Proton permeability of sarcoplasmic reticulum vesicles.

The proton permeability of rabbit skeletal muscle sarcoplasmic reticulum vesicles was investigated by means of membrane potential measurements. Diffusion potentials were generated in sarcoplasmic reticulum vesicles, rat liver microsomes, and sarcoplasmic reticulum phospholipid vesicles by transferring vesicles from low to high pH. Potentials were measured using the voltage-sensitive fluorescent dye 3,3'-dipentyl-2,2'-oxacarbocyanine. Diffusion potentials were generated readily in sarcoplasmic reticulum vesicles using H+ gradients. Generation of a similar potential in phospholipid vesicles or liver microsomes required the presence of an uncoupler of oxidative phosphorylation. H+ and K+ gradient competition experiments indicated that both K+,Na+-permeable and -impermeable sarcoplasmic reticulum vesicles were permeable to H+. These studies suggested that a mechanism of H+ transport exists in sarcoplasmic reticulum that is independent of the K,Na channel, and that transient or steady state membrane potentials may be effected by the movement of H+.


"
The proton permeability of rabbit skeletal muscle sarcoplasmic reticulum vesicles was investigated by means of membrane potential measurements. Diffusion potentials were generated in sarcoplasmic reticulum vesicles, rat liver microsomes, and sarcoplasmic reticulum phospholipid vesicles by transferring vesicles from low to high pH. potentials were measured using the voltage-sensitive fluorescent dye 3,3'-dipentyl-2,2'oxacarbocyanine. Diffusion potentials were generated readily in sarcoplasmic reticulum vesicles using H' gradients. Generation of a similar potential in phospholipid vesicles or liver microsomes required the presence of an uncoupler of oxidative phosphorylation. H' and K+ gradient competition experiments indicated that both K+,Na+-permeable and -impermeable sarcoplasmic reticulum vesicles were permeable to H'. These studies suggested that a mechanism of H+ transport exists in sarcoplasmic reticulum that is independent of the K,Na channel, and that transient or steady state membrane potentials may be effected by the movement of H'.
The vital role of skeletal muscle sarcoplasmic reticulum (SR)' is the rapid release and uptake of Ca", causing muscle contraction and relaxation (MacLennan and Holland, 1975;Tada et at., 1978). The Ca'+-stimulated ATPase that promotes Ca2+ uptake is thought to be electrogenic, as shown by in vitro reconstitution experiments (Zimniak and Racker, 1978). Ca2+ flux across S R theoretically could create a membrane potential that would impede further Ca2' movement unless a mechanism exists which allows transport of a compensating amount of charge. One possible mechanism could involve the K,Na channel of sarcoplasmic reticulum Meissner, 1977, 1978;Miller, 1978). An alternative mechanism, investigated in the present study, is charge compensation by H' (or OH-).
The proton permeability of SR has been investigated previously to a limited extent. With the use of 5,5-dimethyl-2,4oxazolidinedione Nomura and Nakamaru (1976) found that H' equilibrated across the membrane within 20 min, the time necessary for the separation of SR vesicles by centrifugation.
On the other hand, data obtained using the pH indicator bromthymol blue, has been interpreted to indicate that a H'impermeable SR membrane can form a H' gradient during * This work was supported by Research Grant AM18687 from the United States Public Health Service. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. I The abbreviations used are: SR, sarcoplasmic reticulum; diO-CS-(3), 3,3'-dipentyL2,2'-0xacarbocyanine; p-CF1O-CCP, carbonyl cyanide p-trifluoromethoxyphenyl hydrazone; Pipes, l,4-piperazinediethanesulfonic acid. ATP hydrolysis (Madeira, 1979). It has been speculated that the H' gradient provides the motive force for the active uptake of Ca2+ by SR.
The fluorescence emission of the dye 3,3'-dipentyl-2,2'-0~acarbocyanine, diO-C5-(3) has been shown to decrease in response to a K+-induced membrane potential (negative inside) in erythrocytes, phospholipid vesicles, and sarcoplasmic reticulum vesicles (Sims et al. 1974;McKinley and Meissner, 1978). The present work demonstrates on phospholipid membranes the usefulness of diO-C5-(3) in visualizing H'-induced diffusion potentials. Application of this dye to the study of SR and liver microsomes has indicated that sarcoplasmic reticulum contains a H' pathway which is independent of the K,Na channel and does not appear to exist in liver microsomes.  (Meissner, 1974). A "rough" rat liver microsomal fraction was isolated according to the procedure of Fleischer and Kervina (1974). Small, single-walled phospholipid vesicles were prepared by subjecting aqueous suspensions of sarcoplasmic reticulum phospholipids to sonic radiation as described elsewhere (McKinley and Meissner, 1978).

