Heterozygosis and genetic recombination in herpes simplex type 1 virus

Heterozygosis and genetic recombination in herpes simplex type 1 virus

VIROLOGY 82, 323-333 Heterozygosis (1977) and Genetic D. A. RITCHIE,’ Institute Recombination Virus in Herpes Simplex S. MOIRA BROWN,* J. H. ...

938KB Sizes 2 Downloads 15 Views


82, 323-333



and Genetic



Recombination Virus

in Herpes Simplex


of Virology,


of Glasgow,



Type 1


5JR, Scotland

Accepted May 17,1977 Two- and three-factor crosses with temperature-sensitive (ts) and syncytial plaque morphology (syn) mutants of herpes simplex type 1 virus have been used to study the possible role of syn-syn+ mixed plaque-forming virus in genetic recombination. Under the conditions of a standard genetic cross, recombinants first appear about 6 hr after infection, the time of formation of the first infectious progeny virus, and their frequency progressively rises until about 20 hr postinfection. During this period the frequency of mixed plaques remained constant at approximately 5%. The frequency of mixed plaques and recombinants were both increased several-fold when crosses were made in the presence of the DNA synthesis inhibitor Sfluorodeoxyuridine (FUdR). In view of the genetic instability of the partially heterozygous genomes of mixed plaque forming virus this result is interpreted to mean that mixed plaques identify virus which is an intermediate in recombinant formation and that the molecular structure of their genome is probably that of a partial heteroduplex. Measurements of deoxynucleoside triphosphate pools showed that FUdR inhibited the large increase in the dTTP pool size which normally accompanies HEW-1 infection of BHK cells.

cle (Brown et al., 1973). The ability to form a mixed morphology plaque is not due to a simple genetic change as the progeny from a mixed plaque contain very few mixed plaque forming virus and consist predominantly of virus forming pure syn and pure syn+ plaques. Physical and genetic evidence favoured the hypothesis that mixed plaques are produced predominantly by single virions each containing a single haploid genome. Moreover, the heterozygous region is normally relatively short and will extend to additional pairs of alleles only if closely linked to syn; this is illustrated by ts syn x ts+syn+ crosses where the synsyn+ heterozygotes are usually found to be homozygous at the ts locus. The few recorded cases of double heterozygosity occurred for ts markers closely linked to syn (Brown and Ritchie, 1975). These observations suggest a molecular structure in which the heterozygosity is a function of the duplex nature of the DNA molecule


In a previous communication evidence was presented that herpes simplex virus type 1 (HSV-1) could produce virus particles which were partially heterozygous and which possibly represented an intermediate in the recombination process (Brown and Ritchie, 1975). These particles are consistently detected at low frequency (-5%) among the progeny of mixed infections between parents carrying the syncytial (syn) plaque type mutant and its nonsyncytial wild-type allele (syn’) and can be distinguished from the parental virus by their distinctive mixed morphology plaques which have syncytial and nonsyncytial sectors. Since the syn-syn+ mixed plaques are not detectable in the parental virus stocks they must arise from interactions occurring during the replication cy1 Author to whom reprint requests should be addressed. 2 Members of the Medical Research Council Virology Unit. 323 Copyright All rights

8 19’77 by Academic Press, Inc. of reproduction in any form reserved.





(segregation occurring as a consequence of DNA duplication) and the heterozygous region is a short heteroduplex section of an otherwise homozygous molecule. The properties of the syn-syn+ mixed plaques described above are generally similar to those reported for the Heteroduplex or Internal Heterozygotes identified in several phage systems and particularly well studied for phage T4 (Hershey and Chase, 1951; Levinthal, 1954; Edgar, 1961; Sdchaud et al., 1965; Schalitin and Stahl, 1965; Mosig, 1970; Miller, 1975). Internal Heterozygotes arise from events which ultimately lead to the formation of genetically recombinant molecules, semiconservative DNA duplication results in their segregation with the production of homozygous offspring. The rates of formation and destruction (segregation) under normal conditions of infection are considered to be in equilibrium such that the frequency of Internal Heterozygotes is constant in samples matured at all times during infection. The balance can be upset by artificially reducing the amount of phage DNA synthesis with the inhibitor 5-fluorodeoxyuridine (FUdR) when it is found that the frequency of Internal Heterozygotes increases as a consequence of their reduced rate of segregation (Sechaud et al. 1965). A corresponding increase in recombinant frequency occurs under these conditions. The base analogues 5-fluorouracil and 5fluorodeoxyuridine are known to be potent inhibitors of DNA synthesis and infectious particle formation for the herpes viruses, pseudorabies virus (Kaplan and BenPorat, 1961; Reissig and Kaplan, 1962), equine abortion virus (O’Callaghan et al. 1968) and human cytomegalovirus (Goodheart et al. 1963). As a test of the Internal Heterozygote hypothesis for the genetic and physical structure of the genome of HSV-1 mixed plaque forming virus we have used this approach to examine the consequences of the FUdR inhibition of viral DNA synthesis on mixed plaque and recombinant formation. MATERIALS




