Arylaminobenzoates were examined in rabbit colon mounted in an Ussing chamber. The open-circuit transepithelial voltage (Vte) and resistance (Rte) were measured and the equivalent short-circuit current (Isc = Vte/Rte) was calculated. After serosal (s) and mucosal (m) addition of indomethacin (1 mumol/l) Isc was -71 +/- 11 (n = 118) microA/cm2. Amiloride (0.1 mmol/l, m) inhibited this current and reversed the polarity to +32 +/- 4 (n = 118) microA/cm2. In the presence of amiloride and indomethacin, prostaglandin E2 (1 mumol/l, s), known to induce Cl- secretion, generated an Isc of -143 +/- 8 (n = 92) microA/cm2. The arylaminobenzoate and Cl- channel blocker 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB) reduced Isc reversibly with a half-maximal inhibition (IC50) at approximately 0.35 mmol/l and 0.2 mmol/l for mucosal and serosal application respectively. To test whether the poor effect was caused by mucus covering the luminal surface, dose/response curves of the mucosal effect were repeated after several pretreatments. Acidic pH on the mucosal side reduced IC50 to approximately 0.1 mmol/l. A similar effect was observed after N-acetyl-L-cysteine (m) preincubation. Pretreatment with N-acetyl-L-cysteine (m) and carbachol (s), in order to exhaust mucus secretion, and L-homocysteine (m) were more effective and reduced IC50 to approximately 50 mumol/l. To test whether this effect of NPPB was caused by non-specific effects, the two enantiomers of 5-nitro-2-(+/-1-phenylethylamino)-benzoate were tested of which only the (+) form inhibited the Cl- conductance in the thick ascending limb of the loop of Henle (TAL).(ABSTRACT TRUNCATED AT 250 WORDS)