Flow cytometry and sorting of amphibian bladder endocytic vesicles containing ADH-sensitive water channels

Summary The water permeability of ADH target epithelial cells is believed to be regulated by a cycle of exo-endocytosis of vesicles containing functional water channels. These vesicles were selectively labeled in intact frog urinary bladders with an impermeant fluorescent marker, 6-carboxyfluorescein. Vesicle suspensions containing the labeled endosomes were obtained by homogenization and differential centrifugation of bladder epithelial cells. The osmotic permeability of the endocytic vesicles was measured, using a stopped-flow fluorescence technique, in the absence or in the presence of HgCl₂. This permeability was found very high (500 m/sec) and inhibited by 1 mm HgCl₂ (90%), thus confirming the presence of water channels. The labeled endosomes were then separated from the other membrane vesicles by flow cytometry and sorting. Their protein content was analyzed by electrophoresis on ultrathin polyacrylamide gels. Two double bands were found at 71 and 55 kDa as well as a small band at 43 kDa. They respectively correspond to 31, 38 and 10% of the total amount of silver-stained proteins present in the sorted endosomes, while they only represent 2, 4, and less than 1% of the proteins contained in the vesicle suspension, before sorting. These highly enriched proteins (or at least one of them) are likely to be involved in the mechanism of water transport. Associated to their partial purification by differential centrifugation, the sorting of the endosomes by flow cytometry seems a good way to further characterize the water channel.     

Glutaraldehyde fixation preserves the permeability properties of the ADH-induced water channels

Summary Unidirectional and net water movements were determined, in frog urinary bladders, before and after glutraldehyde fixation. Experiments were performed in three experimental conditions: 1) in nonstimulated preparations, 2) after the action of antidiuretic hormone (ADH) and 3) in nonstimulated preparations to which amphotericin B was incorporated from the luminal bath. As previously observed for net water fluxes, the increase in the unidirectional water movement induced by ADH was well preserved by glutaraldehyde fixation. After correction for the effects of unstirred layers and nonosmotic pathways, the observed correlation between the ADH-induced increases in the osmotic (Pf) and diffusional (Pd) permeability coefficients was not modified by the fixative action (before glutaraldehyde: slope 11.19,r:0.87±0.07;n=12; after glutaraldehyde: slope 10.67,r:0.86±0.04,n=39). In the case of amphotericin B, Pf/Pd=3.08 (r: 0.83±0.08), a value similar to that observed in lipid bilayers or in nonfixed toad urinary bladders. It is concluded that 1) The experimental approach previously employed to study water channels in artificial lipid membranes and in amphibian urinary bladders, can be applied to the glutaraldehyde-fixed frog urinary bladder. 2) Glutaraldehyde fixation does not modify the permeability properties of the ADH-induced water channels. 3) Any contribution of exo-endocytic processes or cell regulatory mechanisms to the observed permeability parameters can probably be excluded. 4) Glutaraldehyde-fixed preparations are a good model to characterize these water pathways.     

An Amiloride-sensitive Na+ conductance in the basolateral membrane of toad urinary bladder

Summary Exposing the apical membrane of toad urinary bladder to the ionophore nystatin lowers its resistance to less than 100 cm². The basolateral membrane can then be studied by means of transepithelial measurements. If the mucosal solution contains more than 5mm Na⁺, and serosal Na⁺ is substituted by K⁺, Cs⁺, or N-methyl-d-glucamine, the basolateral membrane expresses what appears to be a large Na⁺ conductance, passing strong currents out of the cell. This pathway is insensitive to ouabain or vanadate and does not require serosal or mucosal Ca²⁺. In Cl-free SO ₄ ^2– Ringer's solution it is the major conductive pathway in the basolateral membrane even though the serosal side has 60mm K⁺. This pathway can be blocked by serosal amiloride (K _i=13.1 m) or serosal Na⁺ ions (K _i 10 to 20mm). It also conducts Li⁺ and shows a voltage-dependent relaxation with characteristic rates of 10 to 20 rad sec^–1 at 0 mV.     

