Salmonellosis: in retrospect and prospect.

Despite years of study and the accumulation of much potentially relevant information, neither the microbial determinants nor the pathophysiological mechanisms of salmonella-induced enteritis are known with precision. Earlier work is reviewed on the experimental pathology of salmonellosis, the pathophysiology of the disease and the biotyping of salmonella strains in closed rabbit ileal loops. The same strains have been confirmed by us to be (i) invasive and diarrhoeagenic, (ii) invasive and non-diarrhoeagenic, and (iii) non-invasive and non-diarrhoeagenic. At least two mechanisms have been put forward to explain fluid exsorption. One implicates prostaglandins released from polymorphonuclear cells (PMNs) interacting with invading organisms, whereas the second involves salmonella enterotoxin(s). This subject is in a state of confusion and requires clarification. The toxin has been shown by some to bear partial likeness to either cholera toxin (although the evidence is in fact contradictory) or Shiga toxin. Since both 'cholera-like' and 'Shiga-like' toxins are produced by all three biotypes in vitro, production of toxin (of whatever class or subclass) cannot per se be the sole explanation for salmonella-induced fluid secretion. In our experiments the onset of fluid secretion in rabbit ileal loops was coincident with the appearance of large numbers of infiltrating PMNs. We have also shown that organisms from all three biotypes, grown for 6 h in iron-containing but not in iron-deficient media, yielded polymyxin B extracts which are enterotoxic in rabbit ileal loops; culture supernatants were negative. Structural damage occurred to villus tips but not crypts in infected loops, succeeded the onset of fluid secretion, and was not reproduced by polymyxin B enterotoxic extracts. Thus salmonella diarrhoea may be a complex phenomenon with multiple determinants which might include the release of endogenous secretagogues and bacterial enterotoxin(s), if such are shown to be synthesized and released in vivo at appropriate times and in appropriate sites. Structural damage to villus tips leading to shortened villi may also contribute to diarrhoea by altering absorption (tip function)/secretion (crypt function) ratios as well as to the expulsion of those organisms which have not migrated to deeper tissues.