1. The role of prostaglandins in mediating bile salt-induced diarrhoea was investigated with a colonic perfusion technique in vivo in rats either untreated or pretreated with the prostaglandin (PG) synthesis inhibitor, indomethacin.
2. Colonic perfusion with sodium deoxycholate (1 and 2 mmol/l) reduced net water and sodium absorption, whereas at a concentration of 5 mmol/l it caused net fluid secretion. Deoxycholate dose-dependently increased protein and deoxyribonucleic acid (DNA) output into the perfusion fluid and, at a concentration of 5 mmol/l, produced histological evidence of colonic mucosal damage (mucus release, goblet cell depletion, patchy epitheliolysis and inflammatory cell infiltration); histological change was less with deoxycholate at 2 mmol/l and did not occur at 1 mmol/l.
3. Output of immunoreactive prostaglandin E2 (PGE2) into the colonic perfusion fluid rose eight-, 10- and 270-fold after deoxycholate at 1, 2 and 5 mmol/l respectively.
4. Colonic perfusion with added PGE2, in concentrations 10 times lower (2.8 nmol/l) and 10 times higher (0.28 μmol/l) than those found in the perfusate after deoxycholate at 5 mmol/l did not alter mucosal function or structure. However, PGE2 in much higher concentration (0.28 mmol/l) reduced net absorption of water and sodium, increased protein output threefold and, as seen with light microscopy, produced excess surface mucus with minimal goblet cell depletion and no tissue damage.
5. Pretreatment with indomethacin reduced the colonic PGE2 output of rats perfused with deoxycholate at 2 and 5 mmol/l by 56 and 87% respectively, but the bile salt-induced changes in net water and sodium transport and DNA output were not significantly affected. The PG synthesis inhibitor reduced protein loss, goblet cell depletion and surface mucus seen after perfusion with deoxycholate at 2 mmol/l, although it did not prevent the more marked structural changes caused by deoxycholate at 5 mmol/l.
6. These results suggest that in rats prostaglandins (i) are not important mediators of the deoxycholate-induced impairment of colonic water and electrolyte transport, (ii) may contribute to the mucus secretion and goblet cell depletion produced by perfusion with deoxycholate in concentrations below those causing gross tissue damage, and that (iii) overt mucosal damage is not an essential prerequisite for prostaglandin release.