pK
a fr
om
: N
eutr
al n
ucl
eobas
e pK
afro
m:
Full
y d
epro
tonat
ed n
ucl
eobas
e
Δ
HP;�
-Τ
ΔS
P;�
Δ
GP;2
98
∆G
° d
δ
1H
e
ΔH
�
-ΤΔ
S N
;� Δ
G N
;29
8
∆G
° d
δ
1H
e
ΔH
D;�
-Τ
ΔS
D;�
ΔG
D;2
98
12
-
- -
- -
0.4
(0
.3)
-0.3
(0
.3)
0.1
(0
.4)
- -
- -
-
13
-
- -
- -
-4.1
(0
.3)
1.5
(0
.3)
-2.6
(0
.4)
- -
- -
-
14
-
- -
- -
0.4
(0
.1)
1.5
(1
.0)
1.9
(1
.0)
- -
- -
-
15
-
- -
- -
-4.6
(0
.4)
1.2
(0
.3)
-3.4
(0
.5)
- -
- -
-
16
-
- -
- -
-4.2
(0
.4)
0.9
(0
.3)
-3.3
(0
.5)
- -
- -
-
α-D
-dd
A (
17
) -1
.7 (
0.1
) 0
.2 (
0.1
) -1
.5 (
0.1
) -
3.7
-1
.7 (
0.1
) 0
.2 (
0.1
) -1
.5 (
0.1
) -
- -
- -
α-D
-dd
G (
18
)
-8.7
6
.0
-2.7
2
.7
2.7
-0
.4 (
0.2
) -0
.9 (
0.2
) -1
.2 (
0.1
) -
9.7
-0
.4 (
0.2
) -0
.9 (
0.2
) -1
.2 (
0.1
)
α-D
-dd
C (
19
)
-2.9
(0
.2)
1.4
(0
.2)
-1.5
(0
.1)
- 4
.1
-2.9
(0
.2)
1.4
(0
.2)
-1.5
(0
.1)
- -
- -
-
α-D
-dd
T (
20
)
- -
- -
- -1
.1 (
0.1
) 0
.6 (
0.1
) -0
.5 (
0.1
) 1
0.1
9
.7
-0.5
(0
.1)
0.3
(0
.1)
-0.2
(0
.1)
α-D
-dA
(2
1)
-5
.0 (
0.4
) 2
.1 (
0.3
) -2
.8 (
0.1
) -
3.6
-5
.0 (
0.4
) 2
.1 (
0.3
) -2
.8 (
0.1
) -
- -
- -
3'-O
Me-
α-D
-dA
(2
2)
-5.8
(1
.5)
0.8
(1
.5)
-4.9
(0
.3)
- -
-6.4
(0
.8)
1.9
(0
.7)
-4.5
(0
.3)
- -
- -
-
3',5
'-d
iOM
e-α
-D-d
A (
23
) -5
.3 (
1.3
) 0
.8 (
1.2
) -4
.7 (
0.3
) -
3.8
-5
.8 (
0.8
) 1
.6 (
0.7
) -4
.1 (
0.3
) -
- -
- -
α-D
-dG
(2
4)
-10
.7 (
2.0
) 6
.4 (
2.0
) -4
.4 (
0.3
) 2
.6
2.5
-4
.5 (
0.5
) 2
.7 (
0.5
) -1
.9 (
0.1
) -
9.5
-3
.4 (
0.5
) 1
.4 (
0.5
) -1
.9 (
0.1
)
α-D
-dC
(2
5)
-7.1
(0
.3)
2.5
(0
.8)
-4.3
(0
.2)
4.1
4
.2
-7.1
(0
.3)
4.0
(0
.5)
-3.1
(0
.2)
- -
- -
-
α-D
-T (
26
) -
- -
- -
-4.0
(0
.2)
2.0
(0
.5)
-2.1
(0
.2)
9.8
9
.8
-4.0
(0
.2)
2.7
(0
.4)
-1.1
(0
.1)
α-L
-dA
(2
7)
-5
.5 (
0.9
) 2
.7 (
0.9
) -2
.8 (
0.3
) -
- -6
.0 (
0.5
) 3
.2 (
0.6
) -2
.9 (
0.4
) -
- -
- -
α-L
-dC
(2
8)
-7.3
(0
.6)
3.0
(0
.6)
-4.2
(0
.2)
- -
-7.2
(0
.5)
4.1
(0
.6)
-3.1
(0
.2)
- -
- -
-
α-L
-T (
29
) -
- -
- -
-4.2
(0
.4)
2.1
(0
.5)
-2.1
(0
.2)
- -
-3.4
(0
.2)
2.5
(0
.4)
-1.0
(0
.1)
β-D
-dd
A (
30
) 9
.2 (
0.1
) -5
.0 (
0.1
)
4.1
(0
.1)
3.6
3
.8
3.5
(0
.1)
-0.9
(0
.1)
2.6
(0
.1)
- -
- -
-
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
59
Tab
le 2
(C
onti
nued
)
Com
pound
Full
y p
roto
nat
ed n
ucl
eobas
e pK
a fr
om
: N
eutr
al n
ucl
eobas
e pK
afro
m:
Full
y d
epro
tonat
ed n
ucl
eobas
e
Δ
HP;�
-Τ
ΔS
P;�
Δ
GP;2
98
∆G
° d
δ
1H
e
ΔH
�
-ΤΔ
S N
;� Δ
G N
;29
8
∆G
° d
δ
1H
e
ΔH
D;�
-Τ
ΔS
D;�
ΔG
D;2
98
β-D
-dd
G (31
) 2
3.6
-1
7.1
6
.7
2.5
2
.5
3.4
(0
.4)
-0.3
(0
.2)
2.9
(0
.1)
9.6
9
.6
1.4
(0
.1)
0.5
(0
.1)
1.8
(0
.1)
5'-O
Me-β
-D-d
dG
(32
) 2
3.4
-1
6.8
6
.8
2.5
2
.5
4.3
(0
.5)
-0.6
(0
.5)
3.6
(0
.4)
9.7
9
.4
2.3
(0
.1)
1.0
(0
.1)
3.3
(0
.1)
β-D
-dd
C (33
) 9
.6 (
0.2
) -4
.9 (
0.1
) 4
.6 (
0.1
) 4
.2
4.3
6
.6 (
0.1
) -3
.0 (
0.1
) 3
.5 (
0.1
) -
- -
- -
β-D
-dd
T (34
) -
- -
- -
5.4
(0
.1)
-2.2
(0
.1)
3.2
(0
.1)
9.8
9
.9
3.5
(0
.1)
-1.1
(0
.1)
2.4
(0
.1)
β-D
-dd
U (35
) -
- -
- -
5.7
(0
.3)
-2.1
(0
.4)
3.6
(0
.5)
- -
- -
-
β-D
-dd
I (36
) -
- -
- -
6.2
(0
.2)
-2.9
(0
.3)
3.3
(0
.4)
- -
- -
-
β-D
-dA
(37
) -0
.7 (
0.1
) -0
.4 (
0.1
) -1
.1 (
0.1
) 3
.5
3.6
-3
.9 (
0.2
) 1
.8 (
0.2
) -2
.1 (
0.3
) -
- -
- -
3'-O
Me-β
-D-d
A (38
) -1
.4 (
0.6
) -0
.7 (
0.6
) -2
.2 (
0.2
) -
- -4
.6 (
0.4
) 1
.4 (
0.4
) -3
.2 (
0.2
) -
- -
- -
3',5
'-d
iOM
e-β
-D-d
A (39
) -1
.0 (
0.6
) -0
.7 (
0.6
) -1
.7 (
0.2
) -
3.5
-3
.6 (
0.4
) 1
.5 (
0.4
) -2
.1 (
0.2
) -
- -
- -
β-D
-dIm
b (40
) 0
.1 (
0.1
) -0
.2 (
0.1
) -0
.1 (
0.1
) 6
.0
6.0
-2
.2 (
0.1
) 0
.8 (
0.1
) -1
.4 (
0.1
) -
- -
- -
β-D
-dG
(41
) 2
.1 (
0.1
) -2
.2 (
0.2
) -0
.1 (
0.2
) 2
.3
2.2
-2
.8 (
0.2
) 1
.1 (
0.2
) -1
.7 (
0.3
) 9
.5
9.5
-4
.9 (
0.2
) 2
.2 (
0.2
) -2
.7 (
0.3
)
β-D
-dC
(42
) 0
.0 (
0.1
) -0
.8 (
0.1
) -0
.8 (
0.1
) 4
.2
4.3
-0
.7 (
0.1
) -0
.5 (
0.1
) -1
.3 (
0.1
) -
- -
- -
β-D
-T (43
) -
- -
- -
-1.4
(0
.2)
0.1
(0
.1)
-1.3
(0
.2)
9.7
9
.8
-1.9
(0
.2)
0.3
(0
.2)
-1.6
(0
.3)
β-D
-dU
(44
) -
- -
- -
-0.6
(0
.2)
-0.4
(0
.1)
-1.1
(0
.2)
9.5
9
.4
-1.3
(0
.2)
-0.1
(0
.3)
-1.4
(0
.4)
5-F
-β-D
-dU
(45
) -
- -
- -
-0.8
(0
.1)
-0.4
(0
.1)
-1.2
(0
.1)
7.8
7
.7
-1.1
(0
.1)
-0.4
(0
.1)
-1.5
(0
.1)
β-L
-dA
(46
) -1
.2 (
0.6
) 0
.1 (
0.6
) -1
.1 (
0.1
) -
- -3
.9 (
0.3
) 1
.8 (
0.4
) -2
.2 (
0.2
) -
- -
- -
β-L
-dG
(47
) 1
.3 (
0.8
) -1
.6 (
0.8
) -0
.3 (
0.1
) -
- -2
.7 (
0.3
) 0
.9 (
0.4
) -1
.8 (
0.2
) -
- -4
.3 (
0.4
) 1
.7 (
0.4
) -2
.6 (
0.2
)
β-L
-dC
(48
) -0
.2 (
0.3
) -0
.6 (
0.3
) -0
.8 (
0.1
) -
- -0
.7 (
0.3
) -0
.5 (
0.3
) -1
.2 (
0.1
) -
- -
- -
β-L
-T (49
) -
- -
- -
-1.2
(1
.1)
0.0
(1
.0)
-1.2
(1
.5)
- -
-2.3
(0
.7)
0.6
(0
.6)
-1.7
(0
.9)
β-D
-A (50
) -0
.2 (
0.1
) -0
.4 (
0.2
) -0
.5 (
0.2
) 3
.5
3.5
-4
.4 (
0.2
) 2
.6 (
0.1
) -1
.8 (
0.2
) -
- -
- -
β-D
-G (51
) 5
.4 (
0.2
) -4
.2 (
0.5
) 1
.5 (
0.5
) 2
.1
2.1
-3
.3 (
0.2
) 1
.8 (
0.2
) -1
.5 (
0.3
) 9
.5
9.6
-7
.6 (
0.5
) 4
.8 (
0.2
) -2
.8 (
0.5
)
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
60
a T
he
erro
rs a
re s
ho
wn i
n p
aren
thes
es.
-T
ΔS
° a
nd
ΔG
°ar
e at
29
8 K
exce
pt
for
17
, 1
8,
30
-3
2 (
28
8 K
) o
win
g t
o d
eco
mp
osi
tio
n a
t ac
idic
pD
36,3
7.
b
∆H
°,
-TΔ
S° a
nd
ΔG
° a
re f
rom
ref.
20
(fo
r 1
2 -
14
), r
ef.
37
(1
5,
16
, 2
1 -
29
, 3
7 -
39
, 4
6 -
49
), r
ef.
36
(1
7 -
20
, 3
0 -
34
), r
ef.4
47
(3
5),
ref
. 3
0 (
40
- 4
5,
50
- 5
5),
ref
. 3
2 (
56
- 6
2),
ref
. 2
7(a
nd
our
unp
ub
lish
ed r
esult
fo
r
acid
ic p
Ds)
fo
r 6
3.
Eal
ier
esti
mat
es f
or
30
- 3
4 i
n t
he
N s
tate
wer
e p
ub
lish
ed20.
The
larg
est
dif
fere
nce
bet
wee
n t
hem
and
the
val
ues
in t
able
1 (
fro
m r
ef.
36
) is
1.4
kJ/
mo
l fo
r Δ
HN
;�
of
33
, w
hic
h i
s nea
rly w
ithin
the
sum
of
the
erro
rs o
f th
e es
tim
ates
. c
The
pla
teau
s in
the
P,
N a
nd
D s
tate
s fo
r ∆
H°,
-TΔ
S° a
nd
ΔG
° o
f 1
8,
20
, 2
4 -
26
, 3
0 -
34
, 3
7,
40
- 4
5,
50
- 6
1
and
63
res
ult
fro
m t
he
fit
of
exp
erim
enta
l p
D-d
epen
den
t val
ues
to
Eq
12
. F
or
22
, 2
3,
27
- 2
9,
38
, 3
9,
46
- 4
9,
the
erro
rs a
re t
ho
se o
f in
div
idual
ΔH
°,
-TΔ
S°
and
ΔG
° a
t a
cert
ain p
D
(ref
. 3
7).
d T
hes
e p
Kas
wer
e ca
lcula
ted
fro
m p
lots
of
exp
erim
enta
l ∆
G°
ver
sus
pD
and
fro
m H
ill
plo
ts o
f p
D v
ersu
s lo
g(∆
∆G
° tot -
∆∆
G°
/ ∆
∆G
°) (
refs
. 3
0,
32
-37
). F
or
18
, 3
1 a
nd
32
at a
cid
ic p
Ds,
the
pK
a c
alcu
late
d f
rom
pD
-dep
end
ent
1H
chem
ical
shif
ts (
col.
6)
was
use
d a
s a
const
rain
t in
the
det
erm
inat
ion o
f ∆
H°,
-TΔ
S° a
nd
ΔG
° i
n P
, N
and
D s
tate
s.
e T
hes
e
(aver
age)
pK
a
val
ues
co
resp
ond
to
the
pD
s at
the
infl
ecti
on p
oin
t(s)
of
plo
ts o
f 1H
chem
ical
shif
ts v
ersu
s p
D (
29
8 K
fo
r 1
7 -
20
, 2
88
K f
or
30
- 3
4)
and
fro
m t
he
corr
esp
ond
ing H
ill
plo
ts.
Com
pound
Full
y p
roto
nat
ed n
ucl
eobas
e pK
a fr
om
: N
eutr
al n
ucl
eobas
e pK
afro
m:
Full
y d
epro
tonat
ed n
ucl
eobas
e
Δ
HP;�
-Τ
ΔS
P;�
Δ
GP;2
98
∆G
° d
δ
1H
e
ΔH
�
-ΤΔ
S N
;� Δ
G N
;29
8
∆G
° d
δ
1H
e
ΔH
D;�
-Τ
ΔS
D;�
ΔG
D;2
98
β-D
-C (
52
) 5
.2 (
0.2
) -3
.3 (
0.4
) 1
.9 (
0.4
) 4
.0
4.1
2
.3 (
0.1
) -0
.8 (
0.5
) 1
.5 (
0.5
) -
- -
- -
β-D
-rT
(5
3)
- -
- -
- 1
.3 (
0.1
) -1
.4 (
0.3
) -0
.1 (
0.3
) 9
.9
9.8
-0
.2 (
0.1
) -0
.2 (
0.3
) -0
.4 (
0.3
)
β-D
-U (
54
) -
- -
- -
2.0
(0
.2)
-1.7
(0
.3)
0.3
(0
.4)
9.6
9
.4
0.3
(0
.1)
-0.2
(0
.1)
0.1
(0
.1)
5-F
−β
-D-U
(5
5)
- -
- -
- 2
.3 (
0.1
) -1
.8 (
0.2
) 0
.5 (
0.3
) 8
.0
7.6
0
.8 (
0.1
) -0
.6 (
0.3
) 0
.2 (
0.3
)
Fo
rmyci
n B
(5
6)
-0.5
(0
.1)
-0.8
(0
.1)
-1.3
(0
.1)
1.4
1
.3
-8.1
(0
.1)
4.8
(0
.1)
-3.3
(0
.1)
8.9
8
.8
-8.8
(0
.1)
5.3
(0
.1)
-3.5
(0
.1)
Fo
rmyci
n A
(5
7)
-2.4
(0
.1)
0.4
(0
.1)
-2.0
(0
.1)
4.5
4
.4
-8.1
(0
.5)
4.7
(0
.4)
-3.4
(0
.1)
- 9
.5
-8.1
(0
.5)
4.7
(0
.4)
-3.4
(0
.1)
9-d
eaza
-A (
58
) -7
.4 (
0.2
) 3
.7 (
0.3
) -3
.6 (
0.1
) 5
.9
6.0
-1
4.2
(0
.4)
9.1
(0
.4)
-5.0
(0
.1)
- -
- -
-
Ψ-i
soC
(5
9)
4.2
(0
.1)
-3.8
(0
.2)
0.5
(0
.1)
3.6
3
.6
-1.9
(0
.1)
0.6
(0
.1)
-1.4
(0
.1)
9.2
9
.0
-7.9
(0
.2)
4.9
(0
.2)
-3.0
(0
.1)
Ψ-U
(6
0)
- -
- -
- 0
.7 (
0.2
) -1
.4 (
0.2
) -0
.6 (
0.1
) 9
.4
9.1
-4
.5 (
0.1
) 2
.2 (
0.1
) -2
.3 (
0.1
)
1-M
e-Ψ
-U (
61
) -
- -
- -
1.1
(0
.1)
-1.8
(0
.1)
-0.7
(0
.1)
9.9
9
.7
-3.0
(0
.1)
1.5
(0
.1)
-1.5
(0
.1)
1,3
-diM
e-Ψ
-U (
62
) -
- -
- -
2.0
(0
.1)
-2.3
(0
.1)
-0.3
(0
.1)
- -
- -
-
β-D
-3'-d
A (
63
) 6
.9 (
0.2
) -3
.2 (
0.3
) 3
.8 (
0.1
) 3
.4
3.5
1
.3 (
0.1
) 0
.5 (
0.1
) 1
.8 (
0.1
) -
- -
- -
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
61
3.7 Estimation of ∆H°, ∆S° and ∆G° of the N� S equilibrium
3.7.1 Methodology
Each calculation performed with PSEUROT (using a single set of experimental temperature-
dependent 3JHH at one particular pD) produces one set of temperature-dependent mole fractions of
the N- and S-type pseudorotamers.
Table 3. ∆H° and -T∆S° contributionsa to ∆G298 of the two-state N�S pseudorotational
equilibrium in mononucleotides 64 - 83
Compound ΔH° ∆S° -T∆S298 ∆G298 % S278 % S358 ∆ %S
β-D-dAMP (64) -5.4 (0.3) -9.9 (0.7) 3.0 (0.2) -2.4 (0.4) 76 65 -11
β-D-dGMP (65) -4.1 (0.2) -6.5 (0.7) 1.9 (0.2) -2.2 (0.3) 73 64 -9
β-D-dCMP (66) -3.2 (0.1) -6.8 (0.5) 2.0 (0.1) -1.2 (0.2) 64 56 -8
β-D-TMP (67) -2.6 (0.1) -4.3 (0.4) 1.3 (0.1) -1.3 (0.2) 65 59 -6
β-D-dUMP (68) -2.7 (0.1) -3.8 (0.8) 1.1 (0.2) -1.6 (0.3) 67 61 -6
β-D-dAMPEt (69) -5.5 (0.2) -7.9 (0.7) 2.4 (0.2) -3.1 (0.3) 81 71 -10
β-D-dGMPEt (70) -4.8 (0.2) -7.0 (0.7) 2.1 (0.2) -2.7 (0.3) 77 68 -9
β-D-dCMPEt (71) -3.6 (0.2) -5.3 (0.9) 1.6 (0.3) -2.0 (0.3) 72 64 -8
β-D-TMPEt (72) -3.0 (0.1) -3.2 (1.1) 1.0 (0.3) -2.0 (0.3) 71 65 -6
β-D-dUMPEt (73) -2.7 (0.2) -2.3 (0.6) 0.7 (0.2) -2.0 (0.3) 71 65 -6
β-D-AMP (74) -4.9 (0.4) -9.1 (0.7) 2.7 (0.2) -2.2 (0.5) 74 63 -11
β-D-GMP (75) -4.5 (0.2) -10.8 (1.6) 3.2 (0.5) -1.3 (0.5) 66 55 -11
β-D-CMP (76) 0.8 (0.2) -0.6 (1.8) 0.2 (0.5) 1.0 (0.6) 40 42 2
β-D-rTMP (77) -0.9 (0.2) -1.6 (0.9) 0.5 (0.3) -0.4 (0.3) 55 53 -2
β-D-UMP (78) 1.2 (0.2) 2.1 (1.2) -0.6 (0.4) 0.6 (0.4) 43 46 3
β-D-AMPEt (79) -6.9 (0.8) -13.6 (1.1) 4.1 (0.3) -2.8 (0.9) 79 66 -13
β-D-GMPEt (80) -5.8 (0.4) -12.3 (1.1) 3.7 (0.3) -2.1 (0.5) 74 62 -12
β-D-CMPEt (81) -1.5 (0.2) -5.3 (1.0) 1.6 (0.3) 0.1 (0.4) 50 47 -3
β-D-rTMPEt (82) -2.5 (0.3) -5.3 (1.2) 1.6 (0.4) -0.9 (0.5) 61 55 -6
β-D-UMPEt (83) -1.6 (0.1) -3.1 (0.6) 0.9 (0.2) -0.7 (0.2) 58 54 -4
a ΔH° (kJmol-1), ΔS° (Jmol-1K-1), -TΔS° (298 K, kJmol-1) and ΔG° (298 K, kJmol-1) have been taken from ref. 23
for 64 - 67, from ref. 447 for 68 and 73 and from ref. 28 for 74 - 83. For 69 - 72, the values in Table 3 are refined
(using their sodium salts) in comparison with the original data in ref. 23 (ammonium salts).
At each pD, the input file for the PSEUROT calculations has been prepared in such a way
that (i) the hyperspace of geometries that is accessible to the constrained P and Ψm values of the
minor pseudorotamer and/or to the constrained Ψm of both N- and S-type conformations was well
covered, and (ii) the errors in the experimental 3JHH were taken into account by generating ≈ 1000
sets of "randomized" temperature-dependent coupling constants for each geometrical constraint.
Therefore, the total number of calculations at each pD was typically ≈ 5000 - 20000. The mole
fractions from each calculation are used to construct a van't Hoff plot. The average slopes and
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
62
intercepts from ≈ 5000 - 20000 van't Hoff plots are used to calculate ∆H° and ∆S° (and their errors)
of the N � S equilibrium of 12 - 83 at the particular pD. The free-energy ∆G° (T), at the
temperature T, of the two-state N � S equilibrium has been calculated either by (i) adding ∆H° and
-T∆S° contributions or (ii) directly from the average [lnaverage (xS / (1 - xS))] of the 5000 - 20000
individual ln (xS / (1 - xS)). The error on ∆G° (T) is given by: σ∆G° = [σ2
∆H° + σ2
-T∆S°] (case (i)) or
σ∆G° = -R.T.lnaverage (xS / (1 - xS)) (case (ii)).
∆H°, ∆S° and ∆G° of the two-state N � S equilibrium of the sugar moiety in compounds 12 - 16, β-
D-ddU (35), β-D-ddI (36) and mononucleotides 64 - 83 (at a single neutral pD), in L-nucleosides
(27 - 29 and 46 - 49) and in D-nucleosides 22, 23, 38 and 39 (at two or three pDs, one in each of
their P, N and D states) are presented in Tables 2 and 3. For all other nucleosides, the
conformational analyses has been performed over part of or the entire 0.5 - 12.0 pD range, and we
only report in Table 2 the limiting values of the thermodynamics of their two-state N � S equilibria
in each of the P, N and D states, which have been derived according to the procedure described in
the present section.
3.7.2 Accuracy of thermodynamics
The error on ∆H°, -TΔS° and ΔG° values corresponds to the standard deviation of the
average of the corresponding individual ∆H°, -TΔS° and ΔG° values. For all compounds, the
standard deviations on ∆H°, -TΔS° and ΔG° values are typically ≈ 0.5 - 1.0 kJ/mol, which have
been estimated during pseudorotational analyses in which the 3JHH error of ≈ 0.1 Hz for all
compounds except for some36 ddNs and abasic sugars (0.2 Hz) was taken into consideration. The
experimental accuracy of pD-dependent ΔH°, -TΔS° and ΔG° values were evident from the fact that
they gave the pKa values of the nucleobase within an accuracy ±0.2 pD unit in average (in
comparison with the literature values) with some exceptions where the ΔG° change as a function of
pD was less than 1.0 kJ/mol (such as pyrimidine nucleosides).
3.7.3 Influence of λN1/9 on the thermodynamics
In their recent investigation of the effect of solvent, pH, temperature and concentration upon
the values of the substituent parameters, λ, Altona et al have shown430 that the 3JHH coupling
constant to methyl in ethylamine and propylamine in D2O increases by 0.16 Hz and 0.26 Hz,
respectively, in going from pD 12.5 (NH2) to pD 7.5 (NH3+), therefore it was suggested that the λ
value of NH2 decreases from 1.10 to 0.82 upon protonation.
In view of this observation, we have examined the influence of the selected λ value for the
glycosyl nitrogen N9 of the nucleobase in β-D-dG (41), in each of its P (pD = 1.0), N (pD = 7.5)
and D (pD = 11.4) states, on ∆H°and ∆S° contributions to the free-energy ∆G° of its two-state N �
S equilibrium. At each pD, we have performed four calculations with PSEUROT. The input for
each of these calculations only differs in the value of λ(N9) which has been taken successively as
0.3, 0.58, 0.88 and 1.2. The individual ∆H° and ∆S° values at each pD for the twelve calculations
are collected in Table 4.
A perusal of the data in Table 4 suggests the followings:
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
63
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
-2.5
-2.0
-1.5
-1.0
-0.5
77
73
69
65
60
55
50
pD
pD0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
77
69
60
50
40
31(E)
β-D-U (54 )
β-D-rT (53 )β-D-A (50 )
β-D-G (51 )
β-D-C (52 )
(D)
β-D-dG (41 )
β-D-dA (37 )
β-D-dC (42 ) β-D-dU (44 )
β-D-T (43 )
β-D-dImb (40 )
5-F-β-D-dU (45 )
5-F-β-D-U (55 )
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
-4.0
-2.0
0.0
92
83
69
50
pD
Formycin B (56 )
Formycin A (57 )
9-deaza-A (58 )
Ψ -isoC (59 )
Ψ -U (60 )
1-Me-Ψ -U (61 )
(F)
1,3-diMe-Ψ -U (62 )
Figure 11 (Cont'd): Experimental ∆G° values (298 K) of N �S equilibria in α-D-ddNs 17 - 20 [Panel (A)], α-D-
dNs 21 and 24 - 26 [Panel (B)], β-D-ddNs 30 - 34 [Panel (C)], β-D-dNs 37, 40 - 45 [Panel (D)], β-D-rNs 50 - 55 [Panel
(E)] and β-D-C-rNs 56 - 62 [Panel (F)] as a function of pD. The sigmoidal plots (except for α-D-ddA, α-D-ddC, α-D-
dA and 1,3-diMe-Ψ-U, pD-independent values) were obtained by fitting the experimental data to Eq 12, giving the
pKa(s) of the nucleobases at the inflection point(s) and limiting ∆G° values in the P, N and D states (Table 2).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
64
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
40
35
31
27
23
20
17
14
12
10
8
7
6
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
77
73
69
65
60
55
50
45
pD
α-D-ddA (17 ) & α-D-ddC (19 )
α-D-ddG (18 )
α-D-ddT (20 )
(A)
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0
-4.5
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
88
86
83
80
77
73
69
65
60
55
50
pD
α-D-dA (21 )
α-D-dG (24 )
α-D-dC (25 )
α-D-T (26 )
(B)
pD
β-D-ddA (30 )
β-D-ddG (31 )
β-D-ddC (33 )
β-D-ddT (34 )
5'-OMe-β-D-ddG (32 )
(C)
Figure 11 (See the legend p. 74)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
65
(i) The largest influence of a change of λ(N9) upon ∆H° is found at alkaline pD [∆H°λ=1.2 -
∆H°λ=0.3 = -1.3 kJmol-1], which is nearly within the sum (0.9 kJmol-1) of the standard deviations of
∆H° values at both pDs. At pD 1.0 and 7.5 ∆H° is virtually pD-independent.
(ii) By comparing the ∆H° values that have been calculated at pD 1.0 and 7.5 using 0.58 as
the value for λ(N9), we find that the N-type pseudorotamers are stabilized by ∆∆H°(P-N, λ=0.58] =
4.6 kJmol-1 in going from the N to the P state of guanin-9-yl in β-D-dG, as the result of the
strenghtening of the anomeric effect (see the discussion in section 4.8).
(iii) If we now assume that λ(N9) is reduced from 0.58 at pD 7.5 to 0.3 at pD 1.0, as suggested by
Altona et al430, and if we compare again the corresponding ∆H° values obtained with this new λ
value, we find that the extent of the stabilization of N-type conformers at pD 1.0 with respect to pD
7.5 is the same as in (i), i.e. ∆∆H°(P-N, λ(pD 1.0)=0.3] = 4.6 kJmol-1.
(iv) If we assume that λ(N9) is not affected by the deprotonation of guanin-9-yl, we find that
S-type pseudorotamers are more stabilized at pD 11.4 than at pD 7.5 by ∆∆H°(D-N, λ=0.58] = -2.2
kJmol-1.
(v) If we now assume that λ(N9) is increased from 0.58 to 0.88 in going from pD 7.5 to pD
11.4, the stabilization of S-type conformers is even greater, i.e. ∆∆H°(D-N, λ(pD 11.4)=0.88] = -2.7
kJmol-1. In conclusion, the extent of the stabilization or destabilization of N-type pseudorotamers
upon protonation and deprotonation of guanin-9-yl in β-D-dG is independent of the value of λ(N9)!
Note however that this conclusion has been only validated for our system, and should be carefully
checked for any other unproven system.
Table 4. Influence of the value of the substituent parameter for the glycosyl nitrogen λ(N9)
in β-D-dG (41) in each of the P, N and D states of the constituent guanin-9-yl upon the
thermodynamicsa of the two-state N �S equilibrium of the constituent sugar moiety
P state (pD = 1.0) N state (pD = 7.5) D state (pD = 11.4)
λ(N9) ΔH°P -TΔS°P ΔG°P ΔH°N -TΔS°N ΔG°N ΔH°D -TΔS°D ΔG°D
0.3 1.8 (0.9) -1.8 (0.9) 0.0 (0.1) -2.6 (0.3) 1.1 (0.3) -1.5 (0.2) -4.6 (0.4) 2.2 (0.4) -2.4 (0.2)
0.58 1.8 (0.9) -2.0 (1.0) -0.2 (0.1) -2.8 (0.3) 1.0 (0.4) -1.8 (0.2) -5.0 (0.4) 2.3 (0.4) -2.7 (0.2)
0.88 1.8 (1.0) -2.2 (1.0) -0.4 (0.1) -3.0 (0.4) 1.1 (0.4) -1.9 (0.2) -5.5 (0.4) 2.5 (0.4) -2.9 (0.2)
1.2 1.8 (1.0) -2.2 (1.0) -0.5 (0.1) -3.2 (0.4) 1.5 (0.3) -1.7 (0.1) -5.9 (0.5) 2.9 (0.5) -3.0 (0.2)
a In kJmol-1. -T∆S° and ∆G° in the P, N and D states are given at 298 K. All thermodynamic parameters have
been calculated using our methodology for each particular value of λ(N9) (Section 3). Standard deviations of
each thermodynamic quantity take into account the influence of the uncertainty (± 0.1 Hz) in experimental 3JHH, as discussed in Section 3.6.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
66
3.7.4 Influence of the nature of the aglycone dictates the thermodynamics
In the case of α-D-ddA (17), α-D-ddC (19) and α-D-dA (21), ∆H°, ∆S° and ∆G° were found
to be virtually pD-independent and the values reported in Table 2 for each of the P, N and D states
have been calculated by averaging the experimental data at all pDs.
On the other hand, the experimental ∆H°, ∆S° and/or ∆G° values of the N � S equilibria in
α-D-ddG (18), α-D-ddT (20), α-D-dG (24), α-D-dC (25), α-D-T (26), β-D-ddNs 30 - 34, β-D-dNs
37 and 40 - 45, β-D-rNs 50 - 55, β-D-ribo-C-nucleosides 56 - 61 and β-D-3'-dA (63) are dictated by
the pD of the aqueous solution30,32,36,37,426, as shown by the sigmoidal plots in Fig 11. The curves
through the experimental points in Fig 11 result from a nonlinear least-squares fitting of the
experimental data to the Henderson-Hasselbach equation (Eq 12), which gives limiting values of the
thermodynamics of two-state N � S equilibria in each of the P and D states and the pKa(s) of the
constituent nucleobase at the inflection point(s). [Note that the errors on the limiting ΔH°, -TΔS°,
and ΔG° values reported in Table 4 actually correspond to standard deviations of averages of
individual experimental ∆H°, ∆S° and ∆G° values at several pDs on the plateau in this particular
state, whereas the standard deviation for ∆H°, ∆S° and ∆G° values at each pD are higher (typically ≈
0.5 - 1.0 kJmol-1), as shown in the Tables of pD-dependent thermodynamics in the original
publications.]
pD = pKa + ]AH[
]A[log
+
= pKa + α
α−1log ..... Eq 12
In Eq 12, α represents the fraction of the protonated species, and it has been calculated from
the change in experimental ∆H°, ∆S° and ∆G° value at a given pD relative to the reference neutral
state divided by the total change in their respective values between the N state and the P or D state.
The pKa(s) of the nucleobase in 18, 20, 24 - 26, 30 - 34, 37, 40 - 45, 50 - 61 and 63 has been
determined either (i) directly from the nonlinear fit of the pD-dependent experimental ∆G° values to
Eq 12, as discussed above, or (ii) from Hill plots of pD as a function of the logarithm of the ratio of
the protonated to the unprotonated species. The Hill plots gave straight lines with Pearson's
correlation coefficients typically above 0.9 and slopes close to 1, which is characteristic of a
protonation � deprotonation equilibrium involving a single protonation site (see Table 2, cols. 5
and 10 for the pKa(s) of nucleobases derived from pD-dependent ∆G° values). The pKa values
determined from pD-dependent ∆G° values of the N � S equilibrium in nucleosides fit very well to
the literature values as well as those obtained from pD-dependent proton chemical shifts.
3.8 New Karplus equation to interpret 3JHF coupling constants
Substitution of a hydroxyl group or hydrogen in nucleosides for a fluorine atom does not
change significantly the steric effect of the functionality, however, the strongly electronegative
fluorine atom involved in powerful gauche interactions governs the overall conformation of the
sugar moiety199,206,210,222,448-449 Fluorine has been widely incorporated into nucleosides to design
new therapeutic agents (e.g. FLT (88): anti-HIV activity450; diFC (87): cytotoxic against Chinese
hamster ovary and tumor cells451; F2"C (112): inhibits the growth of human lymphoblastic cell
lines452) with specific puckering modes453-457. Incorporation of 2'-deoxy-2'-fluoro nucleosides in
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
67
DNA458 stabilizes the antisense DNA:RNA duplex, which adopts an A-form conformation, owing
to the stabilization of N-type sugars by the [F2'-C2'-C1'-O4'] gauche effect459. 2'-fluoro-2'-
deoxyguanosine in a hexadeoxynucleotide induces a A-DNA � Z-DNA conformational
transition460. Incorporation of 2'-fluoro-2'-deoxyarabinothymidine (S-type conformation) in the
Dickerson-Drew dodecamer stabilizes the duplex, but 2'-fluoro-2'-deoxyribothymidine destabilizes
it by inducing the formation an A-B junction in the middle of the duplex64. When nucleosides are
highly substituted (such as difluoronucleosides 84 - 87), it is not to assess the conformation of their
sugar moieties on the basis of a single or a few 3JHH (in 84 - 87, only 3J3'4' is available). 3JHF
coupling constants have only been qualitatively used to assess the structure and conformation of
pyranose derivatives461-463 and fluoronucleosides204,235 owing to the fact that no Karplus-type
equation allows to translate quantitatively experimental 3JHF data into HCCF torsion angles (ΦHF).
Several simplified equations such as 3JHF = A cos2 ΦHF + B cos ΦHF + C taking into account only
the torsion angle dependence were parametrized464-468.
3.8.1 Dataset of (3JHF,ΦHF) pairs for monofluoronucleosides
We have parametrized a Karplus equation for the interpretation of 3JHF coupling constants
using the strategy depicted in Scheme 4. We have initially performed pseudorotational analyses of
temperature-dependent 3JHH for monofluoronucleosides (88 - 98) using the PSEUROT203,427
program (steps 1 - 3 in Scheme 4). In the second step in PSEUROT (step 2 in Scheme 4), νi torsion
angles are translated into the corresponding HCCH torsion angles using simple a relationships:
ΦHH = AH νi + BH. AH and BH were obtained from ab initio calculations using the GAUSSIAN 94
program298 on 88, 90 - 92 and 95 - 98 (Table 3 and experimental Section in ref. 39). The plots of
ΦHH versus νi for 88, 90 - 92 and 95 - 98 as extracted from their ab initio optimized geometries
gave straight lines with correlation coefficients above 0.95.
From the results of these initial pseudorotational analyses, the following conclusions could
be drawn: (i) All compounds are strongly conformationally biased either to the N- or S-type sugar
geometries. (ii) In all nucleosides, the presence of strongly electronegative fluorine atom at C2' or
C3' on the α- or β-face makes the C-F bond adopt a gauche orientation with C1'-O4' or C4'-O4'. (iii)
This gauche effect is stronger over any other stereoelectronic effect and is responsible for the bias of
the sugar conformational equilibrium toward N- or S-type pseudorotamers, which is evidenced by
the following observations: In F3"ddU (91) and F3'ddU (98), the same anomeric effect stabilizes N-
type conformers. However, S-type pseudorotamers are preferred in 91 owing to predominant [F3"-
C3'-C4'-O4'] gauche effect favouring S-type sugars, whereas 98 is locked in N-type conformation
owing to cooperative anomeric effect and [F3'-C3'-C4'-O4'] gauche effect. As in F3"ddU (91), the
pentofuranose sugar in FLT (88) and its derivatives 89 and 90 prefers S-type conformations owing
to the stronger [F3"-C3'-C4'-O4'] gauche effect. In F2'ddU (96), the [F2'-C2'-C1'-N1] fragment is in
gauche orientation both in N- and S-type pseudorotamers, and the stabilization of S- over N-type
sugar is owing to the stronger [F2'-C2'-C1'-O4'] gauche effect over the anomeric effect. In F2"ddU
(97) the combined influence of the [F2"-C2'-C1'-O4'] gauche effect and of the anomeric effect
overrides the counteracting [F2"-C2'-C1'-N1] gauche effect and N-type conformers are preferred.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
68
Basing on the results (PN, PS, Ψm(N), Ψm(S) and temperature-dependent xS) of these
pseudorotational analyses, we have built a dataset composed of 29 pairs of (3JHF,ΦHF) for the
major pseudorotamers of 88 - 98 in two steps (steps 4 and 5 in Scheme 4): (i) From linear
relationships ΦHF = AF νi + BF (based on torsion angles extracted from ab initio optimized
geometries of 88, 90 - 92 and 95 - 98, vide supra) we calculated ΦHF torsion angles from the
endocyclic torsions νi for each HCCF fragment in the major pseudorotamers of 88 - 98. (ii) We
found a linear correlation between temperature-dependent xS and 3JHF values, which allowed us to
calculate limiting 3JHF values for pure N- or S-type (major) pseudorotamers in 88 - 98 (Table 4 in
ref. 39).
An analysis of the statistical distribution of the 29 pairs of (3JHF,ΦHF) for 88 - 98 revealed
that the torsion angles are not distributed evenly over the whole 0 - 360° range, instead most of
them are found in the cis and trans regions. Additionally, neither torsion angles around ≈ ± 90° nor 3JHF above 45.3 Hz were present in this dataset. It is however known that the limiting 3JHF
coupling constants can be ≈ 45 - 47 Hz for the trans substitution pattern of decalins469,470.
Therefore a parametrization based on this dataset would not have allowed to define precisely the
position of the minima (≈ ± 90°) and maxima (≈ 180°) of the Karplus curve in these regions. We
have therefore extended (step 6 in scheme 4, Ref. 39) our dataset by incorporating 28 additional
pairs of (3JHF,ΦHF) from conformationally constrained (cyclic) compounds 99 - 109466,467,469-471.
ΦHF were extracted from the structures of 99 - 109, whose geometry was optimized ab initio with
GAUSSIAN 94298. Among these 28 pairs, 3 (entries #31, 33, 42 in Table 4 in ref. 39) allowed to
precise the (3JHF,ΦHF) values around ± 90° while 9 (entries #43, 44, 47, 48, 51 - 54, 57) were found
in the trans region (with 3JHaxF = 46 Hz in 109).
3.8.2 Parametrization of the Karplus equation
A perusal of the 57 pairs of (3JHF,ΦHF) in Table 4 (in ref. 39) shows variations of up to 16
Hz between the value of 3JHF either in the cis or trans region for the same torsion angle in different
compounds (compare 3JH2"F3" in F3AT (95) and 3JH3'F2' in F2'ddU (96), on one hand; 3JH2'F3" in
AFLT (90) and 3JH4'F3' in FXA (92)). This can be attributed to different substitution patterns on
H2"-C2'-C3'-F3" in 95 and H3'-C3'-C2'-F2' in 96, in one hand, and H2'-C2'-C3'-F3" in 90 versus
H4'-C4'-C3'-F3' in 92, in the other. This led us to conclude that a simple 3-term Karplus-type
equation (i.e. of the form 3JHF = A cos2 ΦHF + B cos ΦHF + C) will not reproduce our experimental
3JHF accurately on the basis of the torsion angle value alone. It has been suggested that the
electronegativity of the substituents on the HCCF fragment affects the value of the 3JHF coupling
constant461,472. In order to account for the influence of the nature and configuration of the
substituents on the HCCF fragments on the value of 3JHF, we have first parametrized (step 7 in
scheme 4) a Karplus-type equation, whose form is that originally proposed by Altona et al (Eq
8a)431 (Section 3.4). In Eq 8a, we have however used λ substituent parameters429 to account for the
effect of the electronegativity of the substituents on the H-C-C-F fragments instead of the original
Huggins group electronegativities (Δχi(g) ).
Through a MonteCarlo fitting procedure, we optimized the values of P1 - P6. Large
discrepancies up to 10 Hz between experimental 3JHF and those back-calculated with help of Eq 8a
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
69
were found, suggesting that its form is not adequate. By examining the variations in 3JHF values for
a specific ΦHF value, we then realized that the coupling constants are also dictated by the values of
HCC and FCC bond angles, as suggested in early qualitative studies466-468. After examining several
forms of equations to reproduce the effects of bond angle, we found that it is best reflected in a
cosine square function of the torsion angle (Eq 8c, Section 3.4). The r.m.s of the fit of our
experimental data to the coupling constants calculated using Eq 8c is 1.38 Hz and the largest
individual error is 2.9 Hz (Table 4 in ref. 39). Thus, the accuracy of our newly developed Karplus
equation (r.m.s. ≈ 1.4 Hz for 3JHF in the trans region ≈ 46 - 47 Hz, i.e. ≈ 3 % error) is comparable to
that used commonly to interpret 3JHH data431 (r.m.s. ≈ 0.4 - 0.5 Hz for 3JHH in the trans region ≈ 8 -
9 Hz, i.e. ≈ 4 % error). Other combinations of parameters in Eq 8c led to similar r.m.s error,
however, plots of 3JHF for particular substitution patterns (in particular highly substituted, with
substituents of high electronegativity) a function of torsion angle showed negative (below -0.5 Hz) 3JHF values for ΦHF ≈ ± 90° (Fig. 1 in ref. 39).
3.8.3 Pseudorotational analyses of 3JHF in fluoronucleosides to validate the Karplus
equation.
In order to demonstrate that Eq 8c can be used as a reliable tool in conformational analysis of
fluoronucleosides, we have subsequently proceeded in four steps:
Plots of 3JHF versus 3JHH for qualitative conformational analysis of 88 - 98: We have first
constructed plots of 3JHF versus 3JHH data for different P values in the 0° to 360° range (at different
Ψm) for mononucleosides 88 - 98 (Figs 3 and 4 in Ref. 39). The position of the experimental
temperature-dependent pairs of (3JHF, 3JHH) for 88 - 98 suggested a high preference of their
pentofuranose moieties for N- and S-type pseudorotamers and the extent of this preference was in
good agreement with that obtained from our initial pseudorotational analyses based on 3JHH data
only.
Pseudorotational analyses on 88 - 98 based on a combination of 3JHF and 3JHH data (Steps 8 -
10 in scheme 4): We have slightly modified (see the experimental Section on "PSEUROT+JHF" in
Ref. 39 and our web site) the original PSEUROT version 3B program in order to make it possible to
perform a pseudorotational analysis using either (i) temperature-dependent 3JHH data or
temperature-dependent 3JHF data alone or (ii) a combination of 3JHH and 3JHF coupling constants.
(i) The results (PN, PS, Ψm(N), Ψm(S) values and temperature-dependent xS as well as r.m.s.
error and individual errors in 3JHF or 3JHH) of the pseudorotational analyses performed on 88 - 98
are compiled in Table 5 in Ref. 39. The comparison of the results from analyses based on 3JHH or
3JHF data only shows that the largest difference between P values is 35° (for AFLT), whereas for
Ψm values it is at most 13° (for AFLT), and for xS values we found a largest error of 13 % unit (for
F2'ddU). The r.m.s. error of both types of analysis is similar and ≤ 0.8 Hz. The largest individual
error between experimental and back-calculated 3JHH and 3JHF is 1.2 Hz and 2.0 Hz, respectively.
(ii) The analyses based on 3JHH and 3JHF data have been performed in two ways. During the
multilinear fitting process with PSEUROT+JHF, the errors in 3JHH and 3JHF data have been either
attributed the same "scale factor (1.0) or in a second series of analyses, the error in 3JHF was scaled
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
70
down to 20 % of that in 3JHH in order to account for the fact that the former are in average ≈ 5 times
greater than the latter. The introduction of scale factors did not affect the geometry of the
pseudorotamers in the final output of the PSEUROT calculation. The analyses based on 3JHH and
3JHF show that the P and Ψm values obtained from the analyses described in step (i) and (ii) are
comparable r.m.s. errors below 1.2 Hz were found.
Qualitative and quantitative pseudorotational analyses on monofluoronucleosides 111, 112
and difluoronucleosides 84 - 87: In order to further verify the validity of Eq 8c, we have applied it
to the conformational analysis of another set of monofluoro (111 and 112) and difluoronucleosides
(84 - 87) through qualitative plots of 3JHF as a function of 3JHH data (vide supra, Fig 5 in Ref. 39)
for phase angles values from 0° to 360° or to the quantitative pseudorotational analysis using our
modified PSEUROT+JHF program (Table 5 in Ref. 39). The position of the experimental pairs of
(3JHF, 3JHH) data suggest that the pentofuranose sugar adopts almost exclusively (or predominantly)
N-type conformations in F3'ddA (or in F2"C). For 84 - 87, conformational analyses based on the
single 3JH3'H4' are at best misleading. Our qualitative plots for 84 - 87 show for the first time that
the conformation of their sugar moieties is strongly biased toward N-E-type pseudorotamers (50° <
P < 90°), as suggested by the clusering of all experimental coupling constants in this region. Our
more elaborate pseudorotational analyses on 3JHH and/or 3JHF data for 84 - 87, 111 and 112 precises
the conclusions drawn on the basis of the above plots. Note that the results of the pseudorotational
analyses are again virtually independent of the type of vicinal coupling constants used (3JHH and/or
3JHF) (i) The sugar moiety in F3'ddA (111) is indeed locked in N-type conformation (≥ 94 % at 298
K) with PN in the range from 11° to 20° while Ψm is between 34° and 37°. The r.m.s. of all analyses
is ≤ 0.7 Hz with -1.3 Hz as the largest error in 3JHH and 3JHF in analyses performed using 3JHH and
3JHF data. (ii) Similarly, N-type pseudorotamers are strongly preferred in F2"C (112) (by > 73%)
with PN ≈ 36° and Ψm(N) ≈ 33° - 37°. The r.m.s. error is ≤ 1.7 Hz for analyses based on all coupling
constants with 3.2 Hz as the largest individual error in 3JHF. (iii) In difluoronucleosides 84 - 87,
C4'-exo to O4'-endo-C4'-exo pseudorotamers (61° < P < 76°, Ψm ≈ 35° - 46°) are favoured by 78 -
94 %. The r.m.s. of all calculations is below 1.5 - 2.0 Hz, while individual errors are at most 3.4 Hz
for 3JHF data.
In conclusion, this is the first report of an accurate Karplus-type equation for the
interpretation of 3JHF coupling constants in terms of torsion angles that can be successfully used to
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
71
define the solution conformational characteristics and relative populations of conformers engaged in
an equilibrium either in nucleosides, nucleotides or organic compounds in general.
4. The quantitation of the stereoelectronic effects in nucleos(t)ides
A study of our first experimental thermodynamic estimates14 of pseudorotations, as early as
1987, on 9-(2'-deoxy-β-D-threo-ribofuranosyl)-adenine and 9-(3'-deoxy-β-D-threo-ribofuranosyl)-
adenine and their comparison with the natural 2'-deoxyadenosine and 3'-deoxyadenosine revealed
that the net result of the gauche effect and the anomeric effect is of major importance in determining
the overall furanose conformation in nucleosides. This has subsequently led us to examine the
interplay of gauche and anomeric effects determining overall sugar conformation in nucleosides, in
general, more in details. The strength of the gauche and anomeric effects operating in 12 - 83 is
reflected in the value of ∆H° contribution to ∆G° of their N � S equilibrium (N � E in 12 and
1420). Our ∆H° estimates correspond to the overall contribution (i.e. stereoelectronic and
counteracting steric effect) of each molecular fragment. -T∆S° of the N �S equilibrium in 12 - 83
shows the difference in the entropy between both pseudorotamers at the temperature T. The effect of
a possible change in the solvation and/or bulk of the nucleobases or other substituents at C2'-C5' in
17 - 83 at different pDs (and/or owing to different configuration at C1') on the modulation of ∆H°
and/or -T∆S° of their N � S equilibria in either of their P or D states compared with the N state
cannot be assessed in a straightforward quantitative manner.
4.1 Quantitation of the anomeric and gauche effects by regression analyses
To quantitate the gauche and anomeric effects that drive the sugar conformation in 12 - 73,
we have initially correlated their structural features with the experimental values of their two-state N
� S equilibria via the regression analyses (A) - (E) (using the program SYSTAT473 (Table 5)).
4.1.1 Stereoelectronic effects in neutral β-D-Ns
The 26 ΔH°N values of the N � E equilibrium (in 12 and 14) or N � S equilibrium in 13, 15
and 16, β-D-ddNs 30 - 35, β-D-dNs 37 - 39 and 41 - 45, β-D-rNs 50 - 55 and β-D-3'-dA (63) have
been dissected into following components: (i) The (steric + gauche) effect of 5'CH2OH; (ii) The
effect of 5'CH2OMe compared with 5'CH2OH; (iii) The [HO2'-C2'-C1'-O4'] gauche effect; (iv) and
(v) The [O2'-C2'-C1'-N9(purine)] and [O2'-C2'-C1'-N1(pyrimidine)] gauche effects for purine and
pyrimidine ribonucleosides, respectively; (vi) The [HO3'-C3'-C4'-O4'] gauche effect; (vii) The
[MeO3'-C3'-C4'-O4'] gauche effect and (viii) - (xiii) the overall (steric + stereoelectronic) effects of
adenin-9-yl, guanin-9-yl, cytosin-1-yl, thymin-1-yl, uracil-1-yl, 5-fluorouracil-1-yl. The ΔH°N
values of β-D-ddI (36) and β-D-dImb (40) were not used, since no other nucleoside has
hypoxanthin-9-yl or imidazol-1-yl as nucleobase. The Pearson's correlation coefficient of regression
analysis (A), performed using these 26 ΔH°N values, is 0.993 and the standard error of estimates 0.6
kJmol-1. The largest individual errors between experimental and back-calculated [using estimates
from regression (A)] ΔH°N values was found for 12 (-1.2 kJmol-1) and β-D-ddA (30) (1.0 kJmol-1).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
72
Table 5. Estimationa of the strengths of the anomeric and gauche effects driving the sugar
conformation in 12 - 73 from regression analysis with the SYSTAT program473
Regression model (A) (B) (C) (D) (E)
SE 5'CH2OH 1.6 (0.4) 1.6 (0.3) 0.1 (0.3) 1.6 (0.3) 1.7 (0.3)
SE 5'CH2OMe 0.7 (0.4) 0.7 (0.4) 0.5 (0.4) 0.9 (0.6) 0.6 (0.4)
AE (A) 0.9 (0.4) 0.9 (0.3) -1.7 (0.3) 1.2 (0.3) 1.0 (0.3)
AE (G) 2.0 (0.5) 2.0 (0.4) -0.6 (0.4) 1.5 (0.4) 2.0 (0.4)
AE (C) 4.4 (0.5) 4.3 (0.4) -3.2 (0.3) 3.7 (0.4) 3.9 (0.4)
AE (T) 3.5 (0.5) 3.5 (0.4) -0.6 (0.3) 2.1 (0.4) 3.6 (0.4)
AE (U) 4.1 (0.5) 4.1 (0.4) - 3.5 (0.6) 4.0 (0.4)
AE (5-FU) 4.1 (0.6) 4.1 (0.5) - 3.2 (0.8) 4.0 (0.5)
GE[HO3'-C3'-C4'-O4'] -6.3 (0.3) -6.4 (0.2) -3.9 (0.2) -5.4 (0.4) -6.3 (0.2)
GE[MeO3'-C3'-C4'-O4'] -6.7 (0.5) -6.7 (0.4) -4.7 (0.4) -6.9 (0.6) -6.7 (0.4)
GE[-1/-2RO3PO3'-C3'-C4'-O4'] - - - - -8.3 (0.3)
GE[O2'-C2'-C1'-O4']] 5.1 (0.7) 5.1 (0.6) - 4.2 (1.0) 5.1 (0.6)
GE[O2'-C2'-C1'-N9(pur.)] -5.9 (0.8) -5.9 (0.7) - -5.6 (1.1) -6.0 (0.7)
GE[O2'-C2'-C1'-N1(pyr.)] -2.5 (0.8) -2.4 (0.7) - -1.6 (1.1) -2.3 (0.7)
Multiple R 0.993 0.993 0.992 0.974 0.992
σ estimates 0.6 0.5 0.4 1.0 0.5
a See the text for the assumptions used to build each regression model. All estimates are given in kJmol-1 (their standard
deviations are indicated in parentheses). SE 5'CH2OH designates the overall (steric + gauche) effect of 5'CH2OH
substituent, whereas SE 5'CH2OMe represents the additonal stabilization of N-type conformations by 5'CH2OMe in
comparison with 5'CH2OH. AE (A), AE (G), AE (C), AE (T), AE (U) and AE (5-FU) give the overall strength of the
(steric + stereoelectronic) effect of adenin-9-yl, guanin-9-yl, cytosin-1-yl, thymin-1-yl, uracil-1-yl and 5-fluoro-uracil-1-
yl in 17 - 73. GE[HO3'-C3'-C4'-O4'], GE[MeO3'-C3'-C4'-O4'] and GE[O2'-C2'-C1'-O4'] are estimates of the strength of
the [HO3'-C3'-C4'-O4'], [MeO3'-C3'-C4'-O4'] and [O2'-C2'-C1'-O4'] gauche effects. GE[-1/-2RO3PO3'-C3'C4'-O4'] in
regression analysis (E) represents the magnitude of the [RO3PO3'-C3'C4'-O4'] gauche effect in β-D-dNMPs 64 - 68 (R =
H, charge = -1 and -2 in 1:1 ratio, Section 7) and β-D-dNMPEt (R = Et, charge is -1) 69 - 73, which has been assumed
to be the same for each nucleobase and independent of the nature of R. GE[O2'-C2'-C1'-N9(pur.)] and GE[O2'-C2'-C1'-
N1(pyr.)] represent the strength of the [O2'-C2'-C1'-N9(pur.)] and [O2'-C2'-C1'-N1(pyr.)] gauche effects in purine and
pyrimidine ribonucleos(t)ides, respectively. The multiple R is the Pearson's correlation coefficient. "σ estimates"
corresponds to the average standard deviation of all the estimated stereoelectronic effects for each regression analysis.
