VARIATION OF HERPES SIMPLEX VIRUS IN PERSISTENTLY INFECTEDTISSUE CULTURES

HARRY C. HINZE AND DUARD L. WALKERDepartment of Medical Microbiology, University of Wisconsin Medical School, Madison, Wisconsin

Received for publication April 5, 1961

ABSTRACT

HINZE, HARRY C. (University of Wisconsin,Madison), AND DUARD L. WALKER. Variationof herpes simplex virus in persistently infectedtissue cultures. J. Bacteriol. 82:498-504. 1961.-Cultures of human conjunctiva, HeLa, and KBcells infected with herpes simplex virus (HFstrain) were cultured for prolonged periods inmedium containing low levels of antibody. Aftercontinuous culture of the infected cell lines for6 months, two major changes were noted in thecharacter of the virus present in the cultures.These changes consisted of an alteration in thetype of cytopathic effect produced by the virusand marked loss of virulence for mice. Furtherstudy of the virus from the cultures revealedthat it was still antigenically similar to theoriginal strain of herpes virus. It was found thatthe variant produced an altered, proliferativetype of cytopathic effect only in the presence ofherpes antibody. The variant multiplied moreslowly in human conjunctiva cells in culturethan did the parent virus, and the variant hadlost practically all capacity for multiplication inthe brain of mice.

Viral carrier states in tissue cultures havebeen described by several workers using a varietyof virus-cell systems (reviewed by Ginsberg(1958)). Several techniques have been foundeffective in establishing cultures in which thereis a sufficiently stable relationship between virusand cell to allow long term culture. As pointedout by Ginsberg (1958), the outstanding featureof most of these carrier states has been a relativeinsusceptibility of the cells resulting from one ormore of several possible mechanisms. Frequentlythis has been accomplished, at least initially, byprotecting the cells with antibody added to theculture medium. In the study reported here itwas found that with the addition of human serumcontaining low levels of antibody to cell cultures

infected with herpes simplex virus it was possibleto continue such cultures for prolonged periods.In contrast to most previous reports on carriercultures where any changes in the system haveusually been attributed to changes in the cells,it was observed in this study that striking changesin the characteristics of the virus appeared afterseveral months of culture. The observed changesin viral characteristics are described in this report.

MATERIALS AND METHODS

Virus. Herpes simplex virus (HF strain),here referred to as herpes virus, was obtainedfrom the American Type Culture Collection.In this laboratory the virus was passed seriallythrough 2 mouse passages and 19 tissue culturepassages. Virus from the fifth tissue culturepassage was used for initial infection of thecarrier cultures.

Tissue cultures. HeLa (Scherer, Syverton, andGey, 1953), KB (Eagle, 1955a), and humanconjunctiva cells (Chang, 1954) were routinelycultivated in 200-ml flat-sided bottles at 37 C.Medium for the growth of all cell lines wasEagle's basal medium (Eagle, 1955b) and 20%horse serum. Cells were dispersed with 0.05%trypsin (Nutritional Biochemicals Corporation,Cleveland, Ohio, 1:300) and were distributed totubes, bottles, and petri dishes as needed.Horse serum was obtained from a local com-

mercial source. Human serum was obtainedfrom a selected group of local donors. Approxi-mately equal amounts of the human serum fromeach donor were pooled and filtered through aSeitz positive pressure filter before use. Thepooled serum was assayed for neutralizing anti-body against herpes virus. A dilution of 1:2 ofthe pooled human serum neutralized approxi-mately 300 plaque-forming units of virus. Allsera were inactivated at 56 C for 30 min beforeuse.

Virus titrations. Titration of stock and variantviruses was done both in tube cultures and by

498

on January 30, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

VARIATION OF HERPES SIMPLEX VIRUS

a plaque technique. Tube cultures seeded with6 X 105 cells were incubated at 37 C until thecells formed confluent sheets. The cultures werewashed once with balanced salt solution andinoculated with serial, 2-fold dilutions of virusin Eagle's basal medium containing 5% horseserum. Cultures were examined microscopicallyfor a period of 7 days for cytopathic effect (CPE)of the virus, and the 50% infective dose (TCID50)was calculated by the method of Reed andMuench (1938).