Reagents
Membrane Potential Measurements-Membrane potentials were generated by establishing ion gradients of permeant ions between the intravesicular cavity and the medium into which the vesicles were diluted. The intravesicular ion composition was formed by fwst equilibrating vesicles for 16 h a t 0°C in a large volume (0.5 to 1.0 mg of SH protein/ml) of 480 mosM K (or Tris)/Pipes, pH 7. Vesicles were sedimented by centrifugation for 30 min a t 35,000 rpm in a Beckman 42.1 rotor and resuspended in incubation medium at a protein concentration of 10 to 20 mg/ml. Vesicles were maintained for 4 h a t 0°C and stored at -65°C before use.
Membrane potentials (negative inside) were visualized with the use of the fluorescent dye 3,3'-dipentyl-2,2'oxacarbocyanine iodide (diO-Cs-(3)) (Sims et al., 1974) as previously described (McKinley and Meissner, 1978). The sign of the membrane potential is reported according to standard convention, i.e. reference (ground) is extravesicular. Fluorescence assays were carried out at 15°C under stirring in a Farrand model 801 fluorometer. Excitation was a t 470 nm and emission was recorded a t 495 nm. Slits used resulted in a half-band width of 2.5 nm. Vesicle concentrations (approximately 15 pg of protein/ml or 15 nmol of phospholipid/ml) were used which produced negligible perturbation of the fluorescence emission during dilution with incubation medium. reticulum (SR) and liver microsomal membranes, sarcoplasmic reticulum phospholipid vesicles plus ion carriers were used in initial experiments as a standard.

Response of di0-C5-(3) Fluorescence to Ion Gradients across Phospholipid Bilayer Membranes--In
The ionophore valinomycin was used to make the bilayer membrane selectively permeable to K' . A large K' gradient was then used to create a negative potential by transferring phospholipid vesicles from high (480 mOSM K/Pipes, pH 7) to low K' concentration (4.8 mosM K/Pipes + 475 mOSM Tris/ Pipes, pH 7). No pronounced change in fluorescence emission was observed unless valinomycin was present (Fig. 1A). The presence of valinomycin allowed the rapid outward movement of K' and subsequent formation of a membrane potential (theoretical Nernst potential = -115 mV). The magnitude of the potential was controlled by varying the K' concentration of the extravesicular medium. In agreement with a previous report (McKinley and Meissner, 1978), fluorescence decreases were nearly a linear function of the logarithm of the extravesicular K' concentration, and hence proportional to the membrane potential formed by dilution (Fig. 2). The fluorescence signal slowly returned to its initial intensity (Fig. lA), suggesting a gradual breakdown of the membrane potential. This breakdown was probably due to the slow diffusion of the relatively impermeable Tris' or Pipes-.
Carbonyl cyanide p-trifluoromethoxyphenylhydrazone (p-CF,O-CCP), an uncoupler of oxidative phosphorylation (Heytler and Prichard, 1962) served as a H' carrier since it allowed selective permeability to H'. Proton gradients were established by diluting phospholipid vesicles from Tris/Pipes medium at pH 6 to Tris/Pipes media of higher pH. In the absence of a H' carrier, a s m d slow decrease in fluorescence emission was followed by a slow return to the base-line (  (Fig, 1B, upper trace). Quenching was lowest for p-CFaO-CCP, and therefore this uncoupler was subsequently used.
Proton gradient-induced membrane potentials produced a smaller dye response than those created by K+ (Fig. 2). The reason for the quantitative difference in dye signal is not clear at present. A pH gradient did not alter the dye response to a K'-induced membrane potential (Fig. 2), suggesting that the difference in dye signal was probably not due to nonspecific  Formation of a positive membrane potential did not affect significantly the fluorescence emission of di0-Cs-(3). Only a minimal signal was seen when K'-induced positive membrane potential changes were initiated by dilution of the vesicles from low K' to high K' medium containing valinomycin (not shown). A positive potential created by a H' gradient (Tris, pH 8 + Tris, pH 6) in the presence of p-CF30-CCP also showed no effect (not shown).
