Preparation of covalently linked complexes between DNA and O 6-alkylguanine-DNA alkyltransferase using interstrand cross-linked DNA

Article abstract of DOI:10.1021/bc300553u

O⁶-alkylguanine-DNA alkyltransferases (AGT) are responsible for the removal of alkylation at both the O⁶ atom of guanine and O ⁴ atom of thymine. AGT homologues show vast substrate differences with respect to the size of the adduct and which alkylated atoms they can restore. The human AGT (hAGT) has poor capabilities for removal of methylation at the O⁴ atom of thymidine, which is not the case in most homologues. No structural data are available to explain this poor hAGT repair. We prepared and characterized O⁶G-butylene-O⁴T (XLGT4) and O⁶G-heptylene-O⁴T (XLGT7) interstrand cross-linked (ICL) DNA as probes for hAGT and the Escherichia coli homologues, OGT and Ada-C, for the formation of DNA-AGT covalent complexes. XLGT7 reacted only with hAGT and did so with a cross-linking efficiency of 25%, while XLGT4 was inert to all AGT tested. The hAGT mediated repair of XLGT7 occurred slowly, on the order of hours as opposed to the repair of O⁶-methyl-2′-deoxyguanosine which requires seconds. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of the repair reaction revealed the formation of a covalent complex with an observed migration in accordance with a DNA-AGT complex. The identity of this covalent complex, as determined by mass spectrometry, was composed of a heptamethylene bridge between the O⁴ atom of thymidine (in an 11-mer DNA strand) to residue Cys145 of hAGT. This procedure can be applied to produce well-defined covalent complexes between AGT with DNA.

Full text of DOI:10.1021/bc300553u

Article
pubs.acs.org/bc
Preparation of Covalently Linked Complexes Between DNA and
O6‑Alkylguanine-DNA Alkyltransferase Using Interstrand Cross-
Linked DNA
Francis P. McManus,‡ Amardeep Khaira,‡ Anne M. Noronha,‡ and Christopher J. Wilds*,‡
‡Department of Chemistry and Biochemistry, Concordia University, 7141 Sherbrooke St. West, Montrea
́
l, QC, Canada H4B 1R6
S
* Supporting Information
ABSTRACT: O6-alkylguanine-DNA alkyltransferases (AGT)
are responsible for the removal of alkylation at both the O6
atom of guanine and O4 atom of thymine. AGT homologues
show vast substrate differences with respect to the size of the
adduct and which alkylated atoms they can restore. The human
AGT (hAGT) has poor capabilities for removal of methylation
at the O4 atom of thymidine, which is not the case in most
homologues. No structural data are available to explain this poor
hAGT repair. We prepared and characterized O6G-butylene-
O4T (XLGT4) and O6G-heptylene-O4T (XLGT7) interstrand
cross-linked (ICL) DNA as probes for hAGT and the
Escherichia coli homologues, OGT and Ada-C, for the formation
of DNA-AGT covalent complexes. XLGT7 reacted only with
hAGT and did so with a cross-linking efficiency of 25%, while XLGT4 was inert to all AGT tested. The hAGT mediated repair of
XLGT7 occurred slowly, on the order of hours as opposed to the repair of O6-methyl-2′-deoxyguanosine which requires seconds.
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of the repair reaction revealed the formation of
a covalent complex with an observed migration in accordance with a DNA-AGT complex. The identity of this covalent complex,
as determined by mass spectrometry, was composed of a heptamethylene bridge between the O4 atom of thymidine (in an 11-
mer DNA strand) to residue Cys145 of hAGT. This procedure can be applied to produce well-defined covalent complexes
between AGT with DNA.
the processivity of DNA polymerase during DNA replication.
O6 MedG adopts a nonwobble base pair with dT causing a GC
to AT transition, while O4 MedT adopts a nonwobble base pair
with dG causing a TA to CG transition. These two
modifications account for 8% of the total DNA damage caused
by N-methyl-N-nitrosourea (MNU).12 Endogenously, O6
MedG is formed at a rate of 10−30 events daily by S-
adenosylmethionine.13
INTRODUCTION
■
Modified nucleic acids with novel properties have had a long-
standing interest and application in the field of bioorganic
chemistry.1 One example of how these probes can be employed
involves the formation of covalent complexes with proteins
where weak binding was previously observed leading to a wide
range of applications.2−4
Formation of covalent complexes for crystallographic
purposes can be employed for various systems due to the
control one can have on the nature of the chemical
modification, which can be tailored to specific protein
complexes.5,6 For example, the structure of hAGT complexed
with DNA has been determined with the aid of modified
nucleic acids.7,8 One of these structures, significant in
enhancing our understanding of the mechanism of hAGT,
contains a chemical cross-link formed between the hAGT active
site and DNA containing the modified nucleoside N1,O6-
ethanoxanthosine.
The formation of O6 MedG is detrimental to cells where the
bioaccumulation of 6 650 O6 MedG modifications in hAGT
deficient tumor cells is sufficient to cause cell lethality.14 Unlike
O6 MedG, O4 MedT is affected by neither the levels of hAGT
nor mismatch repair (MMR), which is why this lesion is more
mutagenic in mammalian cells.15,16 Processing of O4 MedT in
mammalian cells is believed to rely solely on the nucleotide
excision repair (NER) pathway.17
hAGT and Ada-C, the most thoroughly characterized AGT
proteins, repair alkylated DNA by flipping the damaged base
out of the DNA duplex and into the active site where the alkyl
group is transferred from the point of lesion to the active site-
AGT proteins, which are found in all kingdoms of life, are
responsible for the repair of the mutagenic O6-methylguanosine
(O6 MedG) and O4-methylthymidine (O4 MedT) lesions.9−11
The mutagenicity of these lesions rise from the wobble base
pair that these modified bases adopt with respect to their
natural compliments. These wobble base pairs adversely affect
Received: October 9, 2012
Revised: December 19, 2012
Published: January 24, 2013
© 2013 American Chemical Society
224
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
Figure 1. Structures of (A) the O4-thymidine-alkylene-O6-2′-deoxyguanosine cross-links, (B) oligonucleotides XLGT4 (n = 1) and XLGT7 (n = 4),
and (C) AGT-DNA complex formed if XLGT4 and XLGT7 are repaired by AGT.
Cys residue.18 Once alkylated, this suicide protein is degraded
by the ubiquitin pathway in mammalian cells.19 There are
interesting differences in the ability of AGTs to process
alkylation damage. For example, hAGT is unable to process O4
MedT efficiently, unlike OGT and Ada-C.20,21
hAGT has also been employed in an elegant method to
observe the movement of an hAGT fusion protein which reacts
with O6-benzylguanine derivatives.22
Previously, mimics of the ICL generated by 1,7-heptanediol
dimethanesulfonate (hepsulfam) have been studied by our
group.23−25 Results from our previous work indicate hAGT, but
not the E. coli homologues, can repair ICL that involve two O6
atoms of 2′-deoxyguanosine that are directly opposed, in a 1,2
stagger, or in the more clinically relevant 1,3 staggered
orientation. On the other hand, ICL DNA containing a
heptamethylene linkage bridging the O4 atoms of thymidine are
inert to hAGT and the E. coli homologues.
In the current study, the preparation of an ICL containing an
O4-thymidine-alkylene-O6-2′-deoxyguanosine cross-link is de-
scribed. The approach to prepare this ICL involves the method
described by Swann’s group to prepare a ″convertible″
thymidine to introduce an alkyl chain with a terminal alcohol
at the O4 atom of thymidine followed by a Mitsunobu reaction
at the O6 atom of protected 2′-deoxyguanosine to form the
alkylene linkage (Figure 1A).26,27 hAGT, OGT, and Ada-C
were expressed and purified to homogeneity to investigate their
role in the repair of O4-thymidine-alkylene-O6-2′-deoxyguano-
sine ICL XLGT4 and XLGT7 (Figure 1B).
