Chloride and non-selective cation channels in unstimulated trout red blood cells

doi:10.1111/j.1469-7793.1998.213bi.x

Journal List > J Physiol > v.511(Pt 1); Aug 15, 1998

J Physiol. 1998 August 15; 511(Pt 1): 213–224.

doi: 10.1111/j.1469-7793.1998.213bi.x.

PMCID: PMC2231098

Chloride and non-selective cation channels in unstimulated trout red blood cells

Stéphane Egée, Olivier Mignen, Brian J Harvey,^* and Serge Thomas

CNRS, Unité de Recherche en Physiologie Cellulaire, Université de Bretagne Occidentale, Brest, France

^*Cellular Physiology Research Unit, University College Cork, Cork, Ireland

Corresponding author S. Thomas: U.R.P.C., U.F.R. Sciences et Techniques, B.P. 809, 29285 Brest cedex, France. Email: serge.thomas/at/univ-brest.fr

Received October 6, 1997; Accepted May 19, 1998.

This article has been cited by other articles in PMC.

Abstract

The cell-attached and excised inside-out configurations of the patch-clamp technique were used to demonstrate the presence of two different types of ion channels in the membrane of trout red blood cells under isotonic and normoxic conditions, in the absence of hormonal stimulation. The large majority (93 %) of successful membrane seals allowed observation of at least one channel type.
In the cell-attached mode with Ringer solution in the bath and Ringer solution, 145 mm KCl or 145 NaCl in the pipette, a channel of intermediate conductance (15-25 pS at clamped voltage, V_p= 0 mV) was present in 85 % of cells. The single channel activity reversed between 5 and 7 mV positive to the spontaneous membrane potential. A small conductance channel of 5-6 pS and +5 mV reversal potential was also present in 62 % of cells.
After excision into the inside-out configuration (with 145 mm KCl or NaCl, pCa 8 in the bath, 145 mm KCl or NaCl, pCa 3 in the pipette) the intermediate conductance channel was present in 439 out of 452 successful seals. This channel was spontaneously active in 90 % of patches and in the other 10 % of patches the channel was activated by suction. The current-voltage relationship showed slight inward rectification. The channel conductance was in the range 15-20 pS between -60 and 0 mV and increased to 25-30 pS between 0 and 60 mV, with a reversal potential close to zero. Substitution of K⁺ for Na⁺ in the pipette or in the bath did not significantly change the single channel conductance. Dilution of the bathing solution KCl concentration shifted the reversal potential towards the Nernst equilibrium for cations. Substitution of N-methyl-D-glucamine (NMDG) for K⁺ or Na⁺ in the bath almost abolished the outward current whilst the divalent cation Ca²⁺ permeated the channel with a higher permeability than K⁺ and Na⁺. Inhibition of channel openings was obtained with flufenamic acid, quinine, gadolinium or barium. Taken together these data demonstrate that the intermediate conductance channel belongs to a class of non-selective cation (NSC) channels.
In excised patches, under the same control conditions, the conductance of the small conductance non-rectifying channel was 8·6 ± 0·8 pS (n= 12) between -60 and +60 mV and the reversal potential was close to 0 mV. This channel could be blocked by 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB) but not by flufenamic acid, DIDS, barium or gadolinium. Selectivity and substitution experiments made it possible to identify this channel as a non-rectifying small conductance chloride (SCC) channel.

The permeability properties of the red blood cell (RBC) membrane govern its electrolyte and acid/base status and directly influence the transport of oxygen and carbon dioxide within the blood. The organic and inorganic substances influencing the electrolyte and acid/base status diffuse very slowly across the lipid bilayer of the membrane and must therefore follow specific pathways. There are a limited range of transporters by which solutes move across the red blood cell membrane. Among them, ionic channels are present in human and frog RBCs, but very little is known about conductive pathways in erythrocytes in general. The first patch-clamp study of human RBCs was made by Hamill (1981) who showed the appearence of a K⁺ current following inhibition of Ca²⁺ efflux by biochemical treatments previously described by Gardos (1958). Evidence for K⁺ and Cl⁻ channels in the frog erythrocyte was also given by Hamill (1983), indicating two classes of K⁺ channels with different Ca²⁺ sensitivities and a Cl⁻ channel. More recently, an inwardly rectifying K⁺-selective conductance was identified by Virkki & Nikinmaa (1996) in lamprey erythrocytes using the whole-cell patch-clamp technique.

