polymerase engineering: towards the encoded synthesis of ... · polymerase engineering: towards the...
TRANSCRIPT
Polymerase engineering: towards the encoded synthesis
of unnatural biopolymers
David Loakes and Philipp Holliger*
Received (in Cambridge, UK) 18th February 2009, Accepted 6th May 2009
First published as an Advance Article on the web 6th July 2009
DOI: 10.1039/b903307f
DNA is not only a repository of genetic information for life, it is also a unique polymer with
remarkable properties: it associates according to well-defined rules, it can be assembled into
diverse nanostructures of defined geometry, it can be evolved to bind ligands and catalyse
chemical reactions and it can serve as a supramolecular scaffold to arrange chemical groups in
space. However, its chemical makeup is rather uniform and the physicochemical properties of the
four canonical bases only span a narrow range. Much wider chemical diversity is accessible
through solid-phase synthesis but oligomers are limited to o100 nucleotides and variations in
chemistry can usually not be replicated and thus are not amenable to evolution. Recent advances
in nucleic acid chemistry and polymerase engineering promise to bring the synthesis, replication
and ultimately evolution of nucleic acid polymers with greatly expanded chemical diversity within
our reach.
1. Polymerase mechanisms of discrimination
against non-cognate substrates
DNA polymerases perform an astonishing feat of molecular
recognition by incorporating the correct complementary base
opposite its template counterpart, with error rates as low as
10�4 in the case of Thermus aquaticus DNA polymerase I
(Taq), even in the absence of exonucleolytic proofreading.1
The availability of high-resolution structures for different
stages of the polymerase cycle together with elegant work
using synthetic base analogues has begun to illuminate the
subtle interplay of molecular forces and structural determi-
nants on which the correct readout depends.
Elegant work by Kool and others has shown that formation
of the nascent base pair is highly sensitive to geometric
aberrations but surprisingly less dependent on Watson–Crick
hydrogen bonding (provided other forces, e.g. stacking, can
compensate for the loss of binding energy).2 The next step,
extension of the incorporated base, appears to depend not
only on a correct geometric fit and the formation of a
conformationally stable 30-pair but on the presence of cognate
minor groove hydrogen bond donor and acceptor groups.
Furthermore, distortions of cognate duplex geometry by the
newly formed pair cause significant stalling, not just at the
primer 30-end,3 but up to four bases post-extension.4 Polymerase
substrate specificity thus involves passage through three
Medical Research Council, Laboratory of Molecular Biology,Hills Road, Cambridge, Cambridgeshire, UK CB2 0QH.E-mail: [email protected]
David Loakes
David Loakes was born inLondon, England. He receivedhis GRSC graduate degreewhilst working full-time forWellcome, before moving toSussex University (UK) forhis DPhil under the super-vision of Prof. Douglas Young.There he continued his interestin medicinal chemistry workingon the synthesis of cephalosporinanalogues. Following this hecarried out postdoctoral researchwith Dr Daniel M. Brown atthe MRC Laboratory ofMolecular Biology (Cambridge,
UK), working on mutagenic nucleoside analogues. Since beingappointed full-time at the MRC research interests involvenucleic acid chemistry, polymerase function and syntheticbiology.
Philipp Holliger
Philipp Holliger was born inBerne, Switzerland, in 1966.He graduated in BioorganicChemistry and MolecularBiology with Prof. S. A. Bennerat the ETH Zurich beforemoving to the CPE (CambridgeCentre for Protein Engineering)for his PhD and postdoctoralfellowship with Dr G. Winter,FRS. In 2000, he joined theMRC Laboratory of MolecularBiology and was appointedprogram leader in 2005. Hisresearch interests span thefields of chemical biology,synthetic biology, proteinengineering and directedevolution.
This journal is �c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 4619–4631 | 4619
FEATURE ARTICLE www.rsc.org/chemcomm | ChemComm
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
FView Online / Journal Homepage / Table of Contents for this issue
different ‘‘checkpoints’’, involving a readout of distinct
molecular parameters, each of which differentially affect
incorporation, extension and replication.
This ‘‘triple sieve’’ mechanism poses stringent requirements on
the properties of unnatural nucleotide analogues, e.g. hydro-
phobic base analogues (HBAs) or sugar analogues. While many
can be incorporated with good efficiency and specificity due to
energetically favourable interactions, the resulting unnatural base
pairs are often only very inefficiently extended when at the primer
30-end or replicated when acting as a templating base.
While the general features of the specificity mechanisms
described above apply to most polymerases, the relative
importance of the different molecular parameters can vary
significantly between different polymerase families and even
among members of the same family. For example, poly-
merases of the polA family, notably Klenow, are able to
efficiently form base pairs without Watson–Crick hydrogen
bonding (e.g. between difluorotoluene (1) and 4-methyl-
benzimidazole (2)) provided the nascent base pair displays
correct geometry, while the same unnatural bases are utilised
with much reduced efficiency by members of the polY family
of translesion polymerases.5
A striking example of the sophistication of DNA polymerase
molecular discrimination against non-cognate substrates is the
exclusion of non-cognate sugar structures. Ribonucleotides differ
from deoxyribonucleotides only by a 20-hydroxyl group on the
ribose sugar. The addition of an additional hydroxyl group on
the ribose sugar not only renders the product susceptible to
degradation by RNases but also alters the structure of DNA,
which may interfere with the activity of DNA modifying
enzymes. Since the concentration of ribonucleotide triphosphates
(NTPs) within the cell (Escherichia coli) is at about 10-fold higher
than that of deoxyribonucleotide triphosphates (dNTPs),6 DNA
polymerases must be effective at excluding NTPs in order to
maintain the integrity of the DNA genome. Indeed, in E. coli
DNA polymerase I NTPs are excluded from incorporation by a
factor of 104–105-fold, and this stringent discrimination is mainly
encoded in a single residue.7 This ‘‘steric gate’’ residue is
positioned such that it occupies the space where the 20-OH of
an incoming NTP would be placed, thereby sterically excluding
NTPs from the active site.
2. Incorporation and replication of modified
nucleic acids
Great strides have been made in the enzymatic synthesis and
replication of nucleic acids comprising unnatural chemical
modifications simply through elegant chemistry, judicious
use of polymerase and optimisation of reaction conditions
and additives. In general, the incorporation and replication of
modified nucleotide substrates by polymerases (natural or
engineered) comprises a vast body of literature. The work
discussed here therefore represents a blend of the personal
preferences and interests of the authors and is not meant to be
exhaustive.
2.1 Functionalised DNA
The most widely studied group of nucleoside analogues are
those comprising base modifications. Substitutions at the C5
position of either dU or dC, placing the substituent in the
major groove of the double helix, are especially well tolerated
by polymerases and a great variety of C5 substitutions have
been explored and found to be compatible with at least
sporadic incorporation.
Typically, pyrimidines are modified at the C5 position, and
purines at the C7 position (see in particular the extensive
works of Seela et al. for the latter). The number of such
triphosphate derivatives that have been described are too vast
to detail here, but many are dealt with in recent reviews.8–10
The triphosphate derivatives of C5-modified pyrimidines and
C7-modified purines are generally very well tolerated by a
variety of polymerases (primarily DNA polymerases and
reverse transcriptases), thus allowing for the introduction of
additional functional groups into nucleic acids for use in, for
example, aptamers11,12 and detection.13,14 Purines have also
been modified at the C8 position, a modification that generally
induces a syn conformation into the nucleoside which has
consequences for base pairing. As a result, not all C8-modified
purines are good polymerase substrates.15,16
Of particular interest in this group of compounds are
fluorescent dye-labelled nucleotides, as they are an essential
component of several core technologies of modern biology
including sequencing, microarrays and fluorescent in situ
hybridisation (FISH). These and other substituted nucleotides
provide the opportunity to utilise DNA as a molecular
scaffold, i.e. to synthesise nucleic acid polymers displaying
substituents at defined positions along the double helix.
However, despite systematic efforts to optimise substituent
and linker chemistry, many C5-substituted nucleotides, in
particular those bearing large hydrophobic dye molecules,
remain poor polymerase substrates, presumably due to steric
clashes between the large substituents and the polymerase.
Efficient replication of such polymers presents further
challenges as the bulky substituents in the template strand
cause pausing or abortive synthesis.17 Nevertheless, several
groups have reported some success with synthesising nucleic
acid polymers with several or all of the bases bearing
substituents.18–22 The most advanced of these systems was
described by Jager and Famulok, who achieved successful
PCR amplification of short (o100 bp) fragments with a
number of nucleotide analogues functionalised with small
non-fluorescent groups (functionalised DNA (fDNA)) using
Pwo polymerase. Such fDNA was found to stain poorly with
ethidium bromide and displayed a notable shift in the CD
spectrum suggestive of Z-DNA conformation.23,24 Some
reports also briefly commented on the poor solubility in
aqueous solvents of such polymers.18,23 While optimisation
of buffer conditions allowed significant improvements in the
synthesis of fDNA, even greater improvements should be
achievable through the engineering or evolution of polymerases
tailor-made for the synthesis of DNA-polymers bearing
substituents.
