Photorepair of cyclobutane pyrimidine dimers by 8-oxopurine nucleosides

Article abstract of DOI:10.1002/poc.2919

The 8-oxopurine nucleosides 2′,3′,5′-tri-O-acetyl-8-oxo- 7,8-dihydroguanosine (OG) and 2′,3′,5′-tri-O-acetyl- ribosyluric acid (RU) were studied for their ability to mediate the photochemical (λ>300nm) reversion of cyclobutane pyrimidine dimers to their parent pyrimidines thymine and uracil. The bimolecular reactions of these monomers proceeded at very slow rates compared with recently published work using oligonucleotide contexts; nevertheless, it was possible to make comparisons between the efficacy of OG and RU as photocatalysts as a function of pH. Although RU has a lower redox potential and anionic character, it was only equivalent to OG in facilitating thymine dimer photorepair over a broad pH range. Only OG showed pH-dependent behavior with higher activity at pH 8-9 where the base becomes deprotonated. Despite the overall low activity of OG and RU, the results are instructive with respect to a comparison of the two 8-oxopurines, and support the hypothesis that 8-oxopurine nucleosides may have played primordial roles as precursors to modern-day flavins in redox reactions of the RNA world. Copyright

Full text of DOI:10.1002/poc.2919

Research Article
Received: 14 November 2011,
Revised: 20 December 2011,
Accepted: 23 December 2011,
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/poc.2919
Photorepair of cyclobutane pyrimidine dimers
by 8-oxopurine nucleosides
Khiem Van Nguyena and Cynthia J. Burrowsa*
The 8-oxopurine nucleosides 2′,3′,5′-tri-O-acetyl-8-oxo-7,8-dihydroguanosine (OG) and 2′,3′,5′-tri-O-acetyl-ribosyluric
acid (RU) were studied for their ability to mediate the photochemical (l > 300 nm) reversion of cyclobutane
pyrimidine dimers to their parent pyrimidines thymine and uracil. The bimolecular reactions of these monomers
proceeded at very slow rates compared with recently published work using oligonucleotide contexts; nevertheless,
it was possible to make comparisons between the efficacy of OG and RU as photocatalysts as a function of pH.
Although RU has a lower redox potential and anionic character, it was only equivalent to OG in facilitating thymine
dimer photorepair over a broad pH range. Only OG showed pH-dependent behavior with higher activity at pH 8–9
where the base becomes deprotonated. Despite the overall low activity of OG and RU, the results are instructive with
respect to a comparison of the two 8-oxopurines, and support the hypothesis that 8-oxopurine nucleosides may have
played primordial roles as precursors to modern-day flavins in redox reactions of the RNA world. Copyright © 2012
John Wiley & Sons, Ltd.
Keywords: Deoxyribonucleic acid oxidation; photochemistry; prebiotic chemistry; Ribonucleic acid world; thymine dimers
deprotonate the base, or by selecting a different purine with
INTRODUCTION
a lower pKa, such as uric acid. Although the synthesis of the
ribonucleoside 9-ribosyluric acid (RU) is straightforward,[10] the
The RNA world hypothesis proposes that RNA predated DNA and
conversion of RU to a phosphoramidite and incorporation into
an oligomer are not. Therefore, we elected to study the behavior
of monomeric nucleosides toward photorepair of CPDs, realizing
that the efficiency of the intermolecular reaction would be much
less than that of an intrastrand or intraduplex reaction. In this
study, we report the feasibility of repairing the free base CPDs
T = T and U = U by 8-oxo-purine nucleosides OG and RU (Chart 1)
as a function of pH. With respect to prebiotic chemistry, we note
that the OG nucleoside is one step removed from the parent G
via hydroxylation at C8, while RU could be obtained in one addi-
tional step via hydrolytic deamination of the N2 amino group of
OG. Both of these chemical steps would be plausible in the early
Earth environment.