Plaque assays were made using monolayers ofconjunctiva cells in 60-mm petri dishes. Thegrowth medium was removed and the cells werewashed with a balanced salt solution. Afterappropriate dilution in Eagle's medium contain-ing 3% horse serum, 1 ml of the virus suspensionwas added to each culture. Two dishes wereinoculated for each 2-fold dilution. After a 2-hradsorption period, the inoculum was removedand an overlay of 1.5% methylcellulose in me-dium 199 (Morgan, Morton, and Parker, 1950)was applied. Plaques, 2 to 3 mm in diameter,developed in 3 days at 37 C in an atmosphereof 5% carbon dioxide and 95% air, and werereadily counted without staining.Serum neutralization tests. Equal volumes of

serum and virus dilutions were mixed and in-cubated for 1 hr at room temperature. Petridish cultures of conjunctiva cells were inocu-lated with 1 ml of the serum-virus mixtures anda methylcellulose gel overlay was applied as forvirus titrations. The neutralizing titer of theserum was determined by the reduction in plaquenumbers as compared with normal serum con-trols.

RESULTS

Production of herpes carrier cultures. Bottlecultures of HeLa, KB, and conjunctiva cells wereinoculated with 1,000 TCID50 of herpes virus con-tained in 1 ml of balanced salt solution. Thecultures were incubated at 37 C for 1 hr to allowfor adsorption of virus after which 9 ml of growthmedium containing 30% pooled human serumwere added to each bottle. The cultures were in-cubated at 37 C thereafter and medium waschanged every 2 to 4 days to provide optimalconditions for cell growth. The course of infec-tion was followed by daily observations.

Focal lesions appeared in all cultures within24 hr. The lesions enlarged slowly on successivedays but otherwise appeared similar to those

produced by stock virus in the absence of humanserum. Few new lesions appeared at 48 or 72 hr.When the virus had destroyed the majority ofcells in the cultures, the remaining uninfectedcells multiplied sufficiently to form a fairly com-plete sheet of cells covering the side of the culturebottle. Following this outgrowth, cellular de-generation again proceeded slowly producingalternate cycles of degeneration and regrowthof the cultures. Such infected cultures could besubcultured at frequent intervals or kept withoutsubculturing for an indefinite period duringwhich they continued to produce virus. Viruscould be detected consistently in all cultures byinoculation of either cells or uncentrifuged super-natant culture fluid into mice and tissue cultures.After 4 to 5 weeks, the concentration of pooledhuman serum in the culture medium was re-duced to 10% without noticeable change inactivity of the virus in the cultures. Omissionof the human serum from the medium at anytime resulted in rapid and complete destructionof all cells in the culture.

Infected cultures were maintained in thismanner for over 212 years to demonstrate thestability of such a system. Both infected HeLaand KB cultures were maintained for periods upto 1 year without subculturing. Other similarlyinfected cultures were passed serially at 2- to3-week intervals for the duration of the studies.

After continuous culture of the infected celllines for 6 months, two changes were noted in thecharacter of the virus present in the cultures.These changes consisted of (i) an alteration inthe type of CPE produced by the virus both inthe carrier cultures and in normal cell cultures inthe presence of human serum, and (ii) markedloss of virulence for mice.

