Persistent Varicella-Zoster Virus Infection in a Human Rhabdomyosarcoma Cell Line and Recovery of a Plaque Variant

Transcription

1 NFECTON AND MMUNTY, JUlY 1982, p /82/ $02.00/0 Vol. 37. No. 1 Persistent Varicella-Zoster Virus nfection in a Human Rhabdomyosarcoma Cell Line and Recovery of a Plaque Variant JEFFREY P. LTS,l* JAMES VETTE,'t GABREL A. CASTELLANO,' DAVD L. MADDEN,2 AND JOHN L. SEVER2 Microbiological Associates, Bethesda, Maryland ; and nfectioius Diseases Branch, National nstitute of Neurological and Communicatiive Disorders and Stroke, Bethlesda, Maryland Received 18 November 1981/Accepted 26 March 1982 Varicella-zoster virus (VZV) has been found to persistently infect the human rhabdomyosarcoma cell line A204. nfectious center assays and fluorescent antibody staining demonstrated continuous production of infectious VZV and viral antigen. The level of infection determined by fluorescent antibody staining was variable, and usually only a small percentage of the cells were capable of producing plaques in permissive fibroblasts. The extent of infection was similar in cell cultures passaged at split ratios of 1:2 or 1:10 and grown at 33 or 37 C. VZV recovered from A204 cells several months after establishment of the persistent infection had markedly increased syncytia-forming activity as compared with the parental VZV grown in human diploid fibroblast cells and the three monkey kidney-derived cell lines Vero, CV-1, and MA104. The expression of this altered phenotype continued after serial passage of the cell-associated virus in human diploid fibroblast and Vero cells. Consequently, we designated the reisolated VZV as plaque variant A. The buoyant densities of VZV plaque variant A and VZV DNAs in CsCl gradients were indistinguishable. Varicella-zoster virus (VZV), like other human herpesviruses, has a propensity for persisting in the host, most likely in a latent state, for many years after primary infection (12). Studies of persistent infections with herpes simplex virus (2, 3, 13-15) and cytomegalovirus (4, 5, 19) in cell culture elucidated some fundamental mechanisms which allow these herpesviruses to persist and cause recurrent illnesses in humans. Analogous in vitro studies of VZV persistence have not been reported, primarily because VZV is both stringently cell associated and host cell restricted. As a result, little or no infectious VZV is spontaneously released. Significant numbers of infectious particles cannot be obtained by cell disruption of permissive human diploid embryo fibroblast (HDEF), epithelial, or melanoma cells (7, 12). n a recent study, it was reported that VZV can persistently infect Epstein-Barr virus-transformed lymphoblastoid Raji cells (11); however, the interpretation of results with these cells is encumbered by the influence of a resident Epstein-Barr virus genome. We investigated the ability of VZV to replicate in the rhabdomyosarcoma cell line A204 (6), since these cells are continuous and, like HDEF t Present address: College of Dental Medicine, University of Maryland, Baltimore, MD cells, are of mesodermal origin. We found that VZV infection in these cells was self-limited rather than fully permissive or lytic. Consequently, a long-term persistent infection was established in the A204 cell culture. We designated this VZV-infected cell line VZV-A204. We also report in this paper that VZV obtained from the A204 cells induced a marked increase in syncytium formation in HDEF and several monkey kidney-derived cell lines as compared with parental VZV. We designated the VZV plaque variant as VZVpvA. (This paper was presented in part at the 81st Annual Meeting of the American Society for Microbiology, 1 to 6 March 1981, in Dallas, Tex.) MATERALS AND METHODS Cells and virus. Primary HDEF cell lines HR-218 (Flow Laboratories, nc.) and HFS-15 (HEM Research) and the monkey kidney-derived Vero, CV-1, and MA104 cells were grown at 37 C in Eagle minimal essential medium with 10% fetal bovine serum and 1% glutamine. The human rhabdomyosarcoma cell line designated A204 (HEM Research) was propagated in Dulbecco modified Eagle medium with 10% fetal bovine serum and 1% glutamine. Cells and virus were periodically screened for the presence of mycoplasmas by fluorescent staining with 4'-6' diamidino-2-phenylindole (17). 350