Proton Gradients in Sarcoplasmic Reticulum Vesicles a n d Liver Microsomes-Proton gradients generated transient membrane potential changes in native SR vesicles ( ., K' gradients, both vesicle and dilution media at pH 6, with 0.5 p~ valinomycin added; A, K' gradients, vesicle and dilution media a t pH 6 and pH 8, respectively, with 0.5 p~ valinomycin added; 0, H' gradients with 0.5 ~M~-C F .~O -C C P added. [H'] was directly measured as pH prior to addition of vesicles. branes. The initial fluorescence change was of greater magnitude and faster for SR vesicles (as%, 1 to 2 s) than for liver microsomes (14%, 6 s). The rate of initial fluorescence decrease for liver microsomes at intermediatep-CFsO-CCP concentrations, approached that of SR vesicles withoutp-CF:,O-CCP or phospholipid vesicles withp-CF:,O-CCP, with a slight increase in the magnitude of the signal (Fig. 3B, lower trace). The magnitude and time course of the transient fluorescence response for SR vesicles were virtually unaffected by p-CF30-CCP (Fig. 3A, Lower trace). Thus, SR vesicles appear to have higher H' permeability than liver microsomes or phospholipid vesicles.
Proton permeability of K,Na-permeable (Type 4 a n d -impermeable (Type II) Sarcoplasmic Reticulum Vesicles-Isolated sarcoplasmic reticulum consists of vesicles, some of which contain a K,Na channel (type I) and some that do not (type 11) (McKinley and Meissner, 1978). Isotope exchange and membrane potential measurements indicated that the ratio of type I to type I1 was approximately 2:l. While initial attempts at separating type I from type I1 vesicles have not been successful, the two types of vesicles could be distinguished in the present study by taking advantage of their differing K+,Na+ permeability. Only the K,Na-permeable (type I) vesicle fraction formed a negative membrane potential when transferred from K/Pipes medium at pH 7 to Tris/Pipes medium at pH 7 (Fig. 4, upper trace). The impermeability of K' and Tris' in K,Na-impermeable (type 11) vesicles prevented formation of a potential in about one-third of the vesicles (McKinley and Meissner, 1978). Addition of the K' ionophore valinomycin to the Tris/Pipes medium enabled development of a membrane potential in the entire vesicle population, i.e. type I and I1 vesicles (Fig. 4, Lower trace). Accordingly, the lower trace exhibited a larger deflection than the upper one. A negative membrane potential could be generated exclusively in type I1 vesicles by transferring all SR vesicles from K/Pipes to Na/Pipes medium containing the K+-selective ionophore, valinomycin (Fig. 4, middle trace). Under these conditions, no membrane potential formed in   Fig. 3).
K',Na'-permeable (type I) vesicles since K' diffusion out of the vesicles was rapidly compensated by Na' diffusion into the vesicles. We were then able to test whether the permeability of SR to H' was through the K,Na channel or another structure inherent to SR. Proton permeability of type I and type I1 vesicles was assessed by determining the effect of H' gradients on membrane potentials generated by a large initial K' gradient. For a membrane selectively permeable to K' , the development of a membrane potential may be described by the Nernst equation. Only a small number of K' would traverse the vesicles.' In contrast, establishment of opposing K' and H' gradients in vesicles permeable to both K' and H ' would result in rapid movement of many ions until K' and H' gradients of equal size and direction were formed. In order to determine the optimal conditions for K'-H' exchange, vesicles of varying internal K' concentration and buffer capacity were prepared by partially substituting Tris' for K' (Fig. 5). External K' concentrations were lowered accordingly so that a constant magnitude of the initial K' gradients was maintained. The presence of valinomycin allowed formation of a K' diffusion potential in all SR vesicles (see Fig. 4).
It was found that the internal K' concentration and the direction of the H' gradient both influenced the magnitude of membrane potentials (Fig. 5). Optimal fluorescence signals were obtained by diluting SR vesicles from a high K' medium at pH 6 to a low K' medium at pH 8 (pH 6 + pH 8 in Fig. 5). K' and H' gradients were of equal direction and about the same size so that no significant K'-H+ exchange could be observed, and a high membrane potential was formed over the entire K' concentration range. Optimal conditions for observing K'-H' exchange were obtained by diluting vesicles from high K' medium at pH 8 to low K' medium at pH 6 (pH 8 + pH 6 in Fig. 5). Decreased fluorescence signals indicated lower membrane potentials with loss of a significant portion of intravesicular K' within 1 to 2 s, the experimental limit of ' For a 0.1 p diameter vesicle with a membrane capacitance of 1.0 pfarads/cm2, a 60-fold K+ gradient would induce a potential of 100 mV negative inside by outward diffusion of about 200 ions, decreasing the internal K' concentration by 0.6 mM. A potential of comparable magnitude would be generated when vesicles selectively permeable to H' are transferred from a 480 mOSM Tris/Pipes medium at pH 6.2 to one at pH 8. The outward diffusion of 200 H' would raise the internal pH by about 0.01 unit.

