Proton gradients in renal cortex brush-border membrane vesicles. Demonstration of a rheogenic proton flux with acridine orange.

The fluorescence quenching of acridine orange has been used to study the formation and dissipation of acid interior pH gradients in brush-border membrane vesicles from rabbit renal cortex. Acidic interior pH gradients were produced by 1) outwardly directed gradients of Na+ or K+, and 2) the addition of vesicles equilibrated at pH 6.0 to 7.5 buffer. The rate of pH gradient dissipation was stimulated 6.3-fold by the replacement of tetramethylammonium gluconate by tetramethylammonium chloride. A further increase, of 2-fold, was seen upon the addition of carbonyl cyanide-m-chlorophenylhydrazine, demonstrating the existence of a Cl- conductance pathway. In the presence of valinomycin, the replacement of tetramethylammonium gluconate by K gluconate increased the rate of delta pH dissipation by 11-fold, demonstrating the existence of a conductive pathway for protons. This pathway for protons was also shown by the formation of an acidic interior space by an outwardly directed K gradient in the presence of valinomycin. The parallel conductive pathways for H+ and Cl- may dissipate pH and chloride gradients across the luminal membrane of the proximal tubule.

Brush-border membrane vesicles from renal cortex have been shown to possess an electroneutral Na+/H+ antiporter (1, 2). This system is believed to be responsible for Na' absorption and H' secretion in the proximal tubule (3). In addition, these membranes are known to have a conductance pathway for chloride (4-6). Previous work with these transport systems has relied on either the production of pH changes in the exterior space, or measurement of the flux of radiolabeled ions. Recent work has shown that the production of an acidic intravesicular space w i l l cause the fluorescence of acricine dyes to be quenched (7,8). We have used the fluorescence quenching of acridine orange to monitor pH gradient formation by these vesicles. The present studies examine the generation of pH gradients by the Na'/H+ antiporter and the mechanisms by which protons may be translocated across the BBM.' The results demonstrate the existence of a Na'/H' antiporter, and intrinsic conductance pathways for H+ and C1-in BBM vesicles.

EXPERIMENTAL PROCEDURES
Membrane Preparation-Female New Zealand white rabbits, 1-2 kg in weight, were killed by decapitation. Each kidney was perfused via the renal artery with 35 ml of ice-cold 50 mM sucrose, 0.5 mM EDTA, 10 mM Hepes/Tris, pH 7.5 (HSI). The renal cortical tissue was dissected and homogenized in HSI (60 ml/kidney) with an Omnimixer (Sorval, DuPont Instrument Co.) for 4 min. Brush-border membrane fractions were prepared by precipitation with 12 mM MgSOl (6, [9][10][11]. Subsequent homogenization and harvesting by differential centrifugation were performed at 4 "C in 100 mM sucrose, 10 mM Hepes/Tris, pH 7.5. Final membrane pellets were susnended to 30 mg of protein/ml in the indicated buffers with a I-ml syringe fitted with a 27-gauge needle. Salt solutions were added to make the vesicle suspension 150 mM in the desired salt, and the suspension incubated at room temperature for at least 2 h prior to use. A 7-to 10-fold purification of the brush-border marker enzyme alkaline phosphatase was obtained (6).
The fluorescence quenching of acridine orange was used to monitor changes in the transmembrane ApH (8). In all the experiments, 10 pl of BBM vesicles were added to 2 ml of buffer containing 6 p~ acridine orange at 25 "C. Subsequent changes in the fluorescence were monitored with a Perkin-Elmer MPF-44A spectrofluorometer (excitation, 493 nm; emission, 520 nm). The addition of vesicles under conditions where no ApH is formed causes a small fluorescence quenching (presumably, binding of acridine to membranes). Therefore, all quantitative measurements of fluorescence change are based on the total fluorescence measured after pH gradient dissipation (addition of either 2 p~ nigericin when external K' was present, or 25 mM NaC1). Ionophores, when present at initial time, were added from ethanol stocks to the buffer at the same time as vesicles. Samples were mixed with a single pass of a Biolab cuvette mixer immediately after vesicle addition.
General Methods-Protein concentration was determined by the method of Lowry et al. (12), using bovine serum albumin as a standard. Alkaline phosphatase was determined as previously described (6). The observed rate constants (&) for fluorescent recovery following a pH jump were calculated from plots of log (F, -F,) against time, where F, is the fluorescence at infinite time and F! is the fluorescence at time t.
