Recovery of Herpes Simplex Virus From Oculor Tissues of Latently Infected Inbred Mice

Investigative Ophthalmology & Visual Science, Vol. 29, No. 2, February 1988 Copyright Association for Research in Vision and Ophthalmology Recovery of Herpes Simplex Virus From Oculor Tissues of Latently Infected Inbred Mice 5. Z. Abghari and R. Doyle Stulring Evidence for latent infection of ocular tissues following topical corneal inoculation with herpes simplex virus type 1 (HSV) was sought in three strains of inbred mice that differ in susceptibility to HSV stromal keratitis. Corneas of BALB/c, C57BL/6, and mice were inoculated topically with HSV. At 6-8 weeks after inoculation, when no active ocular infection was present, minced whole eyes and trigeminal ganglia were assayed for latent virus. Virus was recovered by explantation from minced eyes of all three strains ( = 2%; BALB/c = %; C57BL/6 = 7%). In order to determine which ocular structures harbored virus, corneas, retinas and choroid-sclera were cultivated separately. Virus was activated from corneas of and BALB/c mice, but not from corneas of C57BL/6 mice. These findings suggest that HSV is capable of establishing latent infection in ocular tissue of inbred mice and that the rate of establishment of latency is under host genetic control. Since neural cell bodies are not present in the cornea, the data suggest that latency is established in cells other than neurons. Invest Ophthalmol Vis Sci 29:239-243,1988 Following primary infection at a peripheral site, herpes simplex virus type 1 (HSV) travels intra-axonally from the site of inoculation to neurons in associated ganglia where, after an initial period of viral replication, it persists in a latent state. 1 " 5 For the eye and adnexa, these are the trigeminal and superior cervical ganglia. 2 ' 3 ' 6 " 9 Although the physical state of viral DNA during latency is not known, a functional definition of latency has evolved: tissue is said to harbor latent virus if virus antigens and particles are not detectable in tissue homogenates, but can be demonstrated by explantation or cocultivation. 2 ' 6-7 ' 9 This functional definition of latency will be used throughout this report. It has been hypothesized that recrudescent disease occurs when, in response to some stimulus, active viral replication occurs in ganglia previously harboring latent infections. Infectious particles are then transmitted along the axons to peripheral sites, where they are released to produce recrudescent disease by replicating in epithelial cells. 3 This hypothesis is known as the "ganglion trigger" theory. 3 " Stimula- From the Department of Ophthalmology, Emory University, Atlanta, Georgia. Supported by NIH Grant EY-597; NIH Departmental Training Grant EY-792 (SZA); a Research Manpower Grant from Research to Prevent Blindness, Inc. (RDS); and a Departmental Grant from Research to Prevent Blindness, Inc. Submitted for publication: March 24, 1987; accepted July, 1987. Reprint requests: R. Doyle Stulting, MD, PhD, Department of Ophthalmology, Emory University, 1327 Clifton Road NE, Atlanta, GA 3322. tion of the trigeminal ganglion has been demonstrated to produce recrudescent ocular disease in animal models. 1 An alternative hypothesis, the "skin trigger' theory, postulates that the virus is released intermittently from the ganglia via peripheral nerves, creating subclinical microfoci of infection at peripheral sites. Various stimuli at the peripheral site might then allow a significant increase in viral replication that is manifested as recrudescent disease.'' A third possibility is that the virus can exist in a latent state in peripheral tissues. In fact, there is experimental evidence to support this hypothesis. 12 " 14 Recent reports have suggested that HSV can establish latent infection in the retina 15 and choroid/sclera 16 of the mouse eye. Following a protocol similar to that of Shimeld et al, 16 in which whole eyes from mice harboring latent HSV infection were cultivated in vitro, we were unable to detect latent virus in ocular tissues. In order to investigate this apparent discrepancy, as well as the discrepancies in identification of ocular structures harboring latent virus reported by the previous investigators, 1516 we conducted further experiments, which are the subject of this report. Virus Materials and Methods HSV type 1, isolate 4 (STU-4), a strain obtained from a patient with herpetic ocular disease, has been previously described. Virus stocks were prepared in 239 Downloaded From: https://iovs.arvojournals.org/ on 11/13/218

