Volume‐regulatory Cl‐ channel currents in cultured human epithelial cells.

1. During osmotic swelling, cultured human small intestinal epithelial cells (Intestine 407) exhibited activation of large Cl‐ currents under the patch‐clamp whole‐cell configuration. The volume‐sensitive Cl‐ conductance was independent of intracellular Ca2+ and cyclic AMP. 2. The anion permeability sequence of the current was SCN‐ > I‐ > Br‐ > Cl‐ > F‐ > gluconate‐, corresponding to Eisenman's sequence I. 3. Cl‐ currents were instantaneously activated by command pulses in a range of ‐120 to +45 mV. At potentials more positive than +50 mV the current showed a time‐dependent inactivation. This inactivation was accelerated by increased depolarization. The instantaneous current‐voltage relationship rectified in the outward direction. 4. A stilbene‐derivative Cl‐ channel blocker, 4‐acetamido‐4'‐isothiocyanostilbene (SITS), inhibited the Cl‐ current at micromolar concentrations. SITS facilitated inactivation at positive potentials. Outward currents were more prominently suppressed by SITS than inward currents. The concentrations required for 50% inhibition (IC50) of outward and inward currents were 1.5 and 6 microM, respectively. The outward and inward currents were equally inhibited by a carboxylate analogue Cl‐ channel blocker, 5‐nitro‐2‐(3‐phenylpropylamino)‐benzoate (NPPB) or diphenylamine‐2‐carboxylate (DPC) at higher doses (IC50 = 25 for NPPB or 350 microM for DPC). Inactivation kinetics at large depolarizations was not affected by NPPB or DPC. 5. The Cl‐ current was blocked by an unsaturated fatty acid, arachidonic acid (IC50 = 8 microM). Arachidonic acid was still effective in the presence of inhibitors of lipoxygenase (nordihydroguaiaretic acid, 10 microM), cyclo‐oxygenase (indomethacin, 10 microM) and protein kinase C (polymyxin B, 30 microM). The Cl‐ current was also sensitive to another cis unsaturated fatty acid, oleic acid, which is not a substrate for oxygenases. A trans isomer of oleate, elaidic acid, and a saturated fatty acid, palmitic acid, were ineffective. 6. Single Intestine 407 cells exposed to a hypotonic solution showed a regulatory volume decrease after initial osmotic swelling. The volume regulation was abolished by SITS, NPPB, arachidonate and oleate, but not by elaidate and palmitate. 7. It is concluded that outwardly rectifying Cl‐ channels, which are sensitive to arachidonic acid, are activated upon osmotic swelling and involved in the subsequent cell volume regulation.

outward direction. 4. A stilbene-derivative Clchannel blocker, 4-acetamido-4'-isothiocyanostilbene (SITS), inhibited the Clcurrent at micromolar concentrations. SITS facilitated inactivation at positive potentials. Outward currents were more prominently suppressed by SITS than inward currents. The concentrations required for 50% inhibition (IC50) of outward and inward currents were 1-5 and 6 EM, respectively. The outward and inward currents were equally inhibited by a carboxylate analogue Clchannel blocker, 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB) or diphenylamine-2-carboxylate (DPC) at higher doses (IC50 = 25 for NPPB or 350 /Lm for DPC). Inactivation kinetics at large depolarizations was not affected by NPPB or DPC. 5. The Clcurrent was blocked by an unsaturated fatty acid, arachidonic acid (IC50 = 8 /M). Arachidonic acid was still effective in the presence of inhibitors of lipoxygenase (nordihydroguaiaretic acid, 1OjtM), cyclo-oxygenase (indomethacin, 10/ M) and protein kinase C (polymyxin B, 30 #M). The Clcurrent was also sensitive to another cis unsaturated fatty acid, oleic acid, which is not a substrate for oxygenases. A trans isomer of oleate, elaidic acid, and a saturated fatty acid, palmitic acid, were ineffective. 6. Single Intestine 407 cells exposed to a hypotonic solution showed a regulatory volume decrease after initial osmotic swelling. The volume regulation was abolished by SITS, NPPB, arachidonate and oleate, but not by elaidate and palmitate. 