Vaccine Development Against Onchocerca volvulus :
Lessons from a Mouse Model

by David Abraham

1. Introduction
An integral component of any effort to develop a vaccine against a human disease is an animal model for the disease. Animal models are used to study the interaction between disease and immunological processes and for the screening of vaccine candidates. One of the significant limitations on studies of the immune response to O. volvulus has been the limited host range. Experimentally, only non-human primates such as chimpanzees and mangabey monkeys have been shown to support complete development of the parasite. Neither of these hosts is of practical use for the initial phases of vaccine development. It was for these reasons that efforts have been made to develop a mouse model for the study of immunity to O. volvulus.
Another handicap with working on O. volvulus is the difficulty in obtaining live fresh larval stages. This difficulty has been addressed through the use of cryopreserved larvae. However, a question to be anwered is "does the immune system recognize and respond to cryopreserved larvae in a manner similar to fresh larvae?".
Several studies have tried unsuccessfully to infect laboratory rodents with O. volvulus. To determine whether mice were resistant to all stages of O. volvulus, diffusion chamber methodologies have been employed. Diffusion chambers are 14 mm Lucite rings sealed with membranes of various pore sizes. Larvae are inserted into the chambers and then implanted subcutaneously. The membrane pore size prevents the larvae from leaving while allowing host cells and serum to enter. It could therefore be determined whether mice are susceptible to the early larval stages, even if they are not supportive of the complete life cycle.

2. Model Development I
The first question addressed in these studies was would cryopreserved and fresh larvae survive and develop in diffusion chambers at equivalent rates. The second fundamental question was whether larvae would survive and develop at equal rates in primate and rodent hosts. Third-stage larvae (L3) were implanted in diffusion chambers in chimpanzees, mangabey monkeys, rhesus monkeys, squirrel monkeys and inbred strains of mice, jirds and rats for 3-63 days. At different times during the experimental period, larvae were recovered and assessed for their viability and development. Survival and growth rates were equal regardless of whether the implanted larvae were fresh or cryopreserved. Survival and growth rates of the implanted larvae did not differ among the primate and rodent hosts tested, with the exception of squirrel monkeys and rats, which were resistant to infection. Molting from L3 to fourth-stage (L4) larvae began on day 3 and continued through day 14 in the primates and rodents. It was therefore concluded that the use of cryopreserved larvae in place of fresh larvae was reasonable and that mice could be used as a model for the study of the first few weeks of infection.

3. Development of Protective Immunity
The next question to be addressed was whether mice could be induced to kill larvae through immune mediated processes. BALB/cBYJ mice were immunized by subcutaneous injection of normal, irradiated, or freeze-thaw killed Onchocerca sp. larvae. The mice received challenge infections of O. volvulus L3 contained in diffusion chambers implanted subcutaneously. At two weeks postinfection, the diffusion chambers were removed and larval survival was assessed. When mice were immunized a single time with 35 krad-irradiated or normal O. volvulus L3, there was a significant reduction in the survival of challenge parasites. However, there was little or no reduction in challenge worm survival when mice were immunized a single time with freeze-thaw killed O. volvulus L3 or L4, or irradiated O. lienalis L3. When a second dose of freeze-thaw killed O. volvulus L3 or irradiated O. lienalis L3 was administered, there was a significant reduction in parasite survival in immunized mice. Immunization with O. volvulus L4 or a combination of L3 and L4 failed to confer protection. These results demonstrated that mice could be immunized against larval O. volvulus and that diffusion chambers were an efficient method for studying protective immunity to this parasite in a mouse model.