24 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / Februory 1988 Vol. 29 Table 1. Recovery of HSV from eyes and trigeminal ganglia of latently infected inbred mice Mouse Number of Number of ganglia Number of eyes strain animals* with latent virus with latent virus BALB/c C57BL/6 1 12 28 1(1%)f 12(1%) 27 (96%) 2 (2%)t 2(%) 2 (7%) * Six weeks after topical inoculation of the left eye with HSV, left trigeminal ganglia and left eyes were removed from animals that initially developed a typical herpetic disease pattern after inoculation, but showed no signs of active infection at the time of this experiment. f Percent of ganglia with latent virus. t Percent of eyes with latent virus. human epidermal carcinoma (HEp-2) cells, titered on vero cells, and stored at -7 C until used. Mice and Inoculations Six to 8 week old BALB/c,, and C57BL/6 mice (Jackson Laboratories, Bar Harbor, ME) were allowed to acclimatize to their new environment for at least 1 week. Before inoculation, animals were anesthetized with methoxyflurane and examined with a slit lamp to exclude those animals with ocular abnormalities. The corneal epithelium of the left eye (5. X 1 5 PFU/ml) or both eyes (1. X 1 5 PFU/ml) of anesthetized animals was inoculated with virus as previously described. Eyes were examined using a slit lamp three times weekly for 1 month. Only animals with evidence of keratitis (dendrites and/or stromal disease) were used in subsequent experiments. Animals were maintained in accordance with the ARVO Resolution on the Use of Animals in Research. Cell Cultures HEp-2 and vero cells were grown in Eagle's minimum essential medium (MEM) and maintained as previously described. Activation of Latent Virus Six to 8 weeks after inoculation, when slit lamp examination showed no active ocular infection, mice were killed by anesthetic overdose. Trigeminal ganglia (TG) and eyes were removed. The eyes were minced into approximately 1 mm pieces in MEM. Whole TG and the minced eyes were then incubated (37 C) in 1. ml growth medium (MEM containing 1% fetal calf serum and antibiotics) to activate latent virus. At 3 day intervals, medium was removed and assayed for virus. 18 Fresh medium (1. ml) was added to the tissue and the incubation was continued. At the end of the incubation period (14 days for TG and 35 days for eyes), tissues that failed to yield virus were homogenized and assayed for virus as previously described. In some experiments, conjunctiva, cornea, retina and sclera/choroid were carefully separated under a dissecting microscope (lens, iris, and ciliary body were discarded). Each tissue was then placed on a subconfluent monolayer of vero cells, which was observed for cytopathic effect (CPE). Every 5 days, each tissue was transferred to a fresh subconfluent vero cell monolayer and the incubation was continued. (Initial attempts to isolate latent virus from dissected eyes as described above for minced eyes failed to yield virus. We suspected that the small tissue fragments did not remain viable long enough to support replication of virus to detectable levels and subsequently were successful using cocultivation on vero monolayers.) Isolation of Infectious Virus Six to 8 weeks after inoculation (on the same days when eyes were obtained for cultivation in vitro), eyes and trigeminal ganglia were removed from 2% (24/124) of randomly selected animals. The presence of infectious virus in these tissues was determined as previously described. Identification of Virus Isolates Virus isolates were serotyped by indirect immunofluorescence using monoclonal antibodies recognizing HSV type 1 (HC-1) 19 and HSV type 2 (H966) 2 as described previously. Results Activation of Latent Virus in Explants of Minced Inoculated Eyes and Intact Trigeminal Ganglia Six to 8 weeks after corneal inoculation with HSV, all mice were found by slit lamp observation to have no evidence of active ocular infection (blepharitis, dendritic keratitis, stromal infiltrates or uveitis). Minced eyes and whole trigeminal ganglia were examined for the presence of latent HSV by cultivation in vitro. HSV was recovered from trigeminal ganglia of all three strains with high frequency (Table 1). The recovery rate from minced eyes was lower ( = 2/1 (2%); BALB/c = 2/12 (%); C57BL/6 = 2/28 (7%)) (Table 1). Kinetics of HSV Activation in the Eye The process of activation of latent virus from the eye was investigated in more detail. Virus release began at 1-24 days of incubation in vitro and persisted for 14 to 25 days (Fig. 1). Peak virus production was greater in DBA mice than in C57BL/6 mice, with Downloaded From: https://iovs.arvojournals.org/ on 11/13/218