Materials-Acridine orange was obtained from Eastman Kodak; all chemicals were the highest purity commercially available. Stock solutions of tetramethylammonium gluconate were made by titrating tetramethylammonium hydroxide pentahydrate to pH 7.5 with dgluconic acid lactone.

RESULTS
Fluorescence Quenching by p H Jumps-The addition of BBM vesicles equilibrated at pH 6.0 to 7.5 buffer (pH-jump) caused quenching of acridine orange fluorescence (Fig. 1). This quenching has previously been shown to be caused by the accumulation of dye into an acidic intravesicular space; the resulting decrease in the concentration of dye in the external space causes a decrease in the observed fluorescence (8). As shown in Fig. 1 and Table I, the stability of pH gradients formed by pH jumps was affected by the ionic

Conductance in
Renal BBM Vesicles composition of the media. The replacement of either external TMA by K' (Fig. 1, truce B ) or internal gluconate by C1- (Fig. 1, truce C) increased the rate of fluorescence recovery in pHjump studies. Thus, mechanisms other than the Na'/H' antiporter (1,2) may also translocate protons across the brushborder membrane.
Dissipation of ApH by Ct-and K+-The rate of ApH dissipation (kObs) following the addition of vesicles at pH 6.0 to 7.5 buffer was measured using various combinations of TMA gluconate, TMA C1, and K gluconate in the vesicles and in the external buffer (Table I). Taking the rate of ApH dissipation seen when TMA gluconate-equilibrated vesicles were added to TMA gluconate buffer as the control rate, the presence of equal internal and externd TMA C1 (150 mM) caused a 6.2-fold increase in the rate of ApH dissipation. An outwardly directed C1-gradient (150 mM TMA C1 inside, 150 mM TMA gluconate outside) increased the rate 9.4-fold (Table  I). Both rates were further increased by the addition of the protonophore, CCCP, but addition of valinomycin was without significant effect. The rate of ApH dissipation was also increased in the presence of external K' , the rate being increased 4.4-fold by a K' gradient (150 mM K gluconate outside, 150 mM TMA gluconate inside) and 3.8-fold with 150 mM K gluconate on both sides. Both of these rates were further increased by the addition of the K' ionophore, valinomycin, while CCCP had a somewhat lesser effect on accelerating H+ dissipation. The stimulation of ApH dissipation by CCCP in the presence of internal C1-, as compared to gluconate, is consistent with the presence of a C1-conductance pathway. Furthermore, the stimulation of the rate of ApH dissipation by valinomycin in the presence of external K' requires that a protonic conductance pathway also be present.
The effect of acridine orange, per se, on the rate of ApH dissipation was examined. At initial time, vesicles equilibrated with 150 mM TMA gluconate buffer at pH 6.0 were added to K gluconate buffer at pH 7.5, with valinomycin. Acridine orange was either initially present or added at various times after addition of the vesicles. Curves for acridine orange fluorescence versus time were superimposable whether acridine orange was initially present or subsequently added. Therefore, the rate of ApH dissipation was unaffected by the presence of acridine orange.  Vesicles equilibrated in 10 mM Hepes/Tris, 100 mM surose at pH 6.0, and 25 "C with I50 mM of the indicated salt. Generation of ApHs from K ' Concentration Gradients-In these experiments, vesicles equilibrated in 150 mM K gluconate buffer at pH 7.5 were added to pH 7.5 buffer containing K, Na, or TMA gluconate. Control experiments showed that the addition of vesicles to 150 mM K gluconate did not generate a ApH (Fig, 2 A ) . The small quenching seen was due to binding of acridine to the added vesicles and was equal to the maximum fluorescence obtained after pH gradient dissipation (see below). The addition of K gluconate vesicles to 150 mM TMA gluconate buffer produced a small quenching of acridine orange fluorescence (Fig. 2B) which slowly returned toward the maximal value. The return was much faster after the addition of 25 mM Na gluconate and 25 mM K gluconate to the external buffer (verticaE arrow, Fig. 2B). The same K' gradient in the presence of valinomycin (Fig. 2C) produced a much larger fluorescence quenching. The formation of a ApH by an outwardly directed K' gradient and valinomycin demonstrates an inward flux of protons through a membrane potential sensitive path. A still larger fluorescence quenching was seen if the K+ gradient was formed in the presence of valinomycin and CCCP (Fig. 2 0 ) or in the presence of nigericin (data not shown). Finally, it was important to show that the fluorescence  CCCP (B, D). Ionophores, when present, were added from ethanol stocks immediately prior to vesicle addition; ethanol concentrations were less than 0.2%. At the indicated times (r), 50 pl of 1 M NaCl was added to collapse the pH gradient and return the fluorescence to the base-line.
