Ring opening photoreactions of cytosine and uracil with ethylamine

Article abstract of DOI:10.1562/0031-8655(2000)071<0243:ROPOCA>2.0.CO;2

The photochemical reactions of cytosine (Cyt) and uracil (Ura) with ethylamine, an analog of the side chain of the amino acid lysine, have been studied. After irradiation of Cyt in aqueous ethylamine at λ = 254 nm, N-(N′-ethylcarbamoyl)-3-aminoacrylamidine (Ia) and N-(N′-ethylcarbamoyl)-3-ethylaminoacrylamidine (Ib) were isolated as products, while irradiation of Ura gave N-(N′-ethylcarbamoyl)-3-aminoacrylamide (IIa) and N-(N′-ethylcarbamoyl)-3-ethylaminoacrylamide (IIb) as products. Studies in which Ia and IIa were incubated with ethylamine at various pH values indicate that Ib and IIb are secondary products produced via thermal reactions of Ia and IIa with ethylamine. Heating of Ia and Ib leads to ring closure with the resultant formation of 1-ethylcytosine; small amounts of 1-ethyluracil are also produced. Heating of IIa and IIb produces 1-ethyluracil as the sole product. Spectroscopic properties were determined for each of these opened ring products, as well as for N-(N′-ethylcarbamoyl)-3-amino-2-methylacrylamidine (III) and N- (N′-ethylcarbamoyl)-3-amino-2-methylacrylamide (IV). Quantum yield measurements showed that Ia was formed with a Φ of 1.6 × 10^-4 at pH 9.8, while Φ for formation of IIa was 7.2 × 10^-4 at pH 11.5. A profile of the relative quantum yield for formation of Ia, determined as a function of pH, showed that the maximum quantum yield occurs at around pH 9.5; the analogous profile for IIa shows a maximum quantum yield at pH 11.3 and above. Acetone sensitization does not produce Ia in the Cyt-ethylamine system, which indicates that the known triplet state of Cyt is not involved in reactions leading to this opened ring product.

Full text of DOI:10.1562/0031-8655(2000)071<0243:ROPOCA>2.0.CO;2

Ring Opening Photoreactions of Cytosine and Uracil with Ethylamine
Author(s): Kellie Hom, Gary Strahan, Martin D. Shetlar
Source: Photochemistry and Photobiology, 71(3):243-253.
Published By: American Society for Photobiology
https://doi.org/10.1562/0031-8655(2000)071<0243:ROPOCA>2.0.CO;2
URL: http://www.bioone.org/doi/full/10.1562/0031-8655%282000%29071%3C0243%3AROPOCA
%3E2.0.CO%3B2
BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and
environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published
by nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries
or rights and permissions requests should be directed to the individual publisher as copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research
libraries, and research funders in the common goal of maximizing access to critical research.

Photochemistry and Photobiology, 2000, 71(3): 243–253
Ring Opening Photoreactions of Cytosine and Uracil with Ethylamine
Kellie Hom, Gary Strahan and Martin D. Shetlar*
Department of Pharmaceutical Chemistry, School of Pharmacy, University of California, San Francisco, CA, USA
Received 17 August 1999; accepted 24 November 1999
ABSTRACT
chemical lesions. Although cytosine (Cyt)† photohydrates,
pyrimidine cyclobutane dimers and pyrimidine(6–4)pyrimi-
done photoproducts (2) are the most often studied chemical
lesions in UV-irradiated DNA, other types of damage may
also be involved in causing the deleterious effects of UV
irradiation on cells. For example, DNA–protein cross-linking
may also play a role in UV-induced cellular damage (3–5).
A variety of proteins cross-link to DNA or polynucleo-
tides upon irradiation by UV light; these include, for ex-
ample, histones (6–13), Escherichia coli DNA polymerase I
The photochemical reactions of cytosine (Cyt) and uracil
(Ura) with ethylamine, an analog of the side chain of the
amino acid lysine, have been studied. After irradiation of
Cyt in aqueous ethylamine at ␭ ؍ 254 nm, N-(N
carbamoyl)-3-aminoacrylamidine (Ia) and N-(N
carbamoyl)-3-ethylaminoacrylamidine (Ib) were isolated
as products, while irradiation of Ura gave N-(N -ethyl-
carbamoyl)-3-aminoacrylamide (IIa) and N-(N -ethylcar-
bamoyl)-3-ethylaminoacrylamide (IIb) as products. Stud-
ies in which Ia and IIa were incubated with ethylamine
at various pH values indicate that Ib and IIb are second-
ary products produced via thermal reactions of Ia and
IIa with ethylamine. Heating of Ia and Ib leads to ring
closure with the resultant formation of 1-ethylcytosine;
small amounts of 1-ethyluracil are also produced. Heat-
ing of IIa and IIb produces 1-ethyluracil as the sole prod-
uct. Spectroscopic properties were determined for each
؅
-ethyl-
-ethyl-
؅
؅
؅
(
14), E. coli RNA polymerase (15–18), fd phage gene-5 pro-
tein (19), single-stranded DNA binding protein from E. coli
20), four of the component proteins of the DNA replication
(
complex of T4 phage (21), A1 heterogeneous nuclear ribo-
nucleoprotein (22), E. coli uracil (Ura)–DNA glycosylase
(23) and rat DNA polymerase
␤ (24, 25).
The potential complexity of irradiated DNA–protein sys-
tems is a major challenge to investigators who try to identify
the amino acid and nucleobase partners in UV-induced
DNA–protein cross-links. Shetlar and coworkers (26–28)
measured the apparent quantum yields for 19 amino acids
and selected di- and tripeptides toward photoaddition to
DNA and homopolynucleotides of the major DNA and RNA
bases. None of the amino acids, when incorporated into a
glycyl dipeptide, can be regarded as unreactive in photoad-
dition to the macromolecules studied in the survey. These
results suggest that any amino acid residue or, perhaps, pep-
tide bond in a protein can in principle form a UV-induced
cross-link with any nucleotide residue in a nucleic acid.
Lysine has been found to be one of the most reactive
amino acids toward photoaddition reactions with nucleobas-
es, nucleosides, DNA and polynucleotides (26–33). It occurs
in relatively high abundance in the histones, an important
class of structural proteins in eukaryotic chromatin. For ex-
ample, it comprises about 11% of the amino acid content in
the histones making up calf thymus nucleosomes (34). In
studies of photocross-linking in DNA–histone complexes,
Saito and coworkers (33) found that lysine residues in his-
tones are photoreactive with thymine (Thy) contained in
of these opened ring products, as well as for N-(N
carbamoyl)-3-amino-2-methylacrylamidine (III) and N-
-ethylcarbamoyl)-3-amino-2-methylacrylamide (IV).
Quantum yield measurements showed that Ia was
؅-ethyl-
(N؅
Ϫ
4
formed with a
formation of IIa was 7.2
the relative quantum yield for formation of Ia, deter-
mined as a function of pH, showed that the maximum
quantum yield occurs at around pH 9.5; the analogous
profile for IIa shows a maximum quantum yield at pH
⌽
of 1.6
؋
؋
10 at pH 9.8, while
10 at pH 11.5. A profile of
⌽
for
Ϫ
4
1
1.3 and above. Acetone sensitization does not produce
Ia in the Cyt–ethylamine system, which indicates that the
known triplet state of Cyt is not involved in reactions
leading to this opened ring product.
INTRODUCTION
Ultraviolet irradiation of living cells damages cellular chro-
matin, which can lead to mutagenesis, carcinogenesis and
cell death (1). Damage to chromatin, mediated by direct ab-
sorption of photons by DNA bases, results in a variety of
†
Abbreviations: Cyt, cytosine; DAD, diode-array detector; DMSO,
dimethylsulfoxide; DMU, 1, 3-dimethyluracil; EA, ethylamine;
EC, 1-ethylcytosine; EU, 1-ethyluracil; LSIMS, liquid secondary
ion mass spectrometry; MA, methylamine; m5Cyt. 5-methylcyt-
To whom correspondence should be addressed at: Department of
Pharmaceutical Chemistry, School of Pharmacy, University of
California, San Francisco, CA 94143-0446, USA.
␣
osine; MeOH, methanol; NAL, N -acetyl-L-lysine; SNEUV,
Southern New England Ultraviolet Company; TMS, tetramethyl-
silane; Thy; thymine; Ura: uracil.
᭧
2000 American Society for Photobiology 0031-8655/00
$5.00ϩ0.00
243

