The reaction of the immune system of fish to vaccination

The studies presented in this thesis deal with the effect of bacterial antigens of Yersinia ruckeri and Aeromonashydrophila on the immune system of carp. The antigens were administered by injection or by bath treatment. The effect on the immune system was studied by measuring the numbers of antibody forming cells (AFCs), the level of serum antibody and the processing of antigen in the lymphoid organs. The antigen dose and the formation of immunological memory were taken into account. Moreover, in view of oral vaccination, a study was carried out on the uptake and transport of macromolecules through the intestinal epithelial cells.Y. ruckeri O-antigen was very immunogenic in carp. Both i.m. and i.p. injection of antigen and direct immersion in antigen solution evoked distinct levels of AFCs and serum antibody (Appendix paper I). The height of the response was dependent on entry route; e.g. it was clearly lower after direct immersion. Contrary to the numbers of AFCs, which showed a fast decrease after a peak at day 10, serum antibody persisted for a long time, especially at the highest antigen dose. It was suggested that a long-term presence of antigen in the lymphoid organs, as observed by immunofluorescence, might account for this feature.Secondary antibody responses in carp injected or bathed in Y.ruckeri O-antigen gave indications for the formation of immunological memory on the first contact with the antigen for both priming routes. Highest secondary responses were obtained in animals in which antibody levels resulting from the priming had decreased to background levels (Appendix paper II).Upon i.m. injection of two A.hydrophila bacterins, formalin killed cells (F-A h ) and heat killed and disrupted cells (H-A h ), considerable serum antibody titers were found in carp. H-A h induced higher levels of antibody than F-A h , which in both cases persisted for more than 8 months (Appendix paper III). Whereas antigen dose clearly affected the height of the response, the peak day depended on the type of bacterin (day 14 and day 20 for all doses of H-A h and F-A h respectively). The effect of mixing the antigen with adjuvant found expression in a second increase of AFCs after a "normal" peak in the primary response, and a prolonged increase of serum antibody levels (Appendix paper VIII).Antisera raised against the two A.hydrophila b acterins showed different agglutinating properties (Appendix paper III). Antibody induced by F-A h was mainly directed to lipopolysaccharide (LPS), whereas with H-A h this was not the case. It was concluded that LPS is an important antigen of A. hydrophila. However, there is at least one other immunogenic component.Fractionated immune (anti F-A h ) and non-immune control sera were tested for immunoglobulin (Ig) by an Elisa and for agglutinating properties. It was shown that Ig was restricted to the first high molecular weight protein peak. In addition to the antibody in this peak also a non-Ig agglutinin was present in lower molecular weight fractions, probably a natural agglutinin.Memory formation was monitored by studying secondary responses upon a primary i.m. injection, carp developed memory to A.hydrophila bacterin (Appendix paper VI). The height of the secondary responses was directly correlated with the priming dose. Moreover, the dose of the booster injection determined the outcome of the formed memory; corresponding priming and boosting doses expressed the highest memory levels. The development of maximum memory took some months, and its presence was still demonstrable at 12 months. The low dose priming induced only a weak memory, which was limited in time.A single bath in A.hydrophila bacterin was not followed by the appearance of serum antibody. However, it did induce formation of memory, for a second bath resulted in clear secondary responses. This memory was maximum at 3 months after priming and was limited in time, for it was gone at 12 months. When a booster was given by i.m. injection instead of immersion, the induced memory was not expressed as a clear secondary response. This feature suggests the importance of local processes during the immune response after bath vaccination (Appendix paper V).In Appendix paper VI a brief morphological description is given of the spleen, head kidney and trunk kidney. An antigen localization study was carried out after i.m. injection of A.hydrophila bacterin, by means of immunofluorescence. The antigen first appeared in the splenic ellipsoids and solitary phagocytic cells of the head and trunk kidney. Later on it gradually concentrated in or near melanocentres (MMCs) in all organs studied. In the later phases antigen was only located in and around MMCs, both intracellularly and bound to cell membranes, and it remained so for a full year. No antigen could be traced in the lymphoid organs following bath immunization. The question arose if the observed processes were of immunological interest, and which were just a non-specific reaction. Therefore the fate of injected inert material (carbon) was studied for comparison (Appendix paper VII). The processing of carbon was comparable to that described for A.hydrophila antigen. However, at two points there was a difference: 1) in the spleen carbon was transported to MMCs by cells leaving the ellipsoids, whereas this was not clear for A.hydrophila antigen, which appeared near and in MMCs bound to the