Extensive photodimerization of non-adjacent pyrimidines

J. Mol. Hiol. (1989) 210, 869-874

LETTERSTOTHEEDITOR

Extensive Photodimerization

of Non-adjacent

Pyrimidines

In a prior study we found that non-adjacent thymidyl residues in the single-stranded alternating copolymer poly[d(G-T)] are subject to photodimerization by germicidal lamp irradiation (A,,,,, 254 nm). The maximum yield of this photoproduct was 1% of the total in this polymer is increased thymine of poly[d(G-T)]. W e now report that dimer formation to 10 to 40% thymine as dimer between non-adjacent pyrimidines, using near-ultraviolet 310 nm) with or without acetone triplet-sensitization. As previously irradiation (A,,, observed for 254 nm irradiation, dimer formation was nearly absent in double-stranded poly[d(G-T) ’ d(C-A)]. These observations extend prior findings by demonstrating high-yield dimerization between non-adjacent pyrimidines via direct irradiation at environmentally relevant wavelengths (2280 nm), and are potentially relevant to the mechanism of the ultraviolet light-induced targeted - 1 frameshift mutation.

Ultraviolet light (u.v.~) induced thyminethymine cvclobutane dimers (TAT) are readily detected “in the single-stranded alternating co-polymer poly[d(G-T)], but not in the double-stranded alternating copolymer poly[d(G-T) . d(C-A)] (Nguyen & Minton, 1988). The lack of adjacent thymidyl residues in poly[d(G-T) . d(C-A)] apparently precludes T* T formation. Single-stranded poly[d(G-T)], however, may be sufficiently flexible to allow apposition of thymidyl residues, separated by one or more intervening residues so t’hat T^T may occur, as follows:

using near-u.v. (without acetone) is lo?& thymine as T^ T (Fig. 1 ), as compared to l:& thymine as T* T using far-u.v. (Nguyen & Minton, 1988). As previously observed for far-u.v.. dimer formation by near-u.v. is strongly suppressed in double-stranded poly[d(G-T) .d(C-A)]. I ncreased dimer formation in single-stranded poly[d(G-T)] by near-u.v., as compared to far-u.v., is attributable to an altered photosteady-state at longer wavelengths. The plateau value of dimers in DNA at’ high fluences is an equilibrium between u.v.-induced dimer formation and u.v.-induced dimer reversal to monomer. Reversal is favored at shorter wavelengths (-240 nm). whereas formation is favored at longer wavelengths (-280 nm) (Deering & Set’low, 1963; Fisher & Johns, 1976: Radany et al., 1981). The presence of acetone during near-u .v exposure increased dimer formation in poly[d(G-T)] to 30?/, thymine as T^ T (Fig. 1; 10,000 J/m2 equivalent to 12 min irradiation). Longer near-u.v. exposures in the presence of acetone (up to 20 min). produced dimer yields up to 40 ‘$A; however, at time point’s after approximately 15 minutes of irradiation, dimer format,ion in poly[d(G-T) . d((‘-A)], was also detected. Since acetone was added immediately prior to irradiation, formation of dimers in poly[d(G-T) . d(C-A)] after 15 minutes was presumably due to progressive denat,uration of the polymer in the presence of the high acetone concentration (not shown). The mechanism of photosensitization successively involves U.V. absorption by acetone, intersystem crossing of the acetonr-excited state from a singlet to triplet, donation of the triplet to DNA, and decay of the DNA t,riplet to produce cyclobutane dimers (Lamola. 1970). As this process is efficient,, and there is little or no photoreversal of cyclobutane dimers at near-u.v. wavelengths. dimerization proceeds to very high levels, until dimer-forming sites are saturated (Lamola, 1970; Rahn & Patrick, 1976; Masnyk et al.. 1989). The chromatographic system employed in Figure 1 provides sensitive quantific*ation of dimers.