TABLE I Effect of opposing H' and K' gradients in SR vesicles on fluorescence emission of diO-C5-(3) In Experiments 1 to 8, SR vesicles containing 240 mOSM K/Pipes
and 240 mOSM Tris/Pipes at pH 6 or 8 were diluted 100-fold into isoosmolal Tris/Pipes or Na/Pipes media at pH 6 or 8. Na/Pipes dilution media contained 0.5 p~ valinomycin. The external K+ in the dilution media was derived from the vesicle media such that the initial K' gradient was maintained constant (100-fold) in Experiments 1 to 8. In Experiments 9 to 12, SR vesicles containing 480 mOSM Tris/ Pipes at pH 6.2 were diluted 250-fold into isoosmolal Pipes media at pH 8 containing the indicated cation. In Experiment 12, 0.5 pM valinomycin was added to the dilution medium. The pH of the dilution media was not significantly changed by the addition of the sample. Fluorescence decreases were obtained by back-extrapolation to the time of vesicle addition (cf- Fig. 3 detection. External cation concentrations were considered to be essentially unchanged since the dilution medium was an infinite bath compared to the total volume of the vesicles. In the presence of an opposing pH gradient, the extent of K' efflux was dependent on internal K' concentration and buffer capacity. Vesicles that initially contained 480 mOSM K/Pipes generated an appreciable membrane potential, presumably because K'-H' exchange resulted in the formation of H' and K' gradients of equal size and direction without complete loss of intravesicular K'. At an initial internal K' concentration of 240 m o m and a Tris/Pipes buffer concentration of 240 mOSM, fluorescence signals approached zero suggesting that most K' rapidly exchanged for H' so that only a small K' gradient (and H' gradient) existed at equilibrium. The extent of K'-H' exchange was dependent upon the buffering capacity of only those vesicles that were permeable to H'. Since the observed fluorescence signal decreased to near zero at an intravesicular K'/Tris' ratio of 1 or less, it appeared that most of the SR vesicles which were either intrinsically permeable to K' (type I) or rendered permeable to K' with valinomycin (type 11) were permeable to H'. In the absence of a pH gradient (pH 8 -+ pH 8 or pH 6 + pH 6 in Fig. 5), fluorescence signals were of intermediate size. Decreased signals with increasing buffer capacity and decreasing K' concentration within the vesicle showed again that the extent of K'-H' exchange was dependent on intravesicular K' concentration and buffer capacity.
The effect of H' gradients on diffusion potentials generated by K' was studied in type I or type I1 vesicles separately. Type I vesicles were examined in Tris medium in the absence of valinomycin (see Fig. 4), as shown in Experiments 1 through 4 of Table I. The optimal conditions determined in Fig. 5 were used, i.e. vesicles of intermediate K/Pipes concentration (240 mOSM) and buffer capacity were used. Dilution of K+-containing vesicles into Tris' media of equal pH yielded fluorescence decreases of 14% at pH 6 and 16% at pH 8. Dilution from K' medium at pH 6 to Tris' medium at pH 8, caused a significant increase in fluorescence signal (14 or 16 -+ 34%):'' In contrast, an opposing pH gradient (K, pH 8 + Tris, pH 6) essentially eliminated the fluorescence signal (14 or 16 -2%). Thus, in most, if not all, K,Na-permeable (type I) vesicles, t.he establishment of a K' diffusion potential was effected by H' movement. Type I vesicles were therefore permeable to both H' and K'.
The permeability of type I1 vesicles was examined in a similar manner, as shown in Experiments 5 to 8 of Table I. Na' was used as the cation in the dilution media instead of Tris' to prevent formation of a membrane potential in type I vesicles and valinomycin was added to make type I1 vesicles permeable to K' (see Fig. 4). As with type I vesicles, in type I1 vesicles, K' and H' gradients of equal direction increased the fluorescence signal and gradients of opposing direction greatly decreased the fluorescent signal relative to those without a pH gradient. The ability of a H' gradient to affect valinomycin-induced K' diffusion potentials in type I1 vesicles suggested that SR vesicles lacking the K,Na channel were nonetheless permeable to H' .