244 Kellie Hom et al.
while samples used for pH profile measurements were contained in
5 mL quartz tubes (SNEUV) mounted in a merry-go-round appa-
DNA, while Kurochkina et al. (35) found that lysine-14 in
the sequence of histone H3 photoreacts with Thy.
1
ratus. Ultraviolet spectra were either measured on the fly using the
rapid detection and data analysis capabilities of a Hewlett-Packard
1040A DAD HPLC or on a Hewlett-Packard HP8452A spectropho-
tometer (Mountain View, CA). Secondary ion mass spectra were run
on a Kratos MS-50 (Ramsey, NJ). Proton NMR spectra were ob-
tained at 300 MHz on a General Electric QE-300 NMR spectrometer
Earlier workers (36) studied ring-opened photoconjugates
produced in irradiated Thy–alkylamine, thymidine–alkyl-
amine and thymidine–lysine systems, irradiated at basic pH.
Upon heating or acidification, these products ring closed to
yield N1 alkyl-substituted Thy. More recently, Shetlar et al.
(Fremont, CA).
(37) studied the photoreactions of Cyt and 5-methylcytosine
In the descriptions of the results of our studies, given below, the
(
m5Cyt) with methylamine (MA) and n-butylamine over a
NMR data are summarized in tabular form. Some of the actual spec-
tra, as well as more detailed results and protocols for the experiments
described below, are given by Hom (40).
range of pH values, including physiological pH, and found
that, after heating or acidification of the photolyzed solu-
tions, the two nucleobases were photoalkylated at the N1
position. In another study, the same group studied the re-
action of 5-bromouracil and related compounds with alkyl-
amines (38). In the case of the m5Cyt reaction with MA, an
intermediate opened ring product was isolated and partially
Isolation of photoproducts: preliminary remarks. The photoprod-
ucts dealt with in this paper have several properties that make them
difficult to work with. In particular, the various opened ring products
are both heat labile and acid labile. In order to obtain successfully
sufficient amounts of these products for characterization, both sam-
ples to be chromatographed and purified products were kept on ice
at all times; in addition, chromatographic solvents were adjusted in
pH to about 7.4. To minimize the effects of product thermal insta-
bility, purification steps were completed in as short a time span as
feasible and characterization was done as soon as possible after iso-
lation. Even under these conditions, some opened ring photoproducts
isomerized.
characterized; this compound, N-(N
amino-2-methylacrylamidine, was converted upon heating to
,5-dimethylcytosine (37). In other studies, Cyt and m5Cyt
Ј-methyl)-carbamoyl-3-
1
␣
were photoreacted with N -acetyl-L-lysine (NAL) and L-ly-
sine at pH 7.5; again, after heating or acidification, photoex-
change products were isolated that were alkylated at the N1
position (39). The fact that Cyt and m5Cyt readily form pho-
toexchange products with alkylamines, NAL and lysine at
physiological pH suggests that similar reactions could occur
in DNA–lysine or DNA–protein systems at biologically rel-
evant pH values.
Reported here are results of investigations of UV-induced
ring opening conjugation reactions of Cyt and Ura with eth-
ylamine (EA). This primary amine was chosen to mimic the
Because of the stickiness of Cyt photoproducts on reverse-phase
HPLC columns and the concomitant broadened and misshapen peak
profiles observed for these compounds when water/methanol
(MeOH) solvents were used for HPLC elution, salts had to be added
to bring these compounds off in reasonable times with more desir-
able peak profiles. However, the chromatographic behavior and sta-
bility of some products appeared to be affected adversely by phos-
phate buffers. Therefore, aqueous lithium chloride, and in some cas-
es, aqueous sodium sulfate were utilized as eluents. The former salt
has an advantage over sodium chloride, as it is more soluble in
alcoholic media.
Isolation of opened ring photoproducts in the Cyt–EA system. Fol-
lowing irradiation of aqueous solutions of Cyt (1 mM) and EA (100
mM) at pH 10, two HPLC peaks, C1 and C2, containing materials
with UV spectral properties characteristic of opened ring com-
pounds, were observed. We found that if irradiation was carried out
⑀-amino moiety of lysine. (This choice reduces the com-
plexity due to the presence of two potentially reactive amino
groups in lysine.) Our results demonstate that opened ring
compounds, similar to those formed in the reaction of Thy
with alkylamines and lysine, are produced in these reactions.
In addition, it was found that the opened ring products of
Cyt and Ura can undergo a secondary thermal reaction that
produces compounds containing two EA moieties.
at pH 7.5 in 70% vol/vol CH CN, then the yield of C1 was boosted
3
by about 40% and the amount of C2 was minimized. The following
conditions were therefore used for the production and isolation of
C1 in amounts sufficient for characterization. An aerated solution of
4
80 mL total volume (1 mM in Cyt, 100 mM in EA, 70% vol/vol
CH CN/H O) was prepared from stock 20 mM aqueous Cyt and 1
3
2
M aqueous EA (pH 7.5) solutions. This solution was irradiated for
5 h in three 180 mL quartz tubes placed inside Vycor shields. The
MATERIALS AND METHODS
1
irradiated solution was concentrated to 150 mL by rotatory evapo-
ration, the pH was adjusted to pH 7.5 and then the resulting solution
was concentrated to 5.5 mL. After readjustment of the pH of the
concentrate to 7.5, the concentrate was immediately fractionated by
HPLC in 1 mL batches on a Capcell SG120 analytical column using
General aspects. Chemicals were obtained from Sigma (St. Louis,
MO) and Aldrich (Milwaukee, WI). Solvents for HPLC were from
Mallinkrodt (St. Louis, MO) and NMR solvents were from Aldrich.
Preparative HPLC were done on a Whatman ODS-3 reverse-phase
column (9.4
ilton PRP-1 reverse-phase column (4.5
size) (Reno, NV) or a Capcell 10 250 mM UG120 column (5
particle size) (Yokohama, Japan). The columns used for analytical
HPLC were a Whatman ODS-3 reverse-phase column (4.7 100
mm, 5 m particle size), a Hamilton PRP-1 reverse-phase (4.5
00 mm, 10 m particle size), a Shiseido Capcell SG-120 reverse-
phase column (4.6 250 mm, 5 m particle size) and a Shiseido
Capcell UG-120 reverse-phase column (4.6 250 mm, 5 m par-
ϫ
250 mm, 10
␮
m particle size) (Clifton, NJ), a Ham-
9
2% water/8% methanol flowing at 1.5 mL/min as the eluent. The
ϫ
200 mm, 10 m particle
␮
material eluting between 2.9 and 6 min was collected, concentrated
to 0.5 mL and repurified in 0.25 mL batches on the same column
using 65% (10 mM LiCl)/35% MeOH as the eluent. The material
eluting between 1.5 and 3.7 min was collected and concentrated to
0.75 mL. This material was desalted by reinjection as one batch on
the same column using 65% water/35% MeOH as the eluent; the
material coming off of the column between 2 and 20 min was col-
ϫ
␮m
ϫ
␮
ϫ
1
␮
ϫ
␮
ϫ
␮
ticle size). Two HPLC systems were used. One consisted of a Rainin
ternary gradient pumping system (Emeryville, CA) and a Hewlett-
Packard 1040A diode-array detector (DAD) (Palo Alto, CA); the
other system consisted of Rainin binary gradient pumping system
and a Kratos Spectroflow 783 absorbance detector (Ramsey, NJ)
whose output was coupled to a Rainin Dynamax data acquisition
and analysis package. Unless noted, HPLC chromatograms were
monitored at 300 nm. All irradiations were done in aerated solution
and carried out at 254 nm in a Southern New England Ultraviolet
Company (SNEUV) Rayonet RP-100 reactor (Hamden, CT) housed
lected. The major and minor
␭max of the purified material were at
298 nm and 228 nm and min was at 254 nm, as measured with the
␭
DAD in the elution solvent mixture.
To isolate the compounds contained in C2, 500 mL of aerated
aqueous solution, 1 mM in Cyt and 100 mM in EA (pH 10), was
irradiated for 15 h and concentrated to 6 mL by rotatory evaporation.
After adjustment of the pH of the concentrate from 6.5 to 7.5, it was
immediately fractionated by HPLC in 2 mL batches on a Hamilton
PRP-1 column (eluents: aqueous 2 mM LiCl [pH 7.4, eluent A] and
MeOH [eluent B]) using a profile that was isocratic at 30% B for
19 min, stepped to 60% B in 1 min and then held isocratic at 60%
B for 45 min; the flow rate was 3 mL/min). Unreacted Cyt eluted
in a cold room (maintained at 4ЊC). Samples for preparative runs
were contained in 180 mL quartz vessels shielded with Vycor filters,

