Recent Surprises in Marek's Disease: New Pathotypes,

Neurological Disease and Vaccines.

R. L. Witter

(USDA-ARS Avian Disease and Oncology Laboratory

East Lansing, MI 48864)

Background:

Pathaotyping tests conducted on MDVs isolated at different times over the past 35 years convincingly (and surprisingly) demonstrate a pattern of mutation to increasing virulence. For example, isolates like the JM strain although able to cause high rates of disease in genetically susceptible, maternal-antibody-negative (MAB-) chickens, cause little or no disease in chickens vaccinated with HVT. Isolates like Md5 cause high rates of disease in HVT-vaccinated chickens but much lower incidence in chickens vaccinated with the bivalent, HVT + Sb-1 vaccine. Isolates like 648A cause high rates of disease even in chickens vaccinated with HVT+SB-1 vaccine. The v pathotype viruses first appeared in the 1960s and were responsible for the catastrophic losses experienced prior to the introduction of vaccines. The vv pathotype viruses first appeared in the late 1970s whereas the vv+ viruses were first noted in the early 1990s. At the present time, most of the viruses isolated from problem flocks in the field type as vv+ or vv.

The procedure for pathotype determination is not standardized between laboratories. In our laboratory, each virus tested requires an average of 150 chickens, including nonvccinated, HVT-vaccinated and bivalent-vaccinated groups in two replicates each held for 9 weeks. This procedure is too slow and too costly for routine diagnostic use. Therefore, other techniques are being examined. Calnek found an apparent relationship between viral pathotype and lymphoid organ size at 6-9 days postinoculation . However, in later tests in which Calnek*s laboratory collaborated with our laboratory, this relationship did not hold up well for all isolates tested (unpublished data). In our laboratory, Gimeno found a fairly strong relationship between viral pathotype and neurological disease, especially classical and acute forms of transient paralysis. This test may be practical as results may be obtained by clinical observation of chickens through 21 days postinoculation. We have also found that pathotypes may, in some case, be differentiated by responses in genetically resistant chickens. However, none of these alternative procedures has yet been validated. Pathotyping remains a difficult and costly procedure.

Results from pathotyping tests may be biased if the test preparation contains other, contaminating viruses. It is important to insure that the virus to be tested is pure , i.e., not contaminated with MD vaccine viruses, avian leukosis virus, reticuloendotheliosis virus, or chicken infectious anemia virus. To this end, it is helpful to notethat these viruses appear to have unexpected biological properties. Field observations indicated that these strains are capable of infecting and inducing disease in mature chickens whereas, previously, old chickens were considered refractory to de novo inoculation with MDV. An age-related resistance to MD has been described. In recent studies at our laboratory, 18-week-old unvaccinated chickens developed severe manifestations of MD following inoculated with several vv+ strains. The birds first experienced classical and acute transient paralysis, and those that survived had high rates of persistent neurological disease and died with typical lymphomas and nerve lesions. On the other hand, we found similar chickens that had been vaccinated at hatch with HVT and exposed once at 6 weeks with the v pathotype JM strain to be totally refractory to challenge with vv+ strains at 18 weeks. Thus, the role of de novo infection with vv+ strains in late outbreaks of MD is still unresolved. Another surprise was the recent reports of MD-related lymphomas in turkeys in France and Israel, which have been assumed to be related to infection with virulent MDV strains. If the vv+ viruses have a natural host range that extends to turkeys, this would be very important but confirmation is required.

Neurological syndromes:

Paralytic clinical signs, probably related to lymphoid infiltration in peripheral nerves, has been a hallmark of MD and prompted the terminology "range paralysis" which remained in use for many years. Although the brain of MDV-infected chickens was known to have microscopic lesions, especially perivascular cuffing, clinical syndromes related to brain lesions was confined to the classical form of transient paralysis. In this syndrome, chickens develop flaccid paralysis of the neck and limbs but recover completely within 24-48 hours. This syndrome was seen often in the 1960s prior to the advent of vaccination and is still noted occasionally in countries where broiler flocks are not routinely vaccinated against MD.

Recently in our laboratory, we observed a surprising and unexpectedly high rate of early mortality careful clinical observation, we noted that death was invariably preceded by flaccid paralysis. Following inoculation of 3- week-old nonvaccinated chickens with vv and vv+ strains of MDV. Upon Microscopic brain lesions included vasculitis and vasogenic edema, especially in the cerebellum, which has been considered diagnostic of transient paralysis. However, the high rates of death (up to 100%) between 9 and 20 days differed from earlier reports and prompted us to consider this a new manifestation that we termed acute transient paralysis. We also found that acute transient paralysis was only induced by the more virulent virus strains in genetically susceptible chickens. Survivors of acute transient paralysis tended to develop persistent neurological signs such as tics, ataxia and torticollis. Viral DNA was found in the brain by polymerase chain reaction assays but was not necessarily associated with the clinical manifestations of acute transient paralysis. Thus we feel that the recent, highly virulent strains of MDV have defined the brain as a previously unrecognized target organ in MDV infection. This conclusion is further supported by the recent work of Cho et al. who reported extensive necrotic brain lesions with accumulations of MDV antigens in chickens inoculated with the Md5 strain, a vv pathotype virus.