quenching of acridine orange, seen in the presence of valinomycin, occurs in response to a ApH under these conditions. This was tested by adding K gluconate-equilibrated vesicles to Na gluconate buffer. In the presence of an inwardly directed Na' gradient, electroneutral Na'/H+ exchange should decrease any ApH. As shown in Fig. 2E, the presence of an inwardly directed Na+ gradient prevented any fluorescence quenching by the outwardly directed K' gradient. External Na' decreased the ApH generated by a K' gradient with valinomycin (compare Fig. 2, C and Fj or with valinomycin and CCCP (compare Fig. 2, D and G). The effect of external Na+ was related to the magnitude of the ApH, as would be expected for a Na'/H+ antiporter.
Na'/H' Antiporter-generated ApHs-Fluorescence quenching was also seen when BBM vesicles equilibrated with Na gluconate at pH 7.5 were added to TMA gluconate medium at pH 7.5 (Fig. 3A). This Na' gradient-dependent quenching is presumed to be the result of Na'/H+ exchange by the Na+ /   H' antiporter ( 1 , 2 ) , and is abolished by the addition of 25 mM Na gluconate (arrow). Parallel studies' have confumed that the Na' gradient-dependent fluorescence quenching has properties identical to those previously reported for the Na+/H' Warnock, D. G., Reenstra, W. W., and Yee, V. J., Am. J. Physiol., in press. antiporter (i.e. inhibition by external Li' and amiloride). As shown in Fig. 3, the magnitude and stability of pH gradients formed by Na+/H+ exchange are affected by the ionic composition of the media. The presence of either internal C1- (Fig.   313 or external K+ (Fig. 3B) decreased the magnitude of Na+ gradient-dependent fluorescence quenching.
The effects of both CCCP and valinomycin on ApH generated by the Na+/H+ antiporter were measured. All experiments were performed in the absence of chloride. As shown in Fig. 4 4 , the addition of vesicles equilibrated in 75 mM Na gluconate and 75 mM K gluconate at pH 7.5 to 75 mM TMA gluconate and 75 mM K gluconate buffer at pH 7.5 caused a large quenching of acridine orange fluorescence. The addition of vesicles in the presence of either CCCP (Fig. 4B) or valinomycin (Fig. 4C) significantly decreased the quenching of acridine orange. The addition of both valinomycin and CCCP (Fig. 4 0 ) caused fluorescence quenching to be reduced to the base-line response caused by vesicle addition.

DISCUSSION
Several studies have shown that a vesicular suspension maintaining a pH gradient, acid interior, causes the fluorescence of acridine orange to be quenched (8, 13). It has been observed that the magnitude of the fluorescence quenching does not correspond to that predicted wholly on the basis of pH-dependent concentration of amines, but can best be explained by a combination of 1) pH-dependent concentration into the internal space and 2) binding to internal sites (8). The combination of these effects increases the sensitivity of acridine orange to small pH gradients and makes it a useful probe for these studies (8). With this probe, we have conformed the ability of BBM vesicles to produce a pH gradient by Na+/H' antiport, i.e. to couple downhill Na' efflux to uphill H+ influx (1, 2). In addition, we have shown that BBM vesicles can maintain pH gradients imposed by pH jumps of the external pH. Furthermore, it is clear that the rate of pH gradient dissipation (H+ or OH-leakage) varies with the ionic composition.
It is readily seen that the presence of chloride accelerates the rate of pH-gradient dissipation (Table I). Previous studies of BBM vesicles have suggested the existence of both C1-/ OH-antiport and rheogenic C1-pathways (4-6, 15). The observation that CCCP increases the rate of ApH dissipation requires that an electrogenic path for some co-ion must also exist in these membranes in order to balance net charge movement. Acceleration is seen with TMA C1 but not with TMA gluconate (Table I); this result demonstrates that the conductance of chloride is greater than the conductance of gluconate. Likewise, the increase in the rate of pH dissipation by valinomycin in the presence of external potassium demonstrates that a protonic conductance exists across BBM vesicles. A protonic conductance pathway was also shown by the formation of a pH gradient from a K+ concentration gradient in the presence of valinomycin (Fig. 3C). Together, the proton and chloride conductance provide a mechanism for ApH dissipation by internal C1-. However, the present results are not inconsistent with the existence of an electrically neutral CI-/OH-exchanger in addition to rheogenic fluxes of H' and C1-.