Photochemistry and Photobiology, 2000, 71(3) 245
from about 5 to 8 min. The desired product eluted from 30 to 60
min. It was collected, concentrated to 5 mL by rotatory evaporation
and repurified on the same column in 2 mL batches. Using 60%
aqueous 2 mM LiCl, pH 7.4/40% MeOH at 3 mL per min as eluent,
the compound of interest eluted from about 8 to 25 min. The col-
lected eluent was concentrated and desalted as above; the material
eluting between 20 and 35 min was collected. The major and minor
␭max of the purified material were at 308 nm and 224 nm and ␭min
was at 254 nm, as determined with the DAD in the elution solvent.
In later experiments, in which smaller amounts of C1 and C2 were
required for determination of UV spectroscopic ⑀ values (see below),
a simpler procedure was developed for isolation of these com-
pounds. For C1, an aerated solution of 175 mL total volume (1 mM
in Cyt, 10 mM in EA, 70% vol/vol CH CN/H O) was prepared from
3
2
stock 20 mM aqueous Cyt and 1 M aqueous EA (pH 7.5) solutions
and irradiated for 15 h as above; about 22% of the parent Cyt was
consumed. After concentration to 1.6 mL and filtration, 200
injected on the 4.6 150 mm Capcell column, using 82/18 (vol/
vol) 10 mM Na SO /MeOH flowing at 1 mL/min as eluent. The peak
␮L was
ϫ
2
4
eluting between 3.5 and 4.5 min, corresponding to C1, was collected,
concentrated and repurified using the same column and elution con-
ditions. Under these conditions, a small peak eluting between 7 and
8.3 min could also be collected (C1Ј). The material in this peak had
an absorption spectrum, as determined by the DAD, that almost
exactly overlaid the spectrum determined for the 3.5 min peak; how-
ever, upon reinjection, most of the material contained within this
fraction coeluted with C1; thus, the material in C1Ј could not be
characterized further. It is likely that the compound contained in C1
Ј
is isomeric with the compound corresponding to C1, perhaps via
isomerism about a double bond.
For C2, the procedure was similar, except that the solution used
was 1 mM in Cyt, 100 mM in EA and at pH 10.5; before concen-
Figure 1. Representative HPLC chromatograms for the Cyt–EA and
Ura–EA systems. Panel A is a chromatogram for the Cyt–EA system
and portrays the results obtained after a solution 1 mM in Cyt and
10 mM in EA at pH 10 was irradiated for 16 h, allowed to stand
for 24 h and then concentrated by a factor of 10 using rotatory
tration, the solution was allowed to stand for 48 h at 8
ЊC. In this
case, the material eluting in the front part of the broad tailing peak
(between 5 and 9.6 min) was collected and corresponded to C2.
An HPLC chromatogram illustrating the separation obtained for
an irradiated Cyt–EA system is shown in Fig. 1A.
evaporation. A 100
with 200 L of 100 mM sodium phosphate buffer, pH 7.4, and 20
L was injected on an analytical Capcell UG120 column. The com-
ponents were eluted with 82/18 (vol/vol) 10 mM Na SO /MeOH at
␮L portion of the resulting solution was diluted
␮
Isolation of opened ring photoproducts in the Ura–EA system. In
a typical run, 500 mL of an aerated aqueous solution of Ura (2 mM)
and EA (10 mM) at pH 11.4 was irradiated in the cold for 40 h;
after this time, about 20% of the parent Ura had been converted to
products. The irradiated solution was concentrated by rotatory evap-
oration to 10 mL (pH 9.0) and filtered through a Millipore HAWP
␮
2
4
a flow rate of 2 mL/min. The substance corresponding to peak 2 is
Cyt, 3 is C1(Ia) and 4 corresponds to C2 (Ib); EC elutes between
peaks 2 and 3. The compound eluting in peak 1 is formed in the
absence of EA; its characterization will be reported in another place.
Panel B is the chromatogram obtained after a solution 1 mM in Ura
and 10 mM in EA at pH 11.3 was irradiated for 40 h. The workup
and chromatographic conditions were identical to those for the Cyt–
filter. It was then fractionated in 1 mL batches on a Capcell 150
.6 mm SG120 column using the following water/acetonitrile gra-
dient with a flow rate of 1.5 mL min: 0 min, 10%, 0.67 min, 10%;
min, 50%; 4 min, 50%; 4.5 min, 10%; 10 min, 10%. Three peaks
ϫ
4
3
EA system, except that 60/40 (vol/vol) H O/MeOH was used as
2
containing material with properties of opened ring products were
collected for characterization. The UV spectrum of the compound
eluent. Peak 1 corresponds to Ura, 2 to EU, peak 3 to U1 (the E
isomer of IIa), 4 to U2 (the Z isomer of IIa) and 5 to U3 (IIb).
contained in U1, the earliest eluting peak (3.3 min) had a ␭max
87 nm and a ␭min 240 nm as measured in the solvent composition
it eluted in, while the compound in U2, a peak eluting at 4.9 min,
showed ␭max 290 and ␭min 240 nm. The third peak (U3) eluted
ϭ
2
ϭ
ϭ
ϭ
eluent. The material corresponding to the peaks eluting between 4–
5 min, 5–6 min and 7–8 min, corresponding to U1, U2 and U3 were
collected; each fraction was concentrated and then repurified, if nec-
essary, using the same column and elution conditions.
An HPLC chromatogram showing the separation obtained for a
particular irradiated Ura–EA system is shown in Fig. 1B.
at 6.1 min and the corresponding compound contained in this frac-
tion had values ␭max and ␭min of 298 and 250, respectively. The
material contained in U1 reacted upon standing to form a mixture
of two compounds that displayed HPLC elution times and UV spec-
tra associated with U1 and U2. Similarly, the material collected in
U2 reacted to form a mixture of compounds with the same retention
times and UV spectra as those corresponding to U1 and U2. Indeed,
within the time required for reinjection of purified U1 (or U2) after
initial purification, detectable amounts of the material associated
with peaks U2 (or U1) had formed, even though the purified mate-
rials in each peak were collected in an ice-cooled beaker and then
stored on ice.
Isolation of opened ring photoproducts in the m5Cyt–MA and
Thy–MA systems. To prepare and isolate the E isomer of N-(NЈ-
methylcarbamoyl)-3-amino-2-methylacrylamidine (III) (37), 150
mL of a solution containing m5Cyt (1 mM) in aqueous MA (10 mM,
pH 11.4) was irradiated for 20 h, rotoevaporated to 0.5 mL and
chromatographed on a Capcell UG 120 preparative column using
81/19 (vol/vol) 10 mM Na SO /MeOH flowing at 4 mL/min as el-
2
4
For experiments in which smaller amounts of U1, U2 and U3
uent. The material eluting between 6.5 and 8 min corresponded to
the desired substance.
were required (e.g. for determination of UV spectroscopic
see below]), an isocratic procedure was developed for isolation of
these compounds. An aerated solution of 175 mL total volume (2
mM in Ura, 10 mM in EA, pH 11.3) was irradiated for 15 h as
above. After concentration to dryness, the residual material was tak-
en up in 1.2 mL of water, to which 800 L of MeOH was then
added; 250 L portions were injected on the 4.6 150 mm Capcell
column, using 60/40 (vol/vol) H O/MeOH flowing at 1 mL/min as
⑀ values
[
In the case of the Thy–MA system, 170 mL of a solution of Thy
(2 mM) in 10 mM aqueous MA (10 mM, pH 11.3) was irradiated
for 20 h in the SNEUV reactor as described above. After concen-
tration of 100 mL of the irradiated solution to 1.0 mL, followed by
filtration, the solution was diluted to 1.2 mL. The opened ring Thy–
MA conjugate IV was isolated using a Capcell UG120 analytical
ϭ
␮
␮
ϫ
column, using 82/18 (vol/vol) H O/MeOH flowing at 1 mL/min as
2
2

