Introduction

Top Abstract Introduction Methods Materials Results Discussion References

The kidney helps to maintain the body's internal environment by secreting a wide variety of organic ions that originate from the diet, from metabolism or from drugs. Secretion occurs in the proximal tubule of the nephron and represents transport in series: entry from the blood across the basolateral membrane of the renal cell and exit into the tubular fluid across the brush border membrane (for reviews see Ross and Holohan, 1983; Pritchard and Miller, 1991). It is accepted that the exit pathway for organic cations (OC⁺) in the mammalian kidney is mediated by an OC⁺/H⁺ exchange mechanism (Holohan and Ross, 1981a; Rafizadeh et al., 1986; Wright, 1985; Ott et al., 1991). The acidity of the tubular fluid is maintained primarily by the activity of a Na⁺/H⁺ exchanger, thus two functionally linked exchangers drive OC⁺ efflux. Organic cation/H⁺ exchange activity occurs across the brush border membrane of other epithelial tissue: intestine (Miyamoto et al., 1988); choroid plexus (Whittico et al., 1990); placenta (Ganapathy et al., 1988) and liver canalicular membrane (Meijer et al., 1991). Furthermore, exchange activity is found in endosomes from several tissues: kidney (Pritchard and Miller, 1991); liver (Van Dyke et al., 1992); chromaffin granules (Isambort et al., 1992; Howell et al., 1989); adrenergic (Liu et al., 1992; Erickson et al., 1992) and cholinergic neurons (Rogers and Parsons, 1992). These endosomal OC⁺/H⁺ exchangers are energetically coupled to proton ATPases that acidify the interior compartment of the endosome. Finally, OC⁺/H⁺ exchangers are found in prokaryotes (Paulsen and Skurray, 1993; Lewis, 1994), where they are energetically coupled to the proton economy of the prokaryotic cell. Thus, OC⁺/H⁺ exchangers are found in numerous locations throughout nature and participate in a variety of physiological functions. What is unknown is whether or not they are all members of a superfamily of transport proteins, or if they are examples of convergent evolution. The objective of our research is to initiate studies aimed at resolving this issue by completing an essential step: the identification of the protein responsible for OC⁺/H⁺ exchange activity in the brush border membrane of the kidney.

Previously, we had used two different reagents, [¹²⁵I]iodoarylazidoprazozin and Az, to photoaffinity label the exchanger in BBMV from dog kidney (Holohan et al., 1992). Based on the ability of known substrates of the exchanger to block photolabeling and on maneuvers known to disrupt exchange activity, we concluded that a 41-kDa protein is, or at least is a component of, the exchanger (Holohan et al., 1992). The results were somewhat equivocal, however, in that there were other proteins that were photolabeled and protected, at least to some extent (Holohan et al., 1992). Therefore, our study was undertaken to reexamine our findings by improving the specificity of the photolabeling reaction. The results show that Az appears to be an ideal photoaffinity labeling reagent; in the dark it binds competitively at the substrate binding site on the transporter, but it binds irreversibly after photoinactivation. Hence, photoaffinity labeling is specific, efficient and sensitive. The results confirm our previous findings that the exchanger is a 41-kDa protein.

	Methods

Top Abstract Introduction Methods Materials Results Discussion References

Isolation of brush border membranes. BBMV were isolated from dog kidney cortex by a divalent cation precipitation method (Kinsella et al., 1979a). Historically a significant number of renal pharmacological studies have been conducted in dogs; however, the use of the dog as an experimental animal is impractical, leaving a gap in the continuity of pharmacological knowledge. To circumvent this dilemma we attempted to use frozen dog kidneys obtained commercially, but we found that the BBMV prepared from these frozen kidneys gave unreliable results; less than one of five preparations had measurable OC⁺/H⁺ exchange activity (data not shown). We came to recognize, however, that functional BBMV could be prepared if the kidneys were removed and frozen in liquid N₂ immediately after the death of the animal. Therefore, all of the results presented were obtained with BBMV isolated from frozen dog kidneys obtained commercially (Pel-Freeze, Rodgers, AK), but harvested under these defined conditions. The purified BBMV were suspended in standard buffer (10 mM HEPES/100 mM K⁺ gluconate/100 mM mannitol, pH 6.0) and stored at 70°C until used. Protein concentration was determined by the method of Bradford (1976) using bovine serum albumin as a standard.