7. It is concluded that outwardly rectifying Cl-channels, which are sensitive to MS 9834 INTRODUCTION Cell volume regulation under hypotonic conditions may be accomplished by separate activation of conductive K+ and Cl-pathways, or by other, usually electroneutral, K+ efflux mechanisms, which allow effluxes of KCl and osmotically obliged water, in a variety of cell species (see Hoffman & Simonsen, 1989;Okada & Hazama, 1989;Grinstein & Foskett, 1990 for recent reviews). In a human intestinal epithelial cell line (Intestine 407), parallel activation of the K+ and Clconductances has been directly observed by both two-microelectrode voltage-clamp and whole-cell patch-clamp studies (Hazama & Okada, 1988). An increase in the cytosolic free Ca2" concentration is responsible for activation of the volume-regulatory K+ channel current in the epithelial cell (Hazama & Okada, 1990a). The properties of volumesensitive Clcurrents have been characterized to some extent (Cahalan & Lewis, 1988;Hazama & Okada, 1988;Hudson & Schultz, 1988;McCann, Li & Welsh, 1989;Worrell, Butt, Cliff & Frizzell, 1989;Estacion, 1991;Yantorno, Carre, Cola-Prados, Krupin & Civan, 1992); however, the mechanisms by which the volume-regulatory Cl-channel is activated have not been directly determined.
Arachidonic acid and its metabolites are known to be important mediators of physiological cellular processes (Irvine, 1982). Recently, it was shown that outwardly rectifying, intermediate-conductance Cl-channels are directly inhibited by arachidonic acid in airway epithelial cells (Anderson & Welsh, 1990;Hwang, Guggino & Guggino, 1990). Lambert (1987) provided the data of cell volume measurements which strongly suggest that the Cltransport pathway responsible for a regulatory volume decrease is sensitive to arachidonate. It is then possible that arachidonic acid directly suppresses the volume-sensitive Clchannel current.
In the present study, the properties of volume-sensitive Clcurrents, including their sensitivity to voltages, Cl-channel blockers and arachidonic acid, were investigated by the whole-cell patch-clamp technique in a small intestinal epithelial cell line (Intestine 407). Preliminary data were presented at The Physiological Society Meeting in Cambridge in July 1991 (Kubo & Okada, 1992).

Cells
A human small intestinal epithelial cell line (Intestine 407), which is known to retain receptors to a variety of intestinal secretagogues (Yada & Okada, 1984;Yada, Oiki, Ueda & Okada, 1989), was cultured in Fischer medium supplemented with 10% newborn calf serum, as described previously (Hazama & Okada, 1988). Suspensions of spherical cells were prepared by detaching from the plastic substrate and culturing with agitation for 10-120 min.
The cells placed in a chamber (0 5 ml) were perfused at a flow rate of around 5 ml/min by gravity 352 VOL UME-REG ULA TOR Y Cl-CURRENT 353 feed from reservoirs or by hydrostatic pressure from syringes. A hypotonic challenge was made by switching the perfusate from an isotonic to hypotonic solution.
Whole-cell current recordings The patch electrodes were fabricated from haematocrit capillaries, as described previously (Kotera, Hashimoto, Ueda & Okada, 1991). Since we had large currents, low-resistance pipettes (around 1 MQ when filled with pipette solutions) were employed to reduce the voltage drop across the residual series resistance. The series conductance (200-600 nS) and capacitance (20-30 pF) were maximally compensated. Voltage-pulses were generated by a computer-loaded pulse generator (Shoshin EM, Type 01-8).
The whole-cell patch-clamp technique (Hamill, Marty, Neher, Sakmann & Sigworth, 1981) was essentially the same as described previously (Hazama & Okada, 1988;Kotera et al. 1991). Wholecell currents were recorded with a patch-clamp amplifier (List EPC7). The currents were converted to digital signals by a pulse-code modulator (Sony PCM-501ES) and stored on tape with a video cassette recorder (Victor HR-085) as well as on diskettes via a personal computer (NEC PC9801DX).