protein as both the informational material and as the set of
molecular catalysts necessary to promote prebiotic chemistry.[1,2]
Although ribozymes have been found to catalyze a wide range
of chemical transformations, examples thus far have required the
assistance of either metal ions or organic cofactors to facilitate
processes outside the realm of chemistry made available by
the normal components of RNA, the four canonical bases plus
the phosphoribose backbone.[3,4] For example, alcohol oxidation
and aldehyde reduction can be catalyzed by ribozymes in the
presence of NAD+/NADH,[5,6] but there are no examples of redox-
active ribozymes utilizing RNA oligomers alone. In recent work,[7]
we proposed that a simple derivative of guanosine (G), namely
8-oxo-7,8-dihydroguanosine (OG), may have been a primordial
redox cofactor for electron transfer processes. We demonstrated
that OG can mimic the function of the flavoenzyme photolyase
to repair cyclobutane pyrimidine dimers (CPD) in single-stranded
or double-stranded DNA or RNA oligonucleotides by what
appears to be a photoinduced electron transfer process. These
results support the hypothesis that a simple chemical transforma-
tion of a base, such as G to OG or C to 5-hydroxyC, could have
launched the evolution of the redox-active cofactors FADH2 or
NADH.[8]
EXPERIMENTAL
Materials
All chemicals were purchased from commercial sources and used
without further purification except where otherwise stated. The 2′,3′,5′-
tri-O-acetylnucleosides OG and RU were synthesized as previously de-
scribed.[10–12] [cis,syn]-Thymine dimer and uracil dimer were synthesized
according to published procedures and purified by reversed-phase high-
performance liquid chromatography (HPLC).[13–15] Nuclear magnetic
One finding of our previous work was that the distance
dependence of CPD photorepair by OG was very steep; the OG
base had to be well stacked in the duplex and within a few
nucleotides of the CPD for efficient repair.[7] A key aspect that
may be inhibiting repair by photoinduced electron transfer is
the separation of charge that occurs when a neutral base such
* Correspondence to: Cynthia J. Burrows, Department of Chemistry, University
of Utah, 315 S. 1400 East, Salt Lake City, UT 84112–0850, USA. E-mail:
burrows@chem.utah.edu
–
•
+
•
as OG donates an electron to form CPD and OG . We postu-
lated that electron transfer would be more facile if the nucleo-
base carried a negative charge in its resting state, as does FADH–
in photolyase.[9] This might be possible by raising the pH to
a K. V. Nguyen, C. J. Burrows
Department of Chemistry, University of Utah, 315 S. 1400 East, Salt Lake City,
UT 84112-0850, USA
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.

K. V. NGUYEN AND C. J. BURROWS
significant decomposition of nucleosides was observed by HPLC
during the course of the irradiation, indicating that no perma-
nent oxidation of OG or RU had occurred. The products of such
oxidation chemistry have been determined previously.[10,22]
We considered three possible mechanisms for the repair pro-
cess: (1) excited state energy transfer, (2) oxidative electron
transfer from T = T to OG, and (3) reductive electron transfer from
OG to T = T. The repair of thymine dimer by direct energy trans-
fer from the photoexcited state of OG and RU (mechanism (1)) is
excluded because the excited singlet state energy of T = T is
much higher than the energy of ~310 nm light used in our
experiments.[23] Although thymine dimer repair may proceed
via an oxidative mechanism, particularly in the presence of redox
active transition metal complexes,[24–27] the enhancement of the
thymine dimer repair yield by OG and RU is more likely occurring
via the reductive mechanism as proposed in our previous
study.[7] In this mechanism, T = T accepts one electron from a
photoexcited purine, followed by rapid bond cleavage of the
strained cyclobutane ring (Fig. 2), and finally back electron trans-
fer to the purine. The reductive electron transfer mechanism is
supported by the significantly lower redox potential of OG com-
pared with G or other bases and by the observed preference for
long-range electron transfer in the 3′–5′ direction in studies of
duplex DNA.[7] One can estimate the Gibbs free energy (ΔGet)
for the electron transfer from the photoexcited OG to T = T at
about À1.30 eV or À125 kJ/mol from a previously reported
method.[23] This reductive electron transfer mechanism has also
been well characterized for photolyase and for related model
systems.[9,26,28–37]
Mechanistic considerations aside, the absolute repair yield of
thymine dimer in these systems is extremely low, only about
1% after 7 h of irradiation. As anticipated, the bimolecular reac-
tion is much less efficient than the case where OG was installed
in proximity to the CPD in double-stranded oligonucleotides.[7]
Thus, the ‘immobilization’ of two species, thymine dimer and 8-
oxopurine, in an oligomer is important to enhance the repair
efficiency.
