*For offprint requests: tel: (916)752-8306; fax: (916)752-1354; tdyilma@ucdavis.edu

SUMMARY

We have prepared a vaccinia virus double recombinant vaccine that affords complete protection against rinderpest, the most important disease of cattle in Africa and Asia. This vaccine was developed through non-traditional methods, therefore thorough and systematic consideration has been given to possible safety issues involved in its general release, including an assessment of risk to target and non-target species and the environment. During testing of the vaccine, seronegative cattle were housed in contact with vaccinated animals to determine transmissibility of the vaccine virus. In addition, non-target species contact was maintained, with goats and chickens moving freely among the vaccinees. Effects of inadvertent vaccine virus contact by immunodeficient individuals, a primary concern where populations are known to have high infection rates with human immunodeficiency virus, were assessed in immunodeficient nude mice and SIV-infected rhesus macaques after vaccine injections. Results of these studies were determined by virus isolation and serology techniques. There were no instances of accidental transmission of the vaccine virus or damage to immunocompromised subjects. These results indicate that the vaccine virus is attenuated by the incorporation of foreign genes. Possibilities for further attenuation by insertion of lymphokine genes are being studied.

Key words: risk assessment, rinderpest, vaccinia virus, safety, target and non-target species, live recombinant vaccines

INTRODUCTION

Rinderpest is the most important livestock disease of Africa and Asia. An acute, febrific, and highly contagious viral disease of ruminants (particularly cattle and buffalo) the disease is characterized by inflammation, hemorrhaging, necrosis, and erosion of the gastrointestinal tract, accompanied by bloody diarrhea, wasting, and death; it is responsible for economic losses in the millions of dollars, as well as the attendant malnutrition and famine that result from outbreaks. This disease was controlled in Africa in the 1970's by an international effort called the Joint Project/15 (JP/15), which vaccinated 124 million cattle with the Plowright tissue culture vaccine (RBOK). However, recent resurgences have demonstrated the logistical difficulties of relying on traditional live vaccines in the regions where this disease is prevalent.

There are many reasons why a new vaccine for this disease is needed, the most important being that vaccines currently employed are all extremely heat-labile and expensive to produce. The Plowright vaccine is effective, but the requirement for skilled personnel in vaccine preparation and administration and for refrigeration facilities--often lacking in the hot and isolated areas where the disease is prevalent--are obstacles to its use. Further one is not able to use a serological test for distinguishing between vaccinated and infected animals with the use of RBOK.

The rinderpest virus (RPV), a member of the paramyxovirus group, is closely related to measles virus of humans, canine distemper virus of dogs, and peste des petits ruminants virus of goats and sheep. Paramyxoviral infection and spread are mediated by the hemagglutinin (H) and fusion (F) proteins of the virus. Antibodies to H and F have been shown to provide protective immunity to other paramyxoviral diseases as well as to rinderpest.

By employing recombinant DNA technology to insert and express the rinderpest H and F genes in vaccinia virus (VV), we have prepared a vaccine that retains the advantages but lacks the logistical disadvantages or safety hazards inherent in traditional live vaccines. This technology is useful for protection against many different diseases in a variety of species: VV replicates in humans, cattle, horses, swine, sheep, goats, mice, and monkeys, and can express one or more incorporated genes of pathogens along with those of VV. The lyophilized form of VV is heat-stable as well as easily produced, stored, transported, and administered by scarification. Each independent area can produce the vaccine and vaccinate animals as needed, without international assistance. A single calf can be used to prepare over 200,000 vaccine doses--a great advantage for use in developing countries (Metzgar, 1985).

To test the efficacy of these recombinant vaccines, protective immune response studies in cattle were conducted in the high containment facility at the Plum Island Animal Disease Laboratory. Seronegative animals were inoculated with 10⁸ plaque forming units (pfu) of the VV recombinant vaccine intradermally. Vaccinated animals were completely protected against challenge with a heavy dose (10³ TCID₅₀) of RPV on day 35 following primary immunization, demonstrating the efficacy of the vaccine. Further work is being conducted in Africa to evaluate efficacy and assess safety of the recombinant.