4.1.2 Effect of the 5'CH2OH versus 5'CH2OMe
According to regression analysis (A), the steric and [O5'-C5'-C4'-O4'] gauche effects of
5'CH2OH drive the conformation of abasic sugars and nucleosides toward the N [∆H°(5'CH2OH) =
1.6 kJmol-1]. Substitution of 5'CH2OH in 15, 31 and 38 for 5'CH2OMe in 16, 32 and 39 slightly
reinforces the drive of sugar conformation toward N-type forms [i.e. ∆H°(5'CH2OMe) >
∆H°(5'CH2OH)]. The stronger influence of 5'CH2OMe than 5'CH2OH might be attributed to the
increased steric bulk of OMe compared with OH. ∆H°(5'CH2OH) is stronger (by 1.2 kJmol-1) than
the experimental ΔH°N value of 12, where only 5'CH2OH controls the sugar conformation. This
might suggest that the effect of 5'CH2OH is stronger in β-D-nucleosides than in 12, owing to the
proximity of the nucleobase in the former. However, the contribution of 5'CH2OH to the drive of
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
73
the sugar conformation in β-D-nucleosides is much weaker than that of all other stereoelectronic
effects.
4.1.3 Effect of the nucleobase
The nucleobase promoted drive of the N � S equilibrium in nucleos(t)ides is the result of
two counteracting contributions: stereoelectronic interactions within the O4'-C1'-N1/9 fragment, i.e.
nO4' →σ∗C1'-N1/N9 interactions and/or electrostatic repulsions between the dipole of the furanose
ring and the C1'-N1/9 dipole, and the inherent steric effect of the nucleobase (Section 2.7).
Stereoelectronic interactions are expected to drive the sugar conformation toward N-type
pseudorotamers, in which the nucleobase is pseudoaxial, whereas the steric effect stabilizes S-type
forms in which the nucleobase adopts a pseudoequatorial orientation (Section 2.7). The fact that
ΔH° estimates resulting from regression analysis (A) for the overall effects of the nucleobases in β-
D-ddNs, β-D-dNs, β-D-rNs and 63 have been found positive shows that the C4'-O4'-C1'-N1/9
fragments adopt preferentially a gauche over trans orientation owing to predominating
stereoelectronic over steric interactions, therefore an anomeric effect is operating. The magnitude of
the overall effect of the nucleobase is dictated by its electronic and steric nature, which are as
follows: adenin-9-yl (0.9 kJmol-1) < guanin-9-yl (2.0 kJmol-1) < thymin-1-yl (3.5 kJmol-1) < uracil-
1-yl (4.1 kJmol-1) ≈ 5-fluoro-uracil-1-yl (4.1 kJmol-1) < cytosin-1-yl (4.4 kJmol-1). The effects of
thymin-1-yl, uracil-1-yl, 5-fluorouracil-1-yl and cytosin-1-yl are stronger than those of adenin-9-yl
and guanin-9-yl because in the former, N1, part of the electron-deficient pyrimidine ring, is more
predisposed to accept electrons through nO4'→σ∗C1'-N orbital interactions than N9 in the latter
which is part of the electron-rich imidazole ring. Similar correlations between the strength of the
anomeric effect involving a substituent X at C2 in heterocyclic six-membered rings and the
electronegativity of X have already been established (Section 1.5). However, it seems that 5-
fluorouracil-1-yl behaves as an exception, since its overall effect is the same as that of uracil-1-yl,
although one would expect fluorine to reinforce the electron-withdrawing character of its glycosyl
nitrogen N1 and therefore its nO4' →σ∗C1'-N1 stereoelectronic interactions.
4.1.4 Gauche effects
The conformation of the pentofuranose moiety in β-D-dNs is driven toward the S owing to
the fact that the [HO3'-C3'-C4'-O4'] (or [MeO3'-C3'-C4'-O4'] in 38 and 39) gauche effect prevails
over the counteracting overall effect of their nucleobases, as evident from the comparison of their
respective ∆H° estimates. The magnitudes of the [HO3'-C3'-C4'-O4'] and [MeO3'-C3'-C4'-O4']
gauche effects is almost the same, which is consistent with the comparable electronegativities of 3'-
OH and 3'-OMe26. In β-D-rNs compared with β-D-dNs, two additional gauche effects operate
within [HO2'-C2'-C1'-O4'] and [HO2'-C2'-C1'-N1/9] fragments. Our regression analysis shows that
∆H° of the [HO3'-C3'-C4'-O4'] and [HO2'-C2'-C1'-O4'] gauche effects in β-D-rNs almost cancel
each other, which is consistent with the fact that the experimental ΔH°N values of the two-state N �
S equilibria in abasic sugars 12 and 14 are identical. Therefore, the actual preference (as reflected in
their respective experimental ΔH°N values) of β-D-rNs for N- or S-type conformations can be
attributed to the fine balance of the overall effect of their nucleobases and of the [HO2'-C2'-C1'-
N1/9] gauche effect. In purine nucleosides β-D-A and β-D-G, N9 is part of the π-electron-rich
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
74
imidazole ring, and its ability to participate to a gauche interaction with O2' is much greater than
that of N1 in the π-electron deficient pyrimidine counterparts 52 - 55. This is reflected in the
stronger [HO2'-C2'-C1'-N9(purine)] gauche effect compared to the [HO2'-C2'-C1'-N1(pyrimidine)]
gauche effect. In sharp contrast, the ability of the purine nucleobase in β-D-A and β-D-G to shift the
pseudorotational equilibrium of the constituent pentofuranose sugar toward N-type conformations is
clearly reduced with respect to that of the pyrimidine ring in 52 - 55. Consequently, ΔH°N values of
the N � S equilibria in purine and pyrimidine β-D-rNs reflect the stabilization of S- and N-type
pseudorotamers, respectively. In β-D-3'-dA, no [HO3'-C3'-C4'-O4'] gauche effect is operating and
the pseudorotational equilibrium is driven toward N-type forms owing to the cooperativity of the
effects of the nucleobase, the 5'CH2OH substituent and of the [HO2'-C2'-C1'-O4'] gauche effect,
which override the counteracting [O2'-C2'-C1'-N9(purine)] gauche effect.
4.1.5 Energetic equivalence of mirror-image β-D-dNs and β-L-dNs
The mirror image relationship of the natural DNA and of the unnatural β-L-counterpart 474-
476 has been proven by the identical X-ray crystal structures of D- and L-d(CGCGCG)
hexamers474,477. The oligomerization of activated guanosine mononucleotides in the presence of
poly(C) template is easily achieved if both the template and the monomers are of the same
handedness, whereas it is much less efficient with the racemic D/L mixture of the monomer, since
monomers of opposite handedness act as chain terminators due to their incorporation at the 2'(3')
end of the oligomer478. The comparison of the thermodynamics of the two-state N � S equilibria in
β-D-dNs 37 and 41 - 43 and in their β-L-dNs counterparts 46 - 49 (either with neutral, fully
protonated or deprotonated nucleobases) shows that D- and L-nucleosides cannot be energetically
distinguished within the timeframe and accuracy of our method based on pseudorotational analyses
of 3JHH, since they exhibit virtually identical pD-dependent conformational preferences (Table 2).
The largest individual difference (0.8 kJmol-1) is found between ΔH°N values of β-D-dG (41) and
β-L-dG (47) with fully protonated guanin-9-yl and it is within the accuracy (σ = 0.8 kJmol-1) of the
experimental ΔH°N value of β-L-dG. This suggests that the magnitudes of the stereoelectronic
anomeric and gauche effects that drive the sugar conformation in D- and L-nucleosides are identical.
In order to improve the statistical significance of the estimates derived from the regression analysis
(A), we have performed a second multilinear regression analysis [regression (B)] basing on an
extended dataset consisting of 30 experimental ΔH°N values (i.e. 26 from the dataset used for
regression (A) as well as the 4 ΔH°N values for β-L-dNs 46 - 49). The estimates for the gauche and
anomeric effects from regression (B) are virtually identical (within ± 0.1 kJmol-1) to those derived
from regression (A) (Table 5, col. 2). The Pearson correlation coefficients of regressions (B) and (A)
are nearly identical (0.993 versus 0.994). The standard error of estimates determined from
regression (B) is 0.5 kJmol-1 [as compared with 0.6 kJmol-1 for regression (A)]. The largest error
between experimental and theoretical ΔH°N values [i.e. calculated using the estimates from
regression (B) or (A)] for nucleosides is found for β-D-ddA (1.0 kJmol-1). For all β-L-dNs
[regression (B)], the error is within 0.3 kJmol-1.
4.1.6 3'-phosphate has stronger gauche effect than 3'-hydroxy
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
75
The magnitudes of the gauche and anomeric effects operating in all β-D/L-nucleosides in β-D-
dNMPs 64 - 68 and β-D-dNMPEts 69 - 73 (Table 5) have been derived from the regression analysis
(E), which was performed using 40 experimental ΔH°N values assuming identical strengths for the
[-1/-2HO3PO3'-C3'-C4'-O4'] and [-1EtO3PO3'-C3'-C4'-O4'] gauche effects in 64 - 68 and 69 - 73.
The estimates resulting from regression analysis (E), its correlation coefficient and the standard
error on the estimates are virtually the same as of regression (B) (Table 5). The average [3'-
phosphate-C3'-C4'-O4'] gauche effect is about -8.3 kJmol-1, i.e. -2.0 kJmol-1 stronger than the
[HO3'-C3'-C4'-O4'] counterpart.
4.1.7 Stereoelectronic effects in neutral α-D-N
In order to quantitate the anomeric and gauche effects driving the conformation of α-
nucleosides in the N state and compare them with those obtained for β-nucleosides [regression
(B)]), we have performed the regression analysis (C) using 17 ΔH°N values for abasic sugars 12, 13,
15 and 16, α-D-ddNs 17 - 20, α-D-dNs 21 - 26 and α-L-dNs 27 - 29. These ΔH°N values have been
correlated with the magnitudes of: (i) the effect of 5'CH2OH; (ii) the additional influence of
5'CH2OMe compared with 5'CH2OH; (iii) the [HO3'-C3'-C4'-O4'] gauche effect, (iv) the [MeO3'-
C3'-C4'-O4'] gauche effect and the overall effects of (v) adenin-9-yl, (vi) guanin-9-yl, (vii) cytosin-
1-yl and (viii) thymin-1-yl. Since, as in the β-anomers, the ∆H° values of the two-state N � S
equilibrium in each of the α-D-/-L- pairs are identical [within the accuracy of the individual
estimates, i.e. with 1.0 kJmol-1 as largest deviation for ΔH°N of α-D-dA (21) versus α-L-dA (27)],
we have considered that the anomeric and gauche effects operate with an identical strength in both
α-D- and -L-series. The correlation coefficient of the regression analysis (C) is 0.992 [0.993 for
regression (B)] and the standard error of the estimates is 0.4 kJmol-1 (0.5 kJmol-1 for regression
(B)]. As a result, the residual errors between theoretical ∆H° values for 12, 13 and 15 - 29
[calculated using the estimates derived from the regression analysis (C)] and the corresponding
experimental ΔH°N values are well within the accuracy of the later.
4.1.8 Weakening of the effects of 5'CH2OH and 5'CH2OMe in α-nucleosides
The comparison of the data in cols. 3 and 4 for regressions (B) (β-nucleosides) and (C) (α-
nucleosides) in Table 5 suggests that in α-nucleosides, the contribution of the effects of 5'CH2OH
and 5'CH2OMe to the drive of the two-state N �S equilibrium is reduced in comparison with the β-
counterparts. This is possibly due to the fact that in the former the 5'-substituent and the nucleobase
are on opposite faces of the pentofuranose ring whereas they are both on the β-face in the later.
4.1.9 Weakening of the effect of the nucleobase in α-nucleosides
The data in Table 5 indicate that the magnitude of the overall effect of the nucleobase is
weaker in α- compared with β-nucleosides by 1.2 kJmol-1 for cytosin-1-yl, 1.4 kJmol-1 for guanin-
9-yl and 2.9 kJmol-1 for thymin-1-yl. Only the overall effect of adenin-9-yl has a comparable
magnitude in the α- and β-series (0.9 kJmol-1 in the β-anomers versus slightly stronger -1.7 kJmol-1
in the α-counterparts). The weakening of the anomeric effect in α-nucleosides implies that the
configuration of the nucleobase at C1' presumably affects the ability of one of the lonepairs of the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
76
O4' oxygen to be engaged in molecular orbital interactions with the σ∗C1'-N1/N9
antibonding orbital.
This may also suggest that the anomeric effect is the result of predominant molecular orbital
interactions over electrostatic repulsions. One may possibly invoke a sp2 hybridization of the O4'
lonepairs orbitals, and an unfavourable orientation of the occupied lobe of the 1nsp2 (p-type) lonepair
with respect to the σ∗C1'-N1/9
orbital in the α-series only (Fig 12). We have shown that in β-
nucleosides the pyrimidine nucleobases drive the two-state N � S pseudorotational equilibrium
toward N-type conformations more efficiently than the purine counterparts (Section 4.4). In the α-
anomers, the ability of the nucleobase to stabilize S-type pseudorotamers is only slightly increased
when the base is changed in the order guanin-9-yl, thymin-1-yl, adenin-9-yl, cytosin-1-yl. The fact
that thymin-1-yl has an intermediate effect cannot be easily explained.
4.1.10 Weaker 3'-hydroxy gauche effect in α- compared with β-D-Ns
The [HO3'-C3'-C4'-O4'] and [MeO3'-C3'-C4'-O4'] gauche effects drive the pseudorotational
equilibrium of the pentofuranose moiety both in α- and β-nucleosides toward S-type conformations,
however their magnitude is much reduced (by 2.4 and 2.0 kJmol-1, respectively) in the α-anomers
compared with the β-counterparts (Table 5, cols 2 and 4). The fact that the nucleobase and the 3'-
substituents do not influence the drive of the pseudorotational equilibrium in α- and β-nucleosides
with the same efficiency is further evidenced by the results of the regression analysis (D), which has
been performed using a dataset of 43 experimental ΔH°N values, including abasic sugars 12 - 16, β-
D-ddNs 30 - 35, β-D-dNs (except 40), β-D-rNs, β-D-3'-dA (63) and β-L-dNs, α-D-ddNs 17 - 20, α-
D-dNs 21 - 26 and α-L-dNs 27 - 29, basing upon the assumption that the magnitudes of the effects
of all nucleobases as well as the gauche effects are independent of the configuration of the sugar
moiety at C1' (for the effects of the nucleobase, we have only assigned the opposite sign in the α-
series compared to the β-counterparts during the fitting procedure) (Table 5, col. 5). The multiple
Pearson's correlation coefficient is 0.974 [0.993 for regression (B)]. The standard error of the
estimates derived from regression (D) is twice higher (1.0 kJmol-1) than for those derived from
regression (B) (0.5 kJmol-1). Whereas the estimates of the anomeric and gauche effects determined
through the regression analyses (B) and (C) allow to reproduce the experimental ΔH°N values of β-
D- and α-D-nucleosides, respectively, fairly accurately, some theoretical ∆H° values calculated on
the basis of the estimates derived from regression (D) are much less accurate, and individual
residual errors between them and the corresponding experimental ΔH°N are -2.1 kJmol-1 for α-D-
ddA, 1.9 kJmol-1 for α-D-T, 1.7 kJmol-1 for α-L-T, 1.8 kJmol-1 for β-D-ddT, ±1.3 kJmol-1 for β-L-
dA, β-D-dA and β-D-ddC.
4.1.11 Limitations of regression analysis to quantitate stereoelectronic effects
For the regression analyses (A) - (E), the following approximations have been made:
(i) The magnitude of the effect of each nucleobase has been assumed to be identical in β-D-
ddNs, β-D-dNs and β-D-rNs (as well as in 63 for adenin-9-yl). This means that any potential
modulation of the electron-withdrawing character of the glycosyl nitrogen and electron-donating
ability of O4' by the presence of 2'- and/or 3'-OH(Me) substituents in β-D-dNs and β-D-rNs in
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
77
comparison with β-D-ddNs counterparts has been neglected. In the case of the glycosyl nitrogen,
this hypothesis is supported by the fact that the change of the pKa value of the nucleobase in β-D-
ddNs compared with the corresponding β-D-dNs and β-D-rNs is rather small (≤ 0.4 pD unit), as
discussed in Section 4.8.
(ii) It has also been assumed that 5'CH2OH and 5'CH2OMe exert the same (steric + gauche)
effect upon the conformation of the sugar moiety in abasic sugars 12 - 16 and in all nucleosides,
independently of their configuration at C1'. Any possible interaction (H-bonding or steric) between
the nucleobase and 5'CH2OH or 5'CH2OMe (section 4.8) in β-D-nucleosides has been neglected.
Figure 12: Possible origin for the
weaker O4'-C1'-N1/9 anomeric effect
in α-D-ddNs 17 - 20 (for the orbital
overlap in β-nucleosides, see Fig. 9,
sections 2.8 & 4). Our interpretation
is based on the fact that the anomeric
effect results from predominant nO4'
σ*C1'-N9 orbital mixing116 over
weaker electrostatic repulsions
(section 4.8). The O4' lonepairs
orbitals are represented using either
the sp2 [i.e. higher energy 1nsp2(p-
type) lonepair with predominant p-
type character and lower energy 2n
sp2(s-type) lonepair with
predominant s-type character] or sp3
[i.e. 1nsp3 and 2n
sp3 lonepairs with
the same energy] hybridization
models123-126,129,130 (section 1.9). (A) & (B) Relative orientations of sp2-hybridized O4' lonepairs with respect to the
C1'-N9 bond in N- (P = 0°, Ψm = 40°) and S-type (P = 160°, Ψm = 40°) sugars of α-D-ddA. β1/2 were calculated
assuming a perfect trigonal symetry. Neither in N- (β1 ≈ 42°) nor in S-type (β1 ≈ 6°) sugars is the 1nsp2 (p-type) orbital
antiperiplanar with the C1'-N9 bond, therefore the anomeric effect is at most weak (in favour of N-type forms!), and
clearly much weaker than in the β-counterparts (i.e. β1 ≈ 162°, Panel (C) in Fig. 9). This model is consistent with our
experimentally observed weaker anomeric effect in α-D-ddNs than in β-D-ddNs (Table 5). (C) & (D) Relative
orientations of sp3-hybridized O4' lonepairs with respect to C1'-N9 bond in the same N- and S-type sugars. The 2nsp3
lonepair and C1'-N9 bond assume an antiperiplanar orientation (β1 ≈ -144°) in the S-type form, but they are
perpendicular in N-type form. Therefore, if O4' lonepairs were sp3 hybridized, one would expect a stabilization of the S-
over N-type form to a similar extent as the stabilization of N-type sugars in β-nucleosides, which is not consistent with
our thermodynamic data for α- versus β-anomers.
(iii) The [HO3'-C3'-C4'-O4'] gauche effect has been attributed the same strength in β-D-dNs,
β-D-rNs and abasic sugars 13 and 14. We have also assumed that the [HO2'-C2'-C1'-O4'] gauche
effect has an equal strength in β-D-rNs, β-D-3'-dA (63) and abasic sugar 14, and that the [MeO3'-
OHOH2C O
HOH2C
N
N
N
N9
N
N
N
NH2
NH2
N9
OHOH2C O
HOH2C
N
N
N
N9
N
N
N
NH2
NH2
N9
C2'
H1'
N9
O4'
C4'
C2'
H1'
N9
O4'
C4'
C2'
H1'
N9
O4'
C4'
C2'
H1'
N9
O4'
C4'
(A)
1nsp2 (p-type)
σ*C1'-N9
(C)
σ*C1'-N9
(B)
β1 = 42ο
β1 = 6ο
(D) β2 = -108ο
β2 = -144ο
N-type sugar S-type sugar
1nsp2 (p-type)
2nsp2 (s-type) 2
nsp2 (s-type)
1nsp2 (p-type)
1nsp3
2nsp3
1nsp3
2nsp3
2nsp3
1nsp3
1nsp3
2nsp3
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
78
C3'-C4'-O4'] gauche effect exerts the same influence upon the pentofuranose conformation in abasic
sugars 15, 16 and adenosine derivatives 22, 23, 38 and 39.
The implications of these assumptions are two-fold: (a) We have not taken into
consideration any possible mutual compensation of different gauche effects in β-D-dNs, β-D-rNs
and abasic sugars 13 - 16 in comparison with β-D-ddNs and 12, respectively. (b) The magnitude of
the above gauche effects has been considered independent of the nature of the nature of the C1'
substituent (either a hydrogen in abasic sugars 12 - 16 or a nucleobase in nucleosides).
4.2 Quantitation of the anomeric and gauche effects by pairwise comparisons
Owing to the above simplifying approximations, the estimates obtained from our regression
analyses can only be used for semi-quantitative purposes. In order to refine these estimates for a
specific class of compounds (or even a single compound), we have also performed a set of pairwise
comparisons between ∆H° of compounds that differ only in one structural feature, in order to assess
the stereoelectronic contribution of that particular feature to the drive of the two-state N � S
pseudorotational equilibrium of the pentofuranose moiety in the compound(s) of interest. The
strategy adopted to perform these pairwise comparisons is depicted in Fig 13, and the resulting
estimates for the anomeric and gauche effects are compiled in Tables 6 - 8. In the following
sections, all discussions are based upon the results of our pairwise comparisons.
4.3 Strengths of ∆H° and -T∆S° to ∆G°of pseudorotation of neutralnucleosides
The rather limited experimental dataset of Altona et al.316,203 showed that the populations of
N- and S-type pseudorotamers of β-D-dA and of other β -D-dNs in the N state remain virtually
unchanged as the temperature increases from 291 K (75 % S) to 333 K (73 % S). They interpreted
this result by assuming that the stabilization of S-type pseudorotamers in. these compounds does not
have an enthalpic origin, but is instead the result of the larger entropy of the S-compared to the N-
type pseudorotamers
The comparison of ∆H° and -T∆S° contributions to ∆G° of the N � S equilibrium in neutral
β-D-dNs and β-L-dNs, derived from our pseudorotational analyses of 3JHH over a wider
temperature range (278 - 358 K) than that used by Altona et al., leads us to conclude that it is the
enthalpy factor, not the entropy, that is mainly responsible for the overall conformational preference
of sugar moieties in β-D-nucleosides.
This conclusion is supported by the following observations:
(i) For β-D/L-dA, 3'-OMe-β-D-dA, 3',5'-diOMe-β-D-dA, β-D-dImb, β-D/L-dG and β-D/L-
T, ∆H° overrides -T∆S°, resulting in the overall stabilization of S-type sugars at 298 K (i.e. negative
∆G° value).
(ii) For β-D/L-dC, β-D-dU and 5-F-β-D-dU, both ∆H° and -T∆S° contributions stabilize S-
type conformations and have nearly the same magnitude. Therefore, in the N state of all β-D/L-dNs,
-T∆S° is never found stronger than ∆H°.
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
79
O
OH
BP/N/D
O
OH
BP/N/D
HO
O
OH
BP/N/D
HO
OH
O
OH
BP/N
MeO
O
OMe
BP/
MeO
O
OH
MeO
BP/N
O
OH
HO
BP/N/D
O
OMe
MeO
BP/N
O
OH
BP/N/D
O
OH
HO
O
OH
HO
OH
O
OH
MeO
O
OMe
MeO
O
OH
BP/N/D
-1/-2HO3PO
OH
O
OH
BP/N/D
-1EtO
3PO
OH
O
OH
BP/N/D
-1/-2HO3PO
O
OH
BP/N/D
-1EtO
3PO
O
OMe
BP/N/D
O
OH
BP/N/D
OH
O
OH
C-B
P/N/D
HO
OH
O
OH
ΔΔ
Ho
17
ΔΔ
Ho
18
ΔΔ
Ho
13
ΔΔ
Ho
14
ΔΔ
Ho
2ΔΔ
Ho
3ΔΔ
Ho
4ΔΔ
Ho
5ΔΔ
Ho
6
ΔΔ
Ho
30
ΔΔ
Ho
15
ΔΔ
Ho
16
ΔΔ
Ho
9
ΔΔ
Ho
10
ΔΔ
Ho
24
ΔΔ
Ho
20
ΔΔ
Ho
23
ΔΔ
Ho
26
ΔΔ
Ho
28
ΔΔ
Ho
8
ΔΔ
Ho
11
ΔΔ
Ho
27
ΔΔ
Ho
29
ΔΔ
Ho
1d
ΔΔ
Ho
1a
ΔΔ
Ho
1b
ΔHo
1c
17 -
20
23
12
13
14
15
16
30, 31 &
33 -
35
50 -
55
38
63
56 -
62
69 -
73
74 -
78
79 -
83
37, 41 -
45
21 &
24 -
26
22
39
32
64 -
68
ΔΔ
Ho
12
ΔΔ
Ho
7
Fig
ure
13
. E
stim
ates
for
the
anom
eric
and
gau
che
effe
cts
in 1
2 -
83 f
rom
pai
rwis
e co
mpar
isons
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
80
Tab
le 6
. E
stim
ates
(∆
∆H°)
for
the
mag
nit
udes
of
the
effe
cts
of
the
nucl
eobas
es a
nd v
ario
us
gau
che
effe
cts
that
dri
ve
the
sugar
confo
rmat
ion i
n
com
pounds
12 -
16 a
nd i
n α
- an
d β
-N-n
ucl
eosi
des
17 -
63 f
rom
pai
rwis
e co
mpar
isonsa
Nucl
eobas
e A
G
Im
C
T
U
5-F
-U
Sta
te
P
N
P
N
D
P
N
P
N
N
D
N
D
N
D
∆∆
H°1
b
0.4
1
.0
- -
- -
- -
- -
- -
- -
-
∆∆
H°1
c -
- -0
.2
0.9
0
.9
- -
- -
- -
- -
- -
∆∆
H°1
d
0.5
0
.6
- -
- -
- -
- -
- -
- -
-
∆∆
H°2
b
8
.8
3.1
2
3.2
3
.0
1.0
-
- 9
.2
6.2
5
.0
3.1
5
.3
- -
-
∆∆
H°3
3
.4
0.2
6
.2
1.3
-0
.8
4.2
1
.9
4.1
3
.4
2.7
2
.2
3.5
2
.8
3.3
3
.0
∆∆
H°4
-0
.6
-4.8
5
.0
-3.7
-8
.0
- -
4.8
1
.9
0.9
-0
.6
1.6
-0
.1
1.9
0
.4
∆∆
H°5
3
.2
0.0
-
- -
- -
- -
- -
- -
- -
∆∆
H°6
3
.2
0.6
-
- -
- -
- -
- -
- -
- -
∆∆
H°1
0
-9.9
-7
.4
-21
.5
-6.2
-6
.3
- -
-9.6
-7
.3
-6.8
-5
.4
-6.3
-
- -
∆∆
H°1
1
0.5
-0
.5
3.3
-0
.5
-2.7
-
- 5
.2
3.0
2
.7
1.7
2
.6
1.6
3
.1
1.9
∆∆
H°1
2
-0.7
-0
.7
- -
- -
- -
- -
- -
- -
-
∆∆
H°1
3
-2.1
-2
.1
-9.1
-0
.8
-0.8
-
- -3
.3
-3.3
-1
.5
-0.9
-
- -
-
∆∆
H°1
4
-0.9
-0
.9
-6.2
-0
.4
0.7
-
- -3
.0
-3.0
0
.1
0.1
-
- -
-
∆∆
H°1
5
-1.2
-1
.8
- -
- -
- -
- -
- -
- -
-
∆∆
H°1
6
-1.1
-1
.6
- -
- -
- -
- -
- -
- -
-
∆∆
H°1
7
-3.3
-3
.3
-2.0
-4
.1
-3.0
-
- -4
.2
-4.2
-2
.9
-3.5
-
- -
-
∆∆
H°1
8
-0.8
-1
.4
- -
- -
- -
- -
- -
- -
-
∆∆
H°1
9
-7.1
-5
.7
- -
- -
- -
- -
- -
- -
-
a
A,
G,
Im,
C,
T,
U a
nd
5-F
-U r
epre
sent
aden
in-9
-yl,
guan
in-9
-yl,
im
idaz
ol-
1-y
l, c
yto
sin-1
-yl,
thym
in-1
-yl,
ura
cil-
1-y
l an
d 5
-flu
oro
-ura
cil-
1-y
l, r
esp
ecti
vel
y.
P,
N a
nd
D d
eno
te
the
pro
tonat
ed,
neu
tral
and
dep
roto
nat
ed s
tate
s o
f 1
7 -
63
, re
spec
tivel
y.
All
est
imat
es (
kJm
ol-1)
wer
e ca
lcula
ted
by s
ub
trac
ting
ΔH°
, Δ
H° N
or Δ
H° D
val
ues
(T
able
1)
of
two
com
po
und
s am
ong 1
2 -
63
, ac
cord
ing t
o t
he
stra
tegy d
epic
ted
in F
ig 1
3.
∆∆
H° 1
a (0
.4 k
Jmo
l-1),
∆∆
H° 7
(-4
.5 k
Jmo
l-1),
∆∆
H° 8
(4
.5 k
Jmo
l-1)
and
∆∆
H° 9
(-0
.5 k
Jmo
l-1)
are
const
ant
fact
ors
cal
cula
ted
by s
ub
trac
ting Δ
H° N
of
abas
ic s
ugar
s 1
2 -
16
. b
∆∆
H° 2
w
as a
lso
cal
cula
ted
by s
ub
trac
ting Δ
H° N
o
f 1
2 f
rom
ΔH° N
of β
-D-d
dI
(36
) (5
.8 k
Jmo
l-1).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
81
(iii) In the N state of β-D-rT, β-D-U and 5-F-β-D-rU, opposing ∆H° (driving to the N) and -
T∆S° nearly cancel each other, resulting in nearly unbiased equilibrium. However, in the N states of
β-D-G and β-D-A, ∆H° (stabilizing S-type sugars) prevails over the opposing -T∆S°, and S-type
sugars are favoured at 298 K. In the N state of β-D-C, ∆H° also prevails over counteracting -T∆S°,
but it stabilizes N-type sugars at 298 K.
Table 7. Quantitation of various gauche effects (∆∆H°) driving the sugar conformation in
mononucleotides 64 - 83 from pairwise comparisonsa
Nucleobase adenin-9-yl guanin-9-yl cystosin-1-yl thymin-1-yl uracil-1-yl
∆∆H°20 -8.9 -7.5 -9.8 -8.0 -8.4
∆∆H°21 -9.0 -8.2 -10.2 -8.4 -8.4
∆∆H°22 -1.5 -1.3 -2.5 -1.2 -2.1
∆∆H°23 -1.6 -2.0 -2.9 -1.6 -2.1
∆∆H°24 -0.1 -0.7 -0.4 -0.4 0.0
∆∆H°25 -0.5 -1.2 -1.5 -2.2 -0.8
∆∆H°26 -2.5 -2.5 -3.8 -3.8 -3.6
∆∆H°27 -1.4 -1.0 2.1 0.5 1.1
∆∆H°28 -2.0 -1.3 -2.3 -1.6 -2.8
∆∆H°29 0.5 -0.4 4.0 1.7 3.9
a All estimates (kJmol-1) have been calculated by subtracting values ΔH°N (Table 2 & 3) of two compounds among β-D-
ddNs 30 , 31 and 33 - 35, β-D-dNs 37 and 41 - 44, β-D-rNs 50 - 54 and mononucleotides 64 - 83 according to the
nomenclature shown in Fig 13.
Table 8. Quantitation of the enthalpy (∆∆H°30) of the (steric + stereoelectronic) effect of
the C-nucleobase, in each of its protonated (P), neutral (N) and deprotonated (D) states,
driving the sugar conformation in C-nucleosides 64 - 83a
Compound P N D
Formycin B (56) -0.9 -8.5 -9.2
Formycin A (57) -2.8 -8.5 -8.5
9-deaza-A (58) -7.8 -14.6 -
Ψ-isoC (59) 3.8 -2.3 -8.3
Ψ-U (60) - 0.3 -4.9
1-Me-Ψ-U (61) - 1.2 -3.4
1,3-diMe-Ψ-U (62) - 1.6 -
a ∆∆H°30
(kJmol-1) has been calculated by subtracting ΔH°N (0.4 kJmol-1, Table 2) of abasic sugar 14 from
those of a particular C-nucleoside among 56 - 62.
(iv) The situation is far less complicated in neutral β-D-ddNs 30 - 36, since for all of them,
∆H° (driving to N) always outweighs the counteracting -T∆S° contribution, which results in a high
preference for N-type pseudorotamers at room temperature. For abasic sugars 13, 15 and 16, the
two-state N � S pseudorotational equilibrium is driven toward S-type conformations mainly by
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
82
∆H°, whereas for 12 ∆H° and -T∆S° contributions cancel each other. In contrast, in 14, the
predominance of N-type conformations results from the cooperative drive of the prevailing -
T∆S° contribution and the minor ∆H° term.
4.4 Nucleobase-dependent anomeric effects in neutral nucleosides
4.4.1 Effect of the 5'CH2OH versus 5'CH2OMe
The strengthening of the overall effect of 5'CH2OMe moiety in abasic sugar 16 in
comparison with 5'CH2OH in 15 is evident from the subtraction of the ΔH°N value of 15 from that
of 16 (i.e. ∆∆H°1a = 0.4 kJmol-1, Fig 13 and Table 6). The comparison of the effect of 5'CH2OMe
in nucleosides 39 and 32 with the influence of 5'CH2OH in their counterparts, 38 and 31 supports
the results of regression analysis (A) regarding the slightly stronger ∆H°(5'CH2OMe) with respect to
∆H°(5'CH2OMe) since ∆∆H°1b (1.0 kJmol-1) and ∆∆H°1c (0.9 kJmol-1) in neutral nucleosides are
indeed slightly larger than ∆∆H°1a for abasic sugars.
4.4.2 The effect of the nucleobase is sugar-dependent
In β-D-ddNs 30, 31 and 33 - 36, an estimate for the overall effect of the constituent
nucleobases can be easily obtained by subtracting (∆∆H°2) the contribution of the 5'CH2OH effect
(i.e. reflected in the experimental ΔH°N value of 12) from experimental ΔH°N values of 30, 31 and
33 - 36 (Table 6). The comparison of ∆∆H°2 values shows that pyrimidines are able to shift the
pseudorotational equilibrium in β-D-ddNs toward N-type conformations more efficiently than
purines, as already suggested by the results of the regression analysis (A), since ∆∆H°2 increases in
the order: guanin-9-yl (3.0 kJmol-1) ≈ adenin-9-yl (3.1 kJmol-1) << thymin-1-yl (5.0 kJmol-1) ≈
uracil-1-yl (5.3 kJmol-1) < cytosine-1-yl (6.2 kJmol-1) .
In β-D-dNs 37 - 45, beside the effect of the nucleobase, both the influence of 5'CH2OH and
the gauche effect of [HO3'-C3'-C4'-O4'] fragment contribute to the drive of the sugar conformation.
Therefore, in order to obtain an estimate for the overall effect of the nucleobase in this series, it is
necessary to subtract (∆∆H°3 in Table 6) from their experimental ΔH°N values an estimate for the
combined effects of the 5'CH2OH and 3'-OH gauche effect. Such an estimate has been found in the
experimental ΔH°N value of abasic sugar 13 (-4.1 kJmol-1). The comparison of ∆∆H°3 values for
the nucleobases in β-D-dNs shows that, as in the β-D-ddNs series, pyrimidines exert a stronger
influence than for purines, i.e. ∆∆H°3 values are in the order: adenin-9-yl (0.2 kJmol-1) < guanin-9-
yl (1.3 kJmol-1) < imidazol-1-yl (1.9 kJmol-1) < thymin-1-yl (2.7 kJmol-1) < 5-fluorouracil-1-yl
(3.3 kJmol-1) ≈ cytosin-1-yl (3.4 kJmol-1) ≈ uracil-1-yl (3.5 kJmol-1). The fact that the effect of
imidazolyl-1-yl is slightly stronger than that of adenin-9-yl and guanin-9-yl is consistent with its
reduced steric bulk in comparison with adenin-9-yl and guanin-9-yl. The subtractions of ΔH°N of
abasic sugar 15 from that of 3'-OMe-β-D-dA (38) (i.e. ∆∆H°5 = 0.0 kJmol-1), on one hand, and of
ΔH°N of 16 from that of 3',5'-diOMe-β-D-dA (39), on the other (i.e. ∆∆H°6 = 0.6 kJmol-1) shows
that the effect of adenin-9-yl is the same in 38, 39 and β-D-dA (37) (0.2 kJmol-1, see above) itself.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
83
For each nucleobase, the comparison of ∆∆H°2 with ∆∆H°3 suggests that in β-D-dNs the effect of
the nucleobase is much weaker than in the β-D-ddNs counterparts. The weaker anomeric effect in
dNs compared to ddNs is explained as follows: (i) The electron-withdrawing character of 3'-OH
reduces the electron density around O4', making O4' lonepair less available for nO4'
→σ∗C1'N1/9
interactions in β-D-dNs compared with β-D-ddNs (Fig 14). If the nucleobase is
substituted for a hydrogen atom, no anomeric effect is operating, therefore the electron density
around O4' is maximal and as a result σC3'H3'
→σ∗
C4'O4' interactions are weaker (because of higher
difference between their energy levels) in abasic sugar 13 compared to β-D-dNs. Similarly as 3'-OH
in dNs is substituted for 3'-phosphate, the 3'-gauche effect is enhanced. This means that the strength
of the 3'-gauche effect increases as an electron-withdrawing group is placed at either C1' or C3'
(Section 4.5). (ii) Conversely, σC3'H3'
→σ∗
C4'O4' interactions in β-D-dNs may result in the lowering of
O
HOH2C
O
HOH2C
OH
OH
N
N
N
NH2
N9
NN
N
NH2
N9
H3'
H3'
σ*C4'O4'
σC3'H3'σ*C4'O4'
σC3'H3'
σ*C1'-N
N-type sugar S-type sugar
1nsp2 (p-type)
1nsp2 (p-type)
σ*C1'-N9
ΔE1 σC3'H3'
σ*C4'O4'
ΔE2
ΔE(GE)
ΔE(AE)
1nsp2 (p-type)
ΔE
___________
__
___________________________________________________________
___________ E1
E2
E3
E4
E5
E6___________________________________
The relative energies of orbitals (E1 - E6) are based on their relative donor-acceptor abilities, and they are constantly modulated by the nature of eachsubstituent at the sugar as well as on the aglycone in free, ionic and complexe state
Since nO4' is a better donor than σC3'H3' and σ*C4'O4' is a better acceptor than σ*C1'N9, therefore ΔE(GE) > ΔE(AE)
(A)
(B)
Figure 14. The interplay of the O4'-C1'-N9 anomeric effect and the [HO3'-C3'-C4'-O4'] gauche effect in β-D-dNs drives the N
� S equilibrium (the 2'-gauche effect in rNs is also a participant in this stereoelectronic interplay, but it is not shown here in
order to retain clarity in the diagram). (A) 1nsp2 →σ∗C1'N1/9 orbital overlap favours N-type sugars (i.e. anomeric effect),
whereas σC3'H3' →σ∗C4'O4' interactions (i.e. [HO3'-C3'-C4'-O4'] gauche effect) stabilize S-type conformations. The
differences in the electronic characters of the substituent at C1', C2' and C3' however decide the donor-aceptor capabilities of
the occupied and vaccant orbitals (i.e. the interplay of the gauche and anomeric effects) (B) The energy difference between
σC3'H3' and σ∗C4'O4' (∆E1) is smaller (hence a better mixing of the orbitals) than that between 1nsp2 and σ∗C1'N1/9 (∆E2),
therefore the gauche effect is more efficient than the anomeric effect This however is dictated on a case-to-case basis by the
nature of a set of substituents at the endocyclic carbons that are present on the sugar ring in a specific nucleoside. The reason
why we do not take 5'CH2OH into consideration is because it is viryually a free rotor where steric and gauche effect have been
found to have negligible contribution to the drive of the sugar conformation (vide supra).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
84
the energy level of the nO4' orbital. Since the efficiency of nO4'
→σ∗C1'N1/9
interactions is inversely
proportional to the difference between their energies, the anomeric effect is weakened in the dNs
compared to ddNs. This means that the modulations of the gauche effect by the anomeric effect, and
of the anomeric effect by the gauche effect are mutually interrelated and interdependent.
In β-D-rNs, the sugar conformation is driven by the interplay of the following
stereoelectronic effects: (a) the effect of the nucleobase, (b) the effect of 5'CH2OH, (c) the gauche
effect of [HO2'-C2'-C1'-O4'], (d) the gauche effect of [HO3'-C3'-C4'-O4'] and (e) the gauche effect
of [HO2'-C2'-C1'-N1/9]. The [HO3'-C3'-C2'-O2'H] gauche effect is negligible, since in N- and S-
type sugars, O3'-C3' is gauche with respect to O2'-C2'. In order to quantitate the effect of the
nucleobase in β-D-rNs, we subtracted from their ΔH°N values the contributions from the effects
detailed in (b) - (e). The combined influence of 5'CH2OH and of the [HO3'-C3'-C4'-O4'] and [HO2'-
C2'-C1'-O4'] gauche effects can be accounted for by subtracting (∆∆H°4) ΔH°N of 14 from that of a
β-D-rN. ∆∆H°4 values increase in the following order: adenin-9-yl < guanin-9-yl < thymin-1-yl <
uracil-1-yl ≈ 5-fluorouracil-1-yl ≈ cytosin-1-yl. In order to obtain an estimate for the overall effect
of the nucleobase in each β-D-rN, it was then necessary to subtract the strength of the [O2'-C2'-C1'-
N1/9] gauche effect from ∆∆H°4. Estimates for the strength of the [O2'-C2'-C1'-N1/9] gauche effect
(∆H°GE[O2'-C2'-C1'-N1/9]) in neutral β-D-rNs (Section 4.6) are as follows: -7.9 kJmol-1 in β-D-A, -
6.7 kJmol-1 in β-D-G, -4.3 kJmol-1 in β-D-C, -4.1 kJmol-1 in β-D-rT and -3.7 kJmol-1 in β-D-U.
The subtraction of ∆H°GE[O2'-C2'-C1'-N1/9] from ∆∆H°4 gives the effect of each nucleobase (AErN)
in β-D-rNs (kJmol-1): 3.0 in β-D-G ≈ 3.1 in β-D-A < 5.0 in β-D-rT ≈ 5.3 in β-D-rU < 6.2 in β-D-C.
The comparison of ∆∆H°2 and AErN values in the neutral state shows that the effect of a nucleobase
is the same in β-D-ddNs and β-D-rNs, owing to the cancellation of the [HO2'-C2'-C1'-O4'] and
[HO3'-C3'-C4'-O4'] gauche effects in the latter.
The fact that the overall effect of the nucleobase in β-D-nucleosides is characteristic of its
electronic and steric nature has three important implications in understanding structure-activity of
nucleic acids:
(i) The change of the electronic character of purine or pyrimidine heterocycles upon their
protonation or deprotonation is expected to result into the modulation of the anomeric effect, which
in turn should be reflected in the shift of bias of the two-state N � S equilibrium toward more N-
type sugars (upon protonation owing to the enhanced anomeric effect) or S-type pseudorotamers
(upon deprotonation due to the weaker anomeric effect). Qualitative studies on the preferred
conformation of the sugar moiety in various adenosine, guanosine and cytidine nucleosides and
nucleotides as a function of pD, as discussed in Section 2.9 are consistent with this simple
reasoning. We, on the other hand, have quantitated for the first time the actual magnitude of the
modulation of the anomeric effect in α- and β-nucleosides upon the change of the pD of the aqueous
solution (Sections 4-6).
(ii) Divalent metal cations are essential in natural processes involving nucleic acids, such as
the replication, transcription and translation of the genetic code479. In the presence of Mg2+
cofactor, the EcoRV restriction endonuclease is able to cleave DNA at one particular recognition
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
85
site with high specificity, whereas when Mg2+ is replaced by Mn2+, both the activity and specificity
of the enzyme are dramatically reduced. Numerous studies have been undertaken to establish the
sites of coordination of metal ions in nucleosides and nucleotides as well as the relative stabilities of
the complexes247,391,393,480-495. Whereas hard metal ions such as Mg2+ mainly bind to the phosphate
oxygen moiety of nucleotides, there is also spectroscopic evidence (NMR and UV) that many
transition metal ions such as Mn2+, Co2+, Ni2+, Cu2+ coordinate to both the phosphate oxygens and
the nucleobase nitrogens of adenosine and guanosine nucleotides either directly or via a coordinated
water molecule487. At the other end of the affinity scale, softer Pt in anticancer drug cis-
Pt(NH3)2Cl2 binds to DNA, by preferential coordination at N7 of the constituent adenin-9-yl and
guanin-9-yl nucleobases393. A slow titration with HgClO4 has shown that the binding of Hg2+ to N3
of thymin-1-yl in the AT-tract of DNA causes the disruption of the Watson-Crick base-pairing, and
is therefore responsible for a conformational transition from a B-type DNA to a new bulge-
containing conformer. As in the case of protonation (Section 4.8), one expects that this metallation
at N7 or N3 will enhance the strength of the anomeric effect of adenin-9-yl, guanin-9-yl or thymin-
1-yl which in turn should be reflected into an increased preference of the sugar moieties in the
respective adenosine, guanosine and thymidine residues for N-type pseudorotamers. The results of
recent studies on the conformational changes induced by the interaction of metal ions with the
nucleobase and the phosphate moieties in guanin-9-yl nucleotides are discussed in Section 8.7.
(iii) The overall effect of a N-nucleobase in a nucleoside upon the conformation of the
constituent sugar moiety consists of the counteracting contributions of (i) stereoelectronic
interactions within O4'-C1'-N1/9 fragment and (ii) the counteracting effect of the nucleobase.
Therefore, to obtain an estimate for the stereoelectronic anomeric effect it is necessary to subtract
from the overall effect of the nucleobase the steric component. Since it is possible to modulate the
relative importance of both contributions by changing the electronic character of the nucleobase, it
will be possible to engineer a nucleoside in which the nucleobase will steer the sugar conformation
through its inherent steric effect alone, and such a nucleoside can therefore be used as a reference
point further on (see our pD-dependent conformational studies on C-nucleosides in Section 6.1
which we used as the basis for the quantitation of the anomeric effect of adenin-9-yl in β-D-dA (37)
and β-D-A (50) and of guanin-9-yl in β-D-dG (41) and β-D-G (51)).
4.5 The gauche effect of the 3'-substituent in neutral dNs
4.5.1 The influence of the C1'-substituent
The subtraction of ΔH°N of abasic sugar 12 from that of 13 yields an estimate for the
strength of the [HO3'-C3'-C4'-O4'] gauche effect in 13 (∆∆H°7 = -4.5 kJmol-1 in Table 6). In order
to compare the magnitude of the [HO3'-C3'-C4'-O4'] gauche effect in 13 and in β-D-dNs, we have
also subtracted ΔH°N of each β-D-ddN 30, 31 and 33 - 35 from that of each β-D-dN counterpart 37
and 41 - 44 (∆∆H°10 in Table 6). ∆∆H°10 values are within the range from -7.4 kJmol-1 to -6.2
kJmol-1, showing that the [HO3'-C3'-C4'-O4'] gauche effect is much more efficient in β-D-dNs than
in abasic sugar 13. This suggests that as O4' is involved in nO4' →σ∗
C1'-N1/N9 stereoelectronic
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
86
interactions, the ability of O4'-C4' bond to adopt a gauche orientation with respect to O3'-C3' is
much greater than in 13. It is however difficult to establish a straightforward quantitative correlation
between the magnitude of the effect of the nucleobase and the corresponding strength of the [HO3'-
C3'-C4'-O4'] gauche effect in β-D-dNs 37 and 41 - 44. The small ∆∆H°9 value (-0.5 kJmol-1) on the
other hand, shows that the [MeO3'-C3'-C4'-O4'] gauche effect in 15 is only slightly more efficient
than the [HO3'-C3'-C4'-O4'] gauche effect in 13. The conclusion remains the same in the β-D-dNs
series, as shown by the small ∆∆H°12 value (-0.7 kJmol-1) reflecting the small additional stabilition
of S-type conformations in 3'-OMe-β-D-dA (38) in comparison with β-D-dA (37).
4.5.2 3'-substituent electronegativity dictates 3'-gauche effect
A comparative conformational analysis of 2'-deoxy-2'-substituted uridine235,199 and adenosine
206,417,418 derivatives has afforded linear relationships between the population of the N-type
pseudorotamer and the electronegativity of the 2'-substituent (Section 2.11). Various hypotheses
regarding the possible origin of the gauche effects in nucleos(t)ides have been discussed (See
Section 1.12-1.14 for the general discussion about the gauche effect in any system). We have
mentioned in above that substitution of 3'-OH in abasic sugar 15 and in β-D-dA (37) for 3'-OMe in
16 and 38 results in a slight increase in the preference of their sugar moieties for S-type
conformations (as experimentally evidenced by the slightly negative ∆∆H°9 and ∆∆H°12 values,
respectively, Table 6). This has been attributed to the modulation of the [MeO3'-C3'-C4'-O4']
gauche effect compared to [HO3'-C3'-C4'-O4'] gauche effect in the latter compared to the former.
The next logical step consisted in addressing the following questions: (i) Is it possible to
modulate the thermodynamics of the N � S equilibrium in a particular 3'-substituted-β-D-ddN (X =
3'-substituent) simply by altering the nature of its 3'-substituent and conversely (ii) is it possible to
predict quantitatively the strength of the [X3'-C3'-C4'-O4'] gauche effect driving the conformation
of any 3'-substituted-β-D-ddN using a simple procedure?
These questions have led us to perform a systematic comparative study26
on the preferred
conformations of the pentofuranose moieties in a series of 3'-substituted-β-D-ddT derivatives [X =
H (34), NH2 (113), OH (43), OMe (114), NO2 (115), OPO3H- (67), F (88)] with the same thymin-9-
yl nucleobase. These 3'-substituted-β-D-ddT derivatives only differ in the nature of 3'-X, which
varies from poorly (i.e. NH2 in 113) to strongly electronegative (i.e. F in 88). 3'-NH2 in 113 should
be half-protonated at ca. pD 7.0 since its pKa is expected to be very similar to that of 2'-NH2 in the
respective nucleoside (for instance, the pKa value of 2'-NH2 in 2'-NH2-β-D-dU has been reported to
be 6.2496,497).
The pseudorotational analyses of temperature-dependent 3JHH extracted from 1H-NMR
spectra of 88 and 113 - 115 and subsequent estimation of ΔH°N, ΔS°N and ΔG°N values of their N �
S equilibria have been performed using the same procedure as for β-D-ddT (34), β-D-T (43) and β-
D-TMP (67) (Section 3.1-3.7, Table 9 for ΔH°N, ΔS°N and ΔG°N values of 43, 67, 88 and 113 - 115.
From the initial comparison of ΔH°N, ΔS°N and ΔG°N values of 34, 43, 67 and 88, 113 - 115, the
following conclusions can be drawn: (i) ΔH°N is the main determinant to the drive of the sugar
conformation since it prevails over the counteracting (in 34, 43, 67 and 88) or cooperative (in all
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
87
other compounds) -ΔS°N contribution to ΔG°N. (ii) In β-D-ddT (34), the N � S equilibrium is
driven almost exclusively toward N-type conformations (16 % S at 278 K) by the cooperative
effects of the nucleobase and of 5'CH2OH, owing to the absence of any electron-withdrawing
substituent at C3'. (iii) In
contrast, the regularly
increasing stabilization of S-
type pseudorotamers in the
order 3'-NH2-β-D-T (113, 20
% at 278 K), β-D-T and β-D-
TMP (43 and 67, 64 % at 278
K), 3'-OMe-β-D-T (114, 74
% at 278 K), 3'-NO2-β-D-T
(115, 81 % at 278 K), 3'-F-β-
D-T (88, 91 % at 278 K), can
be attributed unambiguously
to the increasing preference
for gauche orientation within
[X3'-C3'-C4'-O4'] fragment,
which progressively prevails
over the anomeric effect.
ΔH°N reflects the strength of
the combined influence of the
(steric + stereoelectronic)
contributions of this gauche
effect, increases in the order:
NH2 (113) < OH (43) < OMe
(114) < NO2 (115) < OPO3H-
(67) < F (88).
The [X3'-C3'-C4'-O4'] gauche effect (ΔH°GE) in 43, 67, 88 and 113 - 115 has been
subsequently estimated by subtracting from their ΔH°N values that of β-D-ddT (34) (Table 9), which
allows to account for the combined effects of thymin-9-yl and of 5'CH2OH groups in 43, 67, 88 and
113 - 115. It should be however noted that ΔH°GE values consist of three intractable components: (i)
the stereoelectronic gauche effect, (ii) a term for the electrostatics and solvation, and (iii) the steric
effect of the 3'-substituent.