2.2 Hydrophobic base analogues
A decade ago Kool and co-workers turned the long-held belief
of nucleoside–polymerase recognition on its head by
demonstrating that certain nucleoside isosteres were
4620 | Chem. Commun., 2009, 4619–4631 This journal is �c The Royal Society of Chemistry 2009
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
specifically and efficiently incorporated into duplex DNA
without Watson–Crick hydrogen bonding.25,26 Working with
the thymine isostere difluorotoluene (1, X = F) it was shown
that E. coli DNA polymerase I (Klenow fragment, KF)
inserted the 50-triphosphate of (1, X = F) onto the growing
primer strand specifically opposite dA, and with an efficiency
approaching that of dTTP. Kool argued that hydrogen bond
recognition within the polymerase active site was not essential
provided there was a good geometric fit for the incoming
nucleotide. Kool and co-workers went on to describe the
adenine mimics (2, X = CH)27 and (2, X = N).28 The
triphosphates of both of these analogues are inserted opposite
dT or (1) by KF specifically and with good efficiency. When
these analogues are used as a triphosphate they extend the
primer by one nucleotide, whereas when present in the template
full-length product can be obtained. The analogue (2, X = N)
was used to assess the requirements for minor groove inter-
actions, the nitrogen being equivalent to N3 of adenine. When
(2, X = N) is present in the template full-length product is
readily achieved with KF, whereas (2, X = CH) is much less
efficiently extended, thus demonstrating that minor groove
H-bonding interactions do play an important role in
polymerase extension.
These analogues also offered an opportunity to explore
differences in minor groove H-bonding requirements among
different polymerases. Among the polymerases assayed for their
ability to bypass these non-polar isosteres of dA and dT,
KF(exo–), Taq and HIV-RT were all able to extend the
(1):(2, X = N) base pair, but not the (1):(2, X = CH) base pair,
indicating the requirement for minor groove hydrogen bond
recognition. However, Pola, Polb and T7 DNA polymerases did
not extend the non-hydrogen-bonding analogues.29,30 NMR
studies using 13C-methionine-labelled Polb suggested that the
thymine isostere (1) disrupts closure of the polymerase ternary
complex, which presumably accounts for why this polymerase is
unable to utilise the isostere.31
Single-turnover kinetic analysis with KF(exo–) found that the
presence of (1) decreased catalytic efficiency 30-fold, whilst when
paired with adenine isosteres catalytic efficiency is decreased
103-fold.32 As is the case with most unnatural base analogues,
incorporation and extension is dependent on polymerases lacking
proofreading ability (exo–). When KF(exo+) polymerase is
used, it will remove a terminal dT:(2, X = N) base pair
with the same efficiency as a mismatch, but interestingly the
(1):(2, X = CH) pair is edited much more slowly.33 The
analogues (1) and (2, X = N) when introduced into E. coli as
part of a modified phage genome are bypassed with moderate
efficiency, and with very high efficiency under SOS-induction.
Remarkably, each of the isosteres were read as their equivalent
natural base.34
Kool et al. have further explored the geometric fit of
thymine isosteres by assessing the polymerase specificity of
size analogues of (1) using each of the analogues where X is H,
F, Cl, Br and I and their ability to be incorporated opposite
dA by KF(exo–).35,36 These analogues present a series in which
the base size increases incrementally over the range of 1 A.
Surprisingly, it was the dichloro analogue, which is larger than
the natural thymine base, that was incorporated with the
greatest efficiency. Similar results were observed with the
translesion polymerases Dpo437 and E. coli polymerases II
and IV,38 though in the case of PolIV analogue incorporation
efficiency was much reduced. Using both the deoxyribosyl and
ribosyl analogues of (1, X = H, F, Cl, Br, I) the DNA- and
RNA-dependent synthesis by HIV-RT was explored. DNA
synthesis follows the same pattern as described above, whilst
with the RNA analogues HIV-RT only discriminates against
the smaller analogues.39,40
Isosteric base analogues have also been used to detect the
presence of DNA lesions such as abasic sites, which are
amongst the most common forms of DNA damage. Matray
and Kool et al. demonstrated that the pyrene C-nucleoside (3)
will form a specific ‘‘base pair’’ with an abasic site (4) in duplex
DNA.41,42 The pyrene base occupies a similar space to an entire
Watson–Crick base pair (i.e. is an isostere of the missing base
pair), as shown by an NMR structure of a DNA duplex
containing the (3):(4) pair.43 The 50-triphosphate derivative of
(3) is specifically incorporated opposite to an abasic site with
efficiency similar to that of a natural dNTP by both KF(exo–),
and yeast polZ.44 In the latter case, the efficiency of incorpora-
tion of pyrene exceeds that of the natural dNTP. It is suggested
that the preference for incorporation of pyrene is due to the
highly favourable stacking energy of the large aromatic pyrene
ring system. However, once pyrene has been incorporated into
the primer strand further extension is very inefficient. The large
pyrene ring has also been used as a mechanistic probe for DNA
polymerase bypass of other DNA lesions. Incorporation of
pyrene opposite various template thymine-dimer analogues
occurred preferentially opposite the 30-T of the dimer (except
in the case of trans, syn-photodimer) suggesting that, in this
case, the 30-thymine residue is in a conformation that is
unfavourable to Watson–Crick base pairing.45
In addition to pyrene, other large aromatic base analogues
have been shown to be efficiently incorporated opposite abasic
sites by various DNA polymerases. Berdis and co-workers
have investigated a number of indole nucleoside derivatives
and shown them to be efficiently and specifically incorporated
opposite an abasic site by T4 DNA polymerase and KF with
up B100-fold higher efficiency than dAMP.46–49 Once again,
once incorporated further extension beyond the indole is not
This journal is �c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 4619–4631 | 4621
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
observed. The most efficient analogues were found to be
5-nitro- and 5-phenyl-indole derivatives,50,51 whilst larger52
and less planar53 derivatives were incorporated less efficiently,
as were analogues where p-electron density or p-electronsurface area are reduced compared to 5-nitroindole.54 All of
these support the notion that the energy gain from favourable
stacking interactions with a large p-electron surface area can
compensate for imperfect shape complementarity in the
incorporation step.
2.3 New base pairs
In recent years there has been much interest in identifying
alternative base-pairing systems. A potential advantage of
such a system is that it may allow an expansion of the genetic
code, creating new specific codons for the incorporation of
unnatural amino acids into proteins. Many different
base-pairing systems have been examined. Work in this area
was pioneered by Benner et al., who first examined base pairs
with unnatural Watson–Crick hydrogen-bonding patterns
such as isoguanine (isodG) and isocytosine (isodC).55 This
modified base pair does not impair the duplex structure, and
the nucleotides are substrates for DNA polymerases.
However, isodG does exist as tautomers and as such can form
base pairs with thymine. The use of 7-deazaisodG constrains
the analogue into a single tautomeric state and thus improves
specificity.56 Alternatively, use of 2-thiothymidine, which pairs
only inefficiently with the enol tautomer of isodG, leads to
efficient use of isodC and isodG in PCR reactions, with 98%
retention of the isodC:isodG base pair per PCR cycle.57 The
analogues isodC and isodG have found application in PCR58
and sequencing reactions,59 where they are good polymerase
substrates, though at slightly reduced efficiency compared with
the native nucleotides.
Apart from isodG:isodC Benner has described a family of
nucleoside analogues covering many possible permutations of
the Watson–Crick hydrogen-bonding patterns, some of which
may be suitable as alternative base pairs.60 For example, the
purine analogue dP (5) forms a specific base pair with the
pyrimidine analogue dZ (6), designed to present electron
density in the minor groove to facilitate specific polymerase
recognition. Each of the triphosphates of dP and dZ are
substrates for both A and B family DNA polymerases,
forming specific base pairs with better than 95% retention of
the dZ:dP base pair per cycle of PCR.61
Hirao et al. have designed a number of unnatural
base-pairing systems based on a combination of hydrogen
bond recognition and shape complementarity. The base pairs
consist of a modified adenine and a pseudobase similar in size
to a pyrimidine base. Each base pair occupies about the same
space as a cognate base pair, but they form specific base pairs
with each other while discriminating against pairing with any
of the natural bases due to steric exclusion. For example, the
base pair between the 2-thienylpurine (7, R = NH2, X1 = CH)
and the oxo(1H)pyridine (8, X2 = H) forms specifically as
there would be a steric clash between either the O4 of thymine
or N4 of cytosine and the thiophene ring. In addition, the base
pair is further stabilised by the presence of two hydrogen
bonds, which also enhance selectivity. These nucleobases
form a specific base pair and are a substrate pair for DNA
polymerases.62 Imidazolin-2-one also serves as a very good
partner, with the 2-thienylpurine (7, R = NH2, X1 = CH)
being a better DNA polymerase substrate than the nucleoside
derivative of pyridin-2-one.63 An alternative base pair is one in
which the thienyl ring is replaced by a methyl group (7, R = H)
that forms a specific base pair with difluorotoluene (1) but
forms a superior base pair, both in terms of oligonucleotide
hybridisation and polymerase specificity with the nucleoside
derivative of pyrrole-2-carbaldehyde.64,65 A further base pair,
the imidazo[4,5-b]pyridine derivative of (7, R=NH2, X1 = CH),
forms a specific base pair with the nucleoside derivative of
2-nitropyrrole. PCR reactions using the triphosphate
derivatives of these analogues resulted in B1% of the modified
base pair incorporated into the PCR product after just
20 cycles,66 and even better efficiency obtained using a
pyridoimidazole pairing with a nitropyrrole (see Fig. 2).67
These analogues are also good substrates for RNA
polymerases. Using T7 RNA polymerase the triphosphate
derivative of the ribonucleoside of (8, X2 = H or I or
arylalkynyl) is efficiently incorporated opposite the thienyl-
purine derivatives (7) in a DNA template.68,69 Furthermore,
the RNA transcripts were substrates for translation using an
E. coli cell free system to produce proteins containing
3-chlorotyrosine.70,71 Analogues of the ribonucleotide of (8)
have been incorporated into RNA transcripts bearing
fluorophores,72,73 biotin74 and iodine useful for photo-
crosslinking to proteins.75
Rappaport described an alternative three-base pair set of
nucleotides suitable for replication by T7 DNA polymerase.76
The three base pairs used are A:T, hypoxanthine:cytosine and
6-thiopurine:5-methyl-2-pyrimidinone. Error frequency rates
for these three base pairs were of the order 10�4–10�6, with
the major mismatches occurring between hypoxanthine and
thymine.