Additional evidence for the H' permeability of K,Napermeable (type I)and -impermeable (type 11) SR vesicles was provided by investigating the ability of opposing Na' and K' gradients to nullify membrane potentials generated by H' gradients (Experiments 9 to 12 of Table I). When a H'induced membrane potential was generated by diluting vesicles from Tris' medium at pH 6.2 to Tris' medium of pH 8, a fluorescence decrease of 29% was observed (Experiment 9, Table I, see also Fig. 3). The potential was formed by both type I and type I1 vesicles, since Tris' was impermeable to both types of vesicles. Replacement of external Tris' by Na' or K' had no effect on K,Na-impermeable (type 11) vesicles, but type I vesicles did not develop a significant negative membrane potential due to the influx of K' or Na' through the K,Na channel. The reduced fluorescence signal of only 6% (Experiments 10 and 11 of Table I) indicated that most of the total signal observed in Experiment 9 (29%) was due to type I vesicles. Addition of valinomycin rendered type I1 vesicles permeable to K' and essentially eliminated the remaining signal (Experiment 12 of Table I). Therefore, the signal seen in K' and Na' media in the absence of valinomycin was likely due to type I1 vesicles. Fig. 6 shows the dependence of fluorescence signals on the magnitude of membrane potentials generated by H' and K' gradients in type I1 alone or in type I + I1 vesicles together.
Proton-induced membrane potentials of varying size were established in K,Na-permeable and -impermeable vesicles by dilution from Tris/Pipes, pH 6, to Tris/Pipes media at higher pH (ApH, Tris --* Tris in Fig. 6). Replacing K' for Tris' selected for type I1 vesicles (A@, K + K in Fig. 6). Below pH 8, nearly linear relationships were observed between the magnitude of the pH gradient and the magnitude of fluorescence decreases elicited by type I1 vesicles or type I + I1 vesicles.
Above pH 8, variable dye signals and faster rebounds were seen, presumably due to breakdown of the permeability barrier. K' diffusion potentials were generated in type I1 vesicles (AK, K + Nu + ual in Fig. 6) or type I + I1 vesicles (AK, K -Tris + ual in Fig. 6) by diluting potassium gluconate-filed SR vesicles into sodium or Tris/gluconate media that contained valinomycin. Use of the impermeable nonbuffering gluconate anion (McKinley and Meissner, 1978) minimized the intravesicular buffer capacity, and thus, K' for H' exchange. It was found that a 60-fold K' gradient in gluconate media elicited about a 1 %-fold increase in dye signal over that ' Part of this increase was due to type I1 vesicles which could form a H'-induced membrane potential under these dilution conditions. dients across SR vesicle membranes. H' gradients were established by diluting S R vesicles present in 480 mOSM Tris/ or K/Pipes medium at pH 6 into the same media at higher pH. K' gradients were established by dilution of vesicles containing 480 m o m potassium gluconate at pH 7 into isoosmolal solutions of Tris/ or sodium gluconate at pH 7, in the presence of 0.5 PM valinomycin. The magnitude of H' and K' gradients was established by adjusting the pH or K' concentration of the dilution media, respectively. Fluorescence decreases extrapolated back to the time of vesicle addition are indicated. In traces labeled LIK, K + Tris + val, and ApH, Tris -+ Tris, both types I and I1 were expected to generate K' diffusion or H' diffusion potentials, respectively; in A K , K + Na + Val and ApH, K + K only type I1 vesicles were expected to be polarized.

TABLE I1 Effect of H' gradients on K+-induced membrane potentials in liver microsomes
Liver microsomes containing 240 mOSM K/Pipes and 240 mOSM Tris/Pipes at pH 6 or 8 were diluted 125-fold into isoosmolal Tris/ Pipes media at pH 6 or 8. K' from the sample served to establish an initial 125-fold K' gradient throughout the experiments. Dilution media contained 0.5 PM valinomycin (ual)  in K/Pipes media (see Fig. 4). A ratio of about 2 was observed for the fluorescence responses elicited by K' and H' gradients in type I1 vesicles or type I + I1 vesicles. A similar ratio was obtained for phospholipid vesicles rendered permeable to K' with valinomycin or to H' with p-CF30-CCP (see Fig. 2). It was also of interest that the size of fluorescence signals elicited by K' or H' gradients were similar for phospholipid and SR vesicles (type I and type 11). Similar effectiveness of phospholipid and SR vesicles in eliciting a dye signal supported our contention that a large fraction of type I and type I1 vesicles were permeable to H'. Proton Permeability of Liver Microsomes-Since liver microsomes were found to be less permeable to H' than SR vesicles (see "Proton Gradients in Sarcoplasmic Reticulum Vesicles and Liver Microsomes"), H' gradients should be relatively ineffective in modifying K'-induced membrane potentials. To test this hypothesis, K' diffusion potentials were induced by diluting vesicles from high to low K' media in the presence of valin~mycin.~ A relatively high internal buffer capacity was used to maximize the effect of H' gradients on K' diffusion potentials (cfi Fig. 5). As expected for a membrane relatively impermeable to H' , the fluorescence responses, and therefore membrane potentials, were virtually unaffected by pH or pH gradients (Table  11). When liver microsomes were rendered permeable to H' by the H+-carrier p-CFaO-CCP, H' gradients influenced K' diffusion potentials in a manner typical for SR vesicles.