Aside from providing further insight into the substrate
discrimination patterns of the various AGT, these ICL DNA
have potential use for the preparation of site-specific covalently
linked AGT-DNA adducts. The formation of AGT-oligonucleo-
tide complexes between the various nucleobases has been
shown to occur with 1,2-dibromoethane, attributed to the
formation of an S-(2-bromoethyl) intermediate.28 In our
approach, repair by AGT at the O6 atom of the modified 2′-
deoxyguanosine by the active site Cys of AGT covalently links
AGT to the O4 atom of thymidine by an alkylene linker. The
added facet of this approach is the ease with which the ICL
linker can be varied allowing control over the length and
chemical nature of the tether between the O4 atom of
thymidine and the active site of AGT (Figure 1c).
glass) supports were purchased from Glen Research (Sterling,
Virginia). All other chemicals and solvents were purchased from
the Aldrich Chemical Co. (Milwaukee, WI) or EMD Chemicals
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. (Gibbstown, NJ). NMR spectra were recorded either on a
Varian INOVA 300 MHz NMR spectrometer or on a Varian
1
500 MHz NMR spectrometer at room temperature. H NMR
spectra were recorded at a frequency of 300.0 MHz, and
chemical shifts were reported in parts per million downfield
from tetramethylsilane. 31P NMR spectra (1H decoupled) were
recorded at a frequency of 202.3 MHz with H3PO4 used as an
external standard. Ampicillin, isopropyl β-D-thiogalactopyrano-
side (IPTG), dithiothreitol (DTT), EASY-BLUE, Protein Stain,
and most other biochemical reagents including polyacrylamide
gel materials were purchased from Bioshop Canada Inc.
(Burlington, ON). Ni-NTA Superflow Resin was purchased
from Qiagen (Mississauga, ON). Complete, Mini, EDTA-free
Protease Inhibitor Cocktail Tablets were obtained from Roche
(Laval, QC) Nitrocellulose filters (0.20 μm) were obtained
from Millipore. XL-10 Gold E. coli cells were obtained from
Stratagene (Cedar Creek, TX). T4 polynucleotide kinase
(PNK) and Unstained Protein Molecular Weight Marker
were obtained from Fermentas (Burlington, ON). [γ-32P]ATP
was purchased from Amersham Canada Ltd. (Oakville, ON).
Chemical Synthesis of Modified Nucleosides. 1-{O4-
[3′-O-(t-Butyldimethylsilyl)-5′-O-(4,4′-dimethoxytrityl)-thy-
midinyl]}-4-{O6-[3′-O-(t-butyldimethylsilyl)-5′-O-(4,4′-dime-
thoxytrityl)-N2-phenoxyacetyl-2′-deoxyguanosinyl]}-butane
(1a). To 5′-O-dimethoxytrityl-3′-O-t-butyldimethylsilyl-N2-
(phenoxyacetyl)-2′-deoxyguanosine (0.150 g, 0.183 mmol)
was added 5′-O-dimethoxytrityl-3′-O-t-butyldimethylsilyl-O4-
(hydroxybutyl)-thymidine (0.135 g, 0.185 mmol) and
triphenylphosphine (0.208 g, 7.70 mmol) in dioxane (2 mL)
followed by dropwise addition of diisopropyl azodicarboxylate
(DIAD) (1.521 g, 7.52 mmol) at room temperature. After 30
min, the solvent was evaporated in vacuo, the crude product
taken up in ethyl acetate (100 mL), and the solution washed
with two portions of 3% sodium bicarbonate (2 × 100 mL).
The organic layer was dried over sodium sulfate (4g) and
concentrated to produce a yellow gum. The crude product was
purified by flash column chromatography using a hexane/ethyl
acetate (3:7) solvent system to afford 0.166 g (59%) of product
as a colorless foam. Rf (SiO2 TLC): 0.17 hexane/ethyl acetate
(3:7).
EXPERIMENTAL PROCEDURES
■
Materials. 5′-O-Dimethoxytrityl-N2-phenoxyacetyl-2′-deox-
yguanosine and N,N-diisopropylamino cyanoethyl phosphona-
midic chloride were purchased from ChemGenes Inc.
(Wilmington, MA). 5′-O-Dimethoxytrityl-2′-deoxyribonucleo-
side-3′-O-(β-cyanoethyl-N,N′-diisopropyl)phosphoramidites
and protected 2′-deoxyribonucleoside-CPG (control pore
1H NMR (300 MHz, DMSO-d6, ppm): 10.63 (s, 1H, NH),
8.45 (s, 1H, H8), 7.93 (s, 1H, H6), 7.46−6.79 (m, 31H, Ar),
225
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
6.45−6.40 (dd, 1H, H1′a, J = 6.3 Hz), 6.23−6.17 (dd, 1H,
H1′b, J = 6.6), 5.06 (s, 2H, PhOCH2CO), 4.74−4.70 (m, 1H,
H3′a), 4.68−4.63 (t, 2H, CH2−O6), 4.50−4.46 (m, 1H, H3′b),
4.43−4.39 (m, 2H, CH2−O4), 3.95−3.90 (m, 2H, H4′a and
H4′b), 3.78 (s, 6H, O−CH3), 3.76 (s, 6H, O−CH3), 3.42−3.22
(m, 4H, H5′a, H5″a, H5′b and H5″b), 3.04−2.97 (m, 1H,
H2′a), 2.43−2.28 (m, 3H, H2″a, H2′b and H2″b), 2.03−1.90
(m, 4H, O4−CH2−CH2−CH2−CH2−O6), 1.64 (s, 3H, C5−
CH3), 0.83 (s, 18H, Si−C(CH3)3), 0.06 (s, 6H, Si−CH3), 0.01
(s, 3H, Si−CH3), 0.00 (s, 3H, Si−CH3).
1H NMR (300 MHz, DMSO-d6, ppm) (Figure S1): 10.60 (s,
1H, NH), 8.36 (s, 1H, H8), 7.81 (s, 1H, H6), 7.42−6.71 (m,
31H, Ar), 6.43−6.38 (dd, 1H, H1′a, J = 6.3 Hz), 6.22−6.17
(dd, 1H, H1′b, J = 6.3), 5.36−5.33 (m, 2H, OHa and OHb),
5.04 (s, 2H, PhOCH2CO), 4.63−4.58 (t, 2H, CH2−O6), 4.53−
4.49 (m, 1H, H3′a), 4.39−4.34(m, 3H, CH2−O4 and H3′b),
4.02−3.95 (m, 2H, H4′a and H4′b), 3.74 (s, 6H, O−CH3),
3.72 (s, 3H, O−CH3), 3.71 (s, 3H, O−CH3), 3.32−3.08 (m,
4H, H5′a, H5′b, H5″a and H5″b), 2.94−2.84 (m, 1H, H2′a),
2.42−2.12 (m, 3H, H2″a, H2′b and H2″b), 1.96−1.89 (m, 4H,
O4−CH2−CH2−CH2−CH2−O6).
+
HRMS (ESI-MS) m/z calculated for C86H104N7O15Si2
+
1530.7129: found 1530.7087 [M+H]+.
HRMS (ESI-MS) m/z calculated for C74H76N7O15
1302.5399: found 1302.5392 [M+H]+.
1-{O4-[3′-O-(t-Butyldimethylsilyl)-5′-O-(4,4′-dimethoxytri-
tyl)-thymidinyl]}-7-{O6-[3′-O-(t-butyldimethylsilyl)-5′-O-(4,4′-
dimethoxytrityl)-N2-phenoxyacetyl-2′-deoxyguanosinyl]}-
heptane (1b). To 5′-O-dimethoxytrityl-3′-O-t-butyldimethylsil-
yl-N2-(phenoxyacetyl)-2′-deoxyguanosine (0.700 g, 0.851
mmol) was added 5′-O-dimethoxytrityl-3′-O-t-butyldimethyl-
silyl-O4-(hydroxyheptyl) thymidine (0.600 g, 0.774 mmol) and
triphenylphosphine (0.850 g, 3.25 mmol) in dioxane (8.46 mL)
followed by dropwise addition of DIAD (0.720 g, 3.56 mmol)
at room temperature. After 22 h, the solvent was evaporated in
vacuo, the crude product taken up in ethyl acetate (100 mL)
and the solution washed with two portions of 3% sodium
bicarbonate (2 × 100 mL). The organic layer was dried over
sodium sulfate (4g) and concentrated to produce a yellow gum.
The crude product was purified by flash column chromatog-
raphy using a hexane/ethyl acetate (1:1 → 4:6) solvent system
to afford 0.77 g (70%) of product as a colorless foam. Rf (SiO2
TLC): 0.33 hexane/ethyl acetate (4:6).
