Conjunctivitis in Rabbits Caused by Enterovirus Type 70 (EV70)

Transcription

1 Conjunctivitis in Rabbits Caused by Enterovirus Type 70 (EV70) Marlyn P. Longford,* Marguerite Yin-Murphy,f John C. Barber,4: Hillary K. Heard, and G. John Sranron A rabbit enterovirus 70 (EV70) model infection that closely mimics human enteroviral conjunctivitis was developed. Conjunctivitis occurred 24 hr following topical application of EV70. The conjunctivitis was characterized by tearing, redness, swelling of the eye lids, follicles in the superior palpebral conjunctiva, and dilatation of subconjunctival blood vessels. Histologic examination of conjunctival and corneal tissue taken 1 and 2 days after infection revealed numerous punctate areas devoid of squamous epithelium on the upper palpebral conjunctiva. Also, follicles without germinal centers were observed microscopically in the palpebral and tarsal conjunctiva. Fibroblast infiltration characteristic of wound healing and a sparse mononuclear infiltration was noted by the second day. Peak levels of virus [10 3 to plaque forming units (PFU)/mI] were detected 1 to 2 days after infection and declined to undetectable levels after 3 to 5 days. Interestingly, antiserum to parental EV70 was less effective (8-10-fold) in neutralizing EV70 adapted to animal and tissue culture systems. This finding suggests that an antigenic variant of EV70 arose during adaptation. Fibroblast interferon (IFN/3), which is indicative of viral infection, was detected in tears from 6 of 16 rabbits and declined to undetectable levels 3 days after infection. Serum antibody to EV70 was detectable 8 to 10 days after infection. However, the level of serum antibody was highly variable. The results indicate that the clinical disease, virologic and immunologic courses were similar to that of the human infection. Results suggest that this animal model provides a system for studying the natural antigenic variation of EV70, the natural host defenses of the eye, and antiviral treatments against enteroviral conjunctivitis. Invest Ophthalmol Vis Sci 27:915-9, 1986 Acute hemorrhagic conjunctivitis (AHC) is caused by enterovirus 70 (EV70) and coxsackievirus A24 (CA24). 1 " 6 AHC is highly contagious, usually bilateral, and characterized by rapid onset of a foreign body.sensation in the eye hr after infection. The conjunctivitis is characterized by tearing, vascular dilatation, swelling of the eyelids, and often subconjunctival hemorrhage. Punctate corneal lesions, folliculitis, preauricular lymphadenopathy, and central nervous system (CNS) involvement may also be associated with AHC. 2 " 4 ' 6 Clinical disease and high levels of virus, 5 ' 7 early appearing neutralizing activity (ENA), 58 and interferon (IFN) 7 ' 8 are detectable in tears 24 hr after infection. Virus and IFN in tears decline to undetectable levels 4 to 5 days after infection, and clinical disease usually resolves 7 to 10 days post onset of conjunctivitis. From the Department of Ophthalmology,* Louisiana State University Medical Center, Shreveport, Louisiana, Department of Bacteriology,t University of Singapore, Singapore, and the Departments of Ophthalmology^: and Microbiology, University of Texas Medical Branch, Galveston, Texas. Supported by grant EY from the National Eye Institute, NIH, Bethesda, Maryland. Submitted for publication: January 5, Reprint requests: G. John Stanton, PhD, Department of Microbiology, University of Texas Medical Branch, Galveston, TX An animal model system has not been reported to date for studying the ocular pathogenesis of AHC caused by either EV70 or CA24. This paper reports the development of an ocular model of EV70 conjunctivitis in rabbits that closely mimics AHC in man. These studies also provide evidence suggesting that EV70 can undergo rapid antigenic change. Tissue Culture Materials and Methods Primary rabbit kidney (PRK) cell cultures were prepared using standard methods 9 and maintained on minimum essential medium containing Earle's salts (EMEM) supplemented with 5% fetal bovine serum (FBS), penicillin 100 units/ml, streptomycin 100 /xg/ ml, and mycostatin 100 tig/m\. Human amnionic WISH cells (Flow Laboratories; McLean, VA) rabbit kidney cells (RK13), and monkey VERO cells (American Type Culture Collection; Rockville, MD) were maintained on EMEM containing 5% FBS and antibiotics. Viruses EV70 isolate J648/71 was first adapted to grow in rabbit conjunctival-corneal cells (RCC) and then pas- 915

