INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE KIDNEYS ELIMINATE a wide variety of organic anions (OAs) from the body, and mechanisms associated with the secretion of OAs by cells of the renal proximal tubule have received considerable attention (21). Transport of OAs into proximal cells from the blood, that is, transport across the basolateral (peritubular) membrane of proximal cells, is the active step in trans-tubular secretion (21). Studies of Ullrich and colleagues (30, 32) on the selectivity of peritubular OA transport suggest that there is a single transport process, the "classic" OA transporter, that accepts a broad range of chemical structures and for which p-aminohippurate (PAH) has proven to be the prototypical substrate. Peritubular transport of PAH shows a wide degree of sensitivity to the presence or absence of sodium and potassium in the external medium, as well as to the presence of a wide range of metabolic intermediates, including many substrates of the tricarboxylic acid cycle (21).

In 1987 two research groups independently proposed a model that integrated effectively the broad range of observations on ionic and metabolite dependence of peritubular OA transport (18, 24). The model involves the concerted activity of three parallel transport processes in the peritubular membrane of renal proximal cells: the Na⁺-K⁺-ATPase; a Na⁺-dicarboxylate (Na-DC) cotransporter; and the classic OA transporter (referred to hereafter as the OA/DC exchanger), which mediates the exchange of extracellular OAs for a limited set of intracellular dicarboxylates. Active uptake of OAs is driven by an outwardly directed gradient of dicarboxylates [principally -ketoglutarate (KG)], which, in turn, is supported by the uptake of exogenous (extracellular) KG by the Na-DC cotransporter. Steady-state activity of these functionally linked transport processes is ultimately dependent on Na⁺-K⁺-ATPase activity, which maintains the transmembrane electrochemical gradient for Na⁺. This "tertiary active" transport model has been demonstrated to operate in several different mammalian experimental systems, including isolated basolateral membrane vesicles (18, 19, 24), slices of intact renal cortex (20), and single isolated renal proximal tubules (3, 29). Functionally linked transporters for dicarboxylates and OAs have also been demonstrated in proximal tubules from reptilian (2) and teleost (13) kidneys and from the proximal-like epithelium of crustacean urinary bladder (15). Hence, the functional coupling of parallel transporters appears to be an ancient and highly conserved strategy to support active transport of OAs.

Two aspects of the tertiary active transport model remain untested. First, according to the model, OA uptake is increased as a consequence of transport of exogenous KG. In fact, it is not known whether activity of the Na-DC cotransporter under physiological conditions increases uptake of OAs. Proximal tubule cells can actively accumulate PAH in the absence of exogenous dicarboxylates (e.g., Ref. 8), indicating that cellular metabolism can produce sufficient concentrations of "endogenous" KG to support activity of the OA/DC exchanger. Most studies linking transport of exogenous dicarboxylates to the stimulation of OA transport have used concentrations of exchangeable dicarboxylates much higher than the ~10 µM concentration found in the plasma (17, 23). Although low concentrations of KG have been shown to stimulate OA uptake into renal cells (17), the conditions under which these measurements were made differ markedly from conditions in vivo (in terms of temperature and buffer composition) and were likely to have reduced metabolic production of KG. It is therefore premature to conclude that transport of exogenous KG influences transport of OAs in vivo.

Second, the tertiary active transport model suggests that Na-DC cotransport serves to "recycle" the KG that exchanges for extracellular OAs, thereby sparing the loss of this substrate and helping to maintain the intracellular KG pool (18, 21, 22). Consistent with this notion, net loss of glutarate from proximal cells loaded with the radiolabeled compound is accelerated when the Na-DC cotransporter is inhibited with Li⁺ (20). However, a prediction that arises from the model is that the rate of OA transport should be sensitive to the activity of the Na-DC cotransporter, even in the absence of exogenous KG. There is no such evidence.