Another strategy for orthogonal base pairing through steric
exclusion has been explored by Kool et al. They devised a set
of nucleotides in which a benzene ring has been inserted
between the sugar and the nucleobase, termed xDNA77,78
4622 | Chem. Commun., 2009, 4619–4631 This journal is �c The Royal Society of Chemistry 2009
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
and yDNA79 (depending on the position of the phenyl ring),
and the effect of the additional benzene ring is to widen the
duplex structure. Initial studies to determine whether yDNA
can transmit genetic information showed that DNA
polymerases can insert the correct nucleotide, but at a lower
specificity than for normal DNA.80
The group of Romesberg have investigated a very large
range of chemical space in the search for a third base pair
based on hydrophobic base analogues. A large body of
analogues have been investigated to systematically probe
the steric requirements for polymerase recognition, and for
formation of homo- and hetero-pairs by the hydrophobic base
analogues. The range of chemical space investigated
includes substituted phenyl,81,82 substituted pyridyl83,84 and
isocarbostyril,85–87 as well as pyridones,88 azaindoles89 and
other heterocycles.90 More recently Romesberg et al. have
developed a novel base pair (see Fig. 2) which is replicated by
Klenow fragment (exo–) DNA polymerases with an efficiency
approaching that of a native base pair.91
2.4 Universal bases
Universal base analogues92,93 are generally characterised as
non-hydrogen bonding, hydrophobic base analogues (HBAs)
that pair indiscriminately with all the native nucleobases.94
Generally they are destabilising in a duplex due to loss of
hydrogen bonding and solvation, though this is compensated
for by stacking and hydrophobic interactions, the strength of
which is related to the size of the HBA. Largely due to their
lack of hydrogen bond functionality (but see isosteres above)
universal bases are very poorly recognised by polymerases.95,96
However, Romesberg et al. have shown that various isocarbo-
styril analogues, such as MICS, behave as efficient universal
base analogues, not only in hybridisation terms but also being
recognised indiscriminately by KF, albeit at reduced catalytic
efficiency.85 The universal bases benzimidazole, 5- and
6-nitrobenzimidazole and 5-nitroindole, are incorporated as
their 50-triphosphates opposite native DNA bases by KF and
human Pola 4000-fold more efficiently than a mismatch, with
Pola preferentially incorporating the universal bases opposite
pyrimidines and KF opposite purines.97 However, whilst these
analogues were shown to bind to Moloney murine leukaemia
virus (MMLV)-reverse transcriptase (RT) they were not
incorporated by it. Many HBAs can be incorporated into
DNA under rather forcing conditions, but once incorporated
they are blocking lesions. When present in a template strand
dAMP is preferentially incorporated opposite the universal
base, with polymerases presumably following the A-rule.98
The exception to this is when incorporated opposite a template
lesion, such as an abasic site (see section on isosteres above).
The analogue 3-nitropyrrole has also been shown to be a
substrate for the poliovirus RNA polymerase, preferentially
incorporating it opposite A and U, though at 100-fold reduced
efficiency compared to cognate pairs.99
2.5 Modified sugar–phosphate backbone
While there is a growing body of circumstantial evidence to
support the idea of a primordial RNA-based genetic system, a
plausible and efficient route for the prebiotic synthesis of
ribose has been lacking. This has spurred various groups to
investigate whether simpler sugar, or sugar-substitutes, can be
used to encode genetic information.
Meggers and co-workers have shown that the deoxyribose
ring may be substituted for a simple glycerol unit,100 GNA (9),
which forms stable duplexes with either GNA or (S)-, but not
(R)-GNA, with RNA,101 and does not form a stable duplex
with DNA. Szostack et al. have examined a variety of A and B
family DNA polymerases as well as reverse transcriptases to
determine whether GNA could be replicated as a genetic
system. Using (S)-glycerol NTPs it was shown that
Therminator DNA polymerase was able to perform +1
extension of a DNA primer–template duplex with an efficiency
comparable to incorporation of a native dNTP.102 Furthermore,
Bst DNA polymerase was able to fully extend across a short
region of template GNA, as was an engineered MMLV-RT,
albeit to a lesser extent.103 The fact that Bst DNA polymerase
is able to synthesise DNA from a GNA template suggests that
the polymerase is able to sufficiently stabilise the transient
DNA–GNA duplex for sequence readout from GNA to be
possible.
Another pre-RNA candidate that has been proposed is the
mixed acetal aminal of formyl glycerol, FNA, (10). The
triphosphate derivatives of FNA have been examined as
substrates by DNA polymerases and reverse transcriptases
on a DNA template. Unlike GNA, both enantiomers were
found to be substrates for a variety of polymerases, notably
those with lower fidelity, though the (R)-enantiomer was
preferentially incorporated over the ‘natural’ (S)-isomer.104
One of the most intriguing pre-RNA nucleic acids investi-
gated to date involves an aldo sugar containing only four
carbon atoms, (L)-a-threofuranosyl nucleic acid (TNA), (11),
in which the phosphodiester bonds are attached to the 20- and
30-positions of the threofuranose ring (30 - 20).105 TNA
duplexes show similar thermal stabilities to that of DNA
and RNA, and TNA can also form cross-pairs with either
DNA or RNA. X-Ray crystallography has shown that TNA
can be readily accommodated into B-form duplex DNA,106
but that it is a better mimic for A-form nucleic acids and
therefore forms better cross-pairs with RNA.107 Cytosine
TNA templates specifically direct the non-enzymatic
incorporation of activated rGMP leading to RNA products.108
Szostak and co-workers have examined in some detail the
templating properties of TNA and the incorporation of TNA
triphosphates by a variety of DNA polymerases and reverse
transcriptases. They have shown that certain polymerases are
capable of copying limited regions of template TNA, despite
the differences in the sugar–phosphate backbone between
DNA and TNA.109 Significant improvements in templating
ability were observed in the presence of Mn(II) ions. In order
to assess the effect of Mn(II) ions on fidelity, polymerase
reaction products were sequenced and the error rate compared
with template DNA and template TNA. It was shown that
using the DNA polymerase Sequenase the error rate for TNA
was 1.5-fold higher than for DNA, whilst with the MMLV-RT
Superscript II the TNA error rate was 3-fold increased. The
authors have suggested that evolution of a TNA-dependent
DNA polymerase could lead to improved efficiency, and thus
This journal is �c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 4619–4631 | 4623
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
lead to applications in which TNA was used as genetic
information.
The kinetics of incorporation of TNA triphosphates by
DeepVent(exo–) and Bst and HIV-RT have been reported.
Whilst the kcat and Km (or Vmax/Km) values differ, the trend
is the same, which is that the Km increases with TNA
triphosphates, whilst the kcat decreases.110,111 Interestingly,
the kinetics of incorporation of a second TNA triphosphate
shows a further increase in Km and decrease in Vmax/Km.111
However, the polymerases retain fidelity for the incorporation
of TNA triphosphates.
The polymerase that has been most widely studied for the
incorporation of modified nucleotides and oligonucleotides is
Therminator DNA polymerase, an exo–variant of 91N DNA
polymerase, containing an A485L mutation.112,113 It has also
been shown to be one of the most effective DNA polymerases
for the incorporation of TNA triphosphates. The kinetics of
incorporation are much the same as described above for
DeepVent exo–DNA polymerase, in that the Km is higher
and the kcat reduced compared to dNTPs. Furthermore, when
the primer strand is modified to contain five TNA residues at
its 30-end there is no further significant decrease in efficiency of
incorporation of TNA triphosphates.114 Addition of Mn(II)
ions further improves the kinetics of incorporation as might be
anticipated. Therminator DNA polymerase has been used to
generate TNA sequences as long as 80 nt long. Error rates
were raised compared to the use of normal dNTPs, but low
enough, in principle, for use in SELEX.115,116 The authors
suggest that in vitro evolution could be used to identify a
mutant polymerase capable of faithful production of TNA
oligomers.
A class of modified sugars with interesting properties
that has been widely used for SELEX applications are those
with 20-modifications. Among these 20-F, 20-NH2 and
especially 20-OCH3 have found applications in the aptamer
field as they are reasonably well accepted by T7 RNA
polymerase and by reverse transcriptases and greatly stabilise
aptamers to nuclease degradation.117,118 Indeed, using an
optimised protocol of transcription using the Y693F/H784A
double mutant of T7 RNA polymerase, Burmeister et al.
described the selection of an anti-VEGF aptamer composed
entirely of 20-OCH3 nucleotides, which was stable in plasma
for 496 h and survived autoclaving at 125 1C.119
Locked nucleic acids (LNA) were first described by
Imanishi120 and Wengel,121 and contain a methylene bridge
between 20-O and C40 of the ribose sugar. The presence of the
bridging methylene group locks the sugar into a 30-endo
conformation, reducing the conformational flexibility of the
ring. LNA exhibits enhanced binding towards RNA and
DNA, and incorporation of LNA into DNA induces A-like
conformations. Various DNA polymerases have been assessed
for their ability to either bypass LNA within a DNA template,
or to incorporate LNA triphosphates. When present in a DNA
template LNA is bypassed provided the LNA residues are
distributed along the strand.122 The most efficient polymerases
were found to be KOD(exo–) and Vent(exo–), though
efficiency of replication was reduced compared to native
DNA. LNA triphosphates may be incorporated into DNA
using Phusion and Therminator DNA polymerases,123–125
again providing there are not too many consecutive
substitutions. Pfu exo–DNA polymerase will incorporate
LNA triphosphates, but is unable to further extend the
product.