2 VOL. 37, 1982 The VZV isolate CaQu was originally obtained from Nathalie Schmidt, and was propagated by cell-associated passage of infected HDEF cells (10). The persistent infection of A204 cells was initiated by mixing 2 x 106 trypsinized, VZV-infected HDEF cells displaying 100% cytopathic effects with an equal number of A204 cells. The mixed cells were centrifuged at 1,500 rpm for 5 min, incubated at room temperature for 30 min, and then seeded into a 75-cm2 plastic flask. The cells were passaged twice weekly at a 1:2 split ratio for 3 weeks and once weekly thereafter. All cell counting was done with a hemacytometer, and trypan blueexcluding cells were considered viable. nfectious center and cell-free virus assays. Assay of the number of plaque-forming cells was performed in a manner similar to that previously described (9). Cultures of VZV-A204 cells contained both adherent, weakly attached cells and nonadherent, free-floating clusters of cells. The adherent cells were detached by vigorous pipetting and pooled with nonadherent cells, except during one study in which each population was assayed separately. The number of A204 cells containing infectious VZV was determined by seeding 104 to 107 cells into the culture media of HDEF cells in 150- cm2 flasks incubated at 37 C. Not all of the cells were able to attach to the plastic or the fibroblast cell monolayer; however, all of the cells could settle and contact the monolayer and potentially induce a plaque. No overlay was necessary because of the cell-associated nature of VZV replication. Plaques were counted 8 to 10 days postinfection (p.i.) after crystal violet staining. When plaques were very numerous, a plaque count was made of at least a 40-cm2 area of the cell monolayer with the aid of an inverted microscope. The cells were harvested as described above to obtain cell-free VZV for the infectious center assay. Pooled cells were centrifuged at 1,500 rpm for 5 min and suspended in 2.0 ml of Eagle minimal essential medium with 10% fetal bovine serum per 150-cm2 flask. The cells were disrupted by pulsed-probe sonication for 20 s. Cell debris was removed by centrifugation at 2,000 rpm for 10 min, and the cell-free supernatant was assayed for VZV at several dilutions by adsorption at 37 C to HDEF cells in 150-cm2 flasks for 1 h. Plaques were counted as described above. The assay of infectious centers was performed at 37 C for those experiments in which cell cultures were incubated at 33 C. FAS. The study of VZV antigen expression in HDEF and Vero cells and estimation of the percentage of A204 cells producing VZV antigen(s) was made by indirect fluorescent antibody staining (FAS) of cell smears with VZV immune serum obtained from an individual with recent herpes zoster. Herpes simplex type 1 antibody- and cytomegalovirus antibody-positive sera were tested for reactivity to VZV-A204 cells to assay the specificity of the immunofluorescence. Fluorescein-conjugated goat anti-human immunoglobulin G (gg) (Litton Bionetics) was used as the second antibody. Cell smears from VZV-A204 and A204 cultures were prepared on glass microscope slides and rapidly air dried before being fixed in acetone at -20 C for 10 min. Serum diluted 1:10 in Dulbecco phosphatebuffered saline (PBS) was adsorbed to the fixed cells for 30 min at room temperature. After the slides were rinsed three times for 5 min each in PBS, the conjugate, also diluted 1:10 in PBS, was incubated and PERSSTENT VARCELLA-ZOSTER VRUS NFECTON 351 rinsed in the same manner. Cover slips were mounted with glycerol-pbs (1:1). Cells were viewed in darkfield illumination with a Zeiss microscope with a highpressure mercury lamp. The percentage of VZV antigen-bearing cells was determined after counting 400 to 1,500 cells at 250x magnification. DNA radiolabeling and CsC buoyant density gradients. HDEF cells were infected with VZV recovered from A204 cells (VZVpvA) or parental VZV by mixing infected with uninfected HDEF cells as previously described (10). When cell monolayers contained 80 to 100% cytopathic effects, [3H]thymidine (10 p.ci/ml, 40 to 60 p.ci/mmol) or [14C]thymidine (5,uCi/ml, 57 mci/ mmol) was added to the culture medium, and radiolabeling was carried out for 18 to 36 h. Radiolabeled DNA was extracted from the infected cells by the method of Hirt (8) in a manner similar to that previously described by ltis et al. (10). Briefly, the cell monolayers were washed three times with Tris buffer (ph 7.2), and 4 ml of 10 mm Tris buffer (ph 8.0) containing 1 mm EDTA and 0.6% sodium dodecyl sulfate was added to the flask. The cell lysate was held at 37 C for 10 min before adding 5.0 M NaCl directly to the lysate, which was then gently poured into a 15-ml glass Corex tube and kept at 4 C for 12 to 18 h. The supernatant containing VZV DNA was separated from the precipitate by centrifugation at 17,000 x g for 45 min. Precipitable radioactivity was determined as described below for the CsCl gradient fractions. 