Species-specific and common epitopes on the secreted and surface antigens of Toxocara cati and Toxocara canis infective larvae.

It is widely accepted that the major cause of visceral larva migrans (VLM) in man is Toxocara canis infection. This has been largely based on the detection of antibodies to this species. We have compared the antigens of T. canis and Toxocara cati in order to establish whether assay for the former might be compromised by infection with the latter. Comparisons were made by radioiodination of the surface and excretory/secretory (ES) glycoproteins of the infective larvae of both species, immunoprecipitation with poly- and monoclonal reagents, and SDS-PAGE. The SDS-PAGE profiles of surface antigens of the two species showed few similarities, whereas that of the ES material indicated considerable homology. Serum from infected animals and a human VLM patient exhibited complete cross reactivity, although there was evidence in the mouse of a specific response to one of the components of T. cati ES. Testing of ES against a panel of monoclonal antibodies (MoAbs) confirmed the similarity; all but one of the MoAbs recognized several of the components of both sources of ES. The only exception was MoAb Tcn-2, which did not react with T. cati surface, somatic or ES antigens. This antibody is known to recognize a carbohydrate determinant which is widespread on T. canis glycoproteins. This species-specific determinant, therefore, represents a reversal of the consensus that peptide determinants tend to be the more specific. Finally, the MoAbs were used to examine the exposure of shared epitopes on the surface of intact larvae of T. cati. Again, fine differences in binding by anti-carbohydrate monoclonals were observed when the two species of Toxocara were compared, reflecting a distinction in exposure or orientation of surface molecules on these nematodes. Moreover, these epitopes were absent or variably present on the surface of freshly hatched larvae, and full exposure did not occur until about 24 h post-hatching. This delay in the presentation of epitopes might have implications for the process of infection in sensitized hosts. In conclusion, it is probable that the serological response in man to T. canis is, by current serological methods, indistinguishable in specificity from that induced by T. cati infection, and that the MoAb which we describe could be used to permit discrimination.