Another class of modified sugars that have been studied are
those bearing modification at C40, which have been extensively
examined by Marx and co-workers. Such analogues were
examined to investigate steric effects in the polymerase
active site using various DNA polymerases and reverse
transcriptases.126 The effect of a bulky group at C40 is different
between polymerases, and is dependent to some extent on the
sequence context. The C40-methyl derivative is accepted the
most readily, although kinetics of incorporation with HIV-RT
showed reduced rate of incorporation.127 However, increasing
the size of the alkyl substituent to larger than methyl does
cause the rate of incorporation to reduce markedly, but
increases specificity.128,129
The final class of sugar-modified nucleoside to be considered
here are those based on pyrofuranose. A study of base-pairing
properties of a variety of six-membered carbohydrate mimics
with RNA showed that RNA can cross-pair with a broad
range of nucleic acid structures.130 Amongst those analogues
studied anhydrohexitol nucleic acid (HNA) (12), a homologue
of DNA, was found to be the most stable, forming the most
stable homo duplex in a polyA–polyT system, as well as
forming the most stable cross-pair with RNA (more stable
than RNA:RNA). Crystal structures of HNA:HNA131 and of
HNA:RNA132 reveal that it adopts an A-form duplex, though
with more pronounced major groove, and a more reinforced
backbone hydration, which may account for the enhanced
stability of such duplexes. HNA oligonucleotides have been
used in aptamers,133 and in antisense,134 but are not substrates
for RNase H. Various DNA polymerases were able to catalyse
the incorporation of a single HNA triphosphate, but only
DeepVent(exo–) DNA polymerase could catalyse incorpora-
tion of more than one substitution.135 Up to six substitutions
could be achieved using DeepVent at high enzyme concentra-
tion, though under such conditions fidelity of incorporation
was reduced when more than two HNA triphosphates were
included in the reaction. The HNA triphosphates have also
been tested as substrates for various reverse transcriptases
(RTs).136 Whilst all RTs tested were able to incorporate a
single HNA triphosphate, only the dimeric RTs RAV2 and
HIV-RT were able to incorporate more than one. A series of
mutant HIV-RTs were assayed and the M184V mutant of
HIV-RT was found to be the most efficient at making multiple
incorporations of HNA. HNA incorporated into the AUG
start codon and the UUC codon in mRNA did not prevent
translation.137 Another interesting six-membered ring
replacement for ribose is the cyclohexene nucleoside (CeNA),
(13), which like HNA forms exceptionally stable duplexes with
4624 | Chem. Commun., 2009, 4619–4631 This journal is �c The Royal Society of Chemistry 2009
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
RNA. CeNA has also been used in antisense, but unlike HNA,
CeNA oligomers are substrates for RNase H.138,139
3. Polymerase engineering and evolution
As detailed in the previous section, some mutant polymerases
show dramatic improvements in their capacity to incorporate,
extend and replicate a range of unnatural nucleotide analogues.
However, in many cases further improvements would be
desirable to enable processive synthesis, replication and evolution
of modified nucleic acids. This has provided a strong incentive
to engineer polymerase function by design, screening and
directed evolution.
Polymerase engineering started with simple deletion of
accessory domains of the wild-type polymerase, which sometimes
yielded polymerases with improved properties. For example,
variants of E. coli DNA polymerase I (Klenow) and Taq
polymerase (Stoffel fragment and Klentaq) have been generated
by full or partial deletion of the N-terminal 50–30 exonuclease
domain and show improved stability and fidelity. Mutation of
two essential aspartates in the 30–50 exonuclease domain is
commonly used to generate proofreading defective variants
(e.g.KOD(exo–)). These exo-variants are often the most useful
for the incorporation of unnatural nucleotide substrates, as the
proofreading 30–50 exonuclease will efficiently remove most
unnatural nucleotides. A particularly interesting mutation turned
out to be A485L in the 91N exo–polymerase (Therminator).
Originally designed to allow efficient incorporation of acyclic
nucleotides for terminator sequencing,112,113 it has turned out
to improve the incorporation of a wide spectrum of unnatural
nucleotide substrates (e.g. TNA, see above).
In some cases careful inspection of high-resolution
structures has allowed the rational design of mutants with
improved properties. Examples include mutants of Taq
polymerase with improved properties of dideoxynucleotide
incorporation,140 an increased capability to incorporate
ribonucleotides,141,142 reduced pausing143 as well as altered
fidelity.144 Grafting of new domains has been used to modify
polymerase properties, notably processivity. Insertion of the
thioredoxin binding loops of T7 or T3 DNA polymerase into
the E. coli Pol I Klenow fragment or Taq polymerase,
respectively, leads to an increase in processivity.145,146 Fusion
of the Sulfolobus solfataricus DNA binding protein Sso7d to
Taq and Pfu DNA polymerases also improved processivity in
both enzymes.147 Insertion of AviTagt peptide loops either
side of the primer–template duplex binding cleft of 91N DNA
polymerase allowed enzymatic biotinylation and assembly
with streptavidin into an artificial ‘‘processivity complex’’.
These complexes tethered to a solid surface lead to an
impressive gain in processivity from ca. 20 nt in the unmodified
polymerase to several thousand in the modified polymerase.148
Loeb et al. pioneered the use of in vivo screening by genetic
complementation of E. coli (polA12ts, recA718) (Fig. 1a) with
Taq polymerase, HIV reverse transcriptase (RT) and human
Polb.149–151 Complementation at non-permissive temperatures
was used to probe inter alia the mutability of the polymerase
active site152 and screening of the selected mutants has yielded
polymerase variants with a range of properties including
an increased capability to incorporate ribonucleotides141
or increased fidelity.153 Genetic complementation of
Saccharomyces cervisiae RAD30, RAD52 has been used to
screen variants of human polZ and yielded a polZmutant with
increased activity.154
In vitro screening for polymerase function has long been
used, for example, in drug discovery notably of antiviral
compounds such as nucleotide and non-nucleotide inhibitors
of HIV-1 RT.155 A number of clever high-throughput assay
formats have developed both as semi-quantitative continuous
and end-point assays. For example, several groups have used a
scintillation proximity assay in which a biotinylated primer is
extended with tritium-labelled nucleotide triphosphates and
immobilized on streptavidin coated scintillant beads. Other
assays take advantage of the increased fluorescence intensity
of some intercalating dyes, when bound to the double-
stranded DNA (dsDNA) synthesised by an active polymerase
Fig. 1 Strategies for the selection of polymerase repertoires. (a) Complementation of an E. coli strain with a temperature-sensitive mutant of
DNA polymerase I (polAts). Only active polymerases (Pol1) support colony growth at the non-permissive temperature (37 1C). (b) Proximal display
of polymerase and primer–template duplex substrate on filamentous phage. Primer extension leads to incorporation of a biotinylated nucleotide
which tags the phage displaying the active polymerase (Pol1) for selection. (c) Selection of polymerase function by compartmentalised self-
replication within the aqueous droplets of a water-in-oil emulsion (CSR). Only polymerases active under the selection conditions (Pol1) can
replicate their own encoding gene and generate ‘‘offspring’’, while those that are inactive (Pol2) cannot self-replicate and disappear from the gene
pool.
This journal is �c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 4619–4631 | 4625
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
or detect the incorporation a tagged nucleotide substrate in an
ELISA format.156 Elegant continuous assays have also been
described either based on the change in fluorescence of
single-stranded DNA binding displaced upon primer extension
or on fluorescence resonance energy transfer (FRET) between
acceptor and donor fluorophores located on primer and
template strand or as part of a template hairpin structure.157
Marx et al. have adapted high-throughput liquid-handling to
perform in vitro screening of polymerase function by assaying
duplex DNA synthesis by this hairpin assay157 or simply by
monitoring the fluorescence increase of the intercalating dye
SYBR Green (in primer extension reactions or qPCR). These
in vitro screening methods were used to identify variants of
Taq polymerase with improved mismatch discrimination,158
variants of KlenTaq polymerase with substantial RT
activity159 or with an ability to amplify DNA damaged by
UV radiation.160
While in vivo screening by genetic complementation is a
highly effective and convenient way to identify active variants
from a large library of polymerase mutants, its scope is limited
by the fact that the only property that is screened for is basal
polymerase activity. In vitro screening is potentially much
more flexible as, in principle, any polymerase property can
be optimised, provided it is amenable to a high-throughput
assay. On the other hand, most of the assays rely on the
detection of double-stranded DNA or on the disruption of
template secondary structure by the strand-displacement
activity of the polymerase. In many ways, such assays are
not ideal for the discovery of polymerases with an enhanced
ability to incorporate and replicate unnatural nucleotide
substrates as polymerase activity may be too weak for
significant amounts of dsDNA to be synthesised or the
chemistry of the modified substrates may interfere with dye
binding or fluorescence. Other pitfalls include false positives
e.g. polymerases with an enhanced ability to synthesise
dsDNA from just three nucleotides160 (e.g. avoiding the
unnatural substrate), to preferentially misincorporate tagged
nucleotides or with an enhanced ability to melt template
hairpins. Finally, in vitro screening methods are currently
limited to polymerase libraries an order of magnitude smaller
than those used in genetic complementation.