There are no single ion channel data available to date for fish erythrocyte and the possible role of ion channels in regulating erythrocyte cell volume, for example, has not been addressed. The hypothesis of a chloride conductance in trout erythrocytes was suggested from the observation by Fiévet et al. (1995) of a chloride conductance directly related to the trout band-3 protein when this protein is expressed in Xenopus oocyte. In a recent study, we investigated the membrane conductance of trout red blood cell in steady state or immediately after exposure to hyposmotic medium using the nystatin perforated whole-cell recording mode of the patch-clamp technique (Egée et al. 1997). This study demonstrated the existence of a membrane current which was reversibly stimulated by cell swelling, and 50 % of the membrane current was DIDS sensitive and generated by chloride ions. The remaining conductance was voltage independent, showed a reversal potential close to zero and was likely to have been due to a non-selective cation conductance. The present study aimed to give direct evidence for the ion channels underlying whole-cell conductances, and uses the patch-clamp technique for recording single channel activities in the cell-attached or the excised inside-out configuration. This study is limited to observations in unstimulated cells in order to describe the ionic channels present under steady-state physiological conditions and to provide a basis for future investigations on the role of ionic channels in acid-base and volume regulation after hormonal stimulation or osmotic shock. We demonstrate, for the first time, the existence of two channel types in unstimulated normoxic trout red blood cells: a non-selective cation channel present in 85 % of patch-clamped cells presenting an intermediate conductance (25-30 pS) and slight inward rectification; and a small (6-8 pS), voltage-independent chloride channel present in 62 % of tested cells.

METHODS

Preparation of cells

Rainbow trout (Oncorhynchus mykiss) were obtained from a hatchery located in Sizun (Brittany, France) and held indoors in large fibreglass tanks supplied with running tap water for at least 2 weeks. Experiments were performed between October and May, and the tap water temperature varied in the range 10-15°C. Fish were killed by a sharp blow to the head, and blood was obtained by caudal venipuncture using heparinized syringes. The blood cells were washed three times in saline solution and the buffy coat removed by suction. Cells were then incubated overnight at 4°C in the saline solution to ensure a steady state with respect to ion and water content before experimental treatment. The experiments were carried out at 15°C.

Experimental solutions and drugs

The composition of solutions used in patch pipettes and bathing solutions is described in Table 1. The calcium concentration used in the bathing and the pipette solutions was adjusted to pCa 3 in cell-attached configuration, whilst it was adusted to pCa 8 in the bathing solutions in the excised inside-out configuration. All solutions were equilibrated in an air atmosphere and filtered through 0.2 μm Millipore cellulose disks, and had a final osmolarity of 320 mosmol l⁻¹. A set of reservoirs connected to perfusion pipettes was used to test the effects of different solutions on channel activity in excised patches. Solution changes were performed within a few seconds by manual switching between reservoirs. Amiloride, tetraethylammonium acetate (TEA), flufenamic acid, 4,4′-diisothiocyanostilbene-2,2′-disulphonic acid (DIDS), quinine hydrochloride and N-methyl-D-glucamine chloride (NMDG) were purchased from Sigma. 5-Nitro-2-(3-phenylpropylamino)-benzoate (NPPB) was obtained from Research Biochemicals International. Gadolinium chloride hexahydrate was obtained from Aldrich.

Table 1

Composition of the bathing and pipette solutions (mm)

Current recordings

Characterization of the single channel current was performed in the cell-attached and excised inside-out patch configurations. Single channel currents were recorded by the method of Hamill (1983) using a RK400 patch-clamp amplifier (Biologic, Claix, France), filtered at 0.3 or 1 kHz, digitized (48 kHz) and stored on digital audio tape (DTR 1204, Biologic). For analysis the data were played back and transferred to a computer and analysed by the PAT computer program (J. Dempster, Strathclyde Electrophysiology Software, University of Strathclyde, Glasgow, UK). Patch pipettes (tip resistance, 5-8 MΩ) were prepared from borosilicate glass capillaries (Vitrex, Modulohm, Herlev, Denmark), pulled and polished on a programmable puller (DMZ, Werner Zeitz, Augsburg, Germany). Seals of 3-5 GΩ were obtained by suctions of 10-25 mmHg applied for less than 20 s, using a syringe connected to the patch pipette. Under these conditions, the success rate of obtaining gigaohm seals was 93 %. The sign of the clamped voltage (V_p) refers to the pipette solution with respect to the bath and outward current (positive charges flowing across the patch membrane into the pipette) is shown as an upward deflection in the current traces. In the excised configuration the imposed membrane potential (V_m) is referred to as -V_p. Current-voltage (I-V) curves were constructed by plotting the mean current amplitude for each clamped potential. Open probability (P_o) was determined as the fraction of digitized points above a threshold set midway between the closed and open peaks of current amplitude histograms. P_o was determined from 60 s stable recordings. In these conditions, P_o was defined as the ratio of the total time spent in the open state to the total time of the complete record. Analyses were confined to patches containing one channel event histogram. Conventional 50 % threshold analyses yielded distributions of dwell times that were fitted by multiexponential or power functions consistent with multiple open and closed states. Because spontaneous patch excision often occurred, the recording of channel activity in the cell-attached mode was validated by demonstrable change in channel reversal potential and/or incidence upon excision.

