The inositol 1,4,5-trisphosphate receptor (InsP3R) is the main calcium(Ca2+) release channel in most tissues. Three isoforms have been identified, but only types I and II InsP3R have been characterized. Here we examine the functional properties of the type III InsP3R because this receptor is restricted to the trigger zone from which Ca2+ waves originate and it has distinctive InsP3-binding properties. We find that type III InsP3R forms Ca2+ channels with single-channel currents that are similar to those of type I InsP3R; however, the open probability of type III InsP3R isoform increases monotonically with increased cytoplasmic Ca2+ concentration, whereas the type I isoform has a bell-shaped dependence on cytoplasmic Ca2+. The properties of type III InsP3R provide positive feedback as Ca2+ is released; the lack of negative feedback allows complete Ca2+ release from intracellular stores. Thus, activation of type III InsP3R in cells that express only this isoform results in a single transient, but global, increase in the concentration of cytosolic Ca2+. The bell-shaped Ca2+-dependence curve of type I InsP3R is ideal for supporting Ca2+ oscillations, whereas the properties of type III InsP3R are better suited to signal initiation.

Intercellular communication among certain cell types can occur via ATP secretion, which leads to stimulation of nucleotide receptors on target cells. In epithelial cells, however, intercellular communication is thought to occur instead via gap junctions. Here [Arg8]vasopressin, phenylephrine, and propidium iodide were from Sigma; fluo-3/AM was from Molecular Probes; suramin was from Biomol (Plymouth Meeting, PA); and indodicarbocyanine (Cy5) was from Biological Detection Systems (Pittsburgh). All other chemicals were of the highest quality commercially available.Preparation of Isolated Hepatocytes. Isolated rat hepatocytes were prepared in the Hepatocyte Isolation Core of the Yale Liver Center as described (5, 12). Briefly, rat livers were perfused with Hanks' A medium and then with Hanks' B medium containing 0.05% collagenase (Boehringer Mannheim) and 0.8 unit trypsin inhibitor (Sigma)/U tryptic activity.Livers were then excised, minced, and passed through serial nylon mesh filters, and the resultant cells were washed. These cells were suspended at a concentration of 7.5 x 105 cells per ml in Leibovitz L-15 medium (GIBCO) containing 10% fetal calf serum, 50 units penicillin, and 50 mg streptomycin/ml, and plated onto glass coverslips. Cells were incubated at 37°C and used 2-6 h after plating. When prepared in this fashion, "40% of isolated hepatocytes were not in contact with other hepatocytes, while '60% were in aggregates of two or more. Cell viability by trypan blue exclusion was measured 2 h after plating and exceeded 90%.Preparation of Isolated Bile Duct Units. Bile duct units, which are polarized and physiologically intact segments of bile duct epithelia, also were prepared in the Hepatocyte Isolation Core of the Yale Liver Center as described (13). After livers were perfused with Hanks' B medium as described above, the portal tissue residue was mechanically separated from parenchymal tissue first by shaking and then by forcing the tissue through a syringe to dissociate the remaining hepatocytes. The tissue was then minced in solution C, which contained a-MEM supplemented -with 0.066% collagenase, 0.033% Pronase, 0.006% DNase, 3% fetal calf serum, 0.1% bovine serum albumin, and penicillin/streptomycin at 100,000 units per 100 mg/liter. The minced tissue was then shaken at 37°C for 30 min, minced again, and sequentially filtered. Fragments remaining on the filters were digested for an additional 30 min Abbreviations: [Ca2+]i, cytosolic calcium; Cy5, indodicarbocyanine.

Hormone-induced increases in cytosolic Ca2+ (Cai2+) begin as Cai2+ waves in cells isolated from most types of tissue (1, 11), but whether such waves actually occur in vivo is unknown. To investigate this, we examined vasopressin-induced Cai2+ signals in hepatocytes within the perfused rat liver. Using confocal fluorescence video microscopy, we found that increases in Cai2+ began as waves that usually originated in hepatocytes near central venules, then spread opposite to the direction of blood flow, to hepatocytes near portal venules. We used immunochemistry to determine that the liver vasopressin V1a receptor is most concentrated among hepatocytes in the pericentral region, providing the mechanism by which Cai2+ waves originate there. Pericentral-to-periportal Cai2+ waves may direct peristaltic flow of bile, since Cai2+ induces contraction of the apical pole of hepatocytes and since peristaltic contractions in liver also occur in a pericentral-to-periportal direction. The organization of Cai2+ waves among cells in intact tissue may be a means by which an integrative, organ-level response is provided in response to hormonal stimuli.

Bile duct epithelia play an important role in the formation and conditioning of bile. However, hormonal responses in this epithelial tissue are incompletely understood. Secretin increases ductular secretion through the intracellular messenger adenosine 3',5'-cyclic monophosphate (cAMP), but whether hormones increase cytosolic Ca2+ (Ca2+(i)) in these cells and whether Ca2+(i) regulates duct secretion is unknown. To address these questions, we examined Ca2+(i) signaling in isolated rat bile duct units using ratio microspectrofluorometry and confocal microscopy. We also used videomicroscopy to examine secretion and cell volume in isolated bile duct cells and duct units. Acetylcholine (ACh) and ATP both increased Ca2+(i) in bile duct units and elicited patterns of Ca2+(i) increases and oscillations that were distinct and dose dependent. In contrast, Ca2+(i) was not increased by the hepatocyte Ca2+(i) agonists vasopressin, angiotensin, and phenylephrine or by the exocrine pancreas agonists cholecystokinin (CCK) and bombesin. In addition, secretin did not increase Ca2+(i) in the isolated bile duct units, whereas ACh did not increase Ca2+(i) in isolated hepatocytes. Mobilization of internal, thapsigargin-sensitive Ca2+ stores contributed more than influx of extracellular Ca2+ to the Ca2+(i) increases induced in the duct units, and ATP-induced increases in Ca2+(i) could be blocked by microinjection of heparin but not de-N-sulfated heparin. ACh transiently decreased bile flow in the isolated perfused rat liver, although neither ACh nor ATP altered secretion in isolated ducts or changed the volume of single isolated bile duct cells. These findings demonstrate that bile duct epithelial cells possess both muscarinic and purinergic receptors that activate Ca2+(i) signaling pathways similar to those seen in other types of epithelia, but that the two types of receptors elicit distinct patterns of Ca2+(i) signals. Increases in Ca2+(i) have minimal direct effects on bile duct secretion, although it remains to be determined whether such signals selectively modulate other aspects of bile duct epithelial cell function.

No abstract