1-{O4-[5′-O-(4,4′-Dimethoxytrityl)-thymidinyl]}-7-{O6-[5′-
O-(4,4′-dimethoxytrityl)-N2-phenoxyacetyl-2′-deoxyguano-
sinyl]}-heptane (2b). To compound 1b (0.760g, 0.480 mmol)
was added TBAF (1 M in THF) (0.282 g, 1.01 mmol) in THF
(3 mL). After 30 min, the solvent was evaporated in vacuo and
the crude product taken up in ethyl acetate (50 mL). The
solution was washed with three portions of sodium bicarbonate
(3 × 75 mL) followed by two back-extractions of the sodium
bicarbonate washes with ethyl acetate (2 × 25 mL). The
organic layer was dried over sodium sulfate (4g), evaporated in
vacuo, and purified by flash column chromatography using an
ethyl acetate/methanol (199:1 → 19:1) solvent system to
afford 0.620 g (95%) of product as a colorless foam. Rf (SiO2
TLC): 0.20 ethyl acetate/methanol (49:1).
1H NMR (300 MHz, CDCl3, ppm) (Figure S2): 8.79 (s, 1H,
NH), 8.02 (s, 1H, H8), 8.85 (s, 1H, H6), 7.43−6.78 (m, 31H,
Ar), 6.59−6.55 (dd, 1H, H1′a, J = 6.6 Hz), 6.44−6.37 (dd, 1H,
H1′b, J = 6.3 Hz), 4.83 (m, 1H, OHa), 4.72 (s, 2H,
PhOCH2CO), 4.61−4.57 (m, 3H, CH2−O6 and H3′a),
4.39−4.35 (t, 2H, CH2−O4), 4.22−4.18 (m, 1H, H3′b),
4.16−4.11 (m, 2H, H4′a and H4′b), 3.80 (s, 6H, O−CH3),
3.77 (s, 6H, O−CH3), 3.51−3.33 (m, 4H, H5′a, H5″a, H5′b
and H5″b), 2.99 (m, 1H, OHb), 2.82−2.76 (m, 1H, H2′a),
2.67−2.59 (m, 2H, H2″a and H2′b), 2.32−2.23 (m, 1H,
H2″b), 1.93−1.86 (m, 2H, R−CH2−CH2−O6), 1.76−1.70 (m,
5H, R−CH2−CH2−O4 and C5-CH3),1.58−1.44 (m, 6H, O4−
CH2−CH2−CH2−CH2−CH2−CH2−CH2−O6).
1H NMR (300 MHz, CDCl3, ppm): 8.72 (s,1H, NH), 8.10
(s, 1H, H8), 8.00 (s, 1H, H6), 7.50−6.84 (m, 31H, Ar), 6.52−
6.48 (dd, 1H, H1′a, J = 6.6 Hz), 6.43−6.39 (dd, 1H, H1′b, J =
6.0 Hz), 4.86 (s, 2H, PhOCH2CO), 4.68−4.64 (m, 2H, CH2−
O6), 4.58−4.52 (m, 1H, H3′a), 4.46−4.41 (t, 2H, CH2−O4),
4.19−4.16 (m, 1H, H3′b), 4.05−4.01 (m, 2H, H4′a and H4′b),
3.86 (s, 6H, O−CH3), 3.84 (s, 6H, O−CH3), 3.60−3.31 (m,
4H, H5′a, H5″a, H5′b and H5″b), 2.82−2.73 (m, 1H, H2′a),
2.61−2.48 (m, 2H, H2″a and H2′b), 2.32−2.24 (m, 1H,
H2″b), 2.00−1.95 (m, 2H, R−CH2−CH2−O6), 1.85−1.80 (m,
2H, R−CH2−CH2−O4), 1.66 (s, 3H, C5−CH3), 1.58 (m, 2H,
R−CH2−CH2−CH2−O6), 1.50 (m, 4H, R−CH2−CH2−CH2−
CH2−O4), 0.93 (s, 9H, Si−C(CH3)3), 0.88 (s, 9H, Si−
C(CH3)3), 0.12 (s, 3H, Si−CH3), 0.07 (s, 3H, Si−CH3), 0.00
(s, 6H, Si−CH3).
+
HRMS (ESI-MS) m/z calculated for C77H82N7O15
1344.5869: found 1344.5861 [M+H]+.
1-{O4-[3′-O-(β-cyanoethyl-N,N′-diisopropyl)-5′-O-(4,4′-di-
methoxytrityl)-thymidinyl]}-4-{O6-[3′-O-(β-cyanoethyl-N,N′-
diisopropyl)-5′-O-(4,4′-dimethoxytrityl)-N2-phenoxyacetyl-
2′-deoxyguanosinyl]}-butane (3a). 2a (0.200 g, 0.154 mmol)
and diisopropylethylamine (0.060 g, 0.461 mmol) were
dissolved in THF (1.5 mL). Then, N,N-diisopropylaminocya-
noethylphosphonamidic chloride (0.091 g, 0.383 mmol) was
added dropwise and the reaction allowed to stir at room
temperature for 30 min. This was followed by two subsequent
additions of diisopropylethylamine (2 × 0.012 g, 0.091 mmol)
and N,N-diisopropylaminocyanoethylphosphonamidic chloride
(2 × 0.019 g, 0.081 mmol) with 1 h intervals. Upon
completion, the solvent was evaporated in vacuo, the crude
product taken up in ethyl acetate (25 mL), and the solution
washed with 3% sodium bicarbonate (25 mL) and brine (50
mL). The organic layer was dried over sodium sulfate,
decanted, and evaporated. The product, a colorless powder,
was precipitated from hexanes (0.221 g, 85%). Rf (SiO2 TLC):
0.56, 0.61, 0.66, 0.73 ethyl acetate.
+
HRMS (ESI-MS) m/z calculated for C89H110N7O15Si2
1572.7598: found 1572.7592 [M+H]+.
1-{O4-[5′-O-(4,4′-Dimethoxytrityl)-thymidinyl]}-4-{O6-[5′-
O-(4,4′-dimethoxytrityl)-N2-phenoxyacetyl-2′-deoxyguano-
sinyl]}-butane (2a). To compound 1a (0.150 g, 0.10 mmol)
was added tetra-n-butylammonium fluoride (TBAF) (1 M in
tetrahydrofuran (THF)) (0.059 g, 0.21 mmol) in THF (1 mL).
After 30 min, the solvent was evaporated in vacuo and the
crude product taken up in ethyl acetate (50 mL). The solution
was washed with three portions of sodium bicarbonate (3 × 50
mL) followed by two back-extractions of the sodium
bicarbonate washes with ethyl acetate (2 × 25 mL). The
organic layer was dried over sodium sulfate (4g), evaporated in
vacuo, and purified by flash column chromatography using an
ethyl acetate/methanol (95:5) solvent system to afford 0.083 g
(64%) of product as a colorless foam. Rf (SiO2 TLC): 0.14
ethyl acetate/methanol (95:5).
31P NMR (202.3 MHz, d6-acetone, ppm) (Figure S3):
153.61, 153.41, 153.40, 153.23.
226
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
+
HRMS (ESI-MS) m/z calculated for C92H110N11O17P2
Electrospray Ionization-Mass Spectrometry (ESI-MS)
of Small Molecules and DNA. ESI-MS of XLGT4 and
XLGT7 (Figures S5 and S6) were conducted at the Concordia
University Centre for Biological Applications of Mass
Spectrometry using a Micromass Q-tof2 mass spectrometer
(Waters) equipped with a nanospray ion source. The mass
spectrometer was operated in full scan, negative ion detection
mode. ESI mass spectra for small molecules were recorded on a
LTQ Orbitrap Velos-ETD (Thermo Scientific) in positive ion
mode in acetonitrile.
Exonuclease Digestion. XLGT4 and XLGT7 (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-HCl, pH 8.1, and 2 mM magnesium
chloride) for 4 days at 37 °C. Analysis of the resulting
nucleoside mixtures were performed by reversed phase HPLC
on a Symmetry C-18 5 μm column (0.46 × 15 cm). A linear
gradient of 0−70% buffer B over 30 min was used to elute the
various species (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
(Figures S7 and S8) were compared to the standard nucleotides
which eluted at the following times: dC (4.7 min), dG (7.5
min), dT (8.1 min), dA (8.9 min), O6G-butylene-O4T (18.4
min), and O6G-heptylene-O4T (25.3 min). The ratio of
nucleosides was determined based on known extinction
coefficients, aside from the cross-linked nucleosides where an
extinction coefficient of 6200 M−1 cm−1 was used correspond-
ing to the sum of the extinction coefficients of O4 MedT and O6
MedG (Table S1).
UV Thermal Denaturation Studies of DNA Duplexes.
The molar extinction coefficients for XLGT4 and XLGT7 were
determined using the nearest-neighbor method (M−1 cm−1)
assuming a central G-C match. On the basis of these extinction
coefficients, the control DNA (having a G-C match instead of
the ICL), XLGT4 and XLGT7 were lyophilized to dryness and
reconstituted in 1 mL of 90 mM sodium chloride, 10 mM
sodium phosphate, and 1 mM EDTA buffer (pH 7.0) at a final
concentration of 3 μM. Prior to use, the samples were subjected
to heating at 90 °C for 10 min and slowly cooled back to room
temperature and finally stored at 4 °C overnight. The
denaturation profiles were obtained by heating the samples at
a rate of 0.5 °C min−1 while monitoring the absorbance at 260
nm using a Varian CARY model 3E spectrophotometer fitted
with a six-sample thermostatted cell block and a temperature
controller. Data processing was performed as described by
Puglisi and Tinoco.29
Circular Dichroism (CD) Spectroscopy of DNA
Duplexes. Circular dichroism signatures were obtained on a
Jasco J-815 spectropolarimeter equipped with a Julaba F25
circulating bath using the samples prepared for UV thermal
denaturation studies. The spectra were the average of 5 scans
taken at a rate of 20 nm min−1, with a bandwidth of 1 nm and
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 [ϕ] was obtained from the relation
ϕ = ε/Cl, where ε is the relative ellipticity (mdeg), C is the
molar concentration of the DNA duplex (mol/L), and l is the
path length in cm.
1702.7556: found 1702.7557 [M+H]+.
1-{O4-[3′-O-(β-cyanoethyl-N,N′-diisopropyl)-5′-O-(4,4′-di-
methoxytrityl)-thymidinyl]}-7-{O6-[3′-O-(β-cyanoethyl-N,N′-
diisopropyl)-5′-O-(4,4′-dimethoxytrityl)-N2-phenoxyacetyl-
2′-deoxyguanosinyl]}-heptane (3b). 2b (0.200 g, 0.149
mmol) and diisopropylethylamine (0.058 g, 0.446 mmol)
were dissolved in THF (1.5 mL). Then, N,N-diisopropylami-
nocyanoethylphosphonamidic chloride (0.088 g, 0.372 mmol)
was added and the reaction allowed to stir at room temperature
for 30 min. This was followed by two subsequent additions of
diisopropylethylamine (2 × 0.012 g, 0.091 mmol) and N,N-
diisopropylaminocyanoethylphosphonamidic chloride (2 ×
0.019 g, 0.081 mmol) with 1 h intervals. Upon completion,
the solvent was evaporated in vacuo, the crude product taken
up in ethyl acetate (25 mL), and the solution washed with 3%
sodium bicarbonate (25 mL) and brine (50 mL). The organic
layer was dried over sodium sulfate, decanted, and evaporated.
The product, a colorless powder, was precipitated from hexanes
(0.218 g, 84%). Rf (SiO2 TLC): 0.50, 0.58, 0.67, 0.75 ethyl
acetate.
31P NMR (202.3 MHz, d6-acetone, ppm) (Figure S4):
153.60, 153.40, 153.39, 153.22.
+
HRMS (ESI-MS) m/z calculated for C95H116N11O17P2
1745.8032: found 1745.8059 [M+H]+.
Solid-Phase Synthesis of XLGT4 and XLGT7. DNA
duplexes XLGT4 and XLGT7 were assembled using an Applied
Biosystems model 3400 synthesizer on a 1 μmol scale
employing β-cyanoethylphosphoramidite cycles supplied by
the manufacturer with certain modifications to coupling times
as indicated below.
Commercially available fast deprotecting 3′-O-2′deoxynu-
cleoside phosphoramidites were dissolved in anhydrous
acetonitrile at a concentration of 0.1 M and the modified bis-
3′-O-2′-deoxyphosphoramidites (3a and 3b) at 0.05 M.
Assembly of sequences began with detritylation (3% trichloro-
acetic acid (TCA) in CH2Cl2), followed by phosphoramidite
coupling: commercial 3′-O-2′-deoxyphosphoramidites (2 min)
and modified bis-phosphoramidite (3a and 3b) (30 min);
capping was performed 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. Final removal
of trityl groups was carried out by the synthesizer to yield the
final oligonucleotide on the solid support.
Due to the presence of an O4-alkyl thymidine adduct,
modifications to the standard deprotection protocol were
employed. Deprotection of the oligonucleotide was achieved by
incubating the CPG in 500 μL of 10% 1,8-diazabicyclo[5.4.0]-
undec-7-ene (DBU) in ethanol for 3 days in the dark at room
temperature.
XLGT4 and XLGT7 were purified from preterminated
products by 20% 7 M urea denaturing polyacrylamide gels
(19:1) using 1× TBE [89 mM Tris-HCl, 89 mM boric acid, 2
mM EDTA (pH 8.0)] as running buffer to remove any residual
DBU. Prior to gel separation, 25 O.D. of DNA was lyophilized
and resuspended in 100 μL of formamide. After electrophoresis,
the desired DNA bands, as observed by UV shadowing, were
excised from the gel and submerged overnight in 0.1 M sodium
acetate. The extracted DNA was subsequently desalted using a
C-18 SEP PAK cartridge (Waters Inc.).
Protein Preparation. All AGT homologues were expressed
under the promoter of the pQE30 vector in XL-10 Gold E. coli
cells. The C145S hAGT variant was designed as described
previously using site-directed mutagenesis.24 The XL-10 Gold
227
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
host bacteria were grown in 1 L of LB broth + 100 μg/mL
ampicillin until an OD600 = 0.6 was reached. After the addition
of IPTG, at a final concentration of 0.3 mM, the cells were
incubated at 37 °C for an additional 4 h. The cells were
centrifuged at 6000g at 4 °C for 20 min and resuspended in
lysis buffer [20 mM Tris-HCl (pH 8.0), 250 mM NaCl, and 20
mM β-mercaptoethanol supplemented with Complete, Mini,
EDTA-free Protease Inhibitor Cocktail Tablets] at a final
density of 1 g wet pellet per 5 mL of buffer. The cells were
lysed by French press and the solution clarified by centrifuging
at 17 000g for 45 min at 4 °C. The supernatant was applied to
Ni-NTA column and washed with lysis buffer supplemented
with 20 mM imidazole and the 6×His-tagged AGT proteins
eluted with lysis buffer supplemented with 200 mM imidazole.
The pooled fractions were dialyzed against storage buffer [50
mM Tris-HCl (pH 7.6), 250 mM NaCl, 20 mM β-
mercaptoethanol, and 0.1 mM EDTA] in 8000 Da cutoff
dialysis tubing. Ten milligrams of hAGT and OGT were
obtained per liter of culture and 50 mg of Ada-C was obtained
for the same volume.
Repair Assay. DNA substrates were labeled at the 5′
position using 32P-ATP as previously described.24 The labeling
solution consisting of 200 pmol of dsDNA, 1 μL [γ-32P]ATP
(10 μCi/μL), and 5 units of T4 PNK in 10 μL of 1× PNK
buffer was incubated at 37 °C for 1 h. The reactions were
terminated by boiling the samples for 10 min and slowly
cooling back to room temperature, the samples stored at 4 °C
overnight and diluted 10-fold in Activity Buffer [10 mM Tris-
HCl (pH 7.6), 100 mM NaCl and 1 mM DTT] prior to use.
Repair studies of XLGT4 and XLGT7 were performed by
incubating 2 pmol of DNA and 60 pmol of AGT in a total
volume of 15 μL made up of Activity Buffer and allowed to
react at 37 °C for the desired time.23 Prior to gel loading, the
reactions were terminated by adding 18.2 μL of stop reaction
buffer [81 mM Tris-HCl, 81 mM boric acid, 1.8 mM EDTA,
and 1% SDS (sodium dodecyl sulfate) (pH 8.0) in 80%
formamide] and boiled for 5 min. The samples were analyzed
by 17% 7 M urea denaturing polyacrylamide gel (19:1) run in
1× TBE [89 mM Tris-HCl, 89 mM boric acid, 2 mM EDTA
(pH 8.0)] for 1.5 h at 700 V. The gels were exposed to a
storage phosphor screen, the image captured on a Typhoon
9400 (GE Healthcare, Piscataway, NJ) and the radioactive
counts quantified by ImageQuant (Amersham Biosciences).