Liquid junction potentials

The liquid junction potential (LJP) was defined as the potential of the bath solution with respect to the pipette solution (Barry & Lynch, 1991) and the membrane potential (V_m) was calculated as V_m= -V_p+ LJP where V_p is the reading provided by the patch-clamp amplifier. When bath solutions of different composition were successively applied to the patch membrane, the corresponding changes in liquid junction potentials were corrected using the Henderson equation (JPCalc computer program):

A mathematical equation, expression, or formula that is to be displayed as a block (callout) within the narrative flow. The name of referred object is tjp0511-0213-mu1.jpg

where

and U, C and z represent the mobility, concentration and valency of each ion species (i), and R, T and F are the gas constant, temperature and Faraday constant, respectively. Subscripts ‘b’ and ‘p’ denote bath and pipette solutions, respectively.

Data, expressed as means ±s.e.m., were analysed using standard statistical tests.

RESULTS

Of 553 attempts made to obtain a gigaohm seal, 93 % were successful. The mean duration of a seal was 15-20 min in the cell-attached configuration and 25-35 min in the excised inside-out mode, probably due to the extreme fragility of the red blood cell membrane. Of 514 successful seals, 452 membrane patches showed channel activity; 439 of these exhibited a non-selective cation (NSC) channel of intermediate conductance which was present in cell-attached and/or excised inside-out configurations. A small conductance chloride (SCC) channel was also present in both configurations in 283 patches.

Table 2 summarizes NSC channel conductances between ±60 mV and reversal potentials in cell attached or excised inside-out configurations, with different bathing and pipette-filling solutions.

Table 2

Summary of conductances and reversal potentials

Channel activities in intact cells

Figure. 1A shows an example of the NSC current records obtained with isotonic (320 mosmol l⁻¹), normoxic (oxygen partial pressure, P_O₂= 155 mmHg) and normocapnic (carbon dioxide partial pressure, P_CO₂= 0.3 mmHg) Ringer solution in the bath and KCl solution in the pippette at a range of applied potentials in cell-attached patches. The single channel current-voltage relationship is presented in Fig. 1B (n= 24). Under these conditions the single channel current reversed polarity at the potential (V_rev) of 4.8 ± 1.1 mV and exhibited slight inward rectification. The mean slope conductance between V_rev and -60 mV was 28.9 ± 0.9 pS, and 15.4 ± 0.9 pS between V_rev and +60 mV (n= 24). Replacement of KCl in the pipette by NaCl (n= 9) did not significantly modify these values (Table 2). The leftward shift of the I-V curve observed after substitution of gluconate for Cl⁻ (n= 5) in the pipette (Fig. 1B) (V_rev= 0.0 ± 1.2 mV) can probably be accounted for by the change in the junction potential which, in the case of gluconate, cannot be calculated. An opposite shift in V_rev, however, would be expected in the case of a chloride channel. The channel was also permeable for Ca²⁺ as shown in Table 2 when K⁺ was replaced by 72.5 mm Ca²⁺ (n= 9).

Figure 1

Non-selective cation (NSC) channel in intact cells

Figure 2A shows an example of the presence of the small conductance chloride (SCC) channel superimposed on the NSC channel record. The SCC channel exhibited a linear I-V relationship with a conductance of 5.3 ± 0.7 pS and a calculated reversal potential of 5 mV as shown by the I-V curve presented in Fig. 2B. Due to the small conductance of this channel its activity was tested at V_p=±50 mV and/or ±60 mV.

Figure 2

Small conductance chloride (SCC) channel in intact cells

Two different channels in excised inside-out patches

Following excision from the cell, the small conductance chloride channel and the non-selective cation channel were always present if active in the cell-attached mode (395 out of 439 observations). In the remaining quiesent patches, it was possible (in 44 out of 439 excised patches) to activate NSC channels by imposing a calibrated suction to the pipette using a syringe.