2 916 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1986 Vol. 27 saged in PRK cells, % tissue culture infectious doses (TCID 5 oper/ml) (10). This virus suspension was stored in 1-ml aliquots at 70 C and used as challenge virus during these studies. EV70 isolate SEC 146/71 was passaged in monkey VERO cells and in rabbit skin fibroblast cells. Progeny virus from the second subpassage in each cell type was used in a neutralization assay to test for antigenic change. Virus Quantitation The levels of virus (plaque: forming units (PFU)/ ml) in tear samples were determined on PRK cell cultures prepared in 96-well microtiter plates (Costar; Cambridge, MD). Tear samples were diluted in EMEM supplemented with 2% FBS and antibiotics. One-tenth ml of each dilution was placed on two cultures of PRK cells and incubated for 24 hr at 37 C. Cell cultures were fixed and stained with 30% methanol (MCB Manufacturing Corporation, Inc.; Cincinnati, OH) containing crystal violet (1% w/v) and plaques counted macroscopically. Virus Neutralization Assay Tear samples, serum samples, and antisera were assayed for the presence of neutralizing activity against EV70 in RK13 or WISH cells depending upon the virus used. Serum samples were serially diluted in EMEM supplemented with 2% FBS and antibiotics. Five hundred to 1000 PFU of EV70/ml of medium was mixed with each sample dilution (equal volumes) and incubated for 1 hr at room temperature. One-tenth milliliter of each virus-serum mixture was added to three microtiter plate cultures of WISH or RK13 cells and incubated at 37 C for 5 days. The viral CPE was observed macroscopically after staining with crystal violet, and the titer was calculated." Interferon (IFN) Quantitation Rabbit IFN in samples was detected and quantitated in RK13 cells using a vesicular stomatitis virus microplaque reduction assay. 12 Antiviral activity in rabbit tear fluids was characterized as IFN by its lack of virus specificity, activity on rabbit cells but not human WISH cells, and its acid stability. Rabbit IFN standard was obtained through the Antiviral Substances Program, National Institute of Allergy and Infectious Diseases, NIH, Bethesda, MD. One unit of IFN in this assay was approximately equal to one NIH reference standard unit. Rabbit Model Infection RCC cell adapted EV70 (J648/71) 10 was used to infect the eyes of five Black Velvet (BV) and 21 New Zealand White (NZW) rabbits. The conjunctiva of the upper and lower lids of the left eyes of each rabbit were swabbed (15 to 30 sec) with a cotton applicator saturated with stock virus suspension. Similarly, the contralateral eye (right) of each rabbit was swabbed with culture media. Tear samples ( n\) were collected in glass capillary pipettes from the infected and control eye at the level of the lateral canthus and fornix. 5 Serum samples were collected and slit lamp examinations performed at various times during the course of the infection. Histology of Corneal and Conjunctival Tissue Corneal and attached conjunctival tissue (360, 10 mm rim) was surgically removed from control and infected rabbit eyes 24 hr (three rabbits) and 48 hr (two rabbits) post-application of EV70. All tissues were taken within 30 min of death. Tissues were placed in phosphate buffered saline containing methyl cellulose (1% w/v, ph 7.4), immersed in liquid nitrogen for 30 sec, and stored at -70 C. Four micrometer sections were cut (International Cryostat; Needham Heights, MA) and air-dried on glass slides. The tissue sections were fixed in absolute methanol for 5 min and stained with Giemsa. 9 Animals used in this study were maintained in accordance with the guidelines of the Committee on Animals of the University of Texas Medical Branch, the ARVO Resolution on the Use of Animals in Research, and the guidelines prepared by the Committee on Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council [DHEW (DHHS) publication No. (NIH) 78-23, revised 1978]. Results EV70 Infection of BV and NZW Rabbits Conjunctivitis occurred in BV and NZW rabbits hr after topical application of EV70. Figure 1A shows the control and infected eyes of a BV rabbit 24 hr after application of EV70. The conjunctivitis was characterized by a rapid onset of tearing, redness, and lid edema over a 3-5-hr period. Follicles were noted in the palpebral conjunctiva of the upper lid 10 hr after onset of conjunctivitis but were not detected in the conjunctiva of the lower lid. Also, dilatation of the blood vessels was apparent in the conjunctiva of the upper and lower lids by 10 hr after onset of conjunctivitis. Follicles (without germinal centers) and dilated blood vessels were readily observed by histologic examination (Fig. IB). In addition, histologic examination revealed numerous punctate lesions (Fig. 1C) on