Identification of XLGT7 Repair Product by SDS-PAGE.
600 pmol of hAGT was incubated with 625 pmol of XLGT7 in
15 μL of Activity Buffer for 5 h at 37 °C. The reaction was
terminated by adding 10 μL of SDS Buffer [125 mM Tris-HCl
(pH 7.6), 2% SDS, 30% glycerol and 7.5% 2-mercaptoethanol],
boiling the sample for 5 min, and separated on a 15% SDS-
PAGE. The gel was stained with EASY-BLUE Protein Stain as
per the manufacturer’s protocol.
source voltage 3.5 kV, mass range of 700−1999 m/z in positive
ion mode. The theoretical mass of the hAGT-DNA species was
calculated to be either 25 303.6 Da if the repair event occurred
at the modified 2′-deoxyguanosine (21876.2 Da (hAGT) +
96.2 Da (Linker) + 3331.2 Da (CGAAATTTTCG)) or 25
328.6 Da if the repair event occurred at the modified thymidine
(21 876.2 Da (hAGT) + 96.2 Da (Linker) + 3356.2 Da
(CGAAAGTTTCG)).
C145S hAGT Binding Studies. Binding reactions con-
sisted of 0.5 nM dsDNA, increasing C145S hAGT ranging from
1 to 5 μM in a total 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 incubated for 30
min at room temperature and loaded on a 15% native
polyacrylamide gel (75:1) containing 0.5× TBE. Electro-
phoresis was carried out using 0.5× TBE as running buffer at
21 °C with 250 V for 25 min. The gel was processed as
described above for the repair assay.
Binding parameters for the AGT-DNA interaction were
obtained from the electrophoretic mobility shift assays as
described previously.30 Binding of n moles of AGT protein [P]
to a mole of DNA [D] is described by the equation: nP + D ↔
PnD. Taking the logarithm yields: log ([PnD]/[D]) = n log
[P]free + log Kn. Plotting of log [PnD]/[P] as a function of log
[P] produces a slope that represents the stoichiometry (n), and
the Kn can be obtained from the relationship Kn = 10−(y‑intercept)
.
The reported Kd (dissociation constant) was obtained by taking
the nth root of Kn.
RESULTS
■
Substrate Preparation. The structures of XLGT4 and
XLGT7 are shown in Figure 1. The O6G-butylene-O4T (3a)
and O6G-heptylene-O4T (3b) cross-linked amidites were
prepared as shown in Scheme 1 which began with the
nucleoside adduct containing a tetra- or heptamethylene linker
with a terminal alcohol which was prepared by reacting the
appropriate dialcohol with 3′-O-(tert-butyldimethylsilyl)-5′-O-
(4,4′-dimethoxytrityl)-O4-triazolyl-thymidine.25 The 3′ and 5′
bis-protected thymidine adducts were subsequently reacted
with 5′-O-dimethoxytrityl-3′-O-t-butyldimethylsilyl-N2-(phe-
noxyacetyl)-2′-deoxyguanosine via a Mitsunobu reaction to
give the fully protected dimers 1a and 1b.31 Desilylation of the
dimers gave 2a and 2b and was followed by a phosphitylation
reaction yielding the cross-linked bis-amidites 3a and 3b, which
were employed for solid-phase synthesis. 31P NMR signals at
153.61, 153.41, 153.40, and 153.23 ppm for 3a and 153.60,
153.40, 153.39, and 153.22 ppm for 3b confirmed the presence
of both phosphoramidite groups. As a result of the asymmetry
for the phosphitylated cross-linked nucleosides, four 31P NMR
signals were observed.
ESI-MS and Exonuclease Digestion of DNA. XLGT4
and XLGT7 were analyzed by mass spectrometry to ensure that
the base modifications were present after the deprotection and
purification steps (Figures S5 and S6). The observed masses
were in agreement with calculated values.
XLGT4 and XLGT7 nucleoside compositions were con-
firmed by exonuclease digestion to their constituent nucleosides
with a mixture of snake venom phosphodiesterase and calf
intestinal phosphatase, then analyzed by C-18 reversed phase
HPLC. The elution profiles are shown in Figures S7 and S8.
The presence of modified nucleosides was established by the
appearance of additional peaks, aside from the four standard
nucleosides, on the HPLC chromatogram. The extracted ratios
Identification of XLGT7 Repair Product by ESI-MS.
4000 pmol of hAGT was incubated with 1200 pmol of XLGT7
in 100 μL of Activity Buffer for 3 h at 37 °C. Ten microlilters of
the sample was loaded on a Spursil C-18 column (3 μm, 2.1
mm ×150 mm) using an Agilent 1200 Series LC system at a
flow rate of 0.2 mL/min with the following gradient program:
0−5 min, hold at 5% B; 5−30 min, linear gradient from 5% B to
50% B; 30−40 min, linear gradient from 50% B to 90% B; 40−
42 min, hold at 90% B; 42−50 min, linear gradient from 90% B
to 5% B (buffer A, 0.1% formic acid in water and buffer B, 0.1%
formic acid in acetonitrile). ESI-MS was conducted on a
Micromass Q-Tof Ultima API with the following conditions:
228
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
a
Circular Dichroism Spectroscopy of DNA Duplexes.
The alkylene linkers present in XLGT4 and XLGT7 showed
little effect on overall secondary structure as indicated by the
characteristic B-form DNA signatures observed for all samples
with positive signals centered around 275 nm, cross overs
around 260 nm, and negative signals around 250 nm, as shown
in Figure 3.32,33 These results are in agreement with previously
studied ICL DNA having central O4-thymidine-alkylene-O4-
thymidine and O6-2′-deoxyguanosine-alkylene-O6-2′-deoxygua-
nosine mismatches showing classic B-form CD signatures.25,34
Repair Assay. Sixteen hour repair reactions with the various
AGT and XLGT4 or XLGT7 were conducted to detect the
total amount of repair by the different AGT homologues. The
positive control XLGG7 (O6-2′-deoxyguanosine-heptylene-O6-
2′-deoxyguanosine ICL) incubated with hAGT was used as a
marker, since the various species produced by this reaction are
known and aid in the identification of the various products
formed in our assay, including two slowly migrating bands
corresponding to AGT-DNA complexes that are believed to be
either fully reduced or partially reduced by 2-mercaptoetha-
nol.24
Scheme 1
No appreciable repair was observed by any of the AGT
tested with XLGT4 (data not shown). Repair of XLGT7,
shown in Figure 4, was achieved only by hAGT, where 25% of
a
Reagents and conditions: (i) 5′-O-dimethoxytrityl-3′-O-t-butyldime-
thylsilyl-N2-(phenoxyacetyl)- 2′-deoxyguanosine, triphenylphosphine,
DIAD, dioxane; (ii) tetrabutylammonium fluoride (1 M in THF),
THF; (iii) N,N-diisopropylamino cyanoethyl phosphonamidic chlor-
ide, diisopropylethylamine, THF.
of the component 2′-deoxynucleosides were in good agreement
with the theoretical compositions except for the lower than
anticipated proportion of cross-linked modifications due to
their assumed extinction coefficients, as illustrated in Table S1.
UV Thermal Denaturation Studies of DNA Duplexes.
The stability of XLGT4 and XLGT7 containing duplexes were
determined by UV thermal denaturing studies by monitoring
OD260 at increasing temperatures, as illustrated in Figure 2.
Figure 3. Far-UV circular dichroism spectra of XLGT4, XLGT7, and
control DNA duplexes: Control DNA (red), XLGT4 (green), and
XLGT7 (blue).
the starting material was converted to the AGT-DNA covalent
complex. Moreover, the ratio of the released DNA strand and
Figure 2. Tm curves of XLGT4, XLGT7, and control DNA duplexes.
Control DNA (red), XLGT4 (green), and XLGT7 (blue).
Both modified duplexes displayed sigmoidal denaturation
patterns comparable to the control duplex. The presence of
the ICL stabilized the DNA duplex with the 4 methylene linker
having a greater contribution than the 7 methylene linker. The
melting temperatures for XLGT4, XLGT7, and the control (G-
C pair instead of a G-T cross-link) were 68, 56, and 46 °C,
respectively.