Photoaffinity labeling of the exchanger. The photoincorporation of Az was accomplished as follows: 10 µl of BBMV (10 mg/ml) were added to various concentrations of Az in standard buffer consisting of 10 mM HEPES/100 mM K⁺ gluconate mannitol pH 7.5 in a final reaction volume of 40 µl. Reagents that were tested for their effect on the photolabeling reaction were prepared in the same buffer and added to the reaction mixture. The reaction mixture was preequilibrated for 30 min in the dark at 4°C; the photochemical reaction was initiated by irradiation with a Spectroline high intensity 302 nm UV lamp at a distance of 5 cm for varying periods of time; the reaction was quenched by the removal of the UV source and by the addition of 960 µl of the pH 7.5 buffer described above; the unreacted Az was removed by one wash cycle; the pellet was dissolved in 25 µl of SDS sample buffer (50 mM Tris-HCl, pH 6.8/2% SDS/20% glycerol/100 µM phenylmethylsulphonyl fluoride/2 µM leupeptin/2 µM pepstatin/0.01% bromophenol blue) and subjected to SDS-PAGE (12% polyacrylamide gel) at 35 mA for 60 min in a Mini-Protean II apparatus. After electrophoresis the gels were fixed with a solution of 10% acetic acid/40% methanol for 1 hr, treated with ENTENSIFY, dried on a Bio-Rad (Hercules, CA) model 583 gel dryer, covered with Fuji XAR film and stored at 70°C for 48 hr. The photolabeling method used is this study differs somewhat from the method used previously (Holohan et al., 1992) in that a pH gradient was imposed across the BBMV (pH 6.0 in/pH 7.5 out). In the presence of a pH gradient, the K_m for the substrate is lowered (Sokol et al., 1988).

Measurement of NMN transport. The transport of a prototypic organic cation, NMN, was assayed by a rapid filtration technique at 37°C (Sokol et al., 1985; Kasher et al., 1983). The influx of [¹⁴C]NMN at the concentration designated in the figure legends was initiated by diluting the BBMV (10 µl) 10-fold with the reaction solution (90 µl). The transport reaction was then terminated after 15 sec with 3 ml ice-cold reaction solution, the suspension poured onto prewetted 0.3-mm Millipore (Bedford, MA) filters and the solution removed under vacuum. The filters were rinsed with the reaction solution, transferred to liquid scintillation vials (10 ml; Ecolume, ICN, Costa Mesa, CA), and the amount of radioactivity contained within the vesicles was measured by standard double label liquid scintillation techniques. In some of the experiments the effect of Az on [¹⁴C]NMN transport was measured, therefore a standard double label program was used for all experiments. Corrections for the amount of radioactivity bound to the filters were determined by running filter controls (reaction solution without BBMV) concurrently with each experiment. The specifically mediated transport was determined in each experiment and was calculated as the difference in the transport of NMN in the absence and presence of a 50-fold excess of mepiperphenidol, a competitive inhibitor (Sokol et al., 1988). Putative substrates were identified by their capacity to cis-inhibit NMN transport (i.e., under zero-trans conditions). The data were expressed as percentages of control where the control represents the specifically mediated transport as defined above. Three to four different BBMV preparations were used for each experiment and each experiment was conducted in quadruplicate.