No. 2 HSV IN OCULAR TISSUES OF LATENTLY INFECTED INBRED MICE / Abghori ond Srulring 241 BALB/c mice producing intermediate levels. One eye from a C57BL/6 mouse showed two peaks of virus production on days and 31, suggesting two activation events from the same eye. Recovery of HSV From Eyes of Latently Infected Mice Isolation of Latent Virus From Individual Ocular Tissues Table 2 shows the frequency of virus recovery from trigeminal ganglia and separated ocular tissues of latently infected mice. In this experiment, virus was isolated from trigeminal ganglia from 88-94% of mice of all three strains. Virus was recovered from corneas of 3/ (18%) of DBA mice and 2/18 (12%) of BALB/c mice, but was not isolated from ocular tissues of C57BL/6 mice. In no case was virus isolated from conjunctiva, retina or sclera/choroid. 3 U. BALB/C Identification of Viral Isolates and Search For Active Viral Replication All viral isolates from minced eyes and corneas, as well as 1% of virus isolates from trigeminal ganglia, were serotyped and found to be HSV type 1. To be certain that no infectious virus was present in the animals used for these experiments, 24 trigeminal ganglia, 12 minced eyes, and 12 dissected ocular tissues from inoculated diseased animals selected at random were homogenized and tested for the presence of infectious virus. In no case was infectious virus isolated from these animals. Discussion In humans and other animals, several investigators have successfully reactivated latent HSV from neuronal, but not from non-neuronal, tissues. 1 " 4 ' 6 " 8 ' 21 This has led to the hypothesis that latent HSV infections are restricted to tissues of the nervous system. More recently, however, it has been shown that HSV can be recovered from the site of inoculation in the footpad of guinea pigs and mice during latent infections. 13 ' 14 HSV has also been isolated from the ear skin of latently infected mice in the absence of observable lesions. 12 Incubation Time (days) Fig. 1. Recovery of HSV from eyes of latently infected inbred mice. The quantity of virus recovered from eyes harboring latent HSV (same animals as referred to in Table 1) is shown for varying times of incubation. Eyes were removed, minced and incubated in 1 ml growth media at 37 C for 35 days. Media were changed and assayed for virus two times per week. Three laboratories have recovered HSV from ocular tissues of latently infected mice and humans. 1516 ' 22 In contrast to previous investigators, Easty et al succeeded in recovering HSV from ten of 34 human corneas obtained from patients with quiescent herpetic stromal keratitis undergoing penetrating keratoplasty. 22 Electron microscopy showed viral Table 2. Recovery of HSV from trigeminal ganglia and ocular tissues of latently infected inbred mice Recovery of HSV from Mouse strain Number of specimens* Trigeminal ganglion* Cornea Conjunctiva Retina Sclera and choroid BALB/c C57BL/6 16 16f(94%)t 15 (94%) 15 (88%) 3(18%) 2(12%) * Both trigeminal ganglia and eyes were removed from animals 6 to 8 weeks after topical inoculation of both eyes with HSV. Virus was recovered from trigeminal ganglia by explantation and from sectioned eyes by cocultivation. t Number of specimens with latent virus. % Percent of specimens with latent virus. Downloaded From: https://iovs.arvojournals.org/ on 11/13/218

242 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Februory 1988 Vol. 29 particles in keratocytes. Openshaw, using cocultivation techniques, recovered HSV from 5 of 19 BALB/c mouse eyes 5-7 months after inoculation when disease was quiescent. 15 He found HSV antigens in retinal cells and concluded that these cells must be the site of latent herpetic infection. Shimeld et al recovered virus from cultured whole eyes of 4/16 mice 36 days after anterior chamber inoculation with HSV. 16 They found viral particles in scleral fibroblasts, but were unable to identify living retinal cells at the time of microscopic analysis. Easty et al found HSV in 8/2 anterior segments of animals 123 days after snout infection with zosteriform spread. 