Larger polymerase repertoires can, in principle, be processed
by the use of selection technologies. One of the most
productive selection methods has been phage display, in
particular for the identification of specific peptide and protein
interactions. Both Jestin and Romesberg have adapted phage
display for the selection of polymerase activity by proximal
display of both primer–template substrate and polymerase on
the surface of the phage particle (Fig. 1b).161,162 By selection
for in cis incorporation of a biotin-tagged nucleotide tri-
phosphate into the tethered template–primer duplex Jestin
has isolated variants of Taq polymerase with substantial RT
activity.163 Using a similar approach, Romesberg et al.
have selected variants of the Stoffel fragment of Taq poly-
merase with greatly improved incorporation of ribonucleotide
triphosphates,162 20-OCH3 ribonucleotide triphosphates164 and
a polymerase with 30-fold improved extension of the unnatural
PICS:PICS (14) self-pair.165 The phage selection approach
should in principle be ideal for the identification of polymerases
with enhanced activity with unnatural substrates. The only
caveat being that assay conditions must be compatible with
phage viability, which limits the selection of polymerases from
thermophilic organisms under realistic conditions.
Furthermore, the intramolecular tethering of the substrate
and polymerase may favour the selection of polymerase
variants with low affinity for template–primer duplex and/or
poor processivity. Finally, polymerases are usually not
efficiently secreted and displayed on the phage surface,
although this can be improved by optimisation of the signal
peptide.166 On the other hand, the phage method should be
extremely sensitive and in principle able to detect a single
incorporation event.
We have developed an alternative selection strategy for
the evolution of polymerases, called ‘‘compartmentalised
self-replication’’ (CSR).167 CSR is based on a simple feedback
loop, in which a polymerase replicates only its own encoding
gene with compartmentalisation into the aqueous compartments
of a water-in-oil (w/o) emulsion168 serving to isolate individual
self-replication reactions from each other (Fig. 1c). Thus each
polymerase replicates only its own encoding gene to the
exclusion of those in other compartments (i.e. self-replicates).
In such a system adaptive gains directly (and proportionally)
translate into genetic amplification of the encoding gene.
Therefore, the copy number of a polymerase gene after one
round of CSR is correlated to the catalytic activity of the
encoded polymerase under the selection conditions, with
polymerase genes encoding the most active polymerases best
adapted to the selection conditions dominating the population.
Segregation of self-replication into discrete, physically
separate compartments is critical to ensure linkage of pheno-
type and genotype during CSR. We developed w/o emulsions
that are stable for prolonged periods at temperatures
exceeding 90 1C and allowing selection of polymerases under
a wide range of experimental conditions including PCR
thermocycling.167 CSR has proved to be a productive method
for the selection of polymerase function yielding variants of
Taq polymerase with 410-fold increased thermostability,167
4130-fold increased resistance to heparin167 or a generically
enhanced substrate spectrum.169 Molecular breeding of
polymerase genes from the genus Thermus and CSR selection
also allowed the isolation of polymerases with a striking ability
PCR-amplify damaged DNA, which allowed the increased
recovery of DNA sequences from ice-age specimens.170 CSR
selections from the same polymerase library also yielded a
polymerase with a broad resistance to common environmental
PCR inhibitors or a generic ability to replicate large
hydrophobic base analogues (C. Baar, DL, PH, unpublished
results).
4626 | Chem. Commun., 2009, 4619–4631 This journal is �c The Royal Society of Chemistry 2009
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
CSR makes stringent demands on the catalytic efficiency
and processivity of selected polymerases (requiring replication
of the 42 kb polymerase gene). In order to reduce the
adaptive burden and increase the sensitivity of the method,
we devised short-patch CSR (spCSR), in which only a short,
defined segment of the polymerase gene is replicated and
evolved. spCSR has allowed the isolation of variants of Taq
with an expanded substrate range allowing enhanced
incorporation of 20-substituted nucleotides including the
generation of mixed RNA–DNA, 20-F-DNA amplification
products in PCR156 as well as greatly improved incorporation
of fluorescent dye-labelled nucleotides171 (N. Ramsay,
PH unpublished results).
4. Unnatural biopolymer synthesis
An improved understanding of the molecular mechanisms of
polymerase selectivity together with advances in nucleic acid
chemistry and polymerase engineering and evolution have
brought the synthesis of unnatural nucleic acid polymers
within our grasp.
Different types of polymers pose different challenges to
substrate design and polymerase engineering.
Nucleic acids can be viewed as repeats of tripartite nucleo-
tide building blocks comprising the nucleobase (A), the sugar
scaffold (B) and the backbone linkage (C). Different types of
polymers can comprise alternative compounds replacing or
modifying either of these components.
To simplify discussion, we propose a classification of
unnatural nucleic acid polymers based on this tripartite
structure, Fig. 2. Group A polymers DNA (or RNA) polymers
comprising standard ribofuranose sugars and phosphoester
backbone linkages with modifications to one or all of the four
canonical bases. These can be substituents, for example, to the
50-position of pyrimidines or substitutions of the natural bases
with non-canonical ones (as discussed in 2.1–2.4). Group B
polymers are DNA (or RNA) polymers comprising either
partial or complete replacement of standard ribofuranose
sugars by non-standard structures but retaining canonical
phosphoester backbone linkages (e.g. GNA, TNA) (as
discussed in 2.5). Group C polymers comprise polymers those
where the phosphodiester backbone linkages are replaced by
alternative structures.
There are many examples of partial group A polymers
(in which a natural base is partially replaced by another)
synthesised in the laboratory, notably the incorporation of
unnatural base pairs.57,62,85–87 Other partial group A polymers
are of biotechnological importance as DNA probes used in
fluorescent in situ hybridisation (FISH) or in microarrays are
generated using fluorescent dye substituted nucleotides. While
short probes can be synthesised by a number of polymerases
(e.g. Klenow for random prime dye labelling), longer group A
polymers can be difficult to synthesise in good yield especially
for larger substituents like fluorescent dyes. However, directed
evolution by CSR has yielded polymerases that are able to
completely replace one of the bases with their fluorescent
dye-labelled equivalent (e.g. dA with FITC-12-dA) in PCR.169
Fig. 2 The three panels on the left illustrate the tripartite nucleotide structure, comprising the nucleobase (red), ribofuranose sugar (blue) and
phosphoester backbone linkage (green). Nucleobases are either purine bases (A, G) (left) or pyrimidine bases (T, C) (right) and can bear
substitutions R1 (yellow) at the X (X = C) or exocyclic amino position (purines) or C5 position (pyrimidines). Panels on the right detail
modifications to the canonical structures that are compatible with enzymatic incorporation, such as base substitutions (yellow) (from left to right):
alkyne, enamine, alcohol, imidazole, guanidine and pyridyl groups; alternative base pairing systems (red) (from left to right) such as those
described by Benner (5:6),61 Yokoyama and Hirao67 and Romesberg;91 sugar modifications (blue) (from left to right), such as hexose nucleic acids
(HNA), threose nucleic acids (TNA), glycol nucleic acids (GNA) and 20-modifications (20-F, 20-OCH3) and Locked nucleic acids (LNA);
alternative internucleotide linkages (green) (from left to right): methylphosphonates, phosphorothioates, phosphoramidates, boranophosphates
and selenophosphates. We propose a classification of nucleic acid polymers, whereby group A polymers comprise nucleic acids with modifications
to the base (yellow / red panel), group B polymers comprise modifications to the sugar (blue panel) and group C comprise alternative backbones
(green panel).
This journal is �c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 4619–4631 | 4627
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
Complete group A polymers, in which all of the four bases
bear a substituent have been described by Famulok for a
number of small substituent groups (fDNA).23 Several other
groups have also reported synthesis of group A polymers in
which all four bases are substituted with either biotin or a
fluorescent dye but with poor yields.18–22 Such polymers have
been proposed as the basis of a next-generation sequencing
method, termed exonuclease sequencing. To be practical more
quantitiative synthesis of these types of polymersis needed and
this is likely to require careful design of substituents to
maintain polymer solubility and substantial engineering of
the polymerase primer-template binding interface.
Given the stringency of molecular discrimination against
non-cognate chemistry by DNA polymerases, one would
expect the chemical make-up of natural DNA to be completely
uniform. However, the modifications to the cognate DNA
chemistry found in nature are diverse and surprisingly
wide-spread. They overwhelmingly comprise modifications to
the nucleobases (group A polymers) including, apart from the
well-known post-replicative epigenetic modifications, the
replacement of one of the canonical bases with an unnatural
analogue at the level of replication such as the complete
replacement of dA by 2,6-diaminopurine in the S. elongatus
phage S-2L. These modifications are thought to protect the
viral genome from cellular nucleases and promote preferential
replication of modified phage DNA by the cognate phage
replicases.172
Partial and complete group B polymers have so far not been
found in nature. However, there are several examples of
partial or evn complete group B polymers synthesized in the
laboratory using unnatural substitutes for ribose (e.g. GNA
and TNA) (as discussed in 2.5).101,103,109
Only one natural example of a group C polymer has been
found, that is the aS-modification in Streptomyces.173 However,
completely aS-modified DNA (in which every a oxygen is
replaced by sulfur) has been synthesized in the laboratory.169
The aS-modification is of course a modest alteration of the
standard phosphoester backbone but aS-DNA and mixed
aS-containing aptamers are nevertheless of interest due to their
resistance to many nucleases and their increased binding to
proteins.174 A number of other modifications to the phosphoester
backbone linkage have been described and found to be
compatible with enzymatic synthesis and replication. These
include boranophosphate, selenophosphate, phosphoramidate
or phosphonoester linkages. Boranophosphates are good poly-
merase substrates and have been successfullly used in
SELEX,175 and such aptamers may find clinical application
in the neutron capture therapy of tumors.
