abstract - open access repository of indian...
TRANSCRIPT
Abstract
i
The thesis entitled “Uses of Pyrrole Amino Acid based Scaffolds in Peptidomimetics
and Studies Directed Towards the Total Synthesis of Antascomicin A”, consists of
three chapters:
CHAPTER I: Deals with the synthesis and conformational studies of pyrrole amino acid
based peptidomimetics.
This chapter is divided into two parts:
Part A: Describes the synthesis and conformational studies of linear peptides of pyrrole
amino acid.
Part B: Describes the synthesis and structural studies of pyrrole amino acid containing
peptides.
CHAPTER II: Delineates the synthesis and DNA binding properties of pyrrole amino
acid containing peptides.
CHAPTER III: Describes the Studies Directed Towards the Total Synthesis of
Antascomicin A.
CHAPTER I
Part A
Synthesis and Conformational Studies of Linear Peptides of Pyrrole Amino Acid.
Biological activities of Peptides and proteins depend on their three dimensional
structures in the free state and when interacting with their receptors/acceptors. The
search for relationships between structure and activity has been the principal subject for
designing the selective drugs with no side effects. Initial efforts at finding a connection
Abstract
ii
between structure and activity equated structure with constitution, but it has soon become
evident that configuration was at least of equal importance.
Dehydro amino acids (∆Xaa) are found in many naturally occurring peptide
antibiotics of bacterial origin as well as in some proteins. They have also been used very
extensively as rigid scaffolds in structural studies of peptides. Synthesis of a novel
building block based on a pyrrole amino acid (Paa), 5-(aminomethyl)pyrrole-2-
carboxylic acid (1), its application in peptidomimetic studies as a conformationally
constrained surrogate of the Gly-∆Ala dipeptide residue (2) and the synthesis and
conformational studies of peptides 3 and 4 prepared from this template (Fig-1).
One of the most widely used protocols to restrict conformational degrees of freedom
in small peptides is by covalently stitching the “loose ends” together to fix various torsional
angles at values characteristic of the target structure. Bridging the terminal carbon of an
olefinic analogue of a Gly unit and the Cβ atom of an adjacent ∆Ala residue (2) leads to its
constrained version 1 with fixed ωi, ϕi+1 and probably, ψi+1 torsional angles; 1 was expected
to nucleate interesting structures in peptides 3 and 4.
The target peptidomimetics 3 and 4 were prepared from the pyrrole amino acid (Paa)
which in turn was prepared from pyrrole (Scheme 1)
3 & 4NH
COOMeBocHN
Scheme 1
5 6
NH
BocHNNH
HN
NH
HN
OMe
O
O
O
O
HN
NH
HN
NH
O
O
O
OMeBocHN
O
3
4
H2NN
HN
N
O
O
OH
OH
H
ωiφi+1
2
1
ψi+1
Gly ∆Ala
Figure 1
Abstract
iii
The synthesis of compound 5, i.e. the Boc-protected methyl ester of 1, was started
from pyrrole 6. The 5-formyl pyrrole-2-carboxylic acid methyl ester was prepared from 6 in
three steps following known procedures. A Vilsmeir-Haack reaction on pyrrole 6 furnished
2-formyl pyrrole 7, which was then subjected to Ag2O oxidation followed by treatment with
CH2N2 to give the 2-carbomethoxy group 9 (Scheme 2).
Second Vilsmeir-Haack reaction on pyrrole -2-carboxylic acid methyl ester 9
formylated 5-position 10 (49%) and 4-position 11 (14%). The compound methyl 5-formyl-
pyrrole-2-carboxylate 10 was reduced using NaBH4 in methanol to give an alcohol 12 in
92% yield (Scheme 3).
The alcohol 12 was then treated with CBr4, NaN3 & Ph3P in DMF at ambient
temperature to furnish the desired azido compound 13 in 60% yield. Our next objective
+NH
COOMeOHCNH
COOMe
OHC
POCl3, DMF,CH2-Cl
CH2-Cl
Reflux, 90 min
3.5 : 1
10 11
NaBH4, MeOH, 0 oC, 10 min
92%NH
COOMeHO
Scheme 3
9
10
12
POCl3, DMF,
Ag2O(Alkaline) MeOH:H2O(1:1) R.T, 1h
CH2N2, Ether, 0 oC
NH
NH
CHO NH
COOH
78%1h
67
10 min
85% in two stepsNH
COOMe
Scheme 2
8
9
CH2-Cl
8
CH2-Cl
Reflux
Abstract
iv
was conversion of azide group to primary amine. To achieve this, compound 13 was
treated with Ph3P in MeOH at room temperature. The resulting primary amine was in-
situ protected with Boc2O to furnish 5 in 60% yield (Scheme 4).
Compound 5 was subjected to alkaline hydrolysis using LiOH.H2O in THF-
MeOH-H2O (3:1:1) solution to furnish the carboxylic acid 14. The acid 14 was then
coupled with N-terminal free peptide TFA.H2N-Ala-Leu-Gly-OMe 16 under standard
solution phase peptide coupling using EDCI and HOBt as coupling agent and DIPEA as
base in dry CH2Cl2 solution to afford tripeptide Boc-Paa-Ala-Leu-Gly-OMe 3 in 65%
yield (Scheme 5).