22 HSV antigen was found in anterior uveal cells of these animals. Having failed to recover HSV from 14 whole eyes of mice harboring latent infections, we wondered whether small numbers of infectious particles reactivated within the eye might quickly deplete the population of cells capable of supporting their replication and thus not reach the bathing medium outside of the eye, where they would be detected. In the present series of experiments, the protocol was modified, and virus was detected from minced eyes of all three mouse strains investigated. When ocular tissues were carefully dissected and cultured individually, virus was recovered from cornea, but not conjunctiva, retina or choroid/sclera. Thus, we confirm previous reports indicating that the eye can harbor latent virus. As has previously been found, the time of incubation in vitro required to demonstrate latent virus in ocular structures is considerably longer than that required to demonstrate virus in trigeminal ganglia. Failure to recognize the necessity for prolonged incubation in vitro is perhaps one explanation for the failure of other investigators to demonstrate latent virus in ocular structures. The data suggest that the recovery of HSV from the eyes of latently infected mice is host strain-dependent. However, the differences observed did not reach statistical significance (X 2 ; P =.14), and larger numbers of animals must be examined to confirm the significance of strain-related differences. There are several possible explanations for our localization of latent virus to the cornea in apparent contrast to the data of Openshaw, 15 Shimeld et al, 16 and Easty et al, 22 who found evidence of latent virus in retina, choroid/sclera, and anterior uvea, respectively. It is possible that there is a virus strain-specific predilection to establish latency in host tissues, and the virus strain used in the present experiments is simply more likely to be found in the cornea than in other ocular structures. Secondly, in the case of Shimeld et al and Easty et al, inoculation was by a different route the anterior chamber or snout and other unrecognized differences in the technique of inoculation may also exist. Differences in host mouse strain susceptibility might possibly explain differences between the results of Shimeld et al, Easty et al and those of our laboratory, but Openshaw used BALB/c mice as we did. Finally, one must consider the possibility that latent virus is reactivated from one site within the eye, but replicates at another site during incubation in vitro. For example, when whole eyes or minced globes are incubated in vitro, virus that is released from the cornea might secondarily infect retina, uvea, sclera and other structures. The experimental protocols of Openshaw, Schimeld et al and Easty et al do not exclude the possibility that infectious particles were initially released from a structure different from the one in which virus was identified. In some of the experiments reported herein, ocular tissues were separated before incubation in vitro, so spread of infection from another structure was not a possibility. On the other hand, it is conceivable that we failed to find virus in retina or choroid/sclera because culture conditions did not favor reactivation from these tissues once they were separated from neighboring structures. Is a latent infection truly established in the corneas of experimental mice? Based on the classical, operational definition of latency set forth in the beginning of this report, the answer to this question is yes. The data clearly show that virus cannot be recovered from cell-free homogenates, but can be recovered from isolated corneas and minced eyes cultivated in vitro. The physical state of the viral genome and the mechanism by which the virus remains in (or gains access to) the cornea during latency, however, is not established by these experiments. Possible mechanisms include the following. First, the viral genome may persist within the soma of neuronal cells in the trigeminal ganglia, with intermittent low level release of infectious particles (transported intra-axonally) into the cornea itself. The number of particles may be very small, so that cultivation of the tissue in vitro is required for them to replicate to detectable levels or be released from the tissue matrix. Second, it is possible that the viral genome is present in corneal cells (epithelial cells, endothelial cells, keratocytes, or wandering cells). Either the intermittent low-level release of infectious particles or the transition from a latent to productive infection might explain the prolonged incubation in vitro required for the demonstration of infectious virus. Third, it is possible that handling of the animals prior to administration of anesthetic overdose or the anesthetic overdose itself might stimulate transition from a latent to productive infection in the neuronal cells themselves with release of small numbers of infectious particles into the cornea. Downloaded From: https://iovs.arvojournals.org/ on 11/13/218

