Discovery and synthesis of New UV-induced intrastrand C(4-8)G and G(8-4)C photolesions

Article abstract of DOI:10.1021/ja111304f

UV irradiation of cellular DNA leads to the formation of a number of defined mutagenic DNA lesions. Here we report the discovery of new intrastrand C(4-8)G and G(8-4)C cross-link lesions in which the C(4) amino group of the cytosine base is covalently lin

Full text of DOI:10.1021/ja111304f

COMMUNICATION
pubs.acs.org/JACS
Discovery and Synthesis of New UV-Induced Intrastrand C(4ꢀ8)G and
G(8ꢀ4)C Photolesions
Martin M€unzel, Claudia Szeibert, Andreas F. Glas, Daniel Globisch, and Thomas Carell*
Center for Integrated Protein Science (CIPSM) at the Department of Chemistry, Ludwig-Maximilians-University Munich,
Butenandtstrasse 5ꢀ13, 81377 Munich, Germany
S Supporting Information
b
ABSTRACT: UV irradiation of cellular DNA leads to the
formation of a number of defined mutagenic DNA lesions.
Here we report the discovery of new intrastrand C(4ꢀ8)G
and G(8ꢀ4)C cross-link lesions in which the C(4) amino
group of the cytosine base is covalently linked to the C(8)
position of an adjacent dG base. The structure of the novel
lesions was clarified by HPLCꢀMS/MS data for UV-
irradiated DNA in combination with chemical synthesis
and direct comparison of the synthetic material with irra-
diated DNA. We also report the ability to generate the
lesions directly in DNA with the help of a photoactive
precursor that was site-specifically incorporated into DNA.
This should enable detailed chemical and biochemical
investigations of these lesions.
Figure 1. Chemical structures of UV-induced DNA photoproducts.
V light is carcinogenic because it induces the formation of a
series of lesions in cellular DNA that are in part highly
U
mutagenic.1,2 The most studied UV-induced DNA lesions are
pyrimidine-derived cyclobutane pyrimidine dimers, (6ꢀ4) lesions,
and their Dewar valence isomers.3ꢀ5 Formation of these UV-
induced lesions requires absorption of a UV photon by one
pyrimidine nucleobase followed by reaction with a second pyrim-
idine base out of either the singlet or triplet state. Other than the
pyrimidineꢀpyrimidine UV lesions, very few UV lesions have been
characterized. Most notably, thymidine is able to react with an
adjacent adenine to give a cycloaddition product that rearranges to
produce the TA lesion (Figure 1).6 Here we report the discovery
and independent synthesis of novel UV-induced DNA lesions
formed in d(CpG) or d(GpC) sequences (Figure 2). The lesions
are the first of their kind and feature as a key structural element a
direct link between the C(4)ꢀNH2 of cytosine and the C(8)
carbon atom at the guanine base. As such, they are structurally
related to known dG bulky adducts7,8 and to lesions that have been
identified in damaged DNA that was irradiated with X-rays6,9 or
γ-rays10ꢀ16 or in the presence of photosensitizers.17ꢀ19
additional absorption at 350 nm. In order to characterize the lesion,
we digested the DNA with a mixture of three enzymes (nuclease S1
for 3 h at 37 °C and then antarctic phosphatase and snake venom
phosphodiesterase for 3 h at 37 °C) and analyzed the obtained
nucleoside mixture by high-resolution HPLCꢀMS/MS on a Ther-
mo Finnigan Orbitrap XL instrument (Figure 2). Indeed, aside from
the signals of the canonical bases dC and dG, a new signal with a
high-resolution molecular weight of 492.1717 g/mol was detected at
a detection wavelength of 350 nm. Analysis of the molecular weight
and MS/MS data pointed to the formation of a dCꢀdG cross-link
with a missing central phosphodiester, which was possibly removed
during the digestion procedure. On the basis of these data, we
postulated for the new DNA lesion the structure of the C(4ꢀ8)G
lesion depicted in Figure 2. Upon irradiation of oligonucleotides in
which we had swapped the position of the dG and dC bases, the
corresponding G(8ꢀ4)C product was detected. Additional experi-
ments showed that formation of the lesions is an intrastrand process
and that it occurs with the adjacent base.17,21 We furthermore found
both compounds in irradiated double strands that contained a central
d(CpG) or d(GpC), which may hint at biological significance. For
details, please see the Supporting Information.
