DNA-protein cross-linking: Model systems for pyrimidine-aromatic amino acid cross-linking

Article abstract of DOI:10.1021/ol052876m

We have synthesized simple model systems to explore the possibility of photo-cross-linking between the pyrimidine bases and the side chains of the aromatic amino acids. Thymine/phenylalanine and thymine/tyrosine models gave cross-links, and thymine/tryptophan models gave complex mixtures; the cytosine/phenylalanine model was unreactive. The quantum yields for the model cross-linking reactions were 18-46 times smaller than those for thymine dimer formation. Biphotonic excitation contributes little to the yield of these reactions.

Full text of DOI:10.1021/ol052876m

ORGANIC
LETTERS
2006
Vol. 8, No. 4
681-683
DNA−Protein Cross-Linking: Model
Systems for Pyrimidine−Aromatic Amino
Acid Cross-Linking
Guangxing Sun,‡ Christopher J. Fecko,§ Robert B. Nicewonger,‡
Watt W. Webb,*,§ and Tadhg P. Begley*,‡
Department of Chemistry and Chemical Biology, School of Applied & Engineering
Physics, Cornell UniVersity, Ithaca, New York 14853
www2@cornell.edu; tpb2@cornell.edu
Received November 28, 2005
ABSTRACT
We have synthesized simple model systems to explore the possibility of photo-cross-linking between the pyrimidine bases and the side
chains of the aromatic amino acids. Thymine/phenylalanine and thymine/tyrosine models gave cross-links, and thymine/tryptophan models
gave complex mixtures; the cytosine/phenylalanine model was unreactive. The quantum yields for the model cross-linking reactions were
18−46 times smaller than those for thymine dimer formation. Biphotonic excitation contributes little to the yield of these reactions.
UV irradiation of protein-nucleic acid complexes results in
the cross-linking of the protein to the nucleic acid. The use
of this reaction to study protein nucleic acid interactions has
been limited because it is very inefficient and is accompanied
by extensive protein and nucleic acid damage.1 The resulting
complicated reaction mixtures also make it difficult to iden-
tify specific amino acid/base cross-links; Ser-T, Cys-T, Lys-
T, and Tyr-T are the only cross-links so far characterized.2-4
Essentially, nothing is known about the cross-linking mech-
anisms. Here, we describe simple model systems to explore
the cross-linking of tyrosine, phenylalanine, and tryptophan
to the pyrimidine bases.
side chains, using a short linker to mimic the proximity and
orientation in DNA-protein complexes. We first synthesized
the uracil-tyrosine model 4. Irradiation of 4 resulted in the
formation of the cyclization product 5. This reaction was
much more efficient than the corresponding intermolecular
reaction of uracil 1 and cresol 2 (Scheme 1) and is analogous
Scheme 1
Our model-building strategy involved attaching the pyri-
midine base to the tyrosine, phenylalanine, and tryptophan
‡ Department of Chemistry and Chemical Biology.
§ School of Applied & Engineering Physics.
(1) Meisenheimer, K. M.; Koch, T. H. Crit. ReV. Biochem. Mol. Biol.
1997, 32 (2), 101-140.
(2) Wang, W. Y. Photochemistry and Photobiology of the Nucleic Acids;
Wiley: New York, 1976.
(3) Morrison, H. Bioorganic Photochemistry; Wiley: New York, 1990.
(4) Shetlar, M. D. Frontiers of Photobiology; International Congress
Series, 1993, 1021, 67-72 and references therein.
to the previously reported tyrosine/thymine photolysis which
yielded a similar adduct but also in low yield.5
10.1021/ol052876m CCC: $33.50
© 2006 American Chemical Society
Published on Web 01/19/2006