	Materials

Top Abstract Introduction Methods Materials Results Discussion References

[³H]Azidopine (2,6-dimethyl-4- (2-trifluoromethylphenyl)-1,4-dihydropyridine-3,5-dicarboxylic acid, ethyl, (N-4-azido [3,5- ³H]benzoylaminoethyl) diester with a specific activity of 44 to 50 µCi/nmol was purchased from Amersham Corporation (Arlingtion Heights, IL) as a solution in absolute ethanol. Unlabeled azidopine is not commercially available. NMN with a specific activity of 4.8 mCi/mmol was custom synthesized by ICN (Irvine, CA). The unlabeled N¹-methylnicotinamide chloride, tetraethylammonium chloride, nicardipine and other routine chemical reagents were purchased from Sigma Chemical Co. (St. Louis, MO). Mepiperphenidol [Darstine, 1-(3-hydroxy-5-methyl-4-phenylhexyl-1-methylpiperidinium] bromide was a gift from Merck, Sharp and Dohme (West Point, PA). All the SDS-PAGE reagents were obtained from Bio-Rad.

	Results

Top Abstract Introduction Methods Materials Results Discussion References

Photoaffinity Labeling

Time and temperature. The extent of Az photoincorporation as a function of time and temperature is shown (fig. 1). At 4°C photoincorporation occurred rapidly and efficiently in that the extent of photoincorporation was not increased dramatically by irradiating longer than 30s (fig. 1A). However, longer periods of UV irradiation were needed to achieve comparable levels of photoincorporation at 37°C (fig. 1B). Under either condition the most prominently labeled band was a 41-kDa protein (fig. 1, A and B), and the extent of photoincorporation into that band was diminished by the presence of a known substrate for the exchanger, e.g., mepiperphenidol (Sokol et al., 1988; Holohan et al., 1992; Sokol et al., 1985). Although photolabeling and protection against photolabeling by a substrate were clearly evident at the higher temperature (fig. 1B), the improved efficiency at the lower temperature suggested that conducting the experiments at 4°C was preferable. Hence, all the remaining photolabeling reactions were conducted a 4°C. A 49-kDa protein was prominently labeled as well (fig. 1A). A Coomassie blue stained gel revealed the expected results showing the multiple protein composition of the brush border membrane (fig. 2), yet neither a 41- nor a 49-kDa protein band was detectable by Coomassie blue staining (fig. 2).

View larger version (44K): [in this window] [in a new window]   Fig. 1.   Photoaffinity labeling of as a function of time and temperature. Two sets of BBMV, one at 4°C and the other at 37°C were preequilibrated in the dark for 30 min with 0.4 µM [3H]azidopine in the presence (+) or absence (-) of 5 mM mepiperphenidol. The samples were irradiated for various lengths of time, the photolabeling reaction was quenched, and each sample was analyzed by fluorography after SDS-PAGE (see "Methods"); A, photoaffinity labeling at 4°C; B, photoaffinity labeling at 37°C.

View larger version (56K): [in this window] [in a new window]   Fig. 2.   Coomassie blue stain of the SDS-PAGE gel of the BBMV preparation. BBMV (100 µg) were analyzed by SDS-PAGE, at 35 mA for 60 min in a Mini-Protean II apparatus. After electrophoresis the gel was stained with Coomassie blue G250 for 30 min and destained with 10% acetic acid/40% methanol for 12 hr.