4. Mechanism of Immune Mediated Killing of Larvae
Using this mouse model the mechanism of immune mediated killing of challenge larvae as induced by irradiated L3 was studied. Direct contact between host cells and parasites was required for killing of larvae in immunized hosts. To define the mechanism of immune-mediated killing in this system, the time of influx of cells and cytokines into the infection site were compared to the time challenge infections were killed. The only cell type that was found to increase in diffusion chambers in immunized mice was the eosinophil; maximal levels of eosinophils were coincident with the time of parasite killing. IL-5 was found in diffusion chambers of immunized, but not control, mice coincident with the time of parasite killing. Significant levels of Interferon-g were absent in the diffusion chambers of both groups. Immunized mice were next treated with mAb to eliminate IL-5 or IL-4 to assess the role that these cytokines or their byproducts play in larval killing. Elimination of either IL-5 or IL-4 significantly reduced the protective effects of vaccination against larval O. volvulus. These findings suggests that the protective immune response in mice to larval O. volvulus is dependent on eosinophils (as shown by IL-5 dependency) and IgE or IgG1 (as shown by IL-4 dependency).
Humoral responses of immunized mice to O. volvulus L3 antigens were also examined. ELISA measurements of total serum antibody levels indicated that IgE was the only antibody isotype elevated in mice immunized with O. volvulus L3. IgM from immunized mice was the only isotype that recognized surface antigens on intact O. volvulus L3. IgG1, IgG3, IgE and IgA recognized internal parasite antigens on O. volvulus L3 frozen sections. Western blot analysis of L3 proteins showed that the serum from mice immunized with O. volvulus L3 IgG1, IgG2a/2b, IgA, and IgE, as well as IgM, recognized unique L3 proteins. Antibodies in serum from L3 immunized mice were able to detect O. volvulus adult antigens in a pattern similar to the recognition found in O. volvulus L3. Some L3 antigens were shared by adults, while other antigens were L3 specific. The ELISA, immunohistochemistry and Western blot findings thus demonstrated a complex pattern of parasite antigen recognition by antibodies found in mice immune to the L3 of O. volvulus.

5. Summary of Antigen Screen
The mouse model was also used to screen recombinant antigens supplied by collaborators, as arranged by the Onchocerciasis Task Force of the Edna McConnell Clark Foundation. A variety of criteria were used in the selection of these antigens, all of which indicated that a particular antigen might have vaccine potential. The following immunization protocol was used. Mice were injected subcutaneously with 25 mg of a particular antigen mixed with a water-in-oil block copolymer adjuvant. This was followed four weeks later with an identical booster immunization; challenge infections were given two weeks later in diffusion chambers which were implanted for two weeks. Six out of the 16 antigens tested induced protective immunity in single experiments using challenge L3 which were prepared in Liberia. When these experiments were repeated with L3 from Cameroon protective immunity was not observed with any of the antigens. Results from all of the single antigen immunizations are summarized in the accompanying table.

