Synthesis of DNA interstrand cross-links using a photocaged nucleobase

Article abstract of DOI:10.1002/anie.201108018

The missing linking: BCNU is a chemotherapy drug that generates an ethylene bridge between N¹ of deoxyguanosine and N³ of deoxycytidine. No synthesis of a DNA containing this moiety has been reported until now. A new strategy uses a photocaged nucleobase that, when released, generates a highly reactive intermediate which cross-links the opposing DNA strand in a manner analogous to BCNU (see scheme, NBOC=ortho- nitrobenzyloxycarbonyl). Copyright

Full text of DOI:10.1002/anie.201108018

.
Angewandte
Communications
DOI: 10.1002/anie.201108018
Linking DNA
Synthesis of DNA Interstrand Cross-Links Using a Photocaged
Nucleobase**
Sarah Hentschel, Jawad Alzeer, Todor Angelov, Orlando D. Schꢀrer, and Nathan W. Luedtke*
Bifunctional alkylating agents cause a wide variety of damage
to cellular DNA. Among these, DNA–DNA interstrand
cross-links (ICLs) are the most cytotoxic, because of their
ability to block transcription and replication.[1] ICLs resulting
from endogenous (e.g. malondialdehyde) and exogenous (e.g.
formaldehyde) agents pose threats to human health.[2–4] ICLs
are also generated by anticancer drugs such as cisplatin,
nitrogen mustards, and chloro ethyl nitrosoureas.[5]
1,3-Bis(2-chloroethyl)-1-nitrosourea (BCNU or “Carmus-
tine”) is a widely used chemotherapeutic agent that generates
an ethylene bridge between N1 of deoxyguanosine (dG) and
N3 of deoxycytidine (dC) in a multistep reaction. The putative
mechanism involves chloro alkylation of dG at O6, followed
by cyclization to form an N1-O6-ethanoguanine intermediate
which undergoes attack by N3 of dC to give a N1(dG)-ethyl-
N3(dC) cross-link (Scheme 1).[6,7] Limitations to using BCNU
and other nitrosoureas as chemotherapeutics include cellular
resistance mechanisms mediated by DNA repair enzymes.[8]
Understanding these pathways will lead to improved ther-
apeutic outcomes, but studies aimed at deciphering ICL
repair pathways have been limited by the lack of well-defined
ICLs to serve as repair substrates.[9] Different approaches
have been developed to obtain ICL-containing DNA. In early
studies, duplex DNA was exposed to a vast excess of BCNU,
but this procedure yields only 1–5% of ICL products that are
not amenable to biochemical or high-resolution biophysical
studies owing to the presence of extensive monoalkylation.[6,7]
A more sophisticated approach is bidirectional DNA syn-
thesis using cross-linked dinucleosides.[10–17] Using this proce-
Scheme 1. a) Reaction pathway for BCNU-mediated ICL formation in
DNA; b) Our strategy for the synthesis of DNA–DNA ICLs using
a photocaged O6-(2-chloroethyl)-2’-deoxyguanosine. BD represents a
general base.
dure, a N1-deoxyinosine(dI)-ethyl-N3-thymidine(dT) inter-
strand cross-link was recently reported by Noronha and co-
workers as a structural mimic of the dG–dC cross-link caused
by BCNU.[16] Drawbacks to this approach include requiring
prior knowledge of the exact structure of the desired ICL, as
well as complex protecting-group strategies needed for its
synthesis. An alternative approach to ICL synthesis involves
the incorporation of stable ICL precursors into single-
stranded oligonucleotides using standard DNA synthesis.[18–26]
Following hybridization to a complementary sequence, ICL
formation can be initiated in the modified duplex. Alzeer and
Schꢀrer used this approach to generate a O6(dG)-ethyl-
N3(dT) ICL by incorporating N3-(2-chloroethyl)thymine into
oligonucleotides. Despite such notable progress, no synthesis
of a DNA containing the biologically relevant N1(dG)-ethyl-
N3(dC) cross-link resulting from BCNU treatment has ever
been reported. Such products would provide authentic
samples for biochemical, cell biological, and biophysical
studies, and the novel methodology used in their synthesis
may find other applications in biotechnology.
