The sugar conformation governs (6-4) photoproduct formation at the dinucleotide level

Article abstract of DOI:10.1021/ja077095q

The LNA dinucleotide mimic of TpT whose two-sugar puckers are locked in the C3′-endo conformation selectively produces the corresponding cyclobutane pyrimidine dimer under 254 nm irradiation. In the natural series (TpT) the sugar puckers are in a major C2

Full text of DOI:10.1021/ja077095q

Published on Web 12/11/2007
The Sugar Conformation Governs (6-4) Photoproduct Formation at the
Dinucleotide Level
Ce´line Desnous,† B. Ravindra Babu,‡ Ce´line Moriou,† Javier Ulises Ortiz Mayo,† Alain Favre,§
Jesper Wengel,‡ and Pascale Clivio*,†,#
Institut de Chimie des Substances Naturelles, CNRS, 1 AVenue de la Terrasse, 91190 Gif sur YVette, France,
Nucleic Acid Center, Department of Physics and Chemistry, UniVersity of Southern Denmark, 5230 Odense M,
Denmark, and Laboratoire de Photobiologie Mole´culaire, Institut Jacques Monod, CNRS, UniVersite´ Paris 6 et
Paris 7,
2 Place Jussieu, 75251 Paris cedex 05, France
Received September 13, 2007; E-mail: pascale.clivio@univ-reims.fr
Exposure of DNA to solar UV light induces structural modifica-
tions mostly in the form of covalent bond formation between two
adjacent pyrimidines. As a result, two major types of DNA
photoproducts (PPs), cyclobutane pyrimidine dimers (CPDs) and
pyrimidine (6-4) pyrimidone photoproducts ((6-4) PPs), are formed
(Figure 1).1 These DNA lesions are implicated in photoaging and
skin cancer and induce differential biological effects.2-4
In addition to the intrinsic photophysical properties of the targeted
nucleobases, the formation efficiency of photoproducts is also
known to depend upon the local structure of DNA which can either
promote or prevent the formation of a given PP.5 Consequently,
identification of the precise conformational parameters governing
Figure 2. Fractional amount of 1 and 2 and of their respective PPs as a
PP formation is central to understanding the DNA specific
photoreactivity and therefore its subsequent biological effects.
Thymidylyl(3′-5′)thymidine (TpT, 2) represents the shortest and
most photoreactive single-stranded DNA sequence able to mimic
DNA photoreactivity by giving rise to dipyrimidine PPs.6,7 In
addition, in TpT, the furanose moieties exist in a dynamic
equilibrium between C2′-endo (major) and C3′-endo (minor)
conformations as observed in B-DNA.8
function of irradiation time.
stacking ability and the quantum yield of formation of CPD and
(6-4) PPs.10,11 Therefore, finding a dinucleotide model whose sugar
conformation could selectively and efficiently drive the thymine
photoreactivity was a major challenge. We herein report the striking
photochemical properties of 1, the locked nucleic acid (LNA)
dinucleotide analogue of 212 in which the furanose moieties are
locked in the C3′-endo conformation.
The 254 nm photoreactivity of 1 was compared to that of 2 using
our recently published hplc-based method.11 After 2 min of
irradiation, 65% of 1 had reacted compared to only 15% of 2 (Figure
2). This clearly evidenced the dramatically increased photoreactivity
of 1, compared to 2, induced by the conformational locking of the
sugar moieties. Even more remarkably, the 254 nm irradiation of
1 afforded a single PP (3) (Scheme 1) in contrast to 2 in the limit
1
of H NMR and UV detection. Absence of UV absorption above
240 nm in the UV spectrum of 3 was indicative of a cyclobutane-
1
type structure. Examination of the H NMR spectrum of a crude
irradiation mixture of 1 confirmed the presence of the characteristic
features of a CPD structure: the absence of H6 signals and the
presence of two shielded methyl signals below 1.5 ppm (Figure
S9, Supporting Information).13 The cyclobutane nature of 3 is also
fully consistent with the steady-state reached after 5 min of
irradiation (Figure 2), CPD being known to photoreverse at 254
nm.6 After purification, CPD 3 was isolated in 53% yield together
with 19% of recovered 1. The stereochemistry of 3 was deduced
from NOEs to be cis-syn as in the case of the TpT-derived major
CPD (4) (Figure S13). The quantum yield of formation (Φ) of 3
was evaluated to be (6.8 ( 0.8) × 10-2 mol/einstein using a
previously described procedure11 which considers partial photo-
reversion. Compared to the TpT series,14 this reveals a formation
efficiency of CPD in the LNA series ca. 6 times higher.
Figure 1. Chemical structure of CPD and (6-4) PP at dithymine site in
DNA and in 2 (TpT).
The photochemistry of several TpT analogues has already been
studied to tentatively decipher the relationships linking the sugar-
phosphate backbone conformation to PP formation.9-11 Whereas
the yield of formation and the relative distribution of PPs have been
shown to be clearly influenced by the dinucleotide backbone
structure, only peptide nucleic acids (PNAs), DNA analogues with
no sugar and no phosphate residues, have demonstrated a unique
photoreactivity leading exclusively to CPD.9 Independently, promo-
tion of the population of the C3′-endo sugar conformers in TpT
analogues has been demonstrated to increase the intramolecular base
† Institut de Chimie des Substances Naturelles.
‡ Nucleic Acid Center.
To understand the unprecedented photoreactivity of 1, we
investigated its stacking properties by circular dichroism (Supporting
Information). Taking into account the monomer contribution, the
§ Laboratoire de Photobiologie Mole´culaire.
# Current address: FRE 2715 CNRS, Universite´ de Reims Champagne-Ardenne,
51 rue Cognacq Jay, 51 096 Reims cedex, France.
9
30
J. AM. CHEM. SOC. 2008, 130, 30-31
10.1021/ja077095q CCC: $40.75 © 2008 American Chemical Society

