Correlated morphological and cytochemical investigations of the branchial epithelium of the pinfish, Lagodon rhomboides, have revealed a cell type that is invariably associated with chloride cells. These cells, termed "accessory cells," have been described previously in the teleost pseudobranch (Dunel and Laurent. J. Microsc. Biol. Cell 16:53-74, 1973) but not in the gill proper. Accessory cells are more numerous in pinfish adapted to seawater than to 33% seawater, and in the former participate with chloride cells in the formation of apical crypts. Although accessory cells are much smaller than chloride cells, they possess numerous mitochondria and display an abbreviated labyrinth of plasma membrane-derived tubules. The labyrinth membranes of accessory cells are essentially unreactive, however, when processed for Na-K-ATPase localization by K-nitrophenylphosphatase cytochemistry, whereas chloride cell membranes exhibit copious, ouabain-sensitive reaction products. The zonulae occludentes between accessory cells and chloride cells also appear to be less extensive than those between either of these cells and the flanking pavement cells. These features suggest that accessory cells represent a population of partially differentiated chloride cells.

The activity of the electrolyte transport enzyme, sodium, potassium-activated adenosine triphosphatase (Na+,K+-ATPase), in the gills of the pinfish, Lagodon rhomboides, increased markedly following transfer of fish from brackish water to seawater. Cytochemical localization of Na+,K+-ATPase via its potassium-dependent phosphatase (K+-NPPase) activity in the branchial epithelium of pinfish adapted to seawater demonstrated that chloride cells are the major sites for the enzyme. Subcellularly, the heaviest depositions of reaction product were observed lining the cytoplasmic membrane surfaces of the labyrinth of anastomosing plasma membrane tubules that ramifies throughout the chloride cell cytoplasm. Enzyme activity was demonstrated also on the cytoplasmic surface of the apical crypt membrane and on the cytoplasmic surfaces of vesicles in the cytoplasm subjacent to the crypt. Deletion of potassium from the cytochemical incubation medium or inclusion of 10 mM ouabain abolished the reaction products associated with these membranes. The significance of these cytochemical results is discussed with reference to current hypotheses of chloride cell function.

Regulation of intracellular free calcium ([Ca2+]i) in single epithelial duct cells of isolated rat and guinea pig pancreatic interlobular ducts by secretin, carbachol and cholecystokinin was studied by microspectrofluorometry using the Ca2(+)-sensitive, fluorescent probe Fura-2. Rat and guinea pig duct cells exhibited mean resting [Ca2+]i of 84 nM and 61 nM, respectively, which increased by 50%-100% in response to carbachol stimulation, thus demonstrating the presence of physiologically responsive cholinergic receptors in pancreatic ducts of both species. The carbachol-induced increase in [Ca2+]i involved both mobilization of Ca2+ from intracellular stores and stimulation of influx of extracellular Ca2+. In contrast, neither cholecystokinin nor secretin showed reproducible or sizeable increases in [Ca2+]i. Both rat and guinea pig duct cells showed considerable resting Ca2+ permeability. Lowering or raising the extracellular [Ca2+]i led, respectively, to a decrease or increase in the resting [Ca2+]i. Application of Mn2+ resulted in a quenching of the fluorescence signal indicating its entry into the cell. The resting Ca2+ and Mn2+ permeability could be blocked by La3+ suggesting that it is mediated by a Ca2+ channel.

pH regulation in isolated guinea pig pancreatic interlobular duct segments loaded with the pH-sensitive fluorophore, 5-(6)-carboxy-SNARF-1-acetoxymethyl ester (SNARF-1), was characterized by laser-scanning confocal microscopy. In HCO3(-)-free medium, intracellular pH (pHi) of duct epithelial cells fell by 0.32 +/- 0.06 pH units in the presence of 0.5 mM amiloride and by 0.36 +/- 0.08 pH units in the absence of Na+. In the presence of extracellular HCO3-, pHi acidified in Na(-)-free medium but not in amiloride-containing medium. Superfusion with Cl(-)-free buffers or with buffers containing 4,4'-diisothiocyanostilbene-2,2'-disulfonic acid produced a cytosolic alkalinization of 0.13-0.22 pH units. These observations demonstrate the presence of Na+/H+ exchange, Na(+)-HCO3- cotransport, and Cl-/HCO3- exchange in guinea pig pancreatic ducts. pHi recovered significantly from an NH4Cl pulse in HCO3(-)-free buffers containing amiloride and carbachol (50.4%) or amiloride and secretin (40.6%). This recovery was blocked by the H(+)-adenosinetriphosphatase (H(+)-ATPase) inhibitor bafilomycin A1 and by preincubation of ducts with nocodazole or cytochalasin D. These observations suggest that a vesicular H(+)-ATPase augments Na(+)-dependent H+ extrusion during agonist-stimulated bicarbonate secretion and that activation of this transport mechanism involves cytoskeletal elements.

Regulation of muscarinic receptors in cultured guinea pig pancreatic acini was investigated by assessing the effects of cholinergic agonists on binding of [N-methyl-3H]scopolamine [( 3H]NMS) and on amylase release. Freshly dispersed acini bound [3H]NMS with a Kd of 74 pM and a maximal binding level (Bmax) of 908 fmol/mg DNA. Carbachol (CCh) stimulated amylase secretion and inhibited [3H]NMS binding. Incubation of acini for 30 min with 0.1 mM CCh decreased the subsequent efficacy of CCh in stimulating amylase release by threefold but had no effect on its potency. In contrast, amylase release in response to cholecystokinin octapeptide (CCK-8) was not altered by CCh preincubation. [3H]NMS binding to acini was decreased only 15-20% after 30-min incubation with CCh. However, culture of acini with 0.1 mM CCh decreased [3H]NMS binding by 50% at 3-4 h and by 85-90% at 24 h. This decrease was attributable primarily to a reduction in Bmax. [3H]NMS binding also was decreased to a similar extent by the cholinergic agonists bethanechol and methacholine but not by other secretagogues. The decrease in antagonist binding induced by CCh was dose dependent, with the IC50, 5.8 microM, approximating the EC50 for amylase release, 4.3 microM. Culture of acini for 24 h with CCh abolished subsequent amylase release in response to CCh but not to CCK-8. When CCh was removed from the culture medium after 24 h and acini recultured in its absence, [3H]NMS binding increased with a half-time for recovery of 20-24 h; this recovery was blocked by cycloheximide.(ABSTRACT TRUNCATED AT 250 WORDS)