6-Benzyluracil 6 and 5-benzyluracil 8 were next synthe-
sized to explore thymine/phenylalanine cross-linking. Ir-
radiation of 6 and 8 resulted in the corresponding cyclization
products 7 and 9 (Scheme 2). The structures of 7 and 9 were
To probe the photoreactivity of tryptophan with thymine,
compounds 12-15 were synthesized (Scheme 3). Photolysis
Scheme 3
Scheme 2
of these compounds under anaerobic conditions gave com-
plex reaction mixtures, and none of the expected photocy-
clization products were detected. This was surprising:
because indoles are better reducing agents than benzene, we
anticipated that the trapping of the short-lived pyrimidine
excited state would have been more efficient with those
systems.
The cytosine-phenylalanine model 16 was also studied.
This compound is unreactive under the photolysis conditions
used for the photocyclization of 6 and 8.
1
fully characterized by H NMR, 13C NMR, and MS. These
adducts suggest possible structures for the cross-linking of
phenylalanine to DNA, a reaction reported twenty years ago
for which the adduct structure was not determined.6 Struc-
tures 7 and 9 revealed that both carbons 5 and 6 of the
pyrimidine can participate in cross-link formation. Com-
pounds 10 and 11 failed to undergo photocyclization,
suggesting that the orientation of the base and the amino
acid side chain is critical for cross-linking. There was no
change in the quantum yield for the conversion of 8 to 9 in
the presence and absence of oxygen (Table 1), suggesting
Scheme 4
Table 1. Quantum Yields (QY) for the Photoreactions of
Compounds 5, 7, and 9 and Thymine Photodimer (T)T)
entry
products
conditions
QY × 103
1
2
3
4
5
6
7
8
9
5
degassed
degassed
aerated
aerated
O2 saturated
aerated
aerated
aerated
aerated
1.19
1.60
1.60
2.63 (laser)a
1.58
0.79
1.02
2.24 (laser)a
46.7b
9a
9a
9a
9a
9b
7
7
T)T
We propose the mechanism outlined in Scheme 4 for the
photocyclization of 6. Electron transfer from the phenyl
group to the photoexcited pyrimidine would give the radical
ion pair 17. Protonation of the pyrimidine radical anion at
C6 would give 19. Analogous pyrimidine radical anion
protonation reactions have been previously described.7,8
a Laser peak intensity ∼4 × 109 W/cm2; CW intensity ∼4 × 10-3 W/cm2.
See text for details. b See the last paragraph.
that the cross-linking reaction involves singlet state chem-
istry.
(5) Shaw, A. A.; Falick, A. M.; Shetlar, M. D. Biochemistry 1992, 31,
10976-10983.
(6) Merill, B. M.; Williams, K. R.; Chase, J. W.; Konigsberg, W. H. J.
Biol. Chem. 1984, 259 (17), 10850-10856
(7) Cullis, P. M.; Evans, P.; Malone, M. E. Chem. Commun. (Cambridge,
U.K.) 1996, 8, 985-986.
(8) Deeble, D. J.; Das, S.; Von Sonntag, C. J. Phys. Chem. 1985, 89,
5784-5788.
682
Org. Lett., Vol. 8, No. 4, 2006

Radical coupling would give 21, and rearomatization by
deprotonation would complete the reaction. This proposal
is supported by the observation of deuterium incorporation
from solvent at C6 of the dihydropyrimidine. A similar
mechanistic analysis can be used to explain the formation
of 5 and 9.
It has been reported that irradiation with a high-intensity
UV laser greatly increases the efficiency of protein-nucleic
acid cross-linking because of a biphotonic excitation mech-
anism.9 However, it is difficult to investigate cross-linking
of proteins to nucleic acids in the presence of competing
protein and nucleic acid degradation. Our models provide
an ideal system in which to evaluate the importance of a
biphotonic excitation mechanism because of their cleaner
photochemistry. Using the frequency-tripled output of a Ti:
sapphire regenerative amplifier (265 nm, 150 fs pulselength),
we measured the intensity-dependent yields for the photo-
reactions presented in Scheme 2 (Figure 1). For peak UV
intensities in the range of ∼107-1010 W/cm2, fits of the data
to a power law relationship (yield ) a × intensityb, where
a and b are constants) suggest that the photochemical
conversion of 8 to 9 is predominantly monophotonic (b )
1.03) whereas conversion of 6 to 7 has only a small
biphotonic component (b ) 1.16). The quantum yield for
each reaction at the lowest laser intensities (see Supporting
Information) is close to that measured with continuous-wave
(CW) irradiation, ruling out the possibility that the intensity
dependence is artificially low because of saturation effects.10
NMR analysis was used to confirm that the products of laser
irradiation are the same as those of CW irradiation. At best,
the laser quantum yields are enhanced by only a factor of
2-3 over the CW values (Table 1) despite the difference of
∼1012 W/cm2 in peak intensity, further suggesting that
biphotonic excitation is relatively unimportant for these
photoreactions within the intensity range investigated.
In conclusion, we have synthesized simple model systems
to explore the possibility of photo-cross-linking between the
Figure 1. Product yield (expressed as a fraction of the starting
material) per laser pulse for reactions in Scheme 2 as a function of
UV intensity. Solid lines are the result of fits to a power law
relationship, as described in the text.
pyrimidine bases and the side chains of the aromatic amino
acids. Thymine/phenylalanine and thymine/tyrosine models
gave cross-links and thymine/tryptophan models gave com-
plex mixtures; the cytosine/phenylalanine model was unre-
active. The quantum yields for the model cross-linking
reactions were 18-46 times smaller than the quantum yield
for thymine dimer formation, suggesting that UV damage
to DNA will be a serious problem in the analysis of cross-
linked protein-DNA complexes. We find that biphotonic
excitation contributes little to the yield of these reactions,
even when using high-intensity irradiation.
Acknowledgment. This work was supported by NSF
Grant number CHE-0242328.
Supporting Information Available: Experimental details
for the preparation and characterization of compounds 3-16,
for the photoreactions, and for the quantum yield determina-
tions are provided. This material is available free of charge
via the Internet at http://pubs.acs.org.
(9) Russmann, C.; Stollhof, J.; Weiss, C.; Beigang, R.; Beato, M. Nucl.
Acids Res. 1998, 26, 3967-3970 and references therein.
(10) Nikogosyan, D. N. Int. J. Radiat. Biol. 1990, 57 (2), 233-299.
OL052876M
Org. Lett., Vol. 8, No. 4, 2006
683