Saturation and specificity. Evidence from the photolabeling experiments suggests (fig. 1) that either, or both, the 41- and 49-kDa protein is the exchanger. This conclusion was tested by two criteria: saturation and specificity. As shown (fig. 3A), as the concentration of Az was increased, the extent of photoincorporation into the 41-kDa protein increased concomitantly. Photolabeling of the 49-kDa protein increased as well, but to a much lesser extent (fig. 3A). At the lower concentrations of Az (fig. 3), the 41-kDa protein was the most prominently labeled, suggesting that it has the highest affinity for Az among all the proteins labeled. As the concentration of Az was increased other proteins became labeled as well (fig. 3A), a finding consistent with the anticipated loss of specificity. The extent of photoincorporation was quantified by densitometry (fig. 3B). As shown, photoincorporation into the 41-kDa protein increased with increasing concentrations of Az (fig. 3B). However, saturation was not attained. Unfortunately, higher concentrations of Az could not be tested because of its limited solubility (see "Methods").The extent of photoincorporation into the 49-kDa protein was much less than the photoincorporation into the 41-kDa protein (fig. 3). An analysis of the two photolabeled proteins gave a ratio of 9:1 (41:49 kDa) when the photoaffinity labeling was conducted at 0.1 or 0.3 µM, but decreased to 4:1 at the higher Az concentrations (i.e., >1 µM).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Effect of increasing concentrations of [3H]Azidopine concentration on photoaffinity labeling. A, BBMV were preequilibrated in standard buffer in the dark at 4°C for 30 min with varying concentrations of Az. The samples were irradiated for 60 sec, quenched with ice cold buffer, and analyzed by fluorography after SDS-PAGE (see "Methods"): lane 1, 0.1 µM Az; lane 2, 0.3 µM Az; lane 3, 1.0 µM Az and lane 4, 3.0 µM Az. B, Densitometric analysis of the 41- () and the 49- () kDa proteins as seen in A.

View larger version (43K): [in this window] [in a new window]   Fig. 4.   Effect of a known substrate of the exchanger (acridine orange) on photoaffinity labeling. BBMV were preequilibrated in the presence of 0.25 µM Az in the dark for 30 min at 4°C with or without acridine orange (AO), irradiated for 60 sec and analyzed by fluorography after SDS-PAGE (see "Methods"). Lane 1, without AO; lane 2, 15 µM AO; lane 3, 3 µM AO and lane 4, 0.75 µM AO.

View larger version (31K): [in this window] [in a new window]   Fig. 5.   Inhibition of photoaffinity labeling. Known substrates of the exchanger were tested for their capacity to inhibit photoincorporation of Az into the 41-kDa protein as measured by densitometric analysis after SDS-PAGE. The experiments were conducted by testing concentrations of each substrate at fixed multiples of their known Ki values. The extent of inhibition is presented as percent of control where the 100% value represents the photoincorporation into BBMV in the absence of competitor. The data represent the mean ± S.E. of four experiments. Each experiment was conducted in duplicate with a separate BBMV preparation.

Subunits. The possibility that the 41 and 49 kDa are subunits of the exchanger held together by a disulfide bridge(s) was tested by running the SDS-PAGE experiments under reducing and nonreducing conditions. The rationale for testing this possibility rested on studies showing that both disulfide and sulfhydryl groups are essential for exchange activity (Sokol et al., 1986). As shown (fig. 6), both proteins are present under either condition, and at a 9:1 ratio. These data also show that both proteins were protected against photolabeling by the presence of a known substrate. Thus, if the two are subunits, they do not appear to be held together by a disulfide bridge(s). Additionally, cross-linking experiments were conducted with 5,5-bisdithionitrobenzoate. BBMV were photolabeled with Az and subjected to numerous experimental conditions of varying 5,5-bisdithionitrobenzoate concentrations and times of treatment; no change in the electrophoretic mobility of the photolabeled polypeptides could be detected (data not shown).

View larger version (44K): [in this window] [in a new window]   Fig. 6.   Photoaffinity labeling under reducing and nonreducing conditions. Aliquots of BBMV with or without 250 mM -mercaptoethanol were preequilibrated with 0.3 µM [3H]azidopine at 4°C for 30 min in the absence or presence of 5 mM mepiperphenidol. The samples were irradiated for 60 sec, the reaction was quenched and each sample was analyzed by fluorography after SDS-PAGE (see "Methods"). Lane 1, BBMV photolabeled in the absence of mepiperphenidol under reducing conditions; lane 2, BBMV photolabeled in the presence of mepiperphenidol under reducing conditions; lane 3, BBMV photolabeled in the absence of mepiperphenidol under nonreducing conditions; lane 4, BBMV photolabeled in the presence of mepiperphenidol under nonreducing conditions.