3 No. 6 Fig. 1. A, Control (left) and infected (right) eyes of a BV rabbit 24 hr after infection with EV70. B, Light micrographs of control (left) and infected (right) eyes showing follicles (Fol) and dilated conjunctiva! blood vessels in the palpebral conjunctiva of the upper lid of a BV rabbit 48 hr post infection (V) (Giemsa stain, original magnification X45). C, Light micrographs of normal (left) and infected (right) conjunctival tissue showing a punctate lesion with fibrobtast infiltration just beneath the basement membrane (Giemsa stain, original magnification X430). ENTEROVIRAL CONJUNCTIVITIS IN RADBITS / Longford er ol. B the tarsal and palpebral conjunctiva of all BV and NZW rabbits 24 and 48 hr after onset of conjunctivitis. The lesions were not directly associated with a follicle. Fibroblast infiltration, characteristic of wound healing, was present at the base of the epithelial lesions, and a sparse infiltration of mononuclear leukocytes was evident on the second day after onset of symptoms (Fig. IC). All signs and symptoms of disease were resolved 7-10 days after onset of conjunctivitis. The onset of conjunctivitis was associated with maximal levels of virus in tears (Fig. 2). Virus was detectable in tears hr after infection and increased to maximal levels hr post-infection in BV and NZW rabbits. Virus was detectable in tears for 3 to 5 days. There was no association between the level of virus in tears, length of virus shedding, and level of the serum antibody response. For example, in Figure 2, NZW rabbit A had 1000-fold more detectable EV70 than NZW rabbit B, but the serum antibody response of rabbit A occurred later and to a lower level than rabbit B. There was an association between severity of infection and level of virus in tears (data not shown). The severity of infection and course of virus shedding was similar in BV and NZW rabbits. However, the clinical signs of conjunctivitis were more easily observed in the pigmented BV rabbits. Conjunctival hemorrhages, corneal lesions, and CNS dysfunctions were not observed. Conjunctivitis and virus shedding was not detected in tears collected from control rabbit eyes.

4 918 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1986 Vol O.INCAL SYMPTOMS RABBIT B SERUM ANTBOOY RABBIT A SERUM ANTBOOY po-q b DAYS POST INOCULATION Fig. 2. Clinical course, virologic course and serum antibody response in BV rabbits unilaterally infected with EV70. EV70 isolate J648/71 was adapted to grow in RCC cells 10 and PRK cells. J648/71 was also able to replicate and cause clinical disease in BV and NZW rabbits as described above. Hyperimmune rabbit antisera to parental J648/71 and rabbit adapted J648/71 and monkey antisera to SEC 146/71 were used in neutralization tests against the parental and adapted EV70 isolates (Table 1). Antiserum to the parental J648/71 was about 100- fold less effective in neutralizing J648/71 adapted to grow in RCC/PRK cells and J648/71 isolated from infected rabbits than the parental J648/71. Similarly, antibody to SEC 146/71 was 48-fold less effective in neutralizing J648/71 that had been adapted to RCC/ PRK cells. Also, antisera to J648/71 were less effective (2-16-fold) in neutralizing SEC 146/71 that had been grown in rabbit skin cells and monkey VERO cells. Concomitantly, antiserum from a rabbit infected with J648/71 contained 12-fold less neutralizing antibody to the parental J648/71 than the J648/71 rabbit isolate. The adapted viruses retained their enteroviral physicochemical characteristics, ie, acid stable, ether stable and insensitivity to actinomycin D (data not shown). Discussion Our approach to the development of an animal model for AHC was to systematically adapt EV70 to replicate in primary RCC cell cultures 10 and then to use this virus to reproducibly infect rabbit eyes in vivo. Reproducibility of the Rabbit Model Infection Sixteen NZW rabbits were infected as described above to determine the reproducibility of the virologic course of the EV70 infection. Tear samples were collected daily through 5 days and assayed for virus and IFN. All of the rabbits developed a unilateral conjunctivitis 18 to 24 hr after application of EV70. EV70 was detected in tears 12 hr after infection. Peak levels of virus in tears (mean, PFU/ml; range, to PFU/ml) were detected 24 hr after inoculation (Fig. 3). Virus in tears declined to undetectable levels by day 5. No virus was detected in control (right eye) tear samples. Low levels of IFN (mean, 10' 2 units/ml, range 10 1 to units/ml) were detected in tear samples collected from 6 of 16 rabbits 1 and 2 d after infection (Fig. 3). These results indicate that infection of rabbits with EV70 results in a predictable peak of virus production within 24 hr of application of EV70. Neutralization of EV70 Adapted to Cell Culture and Rabbits ^N LJL Q. CO fjc > 3 o 3 CC LU UJ 2 2 UD DAYS I UJ z o CO b 8 Fig. 3. The levels of EV70 and IFN in tears collected from 16 NZW rabbits (mean ± standard deviation).