Diversity of K+ channels in the basolateral membrane of restingnecturus oxyntic cells

Summary Patch-clamp techniques have been applied to characterize the channels in the basolateral membrane of resting (cimetidine-treated, nonacid secreting) oxyntic cells isolated from the gastric mucosa ofNecturus maculosa. In cell-attached patches with pipette solution containing 100mm KCl, four major classes of K⁺ channels can be distinguished on the basis of their kinetic behavior and conductance: (1) 40% of the patches contained either voltage-independent (a) or hyperpolarization-activated (b), inward-rectifying channels with short mean open times (16 msec fora, and 8 msec forb). Some channels showed subconductance levels. The maximal inward conductanceg _max was 31±5 pS (n=13) and the reversal potentialE _rev was atV _p=–34±6 mV (n=9). (2) 10% of the patches contained depolarization-activated and inward-rectifying channels withg _max=40 ±18 pS (n=3) andE _rev was atV _p=–31±5 mV (n=3). With hyperpolarization, the channels open in bursts with rapid flickerings within bursts. Addition of carbachol (1mm) to the bath solution in cell-attached patches increased the open probabilityP _o of these channels. (3) 10% of the patches contained voltage-independent inward-rectifying channels withg _max=21±3 pS (n=4) andE _rev was atV _p=–24±9 mV (n=4). These channels exhibited very high open probability (P _o=0.9) and long mean open time (1.6 sec) at the resting potential. (4) 20% of the patches contained voltage-independent channels with limiting inward conductance of 26±2 pS (n=3) andE _rev atV _p=–33±3 mV (n=3). The channels opened in bursts consisting of sequential activation of multiple channels with very brief mean open times (10 msec). In addition, channels with conductances less than 6 pS were observed in 20% of the patches. In all nine experiments with K⁺ in the pipette solution replaced by Na⁺, unitary currents were outward, and inward currents were observed only for large hyperpolarizing potentials. This indicates that the channels are more selective for K⁺ over Na⁺ and Cl^–. A variety of K⁺ channels contributes to the basolateral K⁺ conductance of resting oxyntic cells.     

Regulation of K+ channels in the basolateral membrane ofNecturus oxyntic cells

Summary Patch-clamp methods were used to study single-channel events in isolated oxyntic cells and gastric glands fromNecturus maculosa. Cell-attached, excised inside-out and outside-out patches from the basolateral membrane frequently contained channels which had conductances of 67±21 pS in 24% of the patches and channels of smaller conductance, 33±6 pS in 56% of the patches. Channels in both classes were highly selective for K⁺ over Na⁺ and Cl^–, and shared linear current-voltage relations. The 67-pS channel was activated by membrane depolarization, whereas the activity of the 33-pS channel was relatively voltage independent. The larger conductance channels were activated by intracellular Ca²⁺ in the range between 5 and 500nm, but unaffected by cAMP. The smaller conductance channels were activated by cAMP, but not Ca²⁺. The presence of K⁺ channels in the basolateral membrane which are regulated by these known second messengers can account for the increase in conductance and the hyperpolarization of the membrane observed upon secretagogue stimulation.     

Voltage dependence of the basolateral membrane conductance in theAmphiuma collecting tubule

Summary The basolateral potassium conductance of cells of most epithelial cells plays an important role in the transcellular sodium transport inasmuch as the large negative equilibrium potential of potassium across this membrane contributes to the electrical driving force for Na⁺ across the apical membrane. In the present study, we have attempted to establish, theI-V curve of the basolateral membrane of theAmphiuma collecting tubule, a membrane shown to be K⁺ selective. TransepithelialI-V curves were obtained in short, isolated perfused collecting tubule segments. The shunt conductance was determined using amiloride to block the apical membrane Na⁺ conductance. In symmetrical solutions, the shuntI-V curve was linear (conductance: 2.2±0.3 mS·cm^–2). Transcellular current was calculated by subtracting the shunt current from the transepithelial current in the absence of amiloride. Using intracellular microelectrodes, it was then possible to measure the basolateral membrane potential simultaneously with the transcellular current. The basolateral conductance was found to be voltage dependent, being activated by hyperpolarization: conductance values at –30 and –80 mV were 3.6±1.0 and 6.6±1.0 mS·cm^–2, respectively. BasolateralI-V curves were thus clearly different from that predicted by the constant field model. These results indicate that the K⁺-selective basolateral conductance of an amphibian collecting tubule shows inward (anomalous) rectification. Considering the electrogenic nature basolateral Na–K-pump, this may account for coupling between pump-generated potential and basolateral K⁺ conductance.     