Alteration in CPE of virus. The type of cellulardegeneration seen in the carrier cultures duringthe first few months after infection is picturedin Fig. 1. Such lesions were typical of those seenwith herpes virus infection in a variety of celltypes. The unstained lesions consisted of foci ofround, refractile cells varying somewhat in size,and of numerous multinucleated giant cells. Thecells in the center of such lesions appeared to bedestroyed leaving a hole in the cell sheet sur-rounded by a ring of infected cells. After cultiva-tion of the infected carrier lines for several months,the infected areas appeared as slowly enlargingmasses of round refractile cells on a background

1961] 499

on January 30, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

[VOL. 82HINZE AND WALKER

of relatively normal, uninfected cells (Fig. 2).Some cells from these masses eventually werereleased into the medium, but little or no celllysis was apparent. Although such foci continuedto increase in size for 7 to 9 days, histologicalstaining showed no definite indicatibn of increasedmitosis in the cell masses or in the surroundingarea. All cells in the lesions were apparently

infected as indicated by the presence of intra-nuclear inclusions. Similar lesions were readilyproduced in cultures of normal conjunctiva orHeLa cells by inoculating with virus from thecarrier cultures and maintaining the cells in amedium containing pooled human serum. Thecharacteristic proliferative CPE was also shownby virus clones derived from single plaques that

4 4. 4

~~~~~~~~~~~.

X U* a..,

FIG. 1. Typical focus of infection of the stock strain of herpes simplex virus in human conjunctiva cells.Unstained, X100.

FIG. 2. Multiple foci of infection of attenuated variant of herpes simplex virus in human conjunctivacells. Unstained, XIOO.

500

on January 30, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

VARIATION OF HERPES SIMPLEX VIRUS

developed on monolayers of conjunctiva cellsseeded with virus from carrier cultures andcovered with an overlay of Methocel.When human serum free of antiherpes anti-

body, or animal serum containing no antibody,was used in the medium, the CPE produced bythe variant virus was similar to that of the stockherpes virus. In an experiment presented inTable 1, ten human sera and several animalsera known to be free of herpes virus antibodywere individually tested as the serum componentof the medium and were compared with humanand rabbit sera known to contain significantlevels of herpes virus neutralizing antibody. Inthis experiment there was a good correlationbetween the presence of antibody in the mediumand the occurrence of the modified CPE.

Loss of virulence for mice. Virus obtained fromthe carrier cultures during the first few monthsof cultivation readily produced convulsions anddeath in adult Swiss mice within 3 to 5 daysafter intracerebral inoculation. With increasingtime in the carrier cultures, however, the virusshowed progressively less virulence for mice,although maintaining a relatively high titerwhen assayed in tissue culture. Stock herpes viruswas subjected to several passages in conjunctivacells during the period of the present study. Itwas cultivated in a medium free of antibody andwas always harvested at peak of CPE. In con-

trast to the virus from carrier cultures, stockherpes virus grown in conjunctiva cells retainedhigh virulence for mice as well as for conjunctivacells. Table 2 shows the gradual decrease inmouse virulence of the variant virus. At the in-

TABLE 1. Relationship of altered cytopathic effect(CPE) to antibody content of serum in culture

medium

No. of No. ofNo. of Antiherpes cultures cultures

Origin of serum sera virus antibody with with typi-tested in serum altered cal herpes

CPE CPE

Human 10 Negative 0 10Human 9 Positive 9 0Horse 2 Negative 0 2Calf 2 Negative 0 2Lamb 2 Negative 0 2Rabbit 2 Negative 0 2Rabbit 1 Positive 1 0

(pool of 5sera)

TABLE 2. Comparison of stock and variant herpesvirus by tissue culture and mouse titration during a

2-year period

Virus titer

VirusTissue culture Mouse brain

assay assay

Stock herpes virus* 5.5t 6.°tVariant, in culture

(month): 1 5.0 5.56 4.5 3.012 5.0 <1.018 4.75 <1.024 5.5 <1.0

* Titer remained roughly constant during the2-year period in both tissue culture and mice.

t Log1o ID5o per ml.

dicated periods, antibody was removed from themedium of carrier cultures and viral multiplica-tion was allowed to proceed to complete degenera-tion of the cell culture. The virus yield was thenassayed in tube cultures of conjunctiva cells andin mice. Virus obtained from the carrier culturesafter 12 months produced only occasional deathsin mice when inoculated in high concentration.Such deaths occurred over a period of 14 daysafter inoculation in contrast to death in 3 to 5days in mice receiving the virulent stock virus.

Growth of virulent and attenuated herpes virusin mice and in tissue culture. Multiplication ofthe virulent stock virus and attenuated virusfrom the carrier cultures was compared in bothmice and tissue culture in the following way.Groups of 3-week-old Swiss mice were inoculatedintracerebrally with 105 TCIDso of either the stockor attenuated virus. Four mice from each groupwere sacrificed at intervals of 6, 24, 48, 72, 120,and 144 hr after infection. The brains were re-moved, pooled, and ground in a Ten Broeckgrinder to make a 10% suspension in phosphate-buffered saline (pH 7.2). The suspensions werethen titrated in tissue culture. All mice inocu-lated with the attenuated virus remained healthythroughout the experiment. All mice receivingthe stock virus, however, died between 30 and50 hr after infection. As seen in Table 3, thestock virus showed an initial latent period fol-lowed by a period of rapid multiplication in themice, whereas the attenuated virus was undetect-able at 6 hr and failed to reappear during the6-day period.

1961] 501

on January 30, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

HINZE AND WALKER

TABLE 3. Growth of virulent and attenuated variantherpes virus in brains of mice

Virus titer in brainTime afterinoculation*

Virulent strain Attenuated strain

6 hr <1O1 ° <10 °1 day 1O-5.0 <10l.02 days 10-r. <1l.3 days All mice dead <l-- 0

4 days 10-1.o5 days <10- °6 days <10-1.

* Dose: 105 TCID5s intracerebrally.

Comparative growth curves for the two viruseswere carried out in tube cultures of normalconjunctiva cells. Identical sets of cultures con-

taining approximately 2.5 X 105 cells per tubewere inoculated with 10' TCID5o of virulent or

attenuated virus. The cultures were incubatedfor 2 hr at 37 C after which the inoculum was