[*] S. Hentschel, Prof. Dr. N. W. Luedtke
Institute of Organic Chemistry, University of Zurich
Winterthurerstrasse 190, 8057 Zꢀrich (Switzerland)
E-mail: luedtke@oci.uzh.ch
Homepage: http://www.bioorganic-chemistry.com/
Dr. J. Alzeer
Palestine Polytechnic University, Applied Chemistry Department
Hebron (Palestine)
Dr. T. Angelov
Institute of Food, Nutrition and Health, ETH Zꢀrich
Schmelzbergstrasse 9, 8092 Zꢀrich (Switzerland)
Prof. Dr. O. D. Schꢁrer
Department of Pharmacological Sciences and Chemistry
Stony Brook University
New York, 11794 (USA)
Herein we report a new strategy for ICL synthesis using
a photocaged nucleobase as an ICL precursor. Our approach
utilizes an O6-(2-chloroethyl)guanine residue containing
a photolabile ortho-nitrobenzyloxycarbonyl (NBOC) group
at the N2 position. NBOC stabilizes the normally reactive O6-
chloroethyl guanine by acting as an electron-withdrawing
[**] This work was supported by the Swiss National Science Foundation
(Nr. 130074), European Commission (Nr. 005204) and Dr. Helmut
Legerlotz-Stiftung.
Supporting information for this article (details of the synthesis and
analysis) is available on the WWW under http://dx.doi.org/10.1002/
anie.201108018.
3466
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3466 –3469

Angewandte
Chemie
group to the N1 position. This ICL precursor remains stable
during and after its synthetic incorporation into duplex DNA
until NBOC is selectively removed by irradiation at 365 nm.
The resulting free amine at the N2 position activates
formation of a N1,O6-ethanoguanine cyclic intermediate by
donating electron density to the N1 position which displaces
chloride in an intramolecular SN2 reaction. The resulting
N1,O6-ethanoguanine cation is a highly reactive intermediate
that can alkylate a cytosine residue in the opposite strand to
form the ICL in a manner analogous to BCNU itself
(Scheme 1b).
The synthesis of phosphoramidite 6 (Scheme 2) com-
menced with silylation of the 3’-OH and 5’-OH groups of dG
(1). The TES group was used to protect the 3’-OH, while
TBDMS was applied at the 5’-OH position. This strategy
circumvented problematic deprotection reactions involving
3’-O-TBDMS intermediates that failed to give the desired
nucleoside 5. The N2 exocyclic amine was protected with
a photolyzable group in modest yield by addition of 2-
nitrobenzyloxy carbonyl imidazole (NBOC-Im).[27] The O6
position was activated with 2,4,6-triisopropylbenzylsulfonyl
chloride and transformed into O6-(2-hydroxyethyl)-2’-deoxy-
guanosine (3) using quinuclidine as a base. After many failed
attempts, chlorination was accomplished using a method
reported by Wanchai and Warinthorn[28] to furnish the desired
O6-(2-chloroethyl)-2’-deoxyguanosine (4) in moderate yield.
Deprotection of the silyl groups at the 3’-OH and 5’-OH by
fluoride ions was also problematic, but could be accomplished
with a (2:1) mixture of TBAF and p-TsOH. The 5’-OH was
then protected with 4,4’-dimethoxytrityl chloride and the 3’-
OH was activated under standard conditions to give the
desired phosphoramidite 6 in an overall yield of 5% over
8 steps.
Phosphoramidite 6 was incorporated into oligonucleo-
tides using “ultra-mild” DNA synthesis according to pub-
lished procedures.[29,30] Following synthesis, oligonucleotides
were cleaved from the solid support and deprotected by
treatment with diisopropylamine in methanol (1:10) at room
temperature for 15 h. Purification was conducted using
HPLC, and the products were analyzed by MALDI-TOF
mass spectrometry. The observed molecular weight (7769.3)
from the main product of the synthesis was consistent with the
calculated molecular weight (7768.3) of oligonucleotide 7
containing an NBOC group at N2 and a chloroethyl group at
O6 of the modified G residue.
Following synthesis and purification, oligonucleotide 7
was hybridized to complementary oligonucleotides (8–11,
Figure 1) by heat denaturation and slow cooling in an aqueous
phosphate buffer (pH 7.4). The complementary oligonucleo-
tides contained a fluorescent tag (Cy3) on the 5’-end, and
variable bases (N = C, T, A, G) opposite to the ICL precursor.