MATERIALS AND METHODS

Recombinant vaccines. We have previously reported the safety and efficacy of single recombinant VV vaccines that were developed by the insertion of the genes coding for F or H proteins of RPV in the thymidine kinase (TK) region of the VV genome in the WR or Wyeth strains of the virus (Yilma et al., 1988; Giavedoni et al., 1991). Although any nonessential region of the genome can be used as the site of gene insertion, the TK gene locus provides some advantages because recombinants are then TK^-, which serves to further attenuate the already low level of VV pathogenicity and provides a basis for identifying recombinants.

A Wyeth strain VV double recombinant that expresses both the F and the H genes (vRVFH) of RPV was also developed to facilitate the ease of the production and administration of the vaccine (Giavedoni et al., 1991) The TK and the H genes of VV were insertionally inactivated by the incorporation of H and F genes of RPV, respectively, into these sites. This process produces an even more highly attenuated VV recombinant vaccine that expresses authentic H and F proteins of RPV. This vaccine was similarly tested for safety and efficacy at the USDA, Plum Island Animal Disease Center (Giavedoni et al., 1991), the Kenyan Agricultural Research Institute, and the Ethiopian National Veterinary Institute (Wamwayi et al., 1996).

Animals. Cattle, goats and chickens were used to assess the transmissibility and pathogenicity of the vaccine virus across species lines by allowing free contact of vaccinated cattle with these animals. Athymic nude mice and normal mice were inoculated intraperitoneally with vRVFH to assess pathogenicity. The effects of administering vRVFH and the mixture of single recombinants (vRVF + vRVH) by intradermal scarification on the immune systems of rhesus macaques infected with the simian immunodeficiency virus (SIV)were also observed. These monkeys suffer from an immunodeficiency disease virtually identical to the human acquired immunodeficiency syndrome (AIDS); hence, SIV is considered a legitimate model for HIV (Gardner and Luciw, 1989).

Humoral immune responses. Sera were assayed for specific antibodies to RPV using the microtiter virus neutralization test (Rossiter and Jessett, 1982). Antibody titers were expressed as the reciprocal of the highest dilution of serum that gave complete protection against 100 TCID₅₀ of RPV. VV antibodies were assessed by a plaque reduction assay (Giavedoni et al., 1991). Antibody titers were expressed as the reciprocal of the highest dilution of serum that decreased the number of plaques by 50%.

RESULTS

Pock lesions developed as early as four days in all animals vaccinated with the single recombinants vRVF and vRVH, but were limited to the site of inoculation and healed completely by two weeks. Cattle vaccinated with the double recombinant vRVFH showed no pock lesions at the site of inoculation, a strong indication of the level of attenuation (Giavedoni et al., 1991; Wamwayi et al., 1996). All animals vaccinated with the recombinants or RBOK produced SN antibodies to RPV as early as 8 days after vaccination; as expected, control animals had no detectable antibody titer to either RPV or VV.

As low as one TCID₅₀ of RPV induces clinical rinderpest with 100% mortality in unvaccinated animals. Cattle vaccinated with the recombinants or TCRV were completely protected when challenged, even though the inoculation was greater than 1000 times a normally lethal dose of RPV. No anamnestic response could be demonstrated in the groups vaccinated with a cocktail of both recombinants or vRVFH.

All controls, including two unvaccinated animals that were housed with each group in order to assess the transmissibility of VV recombinants from vaccinated to contact animals, tested negative by SN and plaque reduction assays for VV antibody throughout the course of the experiment. A thorough examination also failed to demonstrate pock lesions in the contact or control animals with both groups.

Cattle vaccinated with VV recombinants were completely protected from rinderpest, exhibiting no detectable illness and a normal temperature of 38C following challenge-inoculation with 10⁴ TCID₅₀ of the pathogenic Kabete "O" strain of RPV. The four unvaccinated contacts developed high fever (42C) by day two and died by day six after challenge, showing lesions typical of severe rinderpest. Daily monitoring for two weeks showed no detectable clinical disease in vaccinated animals. A thorough check of contact cattle, goats and chickens gave no indication of vaccine virus transmission; serological and virus isolation assays remained negative.

