O4-Alkyl-2′-deoxythymidine cross-linked DNA to probe recognition and repair by O6-alkylguanine DNA alkyltransferases

Article abstract of DOI:10.1039/c2ob25705j

DNA duplexes containing a directly opposed O⁴-2′- deoxythymidine-alkyl-O⁴-2′-deoxythymidine (O ⁴-dT-alkyl-O⁴-dT) interstrand cross-link (ICL) have been prepared by the synthesis of cross-linked nucleoside dimers which were converted to phosphoramidites to produce site specific ICL. ICL duplexes containing alkyl chains of four and seven methylene groups were prepared and characterized by mass spectrometry and nuclease digests. Thermal denaturation experiments revealed four and seven methylene containing ICL increased the T_m of the duplex with respect to the non-cross-linked control with an observed decrease in enthalpy based on thermodynamic analysis of the denaturation curves. Circular dichroism experiments on the ICL duplexes indicated minimal difference from B-form DNA structure. These ICL were used for DNA repair studies with O⁶-alkylguanine DNA alkyltransferase (AGT) proteins from human (hAGT) and E. coli (Ada-C and OGT), whose purpose is to remove O⁶- alkylguanine and in some cases O⁴-alkylthymine lesions. It has been previously shown that hAGT can repair O⁶-2′-deoxyguanosine- alkyl-O⁶-2′-deoxyguanosine ICL. The O⁴-dT-alkyl- O⁴-dT ICL prepared in this study were found to evade repair by hAGT, OGT and Ada-C. Electromobility shift assay (EMSA) results indicated that the absence of any repair by hAGT was not a result of binding. OGT was the only AGT to show activity in the repair of oligonucleotides containing the mono-adducts O⁴-butyl-4-ol-2′-deoxythymidine and O⁴-heptyl-7-ol- 2′-deoxythymidine. Binding experiments conducted with hAGT demonstrated that the protein bound O⁴-alkylthymine lesions with similar affinities to O⁶-methylguanine, which hAGT repairs efficiently, suggesting the lack of O⁴-alkylthymine repair by hAGT is not a function of recognition.

Full text of DOI:10.1039/c2ob25705j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 7078
www.rsc.org/obc
PAPER
O4-Alkyl-2′-deoxythymidine cross-linked DNA to probe recognition and
repair by O6-alkylguanine DNA alkyltransferases†‡
Francis P. McManus, Derek K. O’Flaherty, Anne M. Noronha and Christopher J. Wilds*
Received 10th April 2012, Accepted 3rd July 2012
DOI: 10.1039/c2ob25705j
DNA duplexes containing a directly opposed O4-2′-deoxythymidine-alkyl-O4-2′-deoxythymidine
(O4-dT-alkyl-O4-dT) interstrand cross-link (ICL) have been prepared by the synthesis of cross-linked
nucleoside dimers which were converted to phosphoramidites to produce site specific ICL. ICL duplexes
containing alkyl chains of four and seven methylene groups were prepared and characterized by mass
spectrometry and nuclease digests. Thermal denaturation experiments revealed four and seven methylene
containing ICL increased the Tm of the duplex with respect to the non-cross-linked control with an
observed decrease in enthalpy based on thermodynamic analysis of the denaturation curves. Circular
dichroism experiments on the ICL duplexes indicated minimal difference from B-form DNA structure.
These ICL were used for DNA repair studies with O6-alkylguanine DNA alkyltransferase (AGT) proteins
from human (hAGT) and E. coli (Ada-C and OGT), whose purpose is to remove O6-alkylguanine and in
some cases O4-alkylthymine lesions. It has been previously shown that hAGT can repair O6-2′-
deoxyguanosine-alkyl-O6-2′-deoxyguanosine ICL. The O4-dT-alkyl-O4-dT ICL prepared in this study were
found to evade repair by hAGT, OGT and Ada-C. Electromobility shift assay (EMSA) results indicated
that the absence of any repair by hAGT was not a result of binding. OGT was the only AGT to show
activity in the repair of oligonucleotides containing the mono-adducts O4-butyl-4-ol-2′-deoxythymidine
and O4-heptyl-7-ol-2′-deoxythymidine. Binding experiments conducted with hAGT demonstrated that the
protein bound O4-alkylthymine lesions with similar affinities to O6-methylguanine, which hAGT repairs
efficiently, suggesting the lack of O4-alkylthymine repair by hAGT is not a function of recognition.
damage is challenging to repair as information on both DNA
strands is affected. Removal of the ICL on one strand of DNA
Introduction
Interference by ICLs of critical cell events involving DNA
unwinding is exploited in cancer chemotherapy regimens
employing bis-alkylating agents.1 The potency of these agents
may be diminished by the ability of cancer cells to repair the
very lesions they induce. Numerous DNA repair pathways
including direct, base- and nucleotide-excision repair (NER),
homologous recombination (HR), non-homologous end joining
and DNA-mismatch repair remove various DNA lesions.2 Some
pathways, such as NER, are complex with broad substrate
specificities whereas direct repair by AGT, which involves one
protein, has a narrow range of damage that it repairs. ICL
leaves behind damage on the opposing strand, complicating
error-free repair from a template strand. Repair pathways such as
NER, HR and translesion DNA synthesis (TLS) have all been
implicated in ICL repair in mammalian cells, however, there is
an increasing realization that ICL processing may depend on the
nature of the lesion making it challenging to generalize how
specific ICLs are repaired.3
An approach pursued by a number of groups to enhance our
understanding of these processes involves the preparation of ICL
DNA substrates by solution and solid-phase synthesis for repair
studies.4 A number of elegant examples have been reported
including the preparation of various phosphoramidites of nucleo-
sides and other molecules to introduce site-specific lesions in
DNA for various experiments.5 Repair and binding experiments
can be conducted by incorporating the oligonucleotides contain-
ing lesions into specific plasmids.6
Department of Chemistry and Biochemistry Concordia University, 7141
Sherbrooke Street West, Montreal, Canada.
E-mail: cwilds@alcor.concordia.ca; Fax: +1-514-848-2424 ext.5798
†In memory of Professor Har Gobind Khorana (1922–2011), acknowl-
edging his legacy to the scientific community.
Due to the instability of certain ICL formed in DNA as a con-
sequence of DNA treatment with bifunctional alkylating agents
it is at times necessary to modify the structure of the ICL by
producing a mimic similar in structure but exhibiting improved
stability to enable biophysical and repair studies. For example,
‡Electronic supplementary information (ESI) available: 1H and 31P
NMR Spectra, ESI-MS Spectra of DNA, C-18 RP-HPLC traces of phos-
phodiesterase digests of DNA, Tm and CD data of mono-adduct DNA,
DNA Repair gels and results by AGTs and EMSA binding data. See
DOI: 10.1039/c2ob25705j
7078 | Org. Biomol. Chem., 2012, 10, 7078–7090
This journal is © The Royal Society of Chemistry 2012

View Article Online
the alkyl sulfamate hepsulfam and the nitrogen mustard mechlor-
ethamine both form similar ICL between two N7 atoms of
guanine in DNA containing a 5′-GNC.7 Alkylation at the N7
atom of guanine destabilizes the nucleoside. For example, the
half-life of the mechlorethamine induced ICL is of the order of
hours, which can limit experiments that may be conducted.8
Our lab recently reported the preparation of a chemically
stable O6-2′-deoxyguanosine-alkyl-O6-2′-deoxyguanosine (O6-
dG-alkyl-O6-dG) ICL via a versatile synthesis strategy that
enables introduction of an alkyl lesion of any desired length and
orientation.9 Repair studies of the O6-dG-alkyl-O6-dG ICL with
AGT proteins were conducted. AGT proteins, which are found
in all kingdoms of life, are responsible predominantly for the
repair of O6-methyl-2′-deoxyguanosine (O6MeG) lesions, which
are renowned for causing point mutations by disrupting normal
Watson–Crick base pairing proving that this protein is integral in
maintaining genomic integrity.10 hAGT, the most thoroughly
characterized AGT protein, repairs alkylated DNA by flipping
the damaged base out of the helix and into its active site where
the alkyl group is transferred from the point of lesion to the
active site Cys145 residue.11 Once alkylated, this protein is
degraded by the ubiquitin pathway.12
Fig. 1 Structures of (a) O4-2′-deoxythymidine-alkyl-O4-2′-deoxythy-
midine interstrand cross-link, (b) duplex where T–T is the ICL, (c) O4-
alkyl-2′-deoxythymidine mono-adducts and (d) DNA sequence where X
contains the mono-adduct.
The formation of O4-methyl-2′-deoxythymidine (O4MeT) in
DNA is highly mutagenic due to the preferential incorporation
of dG opposite this alkylated base during replication by DNA
polymerases, as indicated previously.17 O4MeT is a minor alky-
lation product with respect to its mutagenic O6MeG counterpart.
DNA treated with N-methyl-N-nitrosourea produces 126 times
more O6MeG than O4MeT.20 Though found in relatively low
abundance this lesion is important as it has been observed in the
DNA of smokers and is partly responsible for the adverse effects
of tobacco.21
It was found that hAGT can remove an O6-dG-alkyl-O6-dG
ICL in mismatched13 and 5′-GNC sequence motif DNA, the
latter designed to mimic the lesion formed by hepsulfam.14 We
reported that hAGT reduces hepsulfam cytotoxicity in exper-
iments with CHO cells, suggesting that this drug may form O6-
dG-alkyl-O6-dG ICL given that hAGT repairs O6-alkyated-
guanine. Interestingly, neither AGT homologues from E. coli
(Ada-C or OGT) were capable of repairing these ICL.
DNA mismatches can arise from heteroduplexes formed during
homologous recombination, errors in DNA replication, deamina-
tion and base damage by alkylating agents.15 The exocyclic atoms
of mispaired nucleobases are more susceptible to alkylation as a
result of altered hydrogen bonding patterns that expose these
atoms to the external environment, as observed by the enhanced
reactivity of mispaired thymidine residues to osmium tetroxide.16
Moreover, formation of mutagenic alkylated nucleobases, such as
O4MeT, can cause mispairs when encountered by DNA poly-
merases to form, for example, an O4MeT:dG mismatch.17
In mammalian cells, O4MeT is more toxic than O6MeG both
in normal and repair deficient systems.22 Unlike O6MeG,
O4MeT is not affected by the expression or suppression of
hAGT activity nor by mismatch repair (MMR) explaining its
added toxicity. NER deficient cells exhibit increased sensitivity
to O4MeT indicating that this lesion is possibly repaired via the
NER pathway in mammalian cells.23 E. coli has evolved two
proteins for the direct repair pathway involving the repair of
O4MeT lesions, OGT and the C-terminal of the adaptive
response protein Ada, highlighting the importance of repairing
this lesion in vivo, underscoring an interesting difference in sub-
strate preference between the human and E. coli proteins.24
Formation of pyrimidine–pyrimidine ICL introduced by
bifunctional alkylating agents have been reported. DNA contain-
ing mispaired dC residues treated with mechlorethamine has
been shown to form ICL linking N3 atoms.18 This recent discov-
ery of ICL formation involving two mispaired pyrimidines is
unique and unprecedented. Not much is known about the proper-
ties of these ICL because of their recent appearance and is an
area of research that requires further investigation.
Based on a straightforward synthetic methodology and the
observation that directly opposed O6-dG-alkyl-O6-dG ICL could
be repaired by hAGT, we set out to prepare O4-2′-deoxythymi-
dine-alkyl-O4-2′-deoxythymidine (O4-dT-alkyl-O4-dT) ICL in a
directly opposed motif as well as their respective mono-adducts
(Fig. 1) by adapting methods described by Swann’s group requir-
ing the preparation of convertible nucleosides.19 These ICL were
prepared by solid-phase oligonucleotide synthesis and the
influence of various alkyl linker lengths on duplex stability and
structure, as well as recognition and repair by AGTs from human
and E. coli, were explored.
Results and discussion
Chemical synthesis of modified nucleosides
The structures of the O4-dT-alkyl-O4-dT cross-links and the
mono-adducts are shown in Fig. 1 in addition to the sequences
of the respective cross-links (XLTT4 and XLTT7) and single
stranded oligonucleotides (S4 and S7). The synthesis of the O4-
dT-alkyl-O4-dT cross-linked bis-phosphoramidites 4a and 4b,
containing a four or seven carbon linker, is shown in Scheme 1.
The convertible nucleoside 3′-O-(tert-butyldimethylsilyl)-5′-O-
(4,4′-dimethoxytrityl)-4-triazolyl-2′-deoxythymidine was trans-
formed into adducts 1a or 1b with the sodium salt of 1,4-butane-
diol or 1,7-heptanediol, respectively.19 Reacting either 1a or 1b
with the convertible nucleoside yielded the fully protected
dimers 2a and 2b. The 3′-OH groups of these compounds,
protected as silyl ethers, were liberated by treatment with tetra-
butylammonium fluoride to produce compounds 3a and 3b,
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 7078–7090 | 7079

View Article Online
NMR at 148.08 and 148.28 ppm for 7a and 148.08 and
148.28 ppm for 7b (see ESI‡).
Synthesis of modified and ICL DNA
The solid-phase syntheses of cross-linked duplexes XLTT4 and
XLTT7 were performed on a 1 μmol scale using either bis-phos-
phoramidite 4a or 4b to prepare the oligomers containing the
four or seven carbon cross-link, respectively. Given the lability
of the cross-link, ‘fast-deprotecting’ 3′-O-2′-deoxyphosphorami-
dites were employed (0.1 M in acetonitrile) with phenoxyacetic
anhydride as the capping agent to prevent an undesired N-acetyl-
ation reaction that could occur if acetic anhydride was used.25
Moreover, the directly opposed motif allows for ICL duplex
synthesis without the need to remove additional protecting
groups around the cross-linked site which could compromise the
ICL or the use of 5′-O-2′-deoxyphosphoramidites. The concen-
tration of 3′-O-bis-phosphoramidite 4a or 4b in acetonitrile was
0.05 M, and each were allowed to react for an extended period
(30 min) to ensure that both phosphoramidite functionalities
would couple to two oligonucleotide chains on the solid-
support, thus enabling ICL formation.26 Solid-phase assembly of
the duplex was completed by extension with 3′-O-2′-deoxyphos-
phoramidites at the cross-linked site.
Scheme 1 Reagents and conditions: (i) sodium salt of 1,4-butanediol
(1a) or 1,7-heptanediol (1b), dioxane, (ii) 3′-O-(tert-butyldimethylsilyl)-
5′-O-(4,4′-dimethoxytrityl)-4-triazolyl-2′-thymidine, 1,8-diazabicyclo-
[5.4.0]undec-7-ene, pyridine, (iii) tetrabutylammonium fluoride (1 M in
THF), THF, (iv) N,N-diisopropylaminocyanoethylphosphonamidic
chloride, diisopropylethylamine, THF.
A variation to the standard oligonucleotide deprotection proto-
col was required due to the labile nature of O4-alkylthymine
adducts. These milder deprotection conditions utilized a solution
of 10% 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) in anhydrous
ethanol for 48 h at room temperature to reduce the likelihood of
ICL cleavage linking the O4 atoms upon deprotection.19 The
crude deprotection mixture containing duplexes of XLTT4 and
XLTT7 were purified by polyacrylamide gel electrophoresis,
rather than SAX HPLC, due to difficulties in removing excess
DBU used in the deprotection step. The purified oligonucleotides
were eluted from the gels and desalted. ESI mass spectrometry
analysis of XLTT4 and XLTT7 revealed they had molecular
weights of 6715.6 and 6758.3 Da (expected 6716.5 and
6758.6 Da), consistent with the expected values (see ESI‡). The
composition of the cross-linked oligomers XLTT4 and XLTT7
was further confirmed by digestion to their constituent nucleo-
sides with a mixture of snake venom phosphodiesterase and calf
intestinal phosphatase followed by C-18 reversed phase HPLC
analysis (see ESI‡). The presence of cross-linked nucleosides
was established by the appearance of one additional peak, other
than the four standard nucleosides, that eluted from the HPLC
with retention times of 21.7 for the four carbon and 29.9 min for
the seven carbon O4-dT-alkyl-O4-dT cross-link. The later elution
of the seven carbon cross-link can be attributed to the greater
hydrophobicity of the longer alkyl linker. The ratios extracted
from the chromatogram of the component 2′-deoxynucleosides
and cross-linked nucleosides were in good agreement with the
theoretical compositions (see ESI‡).
which were then phosphitylated with N,N-diisopropylaminocyano-
ethylphosphonamidic chloride in the presence of diisopropyl-
ethylamine to produce bis-phosphoramidites 4a and 4b. 31P
NMR signals at 148.07 and 148.28 ppm for 4a and 148.07 and
148.28 ppm for 4b confirmed the presence of the phosphorami-
dite functionality in these molecules (see ESI‡).
The mono-adduct lesions were introduced at the O4 atom of
dT as shown in Scheme 2. Starting with compound 1a or 1b, the
free terminal alcohol functionality of the adduct was protected
with phenoxyacetyl chloride to yield 5a or 5b. The 3′-OH
groups of these nucleosides were deprotected prior to the phos-
phitylation step by treatment with tetrabutylammonium fluoride
to yield 6a or 6b. They were then converted to phosphoramidites
as confirmed by the presence of two diagnostic signals in 31P
For the mono-adduct lesions, O4-butyl-4-ol-dT and O4-heptyl-
7-ol-dT, introduced in sequences S4 and S7 (see Fig. 1), similar
solid-phase synthesis conditions as described for the ICL sub-
strates were employed. For these oligonucleotides, cleavage and
deprotection conditions involved incubation of the CPG (con-
trolled pore glass) with a 10% solution of DBU in 1,4-butanediol
Scheme 2 Reagents and conditions: (i) phenoxyacetyl chloride, tri-
ethylamine, THF, (ii) tetrabutylammonium fluoride (1 M in THF), THF,
(iii) N,N-diisopropylaminocyanoethylphosphonamidic chloride, diiso-
propylethylamine, THF.
7080 | Org. Biomol. Chem., 2012, 10, 7078–7090
This journal is © The Royal Society of Chemistry 2012

View Article Online
Table 1 Thermodynamic parameters of DNA duplexes
(for S4) or 1,7-heptanediol (for S7) for 5 days at room tempera-
ture to ensure that the adducts were not compromised.19 The
DBU was neutralized with acetic acid before the CPG was
rinsed with acetonitrile (50% in water) to solubilize the fully
deprotected product as a result of the poor solubility of DNA in
the diols. These oligonucleotides were also characterized by ESI
mass spectrometry and nuclease digestion to ensure their compo-
sition (see ESI‡).
DNA
ΔH° (kJ mol−1
)
ΔS° (J mol−1 K−1
)
Control (T–A)
XLTT4
XLTT7
−298 23
−823 71
−735 20
−551 12
−249
−179
7
4
UV thermal denaturation studies of DNA duplexes
The effect of the directly opposed O4-dT-alkyl-O4-dT ICL on
duplex stability was assessed through UV thermal denaturation
experiments. The melting profiles of XLTT4 and XLTT7,
shown in Fig. 2, exhibit a sigmoidal denaturation and hyperchro-
micities comparable to the control duplex containing a T–A base
pair instead of the cross-link. The melting temperatures observed
for the cross-linked oligonucleotides XLTT4 and XLTT7 were
66 and 50 °C, respectively, higher than the control duplex (with
a Tm of 44 °C).
This melting temperature trend was observed earlier by our
group for duplexes containing a directly opposed O6-2′-deoxy-
guanosine-alkyl-O6-2′-deoxyguanosine mismatch present in an
11 bp duplex with increases in Tm of 23 and 33 °C for the heptyl
and butyl linkages, respectively, over the control duplex.27 This
same tendency has also been reported when ICL of various
lengths are introduced between the N3 atoms of thymidine where
an increase in Tm of 16 and 31 °C were observed for heptyl and
butyl linkages over the non-alkylated control.28
Thermodynamic parameters of the ICL and control duplexes
were studied to investigate the origin of the increased thermal
stability observed for the ICL DNA (Table 1). Varying duplex
concentration did not affect the Tm observed for the ICL
duplexes but did influence that of the control, as expected. The
thermodynamic parameters indicate a reduction in the enthalpy
for the ICL containing duplex, which was more pronounced for
the heptyl versus butyl linker. One possible reason may be due
to the disruption of the base pairs around the cross-linked site as
a result of steric effects of the larger linker.
Fig. 3 Far-UV circular dichroism spectra of cross-linked duplexes
XLTT4 (— —), XLTT7 (- - -) and non-cross-linked DNA (—).
The presence of the ICL caused an increase in the entropy of
the duplex due to the preorganized nature of the unimolecular
system thus favouring thermal stability of the DNA. The contri-
bution of the increased entropy outweighed the negative enthal-
pic effect in the ICL system, consistent with what has been
previously observed with ICL formed by trans-diamminedichloro-
platinum(II).29
The influence of alkyl chain length of the mono-adducts at the
O4 atom of 2′-deoxythymidine on stability was also assessed.
There was a 10 °C reduction in duplex stability for the duplexes
containing O4MeT, O4-butyl-4-ol-dT (S4) or O4-heptyl-7-ol-dT
(S7) (48 °C) compared to the non-alkylated control (58 °C)
demonstrating that for these adducts the length of the alkyl chain
attached at the O4 atom of 2′-deoxythymidine had no substantial
effect on duplex stability (see ESI‡). These results are supported
by previous studies demonstrating that O4-ethyl-2′-deoxythm-
idine/2′-deoxyadenosine base pair within a duplex decreased the
T
m by 14 °C with respect to the native control.30
Circular dichroism spectroscopy of DNA duplexes
CD spectra of the cross-linked duplexes XLTT4 and XLTT7
exhibited signatures characteristic of B-form DNA with a posi-
tive signal centered around 275 nm, a crossover around 260 nm
and a negative signal around 250 nm, as shown in Fig. 3.31
These results indicate that the O4-dT-alkyl-O4-dT ICL caused
minimal distortion to the global B-form structure.
Similarly, duplexes containing O4MeT, O4-butyl-4-ol-dT (S4)
or O4-heptyl-7-ol-dT (S7), showed little deviation in their CD
signature relative to the unmodified control duplex demonstrating
minimal distortion from the global B-form structure. These
findings are in agreement with NMR reports on O4MeT contain-
ing duplexes that suggest such lesions do not cause major struc-
tural deformation of the duplex.32
Fig. 2 Absorbance (A260) versus temperature profiles of cross-linked
duplexes XLTT4 (— —), XLTT7 (- - -) and non-cross-linked DNA
(—).
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 7078–7090 | 7081

View Article Online
Repair and binding of AGTs with cross-linked and
mono-alkylated substrates
Experiments probing the repair of larger alkyl adducts at the
O4 atom of dT show a different trend than the ones observed for
methylation (see ESI‡). For the larger lesions, O4-butyl-4-ol-dT
and O4-heptyl-7-ol-dT, only OGT showed any ability to remove
these alkylations to a reasonable level, indicating only OGT out
of all 3 proteins is capable of repairing large lesions at the
O4 atom of thymine.
Repair studies conducted on XLTT4 and XLTT7 cross-links
revealed that none of the AGT proteins from human or E. coli
were able to completely eliminate the cross-links, as observed in
Fig. 4. Moreover, none of the proteins were able to create the
intermediate product, which consists of one repair step leading
to an AGT–ssDNA species.13,14
EMSAwere conducted with XLTT4, XLTT7 and a non-cross-
linked control duplex, which has a T–A match where the T–T
cross-link is located, to determine if the lack of activity of hAGT
towards these lesions was inherent to the binding and detection of
the lesion or if the lack of repair was a function of catalysis.
Results of the EMSA indicate that the hAGT stoichiometry for
the various DNA was the same, with two proteins for every 11 bp
duplex (see ESI‡). These results are consistent with the findings of
others, which stipulates that hAGT binds every 4 nucleotides.33
This stoichiometry was the same for the cross-linked DNA and for
the control demonstrating that the presence of the cross-link does
not affect how hAGT interact together when bound to the DNA.
The dissociation constant of the control duplex was much higher
than the cross-links, which was unanticipated.
To assess the lack of repair of the adducts at the O4 atom of dT
by hAGT, binding assays were conducted. A general trend from
the EMSA could be observed for the binding affinity of hAGT to
the various mono-adduct DNA (see ESI‡). First, hAGT had very
similar binding affinities to all mono-adduct DNA, which was
roughly 1.8 times better than the non-alkylated control. Second,
hAGT bound O4-alkyl-dT lesions with very similar affinity as it
did O6MeG and the size of the lesion at the O4 atom of dT had
no substantial effect on binding for the alkylation sizes tested.
The observed increase in binding affinity for the adduct or ICL
containing DNA and hAGT over their unalkylated controls suggests
hAGT is capable of flipping out the alkylated thymidine and
placing the lesion into its active site as required for repair. These
results are consistent with our previous findings which demonstrate
that hAGT is capable of flipping out cross-linked nucleotides as
demonstrated through the repair of O6-dG-alkyl-O6-dG ICL.13,14
To address the low binding affinity of hAGT for the control
duplex a set of oligos were designed, where the central T in the
Modelling of O4MeT and O4-butyl-4-ol-dT in hAGT active site
control sequence (5′ CGA AAT TTT CG) was replaced by the
̲
other 3 natural 2′-deoxynucleotides, and assessed for binding by
hAGT. Binding studies were conducted over a large range of
protein to obtain a rough estimate of the Kd for the hAGT–DNA
complexes. Results indicate that hAGT preferentially binds to
DNA with a dG nucleotide that is further than 2 nucleotides
from the ends (see ESI‡). This same hAGT binding trend has
been observed for long A–T tracts.34 The lack of a central dG
explains the low binding affinity observed for the control
sequence. The increased binding affinity observed for XLTT4
and XLTT7 over the control may be attributed to the large
hydrophobic adduct at the O4 atoms of the central dT, due to the
presence of the linkers, and not as a result of the thermodynamic
properties of the resulting DNA duplexes. This hydrophobic
interaction between alkyl adduct and hAGT has been proposed
as the basis of substrate discrimination by the protein for O6MeG
and O6-benzyl-dG over dG.35
To understand the underlying substrate discrimination observed
for AGT with respect to the mono-adducts, molecular modelling
was performed (coordinates for these models are provided in the
ESI‡). hAGT is the only AGT that has been co-crystallized with
a DNA substrate providing valuable insight into its active con-
formation. The substrate differences observed between various
AGT homologues suggest substantial variations in the active
sites amongst proteins. Attempts to model methyl and butyl-4-ol
adducts with a homology model of OGT based on hAGT (gener-
ated from PDB ID: 1T38) were performed. However, the mini-
mized structures did not correlate with the experimental
evidence. Our modelling results suggest the methyl adduct fits in
the active site of hAGT without steric clashes (Fig. 5). The
Fig. 4 Denaturing PAGE to demonstrate lack of repair of XLTT4 and
XLTT7 by hAGT, Ada-C and OGT involving 2 pmol of DNA with
60 pmol of AGT. Lane 1, Control DNA; lane 2, XLGG7 + hAGT; lane
3, XLTT4; lane 4, XLTT4 + hAGT; lane 5, XLTT4 + Ada-C; lane 6,
XLTT4 + OGT; lane 7, Control DNA; lane 8, XLGG7 + hAGT; lane 9,
XLTT7; lane 10, XLTT7 + hAGT; lane 11, XLTT7 + Ada-C; lane 12,
XLTT7 + OGT.
Fig. 5 Overlay of O4MeT (cyan) and O4-butyl-4-ol-dT (pink) in the
active site of hAGT (white). The adopted rotamers of Cys145 are shown
for O4MeT (cyan) and O4-butyl-4-ol-dT (pink) and His146 (grey).
7082 | Org. Biomol. Chem., 2012, 10, 7078–7090
This journal is © The Royal Society of Chemistry 2012

View Article Online
model of the butyl-4-ol adduct proposes a steric clash in the
active site of the protein between the γ-carbon of the lesion and
the protein due to the adoption of the gauche conformation at
the β-carbon of the adduct. This interaction causes Cys145 to
adopt the incorrect rotamer for the alkyl transfer reaction to take
place when the butyl-4-ol adduct is present. In the mechanism of
hAGT repair, His146 is responsible for proton abstraction of a
nearby water molecule from Cys145 allowing the formation of
the thiolate anion. In the case where Cys145 is rotated away
from His146 (as observed for the butyl-4-ol lesion) the thiolate
anion will not be produced and the protein will be inactive.
Biocorp (Montreal, QC). Oligonucleotides containing O4MeT
required for repair studies were a kind gift from TriLink Bio-
technologies (San Diego, CA).
Chemical synthesis of nucleosides
The synthesis of phosphoramidites 4a, 4b, 7a and 7b are shown
in Schemes 1 and 2.
3′-O-(tert-Butyldimethylsilyl)-5′-O-(4,4′-dimethoxytrityl)-O4-
(hydroxybutyl)-thymidine (1a)
1,4-Butanediol (1.60 g, 17.8 mmol) was dissolved in THF
(15 mL). Sodium metal (2.35 g, 102 mmol) was added. After
5 h, the salt was extracted to be immediately used in the sub-
Experimental
General methods and materials
5′-O-Dimethoxytrityl-N2-phenoxyacetyl-2′-deoxyguanosine, 3′-
O-dimethoxytrityl-2′-deoxyribonucleoside-5′-O-(β-cyanoethyl-
N,N′-diisopropyl)phosphoramidites and N,N-diisopropylamino-
cyanoethylphosphonamidic chloride were purchased from Chem-
Genes Inc. (Wilmington, MA). 5′-O-Dimethoxytrityl-2′-
deoxyribonucleoside-3′-O-(β-cyanoethyl-N,N′-diisopropyl)phos-
phoramidites and protected 2′-deoxyribonucleoside–CPG sup-
ports were purchased from Glen Research (Sterling, Virginia).
All other chemicals and solvents were purchased from the
Aldrich Chemical Company (Milwaukee, WI) or EMD Chemi-
cals Inc. (Gibbstown, NJ). Flash column chromatography was
performed using silica gel 60 (230–400 mesh) obtained from
Silicycle (Quebec City, QC). Thin layer chromatography (TLC)
was performed using precoated TLC plates (Merck, Kieselgel
60 F254, 0.25 mm) purchased from EMD Chemicals Inc. (Gibbs-
town, NJ). NMR spectra were recorded on a Varian 500 MHz
sequent
reaction.
3′-O-(tert-Butyldimethylsilyl)-5′-O-(4,4′-
dimethoxytrityl)-4-triazolyl-thymidine (2.5 g, 3.52 mmol) was
dissolved in dioxane (10 mL).16 Sodium 4-hydroxybutan-1-olate
was added dropwise. After 15 h, the solvent was evaporated
in vacuo, the crude product was taken up in dichloromethane
(50 mL) and the solution was washed with two portions of
sodium bicarbonate (2 × 50 mL). The organic layer was dried
over sodium sulfate and concentrated to produce a yellow gum.
The crude product was purified by flash column chromatography
using a hexanes–ethyl acetate solvent system (2 : 3) to afford
2.05 g (79.8%) of product as a colorless foam. Rf (SiO2 TLC):
0.24 hexanes–ethyl acetate (2 : 3). λmax(ACN) = 282 nm. 1H NMR
(500 MHz, CDCl3, ppm): 7.98 (s, 1H, H6), 7.30–7.45 (m, 10H,
Ar), 6.85–6.88 (m, 3H, Ar), 6.36 (dd, 1H, H1′, J = 6 Hz), 4.53
(m, 1H, H3′), 4.47 (m, 2H, ArOCH2), 4.02 (m, 1H, H4′), 3.85
(s, 6H, OCH3), 3.75 (t, 2H, CH2), 3.56 (dd, 1H, H5′), 3.31 (dd,
1H, H5′), 2.56 (m, 1H, H2′), 2.27 (m, 1H, H2′), 1.90 (tt, 2H,
CH2), 1.74 (tt, 2H, CH2), 1.64 (s, 1H, OH), 1.57 (s, 3H,
ArCH3), 0.87 (s, 9H, SiC(CH3)3), 0.08 (s, 6H, Si(CH3)2). 13C
NMR (125.7 MHz, CDCl3, ppm): 170.4, 158.7, 156.0 144.4,
139.5, 135.53, 135.51, 130.11, 130.10, 128.2, 127.9, 127.1,
113.24, 113.21, 104.6, 86.7, 86.5, 86.3, 71.0, 67.1, 62.4, 62.2,
42.2, 29.1, 25.7, 25.1, 17.9, 11.7, −4.62, −4.94. IR (thin film);
νmax (cm−1) = 3414, 3061, 2953, 1669, 1509, 1252, 1178, 1109,
1035, 835, 782. HRMS (ESI-MS) m/z calculated for
C41H54N2O8SiNa+ 753.3546: found 753.3555 [M + Na]+.
1
NMR spectrometer at room temperature. H NMR spectra were
recorded at a frequency of 500.0 MHz and chemical shifts were
reported in parts per million downfield from tetramethylsilane.
13C NMR spectra (1H decoupled) were recorded at a frequency
of 125.7 MHz with tetramethylsilane as a reference. 31P NMR
spectra (1H decoupled) were recorded at a frequency of
202.3 MHz with H3PO4 used as an external standard. High
resolution mass spectrometry of modified nucleosides were
obtained using an LTQ OrbitrapVelos – ETD (Thermo Scien-
tific) at the Concordia University Centre for Biological Appli-
cations of Mass Spectrometry. The mass spectrometer was
operated in full scan, positive ion detection mode. Ampicillin,
isopropyl β-D-thiogalactopyranoside (IPTG), and most other
biochemical reagents as well as polyacrylamide gel materials
were purchased from Bioshop Canada Inc (Burlington, ON).
Ni-NTA Superflow Resin was purchased from Qiagen (Missis-
sauga, ON). Complete, Mini, EDTA-free Protease Inhibitor
Cocktail Tablets were obtained from Roche (Laval, QC) Nitro-
cellulose filters (0.20 μm) were obtained from Millipore. XL-10
Gold and Bl21(DE3) E. coli cells were obtained from Stratagene
(Cedar Creek, TX). DpnI, T4 polynucleotide kinase (PNK),
Unstained Protein Molecular Weight Marker and restriction
enzymes EcoRI and KpnI were obtained from Fermentas (Bur-
lington, ON). [γ-32P]ATP was purchased from Amersham
Canada Ltd (Oakville, ON). Phusion Polymerase was obtained
from New England Biolabs (Ipswich, MA). DNA primers for
site directed mutagenesis and cloning were purchased from
3′-O-(tert-Butyldimethylsilyl)-5′-O-(4,4′-dimethoxytrityl)-O4-
(hydroxyheptyl)-thymidine (1b)
1,7-Heptanediol (1.7 g, 11.35 mmol) was dissolved in THF
(10 mL). Sodium metal (1.7 g, 65.2 mmol) was added. After
5 h, the salt was extracted to be immediately used in the sub-
sequent
reaction.
3′-O-(tert-Butyldimethylsilyl)-5′-O-(4,4′-
dimethoxytrityl)-4-triazolyl-thymidine (1.7 g, 1.13 mmol) was
dissolved in dioxane (5 mL).16 Sodium 7-hydroxyheptan-1-olate
was added dropwise. After 15 h, the crude product was taken up
in ethyl acetate (50 mL) and the solution was washed with two
portions of sodium bicarbonate (2 × 50 mL). The organic layer
was dried over sodium sulfate and concentrated to produce a
yellow gum. The crude product was purified by flash column
chromatography using a hexanes–ethyl acetate solvent system
(6 : 4). Further purification of the product was achieved by flash
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 7078–7090 | 7083

View Article Online
column chromatography using an acetonitrile–dichloromethane
(1 : 9 → 3 : 25) solvent system to afford 0.594 g (42.1%) of
product as a colorless foam. Rf (SiO2 TLC): 0.43 hexanes–ethyl
acetate (2 : 3). λmax(ACN) = 282 nm. 1H NMR (500 MHz, CDCl3,
ppm): 7.96 (s, 1H, H6), 7.29–7.34 (m, 10H, Ar), 6.86–6.88 (m,
3H, Ar), 6.37 (dd, 1H, H1′, J = 5.75 Hz), 4.53 (dd, 1H, H3′),
4.24 (t, 2H, ArOCH2), 4.02 (m, 1H, H4′), 3.85 (s, 6H, OCH3),
3.69 (m, 2H, CH2OH), 3.56 (dd, 1H, H5′), 3.30 (dd, 1H, H5′),
2.55 (m, 1H, H2′), 2.25 (m, 1H, H2′), 1.80 (m, 2H, CH2), 1.63
(m, 2H, CH2), 1.57 (s, 3H, CH3Ar), 1.47 (m, 6H, (CH2)3), 0.89
(s, 9H, SiC(CH3)3), 0.08 (s, 6H, Si(CH3)2). 13C NMR
(125.7 MHz, CDCl3, ppm): 170.5, 158.7, 156.0 144.4, 139.4,
135.56, 135.55, 130.12, 130.10, 128.2, 127.9, 127.1, 113.24,
113.21, 104.6, 86.7, 86.5, 86.3, 71.1, 67.3, 62.9, 62.4, 55.3, 42.2,
32.7, 29.8, 29.0, 28.4, 25.8, 25.7, 25.6, 17.9, 11.8, −4.62, −4.94.
IR (thin film); νmax (cm−1) = 3420, 3060, 2930, 1668, 1509, 1252,
1177, 1109, 1035, 834, 782. HRMS (ESI-MS) m/z calculated for
C44H60N2O8SiNa+ 795.4017: found 795.4017 [M + Na]+.
washed with four portions of sodium bicarbonate (4 × 50 mL).
The organic layer was dried over sodium sulfate and concen-
trated to produce a yellow gum. The crude product was purified
by flash column chromatography using an acetonitrile–dichloro-
methane (1 : 9 → 3 : 25) solvent system to afford 0.358 g
(39.5%) of product as a colorless foam. Rf (SiO2 TLC):
0.49 hexanes–ethyl acetate (2 : 3). λmax(ACN) = 282 nm. 1H NMR
(500 MHz, CDCl3, ppm): 7.99 (s, 2H, H6), 7.33–7.49 (m, 20H,
Ar), 6.89–6.91 (m, 6H, Ar), 6.40 (dd, 2H, H1′, J = 12 Hz), 4.54
(dd, 2H, H3′), 4.44 (t, 4H, ArOCH2), 4.03 (m, 2H, H4′), 3.86 (s,
12H, OCH3), 3.58 (dd, 2H, H5′), 3.34 (dd, 2H, H5′), 2.58 (m,
2H, H2′), 2.26 (m, 2H, H2′), 1.81 (m, 4H, CH2), 1.59 (s, 6H,
CH3Ar), 1.48 (m, 6H, (CH2)3), 0.88 (s, 18H, SiC(CH3)3), 0.07
(s, 12H, Si(CH3)2). 13C NMR (125.7 MHz, CDCl3, ppm): 170.5,
158.7, 156.0 144.4, 139.4, 135.5, 130.13, 130.11, 128.2, 127.9,
127.1, 113.24, 113.21, 104.6, 86.7, 86.5, 86.3, 71.1, 67.3, 62.4,
55.3, 42.2, 29.0, 28.5, 25.9, 25.7, 17.9, 11.8, −4.61, −4.94. IR
(thin film); νmax (cm−1) = 3058, 2952, 1671, 1533, 1508, 1327,
1252, 1177, 1035, 834. HRMS (ESI-MS) m/z calculated for
C81H104N4O14Si2Na+ 1435.6985: found 1435.7037 [M + Na]+.
1-{O4-[3′-O-(tert-Butyldimethylsilyl)-5′-O-(4,4′-dimethoxytrityl)-
thymidyl]}-4-{O4-[3′-O-(tert-butyldimethylsilyl)-5′-O-
(4,4′-dimethoxytrityl)-thymidyl]}-butane (2a)
1-{O4-[5′-O-(4,4′-Dimethoxytrityl)-thymidyl]}-4-{O4-[5′-O-
(4,4′-dimethoxytrityl)-thymidyl]}-butane (3a)
1a (0.154 g, 0.205 mmol) and 3′-O-(tert-butyldimethylsilyl)-5′-
O-(4,4′-dimethoxytrityl)-4-triazolyl-thymidine
(0.117
g,
2a (0.658 g, 0.485 mmol) was dissolved in THF (4.8 mL).
TBAF (1 M in THF) (1.97 mL, 1.97 mmol) was added drop-
wise. After 30 min, the crude product was taken up in ethyl
acetate (50 mL) and the solution was washed with two portions
of sodium bicarbonate (2 × 50 mL). The organic layer was dried
over sodium sulfate and concentrated to produce a yellow gum.
The crude product was purified by flash column chromatography
using a methanol–hexanes–ethyl acetate (1 : 5 : 4 → 1 : 4 : 5)
solvent system to afford 0.514 g (93.0%) of product as a color-
less foam. Rf (SiO2 TLC): 0.21 methanol–hexanes–ethyl acetate
(1 : 5 : 4). λmax(ACN) = 282 nm. 1H NMR (500 MHz, CDCl3,
ppm): 7.78 (s, 2H, H6), 7.23–7.37 (m, 20H, Ar), 6.78–6.81 (m,
6H, Ar), 6.34 (dd, 2H, H1′, J = 6 Hz), 4.52 (m, 2H, H3′), 4.40
(m, 4H, CH2), 4.05 (m, 2H, H4′), 3.77 (s, 12H, OCH3), 3.46
(dd, 2H, H5′), 3.35 (dd, 2H, H5′), 2.58 (m, 2H, H2′), 2.25 (m,
2H, H2′), 2.11 (s, 2H, OH),1.85 (m, 4H, CH2), 1.56 (s, 6H,
ArCH3). 13C NMR (125.7 MHz, CDCl3, ppm): 170.4, 158.7,
158.6, 144.5, 139.8, 135.6, 135.5, 130.1, 130.1, 128.1, 128.0,
127.0, 113.3, 105.0, 86.8, 86.5, 86.3, 66.8, 63.5, 60.4, 55.2 42.2,
25.2, 21.1, 14.2, 11.7. IR (thin film); νmax (cm−1) = 3353, 3057,
2953, 1661, 1509, 1431, 1329, 1251, 1177, 1095, 1035, 829,
735. HRMS (ESI-MS) m/z calculated for C66H70N4O14Na+
1165.4786: found 1165.4806 [M + Na]+.
0.164 mmol) was dissolved in pyridine (3 mL). DBU (0.062 g,
0.410 mmol) was added dropwise. After 3 d, the crude product
was taken up in ethyl acetate (100 mL) and the solution was
washed with four portions of sodium bicarbonate (4 × 50 mL).
The organic layer was dried over sodium sulfate and concen-
trated to produce a yellow gum. The crude product was purified
by flash column chromatography using a hexanes–ethyl acetate
(1 : 1 → 2 : 3) solvent system to afford 0.151 g (67.7%) of
product as a colorless foam. Rf (SiO2 TLC): 0.22 hexanes–ethyl
acetate (2 : 3). λmax(ACN) = 282 nm. 1H NMR (500 MHz, CDCl3,
ppm): 7.96 (s, 2H, H6), 7.30–7.46 (m, 20H, Ar), 6.86–6.88 (m,
6H, Ar), 6.37 (dd, 1H, H1′, J = 5.75 Hz), 4.53 (m, 2H, H3′),
4.47 (m, 4H, CH2), 4.01 (m, 2H, H4′), 3.84 (s, 12H, OCH3),
3.57 (dd, 2H, H5′), 3.31 (dd, 2H, H5′), 2.55 (m, 2H, H2′), 2.26
(m, 2H, H2′), 1.93 (m, 4H, CH2), 1.57 (s, 6H, ArCH3), 0.87 (s,
18H, SiC(CH3)3), 0.07 (s, 12H, Si(CH3)2). 13C NMR
(125.7 MHz, CDCl3, ppm): 170.4, 158.7, 155.9 144.4, 139.6,
135.51, 135.50, 130.11, 130.10, 128.2, 127.9, 127.1, 113.24,
113.21, 104.5, 86.7, 86.5, 86.3, 71.1, 66.6, 62.4, 55.2, 42.2,
25.7, 25.3, 17.9, 11.7, −4.62, −4.94. IR (thin film); νmax (cm−1
= 3057, 2954, 1671, 1608, 1509, 1463, 1329, 1252, 1178,
1035, 835, 736. HRMS (ESI-MS) m/z calculated for
C79H98N4O14Si2Na+ 1393.6515: found 1393.6526 [M + Na]+.
)
1-{O4-[5′-O-(4,4′-Dimethoxytrityl)-thymidyl]}-7-{O4-[5′-O-
(4,4′-dimethoxytrityl)-thymidyl]}-heptane (3b)
1-{O4-[3′-O-(tert-Butyldimethylsilyl)-5′-O-(4,4′-dimethoxytrityl)-
thymidyl]}-7-{O4-[3′-O-(tert-butyldimethylsilyl)-5′-O-
(4,4′-dimethoxytrityl)-thymidyl]}-heptane (2b)
2b (0.357 g, 0.263 mmol) was dissolved in THF (2 mL). TBAF
(1 M in THF) (0.63 mL, 0.63 mmol) was added dropwise. After
25 min, the crude product was taken up in ethyl acetate (50 mL)
and the solution was washed with three portions of sodium bicar-
bonate (3 × 50 mL). The organic layer was dried over sodium
sulfate and concentrated to produce a yellow gum. The crude
product was purified by flash column chromatography using a
1b (0.500 g, 0.647 mmol) and 3′-O-(tert-butyldimethylsilyl)-5′-
O-(4,4′-dimethoxytrityl)-4-triazolyl-thymidine
(0.368
g,
0.518 mmol) was dissolved in pyridine (10 mL). DBU (0.197 g,
1.29 mmol) was added dropwise. After 4.5 d, the crude product
was taken up in ethyl acetate (150 mL) and the solution was
7084 | Org. Biomol. Chem., 2012, 10, 7078–7090
This journal is © The Royal Society of Chemistry 2012

View Article Online
methanol–hexanes–ethyl acetate (1 : 5 : 4 → 1 : 4 : 5) solvent
system to afford 0.256 g (82.0%) of product as a colorless foam.
Rf (SiO2 TLC): 0.22 methanol–hexanes–ethyl acetate (1 : 11 : 8).
λmax(ACN) = 282 nm. 1H NMR (500 MHz, CDCl3, ppm): 7.77 (s,
2H, H6), 7.18–7.37 (m, 20H, Ar), 6.79–6.80 (m, 6H, Ar), 6.34
(dd, 2H, H1′, J = 12.5 Hz), 4.50 (m, 2H, H3′), 4.29–4.39 (m,
4H, CH2), 4.05 (m, 2H, H4′), 3.78 (s, 12H, OCH3), 3.46 (dd,
2H, H5′), 3.36 (dd, 2H, H5′), 2.58 (m, 2H, H2′), 2.25 (m, 2H,
H2′), 1.70–1.74 (m, 4H, CH2), 1.56 (s, 6H, ArCH3), 1.39–1.41
(m, 6H, CH2). 13C NMR (125.7 MHz, CDCl3, ppm): 170.6,
158.7, 156.2, 144.5, 139.5, 135.6, 135.5, 130.1, 128.1, 128.0,
127.1, 113.3, 105.0, 86.8, 86.4, 86.1, 71.9, 67.4, 63.4, 55.3,
42.1, 28.5, 25.8, 11.8. IR (thin film); νmax (cm−1) = 3350, 3058,
2932, 1654, 1532, 1508, 1458, 1328, 1251, 1177, 1035, 829,
735. HRMS (ESI-MS) m/z calculated for C69H76N4O14Na+
1207.5255: found 1207.5270 [M + Na]+.
diisopropylaminocyanoethylphosphonamidic chloride (0.085 g,
0.359 mmol) and the reaction was allowed to stir at room temp-
erature for 30 min. Upon completion, the reaction was quenched
by the addition of ethyl acetate (50 mL), and the solution was
washed with sodium bicarbonate (3%, 2 × 25 mL) and brine
(25 mL). The organic layer was dried over sodium sulfate, dec-
anted, and evaporated. The product, a colorless powder, was pre-
cipitated from hexanes (0.146 g, 64.3%). Rf (SiO2 TLC): 0.31,
0.46, 0.62 ethyl acetate. λmax(ACN) = 282 nm. 1H NMR
(500 MHz, d6-acetone, ppm): 7.88–7.92 (s, 2H, H6), 7.48–7.51
(m, 4H, Ar), 7.23–7.40 (m, 14H, Ar), 6.87–6.93 (m, 8H, Ar),
6.30–6.35 (m, 2H, H1′), 4.70–4.76 (m, 2H, H3′), 4.30–4.32 (m,
4H, CH2), 4.18–4.24 (m, 2H, H4′), 3.60–3.91 (m, 20H, CH2OP,
NCH, OCH3), 3.39–3.52 (m, 4H, H5′), 2.77 (t, 2H, CH2CN),
2.53–2.64 (m, 4H, H2′, CH2CN), 2.37–2.42 (m, 2H, H2′),
1.74–1.78 (m, 4H, (CH2)2), 1.58–1.61 (m, 6H, ArCH3)
1.44–1.49 (m, 6H, (CH2)3), 1.18–1.21 (m, 18H, CH3),
1.09–1.11 (m, 6H, CH3). 31P NMR (202.3 MHz, d6-acetone,
ppm): 148.28, 148.07. 13C NMR (125.7 MHz, d6-acetone, ppm):
170.09, 170.08, 158.88, 158.86, 154.7, 145.0, 140.0, 139.9,
135.7, 135.63, 135.61, 135.57, 130.2, 130.1, 128.2, 128.1,
127.9, 127.8, 126.89, 126.86, 118.1, 117.9, 113.1, 103.5, 103.4,
86.62, 86.59, 86.0, 85.9, 85.5, 85.3, 85.2, 73.6, 73.5, 73.1, 73.0,
66.6, 63.2, 63.0, 58.7, 58.5, 54.68, 54.66, 43.1, 43.0, 40.4, 40.3,
29.4, 25.7, 24.04, 24.02, 23.97, 23.96, 23.9, 19.94, 19.89,
19.83, 19.77, 13.6, 11.17, 11.16. IR (thin film); νmax (cm−1) =
3057, 2966, 1672, 1510, 1463, 1329, 1252, 1179, 1036. HRMS
(ESI-MS) m/z calculated for C87H110N8O16P2Na+ 1607.7412:
found 1607.7285 [M + Na]+.
1-{O4-[3′-O-(β-Cyanoethyl-N,N′-diisopropyl)-5′-O-(4,4′-
dimethoxytrityl)-thymidyl]}-4-{O4-[3′-O-(β-cyanoethyl-N,N′-
diisopropyl)-5′-O-(4,4′-dimethoxytrityl)-thymidyl]}-butane (4a)
3a (0.200 g, 0.175 mmol) and diisopropylethylamine (0.068 g,
0.525 mmol) were dissolved in THF (2 mL), followed by N,N-
diisopropylaminocyanoethylphosphonamidic chloride (0.104 g,
0.438 mmol) and the reaction was allowed to stir at room temp-
erature for 30 min. Upon completion, the reaction was quenched
by the addition of ethyl acetate (50 mL), and the solution was
washed with sodium bicarbonate (3%, 2 × 25 mL) and brine
(25 mL). The organic layer was dried over sodium sulfate, dec-
anted, and evaporated. The product, a colorless powder, was pre-
cipitated from hexanes (0.152 g, 56.2%). Rf (SiO2 TLC): 0.12,
0.18, 0.27 ethyl acetate. λmax(ACN) = 282 nm. 1H NMR
(500 MHz, d6-acetone, ppm): 7.83–7.91 (s, 2H, H6), 7.47–7.50
(m, 4H, Ar), 7.21–7.39 (m, 14H, Ar), 6.87–6.91 (m, 8H, Ar),
6.29–6.34 (m, 2H, H1′), 4.70–4.74 (m, 2H, H3′), 4.37 (m, 4H,
CH2), 4.17–4.24 (m, 2H, H4′), 3.57–3.89 (m, 20H, CH2OP,
NCH, OCH3), 3.34–3.51 (m, 4H, H5′), 2.74 (t, 2H, CH2CN),
2.52–2.63 (m, 4H, H2′, CH2CN), 2.34–2.41 (m, 2H, H2′),
1.89–1.91 (m, 4H, (CH2)2), 1.57–1.60 (m, 6H, ArCH3)
1.15–1.20 (m, 18H, CH3), 1.07–1.08 (m, 6H, CH3). 31P NMR
(202.3 MHz, d6-acetone, ppm): 148.28, 148.07. 13C NMR
(125.7 MHz, d6-acetone, ppm): 170.0, 158.9, 158.8, 154.7,
144.97, 145.95, 140.03, 139.97, 135.7, 135.63, 135.60, 135.56,
130.17, 130.15, 128.2, 128.1, 127.87, 127.86, 126.91, 126.88,
118.1, 117.9, 113.1, 103.5, 103.4, 86.64, 86.61, 85.99, 85.9,
85.5, 85.3, 85.2, 73.6, 73.5, 73.1, 73.0, 66.2, 63.2, 63.0, 58.70,
58.69, 54.70, 54.68, 43.1, 43.0, 40.4, 40.3, 29.4, 25.1, 24.05,
24.03, 23.99, 23.93, 19.9, 19.8, 11.18, 11.17. IR (thin film);
νmax (cm−1) = 3031, 2966, 1672, 1534, 1509, 1462, 1329, 1251,
1179, 1035. HRMS (ESI-MS) m/z calculated for
C84H104N8O16P2Na+ 1565.6943: found 1565.6854 [M + Na]+.
3′-O-(tert-Butyldimethylsilyl)-5′-O-(4,4′-dimethoxytrityl)-O4-
(phenoxyacetyloxybutyl)-thymidine (5a)
1a (0.300 g, 0.410 mmol) was dissolved in THF (1.2 mL). Tri-
ethylamine (0.145 g, 1.43 mmol) followed by phenoxyacetyl
chloride (0.122 g, 0.715 mmol) was added dropwise. The reac-
tion was set in an ice bath for 25 min. After 1.5 h, the crude
product was taken up in ethyl acetate (50 mL) and the solution
was washed with three portions of sodium bicarbonate (3 ×
50 mL). The organic layer was dried over sodium sulfate and
concentrated to produce a yellow gum. The crude product was
purified by flash column chromatography using a hexanes–ethyl
acetate (3 : 7) solvent system to afford 0.266 g (75.1%) of
product as a colorless foam. Rf (SiO2 TLC): 0.39 hexanes–ethyl
acetate (1 : 1). λmax(ACN) = 276 nm. 1H NMR (500 MHz, CDCl3,
ppm): 7.98 (s, 1H, H6), 7.26–7.50 (m, 13H, Ar), 6.85–7.09 (m,
5H, Ar), 6.37 (dd, 1H, H1′, J = 6 Hz), 4.69 (s, 2H,
PhOCH2CO), 4.53 (m, 1H, H3′), 4.42 (m, 2H, ArOCH2), 4.30
(t, 2H, COCH2),4.01 (m, 1H, H4′), 3.84 (s, 6H, OCH3), 3.58
(dd, 1H, H5′), 3.31 (dd, 1H, H5′), 2.55 (m, 1H, H2′), 2.26 (m,
1H, H2′), 1.84 (m, 4H, CH2), 1.57 (s, 3H, ArCH3), 0.87 (s, 9H,
SiC(CH3)3), 0.08 (s, 6H, Si(CH3)2). 13C NMR (125.7 MHz,
CDCl3, ppm): 170.3, 169.0, 158.7, 157.8, 155.6, 144.4, 139.7,
135.52, 135.50, 130.11, 130.09, 129.6, 128.2, 127.9, 127.1,
121.8, 114.6, 113.24, 113.21, 104.4, 86.7, 86.5, 86.3, 71.1, 66.5,
65.3, 64.8, 62.4, 55.2, 42.2, 25.7, 25.3, 25.1, 17.9, 11.7, −4.62,
−4.94. IR (thin film); νmax (cm−1) = 3061, 2954, 1760, 1672,
1532, 1509, 1497, 1252, 1178, 835, 783. HRMS (ESI-MS) m/z
1-{O4-[3′-O-(β-Cyanoethyl-N,N′-diisopropyl)-5′-O-(4,4′-
dimethoxytrityl)thymidyl]}-7-{O4-[3′-O-(β-cyanoethyl-N,N′-
diisopropyl)-5′-O-(4,4′-dimethoxytrityl)-thymidyl]}-heptane (4b)
3b (0.170 g, 0.143 mmol) and diisopropylethylamine (0.056 g,
0.430 mmol) were dissolved in THF (2 mL), followed by N,N-
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 7078–7090 | 7085

View Article Online
calculated for C49H60N2O10SiNa+ 887.3914: found 887.3921
[M + Na]+.
128.0, 127.0, 121.8, 114.6, 113.3, 104.9, 86.8, 86.5, 86.4, 72.0,
66.7, 65.3, 64.8, 63.5, 55.2, 42.2, 25.3, 25.1, 11.7. IR (thin
film); νmax (cm−1) = 3342, 3061, 2956, 1758, 1663, 1533, 1509,
1251, 1177, 1090, 755. HRMS (ESI-MS) m/z calculated for
C43H46N2O10Na+ 773.3049: found 773.3055 [M + Na]+.
3′-O-(tert-Butyldimethylsilyl)-5′-O-(4,4′-dimethoxytrityl)-
O4-(phenoxyacetyloxyheptyl)-thymidine (5b)
1b (0.527 g, 0.682 mmol) was dissolved in THF (0.5 mL). Tri-
ethylamine (0.110 g, 1.09 mmol) was added followed by cooling
of the reaction in an ice bath for 10 min. Phenoxyacetyl chloride
(0.175 g, 1.02 mmol) was added dropwise. After 1.5 h, the
solvent was evaporated in vacuo, the crude product was taken up
in dichloromethane (50 mL) and the solution was washed with
two portions of sodium bicarbonate (2 × 50 mL). The organic
layer was dried over sodium sulfate and concentrated to produce
a yellow gum. The crude product was purified by flash column
chromatography using a hexanes–ethyl acetate (4 : 1 → 2 : 3)
solvent system to afford 0.491 g (79.4%) of product as a color-
less foam. Rf (SiO2 TLC): 0.49 hexanes–ethyl acetate (1 : 1).
λmax(ACN) = 276 nm. 1H NMR (500 MHz, CDCl3, ppm): 8.00 (s,
1H, H6), 7.45 (dd, 2H, Ar), 7.25–7.34 (m, 9H, Ar), 6.85–7.04
(m, 7H, Ar), 6.37 (dd, 1H, H1′, J = 6.0 Hz), 4.68 (s, 2H,
PhOCH2CO), 4.53 (dd, 1H, H3′), 4.41 (t, 2H, ArOCH2), 4.25 (t,
2H, CH2OCO), 4.02 (m, 1H, H4′), 3.84 (s, 6H, OCH3), 3.58
(dd, 1H, H5′), 3.30 (dd, 1H, H5′), 2.55 (m, 1H, H2′), 2.26 (m,
1H, H2′), 1.77 (quintet, 2H, CH2), 1.71 (quintet, 2H, CH2), 1.58
(s, 3H, CH3Ar), 1.37–1.46 (m, 6H, (CH2)3), 0.88 (s, 9H,
SiC(CH3)3), 0.07 (s, 6H, Si(CH3)2. 13C NMR (125.7 MHz,
CDCl3, ppm): 170.5, 169.1, 158.7, 157.8, 156.0, 144.4, 139.4,
135.5, 130.11, 130.09, 129.5, 128.2, 127.9, 127.1, 121.7, 114.6,
113.23, 113.21, 104.5, 86.7, 86.5, 86.3, 71.1, 67.2, 65.4, 65.3,
62.4, 55.2, 42.2, 28.8, 28.5, 28.4, 25.8, 25.71, 25.69, 17.9, 11.8,
−4.62, −4.95. IR (thin film); νmax (cm−1) = 3062, 2928, 1760,
1675, 1510, 1464, 154, 1065, 832, 734. HRMS (ESI-MS) m/z
calculated for C52H66N2O10SiNa+ 929.4382: found 929.4384
[M + Na]+.
5′-O-(4,4′-Dimethoxytrityl)-O4-(phenoxyacetyloxyheptyl)-
thymidine (6b)
5b (0.491 g, 0.541 mmol) was dissolved in THF (4.5 mL).
TBAF (1 M in THF) (0.65 mL, 0.65 mmol) was added drop-
wise. After 25 min, the solvent was evaporated in vacuo, the
crude product was taken up in dichloromethane (50 mL) and the
solution was washed with two portions of sodium bicarbonate (2
× 50 mL). The organic layer was dried over sodium sulfate and
concentrated to produce a yellow gum. The crude product was
purified by flash column chromatography using a methanol–
hexanes–ethyl acetate (0 : 1 : 1 → 1 : 19 : 30) solvent system to
afford 0.373 g (86.8%) of product as a colorless foam. λmax(ACN)
1
= 276 nm. H NMR (500 MHz, CDCl3, ppm): 8.00 (s, 1H, H6),
7.29–7.40 (m, 2H, Ar), 7.18–7.27 (m, 9H, Ar), 6.94–6.97 (m,
1H, Ar), 6.87–6.88 (m, 2H, Ar), 6.79–6.81 (m, 4H, Ar), 6.37
(dd, 1H, H1′, J = 6.25 Hz), 4.61 (s, 2H, PhOCH2CO), 4.52 (m,
1H, H3′), 4.34 (t, 2H, ArOCH2), 4.18 (t, 2H, CH2OCO), 4.07
(m, 1H, H4′), 3.78 (s, 6H, OCH3), 3.47 (dd, 1H, H5′), 3.36 (dd,
1H, H5′), 2.60 (m, 1H, H2′), 2.33 (s, 1H, OH), 2.25 (m, 1H,
H2′), 1.72 (quintet, 2H, CH2), 1.64 (quintet, 2H, CH2), 1.56 (s,
3H, CH3Ar), 1.33–1.37 (m, 6H, (CH2)3). 13C NMR
(125.7 MHz, CDCl3, ppm): 170.6, 169.1, 158.6, 157.8, 156.3,
144.5, 139.7, 135.6, 135.5, 130.09, 130.07, 129.5, 128.1, 127.9,
127.0, 121.7, 114.6, 113.2, 105.0, 86.8, 86.5, 86.3, 72.0, 67.3,
65.4, 65.3, 63.5, 55.2, 42.2, 28.8, 28.5, 28.4, 25.8, 25.7, 11.7.
IR (thin film); νmax (cm−1) = 3346, 3060, 2934, 1757, 1663,
1532, 1509, 1251, 1177, 1091, 735. HRMS (ESI-MS) m/z calcu-
lated for C46H52N2O10Na+ 815.3519: found 815.3517
[M + Na]+.
5′-O-(4,4′-Dimethoxytrityl)-O4-(phenoxyacetyloxybutyl)-
thymidine (6a)
3′-O-(β-Cyanoethyl-N,N′-diisopropyl)-5′-O-(4,4′-dimethoxytrityl)-
O4-(phenoxyacetyloxybutyl)-thymidine (7a)
5a (0.200 g, 0.231 mmol) was dissolved in THF (2.4 mL).
TBAF (1 M in THF) (0.277 mL, 0.277 mmol) was added drop-
wise. After 15 min, the crude product was taken up in ethyl
acetate (50 mL) and the solution was washed with two portions
of sodium bicarbonate (2 × 50 mL). The organic layer was dried
over sodium sulfate and concentrated to produce a yellow gum.
The crude product was purified by flash column chromatography
using a methanol–hexanes–ethyl acetate (2 : 11 : 7) solvent
system to afford 0.164 g (94.1%) of product as a colorless foam.
Rf (SiO2 TLC): 0.29 methanol–hexane–ethyl acetate (2 : 11 : 7).
λmax(ACN) = 276 nm. 1H NMR (500 MHz, CDCl3, ppm): 7.79 (s,
1H, H6), 7.18–7.37 (m, 13H, Ar), 6.79–6.96 (m, 5H, Ar), 6.34
(dd, 1H, H1′, J = 6.5 Hz), 4.61 (s, 2H, PhOCH2CO), 4.50 (m,
1H, H3′), 4.36 (m, 2H, ArOCH2), 4.23 (t, 2H, COCH2),4.05 (m,
1H, H4′), 3.77 (s, 6H, OCH3), 3.47 (dd, 1H, H5′), 3.36 (dd, 1H,
H5′), 2.58 (m, 1H, H2′), 2.23 (m, 1H, H2′), 1.75–1.78 (m, 4H,
CH2), 2.04 (s, 1H, OH), 1.55 (s, 3H, ArCH3). 13C NMR
(125.7 MHz, CDCl3, ppm): 170.4, 169.0, 158.7, 158.6, 157.8,
156.1, 144.5, 139.9, 135.6, 135.5, 130.08, 130.07, 129.6, 128.1,
6a (0.200 g, 0.266 mmol) and diisopropylethylamine (0.051 g,
0.398 mmol) were dissolved in THF (2.5 mL), followed by N,N-
diisopropylaminocyanoethylphosphonamidic chloride (0.076 g,
0.319 mmol). The reaction was allowed to stir at room tempera-
ture for 30 min, where TLC revealed the presence of starting
material. Diisopropylethylamine (0.006 g, 0.049 mmol) followed
by N,N-diisopropylaminocyanoethylphosphonamidic chloride
(0.011 g, 0.045 mmol) were added and the reaction was allowed
to proceed for another 30 min. Upon completion, the reaction
was quenched by the addition of ethyl acetate (50 mL), and the
solution was washed with sodium bicarbonate (3%, 2 × 25 mL)
and brine (25 mL). The organic layer was dried over sodium
sulfate, decanted, and evaporated. The product, a colorless
powder, was precipitated from hexanes (0.116 g, 46.2%). Rf
(SiO2 TLC): 0.49, 0.67 hexanes–ethyl acetate (2 : 8). λmax(ACN)
=
1
276 nm. H NMR (500 MHz, d6-acetone, ppm): 7.90–7.94 (s,
1H, H6), 7.46–7.51 (m, 2H, Ar), 7.24–7.39 (m, 9H, Ar),
6.89–6.96 (m, 7H, Ar), 6.31–6.36 (m, 1H, H1′), 4.73–4.75
7086 | Org. Biomol. Chem., 2012, 10, 7078–7090
This journal is © The Royal Society of Chemistry 2012

View Article Online
Solid-phase oligonucleotide synthesis and purification
(m, 3H, H3′, PhOCH2CO), 4.33 (m, 2H, ArOCH2), 4.18–4.24
(m, 3H, H4′ & CH2OCO), 3.60–3.89 (m, 10H, CH2OP, NCH,
OCH3), 3.40–3.50 (m, 2H, H5′), 2.76 (t, 1H, CH2CN),
2.54–2.63 (m, 3H, H2′, CH2CN), 2.35–2.42 (m, 1H, H2′),
1.79–1.81 (m, 4H, (CH2)2), 1.57–1.61 (m, 3H, ArCH3),
1.17–1.23 (m, 9H, CH3), 1.09–1.10 (m, 3H, CH3). 31P NMR
(202.3 MHz, d6-acetone, ppm): 148.28, 148.08. 13C NMR
(125.7 MHz, d6-acetone, ppm): 170.00, 169.99, 168.6, 158.9,
158.2, 154.7, 144.9, 140.1, 140.0, 135.7, 135.64, 135.62,
135.58, 130.18, 130.15, 129.4, 128.2, 128.1, 127.9, 127.0,
126.9, 121.1, 117.9, 114.5, 113.1, 103.5, 103.4, 86.64, 86.60,
86.0, 85.9, 85.5, 85.3, 85.2, 73.1, 66.1, 64.3, 63.2, 63.0, 58.7,
58.5, 54.70, 54.68, 43.1, 43.0, 40.4, 40.3, 25.1, 25.0, 24.05,
24.03, 23.99, 23.9, 19.9, 19.84, 19.79, 13.6, 11.17, 11.16. IR
(thin film); νmax (cm−1) = 3060, 2966, 1758, 1671, 1607, 1508,
1330, 1251, 1180, 1034, 755. HRMS (ESI-MS) m/z calculated
for C52H63N4O11PNa+ 973.4128: found973.4131, [M + Na]+.
DNA duplexes (XLTT4 and XLTT7, whose sequences are
shown in Fig. 1) and single stranded DNA (control strands, O4-
butyl-4-ol-dT (S4) and O4-heptyl-7-ol-dT (S7)), were assembled
on an Applied Biosystems Model 3400 synthesizer on a 1 μmol
scale employing standard β-cyanoethylphosphoramidite cycles
as indicated by the manufacturer with modifications to certain
coupling times as indicated below.
Commercially available 3′-O-2′deoxynucleoside phosphorami-
dites, which were protected with fast deprotecting groups, were
dissolved in anhydrous acetonitrile at a concentration of 0.1 M,
the modified 3′-O-2′-deoxyphosphoramidites (7a and 7b) at
0.15 M and the modified bis-3′-O-2′-deoxyphosphoramidites (4a
and 4b) at 0.05 M. Assembly of sequences began with detrityla-
tion (3% trichloroacetic acid (TCA) in CH2Cl2), followed by
phosphoramidite coupling: commercial 3′-O-2′-deoxyphosphora-
midites (2 min), modified phosphoramidites (7a and 7b)
(10 min) and modified bis-phosphoramidite (4a and 4b)
(30 min); capping was achieved with phenoxyacetic anhydride–
pyridine–tetrahydrofuran (1 : 1 : 8, v/v/v; solution A, and
1-methyl-1H-imidazole–tetrahydrofuran 16 : 84 w/v; solution B)
and oxidation (0.02 M iodine in tetrahydrofuran–water–pyridine
2.5 : 2 : 1) which followed every coupling. Removal of the
3′-terminal trityl group was carried out by the synthesizer to
yield the final oligonucleotide on the solid support.
The oligomer-derivatized CPG support was removed from the
column and placed into screw cap microfuge tubes fitted
with Teflon lined caps. The oligonucleotides containing O4dT
modifications were removed from the support by incubation of
the CPG in 500 μL of a 10% 1,8-diazabicyclo[5.4.0]undec-
7-ene (DBU) solution in the respective alcohol (1,4-butanediol
for O4-butyl-4-ol-dT and 1,7-heptanediol for O4-heptyl-7-ol-dT),
or ethanol for cross-linked oligonucleotides, for 5 days in
the dark at room temperature. The DBU was neutralized
with acetic acid and the DNA solubilised in acetonitrile prior
to being transferred into clean vials to eliminate the CPG.
The DNA was dried down in a speed vacuum and then desalted
using C-18 SEP PAK cartridges (Waters Inc.) prior to
purification.
The cross-link and mono-adduct final products were purified
from pre-terminated products by 20% 7 M urea denaturing poly-
acrylamide gels (19 : 1) using 1X TBE [89 mM Tris–HCl,
89 mM boric acid, 2 mM EDTA (pH 8.0)] as running buffer to
eliminate any residual DBU, which could not be removed by
SAX-HPLC. Prior to gel separation, 25 O.D. of each sample
were dried via a speed-vac concentrator and resuspended in
100 μL of formamide. The bands corresponding to the desired
products were excised from the gels and placed into 15 mL
Falcon tubes and the gel slices submerged in 0.1 M sodium
acetate. The tubes were allowed to shake overnight and the
extracted oligomers desalted using C-18 SEP PAK cartridges.
3′-O-(β-Cyanoethyl-N,N′-diisopropyl)-5′-O-(4,4′-
dimethoxytrityl)-O4-(phenoxyacetyloxyheptyl)-thymidine (7b)
6b (0.205 g, 0.242 mmol) and diisopropylethylamine (0.050 g,
0.390 mmol) were dissolved in THF (1 mL), followed by N,N-
diisopropylaminocyanoethylphosphonamidic chloride (0.073 g,
0.310 mmol). The reaction was allowed to stir at room tempera-
ture for 30 min, where TLC revealed the presence of starting
material. Diisopropylethylamine (0.006 g, 0.049 mmol) followed
by N,N-diisopropylaminocyanoethylphosphonamidic chloride
(0.011 g, 0.045 mmol) were added and the reaction was allowed
to react another 30 min. Upon completion, the reaction was
quenched by the addition of ethyl acetate (50 mL), and the solu-
tion was washed with sodium bicarbonate (3%, 2 × 25 mL) and
brine (25 mL). The organic layer was dried over sodium sulfate,
decanted, and evaporated. The product, a colorless powder, was
precipitated from hexanes (0.079 g, 32.2%). Rf (SiO2 TLC):
1
0.70, 0.83 hexanes–ethyl acetate (2 : 8). λmax(ACN) = 276 nm. H
NMR (500 MHz, d6-acetone, ppm): 7.90–7.93 (s, 1H, H6),
7.46–7.51 (m, 2H, Ar), 7.24–7.39 (m, 9H, Ar), 6.89–6.97 (m,
7H, Ar), 6.31–6.36 (m, 1H, H1′), 4.72–4.75 (m, 3H, H3′,
PhOCH2CO), 4.29 (m, 2H, ArOCH2), 4.16–4.25 (m, 3H, H4′ &
CH2OCO), 3.60–3.90 (m, 10H, CH2OP, NCH, OCH3),
3.40–3.52 (m, 2H, H5′), 2.76 (t, 1H, CH2CN), 2.53–2.63
(m, 3H, H2′, CH2CN), 2.34–2.42 (m, 1H, H2′), 1.71–1.77
(m, 2H, CH2), 1.57–1.68 (m, 5H, CH2 & ArCH3), 1.37–1.44
(m, 6H, (CH2)3), 1.16–1.23 (m, 9H, CH3), 1.09–1.10 (m, 3H,
CH3). 31P NMR (202.3 MHz, d6-acetone, ppm): 148.28,
148.08. 13C NMR (125.7 MHz, d6-acetone, ppm): 170.09,
170.08, 168.6, 158.9, 158.2, 154.7, 145.0, 140.0, 139.9, 135.7,
135.64, 135.62, 135.58, 130.18, 130.16, 129.4, 128.20, 128.15,
127.87, 127.86, 126.91, 126.88, 121.2, 118.1, 117.9, 114.5,
113.2, 103.54, 103.46, 86.64, 86.61, 86.0, 85.9, 85.51, 85.47,
85.3, 73.6, 73.5, 73.1, 66.6, 64.8, 64.6, 63.2, 63.0, 59.7, 58.7,
58.5, 54.71, 54.69, 43.1, 43.0, 40.4, 40.3, 25.7, 25.5, 24.1,
24.04, 24.01, 23.9, 19.91, 19.85, 19.8, 13.7, 11.20, 11.18,
11.14. IR (thin film); νmax (cm−1) = 3060, 2965, 1758, 1671,
1607, 1509, 1252, 1180, 1086, 1035, 755. HRMS (ESI-MS) m/z
calculated for C55H69N4O11PNa+ 1015.4597: found 1015.4620
[M + Na]+.
Oligonucleotide characterization by ESI-MS and nuclease
digestion
ESI mass spectra for oligonucleotides were obtained at the Con-
cordia University Centre for Biological Applications of Mass
Spectrometry using a Micromass Qtof2 mass spectrometer
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 7078–7090 | 7087

View Article Online
Circular dichroism (CD) spectroscopy of DNA duplexes
(Waters) equipped with a nanospray ion source. The mass
spectrometer was operated in full scan, negative ion detection
mode.
Circular dichroism experiments were carried out on a Jasco
J-815 spectropolarimeter equipped with a Julaba F25 circulating
bath as previously reported.14 Briefly, the samples used for the
UV thermal denaturation studies were placed at 4 °C overnight
and subsequently used for circular dichroism. The spectra were
an average of 5 scans at a rate of 20 nm min−1, with a bandwidth
of 1 nm, sampling wavelength of 0.2 nm in a fused quartz cells
(Starna 29-Q-10). Scans were performed between 320 and
220 nm at 10 °C. The molar ellipticity [θ] was obtained from the
relation θ = ε/Cl, where ε is the relative ellipticity (mdeg), C is
the molar concentration of the DNA duplex (moles L−1), and l is
the path length in cm.
The modified oligomers (0.1 A260 units) were analyzed by
exonuclease digestion (snake venom phosphodiesterase:
0.28 units and calf intestinal phosphatase: 5 units, in 10 mM
Tris, pH 8.1 and 2 mM magnesium chloride) for a minimum of
16 h at 37 °C. The resulting mixtures were analyzed by reversed
phase HPLC performed on a Symmetry® C-18 5 μm column
(0.46 × 15 cm) purchased from Waters Inc, Milford, MA. The
column was eluted with a linear gradient of 0–60% buffer
B over 30 min (buffer A, 50 mM sodium phosphate, pH 5.8, 2%
acetonitrile and buffer B, 50 mM sodium phosphate, pH 5.8,
50% acetonitrile). The retention times of the eluted peaks (see
ESI‡) were compared to the standard nucleotides which eluted at
the following times: dC (4.4 min), dG (7.5 min), dT (8.2 min),
dA (9.2 min), O6-MeG (12.0 min), O4-MeT (13.4 min),
O4-butyl-4-ol-dT (15.3 min), O4-heptyl-7-ol-dT (25.6 min) and
the cross-link dimers (21.7 for the four carbon and 29.9 min for
the seven carbon cross-link), and the ratio of nucleosides was
determined. The molecular weights of the modified oligomers
were determined by ESI-MS, which correlated with the calcu-
lated mass.
AGT preparation and purification
All proteins were purified under native conditions under the pro-
moter of the pQE30 vector as described elsewhere.20 The pro-
cedures for site directed mutagenesis and transformations into
XL-10 Gold E. coli cells were conducted as stated in the Strata-
gene manual. Cells containing a plasmid coding for wild-type or
mutant Ada-C, hAGT and OGT were grown in 1 L of LB broth
+ 100 μg mL−1 ampicillin until an OD600 = 0.6 was reached.
Induction of the protein was achieved by adding IPTG at a final
concentration of 0.3 mM. The cells were incubated at 37 °C for
4 h with shaking at 225 rpm, harvested by centrifugation at 6000
× g at 4 °C for 20 min. 5 mL of resuspension buffer [20 mM
Tris–HCl (pH 8.0), 250 mM NaCl, 20 mM β-mercaptoethanol
supplemented with Complete, Mini, EDTA-free Protease Inhibi-
tor Cocktail Tablets] was added for every gram of wet pellet, and
the cells resuspended using a dounce homogenizer. Cell lyses
was carried out using two passes on the French press followed
by centrifuged at 17 000 × g for 45 min at 4 °C. The cell lysate
was introduced on a Ni-NTA column pre-equilibrated with equi-
libration buffer (resuspension buffer lacking the Complete, Mini,
EDTA-free Protease Inhibitor). After extensive washing of the
column with the equilibration buffer supplemented with 20 mM
imidazole the protein was eluted with equilibration buffer sup-
plemented with 200 mM imidazole. Fractions having protein
content were pooled and dialyzed against dialysis buffer [50 mM
Tris–HCl (pH 7.6), 250 mM NaCl, 20 mM β-mercaptoethanol
and 0.1 mM EDTA] using 8000 Da cutoff dialysis tubing. In
general a yield of 8–10 mg of protein was obtained per L of
culture.
UV thermal denaturation of DNA duplexes
The nearest-neighbour method was used to determine the molar
extinction coefficients for all oligonucleotides (with units of
(M−1 cm−1). An equimolar amount of the complementary
strands were mixed together and lyophilized to dryness for both
the mono-adduct and the non-cross-linked duplex. The DNAwas
subsequently resuspended in 90 mM sodium chloride, 10 mM
sodium phosphate, 1 mM EDTA buffer (pH 7.0) to give a final
duplex concentration of 4 μM (OD260 = 0.5). To ensure
proper duplex formation the solution was heated to 90 °C where
it was held for 10 min, cooled slowly to room temperature and
stored at 4 °C overnight prior to use. Degassing of the samples
was achieved using a speed-vac concentrator for 2 min. De-
naturation curves were obtained by monitoring absorbance at
260 nm at a rate of heating of 0.5 °C min−1, using a Varian
CARY Model 3E spectrophotometer fitted with a 6-sample ther-
mostated cell block and a temperature controller. Processing of
the data was performed using the method published by Puglisi
and Tinoco.36
Thermodynamic parameters of the non-cross-linked control
DNA were obtained by performing thermal denaturation studies
in triplicate at concentrations of 0.6, 2.7, 12.6 and 58.5 μM
duplex DNA in 1, 0.5, 0.2 and 0.1 cm cuvettes, respectively. A
plot of ln(Ctot) as a function of 1/Tm, typically used for a non-
self-complementary bimolecular systems, was employed to
obtain ΔH° and ΔS° for the system as described by Puglisi and
Tinoco.36 A similar procedure was employed with the cross-
linked DNA at concentrations of 0.6 and 2.7 μM. The Tm of the
ICL were independent of oligonucleotide concentration due to
the uni-molecular nature of the DNA and therefore the thermo-
dynamic parameters were obtained by plotting ln(K) versus 1/T
at a single concentration of 2.95 μM. The experiments were per-
formed in triplicate (see ESI‡).
Mono-adduct repair assay
A general oligonucleotide sequence was created with a BclI cut
site to monitor the amount of repair of mono-adduct based on
restriction digestion. The sequence of the damage containing
strand was 5′ GGC TXG ATC ACC AG where: X represents dT
(control sequence), O4MeT, O4-butyl-4-ol-dT (S4) or O4-heptyl-
7-ol-dT (S7). The sequence of the complement strand was 5′
CTG GT/G ATC AAG CC where the “/” represent the BclI cut
site.
The damage containing strand was 32P labelled at its 5′ end as
previously described.14 Briefly, a 20 μM solution of DNA
was made in 1X PNK buffer along with 1 μL [γ-32P]ATP
7088 | Org. Biomol. Chem., 2012, 10, 7078–7090
This journal is © The Royal Society of Chemistry 2012

View Article Online
(10 μCi μL−1) and 5 units of T4 PNK. The labelling reaction
was conducted for 1 h at 37 °C after which the reaction was ter-
minated by boiling the sample for 10 min.
C145S hAGT variant binding studies with cross-linked
substrates
Binding reactions were conducted as for the mono-adduct sub-
strates with minor modification due to the different substrate
size. Binding reactions consisted of 0.5 nM dsDNA, increasing
C145S AGT ranging from 1 to 45.5 μM for the control duplex
and from 1 to 6.5 μM for the cross-linked DNA in a total solu-
tion volume of 20 μl of binding buffer. The samples were equili-
brated for 1 h at room temperature and loaded on a 15% native
polyacrylamide gel (75 : 1) and ran with 0.5X TBE, which
allowed the electrophoresis voltage to be increased without
heating the system. Electrophoresis was carried out at 21 °C at
250 V for 25 min.
The monomeric dissociation constant (Kd) and stoichiometry
(n) of AGT binding to DNA were obtained from the electrophor-
etic mobility shift assays as described previously.14 The binding
of n molecules of AGT protein [P] to 1 molecule of DNA [D]
can be expressed by the equation: nP + D ↔ PnD. Taking the
logarithm and rearranging the variables yields: log ([PnD]/[D]) =
n log[P]free + log Kd.
40 pmol of labelled DNA was added to 50 pmol of the comp-
lement strand in a total volume of 20 μL of water making a
2 μM dsDNA solution with 20% excess of the non-damaged
strand. The solution was boiled for 5 min, cooled slowly to room
temperature and placed at 4 °C overnight to ensure proper duplex
formation.
The repair reaction mixtures were constituted of 2 pmol of the
duplex DNA and 10 pmol of AGT in a total volume of 15 μL in
Activity Buffer [10 mM Tris–HCl (pH 7.6), 100 mM NaCl and
1 mM DTT] and allowed to react at 37 °C for 30 min, unless
otherwise indicated. The reaction was terminated by boiling for
10 min. Prior to BclI digestion, MgCl2 was added at a final con-
centration of 10 mM and the reaction mixture allowed to cool to
room temperature. 7.5 units (0.5 μL) of BclI was added to the
mixture and the solution incubated for 1 h at 37 °C. 24 μL of
stop buffer [81 mM Tris–HCl, 81 mM boric acid, 1.8 mM EDTA
and 1% SDS (sodium dodecyl sulfate) (pH 8.0) in 80% forma-
mide] was added, boiled for 10 min and loaded on a 14 cm ×
16 cm, 20% 7 M urea denaturing polyacrylamide gel (19 : 1) for
separation. The gels were run using 1X TBE for 40 min at
400 V and the gels exposed to a storage phosphor screen. The
image was captured on a Typhoon 9400 (GE Healthcare, Piscat-
away, NJ) and the autoradiography counts obtained by Image-
Quant™ (Amersham Biosciences).
The data can be expressed with a plot of log [PnD]/[P] as a
function of log [P]. The slope of the graph represents the stoichi-
ometry (n) and the observed Kd can be obtained by the relation-
ship: Kd = 10−(y−intercept). The monomeric Kd shown in the tables
was obtained by taking the nth root of the observed Kd.
Modelling of O4MeT and O4-butyl-4-ol-dT in hAGT active site
The alkylated DNA was placed in the active site of hAGT based
on PDB entry 1t38 by mutating O6MeG to O4MeT. The place-
ment of O4MeT was verified with PDB entry 1yfh. PDB entry
1t38 was used as the template for hAGT where Ser145 was
mutated back to Cys and the proper rotamer used as based on
PDB entries: 1t39, 1eH6 and 1yfh. This primary model was
used as the template for the addition of the butyl-4-ol adduct,
which was added manually by PyMOL.
The preliminary models underwent conjugate gradient mini-
mization, simulated annealing, and torsion angle dynamics using
Crystallography & NMR System.37 CHARMM (Chemistry at
HARvard Macromolecular Mechanics) topology files were
modified to use the altered nucleotides by adding restraints to
keep the α-carbon of the lesion in a syn conformation with
respect to the N3 atom of the base, as observed in PDB entries:
1t38, 1t39, and 1yfh.38
The models were subjected to 2 sets of 100 steps of gradient
minimization and 200 steps of simulated annealing. The first
cycle was carried out at 500 K with slow cooling at a rate of
4.5 K per cycle of dynamics. The second cycle was carried out
at 350 K with slow cooling at a rate of 2.5 K per cycle of
dynamics. The minimized models were displayed using PyMol
with the absence of a few amino acids for clarity of the active
site.
C145S hAGT variant binding studies with mono-adduct
substrates
Binding reactions of the 14 bp mono-adduct DNA substrates
consisted of 0.5 nM dsDNA (where an excess of 5% of un-
damaged strand was present to force duplex formation) and
increasing C145S hAGT (ranging from 1 to 35.69 μM) in a total
solution volume of 20 μl of binding buffer [10 mM Tris–HCl
(pH 7.6), 100 mM NaCl, 1 mM DTT, 10 μg mL−1 BSA and
2.5% glycerol]. The samples were equilibrated for 30 min at
room temperature and loaded on a 10% native polyacrylamide
gel (75 : 1) and pre-ran with 10 mM Tris-Acetate (pH 7.6) and
100 mM NaCl. Electrophoresis was carried out at 21 °C at
100 V for 45 min.
Cross-link repair assay
Labelling the 5′ nucleosides of the cross-linked oligonucleotides
XLTT4 and XLTT7 as well as the control DNA (non-cross-
linked duplex) were performed as described above. Repair assays
were performed with 60 pmol of AGT protein incubated with
2 pmol of labelled DNA in a total reaction volume of 15 μL
made up of Activity Buffer. The samples were incubated at
37 °C for 16 h. The reaction was ended by adding 18.2 μL of
stop reaction buffer and the mixture separated on a 17% 7 M
urea denaturing polyacrylamide gel (19 : 1). The gels were run
using 1X TBE for 1.5 h at 700 V and the gels exposed to a
storage phosphor screen. The image was captured on a Typhoon
9400 (GE Healthcare, Piscataway, NJ) and the radioactive counts
quantified by ImageQuant™ (Amersham Biosciences).
Conclusions
DNA duplexes containing an opposed O4-2′-deoxythymidine-
alkyl-O4-2′-deoxythymidine ICL were synthesized using a con-
vertible nucleoside strategy to prepare a cross-linked nucleoside
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 7078–7090 | 7089