TFA, CH2Cl2
NH
HN
O
COOMeTFA.H2N
O
i) EDCI, HOBt, CH2Cl2, 0 oC
then 16
DIPEA, CH2Cl2, 0 oC-R.TNH
HN
O
COOMeHN
O
3
NH
BocHN
O
5
LiOH.H2O
THF-MeOH-H2O
0 oC-rt, 48 h
NH
BocHN
O
OH
14
15
0 oC-R.T, 1 h
14
16
65%, 8 h
Scheme 5
i) Ph3P, MeOH, R.T, 3 h
ii) Boc2O, MeOH, R.T, 1 h
60%
NH
COOMeBocHN
12
Scheme 4
CBr4, NaN3, Ph3P
DMF, 0 oC-R.T, 1h
60%
NH
COOMeN3
13
13
5
Abstract
v
The acid Boc-Paa-OH 14 was coupled to TFA.H2N-Leu-Gly-OMe 17 under standard
peptide coupling conditions desired above to furnish the compound 18 Boc-Paa-Leu-Gly-
OMe in 73% yield. Boc deprotection of 18 using trifluoro acid (TFA) in CH2Cl2 solution
gave TFA.Paa-Leu-Gly-OMe 19 which was directly used for the peptide coupling reaction in
the next step. The Boc-Val-OH 20 was treated with TFA.Paa-Leu-Gly-OMe 19 to furnish
the peptidomimetic 4 in 75% yield (Scheme 6).
Detailed NMR studies in CDCl3 and DMSO-d6 revealed that pyrrole amino acid
based peptidomimetic 3 and 4 γ-turn type structure as shown in Figure 1 involving
intramolecular hydrogen bonding between the pyrrole NH and the carbonyl of the
previous residues, BocC=O (3) or ValC=O (4) (Fig-2).
14 17 EDCI, HOBt, CH2Cl2
DIPEA, 0 oC-R.T
73%, 12 h
HN
NH
COOMe
O
NH
BocHN
O
18
18 TFA, CH2Cl2
0 oC-R.T, 1 h
HN
NH
COOMe
O
NH
TFA.H2N
O
19
BocHNOH
O
20
EDCI, HOBt, CH2Cl2, 0 oC
then 19
DIPEA, CH2Cl2, 0 oC-R.T
75%, 12 h
4
Scheme 6
+
Abstract
vi
Figure 2. Schematic representation of the structures of 3 and 4 with the long range rOes seen in their
ROESY spectra.
In conclusion, the rigid scaffold of the pyrrole amino acid described here can serve as
a conformationally constrained template that may find useful application in developing novel
peptidomimetics with interesting structures and useful properties.
BocHNNH
HN
NH
HN
OMe
O
O
O
O
HN
NH
HN
NH
O
O
O
OMeBocHN
O
3 (in DMSO-d6)
4 (in DMSO-d6)
H H
H H
N
HN
NH
HN
OMe
O
O
O
O
3 (in CDCl3 with 12% DMSO-d6)
H H
HNO
OtBu
H
HH
N
HN
NH
OMe
O
O
O
4 (in CDCl3 with 12% DMSO-d6)
H H
HNO
H
HH
NH
H
H-Bond
rOeOtBu
O&
34
Abstract
vii
Part B
Synthesis and Structural Studies of Pyrrole Amino Acid Containing Peptides.
The recognition of β-sheet structures is important for protein-protein interactions.
Small organic molecules with hydrogen bonding pattern and structure complementary to
β-sheets are promising for interception of protein-protein interaction, for stopping protein
aggregation or stabilization or mimicking β-sheet structures. This has led to the design
and synthesis of a large number of conformationally constrained scaffolds, which help to
induce β-turns in short peptides, often a prerequisite to β-hairpin nucleation.
Here we describe the design, syntheses and detailed conformational studies of
three Paa-containing peptides Boc-Paa-Paa-L-Pro-Gly-Ala-Paa-Paa-OMe 2, Boc-Paa-
Paa-D-Pro-Gly-Xaa-Paa-Paa-OMe (3: Xaa = Ala; 4: Xaa = Val) with repeating units of
Paa dimers at both N- and C-termini (Fig-1).
Figure 1
H2NN
OH
OHωi
φi+1
1 (Paa)
ψi+1
1
2
34
5
6
HN
NH
N
O
BocHNNH
OO
NHHN
O
NH
HN
ONH
OHNMeO2C
2
Abstract
viii
It was envisaged that the D-Pro unit with a φ value of +60 ± 20° can induce a
reverse turn that can be further stabilized by noncovalent interactions facilitated by the
near planar disposition of the Paa-dimers at both ends with fixed ωi, ϕi+1 and probably,
ψi+1 torsional angles leading to the formation of hairpin architecture.
The target peptides 2, 3 and 4 were prepared from Boc-Paa-Paa-OMe 5 which in
turn was prepared from pyrrole amino acid (Paa) (Scheme 1).
The peptides 2, 3 and 4 were synthesized from pyrrole amino acid 6 (Paa). The
required dipeptide Boc-Paa-Paa-OMe 5 was prepared following solution phase peptide
coupling procedure. The coupling of Boc-Paa-OH 7 with TFA.H2N-Paa-OMe 8, using
HOBt and EDCI as coupling agents and DIPEA as base in CH2Cl2 gave the dipeptide
Boc-Paa-Paa-OMe 5 in 85% yield (Scheme 2).
