Neuroinvasion by herpes simplex virus. An in vitro model for characterization of neurovirulent strains

Transcription

1 Journal of General Virology (1990), 71, Printed in Great Britain 405 Neuroinvasion by herpes simplex virus. An in vitro model for characterization of neurovirulent strains Tomas Bergstr6m* and Erik Lycke Department of Virology, Institute of Medical Microbiology, University of G6teborg, Guldhedsgatan lob, S G6teborg, Sweden A cell culture technique with rat dorsal root ganglion neurons cultured in a two-chamber system allowing exposure of neuritic extensions or neuronal cell soma to herpes simplex virus (HSV) infection was exploited for studies of viral neuroinvasiveness. Four groups of HSV strains were investigated. Type 1 strains were isolated from cases of either encephalitis or mucocutaneous infections and type 2 strains were from cases of meningitis or genital infections. The neuroinvasiveness of the type 1 encephalitis strains was significantly greater than the neuroinvasiveness calculated for the type 1 strains of cutaneous origin or for either of the type 2 strains. From our evaluation of the rates of neuritic uptake and axonal transport and the rate of replication in dorsal root ganglion cells and skin fibroblasts, we concluded that the differences observed in neuroinvasiveness probably reflected differences in early virus-nerve cell interactions, i.e. virus attachment and fusion with the neuritic plasma membrane. Introduction The pathogenesis of herpes simplex virus (HSV)-induced acute sporadic encephalitis appears to involve functions and dysfunctions of both viral and host origin. Analyses of gel patterns of structural polypeptides of HSV type 1 strains have failed to identify neurovirulent strains (Pereira et al., 1976) and early reports of restriction enzyme analysis could not disclose any common genetic similarity between strains isolated from cases of encephalitis (Buchman et al., 1980; Hammer et al., 1980; Whitley et al., 1982). However recombination of avirulent strains in mice will generate virulent, lethal recombinants (Javier et al., 1986; Sedarati & Stevens, 1987). Also, intertypic recombinants can yield neurovirulent combinations, although most of such recombinants appear to be attenuated independently of the degree of virulence demonstrated by parental strains, suggesting a multigenetic control (Halliburton et al., 1987). In nonvirulent intertypic recombinants with neurovirulence restored by random recombination (Thompson & Stevens, 1983), and in type 1 laboratory strains, gene functions associated with neurovirulence have been localized to different map units of the HSV type 1 genome (Thompson et al, 1983, 1986; Oakes et al., 1986; Ben-Hur et al., 1987; Javier et al., 1988; Day et al., 1988). Consequently, studies on deletion mutants (Meignier et al., 1988) have not suggested any single factor which may be responsible for HSV-induced death of mice. Characterization of HSV-induced neurovirulence is usually based on studies in mice, evaluating the outcome of infection after peripheral and intracerebral inoculation of the virus. A strain-dependent variation has been reported and virus isolated from cases of encephalitis were found to be no more neurovirulent than isolates from non-neural tissues (Dix et al., 1983). In agreement we have noted that strains ofencephalitic cases (the same strains as used in the present study) did not differ from strains isolated from cutaneous lesions with respect to neurovirulence when studied by the intracerebral route of infection. However, the strains of encephalitic origin were significantly more neurovirulent when tested by a peripheral, snout inoculation (T. Bergstr6m et al., unpublished results). Since these differences might well be explained by host-induced interfering reactions unrelated to the nervous system but modifying the infection, we have introduced techniques by which some different facets of neurovirulence can be observed separately and essentially not influenced by host-induced reactions. In the present report, neuroinvasiveness (uptake, transport and replication of HSV in neuronal cells), likely to be an important contribution to the expression of neurovirulence, is studied by an in vitro technique. Methods HSV strains. Ten HSV type 1 isolates from biopsies of enceptialitis [eight reported previously by Sk61denberg et al. (1984) and two cases SGM