5 No. 6 ENTEROVIRAL CONJUNCTIVITIS IN RABBITS / Longford er at. 919 Table 1. Antigenic characteristics of the rabbit adapted EV70 Tiler ofantisera made to EV70 isolate Passage history WISH adapted J648/71* Rabbit adapted J648/71\ SEC 146/7\% J648/71 J648/71 J648/71 SEC 146/71 SEC 146/71 SEC 146/71 Hela 2 WISH 2 RCC, PRK 2 Rabbit, PRK 2 Hela 2, WISH 2 Rabbit skin cells 2 Monkey VERO ND * Serum collected 6-8 wk after rabbits were immunized with WISH cell adapted J648/71 (EV70). t Serum collected from rabbits infected topically with rabbit adapted EV70. % Serum collected from monkey inoculated with Hela cell adapted SEC 146/ 71 (EV70) (prepared by Dr. M. Yin-Murphy). J648/71 was isolated from tears collected 48 h after infection and passaged in PRK cells. We present data indicating that, like humans, 57 ' 8 rabbits infected with EV70 have clinical disease, detectable virus in tears, and IFN in tears within hr after infection. The conjunctivitis in BV and NZW rabbits was similar to AHC in man in that it was characterized by a rapid onset of conjunctivitis over a 3 to 5 hr period. The signs and symptoms of conjunctivitis were redness, tearing, swelling of the eyelids, folliculitis, and dilatation of the subconjunctival blood vessels. Histologic examination of rabbit eye tissue taken 1 and 2 days after infection revealed numerous punctate lesions localized to the upper tarsal and palpebral conjunctiva. In man, punctate lesions have been observed in the cornea only. 3 The conjunctivitis in rabbits resolved completely within 5-7 days after infection. Bilateral infections, subconjunctival hemorrhage, corneal lesions and CNS disease, that may be associated with AHC in man, 1 " 5 were not observed in rabbits. The clinical disease and maximal virus levels in tears of rabbits occurred 24 hr after infection. In addition to virus, IFN was detected in tears collected from 6 of 16 rabbits 1 and 2 days after infection. We have detected IFN in 30% of tears collected from AHC patients. 7 ' 8 Virus and IFN in tears of rabbits declined to undetectable levels by day 5 post-infection. Taken together, these findings suggest that the course of conjunctivitis in association with virus shedding and the IFN response is similar to that observed for humans infected with AHC viruses. 5 ' 7 ' 8 Serum antibody levels in rabbits were highly variable, a finding which is consistent with observations in patients having AHC. 6 This may be due in part to the superficial nature of the infection, a local immunologic response, and/or inability to detect neutralizing antibody using the parental EV70 (see discussion below). In addition, only low levels of ENA 5 ' 8 specific for EV70 was detected in tears of 4 of 21 rabbits. The detection of ENA was not related to severity of disease, level of virus in tears, or level of serum antibody response (data not shown). Since the findings of ENA is variable in man 8 and rabbits and its production is dependent upon an unknown factor(s), further studies are being performed to determine if virus adaptation gives rise to a more severe infection with a more consistent production of ENA in rabbits. Interestingly, during the adaptation of the EV70 isolate (J648/71) to rabbits, it was noted that antisera made to the parental EV70 contained significantly more neutralizing activity to the parental EV70 than the rabbit adapted EV70. How this antigenic change occurred is not known, but it was reproducible with other EV70 isolates in our laboratory. An antigenic change in EV70 due to passage in vitro was first observed by Mirkovic et al. 13 In addition, other investigators have reported antigenic changes among EV70 isolates from different epidemics since Our findings suggest that an antigenic change can also occur during adaptation of EV70 from tissue culture to animals. This finding further supports the idea that EV70 may have evolved from an animal reservoir 1015 and suggests that the mechanism