Rabbit distal colon epithelium: I. Isolation and characterization of basolateral plasma membrane vesicles from surface and crypt cells

Summary A method has been developed for the simultaneous isolation of basolateral plasma membrane vesicles from surface and crypt cells of rabbit distal colon epithelium by sequential use of differential sedimentation, isopycnic centrifugation and Ficoll 400 barrier centrifugation. The protein yield was high (total 0.81 mg/g mucosa) and surface and crypt cell-derived basolateral membrane fractions have been purified 34- and 9-fold with respect to the homogenate. The pattern of marker enzyme enrichments revealed only minor contamination by subcellular organelles. Latency of ouabain-sensitive (Na⁺, K⁺)-ATPase activity prior and after trypsin treatment of membranes indicated a vesicle configuration of sealed right side-out: sealed inside-out: leaky of approximately 211. The presence of sealed vesicles was also evident from the osmotic sensitivity of thed-[1-¹⁴C] mannitol equilibrium space determined with either fraction. Although considerably different in protein profile, surface and crypt basolateral membranes were similar in cholesterol to phospholipid molar ratio and membrane fluidity as determined by steady-state fluorescence polarization.Stopped-flow light scattering experiments revealed a rather low water permeability of the membranes with a permeability coefficient of 6 m/sec at 35°C, which is one order of magnitude lower than reported for small intestinal plasma membranes. Both membrane fractions have been shown to effectively generate outward uphill potassium ion gradients, a process that is energized by ATP and inhibited by the membrane-permeant cardiacglycoside digitoxin. These characteristics are consistent with the activity of a (Na⁺, K⁺) pump operating in inside-out vesicles.     

Osmotic water permeabilities of brush border and basolateral membrane vesicles from rat renal cortex and small intestine

Summary The osmotic water permeabilityP _f of brush border (BBM) and basolateral (BLM) membrane vesicles from rat small intestine and renal cortex was studied by means of stopped-flow spectrophotometry. Scattered light intensity was used to follow vesicular volume changes upon osmotic perturbation with hypertonic mannitol solutions. A theoretical analysis of the relationship of scattered light intensity and vesicular volume justified a simple exponential approximation of the change in scattered light intensity. The rate constants extracted from fits to an exponential function were proportional to the final medium osmolarity as predicted by theory. For intestinal membranes, computer analysis of optical responses fitted well with a single-exponential treatment. For renal membranes a double-exponential treatment was needed, implying two distinct vesicle populations.P _f values for BBM and BLM preparations of small intestine were equal and amount to 60 m/sec. For renal preparations,P _f values amount to 600 m/sec for the fast component, BBM as well as BLM, and to 50 (BBM) and 99 (BLM) m/sec for the slow component. The apparent activation energy for water permeation in intestinal membranes was 13.3±0.6 and in renal membranes, 1.0±0.3 kCal/mole, between 25 and 35°C. The mercurial sulfhydryl reagentpCMBS inhibited completely and reversibly the highP _f value in renal brush border preparations. These observations suggest that in intestinal membranes water moves through the lipid matrix but that in renal plasma membranes water channels may be involved. From the highP _f values of renal membrane vesicles a transcellular water permeability for proximal tubules can be calculated which amounts to 1 cm/sec. This value allows for an entirely transcellular route for water flow during volume reabsorption.     

Ionic channels in the plasma membrane ofSchizosaccharomyces pombe: Evidence from patch-clamp measurements

Patch-clamp studies of the yeastSchizosaccharomyces pombe reveal that the plasma membrane contains a voltage-gated channel mildly selective for potassium over sodium, lithium, and chloride. The channel exhibits several conductances with a maximum of 153 pS. The channel gates in the region of physiologically relevant voltages, being closed at hyperpolarizing and open at depolarizing voltages. It is not inhibited by tetraethylammonium, quinine, or quinidine applied from the cytoplasmic side of the membrane; similarly, ATP and stretch have no effect. The frequency of its occurrence in patches implies that about 35 channels of this kind are present in the plasma membrane of a single cell.     