2 406 T. Bergstrfm and E. Lycke additional ones originating from autopsies of fatal cases of encephalitis] were compared with 10 HSV strains from patients with cutaneous lesions and lacking neurological symptoms. Fourteen HSV type 2 strains for study were isolated from the cerebrospinal fluid from clinical cases of meningitis (T. Bergstr6m et al., unpublished results) and were compared with 10 HSV type 2 isolates from patients with genital infections and without symptoms of meningitis. For reference purposes two laboratory strains, type 1 Mclntyre (A. Nahmias) and type 2 B4327UR (S. Jeansson), were used. Virus stocks were prepared from the second or third passage of the isolates in green monkey kidney cells (GMK AH-1). The stocks were stored in 2 ml vials at -70 C. The concentration of p.f.u, was assc:yed by conventional techniques using GMK AH-1 cells grown in 5 cm plastic dishes with 1% methyl cellulose in Eagle's MEM as the overlaying medium. For the type 1 strains the particle/p.f.u, ratio was 1.1 +_ 0-1 for strains derived from cases with encephalitis and 1,2 _+ 0.3 for strains of cutaneous origin. The corresponding ratios for the type 2 strains were 27-4 _+ 4.4 and 8-2 _+ 3-0 for the strains from patients with meningitis and for strains isolated from genital infections, respectively. A comparison of p.f.u./lds0 ratios yielded for the type 1 strains 6.0 _+ 1.4 and 9.3 _+ 6.4, and for the type 2 strains 0.35 _ and 2.41 _ The LDs0 values were assayed by the intracerebral route of infection in Swiss albino mice. Nerve cell cultures. The application of the two-chamber technique according to Campenot (1977) for studies of virus-nerve cell interactions has been described previously (Ziegler & Herman, 1980; Lycke et al., I984; Lycke & Tsiang, 1987). Briefly, dissociated embryonic dorsal root ganglion cells of Sprague-Dawley rats were cultured iz~ Eagle's MEM supplemented with 10% foetal calf serum and antibiotics on a collagen coat inside a glass cylinder placed in the middle of a plastic dish. The use of mitosis inhibitors prevented outgrowth of fibrobla~ts and other dividing cells and the addition of nerve growth factor stimulated the neuronal cells to differentiate. Neuritic extensions penetrated the methylcellulose and silicone grease, which? ~pt the glass cylinder in place and acted as a diffusion-tight barrier ~:aling off the inner culture chamber (the cylinder) from the outer c~':amber (the dish). The two-chamber system allows the infection of neurites in the outer chamber without simultaneous exposure to HSV of neuronal cell soma in the inner culture compartment. Skinfibroblast cultures. Cells of trypsinized rat embyronic skin were cultured in 5 cm plastic dishes using Eagle's MEM with 2% bovine serum and antibiotics. Experimentalprocedure. Cultures of rat dorsal root ganglion cells well endowed with neurites extending into the outer culture compartment were inoculated in the outer chamber. Each type 1 strain was tested on six cultures, infecting the neurites with 5 x 106 p.f.u, per ml, whereas five cultures were used for each type 2 strain, applying a concentration of 1 x 106 p.f.u, per strain. Virus particles in inoculated cultures were allowed to adsorb for l h at 37 C, when the fluid of the outer chamber was replaced four times with fresh medium. The cultures were then incubated for 24 h, and the fluid and cells of the inner compartment were collected and stored at - 70 C until the yields of infectious virus were assayed by plaque titration. Experiments with HSV type 1 included strains of encephalitic cases as well as strains derived from patients with cutaneous lesions. Similarly, both meningitis-derived and genital strains were tested simultaneously in experiments with HSV type 2. In each experiment one of the two reference strains was included, and in order to compare results of different experiments, the data were adjusted according to the mean values for the reference strains. Replication of virus in neuronal cells was estimated by infecting the nerve cell cultures into the inner culture compartment with p.f.u./ml in 100 ~tl, and assaying the amounts of virus produced inside the glass cylinder after an incubation at 37 C for 24 h. Otherwise the experimental procedure was identical to the one described above. Finally, the production of infectious virus in primary cell cultures of rat skin fibroblasts was determined. Cultures in 5 cm plastic dishes in triplicate were infected with 0.1 p.f.u, per cell in a predetermined volume. After adsorption of the virus to the cells at 37 C for 1 h the culture medium was replaced four times, and the virus and cells were harvested and stored at - 70 C until the number of p.f.u, was assayed. Immunofluorescence. The rate of neuritic transport of six of the encephalitis strains was compared with the rate at which six of the strains from patients with cutaneous lesions were transported. The neurites of five cultures per HSV strain were infected by the addition of p.f.u, to the outer culture compartment. After 5, 30, 60 and 120 min of incubation at 37 C for the adsorption of virus, the neuritic extensions were destroyed by addition of concentrated H2SO 4. By repeated washings with phosphate-buffered saline (PBS) the ph was reset to neutrality and fresh culture medium added. The cultures were incubated at 37 C for a total period of 24 h. The fluid and the glass cylinder of the culture were removed and the cells were left for drying overnight at room temperature. Staining of the cells was performed with a fluorescein isothiocyanate-conjugated HSV type 1 