NH
BocHNHN
NH
COOMe
O
2, 3 & 4
Scheme 1
5
6
HN
NH
N
O
BocHNNH
OO
NHHN
O
NH
HN
ONH
OHNMeO2C
R
3 [R = Me], 4 [R = (CH3)2CH]
LiOH.H2O
THF-MeOH-H2O
0 oC-rt, 48 h
NH
BocHN
O
OH
7
NH
BocHN
O
OMe
6
Abstract
ix
The carboxylic acid Boc-Paa-Paa-OH 9 was coupled with TFA.L-Pro-Gly-OMe
10 and TFA.D-Pro-Gly-OMe 11 using the same conditions desired above to furnish the
tetrapeptides 12 and 13 in 72% and 76% yields respectively (Scheme 3).
The other two tripeptides Boc-Ala-Paa-Paa-OMe 16, Boc-Val-Paa-Paa-OMe 18 were
prepared following the same protocol as described above (Scheme 4).
6TFA, CH2Cl2
NH
COOMeTFA.H2N
8
7 EDCI, HOBt, CH2Cl2, 0 oC
NH
BocHNHN
NH
COOMe
O
then 8
DIPEA, CH2Cl2, 0 oC-R.T
85%5
Scheme 2
NH
BocHNHN
NH
COOH
O
5LiOH.H2O
THF-MeOH-H2O0 oC-r.t, 48 h
9
9 EDCI, HOBt, CH2Cl2, 0 oC
then 10
DIPEA, CH2Cl2, 0 oC-R.T
72%
12
HN
NH
N
O
BocHNNH
OO
NHMeO
O
EDCI, HOBt, CH2Cl2, 0 oC
then 11
DIPEA, CH2Cl2, 0 oC-R.T
76%
13
HN
NH
N
O
BocHNNH
OO
NHMeO
O
9
Scheme 3
Abstract
x
Our next task was to couple the Boc-Paa-Paa-L-Pro-Gly-OMe 12, with the
tripeptide Boc-Ala-Paa-Paa-OMe 16, to get the target peptide 3. Accordingly, the
tetrapeptide 12 upon treatment with LiOH.H2O in THF-MeOH-H2O gave Boc-Paa-Paa-
L-Pro-Gly-OH 19, tripeptide Boc-Ala-Paa-Paa-OMe 16 was treated with TFA in DCM
to give TFA.H2N-Ala-Paa-Paa-OMe 20 which was subsequently coupled with Boc-Paa-
Paa-L-Pro-Gly-OH 19 under standard peptide coupling conditions as discussed above to
furnish Boc-Paa-L-Pro-Gly-Ala-Paa-Paa-OMe 2 in 65% yield (Scheme 5).
LiOH.H2O
THF-MeOH-H2O, 0 oC-rt NH
BocHNHN
NH
O O
HN
OH
O
O
N12
3 h19
16TFA, CH2Cl2
NH
HN
NH
COOMe
O
TFA.H2N
HN
O
20
BocHNOH
O
EDCI, HOBt, CH2Cl2, 0 oC,
then 14
DIPEA, CH2Cl2, 0 oC-R.T
15
16
12 h, 78%
BocHNOH
O
i) EDCI, HOBt, CH2Cl2, 0 oC,
then 14
DIPEA, CH2Cl2, 0 oC-R.T
1712 h, 73%
18
Scheme 4
TFA, CH2Cl2NH
TFA.H2NHN
NH
COOMe
O
5
14
Abstract
xi
The other two peptidomimetics 3 and 4 were synthesized by attaching TFA.H2N-
Ala-Paa-Paa-OMe 20 and TFA.H2N-Val-Paa-Paa-OMe 21 respectively, to the Boc-Paa-
Paa-D-Pro-Gly-OH 22 following the same reaction protocol, as described above for the
synthesis of 2 (Scheme 6).
18TFA, CH2Cl2
NH
HN
NH
COOMe
O
TFA.H2N
HN
O21
19HN
NH
N
O
BocHNNH
OO
NHHN
O
NH
HN
ONH
OHNMeO2C
2
i) EDCI, HOBt, CH2Cl2, 0 oC
then 20
DIPEA, CH2Cl2, 0 oC-R.T
8h, 65%
Scheme 5
13
LiOH.H2O
THF-MeOH-H2O, 0 oC-rt
3 h
HN
NH
N
O
BocHNNH
OO
NHHO
O
22
22 EDCI, HOBt, CH2Cl2, 0 oC
then 20
DIPEA, CH2Cl2, 0 oC-R.T
8 h, 65%
HN
NH
N
O
BocHNNH
OO
NHHN
O
NH
HN
ONH
OHNMeO2C
3
Abstract
xii
NO
NH
O
O NH
NH
NH
NH
NH
O
O
O
NH
O NH
O
NH
NH
O
H
H
HH
H
OMe
Detailed NMR studies in CDCl3 and DMSO-d6 and subsequent constrained MD
simulations revealed that in pyrrole amino acid based peptidomimetic 4, a type-II′ β-turn
is nucleating around the D-Pro-Gly residues. This leads to subsequent stabilization of β-
hairpin formation.
(a) (b)
Figure-2 (a) Schematic representation of the proposed structure of 4 with the long-range ROEs seen in the
ROESY spectrum. (b) One of the 50 energy-minimized structures of 4 sampled during the 300 ps
simulated annealing MD studies.
22 EDCI, HOBt, CH2Cl2, 0 oC
then 21
DIPEA, CH2Cl2, 0 oC-R.T
12 h, 65%
HN
NH
N
O
BocHNNH
OO
NHHN
O
NH
HN
ONH
OHNMeO2C
4
Scheme 6
Abstract
xiii
In this structure, the Val(5)NH is strongly hydrogen bonded to Paa(2)C=O
intramolecularly (i + 3 → i) leading to a 10-membered ring structure of type II′ β-turn,
supported by the φ, ψ angles of D-Pro and Gly (70°, −108° and −106°, 23°, respectively).