Transport

The effect of varying concentrations of Az on the influx of 75 µM [¹⁴C]NMN is shown (fig. 7). Inhibition was observed when Az was on the same side of the membrane as the indicator ion (i.e., under zero-trans conditions). The maximal inhibition achieved was approximately 60 percent. Again, experiments could not be carried out at higher concentrations of Az because of its limited solubility. Consequently, controls consisting of nicardipine, a somewhat more water soluble dihydropyridine, and TEA, a known substrate of the exchanger (Wright, 1985), were tested for their effect on [¹⁴C]NMN influx. Both produced cis-inhibition (fig. 7). The effect of Az on [¹⁴C]NMN influx was analyzed further by a kinetic analysis and was found to be consistent with competitive inhibition (fig. 8). From these data, the K_m for NMN was calculated as 25 µM, a value somewhat greater than that of 10 µM published previously (Sokol et al., 1985). The K_i values for Az, nicardipine, and TEA were determined from their IC₅₀ values (fig. 7), using 25 µM as the K_m of NMN, and were calculated as: 200 nM, 280 nM, and 80 µM for Az, nicardipine, and TEA, respectively.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Cis-inhibition of [14C]NMN influx. The 15 sec influx was measured by adding 90 µl of 75 µM [14C]NMN to a medium of 10 mM HEPES, 100 mM K+ gluconate, 100 mM mannitol, pH 7.5 containing varying concentrations of [3H]azidopine (), nicardipine () or TEA () to 10 µl of BBMV (10 mg/ml) suspended in a medium of the same composition but at pH 6.0. The complete details of the transport assay are presented in Methods. The results are presented as percent control with 100% being the specifically mediated transport of 75 µM [14C]NMN as described (see "Methods") and represent the mean ± S.E. of three experiments conducted in quadruplicate with three separate BBMV preparations. Where error bars are not shown, the S.E. falls within the data point. Inset, A representation of the experimental conditions.

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Eadie-Scatchard analysis of Az inhibition. BBMV were preequilibrated in the dark at 37°C for 5 min in standard buffer, pH 6.0 (see "Methods"). The reaction was initiated by the addition of the same medium, pH 7.5 containing predetermined increasing concentrations of [14C]NMN and Az. The specifically mediated transport of [14C]NMN was measured in the absence () or in the presence of 0.05 µM () or 0.1 µM (Az) (). The data are a representative experiment conducted in quadriplicate.

Irreversible Inhibition

As shown (fig. 9), either UV irradiation alone or the experimental maneuvers used to treat the BBMV with Az under dark conditions resulted in some loss of transport capacity (fig. 9, compare columns 2 and 3 with column 1). The effect of Az under dark conditions appeared to be completely reversible since the transport capacity is not different from the UV control or from control conditions (fig. 9, compare column 3 with columns 1 and 3). These findings agreed with the kinetic analysis showing that Az was a competitive inhibitor in the dark. Conversely, transport was irreversibly lost after photoinactivation (fig. 9, column 4).

View larger version (49K): [in this window] [in a new window]   Fig. 9.   Irreversible inhibition of OC+/H+ exchange activity. The data are a representative experiment conducted in quadruplicate of the effect that Az treatment or UV exposure (UV) either alone or in combination has on NMN transport. Lane 1, BBMV were left at 4°C for 30 min, subjected to one wash cycle, and assayed for [14C]NMN transport. The transport capacity of these BBMV was 1020 pmol/min mg protein and was taken as the 100% value. Lane 2, BBMV were left at 4°C for 30 min, exposed to UV irradiation for 60 sec, subjected to one wash cycle and assayed for [14C]NMN transport. Lane 3, BBMV were preequilibrated in the dark with 0.4 µM Az in at 4°C, subjected to one wash cycle and assayed for [14C]NMN transport. Lane 4, BBMV were preequilibrated in the dark with 0.4 µM Az for 30 min at 4°C, exposed to UV irradiation for 60 sec, subjected to one wash cycle, and assayed for [14C]NMN transport. The wash cycle consisted of diluting the BBMV 10-fold with 10 mM HEPES, 100 mM K+ gluconate, 100 mM mannitol, pH 6.0, collecting the samples by centrifugation at 45,000 × g for 5 min, and resuspending the pellet in original volume of the same buffer. The samples were reequilibrated at 37°C for 5 min before the specifically mediated transport of 75 µM [14C]NMN was measured (see Methods).