The temperature-sensitive



and plaque morphology (sylz) mutants were derived from the wild type (ts+syn’) of HSV-1 (Glasgow strain 17). Their physiological, genetic, and biochemical properties have been described elsewhere (Brown et al., 1973; Subak-Sharpe et al., 1973; Subak-Sharpe et al., 1974; Marsden et al., 1976). The permissive and nonpermissive temperatures for the ts mutants are 31 and 38”, respectively. Cells. BHK (C13) cells growing in ETC medium [Eagle’s medium with 10% (v/v) tryptose phosphate broth and 10% (v/v) calf serum] were used for all growth experiments, for genetic crosses, and for the production and assay of virus stocks (Brown et al., 1973). Genetic crosses. The cross procedure was modified from that used previously (Brown and Ritchie, 1975). Cell monolayers (4 x lo6 tells/50-mm plastic petri dish) were infected with 5 PFU/cell of each parental virus. After absorption for 1 hr at 37”, unadsorbed virus was neutralised with Eagle’s medium containing tryptose phosphate broth (lo%, v/v) and human serum (lo%, v/v). The infected cell monolayers were washed twice with PBSC medium [0.17 M NaCl; 3.4 mM KCl; 10 mM Na,HPO,; 2 mZI4 KH2P04, pH 7.4, containing 10% (v/v) calf serum] and overlaid with 4 ml of ETC with or without FUdR. Following incubation at 31” the cell sheets were gently washed twice with PBSC and harvested into 4 ml of fresh ETC, and the virus was released by sonication. Under standard conditions the virus was harvested 24 hr after infection. For experiments in which samples were to be analysed at intervals throughout the growth cycle, replicate samples of infected cell monolayers were incubated together and at appropriate times plates were harvested in the usual manner. Virus was assayed as described previously (Brown et al., 1973). DNA synthesis. Cell monolayers were infected with 5 PFU/cell of virus as described above and incubated at 31”. Three hours postadsorption, 10 Z&i of [3H]TdR was added to each dish and incubation was continued until 24 hr when the cells were washed twice with PBS and lysed by the addition of 1 ml of lysis buffer (2.5 ml of


20% SDS and 40 ml of 10 x SSC per 100 ml). Following digestion with F’ronase (500 pg/ ml) for 3 to 4 hr at 31”, 0.2 ml of the cell extract was brought to 5 ml with 0.1 x SSC and solid CsCl was added to a final density of 1.718 g/cm3. Samples were centrifuged at 40,000 rpm at lo” for 3 days in a Ti 50 rotor of a Beckman ultracentrifuge. Fractions were collected dropwise from the bottom of the tube onto filter paper discs which were washed twice in cold 10% trichloroacetic acid and dried with ether, and the radioactivity was counted with a liquid scintillation spectrometer. Analysis of triphosphate pools. Measurement of the deoxynucleoside triphosphate pools is uninfected and HSV-l-infected cells followed the method described by Lindberg and Skoog (1970), Skoog (19701, and Jamieson and Bjursell (1976a). Cells were infected with 5 PFU/cell of wild-type virus and replicate dishes were incubated at 31” with or without FUdR as described above. At intervals, medium was removed from the petri dishes and the nucleotide pools were extracted with 60% methanol and then assayed. RESULTS

The Frequency of Mixed Plaques and Recombinants as a Function of Time The frequencies of mixed syn-syn+ plaques and of ts+ recombinants were measured among the progeny virus of threefactor crosses using different ts mutants in combination with the syn alleles. For each cross, assays were made of progeny virus artificially released at intervals throughout the growth cycle to provide a measure of recombinant and mixed plaque frequencies at different stages of virus replication (Table 1). The result observed for recombinant formation confirms our earlier findings that the frequency of recombinants normally rises from about 6 hr after adsorption, the time of appearance of the first infectious progeny, until about 20 hr postinfection (Ritchie, 1973; Subak-Sharpe the recombiet al., 1973). Occasionally, nant frequency appears to drop during the very late stages of infection but we believe this to reflect a statistical fluctuation rather than a reproducible genetic change.