Figure 4. Denaturing PAGE gel illustrating the repair of XLGT7 by
hAGT, Ada-C and OGT using 2 pmol of DNA and 60 pmol AGT at
37 °C for 16 h. Lane 1, Control DNA + hAGT; lane 2, XLGT7; lane 3,
XLGT7 + hAGT; lane 4, XLGT7 + OGT; lane 5, XLGT7 + Ada-C;
lane 6, XLGG7 + hAGT.
229
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
AGT-DNA complex was 1:1, suggesting hAGT only repaired
one of the alkylated atoms involved in the cross-link. This
hAGT repair activity was abolished by introducing the C145S
mutation (data not shown).
The repair rate of XLGT7 and formation of the hAGT-DNA
complex was monitored over the course of 3 h to obtain
information on the kinetics of this reaction, as shown in Figure
5. The results indicate that the repair occurred slowly where
masses, the hAGT-DNA complex occurs between the protein
and the DNA strand alkylated at the O4 atom of thymidine,
indicating hAGT preferentially repairs the alkylated 2′-
deoxyguanosine (Figure 8).
C145S hAGT Binding Studies. Binding assays were
carried out to study the effect of the ICL on hAGT substrate
discrimination in hopes of understanding the preferential repair
observed for XLGT7 over XLGT4 (see Figure S9). Results
presented in Table 1 indicate that hAGT binds both ICL DNA
in a similar 2:1 (protein/DNA) stoichiometry, similar to the
control DNA, suggesting the cross-links have no evident effect
on the mode of hAGT binding.
The Kd values extracted from the EMSA suggest hAGT binds
XLGT7 2-fold better than the control DNA and XLGT4, which
are both bound in the low μM range.
DISCUSSION
■
Directly opposed O6-2′-deoxyguanosine-alkylene-O4-thymidine
ICL were prepared using a combination of solution synthesis to
create the cross-linked phosphoramidite and solid-phase
synthesis to incorporate the cross-link into an oligonucleotide.
The identities of the ICL DNA XLGT4 and XLGT7 were
confirmed by ESI-MS and exonuclease digestion. The nucleo-
side ratios extracted from the digestion were in accordance with
the theoretical composition of the oligonucleotide except for a
lower than anticipated ratio of the cross-linked nucleosides.
These deviations in ratios are attributed to the unknown
extinction coefficient for these novel nucleosides where the sum
of the extinction coefficients of O4 MedT and O6 MedG was
used to get an estimate of the relative ratio of these
modifications. The complimenting nature of the ESI-MS
validates the presence of only one ICL modification per
DNA molecule.
⧫
Figure 5. Time course repair of XLGT7 by hAGT. Substrate ( ),
hAGT-DNA complex ( ), and released DNA strand ( ).
■
▲
only 13% of the AGT-DNA species is present after the three
hour period. As was the case for the repair assay conducted
overnight, the ratio of AGT-DNA complex to released DNA
strand was 1:1.
Identification of XLGT7 repair product by SDS-PAGE
and ESI-MS. Since the radioactivity-based repair assays only
show the DNA species, SDS-PAGE was conducted on the
hAGT mediated repair of XLGT7 to observe the various
protein species formed. The SDS-PAGE shown in Figure 6
The presence of the cross-links increased the thermal
stability of the DNA duplex, where the presence of a four
methylene bridge had a greater stabilizing effect than the seven
as observed by our UV thermal denaturation studies. The
smaller increase in thermal stability brought upon by the larger
ICL is believed to be due to a greater reduction in the enthalpy
for the heptylene versus butylene linker containing DNA, which
may be attributed to the disruption of the base pairs around the
cross-linked site as a result of steric effects of the larger linker.25
A comparable trend has been observed previously by our group
for an O4-thymidine-alkylene-O4-thymidine mismatch where an
increase in melting temperature (Tm) of 9 and 25 °C was
observed over the control duplex for the heptylene and
butylene linkages, respectively.25 This also holds true for an O6-
2′-deoxyguanosine-alkylene-O6-2′-deoxyguanosine mismatch in
a similar oligonucleotide sequence with 12 and 24 °C increases
in Tm for the heptylene and butylene linkages over our control
DNA, respectively.34 Indeed, the presence of an ICL,
irrespective of the length, causes a reduction in the change of
entropy upon denaturation of the duplex due to the organized
nature of the unimolecular system. As previously observed, the
contribution of the entropy factor overrides the negative
enthalpic effect on the ICL system affording a more stable
species.35
Figure 6. 15% SDS-PAGE of hAGT mediated repair of XLGT7.
Repair of 625 pmol XLGT7 by 600 pmol hAGT for 5 h at 37 °C:
unstained protein molecular weight marker (left); XLGT7 + hAGT
reaction (right).
reveals the formation of a protein complex with a greater
molecular weight than unmodified hAGT. On the basis of the
results from Figures 4 and 6, we hypothesize that the new
species corresponds to an hAGT-DNA covalent complex,
which coincides with the calculated mass of (25.3 kDa).
The identity of the hAGT-DNA complex hypothesized from
Figure 6 was determined using ESI-MS. Since the cross-link
involves two different alkylated nucleotides that could
potentially be repaired by hAGT, it was necessary to identify
the hAGT-DNA species in order to establish the site of repair.
Repair at either of the alkylation sites will yield an hAGT-DNA
species of a different mass aiding in identification.
The ESI-MS spectrum of the hAGT mediated repair of
XLGT7 is shown in Figure 7. Two major signals are observed
in the deconvoluted spectrum. The first corresponds to the
unalkylated form of hAGT and the second the hAGT-DNA
covalent complex. On the basis of the observed and calculated
The control DNA, XLGT4, and XLGT7 all assumed a B-
form structure as assessed by far-UV CD. These results are in
accordance with previously studied ICL DNA having cross-
links between the N3 atoms of thymidine and the N4 atoms of
cytidine.36,37 Moreover, NMR studies on ICL DNA containing
an N3-thymidine-butylene-N3-thymidine cross-link revealed no
230
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
Figure 7. Identification of hAGT-DNA covalent complex by ESI-MS.
nucleic acids, could play a part in poor RNA repair by hAGT.
For this reason, the B-form structure of DNA may be
primordial for AGT repair.
Our radioactivity-based repair assay revealed no repair of
XLGT4 by any AGT tested. Lack of repair by OGT and Ada-C
was anticipated based on previous work, which indicates that
both proteins are inactive to ICL DNA.23,25
The absence of any appreciable repair of XLGT4 by hAGT
was surprising considering its ability to repair O6-2′-
deoxyguanosine-butylene-O6-2′-deoxyguanosine ICL
(XLGG4) in a virtually identical sequence.23,24 The difference
in size between thymidine and guanosine may be responsible
for the lack of XLGT4 repair as opposed to XLGG4. Replacing
one 2′-deoxyguanosine by a thymidine reduces the linear
distance between the anomeric carbons of the modified
nucleosides by roughly 1 Å.
A molecular model generated by Fang et al. of hAGT in
complex with XLGG7 suggests the preferential repair of longer
cross-links by hAGT is brought upon by the shape of the hAGT
active site in the DNA-bound conformation, which can be
represented as a tunnel.23 In effect, the larger the distance
between the anomeric carbons of the modified nucleosides, the
greater the ICL repair since the cross-linked nucleobases will sit
at each end of the hAGT active site ″tunnel″. This allows the
linker to lay in the ″tunnel″ while at the same time placing the
α-carbon of the adduct in close proximity to the reactive Cys
residue.
Figure 8. hAGT mediated repair of XLGT7 cross-link.
Table 1. Kd and Stoichiometry of Binding by hAGT to
XLGT4, XLGT7, and Control Duplex
DNA duplex
Kd (μM)
stoichiometry
Control (G-C)
XLGT4
3.55 0.14
3.64 0.20
2.05 0.05
2.03 0.08
1.99 0.17
1.97 0.02
XLGT7
Local denaturation of the DNA duplex upon hAGT binding
may be the cause for the extension of the cross-link, whereby
separating the tethered nucleotides.40 As eluded to in earlier
studies based on the thermal denaturation results, the larger
cross-link causes the disruption of the base pairs around the
lesion itself generating a substrate that is already denatured
around the point of alkylation prior to protein binding.