Vaccination of rhesus macaques with vRVFH or a mixture of vRVF + vRVH resulted in no detectable adverse effects. Both normal and SIV-infected macaques vaccinated with vRVFH developed small pock lesions at the site of inoculation that healed completely by two weeks post-vaccination. No clinical disease or dissemination of VV was observed, however, in either normal or SIV-infected macaques that were followed for one month after vaccination.

Intraperitoneal inoculation of 10⁸ plaque forming units of parental Wyeth or vRVFH into immunodeficient athymic nude mice caused no disease or dissemination of virus in these animals. Mice were observed for 60 days postinoculation.

DISCUSSION

Rinderpest is an excellent candidate for eradication using a VV recombinant vaccine, which has the advantages of thermostability, low cost of production, and ease of vaccination by a number of routes including intradermal, intramuscular, or oral. There is only one serotype of RPV, although there are different strains manifesting different degrees of pathogenicity in the field, and the virus does not exist in a latent or carrier state. a vaccine against one strain will immunize against all, including PPR of sheep and goats (Jones et al., 1993). Use of the rinderpest double recombinant VV in areas of the world where PPRV is endemic would also aid in the control and eradication of PPR.

Another major disadvantage of most conventional vaccines (including the RBOK) is the lack of serological distinction between vaccinated and RPV-infected animals. Thus, cattle that are seropositive to RPV, either as a result of vaccination with RBOK or infection, are barred from lucrative export markets. In contrast, vRVFH-vaccinated cattle are only positive to the F and H protein of RPV and can easily be distinguished from those exposed to all the proteins of the virus, whether by vaccination with RBOK or by infection. We have developed a rapid diagnostic kit based on the nucleoprotein of RPV (N protein) to use in such differential diagnosis of rinderpest exposure, as well as a similar kit for PPRV (Ismail et al., 1994; 1995). With the ability to differentiate vaccinated from infected animals, we can facilitate both the export of vaccinated animals and the global eradication of rinderpest. Epidemiological surveys during outbreaks can accurately track the disease while vaccination programs check its progress.

It is important, however, to consider all possible results, negative as well as positive, from general release of these new preparations (Ada, 1991). Safety concerns have arisen primarily from fears of inadvertent transmission and pathogenicity to non-target species. However, our tests show that the VV recombinant vaccine for rinderpest did not spread from vaccinees to contacts of the same or other species, even though these animals were housed together during the entire period of virus replication. Thorough examination revealed no pock lesions at any point on the control animals, and serological tests for VV and RPV antibodies in controls similarly proved negative.

We have tested vRVFH and the parental Wyeth strain for pathogenicity by inoculating normal and nude mice intraperitoneally; no clinical disease or dissemination of virus was detected in any of the test mice. We have also assessed the effects of vRVFH and the mixture of single recombinants (vRVF + vRVH) on the immune systems of rhesus macaques infected with SIV. These monkeys suffer from an immunodeficiency disease virtually identical to AIDS; hence, SIV is considered a legitimate model for HIV (Gardner and Luciw, 1989). Normal and SIV-infected macaques vaccinated with vRVFH both developed small pock lesions at the site of inoculation that healed completely by two weeks postvaccination. No clinical disease or dissemination of VV was observed, however, in either normal or SIV-infected macaques followed for one month after vaccination.

With the threat of untoward effects always in mind, we have investigated possibilities for further attenuation of the VV recombinant vaccines by insertion of lymphokine genes like interferon-gamma (IFN-) or interleukin-2 (IL-2) (Yilma et al., 1987; Ramshaw et al., 1987; Flexner et al., 1987; Giavedoni, et al., 1992). This gives us a new attenuating approach that will likely be applicable to other live recombinant vaccines. We have constructed fusion proteins of IFN- (human and murine species) with various immunogens, and demonstrated reduced pathogenicity for immunodeficient (nude) mice of several VV recombinants expressing these proteins. Immunodeficient mice were able to clear the recombinant VV at doses as high as 10⁸ pfu, while as few as 100 pfu of the WR strain of VV is normally 100% lethal for nude mice. Mixtures of recombinant VV were always lethal to the mice if one of the recombinant.