By contrast, we have not observed increases in the percentage of mixed plaques with time, and Table 1 shows that their frequency remains approximately constant at all times aRer the onset of maturation. To examine the possibility that the rise in ts+ virus reported in Table 1 results from the faster replication of ts+ compared to ts virus, this experiment was repeated using a mixture of ts+syn+ and tssyn virus to infect the cells. To simulate the conditions of a genetic cross, the wild-type virus was made the minority parent. Table 2 shows that the ratio of wild-type to mutant virus does not systematically increase during the growth cycle when measured either as the ratio of ts+ infectious virus to total infectious virus (column 4) or as the ratio of syn+ to syn virus among the total infectious progeny (column 5). Since the ratio of wild-type to mutant parent apparently remains constant, we conclude that ts+ recombinants arising by recombination between ts mutant parents would not outgrow their parents in that infected cell. The normal rise in the frequency of recombinants is probably the outcome of two processes: (1) the continued and repeated formation of recombinants; and (2) the replication of recombinant genomes. The overall effect will be a progressive increase in the proportion of recombinants as the number of rounds of mating that has taken place within the infected population increases. The Effect of FUdR on Mixed Plaque and Recombinant Formation To determine the genetic consequences of reducing viral DNA synthesis, the frequencies of mixed plaques and of recombinants were determined among the progeny from three-factor crosses made in the presence of FUdR (Table 3). The concentration of FUdR used was chosen to reduce the virus yield but not to eliminate virus replication completely. In each case FUdR at 0, 1.25, or 12.5 pg/ml was present in the growth medium from the end of the 1-hr absorption period until harvest at 24 hr. Under the conditions of infection used, the FUdR reduced the virus yield to a some-






FREQUENCY OF MIXED PLAQUES AND RECOMBINANTS AT INTERVALS DURING THE GROWTH CYCLE” Time* Cross 1 Cross 2 Cross 3 Cross 4 Cross 5 tsGsyn x tstsDsyn x tstsfbyn X tstsAsyn X tstsAsyn X .?s(hr) Lsyn+ Lsyn+ Fsyn+ Bsyn+ Esyn+ M (S)

M (%)


;i 0, 2, 4 6 8 10 12 20 22 24

0 0 0.01 0.01 0.19 0.20 nd 0.20

0 3.5 3.1 3.9 4.1 1.1 nd 1.9

0 0 0.11 0.30 0.35 nd 0.62 0.54

0 3.0 3.6 4.1 nd nd 3.3 5.0

0 0.11 0.14 0.23 0.35 0.47 nd 0.80

Ei 0 5.3 2.8 3.9 5.0 6.0 nd 6.0

M (%I

M (‘70)

M (%) R

0 0.65 0.69 nd 9.3 14.0 nd 26.0

t-E 0 20.0 9.0 nd 8.0 10.0 nd 10.0

0 0.02 0.05 0.14 0.15 nd 0.62 0.42

0 nd’ 4.3 9.0 nd nd 5.0 5.0

a Cells were infected with pairs of ts mutants carrying the syn and syn+ alleles and the progeny virus was released at intervals and assayed to determine the percentage of recombinants (measured as twice that of the selected ts+ recombinant) for the ts markers (RF (8)) and the percentage of syn-syn”r mixed plaques (M (S)). With the conditions used the eclipse phase extends until 4 to 6 hr postinfection. b Hours postinfection. c nd, Not determined. TABLE GROWTH [email protected] (hr)

0 2 4 6 8 12 24



ts+syn+ AND tsIsyn COINFRCTION”







7.0 3.1 3.0 8.6

1.4 7.0 5.0 2.4 3.3 1.0 5.0

x x x x

104 104 104 lo4

9.0 x 104

2.1 x 105 2.9 x 10’

x x x x x x x

104 103 103 lo4 104 105 106



titre 0.20 0.23 0.17 0.28 0.37 0.48 0.17

syn’ 31”

d syn 31 o 0.46 0.41 0.36 0.26 0.60 0.40 0.33

a Cells were infected with a mixture of ts+syn+ and tskyn virus to give a total multiplicity of infection of 10 PFU/cell and with the ts parent in excess. Progeny virus was released at intervals and assayed for total infectious progeny (31”), for t.s+ infectious progeny (389, and for syn and syn+ plaques at 31”. * Hours postabsorption. c Titres given as PFU per lo6 infected cells. d Ratio of .syn+ to syn plaques among total infectious progeny.