This induces nucleation of three additional hydrogen bonds, Paa(2)pyrroleNH →
Val(5)CO, Paa(6)PyrroleNH → Paa(1)CO and Paa(1)PyrroleNH → Paa(6)CO, leading
to the hairpin formation. Comparison of the structures of 3 and 4 makes it evident that
the replacement of Ala by Val in the 5-position helps compound 4 to adopt the hairpin
conformation due to the restricted rotation about χ1 of Val, 57° and 181° in Figure 2(b).
In conclusion, the known propensity of D-Pro-Gly unit to induce type II′ β-turn
coupled with the sheet forming tendency of pyrrole amino acid (Paa) can lead to a very
well defined β-hairpin conformation in short peptides.
CHAPTER II
Synthesis and DNA Binding Properties of Pyrrole amino acid containing peptides
Many research efforts have been aimed at targeting specific sequences in DNA
with synthetic ligands with the idea of designing both drugs and molecular probes for
DNA polymorphism. The minor groove of the double helical DNA is becoming a site of
great interest for developing new drugs since it is the site of non-covalent high sequence
specific interactions for a large number of small molecules. Minor groove binders are
one of the most widely studied class of agents characterized by high level of sequence
specificity and they are still an interesting class of DNA ligands which demonstrates to
possess several biological activities. In fact, they exhibit antiprotozoal, antibiotic,
antiviral properties. Furthermore, some of these have shown antitumor activity and can
be considered for therapeutical applications.
It was found that the flexibility of the ligand molecule allowed a substantial
resetting of the curvature of the ligand, both with respect to the twist along the groove
and also with respect to matching the bottom of the groove. Pyrrole amino acid (Paa),
5-(aminomethyl)pyrrole-2-carboxylic acid 1 is a new class of building block that has
been used in peptidomimetic studies as a structurally restricted surrogate of the Gly-∆Ala
dipeptide isostere (chapter 1). The flexible methylene spacer between the amino group
Abstract
xiv
and the pyrrole ring in 1 is expected to render in peptides derived from it the curvature
that is necessary to bind in the minor groove of DNA. Herein we describe the synthesis
and DNA binding properties of three Paa dimer-based polyamides 2-4 (Figure 1).
Figure1. Pyrrole amino acid (Paa) 1 Paa dimmer based polyamides 2-4.
The peptide 2 was synthesized from Pyrrole amino acid (paa), the required dipeptide
Boc-Paa-Paa-OMe 5 was prepared following solution phase peptide coupling procedure.
6
LiOH.H2O
THF-MeOH-H2O
0 oC-rt, 48 h
NH
BocHN
O
OH
7
NH
BocHN
O
OMe
6
TFA, CH2Cl2
NH
COOMeTFA.H2N
8
NH
H3NHN
NH
COOMe
O
NH
COOHH2N
1
+
2
NH
HN
HN
NH
COOMe
O
H3N
H3N
O
+
+
3
NH
H3NHN
NH
O
+ HN
NH3
+O
OMe
O
4
Abstract
xv
The compound 3 was prepared by coupling Boc-Lys(Boc)-OH 9 with TFA.H2N-Paa-
Paa-OMe 2 using 1-(3-dimethylaminopropyl)-3-ethylcarbodimide hydrochloride (EDCI)
and 1-hydroxybenzotriazole (HOBt) as coupling agents and DIPEA as base in dry
CH2Cl2 solution to furnish the peptide 10 in 65% yield. The Boc protection of 10 was
removed by treating it with trifluroacetic acid (TFA) in CH2Cl2 solution at 0oC to give
compound 3 (Scheme 2).
The other peptide 4 was synthesized by attaching Boc-Paa-Paa-OH 11 to the H2N-
Lys(Boc)-OMe 13 following the same reaction protocol, as described above for the
synthesis of 3 (Scheme 3).