Concern about using live VV recombinant vaccines in the field has focused on the potential for recombination with other poxviruses and/or reversion to the parental type. The probability of such events is extremely low (Ball, 1987), and even if it were to take place, it would only generate the Wyeth strain of VV, a strain of virus which had very few complications (2-6/million vaccinees) during the global eradication of smallpox (Neff, 1965; Lane, 1971). VV has not established itself in nature even after billions of people were vaccinated during the smallpox eradication program, sometimes under the most unhygienic conditions with people and livestock sharing the same quarters.

VV is stable and can persist for prolonged periods under normal environmental conditions, but is not found in the environment. We do not expect the recombinant vaccine, an extremely attenuated strain, to have as good a chance to establish. No carrier state is known to exist for either RPV or VV. vRVFH does not cause pock lesions in cattle, and no shedding occurs from vaccinated animals into the environment. In addition, we have demonstrated for a number of recombinant viruses constructed with either the WR or Wyeth strains of VV an inability to spread to contact groups (Mackett et al., 1985; Yilma et al., 1988; Giavedoni et al., 1991).

In summary, during this past century the VV vaccination procedure has protected hundreds of millions of people worldwide from smallpox, with minimal side effects and no adverse environmental impact despite global release of the VV under varying hygienic conditions. Modern vaccine technology now has given us, through genetic engineering, the techniques to use VV as a live vector to protect animals and people against other unrelated disease agents. The VV recombinant vaccines for rinderpest (vRVH, vRVF, and vRVFH) have been constructed using the Wyeth strain of VV provided by the New York City Board of Health. This strain has been used worldwide for the smallpox eradication program with an extremely low rate of complications, although certain risks, primarily dermal in nature, have been encountered.

The risks assessed during the smallpox eradication campaign ranged from 2 to 6 complications per million vaccinees (Neff, 1965; Lane, 1971). Insertional inactivation of the TK gene for construction of the recombinants, it is estimated, further reduces the risk by a factor of ten, i.e., 2-6 complications per ten million vaccinees (Buller et al., 1985). We have demonstrated that Insertional inactivation of the HA gene of VV further attenuates the virus (Giavedoni et al., 1991). Assuming another tenfold decrease in virulence, one might expect 2-6 complications per 100 million vaccinees. Accidental exposure in this procedure has been calculated at one chance in 300. This means that in order to get 2-6 human complications from use of the VV recombinant rinderpest vaccine in cattle, we would have to involve directly 30 billion people in the vaccinations. Since there are only five billion people on earth, we believe it is safe to say that the vaccine is relatively free from unintended consequences for humans.

ACKNOWLEDGMENTS

The development of the double recombinant vaccine for rinderpest was supported by FAO/UN grant #891487. The development of the single recombinant vaccines for rinderpest and testing in cattle was supported by USAID Cooperative Agreement DAN-4178-a-00-6040-00, USAID-Egypt Cooperative Agreement 263-0152-a-00-1021-00, and NIGMS Graduate Training Program in Biotechnology GMO8343-01A1. The work on SIV and HIV was supported by National Institute of Health grants UO1-A129207 and O1-A1226471. Other support included CFAR Grant AI27732, Base Grant to Calif. Reg. Primate Center RR00169, and Training Grant AI07398. I acknowledge the critical review and extensive editing of this manuscript by my colleague, Dr. Sally Owens. I express appreciation and gratitude to my colleagues, graduate students, and postdoctoral fellows who have contributed so much to the work described in this article.

REFERENCES

Ada G. 1991 Real and imagined dangers. Nature 349:369.

Ball LA. 1987 High-frequency homologous recombination in vaccinia virus. J. Virol. 61:1788-1795.

Buller RML, Smith GL, Cremer K, Notkins AL, Moss B. 1985 Decreased virulence of recombinant vaccinia virus expression vectors is associated with a thymidine kinase-negative phenotype. Nature 317:813-815.

Fenner F, Wittek R, Dumbell KR. 1989 The orthopoxviruses. Academic Press, San Diego.