Murer et al.
(1) studied the Na'/H+ antiporter in BBM vesicles prepared from rat small intestine and renal cortex. They concluded that Na+/H+ exchange occurs by an electrically neutral mechanism rather than by electrically coupled fluxes of Na+ and H+. This result has been c o n f m e d by other studies ( 2 ) , as well as by the present results.
On the other hand, the present studies demonstrate a protonic conductance pathway in BBM vesicles prepared from with regard to the separate effects of valinomycin and CCCP on the transmembrane pH gradient, the present results conf i i their previous finding that the combined effects of valinomycin and CCCP will collapse any ApH generated by the Na+/H+ antiporter. The possibility that electrodiffusional flux of protonated acridine orange could account for the observed protonic conductance and the differences between our results and those of Murer et al. (1) was considered. However, we view this as unlikely because 1) the flux of protonated acridine orange must be extremely small relative to the un-ionized form or else the ApH-dependent accumulation of acridine orange would be severely compromised (7, 14); 2) K' gradients with valinomycin do not cause quenching of acridine orange fluorescence in vesicles known to have low intrinsic proton con-d~c t a n c e ;~ and 3) the rate of pH gradient dissipation is unaffected by the presence of acridine orange. These findings are readily explained by an intrinsic proton conductance in the BBM vesicles, and are not consistent with electrodiffusional flux of protonated acridine orange.
The observation that CCCP increases the rate of ApH dissipation in the presence of K' and the absence of C1- (Table I) is also noteworthy. This fact, as well as that of the inhibition of Na' gradient-dependent ApH formation by CCCP in the presence of K' (Fig. 4B) suggests that a K' permeability also exists in these membranes. Electrical coupling of H' and K' transport via conductive pathways could also explain 1) the decreased ApH formed by Na'/H' antiporter when Na'-loaded vesicles were added to K' instead of TMA' buffer ( Fig. 3B), 2) the formation of a ApH by a K' diffusion gradient in the absence of valinomycin (Fig. 2B), and 3) the increased dissipation of a preformed pH gradient by external K+ (Fig. 1B). It should be noted that under conditions where H' flux was not limiting, the effects of K' on pHgradient dissipation were smaller than the effects of C1- (Table  I). Therefore, the intrinsic permeability of K' is less than that of c1-.
Finally, these findings may be considered in reference to the physiology of the proximal tubule. Early in the proximal tubule, the Na+/H' antiporter secretes protons into the lumen titrating the filtered bicarbonate (3). In the latter portions of the proximal tubule, after reabsorption of most of the bicarbonate, the luminal concentrations of C1-and H' are elevated with respect to the original filtrate. Any mechanism which then effectively translocates H' and C1-across the luminal membrane will allow continued operation of the Na+/H+ antiporter, and accomplish net NaCl absorption (6,15). Therefore, the Na'/H' antiporter may play a central role in Na' absorption along the entire length of the proximal tubule (16).
Our results demonstrate parallel conductance pathways in the BBM for C1-and H' . Thus, these pathways suggest a ' H. C. Lee and J. G. Forte, unpublished observations. mode of H'/C1-co-transport across the luminal membrane, which is electrically coupled. The elevated luminal C1-and H' concentrations in the late proximal tubule could provide the driving force for HC1 uptake. If C1-transport is rheogenic and coupled to rheogenic H' transport, then the cytoplasmic activity of C1-should not be greater than its electrochemical equilibrium value. On the other hand, if the cytoplasmic C1activity is greater than its electrochemical equilibrium value, then an electrically neutral, coupled transport system for C1-, such as HCl cotransport (or Cl-/OH-exchange), must exist in the luminal membrane. Recent measurements have given conflicting results; data from studies using Necturus proximal tubules indicate that cytoplasmic chloride activity exceeds its electrochemical equilibrium value by approximately 2-fold (17, 18), while measurements in bullfrog and rat proximal tubules indicate that C1-is distributed at its electrochemical equilibrium value (19-21). While the conduction pathways for H' and C1-are clearly large in BBM vesicles, the importance of these paths in the proximal tubule cannot be presently determined.