Nature of the water channels in the internodal cells ofNitellopsis

Summary The hydraulic resistance was measured on internodal cells ofNitellopsis obtusa using the method of transcellular osmosis. The hydraulic resistance was approximately 2.65 pm^–1 sec Pa, which corresponds to an osmotic permeability of 101.75 m sec^–1 (at 20°C).p-Chloromercuriphenyl sulfonic acid (pCMPS) (0.1–1mm, 60 min) reversibly increases the hydraulic resistance in a concentration-dependent manner.pCMPS does not have any effect on the cellular osmotic pressure.pCMPS increases the activation energy of water movement from 16.84 to 32.64 kJ mol^–1, indicating that it inhibits water movement by modifying a low resistance pathway.pCMPS specifically increases the hydraulic resistance to exosmosis, but does not influence endosmosis. By contrast, nonyltriethylammonium (C₉), a blocking agent of K⁺ channels, increases the hydraulic resistance to endosmosis, but does not affect that to exosmosis. These data support the hypothesis that water moves through membrane proteins in characean internodal cells and further that the polarity of water movement may be a consequence of the differential gating of membrane proteins on the endo- and exoosmotic ends.     

Active and passive Na+ fluxes across the basolateral membrane of rabbit urinary bladder

Summary The apical membrane of rabbit urinary bladder can be functionally removed by application of nystatin at high concentration if the mucosal surface of the tissue is bathed in a saline which mimics intracellular ion concentrations. Under these conditions, the tissue is as far as the movement of univalent ions no more than a sheet of basolateral membrane with some tight junctional membrane in parallel. In this manner the Na⁺ concentration at the inner surface of the basolateral membrane can be varied by altering the concentration in the mucosal bulk solution. When this was done both mucosal-to-serosal²²Na flux and net change in basolateral current were measured. The flux and the current could be further divided into the components of each that were either blocked by ouabain or insensitive to ouabain. Ouabain-insensitive mucosal-to-serosal Na⁺ flux was a linear function of mucosal Na⁺ concentration. Ouabain-sensitive Na⁺ flux and ouabain-sensitive, Na⁺-induced current both display a saturating relationship which cannot be accounted for by the presence of unstirred layers. If the interaction of Na⁺ with the basolateral transport process is assumed to involve the interaction of some number of Na⁺ ions,n, with a maximal flux,M _max, then the data can be fit by assuming 3.2 equivalent sites for interaction and a value forM _max of 287.8pm cm^–2 sec^–1 with an intracellular Na concentration of 2.0mm Na⁺ at half-maximal saturation. By comparing these values with the ouabain-sensitive, Na⁺-induced current, we calculate a Na⁺ to K⁺ coupling ratio of 1.40±0.07 for the transport process.     

Electrophysiology ofNecturus urinary bladder: II. Time-dependent current-voltage relations of the basolateral membranes

Summary As reported previously (S.R. Thomas et al.,J. Membrane Biol. 73:157–175, 1983) the current-voltage (I–V) relations of the Na-entry step across the apical membrane of short-circuitedNecturus urinary bladder in the presence of varying mucosal Na concentrations are (i) time-independent between 20–90 msec and (ii) conform to the Goldman-Hodgkin-Katz constant field flux equation for a single cation over a wide range of voltages.In contrast, theI–V relations of the basolateral membrane under these conditions are (i) essentially linear between the steady-state, short-circuited condition and the reversal potential (E ^s ); and (ii) are decidedly time-dependent withE ^s increasing and the slope conductance,E ^s , decreasing between 20 and 90 msec after displacing the transepithelial electrical potential difference. Evidence is presented that this time-dependence cannot be attributed entirely to the electrical capacitance of the tissue.The values ofg ^s determined at 20 msec are linear functions of the short-circuit current,I _sc, confirming the relations reported previously, which were obtained using a more indirect approach.The values ofE ^s determined at 20 msec are significantly lower than any reasonable estimate of the electromotive force for K across the basolateral membrane, indicating that this barrier possesses a significant conductance to other ions which may exceed that to K. In addition, these values increase linearly with decreasingI _sc and approach the value of the electrical potential difference across the basolateral membrane observed when Na entry across the apical membrane is blocked with amiloride or when Na is removed from the mucosal solution.A possible explanation for the time-dependence ofE ^s andg ^s is offered and the implications of these findings regarding the interpretation of previous microelectrophysiologic studies of epithelia are discussed.     

