## Abstract

The principal anions transported by colonic epithelium are Cl<latex>$^-$</latex>, HCO<latex>$^-_3$</latex> and organic anions (OA<latex>$^-$</latex>), particularly acetate, butyrate and pyruvate, these last being formed by microbial degradation of carbohydrate. In the normal absorptive rat colon, Cl<latex>$^-$</latex> is transported from lumen to plasma both by the transcellular and paracellular pathways. The transcellular route appears to depend on amiloride-insensitive coupling of Na<latex>$^+$</latex>-Cl<latex>$^-$</latex> at the mucosal (apical) membrane, the Na<latex>$^+$</latex> electrochemical gradient energizing Cl<latex>$^-$</latex> uptake. Intraluminal [HCO<latex>$^-_3$</latex>] rises as Cl<latex>$^-$</latex> is absorbed, and a mucosal Cl<latex>$^-$</latex>-HCO<latex>$^-_3$</latex> exchange carrier has been postulated. In some species (and in distal colon of the rat when sodium-depleted), the putative Na<latex>$^+$</latex>-Cl<latex>$^-$</latex> carrier is absent so that Cl<latex>$^-$</latex> absorption then depends largely on the paracellular electrochemical gradient. Absorption of OA<latex>$^-$</latex> is independent of the transepithelial p.d., is associated with HCO<latex>$^-_3$</latex> secretion and is considerably reduced by acetazolamide. In the absence of Cl<latex>$^-$</latex>, OA<latex>$^-$</latex> supports Na<latex>$^+$</latex> absorption but does not depend on it continuing unchanged when the latter is blocked. Colonic epithelium can become secretory and an example of this state is congenital chloridorrhoea in which an elevated transepithelial p.d. is associated with excessive Cl<latex>$^-$</latex> secretion. Here, it appears that the Na<latex>$^+$</latex>-Cl<latex>$^-$</latex> and Cl<latex>$^-$</latex>-HCO<latex>$^-_3$</latex> carriers are lost and Cl<latex>$^-$</latex> conductance of the mucosal membrane substantially increased. The transepithelial uphill movements of Cl<latex>$^-$</latex> or HCO<latex>$^-_3$</latex> in the absorptive and secretory colon appear to depend on coupling to other ionic flows, and there seems to be no need to postulate active transport of these ions.