antibody (Dako) in a 30-fold dilution for 45 min at 37 C. The cells were washed twice in distilled water, dried and examined in a Zeiss RS microscope. Fluorescent cells were recorded and the number of foci of fluorescence was scored as 1, for one to three foci, 2, for four to six foci and 3, for seven or more foci. Statistics. Wilcoxon's rank sum test and Student's t-test were used. Results Neuroinvasiveness of HS V The amount of infectious virus produced inside the glass cylinders of the cultures after a peripheral, neuritic route of infection was used as a parameter for evaluation of the neuroinvasiveness of an individual strain. The rationale was as follows. The system permits the initiation of infection inside the cylinder only via the neurites extending through the diffusion-tight barrier of the culture, and spread of infection inside the cylinder is mediated by the neuronal cell network. Rats of the same breed and age were used as the source of dorsal root ganglion cells and the cultures were reasonably well standardized with regard to total cell number and number of nerve cells with neurites reaching the outer culture chamber. The 24 h yield, which is the result of one cycle of infection in the culture, would therefore be proportional to the efficiency of uptake and transmission of the virus by neuritic extensions and the number and permissiveness of infected nerve cells. These parameters are listed in Table 1. It is seen that the HSV type 1 strains were considerably more infectious than the type 2 strains, that the mean values for the two groups of type 1 strains differed statistically (P < 0.001); strains associated with clinical encephalitis were signifi- cantly more invasive and/or productive than strains

3 Neuroinvasion by HSV 407 Table 1. HSV produced in nerve cell cultures after the peripheral neuritic route of infection Type Origin Yield* Type E4 E O Type VF1181 VH1728 VH2017 VH3352 VH5962 VH4899 VH774 VG Encephalitis Skin lesion Meningitis Genital * Yield after 24 h, in loglo (p.f.u./ml). t Mean + standard deviation (S.D.) " " " t " "33 _ " " t _ 1.0It isolated from cutaneous lesions. There was no statistical support for a significant difference between the two clusters of type 2 strains derived from cases of meningitis and genital infection. Only four of the 34 strains which were not associated with encephalitis yielded values above 105 p.f.u./ml, compared to eight of the 10 encephalitis strains. Ranking of the type 1 strains indicated that of the 10 encephalitis strains eight were more neuroinvasive than any of the type 1 strains of mucocutaneous origin. Replication of HSV Direct infection of cultures in the inner culture chamber would lead to infection of nerve cells by exposure of the nerve cell soma without requiring neuritic transport of the infection. Results of inner chamber infection, shown in Table 2, indicated that type 1 strains replicated more productively in dorsal root ganglia cells than did type 2 strains (P<0-001). Apparently this difference was connected with a higher permissiveness of rat cells for type 1 as the yields on embryonic rat skin fibroblasts were also generally higher for type 1 than for type 2. The difference observed between the encephalitis strains and the type 1 strains from mucocutaneous lesions was less statistically significant (P < 0.05), and the type 2 strains associated with clinical meningitis seemed not to be more infectious than the strains from genital infections. The results, therefore, do not suggest that a higher rate of replication of the encephalitis type 1 strains in nerve cells was the single crucial factor, but emphasize the importance of neuritic uptake and transportation of the virus. A similar conclusion might be reached by considering the data of Tables 1 and 2 together. Rate o# neuritic transport The plausible difference in the rate of neuritic transport was supported by estimating the transfer rate of infection through the diffusion barrier of the cultures. Cultures inoculated peripherally with six encephalitisassociated and six skin-derived type 1 strains were observed by immunofluorescence for determination of the number of infected neurons. At times ranging from 5 to 120 min post-infection, the neuritic extensions of the outer culture chamber were destroyed to eliminate any further neuritic transport. The cultures were then incubated for a total of 24 h, after which the numbers of fluorescent foci were recorded and scored for each of the strains studied. The relative number of foci will correspond to the efficacy of neuritic infection and transport of the virus through the diffusion-tight barrier in the culture before destruction of the neurites by addition of the acid. In Fig. 1 the mean scores for the two groups of strains examined are plotted against time of incubation before blocking of neuritic transport. The encephalitis strains exhibited a relatively higher rate of infection and transport compared with the strains of cutaneous lesions.