6. Model Development II
As is clear from the table, significant differences were seen in the ability of mice to kill L3 dependent on the source of the larvae. The reason for the change in the source of the larvae was due to political unrest in Liberia. The phenomenon of resistance to immune mediated killing of larvae prepared in Cameroon was also seen in mice immunized with irradiated larvae. This finding precipitated a series of experiments whose goal it was to determine why protective immunity was no longer evident in the mouse model.
Immune mediated killing of challenge larvae was not observed using proven protocols with irradiated larvae, dead larvae, or recombinant antigens nor with new immunization methods using different routes and various adjuvants. It was determined that there was no change in the mouse strain, the x-ray dose, or the materials making up the diffusion chambers. At this point it was concluded that something had changed with the larvae and not with the experimental procedure; it was not known, however, whether the problem in demonstrating protective immunity was in the inability of the larvae to induce protective immunity, or if the failure was because the larvae were resistant to the induced immunity.
It was hypothesized, based on observation of larval-specific immune responses in mice immunized with the cryopreserved larvae, that the deficit in the system was in the ability of the larvae to be killed. It was further hypothesized that the cryopreservation process itself was affecting larval susceptibility to immune-mediated killing. To test these hypotheses, mice were immunized with cryopreserved L3 from Cameroon. Each immunized and control animal received two chambers with challenge larvae. One chamber contained fresh larvae (i.e. not cryopreserved) and the other chamber contained cryopreserved L3 from Cameroon. Diffusion chambers were implanted for 14 days after which they were recovered. The results from this experiment showed that there was a 50% reduction in fresh larval survival in immunized mice whereas cryopreserved worms were not killed. These results were especially compelling because fresh and cryopreserved larvae were implanted in the same animals.
The previous experiment conclusively demonstrated that immunization with cryopreserved L3 from Cameroon was effective at inducing protective immunity to the larval stage. Previous studies using larvae prepared in Liberia, showed that the challenge larvae were killed by day 5 post challenge. All of the experiments described above had diffusion chambers implanted for 14 days, thus allowing almost three times the required amount of time for killing to occur. It was next hypothesized that larvae cryopreserved in Cameroon might require more time to be killed and that they would be killed if they were implanted in vivo for more than two weeks. In the experiment designed to test this hypothesis it was determined that killing of cryopreserved challenge larvae was seen at three and at four weeks post challenge. It was also noted that larval development from L3 to L4 was retarded in immunized mice at the two week time point. Larval developmental retardation in immunized animals has been observed with several other filarial parasites. Therefore, immunity against the cryopreserved L3 from Cameroon was demonstrated by developmental retardation at two weeks and by larval death beginning at three weeks post-challenge.
To further test the effect of cryopreservation on the ability of larvae to induce or be killed by the immune response, the following experiment wasperformed. Mice were immunized with larvae prepared in Cameroon which were cryopreserved using different methods. The mice then received challenge infections consisting of either homologously or heterologously
cryopreserved larvae. In these experiments it was concluded that the cryopreservation method currently in use in Cameroon prepared the larvae such that immunization with them lead to immunity to larvae prepared with all other cryopreservation methods. The reverse was, however, not true; other methods of larval cryopreservation did not result in larvae which
were capable of inducing protective immunity. A cryopreservation method has therefore been identified that will yield larvae which are capable of inducing immunity to fresh larvae and to larvae cryopreserved using a variety of methods. These cryopreserved larvae thus provide the closest representation of that which occurs with fresh larvae.

7. Conclusion
Progress towards developing a vaccine against onchocerciasis has been diverted during the past months because of a failure in the mouse model to demonstrate protective immunity. This problem has been solved and it has been demonstrated that the lack of protective immunity was caused by a change in the manner in which the L3 were being cryopreserved. This diversion has brought to light an issue of critical importance which would have otherwise been overlooked. Experiments have been concluded in which the appropriate method for larval cryopreservation has been determined. The final result is an animal model which will mimic human disease more accurately and thus be of value as a tool in the effort to develop a vaccine against onchocerciasis.


Summary of recombinant antigens in the mouse/diffusion chamber screen (Abraham)
Antigen Source % Reduction in Larval Survival Source of L3
Oncho C-27 McKerrow 0 Liberia
OI-3 Perler 35* Liberia
OI-3 tString 0 Cameroon
OV-7 Lustigman 33* Liberia
OV-7 tString 0 Cameroon
Paramyosin McReynolds 0 Liberia
RAL-2 Unnasch/Chakravarti 45* Liberia
RAL-2 tString 0 Cameroon
OV-9 Lustigman 42* Liberia
OV-9 tString 0 Cameroon
OVSOD Henkle 0 Liberia
OVL3 Lucius 0 Liberia
OV-11 Bradley 42* Liberia
OV-11 tString 0 Cameroon
MOV-14 Bianco 0 Liberia
Oncho-1 McKerrow 0 Liberia
B-20 Bianco 39* Liberia
B-20 tString 0 Cameroon
RAL-6 Unnasch 0 Cameroon
OVGST Henkle 0 Cameroon
RAL-1 Unnasch 0 Cameroon
EGP GG316 Unnasch 0 Cameroon