Repair of XLGT7 was achieved only by hAGT, as was
anticipated based on the ″tunnel″ model. Increasing the length
of the ICL allowed the nucleobases to span the distance of the
hAGT active site, permitting repair to occur. Repair of XLGT7
by hAGT shown in lane 3 of Figure 4 suggests the formation of
a species that corresponds to a AGT-DNA complex, based on
the migration pattern of the XLGG7 control from lane 6, as
well as a fully repaired DNA strand, which migrated similarly to
change in the overall B-form structure of the DNA but did
show widening of the major groove without compromising the
size of the minor groove.38
The conformation of the DNA probe might play a substantial
role in hAGT repair. The repair of O6-methyl-guanosine (O6-
methyl-riboG) in a DNA strand by hAGT is nonexistent as
opposed to its natural substrate, O6MedG which is efficiently
processed by the protein.39 On the basis of hAGT-DNA
cocrystals, the lack of RNA repair by hAGT has been attributed
to a potential steric clash between the 2′ hydroxyl of the
modified ribonucleoside and the α-carbon of Gly131.7 This
only partially addresses the discriminatory pattern. Other
nucleic acid properties may be responsible for this discerning
trend. Notably, differences in the ribose and 2′-deoxyribose
sugar puckers, affecting the overall secondary structure of
231
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
the control DNA in lane 1. The formation of the AGT-DNA
complex and the release of the repaired strand in a 1:1 ratio
implies that repair by hAGT only occurred at one site per DNA
molecule. Repair of XLGT7 by hAGT required hours for
completion as indicated by the time course assay shown in
Figure 5. The C145S variant of hAGT was unable to repair
XLGT7 identifying Cys145 as the reactive residue.
promising. Synthesis of these modified oligonucleotides is
straightforward and allows great control over the length of the
desired tether. As demonstrated herein, hAGT specifically
reacts with the O6 alkyated 2′-deoxyguanosine of the cross-link
allowing the formation of defined covalent complexes that may
find applications in experiments involving the labeling of
hAGT.
To overcome the limitations of the radioactivity-based assay,
which only monitors the DNA species, an SDS-PAGE analysis
of the reaction was performed to analyze the various
proteinaceous materials formed. The absence of a protein
band at 43.8 kDa, which would indicate an hAGT dimer linked
by the seven methylene linker, supports the finding that hAGT
only reacts at one location on XLGT7. A band just above the
25 kDa marker correlates to a covalent species migrating with a
molecular weight corresponding to an AGT-DNA complex in a
1:1 ratio (roughly 25.3 kDa), which is in agreement with our
radioactivity-based repair assay.
ASSOCIATED CONTENT
■
S
* Supporting Information
XLGT4 and XLGT7 characterization data and representative
EMSA binding plots are available. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*Telephone: (514)-848-2424 ext. 5798. Fax: (514)-848-2868.
E-mail: Chris.Wilds@concordia.ca.
The identity of the covalent complex was determined by ESI-
MS. The spectrum shown in Figure 7 displays some unreacted
hAGT (21 875.5 Da) and a peak with a mass of 25 303.3 Da,
corresponding to an hAGT-(CH2)7-CGAAATTTTCG mole-
cule. The absence of a peak at 25 328.6 Da indicates hAGT
reacts selectively at one location on the DNA duplex. To form
this hAGT-(CH2)7-CGAAATTTTCG product, where the
underlined nucleoside depicts the cross-linked position,
hAGT must react on the modified CGAAAGTTTCG strand
thereby releasing the repaired CGAAAGTTTCG oligonucleo-
tide in the process. The absence of any formation of a covalent
complex when reacting hAGT with the control DNA signifies
the reactive moiety on the DNA as the O6 atom of the cross-
linked 2′-deoxyguanosine. The repair events of XLGT7 by
hAGT have been summarized in Figure 8. These results are
consistent with our previous results that demonstrate that
hAGT can repair O6-2′-deoxyguanosine-alkylene-O6-2′-deoxy-
guanosine cross-links but is inert toward O4-thymidine-
alkylene-O4-thymidine cross-links.23−25
The inability of hAGT to repair XLGT4 was not attributed
to poor substrate binding as established from our EMSA results
shown in Table 1. A 2-fold decrease in Kd observed for XLGT7
over XLGT4 and the control DNA is not substantial enough to
support the repair trend. Also, the ICL had no effect on the
binding stoichiometry of hAGT to the DNA. This is in
agreement with the minor groove binding property of hAGT
and the solution structure of similar ICL DNA showing the
cross-link spanning the major groove without influencing the
size of the minor groove.7,38,41
Using our approach, a well-defined hAGT-DNA covalent
complex was generated covalently linking an O4 alkylated
thymidine via a seven methylene linker to hAGT. Unfortu-
nately, the protein−DNA cross-linking reaction was limited to
the human homologue where no complex was observed for
Ada-C and OGT E. coli proteins. No complex was observed
with the four methylene linker containing DNA XLGT4. The
reason for this substrate discrimination by hAGT is not known,
but we hypothesize that the longer cross-link found in XLGT7
allows for easier access to the point of lesion, since (1) the two
modified nucleosides can be separated further apart due to the
increased length of the tether and (2) the larger cross-link
allows for substantial local destabilization of the DNA duplex at
the point of repair granting easier access to hAGT.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
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. F.P.M. is the
recipient of postgraduate fellowships from the Groupe de
́ ́
Recherche Axe sur la Structure des Proteines (GRASP) and
NSERC. Members of the group would like to thank Dr.
Anthony E. Pegg for the plasmids encoding the wild-type Ada-
C, OGT, and hAGT genes.
REFERENCES
■
(1) Egli, M., and Pallan, P. S. (2010) The many twists and turns of
DNA: template, telomere, tool, and target. Curr. Opin. Struct. Biol. 20,
262−275.
(2) Corneillie, T. M., Whetstone, P. A., Lee, K. C., Wong, J. P., and
Meares, C. F. (2004) Converting weak binders into infinite binders.
Bioconjugate Chem. 15, 1389−1391.
(3) Niemeyer, C. M. (2002) The developments of semisynthetic
DNA-protein conjugates. Trends Biotechnol. 20, 395−401.
(4) Niemeyer, C. M. (2010) Semisynthetic DNA-protein conjugates
for biosensing and nanofabrication. Angew. Chem., Int. Ed. Engl. 49,
1200−1216.
(5) Staker, B. L., Hjerrild, K., Feese, M. D., Behnke, C. A., Burgin, A.
B., Jr, and Stewart, L. (2002) The mechanism of topoisomerase I
poisoning by a camptothecin analog. Proc. Natl. Acad. Sci. U. S. A. 99,
15387−15392.
(6) Golan, G., Zharkov, D. O., Grollman, A. P., Dodson, M. L.,
McCullough, A. K., Lloyd, R. S., and Shoham, G. (2006) Structure of
T4 pyrimidine dimer glycosylase in a reduced imine covalent complex
with abasic site-containing DNA. J. Mol. Biol. 362, 241−258.
(7) Daniels, D. S., Woo, T. T., Luu, K. X., Noll, D. M., Clarke, N. D.,
Pegg, A. E., and Tainer, J. A. (2004) DNA binding and nucleotide
flipping by the human DNA repair protein AGT. Nat. Struct. Mol. Biol.
11, 714−720.
(8) Duguid, E. M., Rice, P. A., and He, C. (2005) The structure of
the human AGT protein bound to DNA and its implications for
damage detection. J. Mol. Biol. 350, 657−666.
(9) Spratt, T. E., and Levy, D. E. (1997) Structure of the hydrogen
bonding complex of O6-methylguanine with cytosine and thymine
during DNA replication. Nucleic Acids Res. 25, 3354−3361.
(10) Preston, B. D., Singer, B., and Loeb, L. A. (1986) Mutagenic
potential of O4-methylthymine in vivo determined by an enzymatic
We conclude that the use of ICL DNA as substrates for the
formation of site-specific AGT-DNA covalent complexes is
232
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233

Bioconjugate Chemistry
Article
approach to site-specific mutagenesis. Proc. Natl. Acad. Sci. U. S. A. 83,
8501−8505.
the human DNA repair protein, O6-methylguanine DNA methyl-
transferase (MGMT). Nucleic Acids Res. 34, 1884−1891.
(28) Liu, L., Pegg, A. E., Williams, K. M., and Guengerich, F. P.
(2002) Paradoxical enhancement of the toxicity of 1,2-dibromoethane
by O6-alkylguanine-DNA alkyltransferase. J. Biol. Chem. 277, 37920−
37928.
(11) Pegg, A. E. (2000) Repair of O(6)-alkylguanine by
alkyltransferases. Mutat. Res. 462, 83−100.