In prior studies that employed germicidal lamp irradiation (A,,, 254 nm), T^ T amounbed to a maximum of 1 Y. of the total thymine in poly[d(G-T)] (Nguyen & Minton, 1988). This letter describes formation of T* T between non-adjacent thymidyl residues in much higher yield using near-u.v. ( 2 280 nm: i,,,, 310 nm) in either t’he presence or absence of the triplet-sensitizer acetone. In the presence of acetone the major T^T isomer in poly[d(G-T)] is shown to be cis-syn, with lesser amounts of the tram-syn. In poly[d(G-T)] exposed to near-u.v. without acet’one only the cis-syn isomer is detected. Figure 1 shows the dose dependence of T&T formation in containing poMd(G-T)l> [methyZ-3H]thymine, exposed to near-u.v. in the presence or absence of acetone. Following irradiation the samples were acid-hydrolyzed t’o release constituent bases, and the hydrolysates were chromatographically resolved to separate T^ T from monomeric thymine. The maximum dimer yield t Abbreviations thin-layer cyclobutane

used: chrom&ography; dimer: TFA,

0022~~2~36/~9/240~69-06

u.v..

ultraviolet light; TA T, thymine-thyminr trifluoroacetic acid.

$03.00/O

TLC,

869

Q

1989 Aradernic~

l’rrss

Iknited

II. T.

870

0

50

too Fluence

0

0.5

(J/m*

and K. W. Mintm

Nguyen

I

x 10w4)

Figure 1. Dose dependence of pyrimidine dimer formation in nucleic acids exposed to near-u.v. with or without arrtonr. Kucleic acid samples contained Inrrthy/-3H]thvminr; Cr.5pg DNA in I ml of 100 mM-NaCl. 5 mu-sodium phosphate (pH 7.0) was irradiated at room temperature in either (a) the absence or (b) presence of 20°,c acetone (v/v). The samples were in flame-sealed thinwalled glass 13 mm x 100 mm test, tubes and were not deoxygenated. The emission spectrum of the near-u.v. bulbs (Fotodgne). adjust,ed for att’enuation through one wall of the test tube. was a bell-shaped curve from 280 nm to 360 nm. with a single maximum at 310 nm. The incident iutjensity on the sample n-as 13 W/m’. After irradiation. samples were hydrolyzed with trifluoroacetie acid (TFA). which degrades DNA to its ronstit’uent bases and other by-product,s (Patrick B Rahn. 1976). The hydrolysates were chromatographed to separate T’ T from monomeric thynine using silica gel TLC’ developed with watersaturated ethyl aretat)e/n-propanol (4 : 1. v/v). The plat,e was fractionated and radioact,ivity determined by scint,illation counting. For further details see Nguyen & Minton (198X). Samples: poly[d(G-T)] (m): poly[d(G-T) .d(C-A)] (0): native salmon sperm DKu’A (A).

but poorly separates photoproducts of differing structure (Nguyen & Minton, 1988). Additional st’udies addressed structural identification of the bipyrimidine photoproduct’s occurring in

poly[ d(G-T)] following near-u.v. irradiation with or wit’hout acetone (Figs 2, 3 and 4: Table 1). Using far-u.v.. the primary T^ T isomer is cis-~yl/ with small amounts of trans-syn and &s-an& isomers (Nguyen & Minton. 1988: isomers reviewed in this reference); in contrast. using near-u.v. (no acetone) the only T^ T isomer in poly[d(G-T)] was rix-syrr (Figs 2 and 3; Table 1). One speculat,ive explanation for this observation is as follows: at longer exciting wavelengths there is diminished intrrsyst~em crossing from the lowest, singlet state (S,) to the lowest triplet stat’e (T,) in the polymer (Fisher & Johns. 1970; Lamola &, Eisinger, 1971; Sutherland B Sutherland, 1970: Rahn & Patrick, 1976: Masnyk et a’l., 1989). The singlet lifetime is no longer than nanosecondsin aqueous DN4 at’ room t,emperature. while the triplet lifetime is microseconds or longer (Masnyk et al.. 1989, and references cit)ed therein). Studies on free and aggregated thyminr in solution indicate t’hat dimer formation from t,he short -lived SI is limited t,o pyrimidines that are aggregated at the t,ime of excit*ation, while dimer formation from T, includes pyrimidines that are distant at’ the t’irne of singlet st’ate excitation and intersystem crossing to T, but diffuse towards each other within the relatively long T, lifetime (Fisher & Johns, 1970. 1976). Tf dimrr formation in poly]d((LT)] b? near-u.v.. as opposed tjo far-u.v.. is only via thr singlet state (due t)o diminished intjersyst’em crossing), only stably associated thymidyl residues (equivalent to aggregates) will dimerizr. It is likely that t,he most stable association of two thymidyl residues in poly]d(G-T)] is basestjacking of thymidyl residues with a single intervening guanidyl residue looped out of the base stack into solution. Tt is estimated that 4(5, of thymidyl residues in poly]d(G-T)] are in this configuration at any given time (Nguyen dt Minton. 1988). Tf singlet sta,te dimer formation is limited to this association. only the cis-syr~ isomer will be produced, as is found. If far-u.\-. irradiation of poly[d(G-T)] produces a