More drastic changes to the nucleic acid structure backbone,
in which both phosphoester linkage as well as sugar scaffold are
replaced by alternative chemical structures have been
attempted176 but few are satisfactory replacements. A notable
exception is PNA in which bases are displayed on a aminoethyl-
glycine backbone. PNA displays highly specific hybridization
properties and can direct non-enzymatic replication of
complementary DNA.177 However, enzymatic replication of
PNA would require extensive modification to the polymerase
active site. It is not inconceivable that polymers such as PNAs
could be synthesized on an engineered ribosome.178
5. Conclusion
Advances in polymerase engineering and evolution together
with sophisticated chemistry are poised to enable the enzymatic
synthesis, replication and evolution of unnatural nucleic acid
polymers with an ever widening array of chemical substitutions.
The new sequence and chemical space opened up by these
advances is likely to be a rich source of novel nucleic acid
therapeutics, aptamers and enzymes with useful applications
in medicine, biotechnology, nanotechnology and material
science. These novel polymers will provide insights into the
parameters of the molecular encoding of information leading
to artificial genetic systems based on unnatural chemistry.
Notes and references
1 K. R. Tindall and T. A. Kunkel, Biochemistry, 1988, 27,6008–6013.
2 E. T. Kool, Annu. Rev. Biophys. Biomol. Struct., 2001, 30,1–22.
3 M.-M. Huang, N. Arnheim and M. F. Goodman, Nucleic AcidsRes., 1992, 20, 4567–4573.
4 H. Miller and A. P. Grollman, Biochemistry, 1997, 36,15336–15342.
5 O. Potapova, C. Chan, A. M. DeLucia, S. A. Helquist,E. T. Kool, N. D. Grindley and C. M. Joyce, Biochemistry,2006, 45, 890–898.
6 A. Kornberg and T. A. Baker, DNA Replication, Freeman,San Francisco, 1992.
7 M. Astatke, N. D. Grindley and C. M. Joyce, J. Mol. Biol., 1998,278, 147–165.
8 Y. Brudno and D. R. Liu, Chem. Biol., 2009, 16, 265–276.9 M. Hocek and M. Fojta,Org. Biomol. Chem., 2008, 6, 2233–2241.10 S. H. Weisbrod and A. Marx, Chem. Commun., 2008, 5675–5685.11 M. Hollenstein, C. J. Hipolito, C. H. Lam and D. M. Perrin,
Nucleic Acids Res., 2009, 37, 1638–1649.12 P. Capek, H. Cahova, R. Pohl, M. Hocek, C. Gloeckner and
A. Marx, Chem.–Eur. J., 2007, 13, 6196–6203.13 H. Cahova, L. Havran, P. Brazdilova, H. Pivonkova, R. Pohl,
M. Fojta and M. Hocek, Angew. Chem., Int. Ed., 2008, 47,2059–2062.
14 M. Vrabel, P. Horakova, H. Pivonkova, L. Kalachova,H. Cernocka, H. Cahova, R. Pohl, P. Sebest, L. Havran,M. Hocek and M. Fojta, Chem.–Eur. J., 2009, 15, 1144–1154.
15 C. Lam, C. Hipolito and D. M. Perrin, Eur. J. Org. Chem., 2008,4915–4923.
16 H. Cahova, R. Pohl, L. Bednarova, K. Novakova, J. Cvacka andM. Hocek, Org. Biomol. Chem., 2008, 6, 3657–3660.
17 Z. Zhu, J. Chao, H. Yu and A. S. Waggoner, Nucleic Acids Res.,1994, 22, 3418–3422.
18 S. Brakmann and P. Nieckchen, ChemBioChem, 2001, 2, 773–777.19 Z. Foldes-Papp, B. Angerer, W. Ankenbauer and R. Rigler,
J. Biotechnol., 2001, 86, 237–253.20 Z. Foldes-Papp, B. Angerer, P. Thyberg, M. Hinz, S. Wennmalm,
W. Ankenbauer, H. Seliger, A. Holmgren and R. Rigler,J. Biotechnol., 2001, 86, 203–224.
21 G. Giller, T. Tasara, B. Angerer, K. Muhlegger, M. Amacker andH. Winter, Nucleic Acids Res., 2003, 31, 2630–2635.
22 T. Tasara, B. Angerer, M. Damond, H. Winter, S. Dorhofer,U. Hubscher and M. Amacker, Nucleic Acids Res., 2003, 31,2636–2646.
23 S. Jager and M. Famulok, Angew. Chem., Int. Ed., 2004, 43,3337–3340.
24 S. Jager, G. Rasched, H. Kornreich-Leshem, M. Engeser,O. Thum and M. Famulok, J. Am. Chem. Soc., 2005, 127,15071–15082.
25 S. Moran, R. X.-F. Ren and E. T. Kool, Proc. Natl. Acad. Sci.U. S. A., 1997, 94, 10506–10511.
26 S. Moran, R. X.-F. Ren, S. Rumney and E. T. Kool, J. Am.Chem. Soc., 1997, 119, 2056–2057.
4628 | Chem. Commun., 2009, 4619–4631 This journal is �c The Royal Society of Chemistry 2009
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
27 K. M. Guckian, J. C. Morales and E. T. Kool, J. Org. Chem.,1998, 63, 9652–9656.
28 J. C. Morales and E. T. Kool, J. Am. Chem. Soc., 1999,121, 2323–2324.
29 J. C. Morales and E. T. Kool, J. Am. Chem. Soc., 2000, 122,1001–1007.
30 J. C. Morales and E. T. Kool, Biochemistry, 2000, 39,12979–12988.
31 T. W. Kirby, E. F. DeRose, W. A. Beard, S. H. Wilson andR. E. London, Biochemistry, 2005, 44, 15230–15237.
32 O. Potapova, C. Chan, A. M. DeLucia, S. A. Helquist,E. T. Kool, N. D. Grindley and C. M. Joyce, Biochemistry,2006, 45, 890–898.
33 J. C. Morales and E. T. Kool, Biochemistry, 2000, 39, 2626–2632.34 J. C. Delaney, P. T. Henderson, S. A. Helquist, J. C. Morales,
J. M. Essigmann and E. T. Kool, Proc. Natl. Acad. Sci. U. S. A.,2003, 100, 4469–4473.
35 T. W. Kim and E. T. Kool, J. Org. Chem., 2005, 70, 2048–2053.36 T. W. Kim, J. C. Delaney, J. M. Essigmann and E. T. Kool, Proc.
Natl. Acad. Sci. U. S. A., 2005, 44, 15803–15808.37 S. Mizukami, T. W. Kim, S. A. Helquist and E. T. Kool,
Biochemistry, 2006, 45, 2772–2778.38 A. P. Silverman, Q. Jiang, M. F. Goodman and E. T. Kool,
Biochemistry, 2007, 46, 13874–13881.39 A. P. Silverman and E. T. Kool, J. Am. Chem. Soc., 2007, 129,
10626–10627.40 A. P. Silverman, S. J. Garforth, V. R. Prasad and E. T. Kool,
Biochemistry, 2008, 47, 4800–4807.41 T. J. Matray and E. T. Kool, Nature, 1999, 399, 704.42 L. Dzantiev, Y. O. Alekseyev, J. C. Morales, E. T. Kool and
L. J. Romano, Biochemistry, 2001, 40, 3215–3221.43 S. Smirnov, T. J. Matray, E. T. Kool and C. de los Santos,
Nucleic Acids Res., 2002, 30, 5561–5569.44 H. Hwang and J.-S. Taylor, Biochemistry, 2004, 43, 14612–14623.45 L. Sun, M. Wang, E. T. Kool and J.-S. Taylor, Biochemistry,
2000, 39, 14603–14610.46 A. Sheriff, E. Motea, I. Lee and A. J. Berdis, Biochemistry, 2008,
47, 8527–8537.47 X. Zhang, I. Lee and A. J. Berdis, Org. Biomol. Chem., 2004, 2,
1703–1711.48 E. Z. Reineks and A. J. Berdis, Biochemistry, 2004, 43,
393–404.49 D. Vineyard, X. Zhang, A. Donnelly, I. Lee and A. J. Berdis,Org.
Biomol. Chem., 2007, 5, 3623–3630.50 X. Zhang, I. Lee and A. J. Berdis, Biochemistry, 2005, 44,
13111–13121.51 I. Lee and A. J. Berdis, ChemBioChem, 2006, 7, 1990–1997.52 X. Zhang, A. Donnelly, I. Lee and A. J. Berdis, Biochemistry,
2006, 45, 13293–13303.53 X. Zhang, I. Lee, X. Zhou and A. J. Berdis, J. Am. Chem. Soc.,
2006, 128, 143–149.54 X. Zhang, I. Lee and A. J. Berdis, Biochemistry, 2005, 44,
13101–13110.55 C. R. Geyer, T. R. Battersby and S. A. Benner, Structure
(London), 2003, 11, 1485–1498.56 F. Seela and C. Wei, Helv. Chim. Acta, 1999, 82, 726.57 A. M. Sismour and S. A. Benner, Nucleic Acids Res., 2005, 33,
5640–5646.58 M. L. Abraham, M. Albalos, T. Guettouche, M. J. Friesenhahn
and T. R. Battersby, BioTechniques, 2007, 43, 617.59 J. D. Ahle, S. Barr, A. M. Chin and T. R. Battersby,Nucleic Acids
Res., 2005, 33, 3176–3184.60 S. A. Benner, Acc. Chem. Res., 2004, 37, 784–797.61 Z. Yang, A. M. Sismour, P. Sheng, N. L. Puskar and
S. A. Benner, Nucleic Acids Res., 2007, 35, 4238–4249.62 I. Hirao, T. Fujiwara, M. Kimoto and S. Yokoyama, Bioorg.
Med. Chem. Lett., 2004, 14, 4887–4890.63 I. Hirao, Y. Harada, M. Kimoto, T. Mitsui, T. Fujiwara and
S. Yokoyama, J. Am. Chem. Soc., 2004, 126, 13298–13305.64 T. Mitsui, A. Kitamura, M. Kimoto, T. To, A. Sato, I. Hirao and
S. Yokoyama, J. Am. Chem. Soc., 2003, 125, 5298–5307.65 T. Mitsui, M. Kimoto, A. Sato, S. Yokoyama and I. Hirao,
Bioorg. Med. Chem. Lett., 2003, 13, 4515–4518.66 I. Hirao, T. Mitsui, M. Kimoto and S. Yokoyama, J. Am. Chem.