Volume-sensitive chloride currents
Under whole-cell volume-clamp, single Intestine 407 cells equilibrated with KCl solutions (Table 1B) responded to a hypotonic challenge (86% osmolarity) with osmotic cell swelling (with no subsequent volume recovery) as well as with activation of outward currents at the equilibrium potential to Cl-(Ec1 + 50 mV) and of inward currents at the equilibrium potentials to monovalent cations (ENa, K -mV) and K+ (EK -70 mV), as shown in Fig. 1 A. Under this condition, the outward and inward currents would represent mainly K+ and Cl-currents, respectively (Hazama & Okada, 1988).
When cells were equilibrated with K+-free, CsCl or choline chloride solutions (Table 1A and C), activation of large currents at both positive and negative potentials was found to take place in association with cell swelling induced by a hypotonic challenge (83 % osmolarity), as shown in Fig. 1   were often observed at 0 mV (= Eci). Small inward currents at 0 mV were also observed when the extracellular MgSO4 concentration was reduced to 2 mm to attain symmetrical conditions with respect to Mg2+, which is another candidate for the inward current carrier (duplicate experiments). Therefore, it may be inferred that the small currents at Eci were associated with non-equilibrium, electrokinetic phenomena during vast fluxes of water and ions across the membrane or with small deviation of the membrane potential from Ec, presumably due to a junction (Donnan) potential between the pipette solution and the cytosol. . Ramp command voltages from -100 to + 100 mV were applied for 2-4 s after the steady current activation was observed in association with steady osmotic swelling. Clgradients and extracellular anion species were indicated on current traces in A and B, respectively. The ramp clamp protocol allows accurate evaluation of Erev despite the history dependency of the Clcurrent amplitude (as described later). Inset, relation between the reversal potential (Erev) and the external Clconcentration ([Cl-]0). Each symbol represents the mean Erev value of eight observations with the S.E. of the mean (vertical bar). The straight line has a 55 mV/decade slope. profile and the zero-current (reversal) potential were virtually independent of cationic species (Na+, Cs+ and choline+), the current must be carried by Cl-. These observations were further complemented by measuring the reversal potentials (Erev) at three different extracellular choline chloride concentrations with ramp command voltages (from -100 to +100 mV for 24 s) applied upon steady osmotic swelling ( Fig. 2A). The Erev value shifted by + 55 mV per a 10-fold decrease in the external Clconcentration (inset in Fig. 2A). This is close to the theoretical value for a C1-selective channel. Anion selectivity of the channel was examined by the ramped voltage clamp in VOL UME-REG ULA TOR Y CiJ CURRENT hypotonic bathing solutions in which Clions were totally replaced with other anionic species. From these current-voltage curves observed during maximal current activation upon steady osmotic swelling (Fig. 2B), the reversal potentials and slope conductances were evaluated ( Table 2). The relative anion conductance estimated from the slope conductance was SCN-I--Br--Cl-> F-> gluconate-. The permeability sequence obtained from reversal potentials was SCN-> I-> Br-> Cl-> F-> gluconate-. The corresponding normalized permeability coefficients were estimated to be 1P5, 1P3, 1 2, 1D0, 0 6 and 0 1, assuming that currents were solely carried by anions.