Non-selective cation (NSC) channels in excised inside-out patches

In the excised inside-out configuration with K_int solution in the bath, KCl solution in the pippette, the current-voltage relationship showed a reversal potential (V_rev) of 5.4 ± 1.3 mV and inward rectification (n= 35). The channel slope conductance was 30.0 ± 1.6 pS between -60 mV and V_rev and decreased to 18.0 ± 1.1 pS between V_rev and 60 mV (Fig. 3A). Substitution of NaCl for KCl in the pipette or substitution of Na_int solution for K_int solution in the bath (Fig. 3) did not significantly change the conductance over this voltage range. The cationic versus anionic selectivity was determined by dilution of the bathing solution KCl concentration from 145 mm (K_int solution) to 72.5 mm (K_int_,½ solution). The I-V relationship showed a right shift (Fig. 4A) and V_rev was 17.2 ± 0.7 mV (n= 10, KCl solution in the pipette), close to the Nernst equilibrium potential for cations (17.5 mV). The relative permeability P_cation/P_Cl derived from the Goldman-Hodgkin-Katz relation was 76 ± 12 (n= 10). These results are reinforced by experiments where V_rev shifted to 42.5 ± 3.0 mV (n= 17) when Na⁺ or K⁺ were replaced by NMDG in the bathing solution (Fig. 4B). The apparent P_K/P_NMDG ratio was 8.0 ± 0.5 (n= 15). This channel does not discriminate between Na⁺ and K⁺ (P_K/P_Na= 0.9 ± 0.1, n= 11), and we can therefore denote this channel as a non-selective cation channel. Interestingly the divalent cation Ca²⁺ permeated the channel with a higher permeability than K⁺ or Na⁺ (P_K/P_Ca= 0.6 ± 0.1, n= 11) (Fig. 4B).

Figure 3

Current-voltage relationship corresponding to the data contained in Table 2 of the NSC channel

Figure 4

Current-voltage relationships of substitution or selectivity experiments

Kinetic analysis of the NSC channels in excised inside-out patches

Several active NSC channels were often simultaneously present in the patch membrane. Among the NSC recordings, 62 %, 32 %, 4 % and 2 % exhibited 1, 2, 3, and more than four active channels, respectively. The number of channels simultaneously active was not found to be voltage dependent.Figure 5A shows that P_o was also not a clear function of the membrane voltage. Although NSC channels activity was not affected by pipette suction in cell attached patches, application of suction to quiescent excised inside-out patches produced activation of NSC channels. This effect was irreversible: P_o remained unmodified by repetitive suction pulses (Fig. 5B). Furthermore the activity of NSC channels spontaneously active in inside-out patches was not affected by pipette suction.

Figure 5

Open probability of the NSC channel

The dwell time analysis was performed on patches containing only a single NSC channel opening and the kinetic analysis was made at a holding potential of +30 or -30 mV. Open and closed time distributions were fitted by the sum of two exponentials (n= 21). At V_m= -30 mV, the mean time constants for the closed state (τ_c) were 3.9 ± 0.4 ms and 79.2 ± 9.1 ms and for the open state (τ_o) 6.2 ± 0.8 ms and 74.1 ± 9.4 ms; at V_m=+ 30 mV, the τ_c was 6.3 ± 1.0 ms and 81.3 ± 10.8 ms, and τ_o was 4.4 ± 0.6 ms and 60.7 ± 7.2 ms. The channel did not exhibit ‘run-down’ (time-dependent decrease in P_o) in excised inside-out patches.

Inhibitors or blockers of the NSC channel in excised inside-out patches

Omission of Ca²⁺ from the bath solution had no effect on NSC single channel activity. Amiloride (50 μM), TEA (5 mm) and DIDS (50 μM in the pipette, 100 μM in the bath) failed to modify NSC activity. On the contrary, flufenamic acid induced total inhibition in 50 % of twenty experiments at the concentration of 100 μM as previously described for other NSC channels (Yeh et al. 1995). It has been reported by Gögelein & Capek (1990) that quinine inhibits non-selective cation channels. Quinine added to the bathing solution (n= 10) at 1 mm produced total inhibition of NSC channel activity in 50 % of cases. It is frequently reported that the NSC channels are blocked by gadolinium at concentrations ranging between 10 and 20 μM. Experiments were performed with gadolinium in the patch pipette and the effect was evaluated by comparing the conductance and kinetics obtained with GdCl₃ in the pipette to those obtained in control conditions. Inhibition was always obtained with gadolinium (20 μM), which induced progressive reduction of P_o and number of channels open simultaneously. Total inhibition was reached within 3 min as shown in Fig. 6A. Inhibition was also obtained with barium acetate and was always complete immediately after addition of 5 mm Ba²⁺ (Fig. 6B).