7 EDCI, HOBt, CH2Cl2, 0 oC
NH
BocHNHN
NH
COOMe
O
then 8
DIPEA, CH2Cl2, 0 oC-R.T
85%5
NH
BocHNHN
NH
COOMe
O
5
TFA, CH2Cl2 2
Scheme 1
BocHNOH
O
NHBoc
EDCI, HOBt, CH2Cl2, 0oC
65%
then 2
DIPEA, CH2Cl2, 0oC-R.TNH
HN
NH
COOMe
O
BocHN
HN
O
NHBoc
10
NH
HN
NH
COOMe
O
BocHN
HN
O
NHBoc
10
TFA, CH2Cl23
9
Scheme 2
Abstract
xvi
The binding of these three newly synthesized polyamides containing pyrrole
amino acid (Paa), with duplex DNA was followed by UV melting, CD spectroscopy, an
ethidium bromide exclusion experiment followed by fluorescence spectroscopy along
with a gel electrophoretic mobility assay. Studies showed that all the ligands are able to
bind duplex DNA. In each of these ligands, the number of building blocks is only two
and yet they showed considerable binding affinities. It was found earlier that the bis-
pyrrole peptide did not exhibit any detectable binding with duplex DNA. From
comparative binding studies on bis-, tri-, tetra- and penta-pyrrole peptides of the
distamycin analogues with DNA it was concluded that the minimum number of pyrrole
carboxamide units for the onset of DNA binding, in the absence of the N-terminus amide
is three. In our study, we have observed that ligand 2 is able to bind to ds-DNA with a
considerable apparent binding affinity (Kapp= 2 × 104 M
−1) in spite of the fact that this
ligand contains only two pyrrole-based building blocks. The improvement in the binding
5
LiOH.H2O
THF-MeOH-H2O
0 oC-rt, 48 h
NH
BocHNHN
NH
COOH
O11
FmocHNOMe
O
NHBoc
20% Piperidine, DCMH2N
OMe
O
NHBoc
12
13
11 EDCI, HOBt, CH2Cl2, 0 oC
65%
then 13
DIPEA, CH2Cl2, 0 oC-R.T
NH
BocHNHN
NH
O
HN
NHBocO
OMe
O
14
TFA, CH2Cl2 4NH
BocHNHN
NH
O
HN
NHBocO
OMe
O
14
Scheme 3
Abstract
xvii
affinity may be due to the presence of the methylene spacer between the amino group
and the pyrrole ring in the ligand, which makes the ligand more flexible to fit in the
minor groove of the duplex DNA. Ligand 3 binds to duplex DNA with three times higher
affinity (Kapp= 6 × 104 M
−1) implying that the extra charge on the ligand improves the
binding ability of ligand. Ligand 4 with a positive charge placed at both ends of the
polyamide binds to duplex DNA with twice the affinity of 3 with two positive charges at
the same end. This result clearly indicates that the spatial distribution of the charge plays
a crucial role in the binding event of the ligand. In ligand 4 as the two charges are located
at either end of the polyamide, these charges can neutralize the negative charges of the
DNA phosphate backbone more efficiently giving increased electrostatic interactions.
In summary we have described the synthesis of polyamides using the monomeric
building block pyrrole amino acid (Paa), 5-(aminomethyl)pyrrole-2-carboxylic acid. The
flexible methylene spacer between the amino group and the pyrrole ring in this monomer
was expected to render the curvature in polyamides derived from it that is necessary to
bind in the minor groove of DNA. DNA binding studies revealed that polyamides with
two positive charges bind with affinities of the order of 105. We also observed that the
polyamide with a positive charge placed at either end of the polyamide binds to the DNA
duplex with twice the affinity compared to that with two positive charges at the same
end. These results will help in the more rational design of new minor grove binding
agents.
CHAPTER III
Studies Directed Towards the Total Synthesis of Antascomicin A
FK 506 (tacrolimus) and rapamycin (sirolimus) are potent immunosuppressive agents in
human clinical organ transplantation. The biological mode of action of these promising
natural products have been extensively investigated over the past decade. For example,
FK-506 is known to form an active complex with the cytosolic immunophilin FKBP12,
which subsequently binds to the secondary proteinous target calcenarium. The latter
interaction is responsible for potent immunosuppressive activity of FK-506. On the other
Abstract
xviii
hand, it has recently been discovered that FKBP is present in the brain at levels 10-40
times higher than in the immune system, suggesting the possibility of nervous system
roles for immunophilins. Subsequently, FK-506 has been reported to elicit neurite
outgrowth and neuroprotective effects in neuronal cultures at pico molar concentrations.
Although the mechanism of neuronal activity of FKBP has not yet been clarified, it has
been shown that the neurotrophic properties of FKBP ligands are independent of
calicineurin-mediated effects. Thus, FKBP ligands are now considered as potential lead
compounds for the development of drugs for treatment and cure of neurodegenerative
disorders such as Alzheimer's and Parkinson's diseases. With regard to clinical use,
efforts have been devoted to the design of neurotrophic FBBP ligands that are non-
immunosuppressive.
Macrolide antibiotics antascomicin A-E were found in a fermentation broth of a
strain of Micromonospora, which was isolated from a soil sample collected in China
(Fig-1). The molecular architecture of antascomicin A was determined by extensive
NMR studies. Antascomicin A exhibit potent binding affinity to FKBP12 (IC50 2.0 nM)
and antagonize the immunosuppressive effect of FK506 and rapamycin, but interestingly
it does not show immunosuppressive activity.
Antascomicin A 1 is an enticing target for the synthetic chemist as it has a
complex polyketide structure that includes both lactol and lactam functionalities, 11
stereogenic centers, and a masked 1,2,3-tricarbonyl unit. The wealth of biological
functions and chemical structures prompted us to initiate a program aimed at the total
synthesis of antascomicin A.
Abstract
xix
An RCM approach was contemplated to construct the macrocyclic ring of antascomicin A.
N
OO
O
HO
R2O
O
O
HO
O
n
R1 R3
ANTASCOMICINS
A: R1 = H, R2 = H, R3 = H, n = 2B: R1 = OH, R2 = H, R3 = H, n = 2C: R1 = OH, R2 = H, R3 = H, n = 2D: R1 = H, R2 = H, R3 = H, n = 1E: R1 = H, R2 = H, R3 = OH, n = 2
Figure 1
N
OO
O
HO
HO
O
O
HO
O
N
OO
O
TBSO
TBSO
O
O
TMSO
O
N
OO
O
OH
O
TMSO
O
OTBS
OTBSOH
RCM Reaction
1 2
3
4
+
Scheme 1. Retrosynthetic analysis of antascomicin A (1).
Abstract
xx
Retrosynthetic analysis (Scheme 1) reveals that an acyclic precursor like 2 was
ideally suited to carry out the planned RCM reaction. Triene 2 could be easily assembled
by coupling the C1-C21 unit 3 with the C22-C34 fragment 4.