The high efficiency of photolabeling provided a method of quantifying the exchanger as follows: first, the BBMV were photolabeled in the presence and absence of a known substrate and subjected to SDS-PAGE; second, those areas of the gel containing the labeled proteins (determined by R_f values) were excised and their radioactive content determined. The moles of Az covalently photoincorporated into the exchanger were then calculated from the difference in radioactive content of the proteins photolabeled in the presence and absence of a substrate. (We found that it was technically impractical to excise the 41-kDa band exclusively, and therefore the region of the gel encompassing both the 41- and 49-kDa proteins was removed and counted. Obviously, this means that the measurements are somewhat in error.) By this procedure the transport capacity and the abundance of the exchanger was correlated as shown (table 1). The average value for the abundance of the exchanger in BBMV is 780 ± 140 fmol/mg protein (n = 4) or 0.003% of the total membrane protein.

Variability in transport capacity among BBMV preparations has been noted previously (Kinsella et al., 1979b; Sokol et al., 1985, 1989), although the exact causes were not fully understood. Our study suggested that inactivation was a major cause. Therefore, the effect of heat inactivation on photoaffinity labeling was tested. When BBMV were heated to 65°C for 2 min, the photoaffinity labeling of the 41-kDa protein was lost and [¹⁴C]NMN/H⁺ exchange activity was destroyed concomitantly (data not shown). When the experiments were conducted at a lower temperature (55°C), the loss of photoaffinity labeling and the loss of exchange activity correlated (fig. 10).

View larger version (41K): [in this window] [in a new window]   Fig. 10.   Heat inactivation of transport capacity and photoaffinity labeling. BBMV were heated to 55°C for varying lengths of time and then assayed for [14C]NMN transport and for their capacityo be photoaffinty labeled with Az. The transport reaction was conducted in BBMV that had been preequilibrated in standard buffer pH 6.0 at 37°C for 5 min before assayed. The specifically mediated transport was determined as described (see Methods) and the 100% transport capacity represents the exchange activity of the BBMV before heat treatment (0 time). The photoaffinity labeling was conducted in BBMV that were preequilibrated in the dark with 0.3 µM Az for 30 min at 4°C, exposed to UV irradiation for 60 sec, quenched with ice-cold buffer and analyzed by SDS-PAGE. The moles of Az photoincorporated were calculated from the difference in the radioactive content of the bands photolabeled in the presence and absence of 5 mM mepiperphenidol. The data are a presented as the means ± the S.E. of three experiments conducted in quadruplicate with three separate BBMV preparations.

	Discussion

Top Abstract Introduction Methods Materials Results Discussion References

Organic cation/H⁺ exchangers are found in many different kinds of cells and in many different tissues. The monoamine/H⁺ exchangers from the chromaffin granule and adrenergic neurons have been the most thoroughly characterized. They belong to a family of closely related proteins having a molecular mass of about 80 kDa (Schuldiner, 1994). Conversely, the bacterial transporters belong to a large superfamily of proteins that can be arranged into subgroups, one of which is composed of several members having a molecular mass of 40 to 44 kDa (Schuldiner, 1994). Interestingly, the monoamine/H⁺ exchangers and the bacterial exchangers are distantly related (Schuldiner, 1994), and yet have important differences such as the stoichiometry of the exchange reactions; the monoamine transporters catalyze 2 H⁺: 1 OC⁺ while the bacterial transporters catalyze a 1:1 exchange.