246 Kellie Hom et al.
eluent. Under these circumstances, the desired E isomer of IV eluted
between 4.5 and 6 min.
Determination of total opened ring product yields for the Cyt–EA
and Ura–EA systems. Total opened ring product yields for these
two systems were determined using a protocol with elements similar
to those described immediately above. For the Cyt–EA system, a
solution of Cyt (1 mM) in aqueous EA (100 mM, pH 10.5) was
irradiated for 15 h in the cold. One milliliter samples of unirradiated
parent solution and irradiated samples were rotatory evaporated to
dryness. After the residue from the irradiated sample was taken up
Estimation of
MA and Thy–MA conjugates. Values of
⑀
values for opened ring Cyt–EA, Ura–EA, m5Cyt–
for the opened ring con-
⑀
jugates of Cyt and Ura were estimated at their absorption maxima
and minima using a modified version of the procedure described in
Shetlar et al. (38). Similar determinations were made for analogous
opened ring conjugates of m5Cyt and Thy with MA, namely the E
isomers of N-(N
(
(
Ј
-methylcarbamoyl)-3-amino-2-methylacrylamidine
III)(37) and N-(N -methylcarbamoyl)-3-amino-2-methylacrylamide
IV). A portion of freshly purified conjugate of interest was diluted
in 1 mL of distilled water, 900
placed in a screw-top culture tube, along with 100 ␮
␮
L of the resulting solution was
L of 1 M HCl,
Ј
and heated for 45 min in a boiling water bath. The amounts of EC
and EU present in the resulting solution, after pH adjustment as
above with 100 mM phosphate buffer, were determined as described
in the previous section. The protocol used for determination of total
opened ring product yield in the Ura–EA system was similar. The
system studied in this case was 1 mM in Ura and 10 mM in EA at
pH 11.3; it was irradiated for 40 h in the cold. To determine the
amount of Cyt or Ura that had been consumed in these reactions,
the peak areas for Cyt (or Ura) were determined for both the unir-
radiated and irradiated samples on a Capcell UG120 column, using
100% aqueous Na SO as eluent at a flow rate of 2 mL/min. The
by an appropriate factor in MeOH (or, in one case, doubly distilled
water) and its absorbance was determined at the wavelengths of
maximum and minimum absorbance; each measurement was done
in, at least, triplicate and all dilutions were made with direct dis-
placement pipettors. From the average of the measured values, the
absorbance of the original solution at each of these wavelengths was
determined. Then, to a 1 mL portion of the purified conjugate con-
tained in a Teflon capped screw-top culture tube, was added 20 ␮L
of an internal standard (either 20 mM Ura or 20 mM m5Cyt, de-
pending on which worked best for an individual system) and 111
2
4
␮
L of 1 M HCl. The resulting solutions were heated in a boiling
water bath for 15 min, 30 min and 45 min. After each of these time
intervals, as well as before heating, 100 L samples were withdrawn
from the tube; 50 L of this was then diluted with 100 L of 100
mM sodium phosphate buffer (pH 7.2). The resulting solutions were
analyzed in duplicate by isocratic HPLC by injecting 20 L samples
on the Capcell 0.46 15 cm column at an eluent flow rate of 1
mL/min. The following eluents were used for HPLC analysis of the
various heated systems (N aqueous 10 mM Na SO ): Cyt–EA,
ratio of these areas subtracted from unity indicated the fraction of
parent compound remaining, from which the fraction consumed and
then the amount of parent nucleobase consumed could be calculated.
The total yield of opened ring product, based on reacted parent com-
pound, was obtained by dividing the total amount of 1-ethylpyrim-
idine produced in each system by the amount of parent pyrimidine
nucleobase consumed.
Determination of quantum yields for formation of opened ring
conjugates in the Ura–EA, Cyt–EA and Thy–MA systems. Quantum
yields for formation of opened ring conjugates in the Ura–EA and
Cyt–EA systems were determined at 254 nm using a method some-
what modified from that described by Shaw and Shetlar (45) that
used 1,3-dimethyluracil (DMU) as an actinometric substance. The
application of the method, in this case, relies upon the assumption
that all opened ring conjugates are converted upon heating in acid
to the corresponding 1-alkylcytosine and/or 1-alkyluracil products,
as was demonstrated in work described above. In a typical run, the
nucleobase sample and two actinometric samples were contained in
stirred quartz spectrophotometer cells and, as described in Shaw and
Shetlar (45), irradiated at equivalent positions in front of a Vycor-
shielded germicidal lamp. After irradiation for a period lying be-
tween 45 and 90 min, 0.6 mL of each irradiated sample was re-
moved. The concentration of DMU hydrate in each irradiated acti-
nometric solution was evaluated by HPLC, using a standard working
␮
␮
␮
␮
ϫ
ϭ
2
4
Thy–MA and m5Cyt–MA conjugates, 82 N/18 MeOH; Ura–EA
conjugates, 80 H O/20 MeOH. In each run, the DAD was set to
2
collect data at the absorption maximum of each reactant and product
in the heated mixture. The measured integrated area for each product
peak of interest in a particular heated sample was adjusted by mul-
tiplying by a factor obtained by dividing the area for the internal
standard in a designated sample that had been heated for 15 min by
the corresponding area for the internal standard in the sample of
immediate interest; this insured that possible solvent or sample loss
during heating or other manipulations did not affect the results. The
amounts of each ring closure product of interest (1-alkyluracil, Thy
or Cyt, see below) were determined via conversion of the averaged
integrated area, determined at the appropriate ␭max for the compound
of interest, to concentration using a corresponding working curve of
area at ␭max versus concentration that had been previously construct-
ed for each closed ring product. Working curves that were also used
in the determination of quantum yields for the Cyt–EA, Ura–EA and
Thy–MA systems (see below) were constructed using solutions pre-
pared by dilution of stock solutions of known concentration. The
concentrations of the parent stock solutions were determined spec-
trophotometrically using the following ␭max and ⑀max values from the
curve constructed for a Capcell UG120 4.6
ϫ 150 mm column op-
erating with 80/20 (vol/vol) 10 mM Na SO /MeOH flowing at 1 mL/
2
4
min as eluent; the two resulting values obtained were averaged.
To 0.5 mL of the irradiated nucleobase–EA sample was added
55.5
␮
L of 1 M HCl; this was followed by heating for 45 min in a
L of this material was added 200 ␮L
boiling water bath. To 100
␮
literature: 1-ethylcytosine (EC), ␭max
-ethyluracil (EU), ␭max 268, log ⑀max
mine, ␭max 273, log ⑀max 3.96 (43); 1,5-dimethylcytosine, ␭max
280, log ⑀max 3.89 (44). Summation of the amounts of each
closed ring product in the sample (see below) provided the corre-
sponding amount of the opened ring conjugate that was the precursor
of these compounds. Knowledge of the dilution history of the orig-
inal sample allowed calculation of the amount of opened ring con-
jugate present in the original sample. Division of the absorbance at
each maximum and minimum wavelength by this calculated value
ϭ
274, log ⑀max
ϭ
3.92 (41);
of 100 mM sodium phosphate buffer (pH 7.2); all volume measure-
ments were made with direct displacement pipettors. The concentra-
tion(s) of N-ethylated pyrimidine base product(s) was determined,
using HPLC as above, employing standard working curves to con-
vert integrated area to concentration. After correcting for dilution,
the quantum yields were determined by dividing the concentration
of N-ethylated pyrimidine base product(s) in the irradiated solution
by the averaged concentration of DMU hydrate formed in the acti-
nometric solutions and multiplying by 0.0106, the quantum yield for
formation of DMU hydrate. The quantum yield for formation of the
opened ring Thy–MA conjugate in the Thy–MA system was deter-
mined using a protocol analogous to that described above.
1
ϭ
ϭ
4.04 (42); 1-methylthy-
ϭ
ϭ
ϭ
ϭ
of the concentration of open ring conjugate produced the desired
values. The values obtained are reported in Table 2.
⑀
In the case of the opened ring Ura–EA and Thy–MA conjugates,
the only products obtained upon heating in 0.1 M HCl were the
corresponding 1-alkyluracil. For the Cyt–EA (m5Cyt–MA) opened
ring conjugate, heating under the above conditions led to the cor-
responding EC (1,5-dimethylcytosine) as the major product and to
EU (1-methylthymine) as the only detectable minor product. In all
cases, time course studies were done to insure that the amounts of
reaction products resulting from heating in acidified solution reached
maximum within 15 min; other studies were done on the various
reaction products and the internal standard to demonstrate that they
were stable to heating under the same conditions.
The effect of pH on the quantum yield of formation of opened ring
conjugates in the Ura–EA and Cyt–EA systems.. The pH depen-
dence of the relative quantum yield for formation of opened ring
conjugates in the Ura–EA and Cyt–EA systems was studied using
a slightly modified version of the method described in Shetlar et al.
(38). Solutions that were 2 mM in nucleobase and 25 mM in EA,
but of varying pH, were prepared by mixing stock solutions of the
same concentration of components with pH at or near the ends of
the pH range to be studied. For example, solutions of varying pH
in the Cyt–EA system were prepared by mixing appropriate amounts
of stock solutions at pH 11.7, 6.8 and 5.0. A volume of 3.5 mL of

Photochemistry and Photobiology, 2000, 71(3) 247
each solution was placed in a 15 mL stoppered quartz tube and a
set of tubes was irradiated with light of predominantly 254 nm wave-
length for 60 min on a merry-go-round placed in the SNEUV pho-
toreactor. Each batch of tubes irradiated had a tube among them
with a set pH value (e.g. for the Cyt–EA system, pH 10.1). This
allowed the measured areas from HPLC analysis for samples in dif-
ferent groups to be normalized with respect to the area found for
that standard pH. This procedure circumvented the problem of sam-
ples in different groups receiving differing radiative exposures. The
irradiations were carried out in a cold room maintained at 4ЊC, al-
though the actual temperature in the reactor was somewhat higher
than this. After removal from the reactor, the samples were imme-
diately placed on ice; all subsequent manipulations, up until injec-
tion, were carried out with solutions kept in the ice bucket. Imme-
diately before analysis of a particular sample, one volume of irra-
diated solution was diluted with two volumes of 100 mM sodium
phosphate (pH 7.2). The sample was then analyzed on a Capcell
UG120 4.6
ϫ 150 mm column, using 75/25 (vol/vol) 10 mM
Na SO /MeOH flowing at 1 mL/min as eluent for the Cyt–EA sys-
2
4
tem and 75/25 (vol/vol) H O/MeOH flowing at 1 mL/min for the
2
Ura–EA system The integrated area(s) for the peak(s) corresponding
to the opened ring conjugate of interest was obtained from the DAD;
in each case, the DAD was set to monitor the eluent at a wavelength
at or close to that of the maximum absorbance for the opened ring
conjugate being studied. In the Ura–EA system, peaks corresponding
to U1 and U2 were present. The areas of these two peaks were
summed to obtain the total area. For the Cyt–EA system, two peaks
corresponding to C1, one much larger than the other, were obtained
at pH values of 7.1 and above; these were similarly added to obtain
an area reflective of the total amount of conjugate present. These
sums were normalized to the corresponding peak area for the sample
having the largest combined area; these areas are plotted in Fig. 2.
Although not evident in Fig. 2, opened ring conjugate could be
detected in the Cyt–EA system at a pH as low as 5.6. We were
unable to detect any opened ring conjugates in the Ura–EA system
at pH values of 7.5 or below.
Acetone sensitization and formation of C1 in the Cyt–EA system.
Population of the Cyt triplet state was accomplished using acetone
as a photosensitizer. A 6 mL sample of Cyt (1 mM)–EA (10 mM)
in aqueous solution that was made 5% vol/vol in acetone (pH 10.0)
was irradiated for 16 h in a Pyrex 15 mL tube placed in front of a
Spectronics BLE-1T158 lamp (Westbury, NY). A solution that was
identical, except it contained no acetone, was irradiated in parallel.
Figure 2. The pH profiles for the relative quantum yield of for-
mation of opened ring conjugates in the Ura–EA system (upper pan-
Prior to analysis, 100
00 L of 100 mM sodium phosphate buffer (pH 7.2) and then 200
L of the resulting solution was analyzed as described for the Cyt–
␮L samples of each sample were diluted with
el) and Cyt–EA system (lower panel).
␳ is the normalized value for
2
␮
the quantum yield at a given pH, as compared to the maximum
quantum yield.
␮
EA samples in the preceding section. The resulting chromatograms
were examined for the presence of both C1 and EC.
RESULTS AND DISCUSSION
Purified C2 was also characterized by UV and proton
NMR spectroscopy and by mass spectrometry; the resultant
data are presented in Tables 1 and 2. The UV, proton NMR
and mass spectral data are all consistent with the material in
C2 being N-(NЈ-ethylcarbamoyl)-3-ethylaminoacrylamidine,
one form of which is displayed in Scheme 1 as Ib. The
Characterization of the opened ring photoproducts in
the Cyt–EA system
Purified C1 was analyzed using UV and proton NMR spec-
troscopy and by liquid secondary ion mass spectrometry
(
LSIMS); the data resulting from these studies are given in
Tables 1 and 2. The spectroscopic data are consistent with
the material in C1 being N-(N -ethylcarbamoyl)-3-aminoac-
proton NMR spectrum indicates the presence of two isomers
in the NMR sample, while the mass spectrum showed only
Ј
ϩ
one major (M
ϩ
H) mass peak. While the spin–spin cou-
rylamidine, one form of which is shown as Ia in Scheme 1.
The proton NMR showed that there are two isomeric forms
of C1 contained in the NMR sample, while the mass spec-
pling constants for these compounds could not be deter-
mined, it is likely that they correspond to isomers of Ib that
have a cis–trans relationship about the C –C bond, as found
ϩ
2
3
trum displayed a single major (M
ϩ
H) mass peak. Of
for Ia.
particular interest is the observation that the two compounds
in the NMR sample display significantly different C2H–C3H
spin–spin coupling constants, one having a J of 8 Hz and
Characterization of opened ring photoproducts in the
Ura–EA system
the other 12 Hz. A value of J
uration in which these two protons are trans to one another
as shown in Scheme 1 for Ia), while the value of J 8 Hz
ϭ 12 Hz indicates a config-
After a final repurification, the material contained in the peak
corresponding to U1 was subjected to proton NMR and
LSIMS. As expected from the instability of U1 toward con-
version to U2 and vice versa (see above), the NMR spectrum
(
ϭ
implies that the two protons in this compound are in a cis
configuration relative to one another.