Nevertheless, the repair yields observed in the initial experi-
ments were sufficiently above background to warrant further
study, and we next turned our attention to the uric acid nucleo-
side. We observed that RU is about equally efficient as OG in
repairing thymine dimer at pH 7 (Fig. 1). This observation was
surprising for two reasons: (1) RU has a slightly lower redox po-
tential, about 200 mV below that of OG,[38,39] and it would there-
fore be a more powerful electron donor, all other factors being
equal, and (2) RU exists as an anion at pH 7, which should be a
favorable condition for excited state electron donation.[40] Thus,
the finding that RU is no more effective than OG at intermolecu-
lar photorepair was unanticipated. This observation led us to in-
vestigate the dependence of thymine dimer repair yield on pH
for both RU and OG.
Chart 1. Catalysts and substrates (R = 2′,3′,5′-tri-O-acetylribofuranosyl)
resonance spectra, UV spectra and/or HPLC retention times of synthe-
sized compounds were identical with literature data. Concentrations of
nucleosides and cyclobutane pyrimidine dimer were determined by UV
spectrophotometry using reported extinction coefficients.[16–18]
Photorepair of cyclobutane pyrimidine dimer
A solution containing 0.2 mM CPD (T = T or U = U) and 0.2 mM nucleoside
(OG or RU) in 25 mM NaPi buffer adjusted to the appropriate pH was irra-
diated in a polystyrene cuvette to cut off wavelengths below 300 nm[19]
at ambient temperature (22 ꢀC) with an FS40 UVB lamp (lmax = 313 nm,
Home Phototherapy, OH, USA). In the case of thymine dimer, the irradia-
tion mixture was analyzed by HPLC on a Varian C18 column (5 mm,
250Â4.6 mm) using a linear gradient of 2% to 10% B over 20 min, then
increasing B to 65% over 10 min (A: 0.1%TFA in H2O; B: MeOH). A slightly
different HPLC method with isocratic B at 1% for 15 min and then in-
creasing to 65% B over 10 min was used to analyze the irradiation mix-
ture of uracil dimer repair (A: H2O; B: MeOH). In both cases, the flow rate
was 0.7 mL/min, and the detector was set at 260 nm. The peaks
corresponding to CPD and the repaired product were integrated and
normalized against extinction coefficients to calculate the repair yield,
averaged over three or more independent trials. Because the extinction
coefficient of thymine dimer at 260 nm is not yet reported, this value
was estimated at about 2.8 Â 102 by the formula: e260 = (A260/A220) Â
e220
.
RESULTS AND DISCUSSION
In the present experiments, the 2′, 3′, and 5′ hydroxyl groups of
[10–12]
nucleosides were acetylated
to retard their elution times
compared with CPDs and free base pyrimidines as monitored
by reversed-phase HPLC. Irradiation was carried out in polysty-
rene cuvettes to cut off wavelengths below 300 nm, thus direct-
ing light into the 8-oxopurine chromophore.[7] Both OG and RU
have absorption maxima near 295 nm and absorb significantly
above 300 nm, where normal bases and pyrimidine dimers do
not. The repair of thymine dimer was first investigated at pH 7,
where we found that the presence of an equimolar amount of
OG or RU increased the yield of thymine dimer repair approxi-
mately fourfold compared with background repair, while the
parent nucleoside G did not have any effect (Fig. 1). Although ad-
jacent guanine bases in DNA have been shown to have an effect
on the formation of T = T at shorter wavelengths, the inactivity of
G at l > 300 nm is not surprising given that G displays essentially
no absorbance above 290 nm.[20,21] In addition, we note that no
The background photorepair rate of T = T (no purines present)
is almost independent of pH in the range from 5 to 9 (Fig. 3). It is
also not surprising that the repair yield only slightly increased
from pH 7 to pH 9 in the presence of RU, because with pKa ~ 6,[40]
RU exists in an anionic form across this pH range. However, we
also did not observe a significant increase of thymine dimer re-
pair yield by RU in the pH range from 5 to 7, for which RU is sup-
posed to change from a neutral to an anionic form. This could