4 408 T. Bergstrfm and E. Lycke Table 2. HSV produced in nerve cell and fibroblast cultures after directly infecting the nerve cell soma Type 1 Nerve cells Fibroblasts * 6.45* E E _ 0.46] ]" ]" ]" Type 2 Nerve cells Fibroblasts * 4.75* VF VH V H VH VH VH VH VG t 5.11 _+ 0'22t ]" ]" * Yield after 24 h, in log10 (p.f.u./ml). I Mean + standard deviation (S.D.). Secondly, the pattern of immunofluorescence in the cultures suggested that the encephalitis strains spread more rapidly (P<0.01) within the neuritic network formed between neurons of the culture I Time after infection of blocking axonal transport (rain) Fig. 1. Neuritic infection and transfer of HSV type 1 in rat sensory neurons in culture. Strains isolated from cases of encephalitis and of mucocutaneous origin (six strains of each group) were inoculated into the peripheral chamber of a two-chamber culture system. Nerve cells of rat dorsal root ganglia were cultured inside a small glass cylinder and neurites extended through a diffusion-tight barrier into the peripheral culture component. The system permitted the neuritic extensions to be infected without simultaneously exposing the neuronal cell bodies to the virus. At times ranging from 5 to 120 min post-infection the neuritic extensions were destroyed by treatment with sulphuric acid, whereafter the cultures were further incubated for a total of 24 h. The number of foci of infected neurons was determined by applying FITC-labelled anti-hsv antibodies. The results were scored and plotted against the time of uninterrupted axonal transport. Black columns, mean score with encephalitic strains; open columns, mean score with skin-derived strains. See text for interpretation of score numbers, 1, 2 and 3. Discussion HSV neurovirulence, expressed in sporadic encephalitis of humans or fatal disease of experimental animals, is a complex issue and probably influenced by several genes (Halliburton et al., 1987). One of the major problems in interpretation of neurovirulence in animals involves the difficulty in defining the pathogenetic steps responsible. It is generally accepted that neuroinvasiveness is an important property of neurovirulent strains. In comparison with non-neuroinvasive strains, neuroinvasive ones should be internalized more readily by peripheral nerves and transported in a retrograde direction, and should replicate better in neural tissues. Since it is unavoidable in studies on experimental animals that host-induced reactions must interfere to some extent with the expression of virulence, we have exploited an in vitro technique, originally presented by Campenot (1977), to study neuroinvasion with reduced host-induced influences. We have reported that by our modification of this technique, strains isolated from clinical cases of encephalitis can be proven as more neuroinvasive than strains of other non-encephalitic origins. In mice, it has been reported that non-neuroinvasive strains can spread to the sensory ganglia but do not replicate well in neuronal tissues (Oakes et al., 1986;

5 Neuroinvasion by HSV 409 Thompson et al., 1986). The importance of neural cell permissiveness is not well understood though, and Dix et al. (1983) have claimed that the amount of infectious virus in the infected brain does not reflect the progress of disease directly. Type 2 strains in mice are in general more neurovirulent than type 1 after peripheral inoculation (Schneweis et al., 1984; Halliburton et al., 1987). The difference does not necessarily have to reflect the outcome of neuronal infection but might result as well from different capacities of the virus types to replicate in, e.g. brain glial cells (Vahlne et al., 1980). Apparently, neuroinvasive type 1 strains can replicate well in mouse trigeminal ganglia (Oakes et al., 1986) and mouse neuroblastoma cells (Day et al., 1988). With our rat cell cultures we were able to demonstrate type-related differences of replication rates in neural as well as nonneural rat cells but the neuroinvasive property of the encephalitic type 1 strains seemed not to