3Hlabeled VZVpvA DNA was cosedimented with 14Clabeled parental VZV DNA in a 5-ml CsCl gradient (starting density, g/cm3) by centrifugation at 36,000 rpm in a fixed-angle 50.3 rotor. 3H-labeled VZVpvA and VZV DNAs sedimented with '4C-labeled HDEF cell DNA served as a control to establish that VZV DNA was not cell DNA. The buoyant density of VZV DNA was previously shown to be g/cm3 as compared with a human cell DNA density of g/cm3 (10). Gradient fractions were dripped from the bottom of the tube onto Whatman filter paper disks. The filters were batch precipitated in cold 5% trichloroacetic acid and washed in 95% ethanol and then in acetone. Radioactivity was determined in a Beckman LS8000 liquid scintillation counter with Biofluor (New England Nuclear Corp.). RESULTS VZV infection of A204 cells. nfection of A204 cells with VZV was accomplished with cellassociated virus. This method of infection insured that most of the A204 cells were infected at a multiplicity greater than 1. The infected HDEF cell culture used as the inoculum contained 100% VZV cytopathic effects, and none of the cells excluded trypan blue. We seeded separately a number of infected cells equal to that mixed with the A204 cells and from the same culture to ascertain that no viable HDEF cells were present in the inoculum. No cells were observed attached to the plastic on the following day. We noted previously (unpublished data) that VZV kills virtually every cell in heavily infected HDEF cell cultures. The VZV-infected A204 cells were passaged

3 352 LTS ET AL. FG. 1. Fluorescence photomicrograph of VZV-infected A204 cells after FAS with herpes zoster convalescent serum. (x600). twice weekly at a 1:2 split ratio for 3 weeks. At this time, the FAS and infectious center assays indicated that a small number of VZV-infected cells were present in the cultures. At 6 weeks p.i. we determined the number of infected cells weekly by indirect FAS, infectious center assay, or both. The micrograph (Fig. 1) shows the prominent nuclear and cytoplasmic staining typical of smears of VZV-infected A204 cells. The extent of infection in VZV-A204 cultures was determined in a long-term study from 6 to 39 weeks p.i. The percentage of infectious centers never exceeded 10%, whereas FAS indicated that as many as 17 to 18% of the cells contained VZV antigen at 17 and 33 weeks p.i., respectively (Fig. 2). FAS was the more sensitive assay for estimating the percentage of virus-infected cells (Fig. 2). Cell-free VZV (approximately 2 x 103 to 8 x 103 PFU) was obtained by sonication of VZV-A204 cells at the times indicated. A cellfree virus yield of three logs was consistent with the low level of infection in VZV-A204 cells and with that which can be obtained in permissive HDEF cells (7). nfectious cell-free VZV was released into the culture medium and was not detected in excess of 25 PFU/ml. We performed several growth curve assays with VZV-A204 cells at various times p.i. There was no difference between the growth rate of VZV-A204 cells and of uninfected A204 cells NFECT. MMUN. whether seeded at 2.5 x 105 or 4 x 105 cells per ml (Fig. 3). VZV-A204 and A204 cells seeded at these concentrations generally had a doubling time of 2 to 3 days. The possibility of obtaining a higher level of VZV infection in VZV-A204 or A204 cells was examined. The original experiment described above was repeated in a similar manner to obtain VZV-A20411 cells, except that 2.5 times more VZV-infected HDEF cells were used to infect the A204 cells. Additionally, 2.5 x 106 VZV_ A204 cells were infected with an equal number of VZV-infected HDEF cells to obtain the superinfected VZV-A204(Su) cells. The extent of infection in these two additional cell lines was examined and compared with the original VZV- A204 cells over a 10-week period (Table 