C O M M U N I C A T I O N S
series10a coincide, indicating a very close overall geometry of
stacked species, it can be concluded that (6-4) PP formation results
from minor stacked species.
Scheme 1
Our results clearly bring a new enlightenment to the field of
DNA photoproduct mechanistic studies. In addition, they provide
an easy and efficient access to a cis-syn CPD analogue for
biological studies after incorporation into oligonucleotides. This
study also points up the identical and unique photochemical
behavior of the bis thymine LNA and PNA dinucleotide analogues,
demonstrating their similar base-stacking patterns and therefore
similar preorganized conformation. This could bring new insights
in the comprehension of their unprecedented thermal stabilities
toward complementary DNA and RNA, which have allowed the
development of numerous applications in biotechnology and
therapeutics.17
shape of the CD spectrum of 1 at 20 °C is very similar to the one
of 2 with, however, a slight red-shift (2 nm) of λmax (281 nm) and
λmin (257 nm) and a roughly 3.7-fold increase in [θ] at λmax (Figure
3). Therefore, the overall stacking geometry appears similar in the
two dinucleotides with 1 exhibiting a roughly 4-fold propensity to
base stack than 2.15
Acknowledgment. CNRS and CONACYT are acknowledged
for a doctoral fellowship to C.D. and J.U.O.M., respectively. J.W.
and B.R.B. thank The Danish National Research Foundation and
EU-FP6 (CIDNA; proposal no. 505669-1) for funding.
The strong base-stacking propensity of 1, whose sugar moieties
exclusively exist as C3′-endo conformers substantiates our previous
statement that the percentage of C3′-endo conformers is one of the
major factors influencing the dinucleotide stacking ability.10a,11 In
addition, the absence of trans-syn CPD formation from 1 is in
line with strong base-stacking interactions.7,16
Supporting Information Available: Synthesis procedures, experi-
mental conditions for UV irradiation, HPLC, CD spectrum of 1 at
various temperatures, chromatogramms of the photolysis of an equimo-
lecular mixture of 1 and 2, 1H NMR spectrum of the photolysate of 1
and 1D and 2D NMR spectra of 3. This material is available free of
charge via the Internet at http://pubs.acs.org.
During the first 2 min of irradiation, compound 1 appears (5.7
( 1)-fold more photoreactive than standard TpT (using the quantum
yields ratio). This result is in line with the increased stacked
efficiency of 1, compared to 2, and further confirms the relationship
between stacking and overall photoreactivity.10,11 In the closely
related 2′-R-OMe series, the 5′-end and the 3′-end C3′-endo sugar
conformer population is 75% and 66%, respectively (versus 30%
and 37%, respectively, in 2).10a However in this series, if the base
stacking ability and photoreactivity are simultaneously increased,
compared to TpT, the increase in C3′-endo population does not