be a function of the rate of replication. Attachment, fusion and axonal transport of HSV by neuritic extensions of cultured rat dorsal root ganglion cells have been reported previously (Lycke et al., 1984, 1988). It was found that the neuritic capacity for transporting the viral nucleocapsids was dependent upon microtubular functions (Kristensson et al., 1986) and that the rate of axonal transport was of almost the same order of magnitude as the fast axonal transport of proteins (Lycke et al., 1984). The latter observation suggests that the observed difference in neuroinvasiveness between encephalitic and non-encephalitic type 1 strains might be related to qualitative strainrelated intr, atypic differences which become expressed during the earliest phase of the virus-nerve cell interaction, i.e. attachment and fusion. The cellular receptor(s) of HSV is considered to be heparan sulphate in nature (WuDunn & Spear, 1989) and presumably more than one of the six viral envelope glycoproteins are involved in virus attachment. Rat nerve cells bind HSV type 1 qualitatively and/or quantitatively better than type 2 (Vahlne et al., 1980) but the type-specificity of the cell receptors has not been identified. HSV glycoproteins B, D and H are all supposed to be engaged in virion-cell fusion (Sarmiento et al., 1979; Little et al., 1981 ; Manservigi et al., 1977; Fuller & Spear, 1987) and fusion appears to be the most plausible way by which HSV is internalized into cells, independently of whether the cell is of neural (Lycke et al., 1988) or non-neural origin (Morgan et al., 1968; Fuller & Spear, 1987). We believe that functional alterations of envelope glycoproteins involved in attachment and fusion might account for possible differences in strain neuroinvasiveness. This study was supported by a grant from The Swedish Medical Research Council, grant References BEN-HUR, T., ASHER, Y., TABOR, E., DARAI, G. & BECKER, Y. (1987). HSV-1 virulence for mice by the intracerebral route is encoded by the BamHI-L DNA fragment containing the cell fusion gene. Archives of Virology 96, BUCHMAN, T. G., SIMPSON, T., NOSAL, C., ROIZMAN, B. & NAHMIAS, A. (1980). The structure of herpes simplex virus DNA and its application to molecular epidemiology. Annals of the New York Academy of Sciences 354, CAMPENOT, R. B. (1977). Local control of neurite development by nerve growth factor. Proceedings of the National Academy of Sciences, U.S.A. 74, DAY, S. P., LAUSCH, R. N. & OAKES, J. E. (1988). Evidence that the gene for herpes simplex virus type 1 DNA polymerase accounts for the capacity of an intertypic recombinant to spread from the eye to the central nervous system. Virology 163, DIX, R. D., MCKENDALL, R. R. & BARINGER, J. R. (1983). Comparative neurovirulence of herpes simplex virus type 1 strains after peripheral or intracerebral inoculation of Balb/c mice. Infection and Immunity 40, FULLER, O. & SPEAR, P. G. (1987). Anti-glycoprotein D antibodies that permit adsorption but block infection by herpes simplex virus 1 prevent virion-cell fusion at the surface. Proceedings of the National Academy of Sciences, U.S.A. 84, HALLIBURTON, I. W., HONESS, R. W. & KILLINGTON, R. A. (1987). Virulence is not conserved in recombinants between herpes simplex virus types 1 and 2. Journal of General Virology 68, HAMMER, S. i., BUCHMAN, T. G., D'ANGELO, L. J., KARCHMER, A. W., ROIZMAN, B. & HIRSCH, H. S. (1980). Temporal cluster of herpes simplex encephalitis: investigation by restriction endonuclease cleavage of viral DNA. Journal oflnfectious Diseases 141, JAVIER, R. T., SEDARATI, F. & STEVENS, J. G. (1986). Two avirulent herpes simplex viruses generate lethal recombinants in vivo. Science 234, JAVIER, R. T., IZUM1, K. M. & STEVENS, J. G. (1988). Localization of a herpes simplex virus neurovirulence gene dissociated from high-titre virus replication in the brain. Journal of Virology 62, KRISTENSSON, K., LYCKE, E., ROY'ErA, M., SVENNERHOLM, B. & VAHLNE, A. (1986). Neuritic transport of herpes simplex virus in rat sensory neurons in vitro. Effects of