1). Although 2.5 times more VZV-infected HDEF cells were used to infect the VZV-A20411 cells, the level of infection as compared with VZV- A204 cells was not consistently greater. The greatest percentage of antigen-containing VZV- A204 cells observed over a 39-week period (Fig. 2) was comparable to the levels observed in VZV-A2041 at 2, 3, and 4 weeks p.i. Compared with VZV-A204 cells, the superinfected VZV- A204(Su) cells contained a higher percentage of infectious centers at 2 and 3 weeks p.i. and of antigen-containing cells at 2, 3, and 4 weeks p.i. This higher level of infection was not maintained; the percentage of infectious centers was equivalent at 4 weeks p.i., and the number of antigen-containing cells was approximately the same by 6 weeks p.i. A very low percentage of infectious centers was observed in each cell line by 10 weeks p.i. Viability assays performed at 2, 3, and 4 weeks p.i. showed no apparent correlation between the level of infection and viability, as determined by cell counting with trypan blue. Assay of infection in adherent and nonadherent VZV-A204 cells. VZV-A204 cell cultures contained loosely attached and floating clusters of viable cells which were microscopically indistinguishable from those of the parent A204 cell line. Morphologically, the attached cells were distinct from typical fibroblast cells and comprised both rounded and spindle-shaped cells (Fig. 4). We determined whether the adherent cells contained a significantly different proportion of infected cells as compared with nonadherent cells in the same culture flask. No significant difference in the extent of infection was detected during a 3- week period (weeks 8, 9, and 10 p.i.) by FAS, infectious center assay, or cell-free virus yield. The viability of both cell types was similar and not significantly different from uninfected A204 cells (data not shown). Effect of growth temperature and cell passage split ratio on VZV expression in VZV-A204 cells. We wanted to determine whether persistently

4 VOL. 37, 1982 PERSSTENT VARCELLA-ZOSTER VRUS NFECTON 353 Cn a) a).) 19H 17 A A 7 l x 5 Ā 3D0- -0 V CD 0 CD -0 CD) u) he.-c k 7~ 5F Weeks post-infection FG. 2. Extent of VZV infection in persistently infected A204 cells as determined by FAS (0), infectious center assay (0), and cell-free virus plaque assay (A). infected VZV-A204 cells could be induced to produce a higher level of VZV expression by altering the growth conditions (incubation temperature or split ratio) of the cells. Cummings and coworkers (2) showed that in a lymphoblastoid cell line persistently infected with herpes simplex virus type 1, increased virus production occurs after the cell cultures are shifted from 37 to 34 C. The results of an 8-week study to determine whether cells grown at 33 rather than 37 C would contain a greater proportion of infected cells are shown in Table 2. Growing VZV- A204 cells at 33 rather than 37 C did not increase the low level of VZV infection in these cells. VZV-A204 cells which were obtained directly from the 1:2-passaged cultures were maintained at a 1:10 split ratio starting in week 3 of this study to examine the effect of the passage ratio on virus expression. t has been shown that the passage ratio can affect the level of viral antigen expression in a Vero cell line persistently infected with the papovavirus HD (18). These 1:10- passaged cultures contained more VZV-infected cells than did those passaged at 1:2 during the course of this study; however, in general the proportion of virus-infected cells in cultures passaged at 1:10 did not reach a level greater than that attained in cultures passaged at 1:2 (Fig. 2). VZV-A204 cells are presently being maintained at a split ratio of 1:10, and low, fluctuating levels of infected cells have been demonstrable continuously for 20 months. nterestingly, after week 7 of this study we were no longer able to detect infectious centers or antigen-containing cells in the original VZV- A204 cells which had been maintained for 39 weeks at a 1:2 split ratio (Fig. 2). We continued to assay these cells by FAS for an additional 5 weeks, but cells containing VZV antigen were not detected. Comparison of VZV from A204 cells (VZVpvA) with parental VZV by plaque phenotype and FAS. We had noted on plaque assays with cellfree VZV obtained from VZV-A204 cells at