surface of cells; 2) carbon was only seen intracellularly. Based on these observations it was assumed that the membrane bound antigen in the MMCs had immunological relevance. To test this assumption simultaneous histological and immunological tests were carried out after injection of A. hydrophila antigen (Appendix paper VIII). In those tests it was found that extracellular bound antigen in MMCs was not observed until circulating antibody was present, which is suggestive for the antigen to be complexed by antibody. A further support for this assumption was given by the simultaneous presence of tissue bound Ig and A.hydrophila antigen in MMCs.The possible immunological importance of MMCs is discussed in Appendix papers VI-VIII. There are several indications for MMCs to be early phylogenetic analogues of the mammalian germinal centres.In Appendix paper IX a study is presented on uptake and transport of intact macromolecules in the intestinal epithelium of carp.This study was started as oral administration is an interesting vaccination method. Electronmicroscopical observations showed that two protein tracer-molecules, horse radish peroxidase (HRP) and ferritin were absorbed more extensively in the second than in the first gut segment. Moreover, HRP and ferritin were processed by the enterocytes in a different way. HRP seemed to be taken up by membrane binding and was subsequently transported to the intercellular space, where it contacted the abundant intra-epithelial leucocytes. There was no notable intracellular digestion of HRP. However, ferritin was taken up by pinocytosis and ended up in lysosome-like bodies or supranuclear vacuoles (2nd segment), where it was digested slowly. Although no ferritin was found in the intercellular space, large macrophages, with phagosomes containing ferritin, were present between the epithelial cells. Also some smaller macrophages containing ferritin, have been observed in the lamina propria and even in the spleen. These data were discussed in terms of both selective and non-selective absorption of macromolecules and the possible immunological implications. It was concluded that orally administered antigens reach intra-epithelial leucocytes, and might induce a (local) immune response. As most antigens are transmitted in the second segment, this part of the gut probably has an important immunological function.FINAL CONCLUSIONS- The humoral immune response in carp can be evoked by soluble and particulate bacterial antigens from Yersinia ruckeri and Aeromonas hydrophila.- The height of the humoral response and the antigen dose are directly correlated.- The height of the humoral response is dependent of the route of antigen administration. Injection gives higher serum antibody levels than bath vaccination.- The peak day of the response is not antigen dose dependent, but is influenced by the bacterin type.- Circulating antibody in these animals can be very persistent (e.g. 1 year). This phenomenon is due to continuous stimulation by antigen in or on macrophages (especially melano-macrophages).- Local processes in the skin and gills play a regulatory role in the response after bath vaccination.- Memory develops after both antigen injection and bathing in antigen solution.- The development of memory in fish takes much more time than the antibody response itself. Maximum levels are usually not reached before 3 months after the initial contact with the antigen.- The height of the secondary responses can be used as an indirect method for the quantification of memory. The data obtained can be used for calculating a memory factor.- The height of memory levels achieved and the priming antigen dose are directly correlated.- The expression of memory is affected by several factors, such as residual "priming" antibody at the moment of boosting, the ratio between the first and second antigen dose, and whether or not the priming and boosting routes correspond.- In the early phase of the response mononuclear phagocytes in the kidney and spleen are responsible for antigen handling and stimulation of immuno-competent cells. In the later phase melano-macrophage centres (MMCs) become more important for the stimulation of the response and for memory induction.- MMCs probably are the early phylogenetic analogous of mammalian germinal centres.- Epithelial cells of the second segment of the intestine are important for the initial antigen uptake. Certain antigens are not digested in the cells, but released in the intercellular space (circulation) and taken up by intestinal macrophages. It is unclear if the intestinal lymphocytes react upon these antigens.- Successful application of bacterial vaccines to fish is possible. Injection methods can be used for small numbers of fish. Bath methods are possible for large scale application or young animals. Oral administration is not as effective as injection or immersion, but recent data provide good prospectives for future use.- The best results in terms of protection can be expected when the vaccination route (e.g. bath) corresponds with the route of pathogen entry (e.g. gills and skin).- In tests for the determination of vaccination efficacy the same challenge route as the vaccination route should be preferred.- It will be an advantage when fish, after a vaccination, can develop immunological memory during a period of 2-3 months at optimal temperatures in a relatively clean environment (e.g. specific pathogen free hatcheries).