248 Kellie Hom et al.
Table 1. Proton NMR chemical shift and mass spectral data for the photoproducts formed in the reactions of Cyt and Ura with EA*
Ia
Ib†
IIa
IIb
C1NH2
CONHCO
C2H
9.53 ppm, br, 1; 7.60
ppm, br, 1
8.95 ppm, br; 6.95 ppm, NA
br
NA
NA
NA
Isomer 1: 7.74 ppm, br;
isomer 2: 7.7 ppm
Isomer 1: 7.6 ppm, br; iso-
mer 2: 7.6 ppm
Isomer 1: 5.07 ppm, d, 1,
JC3H–C2H 12 Hz‡; isomer
Isomer 1: 4.14 ppm, d,
isomer 2: 4.55 ppm
Isomer 1: 4.95 ppm, d,
JC3H–C2H 13 Hz; isomer
2: 4.49 ppm, d, JC3H–C2H
8 Hz
Isomer 1: 4.76 ppm, d, 1,
JC3H–C2H 13 Hz; isomer
2: 4.41 ppm, d, 1, JC3H–
2
: 5.88 ppm, d, 1, JC3H–
8
Hz‡
8 Hz
C2H
C2H
C3H
Isomer 1: 8.43 ppm, m, 1, Isomer 1: 6.56 ppm; iso-
mer 2: 7.51 ppm
Isomer 1: 7.45 ppm, m,
JC3H–C2H 13 Hz; isomer
2: 6.82 ppm, m, JC3H–C2H
8 Hz
Isomer 1: 7.48 ppm, m, 1,
JC3H–C2H 13 Hz, JC3H–NHR
8 Hz; isomer 2: 6.76
ppm, m, JC3H–C2H 8 Hz,
JC3H–NHR 13 Hz
JC3H–C2H 12 Hz; isomer
2
: 7.48 ppm, m, 1, JC3H–
8
Hz
C2H
C3NH2
C3NH
6.6 and 6.4 ppm, br, 2.
NA
NA
Isomer 1: 5.38 ppm, br, 2; NA
isomer 2: 6.6 ppm, br
Isomer 1: 9.18 ppm, br, 1: NA
isomer 2: 6.64 ppm, br
Isomer 1: 5.61 ppm, br, 1:
isomer 2: 8.31 ppm, br
C3NHCH CH
NA
Isomer 1: 3.10 ppm, m;
isomer 2: 3.07 ppm, m
NA
Isomer 1: 3.07 ppm, m, 2;
isomer 2: 3.32 to 3.13
ppm, m
2
3
3
C3NHCH CH
NA
1.1 ppm
NA
1.1 ppm
2
CONHCH CH
8.29 ppm, br, 1
Isomer 1: 6.82 ppm, m;
isomer 2: 6.27 ppm, m
Isomer 1: 8.65 ppm, br, 1; Isomer 1: 8.64 ppm, br, 1;
2
3
isomer 2: 8.46 ppm, br,
1
isomer 2: 8.46 ppm, br,
1
CONHCH CH
Isomer x: 3.76 ppm, m, 2; Isomer 1: 3.09 ppm, m;
Isomers 1 and 2: 3.29 to
3.14 ppm, m
Isomers 1 and 2: 3.32 to
3.13 ppm, m
2
3
3
isomer y: 3.22 ppm, m
isomer 2: 3.07 ppm
CONHCH CH
Isomer x: 1.22 ppm, t, 3,
JCH2–CH3 7 Hz; isomer y:
1.1 ppm
Isomer 1: 1.10 ppm, t,
JCH2–CH3 7Hz; isomer 2:
Isomers 1 and 2: 1.18 to
1.06 ppm, m
2
1
7
.12 ppm, t, 3, JCH2–CH3
Hz
1.11 ppm, t, JCH2–CH3
Hz
7
ϩ
ϩ
LSIMS, [M
ϩ
H]
157
185; [M
ϩ
Na]
ϭ
207
158
186
*
All NMR spectra, except where noted by †, were run at 300 MHz in d acetonitrile, using tetramethylsilane (TMS) as an internal standard.
3
For each entry, the resonance of interest is in bold face type. Due to the low signal-to-noise ratio and overlapping of resonances in some
spectra, we report only the integration values and coupling constants (in Hz) that we could measure with certainty. The NMR spectra
denoted by † were run at 300 MHz in d –DMSO, using TMS as an internal standard. The signals belonging to each isomer pair were
6
assigned using spin-decoupling experiments. For the pair of isomers denoted by ‡, the ratio of the area for the isomer with J
to that with J 8 Hz was approximately 2:1. NA means nonapplicable.
ϭ 12 Hz
ϭ
provided evidence that two compounds were present in the
NMR sample of U1, while the mass spectrum suggested that
both components had the same molecular mass. The relevant
NMR and mass spectral data are presented in Table 1. The
behaviors seen for the E and Z forms of the corresponding
photoconjugates of Thy and thymidine with alkylamines
(36). In summary, the data, taken as a whole, are consistent
with the material in the U1 NMR sample being a mixture of
spin–spin coupling constants determined for the C and C3
E and Z isomers of N-(NЈ-ethylcarbamoyl)-3-aminoacrylam-
2
protons for the two isomers (12 Hz and 8 Hz) are consistent
with these two compounds being stereoisomers with an E–
Z relationship about the C –C bond; the major isomer in
ide and suggests that conversion between the two configu-
rations is a process that readily occurs in purified samples
of each isomer.
2
3
the mixture had J
ϭ
12 Hz, indicating that U1 has the E
The purified material corresponding to U3 was also sub-
jected to UV and proton NMR spectroscopy and to LSIMS.
The NMR data, presented in Table 1, suggest that that there
are two compounds eluting in U3; the mass spectrum indi-
cates that the two compounds have the same molecular mass.
The values of the spin–spin coupling constants for the C2
configuration (the form shown as IIa in Scheme 1, termed
subsequently IIa1) and thus, that the material eluting in the
peak corresponding to U2 is the Z form (IIa2, not shown in
Scheme 1). Suggestive evidence supporting this assignment
is obtained from the UV spectroscopic measurements and by
the HPLC behavior of these two compounds. Examination
of the data in Table 2 shows that IIa1 has an absorption
maximum at 286 nm, while the corresponding maximum for
IIa2 is 288 nm; also IIa1 elutes prior to IIa2 on reverse-
phase HPLC. These trends are in agreement with similar
and C protons for the two compounds (see above) suggest
3
that the two compounds are E and Z isomers of N-(N
carbamoyl)-3-ethylaminoacrylamide (IIb). The major form
of IIb in the NMR sample had J 13 Hz, suggesting that
it has the E configuration, the form shown in Scheme 1.
Ј-ethyl-
ϭ

Photochemistry and Photobiology, 2000, 71(3) 249
Table 2. Ultraviolet spectroscopic data for various nucleobase–
amine opened ring conjugates*
Com-
pound
␭max Log ⑀max
␭min
Log ⑀min
␭max Log ⑀max
Ia†
Ia‡
Ib‡
IIa1
IIa2
IIb
230
228
224
4.02
4.07
4.17
258
258
258
234
234
240
256
236
3.28
3.27
3.55
3.10
3.14
3.17
3.29
3.16
300
298
308
286
288
290
300
288
4.62
4.62
4.63
4.45
4.36
4.54
4.62
4.50
III†
IV
230
3.98
*
All values were determined in 98.5/1.5 (vol/vol) MeOH/H O ex-
2
cept for those denoted by daggers; for the cytosine conjugates,
Scheme 1.
this solution was about 0.13 mM in Na SO . Note that measure-
2
4
ments of spectroscopic parameters for IIa1, IIa2, IIb and IV
cannot be made in water-rich solutions as these compounds rap-
idly decompose under these conditions.
7% of the Cyt had been consumed, the total yield of Cyt–
EA opened ring conjugate was about 43%, based on con-
sumed parent Cyt. For the Ura–EA system, about 34% of
the parent Ura was consumed and the total yield of Ura–EA
opened ring conjugate was about 13%. (It should be noted
that meaningful product yields for individual opened ring
products are difficult to obtain, as various of these products
are converted from one form to another, with the amounts
converted depending on the time elapsed between irradiation
and workup and the conditions of workup.)
†
‡
These values were determined in 97/3 (vol/vol) MeOH/H O that
2
was 0.25 mM in Na SO .
2
4
These two sets of values were determined in a 97/3 (vol/vol) H O/
2
MeOH solution that was about 0.25 mM in Na SO .
2
4
Conversion of IIa1 and IIa2 to IIb via a dark reaction
We observed that the process of concentrating photoreaction
mixtures by rotatory evaporation increased the amount of
material eluting in U3 and concomitantly decreased the
amounts of IIa1 and IIa2. The appropriate measurements
were made on samples that were first concentrated to near
dryness, without adjusting pH and then rediluted to their
original volume. Indeed, it was found that the same reaction
occurred in the absence of concentration; when freshly ir-
radiated solution was placed in the cold room and allowed
to stand for 2 weeks, the peak area corresponding to U3
increased dramatically, while those areas corresponding to
IIa1 and IIa2 correspondingly decreased, as compared to
those observed for the freshly irradiated solution. These
shifts in the relative amounts of compounds corresponding
to conjugates containing one molecule of EA and conjugates
containing two molecules of EA are most likely due to dark
reactions in which IIa1 and IIa2 are converted to IIb by
reaction with EA. Similar behavior was observed for Ia, the
corresponding Cyt monoconjugate. The reaction of Ura– and
Cyt–EA monoconjugates with another EA at basic pH to
form conjugates containing two molecules of amine is anal-
ogous to reactions that have been observed in 5-bromoura-
Quantum yields for formation of opened ring
conjugates in the Ura–EA, Cyt–EA and Thy–MA
systems
The quantum yields for formation of opened ring conjugates
were determined for both the aqueous Cyt–EA system and
Ura–EA system at a single pH value, as described in the
Materials and Methods. The values obtained were the fol-
Ϫ
4
lowing: Cyt (1 mM)–EA (10 mM), 1.6
ϫ
10 (pH 9.8); Ura
Ϫ
4
(1 mM)–EA (10 mM), 7.2
ϫ 10 (pH 11.3). (The value for
the Cyt–EA system can be compared to a value of 8.3
ϫ
Ϫ5
10
determined previously (37) at pH 8.7 for the Cyt [2
mM]–MA [100 mM] system.) Still another run was made on
the Cyt (1 mM)–EA (10 mM) system, prepared from stock
EA with pH 7.2 and contained in 70/30 CH CN/H O; the
3
2
Ϫ
3
value obtained in this case was 1.2
ϫ 10 , almost 10-fold
higher than the completely aqueous system. One factor re-
sulting in this increased quantum yield for conjugate for-
mation in the acetonitrile/water system may be that decreas-
ing the water concentration decreases the rate of formation
of hydrate and thus increases the proportion of photoexcited
Cyt that react with EA.
cil– and 5-bromo-2Ј-deoxyuridine–alkylamine systems (38).
This reaction probably involves addition of unprotonated al-
kylamine at the 3 position, thus saturating the 2,3 double
bond; elimination of ammonia would lead to the correspond-
ing final products in which two molecules of amine are in-
corporated.
For comparison purposes, the corresponding yield was
measured for the Thy–MA system (1 mM/10 mM, pH 11.3),
Ϫ
3
for which a value of 6.1
can be compared to the value of 2.8
ϫ 10 was obtained. This value
Ϫ
3
ϫ
10 determined for
the thymidine–MA system (0.2 mM/10 mM) at the same pH
by Saito et al. (36).
Total yields of opened ring conjugates in the Cyt–EA
and Ura–EA systems
The pH profiles for formation of opened ring
conjugates in the Cyt–EA and Ura–EA systems and
some mechanistic implications
The total yield of opened ring conjugate was determined for
two systems, one in which Cyt (1 mM) in aqueous EA (100
mM, pH 10.5) was irradiated for 15 h and the other in which
a solution 1 mM in Ura and 10 mM in EA at pH 11.3 was
irradiated for 40 h. For the Cyt–EA system, in which about
Figure 2 shows that reactivities in the Cyt–EA and Ura–EA
systems display quite different pH profiles. This is not un-