Among the interesting issues is the apparent multiplicity of neurological syndromes now observed in MK. In addition to paralysis associated withlymphoid infiltration of peripheral nerves, we now have classical and acute transient paralysis, and persistent neurological disease. A fourth syndrome, termed late paralysis, was observed at 20-22 days after inoculation and appeared to have many features of transient paralysis. We feel this may be an artifact of the experimental procedure resulting from the failure to infect a proportion of the chickens inoculated intraabdominally at 3 weeks of age. The interactions of these various neurological syndromes with each other and with other manifestations of MD including lymphoma formation is poorly understood, but should be a fruitful area of investigation.

Another interesting issue is the role played by MDV in the brain. Our recent work (5; Gimeno, unpublished data) show that concentrations of MDV DNA in the brain vary widely but can be very high and do not correlate well with antigen load. The endothelial cells, probably the key targets for transient paralysis, do not produce viral antigens as detected with antibodies to pp38 or gB. It may be important to determine which cells in the brain are infected and whether the infection I latent or productive.

Vaccines:

Vaccines and vaccination have been central to efforts to control MD for 30 years. The practical control of MD through the use of attenuated serotype 1 MDV or serotype 3 HVT was considered a true surprise for at least two reasons. First, the dramatic and extensive reduction of losses in the filed exceeded all previous experience with poultry vaccines. Also, this group of vaccines was the first directed towards oncogenic viruses in either man or animals.

The most widely used vaccines for MD are composed of live viruses of either serotype 3, serotype 2, or attenuated variants of serotype 1 administered as suspensions of live chicken embryo cells. The vaccines are administered to chicks at the day of hatch or, because of recently developed technology, to embryos at the 18th day of incubation. Early administration is necessary to insure the development of an immune response before chickens are infected with virulent field strains. The reduction of vaccine efficacy by the presence of maternal antibodies is not sufficient to outweigh the advantages of early administration. The vaccine viruses replicate in host cells, establishing latent infection in b lymphocytes which persists indefinitely. Thus, repeat vaccination to boost immunity is not usually considered necessary even though empirical use of double vaccination is common in certain countries.

The efficacy of vaccine viruses has been compared in various studies. In brief, HVT and certain clones of Rispens are least effective, bivalent vaccines of either HVT + SB-1 or HVT + 301B/1 are more effective, and the original, low passage Rispens virus is most effective. Other attenuated serotype 1 viruses such as R2/23 are also effective but rank behind Rispens.

Immunity is induced quickly, and significant protection can be measured against challenge with virulent MDV by 5 days post vaccination. Both cellular and humoral immune responses are induced, but the cellular responses are considered the more important. Immunity is long lasting, and protection has been measured for as long as 40 weeks in the absence of intervening viral challenge. However, it is possible that external factors may affect the immune response as vaccinal protection is reduced in the presence of other immunodepressive virus infections and it seems likely that many disease outbreaks in mature chickens result from early infection triggered later by some type of environmental stress. The amount of protection provided through vaccination is difficult to quantitate and is subject to a number of variable factors. The amount of protection provided through vaccination is difficult to quantitate and is subject to a number of variable factors. However, it should be considered high compared to other conventional vaccines for poultry or any other animals. When one compares actual MD condemnations with rates prior to the introduction of vaccines, estimated levels of protection in commercial broiler flocks exceed 95%.

Synergism between two different vaccine viruses is the basis of bivalent vaccination such as with HVT+ SB-1 vaccines. This synergism is most apparent between serotypes 2 and 3, and results in greater protection than induced by either virus alone or that can be explained by the sum of the individual vaccine doses. In commercial use, bivalent combinations of serotypes 1 and 3 or trivalent combination of serotypes 1, 2 and 3 are frequently used. The basis for this synergistic effect is poorly understood. However, it represents one approach to augment the effectiveness of existing vaccines and probably deserves more study.

During the past 15 years, much effort has been committed to development of vaccines using organisms modified by recombinant DNA technology. Fowlpox virus (FPV) and HVT have been used to express genes from serotype 1 MDV. Our work with FPV vectors has resulted in constructs that provide significant protection against virulent MDV challenge. This protection can be augmented by combination with live HVT, in a manner analogous to the synergism described above. However, these FPV constructs do not offer protection superior to that provided by conventional vaccine viruses. HVT constructs expressing serotype 1 MDV genes have also been evaluated. As yet, however, there is no such recombinant vaccine available commercially and the enthusiasm once associated with this approach seems to have dimmed. In theory, a deletion mutant serotype 1 MDV vaccine might be a good vaccine candidate. If one deleted or disabled genes associated with either oncogenicity or in vivo replication, the resulting vaccine might be both attenuated and immunogenic. However, Morgan and co-workers deleted the putative oncogene, meq, gene from serotype 1 MDV but found the resulting construct replicated poorly and was an ineffective vaccine.