Soc., 2007, 129, 15549–15555.
67 M. Kimoto, R. Kawai, T. Mitsui, S. Yokoyama and I. Hirao,Nucleic Acids Res., 2009, 37, e14.
68 M. Endo, T. Mitsui, T. Okuni, M. Kimoto, I. Hirao andS. Yokoyama, Bioorg. Med. Chem. Lett., 2004, 14, 2593–2596.
69 T. Mitsui, M. Kimoto, Y. Harada, S. Yokoyama and I. Hirao,J. Am. Chem. Soc., 2005, 127, 8652–8658.
70 I. Hirao, T. Ohtsuki, T. Fujiwara, T. Mitsui, T. Yokogawa,T. Okuni, H. Nakayama, K. Takio, T. Yabuki, T. Kigawa,K. Kodama, T. Yokogawa, K. Nishikawa and S. Yokoyama,Nat. Biotechnol., 2002, 20, 177–182.
71 I. Hirao, M. Kimoto, T. Mitsui, T. Fujiwara, R. Kawai, A. Sato,Y. Harada and S. Yokoyama, Nat. Methods, 2006, 3, 729.
72 R. Kawai, M. Kimoto, S. Ikeda, T. Mitsui, M. Endo,S. Yokoyama and I. Hirao, J. Am. Chem. Soc., 2005, 127,17286–17295.
73 M. Kimoto, T. Mitsui, Y. Harada, A. Sato, S. Yokoyama andI. Hirao, Nucleic Acids Res., 2007, 35, 5360–5369.
74 K. Moriyama, M. Kimoto, T. Mitsui, S. Yokoyama and I. Hirao,Nucleic Acids Res., 2005, 33, e129.
75 M. Kimoto, M. Endo, T. Mitsui, T. Okuni, I. Hirao andS. Yokoyama, Chem. Biol., 2004, 11, 47–55.
76 H. P. Rappaport, Biochem. J., 2004, 381, 709–717.77 H. B. Liu, J. M. Gao, L. Maynard, Y. D. Saito and E. T. Kool,
J. Am. Chem. Soc., 2004, 126, 1102–1109.78 H. B. Liu, J. M. Gao and E. T. Kool, J. Org. Chem., 2005, 70,
639–647.79 A. H. F. Lee and E. T. Kool, J. Am. Chem. Soc., 2005, 127,
3332–3338.80 J. Chelliserrykattil, H. Lu, A. H. F. Lee and E. T. Kool,
ChemBioChem, 2008, 9, 2976–2980.81 G. T. Hwang and F. E. Romesberg, Nucleic Acids Res., 2006, 34,
2037–2045.82 S. Matsuda, A. A. Henry and F. E. Romesberg, J. Am. Chem.
Soc., 2006, 128, 6369–6375.83 Y. Kim, A. M. Leconte, Y. Hari and F. E. Romesberg, Angew.
Chem., Int. Ed., 2006, 45, 7809–7812.84 Y. Hari, G. T. Hwang, A. M. Leconte, N. Joubert, M. Hocek and
F. E. Romesberg, ChemBioChem, 2008, 9, 2796–2799.85 M. Berger, Y. Wu, A. K. Ogawa, D. L. McMinn, P. G. Schultz
and F. E. Romesberg, Nucleic Acids Res., 2000, 28, 2911–2914.86 S. Matsuda and F. E. Romesberg, J. Am. Chem. Soc., 2004, 126,
14419–14427.87 A. M. Leconte, G. T. Hwang, S. Matsuda, P. Capek, Y. Hari and
F. E. Romesberg, J. Am. Chem. Soc., 2008, 130, 2336–2343.88 A. M. Leconte, S. Matsuda, G. T. Hwang and F. E. Romesberg,
Angew. Chem., Int. Ed., 2006, 45, 4326–4329.89 E. L. Tae, Y. Wu, G. Xia, P. G. Schultz and F. E. Romesberg,
J. Am. Chem. Soc., 2001, 123, 7439–7440.90 A. M. Leconte, S. Matsuda and F. E. Romesberg, J. Am. Chem.
Soc., 2006, 128, 6780–6781.91 Y. J. Seo, G. T. Hwang, P. Ordoukhanian and F. E. Romesberg,
J. Am. Chem. Soc., 2009, 131, 3246–3252.92 R. Nichols, P. C. Andrews, P. Zhang and D. E. Bergstrom,
Nature, 1994, 369, 492–493.93 D. Loakes and D. M. Brown, Nucleic Acids Res., 1994, 22,
4039–4043.94 D. Loakes, Nucleic Acids Res., 2001, 29, 2437–2447.95 T. H. Smith, J. Latour, C. Leo, S. Muthini, P. Siebert and
P. S. Nelson, Clin. Chem. (Washington, D. C.), 1995, 41, 10.96 C. L. Smith, A. C. Simmonds, I. R. Felix, A. L. Hamilton,
S. Kumar, S. Nampalli, D. Loakes, F. Hill and D. M. Brown,Nucleosides Nucleotides, 1998, 17, 541–554.
97 M. Chiaramonte, C. L. Moore, K. Kincaid and R. D. Kuchta,Biochemistry, 2003, 42, 10472–10481.
98 C. Crey-Desbiolles, N. Berther, M. Kotera and P. Dumy, NucleicAcids Res., 2005, 33, 1532–1543.
99 D. A. Harki, J. D. Graci, V. S. Korneeva, S. K. B. Ghosh,Z. Hong, C. E. Cameron and B. R. Peterson, Biochemistry, 2002,41, 9026–9033.
100 L. Zhang and E. Meggers, J. Am. Chem. Soc., 2005, 127,74–75.
101 M. K. Schlegel, A. E. Peritz, K. Kittigowittana, L. Zhang andE. Meggers, ChemBioChem, 2007, 8, 927–932.
102 A. T. Horhota, J. W. Szostak and L. W. McLaughlin, Org. Lett.,2006, 8, 5345–5347.
This journal is �c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 4619–4631 | 4629
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
103 C.-H. Tsai, J. Chen and J. W. Szostak, Proc. Natl. Acad. Sci.U. S. A., 2007, 104, 14598–14603.
104 B. D. Heuberger and C. Switzer, J. Am. Chem. Soc., 2008, 130,412–413.
105 K. U. Schoning, P. Scholz, S. Guntha, X. Wu, R. Krishnamurthyand A. Eschenmoser, Science, 2001, 290, 1347.
106 C. J. Wilds, Z. Wawrzak, R. Krishnamurthy, A. Eschenmoserand M. Egli, J. Am. Chem. Soc., 2002, 124, 13716–13721.
107 P. S. Pallan, C. J. Wilds, Z. Wawrzak, R. Krishnamurthy,A. Eschenmoser and M. Egli, Angew. Chem., Int. Ed., 2003, 42,5893–5895.
108 B. D. Heuberger and C. Switzer, Org. Lett., 2006, 8, 5809–5811.109 J. C. Chaput, J. K. Ichida and J. W. Szostak, J. Am. Chem. Soc.,
2003, 125, 856–857.110 V. Kempeneers, K. Vastmans, J. Rozenski and P. Herdewijn,
Nucleic Acids Res., 2003, 31, 6221–6226.111 J. C. Chaput and J. W. Szostak, J. Am. Chem. Soc., 2003, 125,
9274–9275.112 A. F. Gardner and W. E. Jack, Nucleic Acids Res., 1999, 27,
2545–2553.113 A. F. Gardner and W. E. Jack, Nucleic Acids Res., 2002, 30,
605–613.114 A. Horhota, K. Zou, J. K. Ichida, B. Yu, L. W. McLaughlin,
J. W. Szostak and J. C. Chaput, J. Am. Chem. Soc., 2005, 127,7427–7434.
115 J. K. Ichida, A. Horhota, K. Zou, L. W. McLaughlin andJ. W. Szostak, Nucleic Acids Res., 2005, 33, 5219–5225.
116 J. K. Ichida, K. Zou, A. Horhota, B. Yu, L. W. McLaughlin andJ. W. Szostak, J. Am. Chem. Soc., 2005, 127, 2802–2803.
117 C. Wilson and A. D. Keefe, Curr. Opin. Chem. Biol., 2006, 10,607–614.
118 A. D. Keefe and S. T. Cload, Curr. Opin. Chem. Biol., 2008, 12,448–456.
119 P. E. Burmeister, S. D. Lewis, R. F. Silva, J. R. Preiss,L. R. Horwitz, P. S. Pendergrast, T. G. McCauley, J. C. Kurz,D. M. Epstein, C. Wilson and A. D. Keefe, Chem. Biol., 2005, 12,25–33.
120 S. Obika, M. Dorio, D. Nanbu and T. Imanishi, Chem. Commun.,1997, 1643.
121 S. K. Singh, P. Nielsen, A. A. Koshkin and J. Wengel, Chem.Commun., 1998, 455.
122 M. Kuwahara, S. Obika, J.-I. Nagashima, Y. Ohta, Y. Suto,H. Ozaki, H. Sawai and T. Imanishi, Nucleic Acids Res., 2008, 36,4257–4265.