250 Kellie Hom et al.
expected, as prior studies have also shown dissimilarities in
pH profile for the photoexchange reactions of Cyt and
m5Cyt with amines on one hand (37) and the corresponding
reactions of Thy or thymidine on the other hand (36,37). It
should be emphasized that the previous published study on
the reaction of m5Cyt with MA and Cyt with butylamine
(37) produced profiles that were similar in shape to those
found here for the Cyt–EA system; however, measurements
for the earlier study were done by monitoring the amount of
1
-alkylcytosine produced after heating solutions irradiated at
each pH in acid, while the present study was done by di-
rectly monitoring the formation of opened ring conjugates.
The pH profile for the Ura–EA system (shown in the up-
per panel of Fig. 2) is similar to those profiles previously
determined for the thymidine–MA and Thy–MA systems.
Saito and co-workers (36) have interpreted this profile to
indicate that the reactive species entering into the thymi-
dine–MA reaction are an activated anionic form of thymi-
dine and the unprotonated amine. Similar conclusions can
be drawn for the Ura–EA reaction. However, the corre-
sponding pH profile for the Cyt–EA system is qualitatively
different from that observed for the Ura–EA system. As can
be seen in the lower panel of Fig. 2, the pH profile for the
Cyt–EA system has a maximum at a pH of about 9, while
the quantum yield at pH 7.4 (near physiological pH) is about
Figure 3. Mechanistic schemes for the ring-opening reactions of
Ura and Cyt with EA. Figure 3a shows a mechanism for Ura, based
on the mechanism proposed by Saito et al. (36) for the similar re-
action of Thy and related compounds with alkylamines. Fig. 3b dis-
plays a similar mechanism that may be operating at high pH values
for the Cyt–EA photoreaction. Fig. 3c represents a mechanism that
can be proposed to account for the photoreaction of Cyt with pro-
tonated EA at acidic, neutral and slightly basic pH values.
2
2% of the maximal yield. Formation of opened ring con-
as the proportion of unprotonated amine decreases. This
strongly suggests that another pathway must be involved in
producing opened ring conjugate, perhaps operating in par-
allel to the pathway described previously. The fact that the
concentration of protonated amine increases as pH is low-
ered from pH 11.6 to pH 9, with concomitant increased
quantum yield for formation of opened ring conjugate, sug-
gests that protonated amine is involved in this second reac-
tion pathway. However, it is unlikely that this species itself
could act as a nucleophile to attack excited Cyt.
jugate can be detected at a pH as low as 5.6. As one moves
from pH 9 to higher pH values, the reactivity also drops off,
although less rapidly than is the case in moving from pH 9
toward lower pH. An interpretation directly analogous to that
of Saito cannot explain the pH profile corresponding to the
Cyt–EA reaction. In particular, the pK for the ionization of
a
Cyt to form the Cyt anion is around 12.2 (46) and this ion-
ization is at N1 on the pyrimidine ring, rather than N3 as is
the case in N1 alkylated Thy. Indeed, the downward trend
in reactivity of Cyt as pH increases between 9 and 11 is
opposite in direction to that observed over the same pH
range in the corresponding reactions in Ura– and Thy–amine
systems.
Acetone photosensitization does not lead to production
of Ia in the Cyt–EA system
The HPLC analysis of acetone-photosensitized reactions in
the Cyt–EA system showed that neither Ia nor EC, its ring
closure (photoexchange) product, are formed in this system.
This result is in line with the results from studies of other
pyrimidine nucleobase and nucleoside–alkylamine systems
that have shown that acetone photosensitization, followed by
heating in the presence of acid, does not lead to production
of photoexchange products. Among those for which this is
true are the Thy– and Ura–lysine systems (49) and the Thy–,
Cyt– and m5Cyt–MA (36,37) systems. This implies that the
excited-state precursors of the opened ring products, from
which the photoexchange products are ultimately formed, are
not accessible via acetone sensitization and indicates that the
lowest excited triplet states of these compounds are not in-
volved in reactions to form opened ring conjugates analogous
to Ia.
It is interesting to note that photoexcited cytidine also
shows a decreasing reactivity toward hydration to form 5,6-
dihydro-6-hydroxycytidine as the pH decreases in the range
from 9 to 1.5 (47). However, the midpoint of the dropoff in
reactivity occurs near a pH of 4, rather than near 8 as in the
Cyt–EA system. The structure of Cyt itself is analogous to
one of the resonance structures that can be written for Ura
(or Thy) that has lost a proton at N3 (i.e. one with negative
charge delocalized to O4), as shown in Fig. 3a. Thus, reac-
tion of Cyt with unprotonated EA could be considered for-
mally analogous to the reaction of the Ura anion with EA
(Fig. 3b). However, operation of this pathway alone cannot
explain the observed pH profile of reactivity. For example,
because EA has a pK of 10.63 (48), one can calculate that
a
around 98% of EA present in solution will be in the proton-
ated form at pH 9. Indeed, at pH 5.6 (where reaction can
still be detected) the concentration of unprotonated amine in
the 25 mM solutions used for the pH profile studies would
Some general comments concerning the activated state
precursor of the opened ring conjugate in the Cyt–EA
system
be 0.25
␮M, a concentration so small that it would be un-
likely to compete for photoexcited Cyt in a bimolecular pro-
cess. It should also be noted than, in going from pH 11.5 to
pH 9, the quantum yield of reaction actually increases, even
What can be said concerning the nature of the activated state
that reacts with EA to form the Cyt–EA opened ring con-