removed and the cells washed 4 times withbalanced salt solution. One milliliter of Eagle'smedium containing 5% horse serum was addedto each tube and the cultures were again placedat 37 C. At hourly intervals for the first 24 hr,then at 30, 36, and 48 hr after infection, cellsand fluid were harvested from one tube of eachset and frozen at -70 C for titration in tissueculture.

Figure 3 demonstrates two differences in thebehavior of the stock and variant viruses intissue culture. Although the number of infectiousunits inoculated was the same for each virus,at the end of the 2-hr adsorption period theamount of residual detectable virus was more

than one log higher in cultures infected with thevariant virus than in those containing the stockvirus, and this difference persisted throughoutthe eclipse phase. Following the eclipse phasethe variant virus multiplied at a slower rate thanthe stock virus, reaching a peak at 48 hr, whereasthe stock virus attained an equivalent titer at30 hr. Calculations made from the straightportions of the curves give a doubling time ofapproximately 112 min for the stock virus and191.5 min for the variant strain.

Neutralization of attenuated virus by herpesantiserum and immunization of mice with attenu-ated strain. Antiserum prepared in rabbits againstthe virulent herpes strain, and each of two human

N Q.AVIRULENT9 0 STRAIN

3X 0

j-2

10 20 30 40 50HOURS

FIG. 3. Growth of virulent stock virus and theattenuated variant strain of herpes simplex virus inhuman conjunctiva cells in Eagle's medium and5% horse serum. Each culture was inoculated with10 TCIDo of virus. Zero on the horizontal axisrepresents the end of a 2-hr adsorption period at37 C.

sera containing antiherpes antibody neutralizedsimilar numbers of plaque-forming units ofvirulent and attenuated virus. Similarly, anti-serum prepared in rabbits against the variantvirus neutralized approximately equal numbers ofplaque-forming units of virulent and variantvirus. Extensive comparison of the antigenicmakeup of the variant and parent strains by theseveral available methods was not done, but theneutralization tests indicated that, the variantstrain possessed the antigenic characteristics ofherpes virus and that there was no major anti-genic difference between the two strains.Although injection of the attenuated virus

into the brains of mice resulted in little or noyield of infectious virus, it appeared possiblethat such inoculation might, nonetheless, renderthe animals resistant to challenge with virulentvirus. An experiment was performed to test thispossibility. Thirty mice were inoculated intra-cerebrally with 105 TCID5o of living attenuatedvirus. A similar group of mice was inoculatedwith virus heated to 56 C for 30 min. A thirdgroup was inoculated with normal tissue culturefluid and cells and a fourth group was left as un-treated controls. Four weeks after the initialinoculation, all four groups were challengedwith serial 10-fold dilutions of virulent herpesvirus. Five mice were inoculated per virus dilu-tion in each group. The results shown in Table 4indicate clearly that treatment with the living

502 [VOL. 82

on January 30, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

VARIATION OF HERPES SIMPLEX VIRUS

TABLE 4. Immunization of mice with livingattenuated virus

Titer of stand-Immunizing antigen ard challengevirus in test

groups of mice

No immunization.................. 5.0*Uninfected tissue culture cells and

fluid........................... 5.0Heated attenuated virus........... 4.5Living attenuated virus............ <1.0

* Log,o TCID50 per ml.

attenuated virus enabled the mice to resist morethan 104 LDno of the challenge virus. Heating theattenuated virus to 56 C, however, destroyedits immunizing capacity.

DISCUSSION

A gradual but definite change in certaincharacteristics of herpes virus in persistentlyinfected cultures has been observed in thisstudy. The demonstration of neutralization ofthe variant virus by herpes virus antiserum andthe capacity of the variant to immunize miceagainst herpes virus show that this change was

not due to the emergence of an unrelated con-

taminating virus from the cells or medium, butthat it arose from the selection of a variant fromthe virus population originally introduced intothe cultures. Studies reported by Scott and co-

workers (Gray, Tokumaru, and Scott, 1958;Scott and McLeod, 1959; Scott, McLeod, andTokumaru, 1961), and by Finkelstein, Allen,and Sulkin (1959), indicating that most culturesof herpes virus are composed of populationsproducing two or three types of CPE, tend tosupport the assumption that the changes ob-served in the present study resulted from selec-tive processes. The variant observed in thepresent study appeared, in fact, to be similarto Scott's "proliferative strain" (Gray et al.,1958; Scott and McLeod, 1959) in the characterof the CPE produced. There were distinct dif-ferences, however, since the virus of this studyproduced proliferative foci only in the presence

of antiserum, was relatively slow in its multipli-cation rate, and was avirulent for mice. Scott'sproliferative strain, on the other hand, producedproliferative foci in medium free of antibody, was