We have examined the dependence of ΔH°GE on the group electronegativity498-500,501 of 3'-X
using the electronegativity scales developed by Marriott (χMarriott)502, Mullay (χMullay)503,504 and
Inamoto (ιInamoto)505-508 and their coworkers. Marriott's electronegativity scale is based on ab initio
calculations with the GAUSSIAN program298 (at HF/6-31G* level). The actual electronegativity of
X in H-X corresponds to the atomic electron population on H in H-X, as determined from a
Mulliken populations analysis of geometrically optimized H-X. Mullay's group electronegativities
(I)
ΔHo
GE (kJ mol
-1)
-12 -10 -8 -6 -4 -2
Group electronegativity of 3'-substituent χ or τ
0
1
2
3
4
5
(III)
(II)
88
116
114
43
67
115
113
88
116
114
43
67
115
113116
88 114
43
67
115
113
Figure 15: Correlation plots of the group electronegativity of the 3'-X substituent
in 3'-substituted-β-D-ddT derivatives [X = H (34), NH2 (113), OH (43), OMe
(114), NO2 (115), OPO3H- (67), F (88)] as a function of the strength (ΔH°GE) of
the [X3'-C3'-C4'-O4'] gauche effect. The group electronegativity has been
expressed in the scales developed by Mullay (χMullay)503,504 [....., graph (I)],
Inamoto (ιInamoto)505-508 [___, graph (II)] and Marriott (χMarriott)502 [----,
graph (III)]. All plots show straightlines with high Pearson's correlation
coefficients [R = -1.0, -0.98 and -0.96 for graphs (I) - (III), respectively]. In a
recent study42, we have determined the thermodynamics of the N �S
equilibrium [∆H° = 0.0 kJmol-1 (σ = 0.4) , ∆S° = 7.5 Jmol-1K-1 (σ = 1.5) and
∆G° (298 K) = -2.3 kJmol-1 (σ = 0.6)] in 3'-OCF3-β-D-ddT (116) using the
procedure described in Section 3.1-3.7]. ΔH°GE (-5.4 kJmol-1) of the [OCF3-C3'-
C4'-O4'] gauche effect in 116 was subsequently calculated by subtracting ∆H° of
β-D-ddT (34) from its own. From the equations of the straightlines (I) [χMullay
= 2.63 - 0.19ΔH°GE], (II) [ιInamoto = 2.25 -0.08 ΔH°GE] and (III) [χMarriott =
0.26 - 0.022 ΔH°GE], the group electronegativity of 3'-OCF3 in 116 has been
determined: χMullay = 3.66 ± 0.05, ιInamoto = 2.64 ± 0.10 and χMarriott = 0.38
± 0.05.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
88
are based on modified Slater effective nuclear charges, effective principal quantum numbers,
fractional p characters, assuming charge conservation and electronegativity equalization within each
bond in the group. Inamoto's scale ι is an inductive substituent parameter scale, which is derived
from the corresponding group electronegativities showing a strong correlation with trans H,H
coupling constant in monosubstituted ethene fragments. The plots of the ΔH°GE values as a function
of χMarriott502, χMullay
503,504 and ιInamoto505-508 gave all straight lines with high correlation
coefficient (> 0.96) (Fig 15). From the equations of these simple correlation plots, we have
subsequently been able to estimate the group electronegativity of 3'-OPO3H- in β-D-TMP (67), i.e. :
χMarriott = 0.44, χMullay = 4.12 and ιInamoto = 2.8.
Table 9. The enthalpy and entropy contributionsa to the N � S pseudorotational equilibrium of the
pentofuranose moiety in 3'-substituted-β-D-ddT derivatives 43, 67, 88 and 113 – 115 (see ref 26)
Compound Group electronegativity (χ or ι) of
3'-substituent d
ΔH°N ΔS°N −ΤΔS°N ΔG°N Δ%S b
358-278 ΔH°GE
c
Marriott Mullay Inamoto
3'-NH2-ddT (113) 2.6 -1.9 0.6 3.2 +5 -2.8 0.33 3.15 2.47
T (43) -1.8 -0.9 0.3 -1.5 -4 -7.2 0.43 3.97 2.79
3'-Ome-ddT(114) -2.1 1.1 -0.3 -2.4 -4 -7.5 0.44 4.03 2.82
3'-NO2-ddT (115) -2.4 3.7 -1.1 -3.5 -3 -7.8 0.4 4.08 2.75
TMP (67) -2.6 -4.3 1.3 -1.3 -3 -8.0 0.44
e
4.12e 2.8
e
3'-F-ddT (88) -5.9 -2.3 0.7 -5.2 -6 -11.3 0.52 4.73 3.1
a In kJmol-1. ΔS°N and ΔG°N are given at 298K (see the methodology described in Section 3). b Δ %S (358-278)
indicates the change in the population of the S-type pseudorotamer as the temperature is raised from 278 K to 358 K,
and shows the influence of both the gauche effect enthalpy and entropy contributions. At a particular temperature T, the
population of the S-type species is calculated from the corresponding free energy values as follows: %S (T) = 100 * [exp
(- ΔGT/ RT)] / [exp (- ΔGT/ RT) +1]. c ΔH°GE has been calculated by subtracting ΔH°N characterizing the N �S
equilibrium of the sugar moiety in 43, 67, 88 and 113 - 115 from that obtained for β-D-ddT (34) (Table 2). d
The
group electronegativities of the 3'-substituents NH2 (113), OH (43), MeO (114), NO2 (115) and F (88) are given
according to ref. 502-508. e For β-D-TMP (67), the group electronegativities of the OPO3H- substituent have
been back-calculated from its ΔH°GE value using the graphs shown in Fig 15.
S-type and the N-type pseudorotamer populations. Our NMR analysis of a set of 3′-substituted
thymidine derivatives [Scheme 1: for 1 – 7]26
, in which the gauche effect of 5′-CH2OH and the
anomeric effect of the nucleobase remain as the constant factor, has clearly shown that the
conformational preference for the S-type sugar conformation linearly increases with the increase of
the strength of the electronegetivity of the 3′-substitutent. It has been envisioned that as the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
89
electronegetivity of the 3′-substituent increases [H < NH2 < OCF3 < OH < OCH3 < NO2 < F] the
C3′-H3′ becomes further polarised, the difference in the energy levels of the donor σC3′-H3′
and
acceptor σ*
C4′-O4′
orbitals decreases, facilitating the σC3′-H3′
�σ*
C4′-O4′
orbital mixing, thereby
enhancing the 3′-gauche effect promoted S-type conformational preference.
We argued that since we already have a dependable experimental means for the estimation
of K298
NMRG° Δ
ref, we could challenge the theory with this by calculating the total energy difference
(ΔE) of N-type and S-type geometries (ΔES-N) of 1 – 7 (Figure 1) using high level ab initio
calculations. The Gaussian calculations of both N- and S-type pseudorotamers of potential anti-HIV
2′,3′-dideoxy 3′-substituted thymidine derivatives 1 – 7 have been performed using systemetic
variation of different basis functions at Hatree-Fock level. With the lower basis functions such as 6-
31G* or even with 6-311+G* basis sets (Table 1), we failed to observe any colinearity of the ab
initio calculated ΔES-N in the gas phase with our K298
NMRG° Δ . This pilot study showed the necessity of
introducing diffuse function with higher basis set in order to reduce the basis set superposition
error6, which also illustrated the caveats of using a low-level ab initio method for structural
calculation of flexible biomolecules.
The use of higher basis set, such as HF/6-311++G**, showed a marked improvement of
correlation of ΔES-N with the experimental K298
NMRG° Δ for 1 – 7 (Table 1). Thus a plot of ΔES-N (gas
phase HF/6-311++G**) as a function of K298
NMRG° Δ showed a Pearson’s correlation coeffecient of 0.92
(Graph I in Figure 1A). In order to mimick the solvation behaviour of the experimental NMR
analyses, the solution phase (ε = 78.0) ab initio calculations was performed at HF/6-311++G** level
using Onsager solvation model. Remarkably, this gave even better correlation compared to the gas
phase calculations as evidenced from the Graph II in Figure 1A, giving Pearson’s correlation
coeffecient of 0.97, as a result of improved colinearity of the energies from the solution phase ab
initio calculations with the experimental NMR data. Moreover, the single point ab initio
calculations at B3LYP/6-311++G** level of theory in solution phase with Onsager solvation model
have been performed using the optimised geometry at HF/6-311++G** level for 1 – 7 (Table 1).
The plot of ΔES-N calculated from this solution phase B3LYP/6-311++G**//HF/6-311++G**
calculation as a function of experimental K298
NMRG° Δ gives the Pearson’s correlation coeffecient as
0.97 (Figure 1B).
This straightforward correlation of our experimental NMR findings with the present
theoretical ab initio calculations justifiably points to the following observations: (i) it is now
possible to predict the K298
NMRG° Δ of nucleosides quite dependably by simply knowing the ab initio
calculated ΔES-N; (ii) the substituent-dependent steric and stereoelectronic effects on the bias of the
two-state1,2b-e
N � S equilibrium in nucleosides can also be easily predicted from the ab initio
calculations provided a sufficiantly large basis set is used, and (iii) the comparison of Pearson’s
correlation coeffecients in Figure 1 indicates the necessity of mimicking the solvation behaviour of
the experimental NMR measurement condition in these high level ab initio calculations. It is quite
likely that this correlation between the theory and the experiment would give much deeper insight
into the molecular orbital basis of the role of the stereoelectronic forces in modulating the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
90
conformation of nucleosides and nucleotides as well as shed light in their ubiquitous self-assembly
process governing the chemistry of life in general.
4.5.3 Stronger gauche effect in nucleosides than in 1,2-difluoroethane
In 1,2-difluoroethane, an energy stabilization of 2.4 -3.4 kJ/mol has been observed in favour
of the gauche rotamer over the trans counterpart (section 1.13) but in 3'-substituted-2',3'-
dideoxynucleosides we have found that the [X3'-C3'-C4'-O4'] gauche effect ranges from -2.8 kJ/mol
for X = NH2 to -11.3 kJ/mol for X = F (Fig. 15, Table 9). This can be understood in terms of
dihedral constraints imposed on the endocyclic torsions owing to the ring closure nature of the five-
membered pentofuranose nucleosides in contrast to 1,2-difluoroethane where only constraint is
imposed by favourable orbital overlap or bond-bending in the gauche compared to the trans rotamer.
4.6 The 2'-OH effect in ribonucleos(t)ides is nucleobase-dependent
The subtraction of ΔH°N of a particular β-D-dN 37 or 41 - 44 from that of its β-D-rN
counterpart (50 - 54) yields an estimate (∆∆H°11, Table 6) for the combined strength of the [O2'-
C2'-C1'-O4'] and [O2'-C2'-C1'-N1/9] gauche effects, i.e. the overall effect of 2'-OH in β-D-rNs in
comparison with β-D-dNs. ∆∆H°11 values are as follows (in kJmol-1): adenin-9-yl (-0.5) = guanin-
9-yl (0.5) < uracil-1-yl (2.6) ≈ thymin-1-yl (2.7) ≈ cytosin-1-yl (3.0) ≈ 5-fluorouracil-1-yl (3.1).
Thus, the ability of 2'-OH to drive the sugar conformation toward S-type pseudorotamers is greater
in the purine β-D-rNs with respect to the pyrimidine β-D-rNs.
(34)
ΔGNMR
-6 -4 -2 0 2 4
ΔE(S -N)
-16
-12
-8
-4
0
4
(I)
(II)
(34)
ΔGNMR
-6 -4 -2 0 2 4
ΔE(S -N)
-16
-12
-8
-4
0
4(113)
(43)
(114)
(115)
(116)
(88)
(34)
(113)
(113)
(88)
(88)
(43)
(43)
(114)
(114)
(115)
(115)
(116)
(116)
(A) B)
Figure 16. Panel (A) shows the plot of the free energy of the N � S pseudorotational equilibrium (K298
NMRG° Δ , in kJ
mol-1
) as a function of ab initio calculated energy diffrenence between the S- and the N-type pseudorotamers (ΔES-N, in
kJ mol-1
) at HF/6-311++G** level for 2′,3′-dideoxy 3′-substituted (X) thymidine derivatives [with X = H (34), NH2
(113), OH (43), OCH3 (114), NO2 (115), OCF3 (116), F (88)], both in the gas phase [■, dotted line, graph I] as well as
in the solution [�, solid line, graph II] gives straight lines with slope = 1.03 (σ = 0.11), intercept = -1.22 (σ = 0.31) and
R = 0.92 for I and slope = 1.50 (σ = 0.09), intercept = -4.92 (σ = 0.30) and R = 0.97 for II respectively. Panel (B) shows
the similar plot ofK298
NMRG° Δ (in kJ mol
-1) as a function ΔES-N (kJ mol
-1) derived from the single point solution phase
DFT calculations at B3LYP/6-311++G**//HF/6-311++G** level, giving the straight line with slope = 1.41 (σ = 0.11),
intercept = -4.17 (σ = 0.35) and R = 0.97. The fact that ∆ES-N for 116 is clearly smaller than that of 88 supports our
experimental result of the reduced efficiency of the [3'-OCF3-C3'-C4'-O4'] gauche effect in the former in comparison
with the [F3'-C3'-C4'-O4'] gauche effect in the latter.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
91
2'-OH stabilizes N-type pseudorotamers in pyrimidine β-D-rNMPs 76 - 78 (by 1.7 - 4.0
kJmol-1, i.e. ∆∆H°29 in Table 7) and in β-D-rNMPEts 81 - 83 (by 0.5 - 2.1 kJmol-1, i.e. ∆∆H°27 in
Table 7) in comparison with the 2'-deoxycounterparts β-D-dNMPs 66 - 68 and β-D-dNMPEts 71 -
73. In contrast, there is virtually no enthalpy stabilization of the S-type pseudorotamers in purine β-
D-rNMPs 74 and 75 and in β-D-dNMPs 64 and 65, as shown by small ∆∆H°29 values (Table 7).
Finally, in β-D-AMPEt (79) and β-D-GMPEt (80), 2'-OH even destabilizes N-type conformations in
comparison with β-D-dAMPEt (69) and β-D-dGMPEt (70) (by ∆∆H°27 = -1.4 and -1.0 kJmol-1,
respectively). Thus, as in the ribonucleosides, the ability of 2'-OH to stabilize N-type
pseudorotamers is much greater in the pyrimidine than in the purines series.
The subtraction (∆∆H°8 = 4.5 kJmol-1, Table 6) of ΔH°N of abasic sugar 13 from that of 14
yields an estimate for the combined influence of the [HO2'-C2'-C1'-O4'] and [HO2'-C2'-C3'-O3']
gauche effects, the later being negligible (because of gauche orientation in both N and S
conformers). ∆∆H°8 has the same value, but the opposite sign, as ∆∆H°7, showing that the [HO2'-
C2'-C1'-O4'] and [HO4'-C4'-C3'-O3'] gauche effects are cancelling each other, which is consistent
with the results of regression (A) and also with the fact that the enhancement or reduction of
electron-density of O4' owing to variable electronic character of the nucleobase will affect both
[HO2'-C2'-C1'-O4'] and [HO3'-C3'-C4'-O4'] in a similar manner.
Figure 17. The correlation
plot of the overall effect
(AEddN) of adenin-9-yl in β-
D-ddA (30), guanin-9-yl in β-
D-ddG (31), cytosin-1-yl in β-
D-ddC (33), thymin-1-yl in β-
D-ddT (34) and uracil-1-yl in
β-D-ddU (35) as a function of
the overall effect of 2'-OH in
β-D-rNs 50 - 54, β-D-rNMPs
74 - 78 and β-D-rNMPEts 79 -
83 (in the N state). The effect
of the C1'-aglycone in β-D-ddNs was estimated from ∆∆H°2 (Table 6 and Fig 13). The overall effect of 2'-OH was
estimated from ∆∆H°11 for β-D-rNs, ∆∆H°29 for β-D-rNMPs and ∆∆H°27 for β-D-rNMPEts (Table 7 and Fig 13). The
Pearson's corelation coefficient of the straight line is 0.87.
Therefore, if one assumes that the strengths of the [HO3'-C3'-C4'-O4'] and [HO2'-C2'-C1'-
O4'] gauche effects are identical but with the opposite signs (as suggested by the identical ΔH°N
values of 12 and 14), then the nucleobase-dependent -∆∆H°10 values give estimates for the
magnitudes of the [HO2'-C2'-C1'-O4'] gauche effect in β-D-rNs (Table 6). Subtraction of -∆∆H°10
from ∆∆H°11 affords the nucleobase-dependent strength of the [O2'-C2'-C1'-N1/9] gauche effect in
β-D-rNs 50 - 54, i.e.: -7.9 kJmol-1 in β-D-A, -6.7 kJmol-1 in β-D-G, -4.3 kJmol-1 in β-D-C, -4.1
kJmol-1 in β-D-rT and -3.7 kJmol-1 in β-D-U, showing that in purine nucleosides this gauche effect
is much more efficient than in the pyrimidine counterparts.
The interdependency of the overall effect of 2'-OH in β-ribonucleosides and ribonucleotides
is also evidenced by the fact that a plot of the effect of the nucleobase in β-D-ddNs 30 - 35 as a
-4 -2 0 2 4 6
AE ddN (kJ mol-
1
)
2
4
6
8
79 75 7480
50 & 51
82
83
53
54
76
77
52
78
81
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
92
function of ∆∆H°11 (for β-D-rNs 50 - 54), ∆∆H°27 (for β-D-rNMPEts 79 - 83) and ∆∆H°29 (for β-
D-rNMPs 74 - 78) in Fig. 17 gives a straigthline with correlation coefficient of 0.87. As 2'-OH
drives the pseudorotational equilibrium to the N, the strength of the anomeric effect of the
nucleobase increases in a cooperative manner.
4.7 3'-gauche effect modulation by 2'-OH in ribonucleos(t)ide
In ribonucleos(t)ides, 2'-OH influences the drive of the sugar conformation through the [O2'-
C2'-C1'-O4'] and [O2'-C2'-C1'-N1/9] gauche effects. Unlike in β-D-dNs, in β-D-rNs a possible
contribution of H-bonding142 interaction between 2'- and 3'-substituents to the preference for gauche
orientation within [HO3'-C3'-C4'-O4'] fragment cannot be excluded. It has been for instance shown
by 1H-NMR509,510 that there is an intramolecular water bridge between the vicinal 2'-OH and 3'-
phosphate in cAMP in which the 2'-OH hydrogen accepts the lonepair of water oxygen.
We have attempted to delineate27 the magnitudes of the gauche effects involving 2'-OH in β-
D-A (50) in the N state by comparing the conformational preferences of its constituent
pentofuranose sugar with those in β-D-ddA (30), β-D-dA (37), β-D-3'-dA (63), β-D-dAMP (64), β-
D-AMP (74) and abasic sugars 12 - 14 through a set of pairwise subtractions of the ΔH°N values of
their N � S equilibria (Table 6). The main results of this work are as follows:
(i) As discussed in Section 4.6, in abasic sugars 13 and 14, the [HO3'-C3'-C4'-O4'] and
[HO2'-C2'-C1'-O4'] gauche effects cancel each other, their magnitudes being respectively ∆∆H°7 = -
4.5 and ∆∆H°8 = 4.5 kJmol-1.
(ii) The comparison of ∆∆H°10 (-7.4 kJmol-1) and ∆∆H°19 (-5.7 kJmol-1) shows that the
strength of the [HO3'-C3'-C4'-O4'] gauche effect is reduced by 1.7 kJmol-1 in β-D-A in comparison
with β-D-dA, which can be attributed to the effect of 2'-OH in the former.
(iii) The strengths of the [O4'-C4'-C3'-O3'PO3H-1/-2] gauche effect in β-D-dAMP and β-D-
AMP have been estimated by ∆∆H°20 (-8.9 kJmol-1) for the former and by subtracting (-6.2 kJmol-
1) ΔH°N value of β-D-3'-dA from that of β-D-AMP, respectively. Thus the [O4'-C4'-C3'-O3'PO3H-]
gauche effect is less efficient by 2.7 kJmol-1 in β-D-AMP than in β-D-dAMP, showing the effect of
2'-OH in the former.
The fact that 2'-OH weakens the [O3'-C3'-C4'-O4'] gauche effect less efficiently in β-D-A
than in β-D-AMP [compare (i) and (ii)] can be understood as a consequence of the greater freedom
of the lonepair of O2' to participate in the [O2'-C2'-C1'-N9] and [O2'-C2'-C1'-O4'] gauche effects in
β-D-AMP than in β-D-A, where it acts as a donor in the intramolecular hydrogen-bonding
interaction between 2'-OH and 3'-OH511,512. Hydrogen bonding of 2'-OH with N3513 may induce the
strengthening of the [O2'-C2'-C1'-N9] gauche effect in purine β-D-rNs.
In A-type RNA, pseudorotamers are inherently stabilized when 2'-OH is involved in
hydrogen bonding interaction with the nucleobase, and the substitution of 2'-OH in hammerhead
ribozyme for NH2 or F changes the ratio of N- versus S-type pseudorotamers as well as influences
the hydrogen bonding capabilities514-516 of the sugar.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
93
(iv) The overall stereoelectronic effect of 2'-OH has first been estimated in β-D-3'-dA by
subtracting (-2.2 kJmol-1) ΔH°N of β-D-ddA from its own. For β-D-A and β-D-AMP, the strength
of the 2'-OH effect is reflected in the values of ∆∆H°11 (-0.5 kJmol-1) and ∆∆H°29 (0.5 kJmol-1),
respectively. Thus the ability of the effect of 2'-OH to counteract the [O3'-C3'-C4'-O4'] gauche
effect by stabilizing N-type conformations increases in the order: β-D-3'-dA < β-D-A < β-D-AMP.
The 2'-OH stabilization of N-type conformations in β-D-AMP compared with β-D-A and β-D-3'-dA
can be attributed either to a more efficient [O2'-C2'-C1'-O4'] gauche effect or to the weakening of
the [O2'-C2'-C1'-N9] gauche effect.
4.8 Drive of pseudorotation in β-nucleosides by the nature of the nucleobase
4.8.1 The two-state N � S equilibrium is evidenced by pKa values from ∆G°
For all β-D-nucleosides, either in the 2',3'-dideoxy (30 - 34), 2'-deoxy (37 and 40 - 45) or
ribo (50 - 55) series, the plots of ∆H°, -T∆S° and ∆G° values as a function of pD show that the bias
of their two-state N � S equilibrium can be successfully modulated by the protonation and/or
deprotonation of the nucleobase (Fig. 11 and Table 2). The origin of this modulation will be
discussed for each series in the next paragraph. (In β-L-nucleosides, similar trends are observed
since the thermodynamics of the N � S equilibrium in each of the D- and L-enantiomers are
virtually the same owing to the mirror-image relationship, as discussed in Section 4.1(e)). The value
of the pD at the inflection point of the plot of pD-dependent ∆G° values for each compound has
been determined by fitting the experimental data to the Henderson-Hasselbach equation, as
discussed in Section 3.7. These pKa(s) values are virtually identical to the known pKa(s) of the
constituent nucleobases in 30 - 34, 37, 40 - 45 and 50 - 55, as found in the literature165,247,379,382-384
and to other estimates of pKa values independently derived by us from Hill plots of pD-dependent
1H chemical shifts of the anomeric and aromatic protons of the same nucleosides (± 0.4 pD unit,
compare cols. 5 and 6, on one hand, cols. 10 and 11, on the other, in Table 2). This result
unequivocally validates the two-state N �S equilibrium model.
4.8.2 Identical pKas of the nucleobases in 2',3'-dideoxy, 2'-deoxy and ribo series
The pKa values of cytosin-1-yl and thymin-1-yl (determined from pD-dependent 1H
chemical shifts) in β-D-ddC/T, β-D-dC/T and β-D-C/rT are the same (± 0.1 from the average
values), showing the lack of influence of 2'- and/or 3'-OH on their electronic character (See Section
8.8 for the discussion on 3',5'-bisphosphate of guanosine). For adenin-9-yl and guanin-9-yl, the
electron-withdrawing effect of 2'-/3'-OH is only partly reflected in the slight decrease of the pKa
value in the order ddN > dN > rN (i.e. by 0.3 and 0.4 pD unit, respectively). This tendency should
however be considered with caution, since the accuracy of the estimates of the pKa values
determined from pD-dependent chemical shifts is ≈ 0.1 pD unit.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
94
Table 10. The relative populations of syn versus anti rotamers around the glycosidic torsion in
α-D-ddNs 17 - 20, β-D-ddNs 30 - 34, α-D-dNs 21 - 26, and β-D-dNs 37 - 39 and 41 - 43 at various
pDs a from 1D-nOe difference experimentsb.
Compound
pD
ηH1' (%)
(H6/H8 irradiated)
ηH6/H8 (%)
(H1' irradiated)
(ηH1' + ηH6/H8) / 2 % syn
rotamers
α-D-ddA (17) 7.3 2.3 1.9 2.1 20
β-D-ddA (30) 7.0 2.3 2.4 2.4 22
α-D-ddG (18) 7.5 1.8 1.8 1.8 17
β-D-ddG (31) 2.1 1.4 1.5 1.5 14
7.4 1.9 2.4 2.2 21
11.5 2.8 2.8 2.8 26
5'-OMe-β-D-ddG (32) 2.1 1.6 - 1.6 15
6.6 1.7 1.4 1.6 21
11.7 1.3 1.3 1.3 12
α-D-ddC (19) 7.1 2.4 - 2.4 22
β-D-ddC (33) 7.0 1.5 - 1.5 14
α-D-ddT (20) 7.2 3.7 3.2 3.5 33
β-D-ddT (34) 7.0 2.1 2.0 2.1 20
α-D-dA (21) 2.1 2.5 2.8 2.7 25
6.5 5.4 4.0 4.7 44
3'-OMe-α-D-dA (22) 1.6 0.6 0.7 0.7 7
6.9 1.2 0.9 1.1 10
3',5'-diOMe-α-D-dA (23) 1.6 1.0 0.9 1.0 10
6.7 0.9 1.1 1.0 10
β-D-dA (37) 1.6 3.5 4.2 3.9 36
6.9 5.8 5.9 5.9 55
3'-OMe-β-D-dA (38) 1.6 4.1 4.2 4.2 39
7.0 5.5 6.5 6.0 56
3',5'-diOMe-β-D-dA (39) 2.2 2.4 2.4 2.4 22
7.3 1.4 1.2 1.3 12
α-D-dG (24) 1.8 3.8 2.3 3.1 29
7.3 4.7 2.3 3.5 33
11.6 6.1 7.7 6.9 65
β-D-dG (41) 1.8 3.5 3.1 3.3 31
7.7 3.8 4.0 3.9 36
11.1 4.5 4.6 4.6 43
α-D-dC (25) c 2.0 0.9 - 0.9 8
7.1 2.4 - 2.4 22
β-D-dC (42) d 7.3 2.8 2.8 2.8 26
α-D-T (26) 7.1 3.0 - 3.0 28
11.6 3.4 - 3.4 32
β-D-T (43) 6.7 3.7 3.0 3.4 32
11.7 3.1 2.4 2.8 26
a 1D-nOe difference experiments were performed in D2O [298 K except for all guanin-9-yl nucleosides (288 K) due to
their decomposition above 288 K at acidic pDs]. b We used the method of Rosemeyer to calculate the population of
syn rotamers around the glycosidic torsion from homonuclear 1H nOes517. c Selective saturation of H1' could not be
performed at acidic pD for α-D-dC and at neutral pD for α-D-ddC and β-D-ddC due to near isochronous H1' and H5 . d Could not be measured in the acidic solution due to the fact that H1' and H5 in β-D-dC have the same chemical
shift at pD < 3 at 298 K.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
95
4.8.3 Anomeric effect in β-D-ddNs is modulated by the nature of the nucleobase
A comparative analysis of opHΔ , -T o
pSΔ and opGΔ values of 30 - 34 and of the corresponding
oN
HΔ , -T oNSΔ , o
NGΔ and o
DHΔ , -T o
DSΔ and oD
GΔ values (Table 2) shows that opHΔ and o
DHΔ prevail over
the counteracting -T opSΔ and -T o
DSΔ which results into the overall stabilization of N-type sugars in
the P and D states (i.e. positive opGΔ and o
DGΔ ). Protonation of the nucleobase in 30 - 34 steadily
shifts the N � S pseudorotational equilibrium toward more N-type conformation than in the N state.
When the nucleobase is fully protonated, a plateau in thermodynamic values in the P state is reached
[i.e. oNP
G−
ΔΔ (kJmol-1) 1.5 (30), 3.8 (31), 3.2 (32), 1.1 (33)]. In contrast, deprotonation shifts the N
� S equilibrium toward more S-type sugars (with pseudoequatorially oriented nucleobase) than in
the N state [i.e. oND
G−
ΔΔ (kJmol-1) = -1.1 (31), -0.3 (32) and -0.8 (34)]. We have compared the
preferred conformation of the sugar moiety and of the nucleobase in β-D-ddG (31) and 5'-OMe-β-
D-ddG (32) over the whole ≈ 2.0 - ≈ 12.0 pD range in order to examine whether 5'-OH and guanin-
9-yl, by forming an H-bond or guanin-9-yl alone, by adopting different orientations around the
glycosyl torsion at different pDs, contribute to the above oNP
G−
ΔΔ and oND
G−
ΔΔ values. ∆H° of 31 and
32 are nearly identical at any pD suggesting that 5'CH2OH and N3 in 31 do not form a hydrogen
bond. Slightly greater o
NHΔ and
o
NHΔ in 32 than in 31 can be attributed to the increased steric
bulk 5'CH2OMe compared to 5'CH2OH. However, on the overall, N-type sugars are more preferred
at 298 K in 32 than 31 (see their Error! No topic specified. and oD
GΔ values), due to an entropy effect.
[In the P state, slightly larger opHΔ for 31 than for 32 can be easily explained: (i) 1H-NMR spectra of
31 and 32 were recorded at pD 1.9 and 2.0, respectively, but not at lower pDs due to rapid
decomposition. (ii) The standard deviations of ∆H°, ∆S° and ∆G° values at pD 1.9 for 31 (4.5
kJmol-1) and at pD 2.0 for 32 (3.6 kJmol-1) are higher than at other pDs, because 1H-NMR spectra
could only be recorded in a narrow temperature range (274 - 303 K). Thus ΔH°, -TΔS° and ΔG° of
31 and 32 are inevitably less accurate in the P than in the N and D states.] Our 1D-nOe difference
experiments in the P, N and D state of 31 and 32 show that guanin-9-yl assumes an anti orientation
in both compounds at any pD and this preference is almost independent of the pD and of the nature
of 5'-substituent. Moreover, the nucleobases in β-D-ddA (30), β-D-ddC (33) and β-D-ddT (34)
adopt almost exclusively anti orientations (78 - 86 % at room temperature) at neutral pD (Table 10).
The shift of the N � S equilibrium in the protonated and deprotonated states of β-D-ddNs toward
more N- and S-type sugars, respectively, compared with the N state is therefore neither the result of
any interaction between 5'-OH and the nucleobase nor induced by a different orientation of the
nucleobase around the glycosyl torsion at different pDs.
4.8.4 The orbital mixing as the origin of the O4'-C1'-N1/9 anomeric effect
The modulation of the thermodynamics of the two-state N � S equilibrium in β-D-ddNs 30
- 34 by the pD of the aqueous solution can be rationalized in terms of pD-dependent magnitude of
the O4'-C1'-N1/9 anomeric effect: As the nucleobase becomes protonated, a partial positive charge
is created at the glycosyl nitrogen. As a result, nO4' →σ∗C1'-N9 orbital mixing becomes more
favourable because of the lowering of the antibonding σ* orbital (Section 2.8). This favourable
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
96
orbital mixing prompts the nucleobase to adopt a pseudoaxial orientation, which in turn is achieved
in the N-type pseudorotamers. Conversely, as the nucleobase becomes deprotonated in the alkaline
pD, the ability of the glycosyl nitrogen to participate in O4'-C1'-N1/9 stereoelectronic interactions is
reduced, which is evident by the shift of N � S equilibrium to more S-type (with pseudoequatorial
aglycone) in comparison with the N state. Our interpretation is based on the assumption that the
steric bulk of the nucleobase remains to be comparable in the P, D and N states. Of course we
expect a change in the hydration chracteristics in the various cationic or anionic state vis-a-vis N
state, but that cannot be estimated experimentally with the present state-of-the art. Our results also
tend to suggest that the contribution of electrostatic repulsions (see Section 2.8, Fig 8) between the
dipole of the pentofuranose moiety and the C1'-N1/9 dipole to the overall strength of the O4'-C1'-
N1/9 anomeric effect in nucleosides and nucleotides is rather small in comparison with that of
hyperconjugative interactions (Figs 8 and 9).
Indeed, one can reasonably assume that the partial positive charge created at the glycosyl
nitrogen upon protonation of the nucleobase in β-D-ddNs will weaken the C1'-N1/9 dipole, which
in turn should lead to the reduction of the electrostatic repulsions with the O4' lonepair in the P
compared with N state of 30 - 33. Therefore, if the origin of the O4'-C1'-N1/9 anomeric effect in
nucleosides and nucleotides was mainly electrostatic, the preference of the nucleobase for
pseudoaxial orientation in the N-type pseudorotamers should be much reduced in the P state than in
the N state owing to the weakening of the anomeric effect in the former, but our results clearly show
the opposite, i.e. a higher preference for the N-type sugars at acidic compared with neutral pDs.
Similarly, upon deprotonation of the nucleobase in β-D-ddNs 31, 32 and 34, owing to the
presumably higher electron-density at the glycosyl nitrogen, one may also suggest that dipole-dipole
repulsions should be reinforced in comparison with the N or P state, and that the anomeric effect
should be more efficient in the D compared with the N or P state, which is again in contradiction
with the experimentally observed tendency of destabilization of N-type sugars in the alkaline
compared with the neutral solution (see above, section 4.8).
The strengthening of the effect of the nucleobase upon its protonation (or weakening upon
its deprotonation) can be estimated by directly subtracting oN
HΔ from opHΔ (or from o
DHΔ ). Thus
oNP
H−
ΔΔ (kJmol-1) = 3.0 (33, cytosin-1-yl) < 5.7 (30, adenin-9-yl) < 20.2 (31, guanin-9-yl) and 20.1
(32, guanin-9-yl) whereas oND
H−
�ΔΔ (kJmol-1) = -2.0 (31, guanin-9-yl) and (32, guanin-9-yl) ≈ -1.9
(34, thymin-1-yl). The subtraction of ∆∆H°
2 values in the P and N states (or N and D states) gives
identical values as oNP
H−
�ΔΔ and oND
H−
ΔΔ . Thus purine nucleosides respond more to the protonation
of their nucleobase than the pyrimidine counterparts, which is consistent with the fact that upon
addition of one equivalent trifluoroacetic acid the chemical shift of the glycosyl nitrogen is more
affected in the purine than in the pyrimidine series (∆δ (downfield shift of N9) = 6.6 ppm in N1H+-
adenosine and 6.7 ppm in N7H+-guanosine401,403 >> ∆δ (downfield shift of N1) = 1.1 ppm in
N3H+-cytosine404).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
97
4.8.5 No reverse anomeric effect in pentofuranosyl nucleosides
If the reverse anomeric effect were playing any role in the drive of the sugar conformation in
pentofuranosyl β-nucleosides at acidic pD, one should observe an increase in the population of S-
type pseudorotamers, in which the nucleobase adopts a pseudoequatorial orientation, in the acidic
compared to the neutral pD. Our results show the opposite trends, therefore the pD-dependent
conformational preferences of β-D-nucleosides can be attributed to the resulting modulation of the
anomeric effect, not a reverse anomeric effect, as suggested in the case of imidazolium or
pyridinium derivatives of pyranose (Section 1.7).
4.8.6 Variable tunablity of anomeric effect in β-D-ddNs, β-D-dNs and β-D-rNs
In protonated β-D-dA, 3'-OMe-β-D-dA and 3',5'-diOMe-β-D-dA, and deprotonated β-D-dU
and 5-F-β-D-dU, both ∆H° and -T∆S° stabilize S-type conformations either with the same
magnitude or slightly stronger ∆H°. In the D state of β-D-dG and β-D-T, ∆H° overrides the
counteracting -T∆S°, resulting in the overall stabilization of S-type sugars at 298 K. Only in
protonated β-D-dC, -T∆S° is slightly stronger than ∆H° and determines the drive of the N �S
equilibrium toward S-type sugars, whereas in protonated β-D-dImb and β-D-dG, ∆H° (driving to N)
and -T∆S° (driving to S) nearly cancel each other, therefore N and S-type sugars have almost the
same population at 298 K. The balance of ∆H° and -T∆S° terms of the N � S in β-D-rNs is also
dictated by the nature of the nucleobase and the pD: In the D state of β-D-U and 5-F-β-D-rU,
opposing ∆H° (driving to the N) and -T∆S° nearly cancel each other, resulting in nearly unbiased
equilibrium. However, in the D state of β-D-G, ∆H° (stabilizing S-type sugars) prevails over the
opposing -T∆S°, and S-type sugars are favoured at 298 K. In the P state of β-D-G and β-D-C, ∆H°
prevails over counteracting -T∆S°, stabilizing N-type sugars at 298 K. In the P state of β-D-A and D
state of β-D-rT, small negative ∆H° and -T∆S° values favour slightly S-type conformations at 298
K.
In agreement with our observations in the case of β-D-ddNs 30 - 34, as the nucleobase in β-
D-dNs and β-D-rNs is protonated, the two-state N � S equilibrium of the constituent sugar moieties
is also shifted toward N-type conformations [ oNP
G−
ΔΔ (kJmol-1) = 1.0 (37 and 38), 0.4 (39), 1.3 (40),
1.6 (41), 0.5 (42) for β-D-dNs and 1.3 (50), 3.0 (51), 0.4 (52) for β-D-rNs] and conversely, as it is
deprotonated, S-type sugars are further stabilized in comparison with the N state oND
G−
ΔΔ (kJmol-1) =
-1.0 (41), -0.3 (43 - 45) for β-D-dNs and -1.3 (51), -0.3 (53 and 55), -0.2 (54) for β-D-rNs]. The
comparison of oNP
G−
ΔΔ and oND
G−
ΔΔ values for each nucleobase in β-D-ddNs (30, 31, 33 and 34), β-
D-dNs (37, 41 - 43) and β-D-rNs (50 - 53) shows that the change in the ratio of N- and S-type
pseudorotamers in the P and D states in comparison with the N state is much reduced in the 2'-
deoxy and ribo series than in the 2',3'-dideoxy counterparts. This is owing to the fact that both the
overall entropy of the system and the enthalpy of the N � S equilibrium are much less affected by
the pD of the aqueous solution in (37, 41 - 43) compared with (30, 31, 33 and 34) and (50 - 53).
At any pD, the nucleobase in β-D-dNs 37 - 39 and 41 - 43 adopts preferentially an anti
orientation around the glycosyl torsion (except β-D-dA (37) and 3'-OMe-β-D-dA (38) at neutral pD
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
98
since ≈ 1:1 ratio of syn and anti rotamers was found, Table 10), as suggested by our 1D nOe
difference experiments37. These observations let us propose that N3 (in purines) or O2 (in
pyrimidines) does not form a hydrogen bond with 5'-OH. 37 - 39 all have rather similar ∆H°, -T∆S°
and ∆G° values at any pD, in spite of the fact that in 39 the shift from less to more syn rotamers
occurs in going from the N to the P state, not reverse. The fact that 37 and 39 behave in the same
way also implies that in 37 no H-bond between 5'-OH and adenin-9-yl is responsible for the
positive oNP
G−
ΔΔ and oNP
H−
ΔΔ values.
The modulation of the overall stereoelectronic forces (i.e. anomeric effect + gauche effects)
upon protonation and/or deprotonation of the constituent nucleobase in β-D-dNs and β-D-rNs is
reflected in the change in the enthalpy of the N � S equilibrium in the P and/or D compared with N
state, which has been estimated by subtracting their oN
HΔ values from the corresponding opHΔ and
oD
HΔ values, respectively :
oNP
H−
ΔΔ (kJmol-1) = 0.7 (42, cytosin-1-yl) < 2.3 (40, imidazol-1-yl) ≈ 2.6 (39, adenin-1-yl) <
3.2 (37, 38, adenin-1-yl) < 4.9 (41, guanin-9-yl) in β-D-dNs and 2.9 (52, cytosin-1-yl) < 4.2 (50,
adenin-1-yl) < 8.7 (51, guanin-9-yl), in β-D-rNs whereas oND
H−
ΔΔ (kJmol-1)= -0.3 (45, 5-fluoro-
uracil-1-yl) ≈ -0.5 (43, thymin-1-yl) ≈ -0.7 (44, uracil-1-yl) < -2.1 (41, guanin-9-yl) in β-D-dNs and
-1.5 (55, 5-fluoro-uracil-1-yl) = -1.5 (53, thymin-1-yl) ≈ -1.7 (54, uracil-1-yl) < -4.3 (51, guanin-9-
yl) in β-D-rNs (for oND
H−
ΔΔ , the < sign signifies a reduced ability of the deprotonated nucleobase to
destabilize N-type conformations with respect to the N state). [Note that the same values can be
derived from the subtraction of ∆∆H°3 values in the N state from those in the P and D states,
respectively, for β-D-dNs and from the subtraction of ∆∆H°4 values in the N state from those in
the P and D states, respectively, for β-D-rNs (Table 6, Fig 13)]. Thus, in all ddNs (Section 4.8),
dNs and rNs, pyrimidine nucleobases are much less able to transmit the free-energy of its
protonation � deprotonation equilibrium than the purine counterparts. The smaller value of oNP
H−
Δ
of 40 in comparison with those of 37 - 39 and 41 shows the effect of the electron-withdrawing
character of the fused pyrimidine ring upon the imidazole moiety in the purine nucleosides.
oNP
H−
ΔΔ and oND
H−
ΔΔ increase in the following order: β-D-dNs < β-D-rNs << β-D-ddNs
(Table 2). The reduced modulation in the dNs compared to the ddNs is presumably the result of the
opposing influence of the [HO3'-C3'-C4'-O4'] gauche effect in the former, which is also in
agreement with the fact that ∆∆H°2 values (reflecting the overall effect of the nucleobase in for β-D-
ddNs) are much larger than the corresponding ∆∆H°3 values (showing the effect of the base in β-D-
dNs) at any pD (Fig 13, Table 6).
As O4' is involved in the [HO3'-C3'-C4'-O4'] gauche effect, its electrostatic potential is
changed because of σC3'-H3' → σ∗C4'-O4' participation and its ability to be participate to O4'-C1'-N1/9
stereoelectronic nO4' → σ∗C1'-N interactions is reduced. In Table 6, ∆∆H°10
values show that as the
nucleobase is protonated, the strength of the [HO3'-C3'-C4'-O4'] gauche effect in β-D-dNs
increases, whereas it is reverse when the nucleobase is deprotonated.
Again, this is explained by the fact that as the O4' becomes more positively charged owing to
the enhanced nO4' → σ∗C1'-N interactions with the protonated aglycone, the energy level of σ∗C4'-O4'
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
99
is lowered (in comparison with the neutral state), which renders the σC3'-H3' → σ∗C4'-O4' participation
more efficient in the protonated nucleosides.
The modulation of the thermodynamics of N � S equilibrium in β-D-rNs is much smaller
than ddNs counterparts [the relative degree of pD-dependent flexibility follows the following order:
β-D-dNs < β-D-rNs << β-D-ddNs, Table 2] because of the interplay of three gauche effects ([O3'-
C3'-C4'-O4'], [O2'-C2'-C1'-O4'] and [O2'-C2'-C1'-N1/9]) in the former modulating the strength of
the anomeric effect.
Since the [O3'-C3'-C4'-O4'] and [O2'-C2'-C1'-O4'] gauche effects presumably cancel each
other (compare ∆H° of 12 and 14 in Table 2), the greater flexibility of rNs than dNs as a function of
pD may be attributed to the [O2'-C2'-C1'-N1/9] gauche effect. The magnitude of the 2'-OH effect in
β-D-rNs as a function of pD can be estimated by subtracting opHΔ (or o
NHΔ or o
DHΔ ) values of β-D-
dNs counterparts (∆∆H°11
in Table 6) from those of β-D-rNs. The ability of 2'-OH to drive the
pentofuranose conformation toward N-type pseudorotamers increases as follows: adenin-9-yl =
guanin-9-yl <<< uracil-1-yl ≈ thymin-1-yl ≈ cytosin-1-yl ≈ 5-fluorouracil-1-yl (in the N state),
adenin-9-yl << guanin-9-yl << cytosin-1-yl (in the P state) and guanin-9-yl <<< uracil-1-yl ≈
thymin-1-yl ≈ 5-fluorouracil-1-yl (in the D state) (see section 4.6 for correlation of the anomeric
effect and the overall 2'-OH effect in ribonucleosides and nucleotides, Table 6, Figs 13 and 17).
Therefore, in purines 2'-OH stabilizes more S-type conformations than in the pyrimidines at any pD.
This is owing to the reduced [O2'-C2'-C1'-N1(pyrimidine)] gauche effect in the former in
comparison with the [O2'-C2'-C1'-N9(purine)] gauche effect in the latter. In going from the P to the
N and D state, 2'-OH becomes steadily more efficient to drive the sugar conformation in rNs toward
S-type conformations, which are reflected in the values of ∆∆H°11
(Table 6, Figs 13 and 17).
4.8.7 Correlation of the electronic nature of aglycone with pseudorotational state
The monitoring of chemical shift of aromatic and anomeric protons is a dependable marker
to assess the protonation �deprotonation equilibrium, which when measured as a function of pH
yields the pKa
518. For each β-D-ddN 30 - 34, β-D-dN 37 and 40 - 45 and β-D-rN 50 - 55, we have
plotted the change in the chemical shift of the aromatic proton(s) of the constituent nucleobase as a
function of the ∆G° of the two-state N � S equilibrium over the whole pD range (Fig 18) to
examine if there is an interdependency. A straightline could be fitted through all experimental points
of each correlation plot. The Pearson's correlation coefficients (R) are larger than 0.97 for all
nucleosides except for β-D-T (43, R = 0.90) and β-D-U (54, R = 0.95) owing to the errors involved
in the determination of their pD-dependent ∆G° values as a result of the limited change in 3J
HH over
the 278 K - 358 K temperature range. This shows that the electronic changes that take place in the
aglycone as a result of pD-dependent change in the protonation � deprotonation equilibrium is
indeed transmitted to alter the ∆G° of the two-state N � S equilibrium over the whole pD range.
Interestingly, this transmission of the tunable electronic character of the nucleobase to steer the
sugar conformation in β-D-nucleosides takes place through pD-dependent change of the strength of
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
100
ΔGo
(kJ/mol)
-3 -2 -1 0
δH8
(288 K, ppm)
8.1
8.2
8.3
8.4
8.5
8.6
ΔG o (kJ/mol)
-3 -2 -1
δH8
(288 K, ppm)
7.6
8.0
8.4
8.8
9.2
ΔGo
(kJ/mol)
-3 -2 -1 0
δH6
(298 K, ppm)
7.6
7.7
7.8
7.9
8.0
8.1
α-D-ddA (17)
(A )
α-D-ddG (18) α-D-ddC (19)
(B ) (C )
ΔGo
(kJ/mol)
-0.9 -0.6 -0.3 0.0
δH6
(298 K, ppm)
7.2
7.3
7.4
7.5
7.6
7.7
ΔGo (kJ/mol)
-5 -4 -3 -2 -1
δH8
(288 K, ppm)
8.3
8.4
8.5
8.6
ΔGo
(kJ/mol)
-5 -4 -3 -2
δH8
(288 K, ppm)
8.0
8.4
8.8
9.2
α-D-ddT (20)
(D )
α-D-dA (21) α-D-dG (24)
(E ) (F)
ΔGo
(kJ/mol)
-4.4 -4.0 -3.6 -3.2
δH6
(298 K, ppm)
7.7
7.8
7.9
8.0
8.1
8.2
ΔGo (kJ/mol)
-2.0 -1.5 -1.0
δH6
(298 K, ppm)
7.6
ΔGo
(kJ/mol)
1 2 3 4 5 6
δH8
(298 K, ppm)
8.1
8.2
8.3
8.4
8.5
8.6
8.7
α-D-dC (25)
(G )
α-D-dT (26) β-D-ddA (30)
(H ) (I)
ΔGo
(kJ/mol)
2 4 6
δH6
(288 K, ppm)
8.0
8.4
8.8
9.2
ΔG o (kJ/mol)
3 4 5 6
δH8
(288 K, ppm)
7.6
8.0
8.4
8.8
9.2
ΔGo
(kJ/mol)
2 3 4 5 6
δH6
(298 K, ppm)
7.8
7.9
8.0
8.1
8.2
8.3
8.4
β-D-ddG (31)
(J)
5'-OMe-β-D-ddG (32) β-D-ddC (33)
(K ) (L)
ΔGo
(kJ/mol)
2 3 4
δH8
(298 K, ppm)
7.4
7.5
7.6
7.7
7.8
ΔG o (kJ/mol)
-2 -1
δH8
/δH2 (p
pm)
8.0
8.2
8.4
8.6
ΔGo
(kJ/mol)
-1 0
δHa
/δHb
/δHc
(ppm)
6.4
7.2
8.0
8.8
9.6
β-D-ddT (34)
(M )
β-D-dA (37) β-D-dImb (40)
(N ) (O )
ΔGo
(kJ/mol)
-3.2 -2.4 -1.6 -0.8 0.0
δH8
(ppm)
7.5
7.8
8.1
8.4
8.7
9.0
9.3
ΔGo (kJ/mol)
-1.2 -1.0 -0.8
δH6
(ppm)
7.7
7.8
7.9
8.0
8.1
ΔGo
(kJ/mol)
-1.8 -1.6 -1.4 -1.2
δH6
(ppm)
7.35
7.40
7.45
7.50
7.55
7.60
7.65
β-D-dG (41)
(P )
β-D-dC (42) β-D-T (43)
(Q ) (R )
Figure 18 (See the legend p. 116)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
101
ΔGo
(kJ/mol)
-1.5 -1.4 -1.3 -1.2 -1.1 -1.0
δH6
(ppm)
7.5
7.6
7.7
7.8
7.9
ΔGo
(kJ/mol)
-2.1 -1.4 -0.7 0.0
δH8
/δH2
(ppm)
8.0
8.1
8.2
8.3
8.4
8.5
8.6
8.7
β-D-dU (44)
(S)
β-D-A (50)
(U)
ΔGo
(kJ/mol)
-3 -2 -1 0 1
δH8
(ppm)
8.0
8.5
9.0
ΔGo (kJ/mol)
1.4 1.5 1.6 1.7 1.8 1.9 2.0
δH6
(ppm)
7.6
7.8
8.0
8.2
ΔGo
(kJ/mol)
-0.4 -0.3 -0.2 -0.1
δH6
(ppm)
7.4
7.5
7.6(V) (W) (X)
ΔGo
(kJ/mol)
-4.0 -3.6 -3.2 -2.8 -2.4 -2.0
δH8
(ppm)
7.9
8.0
8.1
8.2
8.3
8.4
8.5
ΔGo (kJ/mol)
-5 -4 -3
δH8
(ppm)
7.9
8.0
8.1
8.2
8.3
8.4
8.5
8.6
ΔGo
(kJ/mol)
-4 -3 -2 -1 0 1
δH6
(ppm)
7.6
7.7
(b) (c) (d)
ΔGo
(kJ/mol)
-2 0
δH1'
(ppm)
4.5
4.6
4.7
ΔGo (kJ/mol)
-2 -1
δH6
(ppm)
7.2
7.6
8.0
(e) (f)
ΔGo
(kJ/mol)
-1.6 -1.5 -1.4 -1.3 -1.2 -1.1
δH6
(ppm)
7.6
7.7
7.8
7.9
8.0
8.1
(T)
5-F-β-D-dU (45)
ΔGo
(kJ/mol)
0.0 0.1 0.2 0.3 0.4
δH6
(ppm)
7.4
7.5
7.6
7.7
7.8
7.9
ΔGo (kJ/mol)
0.1 0.2 0.3 0.4 0.5 0.6
δH6
(ppm)
7.6
7.8
8.0
ΔGo
(kJ/mol)
-4 -3 -2 -1
δH2
(ppm)
8.0
8.4
8.8(Y) (Z) (a)
β-D-G (51) β-D-rT (53)β-D-C (52)
β-D-U (54)
Formycin B (56)
5-F-β-D-U (55)
9-deaza-A (58)Formycin A (57) ψ-isoC (59)
1-Me-ψ-U (61)ψ-U (60)
Figure 18: Correlation plots of the chemical shifts of H2/H6/H8/H1' in α-D-ddNs 17 - 20 [Panels (A)-(D)], α-D-dNs
21, 24 - 26 [Panels (E)-(H)], β-D-ddNs 30 - 34 [Panels (I)-(M)], β-D-dNs 37 and 40 - 45 [Panels (N)-(T)], β-D-rNs 50 -
55 [Panels (U)-(Z)] and β-D-C-rNs 56 - 61 [Panels (a)-(f)] (at 298 K unless stated otherwise) as a function of pD-
dependent ∆G° of their N � S equilibrium. The Pearson's correlation coefficients (R) of the straight lines are > 0.96 for
all nucleosides except β-D-T (R = 0.90), β-D-U (R = 0.95), α-D-ddA (R = 0.62), α-D-dA (infinite slope) and formycin
B (R ≈ 0.92) owing to the errors involved in the determination of their pD-dependent ∆G° values as a result of the
limited change in 3JHH over the 278 K - 358 K range.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
102
anomeric and gauche effects. This change of N � S equilibrium further predisposes the
conformation of the phosphate backbone as discussed in section 7. Thus the single-stranded
nucleotide acts like a wire (Nucleotide wire)38,44.
5. Comparison of stereoelectronic effects in α- and β-D-nucleosides
β-D-nucleosides are solely chosen by Nature as the ubiquitous building blocks for the
storage of information in DNA and RNA. Only a few α-nucleosides are found in Nature519-522. The
biological (e.g. antitumor, bacteriostatic, cytostatic) activities of some α-nucleosides have been
reported523-530. Séquin531 has postulated that a double helix featuring stabilizing hydrogen bonds
through Watson-Crick base pairing and base-base stacking as in the natural DNA helix can be
formed between a chain of α-nucleotides and its complementary α-counterpart (chains of opposite
polarity) or its complementary β-counterpart (chains of the same polarity). NMR experiments on α-
hexadeoxyribonucleotides532,533 have subsequently confirmed Séquin's predictions. It has been
shown recently that native β-anomeric sequences containing a single α-anomeric nucleotide
(inserted via 3'-3' or 5'-5' phosphodiester linkage) form duplexes whose structures emulate canonical
B-DNA534, with subtle differences in stability and local structure dictated by the nature of the
nucleobase in the α-nucleotide. A possible relation between the natural selection of β- over α-
nucleosides in DNA and a lack of conformational variability of the pentofuranose moieties in the
latter has been proposed basing on the relative energies of various pseudorotamers in both series175.
However, no experimental comparison of the thermodynamics of the N � S equilibrium in α-
versus β-nucleosides has supported this theoretical work.
We have assessed36,37 for the first time how the interplay of anomeric and gauche effects in
α-nucleosides dictates their sugar conformation. We have initially36 considered only α-D-ddNs,
owing to the fact that only the nucleobase and 5'CH2OH drive their sugar conformation. We have
subsequently37 turned our attention to α-D-dNs in order to analyze the impact of the presence of 3'-
OH on the drive of their sugar conformation in comparison with the dideoxy counterparts. In these
works, we have uniquely shown that the change of the configuration at C1' results in a drastic
modulation of the strengths of stereoelectronic forces as well as of the pD-dependent flexibility of
the sugar conformation in the α- versus β-nucleosides.
5.1 The relative magnitude of the anomeric and gauche effects in Neutral state
5.1.1 Anti orientation of the nucleobase in α/β-D-ddN and α/β-D-dN
Our 1D-nOe difference experiments in the N state of α-D-ddNs 17 - 20 and α-D-dNs 21 - 26
show that except in the case of α-D-dA (44 % syn) and α-D-dG (65 % syn in the D state), the
constituent nucleobases in all α-nucleosides compounds adopt preferentially (67 - 93 %) an anti
orientation around the glycosyl torsion (at 298 K or 288 K) (Table 10). Additionally, the extent of
the stabilization of anti over syn rotamers is nearly independent of the configuration of the sugar
moiety at C1' in α- versus β-D-ddNs and in α- versus β-D-dNs. There are only a few exceptions: In
3'-O-Me-α-D-dA (22), adenyl-9-yl is almost exclusively anti whereas in 3'-OMe-β-D-dA (38) the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
103
population of syn rotamers is higher at any pD, and for α-D-dG (24) (in the D state), in which
guanin-9-yl adopts preferentially a syn orientation whereas in β-D-dG (41) the anti and syn rotamers
are almost equipopulated. These results have three important implications: (i) The observed
differences in the drive of the N � S equilibrium as a function of temperature and change of pD
(see below) in α-D-ddNs compared to their β-counterparts is not correlated with the preferred
orientation of the nucleobase around the C1'-N1/9 bond, since it is independent of the configuration
at C1'. (ii) In α-D-dNs, the anti orientation of the nucleobase prevents it from forming a hydrogen
bond with 3'-OH. (iii) In spite of the fact that syn orientations of adenyl-9-yl are more stable in α-D-
dA (21) than in 3'-OMe-α-D-dA (22) and in 3',5'-diOMe-α-D-dA (23) , the three compounds all
have pD-independent ∆H° values (see below) which discards any effect of the orientation of the
nucleobase in the former.
5.1.2 The balance of ∆H° and -T∆S° in neutral α- and β-D-N
In neutral α-D-ddNs 17 - 20, α-D-dNs 21 - 23, 25, 26 and α-L-dNs 27 - 29, the ∆H°
contribution (favouring S-type conformations) to the free-energy ∆G° of the two-state N � S
equilibrium prevails over the counteracting -T∆S°; Only in the case of α-D-ddG (18) they
cooperate, with slightly stronger -T∆S° than ∆H°. However, the absolute contribution of -T∆S° to
the free-energy ∆G° of the N � S equilibrium in α-nucleosides is negligible (0.2 kJmol-1) only in
the case of α-D-ddA (17). Unlike in dNs and rNs, where the bias of the N � S equilibrium can be
explained by the complex interplay of gauche and anomeric effects, the conformation of the sugar
moiety in ddNs is almost exclusively driven by the overall effect of the nucleobase, the contribution
of the 5'CH2OH group being minimal. Therefore, α-D-ddNs and β-D-ddNs constitute the systems of
choice to examine whether the magnitude of the O4'-C1'-N1/9 anomeric effect is dictated by the
configuration of the pentofuranosyl moiety at C1'. In the β-series (30 - 36), the effect of the
nucleobase drives toward N-type sugars (and cooperates with the much smaller effect of 5'CH2OH),
as evident from the positive ∆H° values (Table 2, Section 4). In α-D-ddNs 17 - 20, owing to the
change of configuration at C1', oN
HΔ values of the N � S equilibrium stabilize S-type conformations,
in agreement with the fact that nO4' →σ∗
C1'-N1/N9 stereoelectronic interactions, favouring S- over
N-type conformations, prevail over the counteracting steric effect of the nucleobase (driving to the
N) as found in the β-anomers. Since oN
HΔ values are the main contribution to oN
G values, S-type
sugars are stabilized at room temperature (vide supra). However, the comparison of oN
GΔ and oN
HΔ
values of the N �S equilibrium in β-D-ddNs 30, 31, 33 and 34 and in α-D-ddNs 17 - 20 shows that
the magnitude of the free-energy and enthalpy stabilization of the N-type sugars in the former is ≈ 2
- 6 times and ≈ 2 - 8 times greater, respectively, than the stabilization of S-type conformers in the
later, depending on the nature of the nucleobase. This means that the pentofuranose moiety in β-D-
ddNs is more prone to adopt multiple conformations as a result of the change of the temperature
than the α-counterparts.
In α-D/L-dNs 21 - 29, oN
HΔ as well as the overall oN
GΔ stabilize more S-type conformations
than in the parent α-D-ddNs (Table 2). This is in agreement with the observation that in the former
both the effect of the nucleobase and the [O3'-C3'-C4'-O4'] gauche effect are expected to prefer S-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
104
type pseudorotamers, whereas in the later, only the effect of the nucleobase is operating. Larger
negative oN
HΔ values for α-D/L-dNs 21 - 29 than for β-D/L-dNs 37 - 39, 41 - 43 and 46 - 49 are
consistent with the fact that in the former both the [O3'-C3'-C4'-O4'] gauche effect and the effect of
the nucleobase stabilize the S-type sugars whereas in the latter the gauche effect counteracts the
effect of the nucleobase driving the conformation toward N-type geometries. However, the
additional enthalpy or free-energy stabilization of S-type conformations in the α-series is less than
or equal to -2.6 kJmol-1 for almost all compounds. oN
HΔ of α-D-dC (25) appears to be the only
exception, since its sugar conformation is driven much more efficiently toward S-type geometries
than that of β-D-dC (42) (Table 2).
In view of the above remarks and since both 5'CH2OH and the steric bulk of the nucleobase
are conserved structural features in both α-and β-D-ddNs, it is reasonable to attribute the lack of
flexibility of the sugar conformation in the former compared with the latter to much less efficient
O4'-C1'-N1/9 anomeric effect in the α-series, owing to the change of configuration at C1'.
5.1.3 Weakening of 5'-substituent effect in α- compared with β-nucleosides
The weakening of the 5'CH2OMe effect in the α- compared to the β-series is shown by the
smaller value of ∆∆H°1d [for 3'-OMe-α-D-dA (22)] compared with ∆∆H°1b [for 3'-OMe-β-D-dA
(38)].
5.1.4 3'-gauche effect weakens anomeric effect in α-D-dN compared to α-D-ddN
The subtraction (∆∆H°13
) of oN
HΔ of abasic sugar 12 from that of a particular α-D-ddN 17 -
20 provides an estimate for the magnitude of the overall effect of the nucleobase in α-D-ddNs. Thus
the ability of the nucleobase to promote the stabilization of S-type conformations increases as
follows (in kJmol-1): guanin-9-yl in 18 (-0.8) < thymin-1-yl in 20 (-1.5) < adenin-9-yl in 17 (-2.1) <
cytosin-1-yl in 19 (-3.3). The subtraction (∆∆H°14
) of oN
HΔ of 13 from that of an α-D-dN 21 or 24 -
26 gives the strength of the effect of the nucleobase in α-D-dNs, assuming that the magnitude of the
[HO3'-C3'-C4'-O4'] gauche effect in α-D-dNs and in abasic sugar 13 is the same. The ability of
∆∆H°14 to stabilize S-type pseudorotamers increases as follows (in kJmol-1): thymin-1-yl (0.1) <
guanin-9-yl (-0.4) < adenin-9-yl (-0.9) < cytosin-1-yl (-3.0). Thus, upon substitution of H3" in α-D-
ddN 17 - 20 for 3'-OH in α-D-dNs 21 and 24 - 26 the effect of each nucleobase is slightly reduced in
the latter compared to the former. This may originate from the change of the electrostatic potential
around O4' in the 2'-deoxy compared to the 2',3'-dideoxy counterparts, as it becomes also involved
in the [HO3'-C3'-C4'-O4'] gauche effect in the former. A similar trend has also been noticed in
going from β-D-ddNs to β-D-dNs. (Section 4). However the weakening of the effect of the
nucleobase in the β-D-dNs (calculated by subtracting ∆∆H°3 from ∆∆H°2) is much more
pronounced than in the α-D-dNs (subtraction of ∆∆H°14 from ∆∆H°13) [the weakening is: 2.9
kJmol-1 for β-D-dA compared to β-D-ddA but only by 1.2 kJmol-1 for α-D-dA compared α-D-ddA,
1.7 kJmol-1 for β-D-dG compared to β-D-ddG but only by 0.4 kJmol-1 for α-D-dG compared α-D-
ddG, 2.8 kJmol-1 for β-D-dC compared to β-D-ddC but only by 0.3 kJmol-1 for α-D-dC compared
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
105
α-D-ddC, 2.3 kJmol-1 β-D-T compared to β-D-ddT but only 1.6 kJmol-1 for α-D-T compared α-D-
ddT].
5.1.5 Weakening of the 3'-gauche effect in α-D/L-dN compared with β-D/L-dN
∆∆H°7 gives an estimate for the [HO3'-C3'-C4'-O4'] gauche effect in abasic sugar 2 (-4.5
kJmol-1) (Table 6). In order to assess the contribution of the [HO3'-C3'-C4'-O4'] gauche effect to the
drive of the sugar conformation in α-D-dNs 21 and 24 - 26, we have subtracted (∆∆H°17) from their
oN
HΔ values those of the parent α-D-ddNs 17 - 20. ∆∆H°17 is weakest in α-D-T (20) (-2.9 kJmol-1)
and strongest in α-D-dC (19) (-4.2 kJmol-1). Thus the ability of the gauche effect of [HO3'-C3'-C4'-
O4'] to stabilize S-type conformations in α-D-dNs is of the same order of magnitude as in abasic
sugar 2 but it is also much reduced than in the β-D-dNs counterparts (compare ∆∆H°17 and ∆∆H°10
in Table 6). A qualitative trend between the reduced magnitudes of the effect of the nucleobase and
of the [HO3'-C3'-C4'-O4'] gauche effect in α- compared to β-D-dNs emerges from the comparison
of ∆∆H°17 and ∆∆H°10, on one hand, of ∆∆H°14 and ∆∆H°3, on the other. The comparison of
∆∆H°18 and ∆∆H°12 values shows that both in α- and in β-nucleosides, the substitution of 3'-OH for
3'-OMe results in a similar small stabilization of S-type conformations.
5.2 The relative magnitude of the anomeric and gauche effects in the ionic states
5.2.1 Virtualy identical pKa values of the nucleobase in α- and β-nucleosides
The pKa values, derived from pD-dependent 1H chemical shifts (Section 3.7) of the
constituent nucleobases in α-D-ddNs versus β-D-ddNs counterparts, on one hand, in α-D-dNs
versus β-D-dNs, on the other, are almost identical (within ± 0.2 and ± 0.3 pD unit in the ddN and
dN series, respectively) (Table 2). This suggests that the electronic character of the nucleobase
remains unchanged as the configuration of the pentofuranose at C1' is inverted. It also implies that
the influence of the configuration of the sugar moiety at C1' upon the magnitude of the O4'-C1'-
N1/9 anomeric effect cannot be attributed to the modulation of the electronegativity of the glycosyl
nitrogen in the α- compared to the β-series.
5.2.2 Predominant enthalpy over entropy in the ionic states of α-D-ddN and -dN
As in the neutral solution, in the P state of α-D-ddA (17), α-D-ddG (18) and α-D-ddC (19), opHΔ s responsible for the free-energy stabilization of S-type pseudorotamers, since it overrides the
counteracting -T opSΔ term. In the D state, enthalpy and entropy have nearly the same strength, but
whereas they counteract each other in α-D-ddT (resulting in nearly unbiased pseudorotational
equilibrium), in α-D-ddG they both stabilize S-type sugars. In the P and D states, as in the N state,
the enthalpy or free-energy stabilization of S-type conformers in α-D-ddNs is much reduced
compared to the corresponding preference for N-type forms in the β-D-ddNs counterparts. Both in
the P and D states of α-D-dNs 21 - 26, ∆H° prevails over the opposing -T∆S° as reflected in the
overall preference for S- over N-type sugars at room temperature.
5.2.3 Weaker anomeric effect in α-D-ddN gives poorer flexibility
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
106
In going from the N to the P state of α-D-ddA (17) and α-D-ddC (19) and from the N to the
D state of α-D-ddG (18), both ∆H° and -T∆S° contributions to ∆G° of their N � S equilibria remain
unchanged and the preference of their constituent sugar moieties for S-type conformations remains
the same (Table 2). This means that the free-energy of their protonation � deprotonation
equilibrium is not transmitted to enhance (for 17 and 19) or weaken (for 18) the N-aglycone effect in
the P state (D for 18) compared to the neutral pD. This is in sharp contrast with what has been found
for the β-D-ddNs counterparts (Section 4.8). pD-dependent conformational preferences in α-D-
ddNs have only been observed upon protonation of guanin-9-yl in α-D-ddG (18) [ oNP
G−
ΔΔ = -1.5
kJmol-1] owing to the enhancement of the N-aglycone substituent effect [ oNP
H−
ΔΔ = -8.3 kJmol-1] and
upon deprotonation of thymin-1-yl oND
G−
ΔΔ = 0.3 kJmol-1] owing to the weakening N-aglycone
substituent effect oND
H−
ΔΔ = 0.6 kJmol-1]. These ∆∆H° and ∆∆G° values are however much smaller
than in the β-counterparts, confirming our above conclusion on relative ability of nucleobase in α-
and β-anomers to steer the sugar conformation.
5.2.4 The interplay of pD-independent ∆H° and ∆S° in α-D-dN
In all α-D-dNs [except α-D-dG (24)], protonation and deprotonation of the nucleobase has
almost no effect on the enthalpy of the two-state N � S equilibrium, showing that the efficiencies of
the effect of the N-aglycone and of the [HO3'-C3'-C4'-O4'] gauche effect are insensitive to the pD.
This result is in agreement with our finding for α-D-ddNs, and it is in sharp contrast with the
situation in β-D-dNs, in which a clear pD-dependent modulation of the anomeric effect has been
observed, as shown by non negligible ∆∆H° values (Section 4.8).
However, whereas in α-D-dA (21) the preference for S-type conformations remains the same
at all pDs, owing to the fact that the entropy term is a constant factor, in the pyrimidine nucleoside
α-D-dC (25), we have observed an entropy stabilization of S-type conformers in the P compared to
the N state [ oNP
)S−
Δ−( T = -1.5 kJmol-1]. Conversely, in α-D-T (26), as thymin-9-yl becomes
deprotonated, N-type pseudorotamers become less unstable than in the neural solution, which is also
the result of an entropy effect [ oND
)S−
Δ−( T = 0.7 kJmol-1]. A comparison of oNP
G−
ΔΔ and oND
G−
ΔΔ values
for α-D-dC, α-D-T and their β-D-counterparts shows that in fact, the entropy in the former
modulates more efficiently the bias of the two-state N � S equilibrium than the combined pD-
dependent enthalpy and entropy in the latter.
In α-D-dG (24), as guanin-9-yl is protonated, both the effect of the N-aglycone [ oNP
)S−
Δ−( T = -
6.2 kJmol-1) and the counteracting -T∆S° contributions [ oNP
)S−
Δ−( T = 3.7 kJmol-1] are stronger than
in the N state, and further stabilize S- and N-type sugars, respectively. On the overall, since oNP
)H−
(Δ
> Δ oNP
)S−
Δ−( T , the population of S-type pseudorotamers is higher in the P state than in the N state.
The strengthening of the N-aglycone effect in going from the N to the P state is however much
weaker in α-D-dG (24) than in α-D-ddG (18), showing the effect of 3'-OH in the former (compare
their oNP
H−
ΔΔ values).
5.2.5 Poor correlation of the nature of aglycone with pseudorotation in α-D-N
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
107
The correlation plots of the pD-dependent 1H chemical shifts as a function of ∆G° of the N
� S equilibria in α-D-ddNs (17 - 20), α-D-dA (21), α-D-dG (24), α-D-dC (25) and α-D-T (26) all
give straight lines (Fig 18, Panels (A) - (H)). The Pearson's correlation coefficients for the plots are
above 0.96 for α-D-ddNs and above 0.99 for the corresponding α-D-dNs, except for α-D-ddA (17,
R = 0.62) and α-D-dA (21), straight line with infinite slope) owing to virtually pD-independent ∆G°
values. For α-D-ddNs, α-D-dA (21) and α-D-dG (24), the slopes of the plots are much larger than in
the case of β-anomers, showing the less efficient transmission of the pD-tunable electronic character
of the nucleobase to drive the sugar conformation. Conversely, the pD-dependent entropy alone is
more efficient to drive the sugar conformation in α-D-dC (25) and α-D-T (26) than the combined
pD-dependent enthalpy and entropy in β-D-dC (42) and β-D-T (43), since, in absolute value, the
slopes of the correlation plots in the former are much smaller than those of the later.
6. Quantitation of the anomeric effects in C- and N-nucleosides
We herein report on our initial24,25 and revised32,33 pD-dependent conformational studies on
C-nucleosides 56 - 62. Basing on the results of these works, we have developed a new method for
the estimation25,33 of the stereoelectronic nO4'
→σ∗C1'-N9 anomeric effect in β-D-A, β-D-dA, β-D-G
and β-D-dG through a set of pairwise comparisons of ∆H° of their N � S equilibria with those of
purine C-nucleosides, which were used as reference points for the quantitation of the counteracting
steric effect of the N-nucleobases.
6.1 The anomeric effect in C-nucleosides
The interplay of nO4'→σ∗C1'-N9 interactions and of the counteracting steric effect determines
the overall effect of a nucleobase upon the sugar conformation in an N-nucleoside. The relative
importance of both contributions can be assessed by monitoring the magnitude of the enthalpy of the
C1'-substituent effect (∆H°C1'-subst.) in comparison with ∆H° of all steric and stereoelectronic forces
driving the N � S equilibrium: If the nucleobase acts as a C-substituent, it preferentially takes up a
pseudoequatorial orientation, which in turn shifts the N � S equilibrium to the S (i.e. negative
∆H°C1'-subst.
value), whereas predominant stereoelectronic interactions stabilize N-type sugars in
which the nucleobase is pseudoaxial (i.e. positive ∆H°C1'-subst.
value). If both terms are of equal
strength and cancel each other, ∆H°C1'-subst.
is zero.
The strength of the (stereoelectronic) anomeric effect alone can only be estimated when the
magnitude of the counteracting steric term is known. If one can engineer a nucleoside X in which
the nucleobase (isosteric to that of the N-nucleoside in which we want to quantitate the anomeric
effect) drives the N � S equilibrium exclusively through its steric effect (stereoelectronic
interactions being minimal), ∆H° of the N � S equilibrium in X can be used as a reference point for
the steric effect of the nucleobase in the N-nucleoside. Making use of this observation, we have
seeked for a family of nucleosides in which the effect of the nucleobase exerts its influence as much
as possible through its steric effect, while the opposing stereoelectronic component remains
negligible.
This has led us to turn our attention to C-nucleosides, since we expected that substitution of
the glycosyl nitrogen in N-nucleosides for C1' in C-nucleosides would result in minimal
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
108
stereoelectronic interactions between O4' and the nucleobase. In the crystal structures of N-
nucleosides, the O4'-C1' bond is shorter (by ≈ 0.03 Å) than the C4'-O4' bond, supporting the
existence of the O4'-C1'-N1/9 anomeric effect (Section 2.8). In contrast, no such clear trend is
observed in the crystal structures of C-nucleosides: in general the difference between O4'-C4' and
O4'-C1' bond lengths is much reduced (∆ ≈ 0.01 Å) compared with N-nucleosides: For 4-thio-ψ-
uridine271 (three molecular structures in the crystalline cell), ψ-isocytidine hydrochloride272, (α-D)
epishowdomycin monohydrate273, 3'-deoxyformycin A hydrochloride274, 5'-chloro-3',5'-dideoxy β-
L-formycin A monohydrate275, 1-benzyl-2'-deoxyshowdomycin276, oxoformycin B277, formycin A
monohydrate278 and α-pseudouridine monohydrate279, C4'-O4' and C1'-O4' bonds have nearly the
same length (∆ ≈ 0.01 Å), whereas C4'-O4' is longer than C1'-O4' in formycin A hydrobromide
monohydrate280 (C4'-O4' = 1.47 Å, C1'-O4' = 1.41 Å), formycin B277 (C4'-O4' = 1.46 Å, C1'-O4' =
1.42 Å) and its hydrochloride281 (C4'-O4' = 1.46 Å, C1'-O4' = 1.43 Å). ψ-uridine crystallizes in the
form of two structures282. In one of them C4'-O4' (1.45 Å) is longer than C1'-O4' (1.42 Å), but in
the other it is the opposite (C4'-O4' = 1.42 Å, C1'-O4' = 1.44 Å).
We have shown that in β-D-rNs 50 - 55, the modulation of the bias of the N � S
equilibrium by the pD of the D2O solution can be attributed to the resulting fine tuning of the effect
of the nucleobase (or to the anomeric effect alone, assuming that the steric effect is a constant factor
at any pD) and of the [HO2'-C2'-C1'-N1/9] gauche effect (Section 4.8). In ribo C-nucleosides 56 -
62, the [HO2'-C2'-C1'-N1/9] gauche effect is absent. Therefore, any differences observed in the
conformational preferences of the sugar moiety in 56 - 62 at various pDs is owing to the different
steric and electronic nature of their C1'-aglycones.
6.1.1 Effect of the C1'-pyrimidine aglycone on the conformation of the sugar
C-nucleosides differ from natural N-nucleosides by the unique carbon-carbon link between
C9 of purines or C5 of pyrimidines and C1' of the pentofuranose sugar. Many of them have been
isolated as antibiotics, and have antiviral and/or anticancer activity166,535-540. Their presence in
tRNAs is absolutely vital to the biochemical function541,542. N-nucleosides are known to carry the
genetic information, but our quantitative and other qualitative studies have also shown that their
conformation, and in turn, biological function, can be engineered by changing the nature of either
the nucleobase, the C2'-C4' substituents or the electronic character of the endocyclic O4' atom upon
their substitution. Very little is known on how the absence of a glycosyl nitrogen in C-nucleosides
affects their conformation and to which extent the stereoelectronic partnership of their nucleobase
and pentofuranose moieties is affected by this modification. It is necessary to examine carefully
whether the absence of N1/9 in the C1'-aglycone really results in the complete cancellation of its
stereoelectronic interaction with O4', as suggested by part of the above data on X-ray crystal
structures of C-nucleosides.
The conformational analysis Ψ-isocytidine (59), its hydrochloride (59b), 1-Me-Ψ-uridine
(61) and 1,3-diMe-Ψ-uridine (62) was initially24 performed in neat D2O solution using the
methodology described in Section 3. The thermodynamics (Table 11) of the N �S equilibria in 59,
59b, 61 and 62, suggest that: (i) Only in Ψ-isocytidine (59), ∆H°, driving to S-type sugars, prevails
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
109
clearly over the counteracting -T∆S°, therefore at 298 K, S-type sugars are preferred. (ii) In the
hydrochloride (59b), ∆H° (driving to the N) and counteracting -T∆S° almost cancel each other, and
the pseudorotational equilibrium is unbiased. (iii) In 1-Me-Ψ-uridine (61) and 1,3-diMe-Ψ-uridine
(62), -T∆S° is the predominant factor driving the conformation to the S over counteracting ∆H°. We
have estimated the effect of C1'-aglycones in 59, 59b, 61 and 62 by subtracting (i.e. ∆∆H°30 in Fig
13) ∆H° of 14 from their ∆H°, which gives (in kJmol-1): -2.5 (59) < 0.5 (61) < 1.6 (62) < 3.5 (59b).
To calculate ∆∆H°30 the strength of the [O3'-C3'-C4'-O4'] and [O2'-C2'-C1'-O4'] gauche effects and
of the 5'CH2OH substituent effect is assumed to be the same in C-nucleosides and in 14. Only
isocytosin-5-yl in 59 adopts preferentially a pseudoequatorial orientation in S-type sugars, however
this preference is rather small. The comparison of ∆∆H°30 values for 59, 61, 62 and 59b suggests
that only in 59, isocytosin-5-yl really acts as a C-substituent to drive the sugar conformation toward
S-type pseudorotamers. In all other cases, ∆∆H°30 is slightly positive, which indicates a drive
toward N-type conformation, and suggests that the steric effect of the nucleobase (stabilizing S-type
sugars) is not the predominant factor contributing to its overall effect. This prompts us to suggest
the existence of stereoelectronic interactions (see below) between the pentofuranose sugar and the
nucleobase, i.e. the nO4'
→σ∗
C1'-C5(sp2) anomeric effect by analogy with what we found for the N-
nucleosides counterparts (Section 4).
Table 11. The ΔH˚ and ΔS˚ of N � S equilibrium in C-nucleosides 56 - 59, 61 and 62 from our
initial studies at a single pD in native D2O solutiona
Compounds ΔH˚ ΔS˚ -TΔS˚ ΔG298 %S278 b %S358 b ∆% S b (358-278 K)
formycin (A) (57)c -7.7 (0.7) -14.7 (0.9) 4.4 -3.3 83 69 -14
formycin (B) (56)c -7.1 (0.6) -13.4 (0.8) 4.0 -3.1 81 68 -13
9-deaza-A (58)c -5.2 (1.2) -5.4 (2.0) 1.6 -3.6 83 75 -8
Ψ-isoC (59)d -2.1 (0.3) -3.0 (2.0) 0.9 -1.2 63 58 -5
Ψ-isoC hydrochloride (59b)d 3.9 (0.2) 12.5 (0.6) -3.7 0.2 45 55 +10
1-Me-Ψ-U (61)d 0.9 (0.2) 5.1 (0.8) -1.5 -0.6 55 58 +3
1,3-diMe-Ψ-U (62)d 2.0 (0.2) 7.9 (0.5) -2.4 -0.4 52 57 +5
a ΔH°, -TΔS° (at 298 K) and ΔG298 are given in kJ/mol, ΔS° in J/molK. b ∆% S (358-278 K) = %S358 - %S278 shows
the change of population of South type sugar owing to the net result of ΔH° and -TΔS° contribution as a function of
temperature c Data taken from ref. 25. d Data taken from ref.24.
The remarkable shift of the N � S equilibrium of 59b with respect to that of 59 to the N
upon protonation of isocytosin-5-yl is easily explained on the basis of the results of our ab initio
calculations on 5-methyl-isocytosine and N1H+-5-methyl-isocytosine suggesting that C5=C6 has a
higher double bond character (since it is shorter) in 59b than in 59. In 59, C5=C6 π-electrons will
easily be delocalized in the pyrimidine ring, and much less accessible for an interaction with nO4'
than in 59b, where the protonated guanidine skeleton results in the more localized C5=C6 π-
electrons, leading to a stronger steroelectronic anomeric effect nO4' →σ∗C1'-C5(sp2) .
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
110
The main results of this work are as follows: (i) Pyrimidine C1'-aglycones in 59, 61 and 62
do not provide good reference points for the estimation of the stereoelectronic O4'-C1'-N1 anomeric
effect in the pyrimidine N-nucleosides, since they do not act exclusively as C-substituents, as shown
by negative ∆∆H°30 values, owing to non negligible nO4' →σ∗
C1'-C5(sp2) stereoelectronic
interactions. (ii) Protonation of isocytosin-5-yl in 59b results in the strengthening of these
stereoelectrionic interactions, as in the case of the anomeric effect in N-nucleosides.
6.1.2 Effect of the C1'-purine aglycone on the drive of the sugar conformation
Our initial conformational analyses25 on purine C-nucleosides formycin B (56), formycin A
(57) and 9-deaza-A (58) have also been performed in D2O solution at neutral pD (see the next
sections for the discussion of the protonation state of 9-deaza-A in this work). The resulting
thermodynamics of their N � S equilibria are presented in Table 11, and their comparison suggests
the main following conclusions: (i) The pseudorotational equilibrium in 56 - 58 is driven toward the
S-type conformation by the ∆H° term, which prevails over the counteracting -T∆S° contribution to
the free-energy and therefore S-type puckered geometries are favoured at room temperature. (ii) The
C1'-aglycone in 56 - 58, in sharp contrast with the situation in the pyrimidine C-nucleosides
counterparts 59 and 61 - 62, acts as an efficient C-substituent pushing the pseudorotational
equilibrium toward S-type conformations. In the purine series, under the present experimental
conditions, stereoelectronic interactions nO4' and the σ*C1'-C9(sp2) orbital are rather inefficient in
comparison with the counteracting steric effect of the C1'-aglycone (see the next paragraph for the
discussion on the case of 9-deaza-A). The strength of the overall C1'-substituent effect upon the
drive of the sugar conformation in 56 - 58 has been estimated by subtracting from their ∆H° values
that of the abasic sugar 14, which yields: ∆∆H°30 = -8.1 kJmol-1 (formycin A) ≈ -7.5 kJmol-1
(formycin B) > -5.6 kJmol-1 (9-deaza-A). The ">" sign indicates the increased ability of the
nucleobase to shift the pseudorotational equilibrium toward S-type conformation through their steric
effect. Therefore, in this work, the C1'-aglycone in formycin B (or formycin A) is the most
pseudoequatorially oriented among those of 56 - 58 and it represents the best reference point for the
estimation of the steric effect of adenin-9-yl in β-D-dA (37) and β-D-A (50) or guanin-9-yl in β-D-
dG (41) and β-D-G (51) (see below).
6.1.3 pD-tunable anomeric effect in C-nucleosides
The thermodynamics of the N � S equilibrium in formycin A (56), formycin B (57), 9-
deaza-A (58), Ψ-isoC (59), Ψ-U (60), 1-Me-Ψ-U (61) and 1,3-diMe-Ψ-U (62) have been
subsequently determined32 over part or the whole 0.5 - 12.0 pD range in order to examine whether,
as suggested above, stereoelectronic nO4' →σ∗C1'-C5/9(sp2) interactions also participate to the drive
of the conformation of the pentofuranose sugar, in a similar manner to the anomeric effect operating
in N-nucleosides (Section 4). We argued, that if an anomeric effect would also exist in C-
nucleosides, it should also be reflected in the shift of the bias of their two-state N �S equilibria
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
111
toward more N- and S-type conformations owing to increased and decreased electron-withdrawing
character of the nucleobase in the P and D states, respectively, compared with the N state.
6.1.4 Transmission of the nature of the C-aglycone drives the N�S equilibrium in C-
nucleosides
In this revised study, we found that ∆H°, -T∆S° and ∆G° values of the N � S equilibrium in
all C-nucleosides 56 - 61, including the purine derivatives 56 - 58, are pD-dependent (Table 232,
Fig. 11, Panel (F)). Just as in the case of β-D-rNs, the energetics of the protonation � deprotonation
equilibrium of the nucleobase are transmitted to steer the sugar conformation through the
modulation of its overall effect. This means that our previous hypothesis of the minimal
contribution of the anomeric effect to the drive of the sugar conformation in purine C-nucleosides
56 - 58 was an oversimplification. Assuming that the steric effect of the nucleobase is constant over
the whole pD range, it is possible to attribute the pD-dependent conformational preferences of the
pentofuranose sugar in 56 - 61 to the tuning of the nO4' →σ∗
C1'-C5(sp2) stereoelectronic interactions.
The experimental pD-dependent ∆H°, -T∆S° and ∆G° values for 56 - 61 have been fitted to
the Henderson-Hasselbach equation to give the limiting values in the P, N and D states and the pKa
of the constituent nucleobases, as described in Section 3. These pKa values are nearly identical to
those found in the literature34,385-390, and to our independent estimates derived from pD-dependent
chemical shifts of aromatic and anomeric protons (Table 2). The efficient communication of
stereoelectronic information between constituent nucleobase and pentofuranose is further evidenced
by the fact that the plots (Fig 18) of the chemical shifts of the aromatic protons as a function of pD-
dependent ∆G° values all give straight lines with high correlation coefficients (> 0.96 except for
formycin B, owing to the smaller change in ∆G° values in the alkaline pD range, Table 2). In each
of the P, N and D states, the pentofuranose moieties in purine C-nucleosides 56 - 58 prefer more S-
type conformations than in the pyrimidine counterparts, in agreement with what we found for β-D-
dNs and β-D-rNs (Table 2).
6.1.5 Estimates for the thermodynamics of the N � S equilibrium
For formycin A (57) and formycin B (56), Ψ-isoC (59), 1-Me-Ψ-U (61) and 1,3-diMe-Ψ-U
(62), almost identical oN
HΔ , -T oNSΔ and o
NGΔ values were found in our earlier24,25 and latest32 studies
[the largest difference is 1.0 kJmol-1 for oN
HΔ of formycin B, which is within the standard deviation
of both estimates]. For 9-deaza-A (58), the comparison of our earlier estimates (Table 11) and those
from the updated study (Table 2) suggested that in the earlier work, the pD of the aqueous solution
was in the acidic range, since in the P state (pD ≤ 5.2), opHΔ , -T o
pSΔ and opGΔ are respectively: -7.4 (σ
= 0.8), 3.7 (σ = 0.6) and -3.6 (σ = 0.4) kJmol-1 whereas in the N state oN
HΔ , -T oNSΔ and o
NGΔ are -
14.2 (σ = 1.5), 9.1 (σ = 1.1) and -5.0 (σ = 0.4) kJmol-1 (Table 2).The above standard deviations,
taken at a single pD in the original publication32, are higher than the errors shown in Table 2. The
difference of 2 kJmol-1 between opHΔ and -T o
pSΔ values of the earlier and latest work is within the
sum of the standard deviations of the estimates.
6.1.6 Enhanced anomeric effect upon protonation of the aglycone
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
112
In the P state, formycin B (56), formycin A (57) and 9-deaza-A (58) prefer S-type
conformations, owing to cooperative opHΔ and -T o
pSΔ values (in 56), or to predominant opHΔ over the
counteracting -T opSΔ (in 57 and 58). In contrast, in the P state of protonated Ψ-isoC, o
pHΔ is slighlty
stronger than the opposing -T opSΔ contribution, which results in the slight stabilization of the N-type
conformers. Protonation of formycin B at N3387,388, formycin A at N3385,386, 9-deaza-A at N3 (as
shown by our pD-dependent 13C chemical shifts in ref. 34) and of Ψ-isoC at N1389 shifts their N �
S equilibria toward more N-type forms until full protonation of the heterocycle is achieved, which is
reflected in the plateau of ∆G°, ∆H° and -T∆S° values in the P state (Table 2, Fig. 11, Panel (F)).
The additional stabilization of N type sugars in the P compared to the N state is given by oNP
G−
ΔΔ
values (in kJmol-1): 2.0 (for 56), 1.4 (57), 1.4 (58) and 1.9 (59). The corresponding oNP
H−
ΔΔ values
(kJmol-1) are as follows: 7.6 (56), 5.7 (57), 6.8 (58) and 6.1 (59). The change of the steric effect of
the nucleobase in 56 - 59 upon protonation cannot account for these oNP
H−
ΔΔ and oND
G−
ΔΔ values: As
the nucleobase is protonated, its steric effect will presumably increase, therefore this should shift the
pseudorotational equilibrium toward more S-type conformations (with pseudoequatorially oriented
nucleobase), however we observe the opposite. This validates our hypothesis of the tuning of nO4'
→σ∗
C1'-C5(sp2) stereoelectronic interactions by the pD of the aqueous solution.
6.1.7 Weaker anomeric effect upon deprotonation of the aglycone
Deprotonation of formycin A at N7 does not affect the bias of its N � S equilibrium
compared with the N state. However, deprotonation of Ψ-isoC (59), at N3, of Ψ-U (60) at N1 and
N3, and of 1-Me-Ψ-U (61) at N3 results into the increased population of S-type sugars in
comparison with the N state, until a plateau in the D state is reached when the nucleobase is fully
deprotonated. The additional stabilization of S-type sugars in the D state compared to the N state is
reflected in the oND
G−
�ΔΔ values (in kJmol-1): -1.6 (for 59), -1.7 (60) and -0.8 (61). The
corresponding oNP
G−
ΔΔ values (kJmol-1) are as follows: -1.6 (59), -1.7 (60) and -0.8 (61).
6.1.8 Quantitation of the anomeric effect in C-nucleosides
We have estimated the enthalpy of the stereoelectronic interactions operating in C-
nucleosides 56 - 62 by subtracting ∆H° of abasic sugar 14 from their ∆H° values (∆∆H°30). The
results of these subtractions are compiled in Table 8. The perusal of ∆∆H°30 values shows that as
purine C-nucleosides 56 - 58 become protonated, the contribution of stereoelectronic interactions to
the overall effect of the C1'-aglycone increases, whereas the effect of deprotonation is insignificant.
For pyrimidine C-nucleosides 60 - 62 in the N state, the steric effect is overriden by the anomeric
effect, whereas in the D state it is reverse. Finally, in protonated Ψ-isoC (59), nO4' →σ∗C1'-C5/9(sp2)
interactions predominate, but as the base becomes neutral and deprotonated, their magnitude is
steadily reduced and they become overriden by the counteracting anomeric effect.
6.1.9 Comparison of pD-induced flexibility in C- and N-nucleosides
The extent of the stabilization of the N-type (or S-type) pseudorotamers in 56 - 61 upon
protonation (deprotonation) is dictated by the electronic character of the nucleobase: oNP
G−
�ΔΔ values
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
113
are of the same order of magnitude in purine C-nucleosides 56 - 58 and in the N-counterparts β-D-A
(50) and β-D-G (51). However, oND
G−
ΔΔ is negligible for 56 and 57 whereas it is -1.3 kJmol-1 in β-D-
G (51). At the other end of the flexibility scale, oNP
G−
ΔΔ for Ψ-isoC (59) is ≈ 5 times larger than for
β-D-C (52) and oND
G−
ΔΔ for Ψ-U (60) is ≈ 8.5 times larger than for β-D-U (54).
6.1.10 Correlation of the effect of the aglycone and its electronic nature
Formycin A (57) and 9-deaza-A (58) differ only in the nature of the fused five-membered
ring. The more efficient deactivating pyrrazolo ring in the former compared with pyrrolo ring in the
later is evident from the lower pKa value of the nucleobase in 57 with respect to 58, and results in
stronger nO4' →σ∗C1'-C5(sp2) stereoelectronic interactions in 57 than in 58 (i.e. less negative ∆G° and
∆H°, Table 2). Although in the N state, the anomeric effect in formycin B (56) and formycin A (57)
has the same strength (compare their oN
GΔ and oN
HΔ values), in the P state the stabilization of the
N3H+ charge is more efficient through delocalization in the amidine moiety in the later than in the
pyrimidone ring in the former, therefore the fused pyrrazolo ring will be more electron-deficient in
56 than in 57 and conversely nO4' →σ∗C1'-C5(sp2) stereoelectronic interactions are stronger in 56 than
in 57. Whereas oN
GΔ and oN
HΔ of formycin B and A are the same, they drive the sugar conformation
in Ψ-U (60) more to the N than in Ψ-isoC (59), showing that replacing the amidine function for an
amide function in the six-membered ring of 60 compared with 59 has much more effect on the sugar
conformation than when this change is made in the remote fused pyrimidone ring of 56 compared
with pyrimidine in 57, which are both not directly attached at C1'. The comparison of ∆G° and ∆H°
of the N � S equilibrium in Ψ-isoC (59) and Ψ-U (60) shows that they stabilize more S-type
conformations in in the former than in the latter, owing to more electron-rich pyrimidine ring in 59
than in 60.
Interestingly, the latest study from our laboratory40 on the preferred conformation of the
constituent pentofuranose sugar in benzene, pyridine and pyrimidine C-nucleosides in aqueous
solution has allowed to establish a clear qualitative correlation between the electronic character of
the C-aglycone and the magnitude of the O4'-C1'-C(aglycone) anomeric effect (Table 12): As the
electron-deficient character of the C-aglycone increases (i.e. in the benzene derivatives, in the order:
anilino (119) < benzyl (117) < α-naphthyl (118)), the nO4'
→σ∗C1'-C(aglycone,sp2) stereoelectronic
interactions are strengthened and counteract more and more efficiently the steric effect, as indicated
by the more positive ∆H° for the drive of the two-state N � S equilibrium toward N-type
conformation.
In the pyridine derivatives, it was shown that the effect of a substituent in the para position
(with respect to the sugar moiety) upon the electronic character of the pyridine ring is negligible,
since negative ∆H° values of the N � S equilibrium in compounds 120 and 122 stabilize S-type
pseudorotamers to the same extent as C1'-benzyl in C-nucleoside 117, owing to relatively weak nO4'
→σ∗C1'-C(aglycone,sp2) orbital mixing. In contrast, it was also found that the inductive (-I) effect of the
ortho substituent promotes a strong O4'-C1'-C(aglycone) anomeric effect, which cancels the
counteracting steric effect, as evident from negligible ∆H° values for compounds 121 and 123.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
114
6.2 Quantitation of anomeric effect in N-nucleosides using C-nucleoside as reference
The drive of the N � S equilibrium in β-D-rNs 50 - 55 toward S-type conformations is the
result of the interplay of six steric and stereoelectronic forces: (i) The effect of 5'CH2OH, (ii) The
stereoelectronic component to the overall effect of the nucleobase (i.e. nO4'
→σ∗
C1'-N9
stereoelectronic
interactions), (iii) The counteracting steric effect of the nucleobase, (iv) the gauche effect of [HO3'-
C3'-C4'-O4'], (v) the gauche effect of [O2'-C2'-C1'-O4'] and (vi) the gauche effect of [O2'-C2'-C1'-
N9] (Section 4). Therefore, the anomeric effect of adenin-9-yl in β-D-A (50) or guanin-9-yl in β-D-
G (51) can be estimated using Eq 13:
AE of adenin-9-yl in β-D-A (50) or guanin-9-yl in β-D-G (51) =
∆H°(β-D-A or β-D-G) - [∆H°(14) + ∆H°GE[O2'-C2'-C1'-N9] + (∆H°ref. C-nucl. - ∆H°(14))] .....Eq 13
In Eq 13, ∆H°GE[O2'-C2'-C1'-N9] represents the strength of the [O2'-C2'-C1'-N9] gauche effect in β-D-A
(50) or β-D-G (51). The subtraction of ∆H°(14) from the experimental ∆H° value of 50 or 51 [∆H°(β-
D-A or β-D-G)] gives an estimate for the resultant of the effect of the nucleobase (stereoelectronic +
steric) and the ∆H°GE[O2'-C2'-C1'-N9] gauche effect in 50 or 51.
Our preliminary conformational study25 suggested that among purine C-nucleosides 56 - 58,
the nucleobase in formycin B (56) and A (57) prefers more pseudoequatorial orientations in S-type
conformations than 9-deaza-adenin-9-yl in 9-deaza-A (58), which was experimentally evidenced by
the larger negative ∆H° values for 56 and 57 in comparison with 58 (Table 11). Therefore, we have
initially quantitated25 the anomeric effect of adenin-9-yl in β-D-A (50) and of guanin-9-yl in β-D-G
(51) in their N state using the average (-7.4 kJmol-1, Table 11) of the ∆H° values of the N � S
equilibrium in formycin B (56) and A (57) as ∆H°ref. C-nucl. in Eq 13. In that work, ∆H°GE[O2'-C2'-C1'-
N9] (-6.3 kJmol-1) was derived from a regression analysis similar to regression (A), based on a total
set of 30 compounds, including β-D-dAMP (64), β-D-AMP (74), β-D-dAMPEt (69), β-D-dGMP
(65), β-D-GMP (75), β-D-dGMPEt (80), β-D-dCMP (66), β-D-CMP (76), β-D-dCMPEt (71), β-D-
dUMP (68), β-D-UMP (78), β-D-TMP (67) and β-D-TMPEt (72). Eq 13 was therefore rewritten as
Eq 13a:
AE of adenin-9-yl in β-D-A (50) or guanin-9-yl in β-D-G (51) in the N state =
∆H°(β-D-A or β-D-G) - [0.4 - 6.3 -7.8] .....Eq 13a
Using -4.6 kJmol-1 and -3.2 kJmol-1 as estimates for ∆H°(β-D-A or β-D-G) of β-D-A (50) and
β-D-G (51) (these values are in agreement with ∆H° values in the N state in Table 2, within the
error of the estimates), we found that the strength of the AE of adenin-9-yl in β-D-A (50) is slightly
weaker (9.1 kJmol-1) than that of guanin-9-yl in β-D-G (51) (10.5 kJmol-1).
In Eq 13a, we considered that formycin A and B in their N state constitute the best reference
points for the quantitation of the steric effect of an N-nucleobase. However, as discussed in Section
6.1, we have recently experimentally evidenced32 that the extent of the stabilization of S-type
conformations in 56 - 58 is dictated by the protonation state of the constituent nucleobase, therefore
the most pseudoequatorially oriented nucleobase is 9-deaza-adenin-9-yl in the N state of 9-deaza-A
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
115
(58) ( oN
HΔ = -14.2 kJmol-1, in the pD range from 8.8 to 12.0, Table 2), not those of formycin A and
B in the N state. Taking this fact into account, we have recently revised the estimates from Eq 13a
using Eq 13b: AE of adenin-9-yl in β-D-A (50) or guanin-9-yl in β-D-G (51) at a certain pD:
pD = ∆H°pD(β-D-A or β-D-G) - [∆H°(14) + ∆H°GEpD[O2'-C2'-C1'-N9] + ( oN
HΔ (58) - ∆H°(14))] .... Eq 13b
O
OH
OH OH
O
OH
OH OH
O
OH
OH OH
NH2
O
OH
OH OH
N
F
O
OH
OH OH
N
Br
HN
O
O
OH
OH OH
NH
OO
OH
OH OH
117 118 119
121
120
122 123
1234
56
1
2345
61
23
456
1
3
6
24
5
123
45
6
OH
OH OH
124
NH
H N
N
O
O
Table 12: Dependence of the Thermodynamicsa of the Two State N � S equilibrium upon the
Electronic Naturea of the C-aglycone.
Benzene-
derivativesb
Pyridine-
derivativesb
Pyrimidine-
derivativesb
(117) (118) (119) (120) (121) (122) (123) (59) (60)
ΔH° -5.4
(0.8)
0.4
(0.6)
-7.3
(1.0)
-4.4
(0.6)
0.4
(0.5)
-5.6
(0.9)
-0.1
(0.5)
-1.8
(0.3)
0.6
(0.2)
ΔS° -8.7
(2.7)
2.7
(1.7)
-13.1
(1.0)
-4.4
(2.0)
5.2
(1.7)
-7.4
(2.0)
0.7
(1.7)
-2.0
(1.1)
4.0
(1.1)
-ΤΔS° 2.6
(0.8)
-0.8
(0.5)
3.9
(0.8)
1.3
(0.6)
-1.5
(0.5)
2.2
(0.6)
-0.2
(0.5)
0.4
(0.3)
-1.2
(0.3)
∆G298 -2.8
(0.5)
-0.4
(0.5)
-3.4
(0.4)
-3.1
(0.4)
-1.1
(0.1)
-3.4
(0.5)
-0.3
(0.2)
-1.4
(0.2)
-0.6
(0.1)
%S298 76 53 80 78 60 80 53 63 57
Actual
AE c -5.8 0.0 -7.7 -4.8 0.0 -6.0 -0.5 -2.2 0.3
a The ΔH°, -ΤΔS° (at 298 K) and ΔG298 are in kJ/mol. The standard deviations (σ) are in parentheses. For 59 (pD =
7.0) and 60 (pD = 6.9), data are taken from ref.32. b In this work, the steric contribution of the substituents in ΔH˚
could not be dissected because of the unavailability of the corresponding saturated system. c The actual anomeric effect
(AE) was obtained by a simple subtraction (i.e. ∆∆H°30
) of the ΔH° of the N � S pseudorotational drive of 1-deoxy-β-
D-ribopentofuranose (14) (ΔH˚ = 0.4 kJ/mol)20 from the ΔH° of a specific C-nucleoside.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
116
(A)
pD
0 2 4 6 8 10 12
ΔΔH1
1
-1.0
-0.5
0.0
0.5
1.0
(B)
pD
0 2 4 6 8 10 12
ΔΔH11
-4
-2
0
2
4
(C)
pD
0 2 4 6 8 10 12
ΔΔH10
-10
-9
-8
-7
(D)
pD
0 2 4 6 8 10 12
ΔΔH1
0
-25
-20
-15
-10
-5
(E)
pD
0 2 4 6 8 10 12
ΔΔH10
+ ΔΔH11
-10
-9
-8
-7
(F)
pD
0 2 4 6 8 10 12
ΔΔH1
0 + ΔΔH1
1
-20
-15
-10
-5
(G)
pD
0 2 4 6 8 10 12
AE (A)
16
18
20
22
24
(H)
pD
0 2 4 6 8 10 12
AE (G)
10
15
20
25
30
35
40
(I)
pD
0 2 4 6 8 10 12
AE (dA)
14
15
16
17
18
19
(J)
pD
0 2 4 6 8 10 12
AE (dG)
12
14
16
18
20
22
Figure 19. Estimates (kJmol-1) for gauche and anomeric effects driving the sugar conformation in β-D-dA (37), β-D-
A (50), β-D-dG (41) and β-D-G (51) from pairwise comparisons (Fig 13, Table 6, Section 6.2). The pD-dependent
strength of the 2'-OH effect (i.e. [HO2'-C2'-C1'-N9] + [HO2'-C2'-C1'-O4'] gauche effects) in β-D-A [Panel (A)] and β-
D-G [Panel (B)] was estimated from ∆∆H°11. The pD-dependent strength of the [HO3'-C3'-C4'-O4'] gauche effect in β-
D-dA [Panel (C)] and β-D-dG [Panel (D)] is reflected in ∆∆H°10 values. The strength of the pD-dependent [O2'-C2'-
C1'-N9] gauche effect in β-D-A [Panel (E)] and β-D-G [Panel (F)] was calculated by adding the respective ∆∆H°10 and
∆∆H°11 at each pD. The modulation of the strength of the anomeric effect of adenin-9-yl in β-D-A and β-D-dA and of
guanin-9-yl in β-D-G and β-D-dG by the pD of the aqueous solution is shown in Panels (G), (I), (H) and (J)
respectively.
Eq 13b allows to calculate the AE in 50 and 51 as a function of pD. ∆H°pD(β-D-A or β-D-G)
represents the enthalpy of the N � S equilibrium of 50 or 51 at a certain pD in the range 0.6 - 12.0.
∆H°GEpD[O2'-C2'-C1'-N9] represents the magnitude of the [O2'-C2'-C1'-N9] gauche effect in 50 or 51
at a certain pD, which has been calculated using a procedure consisting of four steps: (i) We have
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
117
first determined the strength of the overall 2'-OH effect (i.e. [O2'-C2'-C1'-O4'] + [O2'-C2'-C1'-N9]
gauche effects) in 50 and 51 by subtracting from their experimental ∆H°
values at each pD in the 0.6
- 12.0 range those of their 2'-deoxy counterparts β-D-dA (37) and β-D-dG (40), respectively
(∆∆H°11 in Fig 13 and Table 6, Fig 19 for the plot of the change of ∆∆H°11 as a function of pD). (ii)
We have subsequently estimated the change in the strength of the [O3'-C3'-C4'-O4'] gauche effect in
β-D-dA (37) and β-D-dG (40) by subtracting from their experimental ∆H° values those from their
2',3'-dideoxy counterparts β-D-ddA (30) and β-D-ddG (31) (∆∆H°10 in Fig 13, Table 6 and Fig 19
for the plot at each pD). (iii) An estimate for the [O2'-C2'-C1'-O4'] gauche effect in 50 and 51 has
been subsequently obtained from pD-dependent -∆∆H°10 values (Table 6, Fig 13 and Fig 19)
assuming that it has an equal magnitude (and opposite sign) to the [O4'-C4'-C3'-O3'] gauche effect
in β-D-dNs. (iv) By adding the pD-dependent values for ∆∆H°10 and ∆∆H°11, the actual magnitude
of the [O2'-C2'-C1'-N9] gauche effect has been quantitated (Fig 19).
On the basis of these estimates, Eq 13b has been used to give the magnitude of the
stereoelectronic O4'-C1'-N9 anomeric effect of adenin-9-yl in 50 and of guanin-9-yl in 51 over the
whole pD range from 0.6 to 12.0 (Fig 19). Thus, the AE varies from 23.4 to 17.7 kJmol-1 from pD
1.2 to 7.0 for adenosine (50) and changes from 37.5 to 15.6 kJmol-1 from pD 0.6 to 11.6 for
guanosine (51).
In β-D-dA (37) and β-D-dG (41), the drive of the N � S equilibrium toward S-type
conformations is the result of the interplay of only four stereoelectronic forces: (i) The effect of
5'CH2OH, (ii) The stereoelectronic component to the overall effect of the nucleobase (i.e. nO4'
→σ∗
C1'-N9 stereoelectronic interactions), (iii) The counteracting steric effect of the nucleobase, (iv)
the gauche effect of [HO3'-C3'-C4'-O4'] fragment. In order to quantitate the stereoelectronic AE of
adenin-9-yl in 37 and of guanin-9-yl in 41, (ii), it is therefore necessary to subtract from the ∆H°
values of their two-state N �S equilibrium the contributions from (i), (iii) and (iv), according to Eq
14:
AE of 9-adeninyl in β-D-dA (37) or 9-guaninyl in β-D-dG (41) at a certain pD = ∆H°pD(β-D-dA or β-
D-dG) - [∆H°(13) + ( oN
HΔ (58) - ∆H°(14))]..... Eq 14
In Eq 14, ∆H°pD(β-D-dA or β-D-dG) represents the experimental ∆H° value of the N � S
equilibrium in β-D-dA (37) and β-D-dG (41) at the pD of interest, whereas ∆H°(13) and ∆H°(14)
denote the experimental ∆H° value for abasic sugars 13 and 14, respectively. The subtraction of
∆H° value of 13 from ∆H°pD(β-D-dA or β-D-dG) (referred as by ∆∆H°3 in Table 6) gives an estimate
for the pD-dependent overall effect of the nucleobase in 37 or 41. During this quantitation, the same
reference point (9-deazaadenin-9-yl in 9-deaza-A (58) in the N state) has been used to estimate the
steric effect of the nucleobase in 37 and 41. The AE of adenin-9-yl in β-D-dA (37) is weakened
from 18.0 to 14.8 kJmol-1 in going from pD 0.9 to pD 7.0, whereas the AE of guanin-9-yl in β-D-
dG (41) is reduced from 20.7 kJmol-1 to 13.8 kJmol-1 from pD 0.9 to pD 11.6.
An alternative strategy based on the cancellation of the stereoelectronic interactions between
the cyclopentane ring and the nucleobase in a carbocylic nucleoside to estimate the anomeric
effect of the nucleobase in N-nucleosides
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
118
In a recent work543, an alternative strategy has been developed to estimate the strength of the
stereoelectronic anomeric effect in purine nucleosides. The thermodynamics of the two-state N �S
equilibrium (suggested by ab initio calculations) in the pyrazolo[4,3-c]pyridine-carbaribo-C-
nucleoside (124) were estimated using our methodology. It was found that the cyclopentane ring in
124 adopts preferentially S-type conformation (∆H° = -11.6 kJmol-1) as does anionic 9-
deazadenosine (∆H° = -14.2 kJmol-1)32. The steric effect (∆H°steric = -12.0 kJmol-1) of the C1'-
pyrazolo[4,3-c]pyridine aglycone in 124 was estimated by subtracting ∆H° of 14 from that of 124.
The C-nucleobase in 124 was assumed to be isosteric to adenin-9-yl in β-D-ddA, β-D-dA and β-D-
rA and to guanin-9-yl in β-D-ddG, β-D-dG and β-D-rG. Subtraction of ∆H° of 12 and of ∆H°steric
from ∆H° of β-D-ddA and β-D-ddG yielded estimates for the anomeric effect (16.4 kJmol-1) of
adenin-9-yl and guanin-9-yl in these ddNs. The anomeric effect of adenin-9-yl in β-D-dA (14.7
kJmol-1) and of guanin-9-yl in β-D-dG (16.4 kJmol-1) was subsequently estimated by subtracting
∆H° of 124 and the strength of the [HO3'-C3'-C4'-O4'] gauche effect from ∆H° of each dN. The
anomeric effect of adenin-9-yl in β-D-A (13.9 kJmol-1) and of guanin-9-yl in β-D-G (15.3 kJmol-1)
was calculated by subtracting ∆H° of 124 and the strength of the [HO2'-C2'-C1'-N9] gauche effect
from ∆H° of each rN. These estimates are of the same order of magnitude as our own33, except for
adenosine, where a difference of 3.8 kJmol-1 has been found. This means that the 9-deazaadenin-9-
yl aglycone in the D state indeed takes up a maximal pseudoequatorial orientation in 57 and hence
serves a correct reference point for the estimation of the steric of purine N-nucleobase.
7. The interdependency of the sugar and phosphate conformation
The sugar moiety, the nucleobase and the phosphate backbone constitute the three structural
elements making ploynucleotide chains. We have discussed in Sections 2 - 6 how in nucleosides the
change of the electronic character (for instance upon the change of the pD of the aqueous solution or
complexation with metal ions) and steric bulk of the substituents at C1' - C5' allows to engineer
certain conformational preferences of the constituent pentofuranose sugar, as the result of tunable
stereoelectronic gauche and anomeric effects. The phosphate backbone is not itself an isolated
structural element, as suggested by the qualitative correlation between preferred orientation around
the ε torsion and sugar puckering modes in mononucleotides and RNA trimers. We report the results
of our investigations23,28 that have adressed the following fundamental questions: Does the sugar
conformation dictate the phosphate backbone torsions? Is there any preferred phosphate torsion that
steers the sugar conformation in a certain manner? Are there any correlated interdependencies of
endocyclic sugar torsions with the preferred phosphate torsions? In order to further understand the
forces that govern the stabilization of the tertiary structure of oligo- and polynucleotides, we have
examined and dissected the nature of fundamental intranucleotidyl interactions that contribute to the
drive of the sugar-phosphate backbone in DNA and RNA using simple model systems, i.e.
nucleoside 3'-ethylphosphates, both in the 2'-deoxy (69 - 73) and ribo series (79 - 83) in which the
effect of internucleotidyl base-base stacking on the drive of the sugar-phosphate backbone is
completely eliminated.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
119
7.1 Methods to assess the preferred conformation across the phosphate backbone
The dependence of 3JC2'P3', 3JC4'P3' and 3JH3'P3' coupling constants on the torsion angle ε
(Fig. 20) is described by the following Karplus-type equations223,435,544-546 (Eqs 15a - 15c): 3J
H3'P3' = 15.3 cos2(ε + 120)- 6.2 cos(ε + 120) + 1.5 ..... Eq 15a
3JC4'P3'
= 9.1 cos2(ε)- 1.9 cos(ε) + 0.8 ..... Eq15b
3JC2'P3'
= 9.1 cos2(ε - 120)- 1.9 cos(ε - 120) + 0.8 ..... Eq 15c
In Eqs 15a - 15c, a perfect trigonal symmetry for the position of C2', C4' and H3' with respect
to the C3'-O3' bond is assumed. In our interpretation23,28 of the experimentally measured 3JH3'P3'
,
3JC4'P3'
and 3JC2'P3'
coupling constants for nucleosides 3'-monophosphates 64 - 73 and their 3'-
ethylphosphates 74 - 83, we have considered a two-state εt � ε- equilibrium for the following
reasons: (i) ε+ rotamers are not found in crystal structures of nucleotides, suggesting that this state is
forbidden435 on account of steric and electrostatic repulsions547-550 between O4' and phosphoryl
oxygen. (ii) Additionally, for ε+ rotamers, one expects 3JH3'P3' ≈ 23 Hz223,435, but our experimental
values23,28 are ≈ 7 - 8 Hz, therefore it is likely that the population of ε+ rotamers is negligible.
A rough estimate for the population of ε- rotamers can also be obtained551 from 4JH2'P3' using
Eq 16: x(ε-) = 4JH2'P (obs) / 4JH2'P (-) .... Eq 16, Where 4JH2'P (obs) represents the
experimental value of 4JH2'P coupling constant for the compound of interest and 4JH2'P (-)
designates the corresponding coupling constant in a pure gauche- (ε-) conformation (i.e. 2.3 Hz551).
The experimental time-averaged 3JH5'P5', 3JH5"P5' and 3JC4'P5' coupling constants can be used
to estimate the population of the staggered rotamers around the C5'-O5' bond via simple linear
relationships435 (Eqs 17a - 17b):
% βt = 100 x [25.5 - (3JH5'P5' - 3JH5"P5')] / 20.5 .... Eq 17a
% βt = 100 x (3JC4'P5' - 0.73 / 10.27) .... Eq 17b
When P5'-O5'-C5'-C4'-H4' lie in the same plane and form a W-type conformation551-553 [i.e.
(βt, γ+) rotamer], the 3JH4'P5' coupling constant is approximately 3 Hz and can be used as a marker
to recognize this conformer, since in all other cases the coupling constant will be much reduced. 3JH4'H5' and 3JH4'H5" are translated into the populations of the staggered γ+ [x(γ+)], γ- [x(γ-)]
and γt [x(γt)] rotamers554 using Eqs 18a - 18d which have been parametrized on the basis of
crystallographic data: 3JH4'H5' = 2.4 x(γ+) + 10.6 x(γ-) + 2.6 x(γt) .... Eq 18a
3JH4'H5" = 1.3 x(γ+) + 3.8 x(γ-) + 10.5 x(γt) .... Eq 18b
x(γ+) + x(γ-) + x(γt) = 1 .... Eq 18c
% γ+ = 100 x [13.3 - (3JH4'H5' + 3JH4'H5")] / 9.7 .... Eq 18d
Using Eq. 19, the value of the torsion angle δ [C5'-C4'-C3'-O3'] can be derived from that of ν3
for each of the N- and S-type pseudorotamers, since ν3 is itself related to the respective PN, Ψm(N),
PS and Ψm(S) values via Eq. 6a. Finally, owing to the low natural abundance of 17O (0.04 %), the
preferred α and ζ conformations in nucleotides cannot be determined from spin-spin coupling
constants: δ ≈ ν3 + 120° .... Eq 19
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
120
7.2 No correlation of sugar and phosphate conformation in 2'-dN-3'-ethylphosphates
We have elucidated33 the conformational preferences of the pentofuranose moieties in β-D-
dNMPs (64 - 68) and their 3'-ethylphosphates (69 - 73) using the methodology described in Section
3. The thermodynamics of the N � S equilibrium in 64 - 73 in the neutral solution are compiled in
Table 3. In order to assess the preferred orientation around the ε torsion in β-D-dNMPEts 69 - 73 in
the 278 K - 358 K range, we have developed the program epsilon, which allows to translate
experimental 3JC2'P3', 3JC4'P3' and 3JH3'P3' coupling constants into the geometries and relative
populations of the rotamers engaged in the two-state εt � ε- equilibrium. These vicinal coupling
constants did not significantly change (≤ 0.5 Hz) as the temperature was raised from 278 K to 358
K, as the result of nearly temperature-independent mole fractions of the εt and ε- rotamers (≈ 1:1
ratio). Similarly, 3JCH2P3' and 3JCH3P3' were nearly the same over the whole temperature range,
suggesting that the population of βt rotamers (≈ 50%) is not affected by the temperature. The
comparative analysis of ∆H°, -T∆S° and ∆G° values of 64 - 73 in Table 3 leads to the main
following conclusions: (i) In all cases, ∆H° prevails over the counteracting -T∆S° term and it is the
main factor responsible for the overall (∆G°) stabilization of S-type pseudorotamers at 298 K. (ii) S-
type pseudorotamers in β-D-dNMPs 64 - 68 are slightly more preferred than the β-D-dNs 37 and 41
- 44 counterparts as shown by more negative ∆G° values for the former in comparison with the
latter. This can be attributed to the slightly more electronegative 3'-OPO3H- group in the latter with
respect to 3'-OH in the former. The additional stabilization of S-type pseudorotamers in 64 - 68
through [-1/-2HO3PO3'-C3'-C4'-O4'] gauche effect is shown by the ∆∆H°22 values in Table 7 (Fig
13) which are slightly nucleobase-dependent and within the range from -2.5 to -1.2 kJmol-1. (iii)
Similarly, S-type pseudorotamers are slightly more favoured in β-D-dNMPEts 69 - 73 than in β-D-
dNs 37 and 41 - 44. This stabilization throught the stronger [-EtO3PO3'-C3'-C4'-O4'] gauche effect
is shown in the ∆∆H°23 values in Table 7 (Fig 13). (iv) For each nucleobase, the [-1/-2HO3PO3'-C3'-
C4'-O4'] gauche effect in 64 - 68 has been estimated by subtracting from their experimental ∆H°
values those of the corresponding β-D-ddNs 30, 31 and 33 - 35, yielding: ∆∆H°20 (kJmol-1) = -8.9
(adenin-9-yl), -7.5 (guanin-9-yl) = -9.8 (cytosin-1-yl), -8.0 (thymin-1-yl), -8.4 (uracil-1-yl). Thus
∆∆H°20 varies by ≈ ± 1kJmol-1 from the average value depending upon the nature of the
nucleobase. (v) Using the same strategy, we have quantitated the [-EtO3PO3'-C3'-C4'-O4'] gauche
effect in 69 - 73: ∆∆H°21 (kJmol-1) = -9.0 (adenin-9-yl), -8.2 (guanin-9-yl) = -10.2 (cytosin-1-yl), -
8.4 (thymin-1-yl), -8.4 (uracil-1-yl). Thus the comparison of ∆∆H°20 with ∆∆H°21 values shows that
the stereoelectronic gauche effects within [-1/-2HO3PO3'-C3'-C4'-O4'] and [-EtO3PO3'-C3'-C4'-O4']
fragments operate with nearly the same magnitude in the 2'-deoxynucleotides 64 - 68 and 69 - 73.
This is also proven by the negligible ∆∆H°24 values in Table 7.
We have also estimated the magnitudes of the gauche and anomeric effects operating in all β-
D/L-nucleosides (i.e. the dataset used for regression analysis (B))
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
121
as well as β-D-dNMPs 64 - 68 and β-D-dNMPEts 69 - 73 (Table 5, Section 4.1) from regression
(E), which has been performed using a dataset consisting of 40 experimental ∆H° values assuming
identical strengths for
the [-1/-2HO3PO3'-C3'-
C4'-O4'] and [-
1EtO3PO3'-C3'-C4'-O4']
gauche effects in 64 -
68 and 69 - 73. The
estimates resulting from
regression analysis (E),
its correlation
coefficient and the
standard error on the
estimates are virtually
the same as of
regression (B). The
average [3'-phosphate-
C3'-C4'-O4'] gauche
effect is about ≈ -8.3
kJmol-1, i.e. -2.0 kJmol-
1 stronger than the
[HO3'-C3'-C4'-O4']
counterpart.
Thus this work
shows that: (i) there is no straightforward correlation between preferred orientation across C3'-O3'
bond and most stable sugar puckering mode in the 2'-deoxynucleotides, (ii) the additional
stabilization of S-type pseudorotamers in 64 - 68 and 69 - 73 with respect to β-D-dNs 37 and 41 -
44 is nearly the same owing to equally efficient 3'-gauche effects.
7.3 Interaction of 2'-OH with vicinal 3'-phosphate in ribonucleotides
The thermodynamics28 of the N � S equilibrium in β-D-rNMPs 74 - 78 and β-D-rNMPEts
79 - 83 based upon pseudorotational analyses of temperature-dependent 3JHH coupling constants are
presented in Table 3. The pairwise comparison of ∆H° values of 74 - 83 with those of nucleosides
and other 2'-deoxynucleotides leads to following conclusions: (i) For all ribonucleotides (except β-
D-CMPEt (81), for which they are of equal strength and cancel each other), ∆H° contribution to the
free-energy ∆G° of the N � S equilibrium stabilizes S-type conformations over the counteracting
entropy, which prefers N-type sugars. (ii) The N � S equilibrium in β-D-rNMPs 74 - 78 is driven
more toward S-type conformations than in the parent β-D-rNs 50 - 55, as evident from negative
∆∆H˚25 values. ∆∆H˚25 (kJmol-1) is in the range -1.2 < ∆∆H˚25 < -0.5 for purines and -2.2 <
∆∆H˚25 < -0.8 for pyrimidines. This can be attributed to the stronger [-1/-2HO3PO3'-C3'-C4'-O4']
O
O
O
P
O
-O O
OH
O
PO
-O
C4'
PO3
H5"H5'
C4'
O3P
H5"H5'
C4'
PO3
H5" H5
'
C2' C4'
H3'
PO3
C2' C4'
H3'
O3P
PO3
C2' C4'
H3'
O5'
H5' H5"
O4' C3'
H4'
O5'
H5' H5"
C3' H4'
O4'
O5'
H5' H5"
H4' O4'
C3'
ON
OH
O
O
ON
OH
NH
NN
N
O
NH2
β
χ
α
εδ
γ
A
C
ζ
5'
3'
(n-1)
(n+1)
(n)
-------------------------------------------------
------------------------------------------------------
εt
ε-
Torsion angle
γt
ε+
εt
= 180º
γ−
ε-
= -60º
γ (O5'-C5'-C4'-C3')
β (P5'-O5'-C5'-C4')
βt
= 180º
γt
= 180º
ε+
= +60º
γ-
= -60º
γ+
γ+
= +60º
Torsion angle ε (C4'-C3'-O3'-P)
β-
= -60º β+
= +60º
Torsion angle
Figure 20. The description of the phosphate backbone conformation in
nucleosides and nucleotides by the α [(n-1)O5'-P5'-O5'-C5'], β [P5'-O5'-C5'-C4'], γ
[O5'-C5'-C4'-C3'], δ [C5'-C4'-C3'-O3'], ε [C4'-C3'-O3'-P3'] and ζ [C3'-O3'-P3'-
O5'(n+1)] torsion angles.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
122
gauche effect in β-D-rNMPs compared with the [HO3'-C3'-C4'-O4'] gauche effect in the former.
(iii) In β-D-rNMPEts 79 - 83, S-type conformations are also more stable than in β-D-rNs 50 - 55 by
∆∆H˚26 = -2.5 kJmol-1 for purines and ∆∆H˚26 ≈ -3.7 kJmol-1 for pyrimidines, owing to the
stronger [-EtO3PO3'-C3'-C4'-O4'] gauche effect in 79 - 83 in comparison with the [HO3'-C3'-C4'-
O4'] gauche effect in 50 - 55. (iv) The [-EtO3PO3'-C3'-C4'-O4'] gauche effect in β-D-rNMPEts is
clearly stronger (by -2.8 to -1.3 kJmol-1) than the [-1/-2HO3PO3'-C3'-C4'-O4'] gauche effect in β-D-
rNMPs, as shown by ∆∆H˚28 values, whereas in the 2'-deoxy counterparts, they are the same (as
shown by nearly identical ∆∆H˚20 and ∆∆H˚21). (v) The overall 2'-OH effect drives the sugar
conformation toward S and N-type geometries in purine and pyrimidine β-D-rNMPEts, respectively,
as shown by the comparison of their conformational preferences with the β-D-dNMPEts
counterparts (i.e. ∆∆H˚27). In contrast, 2'-OH drives the conformation of the sugar moiety in β-D-
rNMPs to more N in comparison with β-D-dNMPs, as shown by negligible or positive ∆∆H˚29
values.
(vi) As the temperature is increased from 278 K to 358 K, the population of S-type conformers
decreases and the population of εt rotamers in β-D-rNMPEts 79 - 83 increases in a coopeative
manner, as shown by the concomittant change in 3JC2'P3', 3JC4'P3' and 3JH3'P3' coupling constants.
Table 13.The thermodynamics of the two-state N �S and εt � ε- pseudorotational equilibria in β-
D-rNMPEts 79 - 83 showing the identical values (within experimental error) for ΔG298 of N � S
pseudorotational and εt � ε- equilibria (see Fig 21 for the correlation plot).
Estimation of the drive of εt � ε- conformational equilibria derived from
temperature-dependent 3JHP and 3JCP
ΔHºε
a ΔSºε
a -TΔS˚ b ΔG
298
%ε-278 c
%ε-358 c Δ%ε- d
β-D-AMPEt (79) -6.6 (0.9) -14 (3) 4.2 -2.4 76 63 -13
β-D-GMPEt (80) -5.8 (0.9) -12 (3) 3.6 -2.2 74 62 -12
β-D-CMPEt (81) -2.8 (0.9) -9 (4) 2.7 -0.1 53 46 -7
β-D-rTMPEt (82) -3.8 (0.8) -8 (3) 2.4 -1.4 66 58 -8
β-D-UMPEt (83) -3.3 (0.7) -7 (2) 2.1 -1.2 64 57 -7
a ΔH˚ (kJ mol-1) and ΔS˚ (J mol-1 K-1) are the average values (standard deviations are given in brackets) and
were calculated from individual van't Hoff plots using populations of N and S pseudorotamers from several
individual PSEUROT analyses. ΔHºε and ΔSº
ε were calculated from 15 van't Hoff plots using populations of εt
and ε- rotamers. The signs of thermodynamic parameters are arbitrarily chosen in such a way that the positive
values indicate the drive of N � S and εt �ε- equilibria to N and εt, whereas the negative signs describe the
drive to S and ε-, respectively. b -TΔS˚ (kJ mol-1) term is given at 298 K. c The population of the S and ε-
conformers were calculated using the relation: %S (T) or %ε- (T) = 100 * [exp (- ΔGT/ RT)] / [exp (- ΔGT/ RT)
+1]. d Δ%ε- = %ε-358 - %ε-278.
(vii) The interdependency of the conformation of the sugar and phosphate moieties in β-D-
rNMPEts 79 - 83 is evidenced by the fact that the free-energies (∆G298) of the two-state N � S and
εt � ε- equilibria are the same (Table 2 and Table 13), within the experimental error of our
measurements and calculations (±0.5 kJ/mol). This is further evidenced by a simple correlation plot
of the temperature-dependent populations of ε- rotamers as a function of the temperature-dependent
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
123
population of the S sugar pseudorotamers (Fig 21, Panels (A1) - (E1)) and of ∆G° of the εt � ε-
equilibrium versus ∆G° of the N � S equilibrium for all β-D-rNMPEts (Fig 21, Panels (A2) - (E2)),
showing their straightforward correlation with the correlation coefficients between 0.79 and 0.98 (in
pyrimidine nucleotides, the relatively smaller values for the correlation coefficients are the result of
the limited change in ∆G° or xε- or xS as a function of temperature in comparison with the purines
counterparts).
(viii) The unique interaction of 2'-OH with the lonepair of the vicinal heteroatom (e.g. O3')
is able to act as a molecular switch between (N,εt) � (S,ε-) conformational equilibria in 79 - 83. An
alternative H-bond directly between 2'-OH and the 3'-phosphate oxygen is ruled out because they
are too far away from each other, even on rotation around α and ζ torsions. This 2'-OH....O3'
interaction stabilizes the S and ε- conformers ("On-Off" switch) in a cooperative manner over N and
εt, and it is experimentally evidenced by the fact that the difference in the chemical shifts of the
CH2 protons of 3'-ethylphosphate in 79 - 83 is steadily reduced as the temperature is increased (until
β-D-AMPEt (79)
% S
64 68 72 76 80 84
% ε
−
56
60
64
68
72
76
80
ΔGo
(N/S)
-3.2 -2.8 -2.4 -2.0 -1.6 -1.2
ΔG
o(ε
t/ε
−)
-3.5
-3.0
-2.5
-2.0
(A1) (A
2)
β-D-AMPEt (79)
% S
60 64 68 72 76
% ε
−
60
64
68
72
76 (B1)
β-D-GMPEt (80)
ΔGo
(N/S)
-2.8 -2.4 -2.0 -1.6 -1.2
ΔG
o(ε
t/ε
−)
-3.0
-2.5
-2.0
-1.5 (B2)
β-D-GMPEt (80)
β-D-CMPEt (81)
% S
46 48 50
% ε
−
44
48
52
56
ΔGo
(N/S)
-0.5 0.0 0.5
ΔG
o(ε
t/ε
−)
-1.0
-0.5
0.0
0.5(C
1) (C
2)
β-D-CMPEt (81)
% S
52 56 60 64
% ε
−
56
60
64
68
72
(D1)
β-D-rTMPEt (82)
ΔGo
(N/S)
-1.0 -0.5
ΔG
o(ε
t/ε
−)
-2.0
-1.5
-1.0(D
2)
β-D-rTMPEt (82)
β-D-UMPEt (83)
% S
54 57 60
% ε
−
54
57
60
63
66
ΔGo
(N/S)
-0.8 -0.6 -0.4
ΔG
o(ε
t/ε
−)
-1.5
-1.2
-0.9
-0.6(E1) (E
2)
β-D-UMPEt (83)
Figure 21. The correlation plots of the temperature-dependent population of the ε- phosphate backbone
rotamers versus the population of S-type conformers (at 298 K) [Panels (A1) - (E1)] and of ∆G°(εt/ε-) of
εt � ε- equilibrium versus ∆G°
(N/S) of N �S equilibrium [Panels (A2) - (E2)] in β-D-rNMPEts 79 - 83.
All plots show straight lines (s = slope, i = intercept, R = Pearson's correlation coefficient): For β-D-
AMPEt, s = 1.02, i = -5.78, R = 0.98 (A1) and s = 1.01, i = 0.62, R = 0.95 (A2), For β-D-GMPEt, s =
0.95, i = 4.10, R = 0.98 (B1) and s = 0.94, i = -0.25, R = 0.95 (B2), For β-D-CMPEt, s = 1.44, i = -19.9,
R = 0.79 (C1) and s =1.39, i = -0.25, R = 0.80 (C2), For β-D-rTMPEt, s = 1.37, i = -16.73, R = 0.93
(D1) and s = 1.30, i = -0.30, R = 0.86 (D2), For β-D-UMPEt, s = 1.96, i = -49.5, R = 0.94 (E1) and s =
1.94, i = 0.12, R = 0.85 (E2).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
124
both protons become isochronous) owing to the increase in the population of N,εt-type conformers
in which no H-bond between 2'-OH and the vicinal phosphate occurs. The N � S equilibrium is
unbiased (ΔG298 = 0.1 kJ mol-1) for β-D-CMPEt (81) because of the strongly opposing anomeric
effect. (ix) The in-line attack of 2'-OH to the vicinal phosphate in the RNA self-cleavage
reaction10,555, giving a 2',3'-cyclic phosphate via a transient trigonal bipyramidal phosphorane as
intermediate, requires a S,ε- conformational state246,556,557. This cleavage is more facile whith
cytosin-1-yl as the nucleobase at the cleavage site than any other nucleobase558, presumably owing
to a smaller activation energy barrier for N- to S-type sugar interconversion. We, on the other hand,
have found that ∆G° of the (N,εt) � (S,ε-) conformational equilibrium in pyrimidine β-D-rNMPEts
is much reduced (≈ 0 kJmol-1 for β-D-CMPEt) than for the purine counterparts. (x)
Owing to the internucleotidyl stacking, the preferred conformational state in RNA-RNA559 or RNA-
DNA560 duplex is (N,εt). However, when these stacking interactions are absent such as in the single
stranded hairpin loop, (S,ε-) and gg C4'-C5' conformers are favoured, as in our model systems β-D-
rNMPEts 79 - 83.
8. Application of stereoelectronic effects in oligonucleotides
8.1 Design of antisense oligonucleotides via the gauche engineering
What is the structure of DNA-RNA hybrid?
The target for the antisense strand is the RNA. The mechanism of antisense effects of the
antisense NA involve either RNase H mediated cleavage of the RNA strand in the hybrid duplex
duplex561, or the physical blocking of the translation machinary562. Much is understood about the
RNase H promoted RNA excission from the hybrid DNA-RNA duplex than the physical blocking.
In order to be able to optimally design the antisense strand using the RNase H promoted RNA
excission of the DNA-RNA hybrid, it is important to understand first the structure of the DNA-
RNA hybrid both in the solution and in the solid state. The summary of various studies performed
in many labs can be divided into three categories, and they are as follows: (i) The RNA and DNA
strands of the chimeric duplex are similar to the corresponding A-RNA and the B-DNA563,564. (ii)
The RNA strand of the duplex is similar to A-RNA and the DNA strand is somewhat intermediate
between A- and B-forms of DNA560,563,564. (iii) In one crystalline duplex, the conformation of both
ribose and deoxyribose sugars were found to be in C3'-endo (i.e. N-type) conformation, but the
sugar residues in the DNA strand underwent a conformational transition to C2'-endo (i.e. S-type) in
solution565. The occurrence of all the above variations in the DNA-RNA hybrid clearly shows that
the RNA sugars are invariably C3'-endo, but the DNA sugars are flexible and can indeed take up
various conformations (O4'-endo to C1'-exo to C2'-endo, 72° < 180°).
Configuration-dependent drive of the sugar conformation
The effects of covalently linked substituents at C2', C3' and C5' upon the conformation of
the sugar moieties of the resulting nucleosides and the overall stability in modified oligonucleotides
have been extensively reviewed46,49. It has been demonstrated first by Remin et al 405 that inversion
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
125
of configuration at C2' from ribo to ara-cytidine induces further stabilization of the S-type
conformation (40% S for cytidine and 50% S for ara-C at 24°C in D2O at pH 7). This preference for
S-type conformation is dramatically increased as the hydroxyl groups are ionized at pH ≥ 13
because of intramolecular 5'-OH...O2'- H-bond (45%S for cytidine and 90%S for ara-C). Consistent
with this study, it has been found226 that the electronegative substituents such as 2'-F at C2' on the
α-face (i.e. 2'F-2'-deoxyribo analog) enhances the preference of the sugar moiety for N-type
pseudorotamers, whereas C2'-F in the β-face (i.e. 2'F-2'-deoxyara-analog) enhances the preference
of the sugar moiety for S-type pseudorotamers, thereby showing that the gauche effect is indeed
configuration-dependent.
The engineered conformational preference of nucleosides preorganizes the oligo conformation.
It has been shown64 that two South thymidines in a 12mer duplex greatly stabilized the
double helix, whereas two North thymidines destabilized it by inducing an A-B junction in the
middle of the duplex. The introduction of the North nucleosides into oligo-DNA will induce an
RNA-type geometry, and consequently improves its binding capability with the complementary
target RNA (but not the RNase H cleavage property; see below!). Consequently, the incorporation
of 2'F-2'-deoxyribonucleosides into a DNA strand458 stabilizes the DNA:RNA heteroduplex by
≈2°C per modification. The modified antisense oligodeoxynucleotide adopts an A-form
conformation, as a result of the drive of the sugar conformation toward N-type pseudorotamers by
the [F2'-C2'-C1'-O4'] gauche effect458,459.
The antisense NA-RNA hybrid should mimic the DNA-RNA hybrid conformation for good RNase H
cleavage
Consistent with the studies of Remin et al 405 that the ara-nucleosides have intrinsic
preference for the South-type conformation compared to ribo counterpart, fluoroarabinonucleic
acids (2'F-ANA)-RNA and arabinonucleic acid(ANA)-RNA duplexes showed a close CD
resemblance566 to the native DNA-RNA counterpart (suggesting that 2'F-ANA and ANA have
preorganized structures similar to DNA, and their hybrids with RNA have very similar A-like
helical conformations as the native DNA-RNA hybrid), but their melting studies showed that the
former had higher thermodynamic stability compared to latter. Consequently, the 2'F-ANA-RNA
duplex was cleaved by RNase H much faster than the ANA-RNA duplex. The susceptibility of the
2'F-ANA-RNA hybrids to RNase H was found to be very similar to that observed for native DNA-
RNA and DNA-thioate-RNA hybrids. This study showed that lower thermal stability of the ANA-
RNA hybrids translated into poorer digestion by RNase H although they had similar A-type helix,
whereas 2'F-ANA-RNA, native DNA-RNA and DNA-thioate-RNA have also similar conformation
but higher stabilities than ANA-RNA duplex, therefore they all showed faster cleavage. In
contradistinction, 2'F-2'-deoxyribonucleosides has a predominant N-type conformation as the native
ribonucleosides, and therefore 2'F-RNA has a preorganized RNA-type structure, which forms A-
type hybrid with a target RNA as the native RNA-RNA duplex, and, as expected, none of them
serves as a substrate to RNase H.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
126
The antisense NA-RNA heteroduplex is more efficiently stabilized by 2'F-2'-deoxyribo- than
by 2'-O-methyl ribonucleosides567, which is consistent with the stronger gauche effect induced by
the former, owing to the higher electronegativity of fluorine compared with hydroxy458. The
presence of North 2'-fluoro-2'-deoxyguanosine in a modified hexadeoxynucleotide enables the
transition between A-DNA conformation at low ionic strength and Z-DNA at high salt
concentrations460. 2'-O-methyl-ribonucleosides567 when incorporated into an oligodeoxynucleotide
improves its binding affinity with the complementary RNA strand, but incorporation of 2'-SMe-46,49
or 2'-amino-2'-deoxyriboucleosides497 into the antisense strand results in the destabilization of the
resulting antisense-NA-RNA duplex because of the propensity of these modified nucleosides to
adopt S-type conformations, owing to the poorer electronegativity of 2'-SMe or 2'-NH2 group
comapred to 2'-OH.
Requirements for the design of the antisense strand for RNase H cleavage of the hybrid
The above discussion point to two distinct requirements for the design of the antisense
strand complementary to RNA target: (i) The antisense strand should mimic the preorganized
conformation of a native DNA strand in order to form a hybrid duplex with a target RNA with a
structure closely similar to the native DNA-RNA hybrid, and (ii) the affinity of this preorganized
antisense strand to the native RNA should be at least as high as the native DNA-RNA strand in
order to be a good substrate to RNase H. In contradistinction to these requirements, the necessity
for the antisense strand to mimic the preorganized conformation of a native DNA strand is often
overlooked. A major consideration is still given49 in the design of antisense strand is the increase of
the thermal stability of the antisense NA-RNA hybrid. Despite this, the great majority of the sugar
and backbone-modified nucleoside building blocks (for example, -S-CH2-O-CH2- or -CH2-NCH3-
O-CH2- backbone) have a preferred N-type conformation, thereby their antisense oligo is expected
to have a preferred RNA-type, not the DNA-type, preorganized structure. The increase of melting
point observed with the hybrid of these antisense NA with RNA target is therefore presumably
owing to RNA-RNA duplex type structure, which are not as good substrates for RNAse H as DNA-
RNA duplex. Regarding the relative stability of various types of duplexes, it is noteworthy that it
increases in the following order: DNA-DNA < DNA-RNA < RNA-RNA.
These studies are consistent with our own studies568 with ten different 5'-fluorophore
tethered DNA-RNA hybrids as substrates for RNAse H: We showed that among all these ten
different 5'-tethered chromophores, 5'-phenazine and 5'-phenanthridene tethered DNA-RNA hybrids
had the shortest half-life of digestion by RNase H because these antisense oligos showed higher
thermal stability (5-10°C higher than the native counterpart) and least deviation of the global
structure from the native heteroduplex.
Requirements for 3'-modification for engineering antisense DNA strand (the Gauche engineering)
It has been shown by our lab in 199426 that the distinct conformational preferences observed
for the pentofuranosyl moieties in various 3'-substituted (X)-2',3'-dideoxynucleosides [X = H, NH2,
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
127
OH, OMe, NO2, OPO3H-1/-2 and F] are closely related to the strength of the 3'-gauche effect exerted
by the 3'-substituent, X. This work demonstrated that the magnitude of the enthalpy of the 3'-gauche
effect increases with the increase of the electronegativity of the 3'-substituent. Thus, as the
electronegativity of the 3'-substituent decreases [F > OPO3H-1/-2 > NO2 > OMe > OH > NH2], the
preference for the North-type pseudorotamer increases [F < OPO3H-1/-2 < NO2 < Ome < OH <
NH2]. Thus 3'-amino-2',3'-dideoxynucleosides have been shown to possess a preferred N-type
conformation26,569 which indeed is conserved in the sugar puckers (66 of the 72 nucleotides per
asymmetric unit adopting C3'-endo conformation) in the N3' → P5' phosphoramidate 12mer DNA
duplex as evident by the A-type duplex conformation62. Moreover, the CD-spectra of the N3' P5'
phosphoramidate 12mer DNA duplex are consistent with adoption of an A-RNA conformation570.
8.2 Fused nucleosides to engineer preorganized DNA structures
We have seen above that the engineered conformational preference of the flexible
pentofuranose ring in nucleosides is capable of preorganizing the conformation of the
oligonucleotide duplex. This can be simply achieved by engineering the stereoelectronics of the
nucleosides, namely the gauche and the anomeric effects as described in the earlier chapters.
Clearly, such stereoelectronic engineering gives us the possibility to steer a specific conformation
depending upon the conformation of the target strand. The challenge in the stereoelectronic
engineering is to direct the sugar conformation to a specific low energy minima and stabilize it by
overcoming the intrinsic temptation to fall into other energy minima since the interconversions
between various pseudorotamers take place through a low energy barrier. Alternatively, one can
introduce appropriate rigidity to the pentofuranose or to the cyclopentane system in a nucleoside
analogue and trap it into one specific conformation. Introduction of these fused systems in oligos
have produced preorganization and appropriate rigidity of the single strand, which has been found to
play a critical role for the formation of stable duplexes. Several such systems have been so far
designed, and the readers are directed to a recent review article49.
8.3 Enzyme recognition by fused carbocyclic nucleosides of fixed conformation
Recently, a conformationally constrained 2E (N)-methano-carbocyclic AZT mimic [(N)-
4',6'-methano-carba-3'-azidothymidine-5'-triphosphate] was shown to be equipotent to AZT-5'-
triphosphate in inhibiting the target enzyme, reverse transcriptase, whereas the antipodal 3E
carbocyclic analogue [(S)-1',6'-methano-carba-3'-azidothymidine-5'-triphosphate] was inactive571.
When carbocyclic nucleosides are incorporated into a modified oligodeoxynucleotide59,572,
they are able to mimic the conformation of the neighbouring nucleotides owing to the relatively
more flexible nature of the cyclopentane ring (see below). In contrast, the incorporation of the
conformationally constrained 2E carbathymidine analogue [(N)-4',6'-methano carbathymidine] was
able to force a conformational change in the oligonucleotide that resulted in an enhanced stability of
DNA-RNA heteroduplexes573,574.
8.4 The conformational transmission in the self-cleaving Lariat-RNA
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
128
Small synthetic lariat RNAs 129a, 130a and 131a have been found557 to undergo site
specific self-cleavage to give an acyclic branched-RNA with 2',3'-cyclic phosphate and a
5'-hydroxyl termini as in 129b, 130b and 131b, which are reminiscent of the products formed in
some catalytic RNAs. Noteworthy is the fact that in these self-cleavage reactions of 129a, 130a and
131a one specific phosphodiester bond between G and C residues underwent intramolecular
transesterification reaction amongst all other phosphates moieties in the molecule (akin to the
splicing of pre-mRNA to give functional mRNA), suggesting that the self-cleaving transesterified
G C phosphodiester has enormously different chemical reactivity in comparison with all other
phosphates in the molecule. These lariat-RNAs 125, 126, 129a, 130a and 131a are much smaller
than the natural catalytic RNAs such as the hammerhead ribozyme (k = ~1 min-1 at 37 °C), and
their rate of the self-cleavage is also much slower (k = 0.25 x 10-4 min-1 for lariat hexamer 129a,
and 0.16 x 10-3 min-1 for lariat heptamer 130a at 22 °C). We have shown that the trinucleotidyl
loop in the tetrameric and pentameric lariat-RNAs245 126 is completely stable whereas the
tetranucleotidyl or pentanucleotidyl loop in the hexameric 129a or heptameric 130a lariat-RNA
RNA246,556,575 does indeed have the required local and global conformation promoting the self-
cleavage. It has been also shown that simple 2' 5' or 3' 5'-linked cyclic RNAs, 127 and 128,
respectively, are completely stable and their structures are considerably different from the self-
cleaving lariat-RNAs such as 129a or 130a. In our search to explore the optimal structural
requirement for the self-cleavage reaction of RNA, we showed that the unique 3'-ethylphosphate
function in 131a (in which the branch-point adenosine has a 2' 5'-linked tetranucleotidyl loop and
a 3'-ethylphosphate moiety, mimicking the 3'-tail of the lariat-hexamer 129a) is the key structural
feature that orchestrates its self-cleavage reaction (k = 0.15 x 10-4 min-1 at 19 °C) compared to the
stable 2' 5'-linked cyclic RNA 127. A comparative study of the temperature dependence of the N
� S equilibrium for the lariat-tetramer 131a and the 2' 5'-linked cyclic tetramer 127 shows that the
A1 residue in the former is in 92% S-type conformation at 20 °C, whereas it is only in 55% S in the
latter. This displacement of the N � S pseudorotational equilibrium toward the S-type geometry is
due to the enhanced gauche effect of the 3'-OPO3Et- group at the branch-point adenosine in 131a
compared to 3'-OH group in 127. This 3'-OPO3Et- group promoted stabilization of the S geometry
at the branch-point by ΔH ≈ 4 kcal.mol-1 in 131a is the conformational driving force promoting its
unique self-cleavage reaction. The comparison of ΔH° and ΔS° of the N � S pseudorotational
equilibria in 131a and 127 (see Table 14) clearly shows the remarkable effect of the
3'-ethylphosphate group in 131a in being able to dictate the conformational changes from the sugar
moiety of the branch-point adenosine to the entire molecule (conformational transmission). Thus
the S conformation in A1, U2 and C6 sugar moieties is clearly thermodynamically more stabilized
while it is considerably destabilized in G3 owing to the 3'-ethylphosphate group in 131a compared
to 127. This is a clear experimental evidence of conformational transmission through a crankshaft
motion originating from the gauche effeect exerted by the 3'- ethylphosphate moiety at the A1 sugar.
It is interesting to note that the magnitude of enthalpy and entropy for the North to South transition
of the A1 sugar in 127 is comparable to the enthalpy and entropy of transition between the A- and
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
129
B-form of the lariat hexamer556,557. This self-cleaving tetrameric lariat-RNA 131a to 131b is the
smallest lariat-RNA molecule hitherto known to undergo the self-cleavage reaction, and hence it is
the simplest model of the active cleavage site of the natural self-cleaving catalytic RNA.
8.5 The conformational transmission by dihydrouridine in RNA
It has been recently shown576 that S-type conformations of the pentofuranose sugar in
dihydrouridine (D) 3'-monophosphate are more stable (by ≈ 1.5 kcalmol-1) than in the
corresponding β-D-UMP (78). This effect is amplified for the D residue (5.3 kcalmol-1) in the
ApDpA trimer (compared to ApUpA), and is also further transmitted to the neighbouring 5'-
terminal A (3.6 kcalmol-1). This observation is consistent with the fact that the C5-C6 saturation of
the uracil-1-yl aglycone to the dihydrouracil-1-yl moiety reduces the electron-withdrawing character
of the former, and hence weakens the n →σ* interactions of O4'-C1'-N1 moiety. This is consistent
P
O
O
N
O
N
O-
O
O
O
NO
P- O
O
OO
O
O
OPO
P- O
O
HO
ON
O-
OH
N
N
HN
HO
HO
O
O
O
NH2
N
N
NH2
H2NN
H
N N
O
PO
H2N
O
N
O
N
O-
O
O
O
N
O
P- O
O
O
OO
O
P- O
O
HO
O
N
HN
HO
O
O
NN
NH2
H2N
N
H
N
N
O
N
O
P
O
O-
O
NH
HO
O
O
OH
N
O
HO
O P
O O-
NO
P O
O
N
O-
O
O
N
O
O
P- O
OH
O
O
OP
- O
O
HO
O
N
N
HN
HO
O
O
ON
N
NH2
H2NN
H
N
N
O
O
P- O
O
O
HO
O
N
N
H2N
O
P O
HO
N
O-
O
O
O
NO
P- O
O
O
O
O
P
- O
O
HO
O
N
N
HN
HO
O
O
ON
N
NH2
H2NN
H
N
N
O
O
P- O
O
O
HO
O
N
N
H2N
O
P
128
C6
G3
A1
U2
U2
C6
G3
A1
127
126125
A1
G3
U2
C4
G3
A1
U2
C5
U4
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
130
P- O
O
PO
NH2
O
O
NH2
O
N
N
O-
O
O
O
N
NO
P OO
PO
O-
P
- O
O
N
O-
O
O
O
O
NO
P- O
O
O
O
O
O
NH
HO
P
- O
O
HO
O
N
O
N
HN
HO
O
O
O
O O
HO
N
N
N
NH2
H2NN
H
N
N
O
O
P- O
O
O
HO
O
N
N
H2N
O
OHO
O
P
O
O-
O
NO
O
O
PO
PO
- OOH
O
N
O-
N
O
HN
HO
ONH
O
O
HO
O
N
NNH2
H2NN
H
N
N
O
O
P
- O
O
O
O
HO
ON
N
H2N
O
O
HO
N
OHO
O
P
O
O-
NO
O
HOO
N
HN
O
O
P- O
O
PO
NH2
O
O
NH2
O
N
N
O-
O
O
O
N
NO
P OO
PO
O-
P
- O
O
N
O-
O
O
O
O
NO
P- O
O
O
O
O
O
NH
HO
P
- O
O
HO
ON
O
N
HN
O
O
O
O
O O
HO
N
N
N
NH2
H2NN
H
N
N
O
O
P- O
OOH
HO
O
N
N
H2N
O
OHO
O
P
O
O-
O
NO
O
OPO
PO
- OOH
O
N
O-
N
O
HN
O
ONH
O
O
HO
O
N
NNH2
H2NN
H
N
N
O
O
P
- O
O
O
O
HO
ON
N
H2N
O
O
HO
N
OHO
O
P
O
O-
NO
OHHO
O
N
HN
O
O
P O
O
N
O-
O
O
O
NO
P- O
O
O
O
O
P
- O
O
HO
O
N
N
HN
HO
O
O
ON
N
NH2
H2NN
H
N
N
O
O
P- O
O
O
HO
O
N
N
H2N
O
P OCH3CH2O
PO
-
O
N
O-
O
O
O
NO
P- O
O
O
O
O
P
- O
O
OH
O
N
N
HN
O
O
O
ON
N
NH2
H2NN
H
N
N
O
O
P- O
OOH
HO
O
N
N
H2N
O
OCH3CH2O
PO
-
O
130a
131a
130b
131b
129b
at 22o C
k = 0.25 x 10-4
m-1
k = 0.16 x 10-3
m-1
k = 0.15 x 10-4
m-1
Cleavage
site
Cleavage
site
Cleavage
site
Some Synthetic Self-cleaving RNAs
at 19o C
at 22o C
129a
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
131
Table 14: The thermodynamicsa of for the N S equilibrium in 127, 129a & 131a
Compound Residue ΔH°
(kcal.mol-1)
ΔS°
(cal.K-1.mol-1)
-TΔS°b
(kcal.mol-1)
ΔG°
(kcal.mol-1)
Lariat- A1 -5.8 ± 1.3 -15.9 ± 3.8 4.7 ± 1.1 -1.1 ± 2.4
3'-Et-Phos U2 -7.1 ± 2.2 -21.3 ± 6.4 6.3 ± 1.9 -0.8 ± 4.1
RNA G3 12.3 ± 2.8 36.2 ± 8.2 -10.6 ± 2.4 1.7 ± 5.2
131a C6 1.2 ± 0.8 3.2 ± 2.5 -0.9 ± 0.7 0.3 ± 1.6
3'-OH A1 -1.8 ± 0.5 -5.5 ± 1.7 1.6 ± 0.5 -0.2 ± 1.1
RNA U2 -1.9 ± 1.2 -2.6 ± 3.8 0.8 ± 1.1 -1.2 ± 2.3
127 G3 6.1 ± 1.1 17.0 ± 3.2 -5.0 ± 0.9 1.2 ± 2.0
C6 6.6 ± 1.1 18.7 ± 3.4 -5.5 ± 1.0 1.1 ± 2.1
Hexameric A1 -1.2 ± 1.1 -0.7 ± 3.6 0.2 ± 1.0 -1.0 ± 2.2
Lariat- U2 -2.2 ± 1.0 -4.3 ± 3.0 1.3 ± 0.9 -1.0 ± 1.9
RNA G3 7.3 ± 1.2 22.1 ± 3.9 -6.5 ± 1.1 0.8 ± 2.4
129a C6 5.0 ± 1.1 16.2 ± 3.4 -4.8 ± 1.0 0.3 ± 2.1
a 90% confidence. b The -TΔS term is computed for 293 K.
with our observation that the strength of the anomeric effect of an aglycone increases as it becomes
more electron-withdrawing, and decreases as it becomes electron-rich. The transmission of the
stabilization of S-type conformations from the D nucleotide moiety to the 5'-terminal A moiety in
the ApDpA trimer most probably takes place through a crankshaft motion across the sugar-
phosphate backbone. As it has been stated above, the earlier example of this type of stereoelectronic
transmission through the crankshaft motion across the sugar-phosphate backbone has been
documented for the tetranucleotidyl lariat-RNA loop, in which placement of a 3'-ethylphosphate
moiety at the branch point not only stabilized the S-type conformation of the constituent sugar of the
nucleotide at the branch-point but also affected the conformation of all four other loop
nucleotides246,556,556.
Similar modifications of the electronic nature of five most common aglycones, say by
methylation or by any hypermodification (as found in tRNAs) or by complexation with any ligand
such as metal ion, drug or polypeptide will profoundly affect the local structure of the DNA and
RNA, which may itself locally alter the hydration pattern, H-bonding or electrostatics, and any one
of these, in turn, may serve as a recognition signal for a specific interaction and function.
8.6 Preferential recognition of 3'-anthraniloyladenosine by Elongation Factor Tu
The functional role of transfer RNA (tRNA) as adapter in translating the genetic code is well
recognized. The protein biosynthesis occurs via stepwise manner – (i) Initiation, (ii) Chain
elongation and (iii) Termination. The initiation step requires the assembly of a ribosome-mRNA-
tRNA ternary complex. During this process, amino acids are linked to tRNA by aminoacyl-tRNA
synthetases to form aminoacyl-transfer RNA (aa-tRNA). The binding of the aa-tRNA to the
ribosome is catalysed by the elongation factor Tu (EF-Tu), a GTP binding protein. The EF-Tu, in its
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
132
active form (EF-Tu.GTP), can specifically recognize the aa-tRNAs from the non-aminoacylated
tRNAs.
This precise recognition process depends on the complimentary structural features of the aa-
tRNA scaffold. From this viewpoint, the development of tRNA-mimics is an interesting approach to
understand the functional aspects of this biological process. Small molecule can mimic only a part
of a large tRNA. There are three such molecules, having tRNA mimicry, found in literatures – (i)
antibiotic puromycin [(a) Welch, M.; Majerfeld, I.; Yarus, M. Biocehmistry 1997, 36, 6614. (b)
Monroe, R. E.; Marcker, K. A. J. Mol. Biol. 1967, 25, 347], which corresponds to the 3'-end of a
tyrosinylated tRNA and found to be a powerful inhibitor in protein biosynthesis. However, it failed
to interact with EF-Tu or aaRS (ii) micro-tRNA, CCAOH trinucleotides, which has been shown to
mimic tRNA for aaRS recognition and can even be aminoacylated [Jovine, L.; Djordjevic, S.;
Rhodes, D. J. Mol. Biol. 2000, 301, 401] and (iii) 3'-O-anthraniloyladenosine (used in our present
study, see below). The differential recognition of 3'-O-anthraniloyladenosine (having 2'-endo sugar
conformation and puromycin (having 3'-endo sugar conformation) towards EF-Tu has enormous
biological significance.
Anthranilic acid charged yeast tRNAPhe or E. coli tRNAVal are able to form a stable
complex with EF-Tu*GTP, hence the
2'- and 3'-O-anthraniloyladenosines
and their 5'-phosphate counterparts
have been conceived to be the smallest
units that are capable to mimic
aminoacyl-tRNA577-582. Since the 3'-
O-anthraniloyladenosine (134) also
binds more efficiently to EF-Tu*GTP
complex compared to its 2'-isomer
(132), we delineated the stereoelectronic features that dictate the conformation of 3'-O-
anthraniloyladenosine (134) and its 5'-phosphate (135) vis-a-vis their 2'-counterparts (132 and 133)
as well as address how their structures and thermodynamic stabilizations are different from
adenosine and 5'-AMP. It has been found43 that the electron-withdrawing anthraniloyl group exerts
gauche effects of variable strengths depending upon whether it is at 2' or at 3' because of either
participation or absence of the [O2'-C2'-C1'-N9] gauche effect, thereby steering the pseudorotation
of the constituent sugar moiety either to the N-type or S-type conformation.
The 3'-O-anthraniloyladenosine 5'-phosphate has a relatively more stabilized S-type conformation
(ΔG° = -4.6 kJ/mol) than 3'-O-anthraniloyladenosine itself (ΔG° = -3.9 kJ/mol), whereas the ΔG°
for 2'-O-anthraniloyladenosine and its 5'-monophosphate are respectively -0.9 and -1.8 kJ/mol,
suggesting that the 3'-gauche effect of 3'-O-anthraniloyl group is stronger than that of 2'-O-
anthraniloyl in the drive of the sugar conformation. Since the EF-Tu can specifically recognize the
aminoacylated-tRNA from the non-charged tRNA, we have assesssed the free-energy (ΔG°) for this
recognition switch to be at least ≈ -2.9 kJ/mol by comparison of ΔG° of N � S pseudorotational
O
ROH2CO
ROH2C
O
HO
NN
N
NN
N
N
NH2
NH2
N
O
OH
O
H2N
O
NH2
132: R = H
133: R = OPO3H-1/-2
134: R = H
135: R = OPO3H-1/-2
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
133
equilibrium 3'-O-anthraniloyladenosine 5'-phosphate and 5'-AMP. The 3'-O-anthraniloyladenosine
and its 5'-phosphate are much more flexible than the isomeric 2'-counterparts as evident from the
temperature-dependent coupling constant analysis. The relative rate of the transacylation reaction of
2'(3')-O-anthraniloyladenosine and its 5'-phosphate is cooperatively dictated by the two-state N �S
pseudorotational equilibrium of the sugar, which in turn is controlled by a balance of the
stereoelectronic 3'- and the 2'-gauche effects as well as by the pseudoaxial preference of the 3'-O or
2'-O-anthraniloyl group. The reason for the larger stabilization of the 2'-endo conformer for 3'-O-
anthraniloyladenosine and its 5'-phosphate lies in the fact that the C3'-O3' bond takes up an optimal
gauche orientation with respect to C4'-O4' bond dictating the pseudoaxial orientation of 3'-
anthraniloyl residue, which can be achieved only in the S-type sugar conformation with adenin-9-yl
and the 2'-OH groups in the pseudoequatorial geometry, compared to the preferred C3'-endo sugar
with pseudoaxial aglycone and 2'-OH found in 3'-terminal adenosine moiety in the helical 3'-CCA
end of uncharged tRNA.
8.7 Conformational changes in nucleotides induced by interaction with metal ions
(i) Effect of Zn2+ or Hg2+ binding to the nucleobase on the sugar conformation in β-D-dG (41)
It has been shown by us30 that protonation at N7 in β-D-dG (41) shifts the N �S
pseudorotational equilibrium toward more N by ≈ 15% with respect to the neutral state, which was
attributed to the strengthening of the nO4' →σ∗
C1'-N9 anomeric effect at acidic pD owing to the
reduced electron-density in the imidazole ring compared with the neutral pD (Section 4.8). In
contrast, it has been found395 that binding of Zn2+ with N7 of the constituent guanin-9-yl in β-D-2'-
dG has a negligible effect on the population of the preferred S-type pseudorotamers. Similarly, upon
addition of 0.2 equivalent Hg2+, only a slight shift (by ≈ 4 % unit) of the two-state N � S
pseudorotational equilibrium toward more N-type has been found. These results suggest that
binding of Zn2+ or Hg2+ with N7 do not modulate efficiently the strength of the O4'-C1'-N9
anomeric effect. This is in agreement with the
observation395 that the resulting perturbation
(with respect to the uncomplexed/neutral
state, Section 2.9.3) of the electronic character
of the imidazole moiety is much reduced than
when N7 becomes protonated: Indeed, the
15N chemical shift of N7 in β-D-2'-dG
changes much less upon its complexation with Zn2+ (∆δ15N(N7) = 3.4 ppm) than upon its
protonation401 [∆δ15N(N7) = 46 ppm]. In the same study, it was also noted that hard Mg2+ did not
form any stable complex with N7 of the guanin-9-yl nucleobase in β-D-2'-dG.