Photochemistry and Photobiology, 2000, 71(3) 251
jugate? Some of the photophysical properties of two excited
states of Cyt have been determined, namely the singlet state
populated by direct absorption (50) and a triplet state that is
populated via acetone sensitization (51–53) and presumably,
through inefficient intersystem crossing as well. The known
triplet state, or any activated states arising from this triplet,
can be ruled out as precursors because of our observation
that acetone does not sensitize the formation of opened ring
conjugate in Cyt–amine systems (see discussion above).
However, based on presently available information, the ex-
cited singlet cannot be definitively ruled out. From the value
for the quantum yield for the Cyt–EA ring opening reaction
could be speculated that the reaction between cytosine and
ethylamine involves yet another activated state of the nu-
cleobase (e.g. a longer-lived nonfluorescent singlet state pro-
duced in significant amount by either direct light absorption
or through internal conversion from the initially excited sin-
glet) or a triplet state inaccessible via acetone sensitization.
An example of such an activated state might be one asso-
ciated with a tautomeric form of Cyt. It is interesting that
there is both experimental (55) and computational evidence
(56) suggesting that significant concentrations of tautomeric
isomers of Cyt are present in equilibrium with the amino
keto form in aqueous solution. For example, calculations
(56) indicate that, in aqueous solution, neither the transition
Ϫ
4
given in a previous section (1.6
ϫ 10 at pH 9.8 in a 10
mM EA solution), a rough estimate can be made of the min-
energy of the lowest ␲–␲* transition nor the oscillator
imum lifetime that a prospective reactive precursor for the
strength of the predominant conformer of the amino hydroxy
tautomer of Cyt differ greatly from that of the amino keto
isomer of Cyt. Thus, these calculations suggest that absorp-
tion of 254 nm light will produce excited states of both the
amino keto and the amino hydroxy tautomers. Further spec-
ulation along these lines, however, is unwarranted until more
experimental information about the physical and reactive
properties of such tautomers and their excited states becomes
available. It should be emphasized, however, that any suc-
cessful mechanistic interpretation of the photoreaction of Cyt
with EA to form opened ring conjugate must predict the
shape of the pH profile for reactivity as shown in Fig. 2; in
particular, it must explain the increase in the quantum yield,
as the proportion of total amine that is protonated increases
between pH 11 and 9, and the rapid drop in relative quantum
yield as the pH decreases from 9 (a region where the pro-
portion of protonated EA increases only slightly, eventually
becoming essentially unity).
opened ring product could have. If reaction of singlet excited
9
Cyt with EA is taken as diffusion controlled (k
ϭ 7.4 ϫ 10
L/mol s (54)), then the maximal rate of reaction would be
7
Ϫ1
7
.4
ϫ 10 [C*] s , where [C*] is the steady-state concentra-
tion of activated Cyt species. Because the observed quantum
Ϫ
4
yield for reaction is 1.6
ϫ 10 , the rate of deactivation of
this activated precursor in the presence of amine would be
11
Ϫ1
on the order of 4.6
ϫ 10 [C*] s . From this number, it can
be calculated that the lifetime of the reactive state (in the
presence of 10 mM EA) would be on the order of 2.2 ps
under these conditions. It is interesting to note that the life-
time of photoexcited singlet state of Cyt (in the absence of
amine) has been estimated as being on the order of 2 ps
(50). Thus, based on this calculation that assumes optimal
reactivity (reaction upon every collision of Cyt* [hereafter
abbreviated as C*] with EA), the known Cyt excited singlet
state cannot be excluded as a precursor of the Cyt–EA
opened ring conjugate. However, at pH 9.8 an appreciable
fraction (86%) of EA molecules will be protonated. This
presents a dilemma, for if C* is the precursor of opened ring
conjugate, then it must be able to react with protonated
amine in order for the rate of reaction to form these com-
pounds to be sufficiently large to yield a quantum yield of
Biological considerations
The biological importance of the types of reactions described
in this paper remains to be determined. When Cyt and Ura
are contained in DNA or RNA complexed with proteins,
photohydration is a reaction likely to be competitive with
formation of opened ring conjugates. Photohydrates are ma-
jor products produced when Cyt is irradiated in the context
of DNA (contribution by Cerutti and Vanderhoek, in Patrick
and Rahn (57)). However, the ring-opening photoreactions
with lysine in such complexes could be significantly more
competitive with photohydration than would be the case in
solutions of free nucleobase and amino acid, as diffusion
together of reaction partners would not be a factor. Also, the
effective concentration of water in the neighborhood of a
Cyt residue at the nucleic acid–protein interface in such a
complex should be considerably smaller, thus lowering the
efficiency of formation of photohydrates. If cross-links in-
volving Cyt moieties in DNA and lysine residues in DNA-
binding proteins are formed (e.g. histones) and if these cross-
links are stable for a period of time comparable to that in
which, for example, transcriptional events in chromatin oc-
cur, then there could be biological ramifications. In partic-
ular, transcriptional and replication processes requiring dis-
placement of protein or conformational changes in a DNA–
protein complex could be affected by the presence of such
cross-links. Alternatively, ring closure would lead to loss of
a Cyt residue in the DNA sequence, accompanied by con-
Ϫ
4
1
.4 ϫ 10 . One way by which such a reaction might occur
is shown in Fig. 3c, in which an amine nucleophile is gen-
erated via transfer of a proton to C*, followed by reaction
ϩ
of the resulting protonated cytosine excited state CH* with
unprotonated amine; this latter step could be a combination
reaction of the participants in the proton transfer within the
solvent cage in which they are produced. This type of pro-
cess has attractive aspects, as it would explain the drop-off
in reactivity as the pH increases from 9 to 11.3 as being due
to decrease in the amount of protonated amine as pH in-
creases. In this case the overall mechanism would consist of
parallel pathways described by Fig. 3b and Fig. 3c, with
pathway 3c dominating at pH 9 and the slower reaction of
cytosine itself with unprotonated amine (pathway 3b) be-
coming the dominant reaction pathway at high pH values.
However, even beyond the requirement that C* react with
ethylamine (protonated or nonprotonated) upon every en-
counter between these two species, this mechanistic descrip-
tion in its present form has severe limitations as it does not
explain the decrease in relative quantum yield as the pH
moves from 9 to 5.6, as shown in Fig. 2.
It is possible that a completely different process could be
an important contributor to formation of Ia. For example, it

252 Kellie Hom et al.
1
1
7. Wu, G.-W., Z. Hillel and C. S. Park (1983) A rapid mixing–
photocrosslinking technique to study the dynamics of nucleic
acid–protein interactions. Anal. Biochem. 128, 481–489.
8. Jeppesen, C., K. F. Jensen and P. E. Nielsen (1988) A specific
and efficient photoreaction between E. coli RNA polymerase
version of a charged lysine residue in the protein partner
into an uncharged moiety (in which a Cyt or Ura moiety
incorporates the ⑀ amino group). Thus, ring closure reactions
that would remove DNA–protein cross-links could also have
biological significance.
and T
in the lac UV5 or deo P1 promoter. Nucleic Acids Res.
ϩ1
1
6, 9545–9555.
1
2
9. Anderson, R. A., Y. Nakashima and J. E. Coleman (1975)
Chemical modifications of functional residues of fd gene 5
DNA-binding protein. Biochemistry 14, 907–17.
0. Merrill, B. M., K. R. Williams, J. W. Chase and W. H. Kon-
igsberg (1984) Photochemical cross-linking of the Escherichia
coli single-stranded DNA-binding protein to oligodeoxynucleo-
tides. J. Biol. Chem. 259, 10850–10856.
1. Hockensmith, J. W., W. L. Kubasek, W. R. Vorachek and P.
H. von Hippel (1986) Laser cross-linking of nucleic acids to
proteins: methodology and first applications to the phage T4
replication system. J. Biol. Chem. 261, 3512–3518.
2. Merrill, B. M., K. L. Stone, F. Cobianchi, S. H. Wilson and K.
R. Williams (1988) Phenylalanines that are conserved among
several RNA-binding proteins form part of a nucleic acid bind-
ing pocket in the A1 heterogeneous nuclear ribonucleoprotein.
J. Biol. Chem. 263, 3307–3313.
Acknowledgments Research support from the NIH (grant
GM23526) is gratefully acknowledged. Also acknowledged is the
Bio-organic, Biomedical Mass Spectrometry Resource (A. L. Bur-
lingame, Director), supported by NIH Division of Research Resourc-
es Grant RR 01614. K.H. acknowledges the helpful advice and en-
couragement of Drs. Vladimir Basus, Lech Celewicz, Lee-Hong
Chang, Phyllis Kosen, David Maltby, Anthony Shaw and the late
Devarajan Sundaraman. This paper is based in part on a dissertation
submitted by K.H. in partial fulfillment of the requirements for the
Ph.D. in Pharmaceutical Chemistry at the University of California,
San Francisco.
2
2
REFERENCES
1
. Friedberg, E. C., G. C. Walker and W. Siede (1995) DNA Re-
pair and Mutagenesis. ASM Press, Washington, DC.
. Cadet, J. and P. Vigny (1990) The photochemistry of nucleic
acids. In Bioorganic Photochemistry, Vol. 1 (Edited by H. Mor-
rison), pp. 1–272. John Wiley and Sons, New York.
2
3. Bennett, S. F., O. N. Jensen, D. F. Barofsky and D. W. Mos-
baugh (1994) UV-catalyzed crosslinking of Escherichia coli
Ura-DNA glycosylase to DNA: identification of amino acid res-
idues in the single-stranded DNA binding site. J. Biol. Chem.
2
3
. Shetlar, M. D. (1980) Cross-linking of proteins to nucleic acids
by ultraviolet light. Photochem. Photobiol. Rev. 5, 105–197.
. Saito, I. and H. Sugiyama (1990) Photoreactions of nucleic acids
and their constituents with amino acids and related compounds.
In Bioorganic Photochemistry, Vol. 1 (Edited by H. Morrison),
pp. 317–340, John Wiley and Sons, New York.
2
69, 21870–21879.
2
4. Prasad, R., A. Kumar, S. G. Widen, J. R. Casas-Finet and S. H.
Wilson (1993) Identification of residues in the single-stranded
DNA-binding site of the 8 kDa domain of rat DNA polymerase
4
␤
by UV cross-linking. J. Biol. Chem. 268, 22746–22755.
5. Connor, D. A., A. M. Falick, M. C. Young and M. D. Shetlar
1998) Probing the binding region of the single-stranded DNA-
binding domain of rat DNA polymerase using nanosecond-
2
5
6
7
. Meisenheimer, K. and T. Koch (1997) Photocross-linking of
nucleic acids to associated proteins. Crit. Rev. Biochem. Mol.
Biol. 32, 101–140.
. Martinson, H. G., M. D. Shetlar and B. J. McCarthy (1976)
Histone–histone interactions within chromatin: cross-linking
studies using ultraviolet light. Biochemistry 15, 2002–2007.
. Kunkel, G. R. and H. G. Martinson (1978) Histone–DNA in-
teractions within chromatin: isolation of histone from DNA–
histone adducts induced in nuclei by UV light. Nucleic Acids
Res. 5, 4263–4272.
(
␤
pulse laser-induced cross-linking and mass spectrometry. Pho-
tochem. Photobiol. 68, 299–308.
6. Shetlar, M. D., J. Christensen and K. Hom (1984) Photochem-
ical addition of amino acids and peptides to DNA. Photochem.
Photobiol. 39, 125–133.
2
2
7. Shetlar, M. D., K. Hom, J. Carbone, D. Moy, E. Steady and M.
Watanabe (1984) Photochemical addition of amino acids and
peptides to homopolyribonucleotides of the major DNA bases.
Photochem. Photobiol. 39, 135–140.
8. Sperling, J. and R. Sperling (1978) Photochemical cross-linking
of histones to DNA nucleosomes. Nucleic Acids Res. 5, 2755–
2
8. Shetlar, M. D., J. Carbone, E. Steady and K. Hom (1984) Pho-
tochemical addition of amino acids and peptides to polyuridylic
acid. Photochem. Photobiol. 39, 141–144.
2773.
9
. Mandel, R., G. Y. Kolomiitseva and J. G. Brahms (1979) DNA–
protein interactions in nucleosomes and in chromatin: structural
studies of chromatin stabilized by ultraviolet-light-induced
crosslinking. Eur. J. Biochem. 96, 257–266.
2
9. Smith, K. C. (1969) Photochemical addition of amino acids to
14
C-uracil. Biochem. Biophys. Res. Commun. 34, 354–357.
3
0. Schott, H. and M. D. Shetlar (1974) Photochemical addition of
1
1
1
0. Cao, T. M. and M. T. Sung (1982) Ultraviolet light induced
preferential cross-linking of histone H3 to deoxyribonucleic acid
in chromatin and nuclei of chicken erythrocytes. Biochemistry
amino acids to thymine. Biochem. Biophys. Res. Commun. 59,
1
112–1116.
1. Shetlar, M. D., H. N. Schott, H. G. Martinson and E. T. Lin
1975) Formation of thymine–lysine adducts in irradiated DNA–
3
3
21, 3419–3427.
(
1. Kurochkina, L. P. and G. Y. Kolomiitseva (1985) Participation
of histones H3 and H1 in the formation of a thymine–lysine
cross-link in UV irradiation of deoxyribonucleoprotein. Bio-
chem. SSR 50, 1219–1223.
2. Kurochkina, L. P. and G. Y. Kolomiitseva (1989) Isolation of
modified histone H3 from ultraviolet-irradiated deoxyribonu-
cleoprotein by reversed-phase high performance liquid chro-
matography. Anal. Biochem. 178, 86–92.
lysine systems. Biochem. Biophys. Res. Commun. 66, 88–93.
2. Saito, I., H. Sugiyama, S. Ito, N. Furukawa and T. Matsuura
(1981) A novel photoreaction of thymidine with lysine. Photo-
induced migration of thymine from DNA to lysine. J. Am.
Chem. Soc. 103, 1598–1600.
33. Saito, I., H. Sugiyama and T. Matsuura (1983) Isolation and
characterization of a thymine–lysine adduct in UV-irradiated
nuclei. The role of thymine–lysine photoaddition in photo-cross-
linking of proteins to DNA. J. Am. Chem. Soc. 105, 6989–6991.
34. McGhee, J. D. and G. Felsenfeld (1980) Nucleosome structure.
Annu. Rev. Biochem. 49, 1115–1156.
35. Kurochkina, L. P., A. A. Komissarov and G. Y. Kolomiitseva
(1987) Localization of the lysine residue in histone H3 forming
a thymine–lysine cross-link when deoxyribonucleoprotein is ir-
radiated with UV light. Biochem. SSR 52, 1457–1461.
36. Saito, I., H. Sugiyama and T. Matsuura (1983) Photoreaction of
thymidine with alkylamines. Application to selective removal of
thymine from DNA. J. Am. Chem. Soc. 105, 956–962.
37. Shetlar, M. D., K. Hom, S. Distefano, K. Ekpenyong and J.
Yang (1988) Photochemical reactions of cytosine and 5-meth-
1
1
1
3. Kurochkina, L. P. and G. Y. Kolomiitseva (1996) UV cross-
linking of the globular domain of histone H1 with DNA in de-
oxyribonucleoprotein. Biochem. Moscow 61, 757–763.
4. Markovitz, A. (1972) Ultraviolet light-induced stable complexes
of DNA and DNA polymerase. Biochim. Biophys. Acta 281,
522–534.
5. Strniste, G. F. and D. A. Smith (1974) Induction of stable link-
age between the deoxyribonucleic acid-dependent polymerase
and the d(A–T) :d(A–T) by ultraviolet light. Biochemistry 13,
n
n
485–493.
1
6. Hillel, Z. and G.-W. Wu (1978) Photochemical cross-linking
studies on the interaction of Escherichia coli RNA polymerase
with T7 DNA. Biochemistry 17, 2954–2961.