Figure 6

NSC channel inhibition by gadolinium

Small conductance chloride (SCC) channels

In excised patches, under control conditions (e.g. K_int solution in the bath and KCl solution in the pipette) the conductance of the small non-rectifying channel was 8.6 ± 0.8 pS (n= 12) between -60 and +60 mV.Figure 7 shows that in these symmetrical KCl conditions the reversal potential was close to 0 mV. On the contrary, when the bathing solution was low in K⁺ and Cl⁻ (K_int_,½ solution) the reversal potential was shifted towards -10 mV. This is indicative of an anionic over cationic selectivity, which is confirmed by the fact that replacement of KCl by NMDG chloride in the bath did not modify the I-V relationship nor the reversal potential (data not shown). Additional evidence for anion selectivity was provided by experiments in which the the pipette was filled with a potassium gluconate solution. In this case, the outward current was totally abolished (Fig. 7).

Figure 7

Single-channel current-voltage relationships of the SCC channel recorded from excised inside-out patches

The dwell time analysis was performed on patches containing only a single NSC channel opening and the kinetic analysis was made at a holding potential of +60 or -60 mV. At V_m= -60 mV, τ_c was 3.9 ± 1.1 ms (n= 6) and τ_o was 7.8 ± 0.8 ms (n= 6); at V_m=+ 60 mV, τ_c was 2.7 ± 0.2 ms (n= 6) and τ_o was 7.5 ± 1.2 ms (n= 6). The stilbene derivative DIDS (50 μM) had no significant effect on the recorded currents when applied in the pipette on the external side of the membrane. On the contrary, an inhibitory effects of NPPB (50 μM) was detected since the SCC channel was always absent on the current recordings in the presence of this inhibitor of chloride channels. SCC channel activity was not modified by bath addition of flufenamic acid, DIDS, or quinine nor by pipette addition of DIDS or Gd³⁺. The present data are sufficient to denote this channel as a chloride channel.

DISCUSSION

Very little is known about ionic conductances in erythrocytes. The first patch-clamp study of human RBCs was made by Hamill (1981) who showed the appearence of a K⁺ current following inhibition of Ca²⁺ efflux by biochemical treatments previously described by Gardos (1958). The so-called Gardos channel is an intermediate-conductance channel (Hamill, 1981,Hamill, 1983;Gryorczyk et al. 1984;Latorre et al. 1989;Christophersen, 1991) which remains relatively unexplored because of the technical difficulties inherent to patch-clamp analysis in human red blood cells. A voltage-gated, non-selective cation channel involved in the increased transport of cations was also shown in intact human red cells suspended in a depolarizing medium (Christophersen & Benekou, 1991). In frog erythrocytes,Hamill (1983) described specific K⁺ and Cl⁻ ion-selective membrane channels activated during osmotic swelling.

We investigated single ion channels in red blood cells exposed to normoxic and normocapnic conditions in the absence of osmotic or hormonal stimulation. To date there is no evidence for single ion channel activity in fish erythrocytes and, furthermore, the role of ion channels in regulating erythrocyte acid-base and electrolyte homeostasis, in general, has not been addressed. This work is the first direct evidence of single ion channel activity in fish RBCs. In view of the large number of cells used in the present study, it is most likely that the two channel types described are important conductive pathways involved in the maintenance of the RBC ionic homeostasis under physiological conditions.

The non-selective cation channel, present in 83 % of membrane patches, is the most characteristic channel in unstimulated trout RBCs. Its presence was predictable according to our recent work using the nystatin perforated whole-cell recording mode of the patch-clamp technique (Egée et al. 1997), which indicated the presence of a DIDS- and Cl⁻-insensitive, voltage-independent non-selective cation conductance corresponding to 50 % of the total membrane current under isosmotic conditions.

The amplitude of single channel currents and sensitivity to quinine observed for the NSC channel are similar to other non-selective cation channels recorded in various cell types such as rat distal colon (Gögelein & Capek, 1990), rat exocrine pancreatic cells (Sturgess et al. 1987), cultured strial marginal cells from gerbil (Yeh et al. 1995), or outer hair cells of guinea-pig cochlea (Van den Abbeele et al. 1994). The sensitivity to quinine is a common feature also shared with the Gardos K⁺ channels of human RBCs (Reichstein & Rothstein, 1981) and with the lamprey erythrocyte (Gusev et al. 1992); however the Gardos channel is not blocked by flufenamic acid or Gd³⁺.