No. 2 HSV IN OCULAR TISSUES OF LATENTLY INFECTED INBRED MICE / Abghori ond Srulring 243 Recent experiments using in situ hybridization techniques are beginning to shed light on the molecular mechanism(s) of latency. The data indicate that latently infected sensory neurons contain viral DNA, but differ from productively infected cells by the lack of viral protein synthesis, limited transcription, the presence of viral RNA in the nucleus but not cytoplasm, and the presence of antisense ICP-O transcripts. 23 Further experimentation will be required to determine whether similar molecular events occur in the corneas of latently infected animals and to identify cells harboring the HSV genome during latency. Key words: cornea, herpes simplex virus, inbred mice, keratitis, latency, mice, reactivation, virus Acknowledgments We are grateful to Ms. Janice C. Kindle for her expert technical assistance and Dr. L. Pereira, State of California, Department of Health Services, Berkeley, California for supplying the monoclonal antibodies. References 1. Cook ML and Stevens JG: Pathogenesis of herpetic neuritis and ganglionitis in mice: Evidence for intra-axonal transport of infection. Infect Immun 7:272, 1973. 2. Baringer JR and Swoveland P: Persistent herpes simplex virus infection in rabbit trigeminal ganglia. Lab Invest 3:23, 1974. 3. Stevens JG: Latent herpes simplex virus and the nervous system. Curr Top Microbiol Immunol 7:31, 1975. 4. Tullo AB, Easty DL, Hill TJ, and Blyth WA: Ocular herpes simplex and the establishment of latent infection. Trans OphthalmolSocUK 12:15, 1982. 5. Klein RJ: Initiation and maintenance of latent herpes simplex virus infections: The paradox of perpetual immobility and continuous movement. Rev Infect Dis 7:21, 1985. 6. Stevens JG, Nesburn AB, and Cook ML: Latent herpes simplex virus from trigeminal ganglia of rabbits with recurrent eye infection. Nature (New Biol) 235:216, 1972. 7. Nesburn AB, Cook ML, and Stevens JG: Latent herpes simplex virus. Arch Ophthalmol 88:412, 1972. 8. Baringer JR and Swoveland P: Recovery of herpes simplex virus from human trigeminal ganglions. N Engl J Med 288:648, 1973. 9. Price RW, Katz BJ, and Notkins AL: Latent infection of the peripheral ANS with herpes simplex virus. Nature 257:686, 1975. 1. Nesburn AB, Dickinson R, and Radnoti M: The effect of trigeminal nerve and ganglion manipulation on recurrence of ocular herpes simplex in rabbits. Invest Ophthalmol 15:726, 1976. 11. Hill TJ and Blyth WA: An alternative theory of herpes simplex recurrence and a possible role for prostaglandins. Lancet 1:397, 1976. 12. Hill TJ, Harbour DA, and Blyth WA: Isolation of herpes simplex virus from the skin of clinically normal mice during latent infection. J Gen Virol 47:25, 198. 13. Scriba M: Persistence of herpes simplex virus (HSV) infection in ganglia and peripheral tissues of the guinea pig. Med Microbiol Immunol 169:91, 1981. 14. Subak-Sharpe JH, Al-Saadi SA, and Clements GB: Herpes simplex virus type 2 establishes latency in the mouse footpad and in the sensory ganglia. J Invest Dermatol 83(Suppl 1):67, 1984. 15. Openshaw H: Latency of herpes simplex virus in ocular tissue of mice. Infect Immun 39:96, 1983. 16. Shimeld C, Tullo AB, Hill TJ, Blyth WA, and Easty DL: Spread of herpes simplex virus and distribution of latent infection after intraocular infection of the mouse. Arch Virol 85:5, 1985.. Abghari SZ, Stulting RD, Nigida SM, Downer DN, and Nahmias AJ: Spread of herpes simplex virus and the establishment of latency after corneal infection in inbred mice. Invest Ophthalmol Vis Sci 27:77, 1986. 18. Schinazi RF, Peters J, Williams CC, Chance D, and Nahmias AJ: Effect of combination of acyclovir with vidarabine or its 5' monophosphate on herpes simplex virus in cell culture and in mice. Antimicrob Agents Chemother 22:499, 1982. 19. Pereira L, Dondero DV, Gallo D, Devlin V, and Woodie JD: Serological analysis of herpes simplex types 1 and 2 with monoclonal antibodies. Infect Immun 35:363, 1982. 2. Roizman B, Norrild B, Chan C, and Pereira L: Identification and preliminary mapping with monoclonal antibodies of a herpes simplex virus 2 glycoprotein lacking a known type 1 counterpart. Virology 133:242, 1984. 21. Baringer JR: Herpes simplex virus infection of nervous tissue in animals and man. Prog Med Virol 2:1, 1975. 22. Easty DL, Schimeld C, Claoue CMP, and Menage M: Herpes simplex virus isolation in chronic stromal keratitis: Human and laboratory studies. Curr Eye Res 6:69, 1987. 23. Stevens JG, Wagner EK, Devi-Rao GB, Cook ML, and Feldman LT: RNA complimentary to a herpesvirus a gene mrna is prominent in latently infected neurons. Science 235:156, 1987. Downloaded From: https://iovs.arvojournals.org/ on 11/13/218