indicate that despite a favorable redox potential, the excited state
lifetimes of RU and RUˉ are both too short to permit effective in-
termolecular electron transfer. In contrast to RU, the activity of
wileyonlinelibrary.com/journal/poc
Copyright © 2012 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. (2012)

PHOTOREPAIR OF PYRIMIDINE DIMERS BY 8-OXOPURINES
Figure 1. (A) HPLC traces analysis of T = T versus repaired T after 5 h irradiation at pH 7; 2′,3′,5′-tri-O-acetylnucleosides eluted 20 min later. (B) Plot of
T = T repair yield as a function of irradiation time
photoexcited state lifetimes were thought responsible for this
phenomenon.[15,34] In studies of duplex versus monomer sys-
tems, base stacking and base pairing are important factors in
determining the excited state lifetimes, and are therefore
expected to play significant roles in the ability of base chromo-
phores to undergo collisional charge transfer with CPDs.[42–44]
However, there is presently no information about the excited
state lifetimes of OG or RU. Clearly, additional photophysical
studies will be necessary to fully understand the complex factors
contributing to the lower efficiency of RU compared with OG in
repairing thymine dimer at high pH.
Repair of uracil dimer, a common type of CPD found in RNA,
by 8-oxopurine nucleosides was also investigated. A previous
Figure 2. Proposed mechanism for the enhancement of CPD repair by
study demonstrated that repair of thymine dimer was faster than
uracil dimer when mediated by the enzyme photolyase because
of a strain effect induced by the two cis methyl groups.[45] How-
ever, a more recent report produced an opposite result in a
model system in which uracil dimer was repaired faster than
thymine dimer in flavin-conjugated model compounds, and an
explanation was based on stereoelectronic effects.[46] In a previ-
ous work,[7] our laboratory observed slower repair for a CPD in an
A-form helix (U = U in RNA/DNA or RNA/RNA duplexes) com-
pared with a CPD in a B-form helix (T = T in a DNA/DNA duplex)
by OG. We attributed this to the difference in base stacking
between the two helical forms; an adjacent OG-containing base
pair stacks poorly on a U = U lesion in A-form RNA. Here, by using
a nucleoside repair model, we are able to compare the repair
efficiency of OG between these two types of CPD in an identical
environment, absent base stacking.
8-oxopurine nucleosides
Figure 3. Yield of thymine dimer repair after 5 h as a function of pH for
various additives; red: OG, violet: RU, blue: none
At pH 7, we observed the repair of U = U increased by three-
fold in the presence of an equimolar amount of OG (Table 1).
In addition, the repair yield was not significantly different with
that of thymine dimer. Unexpectedly, we found the background
OG dramatically increased from pH 7 to pH 9 (Fig. 3), which is in
good agreement with its pKa of 8.6.[41] At pH 9, the repair yield of
T = T by OG reached to about 8%, which is a 30-fold increase
compared with background repair.
Table 1. T = T and U = U repair yields in the presence of OG
as a function of pH. Errors are estimated at Æ10%
The instability of 8-oxopurine nucleosides at higher pH pre-
cluded investigation of repair at pH > 9. Despite that limitation,
the pH dependence observed further supports a reductive
mechanism for thymine dimer repair by OG and agrees with
previous work showing that flavins in an anionic form repair
thymine dimer more efficiently.[9,33] The difference in activity
between OG and RU may also imply that the purine redox poten-
tial is not the only factor determining the thymine dimer repair
efficiency. Indeed, previous studies showed that not all flavin
derivatives are able to repair thymine dimer, and their different
Entry
Substrate
Irradiation time (h)
Yield (%)
pH 7
pH 9
1
2
3
4
T = T
T = T, OG
U = U
5
5
5
5
0.2
0.7
0.3
1
0.2
8
5
U = U, OG
12
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

K. V. NGUYEN AND C. J. BURROWS
[8] H. Yanagawa, Y. Ogawa, M. Ueno, J. Biol. Chem. 1992, 267,
13320–13326.
[9] R. F. Hartman, S. D. Rose, J. Am. Chem. Soc. 1992, 114, 3559–3560.
[10] K. V. Nguyen, J. G. Muller, C. J. Burrows, Tetrahedron Lett. 2011, 52,
2176–2180.
repair of U = U was somewhat pH dependent compared with
T = T, and it reached to 5% at pH 9 (Table 1, entry 3). We cannot
explain this unusual observation at the present time. In any case,
the presence of OG still increased the repair yield for U = U about
twofold. From these data, it is reasonable to conclude that OG
enhances the intrinsic repair yield of uracil dimer with similar
efficiency to that of thymine dimer in the absence of a specific
helical environment.
[11] C. Grünewald, T. Kwon, N. Piton, U. Förster, J. Wachtveitl, J. W.
Engels, Bioorg. Med. Chem. 2008, 16, 19–26.
[12] J. C. Niles, J. S. Wishnok, S. R. Tannenbaum, Chem. Res. Toxicol. 2000,
13, 390–396.
[13] A. J. Varghese, Biochemistry 1971, 10, 4283–4290.
[14] D. L. Wulff, G. Fraenkel, Biochim. Biophys. Acta 1961, 51, 332–339.
[15] S. E. Rokita, C. T. Walsh, J. Am. Chem. Soc. 1984, 106, 4589–4595.
[16] M. A. Herbert, J. C. Leblanc, D. Weinblum, H. E. Johns, Photochem.
Photobiol. 1969, 9, 33–43.
CONCLUSIONS
In conclusion, we showed that the 8-oxopurine nucleosides OG
and RU are able to mediate the repair of cyclobutane pyrimidine
dimers T = T and U = U, even in a bimolecular reaction. The
observations are in accord with the mechanism proposed in
our previous study[7] in which CPD is repaired reductively by
accepting one electron from the photoexcited state of the
purine. Unexpectedly, deamination of the OG base to form RU
does not provide a more active photocatalyst, despite the lower
redox potential of RU and its anionic nature, suggesting that
other factors, such as the excited state lifetimes of purines and
flavin mimics, may play equally important roles in the process.
In contrast, OG did display the expected pH-dependent behav-
ior, and the photorepair of CPDs was enhanced as the pH
approached the pKa value of 8.6. These studies also showed that
T = T and U = U underwent repair at similar rates in the absence
of a helical environment, although the overall levels are suffi-
ciently low as to not provide an accurate assessment. Neverthe-
less, the results are instructive with respect to a comparison of
OG and RU, and support the hypothesis that 8-oxopurine nucleo-
sides may have played primordial roles as precursors to modern-
day flavins in redox reactions of the RNA world.
[17] R. Trivedi, L. Rebar, K. Desai, L. Stong, Clin. Chem. 1978, 24, 562–566.
[18] P. G. Slade, N. D. Priestley, K. D. Sugden, Org. Lett. 2007, 9, 4411–4414.
[19] D. J.-F. Chinnapen, D. Sen, Proc. Natl. Acad. Sci. 2004, 101, 65–69.
[20] M. R. Holman, T. Ito, S. E. Rokita, J. Am. Chem. Soc. 2006, 129, 6–7.
[21] Z. Pan, M. Hariharan, J. D. Arkin, A. S. Jalilov, M. McCullagh, G. C.
Schatz, F. D. Lewis, J. Am. Chem. Soc. 2011.
[22] W. Luo, J. G. Muller, E. M. Rachlin, C. J. Burrows, Org. Lett. 2000, 2,
613–617.
[23] D. J. F. Chinnapen, D. Sen, J. Mol. Biol. 2007, 365, 1326–1336.
[24] O. Kruger, U. Wille, Org. Lett. 2001, 3, 1455–1458.
[25] A. A. Voityuk, M.-E. Michel-Beyerle, N. Roesch, J. Am. Chem. Soc.
1996, 118, 9750–9758.
[26] F. Boussicault, O. Kruger, M. Robert, U. Wille, Org. Biomolec. Chem.