Synthesis of C1-C21 unit
Our synthesis commenced with the mono-benzylation of commercially available
pentane-1,5-diol 5. Among two free hydroxyl groups of pentane-1,5-diol, one was
selectively protected using one equivalent each of sodium hydride and benzyl bromide in
THF to give mono-benzyl protected pentate-1,5-diol 6 in 68% yield. The Swern
oxidation of 6 provided aldehyde 7, in 92% yield. Asymmetric aldol addition of the
titanium enolate derived from the N-propanoyl oxazolidinethione 8 to aldehyde 7, gave
the ‘non-Evans’ syn aldol product 9 as the only isolable diastereomer in 78% yield.
Reduction of 9 with sodium borohydride in ethanol gave the diol 10 in 90% yield. Diol 10
was subjected to disilylation by using TBSCl, imidazole and catalytic amount of DMAP
in DMF to furnish the di-TBS ether 11 and finally debenzylation by hydrogenation gave
12 in 72% yield from 10 (Scheme 2).
OHHO OHBnO
HBnO
5 6
O
(COCl)2, DMSO, Et3N
CH2Cl2, -78 oC
NaH,BnBr,TBAI
THF, 0 oC-R.T
68%
12 h
7
O N
S O
Bn
BnO H
O
O N
S O
Bn
OBn
OH
TiCl4, DIPEA, CH2Cl2,
-78 oC, 15 min, then 7,
-78 oC to 0 oC, 3 h (78%)
8
79
NaBH4, EtOH,
0 oC, 10 min
90%
Abstract
xxi
Compound 12 was converted to acid 14 in a two-step oxidation protocol. First primary
alcohol 12 was converted to an aldehyde 13 in 94% yield, using Swern oxidation
procedures, which, without further purification and characterization was oxidized to acid
14 with 1.3 equivalents of NaClO2 and 1.3 equivalents of NaH2PO4 in t-BuOH and 2-
methyl-2-butene. Compound 14 was used to carry out N-acylation of the chiral
oxazolidinone 15, derived from D-Phe, under the mixed anhydride method to furnish 16
in 85% yield. Diastereoselective alkylation of the Na-enolate of 16 with MeI was
followed by reductive removal of the chiral auxiliary to give alcohol 18 as the only
isomer in 46% overall yield (Scheme 3).
OBn
OH
HO
10
OH
OTBS
TBSO
1. TBDMSCl, imidazole, DMAP (cat)
DMF, 0 oC to rt, 24 h
2. H2, Pd/C, EtOAc
72% in two steps
12
Scheme 2
12 OH
OTBS
TBSO
O
OTBS
TBSO OH
HN O
O
Bn
N
OTBS
TBSO
O
O
O
Bn
1. Swern oxidation
2. NaClO2, NaH2PO4,
2-methyl-2-butene, tBuOH
0 oC to rt
86% in two steps 14
Piv-Cl, Et3N, THF
then 15, LiCl, THF,
−20 oC to rt, 4 h
85%
15
16
1. NaHMDS, THF,
then MeI, −78 oC, 2 h
2. LiBH4, Et2O, H2O
0 oC, 10 min
46% in two steps 18
Scheme 3
Abstract
xxii
Swern oxidation of 18 gave an intermediate aldehyde 19, which was reacted with
the Li-enolate of ethyl acetate to give the β-hydroxy ester 20 as a mixture of isomers in
85% yield. Saponification of 20 and subsequent coupling with L-pipecolic acid methyl
ester furnished the amide 22 in 67% overall yield (Scheme 4).
Oxidation of the β-hydroxy amide 22 with Dess-Martin periodinane (DMP) gave
an ‘1,2,3-triketo’ intermediate, which was subjected to desilylation resulting in the
spontaneous formation of the hemiketal 24 in 60% yield. Disilylation of 24 gave a di-
TMS-ether intermediate 25. Brief exposure of this intermediate di-TMS-ether to mild
acid selectively deprotected the primary hydroxyl group to furnish the TMS-ether of the
hemiketal 26 in 70% yield. Finally oxidation of 26 gave the aldehyde 27 in 96% yield.
18
OTBS
TBSO
OH O
N
CO2Me
OTBS
TBSOCO2Et
OH
20
1. LiOH, THF-MeOH-H2O (3:1:1)
0 oC to rt, 4 h
2. L-Pip-OMe, EDCI, HOBt, DIPEA
CH2Cl2, 0 oC to rt, 3 h
67% in two steps 22
1. Swern oxidation
2. LDA, CH3CO2Et, THF
−78 to 0 oC, 30 min
85% in two steps
Scheme 4
22
O
OO
NOMe
O
HO
OH
1. DMP, py, CH2Cl2,
rt, 2 h
2. HF, CH3CN, 3 h
60% in two steps
1. TMSOTf, 2,6-lutidine,
CH2Cl2, 0 oC to rt, 2 h
2. 0.1(N) HCl, THF, rt,
15 min
70% in two steps
24
Abstract
xxiii
Synthesis of the C17-C21 (35) fragment
The synthesis of the C17-C21 (35) fragment started with the pentane-1,5-diol 5.