substances interacting with microtubular function and axonal flow [nocodazole, taxol and erythro-9-3(2-hydroxynonyl)adenine]. Journal of General Virology 67, LITrLE, S. P., JOFIE, J. T., COURTNEY, R. J. & SCHAEFER, P. A. (1981). A virion-associated glycoprotein essential for infectivity of herpes simplex type 1. Virology 115, LYCKE, E. & TSlANG, H. (1987). Rabies virus infection of cultured rat sensory neurons. Journal of Virology 61, LYCKE, E., KRISTENSSON, K., SVENNERHOLM, B., VAHLNE, A. & ZIEGLER, R. (1984). Uptake and transport of herpes simplex virus in neurites of rat dorsal root ganglia cells in culture. Journal of General Virology 65, LYCKE, E., HAMARK, B., JOHANSSON, i., KROTOCHVIL, A., LYCKE, J. & SVENNERHOLM, B. (1988). Herpes simplex virus infection of the human sensory neuron. An electron microscopy study. Archives of Virology 101, MANSERVIGI, R., SPEAR, P. G. & BUCHMAN, A. (1977). Cell fusion induced by herpes simplex virus is promoted and suppressed by different viral glycoproteins. Proceedings of the National Academy of Sciences, U.S.A. 74, MEIGNIER, B., LONGNECKER, R., MAVROMARA-NAZOS, P., SEARS, A. E. & ROIZMAN, B. (1988). Virulence of and establishment of latency by genetically engineered deletion mutants of herpes simplex virus 1. Virology 162, MORGAN, C., ROSE, H. i. & MEDNIS, B. (1968). Electron microscopy of herpes simplex virus. I. Entry. Journal of Virology 2, OAKES, J. E., GRAY, W. L. & LAUSCH, R. N. (1986). Herpes simplex virus type 1 DNA sequences which direct spread of virus from cornea to central nervous system. Virology 150,

6 410 T. Bergstrfm and E. Lycke PEREIRA, L., CASSAI, E., HONESS, R. W., TERM, M. & NM-IMIaS, A. (1976). Variability in the structural polypeptides of herpes simplex virus 1 strains: potential application in molecular epidemiology. Infection and Immunity 13, SARMIENTO, M., HAFFEY, M. & SPEAR, P. G. (1979). Membrane proteins specified by herpes simplex viruses. III. Role of glycoprorein VP 7 (B2) in virion infectivity. Journal of Virology 29, SCHNEWEIS, K. E., FORSTBAUER, H., OLBRICH, M. & TAG, M. (1984). Pathogenesis of genital herpes simplex virus infection in mice. III. Comparison of the virulence of wild and mutant strains. Medical Microbiology and Immunology 173, SEDARATI, F. & STEVENS, J. G. (1987). Biological basis for virulence of three strains of herpes simplex virus type 1. Journal of General Virology 68, SK6LDENBERG, B., FORSGREN, M., ALESTIG, K., BERGSTROM, T., BURMAN, L., DAHLQUIST, E., FORKMAN, A., FRYDEN, A., LOVGREN, K., NORLIN, K., NORRBY, R., OLDING-STENQVIST, E., STIERNSTEDT, G., UI-INOO, I. & DE VM-IL, K. (1984). Acyclovir versus Vidarabine in herpes simplex encephalitis. Lancet ii, I. THOMPSON, R. L. & STEVENS, J. G. (1983). Biological characterization of a herpes simplex intertypic recombinant which is completely and specifically non-neurovirulent. Virology 131, THOMPSON, R. L., WAGNER, E. K. & STEVENS, J. G. (1983). Physical location of a herpes simplex virus type-1 gene function(s) specifically associated with a 10 million-fold increase in HSV neurovirulence. Virology 131, THOMPSON, R. L., COOK, M. U, DEvI-RAo, G. B., WAGNER, E. K. & STEVENS, J. G. (1986). Functional and molecular analyses of the avirulent wild-type herpes simplex virus type 1 strain KOS. Journal of Virology 58, VAHLNE, m., SVENNERHOLM, B., SANDBERG, M., HAMBERGER, A. & LYCKE, E. (1980). Difference in attachment between herpes simplex virus type l and 2 viruses to neurons and glial cells. Infection and Immunity 28, WrUTLEY, R., LAKEMAN, A. D., NMtrala.S, A. & ROlZI, t~n, B. (1982). DNA restriction-enzyme analysis of herpes simplex virus isolates obtained from patients with encephalitis. New England Journal of Medicine 3117, WuDuNN, D. & SPEAR, P. G. (1989). Initial interaction of herpes simplex virus with cells is binding to heparan sulfate. Journal of Virology 63, ZIEGLER, R. J. & HERMAN, R. E. (1980). Peripheral infection in culture of rat sensory neurons by herpes simplex virus. Infection and Immunity 28, (Received 21 August 1989; Accepted 5 October 1989)