may have been associated with an antigenic shift in the parental virus. Additional studies are being done to determine if the animal adapted EV70 will revert to the parental EV70 after adaptation to cell culture. In summary, our results indicate that EV70 can cause a conjunctivitis in rabbits that is similar to the conjunctivitis caused in man. This reproducible animal infection can be used to study the course of the infection and host defense responses of the eye, investigate the mechanism(s) associated with the evolution and natural selection of new antigenic variants of EV70, and for testing the efficacy of vaccines and therapeutic agents like arildone, IFN and antibody 16 against this enteroviral infection. Key words: enterovirus type 70, conjunctivitis/conjunctiva, animal model, antigenic shift, eye infection/eye disease, interferon, host defenses/immunologic response

6 9 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1986 Vol. 27 Acknowledgments The authors thank Anita Villarreal-Pitts and Donna Clegg for their expert technical assistance. We also thank Dr. Alvin R. Solomon, Jr. (Department of Dermatology, University of Texas Medical Branch at Galveston, TX) for discussion and examination of the conjunctiva and corneal sections. References 1. Chatterjee S, Quarcoopome CO, and Apenteng A: Unusual type of epidemic conjunctivitis in Ghana. Br J Ophthalmol 54:628, Kono R, Sasagawa A, Ishi K, Sugiura B, Ochi M, Matsumiya H, Uchida Y, Kameyama K, Kaneko M, and Sakurai N: Pandemic of new type of conjunctivitis. Lancet i:l 191, Yin-Murphy M and Hai LK: Viral epidemic conjunctivitis in Singapore ( ). In Viral Diseases in Southeast Asia and the Western Pacific, MacKenzie S, editor. New York, Academic Press, 1982, pp Pramanki DD: Joy bangla, an epidemic of conjunctivitis in India. Practitioner 7:805, Langford MP, Stanton GJ, Barber JC, and Baron S: Early-appearing antiviral activity in human tears during a case of picornavirus epidemic conjunctivitis. J Infect Dis 139:653, Kono R, Miyamura K, Tajiri E, Shiga S, Sasagawa A, Irani PR, Katrak SK, and Wadia NH: Neurologic complications associated with acute hemorrhagic conjunctivitis virus infection and its serological confirmation. J Infect Dis 129:590, Langford MP, Yin-Murphy M, Ho YM, Barber J, Baron S, and Stanton GJ: Human fibroblast interferon in tears of patients with picornavirus epidemic conjunctivitis. Infect Immun 29:995, Langford MP, Barber JC, Sklar VEF, Clark HI SW, Patriarca PA, Onarato IM, Yin-Murphy M, and Stanton GJ: Virus-specific, early appearing neutralizing activity and interferon in tears of patients with acute hemorrhagic conjunctivitis. Curr Eye Res 4:233, Hsiung GD and Henderson JR: Diagnostic Virology. New Haven, Yale University Press, 1964, pp Langford MP and Stanton GJ: Replication of acute hemorrhagic conjunctivitis viruses in conjunctival-corneal cell cultures of mice, rabbits and monkeys. Invest Ophthalmol Vis Sci 19:1477, Reid DWB: Statistical methods in virology. In Textbook of Virology for Students and Practitioners of Medicine and Other Health Sciences, Rhodes AJ and Van Rooyen CE, editors. Baltimore, The Williams and Wilkins Co., 1968, pp Langford MP, Weigent DA, Stanton GJ, and Baron S: Virus plaque reduction assay for interferon: microplaque and regular macroplaque reduction assays. Methods Enzymol 78:339, Mirkovic RR, Kono R, Yin-Murphy M, Schmidt NJ, and Melnick JL: Enterovirus type 70: the etiologic agent of pandemic acute hemorrhagic conjunctivitis. Bull WHO 49:341, Kawamoto H: Antigenic analysis of acute hemorrhagic conjunctivitis viruses (enterovirus type 70). Microbiol Immunol 23: 859, Yoshii T, Natorik, and Kono R: Replication of enterovirus 70 in non-primate cell cultures. J Gen Virol 36:377, Langford MP, Carr DJJ, and Yin-Murphy M: Activity of arildone with or without interferon against acute hemorrhagic conjunctivitis viruses in cell culture. Antimicrob Agents Chemother 28(5): 1985.