Ion channels in the membrane ofChara inflata

Summary Voltage-clamped steps in the electric potential difference (PD) across the membrane in cells of the green alga,Chara inflata, cause voltage- and time-dependent current flows, interpreted to arise from opening and closing of various types of ion channel in the membrane. With cells in the light, these channels are normally closed, and the resting PD is probably determined by the operation of an H⁺ efflux pump. Positive steps in PD from the resting level often caused the opening of K⁺ channels with sigmoid kinetics. The channels began to show opening when the PD–120 mV for an external concentration of K⁺ of 1.0mm. Return of the PD to the resting level caused closing of the channels with complex kinetics. Various treatments of the cell could cause these K⁺ channels to open, and remain open continuously, with the PD then lying closer to the Nernst PD for K⁺. The K⁺ channels have been identified by the blocking effects of TEA⁺. Another group of channels, probably Cl^– and Ca²⁺ associated with the action potential open when the PD is stepped to values less negative than –50 mV. Negative steps from the resting PD cause the slow opening, with a time course of seconds, of yet another type of channel, probably Cl^–.     

Regulation of the basolateral potassium conductance of theNecturus proximal tubule

Summary Two methods, the measurement of the response of the basolateral membrane potential (V _bl) of proximal tubule cells ofNecturus to step changes in basolateral K⁺ concentration, and cellular cable analysis, were used to assess the changes in basolateral potassium conductance (G _K) caused by a variety of maneuvers. The effects of some of these maneuvers on intracellular K⁺ activity (a _K ⁱ ) were also evaluated using double-barreled ion-selective electrodes. Perfusion with 0mm K⁺ basolateral solution for 15 min followed by 45 min of 1mm K⁺ solution resulted in a fall in basolateral potassium (apparent) transference number (t _K),V _bl anda _K ⁱ . Results of cable analysis showed that total basolateral resistance,R _b , rose. The electrophysiological effects of additional manipulations, known to inhibit net sodium reabsorption across the proximal tubular epithelium ofNecturus, were also investigated. Ouabain caused a fall int _K accompanied by large decreases ina _K ⁱ andV _bl. Lowering luminal sodium caused a fall int _K and a small reduction inV _bl. Selective reduction of peritubular sodium, a maneuver that has been shown to block sodium transport from lumen to peritubular fluid, also resulted in a significant decrease int _K. These results suggest thatG _K varies directly with rate of transport of the sodium pump, irrespective of the mechanism of change in pump turnover.Part of this material has been presented at the 10th International Conference on Biological Membranes (Cohen & Giebisch, 1984).     

Papaverine reduces the sodium permeability of the apical membrane and the Potassium permeability of the basolateral membrane in isolated frog skin

Summary The effect of papaverine, an inhibitor of the phosphodiesterase responsible for breakdown of cAMP, on the transepithelial sodium transport across the isolated frog skin was investigated.Serosal addition of papaverine caused initially an increase in the short-circuit current (SCC), a doubling of the cellular cAMP content and a depolarization of the intracellular potential under SCC conditions (V _scc).The initial increase in the SCC was followed by a pronounced decrease both in the SCC and in the natriferic action of antidiuretic hormone (ADH), but papaverine had no inhibitory effect on the ability of ADH to increase the cellular cAMP content. As SCC declines, no hyperpolarization was observed.The I/V relationship across the apical membrane during the inhibitory phase, revealed that papaverine reduces the sodium permeability of the apical membrane (P _Na ^a )as well as intracellular sodium concentration. These observations and the previously noted effect of papaverine on V _scc indicates that papaverine must have an effect on the cellular Cl or K permeability.The basolateral Na,K,2Cl cotransporter was blocked with bumetanide, which should bring the cellular chloride in equilibrium. Bumetanide had no effect on basal SCC and V _scc. When papaverine was added to skins preincubated with bumetanide, the effect of papaverine on SCC and V _scc was unchanged. Therefore, the depolarization of V _scc, observed during the papaverine induced inhibition of the SCC, must be due to a reduction in the cellular K permeability.In conclusion, it is suggested that papaverine reduces the sodium permeability of the apical membrane and the potassium permeability of the basolateral membrane of the frog skin epithelium.     