what variable extent, but of the order of 30% of the untreated control (Table 3, column 3). The possibility of unreplicated parental virions forming a major component of the virus yield was eliminated by washing the infected cells after adsorption and neutralising unadsorbed virus with antiserum. With the exception of one experiment which exhibited an unusually high control value, the results consistently show that

the frequency of mixed plaques is increased in the presence of FUdR (Table 3, columns 4 and 5). The mixed plaque frequency is much the same when measured among the total progeny (31” titration) or among the selected t.s+ recombinant progeny (38” titration). This increase varies generally between two- and tenfold and is not noticeably higher among the samples treated with 12.5 pglml of FUdR. A consistent, though somewhat smaller, increase (twofold to sixfold) was observed for the production of ts+ recombinant virus (Table 3, column 6). The time course of ts+ recombinant and mixed plaque production under conditions of FUdR inhibition is shown in Fig. 1. FUdR enhances recombinant production by increasing the overall rate of recombination to produce a curve with an increased slope when compared with the untreated control (Fig. 1A). The drop in recombinant frequency occasionally observed is quite pronounced in this particular experiment. By contrast, the effect of FUdR on mixed plaque production was to convert a constant proportion of mixed plaques into one that progressively increased with time after infection (Fig. 1B). These results confirm that the effect of FUdR is different for mixed plaque formation than for recombinant formation al-


EFFECT OF FUdR cross ts’syn

x tsFsyn+


x tsDsyn+


x tsGsyn+


x tsJsyn+

tslsyn tsJsyn

Virus yieil

0 1.25 0 1.25 0 1.25 12.5 0 1.25 0 1.25 12.5 0 1.25 0 1.25

x tsFsyn+


TABLE 3 ON FREQUENCY OF MIXED PL.teuEs AND [email protected]

FUdR (pg/ml)

X tsGsyn+



4.2 x lo7 3.8 x 106

3.1 x 2.0 x 1.0 x 3.0 x 2.0 x 8.0 x 4.0 x 3.0 x 9.0 x 2.0 x 3.3 x 1.0 x 1.8 x 6.0 x

108 106 108 10’ 10’ 106 106 10’ 106 10’ 106 106 106 lo5

M (81 31”

M (8) 38

4.8 21.0

6.0 20.0








5.0 2.7 4.0 8.1

7.7 5.0 12.0 1.4


9.5 10.0

7.3 11.7 2.6 13.6 4.8

2.4 6.5

1.8 16.0


RF (%I

5.3 13.0 7.7 12.0 14.0 3.6 8.0 1.3 4.7 7.2

9.0 15.0 13.3 35.0

a From each cross the progeny was titrated for the percentage ofsyn-syn+ mixed plaques at 31” (M (%I 31”) and at 38” (M (8) 38”) and for the percentage of recombinants for the ts markers (RF (o/o)). 12 A

o--o 0




16 L 20 0 time post absorption






FIG. 1. The effect of FUdR on the production of recombinants and mixed plaques. The cross tsAsyn x tsGsyn+ was set up as described in Materials and Methods and replicate plates of infected cells were incubated at 31” in the presence (--) and absence (-----) of FUdR (1.25 pg/ml), and at intervals throughout the growth cycle individual samples were harvested for assay of total infectious progeny, ts+ recombinants, and mixed syn-syn+ plaques. (A) Frequency of recombinants for the ts markers (calculated as twice the percentage of the ts+ recombinant&; (B) percentage of syn-syn+ mixed plaques among the total infectious progeny.

though both effects increase the measured parameter. Inhibition FUdR

of Viral




The consequence of FUdR treatment on DNA synthesis in virus-infected cells is illustrated in Fig. 2, with more detailed results given in Table 4. Infected cells were labelled with [3H]TdR from 3 to 24 hr

postadsorption and cell extracts were centrifuged to equilibrium in CsCl to separate cellular and viral DNA. FUdR treatment consistently reduced the total radioactivity incorporated into viral DNA. Although somewhat variable, viral DNA synthesis was reduced to about 40-60% of the control value following treatment with 12.5 pg/ml of FUdR and to about 65-95% for 1.25 pgl ml (Table 4, column 3). Furthermore, the



HSV-1 infection of BHK/C13 cells causes a large increase in the size of the d’M’P pool, while the dATP pool size is reduced (Jamieson and Bjursell, 1976a). FUdR treatment of virus-infected cells produced increases in the dATP, dCTP, and dGTP pools. However, the dTTP pool size was markedly reduced. For uninfected cells FUdR treatment resulted in the same pattern of changes in the sizes of the dNTP pools. The shortage of dTTP resulting from inhibition of thymidylate synthetase function is presumably the cause of the FUdRinduced inhibition of virus DNA synthesis. The failure of these concentrations of FUdR to abolish completely the d’l.TP pool no doubt accounts for the remaining DNA synthesis in the virus-infected cells. For the uninfected cells, FUdR produced slight increases of the dATP and dGTP pools whilst leaving the remaining pools unaffected.