Flexner C, Hugin A, Moss B. 1987 Prevention of vaccinia virus infection in imunodeficient mice by vector-directed IL-2 expression. Nature 330:259-262.

Gardner MB, Luciw PA. 1989 Animal models of AIDS. FASEB J 3:2593-2606.

Giavedoni L, Jones L, Mebus C, Yilma T. 1991 a vaccinia virus double recombinant expressing F and H genes of RPV protects cattle against rinderpest and causes no pock lesions. Proc Nat Acad Sci USA 88:8011-8015.

Giavedoni L, Jones L, Gardner, MB, Gibson HL, Ng CTL, Barr PJ, Yilma T. 1992 Vaccinia virus recombinants expressing chimeric proteins of human immunodeficiency virus and interferon-gamma are attenuated for nude mice. Proc Natl Acad Sci USA. 89:3409-3413.

Ismail T, Ahmad S, D'Souza-Ault M, Bassiri M, Saliki J, Mebus C, Yilma T. Cloning and expression of the nucleocapsid gene of virulent kabete o strain of rinderpest virus in baculovirus: use in differential diagnosis between vaccinated and infected animals. Virology 1994;198:138-147.

Ismail T, Yamanaka M, Jeremiah T. Saliki, El-Kholy A, Saliki J, Mebus C, Yilma T. Cloning and expression of the nucleocapsid gene of peste des petits ruminants virus in baculovirus for use in serological diagnosis. Virology 1995;208: 776-778.

Jones L, Giavedoni L, Saliki JT, Brown C, Mebus C, Yilma T. 1993 Protection of goats against peste des petits ruminants with a vaccinia virus double recombinant expressing the F and H genes of RPV. Vaccine 11:961-964.

Lane JM, Miller JS, Neff JM. 1971 Smallpox and smallpox vaccination policy. Ann. Rev. Med. 22: 251-272.

Mackett M, Yilma T, Rose JK, Moss B. 1985 Vaccinia virus recombinants: expression of VSV genes and protective immunization of mice and cattle. Science 227:433-435.

Metzgar DP. 1985 Smallpox vaccines in vivo production and testing. In: Quinnan GV (ed) Vaccinia Viruses as Vectors for Vaccine Antigens. Elsevier-North Holland, NY, pp 109-111.

Miller RB. 1991 Proposed field trial in Pennsylvania of a live experimental vaccinia-vector recombinant rabies vaccine for raccoons. USDA/APHIS.

Neff J. 1985 Vaccinia virus vaccines: virulence and attenuation of vaccinia strain variation. In: Quinnan, GV (ed) Vaccinia Viruses as Vectors for Vaccine Antigens. Elsevier-North Holland, NY, pp 69-75.

Ramshaw IA, Andrew ME, Phillips SM, Boyle DB, Coupar BE. 1987 Recovery of immunodeficient mice from a vaccinia virus/Il-2 recombinant infection. Nature 329:545-546.

Rossiter, PB, Jessett DM. 1982 Microtiter Techniques for the Assay of Rinderpest Virus and Neutralizing Antibody. ResVet Sci 32:253-256.

Wamwayi HM, Wafula JS, Kariuki DP, Jones L, Yilma T. Confined field trial of a vaccinia virus double recombinant vaccine for rinderpest (vRVFH) in Kenya: I. Determination of the minimum immunizing protective dose in African Zebu cattle (in review).

Yilma T, Anderson K, Brechling K, Moss B. 1987 Expression of an adjuvant gene (BoIFN-gamma) in infectious vaccinia virus recombinants. In: Chanock RM, Giensburg H, Lerner R, Brown F (eds) Vaccines 87. Modern Approaches to New Vaccines including Prevention of AIDS. Cold Spring Harbor Laboratory. Cold Spring Harbor, NY, pp 393-394.

Yilma T, Hsu D, Jones L, Owens S, Grubman M, Mebus C, Yamanaka M, Dale B. 1988 Protection of cattle against rinderpest with infectious vaccinia virus recombinants expressing the HA or F gene. Science 242:1058-1061.