Effects of tetrodotoxin and ion replacements on the short-circuit current induced by Escherichia coli heat stable enterotoxin across small intestine of the gerbil (Gerbillus cheesmani)

Enterocyte chloride and water secretion into the small intestine after enterotoxin challenge: unifying hypothesis or intellectual dead end?

Many forms of diarrhoeal disease, particularly so called "secretory" diarrhoeal disease are thought to arise by the active secretion of chloride ion from the enterocytes, creating an osmotic gradient for fluid movement into the small intestinal lumen. This model implies that normally occurring intestinal secretion is catastrophically enhanced by bacterial enterotoxins. This review advocates that neither normal nor abnormal intestinal secretion from the enterocytes occurs and that no competent proof for chloride secretion exists. Prior to 1970, the physiological evidence failed to support the concept of the formation of intestinal juice as a normal intestinal event. The concept was later revived to explain the high rate of fluid entry into the lumen after exposure to cholera toxin. Much evidence has been advanced for the chloride secretion hypothesis, the dominant secretory paradigm after 1974, but is the evidence sufficiently compelling for it to be regarded as proving the chloride secretory model? The evidence falls into four categories and a fifth conjectural argument that proposes that an abnormal chloride ion channel in cystic fibrotic sufferers confers a natural selective advantage by preventing diarrhoeal disease. Secretion is putatively demonstrated by 1) showing that mass transfer of fluid is into the lumen (secretion) and not merely a failure to transport out of the lumen (failed absorption). Support is offered by 2) chloride ion flux measurements in vitro in Ussing chambers and by 3) short-circuit current measurements that are consistent with and purport to show chloride ion movement into the lumen. In addition, 4) pharmacological agents are identified that affect short-circuit current and these are assumed to be anti-secretory, consistent with the biochemical mechanism for secretion, confirmed wherever possible by mouse knock-out models. Finally, the proxy methods used to study water movement such as elevated short-circuit current measurements show these to be absent in cystic fibrotic patients. The enterocyte secretion hypothesis is challenged here on the basis of an examination of the methods used to show secretion, particularly after exposing the small intestine to heat stable enterotoxin (STa) from E. coli. STa is thought to be secretory because fluid entry into the lumen is claimed, enhanced isotopic flux of chloride ion towards the lumen occurs, an increase in short-circuit current is found, preventable by various drugs that are deemed likely to be anti-secretory and also because the short-circuit current changes after STa are not seen in cystic fibrotic patients. Using volume recovery in vivo, STa is found not to be secretory but only anti-absorptive. Hence, other techniques used to show secretion are not fit for that purpose. If STa is identified as secretory and yet no secretion occurs, how reliable is the evidence for other toxins being secretory when these methods are used? This review concludes that chloride ion secretion is unproven. A review of the literature indicates that secretion occurs not because epithelial cells actively pump water but by interdiction of fluid absorption, increased conductivity through tight junctions and an increased hydrostatic driving force through elevated capillary pressure. The exclusive focus on chloride secretion may explain the failure to develop antisecretory drugs over the last three decades.