(ii) Conformational changes observed across the sugar-phosphate backbone and in the
pentofuranose sugar in 5'- and 3'-methylmonophosphates of β-D-dG as the result of the interaction
of the nucleobase/phosphate moiety with Mg2+, Zn2+ and Hg2+
As an extension of the above work, Polak and Plavec have recently shown583 that the preferential
binding of Mg2+ with the phosphate oxygen atoms and C6=O carbonyl group in 5'- and 3'-
O
O
HO
O
P
MeO
-O
NH
NN
N
O
NH2
O
OH
O
NH
NN
N
O
NH2
PO
O- OMe
136: MepdG 137: dGpMe
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
134
methylmonophosphates of β-D-dG (i.e. MepdG (136) and dGpMe (137), respectively) does not alter
the bias of the N �S pseudorotational equilibrium toward S-type conformations (At 298K, 68% S
for MepdG and 75% S for dGpMe).
On the other hand, the interactions of Zn2+ or Hg2+ with N7 resulted in the small shift of the sugar
conformation in MepdG and dGpMe toward N (by ≈ 1 - 5%), as the result of the slight
strengthening of the anomeric effect (vide supra and Section 4.8). It was also observed that the
binding of the softer Zn2+ causes a larger shift of the syn � anti equilibrium toward anti than
complexation with hard Mg2+. Whereas the preferred orientation around the C4'-C5' (γ torsion
angle, in dGpMe) and O5'-C5' (β torsion angle, in MepdG) remain virtually unchanged in the metal
ion complexed state in comparison with the free state, interaction of hard Mg2+ or softer Zn2+ or
Hg2+ resulted in a small shift of the εt �ε- equilibrium toward εt rotamers. Thus, this work has
qualitatively shown that the small change of the electronic character of the nucleobase in dGpMe
upon the complexation of N7 with a soft metal ion resuts in relatively minor strengthening of the
O4'-C1'-N9 anomeric effect pushing the sugar conformation toward slightly more N-type forms, and
this additional
preference for N-
type sugars is in
turn transmitted to
control the
conformation of the
phosphate-
backbone by further
stabilizing εt
rotamers.
(iii) The change of
the electronic
character upon
Pt2+
binding to
guanin-9-yl nucleotide is transmitted to drive the conformation of the local sugar-phosphate
backbone
The anticancer drug cisplatin interferes with replication and transcription processes to form
crosslinks as major lesions by reacting with the N7 of guanines and adenines in DNA. The dynamic
microstructure alteration as a result of cisplatin binding to N7 of guanines in DNA has been recently
assessed45 through the multinuclear temperature-, pD- and concentration-dependent NMR study of
the effect of Pt2+ complexation to 2'-deoxyguanosine 3',5'-bisethylphosphate (138) and its ribo
analogue (140). Compounds 138 and 140 serve as models mimicking the central nucleotide moiety
in a trinucleoside diphosphate in order to shed light on how the nature and strength of
intramolecular stereoelectronic effects change as a result of Pt2+ binding in complete absence of
O
O
O
P
O
-
O O
R
O
P
O
-
O
NH
NN
N
O
NH2
H2C
CH3
CH2
H3C
Pt
NH3H3N
N7N7
O
O
O
P
O
-O O
R
O
P
O
-O
H2C
CH3
CH2
H3C
β
χ
α
ε
δ
γ
ζ
ζ−1
ε−1
α+1
β+1
138: EtpdGpEt (R = H)140: EtpGpEt (R = OH)
139: cis-(NH3)2Pt(EtpdGpEt)2141: cis-(NH3)2Pt(EtpGpEt)2
142: EtpdRpEt (R = H)143: EtpRpEt (R = OH)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
135
intramolecular base-base stacking interactions. The study revealed that the N �S pseudorotational
equilibrium shifts towards more N-type conformers by Δ∆G° = 2.0 kJmol-1 and 2.6 kJmol-1,
respectively, thereby showing that free-energy of complexation is transmitted to drive the sugar
conformation. The increase in the population of N-type conformers was rationalized with the
strengthening of the anomeric effect in both 2'-deoxy and ribo nucleotides upon the formation of Pt-
N7 bond which promotes nO4' →σ∗
C1'-N9 orbital interactions due to the reduction of π-electron
density in imidazole part of guanine.
The additional stabilization of N-type conformers in Pt2+ complexes of ribonucleotide is due
to the tuning of gauche effect of [N9-C1'-C2'-O2'] fragment, which is absent in 2'-deoxyribo
counterparts. The platination of N7 favors N1 deprotonation in 139 and 141 by ΔpKa of 0.7 and 0.9
units in comparison to parent nucleotides 138 and 140, respectively. The N�S pseudorotational
equilibrium in 138 – 141 showed classical sigmoidal dependence as a function of pH with pKa
values at the inflection points.
Upon N1 deprotonation, the N �S pseudorotational equilibrium was shifted toward more S-
type conformations by Δ∆G°D-N = -0.3 kJmol-1 in 138, -0.9 kJmol-1 in 139, -0.4 kJmol-1 in 140 and
-2.5 kJmol-1 in 141 which showed that the thermodynamics of N1 deprotonation in guanin-9-yl
nucleosides is transmitted to drive N �S equilibrium towards S–type pseudorotamers. The
anomeric effect is weakened upon N1 deprotonation due to the increased π-electron density in the
imidazole part of the guanin-9-yl moieties and the population of S-type pseudorotamers increases.
The Pt2+-complex bound to N7 has been found to promote the redistribution of the π-electronic
density from anionic deprotonated N1 which results in the larger increase in the population of S-
type conformers in 139 and 141 in comparison to parent nucleotides 138 and 140. The formation of
Pt-N7 bond in bifunctional complexes 139 and 141 simultaneously causes a shift of syn �anti
equilibrium towards anti by 43 and 63 %, and the increase of the population of εt conformers by 20
and 32 % at 278K, respectively. Only minor conformational redistributions along β, γ, β+ and ε-
torsion angles have been observed which suggests weak conformational cooperativity between the
shift in the N �S pseudorotational equilibrium towards N-type conformers and the conformational
changes across phosphate torsion angles other than ε as a result of platination of guanin-9-yl. In
comparison to nucleotide phosphodiesters, apurinic 3',5'-bisethylphosphate sugars 142 and 143
showed no interaction with Pt2+ and therefore no conformational changes.
It is thus clear that any intermolecular interaction between nucleic acid and a ligand (other
than soft metal ion as described above) is expected to produce effects similar to partial electron-
deficiency as arising from the protonation of the aglycone (such as in the case of a lone-pair
donating aglycone involved in H-bonding or through-space charge transmission in stacked
complexes with intercalated drugs), or to partial electron-enrichment as arising from the
deprotonation of the aglycone (such as ionization of the aglycone by the basic sites of polypeptides
in general or a phosphate moiety in the vicinity or by a relatively hard metal ion).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
136
8.8 RNA as molecular wire
The intrinsic dynamics and architectural flexibility of nucleic acids resulting into specific function
are the result of cooperative interplay of pentofuranose, nucleobase and phosphodiester moieties.
We have shown in Sections 4 - 7 that the interplay of various stereoelectronic gauche and anomeric
effects and associated steric effects energetically drives the sugar conformation, which in turn is
dictated by the electronic nature of the aglycone and other substituents on the sugar ring.
In our latest work44a,b, we have employed guanosine 3'-ethylphosphate, 5'-methylphophate
[MepGpEt (144a)] and adenosine 3',5'-bisphosphate [MepApEt (144b)], as a model mimicking the
central nucleotide moiety in a trinucleoside diphosphate in order to shed light on how the strength of
intramolecular stereoelectronic effects is modulated by the change of protonation �deprotonation
equilibrium of the aglycone in the complete absence of any intramolecular base-base stacking (For
our work on 2'-deoxy23 and ribonucleoside28 3'-ethylphosphates at the neutral pD, see Section 7).
The work uniquely shows a complete interdependency of conformational preference of sugar and
phosphate backbone in 144a and 144b as the protonation � deprotonation equilibrium of the
aglycone changes as a function of pD. Moreover, similar thermodynamic studies (only at two
extreme pDs i.e. at the neutral and the protonated states) of cytosine 3',5'-bisphosphate [MepCpEt
(144c)] have been performed (unpublished work) in order to compare with 144a and 144b to
demonstrate the tunibility of nucleobases in 144a – c.
In oligonucleotides the protonation of nucleobase directly affects their hydrogen-bonding
capabilities and therefore induces a change in overall three-dimensional structure (Section 2.9).
O
O
H
H
OH
O
H
H OH
NH
NN
N
O
NH2
N
NN
N
NH2
R
P
H3CH2CO
O−
OO
P
H3CH2CO
O−
O
O
R
O−
P
OCH3
O
O
O−
P
H3CO
O
N
N
NH2
O
(144b) R =
(145) R = H
(144c) R =
(144a) R =
South: 3'-exo-2'-endoNorth: 3'-endo-2'-exo
0° < P < 36°
Ψm
= 38.6 ± 3° 144° < P < 190°
Ψm
= 38.6 ± 3°
Schematic representation of the dynamic two-state North (N) � South (S) pseudorotational
equilibrium [P = phase angle and Ψm = puckering amplitude] of the constituent β-D-pentofuranose
moiety in gaunosine 3',5'-bisphosphate [MepGpEt (144a)], adenosine 3',5'-bisphosphate [MepApEt
(144b)] and cytidine 3',5'-bisphosphate [MepCpEt (144c)], as well as their apurinic counterpart
[Mep(ab)pEt (145)].
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
137
In these studies we have shown that changing the electronic character of aglycone through
protonation at basic site (N7 of guanin-9-yl, N1 of adenin-9-yl and N3 of cytosin-1-yl) not only
modulates the shift of N �S pseudorotational equilibrium of its constituate sugar by strengthening
the anomeric effect but is also transmitted to the sugar-phosphate backbone to influence the
conformational characteristics of backbone in 144a – c, which mimic the model of trinucleoside (3'-
5') diphosphate. This tunable transmission which has been observed, compared to the abasic
counterpart [Mep(ab)pEt] 145, to be much stronger at the 3'-phosphate backbone compared to 5'-
end justifies RNA as "molecular wire".
(i) pD-dependent shift of N � S pseudorotational equlibrium
In this endeavour, the mole fractions of N- and S-type conformers have been plotted as the
function of temperature in the van't Hoff type analyses to obtain ΔH° and ΔS° and subsequently to
calculate the total change in free energy at 298 K ( o K298
ΔG ) of the N � S pseudorotational
equlibrium at each of the seven pDs ranging from 6.7 to 1.0 (Table 15 and 16).
As the medium becomes more acidic the anomeric effect becomes more strengthened (See
Section 4.8) by enhancing the nO4' → σ∗C1'-N
orbital interaction owing to the reduced electron
density at N9 and due to which the N �S pseudorotational equilibrium for MepGpEt (144a) and
MepApEt (144b) is gradually shifted towards N-type pseudorotamers [79% S (for 144a) and 76%
S (for 144b) in the neutral state to 55% S (for 144a) and 67% S (for 144b) in the protonated state
respectively] as reflected from change of o K298
ΔG from -3.3 kJ mol
-1 (for 144a) and –2.8 kJ mol
-1
(for 144b) at pD 6.7 to -0.1 kJ/mol and -1.7 kJ mol-1
(for 144b) at pD 1.0 [Tables 15 and 16]. The
plots of pD-dependent o K298
ΔG values of the N �S equilibrium in MepGpEt (144a) as well as
MepApEt (144b) [Panel (B) in Fig 22] have the typical sigmoidal shapes, as found for β-D- 2',3'-
dideoxy, 2'-deoxy, ribo N- or C-nucleosides and some α-D-2',3'-dideoxy and 2’-deoxy-nucleosides
(Section 3.7, Fig 11). The value of the pD at the inflection point of the plots of pD-dependent o
K298
ΔG of N �S equilibrium in 144a and 144b were found to be 2.4 ± 0.1 and 3.8 ± 0.2
respectively, which are identical to the pKa of the constituent guanin-9-yl and adenine-9-yl
nucleobases, as determined independently from the plots of pD-dependent H8 chemical shift for
144a and that of both H8 and H2 chemical shifts for 144b [Panel (A) in Fig 22] [see the legend of
Fig 22 for details, Eq. 12 and Section 3.7.4]. As a control experiment, the difference in 3JHH values
between neutral and acidic pDs at 298K was found to be negligible in apurinic phosphodiester 145,
showing that in the absence of aglycone, the N � S equilibrium is unbiased and remains unchanged
at all pDs compared to 144a and 144b.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
138
Table 15. Pseudorotational analyses and EPSILON calculations using temperature-dependent 3JHH
and 3JCPa at seven different pDs (1.0-6.7), and determination of pD-dependent thermodynamics of
the two-state N →← S and εt � ε- equilibrium in MepGpEt (144a).
Thermodynamics of N →← S
equilibrium
Thermodynamics of
εt ε- equilibrium
pD
o K298
ΔG =
-R*298*ln (xN / xS) %S
o K298
ΔG =
-R*298*ln (xεt/xε
-) %ε-
1.0b
-0.0 (0.1) -b 0.3 (0.2) -
b
1.6 -0.3 (0.2) 53 0.2 (0.2) 48
2.0 -0.7 (0.2) 58 0.0 (0.5) 50
2.4 -1.4 (0.2) 64 -1.2 (0.2) 62
2.7 -2.3 (0.2) 72 -1.7 (0.2) 66
3.0 -2.4 (0.2) 72 -1.8 (0.2) 67
6.7 -3.2 (0.2) 78 -2.0 (0.2) 69
a For the experimental procedure used to calculate the values, see Section 3 and ref. 44a. b
No calculations performed due to unavailability of 3JHH owing to spectral overlap. ΔG° (at
298 K) value have been extrapolated.
Table 16. Pseudorotational analyses and EPSILON calculations using temperature-dependent 3JHH
and 3JCPa at nine different pDs (1.0-7.9), and determination of pD-dependent thermodynamics of
the two-state N →← S and εt � ε- equilibrium in MepApEt (144b).
Thermodynamics of N →← S
equilibrium
Thermodynamics of
εt � ε- equilibrium
pD
o K298
ΔG =
-R*298*ln (xN / xS) %S
∆Go (at 298 K) =
-R*298*ln (xεt/xε
-) %ε-
1.0 -1.7 (0.2) 67 -1.5 (0.2) 65
2.0 -2.0 (0.2) 69 -1.6 (0.2) 66
2.5 -2.0 (0.2) 69 -1.6 (0.2) 66
3.0 -2.1 (0.2) 70 -1.6 (0.2) 66
3.5 -2.2 (0.2) 71 -1.7 (0.2) 66
4.0 -2.4 (0.2) 73 -1.8 (0.2) 67
4.5 -2.7 (0.2) 75 -1.9 (0.2) 68
4.8b
-2.8 (0.2) -b -2.0 (0.2) -
b
7.9 -2.8 (0.2) 76 -1.9 (0.2) 68
a For the experimental procedure used to calculate the values, see Section 3 and ref. 44b.
b
No calculations performed due to unavailability of 3JHH owing to spectral overlap. ΔG° (at
298 K) value have been extrapolated.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
139
(ii) pD-dependent shift of the εt � ε- equlibrium
The temperature dependent 3JH3'P3',
3JC4'P3', 3JC2'P3' have been used to calculate the bias of
conformational equlibrium across C3'-O3' bond (ε torsion). Since the ε+ conformer is energetically
forbidden and is not found in crystal data we have interpreted the experimental coupling constants
in terms of the two-state εt � ε- equilibrium (Section 7.1). Logarithm of ratio of the resulting mole
fractions of ε- and εt were plotted as a function of temperature in a van't Hoff type analyses to obtain
ΔH° and ΔS° and subsequently to calculate the total change in free energy at 298 K ( o K298
ΔG ) of
εt � ε- equlibrium at each of the seven pDs ranging from 6.7 to 1.0 (Table 15). 69% (for 144a) and
68% (for 144b) in neutral state to 48% (for 144a) and 65% (for 144b) in protonated state
respectively. The corresponding shift of o K298
ΔG is from -2.1 kJ mol
-1 (for 144a) and -1.9 kJ mol
-1
(for 144b) in neutral state to +0.3 kJ mol-1
(for 144a) and -1.5 kJ mol-1
(for 144b) in protonated
state [Tables 15 and 16]. Interestingly, the plots of pD-dependent o K298
ΔG values of the εt �ε-
equlibrium in MepGpEt (144a) as well as in MepApEt (144b) also give a sigmoidal curve [Panel
(C) in Fig. 22], as found for the plot of the corresponding o K298
ΔG values of the N � S equilibrium
of the constituent pentofuranose sugar. The values of pD at the inflection point (i.e. pD = 2.3 for
144a and pD = 3.7 for 144b) of the graphs shown Panel (C) in Fig 22, and is nearly identical (i.e.
within the accuracy of the measurements) to the pKa values (2.4 for 144a and 3.9 for 144b
respectively) of the guanin-9-yl nucleobase in MepGpEt. Thus, it can be concluded that the pD-
dependent reorientation of the 3'-ethylphosphate moiety across the C3'-O3' bond (reflected
in o K298
ΔG values of the εt � ε- equlibrium) is directly dictated by the pKa of the constituent C1’-
aglycone in MepGpEt (144a) and MepApEt (144b).
(A)
pD
0 1 2 3 4 5 6 7 8
δH
8.0
8.2
8.4
8.6
8.8
9.0
9.2
(B)
pD
0 1 2 3 4 5 6 7 8
ΔG
o(N/S) at 298 K
-3.6
-3.0
-2.4
-1.8
-1.2
-0.6
0.0
0.6
(C)
pD
0 1 2 3 4 5 6 7 8
ΔG
o(ε
t/ε
−) at 298 K
-2.4
-1.8
-1.2
-0.6
0.0
0.6
δH8G of MepGpEt
δH8A of MepApEt
δH2A of MepApEt
MepGpEt
MepApEt
MepGpEt
MepApEt
Figure 22. Panel (A) shows the plots of H8 chemical shift of the constituent guanine-9-yl nucleobase (δH8G in ppm) in
144a as well as H8 and H2 chemical shift of the constituent Adenin-9-yl nucleobase (δH8A and δH2A respectively in
ppm) in 144b, as a function of pD at 298K. Panels (B) and (C) show the plots of the experimental ΔG° (in kJ mol-1
) of
the N � S equilibrium of the constituent pentofuranose moiety and that of ε
t
� ε− equilibrium of the 3'-ethylphosphate
group in 144a and 144b respectively, as a function of pD at 298K. The following pKa values have been obtained: 2.4 ±
0.1 (for δH8 in 144a), 4.0 ± 0.1 and 3.9 ± 0.02 (for δH8 and δH2 respectively in 144b) in Panel (A); 2.4 ± 0.1 (for
144a) and 3.8 ± 0.2 (for 144b) in Panel (B); 2.3 ± 0.4 (for 144a) and 3.7 ± 0.2 (for 144b) in Panel (C).
These results constitute the first experimental evidence for transmission of electronic
information to steer the conformation of the 3'-phosphate as a result of change of electronic
properties of the nucleobase (protonation/deprotonation) via the tuning of the sugar conformation in
a nucleoside 3',5'-bisphosphates (such as in 144a and 144b) (see the correlation plots below under
subsection (iii)). In the case of our reference compound, i.e. apurinic phosphodiester [Mep(ab)pEt
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
140
(145)] no change in 3JHP and 3JCP coupling constant values has been observed compared to that of
purinic phophodiester 144a and 144b over the whole pD range which allows us to conclude that the
conformation across C3'-O3' remains unchanged over the whole pD range in abasic compound
(145) since there is no nucleobase to be protonated in 145.
(iii) The cooperative shift of the (N,εt) �(S,ε-) equilibrium as the result of protonation of guanin-9-
yl in MepGpEt (144a) and MepApEt (144b) are evidenced by [∆G°(N/S) vs ∆G°(ε t/ε-)] or [∆G°(N/S)
or (ε t/ε-) vs δH8] correlation plots
The correlation plots of pD-dependent o K298
ΔG values of the N�S equilibrium of the
pentofuranose sugar as a function of pD-dependent o K298
ΔG values of the εt � ε- equilibrium of the
3'-phosphate moieties in 144a and 144b give a straight line with a high Pearson correlation
coefficient [R = 0.98 for both 144a and 144b, Panel (A) in Fig 23]. This means that as the
constituent guanin-9-yl in 144a and adenine-9-yl in 144b are gradually protonated in the acidic
medium, the modulation of the strength of the anomeric effect shifts the N� S equilibrium toward
N, which in turn dynamically shifts the εt �ε- equilibrium toward more εt in comparison with the
neutral pD.
(A)
ΔGo
(εt
/ε−
) at 298 K
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5
ΔG
o(N/S) at 298 K
-4
-3
-2
-1
0
1
2
(C)
δH
8.0 8.2 8.4 8.6 8.8 9.0
ΔG
o(ε
t/ε
−) at 298 K
-3.6
-2.4
-1.2
0.0
1.2
(B)
δH
8.0 8.2 8.4 8.6 8.8 9.0
ΔG
o(N/S) at 298 K
-3.6
-2.4
-1.2
0.0
1.2 δH8G of MepGpEt
δH8A of MepApEt
δH2A of MepApEt
MepGpEt
MepApEt
δH8G of MepGpEt
δH8A of MepApEt
δH2A of MepApEt
Figure 23. Panel (A) shows the plot of o
K298
ΔG (in kJ mol-1
) of the N � S pseudorotational equilibrium [ΔG
o(N/S) at 298
K] as a function of the o
K298
ΔG (in kJ mol
-1) of ε
t
� ε− equilibrium [ΔG
o(εt/ε−) at 298 K] of the 3'-ethylphosphate group at
298 K for 144a and 144b; R = 0.98 (for 144a) and 0.98 (for 144b). Panels (B) shows the plot of o
K298
ΔG (in kJ mol
-1)
of the N � S pseudorotational equilibrium as a function of the chemical shift of the constituent guanine-9-yl nucleobase
(δH8G in ppm) in 144a as well as H8 and H2 chemical shift of the constituent adenin-9-yl nucleobase (δH8A and δH2A
respectively in ppm) in 144b at 298 K; R = 1.00 (for 144a), 0.94 and 0.98 (for δH8A and δH2A respectively in 144b).
Panels (C) shows the plot of o
K298
ΔG (in kJ mol
-1) of the ε
t
� ε− equilibrium
as a function of the chemical shift of the
constituent guanine-9-yl nucleobase (δH8G in ppm) in 1 as well as H8 and H2 chemical shift of the constituent adenin-
9-yl nucleobase (δH8A and δH2A respectively in ppm) in 2 at 298 K; R = 0.99 (for 144a), 0.94 and 0.97 (for δH8A and
δH2A respectively in 144b).
An additional evidence for the transmission of the free-energy of the
protonation�deprotonation equilibrium of the nucleobase to steer phosphate conformation through
the change of the sugar conformation moiety is that the plots of the pD-dependent o K298
ΔG values
of the N�S equilibrium [Panel (B) in Fig 23] or of the pD-dependent o K298
ΔG values of the εt �ε
-
equilibrium [Panel (C) in Fig 23] in 144a as well as in 144b as a function of the pD-dependent
chemical shift of aromatic protons (H8 in 144a as well as both H8 and H2 in 144b) give straight
lines with high correlation coefficients (R ≥ 0.94).
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
141
Figure 24. The demonstration of single stranded RNA as molecular wires using the model MepGpEt (144a).
Transmission of the free energy of the protonation � deprotonation equilibrium of the guanin-9-yl group in 144a drives
the sugar-phosphate conformations through three consecutive stereoelectronic tunings (Newman projections: a - c; AE =
anomeric effect, GE = gauche effect). Appropriate orbital overlap and the energy difference between the donor and
acceptor orbitals (see Fig 25), dictated by various sugar atoms and substituents, drive the sugar-phosphate conformation
through the interplay of gauche and anomeric effects. All orbitals are shown by straight arrows. Smaller curved arrows
show the preferred torsional orientation, whereas the larger curved arrows indicate the mixing of donor and acceptor
orbitals.
The energy stabilization through either GE or AE, manifested by the interaction between the participating
donor and the acceptor orbitals, is directly proprtional to the square of the overlap between the donor and the acceptor
orbitals (i.e. S2) as well as inversely proportional to their energy differences (stabilization by AE or GE = S
2/ΔE). When
the aglycone is protonated at N7 (Panel A), the N-type pseudorotamers are preferred owing to the strengthening of the
AE(O4'-C1'-N9), as shown by the overlap between the 1nsp2 orbital of one of the O4' lone pairs [1nsp2 (p-type, O4')]
and the antibonding orbital of the C1'-N9 bond [σ*C1’-N9]. This is illustrated through a Newman projection (Panel 3a),
where O4' lone pair orbitals [preferentially the higher energy 1nsp2 (p-type), not the lower energy 2nsp2 (S-type)] and
the nO4' (P-type) → σ∗C1'-N9 overlap stabilizes the N-type over the S-type sugars. For an N-type sugar, nO4' →
σ∗C1'-N9 interactions are possible owing to a near antiperiplanar orientation of 1nsp2 (p-type) with respect to the
σ*C1'-N9 bond [Φ(1nsp2(p-type)-O4'-C1'-N9) ≈ 159°], which takes place when the aglycone is pseudoaxial. In
contrast, they are much weaker when the aglycone is pseudoequatorial in the S-type sugars owing to relatively large
O
HOH2C
O
N
N9
H3'
P
O
-O O
H3CH2C
O
HOH2C
O
N
N9
H3'
P
O
-O
O
CH2CH3
+H+
-H+
1nsp2 (p-type, O4')
σC3'H3'
σ*C1'-N9
σ*C4'O4'
σ*O3'P3'
1nsp2 (p-type, O3')
εt
1nsp2 (p-type, O4')
σC3'H3'
σ*C1'-N9
σ*C4'O4'
σ*O3'P3'
1nsp2 (p-type, O3')
ε-ζ
-
α-
ζ-
α-
H+
H4'
C5'O4'
C2'
C2'
O4'
C5'
H3'
C1'
C1'
C2'
N9 (N7H+)
H1'
O4'
C4'
C2'
N9
H1'
O4'
C4'
+H+
-H+
+H+
-H+
OPO3CH2CH3 OPO3CH2CH3
+H+
-H+
Φ = -101
o
σ* C4'O4'
σ C3'H3'
Φ = -45
o
σ* C4'O4'
σ C3'H3'
Φ = 159o
2nsp2 (s-type)
1nsp2 (p-type)
Φ = 128ο
2nsp2 (s-type)
1nsp2 (p-type)
(a) AE (O4'-C1'-N9)
(b) GE (O4'-C4'-C3'-O3')
(c) AE (O3'-P3'-OEt)
Φ = -156
o
2nsp2 (s-type)
1nsp2 (p-type)
OCH2CH3
O3'C3'
O-
O
H3'
H4'
Φ = -154
o
2nsp2 (s-type)
1nsp2 (p-type)
OCH2CH3
O3'C3'
O-
O
Change of Electronic
Character of the Aglycone
One-way Transmission
OFFON
ON OFF
εt : ON ε
- : OFF
(A) (B)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
142
deviation from antiplanarity, as indicated by smaller Φ values [Φ(1nsp2(p-type)-O4'-C1'-N9) ≈ 128°]. On the other
hand, in the neutral state (Panel 3B) the S-type pseudorotamers are relatively preferred owing to the absence of an O4'-
C1'-N9 anomeric effect as well as due to the stabilization through the O3'-C3'-C4'-O4' gauche effect (i.e. the overlap of
σC3'-H3' with σ*C4'-O4'. In the Newman projection (Panel 3b), the most efficient overlap of the best donor (σC3'-H3') and
best acceptor orbitals (σ*C4'-O4') is indicated by a preference for the gauche orientation [smaller value for Φ(σC3'-H3'-
C3'-C4'-σ*C4'-O4') ≈ -45°] in the S-type sugar over the trans conformer [higher value of Φ(σC3'-H3'-C3'-C4'-σ*C4'-O4') ≈ -
101°] in the N-type sugar. Thus, in the S-type conformation, a reduction of charge density at the O3' lone pair [1nsp2(p-
type)] takes place owing to its maximal interaction with σ*C4'-O4' (presumably a combination of through-bond and
through-space effects are involved in this process), thereby weakening the O3'-P3'-OCH2CH3 anomeric effect in the S/ε-
state. Owing to the nO4' → σ∗C1'-N9 overlap (Panel 3A), the O4' lone pair is relatively more delocalized in the N-type
conformation, which places the O3'-C3'-C4'-O4' fragment in the trans orientation, preventing the gauche effect from
being fully operational (the reverse is true when the anomeric effect is weakened in the S-type conformation, Panel 3B).
Since the 3'-GE is not fully operational in the N-type sugar conformation, the charge density at the O3' lone pair [1nsp2
(p-type, O3')] is fully available to act as a donor and interact through the anomeric effect with the antibonding σ*P3'-
O(ester) orbital [AE(O3'-P3'-OCH2CH3], when C3'-O3' is in εt, O3'-P3' in ζ
−
, and P3'-O5' in α−
conformations. The
Newman projection (Panel 3c) shows that the overlap between the O3' lone pair orbitals and the σ*P3'-O(ester) orbital [nO3'
→ σ* P3'- O(ester) orbital mixing] stabilizes εt over ε
-. This is not only due to an antiperiplanar orientation of 1nsp2(p-
type) with respect to the P3'-O(ester) bond, as Φ(1nsp2(p-type)-O3'-P3'-OCH2CH3) is nearly the same for the two cases,
but largely owing to the greater electron density availible at nO3', arising from the absence of 3'-GE in the N-type sugar.
Therefore, these works constitute the first quantitative evidence for the transmission of the
electronic character of the nucleobase to drive the 3'-phosphate conformation in a ribonucleoside
3',5'-bis-phosphate as the direct consequence of the concomittant modulation of the bias of the
pseudorotational equilibrium of the constituent sugar moiety.
(iv) The mechanism of the transmission of anomeric effect of the nucleobase to 3’-gauche effect and
further on to steer the 3’O-P-O(ester) anomeric effect
In our earlier work on 2'-deoxy23 and ribonucleoside28 3'-ethylphosphates at the neutral pH
(Section 7), we found a non-equivalent methylene protons of the 3'-ethyphosphate moiety, which
turned out to be a temperature-dependent feature. At higher temperature, this non-equivalency
disappeared and the methylene protons showed NMR time average chemical shifts with similar
multiplicity as those of 2'-deoxy counterparts (owing to the absence of 2'-OH promoted hydrogen
bonding). This observation demonstated that the nonequivalency of the methylene protons in
ribonucleotides was owing to the 2'-OH promoted hydrogen bonding with the vicinal O3'. We
attributed the concerted sugar-phosphate backbone conformational change solely to the 2'-OH
effect, the strength of which was modulated by temperature (the “on-off switch”). In our work with
pD-dependent conformational studies with MepGpEt (144a) and MepApEt (144b), we found
similar temperature-dependent pattern and intensity of multiplicities of methylene protons of the 3'-
ethylphosphate moiety at all pD as found for the neutral state of guanosine 3'-ethylphosphate (80)
and adenosine 3'-ethylphosphate (79) respectively, thereby suggesting that the 2'-OH hydrogen-
bonding remains the same in 144a and 144b over the whole pD range (pD 1.0 to 6.7) at room
temperature (298 K). This means that all changes of free-energies observed at 298 K (Table 15) for
144a and 144b as a function of pD is attributed to the free-energy changes of the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
143
protonation�deprotonation equilibrium of the aglycones (i.e. guanin-9-yl and adenin-9-yl) to drive
the sugar-phosphate backbone in a concerted manner. As the pD-tunable change of the electronic
character of the nucleobase tunes the strength of the anomeric effect, an increased preference of the
N-type sugar conformation is imposed because of enhanced nO4'→σ∗
C1'-N orbital interaction,
which, in turn, affects the strength of the [O3'-C3'-C4'-O4'] gauche effect by retuning the energy
levels of the donor and aceeptor orbitals in the σC3'-H → σ∗C4'-O4' interaction. The extent of σC3'-H
→ σ∗C4'-O4' participation influences the electron density at O3', which in turn modulates the O3'-
P3'-O anomeric effect (a concerted transmission). Readers are directed to Fig 24 for a detailed
discussion of the molecular orbital diagram based interpretation of the concerted pD-dependent
change of sugar-phosphate backbone conformation (taking 144a as a model). We also suggest that
the readers take a close look at Fig 25 for the corresponding energy diagram, which show that the
energy levels of the orbitals involved in the anomeric and gauche effects are based on their relative
acceptor/donor abilities in a purely qualitative manner. In Fig 25, various donor and acceptors
orbitals are connected by the dotted lines to show their interdependent modulation and steering of
different orbital mixing. This translates itself in terms of relative strength of gauche and anomeric
effects and the preferred confomational states which make the RNA to act as a molecular wire. The
electron flow is unidirectional from the nucleobase to the phosphate, which is relected by the fact
that the hybrid orbital produced by the O3'-P3'-O(ester) anomeric effect is at a lower energy level
than the corresponding hybrid orbital resulting from O4'-C1'-N anomeric effect.
It is noteworthy that the proof of the one-way (i.e. from nucleobase to phosphate through
sugar) transmission of electronic information in MepGpEt (144a) is evident by the fact that the pKa
value of the guanin-9-yl aglycone is 2.4, which is the same as the 2',3'-dideoxuguanosine (pKa 2.5),
2'-deoxyguanosine (pKa 2.3) and guanosine (pKa 2.1) within the error limits of our experimentals.
In fact, the same pKa value has been found for each nucleobase in corresponding nucleosides
and nucleotides in either 2',3'-dideoxy, 2'-deoxy and ribo configuration as well as in MepGpEt
(144a) and MepApEt (144b), as discussed in Section 4.8, show the minimal influence of the
sbstituents at C2'/3'/5' on the electronic character of their constituent nucleobase at C1'. The final
proof of the operation of O3'-P3'-O anomeric effect could only be experimentally obtained if we
could only measure the ζ and α torsions across the 3'-phosphate backbone and the preferential O3'-
P3'-O bond angle.
(v) pD-dependent conformational change across β torsion is negligible
The conformational equilibrium across the torsion angle β [C4'-C5'-O5'-P5'] for 144a has been
calculated by using temperature dependent 3JC4'P5' as well as the sum of 3JH5'P5' and 3JH5''P5'
(Section 7.1). The analysis of both sets of data provided comparable results with discrepancies
below 3%, which is within the accuracy of Karplus equation used, showing that the population of βt
.
Chatt
opadhya
ya e
t al,
"S
tere
oel
ectr
onic
Eff
ects
in N
ucl
eosi
des
& N
ucl
eoti
des
and t
hei
r S
truct
ura
l Im
pli
cati
ons"
,
Dep
t of
Bio
org
anic
Chem
istr
y, B
ox 5
81, U
ppsa
la U
niv
ersi
ty, S
-75123 U
ppsa
la, S
wed
en, V
er 1
60205 j
yoti
@boc.
uu.s
e
144
AN
AP
CP
CN
PN
BP
BN
DP
DN
N1
P1
Neu
tral
Sta
te
ΔE
[O3
'-P
3'-
O(e
ste
r) A
E]
ΔE
[O4
'-C
1'-
N9
AE
]
σ*C
1'-
N9
σ*C
4'O
4'
1n
sp2
(p
-typ
e, O
3')
σ*P
3'-
O(e
ste
r)σ
C3'H
3'
ΔE
(GE
)
ΔΔ
E(G
E)
ΔΔ
E(A
E)
1n
sp2
(p
-typ
e, O
4')
ΔE
Proto
nate
d S
tate
F
igu
re 2
5.
The
rela
tive
do
no
r-ac
cep
tor
abil
itie
s o
f var
ious
orb
ital
s (i.e
. th
e re
lati
ve
emp
iric
al e
ner
gie
s) a
re m
od
ula
ted
by t
he
nat
ure
of
each
sub
stit
uen
t at
the
sugar
and
o
f th
e p
ho
sphat
e b
ackb
one
as w
ell
as o
n th
e ag
lyco
ne
in fr
ee,
ionic
an
d co
mp
lex st
ate.
S
ince
th
e el
ectr
onic
st
ate
of
the
agly
cone
mo
dula
tes
the
sugar
confo
rmat
ion w
hic
h i
n t
urn
mo
dula
tes
the
pho
sphat
e to
rsio
n,
we
hav
e p
lace
d t
he
1
nsp
2 (p
-type,
O3’) o
rbit
al a
t a
rela
tivel
y l
ow
er e
ner
gy l
evel
than
1
nsp
2 (p
-type,
O4’).
As
the
O4
’-C
1’-
N9
ano
mer
ic e
ffec
t st
arts
op
erat
ing,
the
stre
ngth
of
the
[O3
’-C
3’-
C4
’-O
4’]
gau
che
effe
ct (σ
C3’-
H3’ →
σ*
C4’-
O4’)
co
unte
ract
s th
e an
om
eric
eff
ect
ow
ing t
o t
he
rela
tivel
y l
ow
er e
ner
gy o
f σ
*C
4’-
O4’
[AN b
eco
mes
AP s
tate
when
neu
tral
nucl
eob
ase
(N)
bec
om
es p
roto
nat
ed (
P)]
. A
s th
e nucl
eob
ase
in t
he
(N)
stat
e ta
kes
up
the
(P)
stat
e in
eit
her
Mep
Gp
Et
(14
4a
) o
r in
Mep
Ap
Et
(14
4b
), σ
*C
1’-
N9 b
eco
mes
a b
ette
r ac
cep
tor
and
the
O4
’-C
1’-
N9
ano
mer
ic e
ffec
t is
str
ength
ened
, an
d t
hat
mak
es [
O3
’-
C3
’-C
4’-
O4
’] g
auch
e ef
fect
mo
re e
ffec
tive
[σC
3’-
H3’
(BN b
eco
mes
BP s
tate
) →
σ*
C4’-
O4’, i.e
. m
ore
eff
ecti
ve
orb
ital
mix
ing o
f BP w
ith A
P,
see ΔΔ
H°
10 i
n F
ig 1
3 a
nd
Tab
le 6
). H
ow
ever
, th
e an
om
eric
eff
ect
is s
tro
nger
than
the
gau
che
effe
ct,
ther
efo
re w
e se
over
all
stab
iliz
atio
n o
f m
ore
N-t
yp
e su
gar
s (ΔΔ
E(G
E) <
ΔΔ
E(A
E), T
able
15
)
in t
he
P s
tate
co
mp
ared
to
in N
sta
te.
The
O3
’-P
3’-
O(e
ster
) an
om
eric
eff
ect
is w
eaker
in t
he
S-t
yp
e p
seud
oro
tam
ers
(at
aro
und
neu
tral
pH
, i.e.
in C
N s
tate
) th
an i
n t
he
N-t
yp
e co
unte
rpar
ts (
at a
round
aci
dic
pH
, i.e.
in C
P s
tate
) b
ecau
se t
he σ
C3’-
H3’
orb
ital
over
lap
s w
ith t
he σ
*C
4’-
O4’
(i.e
. [O
3’-
C3
’-C
4’-
O4
’ gau
che
effe
ct]
is s
tro
nger
in
the
form
er s
tate
), r
educi
ng t
he
elec
tro
n d
ensi
ty a
t O
3’
(B).
This
mea
ns
that
the
1
nsp
2 (p
-type,
O3’) l
onep
air
is r
elat
ivel
y l
ess
avai
lab
le i
n S
-typ
e co
nfo
rmat
ion (
at n
eutr
al
pH
) to
inte
ract
wit
h t
he σ
*P
3’-
O(e
ster
) th
an i
n t
he
N-t
yp
e co
nfo
rmat
ion (
at a
cid
ic p
H),
whic
h s
ho
ws
that
the
O3
’-P
3’-
O(e
ster
) an
om
eric
eff
ect
is w
eaker
in S
-typ
e
confo
rmat
ion.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
145
rotamer is pD independent. Hence, the transmission of the information on the change of the
electronic character from nucleobase to 5'-phosphate ends at the γ torsion Similarly, the change of βt
rotamer is pD independent and that of γ torsion is very much within the limit of error of the analyses
in 144b (even less than 144a).
(vi) Conformational equilibrium across C1'-N9 (χ) bond
The populations of the syn and anti conformations involved in the two-state syn �anti
conformational equilibrium across the glycosyl bond in 144a have been calculated from nOe
enhancements experimentally measured at H1’ upon saturation of H8 using the semi-quantitative
equation derived by Rosemeyer and Seela517. The change in population of anti conformer from
neutral (6.7) to acidic (1.0) pDs is observed to be very small viz. 49% to 63% respectively.
(vii) 1H and 31P Chemical shift correlations
The correlation plot between the change in population of S-type sugar conformation owing to the
stronger anomeric effect in 144a upon protonation at N7 of guanin-9-yl and the resultant pD-
dependent H8 chemical shift has been shown in Panel (B) of Fig 26. Inspection of Panels (A) - (D)
in Fig 26 also shows that the percentage conformational change in either 3'-sugar-phosphate
backbone conformation (ε- conformer) or the 5'-end (population of γ+ conformer) in 144a has been
nicely corroborated with the gradual downfield chemical shift of H8 of heterocyclic base due to
protonation. However, the change of γ torsion as a function of pD is very much within the limit of
error of the analyses in 144b (even less than 144a).
Figure 26. (A) The plot of percentage population of γ+ conformer (calculated w.r.t. original assignment, taking H5'
downfield and H5" upfield) at 298K as a function of H8 chemical shift of guanin-9-yl in 144a at different pDs ranging
from 1.6 to 6.7 showing the straight line (R = 0.91) with a slope of 7.37 (σ = 0.99) and an intercept of 11.88 (σ = 8.37).
(B) The plot of %S at 298K as a
function of H8 chemical shift of
guanin-9-yl in 144a at six pD
values ranging from 1.6 to 6.7
showing the straight line (R =
1.00) with a slope of -27.54
(σ = 0.39) and an intercept of
299.00 (σ = 3.31). (C) The plot of
percentage population of ε-
conformer at 298K in 144a as a
function of H8 chemical shift of
its guanin-9-yl at six pDs ranging
from 1.6 to 6.7 showing the
straight line (R = 0.98) with a
slope of -25.18 (σ = 1.43) and an
intercept of 273.27 (σ = 12.05). (D) The plot of percentage
population of βt conformer in 144a at 298K as a function of H8 chemical shift of guanin-9-yl at six pDs ranging from
1.6 to 6.7 showing the straight line (R = 0.96) with a slope of -2.93 (σ = 0.25) and an intercept of 104.25 (σ = 2.07).
Thus the tranmission of the electronic character of the nucleobase to drive the phosphate backbone
conformation via tuning of the pentofuranose conformation has been quantitatively established.
Moreover, the plots in Panels (B) and (C) of Fig 23 and in Panels (A) and (B) of Fig 27 indicate a
direct correlation between 31P3' chemical shift respectively with both ∆G°298Κ (N � S) and
(A)
δH8 (ppm)
7.8 8.0 8.2 8.4 8.6 8.8 9.0
%γ
+
68
70
72
74
76
78
80
(B)
δH8 (ppm )
7.8 8.0 8.2 8.4 8.6 8.8 9.0
%S
50
55
60
65
70
75
80
85
90
(C)
δH8 (ppm)
7.8 8.0 8.2 8.4 8.6 8.8 9.0
%ε
−
45
50
55
60
65
70
75
80
(D)
δH8 (ppm)
7.8 8.0 8.2 8.4 8.6 8.8 9.0
%β
t
77
78
79
80
81
82
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
146
∆G°298Κ (εt � ε-) over the entire pD range for 144a and 144b which can be considered as another
proof of the transmission in terms of conformational energetics from base to phosphate backbone
through nucleotidyl wire in our trinucleotide model system.
(A)
ΔGo
(N/S) at 298 K
-3.6 -3.0 -2.4 -1.8 -1.2 -0.6 0.0
δ3
1
P
-1.65
-1.60
-1.55
-1.50
-1.45
-1.40
(B)
ΔGo
(εt
/ε−
) at 298 K
-2.4 -1.8 -1.2 -0.6 0.0 0.6
δ3
1
P
-1.65
-1.60
-1.55
-1.50
-1.45
-1.40MepGpEt
MepApEt
MepGpEt
MepApEt
Figure 27. Panel (A) shows the plot of ΔG° (in kJ mol-1
) of the N � S pseudorotational equilibrium as a function of the
phosphorous chemical shift (δ31
P in ppm) for MepGpEt (144a) and MepApEt (144b) at 298 K; R = 0.97 (for 144a) and
0.96 (for 144b). Panel (B) shows the plot of ΔG° (in kJ mol-1
) of the εt
� ε− equilibrium
as a function of the
phosphorous chemical shift (δ31
P in ppm) for 144a and 144b at 298 K; R = 0.95 (for 144a) and 0.91 (for 144b).
(viii) Tunibility of aglycones and tranmission of the electronic character to drive the phosphate
backbone conformation
This aglycone dependent conformational transmission of sugar-phosphate backbone via
pentofuranose depends upon the tunibility of aglycone vis-à-vis conformational modulation of sugar
geometry. Our control studies with MepCpEt (144c) at neutral and protonated state showed that the
relative conformational tunibility is in order MepGpEt (144a) > MepApEt (144b) > MepCpEt
(144c) [Fig 28].
(A)
0.944
0.215 0.198
0
0.3
0.6
0.9
1.2
ΔΔδ(P-N)
ppm
EtpGpMe
EtpApMe
EtpCpMe
(B)3.2
1.1 0.9
0
1
2
3
4
ΔΔG(P-N) (N/S)
kJ mol-
1
(C)
0.0
0.4
2.3
0
0.5
1
1.5
2
2.5
ΔΔG(P-N) (εt/ε
−
)
kJ mol-
1
Figure 28. The relative conformational tunibility of MepGpEt (144a), MepApEt (144b) and MepCpEt (144c) between
neutral (N) and protonated (P) states at 298 K. Panel (A) shows the relative change of chemical shift of aromatic protons
[ΔΔδ (P-N)] in 144a – c; Panel (B) and (C) show the relative change of free energy of the N � S pseudorotational
equilibrium [ΔΔG°(P-N) (N/S), in kJ mol-1
] and that of εt
� ε− equilibrium [ΔΔG°(P-N) (ε
t/ε−
), in kJ mol-1
] in 144a – c. The
relative order of tunibility is 144a > 144b > 144c
8.9 The importance of O4' in the self-organization of oligo-DNA
In order to understand the implication of the endocylic oxygen (i.e. O4') in the origin of the
stereoelectronic gauche and anomeric effects, we have determined584 the solution conformation of a
12mer oligo-DNA in which a specific sugar has been replaced by a carbocyclic analogue,
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
147
k1
k-1
Watson-CrickbasepairedDuplex (Ia)
HoogsteenbasepairedDuplex (Ib)
Aristeromycin (i.e. A6 residue). This
modified oligo-DNA, 5'-
d(C1G2C3G4A5A6T7T8C9G10C11G12)-3'
(Duplex I) with the sugar moiety of A6
replaced by aristeromycin has been compared with the natural unmodified counterpart, the
Dickerson-Drew dodecamer (5'-d(CGCGAATTCGCG)-3').
Table 16. Thermodynamics of duplex to single strand melting by UV spectroscopy for the native
and modified dodecamersa from the plot of 1/Tm vs ln(concentration) at 8, 12, 16, 20 and 24 mM
concentration (at least two measurements have been made at each concentration).
Native dodecamer
(kcalmol-1)
Modified dodecamer
(kcalmol-1) ∆H° -82 (± 1) -79 (±1)
-T∆S° -64 (±1) -63 (±1)
∆G°298 -18 (±1) -16 (±1)
a See ref. 585 for an independent measurement.
Table 17. Thermodynamicsa of the exchange process between Hoogsteen and Watson-Crick
basepaired duplexes.
kJmol-1 (Ia)(Ib) 1k⎯⎯ →⎯−
(Ib)(Ia) 1k⎯⎯→⎯
Ea 155 ±13 167 ±14
∆Gθ 117 124
∆Hθ‡
153 ± 13 164 ± 14
-T∆Sθ‡
-36 -40
a The rate constants (k1 and k-1) for the two-state exchange process between Watson-Crick basepaired duplex (Ia) and
Hoogsteen basepaired duplex (Ib) were calculated from the initial slope of build-up rates of NOESY and ROESY
exchange crosspeaks at different mixing times at 7 different temperatures. The rate constants (k1 and k-1) were then used
in an Arrhenius plot to calculate the energy of activation (Ea) which in turn gives the enthalpy (∆Hθ‡
) and enthropy
(∆Sθ‡
) contributions to the free energy of activation (∆Gθ) (see ref. 586 and 587).
The duplex (I) was found to exist in dynamic equilibrium between two different
interconverting conformations, (Ia) and (Ib), with equilibrium constant K = k1 / k-1 = 0.56 ± 0.08
in the temperature range of 287 - 308K. Thus the free-energy of stabilization (ΔΔG°298) of Watson-
Crick basepaired duplex (Ia) realtive to Hoggsteen basepaired duplex (Ib) is 1.4 kJmol-1, which is
calculated from -RT* lnK. In the (Ia)-form, the duplex adopts a canonical B-type conformation
where all basepairs are of the Watson-Crick type, whereas in the (Ib)-type structure, the basepairs
formed between A6 and T7 are of the Hoogsteen type (the N1 of the adenin-9-yl base in
aristeromycin is looking at the major groove), and the rest of the molecule however adopts a B-type
conformation. The thermodynamics and the kinetics of the Watson-Crick basepaired (Ia) duplex �
Hoogsteen basepaired (Ib) duplex equilibrium have also been investigated, and are shown in Tables
16 and 17, respectively. It can be seen that the difference between the thermodynamics of the self-
assembly of the modified DNA duplex and the native counterpart is negligible, well within the error
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
148
limits. This means that there is virtually no energy penalty for replacement of O4' by CH2, which is
evident from the fact that the equilibrium populations of Watson-Crick and Hoogsteen duplexes are
nearly the same. This suggests that there is no preference for furanose-based duplex over the
cyclopentane-based duplex in energetic terms. However, the largest structural implication in the
cyclopentane-based duplex (Ib) is the change of the local structure around the Hoogsteen
basepaired junction compared to the Watson-Crick basepaired counterpart in that the N1 of the
adenin-9-yl moieties in aristeromycin (i.e. A6) is now turned toward the major groove compared to
N7 in the corresponding Watson-Crick basepaired duplex (Ia), which has considerable implication
regarding the hydration behaviour as well as in the ligand binding capability of DNA duplex (Ib)
comapred to (Ia), which is under investigation in the author's laboratory.
9. References
(1) Kirby, A. J. In The Anomeric Effect and Related Stereoelectronic Effects at Oxygen; Springer Verlag: Berlin,
1983; Vol. 15.
(2) Deslongchamps, P. InStereoelectronic Effects in Organic Chemistry; Pergamon Press: Oxford, 1983.
(3) Tvaroska, I.; Bleha, T. In Advances in Carbohydrate Chemistry and Biochemistry; R. S. Tipson and D.
Horton, Ed.; Academic Press: San Diego, 1989; Vol. 47; pp 45.
(4) Juaristi, E.; Cuevas, G. Tetrahedron 1992, 48, 5019.
(5) Thatcher, G. R. J. In The Anomeric Effect and Associated Stereoelectronic Effects; G. R. J. Thatcher, Ed.;
American Chemical Society: Washington, DC, 1993; Vol. 539; pp 6.
(6) Graczyk, P. P.; Mikolajczyk, M. In Topics in Stereochemistry; Eliel, E. L. and Wilen, S.H., Ed.; John Wiley
& Sons, Inc.: New York, 1994; Vol. 21; pp 159.
(7) Perrin, C. L. Tetrahedron 1995, 51, 11901.
(8) Taira, K.; Uebayasi, M.; Maeda, H.; Furukawa, K. Protein Engineering 1990, 3, 691.
(9) Storer, J. W.; Uchimaru, T.; Tanabe, K.; Uebayasi, M.; Nishikawa, S.; Taira, K. J. Am. Chem. Soc. 1991,
113, 5216.
(10) Zhou, D.-M.; Taira, K. Chem. Rev. 1998, 98, 991.
(11) Bizzozero, S. A.; Dutler, H. Bioorg. Chem. 1981, 10, 46.
(12) Post, C. B.; Karplus, M. J. Am. Chem. Soc. 1986, 108, 1317.
(13) Taira, K. Bull. Chem. Soc. Jpn. 1987, 60, 1903.
(14) Koole, L. H.; Buck, H. M.; Nyilas, A.; Chattopadhyaya, J. Can. J. Chem. 1987, 65, 2089.
(15) Koole, L. H.; Buck, H. M.; Bazin, H.; Chattopadhyaya, J. Tetrahedron 1987, 43, 2989.
(16) Koole, L. H.; Moody, H. M.; Buck, H. M.; Grouiller, A.; Essadiq, H.; Vial, J.-M.; Chattopadhyaya, J. Recl.
Trav. Chim. Pays-Bas 1988, 107, 343.
(17) Plavec, J.; Koole, L. H.; Sandström, A.; Chattopadhyaya, J. Tetrahedron 1991, 47, 7363.
(18) Plavec, J.; Buet, V.; Grouiller, A.; Koole, L.; Chattopadhyaya, J. Tetrahedron 1991, 47, 5847.
(19) Plavec, J.; Koole, L. H.; Chattopadhyaya, J. J. Biochem. Biophys. Meth. 1992, 25, 253.
(20) Plavec, J.; Tong, W.; Chattopadhyaya, J. J. Am. Chem. Soc. 1993, 115, 9734.
(21) Plavec, J.; Garg, N.; Chattopadhyaya, J. J. Chem. Soc. Chem. Comm. 1993, 1011.
(22) Plavec, J.; Fabre-Buet, V.; Uteza, V.; Grouiller, A.; Chattopadhyaya, J. J. Biochem. Biophys. Meth. 1993, 26,
317.
(23) Plavec, J.; Thibaudeau, C.; Viswanadham, G.; Sund, C.; Chattopadhyaya, J. J. Chem. Soc. Chem. Com. 1994,
781.