5 354 LTS ET AL t Hours post-seed FG. 3. Comparison of the growth rate of persistently infected A204 cells (VZV-A204) and uninfected A204 cells. VZV-A204 cells seeded at 2.5 x 105 (0) or 4 x 105 (A) cells per ml and uninfected A204 cells seeded at 2.5 x 105 (0) or 4 x 105 (A) cells per ml. various times p.i. that the majority of plaques were variant or atypical because syncytium formation was more pronounced as compared with any wild-type or parental VZV plaque we had previously observed. We determined that VZVpvA caused the pronounced syncytial activity in HDEF, Vero, CV-1, and MA104 cells. Because of this difference in plaque morphology in these cell lines, we concluded that a VZV plaque variant (designated VZVpvA) had been selected for or induced by passage in A204 cells. FAS was performed to confirm that VZVpvA was antigenically similar to parental VZV. Herpes simplex virus antibody- and cytomegalovirus antibody-positive sera, tested to demonstrate specificity, did not react to VZVpvA antigen by FAS. mmunofluorescence micrographs showed that herpes zoster convalescent serum reacted with VZVpvA and parental VZV antigens in HDEF and Vero cells (Fig. 5). The syncytial plaque phenotype of VZVpvA is readily apparent. Uninfected A204 cells were unreactive to the herpes zoster serum. Buoyant density comparison of parental and VZVpvA DNAs. We directly compared the buoyant densities of VZV and VZVpvA DNAs to more definitively differentiate the wild-type virus from the variant. However, we were not able to detect any difference in the buoyant density between VZVpvA and parental VZV DNAs by CsCl isopycnic centrifugation (Fig. 6). DNA obtained from VZVpvA at 7 months and 1 year p.i. (Fig. 6A and C, respectively) was indistinguishable in buoyant density from the parental VZV DNA. The radiolabeled DNA profiles (Fig. TABLE 1. Level of VZV infectivity in three different VZV-infected A204 cell cultures % Positive at wk p.i.: Cell-free Assay Cell line virus at wk p.i. nfectious center VZV-A204" VZV-A204b VZV-A204(Su)' Fluorescent cell VZV-A VZV-A VZV-A204(Su) Viable cell VZV-A VZV-A VZV-A204(Su) Control A NFECT. MMUN. Cell free virus VZV-A x 103 VZV-A x 103 VZV-A204(Su) 3.2 x 102 a The original VZV-infected A204 cell line was already 12 weeks p.i. at the beginning of this experiment. b A second VZV-infected C A204 cell line. Original VZV-infected A204 cell line superinfected with the same VZV isolate (CaQu).

6 VOL. 37, 1982 PERSSTENT VARCELLA-ZOSTER VRUS NFECTON 355 ~ ~ ~ ~~ A, 7- FG 4 Photomtcrograph showing the morphology of perststently infected VZV-A204 cells in culture. Rounded and spindle-shaped cells are apparent. Uninfected A204 cells are morphologically indistinguishable. (x 130) 6B and D) showed the characteristic buoyant position of VZVpvA and parental VZV DNAs relative to HDEF cell DNA previously reported (10). The one-fraction difference (Fig. 6A) was not considered significant because it was consistent with the isotope effect caused by the heavier 14 C radionuclide (1). DSCUSSON The data presented in this report indicate that VZV can persistently infect the human rhabdomyosarcoma cell line A204. Levonton-Kriss et al. (11) previously reported persistent VZV infection in Epstein-Barr virus-transformed Rajii cells; however, to our knowledge VZV-A204 is the first nonlymphoid persistent VZV infection in cell culture. The continued presence of infectious virus and antigen expression in VZV-A204 cells for more than 19 months signifies a persistent infection, although it is possible that latent VZV DNA (full or partial genome) may reside in some of the cells. The reason for the low level of VZV infection in A204 cells is not clear. Although approximate- TABLE 2. Effect of incubation temperature and cell split ratio on level of infection in VZV-A204 cells' Wk ratio' Split temp Growth ( C) centers nfectious (%) 1 1:2 1: : : : : : : : : : :2 37 <7.14 x : : : x 1-4 1: : :2 37 <5.00 x : a This experiment was carried out for an 8-week period 7 months after the A204 cells were infected with VzV. b Cells were passaged at 1:2 or 1:10 by removing 50 or 90% of the cells and replenishing the culture with medium to the original volume. ly 34% of the cells were shown to be capable of expressing VZV antigen (Table 1), this level of infection was not maintained. Using the assay of Rubenstein et al. (16), we did not detect interferon in the supernatant