One of the hydroxyls of 5 was protected as tert-butyldimethylsilyl ether by using TBSCl
and NaH in THF at 0°C to room temperature for 3-4 h to afford compound 28 in 72%
yield. The aldehyde 29 was obtained by the oxidation of the primary hydroxyl group in
28 by employing the Swern oxidation condition. The aldehyde 29 was reacted with CHI3
in the presence of anhydrous CrCl2, generated in situ from CrCl3 and LAH, following the
Takai protocol to introduce an E-vinyl iodide moiety 30, in an overall yield of 73% after
two steps. Acid-catalyzed desilylation of 30 gave 31 in 81% yield. Tosylation of 31 was
followed by nucleophilic substitution of the tosylate group by PhSe−, generated in-situ by
sodium borohydride reduction of diphenyl diselenide, giving the phenylselenide 33, in
78% overall yield, which was then subjected to oxidation-elimination process. Oxidation
of selenide 33 using mCPBA and subsequent β-elimination of the resulting selenoxide
furnished the diene 35 in 71% yield (Scheme 6).
O
OO
NOMe
O
TMSO
OH
O
OO
NOMe
O
TMSO
H
O26
SO3-py, DMSO
Et3N, CH2Cl2,
0 oC, 1 h
96%
27
Scheme 5
TBSO OH
28
HO OH
TBS-Cl, NaH, THF
0 oC-r.t, 3h, 72%
5
Abstract
xxiv
Coupling of C1-C16 Moiety (27) and C17-C21 (35):
Coupling of C1-C16 (27) and C17-C21 (35) fragments to build the target C1-C21
moiety is shown in Scheme 7. A Nozaki-Hiyama-Kishi coupling was employed to carry
out the coupling, giving the coupled product 36 in 62% yield as a mixture of isomers.
Finally, Dess-Martin oxidation of 36 furnished dienone 37, the target C1-C21 fragment of
antascomicin A, in 85% yield.
1. Swern oxidation
2. CHI3, CrCl2, THF,
0 oC to rt, 1 h
73% in two steps
TBSOI
CSA, CH2Cl2, MeOH
0 oC, 30 min
81%
HOI
30
31
1. TsCl, Et3N, DMAP (cat),
CH2Cl2, 0 oC to rt, 10 h
2. PhSeSePh, NaBH4, EtOH,
THF, 0 oC to rt, 4 h
79% in two steps
PhSeI
33
1. mCPBA, CH2Cl2, −20 oC, 2 h
2. DIPA, CCl4, reflux, 4 h
71% in two steps
I
35
28
33
Scheme 6
27NOZAKI KISHI COUPLING
CrCl2, NiCl2 (0.1%), DMSO
N
OO
O
OMe
O
TMSO
HO
67%
R.T, 5 h
+ 35
36
Abstract
xxv
Synthesis of the C22-C34 fragment
For the synthesis of the C22-C34 fragment 4 of antascomicin A, D-quinic acid was
chosen as the starting material. Scheme 8 outlines the details of the synthesis. D-(-)-
Quinic acid 38 was transformed into the intermediate 6 in six steps and in 33% overall
yield following the procedure reported earlier.
OH
HO
HO
O
OH
D(-)QUINIC ACID
PhCHO, PTSA
Benzene, Reflux
OO
OH
OO
Ph
KH, CS2, MeI,
Bu3SnH, AIBN,
Toluene, Reflux, 3 h64% in two steps O
O
OO
Ph
NBS, AIBN (cat)
Benzene, Reflux, 2 h
Bu3SnH, AIBN
OO
BzO
K2CO3(Cat), MeOH
BzO
HO CO2Me
OH
0 oC, 10 min
70%
68% in two steps85%
12 h
40
38
39
THF, 0 oC - R.T, 2 h
OO
O
OO
Ph
40
CS-Me
S
41
Br
42
42
Toluene, Reflux, 3 h OO
BzO
43
Scheme 8
44
Scheme 7
DMP, DCM, R.T
N
OO
O
OMe
O
TMSO
O
85%
1 h36
37
Abstract
xxvi
Silylation of 44 was followed by reduction to furnish the diol 46 in 76% yield.
Selective acylation of the primary hydroxyl group, silylation of the secondary hydroxyl
and deprotection of the acetate gave the disilylated intermediate 49 in 78% yield in three
steps. Finally oxidation of 49 furnished the aldehyde 50 in 95% yield (Scheme 9).
Horner-Wadsworth-Emmons olefination of 50 with the (1S,2R)-(+)-
norephedrine-derived keto-phosphonate 51 using LiCl-DIPEA provided an α,β-
unsaturated amide intermediate, which was hydrogenated to furnish the N-acylated chiral
auxiliary 53 (Scheme 10).
46
HO
TBSOOAc
AcCl, 2,4,6,CollidineTBS-OTf, 2,6-lutidine
K2CO3, MeOH
TBSO
TBSOOH
DCM, -78 oC, 3 h
88% 47
DCM, 0oC, 5 min, 97%
TBSO
TBSOOAc
48
0 oC-R.T,1 h, 98%
49
TBSO
TBSOH
O
SO3-Py Complex
DCM, DMSO, NEt3
0 oC
95%
50Scheme 9
TBS-OTf
2, 6lutidine, DCM, 0 oC
94%
10 min
BzO
TBSO CO2Me
DIBAL, DCM, -78 oC-0 oC
HO
TBSOOH
86%
2 h
45
44
46
Abstract
xxvii
The two chiral centers in compound 61 were planned to be generated by an Evans
aldol reaction between 53 and the aldehyde 60, which in turn was prepared from
commercially available methyl (R)-(-)-3-hydroxy-2-methylpropionate 54. Acid catalyzed
benzylation of 54, in the presence of O-benzyltrichloroacetamidate and catalytic amount
of triflic acid in a mixture of cyclohexane and methylene chloride (2:1) at 0oC afforded
the benzyl ether 55 in 87% yield. Treatment of 55 with LiAlH4 in ether at 0oC gave the
alcohol 56 (98% yield). Compound 56 was then transformed into 60 in four steps, as
shown below, and in 65% overall yield (Scheme 11).