Photochemistry and Photobiology, 2000, 71(3) 253
ylcytosine with methylamine and n-butylamine. Photochem.
Photobiol. 47, 779–786.
8. Shetlar, M. D., R. B. Rose, K. Hom and A. A. Shaw (1991)
pyrimidines. Determination of the lowest excited singlet and
triplet state ionization constants. J. Chim. Phys. 81, 21–31.
47. Burr, J. G., E. H. Park and A. Chan (1972) Nature of the re-
active species in the photohydration of uracil and cytosine de-
rivatives. J. Am. Chem. Soc. 94, 5866–5872.
48. Perrin, D. D. and B. Dempsey (1974) Buffers for pH and Metal
Ion Control. Chapman and Hall, London.
49. Shetlar, M. D., J. A. Taylor and K. Hom (1984) Photochemical
exchange reactions of thymine, uracil and their nucleosides with
selected amino acids. Photochem. Photobiol. 40, 299–308.
3
3
4
Ring opening photoreactions of 5-bromouracil and 5-bromo-2
deoxyuridine with selected alkylamines. Photochem. Photobiol.
3, 595–609.
9. Dorwin, E. L., A. A. Shaw, K. Hom, P. Bethel and M. D. Shetlar
1988) Photoexchange products of cytosine and 5-methylcyto-
sine with N- -acetyl-l-lysine and l-lysine. J. Photochem. Pho-
Ј-
5
(
␣
tobiol. B 2, 265–278.
0. Hom, K. (1991) Photochemical reactions of 2Ј-deoxycytidine,
5
0. Ruzsicska, B. P. and D. G. E. Lamaire (1995) DNA photochem-
istry. In CRC Handbook of Organic Photochemistry and Pho-
tobiology (Edited by W. M. Horspool and P.-S. Song), pp.
2Ј-deoxyuridine and related pyrimidine nucleosides and nucleo-
bases with ethylamine. Ph.D. Dissertation, University of Cali-
fornia, San Francisco.
1
289–1315. CRC Press, Boca Raton, FL.
5
5
5
1. Zuo, Z.-H., S.-D. Yao, J. Luo, W.-F. Wang, J.-S. Zhang and N.-
Y. Lin (1992) Laser photolysis of cytosine, cytidine and dCMP
in aqueous solution. J. Photochem. Photobiol. B 15, 215–222.
2. Wang, W.-F., J. Luo, Z.-H. Zuo, S.-D. Yao, Z.-D. Zheng, J.-S.
Zhang and N.-Y. Lin (1993) Novel transient species of DNA
bases from laser photolysis. Sci. China B 36, 1466–1475.
3. Lin, N. Y. (1996) Excitation transfer, charge transfer and hy-
drogen abstraction reaction, kinetic studies of modification of
DNA components. Radiat. Phys. Chem. 47, 209–212.
4
1. Doub, L. and U. Krolls (1970) Dihydrocytosines and cytosines
from 3-aminopropionitriles. J. Heterocycl. Chem. 7, 527–535.
2. Yamauchi, K. and M. Kinoshita (1973) N-alkylation of thymine
and uracil with trialkyl phosphates. J. Chem. Soc. Perkin Trans.
I, 391–392.
4
4
4
3. Pecorari, P., G. Vampa, A. Albasini, M. Rinaldi, M. Melegari
and M. P. Costi (1988) Preparation and physiochemical prop-
erties of uracil derivatives with potential biological activity. Il.
Farm. Ed. Sci. 43, 311–318.
4. Fox, J. J., D. van Praag, I. Wempen, I. L. Doerr, L. Cheong, J.
E. Knoll, M. L. Eidenoff, A. Bendich and G. B. Brown (1959)
5
4. Murov, S. L., I. Carmichael and G. L. Hug (1993) Handbook
of Photochemistry, 2nd ed. Marcel Dekker, New York.
5. Dreyfus, M., O. Bensaude, G. Dodin and J. E. Dubois (1976)
Tautomerism in cytosine and 3-methylcytosine. A thermody-
namic and kinetic study. J. Am. Chem. Soc. 98, 6338–6349.
6. Broo, A. and A. Holm e´ n (1997) Calculations and characteriza-
tion of the electronic spectra of DNA bases based on ab initio
MP2 geometries of different tautomeric forms. J. Phys. Chem.
A 101, 3589–3600.
5
Thiation of nucleosides. II. Synthesis of 5-methyl-2
tidine and related pyrimidine nucleosides. J. Am. Chem. Soc.
1, 178–187.
Ј-deoxycy-
8
5
4
5. Shaw, A. A. and M. D. Shetlar (1990) 3-Ureidoacrylonitriles:
novel products from the photoisomerization of cytosine, 5-meth-
ylcytosine and related compounds. J. Am. Chem. Soc. 112,
7737–7742.
57. Patrick, M. and R. O. Rahn (1976) Photochemistry of DNA and
polynucleotides: photoproducts. In Photochemistry and Photo-
biology of Nucleic Acids, Vol. 2 (Edited by S. Y. Wang), pp.
35–95. Academic Press, New York.
4
6. Parkanyi, C., D. Bouin, D.-C. Shieh, S. Tunbrant, J.-J. Aaron
and A. Tine (1984) The effect of pH on the electronic absorp-
tion, fluorescence and phosphorescence spectra of purines and