The molecular properties of the OC⁺/H⁺ exchangers in the brush border membranes of epithelial tissues have not been described in any depth. Photoaffinity labeling of canalicular membranes from rat liver identified two potential OC⁺/H⁺ exchangers: one with a molecular mass of 78 kDa and the other with a molecular mass of 38 kDa (Meijer et al., 1991). Interestingly, in this tissue OC⁺/H⁺ exchange activity of 1:2 and 1:1 has been described. The issue under study is OC⁺/H⁺ exchange in kidney. Thus far, only electroneutral OC⁺/H⁺ exchange activity has been described, although the possibility of 1:2 (OC⁺/H⁺) stoichiometry has been noted (Wright, 1985).Previously, we tentatively identified a 41-kDa protein in BBMV from dog kidney by photoaffinity labeling (Holohan et al., 1992). Our study confirms and extends those earlier findings.

The results show that photoaffinity labeling with Az is an efficient, specific and sensitive method for identifying and quantifying the OC⁺/H⁺ exchanger in BBMV from dog kidney. The efficiency of photolabeling is shown by the findings that maximal photoincorporation is achieved within minutes at 4°C (fig. 1) even though the exchanger is only a minor component of the BBMV (fig. 2). Our experiments were conducted under conditions where a pH gradient was imposed across the BBMV, and at a lower temperature than had been used previously (4°C vs. 37°C). One possible explanation for the improved efficiency is that when a pH gradient is imposed across the membrane, the transporter cycles between high affinity and low affinity states (Holohan and Ross, 1980; Holohan and Ross, 1981 a, b; Sokol et al., 1988). Hence the high affinity site maybe stabilized at the exterior face of the BBMV at 4°C, thereby favoring the photoaffinity labeling reaction. It is possible to photolabel in the absence of a pH gradient, but the extent of photoincorporation is noticeably diminished (data not shown). Therefore, under the experimental conditions used, Az is efficiently incorporated into a protein(s) that is not a major component of the brush border membrane (fig. 2).

The specificity of photolabeling is shown also by the finding that known substrates protect against (or inhibit) photoincorporation, and that the extent of protection is related to the affinity of the exchanger for that particular substrate (figs. 4 and 5). Moreover, if the exchange activity is destroyed by heating, the capacity to photolabel the 41-kDa protein is lost concomitantly (fig. 10). The sensitivity of photolabeling reaction is demonstrated by the fact that the 41-kDa protein is specifically labeled by concentrations of Az less than 100 nM whereas other proteins are not (fig. 3). However, saturation of photolabeling was not attained. One possible cause of why saturation is not observed is because concentrations of Az greater than those used are needed, but cannot be prepared. The reason is that Az is obtained as a solution in absolute ethanol and in order to achieve higher concentrations, ethanol would be present in the reaction solution at concentrations (i.e., >3%) that are incompatible with the catalytic integrity of the exchanger (Sokol et al., 1989), and as demonstrated in figure 10, catalytic activity is needed to observe photoaffinity labeling. An attempt was made to circumvent this problem by concentrating the Az by evaporation. This maneuver produced a chemically unstable Az (data not shown), and therefore these efforts were abandoned.