similar in multiplication rate to lines of virus

producing other patterns of CPE, and retainedits virulence for mice (Scott et al., 1961).

In another study concerned with herpes sim-plex in carrier cultures, Hoggan and Roizman(1959) found that a variant emerged after con-tinued passage of herpes virus in FL cells. Theirvariant also was similar to one described byScott and co-workers, but it was one that pro-duced large multinucleated giant cells ratherthan the proliferative foci seen in the presentstudy. There was, perhaps, some difference fromScott's "giant-cell line" of virus, in that thesyncytial CPE of the variant of Hoggan andRoizman was produced in the presence of anti-body (Hoggan and Roizman, 1959; Hoggan,Roizman, and Roane, 1961), whereas that ofScott was seen in the absence of antibody.Change in virulence or in other characteristics

of the virus has not been prominently mentionedin studies of virus carrier systems (Ginsberg,1958). Wheeler and Canby (1959), for instance,studied carrier cultures of herpes virus usingantibody containing culture procedures not verydifferent from those used by us and by Hogganand Roizman, and yet they apparently did notobserve changes in the characteristics of thevirus. In some carrier systems studied by otherinvestigators, the virus has ceased to produceCPE, but this has usually been attributed to anincrease in cell resistance by cell selection or someother mechanism, such as interference. In viewof the variability of viruses, however, and thetendency of culture systems to select for variants,it seems surprising that changes in the viruspopulation were not observed frequently in thecourse of the several studies now reported. Ourstudy and that of Hoggan and Roizman indicatethat change in the virus population can occurwith herpes virus, and it seems likely that similarvariation will be found in carrier cultures in-volving other viruses if it is looked for.Major attention in the present study was given

to the change in viral characteristics, and suffi-cient data have not yet been collected on otheraspects of the cell-virus relationship to allowsatisfactory evaluation of the importance of thevariant virus characteristics in maintenance ofthe carrier state. It is likely, however, since thevirus made its appearance within the selectivepressures of the carrier cultures, that one ormore of its characteristics may be important tothe stability of the carrier system.

1961] 503

on January 30, 2021 by guesthttp://jb.asm

.org/D

ownloaded from

HINZE AND WALKER

ACKNOWLEDGMENT

This work was supported by a research grant(E 1214) from the National Institute of Allergyand Infectious Diseases, U. S. Public HealthService.

LITERATURE CITED

CHANG, R. S. 1954. Continuous subcultivation ofepithelial-like cells from normal humantissues. Proc. Soc. Exptl. Biol. Med. 87:440-443.

EAGLE, H. 1955a. Propagation in a fluid mediumof a human epidermoid carcinoma, strain KB.Proc. Soc. Exptl. Biol. Med. 89:362-364.

EAGLE, H. 1955b. Nutrition needs of mammaliancells in tissue culture. Science 122:501-504.

FINKELSTEIN, R. A., R. ALLEN, AND S. E. SULKIN.1959. Inhibition of herpes simplex virus bynormal serum: Its relationship to the pro-

perdin system. J. Infectious Diseases 104:184-192.

GINSBERG, H. S. 1958. The significance of theviral carrier state in tissue culture systems.Progr. in Med. Virol. 1:36-58.

GRAY, A., T. TOKUMARU, AND T. F. McN. ScoTT.1958. Different cytopathogenic effects ob-served in HeLa cells infected with herpessimplex virus. Arch. ges. Virusforsch. 8:59-76.

IOGGAN, M. D., AND B. ROIZMAN. 1959. Theisolation and properties of a variant of herpessimplex producing multinucleated giant cells

in monolayer cultures in the presence ofantibody. Am. J. Hyg. 70:208-219.

HOGGAN, M. D., B. ROIZMAN, AND P. R. ROANE.1961. Further studies of variants of herpessimplex virus that produce syncytia or pock-like lesions in cell cultures. Am. J. Hyg.73:114-122.

MORGAN, J. F., H. J. MORTON, AND R. C. PARKER.1950. Nutrition of animal cells in tissueculture. I. Initial studies on a syntheticmedium. Proc. Soc. Exptl. Biol. Med. 73:1-8.

REED, L. J.. AND H. MUENCH. 1938. A simplemethod of estimating fifty per cent endpoints.Am. J. Hyg. 27:493-496.

SCHERER, W. F., J. T. SYVERTON, AND G. 0. GEY.1953. Studies on the propagation in vitro ofpoliomyelitis viruses. IV. Viral multiplicationin a stable strain of human malignant epi-thelial cells (strain HeLa) derived from anepidermoid carcinoma of the cervix. J. Exptl.Med. 97:695-709.

SCOTT, T. F., AND D. L. MCLEOD. 1959. Cellularresponses to infection with strains of herpessimplex virus. Ann. N. Y. Acad. Sci. 8:118-128.

SCOTT, T. F. McN., D. L. MCLEOD, AND T. TOKU-MARU. 1961. A biologic comparison of twostrains of Herpesvirus hominis. J. Immunol.86:1-12.

WHEELER, C. E., AND C. M. CANBY. 1959. In-fluence of specific antibody on herpes simplexinfections in tissue culture. A. M. A. Arch.Dermatol. 79:86-95.

r504 [VOL. 82

on January 30, 2021 by guesthttp://jb.asm

.org/D

ownloaded from