(24) Thibaudeau, C.; Plavec, J.; Watanabe, K. A.; Chattopadhyaya, J. J. Chem. Soc. Chem. Com. 1994, 537.
(25) Thibaudeau, C.; Plavec, J.; Chattopadhyaya, J. J. Am. Chem. Soc. 1994, 116, 8033.
(26) Thibaudeau, C.; Plavec, J.; Garg, N.; Papchikhin, A.; Chattopadhyaya, J. J. Am. Chem. Soc. 1994, 116, 4038.
(27) Plavec, J.; Thibaudeau, C.; Chattopadhyaya, J. J. Am. Chem. Soc. 1994, 116, 6558.
(28) Plavec, J.; Thibaudeau, C.; Viswanadham, G.; Sund, C.; Sandström, A.; Chattopadhyaya, J. Tetrahedron
1995, 51, 11775.
(29) Plavec, J.; Thibaudeau, C.; Chattopadhyaya, J. Pure Appl. Chem. 1996, 68, 2137.
(30) Thibaudeau, C.; Plavec, J.; Chattopadhyaya, J. J. Org. Chem. 1996, 61, 266.
(31) Chattopadhyaya, J. Nucleic Acids Res. Symposium Ser. 1996, 35, 111.
(32) Luyten, I.; Thibaudeau, C.; Sandström, A.; Chattopadhyaya, J. Tetrahedron 1997, 53, 6433.
(33) Luyten, I.; Thibaudeau, C.; Chattopadhyaya, J. J. Org. Chem. 1997, 62, 8800.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
149
(34) Luyten, I.; Thibaudeau, C.; Chattopadhyaya, J. Tetrahedron 1997, 53, 6903.
(35) Thibaudeau, C.; Chattopadhyaya, J. Nucleosides & Nucleotides 1997, 16, 523.
(36) Thibaudeau, C.; Földesi, A.; Chattopadhyaya, J. Tetrahedron 1997, 53, 14043.
(37) Thibaudeau, C.; Földesi, A.; Chattopadhyaya, J. Tetrahedron 1998, 54, 1867.
(38) Thibaudeau, C.; Chattopadhyaya, J. Nucleosides & Nucleotides 1998, 17, 1589.
(39) Thibaudeau, C.; Plavec, J.; Charropadhyaya, J. J. Org. Chem. 1998, 63, 4967.
(40) Luyten, I.; Matulic-Adamic, J.; Beigelman, L.; Chattopadhyaya, J. Nucleosides & Nucleotides 1998, 17,
1605.
(41) Thibaudeau, C.; Kumar, A.; Bekiroglu, S.; Matsuda, A.; Marquez, V. E.; Chattopadhyaya, J. J. Org. Chem.
1998, 63, 5447.
(42) Thibaudeau, C.; Nishizono, N.; Sumita, Y.; Matsuda, A.; Chattopadhyaya, J. Nucleosides & Nucleotides, 18,
1035-1053 (1999).
*(43) Acharya, P.; Nawrot, B.; Sprinzl, M.; Thibaudeau, C.; Chattopadhyaya, J. J. Chem. Soc. Perkin Trans. II,
p1531-1536 (1999).
*(44) (a) Acharya, P.; Trifonov, A.; Thibaudeau, C.; Földesi, A.; Chattopadhyaya, J. Angew Chem. Int. Ed. 38,
3645-3650 (1999); (b) Velikian, I., Acharya, P.; Trifonov, A.; Földesi, A.; Chattopadhyaya, J. J. Phys. Org.
Chem.. 13, 300-305 (2000).
*(45) Polak, M.; Plavec, J.; Trifonova, A.; Földesi, A.; Chattopadhyaya, J. JC.S. Perkin I, p2835-2843 (1999).
(46) Griffey, R. H.; Lesnik, E.; Freier, S.; Sanghvi, Y. S.; Teng, K.; Kawasaki, A.; Guinosso, C.; Wheeler, P.;
Mohan, V.; Cook, P. D. In Carbohydrate Modifications in Antisense Research; New Twists on Nucleic
Acids; Structural Properties of Modified Nucleosides Incorporated into Oligonucleotides V.S. Sanghvi &
P.D. Cook, Ed.; American Chemical Society: Washington DC, 1994; Vol. 580; pp 212,
(47) de Mesmaeker, A.; Häner, R.; Martin, P.; Moser, H. E. Acc. Chem. Res. 1995, 28, 366.
(48) Crooke, S. T. Med. Res. Rev. 1996, 16, 319.
(49) Freier, S. M.; Altmann, K.-H. Nucleic Acids Res. 1997, 25, 4429.
(50) Egli, M. Angew. Chem. Internat. Edit. 1996, 35, 1894.
(51) Zong, G.; Matsukura, M. Int. J. Immunotherapy 1990, 6, 187.
(52) Uhlmann, E.; Peyman, A. Chem. Rev. 1990, 90, 543.
(53) Altmann, K.-H.; Bévierre, M.-O.; De Mesmaeker, A.; Moser, H. E. Bioorg. & Med. Chem. Lett. 1995, 5, 431.
(54) Schmit, C.; Bévierre, M.-O.; De Mesmaeker, A.; Altmann, K.-H. Bioorg. & Med. Chem. Lett. 1994, 4, 1969.
(55) Brown, S. C.; Thomson, S. A.; Veal, J. M.; Davis, D. G. Science 1994, 265, 777.
(56) Betts, L.; Josey, J. A.; Veal, J. M.; Jordan, S. R. Science 1995, 270, 1838.
(57) Lubini, P.; Zürcher, W.; Egli, M. Chem. & Biol. 1994, 1, 39.
(58) Boggon, T. J.; Hancox, E. L.; McAuley-Hecht, K. E.; Connolly, B. A.; Hunter, W. N.; Brown, T.; Walker, R.
T.; Leonard, G. A. Nucleic Acids Res. 1996, 24, 951.
(59) Portmann, S.; Altmann, K.-H.; Reynes, N.; Egli, M. J. Am. Chem. Soc. 1997, 119, 2396.
(60) Gao, X.; Jeffs, P. W. J. Biomol. NMR 1994, 4, 17.
(61) Blommers, M. J. J.; Pieles, U.; de Mesmaeker, A. Nucleic Acids Res. 1994, 22, 4187.
(62) Tereshko, V.; Gryaznov, S.; Egli, M. J. Am. Chem. Soc. 1998, 120, 269.
(63) Berger, I.; Tereshko, V.; Ikeda, H.; Marquez, V. E.; Egli, M. Nucleic Acids Res. 1998, 26, 2473.
(64) Ikeda, H.; Fernandez, R.; Wilk, A.; Barchi Jr., J. J.; Huang, X.; Marquez, V. E. Nucl. Acids Res. 1998, 26,
2237.
(65) Edward, J. T. Chem. and Ind. 1955, 1102.
(66) Lemieux, R. U.; Chu, P. In Abstracts of Papers, 133rd Meeting of the American Chemical Society,; American
Chemical Society: San Francisco, CA, 1958; på 31N.
(67) Lemieux, R. U. In Molecular Rearrangements, Pt. 2; de Mayo P.., Ed.; Interscience Publishers: New York,
1964; pp 709.
(68) Biali, S. E. J. Org. Chem. 1992, 57, 2979.
(69) Booth, H.; Everett, J. R. J. Chem. Soc. Perkin II 1980, 255.
(70) Lemieux, R. U.; Kullnig, R. K.; Bernstein, H. J.; Schneider, W. G. J. Am. Chem. Soc. 1958, 80, 6098.
(71) Lemieux, R. U. In Profiles, Pathways and Dreams; J. I. Seeman, Ed.; American Chemical Society:
Washington DC, 1990; pp 89.
(72) Franck, R. W. Tetrahedron 1983, 39, 3251.
(73) Booth, G. E.; Ouellette, R. J. J. Org. Chem. 1966, 31, 544.
(74) Anderson, C. B.; Sepp, D. T. J. Org. Chem. 1967, 32, 607.
(75) Booth, H.; Khedhair, K. A.; Readshaw, S. A. Tetrahedron 1987, 43, 4699.
(76) de Hoog, A. J.; Buys, H. R.; Altona, C.; Havinga, E. Tetrahedron 1969, 25, 3365.
(77) Pierson, G. O.; Runquist, O. A. J. Org. Chem. 1968, 33, 2572.
(78) Anderson, C. B.; Sepp, D. T. Chem. & Ind. (London) 1964, 2054.
(79) de Hoog, A. J.; Havinga, E. Recl. Trav. Chim. Pays-Bas 1970, 89, 972.
(80) El-Kafrawy, A.; Perraud, R. C. R. Acad. Sci. Ser. C. 1975, 280, 1219.
(81) Barbry, D.; Couturier, D.; Ricard, G. J. Chem. Soc. Perkin II 1982, 249.
(82) Eliel, E. L.; Hargrave, K. D.; Pietrusiewicz, K. M.; Manoharan, M. J. Am. Chem. Soc. 1982, 104, 3635.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
150
(83) Eliel, E. L.; Giza, C. A. J. Org. Chem. 1968, 33, 3754.
(84) Anderson, C. B.; Sepp, D. T. Tetrahedron 1968, 24, 1707.
(85) Anderson, C. B.; Sepp, D. T. J. Org. Chem. 1968, 33, 3272.
(86) Juaristi, E.; Cuevas, G. Tetrahedron Lett. 1992, 33, 2271.
(87) Booth, H.; Dixon, J. M.; Khedhair, K. A. Tetrahedron 1992, 48, 6161.
(88) Schneider, H.-J.; Hoppen, V. J. Org. Chem 1978, 43, 3866.
(89) Eliel, E. L.; Allinger, N. L.; Angyal, S. J.; Morrison, G. A. In Conformational Analysis; Wiley: New York,
1965.
(90) Jensen, F. R.; Bushweller, C. H.; Beck, B. H. J. Am. Chem. Soc. 1969, 91, 344.
(91) Booth, H.; Grindley, T. B.; Khedhair, K. A. J. Chem. Soc. Chem. Comm. 1982, 1047.
(92) Booth, H.; Khedhair, K. A. J. Chem. Soc. Chem. Comm. 1985, 467.
(93) Eliel, E. L. Acc. Chem. Res. 1970, 3, 1.
(94) Lemieux, R. U. Pure Appl. Chem. 1971, 25, 527.
(95) Eliel, E. L. Angew. Chem. 1972, 84, 779.
(96) Bailey, W. F.; Eliel, E. L. J. Am. Chem. Soc. 1974, 96, 1798.
(97) Penner, G. H.; Schaefer, T.; Sebastian, R.; Wolfe, S. Can. J. Chem. 1987, 65, 1845.
(98) Denmark, S. E.; Dappen, M. S.; Sear, N. L.; Jacobs, R. T. J. Am. Chem. Soc. 1990, 112, 3466.
(99) Irwin, J. J.; Ha, T.-K.; Dunitz, J. D. Helv. Chim. Acta 1990, 73, 1805.
(100) Luger, P.; Kothe, G.; Vangehr, K.; Paulsen, H.; Heiker, F. R. Carbohydr. Res. 1979, 68, 207.
(101) Paulsen, H.; Luger, P.; Heiker, F. R. In The Anomeric Effect, Origin and Consequences; W.A. Szarek and D.
Horton, Ed.; American Chemical Society: Washington DC, 1979; Vol. 87; pp 63.
(102) Durette, P. L.; Horton, D. Carbohyd. Res. 1971, 18, 403.
(103) Durette, P. L.; Horton, D. J. Org. Chem. 1971, 36, 2658.
(104) Durette, P. L.; Horton, D. Carbohydr. Res. 1971, 18, 389.
(105) Praly, J.-P.; Lemieux, R. U. Can. J. Chem. 1987, 65, 213.
(106) Lemieux, R. U.; Morgan, A. R. Can. J. Chem. 1965, 43, 2205.
(107) Paulsen, H.; Györgydeák, Z.; Friedmann, M. Chem. Ber. 1974, 107, 1590.
(108) Hosie, L.; Marshall, P. J.; Sinnott, M. L. J. Chem. Soc. Perkin Trans II 1984, 1121.
(109) Perrin, C. L. Pure Appl. Chem. 1995, 67, 719.
(110) Vaino, A. R.; Chan, S. S. C.; Szarek, W. A.; Thatcher, G. R. J. J. Org. Chem. 1996, 61, 4514.
(111) Perrin, C. L.; Armstrong, K. B. J. Am. Chem. Soc. 1993, 115, 6825.
(112) Perrin, C. L.; Fabian, M. A.; Armstrong, K. B. J. Org. Chem. 1994, 59, 5246.
(113) Finch, P.; Nagpurkar, A. G. Carbohydr. Res. 1976, 25, 275.
(114) Fabian, M. A.; Perrin, C. L.; Sinnott, M. L. J. Am. Chem. Soc. 1994, 116, 8398.
(115) Romers, C.; Altona, C.; Buys, H. R.; Havinga, E. In Topics in StereochemistryE.L. Eliel and N. L. Allinger,
Ed.; Wiley Interscience: New York, 1969; Vol. 4; pp 39.
(116) Lucken, E. A. C. J. Chem. Soc. 1959, 2954.
(117) Perrin, C. L.; Armstrong, K. B.; Fabian, M. A. J. Am. Chem. Soc. 1994, 116, 715.
(118) Wiberg, K. B.; Marquez, M. J. Am. Chem. Soc. 1994, 116, 2197.
(119) Hutchins, R. O.; Kopp, L. D.; Eliel, E. L. J. Am. Chem. Soc. 1968, 90, 7174.
(120) Box, V. G. S. Heterocycles 1990, 31, 1157.
(121) Fuchs, B.; Schleifer, L.; Tartakovsky, E. Nouv. J. Chim. 1984, 8, 275.
(122) Schleifer, L.; Senderowitz, H.; Aped, P.; Tartakovsky, E.; Fuchs, B. Carbohydr. Chem. 1990, 206, 21.
(123) Sweigart, D. A. J. Chem. Educ. 1973, 50, 322.
(124) David, S.; Eisenstein, O.; Hehre, W. J.; Salem, L.; Hoffmann, R. J. Am. Chem. Soc. 1973, 95, 3806.
(125) Laing, M. J. Chem. Educ. 1987, 64, 124.
(126) Brundle, C. R.; Turner, D. W. Proc. Roy. Soc. 1968, A307, 27.
(127) Reed, A. E.; Schleyer, P. v. R. Inorg. Chem. 1988, 27, 3969.
(128) Salzner, U. J. Org. Chem. 1995, 60, 986.
(129) Cossé-Barbi, A.; Dubois, J.-E. J. Am. Chem. Soc. 1987, 109, 1503.
(130) Cossé-Barbi, A.; Watson, D. G.; Dubois, J. E. Tetrahedron Lett. 1989, 30, 163.
(131) Salzner, U.; von Ragué Schleyer, P. J. Org. Chem. 1994, 59, 2138.
(132) Fuchs, B.; Ellencweig, A.; Tartakovsky, E.; Aped, P. Angew. Chem. Internat. Edit. 1986, 25, 287.
(133) Kessler, H.; Möhrle, H.; Zimmermann, G. J. Org. Chem. 1977, 42, 66.
(134) Wolfe, S. Acc. Chem. Res. 1972, 5, 102.
(135) Phillips, L.; Wray, V. J. Chem. Soc. Chem. Comm. 1973, 90.
(136) Brunck, T. K.; Weinhold, F. J. Am. Chem. Soc. 1979, 101, 1700.
(137) Juaristi, E. J. Chem. Educ. 1979, 56, 438.
(138) Zefirov, N. S.; Gurvich, L. G.; Shashkov, A. S.; Krimer, M. Z.; Vorob'eva, E. A. Tetrahedron 1976, 32, 1211.
(139) Zefirov, N. S. Tetrahedron 1977, 33, 3193.
(140) Zefirov, N. S.; Samoshin, V. V.; Subbotin, O. A.; Baranenkov, V. I.; Wolfe, S. Tetrahedron 1978, 34, 2953.
(141) Wiberg, K. B.; Murcko, M. A.; Laidig, K. E.; MacDougall, P. J. J. Phys. Chem. 1990, 94, 6956.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
151
(142) Murcko, M. A.; DiPaola, R. A. J. Am. Chem. Soc. 1992, 114, 10010.
(143) Wiberg, K. B. Acc. Chem. Res. 1996, 29, 229.
(144) Abe, A.; Inomata, K. J. Mol. Struct. 1991, 245, 399.
(145) Craig, N. C.; Piper, L. G.; Wheeler, V. L. J. Phys. Chem. 1971, 75, 1453.
(146) Kaloustian, M. K.; Dennis, N.; Mager, S.; Evans, S. A.; Alcudia, F.; Eliel, E. L. J. Am. Chem. Soc. 1976, 98,
956.
(147) Eliel, E. L.; Juaristi, E. J. Am. Chem. Soc. 1978, 100, 6114.
(148) Harris, W. C.; Holtzclaw, J. R.; Kalasinsky, V. F. J. Chem. Phys. 1977, 67, 3330.
(149) Durig, J. R.; Liu, J.; Little, T. S.; Kalasinsky, V. F. J. Phys. Chem. 1992, 96, 8224.
(150) Hirano, T.; Nonoyama, S.; Miyajima, T.; Kurita, Y.; Kawamura, T.; Sato, H. J. Chem. Soc. Chem. Comm.
1986, 606.
(151) Fernholt, L.; Kveseth, K. Acta Chem. Scand. 1980, A34, 163.
(152) van Schaick, E. J. M.; Geise, H. J.; Mijlhoff, F. C.; Renes, G. J. Mol. Struct. 1973, 16, 23.
(153) Freisen, D.; Hedberg, K. J. Am. Chem. Soc. 1980, 102, 3987.
(154) Dixon, D. A.; Smart, B. E. J. Phys. Chem. 1988, 92, 2729.
(155) Kazerouni, M. R.; Hedberg, L.; Hedberg, K. J. Am. Chem. Soc. 1997, 119, 8324.
(156) Zefirov, N. S. Zhur. Org. Khim. 1970, 6, 1761.
(157) Zefirov, N. S.; Blagoveshchensky, V. S.; Kazimirchik, I. V.; Surova, N. S. Tetrahedron 1971, 27, 3111.
(158) Hoffmann, R. Acc. Chem. Res. 1971, 4, 1.
(159) Epiotis, N. D.; Sarkanen, S.; Bjorkquist, D.; Bjorkquist, L.; Yates, R. J. Am. Chem. Soc. 1974, 96, 4075.
(160) Gleiter, R.; Kobayashi, M.; Zefirov, N. S.; Palyulin, V. A. Dokl. Chem. 1977, 235, 396.
(161) Dionne, P.; St-Jacques, M. J. Am. Chem. Soc. 1987, 109, 2616.
(162) Craig, N. C.; Chen, A.; Suh, K. H.; Klee, S.; Mellau, G. C.; Winnewisser, B. P.; Winnewisser, M. J. Am.
Chem. Soc. 1997, 119, 4789.
(163) Runtz, G. R.; Bader, R. F. W.; Messer, R. R. Can. J. Chem. 1977, 55, 3040.
(164) Muir, M.; Baker, J. Molec. Phys. 1996, 89, 211.
(165) (a) Saenger, W. In Principles of Nucleic Acid Structure; Springer Verlag: Berlin, 1988. (b) ibid, pp93-96 for
anomeric effect across the 3'O-P-O-ester fragment in the sugar-phosphate backbone and references cited
therein.
(166) Watanabe, K. A. In Chemistry of Nucleosides and Nucleotides; L. B. Townsend, Ed.; Plenum Press: New
York, 1994; Vol. 3; p 421.
(167) IUPAC-IUB Eur. J. Biochem. 1983, 131, 9.
(168) IUPAC-IUB Pure Appl. Chem. 1983, 55, 1273.
(169) Angyal, S. J. In The Composition of Reducing Sugars in Solution, R.S. Tipson and D. Horton, Ed.; Academic
Press: London, 1984; Vol. 42; p 15.
(170) Squillacote, M.; Sheridan, R. S.; Chapman, O. L.; Anet, F. A. L. J. Am. Chem. Soc. 1975, 97, 3244.
(171) Anderson, J. E. Top. Curr. Chem. 1974, 45, 139.
(172) Kilpatrick, J. E.; Pitzer, K. S.; Spitzer, R. J. Am. Chem. Soc. 1947, 69, 2483.
(173) Strauss, H. L. Ann. Rev. Phys. Chem. 1983, 34, 301.
(174) de Leeuw, H. P. M.; Haasnoot, C. A. G.; Altona, C. Isr. J. Chem. 1980, 20, 108.
(175) Swarna Latha, Y.; Yathindra, N. Biopolymers 1992, 32, 249.
(176) Carreira, L. A.; Jiang, G. J.; Person, W. B.; Willis, J. N. J. J. Chem. Phys. 1972, 56, 1440.
(177) Hall, L. D.; Steiner, P. R.; Pedersen, C. Can. J. Chem. 1970, 48, 1155.
(178) Cremer, D.; Pople, J. A. J. Am. Chem. Soc. 1975, 97, 1354.
(179) Geise, H. J.; Altona, C.; Romers, C. Tetrahedron Lett. 1967, 15, 1383.
(180) Altona, C.; Geise, H. J.; Romers, C. Tetrahedron 1968, 24, 13.
(181) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc. 1972, 94, 8205.
(182) Haasnoot, C. A. G.; de Leeuw, F. A. A. M.; de Leeuw, H. P. M.; Altona, C. Biopolymers 1981, 20, 1211.
(183) Diez, E.; Esteban, A. L.; Bermejo, F. J.; Altona, C.; de Leeuw, F. A. A. M. J. Mol. Struct. 1984, 125, 49.
(184) de Leeuw, F. A. A. M.; van Kampen, P. N.; Altona, C.; Diez, E.; Esteban, A. L. J. Mol. Struct. 1984, 125, 67.
(185) Diez, E.; Esteban, A. L.; Bermejo, F. J.; Rico, M. J. Phys. Chem. 1980, 84, 3191.
(186) Diez, E.; Esteban, A. L.; Guilleme, J.; Bermejo, F. L. J. Mol. Struct. 1981, 70, 61.
(187) Davies, D. B.; Sadikot, H. Biopolymers 1983, 22, 1843.
(188) Sarma, R. H.; Lee, C.-H.; Evans, F. E.; Yathindra, N.; Sundaralingam, M. J. Am. Chem. Soc. 1974, 96, 7337.
(189) Lüdemann, H.-D.; Röder, O.; Westhof, E.; Goldammer, E. v.; Müller, A. Biophys. Struct. Mechanism 1975,
1, 121.
(190) Rossi, C.; Pichi, M. P.; Tiezzi, E. Magn. Reson. Chem. 1990, 28, 348.
(191) Davies, D. B.; Danyluk, S. S. Biochemistry 1974, 13, 4417.
(192) Davies, D. B.; Danyluk, S. S. Biochemistry 1975, 14, 543.
(193) Plach, H.; Westhof, E.; Lüdemann, H.-D.; Mengel, R. Eur. J. Biochem. 1977, 80, 295.
(194) de Leeuw, H. P. M.; de Jager, J. R.; Koeners, H. J.; van Boom, J. H.; Altona, C. Eur. J. Biochem. 1977, 76,
209.
(195) Westhof, E.; Plach, H.; Cuno, I.; Lüdemann, H.-D. Nucleic Acids Res. 1977, 4, 939.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
152
(196) Ekiel, I.; Darzynkiewicz, E.; Dudycz, L.; Shugar, D. Biochemistry 1978, 17, 1530.
(197) Klimke, G.; Cuno, I.; Lüdemann, H.-D.; Mengel, R.; Robins, M. J. Z. Naturforsch. 1979, 34c, 1075.
(198) Ekiel, I.; Remin, M.; Darzynkiewicz, E.; Shugar, D. Biochim. Biophys. Acta 1979, 562, 177.
(199) Guschlbauer, W.; Jankowski, K. Nucleic Acids Res. 1980, 8, 1421.
(200) Guschlbauer, W. Biochim. Biophys. Acta 1980, 610, 47.
(201) Klimke, G.; Cuno, I.; Lüdemann, H.-D.; Mengel, R.; Robins, M. J. Z. Naturforsch. 1980, 35c, 853.
(202) Klimke, G.; Cuno, I.; Lüdemann, H.-D.; Mengel, R.; Robins, M. J. Z. Naturforsch. 1980, 35c, 865.
(203) Haasnoot, C. A. G.; de Leeuw, F. A. A. M.; de Leeuw, H. P. M.; Altona, C. Org. Magn. Reson. 1981, 15, 43.
(204) Lipnick, R. L.; Fissekis, J. D. Biochim. Biophys. Acta 1980, 608, 96.
(205) de Leeuw, F. A. A. M.; Altona, C. J. Chem. Soc. Perkin II 1982, 375.
(206) Uesugi, S.; Kaneyasu, T.; Imura, J.; Ikehara, M.; Cheng, D. M.; Kan, L.-S.; Ts'o, P. O. P. Biopolymers 1983,
22, 1189.
(207) Yamamoto, Y.; Yokoyama, S.; Miyazawa, T.; Watanabe, K.; Higuchi, S. FEBS Lett. 1983, 157, 95.
(208) Cheng, D. M.; Kan, L.-S.; Ts'O, P. O. P.; Uesugi, S.; Takatsuka, Y.; Ikehara, M. Biopolymers 1983, 22, 1427.
(209) de Leeuw, F. A. A. M.; Altona, C. J. Comp. Chem. 1983, 4, 428.
(210) Chiasson, J. B.; Jankowski, K.; Brostow, W. J. Bioelectricity 1983, 2, 93.
(211) Westhof, E.; Sundaralingam, M. J. Am. Chem. Soc. 1983, 105, 970.
(212) Birnbaum, G. I.; Brisson, J. R.; Krawczyk, S. H.; Townsend, L. B. J. Am. Chem. Soc. 1985, 107, 7635.
(213) Rinkel, L. J.; Altona, C. J. Biomol. Struct. Dynamics 1987, 4, 621.
(214) Orbons, L. P. M.; van der Marel, G. A.; van Boom, J. H.; Altona, C. Eur. J. Biochem. 1987, 170, 225.
(215) Hosur, R. V.; Govil, G.; Miles, H. T. Magn. Reson. Chem. 1988, 26, 927.
(216) Swapna, G. V. T.; Jagannadh, B.; Gurjar, M. K.; Kunwar, A. C. Biochem. Biophys. Res. Commun. 1989, 164,
1086.
(217) Altona, C. In Theor. Biochem. & Mol. Biophys.; D. L. Beveridge and R. Lavery, Ed.; Adenine Press: 1990;
Vol. 1 (DNA); pp 1.
(218) Pieters, J. M. L.; de Vroom, E.; van der Marel, G. A.; van Boom, J. H.; Koning, T. M. G.; Kaptein, R.;
Altona, C. Biochemistry 1990, 29, 788.
(219) Jagannadh, B.; Reddy, D. V.; Kunwar, A. C. Biochem. Biophys. Res. Commun. 1991, 179, 386.
(220) Everaert, D. H.; Peeters, O. M.; de Ranter, C. J.; Blaton, N. M.; van Aerschot, A.; Herdewijn, P. Antivir.
Chem. Chemotherapy 1993, 4, 289.
(221) Wijmenga, S. S.; Mooren, M. M. W.; Hilbers, C. W. In NMR of Macromolecules; G. C. K. Roberts, Ed.;
Oxford University Press: Oxford, 1993; pp. 217-288.
(222) van den Boogaart, J. E.; Kalinichenko, E. N.; Podkopaeva, T. L.; Mikhailopulo, I. A.; Altona, C. Eur. J.
Biochem. 1994, 221, 759.
(223) Mooren, M. M. W.; Wijmenga, S. S.; van der Marel, G. A.; van Boom, J. H.; Hilbers, C. W. Nucleic Acids
Res. 1994, 22, 2658.
(224) Verberckmoes, F.; Esmans, E. L. Spectrochim. Acta 1995, 51A, 153.
(225) Rosemeyer, H.; Seela, F. J. Chem. Soc. Perkin Trans II 1997, 2341.
(226) Barchi, J. J. J.; Jeong, L.-S.; Siddiqui, M. A.; Marquez, V. E. J. Biochem. Biophys. Meth. 1997, 34, 11.
(227) de Leeuw, F. A. A. M.; Altona, C.; Kessler, H.; Bermel, W.; Friedrich, A.; Krack, G.; Hull, W. E. J. Am.
Chem. Soc. 1983, 105, 2237.
(228) Church, T. J.; Carmichael, I.; Serianni, A. S. J. Am. Chem. Soc. 1997, 119, 8946.
(229) Haasnoot, C. A. G.; Liskamp, R. M. J.; van Dael, P. A. W.; Noordik, J. H.; Ottenheijm, H. C. J. J. Am. Chem.
Soc. 1983, 105, 5406.
(230) Pavelcik, F.; Luptakova, E. Collect. Czech. Chem. Commun. 1990, 55, 1427.
(231) Varela, O.; Marino, C.; Cicero, D. An. Assoc. Quim. Argent. 1991, 79, 107.
(232) Cremer, D.; Pople, J. A. J. Am. Chem. Soc. 1975, 97, 1358.
(233) Pitzer, K. S.; Donath, W. E. J. Am. Chem. Soc. 1959, 81, 3213.
(234) Hoffmann, R. A.; van Wijk, J.; Leeflang, B. R.; Kamerling, J. P.; Altona, C.; Vliegenthart, J. F. G. J. Am.
Chem. Soc. 1992, 114, 3710.
(235) Cushley, R. J.; Codington, J. F.; Fox, J. J. Can. J. Chem. 1968, 46, 1131.
(236) Fang, K. N.; Kondo, N. S.; Miller, P. S.; Ts'o, P. O. P. J. Am. Chem. Soc. 1971, 93, 6647.
(237) Schleich, T.; Blackburn, B. J.; Lapper, R. D.; Smith, I. C. P. Biochemistry 1972, 11, 137.
(238) Tran-Dinh, S.; Cachaty, C. Biochim. Biophys. Acta 1973, 335, 1.
(239) Allore, B. D.; Queen, A.; Blonski, W. J.; Hruska, F. E. Can. J. Chem. 1983, 61, 2397.
(240) West, R. T.; Garza, L. A. I.; Winchester, W. R.; Walmsley, J. A. Nucleic Acids Res. 1994, 22, 5128.
(241) Feigon, J.; Wang, A. H.-J.; van der Marel, G. A.; van Boom, J. H.; Rich, A. Nucleic Acids Res. 1984, 12,
1243.
(242) Tran-Dinh, S.; Taboury, J.; Neumann, J.-M.; Huynh-Dinh, T.; Genissel, B.; Langlois d'Estaintot, B.; Igolen, J.
Biochemistry 1984, 23, 1362.
(243) Davis, P. W.; Hall, K.; Cruz, P.; Tinoco Jr., I.; Neilson, T. Nucleic Acids Res. 1986, 14, 1279.
(244) Davis, P. W.; Adamiak, R. W.; Tinoco, I. J. Biopolymers 1990, 29, 109.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
153
(245) Agback, P.; Sandström, A.; Yamakage, S.-i.; Sund, C.; Glemarec, C.; Chattopadhyaya, J. J. Biochem.
Biophys. Meth. 1993, 27, 229.
(246) Agback, P.; Glemarec, C.; Yin, L.; Sandström, A.; Plavec, J.; Sund, C.; Yamakage, S.-i.; Viswanadham, G.;
Rousse, B.; Puri, N.; Chattopadhyaya, J. Tetrahedron Lett. 1993, 34, 3929.
(247) Izatt, R. M.; Christensen, J. J.; Rytting, J. H. Chem. Rev. 1971, 71, 439.
(248) Brahms, S.; Fritsch, V.; Brahms, J. G.; Westhof, E. J. Mol. Biol. 1992, 223, 455.
(249) Westhof, E.; Sundaralingam, M. J. Am. Chem. Soc. 1980, 102, 1493.
(250) Sasisekharan, V. Jerus. Symp. Quant. Chem. Biochem. 1973, 5, 247.
(251) Coulter, C. L. J. Am. Chem. Soc. 1973, 95, 570.
(252) Reddy, B. S.; Saenger, W. Acta Cryst. 1978, B34, 1520.
(253) Saenger, W.; Eckstein, F. J. Am. Chem. Soc. 1970, 92, 4712.
(254) Lee, C.-H.; Sarma, R. H. J. Am. Chem. Soc. 1976, 98, 3541.
(255) Ku Chwang, A.; Sundaralingam, M. Acta Cryst. 1974, B30, 1233.
(256) Coulter, C. L. Acta Cryst. 1969, B25, 2055.
(257) Varughese, K. I.; Lu, C. T.; Kartha, G. J. Am. Chem. Soc. 1982, 104, 3398.
(258) Lai, T. F.; Marsh, R. E. Acta Cryst. 1972, B28, 1982.
(259) Shikata, K.; Ueki, T.; Mitsui, T. Acta Cryst. 1973, B29, 31.
(260) Gurskaya, G. V.; Tsapkina, E. N.; Skaptasova, N. V.; Lindeman, A. A.; Struchkov, Y. T. Dokl Akad. Nauk.
USSR 1986, 291, 854.
(261) Birnbaum, G. I.; Giziewicz, J.; Gabe, E. J.; Lin, T.-S.; Prusoff, W. H. Can. J. Chem. 1987, 65, 2135.
(262) Camerman, A.; Mastropaolo, D.; Camerman, N. Proc. Natl. Acad. Sci. USA 1987, 84, 8239.
(263) van Roey, P.; Salerno, J. M.; Duax, W. L.; Chu, C. K.; Ahn, M. K.; Schinazi, R. F. J. Am. Chem. Soc. 1988,
110, 2277.
(264) Parthasarathy, R.; Kim, H. Biochem. Biophys. Res. Commun. 1988, 152, 351.
(265) Dyer, I.; Low, J. N.; Tollin, P.; Wilson, H. R.; Howie, R. A. Acta Cryst. 1988, C44, 767.
(266) van Roey, P.; Schinazi, R. F. Antiviral Chem. & Chemotherapy 1990, 1, 93.
(267) Camerman, N.; Mastropaolo, D.; Camerman, A. Proc. Natl. Acad. Sci. USA 1990, 87, 3534.
(268) Chu, C. K.; Bhadti, V. S.; Doboszewski, B.; Gu, Z. P.; Kosugi, Y.; Pullaiah, K. C.; van Roey, P. J. Org.
Chem. 1989, 54, 2217.
(269) Birnbaum, G. I.; Lin, T.-S.; Prusoff, W. H. Biochem. Biophys. Res. Commun. 1988, 151, 608.
(270) Koole, L. H.; Plavec, J.; Liu, H.; Vincent, B. R.; Dyson, M. R.; Coe, P. L.; Walker, R. T.; Hardy, G. W.;
Rahim, S. G.; Chattopadhyaya, J. J. Am. Chem. Soc. 1992, 114, 9936.
(271) Barnes, C. L.; Hawkinson, S. W.; Wigler, P. W. Acta Cryst. 1980, B36, 2299.
(272) Birnbaum, G. I.; Watanabe, K. A.; Fox, J. J. Can. J. Chem. 1980, 58, 1633.
(273) Barrett, A. G. M.; Broughton, H. B. J. Org. Chem. 1984, 49, 3673.
(274) Mc Kenna, R.; Neidle, S.; Serafinowski, P. Acta Cryst. 1987, C43, 2358.
(275) Mc Kenna, R.; Neidle, S.; Serafinowski, P. Acta Cryst. 1990, C46, 2448.
(276) Mumper, M. W.; Aurenge, C.; Hosmane, R. S. J. Biomol. Struct. Dynamics 1994, 11, 1107.
(277) Koyama, G.; Nakamura, H.; Umezawa, H.; Iitaka, Y. Acta Cryst. 1976, B32, 813.
(278) Prusiner, P.; Brennan, T.; Sundaralingam, M. Biochemistry 1973, 12, 1196.
(279) Rohrer, D.; Sundaralingam, M. J. Am. Chem. Soc. 1970, 92, 4950.
(280) Koyama, G.; Umezawa, H.; Iitaka, Y. Acta Cryst. 1974, B30, 1511.
(281) Singh, P.; Hodgson, D. J. Acta Cryst. 1977, B33, 3189.
(282) Hempel, A.; Lane, B. G.; Camerman, N. Acta Cryst. 1997, C53, 1707.
(283) Krol, M. C.; Huige, C. J. M.; Altona, C. J. Comp. Chem. 1990, 11, 765.
(284) Marquez, V. E. In Advances in Antiviral Drug Design; E. de Clercq, Ed.; JAI Press, Inc.: Greenwich, CT,
1996; Vol. 2; pp 89.
(285) Borthwick, A. D.; Biggadike, K. Tetrahedron 1992, 48, 571.
(286) Marquez, V. E.; Lim, M.-I. Med. Res. Rev. 1986, 6, 1.
(287) Altmann, K.-H.; Imwinkelried, R.; Kesselring, R.; Rihs, G. Tetrahedron Lett. 1994, 35, 7625.
(288) Shealy, Y. F.; O'Dell, C. A. Tetrahedron Lett. 1969, 27, 2231.
(289) Marumoto, R.; Yoshioka, Y.; Furukawa, Y.; Honjo, M. Chem. Pharm. Bull. 1976, 24, 2624.
(290) Kinoshita, K.; Yaginuma, S.; Hayashi, M.; Nakatsu, K. Nucleosides & Nucleotides 1985, 4, 661.
(291) Popescu, A.; Hörnfeldt, A.-B.; Gronowitz, S.; Johansson, N. G. Nucleosides & Nucleotides 1995, 14, 1639.
(292) Siddiqui, M. A.; Ford, H. J.; George, C.; Marquez, V. E. Nucleosides & Nucleotides 1996, 15, 235.
(293) Rodriguez, J. B.; Marquez, V. E.; Nicklaus, M. C.; Mitsuya, H.; Jr., B. J. J. Med. Chem. 1994, 37, 3389.
(294) Kishi, T.; Muroi, M.; Kusaka, T.; Nishikawa, M.; Kamiya, K.; Mizuno, K. Chem. Pharm. Bull. 1972, 20, 940.
(295) Kálman, A.; Koritsánszky, T.; Béres, J.; Sági, G. Nucleosides & Nucleotides 1990, 9, 235.
(296) Biggadike, K.; Borthwick, A. D.; Exall, A. M.; Kirk, B. E.; Roberts, S. M.; Youds, P.; Slawin, A. M. Z.;
Williams, D. J. J. Chem. Soc. Chem. Comm. 1987, 255.
(297) Marumoto, R. AIDS-Iyakuhin Kaihatsu-Kenkyu-Hokuku (The Japan Health Sciences Foundation) 1991, 3.
(298) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.;
Keith, T.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.;
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
154
Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C.
Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox,
D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A.
Gaussian 94 Revision C.3; Gaussian, Inc.: Pittsburgh PA, 1995.
(299) Govil, G.; Saran, A. J. Theor. Biol. 1971, 33, 399.
(300) Lugovskoi, A. A.; Dashevskii, V. G. Mol. Biol. (USSR) 1972, 6, 354.
(301) Lugovskoi, A. A.; Dashevskii, V. G.; Kitaigorodskii, A. I. Mol. Biol. (USSR) 1972, 6, 361.
(302) Lugovskoi, A. A.; Dashevskii, V. G.; Kitaigorodskii, A. I. Mol. Biol. (USSR) 1972, 6, 494.
(303) Levitt, M.; Warshel, A. J. Am. Chem. Soc. 1978, 100, 2607.
(304) Olson, W. K.; Sussman, J. L. J. Am. Chem. Soc. 1982, 104, 270.
(305) Olson, W. K. J. Am. Chem. Soc. 1982, 104, 278.
(306) Pearlman, D. A.; Kim, S.-H. J. Biomol. Struct. Dynamics 1985, 3, 99.
(307) Röder, O.; Lüdemann, H.-D.; von Goldammer, E. Eur. J. Biochem. 1975, 53, 517.
(308) Plavec, J.; Roselt, P.; Földesi, A.; Chattopadhyaya, J. Mag. Res. Chem. 1998, 36, 732.
(309) Jalluri, R. K.; Yuh, Y. H.; Taylor, E. W. In The Anomeric Effect and Associated Stereoelectronic Effects; G.
R. J. Thatcher, Ed.; American Chemical Society: Washington, DC, 1993; Vol. 539; pp 277.
(310) Allen, F. H.; Davies, J. E.; Galloy, J. J.; Johnson, O.; Kennard, O.; MacRae, C. F.; Mitchell, E. M.; Mitchell,
G. F.; Smith, J. M.; Watson, D. G. J. Chem. Inf. Comput. Sci. 1991, 31, 187.
(311) Lo, A.; Shefter, E.; Cochran, T. G. J. Pharm. Sc. 1975, 64, 1707.
(312) Fox, J. J.; Shugar, D. Biochim. Biophys. Acta 1952, 9, 369.
(313) Tran-Dinh, S.; Thiéry, J.; Guschlbauer, W.; Dunand, J.-J. Biochim. Biophys. Acta 1972, 281, 289.
(314) Sarma, R. H.; Mynott, R. J. J. Am. Chem. Soc. 1973, 95, 1641.
(315) Remin, M.; Shugar, D. J. Am. Chem. Soc. 1973, 95, 8146.
(316) Altona, C.; Sundaralingam, S. J. Am. Chem. Soc. 1973, 95, 2333.
(317) Tran-Dinh, S.; Guschlbauer, W. Nucleic Acids Res. 1975, 3, 873.
(318) Hruska, F. E.; Wood, D. J.; Singh, H. Biochim. Biophys. Acta 1977, 474, 129.
(319) Seela, F.; Becher, G.; Rosemeyer, H.; Reuter, H.; Kastner, G.; Mikhailopulo, I. A. Helv. Chim. Acta 1999, 82,
105.
(320) Hillen, W.; Gassen, H. G. Biochim. Biophys. Acta 1978, 518, 7.
(321) Egert, E.; Lindner, H. J.; Hillen, W.; Böhm, M. C. J. Am. Chem. Soc. 1980, 102, 3707.
(322) Uhl, W.; Reiner, J.; Gassen, H. G. Nucleic Acids Res. 1983, 11, 1167.
(323) O'Leary, D. J.; Kishi, Y. J. Org. Chem. 1994, 59, 6629.
(324) Birnbaum, G. I.; Blonski, W. J. P.; Hruska, F. E. Can. J. Chem. 1983, 61, 2299.
(325) Cadet, J.; Ducolomb, R.; Taieb, C. Tetrahedron Lett. 1975, 40, 3455.
(326) Birnbaum, G. I.; Hruska, F. E.; Niemczura, W. P. J. Am. Chem. Soc. 1980, 102, 5586.
(327) Bergström, D. E.; Zhang, P.; Toma, P. H.; Andrews, P. C.; Nichols, R. J. Am. Chem. Soc. 1995, 117, 1201.
(328) Post, M. L.; Birnbaum, G. I.; Huber, C. P.; Shugar, D. Biochim. Biophys. Acta 1977, 479, 133.
(329) Sundaralingam, M. J. Am. Chem. Soc. 1971, 93, 6644.
(330) Rohrer, D. C.; Sundaralingam, M. J. Am. Chem. Soc. 1970, 92, 4956.
(331) Remin, M.; Ekiel, I.; Shugar, D. Eur. J. Biochem. 1975, 53, 197.
(332) Seto, H.; Otake, N.; Yonehara, H. Tetrahedron Lett. 1972, 38, 3991.
(333) Ruzic-Toros, Z. Acta Cryst. 1979, B35, 1277.
(334) Crowfoot Hodgkin, D.; Lindsey, J.; Sparks, R. A.; Trueblood, K. N.; White, J. G. Proc. Roy. Soc. 1962,
A266, 494.
(335) Brink-Shoemaker, C.; Cruickshank, D. W. J.; Crowfoot Hodgkin, D.; Kamper, M. J.; Pilling, D. Proc. Roy.
Soc. 1964, A278, 1.
(336) Hawkinson, S. W.; Coulter, C. L.; Greaves, M. L. Proc. Roy. Soc. 1970, A318, 143.
(337) Savage, H. F. J.; Lindley, P. F.; Finney, J. L.; Timmins, P. A. Acta Cryst. 1987, B43, 280.
(338) Lenhert, P. G. Proc. Roy. Soc. 1968, A303, 45.
(339) Hamor, T. A.; O'Leary, M. K.; Walker, R. T. Acta Cryst. 1977, B33, 1218.
(340) Shefter, E.; Kotick, M. P.; Bardos, J. J. Pharm. Sci. 1967, 56, 1293.
(341) Ide, H.; Shimizu, H.; Kimura, Y.; Sakamoto, S.; Makino, K.; Glackin, M.; Wallace, S. S.; Nakamuta, H.;
Sasaki, M.; Sugimoto, N. Biochemistry 1995, 34, 6947.
(342) Gutowski, G. E.; Chaney, M. O.; Jones, N. D.; Hamill, R. L.; Davis, F. A.; Miller, R. D. Biochem. Biophys.
Res. Commun. 1973, 51, 312.
(343) Armstrong, V. W.; Dattagupta, J. K.; Eckstein, F.; Saenger, W. Nucleic Acids Res. 1976, 3, 1791.
(344) Cline, S. J.; Hodgson, D. J. Biochim. Biophys. Acta 1979, 563, 540.
(345) Post, M. L.; Huber, C. P.; Birnbaum, G. I.; Shugar, D. Can. J. Chem. 1981, 59, 238.
(346) Piper, I. M.; MacLean, D. B.; Faggiani, R.; Lock, C. J. L.; Szarek, W. A. Can. J. Chem. 1985, 63, 2915.
(347) Serianni, A.; Barker, R. J. Org. Chem. 1984, 49, 3292.
(348) Serianni, A. S.; Chipman, D. M. J. Am. Chem. Soc. 1987, 109, 5297.
(349) Kline, P. C.; Serianni, A. S. J. Am. Chem. Soc. 1990, 112, 7373.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
155
(350) Raap, J.; van Boom, J. H.; van Lieshout, H. C.; Haasnoot, C. A. G. J. Am. Chem. Soc. 1988, 110, 2736.
(351) Ellervik, U.; Magnusson, G. J. Am. Chem. Soc. 1994, 116, 2340.
(352) Teng, K.; Cook, P. D. J. Org. Chem. 1994, 59, 278.
(353) Cyr, N.; Perlin, A. S. Can. J. Chem. 1979, 57, 2504.
(354) Le Doan, T.; Perrouault, L.; Praseuth, D.; Habhoub, N.; Decout, J.-L.; Thouong, N. T.; Lhomme, J.; Hélène,
C. Nucleic Acids Res. 1987, 15, 7749.
(355) Moser, H. E.; Dervan, P. B. Science 1987, 238, 645.
(356) Povsic, T. J.; Dervan, P. B. J. Am. Chem. Soc. 1989, 111, 3059.
(357) Xodo, L. E.; Manzini, G.; Quadrifoglio, F.; van der Marel, G. A.; van Boom, J. H. Nucleic Acids Res. 1991,
19, 5625.
(358) Brown, T.; Leonard, G. A.; Booth, E. D.; Kneale, G. J. Mol. Biol. 1990, 212, 437.
(359) Boulard, Y.; Cognet, J. A. H.; Gabarro-Arpa, J.; Le Bret, M.; Sowers, L. C.; Fazakerley, G. V. Nucleic Acids
Res. 1992, 20, 1933.
(360) Wang, C.; Gao, H.; Gaffney, B. L.; Jones, R. A. J. Am. Chem. Soc. 1991, 113, 5486.
(361) Macaya, R. F.; Gilbert, D. E.; Malek, S.; Sinsheimer, J. S.; Feigon, J. Science 1991, 254, 270.
(362) Gao, X.; Patel, D. J. J. Am. Chem. Soc. 1988, 110, 5178.
(363) Dolinnaya, N. G.; Fresco, J. R. Proc. Natl. Acad. Sci. USA 1992, 89, 9242.
(364) Dolinnaya, N. G.; Braswell, E. H.; Fossella, J. A.; Klump, H.; Fresco, J. R. Biochemistry 1993, 32, 10263.
(365) Robinson, H.; van der Marel, G. A.; van Boom, J. H.; Wang, A. H.-J. Biochemistry 1992, 31, 10510.
(366) Liu, K.; Miles, H. T.; Frazier, J.; Sasisekharan, V. Biochemistry 1993, 32, 11802.
(367) Jaishree, T. N.; Wang, A. H.-J. Nucleic Acids Res. 1993, 21, 3839.
(368) Gehring, K.; Leroy, J.-L.; Guéron, M. Nature 1993, 363, 561.
(369) Leroy, J.-L.; Gehring, K.; Kettani, A.; Guéron, M. Biochemistry 1993, 32, 6019.
(370) Garcia, B.; Palacios, J. C. Ber. Bunsenges. Phys. Chem. 1988, 92, 696.
(371) Taylor, R.; Kennard, O. J. Mol. Struct. 1982, 78, 1.
(372) Jardetzky, C. D.; Jardetzky, O. J. Am. Chem. Soc. 1960, 82, 222.
(373) Angell, C. L. J. Chem. Soc. 1961, 504.
(374) Tsuboi, M.; Kyogoku, Y.; Shimanouchi, T. Biochim. Biophys. Acta 1962, 55, 1.
(375) Katritzky, A. R.; Waring, A. J. J. Chem. Soc. 1962, 1540.
(376) Lord, R. C.; Thomas, G. J. J. Spectrochim. Acta 1967, 23A, 2551.
(377) Clauwaert, J.; Stockx, J. Z. Naturforsch. 1968, 23B, 25.
(378) Wagner, R.; von Philipsborn, W. Helv. Chim. Acta 1970, 53, 299.
(379) Christensen, J. J.; Rytting, H.; Izatt, R. M. Biochemistry 1970, 9, 4907.
(380) Poulter, C. D.; Anderson, R. B. Tetrahedron Lett. 1972, 36, 3823.
(381) Dunn, D. B.; Hall, R. H. In Handbook of Biochemistry and Molecular Biology; G. D. Fasman, Ed.; C.R.C.
Press: Cleveland, Ohio, 1975; Vol. I; pp 65.
(382) Kremer, A. B.; Mikita, T.; Beardsley, G. P. Biochemistry 1987, 26, 391.
(383) Wempen, I.; Fox, J. J. J. Am. Chem. Soc. 1964, 86, 2474.
(384) Cushley, R. J.; Wempen, I.; Fox, J. J. J. Am. Chem. Soc. 1968, 90, 709.
(385) Ward, D. C.; Reich, E.; Stryer, L. J. Biol. Chem. 1969, 244, 1228.
(386) Cho, B. P.; McGregor, M. A. Nucleosides & Nucleotides 1994, 13, 481.
(387) Koyama, G.; Umezawa, H. J. Antibiotics 1965, A18, 175.
(388) Robins, R. K.; Townsend, L. B.; Cassidy, F.; Gerster, J. F.; Lewis, A. F.; Miller, R. L. J. Heterocycl. Chem.
1966, 3, 110.
(389) Chu, C. K.; Wempen, I.; Watanabe, K. A.; Fox, J. J. J. Org. Chem. 1976, 41, 2793.
(390) Chambers, R. W. Prog. Nucleic Acids Res. Mol. Biol. 1966, 5, 349.
(391) Smith, R. M.; Martell, A. E.; Chen, Y. Pure Appl. Chem. 1991, 63, 1015.
(392) Marzilli, L. G.; de Castro, B.; Solorzano, C. J. Am. Chem. Soc. 1982, 104, 461.
(393) Marzilli, L. G.; Kline, T. P.; Live, D.; Zon, G. In Metal-DNA Chemistry; T. D. Tullius, Ed.; American
Chemical Society: Washington DC, 1989; pp 119.
(394) InInteractions of Metal Ions with Nucleotides, Nucleic Acids, and Their Constituents; A. Sigel and H. Sigel
ed.; M. Dekker, Inc.: New York, 1996; Vol. 32 (Metal Ions in Biological Systems).
(395) Polak, M.; Plavec, J. Nucleosides & Nucleotides 1998, 17, 2011.
(396) Puglisi, J. D.; Wyatt, J. R.; Tinoco, I. J. Biochemistry 1990, 29, 4215.
(397) Legault, P.; Pardi, A. J. Am. Chem. Soc. 1994, 116, 8390.
(398) Legault, P.; Pardi, A. J. Am. Chem. Soc. 1997, 119, 6621.
(399) Singleton, S. F.; Dervan, P. B. Biochemistry 1992, 31, 10995.
(400) Tran-Dinh, S.; Guschlbauer, W.; Guéron, M. J. Am. Chem. Soc. 1972, 94, 7903.
(401) Remaud, G.; Zhou, X.-x.; Welch, C. J.; Chattopadhyaya, J. Tetrahedron 1986, 42, 4057.
(402) Schuster, I. I.; Roberts, J. D. J. Org. Chem. 1979, 44, 3864.
(403) Remaud, G.; Welch, C.; Zhou, X.-x.; Chattopadhyaya, J. Nucleosides & Nucleotides 1988, 7, 167.
(404) Markowski, V.; Sullivan, G. R.; Roberts, J. D. J. Am. Chem. Soc. 1977, 99, 714.
(405) Remin, M.; Darzynkiewicz, E.; Dworak, A.; Shugar, D. J. Am. Chem. Soc. 1976, 98, 367.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
156
(406) Froehler, B. C.; Wadwani, S.; Terhorst, T. J.; Gerrard, S. R. Tetrahedron Lett. 1992, 33, 5307.
(407) Manoharan, M., unpublished results.
(408) Ramzaeva, N.; Seela, F. Helv. Chim. Acta 1996, 79, 1549.
(409) Seela, F.; Thomas, H. Helv. Chim. Acta 1995, 78, 94.
(410) Buhr, C. A.; Wagner, R. W.; Grant, D.; Froehler, B. C. Nucleic Acids Res. 1996, 24, 2974.
(411) Hashimoto, H.; Nelson, M. G.; Switzer, C. J. Am. Chem. Soc. 1993, 115, 7128.
(412) Lesnik, E.
(413) Gryaznov, S.; Schultz, R. G. Tetrahedron Lett. 1994, 35, 2489.
(414) Sanghvi, Y. S.; Hoke, G. D.; Freier, S. M.; Zounes, M. C.; Gonzalez, C.; Cummins, L.; Sasmor, H.; Cook, P.
D. Nucleic Acids Res. 1993, 21, 3197.
(415) Dunkel, M.; Cook, P. D.; Acevedo, O. L. J. Heterocycl. Chem. 1993, 30, 1421.
(416) Imazawa, M.; Ueda, T.; Ukita, T. Chem. Pharm. Bull. (Tokyo) 1975, 23, 604.