The NSC channel was shown to conduct Ca²⁺ and calcium permeation is a feature of stretch-activated cation non-selective channels in endothelial cells (Lansman et al. 1987), choroid plexus epithelium (Christensen, 1987), oocytes (Methfessel et al. 1986;Taglietti & Toselli, 1988;Moody & Bosma, 1989;Yang & Sachs, 1989) and renal proximal tubule (Filipovic & Sackin, 1991). Given the large selectivity of the NSC channel for calcium, its role could be to permit permeation of Ca²⁺ ions likely to activate calcium-dependent processes.Christensen (1987) was the first to hypothesize that the non-selective cation channels may act as the primary signal of cell swelling, allowing the influx of Ca²⁺, which in turn activates calcium-dependent K⁺ or Cl⁻ channels effective in volume regulation. In this case the NSC channel would not act as a sensor of cell volume changes but would lead to accumulation of an intracellular chemical signal. This role for non-selective cation channels has been shown for other cell types such as neuroblastoma cells (Falke & Misler, 1989), choroid epithelium (Christensen, 1987) and rat liver cells (Bear, 1990) but the possible importance of this channel for cell volume regulation depends on the magnitude of the change in cytosolic calcium. One problem in studying non-selective cation channels is the relative lack of specific pharmacological agents. Gadolinium may be a specific blocker for stretch-activated channels and the sensitivity to gadolinium reported in the present work is similar to that reported for the stretch-activated channel in oocytes (Yang & Sachs, 1989) and in Necturus renal proximal tubule (Filipovic & Sackin, 1991). Although the NSC channel shares some pharmacological and biophysical characteristics with the family of the stretch-activated cation channels with conductances ranging from 25 to 35 pS and signifiant permeability to divalent cations, it remains difficult to assign it to a class of mechanosensitive channels. Indeed, the kinetics of the stetch-activated channels are usually affected by the external concentration of calcium ions, which is not the case for the NSC channel. In addition, the term ‘stretch-activated channels’ refers to those channels that become more active after the direct application of pressure through a patch-clamp pipette in the cell attached or excised patch configuration (Sackin, 1995;Hamill & McBride, 1995). This is not the case for the NSC channel given the inefficiency of suction to induce any modification in the kinetics of channels spontaneously active in cell-attached membranes. However, an irreversible activation of NSC channels was observed in excised inside-out patches. Once the channel was activated by suction, no further change in P_o could be induced by release of suction or subsequent repetitive application of suction. As reviewed by Hamill & McBride (1997), it has been demonstrated that a channel may display hyper- or hypomechanosensitivity depending upon the mechanical stresses associated with sealing. Therefore we cannot exclude the possibility that sealing may alter channel activity by mechanically stressing the cytoskeleton in the range of suctions that we applied before sealing (between 10 and 25 mmHg). However, even though activation of some NSC channels can be accounted for by the process of sealing itself, the whole-cell recording of a voltage-independent non-selective cation conductance corresponding to 50 % of the total membrane current under isosmotic conditions (Egée et al. 1997) sustains the conclusion that the above-described NSC channels correspond to resting conductances. As the surface of the membrane patches ranges between 2.5 and 3.5 μm², and the calculated mean surface area of the red cells is 300 μm², the number of NSC channels is approximately 100 copies per cell in resting conditions.

The physiological role of the NSC channels is unclear because they allow Na⁺ influx as well as K⁺ efflux. At the membrane potential prevailing in trout RBCs (around -25 mV) both the electrical and chemical gradients are inward and result in a flow of Na⁺ into the RBCs, which would tend to depolarize the membrane potential. If this is the case, then it would place an unnecessary load on the Na⁺-K⁺-ATPase pump, unless Na⁺ is involved in a Na⁺-substrate exchange. However, it is established that the membrane Na⁺-H⁺ exchanger can be activated exclusively by β-adrenergic stimulation. Thus, in the absence of NSC channels, the membrane of RBCs could be considered as totally impermeable to small cations. An entry of cation is necessary for regulating the cell volume in case of shrinkage, and this role is usually played by the Na⁺-H⁺ exchanger in most cell types. The NSC channels are good candidates to play this role in trout RBCs.

The small conductance chloride channel, present in 55 % of membrane patches, was not DIDS inhibitable. Therefore, the membrane chloride conductance previously described in trout RBCs under whole-cell recording and activated by hyposmotic shock (Egée et al. 1997) cannot be accounted for by the activity of the SCC channel. In cell-attached patches, no DIDS-sensitive single channel currents could be detected under isosmotic conditions. It is possible, therefore, that the SCC channel generates the resting Cl⁻ conductance and that the DIDS-sensitive current is a feature of volume regulation.