The safety of MD vaccines has not been of major concern. HVT seems totally avirulent. Serotype 2 vaccines are avirulent but in certain strains of chickens can enhance tumorigenesis caused by avian leukosis virus. This has ceased to be a concern due to the eradication of avian leukosis virus from susceptible White Leghorn stocks. Serotype 1 vaccines, if sufficiently attenuated, are totally avirulent. However, it is necessary to how the absence of reversion through the use of backpassage studies. Most of the attenuated virus do not revert to virulence. However, the Md11/75C strain regained partial virulence upon 5 backpassages in chickens. This virulence could be eliminated by further passage in call culture.

If serial passage in cell culture is continued long enough, the virus looses its ability to replicate in the chicken and is ineffective as a vaccine. If serial passage is continued only long enough to attenuate virulence, a number of other properties also change indicating the presence of multiple mutations. On the assumption that lower passage serotype 1 vaccines might make better vaccines, e endeavored to use the 132 base pair repeat marker as an early indicator of attenuated clones. Unfortunately, this attempt failed. We obtained several clones with altered 132bpr phenotypes at less than 40 passages but all ere virulent. This illustrated that the 132bpr marker is not tightly linked with oncogenicity as originally thought. Markers to identify avirulent or nononcogenic clones early in serial passage are still elusive.

Attenuation by serial passage in cell culture and gene deletion by recombinant DNA technology are both ways to achieve mutants of serotype 1 MDV, some of which could be useful vaccine candidates. However, kung and associates found that co-infection of MDV with retroviruses may result in the insertion of retroviral sequences, usually the LTR, into the herpesvirus genome thus achieving a type of insertional mutagenesis. One such mutant, RM1, resulted from the co-infection of the JM strain of MDV with reticuloendotheliois virus . This virus contained a retroviral insert at the junction of the unique short and internal repeat regions, impacting the function of SORF2. This virus was attenuated for oncogenicity but replicated much like low passage wild type virus. Curiously, this virus provided excellent protection against challenge with virulent MDV in preliminary trials. Perhaps retroviral induced mutagenesis I an alternative route to the generation of modified live vaccines for MD.

Whether the next generation of MD vaccines? This question is all the more pertinent as one considers that Rispens has been used in the United States for 5 years already and no better products appear to be in the pipeline. If MDV continues to evolve to greater virulence thereby resulting in reduced efficacy of existing vaccines, will we have a better product ready? The outlook is not very promising. Overall effort in MD vaccine development appears to have lessened. Recombinant vaccines have been disappointing. Natural mutants with value as vaccines are probably rare and discovery of new strains is problematic. Even if a better vaccine can be developed, the poultry industry will be reluctant to pay for it unless loses rise significantly above current levels. Indeed, the failure of the industry to support a price structure for MD vaccines that insures profitability and encourages further research and development by vaccine manufacturers is disturbing and does not bode well for the future. These dilemmas may well become the dominant issue as MD enters the new millennium.

Conclusions:

The future of MDV seems to be assured - it will continue to survive and prosper. The virus has the capacity to mutate as needed to insure this end. Eradication is technically possible but will be too expensive to be practical in the foreseeable future. Selection pressure to encourage further mutation will be provided by extensive reliance on our most effective vaccines. Only relatively complicated interventions can change this scenario.

The future of the nature of the disease is towards more complexity. We will continue to see lymphomas but the traditional forms of the disease will be augmented by more evidence of immunodepression, neurological disease, blindness, and their respective interactions. Older birds may become more susceptible, and the host range may expand to other species such as turkeys. The broiler industry in the United States will continue to exacerbate the problem by subordinating biosecurity to short-term profits, thus insuring that most broiler clocks will receive a heavy, early exposure to MDV.

The future of vaccines is not clear. In ovo administration will become institutionalized for broilers and may also extend to layers and breeders, however, this may not necessarily reduce disease incidence. A retrenchment of MD research towards basic molecular biology may ultimately pay off with better understandings of immune response and better techniques for deriving candidate vaccine viruses, but this will require time.

The future of the chicken is also uncertain. On one hand, breeders are quietly working to improve genetic resistance to MD using conventional approaches. Quantitative trait loci identified by genome mapping technologies appear to influence MD susceptibility and are being evaluated in commercial stocks. Modification of the chicken germline by gene transfer to create specific resistance to MDV is still under discussion, but the technology needed to even test this approach is not yet in hand. There is a fear that any gains in genetic resistance may be offset by selection for greater productivity that will inevitably bring additional physiological stress.

What will be the future economic impact of MD? We have made the argument that continued evolution of the virus will certainly result in a future disaster. How do we know this is true? Broiler condemnation trends in the United States continue to be down. How much public money should be invested now to prevent future problems? How much will be invested? How much is at stake if we guess incorrectly? Answers to these questions will affect the poultry industry in the United States and the world.

Meanwhile, the quest for an ultimate solution is ongoing. Perhaps a truly integrated approach involving vaccines, biosecurity, genetics and management will be effective. If multiple barriers are raised to the virus. Evolution from single mutations will be less likely. Any solution will have an associated cost that will defer its implementation until the industry determines it is profitable to do so. The real questions are when will we have a truly effective procedure and will the decision to implement be based on long- or short-term considerations.