Gastric parietal cell apical membranes must protect the cell from the extremely low pH and wide variations in osmolality of the gastric juice. To characterize the permeability properties of gastric apical membranes, we have measured passive permeabilities to water, protons, NH3, and small nonelectrolytes of membrane vesicles derived from parietal cells of fasted animals and fed animals. Both preparations are known to be highly enriched in H+/K(+)-ATPase, the enzyme responsible for acidifying the gastric contents. The preparations behaved as single populations, and their permeability properties were similar in all respects, permitting pooling of the results. This similarity suggests that insertion of tubulovesicles into the apical membrane does not change the behavior of the lipid bilayer. Osmotic water permeability (Pf) averaged (mean +/- SD) (2.8 +/- 0.3) x 10(-4) cm/s, a value 10-fold lower than that obtained in lecithin large unilamellar vesicles (LUV) and similar to that obtained in other water-tight epithelia. Similarly, ammonia permeability (PNH3) was low [(4.4 +/- 2.3) x 10(-3) cm/s] and 10 times below that of lecithin LUV. By contrast, proton permeability (PH+) was surprisingly high (0.030 +/- 0.011 cm/s) and similar to that of lecithin LUV. These results suggest that the pathway for proton permeation differs from that of water and NH3. Nonelectrolyte permeabilities were strikingly similar to those obtained in another water-tight epithelium, the toad urinary bladder. Moreover, these permeabilities followed Overton's rule in that permeability varied in accordance with the oil-water partition coefficient. We conclude that the gastric apical membrane, like that of several renal epithelia, is relatively water-tight and exhibits low permeabilities to small nonelectrolytes. These properties are likely to be essential to the ability of this membrane to perform its barrier function.

Synaptic plasticity is an important feature of neural networks involved in the encoding of information. In the analysis of long-term potentiation and long-term depression, several examples have emerged in which this plasticity is itself modulated. This higher-order form of plasticity has been referred to as 'metaplasticity', a modification of synapses reflected as a change in the ability to induce or maintain plasticity. These observations raise the question of the possible advantage of regulating the intrinsic plastic properties of a synapse. The neural circuit mediating the siphon withdrawal reflex in Aplysia provides a useful network in which to examine this question directly. Inhibitory synapses in this circuit (from L30 neurons) exhibit a variety of forms of activity-dependent short-term synaptic enhancement which contribute to dynamic gain control in the siphon withdrawal reflex. Here we report that tail shock, an extrinsic modulatory input of known behavioural relevance, induces differential metaplasticity at this synapse, attenuating its ability to exhibit short-term synaptic enhancement after presynaptic activation (augmentation and post-tetanic potentiation), while leaving intact its capacity for enhancement during activation. This attenuation of inhibition at the synaptic level seems to mediate comparable attenuation of inhibitory modulation at both network and behavioural levels.

Renal collecting duct and thick ascending limb, as well as stomach, exhibit strikingly low permeabilities to water and solutes. However, the apical membrane characteristics responsible for these unique permeabilities remain unknown. While the lipid composition of artificial membranes governs membrane permeability, exoplasmic and cytoplasmic leaflets of biological apical membranes exhibit striking asymmetries in lipid composition. This asymmetry, as well as the presence of membrane proteins, may be critical to barrier function. To determine the role of bulk lipid composition in apical membrane barrier function, we compared permeabilities to water (Pf), protons, ammonia, and several small nonelectrolytes of gastric apical membrane vesicles [native gastric vesicles (NGV)] and liposomes prepared from lipids quantitatively extracted from these vesicles [gastric lipid large unilamellar vesicles (LUV)]. Permeabilities were measured on a stopped-flow fluorimeter by monitoring self- or pH-sensitive quenching of entrapped carboxyfluorescein. NGV exhibited low Pf (2.8 +/- 0.3 x 10(-4) cm/s) while gastric lipid LUV Pf averaged 1.2 +/- 0.1 x 10(-3) cm/s, a fourfold increase compared with the value in NGV. Gastric lipid LUV also demonstrated higher permeabilities to protons, ammonia, propylene glycol, butyramide, ethanolamine, and acetamide compared with values in NGV. In contrast, gastric lipid LUV exhibited the same or lower permeabilities to urea, glycerol, and ammonia compared with values in NGV. We conclude that lipid composition alone can reconstitute membrane permeabilities to some, but not all, molecules. These results indicate that bilayer asymmetry may be required for the unique permeability of "water-tight" apical membranes and reveal different barrier mechanisms for water and protons, as opposed to ammonia, urea, and glycerol.