The lesion was discovered in experiments in which we irradiated
small oligonucleotides with UV light (254 nm) in a glovebox.
Whereas irradiation of oligonucleotides with a central d(TpT) or
d(TpC) sequence furnished as expected the corresponding (6ꢀ4)
lesions, dG-rich oligonucleotides with a central dC under exactly the
same conditions20 furnished a small but significant amount of a new
oligonucleotide with an unknown lesion (for yields, see Table S1 in
the Supporting Information). This lesion was found to possess an
In order to prove that our proposed structure is correct, we started
the total synthesis of the expected digestion product without the
Received: December 15, 2010
Published: March 22, 2011
r
2011 American Chemical Society
5186
dx.doi.org/10.1021/ja111304f J. Am. Chem. Soc. 2011, 133, 5186–5189
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 1. Total Synthesis of the CꢀG Photoproducts
without Central Phosphatea
Figure 2. Structure of the C(4ꢀ8)G photolesion and the HPLC
chromatogram at 260 nm after irradiation and enzymatic digestion of
ODN1 [50-d(GGC GG)-30]. The insets show the UV chromatogram at
350 nm (blue) and the HRMS trace at 493.1790 ( 0.006 Da (green,
[M þ Hþ]þ). The CꢀG photolesion in its dephosphorylated form can
be detected in the HRMS trace at 32 min. The red PO2ꢀ unit was most
likely cleaved out during enzymatic digestion. X = unknown products.
central phosphodiester, as shown in Scheme 1. A Buchwaldꢀ
Hartwig reaction was utilized as the key step for the assembly, which
was challenging because Pd-catalyzed CꢀN bond formations with
cytosine derivatives had limited precedent.22 For the coupling we had
to protect all further reactive groups of the dG unit.23ꢀ25 In order to
streamline the synthesis, we chose to employ a protective-group
strategy that allowed final removal of all of the protective groups in a
single step.
a Conditions: (a) NBS, 94%; (b) TBS-Cl, imid., 98%; (c) TMS-EtOH,
PPh3, DIAD, 75%; (d) (ClMe2SiCH2)2, LiHMDS, 79%; (e) TBS-dC,
(()-BINAP, LiHMDS, Pd2(dba)3, 33%; (f) TBAF, 50%.
Scheme 2. Synthesis of 4-NHOHdC Phosphoramidite 10a
The starting point of the synthesis was deoxyguanosine (1), which
was first brominated at the C(8) position with NBS and then TBS-
protected to give compound 2.23 Alkylation of the C(6)dO bond
with TMS-ethanol under Mitsunobu conditions furnished compound
3, which was treated with bis(chlorodimethylsilyl)ethane22 and
LiHMDS to provide the fully protected 8-bromo-dG derivative 4.
Pd-catalyzed BuchwaldꢀHartwig coupling of this compound with
TBS-protected cytidine furnished 5 in 33% yield along with 55%
recovery of 3. Final deprotection of compound 5 yielded the desired
putative lesion 6as a dinucleotide with indeed a strong UV absorption
at 350 nm (for the UV spectrum, please see Figure S2 in the
Supporting Information). Co-injection of the synthetic material with
digested irradiated DNA in the HPLCꢀMS experiment showed that
the retention time, the high-resolution molecular weight, and the UV
absorption characteristics of the synthetic compound and the ques-
tionable digestion product were identical (see Figure S1). We
consequently conclude that the synthesized compound 6 is indeed
the C(4ꢀ8)G or G(8ꢀ4)C photoproduct without the central
phosphate backbone, thereby proving that the C(4ꢀ8)G lesion
has the structure depicted in Figure 2.