123 R. N. Veedu, B. Vester and J. Wengel, ChemBioChem, 2007, 8,490–492.
124 R. N. Veedu, B. Vester and J. Wengel, Nucleosides, NucleotidesNucleic Acids, 2007, 26, 1207–1210.
125 R. N. Veedu, B. Vester and J. Wengel, J. Am. Chem. Soc., 2008,130, 8124–8125.
126 F. Di Pasquale, D. Fischer, D. Grohmann, T. Restle, A. Geyerand A. Marx, J. Am. Chem. Soc., 2008, 130, 10748–10757.
127 M. Strerath, J. Cramer, T. Restle and A. Marx, J. Am. Chem.Soc., 2002, 124, 11230–11231.
128 D. Summerer and A. Marx, Angew. Chem., Int. Ed., 2001, 40,3693–3695.
129 D. Summerer and A. Marx, J. Am. Chem. Soc., 2002, 124,910–911.
130 L. Kerremans, G. Schepers, J. Rozenski, R. Busson, A. VanAerschot and P. Herdewijn, Org. Lett., 2001, 3, 4129–4132.
131 R. Declercq, A. Van Aerschot, R. J. Read, P. Herdewijn andL. Van Meervelt, J. Am. Chem. Soc., 2002, 124, 928.
132 T. Maier, I. Przylas, N. Strater, P. Herdewijn and W. Saenger,J. Am. Chem. Soc., 2005, 127, 2937–2943.
133 G. Kolb, S. Reigadas, C. Boiziau, A. Van Aerschot,A. Arzumanov, M. J. Gait, P. Herdewijn and J.-J. Toulme,Biochemistry, 2005, 44, 2926–2933.
134 H. Kang, M. H. Fisher, D. Xu, Y. J. Miyamoto, A. Marchand,A. Van Aerschot, P. Herdewijn and R. L. Juliano, Nucleic AcidsRes., 2004, 32, 4411–4419.
135 K. Vastmans, S. Pochet, A. Peys, L. Kerremans, A. Van Aerschot,C. Hendrix, P. Marliere and P. Herdewijn, Biochemistry, 2000, 39,12757–12765.
136 K. Vastmans, M. Froeyen, L. Kerremans, S. Pochet andP. Herdewijn, Nucleic Acids Res., 2001, 29, 3154–3163.
137 I. N. Lavrik, O. N. Avdeeva, O. A. Dontsova, M. Froeyen andP. Herdewijn, Biochemistry, 2001, 40, 11777–11784.
138 J. Wang, B. Verbeure, I. Luyten, M. Froeyen, C. Hendrix,H. Rosemeyer, F. Seela, A. Van Aerschot and P. Herdewijn,Nucleosides, Nucleotides Nucleic Acids, 2001, 20, 785–788.
139 B. Verbeure, E. Lescrinier, J. Wang and P. Herdewijn, NucleicAcids Res., 2001, 29, 4941–4947.
140 Y. Li, V. Mitaxov and G. Waksman, Proc. Natl. Acad. Sci.U. S. A., 1999, 96, 9491–9496.
141 P. H. Patel and L. A. Loeb, J. Biol. Chem., 2000, 275,40266–40272.
142 M. Astatke, K. Ng, N. D. Grindley and C. M. Joyce, Proc. Natl.Acad. Sci. U. S. A., 1998, 95, 3402–3407.
143 K. B. Ignatov, A. A. Bashirova, A. I. Miroshnikov andV. M. Kramarov, FEBS Lett., 1999, 448, 145–148.
144 T. A. Kunkel and K. Bebenek, Annu. Rev. Biochem., 2000, 69,497–529.
145 E. Bedford, S. Tabor and C. C. Richardson, Proc. Natl. Acad. Sci.U. S. A., 1997, 94, 479–484.
146 J. F. Davidson, R. Fox, D. D. Harris, S. Lyons-Abbott andL. A. Loeb, Nucleic Acids Res., 2003, 31, 4702–4709.
147 Y. Wang, D. E. Prosen, L. Mei, J. C. Sullivan, M. Finney andP. B. Vander Horn, Nucleic Acids Res., 2004, 32, 1197–1207.
148 J. G. Williams, D. L. Steffens, J. P. Anderson, T. M. Urlacher,D. T. Lamb, D. L. Grone and J. C. Egelhoff, Nucleic Acids Res.,2008, 36, e121.
149 M. Suzuki, D. Baskin, L. Hood and L. A. Loeb, Proc. Natl. Acad.Sci. U. S. A., 1996, 93, 9670–9675.
150 B. Kim, T. R. Hathaway and L. A. Loeb, J. Biol. Chem., 1996,271, 4872–4878.
151 J. B. Sweasy and L. A. Loeb, Proc. Natl. Acad. Sci. U. S. A., 1993,90, 4626–4630.
152 E. Loh and L. A. Loeb, DNA Repair, 2005, 4, 1390–1398.153 E. Loh, J. Choe and L. A. Loeb, J. Biol. Chem., 2007, 282,
12201–12209.154 E. Glick, K. L. Vigna and L. A. Loeb, EMBO J., 2001, 20,
7303–7312.155 S. G. Sarafianos, B. Marchand, K. Das, D. M. Himmel,
M. A. Parniak, S. H. Hughes and E. Arnold, J. Mol. Biol.,2009, 385, 693–713.
156 J. L. Ong, D. Loakes, S. Jaroslawski, K. Too and P. Holliger,J. Mol. Biol., 2006, 361, 537–550.
157 D. Summerer and A. Marx, Angew. Chem., Int. Ed., 2002, 41,3620–3622; D. Summerer and A. Marx, Angew. Chem., Int. Ed.,2002, 41, 3516.
158 D. Summerer, N. Z. Rudinger, I. Detmer and A. Marx, Angew.Chem., Int. Ed., 2005, 44, 4712–4715.
159 K. B. Sauter and A. Marx, Angew. Chem., Int. Ed., 2006, 45,7633–7635.
160 C. Gloeckner, K. B. Sauter and A. Marx, Angew. Chem., Int. Ed.,2007, 46, 3115–3117.
161 J. L. Jestin, P. Kristensen and G. Winter, Angew. Chem., Int. Ed.,1999, 38, 1124–1127.
162 G. Xia, L. Chen, T. Sera, M. Fa, P. G. Schultz andF. E. Romesberg, Proc. Natl. Acad. Sci. U. S. A., 2002, 99,6597–6602.
163 S. Vichier-Guerre, S. Ferris, N. Auberger, K. Mahiddine andJ. L. Jestin, Angew. Chem., Int. Ed., 2006, 45, 6133–6137.
164 M. Fa, A. Radeghieri, A. A. Henry and F. E. Romesberg, J. Am.Chem. Soc., 2004, 126, 1748–1754.
165 A. M. Leconte, L. Chen and F. E. Romesberg, J. Am. Chem. Soc.,2005, 127, 12470–12471.
166 J. L. Jestin, G. Volioti and G. Winter, Res. Microbiol., 2001, 152,187–191.
167 F. J. Ghadessy, J. L. Ong and P. Holliger, Proc. Natl. Acad. Sci.U. S. A., 2001, 98, 4552–4557.
168 D. S. Tawfik and A. D. Griffiths, Nat. Biotechnol., 1998, 16,652–656.
169 F. J. Ghadessy, N. Ramsay, F. Boudsocq, D. Loakes, A. Brown,S. Iwai, A. Vaisman, R. Woodgate and P. Holliger,Nat. Biotechnol.,2004, 22, 755–759.
170 M. d’Abbadie, M. Hofreiter, A. Vaisman, D. Loakes,D. Gasparutto, J. Cadet, R. Woodgate, S. Paabo andP. Holliger, Nat. Biotechnol., 2007, 25, 939–943.
171 J. Ong, PhD Thesis, Cambridge University, 2004.
4630 | Chem. Commun., 2009, 4619–4631 This journal is �c The Royal Society of Chemistry 2009
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online
172 J. H. Gommers-Ampt and P. Borst, FASEB J., 1995, 9,1034–1042.
173 L. Wang, S. Chen, T. Xu, K. Taghizadeh, J. S. Wishnok,X. Zhou, D. You, Z. Deng and P. C. Dedon, Nat. Chem. Biol.,2007, 3, 709–710.
174 A. Somasunderam, M. R. Ferguson, D. R. Rojo,V. Thiviyanathan, X. Li, W. A. O’Brien and D. G. Gorenstein,Biochemistry, 2005, 44, 10388–10395.
175 S. M. Lato, N. D. Ozerova, K. He, Z. Sergueeva, B. R. Shaw andD. H. Burke, Nucleic Acids Res., 2002, 30, 1401–1407.
176 P. E. Nielsen, Annu. Rev. Biophys. Biomol. Struct., 1995, 24,167–183.
177 J. G. Schmidt, L. Christensen, P. E. Nielsen and L. E. Orgel,Nucleic Acids Res., 1997, 25, 4792–4796.
178 O. Rackham and J. W. Chin, Nat. Chem. Biol., 2005, 1,159–166.
This journal is �c The Royal Society of Chemistry 2009 Chem. Commun., 2009, 4619–4631 | 4631
Dow
nloa
ded
by P
ortla
nd S
tate
Uni
vers
ity o
n 04
Feb
ruar
y 20
12Pu
blis
hed
on 0
6 Ju
ly 2
009
on h
ttp://
pubs
.rsc
.org
| do
i:10.
1039
/B90
3307
F
View Online