Ph
N
Me
O
O O
P(O)(OEt)2
Ph
N
Me
O
O O
OTBS
OTBS
TBSO
TBSOO
H50
1. 10, LiCl, DIPEA, CH3CN, rt, 6 h
2. H2, Pd-C, EtOAc, rt, 1 h71% in two steps
53
51
Scheme 10
OHMeO
O
Cl3C OBn
NH
Cyclohexane, DCM
0 oC-R.T, 5 h, 87%
LAH, Et2O, 0 oC
LiBH4, Et2OOBnHO
OBnH
O
i) Swern oxidation
Swern oxidation
ii) Ph3P=CH-CO2Et
80%
DCM, R.T, 12 h
54
TfOH (cat)
55
OBnMeO
O 30 min, 98%
56 OBnEtO
O
0 oC- R.T, 48 h
84%58 59
59
60
Scheme 11
56
Abstract
xxviii
Enolate of 53 was reacted with the aldehyde 60 to furnish the aldol product as a
single isomer 61. The stereochemistries of the newly generated chiral centers were
assigned on the basis of earlier reported work. The aldol adduct was then silylated and
the chiral auxiliary was removed by reduction with lithium triethylborohydride to furnish
63 in 40% yield in three steps. A two-step protocol was followed to convert the
hydroxymethyl of 63 to a methyl group – tosylation followed by nucleophilic
substitution of the tosylate group with hydride – to furnish 65 in 70% yield.
Debenzylation of 65 by catalytic hydrogenation, oxidation of the resulting primary
hydroxyl group, one-carbon Wittig olefination and finally an acid-catalyzed desilylation
step furnished the target fragment 4 in 75% yield (Scheme 12).
63
OTBS
OTBSOH
BnO
OTES
Ph
N
Me
O
O O
OTBS
OTBS
53
1. Bu2BOTf, Et3N, CH2Cl2, −78 oC
then 60, −78 oC to rt, 12 h
2. TESOTf, 2,6-lutidine, CH2Cl2,
0 oC, 5 min
3. LiEt3BH, Et2O, 0 oC, 30 min
40% in three steps
63
1. TsCl, Et3N, DMAP (cat)
CH2Cl2, 0 oC to rt, 24 h
2. LiEt3BH, THF, −20 to
0 oC, 5 h
70% in two steps
OTBS
OTBS
BnO
OTES
65
1. H2, Pd-C, EtOAc, rt, 1 h
2. SO3-py, Et3N, DMSO, CH2Cl2, 0 oC, 1 h
3. (Ph3P-CH3)I, NaNH2, Et2O, 0 oC, 10 min
4. CSA (cat), MeOH, CH2Cl2, 0 oC, 5 min
75% in four steps
OTBS
OTBS
OH
4
65
Scheme 12
Abstract
xxix
Coupling of C1-C21 Moiety (37) and C22-C34 Moiety (4):
Having synthesized both the fragments 37 and 4, the stage was know ready to
couple both the fragments to get compound 2. Initially our target was to deprotect the
methyl ester of compound 37 to get acid 3. However, all attempts to convert acid 3 from
the corresponding compound 37 by several methods did not give the corresponding
product. Our efforts to convert α,β-enone ester 37 to acid 3 by several methods like
LiOH, Ba(OH)2, TMSOK, TMSI are ended with decomposition of compound 37. Hence
we were forced to convert methyl ester 37 to an allyl ester by trans esterification method.
Next allyl ester 66 was successfully converted to acid 3 by using catalytic amount of
Pd(Ph3P)4 and morpholine in THF gave compound 3 in 60% yield (Scheme 13).
Compound 3 and 4 were successfully coupled by using 1 equivalent of DCC as a
coupling agent and catalytic amount of DMAP at -20oC in CH2Cl2 to furnish the coupled
product triene 2 in 60% yield. The triene 2 was now ready for the crucial Ring Closing
metathesis reaction (RCM). We thought that, the ring closing would give the desired
product, protected antascomicin A 67. But the RCM reaction method using 0.1 equivalents
of 2nd generation Grubbs catalyst in 10-2 M CH2Cl2 in dilution conditions failed to give the
N
OO
O
OMe
O
TMSO
O
Bu2SnO, Allyl alchohol
N
OO
O
O
O
TMSO
O
Reflux, 24 h
37
52%
66
66 Pd(PPh3)4, Morpholine, THF
60%
R. T, 3 h
N
OO
O
OH
O
TMSO
O
3Scheme 13
Abstract
xxx
desired product 67 and gave rise to an unwanted product (Scheme 14). Our efforts with other
solvents like benzene and toluene also could not provide the expected product. The synthesis
suffered a major setback at this point and our achievement became restricted up to the
synthesis of compound 2 leaving the total synthesis still elusive.
3DCC, DMAP, DCM, -20oC, 7 h
60%
N
OO
O
TBSO
TBSO
O
O
TMSO
O
4+
2
2
Scheme 14
N
OO
O
TBSO
TBSO
O
O
TMSO
O
DCM (10-2M), R.T, 24 h
2nd Generation Grubbs catalyst, 10 mol%
//
67