We next wanted to learn more about the mechanism leading to the
formation of the novel lesions, and we also wanted to gain further
support for the proposed structures. We believe that UV irradiation of
cytosine-containing DNA leads to the formation of the cytosine base
in the excited triplet state. In this state, the dC(4)ꢀNH2 group is
known to possess substantial spin density,26ꢀ28 which for example
allows the formation of the well-studied deoxycytosine-containing
(6ꢀ4) lesions.4 Indeed, the observation that the novel lesion forms
after irradiation of DNA preferentially under exclusion of oxygen in a
a Conditions: (a) TBS-Cl, imid., 83%; (b) triisopropylbenzenesul-
fonyl chloride, NaH, 99%; (c) NH3OHCl, DBU, 34%; (d) Ac2O, 97%;
(e) HF pyr, pyridine, 63%; (f) DMT-Cl, pyridine, 83%; (g) bis-
3
(diisopropylamino)(2-cyanoethoxy)phosphine, diisopropyltetrazolide,
55%. C* = 4-NHOHdC.
glovebox is a strong hint that a triplet state is involved in lesion
formation. The dC(4)ꢀNH2 “radical” could in the next step attack
the C(8) position of the adjacent dG base, in accord with the well-
known fact that radicals tend to react with the dG base at C(8) to give,
for example, the mutagenic lesion 8-oxo-dG.29ꢀ31 Final oxidation of
5187
dx.doi.org/10.1021/ja111304f |J. Am. Chem. Soc. 2011, 133, 5186–5189

Journal of the American Chemical Society
COMMUNICATION
In the digested sample, we observed next to the canonical bases two
new signals, both of which had the unusual long-wavelength absorp-
tion. Both compounds had molecular weights consistent with selective
formation of the expected C(4ꢀ8)G lesion. One signal corresponded
to the already known C(4ꢀ8)G compound with the open backbone.
The second signal originated from an incomplete digestion product
consisting of the C(4ꢀ8)G lesion including the phosphate plus an
adjacent thymidine. Surprisingly, no formation of a d(TC) adduct was
observed, showing the preferred reaction of dC(4)ꢀNH with an
3
adjacent guanine. This result supports the idea that a UV-induced
dC(4)ꢀN-centered cytosine radical can attack the C(8) position of
an adjacent dG base to form the CG photoproducts. The hydro-
xylamine-dC precursor furthermore allows the direct production of
oligonucleotides containing the novel C(4ꢀ8)G lesion. It can be site-
specifically incorporated in sufficient amounts and excellent purity.
This will enable detailed biochemical investigation of the unusual new
intrastrand cross-link lesions.36 Furthermore, our total synthesis
allows the preparation of isotope-labeled lesion analogues for direct
quantification of the compound in cellular DNA.37
In summary, we have reported the discovery of new UV-
induced lesions that are formed by oxidation of an initially
formed dCꢀdG cross-link to give a highly conjugated hetero-
cyclic ring system. The new lesions are formed in d(CpG) and
d(GpC) sequences in both single strands and duplexes. They
feature new absorption bands above 350 nm, which allow ready
detection. The lesions can easily be site-specifically prepared in
DNA by irradiation of DNA containing a hydroxylamine-dC
precursor situated next to a dG base.
Figure 3. (A) HPLC chromatogram of ODN2 directly after irradiation.
(B) HPLC chromatogram of purified ODN2 containing the C(4ꢀ8)G
photolesion. (C) HPLC chromatogram and MS traces (negative-ion
mode) after irradiation, HPLC purification, and enzymatic digestion
of ODN2.
’ ASSOCIATED CONTENT
S
Supporting Information. Irradiation of oligonucleotides,
b
preparation of compounds 1ꢀ10, oligonucleotide synthesis and
purification, enzymatic digestion, and detailed HPLCꢀMS
data. This material is available free of charge via the Internet at
http://pubs.acs.org.
the reaction product by traces of oxygen would then furnish the highly
conjugated (4ꢀ8) lesion. On the basis of the reaction conditions, we
do not believe that the formation of the lesion is initiated by a radical
such as HO or CO3 , as was proposed for similar G(8ꢀ3)T,
’ AUTHOR INFORMATION
3
3
G(8ꢀ5)C, and G(8ꢀ5)mC lesions.17,32,33 For a mechanistic pro-
Corresponding Author
thomas.carell@cup.uni-muenchen.de
posal, please see the Supporting Information. In order to gain support
for this hypothesis, we prepared a dC(4)ꢀNH radical dC precursor
3
and incorporated it into DNA. We chose the hydroxylamine-dC
derivative,34,35 which should give the dC(4)ꢀNH radical upon
’ ACKNOWLEDGMENT
3
irradiation by cleavage of the rather weak NꢀO bond.
We thank Amra Mulahasanovic (LMU Munich) for prepara-
tive work. Financial support was obtained from the DFG (CA
275/8-4 and SFB749). M.M. thanks the Fonds der Chemischen
Industrie for a Kekulꢀe Fellowship.
The synthesis of the dC hydroxylamine precursor phosphorami-
dite was accomplished in seven steps, as shown in Scheme 2. After
TBS protection of deoxyuridine 7, the 4-position was activated as a
sulfonate to yield compound 8. Nucleophilic aromatic substitution
with hydroxylamine hydrochloride and subsequent acetylation and
deprotection with HF in pyridine furnished compound 9. Free
nucleoside 9 was converted to the phosphoramidite 10 by DMT
protection and subsequent phosphitylation. The incorporation into
oligonucleotides was possible using standard phosphoramidite chem-
istry. For the lesion formation experiment, we prepared the oligo-
nucleotide ODN2 in which the precursor was situated next to a 30-dG
base and a 50-dT (Scheme 2). Using this setup, we intended to get
’ REFERENCES
(1) Cadet, J.; Sage, E.; Douki, T. Mutat. Res. 2005, 571, 3–17.
(2) Friedel, M. G.; Cichon, M. K.; Carell, T. CRC Handbook of
Organic Photochemistry and Photobiology, 2nd ed.; CRC Press: Boca
Raton, FL, 2004.
(3) Glas, A. F.; Kaya, E.; Schneider, S.; Heil, K.; Fazio, D.; Maul,
M. J.; Carell, T. J. Am. Chem. Soc. 2010, 132, 3254–3255.
(4) Heil, K.; Pearson, D.; Carell, T. Chem. Soc. Rev. [Online early
access]. DOI: 10.1039/C000407N. Published Online: Nov 15, 2010.
(5) Cadet, J.; Vigni, P. In Photochemistry and the Nucleic Acids;
Morrison, H., Ed.; Bioorganic Photochemistry, Vol. 1; Wiley:
New York, 1990; Chapter 1.
direct insight into the reactivity of dC(4)ꢀNH (dG vs dT).
3
Irradiation of ODN2 with a mercury lamp for 14 h gave only
one new distinct oligonucleotide with a characteristic absorption at
>300 nm (Figure 3a). The irradiated DNA strand was purified
(Figure 3b), digested, and analyzed by HPLCꢀMS (Figure 3c).
(6) Zhao, X.; Taylor, J. S. Nucleic Acids Res. 1996, 24, 1561–1565.
5188
dx.doi.org/10.1021/ja111304f |J. Am. Chem. Soc. 2011, 133, 5186–5189

Journal of the American Chemical Society
COMMUNICATION
(7) Schorr, S.; Schneider, S.; Lammens, K.; Hopfner, K. P.; Carell, T.
Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 20720–20725.
(8) Hoffmann, G. R.; Fuchs, R. P. Chem. Res. Toxicol. 1997,
10, 347–359.
(9) Box, H. C.; Dawidzik, J. B.; Budzinski, E. E. Free Radical Biol.
Med. 2001, 31, 856–868.
(10) Delatour, T.; Douki, T.; Gasparutto, D.; Brochier, M. C.; Cadet,
J. Chem. Res. Toxicol. 1998, 11, 1005–1013.
(11) Gu, C.; Wang, Y. Biochemistry 2004, 43, 6745–6750.