In steady-state conditions, H⁺ and Cl⁻ distribution across the RBC membrane is purely passive and, in accordance with a Donnan equilibrium, the intraerythrocytic proton concentration results from a physico-chemical equilibrium where Na⁺ and K⁺ movements across the RBC membrane are extremely slow and the Cl⁻ and HCO₃⁻ movements are approximately 1 million times faster. Thus, intracellular pH (pH_i) passively follows the variations of extracellular pH (pH_o). The transmembrane distributions of proton, chloride and bicarbonate are covariant and the ratios [H⁺]_o/[H⁺]_i, [Cl⁻]_i/[Cl⁻]_o and [HCO₃⁻]_i/[HCO₃⁻]_o are similar, which makes it possible to calculate the intracellular pH by measuring intra- and extracellular chloride concentrations and external pH. It is a matter of fact that the intracellular pH measured by an intracellular microelectrode (Guizouarn et al. 1993) or the freeze-thaw technique (Motais et al. 1989) are in agreement with the theory; the physiological values of pH_o= 7.90, pH_i= 7.43, [Cl⁻]_o= 120 mm and [Cl⁻]_i= 45 mm correspond to the ratios [H⁺]_o/[H⁺]_i= 0.34 and [Cl⁻]_i/[Cl⁻]_o= 0.37. Moreover, using the values for extra- and intracellular concentrations for Na⁺, K⁺ and Cl⁻ given by Fiévet et al. (1988) and Motais et al. (1989), it is possible to calculate the reversal potentials E_Na=+69 mV, E_K= -96 mV and E_Cl=-24 mV. The membrane potential measured using intracellular microelectrodes (-25 mV,Guizouarn et al. 1993) is similar to the reversal potential for Cl⁻. Under these conditions the functional significance of the small conductance chloride channel may be to set the RBC resting potential at E_Cl so that there is no electrochemical gradient for conductive anion movement. This voltage clamping of E_Cl may optimize the role of the Cl⁻/HCO₃⁻ exchange mechanism in regulating pH_i.

In conclusion, we have presented the first evidence for single ion channel activity in fish red blood cells. The channels are candidates for setting the resting membrane conductance (small conductance chloride channel) and a pathway for transmembrane Ca²⁺ signalling (non-selective cation channel). The role of these and other ion channels in mediating hormonal and volume regulatory responses is currently under investigation.