Voltage dependence
The voltage dependence of the volume-sensitive Clcurrent was first investigated by applying positive and negative pulses alternately with 20 mV increments between -120 and + 120 mV' (Fig. 3A) at 7 s intervals. Basal whole-cell currents in unstimulated cells under isotonic conditions were very low (< ± 50 pA at + 100 mV) and remained stable (Fig. 3B). After a hypotonic challenge stationary currents were instantaneously activated upon applications of voltage pulses of less than ± 60 mV. The currents exhibited time-dependent inactivation at potentials more positive than + 60 mV, whereas time-dependent activation of the currents appeared at potentials more negative than -60 mV (Fig. 3 C). Inactivation became progressively faster as the potential was clamped to increasingly positive voltages. Figure 4A shows tail currents measured at different voltage levels (-90 to + 105 mV) after full activation of the osmotically induced current at -105 mV. The tail currents at potentials more negative than +45 mV had an instantaneous and steady profile, whereas at more positive potentials the currents rapidly inactivated after instantaneous activation (Fig. 4A). Furthermore, no time-dependent activation occurred at potentials more negative than -60 mV. In contrast, after full inactivation of the currents by large positive pulses (+ 100 mV), the tail currents showed time-dependent activation kinetics at negative potentials ( Fig. 4B) and positive potentials less than +45 mV (Fig. 40).
The magnitude of pulse-induced instantaneous currents was dependent upon the pre-potential level. The more negative the conditioning pulse., the greater the instantaneous current was evoked by a constant positive command potential (Fig.  4D). Such history dependency was observed even 15 s after restoring the holding potential to 0 mV (Fig. 4E). Then, the apparent time-dependent activation observed at < -60 mV with the pulse protocol alternated every 7 s (Fig. 3 C and D) would have in fact represented the release from inactivation induced by previous large positive pulses. Thus, the accurate current-voltage relationship cannot be evaluated by the alternating pulse protocol (Fig. 3A) or the ramp clamp protocol (as applied in Fig. 2) because of the history dependency. Therefore, the instantaneous whole-cell current-voltage relationship was re-evaluated from the tail currents measured at different voltages after full activation by a pre-pulse of -105 mV (as in Fig. 4A). The resultant relationship showed rectification in the outward direction with slope conductances of around 120 and 40 nS at + 90 and -75 mV, respectively (Fig. 5).  Fig. 1A or pCa 9 in Figs lB and C, 2, 3 C, 4 and 5). Also, essentially similar volume-sensitive Clcurrents were observed when cytosolic Ca2+ was strongly chelated with 5 mm BAPTA (Fig. 3D). This is in good agreement with our previous results obtained by dialysing with 10 mm EGTA (Hazama & Okada, 1988).
In the presence of a Ca2+ ionophore (ionomycin, 1 ftM), Clcurrents were activated under isotonic conditions (Fig. 3E). However, the peak currents were much smaller than Clcurrents induced by osmotic swelling, and, moreover, the kinetic pattern of Ca2+-activated currents was quite different from that of volume-sensitive currents. Ionomycin-induced currents exhibited activation upon depolarizations and inactivation upon hyperpolarizations.

Cyclic AMP independence
When cyclic AMP (1-2 mM) was added to the choline chloride pipette solution, the basal currents under isotonic conditions were increased by more than three times (from 33+5 to 98+7pA at +lO0mV and from -23+7 to -101+13 pA at -100 mV, n = 10). In the presence of intracellular cyclic AMP neither outward nor inward currents exhibited time-dependent inactivation or activation, and the current-voltage relationship was almost linear. Upon osmotic swelling activation of large Clcurrents was normally observed with cyclic AMP (five observations).
As shown in Fig. 3F, the magnitude and pattern of Clcurrents activated by osmotic swelling were virtually unaffected by prior administration of an inhibitor of cyclic AMP-dependent protein kinase, H-8 (Hidaka, Inagaki, Kawamoto & Sakai, 1984).

Sensitivity to Clchannel blockers
Volume-sensitive Clcurrents were rapidly suppressed by a stilbene-derivative Clchannel blocker, SITS, at 1-100 /LM. The SITS effect was fairly reversible (about 70 % at 30 EM) after wash-out ( Fig. 6A and B). In the presence of SITS, inactivation of the current was accelerated in both its rate and the threshold voltage level (Fig.  6C). The outward current was more sensitive to SITS than the inward current. The half-maximum inhibition doses (IC50) were 1-5 and 6 /M for the outward and inward currents, respectively (Fig. 7).