The photoaffinity labeling experiments support the conclusion that Az binds at the transport, or substrate binding site, on the exchanger; i.e., Az is a substrate. Unfortunately, it is not possible to test this directly because the highly lipophilic nature of Az gives background radioactivity several orders of magnitude greater than the amount of radioactive material moved by the transport. Therefore, the interpretation that Az binds at the substrate binding site on the exchanger had to be established indirectly by examining the effect Az has in the dark on the transport of an indicator OC⁺ (i.e., [¹⁴C]NMN) by a double labeling technique. A disadvantage of the method is that the low specific activity of the [¹⁴C]NMN diminishes the precision of the measurements. Nevertheless, the results are consistent with the conclusion that Az is a competitive inhibitor (fig. 8). However, these findings do not allow us to conclude unequivocally that Az is a substrate. It is conceivable that Az binds at the substrate binding site on the exchanger, but is not translocated across the membrane. Nonetheless, the results are consistent with the conclusion that Az is an ideal photoaffinity labeling reagent; it binds reversibly in the dark at the substrate binding site, but irreversibly after photoinactivation. These properties were exploited to quantify the exchanger. As would be expected, the transport of a prototypical OC⁺ (i.e., [¹⁴C] NMN) is irreversibly lost after photoactivation, but not before (fig. 9). These findings strengthen the conclusion that the photolabeling reaction is remarkably efficient. Under the experimental conditions (400 nM Az, K_i of 200 nM, and assayed for the transport of 75 µM [¹⁴C]NMN, K_m of 25 µM) the transport capacity should be inhibited by 61% if each transport protein molecule that has an Az molecule bound in the dark becomes photoinactivated. The actual loss of transport capacity is 46%; the efficiency of photolabeling is calculated as actual/theoretical and gives a value of 75%. The photolabeling method also provides a way of quantifying the exchanger. We calculate a density of 780 fmol/mg protein which represents 0.003% of the membrane protein. This value is consistent with the finding that the 41-kDa protein is not detectable by Coomassie blue staining (fig. 2). To the best of our knowledge this study is the first report of any estimate of the abundance of the exchanger in kidney BBMV or in any other epithelial tissue. Recently, Kimura et al. (1995) synthesized a photoaffinity analog of cimetidine (a known substrate of the exchanger) which they used to photolabel BBMV from rat kidney. They concluded that the OC⁺/H⁺ exchanger is a 36-kDa protein, but they did not attempt to quantify the level of expression. They also concluded that the 36-kDa protein in rat kidney is different from the 41-kDa protein in dog kidney. The basis for their conclusion is not clear to us. It may be that the difference in molecular mass simply represents a difference in the extent of glycosylation.

The results show that a 49-kDa protein is photolabeled in addition to the 41-kDa protein. The photoincorporation into the larger protein is paradoxical; it has low affinity for Az (fig. 3), yet the labeling is specific (fig. 4). The effect of Az concentration on photoincorporation (fig. 3) suggests that either the 49-kDa protein is present in very small amounts, or it has low affinity for Az, or both. The shape of the curve (fig. 3B) describing photoincorporation into the 49-kDa protein appears to be linear, suggesting nonspecific labeling. Yet the photoincorporation is blocked by the presence of a substrate for the exchanger (figs. 1, 4 and 6). Furthermore, if photolabeling of the 41-kDa protein is destroyed by heating (fig. 10), photolabeling of the 49-kDa protein is lost concomitantly (data not shown). Photolabeling of the 49-kDa protein has characteristics consistent with those expected of specific binding (specificity and saturation) which suggests that it may be a component of the exchanger. However, the detection of the 49-kDa protein is highly variable. When the experiments are conducted at 37°C, the 49-kDa band is rarely, if ever, seen, in agreement with our previous results (Holohan et al., 1992). When the labeling experiments are conducted at 4°C, the 49-kDa band is found most of the time, but not always. Thus, the relationship between the 41- and 49-kDa proteins remains unresolved. Consequently, the possibility that either the exchanger is a heteroligomer or that there is more than one exchange protein remains unanswered. Regardless, the evidence supports the conclusion that the 41-kDa protein is, or at least is a part of, the OC⁺/H⁺ exchanger. The molecular mass of the exchanger and the stoichiometry of the reaction that is catalyzed (Sokol et al., 1985) suggests that it may be more closely related to the bacterial transporters than to the monoamine/H⁺ exchangers found in the nervous system.