from a VZV-A204 culture with approximately 16% antigen-containing cells (data not shown). Cellular growth curve studies indicated that the presence of persistent VZV expression did not lead to a detectable alteration of the cell growth rate, since the growth of VZV- A204 tumor cells was comparable with that of control A204 cells. ncreasing the level of infection in VZV-A204 cells would greatly improve our ability to study these cells; however, attempts to do this by growing them at 33 C, passing them at a 1:10 split ratio, and superinfecting them were not successful. The higher level of infection generally determined by FAS compared with the infectious center assay suggests that VZV expression in some cells did not lead to complete virion production. The plaque-inducing efficiency of infected cells was probably close to 100%, since nearly all of the cells were capable of settling onto the monolayer. The persistent VZV infec-

7 356 LTS ET AL. NFECT. MMUN. > -x _. v).- H _._8 r _ L )c: to = C. a) = -ov -0 ' o= ) C.) -o _ 0 _ f >a 0D -

8 VOL. 37, 1982 PERSSTENT VARCELLA-ZOSTER VRUS NFECTON x E ch CX 08 x E H N x.36 E C) 30v Fraction number FG. 6. Comparison of the buoyant density of parental VZV and VZVpvA DNAs by CsC gradient isopycnic centrifugation. Centrifugation was performed at 36,000 rpm at 20 C for 65 h in a Beckman 50.3 fixed-angle rotor. (A) 3H-labeled VZVpvA DNA (0) comixed with "4C-labeled parental VZV DNA (0). VZVpvA DNA was prepared from virus isolated from A204 cells 7 months p.i. and grown in HDEF cells. (B) 3H-labeled parental VZV DNA (0) comixed with 14C-labeled HDEF cell DNA (0). (C) 3H-labeled VZVpvA DNA (0) comixed with 14C-labeled parental VZV DNA (0). VZVpvA DNA was prepared from virus isolated from A204 cells 1 year after infection and grown in HDEF cells. (D) 3H-labeled VZVpvA DNA (0) comixed with t4c- labeled HDEF cell DNA (0). tion of A204 cells did not appear to alter the number of viable cells as compared with A204 control cells. Even at rather high levels of VZV antigen expression (Table 1), cell viability was comparable to uninfected A204 cells. Thus, we could not determine whether VZV expression affects A204 cell viability. The effect, if any, would seem to be minimal. t is possible that VZV expression in an infected cell fluctuates and may not always cause cell death. The rapid growth of the cells may offset the limited or prolonged replication of VZV in A204 cells. We have shown that VZV reisolated from VZV-A204 cells is a plaque variant when compared with the input parental VZV. The nature of the altered phenotype of VZVpvA is quantitative rather than qualitative, since wild-type VZV normally induces a limited amount of syncytium formation in cell culture. We doubt that preparations of cell-free VZVpvA were homogenous, because plaque purification of VZV has not been achieved due to the stringent cell-associated nature of the virus. VZVpvA serially passaged in HDEF and Vero cells maintained its variant syncytial phenotype. The fact that VZVpvA was capable of inducing pronounced syncytium formation in four cell lines (HDEF, Vero, CV-1, and MA104) in which parental VZV was not

9 358 LTS ET AL. indicates that the altered phenotype is not host cell induced. Depending on the level of VZV expression, VZV-A204 cell cultures contained various numbers of syncytia evident by light microscopy and FAS, although these were much less pronounced in VZV-A204 cells than in HDEF or Vero cells. mmunofluorescent staining in HDEF, Vero, and A204 cells demonstrated the gross antigenic similarity between the parental and variant VZV. VZVpvA and wild-type VZV did not grow well at 39 C in HDEF or Vero cells; however, the syncytial plaque phenotype of VZVpvA, although markedly reduced, was apparent at 39 C. Likewise, no significant differences in plaquing between VZVpvA and wild-type VZV were noted at 33 C (J. P. ltis, unpublished data). The variant plaque phenotype was induced by VZVpvA but not by wild-type VZV in HDEF and in three monkey kidney cell lines. There are two possible explanations for the origin of this variant. (i) Some A204 cell-induced modification of VZV DNA may have occurred, or (ii) a unique VZV variant was selected from a heterogeneous population of virions in the original VZV inoculum. CsCl gradient analysis did not detect any difference in the buoyant density of VZVpvA