The resulting duplexes were irradiated for 3 min with 365 nm
laser light to remove the NBOC group.[31,32] Following
irradiation, the oligonucleotides were incubated at 378C for
17 h, and the products resolved on a 15% denaturating
polyacrylamide gel (Figure 1). All four of the irradiated DNA
sequences gave single products having reduced electropho-
retic mobilities, consistent with the formation of duplex
DNAs containing ICLs (Figure 1). Higher yields were
obtained when pyrimidines (dT and dC) were positioned
opposite to the ICL precursor as compared to purines (dA
and dG). To characterize the products generated from
Scheme 2. Synthesis of a photocaged O6-chloroethyl phosphoramidite
6: a) TBDMS-Cl, TES-Cl, imidazole, DMF, ꢀ108C!RT, 20 h, 98%;
b) NBOC-Im, [18]crown-6, NaH, THF, RT, 3 h, 55%; c) TiPBS-Cl, Et3N,
4-DMAP, CH2Cl2, 08C, 3 h 81%; d) quinuclidine, ethylene glycol, THF,
08C, 4 h, 72%; e) PPh3, Cl3CCN, THF, 08C, 20 min, 51%; f) TBAF, p-
TsOH, MeOH, THF, H2O, 08C!RT, 26 h, 53%; g) DMTr-Cl, pyridine,
RT, 2 h, 80%; h) (iPr2N)2POC2H4CN, 5-(ethylthio)-1H-tetrazole, CH2Cl2,
RT, 20 min, 80%. DMAP=dimethylaminopyridine, TBDMS-Cl=tert-
butyldimethylsilyl chloride; TES-Cl=chlorotriethyl silane; TiPBS-
Cl=2,4,6-triisopropylbenzene sulfonyl chloride; TBAF=tetra-n-butyl-
ammonium fluoride; p-TsOH=p-toluenesulfonic acid; DMTr-Cl=4,4’-
dimethoxytrityl chloride.
Figure 1. a) Schematic representation of cross-linking reactions.
b) Analysis of ICL formation by denaturating polyacrylamide gel
electrophoresis (DPAGE). Imaging was conducted using Cy3 fluores-
cence emission.
Angew. Chem. Int. Ed. 2012, 51, 3466 –3469
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3467

.
Angewandte
Communications
reactions involving pyrimidines, HPLC was used to isolate the
ICL duplex DNA 12 and 13 (Figure 2). According to MALDI
TOF mass spectrometry, the molecular weights of these ICL
DNA products (12: 15245.0, 13: 15263.0) were in excellent
agreement with the calculated values (12: 15245.2, 13:
Figure 3. Enzymatic digestion analyses of cross-linked oligonucleotides
12 (a) and 13 (b) by HPLC. Wavelength of detection: 260 nm.
Figure 2. HPLC analysis of cross-link formation following photodepro-
tection. a) Chromatogram of a crude reaction mixture containing ICL
product 12 resulting from oligonucleotides 7 and 8. b) Chromatogram
of a crude reaction mixture containing ICL product 13 and correspond-
ing starting materials 7 and 9. Bu=buffer components. Wavelength of
detection: 260 nm.
Scheme 3. Synthesis of N1(dG)-ethyl-N3(dC) dinucleoside 15 according
to published procedures.[33] a) NaI, DMSO, 558C, 11 d, 10%. dRib=2’-
deoxyribose.
15259.0). By contrast, DNA samples that were prepared
without irradiation at 365 nm gave no detectable ICL
formation (Figure 1). Together these results demonstrate
that NBOC removal is a prerequisite for efficient ICL
formation.
complex product mixtures resulting from the addition of
BCNU to duplex DNA.[6,7] Our orthogonal strategy for the
synthesis ICL DNA utilizes the same mechanistic route as
proposed by Ludlum, and it provides the exact same N1(dG)-
ethyl-N3(dC) adduct that had been isolated from enzymatic
digests of genomic DNA treated with BCNU.[6,7] Our results
therefore give direct support for the proposed ICL mecha-
nism while providing the first total synthesis of a DNA
containing the biologically relevant N1(dG)-ethyl-N3(dC)
cross-link. In stark contrast to BCNU-treatment of duplex
DNA, our approach generates homogenous ICL duplex DNA
in sufficient quantities for future biophysical and biological
analyses.
In conclusion, we have developed a novel strategy for the
synthesis of site-specific ICLs using a N2-photocaged-O6-(2-
chloroethyl)guanine. While photocaging and release strat-
egies are widely utilized for tuning molecular-recognition
interfaces,[32] our results provide, to our knowledge, the first
example of a photoinitiated reaction between biological
macromolecules accomplished by release of a latent electro-
phile. This same strategy should be applicable in other
situations where highly reactive alkylation intermediates are
utilized for site-specific covalent modifications in aqueous
solution.
To evaluate the structures of cross-links present in 12 and
13, the purified oligonucleotides were digested with snake
venom phosphodiesterase, exonuclease III, and shrimp alka-
line phosphatase. Crude products from each reaction mixture
were analyzed by HPLC and mass spectrometry (Figure 3). In
addition to the four canonical nucleosides, the ICL dinucleo-
tides “15” and “16” were present in the chromatograms.
Dinucleotide 15 exhibited an identical retention time and
high-resolution mass spectrum as an authentic sample of
N1(dG)-ethyl-N3(dC) that was prepared from O6-(2-fluo-
roethyl)-2’-deoxyguanosine (Scheme 3). Co-injection of 15
with a crude digestion mixture further verified the structure of
the ICL formed (Supporting Information). The high-resolu-
tion mass spectrum of 16 was consistent with a G–T cross-link
(Supporting Information). Together these results demon-
strate that DNA containing a O6-(2-chloroethyl)-2’-deoxy-
guanosine residue can react with a C, T, A ,or G residue in the
opposite strand, and suggest that the putative N1,O6-ethano-
guanine cationic intermediate formed after NBOC removal is
an extremely reactive and non-specific electrophile.
The proposed mechanism for G–C cross-link formation by
BCNU was first proposed by Ludlum and co-workers in 1982
(Scheme 1),[34] but it has never been rigorously demonstrated.
Previous mechanistic studies have focused on the analysis of
3468 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3466 –3469

Angewandte
Chemie
[7] D. B. Ludlum, Mutat. Res. 1990, 233, 117.
Experimental Section
[8] L. Panasci, J.-P. Paiement, G. Christodoulopoulos, A. Belenkov,
A. Malapelsa, R. Aloyz, Clin. Cancer Res. 2001, 7, 454.
[9] A. Guainazzi, O. D. Schꢀrer, Cell. Mol. Life Sci. 2010, 67, 3683.
[10] B. Bhat, N. J. Leonard, H. Robinson, A. H.-J. Wang, J. Am.
Chem. Soc. 1996, 118, 10744.
[11] E. A. Harwood, S. T. Sigurdsson, N. B. F. Edfeldt, B. R. Reid,
P. B. Hopkins, J. Am. Chem. Soc. 1999, 121, 5081.
[12] H.-Y. Li, Y.-L. Qiu, E. Moyroud, Y. Kishi, Angew. Chem. 2001,
113, 1519; Angew. Chem. Int. Ed. 2001, 40, 1471.
[13] D. M. Noll, A. M. Noronha, P. S. Miller, J. Am. Chem. Soc. 2001,
123, 3405.
[14] A. M. Noronha, D. M. Noll, C. J. Wilds, P. S. Miller, Biochemistry
2002, 41, 760.
[15] C. J. Wilds, A. M. Noronha, S. Robidoux, P. S. Miller, J. Am.
Chem. Soc. 2004, 126, 9257.
[16] C. J. Wilds, F. Xu, A. M. Noronha, Chem. Res. Toxicol. 2008, 21,
686.
[17] F. P. McManus, Q. Fang, J. D. M. Booth, A. M. Noronha, A. E.
Pegg, C. J. Wilds, Org. Biomol. Chem. 2010, 8, 4414.
[18] W. R. Kobertz, J. M. Essigmann, J. Am. Chem. Soc. 1997, 119,
5960.
DPAGE analysis: Fresh DNA solutions were prepared in 20 mm
phosphate buffer (pH 7.4) with 270 mm NaCl and 5.4 mm KCl (total
volume 20 mL) containing 1.2 mm of 7, and 1 mm of 8, 9, 10, or 11. To
anneal the DNA strands, samples were heated for 1 min to 808C and
slowly cooled to room temperature (2 h). Samples were then
irradiated with 365 nm light (3500 mWcmꢀ2) using a LED-Pen laser
(Abecon AG) for 3 min at room temperature and incubated for 17 h
at 378C. 5 mL of each sample was diluted with 10 mL of loading buffer
(50% formamide, 9 mm Tris/boric acid, 0.2 mm EDTA, 6m urea,
traces of bromphenol blue) and an additional 5 mg urea was added.