Table 1 l’hymine-thymine

Irradiation

Sample Pal\-[d(GT)I$ Poly[d(G-T)]j Poly[d(Q-T)]: Thyminr (lo-’ M) Thymine (10 -’ M) Sative DNA11 Denatured DNA11 Satire DNA11 Iknatured DNq t

A. cys-syn;

13, trans-syn;

C’, cis-anti;

(mn)

cyclobutane

Triplet sensitizer

A

NOW None Acetone (209o,. v/r) Acetone (JO”/o. v/v) Acetone (83%. v/r) None None Acetophenone (2 mu) Acetophenone (2 mx)

87 100 86 2% 20 >QQ 91 100

11 Escherichia

coli DNA.

x7

isomer yields

Ngup1

$ hlir1tot1

( testqg

This works This works .Jennings et (II. (1970) Jennings et (11.(1970) Ron-Hur & Ren-Ishai (1968) Ren-Hur & Ren-Ishai (1968) Rahn & Landry (1971) Kahn &, Landry (1971)

I). tmns-an&

$ Single-stranded alternating copolymer. 9: Values derived by transmittance densitometry of autoradiogram work.

dimer

shown in Fig..Jo of Nguyen & .Minton (I$+#) and Pig. 3 of this

871

Letters to the Editor

r

TAT(A)

lb)

PyC -

T”C -I

0 I

20

40

60 Fraction

I

IO

20

30

number

Figure 2. Separation of dimer species in poly[d(G-T)] exposed to near-uv. (a) Poly[d(G-T)]. labeled with [methyL3H]thymine, was irradiated as for Fig. 1 and TFA-hydrolyzed, and the hydrolysate was developed on Whatman no. 3 filter paper by descending chromatography with n-butanollacetic acid/water (80 : 12 : 30. by vol.). Fractions were cut. eluted into 1 ml water and liquid samples were scintillation counted. The major photoproduct peak co-chromatographed with the cis-syn T^ T and PyC markers. To separate these products eluates of fractions 17 to 20 were pooled. concentrated and resolved in the system shown in (b); consisting of Whatman no. 3 filter paper developed by descending chromatography with t-butanollmethyl ethyl ketone/water/58 o/o ammonium hydroxide (4 : 3 : 2 : 1, by vol.). The paper was cut into fractions and scintillation counted. All radioactivity co-chromatographed with cis-syn T*T marker. For further details of procedure see Nguyen & Minton (1988). Markers: T^ C, thymine-cytosine cyclobutane dimer from native salmon sperm DNA; T T(A), cis-syn thymine-thymine cyclobutane dimer; PyC, 64’-[pyrimidin-2’-one]-thymine; T^ T(B), trans-syn thymine-thymine cyclobutane dimer; T. thymme. All photoproduct markers in Figs 2, 3 and 4. including T^T isomers. were obtained as described (Nguyen & Minton, 1988). In addition. T^T isomers were also generated by near-u.v. irradiation of lo-’ M-[methyL3H]thymine solution, 2Oqi, (v/v) acetone, which produces the cis-syn, trans-syn, &-anti and trans-anti isomers of T* T (Jennings et al., 1970). Samples: unirradiated poly[d(G-T)] (0); polg[d(G-T)] irradiated in the absence of acetone to 7 x lo5 *J/m’ (A); poly[d(G-T)] irradiated in the presence of ZOO/,, acet’one (v/v) to 1.7 x lo4 J/m’ (m).