significantly alter PP distribution.10 In contrast, in the case of 1
where both sugar puckers are fully constrained in the conforma-
tionally rigid C3′-endo sugar pucker, (6-4) PP formation is impeded.
Consequently, it is likely that the C2′-endo sugar conformation is
necessary for (6-4) PP formation. Moreover, this study demonstrates
for the first time that in the dinucleotide series, CPD and (6-4) PP
result from distinct photoreactive stacked nucleobase patterns. In
addition, since the CD λmax and λmin in the LNA and the 2′-R-OMe
References
(1) Cadet, J.; Courdavault, S.; Ravanat, J.-L.; Douki, T. Pure Appl. Chem.
2005, 77, 947-961.
(2) Trautinger, F. Clin. Dermatol. 2001, 26, 573-577.
(3) (a) Ichihashi, M.; Ueda, M.; Budiyanto, A.; Bito, T.; Oka, M.; Fukunaga,
M.; Tsuru, K.; Horikawa, T. Toxicology 2003, 189, 21-39. (b) Matsumura,
Y.; Ananthaswamy, H. N. Front. Biosci. 2002, 7, d765-783.
(4) Lo, H.-L.; Nakajima, S.; Ma, L.; Walter, B.; Yasui, A.; Ethell, D. W.;
Owen, L. B. BMC Cancer 2005, 5, 135-142.
(5) Becker, M. M.; Wang, Z. J. Mol. Biol. 1989, 210, 429-438.
(6) Johns, H. E.; Pearson, M. L.; LeBlanc, J. C.; Helleiner, C. W. J. Mol.
Biol. 1964, 9, 503-524.
(7) Douki, T.; Court, M.; Sauvaigo, S.; Odin, F.; Cadet, J. J. Biol. Chem.
2000, 275, 11678-11685.
(8) Cheng, D. M.; Sarma, R. H. J. Am. Chem. Soc. 1977, 99, 7333-7348.
(9) Clivio, P.; Guillaume, D. Tetrahedron Lett. 1998, 39, 6881-6884.
(10) (a) Ostrowski, T.; Maurizot, J.-C.; Adeline, M.-T.; Fourrey, J.-L.; Clivio,
P. J. Org. Chem. 2003, 68, 6502-6510. (b) Santini, G. P. H.; Pakleza,
C.; Auffinger, P.; Moriou, C.; Favre, A.; Clivio, P.; Cognet, J. A. H. J.
Phys. Chem. B. 2007, 111, 9400-9409.
(11) Moriou, C.; Thomas, M.; Adeline, M.-T.; Martin, M.-T.; Chiaroni, A.;
Pochet, S.; Fourrey, J.-L.; Favre, A.; Clivio, P. J. Org. Chem. 2007, 72,
43-50.
(12) The thymine LNA dinucleotide 1 was synthesized as described in the
Supporting Information.
(13) Koning, T. M. G.; van Soest, J. J. G.; Kaptein, R. Eur. J. Biochem. 1991,
195, 29-40.
(14) Φ of 4 and 5 and the t,s I CPD produced from 2 is (1.1 ( 0.05) × 10-2
,
(0.1 ( 0.05) × 10-2, and (0.2 ( 0.05) × 10-2 mole/einstein, respectively.6
(15) Warshaw, M. M.; Cantor, C. R. Biopolymers 1970, 9, 1079-1103.
(16) Douki, T. J. Photochem. Photobiol., B 2006, 82, 45-52.
(17) (a) Elayadi, A. N.; Corey, D. R. Curr. Opin. InVestig. Drugs 2001, 2,
558-561. (b) Petersen, M.; Wengel, J. Trends Biotechnol. 2003, 21, 74-
81. (c) Demidov, V. V. Trends Biotechnol. 2003, 21, 4-7. (d) Vester,
B.; Wengel, J. Biochemistry 2004, 43, 13233-13241. (d) Karkare, S.;
Bhatnagar, D. Appl. Microbiol. Biotechnol. 2006, 71, 575-586.
Figure 3. CD difference spectra between 1 and its corresponding
mononucleoside thymidine LNA (TL) and between 2 and thymidine (T)
multiplied by a factor of 3.67 at 20 °C, in a buffer containing 0.01 M Na
phosphate, 0.1 M NaCl, pH 7.0.
JA077095Q
9
J. AM. CHEM. SOC. VOL. 130, NO. 1, 2008 31