20 25 30 fraction number


FIG. 2. Virus and host cell DNA synthesis in the presence and absence of FUdR. Cells were infected with a mixture of tsDsyn and tsAsyn+ virus under standard cross conditions in the presence of FUdR l(A) 0 pglml; (B) 1.25 pglml; (0 12.5 kg/ml], and the DNA was labelled with [3H]TdR from 3 to 24 hr postabsorption as described in Materials and Methods. Cell extracts were centrifuged to equilibrium in CsCl and fractions assayed for acid-insoluble radioactivity. The denser virus DNA bands to the left of each gradient.

incorporation of radioactivity into viral DNA relative to that into cellular DNA was also usually reduced, particularly following exposure to 12.5 pglml of FUdR (Table 4, column 5). We also note that in most cases, FUdR treatment led to an increased uptake of r3HlTdR into host cell DNA (Table 4, column 4). This effect was observed for both uninfected and virusinfected cells. To determine a little more precisely the effects of FUdR on DNA synthesis of HSV-l-infected cells, the concentrations of the four deoxynucleoside triphosphate pools were measured in the presence and absence of the inhibitor (Fig. 3). This analysis confirmed an earlier observation that


The analysis of the work presented here depends on the successful inhibition of HSV-1 DNA synthesis by FUdR. The results given in Fig. 2 and Table 4 demonstrate that incubation of the HSV-l/ BHK21(C13) system in the continued presence of FUdR leads to a significant but incomplete depression of viral DNA synthesis as measured by the incorporation of i3HlTdR into acid-insoluble material. The synthesis of infectious progeny particles is similarly inhibited (Table 3). Much the same inhibition was observed whether 1.25 or 12.5 pglml of FUdR was used and this suggests that the observed depression is the maximum that can be achieved by inhibiting the activity of the intracellular thymidylate synthetase. The remaining synthesis derives presumably from the production of dTMP via the salvage pathway catalysed by thymidine kinase, which is resistant to FUdR inhibition. The incomplete inhibition of viral DNA synthesis and infectious particle production reported here for HSV-1 is much less pronounced than the much more severe inhibition which has been observed for the herpesviruses pseudorabies (Kaplan and BenPorat, 1961; Reissig and Kaplan, 1962) and



IN HSV-1 4


FUdR (cLg/ml)

ts +syn +

ts +syn +


x tsDsyn+


x tsDsyn+


x tsJsyn+


x tsJsyn+





Viral DNA* (cpm X 10W4)

0 1.25 12.5 0 1.25 12.5 0 1.25 12.5 0 12.5 0 1.25 0 12.5

23.4 16.1 13.1 14.4 13.8 5.3 30.5 19.6 14.9 9.6 3.9 26.5 19.3 12.2 7.5

0 1.25 0 1.25


(69) (58) (96) (37) (64) (49) (41) (73) (61)

Cell DNA* (cpm X 103

Virus DNAc cell DNA

48.7 63.6 65.1 45.9 42.4 70.4 76.3 47.2 49.4 49.9 83.5 67.8 60.5 23.3 45.5

0.49 0.25 0.21 0.31 0.33 0.08 0.40 0.41 0.30 0.19 0.05 0.39 0.32 0.53 0.16

51.1 89.4 56.9 74.6


a Details are given in Materials and Methods and in legend to Fig. 2. * Radioactivity in the peak fractions of viral and cellular DNA were summed to give total radioactivity incorporated into viral and cellular DNA. Numbers in parentheses give counts incorporated after FUdR treatment as a percentage of the untreated control. c Ratio of counts incorporated into virus and cellular DNA. d Uninfected cells were labelled for a period equal to that for the virus-infected samples.

equine abortion virus (O’Callaghan et al., 1968). Measurement of the deoxynucleoside triphosphate pools reveals that FUdR treatment prevents the large increase in the size of the dlTP pool which normally accompanies HSV-1 infection; in fact the d’ITP pool size is the same as for the uninfected and untreated cells (Newton et al., 1962; Jamieson and Bjursell, 1976a). We assume this shortage of dTTP in HSV-linfected, FUdR-treated cells to be the immediate cause of the reduced DNA synthesis. This view is consistent with the observations of Jamieson and Bjursell (197613) who suggested that in dPyK-HSVinfected resting cells the existing d’ITP pool size was insufficient to sustain viral DNA synthesis. Interestingly, the absolute amounts of radioactivity incorporated into cellular DNA often showed a rise with FUdR treatment; in our view this probably reflects a higher specific activity of d’I’TP because of