significant fraction of dimers via T,, formation of T^T is not constrained to the association noted above, and might include less-frequent contacts yielding the tram-syn and G-anti isomers, as is found (Table 1). Near-uv. radiation of poly[d(G-T)] in the presence of the triplet-sensitizer acetone yielded cis-syn and tmns-syn isomers (Figs 2, 3 and 4; Table 1). As dimer formation under these conditions is almost solely via the long-lived triplet state, and would be expect’ed to include less-common isomeric form(s), it is noteworthy that no cis-anti isomer was detected. A potential explanation is that the reduction in solution dielectric constant by the addition of 20% acetone diminishes base-stacking interaction (Jennings et al., 1970). The &-anti T^ T isomer is likely to arise between bases from different strands, that are transiently intercalated via a weak hydrophobic interaction (Nguyen & Minton, 1988), which may be abolished by acetone. Both the cis-syn and tram-syn isomers are expected to arise from nearby bases t’hat are covalently linked to each other via

the sugar-phosphate backbone; t,he drop in solution polarity might not greatly affect t,he frequency with which these residues encounter each ot~her. In prior studies using 254 nm u.v., the relatively low maximum yield of 1 “1; thymine as T^ T in poly[d(G-T)] necessitated a variety of control experiments to rule out alternative explanations for dimer formation, such as the presence of adjacent pyrimidines contaminating the alternating copolymer sequence (Nguyen & Minton, 1988). Although the same control studies as previously performed (Nguyen & Minton, 1988) were carried out during the present study (not shown), their importance is greatly diminished. The much higher yield of T^ T using near-u.v.. either with or without triplet-sensitization, eliminates the possibility that rare undetected anomalies in the poly[d(G-T)] substrate could be responsible for cpclobutane dimer formation, and strengthens the conclusion that such dimers must arise between non-adjacent thymidyl residues. We have previously suggested t’hat 1)NA synthe-

872

H. T. Nguyen and K. W. Minton,

T

-T-

B -D

A

_

-

-A-

0

22

3

Figure 3. T”T isomers in poly[d(G-T)] exposed t o near-u.v. Poiy]d(G-T)], labeled with [metklll-3H]thynlirle. was irradiat,ed as for Fig. 1. The samples were TFA-hydrolyzed and the hydrolysate was developed on cellulose F TLC (Merck) with n-butanollacetic acid/water (80: 12: 30. by vol.). The plate was sprayed with En3Hance (New England Nuclear) and fluorographed overnight. Markers: A. cis-syn T’ T isomer; B. trcc.ns-syn T’ T isomer: C, c&anti T” T isomer; D, frays-anti TT isomer; T. thymine monomer. Samples: M, markers; 1, unirradiated poly[d(G-T)]. 5 x IO5 atsjmin; 2. poly[d(G-T)] irradiated without acetone to 75 x lo5 J/m’. 5 x 10’ &s/min: 3, poly[d(G-T)] irradiated in 20:/, acetone (v/v) to 8.3 x IO3 J/m2, 2 x lo5 cts/min; 4, T-T(A) peak eluted from Fig. 2(a), approx. 25,000 cts/min; 5. TAT(B) peak eluted from Fig. 2(a), approx. 25,000 cts/min.

sis proceeding non-adjacent

across an unrepaired pyrimidines might

dimer between result in a - I

frameshift mutation, since the int,ervening residue is physically excluded from base-stacking. and conseyuently ma\- not’ be apparent to the replication apparatus. This mechanism could occur in Z&IO in chromosomal regions with single-stranded charac-

ter, such as highly transcribed Minton. 1988). The carcinogenic

genes (Nguyen Hr effects of solar

radiation are caused by wavelengths between 280 and 320 nrn (Harm,

between near-u.v. process.

1980); t.he occurrence

of dimers

non-adjacent thymidyl residues at wavelengths may be relevant to this

Figure 4. Photoreversal of T’T isomers in poly[d(G-T)] exposed to near-u.v. Pyrimidine dimers. unlike pyrimidinr adducts. are reversible to thymine monomers by additional U.V. irradiation at 254 nm: and this photoreversibility serves as a test for their presence (Patrick & Rahn, 1976). Bands corresponding to cis-syn and tram-syn T’ T isomers in lane 3 of Fig. 3 were scraped off the TLC plate, eluted int.o water. concent.rated and developed on silica gel G60 TLC (Merck) plates in the first dimension with the lower layer of chloroform/methanol/water (4 : 2 : 1, by vol.) to which 5 ml of absolute met’hanol had been added per 100ml of lower phase: and developed in the second dimension with ethyl acetate/ isopropanol/water (75 : 16 : 9, by vol.). The plate was sprayed with En3Hance and fluorographed overnight,. For details of procedure, see Nguyen & Minton (1988). Markers: Open circles are the locations of markers, designat,ed A to I). and T. as in Fig. 3. The vertical line bisecting the thymine monomer marker indicates the mobility of thymine monomer in the 2nd dimension. Samples: Upper left., the cis-syn (A) band from Fig. 3, lane 3; approx. 35,000 cts/min total. Upper right: Same as upper left, except that following chromatography in the 1st dimension, the plate was exposed to 25,009 J/m2 of 254 nm u.v. using a germicidal lamp (General Electric G8T5) with a dose rate of 10 W/m2 to phoboreverse thymine dimers to monomers. The plate was then developed in the 2nd dimension. Note that A is absent, while a new spot with the mobility of thymine (arrow) is present. Lower left, t,he trans-ayn (B) band from Fig. 3, lane 3; approx. 35,000 &/nun total. Lower right, same as lower left. except that following chromatography in the 1st dimension, the plate was exposed to 25,000 J/m2 of 254 nm U.V. to photoreverse thymine dimers to monomers. The plate was then developed in the 2nd dimension. Kate that B is absent while a new spot with the mobility of thymine (arrow) is present,. The same results as illustrated in this Figure were also obtained for the cis-syn (A) isomer from .poly[d(G-T)], irradiated in the absence of acetone and purified in the system described for Fig. 3 (not shown).

873

Letters to the Editor

4

T

D

D

0 0 0 0 ‘_

0

C

a

A

@

2nd

C

0

/

c‘

0

J rigin

0

:-

b

dimension

Origin

2nd

\

1

A

T

r

\

D

/ D

0

0

B

B

C A

C

0

‘T>

I rigin

2nd

0

A

0

0 x ”

0

0

dimension

Origin

Fig. 4.

l

dimension

2nd

dimension

H. T. Nguyen

874 This work was supported

by IXUHS

and K. W. Minton

grant R074AK.

Harm,

WI. (1980). Radiation,, pp.

In Riological Eflects of 17Ztm&olet 18&195, Cambridge University Press.

Hong T. Nguyen Kenneth W. Mintont

Cambridge. Jennings, B..

Department of Patholog) F. E. HBbert School of Medicine [Tniformed Services ITniversitp of the Health Bethesda. Maryland 20814.4799,U.S.A.

Lamola, A. Appl. Chem. 24, 599~-610. Lamola, A. & Eisinger. ,J. (1971). Biochirn. Biophyls. Acta. 240, 313-325. Masnyk. T.. Nguyen, H. & Minton. K. (1989). ./. Ljiol. Chem. 264, 2482-2488. Nguyen. H. & Mint,on, K. (1988). J. ,220l. Viol. 200. 681~-693. Patrick, M. & Rahn. R. (1976). In Photochemistry and Photobiology of Nucleic Acids (Wang. S. Y.. rd.), vol. 2, pp. 35-95. Academic Press, New York. Radany. E.. Love, ,J. &, Friedberg. E. (1981). In Chromosome Damage and Rep&r (Seeberg. E. 8 Kleppr. K.. eds). pp. 91-95, Plenum Publishing Corp., Sew York. Rahn, R. & Landrp. I,. (1971). Biochin~. Hiopllp. Acta. 247. 1977206. Rahn, R. & Patrick, M. (1976). In Phjotochemiatry and Ph,otob%ology of Nucleic Acids (Wang, S. Y.. ed. ). vol. 2, pp. 97-145, Academic Press. Kew York. Sutherland. B. K: Sutherland. J. (1970). Biopolymers. 9.

Received 20 March

Photochem

Sciences

1989. and in revised form 11
1989 t Author

for correspondence.

References Ben-Hur,

E. & Ben-Ishai, R. (1968). Biochim. Biophys. 166, 9-15. Deering, R. & Setlow. R. (1963). Biochim. Biophys. Acta. 68, 526-534. Fisher. G. &
429444. Fisher.

G. & Johns, H. (1976). In Photochemistry and of Nucleic Acids (Wang, S. Y.. ed.), vol. 1, pp. 225-293, Academic Press, P;ew York. Photobiology

Edited

639-653.

by S. Brenner

Pastra,

Photobiol. (1970). Pure

S. &

Wellington,

J.

(1970).

11. 215-226.