DISCUSSION
This study has shown that K,Na-permeable (type I) andimpermeable (type 11) S R vesicles are both permeable to H', both types of SR vesicles being significantly more permeable to H' than SR phospholipid vesicles or liver microsomes. The proton permeability of K,Na-impermeable SR vesicles indicated that H' movement is likely facilitated in SR by a mechanism independent of the K,Na channel.
The fluorescent dye di0-Cs-(3) has been found useful in investigating ion gradient-induced membrane potential changes. Transfer of phospholipid vesicles from pH 6 to pH 8 media showed a fluorescence signal of low magnitude indicative of only slight H' permeability. An increase in H' permeability, artificially induced by the H' carriersp-CF30-CCP or 2,4-dinitrophenol, resulted in rapidly formed membrane potentials which were a linear function of the pH gradient. The linearity of the fluorescence signal toward the pH gradients validated use of the dye as an indicator of the magnitude of H'-induced membrane potentials in SR vesicles and liver microsomes. The ability of membrane vesicles to generate H' diffusion potentials in response to pH gradients suggested that movement of H' observed here was free and not coupled with other ions.
The H' permeability of SR vesicles and liver microsomes was also investigated by considering the mechanism of net ion flow. The H' permeability of K,Na-permeable (type I) or -impermeable (type 11) SR vesicles was-investigated by taking advantage of their differing K+,Na' permeabilities. Lack of an efficient inhibitor for the K,Na channel and the heterogeneous SR vesicle population complicated these studies. Studies of opposing H' and K' gradients showed reduced fluorescence signals indicating that K+-H' exchange was completed in K,Na-permeable or -impermeable SR vesicles within 1 to 2 S, the experimental limit of detection. Liver microsomes rendered permeable to K+ with the K' ionophore valinomycin required in addition the H' carrier p-CF30-CCP to mediate rapid exchange of K' for H' . Therefore, liver microsomes appeared to lack the putative structure which rendered SR readily permeable to H' . SR vesicles were typically about 0.1 p in diameter (Meissner, 1975) and contained approximately 150 mM K' (Table I), so that average K' and H' fluxes of at least 3 X mol/s were necessary. Since the extracellular H' concentration was about M, the permeability coeffcient for H' would be at least cm/s. The high H' permeability of both K,Na-permeable and -impermeable vesicles suggested that H' moved independent of the K,Na channel. It could not be excluded, however, that the K,Na channel was permeable to H' . In fact, previous membrane potential measurements showed that the K,Na channel is permeable to organic cations which have a cross-section of about 4.5 x 6 A (McKinley and Meissner, 1978).
Proton-induced fluorescence signals were similar for phospholipid vesicles with uncoupler and SR vesicles without uncoupler. In type I + I1 S R vesicles, an opposing H' gradient 'The majority of the liver microsomes were impermeable to K' (unpublished studies). destroyed greater than 90% of the signal seen when K'-H' exchange was minimal. Similarly, most of the signal was lost in type I1 vesicles due to an opposing H' gradient. Thus, a majority of type I and type I1 vesicles appeared to be inherently permeable to H' . Since SR vesicles were about 0.1 pm in diameter there were approximately 50 vesicles/pm' of SR surface area. Assuming a random distribution, the density of H'-permeable pathways in SR in vivo could be estimated to be at least 50/pm2. A possible physiological function of a H'permeable pathway may be that, like the K,Na channel (McKinley and Meissner, 1978;Meissner, 1979), it allows rapid cation movement across the SR membrane to counter Ca2' fluxes during muscle contraction and relaxation. The protonbuffering capacity and K' concentration of the SR cisternae would determine the extent of H' and K' transfer.