(12) Hong, H.; Cao, H.; Wang, Y. Nucleic Acids Res. 2007,
35, 7118–7127.
(13) Romieu, A.; Bellon, S.; Gasparutto, D.; Cadet, J. Org. Lett. 2000,
2, 1085–1088.
(14) Su, D. G.; Taylor, J. S.; Gross, M. L. Chem. Res. Toxicol. 2010,
23, 474–479.
(15) Zeng, Y.; Wang, Y. J. Am. Chem. Soc. 2004, 126, 6552–6553.
(16) Zhang, Q.; Wang, Y. J. Am. Chem. Soc. 2003, 125, 12795–12802.
(17) Crean, C.; Uvaydov, Y.; Geacintov, N. E.; Shafirovich, V.
Nucleic Acids Res. 2008, 36, 742–755.
(18) Liu, Z.; Gao, Y.; Wang, Y. Nucleic Acids Res. 2003, 31,
5413–5424.
(19) Liu, Z.; Gao, Y.; Zeng, Y.; Fang, F.; Chi, D.; Wang, Y.
Photochem. Photobiol. 2004, 80, 209–215.
(20) Maul, M. J.; Barends, T. R.; Glas, A. F.; Cryle, M. J.; Domratcheva,
T.; Schneider, S.; Schlichting, I.; Carell, T. Angew. Chem., Int. Ed. 2008,
47, 10076–10080.
(21) Douki, T.; Laporte, G.; Cadet, J. Nucleic Acids Res. 2003,
31, 3134–3142.
(22) Takamura-Enya, T.; Enomoto, S.; Wakabayashi, K. J. Org.
Chem. 2006, 71, 5599–5606.
(23) Gillet, L. C.; Sch€arer, O. D. Org. Lett. 2002, 4, 4205–4208.
(24) Lakshman, M. K. J. Organomet. Chem. 2002, 653, 234–251.
(25) Wang, Z.; Rizzo, C. J. Org. Lett. 2001, 3, 565–568.
(26) Hawkins, C. L.; Davies, M. J. Chem. Res. Toxicol. 2001,
14, 1071–1081.
(27) Hawkins, C. L.; Davies, M. J. Chem. Res. Toxicol. 2002,
15, 83–92.
(28) Naumov, S.; Barthel, A.; Reinhold, J.; Dietz, F.; Geimer, J.;
Beckert, D. Phys. Chem. Chem. Phys. 2000, 2, 4207–4211.
(29) Cadet, J.; Douki, T.; Ravanat, J. L. Nat. Chem. Biol. 2006,
2, 348–349.
(30) Hsu, G. W.; Ober, M.; Carell, T.; Beese, L. S. Nature 2004,
431, 217–221.
(31) Sch€arer, O. D. Angew. Chem., Int. Ed. 2003, 42, 2946–2974.
(32) Box, H. C.; Budzinski, E. E.; Dawidzik, J. B.; Wallace, J. C.;
Iijima, H. Radiat. Res. 1998, 149, 433–439.
(33) Cao, H.; Wang, Y. Nucleic Acids Res. 2007, 35, 4833–4844.
(34) M€unzel, M.; Lercher, L.; M€uller, M.; Carell, T. Nucleic Acids Res.
2010, 38, e192.
(35) Freese, E.; Bautz, E.; Freese, E. B. Proc. Natl. Acad. Sci. U.S.A.
1961, 47, 845–855.
(36) Gu, C.; Wang, Y. Biochemistry 2005, 44, 8883–8889.
(37) M€unzel, M.; Globisch, D.; Br€uckl, T.; Wagner, M.; Welzmiller,
V.; Michalakis, S.; M€uller, M.; Biel, M.; Carell, T. Angew. Chem., Int. Ed.
2010, 49, 5375–5377.
5189
dx.doi.org/10.1021/ja111304f |J. Am. Chem. Soc. 2011, 133, 5186–5189