(12) Kleibl, K. (2002) Molecular mechanisms of adaptive response to
alkylating agents in Escherichia coli and some remarks on O(6)-
methylguanine DNA-methyltransferase in other organisms. Mutat. Res.
512, 67−84.
(29) Puglisi, J. D., and Tinoco, I., Jr. (1989) Absorbance melting
curves of RNA. Methods Enzymol. 180, 304−325.
(30) Fried, M. G., Kanugula, S., Bromberg, J. L., and Pegg, A. E.
(1996) DNA binding mechanism of O6-alkylguanine-DNA alkyl-
transferase: stoichiometry and effects of DNA base composition and
secondary structure on complex stability. Biochemistry 35, 15295−
15301.
(31) Mitsunobu, O. (1981) The use of diethyl azodicarboxylate and
triphenylphosphine in synthesis and transformation of natural
products. Synthesis 1, 1−28.
(13) Rydberg, B., and Lindahl, T. (1982) Nonenzymatic methylation
of DNA by the intracellular methyl group donor S-adenosyl-L-
methionine is a potentially mutagenic reaction. EMBO J. 1, 211−216.
(14) Rasouli-Nia, A., Sibghat-Ullah, Mirzayans, R., Paterson, M. C.,
and Day, R. S., 3rd. (1994) On the quantitative relationship between
O6-methylguanine residues in genomic DNA and production of sister-
chromatid exchanges, mutations and lethal events in a mer- human
tumor cell line. Mutat. Res. 314, 99−113.
(32) Johnson, W. C., Jr. (1985) Circular dichroism and its empirical
application to biopolymers. Methods Biochem. Anal. 31, 61−163.
(33) Johnson, W. C., Jr. (1996) Circular Dichroism Instrumentation.
In Circular Dichroism and the Conformational Analysis of Biomolecules
(Fasman, G. D., Ed.) pp 433, Plenum Press, New York.
(34) Booth, J. D., Murphy, S. P., Noronha, A. M., and Wilds, C. J.
(2008) Effect of linker length on DNA duplexes containing a
mismatched O6-2′-deoxyguanosine-alkyl interstrand cross-link. Nucleic
Acids Symp. Ser. 52, 431−432.
(35) Hofr, C., and Brabec, V. (2005) Thermal stability and energetics
of 15-mer DNA duplex interstrand crosslinked by trans-
diamminedichloroplatinum(II). Biopolymers 77, 222−229.
(36) Wilds, C. J., Noronha, A. M., Robidoux, S., and Miller, P. S.
(2004) Mispair-aligned N3T-alkyl-N3T interstrand cross-linked DNA:
synthesis and characterization of duplexes with interstrand cross-links
of variable lengths. J. Am. Chem. Soc. 126, 9257−9265.
(37) Noronha, A. M., Noll, D. M., Wilds, C. J., and Miller, P. S.
(2002) N(4)C-Ethyl-N(4)C cross-linked DNA: synthesis and
characterization of duplexes with interstrand cross-links of different
orientations. Biochemistry 41, 760−771.
(38) Webba da Silva, M. W., Bierbryer, R. G., Wilds, C. J., Noronha,
A. M., Colvin, O. M., Miller, P. S., and Gamcsik, M. P. (2005)
Intrastrand base-stacking buttresses widening of major groove in
interstrand cross-linked B-DNA. Bioorg. Med. Chem. 13, 4580−4587.
(39) Pegg, A. E., Morimoto, K., and Dolan, M. E. (1988)
Investigation of the specificity of O6-alkylguanine-DNA-alkyltransfer-
ase. Chem. Biol. Interact. 65, 275−281.
(15) Pauly, G. T., and Moschel, R. C. (2001) Mutagenesis by O(6)-
methyl-, O(6)-ethyl-, and O(6)-benzylguanine and O(4)-methylthy-
mine in human cells: Effects of O(6)-alkylguanine-DNA alkyltransfer-
ase and mismatch repair. Chem. Res. Toxicol. 14, 894−900.
(16) Altshuler, K. B., Hodes, C. S., and Essigmann, J. M. (1996)
Intrachromosomal probes for mutagenesis by alkylated DNA bases
replicated in mammalian cells: A comparison of the mutagenicities of
O4-methylthymine and O6-methylguanine in cells with different DNA
repair backgrounds. Chem. Res. Toxicol. 9, 980−987.
(17) Klein, J. C., Bleeker, M. J., Roelen, H. C., Rafferty, J. A.,
Margison, G. P., Brugghe, H. F., et al. (1994) Role of nucleotide
excision repair in processing of O4-alkylthymines in human cells. J.
Biol. Chem. 269, 25521−25528.
(18) Verdemato, P. E., Brannigan, J. A., Damblon, C., Zuccotto, F.,
Moody, P. C., and Lian, L. Y. (2000) DNA-binding mechanism of the
Escherichia coli ada O(6)-alkylguanine-DNA alkyltransferase. Nucleic
Acids Res. 28, 3710−3718.
(19) Srivenugopal, K. S., Yuan, X. H., Friedman, H. S., and Ali-
Osman, F. (1996) Ubiquitination-dependent proteolysis of O6-
methylguanine-DNA methyltransferase in human and murine tumor
cells following inactivation with O6-benzylguanine or 1,3-bis(2-
chloroethyl)-1-nitrosourea. Biochemistry 35, 1328−1334.
(20) Sassanfar, M., Dosanjh, M. K., Essigmann, J. M., and Samson, L.
(1991) Relative efficiencies of the bacterial, yeast, and human DNA
methyltransferases for the repair of O6-methylguanine and O4-
methylthymine. suggestive evidence for O4-methylthymine repair by
eukaryotic methyltransferases. J. Biol. Chem. 266, 2767−2771.
(21) Zak, P., Kleibl, K., and Laval, F. (1994) Repair of O6-
methylguanine and O4-methylthymine by the human and rat O6-
methylguanine-DNA methyltransferases. J. Biol. Chem. 269, 730−733.
(22) Keppler, A., Gendreizig, S., Gronemeyer, T., Pick, H., Vogel, H.,
and Johnsson, K. (2003) A general method for the covalent labeling of
fusion proteins with small molecules in vivo. Nat. Biotechnol. 21, 86−
89.
(40) Federwisch, M., Hassiepen, U., Bender, K., Rajewsky, M. F., and
Wollmer, A. (1997) Recombinant human O6-alkylguanine-DNA
alkyltransferase induces conformational change in bound DNA.
FEBS Lett. 407, 333−336.
(41) Webba da Silva, M., Noronha, A. M., Noll, D. M., Miller, P. S.,
Colvin, O. M., and Gamcsik, M. P. (2002) Solution structure of a
DNA duplex containing mispair-aligned N4C-ethyl-N4C interstrand
cross-linked cytosines. Biochemistry. 41, 15181−15188.
(23) Fang, Q., Noronha, A. M., Murphy, S. P., Wilds, C. J., Tubbs, J.
L., Tainer, J. A., et al. (2008) Repair of O6-G-alkyl-O6-G interstrand
cross-links by human O6-alkylguanine-DNA alkyltransferase. Bio-
chemistry 47, 10892−10903.
(24) McManus, F. P., Fang, Q., Booth, J. D., Noronha, A. M., Pegg,
A. E., and Wilds, C. J. (2010) Synthesis and characterization of an
O(6)-2′-deoxyguanosine-alkyl-O(6)-2′-deoxyguanosine interstrand
cross-link in a 5′-GNC motif and repair by human O(6)-
alkylguanine-DNA alkyltransferase. Org. Biomol. Chem. 8, 4414−4426.
(25) McManus, F. P., O’Flaherty, D. K., Noronha, A. M., and Wilds,
C. J. (2012) Preparation of O4-alkyl-2′-deoxythymidine cross-linked
DNA to probe recognition and repair by O6-alkylguanine-DNA
alkyltransferases. Org. Biomol. Chem. 10, 7078−7090.
(26) Xu, Y. Z., and Swann, P. F. (1990) A simple method for the
solid phase synthesis of oligodeoxynucleotides containing O4-
alkylthymine. Nucleic Acids Res. 18, 4061−4065.
(27) Shibata, T., Glynn, N., McMurry, T. B., McElhinney, R. S.,
Margison, G. P., and Williams, D. M. (2006) Novel synthesis of O6-
alkylguanine containing oligodeoxyribonucleotides as substrates for
233
dx.doi.org/10.1021/bc300553u | Bioconjugate Chem. 2013, 24, 224−233