A carboxylate analogue Clchannel blocker, NPPB, also dose-dependently inhibited the current (Figs 7 and 8). The NPPB effect was fully reversible. The outward and inward currents were equally affected with an IC50 of 25 #M. The inactivation kinetics was not affected by this drug. Another carboxylate analogue Clchannel blocker, DPC, similarly suppressed the volume-sensitive Clcurrent at high doses (IC50 350 #M) (Fig. 7).

Sensitivity to arachidonic acid
Based on the data of cell volume measurements in Ehrlich ascites tumour cells, Lambert (1987) strongly suggested that the Cltransport pathway responsible for the regulatory volume decrease is sensitive to an unsaturated fatty acid, arachidonic albumin, which is known to bind fatty acids (Spector, Fletcher & Ashbrook, 1969).
In the presence of inhibitors of cyclo-oxygenase and lypoxygenase, indomethacin and NDGA (Rainsford, 1988), arachidonic acid was still effective in inhibiting the volume-sensitive Clcurrents ( Fig. 9 C). Arachidonic acid also blocked the current in cells loaded with an inhibitor of protein kinase C, polymyxin B (Mazzei, Katoh & Kuo, 1982), as shown in Fig. 9D.
Another cis unsaturated fatty acid, oleic acid, which is not a substrate for oxygenases, also blocked the volume-sensitive Clcurrent (Fig. 9E) in a dosedependent manner (Fig. 10). In contrast, a trans isomer of oleate, elaidic acid, and a saturated fatty acid, palmitic acid, did not inhibit the Clcurrent (10-50 /tM, three and four observations).

Involvement in the regulatory volume decrease
Single-cell size measurements showed that Intestine 407 cells exhibit a regulatory volume decrease after transient osmotic swelling upon a hypotonic challenge (Fig.  11 A, 0), as found in the suspended cells by electronic mean cell size measurements (Hazama & Okada, 1988) and in the monolayer cells by Fura-2 concentration measurements (Okada, Hazama & Yuan, 1990). The regulatory volume decrease was impaired by the extracellular application of a Cl-channel blocker (SITS or NPPB). A cis unsaturated fatty acid, arachidonic acid or oleic acid, also inhibited the volume regulation, whereas a trans isomer of oleate, elaidic acid, had no effects (Fig. 1 tB) Fig. 6. Effects of SITS on volume-sensitive Clcurrents. The currents were recorded by applying alternating pulses (O to + 40 mV) in A and B or command pulses (as in Fig. 3A) in C to single Intestine 407 cells equilibrated with the CsCl solutions. Hypotonic challenges were made at the arrow in A, and 7 to 16 min before recordings in C. The zero current level is indicated by the arrowhead (A, B) or horizontal line (C). to block the volume regulation under hypotonic conditions (duplicate observations). Therefore, it appears that the cell volume regulation under hypotonic conditions is abolished by blocking the volume-sensitive Clcurrent in Intestine 407 cells.  Fig. 9. Effects of arachidonic acid (A-D) and oleic acid (E) on volume-sensitive C1currents. The currents were measured by applying alternating pulses (0 to +40 mV) to single Intestine 407 cells equilibrated with CsCl solutions. The hypotonic challenge was made at the arrow. Upon wash-out of arachidonic acid, bovine serum albumin (BSA, 1 mg/ml) was added in B but not in A. In C, NDGA (10 /UM) and indomethacin (10 /SM) were added to the hypotonic perfusate before administration of arachidonic acid. In D, polymyxin B was added to the pipette solution. The zero current level is indicated by the arrowhead. The data represent three to nine similar experiments. (circles) on the inward (open symbols) and outward Clcurrents (filled symbols). The currents were observed at +40 mV in single osmotically swollen Intestine 407 cells equilibrated with the CsCl solutions before and after application of fatty acids. The ordinate represents the percentage of peak currents at maximum inhibition against steady currents before applications of fatty acids. Each symbol represents the mean value of four to eleven experiments with the s.E. of the mean (vertical bar). Since small oil droplets were often found in the bathing solutions containing oleate (but not arachidonate) of > 10 UM, the effective concentrations of oleic acid would be smaller than those indicated.