View Article Online
9 C. J. Wilds, J. D. Booth and A. M. Noronha, Tetrahedron Lett., 2006, 47,
9125–9128.
10 A. E. Pegg, Mutat. Res., 2000, 462, 83–100.
11 D. S. Daniels, T. T. Woo, K. X. Luu, D. M. Noll, N. D. Clarke and
A. E. Pegg, et al., Nat. Struct. Mol. Biol., 2004, 11, 714–720.
12 K. S. Srivenugopal, X. H. Yuan, H. S. Friedman and F. Ali-Osman, Bio-
chemistry, 1996, 35, 1328–1334.
13 Q. Fang, A. M. Noronha, S. P. Murphy, C. J. Wilds, J. L. Tubbs,
J. A. Tainer, G. Chowdhury, F. P. Guengerich and A. E. Pegg, Biochemis-
try, 2008, 47, 10892–10903.
dimer that was converted into a bis-phosphoramidite. Solid
phase synthesis was employed to incorporate the cross-link into
a specific orientation and position in a DNA duplex. This
method is versatile and allows for various linker lengths to be
introduced for a variety of studies including DNA repair. ICL
duplexes containing a four and seven carbon linker evaded repair
from various AGT proteins from human and E. coli.
Studies to investigate AGT proteins that can be engineered to
repair these ICL and their repair by other systems are ongoing to
enhance our understanding of how ICL lesions may be pro-
cessed. Furthermore this may result in increasing the potency of
bifunctional chemotherapeutic agents which create ICL lesions
and hopefully limit their mutagenicity and other side effects
such as the onset of resistance to these drugs.
14 F. P. McManus, Q. Fang, J. D. Booth, A. M. Noronha, A. E. Pegg and
C. J. Wilds, Org. Biomol. Chem., 2010, 8, 4414–4426.
15 (a) A. S. Waldman, BioEssays, 2008, 30, 1163–1171;
(b) M. F. Goodman, S. Creighton, L. B. Bloom and J. Petruska, Crit. Rev.
Biochem. Mol. Biol., 1993, 28, 83–126; (c) T. A. Kunkel and
K. Bebenek, Annu. Rev. Biochem., 2000, 69, 497–529; (d) T. Lindahl,
Nature, 1993, 362, 709–715.
16 A. Bhattacharyya and D. M. Lilley, J. Mol. Biol., 1989, 209, 583–597.
17 B. D. Preston, B. Singer and L. A. Loeb, Proc. Natl. Acad. Sci. U. S. A.,
1986, 83, 8501–8505.
18 P. Rojsitthisak, N. Jongaroonngamsang, R. M. Romero and I. S. Haworth,
PLoS One, 2011, 6, e20745.
Acknowledgements
19 Y. Z. Xu and P. F. Swann, Nucleic Acids Res., 1990, 18, 4061–4065.
20 M. E. Dolan and A. E. Pegg, Carcinogenesis, 1985, 6, 1611–1614.
21 S. S. Hecht, Chem. Res. Toxicol., 1998, 11, 559–603.
22 (a) G. T. Pauly and R. C. Moschel, Chem. Res. Toxicol., 2001, 14, 894–
900; (b) K. B. Altshuler, C. S. Hodesand and J. M. Essigmann, Chem.
Res. Toxicol., 1996, 9, 980–987.
23 J. C. Klein, M. J. Bleeker, H. C. Roelen, J. A. Rafferty, G. P. Margison
and H. F. Brugghe, et al., J. Biol. Chem., 1994, 269, 25521–25528.
24 (a) P. M. Potter, M. C. Wilkinson, J. Fitton, F. J. Carr, J. Brennand and
D. P. Cooper, et al., Nucleic Acids Res., 1987, 15, 9177–9193;
(b) S. R. Paalman, C. Sung and N. D. Clarke, Biochemistry, 1997, 36,
11118–11124; (c) M. C. Wilkinson, P. M. Potter, L. Cawkwell,
P. Georgiadis, D. Patel and P. F. Swann, et al., Nucleic Acids Res., 1989,
17, 8475–8484; (d) Q. Fang, S. Kanugula, J. L. Tubbs, J. A. Tainer and
A. E. Pegg, J. Biol. Chem., 2010, 285, 8185–8195.
25 Q. Zhu, M. O. Delaney and M. M. Greenberg, Bioorg. Med. Chem. Lett.,
2001, 11, 1105–1107.
26 (a) C. J. Wilds, F. Xu and A. M. Noronha, Chem. Res. Toxicol., 2008, 21,
686–695; (b) C. J. Wilds, E. Palus and A. M. Noronha, Can. J. Chem.,
2007, 85, 249–256.
27 J. D. Booth, S. P. Murphy, A. M. Noronha and C. J. Wilds, Nucleic Acids
Symp. Ser., 2008, 52, 431–432.
28 C. J. Wilds, A. M. Noronha, S. Robidoux and P. S. Miller, J. Am. Chem.
Soc., 2004, 126, 9257–9265.
We are grateful to Dr Anthony E. Pegg (Penn. State University)
for the plasmids encoding the wild-type Ada-C, OGT and hAGT
genes. This research was supported by grants from the Natural
Sciences and Engineering Research Council of Canada
(NSERC), the Canada Foundation for Innovation (CFI) and the
Canada Research Chair (CRC) program. Both FPM and DKO
were recipients of postgraduate fellowships from NSERC.
Notes and references
1 (a) N. Magana-Schwencke, J. A. Henriques and R. E. Moustachhi, Proc.
Natl. Acad. Sci. U. S. A., 1982, 79, 1722–1726; (b) P. D. Lawley and
D. H. Phillips, Mutat. Res., Fundam. Mol. Mech. Mutagen., 1996, 355,
13–40; (c) S. R. Rajski and R. M. Williams, Chem. Rev., 1998, 98,
2723–2796.
2 (a) M. L. Dronkert and R. Kanaar, Mutat. Res., 2001, 486, 217–247;
(b) P. J. McHugh, V. J. Spanswick and J. A. Hartley, Lancet Oncol.,
2001, 2, 483–490; (c) M. Christmann, M. T. Tomicic, W. P. Roos and
B. Kaina, Toxicology, 2003, 193, 3–34; (d) A. Sancar, L. A. Lindsey-
Boltz, K. Unsal-Kacmaz and S. Linn, Annu. Rev. Biochem., 2004, 73,
39–85.
29 C. Hofr and V. Brabec, Biopolymers, 2005, 77, 222–229.
30 C. Fabrega, R. Eritja, N. D. Sinha, M. K. Dosanjh and B. Singer, Bioorg.
Med. Chem., 1995, 3, 101–108.
31 (a) W. C. Johnson Jr., Methods Biochem. Anal., 1985, 31, 61–163;
(b) W. C. Johnson Jr., in Circular Dichroism and the Conformational
Analysis of Biomolecules, ed. G. D. Fasman, Plenum Press, New York,
1996, p. 433.
32 M. W. Kalnik, M. Kouchakdjian, B. F. Li, P. F. Swann and D. J. Patel,
Biochemistry, 1988, 27, 100–108.
33 (a) C. A. Adams, M. Melikishvili, D. W. Rodgers, J. J. Rasimas,
A. E. Pegg and M. G. Fried, J. Mol. Biol., 2009, 389, 248–263;
(b) M. Melikishvili, J. J. Rasimas, A. E. Pegg and M. G. Fried, Biochem-
istry, 2008, 47, 13754–13763.
3 (a) D. M. Noll, T. M. Mason and P. S. Miller, Chem. Rev., 2006, 106,
277–301; (b) K. M. McCabe, S. B. Olson and R. E. Moses, J. Cell.
Physiol., 2009, 220, 569–573; (c) T. V. Ho and O. D. Schärer, Environ.
Mol. Mutagen., 2010, 51, 552–566.
4 A. Guainazzi and O. D. Schärer, Cell. Mol. Life Sci., 2010, 67,
3683–3697.
5 (a) K. Stevens, D. D. Claeys, S. Catak, S. Figaroli, M. Hocek,
J. M. Tromp, S. Schürch, V. Van Speybroeck and A. Madder, Chem.–Eur.
J., 2011, 17, 6940–6953; (b) K. Hattori, T. Hirohama, S. Imoto,
S. Kusano and F. Nagatsugi, Chem. Commun., 2009, 6463–6465;
(c) T. Angelov, A. Guainazzi and O. D. Schärer, Org. Lett., 2009, 11,
661–664; (d) H. Huang, H. Y. Kim, I. D. Kozekov, Y. J. Cho, H. Wang
and A. Kozekova, et al., J. Am. Chem. Soc., 2009, 131, 8416–8424;
(e) S. Hentschel, J. Alzeer, T. Angelov, O. D. Schärer and N. W. Luedtke,
Angew. Chem., Int. Ed., 2012, 51, 1–5.
34 M. G. Fried, S. Kanugula, J. L. Bromberg and A. E. Pegg, Biochemistry,
1996, 35, 15295–15301.
35 D. S. Daniels, C. D. Mol, A. S. Arvai, S. Kanugula, A. E. Pegg and
J. A. Tainer, EMBO J., 2000, 19, 1719–1730.
6 E. M. Hlavin, M. B. Smeaton, A. M. Noronha, C. J. Wilds and
P. S. Miller, Biochemistry, 2010, 49, 3977–3988.
36 J. D. Puglisi and I. Tinoco Jr., Methods Enzymol., 1989, 180, 304–
325.
37 A. T. Brunger, P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros and
R. W. Grosse-Kunstleve, et al., Acta Crystallogr., Sect. D: Biol. Crystal-
logr., 1998, D52, 30–42.
7 (a) J. O. Ojwang, D. A. Grueneberg and E. L. Loechler, Cancer Res.,
1989, 49, 6529–6537; (b) Q. Dong, D. Barsky, M. E. Colvin,
C. F. Melius, S. M. Ludeman and J. F. Moravek, et al., Proc. Natl. Acad.
Sci. U. S. A., 1995, 92, 12170–12174; (c) R. T. Streeper, R. J. Cotter,
M. E. Colvin, J. Hilton and O. M. Colvin, Cancer Res., 1995, 55,
1491–1498.
38 B. R. Brooks, R. E. Bruccoleri, B. D. Olafson, D. J. States,
S. Swaminathan and M. Karplus, J. Comput. Chem., 1983, 4, 187–217.
8 S. Kallama and K. Hemminki, Chem.-Biol. Interact., 1986, 57, 85–96.
7090 | Org. Biomol. Chem., 2012, 10, 7078–7090
This journal is © The Royal Society of Chemistry 2012