lesser dilution of the [3HlTdR label by de nouo synthesised dTMP. If viral DNA is synthesised from the same precursor pools as host DNA and the d’ITP pool has higher specific activity, then the actual amount of HSV-1 viral DNA made could be correspondingly lower in real terms. However, the precise level of the DNA inhibition is not particularly pertinent in the present study. Since the rise in mixed plaque frequency resulting from FUdR treatment occurs under conditions which reduce the total virus yield (Table 3), the possibility that all mixed plaques arise from virion aggregates or multiploid particles seems very unlikely. In an earlier paper we suggested that the production of mixed plaques may be related to the recombination process since mixed plaques arising from two-factor crosses usually segregate to give one parent and one recombinant. The concomitant increase in both mixed plaques and ts+ recombinants with FUdR treatment is



time post odsorptlon


FIG. 3. The effect of FUdR on the deoxynucleoside triphosphate pools of uninfected and ts+ virusinfected cells. Infected and control cells were incubated at 31” with or without FUdR at 12.5 pglml. At intervals samples were removed and assayed for the four deoxynucleoside triphosphates. Uninfected cells with FUdR (Cl- - - -0) and without FUdR (O- - - -0); infected cells with FUdR (m-m) and without FUdR (0-O). (A) dGTP, (B) dCTP; (C) dATP; (D) d’lTP.

consistent with this view. Under normal growth conditions the frequency of mixed plaque forming virus remains constant throughout the posteclipse phase of replication even though the yield of virus progeny increases by several orders of magnitude (Tables 1 and 2). Two general possibilities might account for this observation. (1) Mixed plaque forming virus genomes are newly formed only during the eclipse phase and thereafter replicate and are matured at the same rate as other virus genomes. This proposition fails to account for the observed genetic instability of the mixed plaque forming genotype. Although a limited degree of such “conservative” replication cannot be ruled out, our published results on 1032 progeny plaques from 26 mixed plaques showed only 36 (about 3.5%) to again have mixed morphology, the majority being pure syn or syn+ (Brown and Ritchie, 1975). It will be recalled that about 5% mixed morphology plaques normally arise in mixed infec-

tions. Thus these 36 mixed plaques would be expected to have arisen de nouo from pure segregant DNA molecules during the growth of the parental mixed plaques. On this proposition the rise in mixed plaque frequency during FUdR treatment (Fig. 1B) additionally demands either a selectively increased rate of replication for mixed plaque forming progeny, or FUdRinduced extension of the eclipse phase, or FUdR-induced increase in formation of mixed plaque forming progeny during the eclipse phase. Lastly, the proposition does not imply a connection between mixed plaque formation and recombination. (2) Mixed plaque forming genomes are continuously being produced and destroyed during DNA replication such that the turnover rate under the standard experimental conditions (i.e., untreated control) is constant. This proposition is consistent with the segregation data and also with the current view of heterozygote formation and of the mechanism of genetic recombination (Mosig, 1970; Miller, 1975). On this hypothesis the rise in mixed plaque frequency during FUdR treatment would be explained by an increased rate of production of heteroduplex genomes, or by a decreased rate of segregation through replication of DNA molecules, or by some combination of the two such that the overall rate of production exceeds the rate of loss. As we have no experimental evidence of any changes in the size of the genome (mating) pool we assume a constant pool size. The rise in the recombinant frequency with yield of progeny virus (i.e., with time after infection) under standard experimental conditions was interpreted to mean that repeated acts of mating between DNA molecules occurred during the growth cycle within the HSV-l-infected cell (Ritchic, 1973; Subak-Sharpe et al., 1973; Subak-Sharpe et al., 1974). This leads to an increased probability for any molecule to be recombinant as the infection progresses. [This interpretation has also been put forward by Williams et al. (1974) to explain a similar phenomenon observed in crosses with adenovirus type 5.1 The increased fraction of recombinants observed








sYn I ---__-..__ -+‘---____

syn 5yn -7. ---++




SYn ----+sCn-------I-

+ c





----L..__ $ _____,_ +


FIG. 4. Scheme for the formation and segregation of syn-syn+ heterozygotes from a syn x syn+ cross. Recombination is considered to occur at any site on the molecule by a mechanism leading to the exchange of genetic material on either side of a relatively short “overlap” region. (A) Heteroduplex heterozygote. The syn locus is at a single genetic site and syn-syn+ heterozygotes arise when the “overlap” region covers the syn locus. Segregation causes loss of heterozygosity and formation of homozygous syn and syn+ progeny. (B) Redundancy heterozygote. The syn locus is in a redundant (repeat) region surrounding a unique sequence and is represented twice on the molecule. Recombination rearranges the syn sites to form a redundancy heterozygote which on replication produces progeny which are daughter redundancy heterozygotes. More complicated arrangements could arise on this model, e.g., a crossover, leading to formation of a heteroduplex heterozygote, might occur within a redundant region to give a molecule with three copies of one allele and one of the other.

under conditions of FUdR-inhibition of HSV-1 DNA synthesis could arise from an increased rate of recombinant production, or from a longer average sojourn of any DNA molecule in the mating pool, or from a selective rise in the rate of recombinant replication. Control experiments in which a mixture of ts and ts + virus was allowed to infect the same cells clearly showed that the rate of synthesis of the two genotypes was much the same (Table 2). This observation rules out the last possibility. Mixed plaques arising from two-factor crosses contain progeny virus which segregate to yield usually one recombinant and only one of the parental genomes (Brown and Ritchie, 1975). This evidence coupled with the results presented above provide a consistent picture of virus genomes mating to produce recombination intermediates which are partially heterozygous (heteroduplex) and which subsequently segregate by DNA replication, to give homozygous (homoduplex) progeny, some parental and some recombinant. Normally the rates of formation and segregation of heterozygotes are in equilibrium but if segregation is restricted (e.g., by limiting DNA repli-

cation through FUdR treatment), without equivalent restriction on mating, then heterozygous genomes accumulate. It would be expected that FUdR treatment would lead to an increase in the frequency of recombinant intermediates and of recombinants; all syn-syn+ heterozygotes will give rise to recombinants between syn. and another locus on replication except for that proportion in which the heterozygous region spans both loci. This in fact is the observed result. The expected effect of FUdR treatment on syn-syn+ heterozygote formation is a rise above the normal experimental value of about 5%, while the expected effect on recombination frequency must always be a function of the distance separating the two loci. A structure for the mixed plaque genome in which one strand of the DNA duplex is syn and the other is syn+ is consistent with two basic properties of such virus: (1) the constant instability of the mixed plaque character, and (2) the increase in mixed plaque frequency when DNA synthesis is reduced (Fig. 4A). There is an alternative possible structure which would invoke redundancy, i.e., in which



GOODHEART, C. R., FILBERT, J. E., and MCALLIBTER, both strands are syn on one repetitious R. M. (1963). Human cytomegalovirus. Effects of sequence and both strands are syn’ on the 5-fluorodeoxyuridine on viral synthesis and cytohomologous repeat (Fig. 4B) (S&haud et pathology. Virology 21, 530-532. al., 1965). This structure does not ob- GRAFSTROM, R. H., ALWINE, J. C., STEINHART, W. viously arise from the two basic properties L., and HILL, C. W. (1974). Terminal repetitions stated above, although the known physical in herpes simplex virus type 1 DNA. Cold Spring structure of the HSV-1 genome (Grafstrom Harbor Symp. Quant. Biol. 39, 679-682. et al., 1974; Sheldrick and Berthelot, 1974; HAYWARD, G. S., JACOB, R. J., WADSWORTH, S. C., and ROIZMAN, B. (1975). Anatomy of herpes simHayward et al., 1975; Wadsworth et al., plex virus DNA: Evidence for four populations of 1975; Delius and Clements, 1976; Wilkie molecules that differ in the relative orientations and Cortini, 1976) does not rule it out unof their long and short components. hoc. Nat. less it can be shown that the syn locus lies Acad. Sci. USA 72, 4243-4247. within a unique region. As the repeat seHERBHEY, A. D., and CHASE, M. (1951). Genetic quences in HSV-1 DNA are nonpermuted, recombination and heterozygosis in bacterioreplication would not automatically lead phage. Cold Spring Harbor Symp. Quant. Biol. to segregation and such heterozygotes 16, 471-479. would be inherited. Any reassortment JAMIESON, A. T., and BJURSELL, G. (1976a). Deoxyeither from recombination or, for example, ribonucleoside triphosphate pools in herpes simfrom the maturation of concatemer or cirplex type 1 infected cells. J. Gen. Virol. 31, 101113. cular intermediates would not be expected to reduce the yield of daughter mixed JAMIESON, A. T., and BJURSELL, G. (1976b). Deoxyribonucleoside triphosphate pools in cells infected plaques to below 50% of the total progeny with deoxypyrimidine kinaseless herpes simplex of a given mixed plaque. This is not the virus. J. Gen. Virol. 31, 115-123. situation which we have observed. Precise KAPLAN, A. S., and BEN-P• BAT, R. (1961). The aclocation of the syn locus should help totion of 5-fluorouracil on the nucleic acid metabowards the final resolution of the precise lism of pseudorabies virus-infected and non-instructure of syn-syn+ heterozygotes, and fected rabbit kidney cells. Virology 13, 78-92. determination of the precise location is in LEVINTHAL, C. (1954). Recombination in phage T2: progress. Its relationship to heterozygosis and growth. Ge-