Rabbit distal colon epithelium: III. Ca2+-activated K+ channels in basolateral plasma membrane vesicles of surface and crypt cells

Summary In the mammalian distal colon, the surface epithelium is responsible for electrolyte absorption, while the crypts are the site of secretion. This study examines the properties of electrical potential-driven⁸⁶Rb⁺ fluxes through K⁺ channels in basolateral membrane vesicles of surface and crypt cells of the rabbit distal colon epithelium. We show that Ba²⁺-sensitive, Ca²⁺-activated K⁺ channels are present in both surface and crypt cell derived vesicles with half-maximal activation at 5×10^–7 m free Ca²⁺. This suggests an important role of cytoplasmic Ca²⁺ in the regulation of the bidirectional ion fluxes in the colon epithelium.The properties of K⁺ channels in the surface cell membrane fraction differ from those of the channels in the crypt cell derived membranes. The peptide toxin apamin inhibits Ca²⁺-activated K⁺ channels exclusively in surface cell vesicles, while charybdotoxin inhibits predominantely in the crypt cell membrane fraction. Titrations with H⁺ and tetraethylammonium show that both high-and low-sensitive⁸⁶Rb⁺ flux components are present in surface cell vesicles, while the high-sensitive component is absent in the crypt cell membrane fraction. The Ba²⁺-sensitive, Ca²⁺-activated K⁺ channels can be solubilized in CHAPS and reconstituted into phospholipid vesicles. This is an essential step for further characterization of channel properties and for identification of the channel proteins in purification procedures.     

Basolateral membrane potassium conductance of A6 cells

Summary To study the properties of the basolateral membrane conductance of an amphibian epithelial cell line, we have adapted the technique of apical membrane selective permeabilization (Wills, N.K., Lewis, S.A., Eaton, D.C., 1979b, J. Membrane Biol. 45:81–108). Monolayers of A6 cells cultured on permeable supports were exposed to amphotericin B. The apical membrane was effectively permeabilized, while the high electrical resistance of the tight junctions and the ionic selectivity of the basolateral membrane were preserved. Thus the transepithelial current-voltage relation reflected mostly the properties of the basolateral membrane. Under basal conditions, the basolateral membrane conductance was inward rectifying, highly sensitive to barium but not to quinidine. After the induction of cell swelling either by adding chloride to the apical solution or by lowering the osmolarity of the basolateral solution, a large out-ward-rectifying K⁺ conductance was observed, and addition of barium or quinidine to the basolateral side inhibited, respectively, 82.4±1.9% and 90.9±1.0% of the transepithelial current at 0 mV. Barium block was voltage dependent; the half-inhibition constant (K _i) varied from 1499±97 m at 0 mV to 5.7±0.5 m at –120 mV.Cell swelling induces a large quinidine-sensitive K⁺ conductance, changing the inward-rectifying basolateral membrane conductance observed under basal conditions into a conductance with outward-rectifying properties.     

Effect of osmotic gradient on ADH-induced intramembranous particle aggregates in toad bladder

Summary Paired toad urinary bladders were prepared without or with an osmotic gradient (175 mosm) across them, stimulated for 2.5 (n=6), 5 (n=6), 30 (n=6) or 60 (n=6) min with ADH (20 mU/ml), and studied by freeze-fracture electron microscopy. Water permeability at these times was assessed in additional bladders (n=6 for each case) after tissue fixation according to the technique of Eggena. After both 60 and 30 min of ADH stimulation, the presence of a gradient compared with the absence of one was associated with fewer aggregates (242±35vs. 382±14 ×235 m^–2 at 60 min,P<0.01; 279±36vs. 470±51 ×235 m^–2 at 30 min,P<0.01) and lower water permeability (8.4±1.1vs. 18.8±1.8g×min^–1×cm^–1 ×mosm ^–1 at60min,P<0.005; 9.2±1.0vs. 22.0±2.1 g ×min^–1×cm^–2×mosm ^–1 at 30 min,P<0.001). In addition, with a gradient both maximum water permeability and maximum aggregate frequency were reached nearly together; a similar correspondence occurred without a gradient. We conclude that in the presence of an osmotic gradient both the ADH-associated aggregates and the water permeability response to ADH are prevented from reaching the higher levels observed in bladders not exposed to a gradient.