Samples were heated at 808C for 2 min and loaded onto a 15%
polyacrylamide gel (19:1 ratio of acylamide to bisacrylamide con-
taining 8m urea and 14 mm Tris/boric acid, 0.3 mm EDTA) and
resolved for 80 min at 100 V. The gels were then imaged using a flat-
bed scanner equipped for Cy3 detection (Typhoon 9400, GE Health-
care Bioscience-AB).
Preparative-scale synthesis of cross-linked DNA: Oligonucleo-
tide 7 (6.5 mm) was mixed with oligonucleotide 8 or 9 (4 mm) in 20 mm
phosphate buffer (pH 7.4) containing 420 mm NaCl, 5.4 mm KCl, and
8 mm NaI (total volume: 1 mL). The samples were annealed and
photodeprotected as described above. Products were resolved using
Varian Pro Star HPLC system and a C-18 reverse-phase column
(YMCbasic, B-22-10P 150 ꢁ 10 mm) using a linear gradient of 3–40%
CH3CN in 0.1m triethylammonium acetate buffer (pH 7) over 26 min.
The fractions corresponding to cross-linked DNA 12 and 13 were
collected and lyophilized. MALDI-TOF mass spectrometry: 15245.2
(12, calcd 15245.2); 15263.0 (13, calcd 15259.0).
[19] I. S. Hong, M. M. Greenberg, J. Am. Chem. Soc. 2005, 127,
10510.
[20] J. Alzeer, O. D. Schꢀrer, Nucleic Acids Res. 2006, 34, 4458.
[21] Z. Qiu, L. Lu, X. Jian, C. He, J. Am. Chem. Soc. 2008, 130, 14398.
[22] T. Angelov, A. Guainazzi, O. D. Schꢀrer, Org. Lett. 2009, 11, 661.
[23] X. Peng, I. S. Hong, H. Li, M. M. Seidman, M. M. Greenberg,
J. Am. Chem. Soc. 2008, 130, 10299.
[24] K. Hattori, T. Hirohama, S. Imoto, S. Kusano, F. Nagatsugi,
Chem. Commun. 2009, 6463.
[25] M. Op de Beeck, A. Madder, J. Am. Chem. Soc. 2011, 133, 796.
[26] A. Guainazzi, A. J. Campbell, T. Angelov, C. Simmerling, O. D.
Schꢀrer, Chem. Eur. J. 2010, 16, 12100.
[27] A. Stutz, C. Hçbartner, S. Pitsch, Helv. Chim. Acta 2000, 83,
2477.
[28] P. Wanchai, C. Warinthorn, Tetrahedron Lett. 2006, 47, 6821.
[29] L. C. J. Gillet, J. Alzeer, O. D. Schꢀrer, Nucleic Acids Res. 2005,
33, 1961.
Received: November 14, 2011
Published online: February 17, 2012
Keywords: alkylation · DNA · nitrosourea · photocage
.
[1] O. D. Schꢀrer, ChemBioChem 2005, 6, 27.
[2] L. J. Niedernhofer, J. S. Daniels, C. A. Rouzer, R. E. Greene,
L. J. Marnett, J. Biol. Chem. 2003, 278, 31426.
[3] M. P. Stone, Y.-J. Cho, H. Huang, H.-Y. Kim, I. D. Kozekov, A.
Kozekova, H. Wang, I. G. Minko, R. S. Lloyd, T. M. Harris, C. J.
Rizzo, Acc. Chem. Res. 2008, 41, 793.
[4] C. Auerbach, M. Moutschen-Dahmen, J. Moutschen, Mutat. Res.
1977, 39, 317.
[30] R. T. Pon, S. Yu, Tetrahedron Lett. 1997, 38, 3327.
[31] A. Patchornik, B. Amit, R. B. Woodward, J. Am. Chem. Soc.
1970, 92, 6333.
[32] G. Mayer, A. Heckel, Angew. Chem. 2006, 118, 5020; Angew.
Chem. Int. Ed. 2006, 45, 4900.
[5] D. M. Noll, T. M. Mason, P. S. Miller, Chem. Rev. 2006, 106, 277.
[6] P. L. Fischhaber, A. S. Gall, J. A. Duncan, P. B. Hopkins, Cancer
Res. 1999, 59, 4363.
[33] W. J. Bodell, K. Pongracz, Chem. Res. Toxicol. 1993, 6, 434.
[34] W. P. Tong, M. C. Kirk, D. B. Ludlum, Cancer Res. 1982, 42, 3102.
Angew. Chem. Int. Ed. 2012, 51, 3466 –3469
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 3469