(417) Uesugi, S.; Miki, H.; Ikehara, M.; Iwahashi, H.; Kyogoku, Y. Tetrahedron Lett. 1979, 42, 4073.
(418) Ikehara, M. Heterocycles 1984, 21, 75.
(419) Koole, L. H.; Wu, J.-C.; Neidle, S.; Chattopadhyaya, J. J. Am. Chem. Soc. 1992, 114, 2687.
(420) Xi, Z.; Glemarec, C.; Chattopadhyaya, J. Tetrahedron 1993, 49, 7525.
(421) Fraser, A.; Wheeler, P.; Dan Cook, P.; Sanghvi, Y. S. J. Heterocycl. Chem. 1993, 30, 1277.
(422) Hossain, N.; Papchikhin, A.; Garg, N.; Fedorov, I.; Chattopadhyaya, J. Nucleosides & Nucleotides 1993, 12,
499.
(423) Glemarec, C.; Reynolds, R. C.; Crooks, P. A.; Maddry, J. A.; Akhtar, M. S.; Montgomery, J. A.; Secrist, J. A.
I.; Chattopadhyaya, J. Tetrahedron 1993, 49, 2287.
(424) van Aerschot, A.; Everaert, D.; Balzarini, J.; Augustyns, K.; Jie, L.; Janssen, G.; Peeters, O.; Blaton, N.; de
Ranter, C.; de Clerq, E.; Herdewijn, P. J. Med. Chem. 1990, 33, 1833.
(425) Sanghvi, Y. S.; Vasseur, J.-J.; Debart, F.; Cook, P. D. Collect. Czech. Chem. Commun. 1993, 58, 158.
(426) Thibaudeau, C.; Chattopadhyaya, J. unpublished result
(427) de Leeuw, F. A. A. M.; van Wijk, J.; Altona, C. PSEUROT Version 5.4, Leiden University, Leiden,
Netherlands 1992,
(428) Hägele, G. DAISY-Program-Package for Simulation and Automated Analysis of High Resolution NMR
Spectra on Bruker's Aspect X32 and Silicon Graphics Indy computers 1991-1996.
(429) Altona, C.; Ippel, J. H.; Westra Hoekzema, A. J. A.; Erkelens, C.; Groesbeek, M.; Donders, L. A. Magn.
Reson. Chem. 1989, 27, 564.
(430) Altona, C.; Francke, R.; de Haan, R.; Ippel, J. H.; Daalmans, G. J.; Westra Hoekzema, A. J. A.; van Wijk, J.
Magn. Reson. Chem. 1994, 32, 670.
(431) Haasnoot, C. A. G.; De Leeuw, F. A. A. M.; Altona, C. Tetrahedron 1980, 36, 2783.
(432) Harte, W. E. J.; Starrett, J. E. J.; Martin, J. C.; Mansuri, M. M. Biochem. Biophys. Res. Commun. 1991, 175,
298.
(433) van Roey, P.; Salerno, J. M.; Chu, C. K.; Schinazi, R. F. Proc. Natl. Acad. Sci. USA 1989, 86, 3929.
(434) Sanghvi, Y. S.; Hanna, N. B.; Larson, S. B.; Robins, R. K.; Revanker, G. R. Nucleosides & Nucleotides 1987,
6, 761.
(435) Lankhorst, P. P.; Haasnoot, C. A. G.; Erkelens, C.; Altona, C. J. Biomol. Struct. Dynamics 1984, 1, 1387.
(436) Imai, K.; Osawa, E. Magn. Reson. Chem. 1990, 28, 668.
(437) For a recent review, see the contributions in Magn. Res. Chem. 1996, 34, S1-S186.
(438) Karplus, M. J. Chem. Phys. 1959, 30, 11.
(439) Huggins, M. L. J. Am. Chem. Soc. 1953, 75, 4123.
(440) Karplus, M. J. Am. Chem. Soc. 1963, 85, 2870.
(441) Barfield, M.; Smith, W. B. J. Am. Chem. Soc. 1992, 114, 1574.
(442) Marshall, J. L.; Walter, S. R.; Barfield, M.; Marchand, A. P.; Marchand, N. W.; Segre, A. L. Tetrahedron
1976, 32, 537.
(443) Jaworski, A.; Ekiel, I.; Shugar, D. J. Am. Chem. Soc. 1978, 100, 4357.
(444) de Leeuw, F. A. A. M.; van Beuzekom, A. A.; Altona, C. J. Comp. Chem. 1983, 4, 438.
(445) Diez, E.; San-Fabian, J.; Guilleme, J.; Altona, C.; Donders, L. A. Mol. Phys. 1989, 68, 49.
(446) Donders, L. A.; de Leeuw, F. A. A. M.; Altona, C. Magn. Reson. Chem. 1989, 27, 556.
(447) Plavec, J. Ph.D. Thesis, Department of Bioorganic Chemistry, Uppsala University, Sweden 1995.
(448) Blandin, M.; Son, T.-D.; Catlin, J. C.; Guschlbauer, W. Biochim. Biophys. Acta 1974, 361, 249.
(449) Blandin, M.; Jankowski, K. Rev. Latinoameric. Quim. 1983, 14, 45.
(450) Huang, J.-T.; Chen, L.-C.; Wang, L.; Kim, M.-H.; Warshaw, J. A.; Armstrong, D.; Zhu, Q.-Y.; Chou, T.-C.;
Watanabe, K. A.; Matulic-Adamic, J.; Su, T.-L.; Fox, J. J.; Polsky, B.; Baron, P. A.; Gold, J. W. M.; Hardy,
W. D.; Zuckerman, E. J. Med. Chem. 1991, 34, 1640.
(451) Heinemann, V.; Hertel, L. W.; Grindey, G. B.; Plunkett, W. Cancer Res. 1988, 48, 4024.
(452) Brox, L. W.; LePage, G. A.; Hendler, S. S.; Shannahoff, D. H. Cancer Res. 1978, 34, 1838.
(453) Tsuchiya, T. Adv. Carbohydr. Chem. Biochem. 1990, 48, 91.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
157
(454) Marquez, V. E.; Lim, B. B.; Barchi, J. J. J.; Nicklaus, M. C. In Nucleosides and Nucleotides as Antitumor
and Antiviral Agents; C. K. B. Chu, D.C., Ed.; Plenum Press: New York, 1993; pp 265. (455) Herdewijn, P.; Pauwels, R.; Baba, M.; Balzarini, J.; de Clerq, E. J. Med. Chem. 1987, 30, 2131. (456) Herdewijn, P.; van Aerschot, A.; Kerremans, L. Nucleosides & Nucleotides 1989, 8, 65. (457) Martin, J. A.; Brushnell, D. J.; Duncan, I. B.; Dunsdon, S. J.; Hall, M. J.; Machin, P. J.; Merret, J. H.; Parkes,
K. E. B.; Roberts, N. A.; Thomas, G. J.; Galpin, S. A.; Kinchington, D. J. Med. Chem. 1990, 33, 2137. (458) Kawasaki, A. M.; Casper, M. D.; Freier, S. M.; Lesnik, E. A.; Zounes, M. C.; Cummins, L. L.; Gonzalez, C.;
Cook, P. D. J. Med. Chem. 1993, 36, 831. (459) Monia, B. P.; Lesnik, E. A.; Gonzalez, C.; Lima, W. F.; McGee, D.; Guinosso, C. J.; Kawasaki, A. M.; Cook,
P. D.; Freier, S. M. J. Biol. Chem. 1993, 268, 14514. (460) Fazakerley, G. V.; Uesugi, S.; Izumi, A.; Ikehara, M.; Guschlbauer, W. FEBS Lett. 1985, 182, 365. (461) Phillips, L.; Wray, V. J. Chem. Soc. (B) 1971, 1618. (462) Hall, L. D.; Manville, J. F. Chem. Comm. 1968, 37. (463) Foster, A. B.; Hems, R.; Hall, L. D. Can. J. Chem. 1970, 48, 3937. (464) Gopinathan, M. S.; Narasimhan, P. T. Molec. Phys. 1971, 21, 1141. (465) Govil, G. Molec. Phys. 1971, 21, 953. (466) Ihrig, A. M.; Smith, S. L. J. Am. Chem. Soc. 1970, 92, 759. (467) Williamson, K. L.; Li Hsu, Y.-F.; Hall, F. H.; Swager, S.; Coulter, M. S. J. Am. Chem. Soc. 1968, 90, 6717. (468) Williamson, K. L. W.; Li, Y.-F.; Hall, F. H.; Swager, S. J. Am. Chem. Soc. 1966, 88, 5678. (469) Berthelot, J.-P.; Jacquesy, J.-C.; Lesivalles, J. Bull. Soc. Chim. Fr. 1971, 5, 1896. (470) Lankin, D. C.; Chandrakumar, N. S.; Rao, S. N.; Spangler, D. P.; Snyder, J. P. J. Am. Chem. Soc. 1993, 115,
3356. (471) Emsley, J. W.; Phillips, L.; V., W. In Progress in Nuclear Magnetic Resonance Spectroscopy; J. W. Emsley,
Feeney, J., Sutcliffe, L.H., Ed.; Pergamon Press: New York, 1977; Vol. 10(3/4); pp 229. (472) Abraham, R. J.; Cavalli, L. Molec. Phys. 1965, 9, 67. (473) Wilkinson, L. SYSTAT: The System for Statistics Systat Inc.: Evanston, IL, 1989. (474) Urata, H.; Shinohara, K.; Ogura, E.; Ueda, Y.; Akagi, M. J. Am. Chem. Soc. 1991, 113, 8174. (475) Urata, H.; Ogura, E.; Shinohara, K.; Ueda, Y.; Akagi, M. Nucl. Acids Res. 1992, 20, 3325. (476) Hashimoto, Y.; Iwanami, N.; Fujimori, S.; Shudo, K. J. Am. Chem. Soc. 1993, 115, 9883. (477) Doi, M.; Inoue, M.; Tomoo, K.; Ishida, T.; Ueda, Y.; Akagi, M.; Urata, H. J. Am. Chem. Soc. 1993, 115,
10432. (478) Joyce, G. F.; Visser, G. M.; van Boeckel, C. A. A.; van Boom, J. H.; Orgel, L. E.; van Westrenen, J. 1984,
310, 602. (479) Eichhorn, G. L. In Inorganic BiochemistryElsevier/North Holland: New York, 1973; pp 1210. (480) Buchanan, G. W.; Stothers, J. B. Can. J. Chem. 1982, 60, 787. (481) Dabrowiak, J. C.; Greenaway, F. T.; Longo, W. E.; van Husen, M.; Crooke, S. T. Biochim. Biophys. Acta
1978, 517, 517. (482) Buncel, E.; Boone, C.; Joly, H. Inorg. Chim. Acta 1986, 125, 167. (483) Shankar Nandi, U.; Wang, J. C.; Davidson, N. Biochemistry 1965, 4, 1687. (484) Kan, L. S.; Li, N. C. J. Am. Chem. Soc. 1970, 92, 4823. (485) Young, P. R.; Shankar Nandi, U.; Kallenbach, N. R. Biochemistry 1982, 21, 62. (486) Beveridge, S. J.; Walker, W. R. Aust. J. Chem. 1974, 27, 2563. (487) Sigel, H. In Metal-DNA Chemistry; T. D. Tullius, Ed.; American Chemical Society: Washington DC, 1989;
pp 159. (488) Frøystein, N. Å.; Sletten, E. J. Am. Chem. Soc. 1994, 116, 3240. (489) Simpson, R. B. J. Am. Chem. Soc. 1964, 86, 2059. (490) Kajiwara, T.; Ito, T. J. Chem. Soc. Chem. Comm. 1994, 1773. (491) Kajiwara, M.; Sugiyama, M.; Kimura, E. J. Chem. Soc. Chem. Comm. 1994, 1747. (492) Chu, G. Y. H.; Mansy, S.; Duncan, R. E.; Tobias, R. S. J. Am. Chem. Soc. 1978, 100, 593. (493) Eichhorn, G. L.; Clark, P. J. Am. Chem. Soc. 1963, 85, 4020. (494) Gruenwedel, D. W.; Cruikshank, M. K. Biochemistry 1990, 29, 2110. (495) Sarker, M.; Chen, F.-M. Biophys. Chem. 1991, 40, 135. (496) Verheyden, J. P. H.; Wagner, D.; Moffatt, J. G. J. Org. Chem. 1971, 36, 250. (497) Aurup, H.; Tuschl, T.; Benseler, F.; Ludwig, J.; Eckstein, F. Nucleic Acids Res. 1994, 22, 20. (498) Wells, P. R. Prog. Phys. Org. Chem. 1968, 6, 111. (499) Bratsch, S. G. J. Chem. Educ. 1985, 62, 101. (500) Sanderson, R. T. J. Chem. Educ. 1988, 65, 112. (501) Bratsch, S. G. J. Chem. Educ. 1988, 65, 223. (502) Marriott, S.; Reynolds, W. F.; Taft, R. W.; Topsom, R. D. J. Am. Chem. Soc. 1984, 49, 959. (503) Mullay, J. J. Am. Chem. Soc. 1984, 106, 5842. (504) Mullay, J. J. Am. Chem. Soc. 1985, 107, 7271. (505) Inamoto, N.; Masuda, S. Chem. Lett. 1982, 1003. (506) Inamoto, N.; Masuda, S. Chem. Lett. 1982, 1007.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
158
(507) Inamoto, N.; Masuda, S.; Tori, K.; Yoshimura, Y. Tetrahedron Lett. 1978, 46, 4547.
(508) Inamoto, N.; Masuda, S. Tetrahedron Lett. 1977, 3287.
(509) Bolton, P. H.; Kearns, D. R. Biochim. Biophys. Acta 1978, 517.
(510) Bolton, P. H.; Kearns, D. R. J. Am. Chem. Soc. 1979, 101, 479.
(511) Birnbaum, G. I.; Giziewicz, J.; Huber, C. P.; Shugar, D. J. Am. Chem. Soc. 1976, 98, 4640.
(512) Ducruix, A.; Riche, C.; Pascard, C. Acta Cryst. 1976, B32, 2467.
(513) Sundaralingam, M.; Rao, S. T. Int. J. Quantum Chemistry: Quantum Biology Symposium 1983, 10, 301.
(514) Williams, D. M.; Pieken, W. A.; Eckstein, F. Proc. Natl. Acad. Sci. USA 1992, 89, 918.
(515) Pieken, W. A.; Olsen, D. B.; Benseler, F.; Aurup, H.; Eckstein, F. Science 1991, 253, 314.
(516) Kulinska, K.; Laaksonen, A. J. Biomol. Struct. Dynamics 1994, 11, 1307.
(517) Rosemeyer, H.; Tóth, G.; Golankiewicz, B.; Kazimierczuk, Z.; Bourgeois, W.; Kretschmer, U.; Muth, H.-P.;
Seela, F. J. Org. Chem. 1990, 55, 5784.
(518) Sowers, L. C.; Fazakerley, G. V.; Kim, H.; Dalton, L.; Goodman, M. F. Biochemistry 1986, 25, 3983.
(519) Kaplan, N. O.; Ciotti, M. M.; Stolzenbach, F. E.; Bachur, N. R. J. Am. Chem. Soc. 1955, 77, 813.
(520) Suzuki, S.; Suzuki, K.; Imai, T.; Suziki, N.; Okuda, S. J. Biol. Chem. 1965, 240, 554.
(521) Suzuki, K.; Nakano, H.; Suzuki, S. J. Biol. Chem. 1967, 242, 3319.
(522) Dinglinger, F.; Renz, P. Hoppe-Seyler's Z. physiol. Chem. 1971, 362, 1157.
(523) LePage, G. A.; Junga, I. G. Molec. Pharmacol. 1967, 2, 37.
(524) LePage, G. A. Can. J. Biochem. 1968, 46, 655.
(525) Perry, A.; LePage, G. A. Cancer Res. 1969, 29, 617.
(526) Nakai, Y.; LePage, G. A. Cancer Res. 1972, 32, 2445.
(527) Tamaoki, T.; LePage, G. A. Cancer res. 1975, 35, 1015.
(528) Bobek, M.; Bloch, A. J. Med. Chem. 1975, 18, 784.
(529) Falke, D.; Range, K.; Arendes, J.; Müller, W. E. G. Biochim. Biophys. Acta 1979, 563, 36.
(530) Christensen, L. F.; Broom, A. D.; Robins, M. J.; Bloch, A. J. J. Med. Chem. 1972, 15, 735.
(531) Séquin, U. Experientia 1973, 29, 1059.
(532) Morvan, F.; Rayner, B.; Imbach, J.-L.; Chang, D.-K.; Lown, J. W. Nucleic Acids Res. 1987, 15, 4241.
(533) Morvan, F.; Rayner, B.; Imbach, J.-L.; Lee, M.; Hartley, J. A.; Chang, D.-K.; Lown, J. W. Nucleic Acids Res.
1987, 15, 7027.
(534) Aramini, J. M.; van de Sande, J. H.; Germann, M. W. Biochemistry 1997, 36, 9715.
(535) Townsend, L. B. In Handbook of Biochemistry and Molecular Biology; G. D. Fasman, Ed.; C.R.C. Press:
Cleaveland, Ohio, 1975; Vol. I; pp 271-401.
(536) Hanessian, S.; Pernet, A. G. Adv. Carbohydr. Chem. Biochem. 1976, 33, 111.
(537) Daves, G. D.; Cheng, C. C. Prog. Med. Chem. 1976, 13, 303.
(538) Hacksell, U.; Daves, G. D. Prog. Med. Chem. 1985, 22, 1.
(539) Daves, G. D. Acc. Chem. Res. 1990, 23, 201.
(540) Burchenal, J. H.; Ciovacco, K.; Kalaher, K.; O'Toole, T.; Kiefner, R.; Dowling, M. D.; Chu, C. K.; Watanabe,
K. A.; Wempen, I.; Fox, J. J. Cancer Res. 1976, 36, 1520.
(541) Cortese, R.; Kammen, H. O.; Spengler, S. J.; Ames, B. N. J. Biol. Chem. 1974, 249, 1103.
(542) Samuelsson, T.; Boren, T.; Johansen, T. I.; Lustig, F. J. Biol. Chem. 1988, 263, 13692.
(543) Polak, M.; Mohar, B.; Kobe, J.; Plavec, J. J. Am. Chem. Soc. 1998, 120, 2508.
(544) Davies, D. B.; Sadikot, H. Org. Magn. Reson. 1982, 20, 180.
(545) Lankhorst, P. P.; Haasnoot, C. A. G.; Erkelens, C.; Westerink, H. P.; van der Marel, G. A.; van Boom, J. H.;
Altona, C. Nucleic Acids Res. 1985, 13, 927.
(546) Plavec, J.; Chattopadhyaya, J. Tetrahedron Lett. 1995, 36, 1949.
(547) Dhingra, M. M.; Saran, A. Biopolymers 1982, 21, 859.
(548) Sundaralingam, M. Biopolymers 1969, 7, 821.
(549) Yokohama, S.; Inagaki, F.; Miyazawa, T. Biochemistry 1981, 20, 2981.
(550) Damha, M. J.; Ogilvie, K. K. Biochemistry 1988, 27, 6403.
(551) Blommers, M. J. J.; Nanz, D.; Zerbe, O. J. Biomol. NMR 1994, 4, 595.
(552) Hruska, F. E.; Mak, A.; Singh, H.; Sugar, D. Can. J. Chem. 1973, 51, 1099.
(553) Sarma, R. H.; Mynott, R. J.; Wood, D. J.; Hruska, F. E. J. Am. Chem. Soc. 1973, 95, 6457.
(554) Haasnoot, C. A. G.; de Leeuw, F. A. A. M.; de Leeuw, H. P. M.; Altona, C. Recl. Trav. Chim. Pays-Bas
1979, 98, 576.
(555) Uhlenbeck, O. C. Nature 1987, 328, 596.
(556) Rousse, B.; Puri, N.; Viswanadham, G.; Agback, P.; Glemarec, C.; Sandström, A.; Sund, C.; Chattopadhyaya,
J. Tetrahedron 1994, 50, 1777.
(557) Rousse, B.; Sund, C.; Glemarec, C.; Sandström, A.; Agback, P.; Chattopadhyaya, J. Tetrahedron 1994, 50,
8711.
(558) Ruffner, D. E.; Stormo, G. D.; Uhlenbeck, O. C. Biochemistry 1990, 529, 10695.
(559) Varani, G.; Cheong, C.; Tinoco, I. Biochemistry 1991, 30, 3280.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
159
(560) Maltseva, T. V.; Agback, P.; Repkova, M. N.; Venyaminova, A. G.; Ivanova, E. M.; Sandström, A.; Zarytova,
V. F.; Chattopadhyaya, J. Nucleic Acids Res. 1994, 22, 5590.
(561) Stein, C. A.; Subasinghe, C.; Shinozuka, K.; Cohen, J. S. Nucl. Acids Res. 1988, 16, 3209.
(562) Knudsen, H.; Nielsen, P. E. Nucl. Acids Res. 1996, 20, 494.
(563) Chou, S.-H.; Flynn, P.; Reid, B. Biochemistry 1989, 28, 2435.
(564) Katahira, M.; Lee, S. J.; Kobayashi, Y.; Sugeta, H.; Kyogoku, Y.; Iwai, S.; Ohtsuka, E.; Benevides, J. M.;
Thomas, G. J. J. J. Am. Chem. Soc. 1990, 112, 4508.
(565) Wang, A. H.-J.; Fujii, S.; van Boom, J. H.; van der Marel, G. A.; van Boeckel, S. A. A.; Rich, A. Nature
1982, 299, 601.
(566) Damha, M. J.; Wilds, C. J.; Noronha, A.; Brukner, I.; Borkow, G.; Arion, D.; Parniak, M. A. J. Am. Chem.
Soc. 1998, 120, 12976.
(567) Lesnik, E. A.; Guinosso, C. J.; Kawasaki, A. M.; Sasmor, H.; Zounes, M.; Cummins, L. L.; Ecker, D. J.;
Cook, P. D.; Freier, S. M. Biochemistry 1993, 32, 7832.
(568) Puri, N.; Chattopadhyaya, J. Tetrahedron 1999, submitted,
(569) Ding, D.; Gryaznov, S. M.; Lloyd, D. H.; Chandrasekaran, S.; Yao, S.; Ratmeyer, L.; Pan, Y.; Wilson, W. D.
Nucl. Acids Res. 1996, 24, 354.
(570) Gryaznov, S. M.; Lloyd, D. H.; Chen, J.-H.; Schultz, R. G.; DeDionisio, L. A.; Ratmeyer, L.; Wilson, W. D.
Proc. Natl. Acad. Sci. USA 1995, 92, 5798.
(571) Marquez, V. E.; Ezzitouni, A.; Russ, P.; Siddiqui, M. A.; Ford, H. J.; Feldman, R. J.; Mitsuya, H.; George,
C.; Barchi, J. J. J. J. Am. Chem. Soc. 1998, 120, 2780.
(572) Denisov, A. Y.; Bekiroglu, S.; Maltseva, T. V.; Sandström, A.; Altmann, K.-H.; Egli, M.; Chattopadhyaya, J.
J. Biochem. Biophys. Meth. 1998, 16, 569.
(573) Altmann, K.-H.; Kesselring, R.; Francotte, E.; Rihs, G. Tetrahedron Lett. 1994, 35, 2331.
(574) Marquez, V. E.; Siddiqui, M. A.; Ezzitouni, A.; Russ, P.; Wang, J.; Wagner, R. W.; Matteuci, M. D. J. Med.
Chem. 1996, 39, 3739.
(575) Agback, P.; Glemarec, C.; Sandström, A.; Yin, L.; Plavec, J.; Sund, C.; Yamakage, S.-i.; Viswanadham, G.;
Rousse, B.; Puri, N.; Chattopadhyaya, J. In Proceedings of the 8th Conversations; Adenine Press, New York:
1994; pp 293.
(576) Dalluge, J. J.; Hashizume, T.; Sopchik, A. E.; Closkey, J. A.; Davis, D. R. Nucleic Acids Res. 1996, 24, 1073.
(577) Nawrot, B.; Milius, W.; Ejchart, A.; Limmer, S.; Sprinzl, M. Nucl. Acids Res. 1997, 25, 948.
(578) Taiji, M.; Yokohama, S.; Miyazawa, T. J. Biochem. 1985, 98, 1447.
(579) Sprinzl, M.; Kucharzewski, M.; Hobbs, J. B.; Cramer, F. Eur. J. Biochem. 1977, 78, 55.
(580) Janiak, F.; Dell, V. A.; Abrahamson, J. K.; Watson, B. S.; Miller, D. L.; Johnson, A. E. Biochemistry 1990,
29, 4268.
(581) Potts, R. O.; Ford, N. C. J.; Fournier, M. J. Biochemistry 1981, 20, 1653.
(582) Limmer, S.; Vogtherr, M.; Nawrot, B.; Hillenbrand, R.; Sprinzl, M. Angew. Chem. Internat. Edit. 1997, 36,
2485.
(583) Polak, M.; Plavec, J. Eur. J. Inorg. Chem. 1999, 547.
(584) Isaksson, J.; Maltseva, T.; Agback, P.; Kumar, A.; Zamaratski, E.; Chattopadhyaya, J. 1999, Manuscript in
preparation.
(585) Fujiwara Nucl. Acids Res. Symposium Ser. 1997, 37, 51.
(586) Overmars, F. Ph.D. Thesis Thesis, Leiden University, The Netherlands, 1997.
(587) Mc Connell, H. M. J. Chem. Phys. 1958, 28, 430.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
160
(10.4) Latest articles (1998 – 2002) on the aspects of Anomeric (AE)
and Gauche (GE) Effects and discussions/comments on these works
(10.4.1) Review: Perrin, C. L. Acc. Chem. Res. 2002, 35, 28.
A detailed discussion on the stabilities of product formed on hydrolysis of five- and seven-
membered cyclic amides. Product studies showed that the substantial cleavage of the bond that is
antiperiplannar to only one lone pair of electrons and syn to the other with transition-state
stabilization of 2 kcal mol-1
.
(10.4.2) NMR titration studies (Perrin, C. L. J. Am. Chem. Soc. 1999, 121, 6911) showed the
absence of reverse AE (see also Kirby, A. J. J. Chem. Soc., Chem. Commun. 1998, 1695, Szarek, A.
J. Org. Chem. 2001, 66, 1097, Pinto, B. M. J. Org. Chem. 2000, 65, 220) in cationic N-
(glycopyranosyl)imidazoles and their tetra-O-acetyl derivatives (ΔΔGβ-α is negative, thereby
showing greater preference for the axial position of a protonated imidazolyl group than of an neutral
group).
(10.4.3) The net effect (both steric and stereoelectronic) of substitution (X, where X = OH, O-alkyl,
O-acetyl and F) in a hexapyranoside (J. Chem. Soc., Perkin Trans. 2, 2002, 337) indicates reduced
reactivity (in terms of the rate of acid-catalysed hydrolysis) at the anomeric center compared to the
parent tetrahydropyranyl acetal.
(10.4.4) Ab initio (MP2/6-31G*) studies of pseudorotation and conformational stabilities of
pyrrolidine (Carballeria, L.; Perez-Juste, I. J. Chem. Soc., Perkin Trans. 2, 1998, 1339) showed that
pseudorotation path is preferred for inter-conversion between the N-H axial and equatorial form
with a barrier ~0.6 kcal mol-1
supporting the experimental microwave spectroscopy results.
(10.4.5) Computational studies (Senyurt, N.; Aviyente, V. J. Chem. Soc., Perkin Trans. 2, 1998,
1463) by ab initio (HF/6-31G*) of AE in 2-[(4-substituted phenyl)seleno]-1,3-dithianes with NO2,
H, CH3, OCH3 and N(CH3)2 as substituents showed the preference for axial conformer with
enhanced electron withdrawing groups. NBO analyses showed that delocalization involving the
sulfur and selenium lone-pairs and the σ*C2-Se plays and important role in stabilizing the axial
conformer.
(10.4.6) NMR and ab initio studies (Serianni, A. S. J. Am. Chem. Soc. 1997, 119, 8933) for 2-
deoxy-β-D-glycero-tetrofuranose showed that it favors the S-form in solution (89% 4T3, 11% E2).
Protonated form showed exclusive preference for S-form (E3) with higher energy barrier than the
neutral form.
(10.4.7) Conformational studies with constrained nucleosides (Lowary, T. L. J. Org. Chem. 2001)
showed that furanose ring conformation in all compounds is locked either into E3 or 0E. It has also
been shown that a nucleoside containing a conformationally locked furanose ring does influence the
conformation of adjacent neighbor.
(10.4.8) Comparison of conformatinal analysis by theoretical studies (at various level with ab initio
and DFT methods) with that of experimentally measured by NMR for Methyl 3-O-Methyl-α-D-
arabinofuranoside (Lowary, T. L. J. Am. Chem. Soc. 2001, 123, 8811) has been performed.
Moreover, computational studies (at HF/6-31G* and B3LYP/6-31G*) of conformational analyses of
Methyl-α-D-arabinofuranoside (Lowary, T. L. J. Am. Chem. Soc. 1999, 121, 9682) showed 3E as the
lowest energy N-type conformer and either 2E or E1 (depending upon the level of theory used) as
lowest energy S-type conformer which is also the global minimum.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
161
(10.4.9) Conformational analyses of carba sugar analogues of methyl α-D-arabinofuranosides and
methyl β-D-arabinofuranosides have been reported (Lowary, T. L. J. Org. Chem. 2001, 66, 8961). It
showed that furanose ring conformation in these compounds is biased towards N-type.
(10.4.10) The magnitude of the one-bond coupling constant between C1 and H1 in 2,3-anhydro-O-
furanosides has been shown (Lowary, T. L. J. Org. Chem. 2001, 66, 4549) to be sensitive to the
stereochemistry at the anomric center. A panel of 24 compounds was studied and in case where
anomeric hydrogen is trans to the epoxide moiety, 1JC1-H1 = 163 – 168 Hz; whereas for cis
orientation of this hydrogen with respect to the oxirane ring, 1JC1-H1 = 171 – 174 Hz. However, for
2,3-anhydro-S-glycosides, the 1JC1-H1 is insensitive to the C1 stereochemistry.
(10.4.11) NMR and molecular modeling studies of 8-Aza-3-deazaguanine (Plavec, J. J. Chem. Soc.,
Perkin Trans. 2, 2001, 1433) showed stabilization of N-type conformers by ΔΔHo of 3.1 kJ mol
-1,
which has been attributed to the strengthening of O4'-C1'-N9 anomeric effect (strength of 19.5 kJ
mol-1
).
(10.4.12) NMR and ab initio studies (Plavec, J. J. Chem. Soc., Perkin Trans. 2, 2000, 255) showed
The S-C-N anomeric effect is stronger in purine than in pyrimidine 4'-thionucleosides (thymine <
cytosine < guanine < adenine), which is opposite to natural 4'-oxonucleosides. However, the
strength of AE in 4'-thionucleosides is weaker compared to that in natural nucleosides.
(10.4.13) Steric and stereoelectronic effects of 7- and 8-substituted 7-deaza-2'-deoxy-adenosine and
-guanosine on the sugar pucker as well as conformation about C4'-C5' bond have been studied
(Seela, F. J. Chem. Soc., Perkin Trans. 2, 1997, 2341): (i) higher electron-attracting effect of the
substituents drives N/S equilibrium of the pentofuranose towards N-type conformation and (ii)
higher electron-withdrawing effect of the 7-substituents, higher γ+ across C4'-C5'.
(10.4.14) Conformational analysis of 2-halocyclohexanones by NMR, theoretical and solvation
studies (Yoshinaga, F. J. Chem. Soc., Perkin Trans. 2, 2002, 1494) showed axial conformation of 2-
fluoro compound in vapour phase is stabilized (ΔEeq – ax = 0.45 kcal mol-1
) whereas equatorial
conformer predominates for other haloketones (ΔEeq – ax = 1.05, 1.50 and 1.90 kcal mol-1
for chloro,
bromo and iodo compounds respectively).
(10.4.15) Detailed MO calculations of cyclohexane, 1,3-Dioxane, 1,3-Oxathiane and 1,3-Dithiane
showed (Alabugin, I. V. J. Org. Chem. 2000, 65, 3910) the importance of hyperconjugative
interaction (through the balance of three effects: σC-X → σ*C-Heq, σ*C-Heq → σC-X and nP → σ*C5-Heq
interactions) to explain the relative elongation of equatorial C5-H bonds. The role of nP → σ*C5-Heq
interaction is important in dioxane. In diathiane, distortion of the ring by long C – S bonds increases
preferential overlap of σC5-Heq and σ*C-S orbitals.
(10.4.16) Recently a report has appeared (S.F. Wnuk, D.R. Companioni, V.Neschadimenko, and
M.J. Morris, J. Org. Chem. (Web: http://dx.doi.org/10.1021/jo020428b) on the competition
between steric and -fluorine effects, including the impact of fluorine regio- and stereochemistry, on
radical deoxygenations of 2'(3')-O-phenoxythiocarbonyl (PTC) esters of fluoropentofuranosyl-
adenine nucleosides. Thus it has been found that steric effects are decisive for determination of the
stereoslectivity of transfer of deuterium from tributyltin hydride to fluoropentofuranosyl radicals
generated from these adenine nucleoside derivatives. In all cases, deuterium abstraction occurs at
the less hindered α face of the sugar ring trans to the heterocyclic base. However, this α face
stereoselectivity is enhanced by the anti effect of a vicinal fluorine substituent with an arabino or
xylo orientation (on the α face of the ring). A smaller anti effect is still apparent with a vicinal
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
162
fluorine on the α face (ribo orientations). Complex stereoelectronic/steric interactions might be
involved with these furanose rings that have electronegative (F, O, N) substituents.
(10.5) The sensitivity of the RNase H discriminates the local structure changes owing to
conformational transmission induced by 3'-endo sugar constrained nucleotides in the
antisense strand of the antisense-RNA hybrid duplex.
RNase H an ubiquitous enzyme cleaves the RNA in a DNA/RNA hybrid. Based on the
dissociation constant determined by competitive inhibition analysis, the RNase H can bind to all
duplexes, irrespective of their conformational preorganization [DNA/RNA > RNA/RNA >
DNA/DNA]. The catalytic cleavage by the enzyme however demands a flexible hybrid duplex
structure whose overall geometry should be close to an A type helix. For example, in the native
DNA-RNA duplex, the conformation of the DNA strand is B type (all nucleotides are in O4’-endo
conformation) whereas the RNA strand adopts a conformation, which is very similar to single
stranded A-RNA type (all nucleotides are in C3’-endo conformation). It has emerged that the RNase
H cleavage site retains this B-type-DNA/A-type RNA conformation in order to be substrate for
cleavage reaction by RNase H. This conformational criteria has been so far difficult to achieve with
the modified residues in the antisense strand at the cleavage site. We have thus witnessed the
emergence of the gapmer or the mixmer strategies, which not only allows the cleavage of the
complementary RNA strand, but also enhances the thermodynamic stability of the hybrid duplex.1
The tolerance of the modified nucleotide residues in the antisense strand in the above two strategies
are very sensitive to the type of modifications used, both because of conformational pre-
organization and the intrinsic flexibility required of the hybrid duplex for catalytic cleavage by
RNase H.
Most of the North conformationally constraind nucleosides upon introduction (full
modification or partial modification) to Antisense OligoNucleotide (AON) render their hybrid
duplex with RNA (AON/RNA) to a rigid RNA/RNA type duplex, which were found to be
insensitive towards the RNase H promoted cleavage, although the stability of the resulting hybrid
duplexes are enhanced1. The reasons for this RNase H insensitivity to conformationally-constrained
containing antisense complex with complementary RNA (both in the gapmer as well as in the
mixmer strategies) is beginning to be understood. Remarkably, some conformational alterations
brought about by the modifications is transmitted to the neighbouring nucleotides, which enzyme
can sense, but spectroscopic techniques such as NMR or CD can not detect2. A slow consensus is
however emerging as to how far the rigidity of the N-type conformationally constrained nucleotides
in the AON can propagate to alter the conformational characters of the neighboring S-type residues
in the hybrid duplex. This information is important because the alteration of S-type character in the
neighbors to an N-type forces those nucleotides in the AON/RNA duplex to adopt RNA/RNA type
conformation, which prevents it from the RNase H cleavage. The uniquness of RNase H promoted
cleavage of AON/RNA hybrids holding varying number of conformationally constrained North-East
(N/E)-type nucleotides, can be employed to address these issue which we have shown recently with
the oxetane incorporated nucleotides in the antisense-RNA hybrid duplex 2.
We have introduced single, double, triple N/E conformationaly constrained nucleoside, [1-
(1',3'-O-anhydro-β-D-psico-furnosyl)thymine] (T), to a 15 mer AON and targeted to a 15 mer RNA.
CD failed to detect any structural perturbances of the modified hybrids compared to the native
counterpart; however, the RNase H cleavage pattern clearly showed that the local conformational
changes spanning a total of 5 nucleotides including the modification towards the 5'-end of AON (3'-
end of RNA). This was evident from the fact that the 5 nucleotide region of the RNA strand,
beginning from the nucleotide opposite to the T modification, became completely inactive to the
catalytic cleavage reaction by RNase H2. Since the site just after the 5 nucleotides were accessible
for enzymatic cleavage (Figure 10), and the fact that the binding and cleavage sites are different for
RNase H3, it is evident that the structurally altered duplex region was suitable for enzyme binding
but not for cleavage. By suitably placing the just three T modifications we have shown that the
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
163
single cleavage site can be engineered in the 15mer AON/RNA hybrid duplex (Figure 1). This work
clearly shows that owing to the N-type or N/E-type conformational constrain introduced at the
oligonucleotide level, leads to microenvironmental conformational alterations in the neighboring 4
nucleotides toward the 5’-end. This conformational transmission can be effectively mapped by
RNase H, when CD and NMR fails to show any conformational heterogeinity in the local structure.
There is only limited information available regarding the tolerance of RNase H towards local
conformational transmission of AONs holding other conformationally constrained nucleotides. A
recent report4 outlined the RNase H tolerence of AON/RNA hybrids modified with North-
conformationally constrained LNA monomers. Introduction of one LNA itself was found to alter
global helical structure5 (detected by CD) and three LNA monomers were sufficient to create a
RNA/RNA type AON/RNA hybrid duplex. None of these modifications are however reported to
elicit any RNase H response, although considerable enhnacement in the thermodynamic stability
was observed. Mixmer AONs with 3-5 deoxynucleotide gaps were found to be also insensitive
towards RNAse H promoted cleavage once hybridized to RNA4. This clearly shows that
microstructural alterations brought by the LNA modification propogates beyond 5 nucleotids in
contradistinction to the oxetane modifications.
An important question in this context is how far down the polynucleotide chain the
conformational transmission propagates in the AON/RNA hybrid duplex2? Clearly, an appropriate
answer should make it possible to use of the LNAs both as Tm enhancer and a substrate for RNAse
H, as it has been achieved in a recent systematic study with a series of LNA incorporated AON-
RNA duplexes in which the size of the deoxynucleotide gap (required for RNase H cleavage) in the
mixmer as well as in the gapmer has been varied5. This study has shown that the gapmer/mixmer of
LNA/RNA duplexes having 6 - 10 deoxynucleotide gaps was cleaved by RNase H. Quite expectedly2,
it was also found that the maximum cleavage efficiency was observed as the size of the
deoxynucleotide gap was increased to 10, presumably because it adopted a substantial degree of
DNA/RNA type structure5. This shows that the conformational transmission of LNA stretches upto 7
nucleotides to make it conform to RNA/RNA type structure, which is resistant to RNase H, whereas
for oxetane modifications, we have shown it stretches upto 5 nucleotides2
2’-O-methylnucleosides (3J1’2’ = 5.2 Hz,
3J3’4’ = 4.7 Hz,) show a slightly preferred N-type
conformation in the North-South pseudorotational equilibrium. This slight preference for N-type
conformation is attributed to the steric repulsion between the aglycone and the 2’-O-methyl group in
the C2’-endo (South form)6. In the oligomer there would be additional steric clash of 2’-O-alkyl
group with the 3’-phosphate, which drives it to more North conformation. Therefore, unlike the
other North-constrained nucleotides (LNA, oxetane), the conformational influence by 2’-O-alkyl
nucleotides, which can be detected by RNase H, to the neighbours are less pronounced, and it was
found to be sequence dependent7. This is clearly visible in the RNase H digestion pattern of 2’-O-
alkyl modified AON/RNA gapmer hybrids, where the microstructural alterations brought by the
modification, which blocks the cleavage was found to vary from 3 to 5 nucleotides including the
modification7. This means that the conformational transmission of the North-type 2’-O-
alkylnucleotides in the AON/RNA/RNase H ternary complex is less pronounced than in other
conformational constrained counterparts. Unlike the North constrained nucleotides, the
conformational status of 2’-O-methylnucleosides and its analogs in the antisense strand or in the
hybrid duplex with RNA can be envisioned as sensitive to both sequence make-up as well as
whether they are in the complex form with RNase H or not. It has been recently shown that the
modified AONs containing acyclic interresidue units (the 2’,3’-secouridine or a butanediol linker in the
modified AONs containing 2’F-ANA) supports RNase H-promoted cleavage of complementary RNA
(Damha et al., J. Am. Chem. Soc. 125, 654-661 (2003). These conformationally labile AONs shows
some remarkable benefits in the enzymatic recognition of the AON/RNA hybrids, because of the
flexible nature of their sugar-phosphate backbone conformation, which is consistent with our
observation of the oxetane-modified AON/RNA recognition and cleavage by RNase H. It should be
noted that the introduction of a 2’,3’-secouridine moitey reduces the Tm by ca 10ºC whereas a butanediol
linker reduces the Tm by ca 6ºC.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
164
(A) AON (1) AON (2) AON (4) AON (5)AON (3) AON (6)
5'-r(G A A G A A A A A A U G A A G)-3'3'-d(C T T C T T T T T T A C T T C)-5'
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
5'-r(G A A G A A A A A A U G A A G)-3'3'-d(C T T C T T T T T T A C T T C)-5'
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
11
1213
5'-r(G A A G A A A A A A U G A A G)-3'3'-d(C T T C T T T T T T A C T T C)-5'
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1112
5'-r(G A A G A A A A A A U G A A G)-3'3'-d(C T T C T T T T T T A C T T C)-5'
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
12
5'-r(G A A G A A A A A A U G A A G)-3'3'-d(C T T C T T T T T T A C T T C)-5'
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
8
5'-r(G A A G A A A A A A U G A A G)-3'3'-d(C T T C T T T T T T A C T T C)-5'
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
8
7
AON (2)
AON (3)
AON (4)
AON (5)
AON (6)
8 1011
12139765
AON (1)
9
10
7
Figure 10. (A) The PAGE analysis of RNase H hydrolysis of the hybrid duplexes, AON (1)-
(6) / RNA. Time after the
addition of the enzyme is shown on the top of each gel lane. The length and sequence of 5’-32
P-labeled RNAs formed up
on enzymatic cleavage are shown on the left and right side of the gel and it was deduced by comparing the migration of
products with those oligonucleotides generated by parial digetion of the target RNA by Snake venom phosphodiesterase
(SVPDE). (B) RNase H cleavage cleavage pattern of the hybrid duplexexs. Long and short arrows represents major and
minor sits respectively (after 2h of incubation). Boxes represents the parts of the RNA sequence insensitive towards
RNase H cleavage.
References:
1. (a) M. Manoharan, Biochim Biophys Acta., 1489, 117(1999). (b) P. Herdewijn, Biochim
Biophys Acta., 1489, 167(1999).
2. (a) P. I.Pradeepkumar, E. Zamaratski, A. Földesi, J. Chattopadhyaya. Tetrahedron Lett., 41,
8601 (2000). (b) P. I. Pradeepkumar, E. Zamaratski, A. Földesi, J. Chattopadhyaya. J. Chem.
(B)
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
165
Soc., Perkin Trans. 2, 3, 402 (2001). (c) P. I. Pradeepkumar, J. Chattopadhyaya. J. Chem. Soc.,
Perkin Trans. 2, 11, 2074 (2001).
3. S. Kanaya, Enzymatic activity and protein stability of E. coli ribonuclease HI, in
Ribonucleases H, Crouch, R. J., editors; INSERM: Paris, Chapter 1, pp. 1-37 (1998).
4. J. Kurreck, E. Wyszko, C. Gillen, V. A. Erdmann, Nucleic Acids Res., 30, 1911(2002).
5. K.Bodensgaard, M. Petersen, S.K. Singh, V.K. Rajwanshi, R. Kumar, J. Wengel, J.P.
Jacobsen. Chem. Eur. J, 6, 2687 (2000).
6. G. Kawai, Y. Yamamoto, T. amimura, T. Masegi, M. Sekine, T. Hata, T. Iimori,
T.Watanabe, T. Miyazawa, S. Yokoyama, Biochemistry, 31, 1040 (1992).
7. (a) H. Inoue, Y. Hayase, S. Iwai, E. Ohtuka., FEBS Lett., 215, 327 (1987). (b) Y-T.Yu, M-D.
Shu, J. A. Steiz, RNA., 3, 324 (1997). (c) B.Larrouy, C. Boiziau, B.Saproat, J-J.Toulme,
Nucleic Acids Res., 23, 3434 (1995).
(10.6) The influence of flouro substitution at the sugar moiety in modulating the furanose
conformation of flourinated nucleosides.
The furanose conformation of a nucleoside is determined by interplay of stereoelectronic
gauche and anomeric effects (vide infra) tuned by different sugar substituents. The strength of the
gauche and anomeric effects is determined by the electronic nature of the nucleobase, and the
electronegetivity, position, configuration of substituents in the furanose moiety. Thus highly
electronegative flouro substitution in the sugar moiety can predominantly drive its overall North
(N)� South (S) equilibrium. The various flourinated nucleosides have therefore been designed and
analysed for their ability to conformationally preorganize the furanosyl moiety in solution.
Marquez et al. have performed a systematic study of various fluorinated dideoxy uridines by
NMR and pesudorotational analysisis1. In 2'-flouro-2', 3'-didexy-ara-uridine (2'F-up, 1) and 3'-
flouro-2', 3'-didexy uridine (3'F-down, 4) the F-gauche effect dominates over the opposing O4'-C1
'-
N anomeric effect, thereby the furanose ring adopts the South.type conformation in solution.
However in the 2'F-down (2) and 3'F–up (3) dideoxy uridines, the F-gauche effect along with the
anomeric effect plays in tandem to lock the sugar conformation in to the Nothern hemisphere (N-
type) of the pseudorotational cycle. It is worth noting that the 2'F-down dideoxy uridine (2)
exclusively prefers the N-type conformation at 25 oC with a pseudorotatonal phase angle of 19
o.
In the case of the 2'-flouro-2'-deoxyarabinoribofuranosyl thymine (2'-F up, 5) the gauche
effect of O4'-C1
'-C2'-F and F-C2
'-C1
'-N fragments prevails over the anomeric effect and the sugar
adopts the C2'-endo/ C1'-exo conformation in solution.
2,3 The 2
'-flouro-2'-deoxyribofuranosyl
thymine (2'F-down, 6) prefers the N-type sugar conformation due to the combined influence of
gauche and anomeric effects. However, in the X-ray and NMR structure of the oligonucleotide
hybrid duplexes containing the 2'F-up thymidine units showed the O4'-endo conformation for the
furanose ring.4-6
This is explained on the basis of steric clash between the 2'F atom and the C6-
carbon in the South sugar.
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications", Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
166
O
H3'' H''
HO
NH
N
O
O
FH3'
O
H3'' F
HO
NH
N
O
O
H3'H2'
O
H3'' H2''
HO
NH
N
O
O
FH
'O
F H2''
HO
NH
N
O
O
H3'H2'
J3'-4' = 5.2 Hz
J3''-4' = 8.6 Hz
J1'-2'' = 3.3Hz
P = 131o
South -type
J3'-4' = 11.3 Hz
J3''-4' = 4.8 Hz
J1'-2' = 0.9 Hz
P = 19o
North -type
J3''-4' = 2.4 Hz
J1'-2' = 2.1 Hz
J1'-2'' = 8.2 Hz
P = 27o
North -type
J3'-4' = 1.15 Hz
J1'-2' = 9.0 Hz
J1'-2'' = 5.7 Hz
P = 169o
South -type
Barchi, J.J.; Jeong, L.K.; Siddiqui, M.A.; Marquez, V.E. J. Biochem. Biophys. Methods.1997, 34, 11.
2', 3' monoflorinated dideoxyuridines
1 2 34
O
HO H2''
HO
NH
N
O
O
F
J3'-4' = 4.5 Hz
J1'-2'' = 4 Hz
South -type
5
O
HO F
HO
NH
N
O
O
J3'-4' = 11.8 Hz
North -type
6
2' monoflorinated thymidines
O
HO
HO
F
N
NN
N
NH2
O
HO F
HO
N
NN
N
NH2
O
H3'' F
HO
N
NN
N
NH2
O
H3''
HO
F
N
NN
N
NH2
H3'
Ikeda, H.; Fernandez, R.; Wilk, A.; Barchi, J.J.;
Huang, X.; Marquez, V.E. Nucleic. Acids. Res.
1998, 26, 2237.
Wilds, J.C.; Damha, M.J. Nucleic. Acids. Res.
2000, 28, 3625.
H3'
South -type
P = 130o
North -type
J1'-2' = 2.82 Hz
P = 0o
South -type North -type
J1'-2' = 3.3Hz
P = 130o
J1'-2' = 0 Hz
P = 0o
2' monoflorinated deoxy and dideoxy adenosines
Ford, H.; Lan Mu, F.D.; Siddiqui, M.A.; Nicklaus, M.C.; Anderson, L.; Marquez, V.E.; Barchi, J.J.
Biochemistry. 2000, 39, 2581.
O
HO
HO
F
6 7 89
10
NH
N
N
O
NH2N
Br
O
HO
HO
F
N
N
N
NH2
NH2N
Br
11
2' monoflorinated deoxy 3-bromopyrazolo[3,4,-d] pyramidines
J3'-4' = 7.2 Hz
J1'-2'' = 6.3 Hz
J3'-4' = 7.5 Hz
J1'-2'' = 5.98 Hz
North -type North -type
P = -2.1o
P = -2.1o
He, J.; Mikhailopulo, I.; Seela, F. J. Org. Chem. 2003, 68, 5519.
To unravel the structural preference of North-constrained deoxy and dideoxy 2'F-adenosine
at the active site of adenosine deaminase, a detailed conformational analysis of various fluorinated
adenosine analogues has been reported recently.7 In the case of 2'-flouro-2',3'-
Chattopadhyaya et al, "Stereoelectronic Effects in Nucleosides & Nucleotides and their Structural Implications",
Dept of Bioorganic Chemistry, Box 581, Uppsala University, S-75123 Uppsala, Sweden, Ver 160205 [email protected]
167
dideoxyarabinofuranosyl adenosine (F-up, 8) and in the corresponding ribo analogue (F-down, 9)
only O4'-C1
'-C2'-F is operative along with the anomeric effect. In 8 this F-.gauche effect
predominates over the anomeric effect and the sugar adopts the C1'-exo conformation (P =130o,
81% S-type). The assistance of anomeric effect in 9 locks the sugar conformation in to an exclusive
N-type (P= 0o, 99% N-type). The situation is more complex in the case of 2'-flouro -2'-
deoxyarabinofuranosyl adenosine (6, F-up) and 2'-flouro -2'-deoxyribofuranosyl adenosine (7, F-
down) owing to the presence of 3'-OH. Here the additional O4'-C4
'-C3'-O3' gauche effect plays a
crucial role in dictating the sugar conformation. In 2'F-down compound 6 the 2'F -gauche effect and
anomeric effect overcome the O4'-C4
'-C3'-O3' gauche effect and the major pesudorotomer adapts
the N-type conformation with 76% population. While in 2'F-up compound 7 the sugar adapts
predominantly the C1'-exo pseudorotomer owing to the dominal influence of gauche over anomeric
effect (P =130o, % of S-type = 64%)
Sofar the studies on the stereoelectronic effects in nucleosides and nucleotides showed that 2' (or
2") and/or 3' (or 3") F/O mediated GE is stronger than AE. A recent report from Seela's group showed,8
for the first time, that the AE can be stronger than the F/O mediated GE in dictating the furanose
conformation of the nucleosides. Thus, they have shown that they could overcome the strong GE of 2'-F
substituent in the β-face (ara configuration) by the introduction of the extra nitrogen atom at the 6 position
of a pyrimidine moity (6-azauracil-1-yl moiety) or at the 8 position of the purine moiety [3,4-d]pyrimidine-
1-yl aglycone (3-bromo or 3-H gave almost identical N/S population). The two nucleosides, 3-
bromopyrazolo [3,4-d]pyrimidine-2'-deoxy-2"-flouro and its ara counterpart 10 and 11 showed exclusive
N-type conformation in solution and solid state (with pseudorotational phase angle –2.10),8 which is
unexpected in view of the earlier studies with 2'(β)-flouro substituted nucleosides. The strong electron
wihdrawing nature of nucleobase in these nuclosides enhances the nO4' → σ*C1'-N1 interaction
(anomeric effect) and it prevails over the gauche interactions results in the existence of C3'-endo
(N-type) pseudorotomer in solution.
References:
1. Barchi, J.J.; Jeong, L.K.; Siddiqui, M.A.; Marquez, V.E. J. Biochem. Biophys.
Methods.1997, 34: 11.
2. Ikeda, H.; Fernandez, R.; Wilk, A.; Barchi, J.J.; Huang, X.; Marquez, V.E. Nucleic. Acids.
Res.1998, 26, 2237.
3. Wilds, J.C.; Damha, M.J. Nucleic. Acids. Res.2000, 28, 3625.
4. Berger, I.; Tereshko, V.; Ikeda, H.; Marquez, V.E.; Egli, M. Nucleic. Acids. Res.1998, 26,
2473.
5. Trempe. J.F.; Wilds, C.J.; Denisov, A.Y.; Pon, R.T.; Damha, M.J.; Gehring, K. J. Am.
Chem. Soc.2001, 123, 4896.
6. Denisov, A.U.; Noronha, A.M.; Wilds, C.J. Trempe, J-F.; Pon, R.T.; Gehring, K.; Damha,
M.J. Nucleic. Acids. Res. 2001, 29, 4284.
7. Ford, H.; Lan Mu, F.D.; Siddiqui, M.A.; Nicklaus, M.C.; Anderson, L.; Marquez, V.E.;
Barchi, J.J. Biochemistry. 2000, 39, 2581.
8. He, J.; Mikhailopulo, I.; Seela, F. J. Org. Chem. 2003, 68, 5519.