In conclusion, the bacterial toxins are completely unlike the MTV superantigens in primary sequence and structure. The former are soluble globular proteins which do not have to be proteolytically cleaved before they act. The latter are synthesized as type II membrane proteins and may be clipped before they reach the cell surface and act to stimulate T cells. Table III summarizes the similarities and differences between the two sets of superantigens. The most notable quality of these molecules is that both sets of families have developed strategies whereby they bind to Class II and engage V beta. As far as the microorganisms which produce them are concerned, these two properties appear to be essential since they are absolutely conserved over proteins of a number of different structures. Several questions can now be addressed as follows. a. Why do all known superantigens bind to Class II? For the microorganism which produces them, the function of superantigens appears to be T-cell and perhaps directly or indirectly B-cell and macrophage stimulation. Activation of virgin T cells requires engagement with antigen plus MHC on professional antigen-presenting cells. Unlike other cell surface proteins, for example Class I, most Class II in animals is expressed on such cells. Therefore it is likely that superantigens have evolved to engage Class II because presentation to T cells by Class II-bearing cells offers the superantigen the best chance of activating its target T cells. b. Why do superantigens engage TCR V beta and not V alpha or CD3? It is possible that superantigens bind to the V beta portion of the TCR rather than V alpha because the latter does not have a consistently well exposed face for engagement. The fact that it is perhaps relatively easier to produce anti-V beta rather than anti-V alpha antibodies supports this idea. We have shown that N-glycosylation of V beta can interfere with recognition by vSAGs (Pullen et al. 1991), perhaps glycosylation of V alpha tends to conceal otherwise available sites. As far as C beta, C alpha or CD3 engagement is concerned, this may be just too dangerous for MTVs. The role of MTVs SAgs in the life history of the virus seems to be to stimulate T cells in the suckling recipient and thereby create a pool of activated lymphocytes in which the virus may survive until the mouse gives birth and transmits the virus to her own progeny (Hainaut et al. 1990, Golovkina et al. 1992).(ABSTRACT TRUNCATED AT 400 WORDS)