Properties of volume-sensitive Cl-channel currents
The volume-sensitive Cl-channel in Intestine 407 cells has unique voltagesensitivities. The instantaneous current shows outward rectification (Fig. 5). The Clchannel current could be maintained in the activated state over the physiological range of membrane potentials but was quickly inactivated at large depolarizations (over + 50 mV) (Fig. 4A). Inactivation becomes more rapid with increasing degree of depolarization (Figs 3C and 4A). The Clcurrent is dependent on the previous conditions inasmuch as greater instantaneous currents were evoked by constant command pulses in the cells to which more negative conditioning pulses had been applied (even after interruption for more than 15 s; Fig. 4E).
The anion selectivity determined from reversal potentials (Fig. 2B, Table 2) was SCN-> I-> Br-> Cl-> F-> gluconate-, which corresponds to Eisenman's sequence I (Wright & Diamond, 1977), suggesting that the anion channel contains weak binding sites. This result is in good agreement with the permeability sequence of volume-sensitive anion currents in a colonic epithelial cell line (Worrell et al. 1989).
Cltransport pathways involved in the regulatory volume decrease were reported to be sensitive to a stilbene-derivative Cl-channel blocker (SITS or DIDS) in epithelial MDCK cells (Rothstein & Mack, 1990). In Intestine 407 cells, the volumesensitive Clchannel was found to show a high sensitivity to SITS (Figs 6 and 7). Recently, a carboxylate analogue Clchannel blocker, 9-anthracenecarboxylic acid (9-AC), was found to inhibit Cl-pathways activated upon osmotic swelling in guineapig jejunal enterocytes (MacLeod & Hamilton, 1991 b). Another carboxylate analogue, NPPB, was also found to suppress volume-sensitive Clcurrents in a ciliary epithelial cell line (Yantorno et al. 1992). Consistently, NPPB was effective in inhibiting volume-sensitive Clcurrents in Intestine 407 cells (Figs 7 and 8).
Hypotonic swelling is known to cause an increase in the cytosolic free Ca21 concentration in Intestine 407 cells (Hazama & Okada, 1990 a, b). Intracellular cyclic AMP increases were also observed in association with osmotic swelling in other cell species (Watson, 1990;Baquet, Miejer & Hue, 1991). Therefore, there is a possibility that either or both second messengers are involved in activation of volume-sensitive Cl-channels. In Intestine 407 cells, however, swelling-induced Clcurrents were found to be totally independent of cytosolic Ca2+ (Fig. 3D) and cyclic AMP. Different types of Clcurrents were activated by increases in cytosolic Ca21 (Fig. 3E) and cyclic AMP, as was the case in a colonic epithelial cell line (Cliff & Frizzell, 1990).
At present, the factors responsible for activation of the Cl-channel are not known. Cell membrane expansion itself might not directly activate the Clchannel, because the Clcurrent activation had a time lag of 5-10 s behind the onset of visible osmotic swelling (Fig. 1).

Direct inhibition by arachidonic acid
In recent years a compelling body of evidence has been amassed that an unsaturated fatty acid, arachidonic acid, which makes up a major component of the cell membrane phospholipids (Irvine, 1982), directly or indirectly modulates ion channels. The present study showed that arachidonic acid inhibits the volumesensitive Clcurrents (Figs 9 and 10).
There are possibilities that some oxygenase metabolites of arachidonic acid (Needleman et al. 1986) or protein kinase C activated by arachidonic acid (Nishizuka, 1988) mediate the arachidonate effect. However, even in the presence of inhibitors of lipoxygenase, cyclo-oxygenase and protein kinase C, arachidonic acid still blocked the volume-sensitive Clcurrent ( Fig. 9 C and D). Furthermore, another cis unsaturated fatty acid, oleic acid, which is not a substrate for oxidases and ineffective in generating oxygen free radicals (Chan, Chen & Yu, 1988), also blocked the Clcurrent (Fig. 9E). Taken together, arachidonic acid appears to inhibit directly the volume-sensitive Clchannel in Intestine 407 cells.