ACKNOWLEDGMENTS We wish to thank Roche Products Limited, Welwyn Garden City, for the gift of 5fluorodeoxyuridine, the Science Research Council for their tinancial support and Miss D. MacInnes for excellent technical assistance. REFERENCES BROWN, S. M., and RITCHIE, D. A. (1975). Genetic

studies with herpes simplex virus type 1. Analysis of mixed plaque-forming virus and its bearing on genetic recombination. Virology 64, 32-42. BROWN, S. M., RITCHIE, D. A., and SIJBAK~HARPE, J. H. (1973). Genetic studies with herpes simplex virus type 1. The isolation of temperature-sensitive mutants, their arrangement into complementation groups and recombination analysis leading to a linkage map. J. Gen. Virol. 18, 329-346. DELIUS, H., and CLEMENTS, J. B. (1976). A partial denaturation map of herpes simplex virus type 1 DNA: Evidence for inversions of the unique DNA regions. J. Gen. Virol. 33, 125-134. EDGAR, R. S. (1961). High negative interference and heterozygosis: A study of the mechanism of recombination in bacteriophage. Virology 13, 1-12.

netics 39, 169-184. LINDBERG, U., and SKOOC,L. (1970). A method for

the determination of dATP and d’R’P in picomole amounts. Anal. Biochem. 34, 152-160. MARSDEN, H. S., CROMBIE, I. K., and SIJBAKSHARPE, J. H. (1976). Control of protein synthesis in herpesvirus-infected cells: Analysis of the polypeptides induced by wild type and sixteen temperature sensitive mutants of HSV strain 17. J. Gen. Virol. 31, 347-372. MILLER, R. C. (1975). Replication and molecular recombination of T-phage. Annu. Rev. Microbial. 29, 355-376. MOSIG, G. (1970). Recombination in bacteriophage T4. Advan. Genet. 15, l-49. NEWTON, A., DENDY, P. P., SMITH, C. L., and WILDY, P. (1962). A pool size problem associated with the use of tritiated thymidine. Nature (London) 194, 886-887. O’CALLAGHAN, D. J., HYDE, J. M., GENTRY, G. A., and RANDALL, C. C. (1968). Kinetics of viral de-

oxyribonucleic acid, protein and infections particle production and alterations in host macromolecular synthesis in equine abortion (herpes) virus infected cells. J. Virol. 2, 793-804. REISSIG, M., and KAPLAN, A. S. (1962). The morphology of non-infectious pseudorabies virus pro-




duced by cells treated with 5fluorouracil. VirolBiosciences” (G. Rasp& ed.), Vol. 11, pp. 205-218. ogy 16, 1-8. Pergamon Press, Vieweg. RITCHIE, D. A. (1973). Genetic analysis of animal SUBAK-SHARPE, J. H., BROWN, S. M., RITCHIE, D. viruses. Brit. Med. Bull. 29, 247-252. A., TIMBURY, M. C., MACNAB, J. C. M., MARSDEN, SBCHAUD, J., STREISINGER, G., EMRICH, J., NEWH. S., and HAY, J. (1974). Genetic and biochemical TON, J., LANFORD, H., and STAHL, M. M. (1965). studies with herpes virus. Cold Spring Harbor Chromosome structure in phage T4. II. Terminal Symp. Quant. Biol. 39, 717-730. redundancy and heterozygosis. Proc. Nat. Acad. WADSWORTH, S. C., JACOB, R. J., and ROIZMAN, B. Sci. USA 54, 1333-1339. (1975). Anatomy of herpes simplex virus DNA. II. SHALITIN, C., and STAHL, F. W. (1965). Additional Size composition and arrangement of inverted terevidence for two kinds of heterozygotes in phage minal repetitions. J. Virol. 15, 1487-1497. T4. Proc. Nat. Acad. Sci. USA 54, 1340-1341. WILKIE, N. M., and CORTINI, R. (1976). Sequence SHELDRICK, P., and BERTHELOT, N. (1974). Inverted arrangement in herpes simplex virus type I DNA: repetitions in the chromosomes of herpes simplex Identification of terminal fragments in restriction virus. Cold Spring Harbor Symp. Quant. Biol. 39, endonuclease digests and evidence for inversions 667-678. in redundant and unique sequences. J. Virol. 20, SKOOG, L. (1970). An enzymatic method for the de211-221. termination of dCTP and dGTP in picomole WILLIAMS, J. F., YOUNG, C. S. H., and AUSTIN, P. E. amounts. Eur. J. Biochem. 17, 202-208. (1974). Genetic analysis of human adenovirus type SUBAK-SHARPE, J. H., BROWN, S. M., RITCHIE, D. 5 in permissive and non-permissive cells. Cold A., TIMBURY, M. C., and HALLIBURTON, I. W. Spring Harbor Symp. Quant. Biol. 39,427-437. (1973). Herpes virus genetics. In “Advances in the