References

Barry, PH; Lynch, JW. Liquid junction potentials and small cell effects in patch-clamp analysis. Journal of Membrane Biology. 1991;12:101–117. [PubMed]
Bear, CE. A nonselective cation channel in rat liver cells is activated by membrane stretch. American Journal of Physiology. 1990;258:C421–428. [PubMed]
Christensen, O. Mediation of cell volume regulation by Ca²⁺ influx through stretch-activated channels. Nature. 1987;330:66–68. [PubMed]
Christophersen, P. Ca-activated K channel from human erythrocyte membranes: single channel rectification and selectivity. Journal of Membrane Biology. 1991;119:75–83. [PubMed]
Christophersen, P; Benekou, P. Evidence for a voltage-gated, non-selective cation channel in the human red cell membrane. Biochimica et Biophysica Acta. 1991;1065:103–106. [PubMed]
Egée, S; Harvey, BJ; Thomas, S. Volume-activated DIDS-sensitive whole-cell chloride currents in trout red blood cells. The Journal of Physiology. 1997;504:57–63. [PubMed]
Falke, LC; Misler, S. Activity of ion channels during volume regulation by clonal N1E115 neuroblastoma cells. Proceedings of the National Academy of Sciences of the USA. 1989;86:3919–3923. [PubMed]
Fiévet, B; Claireaux, G; Thomas, S; Motais, R. Adaptative respiratory response of trout to acute hypoxia. III. Ion movements and pH changes in the red blood cell. Respiration Physiology. 1988;74:99–114. 10.1016/0034-5687(88)90144-2. [PubMed]
Fiévet, B; Gabillat, N; Borgese, F; Motais, R. Expression of band 3 anion exchanger induces chloride current and taurine transport: structure-function analysis. EMBO Journal. 1995;14:5158–5169. [PubMed]
Filipovic, D; Sackin, H. A calcium-permeable stretch-activated cation channel in renal proximal tubule. American Journal of Physiology. 1991;260:F119–129. [PubMed]
Gardos, G. The function of calcium in the potassium permeability of human erythrocytes. Biochimica et Biophysica Acta. 1958;30:653–654. 10.1016/0006-3002(58)90124-0. [PubMed]
Gögelein, H; Capek, K. Quinine inhibits chloride and non-selective cation channels in isolated rat distal colon cells. Biochimica et Biophysica Acta. 1990;1027:191–198. [PubMed]
Gryorczyk, R; Schwartz, W; Passow, H. Ca²⁺-activated K⁺ channels in human red cells: Comparison of single-channel currents with ion fluxes. Biophysical Journal. 1984;45:693–698. [PubMed]
Guizouarn, H; Harvey, BJ; Borgese, F; Gabillat, N; Garcia-Romeu, F; Motais, R. Volume-activated Cl⁻-independent and Cl⁻-dependent K⁺ pathways in trout red blood cells. The Journal of Physiology. 1993;462:609–626. [PubMed]
Gusev, GP; Sherstobitov, AO; Skulskii, IA. Potassium transport in Lamprey (Lampreta fluviatilis) erythrocytes: Evidence for K⁺ channels. Comparative Biochemistry and Physiology. 1992;101A:569–573. [PubMed]
Hamill, OP. Potassium channel currents in human red blood cells. The Journal of Physiology. 1981;319:97–98.
Hamill, OP. Potassium and chloride channels in red blood cells. In: Sakmann B, Neher E. , editors. Single Channels Recording. New York: Plenum Press; 1983. pp. 451–471.
Hamill, OP; McBride, DW. Mechanoreceptive membrane channels. American Scientist. 1995;83:30–37.
Hamill, OP; McBride, DW. Induced membrane hypo/hyper-mechanosensitivity: a limitation of patch-clamp recording. Annual Review of Physiology. 1997;59:621–631. 10.1146/annurev.physiol.59.1.621. [PubMed]
Lansman, JB; Hallam, TJ; Rink, TJ. Single stretch-activated ion channels in vascular endothelial cells as mechanotransducers. Nature. 1987;325:811–813. 10.1038/325811a0. [PubMed]
Latorre, R; Oberhauser, A; Labarca, P; Alvarez, O. Varieties of calcium-activated potassium channels. Annual Review of Physiology. 1989;51:385–399. 10.1146/annurev.ph.51.030189.002125. [PubMed]
Methfessel, C; Witzmann, V; Takahashi, T; Mishina, M; Numa, S; Sackmann, B. Patch-clamp measurements on Xenopus laevis oocytes: currents through endogenous channels and implanted acetylcholine and sodium channels. Pflügers Archiv. 1986;407:577–588.
Moody, WJ; Bosma, MM. A non-selective cation channel activated by membrane deformation in oocytes of the ascidian Boltenia villosa. Journal of Membrane Biology. 1989;107:179–188. [PubMed]
Motais, R; Fiévet, B; Garcia-Romeu, F; Thomas, S. Na⁺/H⁺ exchange and pH regulation in red blood cells: Role of uncatalized H₂CO₃ dehydration. American Journal of Physiology. 1989;256:C728–73. [PubMed]
Reichstein, E; Rothstein, A. Effects of quinine on Ca²⁺-induced K⁺ efflux from human red blood cells. Journal of Membrane Biology. 1981;59:57–63. [PubMed]
Sackin, H. Mechanosensitive channels. Annual Review of Physiology. 1995;57:333–353. [PubMed]
Sturgess, NC; Hales, CN; Ashford, MLJ. Calcium and ATP regulate the activity of a non selective cation channel in a rat insulinoma cell line. Pflügers Archiv. 1987;409:607–615.
Taglietti, J; Toselli, M. A study of stretch-activated channels in the membrane of frog oocyte: interaction with Ca²⁺ ions. The Journal of Physiology. 1988;407:311–328. [PubMed]
Van de Abbeele, T; Tran Ba Huy, P; Teulon, J. A calcium-activated nonselective cationic channel in the basolateral membrane of outer hair cells of the guinea pig cochlea. Pflügers Archiv. 1994;417:56–63.
Virkki, LV; Nikinmaa, M. Conductive ion transport across the erythrocyte membrane of lamprey (Lampreta fluviatilis) in isotonic conditions is mainly via an inwardly rectifying K⁺ channel. Comparative Biochemistry and Physiology. 1996;115A:169–176.
Yang, XC; Sachs, F. Block of stretch-activated ion channels in Xenopus oocytes by gadolinium and calcium ions. Science. 1989;243:1068–1071. [PubMed]
Yeh, T; Van den Abbeele, T; Marianovski, R; Herman, P; Tran Ba Huy, P. Luminal non-selective cation and outwardly rectifying chloride channels in cultured strial marginal cells from gerbil. Hearing Research. 1995;90:79–88. [PubMed]