and parental VZV DNAs; however, only a relatively gross genomic G-C variation would be detected by this procedure. Restriction endonuclease cleavage analysis of VZVpvA DNA, currently in progress, may indicate whether cleavage site alterations are present as a result of genomic variation. The increased formation of syncytia caused by VZVpvA is probably related to the production of a unique or modified virion glycoprotein necessary for cell fusion. Further studies with polyacrylamide gel electrophoresis of immunoprecipitates of VZVpvA- and wild-type VZV-coded glycoproteins should help to determine the nature of the variant VZV phenotype. The VZV-A204 cell line is a useful model, heretofore unavailable, for studying the mechanism(s) of VZV persistence in cell culture. f significant levels of infection can be achieved in VZV-A204 cells, e.g., by phorbol ester or corticosteroid induction, then a convenient resource will be readily available for clinical and basic studies of this important human pathogen. We hope that additional studies with VZVpvA will contribute to our understanding of wild-type vzv. ACKNOWLEDGMENTS We thank Carol Cleghorn for technical assistance and June Jackson for preparation of the manuscript. We also thank Eugene Majors for helpful discussion of the manuscript. NFECT. MMUN. This study was supported by Public Health Service contract no. NO1-NS from the National nstitute of Neurological and Communicative Disorders and Stroke. LTERATURE CTED 1. Cassai, E., and S. Bachenheimer Effect of isotopic label on buoyant density determinations of viral DNA in the preparative ultracentrifuge. J. Virol. 11: Cummings, P. J., R. J. Lakomy, and C. R. Rinaldo, Jr Characterization of herpes simplex virus persistence in a human T lymphoblastoid cell line. nfect. mmun. 34: Doller, E., J. Aucker, and A. Weissbach Persistence of herpes simplex virus type 1 in rat neurotumor cells. J. Virol. 29: Furukawa, T Cell cycle-dependent chronic infection of human cytomegalovirus in human osteogenic sarcoma cells. J. Gen. Virol. 45: Furukawa, T., N. Yoshimura, L. Jean, and S. S. Plotkin Chronically persistent infection with human cytomegalovirus in human lymphoblasts. J. nfect. Dis. 139: Giard, D. J., S. A. Aaronson, G. J. Todaro, P. Arnstein, J. H. Kersey, H. Dosik, and W. Parks Propagation of human tumors: establishment of cell lines derived from a series of solid tumors. J. Natl. Cancer nst. 51: Grose, C., and P. A. Brunell Varicella-zoster virus: isolation and propagation in human melanoma cells at 36 and 32 C. nfect. mmun. 19: Hirt, B Selective extraction of polyoma DNA from infected mouse cell cultures. J. Mol. Biol. 26: ltis, J. P., T.-S. Lin, W. H. Prusoff, and F. Rapp Effect of 5-iodo-5'-amino-2',5'-dideoxyuridine on varicella-zoster virus in vitro. Antimicrob. Agents Chemother. 16: ltis, J. P., Oakes, J. E., R. W. Hyman, and F. Rapp Comparison of the DNAs of varicella-zoster viruses isolated from clinical cases of varicella and herpes zoster. Virology 82: Levonton-Kriss, S., T. Gotlieb-Sematsky, A. Vonsover, and A. Smetana nfection and persistence of varicella-zoster virus in lymphoblastoid Raji cell line. Med. Microbiol. mmunol. 167: Nahmias, A. J., W. R. Dowdle, and R. F. Schinazi The human herpesvirus. Elsevier/North-Holland Publishing Co., New York. 13. Nishiyama, Y., and F. Rapp Regulation of persistent infection with herpes simplex virus in vitro by hydrocortisone. J. Virol. 31: Rinaldo, C. R., Jr., B. S. Richter, P. H. Black, and M. S. Hirsch Persistent infection of human lymphoid and myeloid cell lines with herpes simplex virus. nfect. mmun. 25: Roumillat, L. F., P. M. Feorino, D. D. Caplan, and P. D. Lukert Analysis and characterization of herpes simplex virus after its persistence in a lymphoblastoid cell line for 15 months. nfect. mmun. 29: Rubenstein, S., P. C. Familletti, and S. J. Pestka, Convenient assay for interferons. J. Virol. 37: Russell, W. C., C. Newman, and D. H. Williamson A simple cytochemical technique for demonstration of DNA in cells infected with mycoplasma and viruses. Nature (London) 253: Steffen, M., P. Krieg, M. Pernfuss, E. Sauer, V. Eisinger, and G. Sauer Growth dynamics of a latent primate papovavirus. J. Virol. 35: Tocci, M. J., and S. C. St. Jeor Persistence and replication of the human cytomegalovirus genome in lymphoblastoid cells of B and T origin. Virology 96: