Biochimica etBiophysica Acta, 781 (1984) 257-264 257 Elsevier

BBA91327

T E R M I N A T I O N S OF DNA SYNTHESIS ON 'PROFLAVINE AND L I G H T ' - T R E A T E D ~X174 S INGLE-STRANDED DNA

JACQUES PIETTE a.., CLAIRE-MICHELLE CALBERG-BACQ a, MIGUEL LOPEZ a and ALBERT VAN DE VORST b

a Laboratories of General and Medical Microbiology and b Experimental Physics, University of Libge, Sart-Tilman, B-4000 Liege (Belgium)

(Received July 29th, 1983)

Key words: Proflavine; Blocking lesion; DATA synthesis," DNA damage; Chain termination; (0X174 DNA)

Bacteriophage ~X174 single-stranded DNA molecules were primed with five different restriction fragments and irradiated with visible light in the presence of profiavine. This photodamaged DNA was used as template for the in vitro complementary chain synthesis by E. coil DNA polymerase I (Klenow fragment). Chain terminations were observed by polyacrylamide gel electrophoresis of the synthesized products and localized by comparison with standard sequencing pedormed simultaneously on the untreated template. 90% of the chain terminations occurred one nucleotide before a guanine residue in the template strand. More than 80% of the sequenced guanine residues were blocking lesions demonstrating the absence of 'hot-spots' for the photodamaging effect of proflavine. At a defined position, the chain termination frequency increased linearly with the irradiation time and was directly influenced by the proflavine concentration present. An important part of lesions resulted from the action of singlet oxygen produced by excited proflavine as shown by the effect that both NaN s and ZH20 exerted on the reaction. The induced blocking lesions must be important in vivo since no complete replicative forms could he extracted from cell infected with bacteriophages inactivated by 'proflavine and light' treatment.

Introduction

Natural structural features or induced modifica- tions in the template chain can affect the kinetics of nucleic acid synthesis by the different poly- merases. The secondary structure of a single- stranded template acts as a kinetic barrier to delay the chain extension catalyzed by the D N A poly- merase of T4 bacteriophage [1]. In the primary structure, the highly modified uridine residue that occurs naturally in Drosophila ribosomal 18 S RNA induces the reverse transcriptase to stop c D N A

* To whom correspondence should be adressed at: Service de Microbiologie G~n~rale et M&lieale, Institut de Pathologic (B23) Universit/~ de Li/~ge Sart-Tilman, B-4000 Liege, Bel- gium.

synthesis one base before the modified nucleoside [2]. Also the presence of N 2 methylguanine causes a 3 min pause in the cDNA elongation on Escherichia coli 16 S RNA [3] and double-methyl- ated adenine in eukaryotic 18 S rRNA interrupts e D N A transcription [4]. Using damaged single- stranded D N A from bacteriophage t3X174, it is demonstrated that the D N A polymerase I from E. coli terminates D N A synthesis with no nucleotide in front of the lesion: i.e., a pyrimidine dimer after ultraviolet irradiation or a guanine-carcinogen ad- duct after N-acetoxy-2-acetyl aminofluorene treat- m e n t [5,6]. On N-acetoxy-2-aminofluorene-treated OX174 D N A templates, chain terminations occur either one nucleotide before the adduct or on the lesion itself depending upon the polymerizing en- zyme used [7,8].

0167-4781/84/$03.00 © 1984 Elsevier Science Publishers B.V.

258

The enzymatic mechanisms that govern ter- minations or pauses at specific sites on a template remain to be elucidated but it is clear that syn- chronous and uniquely primed synthesis of DNA on a treated template by E. coli polymerase repre- sents a useful technique to localize lesions in DNA. Using this approach, Piette and Moore [9] have demonstrated that the main target in DNA for the photodynamic action of the acridine dye prof- lavine are the guanine residues. This study was performed with DNA polymerase I large fragment which permits detailed studies of the polymeriza- tion reaction without the complications that arise from the 5'-3' exonuclease activity [10,11].

In this paper, it will be shown that the synthesis of the strand complementary to the OX174 viral strand is mainly blocked before a guanine residue on a proflavine photosensitized template, whatever the primer used to initiate the reaction. To in- vestigate the role that singlet oxygen (1 O2) plays in the induction of the blocking lesions, the chain termination is measured in the presence of either NaN 3, a known 102 quencher, or 2H20 which increases the ~O 2 lifetime [12]. In addition, the biological importance of the abortive DNA synthesis during phage infection is demonstrated.

Materials and Methods

Preparation of the primed single-stranded DNA from OX174 (OXssDNA). OXssDNA was prepared from the virions according to the techniques de- scribed by Moore et al. [6] or ourselves [13]. Fol- lowing phenol extraction, it was dialysed against 50 mM Tris-HC1 buffer, pH 7.5/1 mM EDTA, denatured in 50% (v/v) formamide (at 50°C for 10 min), purified by chromatography through a Sep- harose 4B column equilibrated with the above buffer, and finally precipitated with 2 vol. ethanol. QXss concentration was estimated on the basis of an E M value for the DNA phosphorus of 8250 at 260 nm and in 0.2 M NaCI.

The double-stranded DNA form of OX174 (13XRFI DNA) was prepared according to the same authors [6,13], it was digested with either restriction endonuclease HaeII or HaeIII (from Boehringer, Mannheim) and the fragments were purified as mentioned previously [5,6], or re- covered from the agarose gel by electroelution. All

the DNA materials were checked for purity by analytical electrophoresis on agarose gels [14]. To be used as template the OXssDNA was annealed with a purified denatured restriction fragment as described by Moore et al. [5,6].

Photosensitization of the primed OXssDNA. The primed OXssDNA (40 btM (P) in ssDNA) was mixed with proflavine (hemisulfate from British Drug House) at room temperature and under dim red light. The proflavine amount was varied from 4 to 30 ttM as estimated on the basis of an E M value at 444 nm of 33.500.

The mixtures were prepared and irradiated in the polymerizing buffer (i.e., 50 mM Tris-HC1 buffer, pH 8.0/5 mM dithiothreitol/8 mM MgCI2). Irradiation was performed at low fluence from 10 to 50 rain at X > 290 nm using either an HQE (Osram 42 W) or and XBO (Osram 150 W) lamp giving at irradiation position a fluence rate of 22 W. m-2.

Synthesis of the complementary strands. Poly- merase I lacking 5'-3' exonuclease activity (Klenow fragment) was purchased from Boehringer (Mann- heim). On the templates non-irradiated or irradia- ted for various periods of time, the DNA polymer- izing reaction was carried out according Piette and Moore [9] and using 1 #Ci [a-32P]dATP (800 Ci /mmol from New England Nuclear). Either HaeII or HaeIII enzyme (0.5 U) was added (for 45 min at 37°C) at the end of the reaction depend- ing upon the restriction fragment used as primer.

On a non-irradiated template, the sequencing of the DNA complementary chain was performed following the technique described by Sanger et al. [15[. The synthesized products were denatured [9] and separated by high voltage electrophoresis on either 15 or 20% polyacrylamide gels. The autora- diographs [9] were quantitatively analysed by densitometry [14].

Photosensitization of 3H-labelled OX174 phages. Phages were prepared according to the technique described earlier [13], except for the presence of [methyl-3H]thymidine (from IRE, Fleurus, Bel- gium) and the use of the host bacterium E. colic HF4733 which allowed us to grow the 13X wild type phage. The phages (1.8-1013 pfu /ml) sus- pended in borate (pH 9.2, saturated at 4°C) were first incubated overnight at 25°C and in the pres- ence of 0.45 /~M proflavine. After a 100-times

259

Synthesized strand

l,Ll ,L ,L ,L 5' G A G C C A T A C C G C T G A T T C T G C G T T T G C T G A T G A A C T A A G 3'

3' C T C G G T A T G G C G A C T A A G A C G C A A A C G A C T A C T T G A T T C 5'

Template strand (3711)

Fig. 1. Position of the chain termination sites. The template strand primed with HaelI fragment 1 was irradiated in the presence of 18.2 /~M proflavine. On the synthesized strand, the arrows indicate the positions of the bands (on a 20$ polyacrylamide gel) the intensity of which increased with the irradiation time,

dilution in borate, half of the suspension was irradiated for 3 rain at a fluence rate of 22 W • m - 2. Both irradiated and non-irradiated samples were

A C [ S T

min irradiation , , % _ _

0 5102030

Fig. 2. Polyacrylamide gel analysis (15% gel) of the products synthesized by DNA polymerase I on OXssDNA primed with HaelII fragment 4 and irradiated in the presence of 18.2 #M proflavine. Visible light irradiation was 22 W- m- 2 for 0, 5, 10, 20 and 30 rain as indicated. Lanes ACGT referred to the sequence standards synthesized by polymerase I on the un- treated primed template in the presence of the dideoxy chain terminating nucleotides. The arrow points to the lesion consid- ered in Fig. 4.

used to infect E. coli C cells: O X R F was extracted after 4 h of infection and analysed by velocity sedimentation through a 5-20% linear sucrose gradient [14].

Results

Two types of restriction fragments were used in this work in order to prime the synthesis of the complementary D N A chains. The first set was fragments 1 and 2 generated after I~XRFI diges- tion by HaelI, they allow an analysis of the re- gions from 3763 as numbered by Sanger et al. [16] (gene H) and from 1415 (gene F) with fragment 1 and fragment 2, respectively. The second set was fragments 1, 2 and 4 obtained by I~IXRFI diges- tion with HaelI I . These fragments primed D N A synthesis in the nucleotide sequences starting at 1776 (gene F) with fragment 1, at 3129 (gene H) with fragment 2 and at 1173 (gene F) with frag- ment 4. In our experimental conditions, series of about 80 nucleotides were defined on a gel, their sequences were found to be in agreement with the revised 12tX174 D N A sequence published by Sanger et al. [16].

Localization of DNA synthesis termination sites Using 121XssDNA primed with HaelI-generated

fragment 1 or 2 as substrates for the photosensiti- zation reaction mediated by proflavine, the autora- diographs revealed several bands the intensity of which increased which the irradiation time and which did not appear when the non-irradiated complexes were used (Fig. 1). These bands corre- sponded to those previously observed when IDXs- sDNA-prof lavine complexes were irradiated be-

260

Synthesized strand

5' A G T T C C A T C A A C A T C A T A G C C A A G A T G C C C A G A G A T T A 3'

3' T C A A G G T A G T T G T A G T A T C G G T T C T A C G G G T C T C T A A T 5'

Template strand (primed with fragment I) (1723)

Synthesized strand

5' A G G A A C A T T A G A G C C T T G A A T G G C A G A T T T A A T A C C A G C A 3,

3' T C C T T G T A A T C T C G G A A C T T A C C G T C T A A A T T A T G G T C G T 5'

Template strand (primed with fragment 2) (3077)

Synthesized strand

5' A T G G A G A A A G A C G G A G A G C G C C A A C G G C G T C C A T C T C G A A G G A G T C 3'

3' T A C C T C T T T C T G C C T C T C G C G G T T G C C G C A G G T A G A G C T T C C T C A G 5'

Template strand (primed with fragment 4) (1118)

Fig. 3. Position of the chain termination sites in the synthesized DNA (arrows) and the deduced lesions in the photosensitized template (*). The number (from Sanger et al. [16]) of the last nucleotide in the template region considered is indicated.

fore being annealed with the same HaelI restric- tion fragments [9]. It thus appeared that the pho- tosensitization reaction does not interfere with the annealing procedure.

when IDXssDNA primed with either HaelII- generated fragment 1, 2 or 4, was mixed with proflavine and irradiated for various periods of time, analysis of the products synthesized by DNA polymerase I on such templates revealed the ap- pearance of bands (Fig. 2). The bands whose in- tensity increased with the irradiation time oc- curred mainly one base before a cytosine residue in the synthesized strand and corresponded to a lesion at the level of the guanine residue in the template strand. The determination of the various termination sites in relation with the lesions in- duced by proflavine photoreaction are shown in Fig. 3.

From the data obtained using five different restriction fragments, it appeared that on the 42 guanines which were sequenced, 35 residues (83%) were altered by the photosensitization reaction in such a way that they were able to block the DNA polymerization. This result clearly points to the lack of 'hot-spots' for the photodamaging effect of proflavine. The percentage of chain termination before the other bases was much lower: only 10% of the cytosine residues (3 in 29 total) and 2% (1 in 45 total) of the adenine residues in the template chain could induce a stop in DNA synthesis.

Influence of the proflavine concentration on the termination frequency

The scanning of the autoradiographs indicates that the termination frequency in front of one altered site (i.e., the darkening of the correspond-

ing band) increased linearly with the irradiation time (Fig. 4A). The alteration rate of a guanine residue during photosensitization can thus be ex- pressed by the slope of the straight line obtained when the termination frequency at one residue is plotted vs. the irradiation time. This has been done when OXssDNA primed with HaelII fragment 4 was irradiated in the presence of increasing amounts of proflavine (from 4 to 27.4 #M). The alteration rate was thus found to largely increase up to a proflavine concentration equal to 18/~M; it tended to decrease in the presence of a higher proflavine amount. Fig. 4B illustrates the phenom- enon in the case of the stop site occurring at residue 1140 but scanning of the other bands gave similar results.

261

Modulation of the singlet oxygen pathway: effect of NaN 3 and ell20 on the termination frequency

When a defined photosensitization reaction was carried out in the presence of various concentra- tions of sodium azide (from 0.1 to 10 mM), the analysis of the products synthesized by DNA polymerase I revealed that the band intensity de- creased when the amount of NaN 3 present in- creased (Fig. 5). This observation was made with templates constructed either with HaelI fragment 1 or with HaelII fragment 4. From the scanning of the autoradiographs it appeared that the termina- tion frequency at all sites could be affected by the presence of NaN 3 in the irradiated mixture.

In contrast, when 2H20 was used to replace H 2 0 in the medium during the irradiation of the

=-- 80

g ~ • - 4 0 ~ , .~_ :.~

E~- 20 ...x ~- x ,

o 20 3o 5o Irradiation time (rain)

mM NaN 3

A C G f 0 0.1 1 5 10

.8 1.z / o\ 1.0

~ 0.6 o ~ _~0.4 t /

021 .< 10 20 30

PF concenlration pM

Fig. 4. Rate of termination frequency at guanine residue No. 1140. The OXssDNA was primed with HaeIII fragment 4 (see Figs. 2 and 3). The darkening of the band was expressed in arbitrary units by the height of the peak on the scanned autoradiograph. A: the termination frequency was plotted vs. the irradiation time (in min); the proflavine amount added to the primed template was 27.4 #M. B: the slope of graph A was plotted vs. the proflavine amount present (from 3.65 to 27.4 p.M).

Fig. 5. NaN 3 effect. Polyacrylamide gel analysis of the products synthesized by DNA polymerase I on OXssDNA primed with HaeII fragment 1. Photosensitization was carried out by a 30 min irradiation in the presence of 8.1 #M proflavine and the irradiation medium was supplemented with various NaN 3 con- centrations from 0.1 to 10 mM as indicated. The arrows point to bands whose intensity clearly decreased with increasing NaN 3 concentrations. Lanes ACGT are standards as in Fig. 2.

262

template, the rate of induction of DNA synthesis terminations was increased by a factor of 2 (Fig. 6: the ratio between the slopes 2 H 2 0 / H 2 0 was 2.7 in A and 1.9 in B). As above, the data were quite similar whatever the primer used.

Importance of the DNA synthesis stop sites during the OX174 phage infectious process

To investigate to what extent the D N A synthe- sis blocking lesions were important ' in vivo', E. coli cells were infected either with non-irradiated or with irradiated 3H-labelled phage-proflavine complexes and the replicative form of the infecting D N A was extracted in both cases. The proflavine amount was 0.45 ffM and after 3 min of irradia- tion the surviving phage fraction was 9%. It had been shown previously that such photodynamic treatment did not significantly modify the phage adsorption nor the D N A penetration into the host-bacteria [13]. Fig. 7 shows the sedimentation profiles of the OXDNAs extracted from the infected cells: a large peak corresponding to

X 5O

~Z,O × (a) / /

.~ lO

"5 50i (b)

£ <o o 0 5 10 15 20

IPradiafion time (min)

Fig. 6. 2H20 effect. The chain termination frequency (in arbitrary units) on a OXssDNA template primed with HaelI

fragment 2 was measured (in arbitrary units) at the level of the guanine residue No. 1388 (in A) and No. 1380 (in B). The proflavine concentration present was 8.1 ,tt M and the irradiated mixture was prepared in either 2H20 ( x - - × ) or H 2 0 (0 o).

140, / l ~- 1280

' 100 200 3 ,r- 2

£ ~o ] ~2o "

20 ~~ ,

O 10 20 Fraction number

Fig. 7. Abortive DNA synthesis 'in vivo'. Velocity sedimenta- tion profiles of OXDNA extracted from E. coil C cells infected with 3H-labelled phage treated with proflavine (0.45 ffM) and either irradiated for 3 min ( x - ×) or non-irradiated (0 0). The radioactivity (dpm) was plotted vs. the fraction number. The sedimentation was right to left.

OXRFI DNA was observed in the extracts from cells infected with the non-irradiated OX174-prof- lavine complexes, whereas this peak was absent if the phages had been irradiated. Moreover, aliquots of each fraction were analysed by agarose gel electrophoresis [14] and OXRFI bands were only detected in the fractions corresponding to the RFI peak isolated from the control cells (data not shown). It thus appeared that blocking the com- plementary chain synthesis was a major factor in the loss of infectivity of proflavine photosensitized OXssDNA.

Discussion

Primed OXssDNA is both an especially good system to study alterations in DNA synthesis [5,9.17] and a very well known substrate for the photodynamic activity of proflavine [13,18]. It is demonstrated here that numerous chain termina- tions are observed when a population of Ox- ssDNA molecules treated with proflavine and light are used as templates for E. coli polymerase I (Klenow fragment) independently of the DNA region considered. The stop sites are interpreted as localized one nucleotide before an altered guanine residue in the template since this explanation accounts for 90% of the bands observed. As the remaining terminations do not occur in front of a

• 263

template guanine they are thought to arise before another base (C and sometimes A) which would be much more rarely altered than guanine by the photodynamic treatment.

That guanine is mainly involved in chain termination induction is in agreement with the large extent of chemical modification of the guanine residues observed in DNA treated with proflavine and light [18]. This modification might involve the opening of the guanine imidazole ring and it has been shown in the case of 7-methyl- guanine that the ring-opened form of the damaged base efficiently blocks DNA chain elongation [19]. Irradiated proflavine has been demonstrated to produce singlet oxygen [20] and among the DNA constituants the guanine residues are the most sensitive to the action of this activated oxygen species [21]. Singlet oxygen is certainly involved in the photodynamic induction of the blocking le- sions as shown by both 2H20 and NaN 3 effects. Although the lifetime of 102 is 10-times higher in 2H20 than in H20 [22] the biological reactions implicating 102 are, as observed here, generally increased by a factor of 2 when 2H20 replaces H20. The chain termination induction also regu- larly decreases in the presence of increasing con- centrations of the singlet oxygen quencher N a N 3.

However, if the photodamage was due only to 102 action, the reaction would be totally suppressed by 10 mM NaN 3 [23]. This does not seem to be the case so that the other photochemical pathway ini- tiated by excited proflavine might also be respon- sible in part for the induction of the blocking lesions. There are indeed indications that this radicalar pathway has also the guanine residues as main target [24].

As more than 80% of the guanine sequenced can induce chain termination after phototreatment with proflavine, it is clear that there is no pre- ferred site for the damage to occur. The presence of hot spots for pausing of DNA polymerase a and E. coli polymerase 1, has been correlated with the position of stable hairpins in the template DNA secondary structure [25]. Highly localized hot spots are also found for cDNA elongation blocks on rRNA phototreated with HMT (4'-hy- droxymethyl-4,5',8-trimethylpsoralen, 2). In that case it is the binding of the drug which occurs at definite sites whose location is governed by the

secondary structure of the nucleic acid. No specific binding sites are observed for proflavine on ~Xs- sDNA. Under the experimental conditions used here, the increased rate of chain termination in- duction observed when varying the proflavine con- centration corresponds to increasing amounts of intercalated proflavine molecules (up to r = 0.25 where r is the molar ratio between bound prof- lavine and DNA phosphorus) [26]. The decreasing part of the graph in Fig. 4 must be due to the presence of either weakly-bound or free dye mole- cules. It is clear that all photodynamic reactions induced by intercalating dyes show the same type of dye concentration influence [26].

Complete block of the polymerase one nucleo- tide before the altered base has been related with the presence of pyrimidine dimers or bulky lesions like carcinogen adducts [5,6]. However, it is now known that different polymerases stop either be- fore or in front of the same lesion [7,8] and that replacement of Mg 2÷ by Mn 2+ in the polymerizing buffer also modifies the termination position rela- tively to the lesion [7,27].

It has also been shown recently that blocking lesions on single-stranded templates are bypassed by the polymerase replicating double-stranded DNA [28]. However, the first event in the infec- tious cycle of OX174 is the synthesis of the DNA chain complementary to the viral chain. It results that the damages induced on the ssDNA are im- portant in vivo and indeed no complete replicative form could be isolated from cells infected with photoinactivated phages. The same experiment has been done to show that the presence of psoralen adducts produced in Rous Sarcoma virus RNA is correlated with the loss of virus infectivity and a decreased length of viral DNA synthesized in vivo [291.

Acknowledgement

Dr. J. Piette is 'Chercheur qualifi~' from the Belgian National Fund for Scientific Research (Brussels).

References

1 Huang, C.C. and Hearst, J.E. (1980) Anal. Biochem. 103, 127-139

264

2 Youvan, D.C and Hearst J.E. (1981) Nucleic Acid Res. 9, 1723-1741

3 Youvan, D.C. and Hearst, J.E. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 3751-3754

4 Hagenbiichle, O., Santer, M. and Steitz, J.A. (1978) Cell 13, 551-563

5 Moore, P. and Strauss, B.S. (1979) Nature (London) 278, 664-666

6 Moore, P.D., Rabkin, S.D. and Strauss, B.S. (1980) Nucleic Acid Res. 8, 4473-4484

7 Moore, P.D., Bose, K.K., Rabkin, S.D. and Strauss, B.S. (1981) Proc. Natl. Acad. Sci. U.S.A. 78, 110-114

8 Moore, P.D., Rabkin, S.D., Osborn, A.L., King, C.M. and Strauss, B.S. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 7166-7170

9 Piette, J. and Moore, P.D. (1982) Photochem. Photobiol. 35, 705-708

10 Setlow, P., Brutlag, D. and Kornberg, A. (1972) J. Biol. Chem. 247, 224-231

11 Mc Clure, W.R. and Chow, Y. (1980) Methods Enzymol. 64, 277-292

12 Foote, C.S. (1979) in Biochemical and Clinical Aspects of Oxygen (Caughey, W.S., ed.) pp. 603-626, Academic Press, New York

13 Piette, J., Calberg-Bacq, C.M. and Van de Vorst, A. (1979) Photochem. Photobiol. 30, 369-378

14 Piette, J., Lopez, M., Calberg-Bacq, C.M. and Van de Vorst, A. (1981) Int, J. Radiat. Biol. 40, 427-433

15 Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. U.S.A 74, 5463-5467

16 Sanger, F., Coulson, A.R., Friedmann, T., Air, G.M., Bar- rell, B.G., Brown, N.L., Fiddes, J.C., Hutchinson III, C.A.,

Slocombe. P.M. and Smith, M. (1978) J. Moi. Biol. 125, 225-246

17 Burke, J.F., Plummer, J., Huherman, J.A. and Evans, M.J. (1980) Biochim. Biophys. Acta 609, 205-223

18 Piette, J., Calberg-Bacq, C.-M. and Van de Vorst, A. (1981) Photochem. Photobiol. 33, 325-333

19 Boiteux, S. and Laval, J. (1983) Biochem. Biophys. Res. Commun. 110, 552-558

20 Houba-H6rin, N., Calberg-Bacq, C.-M., Pierre, J. and Van de Vorst, A. (1982) Photochem. Photobiol. 36, 297-306

21 Cadet, J. and Teoule, R. (1978) Photochem. Photobiol. 28, 661-667

22 Nilsson, R. and Kearns, D.R. (1973) Photochem. Photobiol. 17, 65-68

23 Kraljic, I. and Sharpatyi, V.A. (1978) Photochem. Photo- biol. 28, 583-586

24 Cadet, J., Decarroz, C., Wang, S.Y. and Midden, W.R. (1982) Am. Soc. Photobiol. 10th Ann. Meeting Abst. WAM-C6.

25 Kaguni, L.S. and Clayton, D.A. (1982) Proc. Natl. Acad. Sci. U.S.A. 79, 983-987

26 Piette, J., Calberg-Bacq, C.-M. and Van de Vorst, A. (1979) Rad. Environ. Biophys. 16, 125-134

27 Rabkin, S.D., Moore, P.D. and Strauss, B.S. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 1541-1545

28 Piette, J. and Hearst, J.E. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 5540-5544

29 Swanstrom, R., Hallick, L.M., Jackson, J., Hearst, J.E. and Bishop, J.M. (1981) Virology 113, 613-622