The present study showed that volume-sensitive Clcurrents rectify in the outward direction and are not activated by Ca2+ and cyclic AMP. The single channel conductance was not as yet determined in Intestine 407 cells. However, this was reported to be 23 pS in Ehrlich ascites tumour cells (Hudson & Schultz, 1988) and 75 pS in a colonic epithelial cell line (Worrell et al. 1989).
As in the outwardly rectifying Clchannel of airway epithelial cells (Anderson & Welsh, 1990;Hwang et al. 1990), volume-sensitive Cl-channels in human intestinal epithelial cells were sensitive to arachidonic acid. This is in contrast with smallconductance ohmic Clchannel currents in gastric parietal cells, which are activated by arachidonate (Sakai, Okada, Morii & Takeguchi, 1992).
Volume-sensitive Clcurrents exhibit inactivation with time at large depolarizations (Worrell et al. 1989; the present study). Similar voltage-dependent inactivation kinetics was observed in the outwardly rectifying Clchannel current in excised patches (Shoemaker, Frizzell, Dwyer & Farley, 1986;McCann et al. 1989;Tabeharani, Jensen, Riordan & Hanrahan, 1989) and under the whole-cell mode (McCann et al. 1989). In contrast, the Ca2+-activated Clchannel current shows activation kinetics upon depolarization and inactivation kinetics upon hyperpolarization (Evans & Marty, 1986;Taleb et al. 1988;Cliff & Frizzell, 1990;also see Fig. 3E in the present study). Neither marked time-dependent activation nor inactivation was found in the cyclic AMP-activated ohmic Clchannel (Cliff & Frizzell, 1990;. On balance, it appears that the properties of volume-sensitive Cl-channel currents bear resemblance to those of intermediate-conductance, outwardly rectifying Clchannel currents in many aspects. The outwardly rectifying Clchannel recorded in excised membranes is known to be activated by protein kinase A-mediated phosphorylation (Li et al. 1988). However, volume-sensitive whole-cell Clcurrents could be activated even in the presence of an inhibitor of protein kinase A (Fig. 3F). Therefore, there is a possibility that the same channel can be activated by osmotic swelling under the whole-cell mode and by protein kinase A under the excised mode. Further studies by both single-channel and whole-cell recordings are needed to determine the precise relationship between the volume-sensitive Clchannel and the intermediate-conductance, outward rectifier Clchannel.
3M. KUBO AND ad. OKADA Physiological relevance Increases in the Cl-permeability induced by osmotic cell swelling are known to be essential for subsequent volume regulation in a number of cell species (Hoffman & Simonsen, 1989;Okada & Hazama, 1989;Grinstein & Foskett, 1990). In Intestine 407 cells effective blockers of the volume-sensitive Clchannel such as SITS, NPPB, arachidonic acid and oleic acid were found to abolish the regulatory volume decrease (Fig. 11). In contrast, elaidate and palmitate which did not inhibit the Cl-channel were ineffective in inhibiting the volume regulation. Therefore, it appears that the Clchannel is volume-regulatory.
Cell volume regulation would be a prerequisite for the physiological function of intestinal epithelial cells, because their vigorous transport activities produce an osmotic gradient across the cell membrane due to active accumulation of osmotically active solutes within the cells (Okada, 1979). A regulatory volume decrease has actually been observed in isolated villus enterocytes after swelling upon Na'-dependent absorption of organic solutes (MacLeod & Hamilton, 1991 a).
The physiological significance of Clchannel modulation by arachidonic acid is not clear. Arachidonic acid and its metabolites are known to be involved in the pathophysiology of several intestinal diseases including intestinal inflammation and anaphylaxis (Powell, 1991). The arachidonic acid-induced inhibition of volumeregulatory Clchannels might be then at least in part involved in the pathogenesis of these intestinal diseases.