2004, 2, 2742–2750.
[27] P. J. Dandliker, R. E. Holmlin, J. K. Barton, Science 1997, 275,
1465–1468.
[28] S. Breeger, U. Hennecke, T. Carell, J. Am. Chem. Soc. 2004, 126,
1302–1303.
[29] C. Chatgilialoglu, M. Guerra, P. Kaloudis, C. Houée-Lévin, J. L. Marignier,
V. Swaminathan, T. Carell, Chem. Eur. J. 2007, 13, 8979–8984.
[30] D. Fazio, C. Trindler, K. Heil, C. Chatgilialoglu, T. Carell, Chem. Eur. J.
2011, 17, 206–212.
[31] K. Heil, D. Pearson, T. Carell, Chem. Soc. Rev. 2011.
[32] A. Schwögler, L. Burgdorf, T. Carell, Angew. Chem. Int. Ed. 2000, 39,
3918–3920.
[33] R. Epple, E.-U. Wallenborn, T. Carell, J. Am. Chem. Soc. 1997, 119,
7440–7451.
[34] M. S. Jorns, J. Am. Chem. Soc. 1987, 109, 3133–3136.
[35] W.-J. Tang, Q.-H. Song, H.-B. Wang, J.-Y. Yu, Q.-X. Guo, Org. Biomol.
Chem. 2006, 4, 2575–2580.
[36] O. Wiest, C. B. Harrison, N. J. Saettel, R. Cibulka, M. Sax, B. König, J.
Org. Chem. 2004, 69, 8183–8185.
[37] S. R. Yeh, D. E. Falvey, J. Am. Chem. Soc. 1991, 113, 8557–8558.
[38] M. G. Simic, S. V. Jovanovic, J. Am. Chem. Soc. 1989, 111, 5778–5782.
[39] S. Steenken, S. V. Jovanovic, M. Bietti, K. Bernhard, J. Am. Chem. Soc.
2000, 122, 2373–2374.
Acknowledgements
We thank Prof. Jack Simons (University of Utah) for helpful
discussions and the US National Science Foundation (CHE-
0809483) for support of this work.
REFERENCES
[1] W. Gilbert, Nature 1986, 319, 618–618.
[40] H. S. Forrest, D. Hatfield, J. M. Lagowski, J. Chem. Soc. 1961, 963–968.
[41] S. J. Culp, B. P. Cho, F. F. Kadlubar, F. E. Evans, Chem. Res. Toxicol.
1989, 2, 416–422.
[2] T. R. Cech, in RNA Worlds: From Life’s Origins to Diversity in Gene
Regulation (Eds: J. F. Atkins, R. F. Gesteland, T. R. Cech), Cold Spring
Harbor Press, Cold Spring Harbor, NY, 2011.
[42] C. E. Crespo-Hernandez, B. Cohen, B. Kohler, Nature 2005, 436,
[3] X. Chen, N. Li, A. D. Ellington, Chem. Biodiversity 2007, 4, 633–655.
[4] J. C. Cochrane, S. A. Strobel, RNA 2008, 14, 993–1002.
[5] S. Tsukiji, S. B. Pattnaik, H. Suga, Nature Struct. Biol. 2003, 10,
713–717.
1141–1144.
[43] C. E. Crespo-Hernandez, K. de La Harpe, B. Kohler J. Am. Chem. Soc.
2008, 130, 10844–10845.
[44] A. Kumar, M. D. Sevilla, Chem. Rev. 2010, 110, 7002–7023.
[45] S. T. Kim, A. Sancar, Biochemistry 1991, 30, 8623–8630.
[46] J. Butenandt, R. Epple, E.-U. Wallenborn, A. P. M. Eker, V. Gramlich,
T. Carell, Chem. Eur. J. 2000, 6, 62–72.
[6] S. Tsukiji, S. B. Pattnaik, H. Suga, J. Am. Chem. Soc. 2004, 126,
5044–5045.
[7] K. V. Nguyen, C. J. Burrows, J. Am. Chem. Soc. 2011, 133, 14586–14589.
wileyonlinelibrary.com/journal/poc
Copyright © 2012 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. (2012)