We addressed these two issues in the present study by examining the effect of Na-DC cotransporter activity on the initial rate of OA transport. Epifluorescence microscopy and fluorescein (FL), a fluorescent OA, were used to measure activity of the peritubular OA transporter in single isolated rabbit proximal tubules (28). Our results suggest that, under conditions similar to those found in vivo, Na-DC cotransport does play a significant role in sustaining activity of the OA/DC exchanger, even in the absence of exogenous dicarboxylates.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Chemicals

Solutions

Animals and Proximal Tubule Preparation

Measuring FL Uptake Into Rabbit Proximal Tubules

Figure 1A shows a fluorescence profile of FL exchange in the chamber, resulting from the switch from an FL-free buffer to one containing 1 µM FL. For this profile, the microscopic field did not include a tubule, and the microscope was focused at the same level as that used for the profile shown in Fig. 1B. Before adding FL, the signal was relatively low, 480 ± 30 photons/s, representing instrument background. Background fluorescence increased rapidly to 364,000 ± 960 (SD) photons/s when FL was added to the bath, yielding a signal-to-noise ratio of ~750. This "solution background" was very stable; there was no significant change in slope during the period measured, and the standard deviation of the signal was less than 0.3% of the average. The fluorescence measured was a function of both monochrometer slit width and microscopic field size, which varied between experiments but not within an experiment. As a result, the signal-to-noise ratio remained constant between experiments.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Fluorescence output measured in the microscope-based photomultiplier system that occurred upon switching from a superfusion buffer containing no fluorescein (FL) to one containing 1 µM FL. A: a fluorescence profile obtained when the microscope was focused on a region of the chamber that did not contain a tubule (the plane of focus was at the same level used to obtain the profile shown in B). Open circles, fluorescence output prior to switching to the FL-containing solution, i.e., instrument background (instr. bkgd.), in this case, 480 ± 30 (SD) photons/s. After switching to a solution containing 1 µM FL, there was a rapid increase (t0.5 < 2 s) in background fluorescence to a level of 3.6 × 105 ± 960 (SD) photons/s (solid circles). This "solution background" did not change with time. Dotted line, linear regression of fluorescence vs. time for t = 10 to t = 35 s (slope of 29 photons/s2; r2 = 0.05, P > 0.1). B: a fluorescence profile obtained when the microscope was focused on a portion of a proximal S2 renal tubule. Instrument settings (monochrometer slit width and microscope field) were the same as those used in A. Open circles, fluorescence output prior to switching to the FL-containing solution (i.e., combined influence of instrument background and tubule autofluorescence), in this case, 2,580 photons/s. After switching to a solution containing 1 µM FL, there was a rapid increase in fluorescence (solid circles), resulting from the exchange of solutions in the bath, and a slower increase in tubule fluorescence indicating accumulation of FL by the tubule. Dotted line, linear regression of fluorescence vs. time for t = 10 to t = 35 s (slope of 4,790 photons/s2; r2 = 0.99).

Figure 1B presents a fluorescence profile to show how initial rates of FL uptake were calculated. For this profile, the microscope was focused on a tubule (midpoint of tubule depth) in the chamber. Tubule autofluorescence at FL wavelengths was 2,580 ± 49 (SD) photons/s, a 5.4-fold increase in signal over instrument background. Upon switching to a superfusion buffer containing 1 µM FL, the increase in signal had two evident phases: a rapid increase in fluorescence due to the addition of FL to the bath and a slower increase in fluorescence indicating accumulation of FL in the tubule. Because the half time (t_0.5) for solution exchange in the chamber was 1.5 to 2 s, the first 5-10 s of the fluorescence record (i.e., ~5 halftimes) was discarded following the switch to a buffer containing FL. The next 25 s of the record was linear, as shown by the t = 10-35 s portion of the trace in Fig. 1B. The slope of this line was calculated and represents the initial rate of FL uptake. For example, using this standard approach, the slope calculated for Fig. 1B was 4,790 photons/s², compared with a standard deviation of 960 photons/s for the solution background (Fig. 1A).

Measuring the Stimulation or Inhibition of FL Uptake

Cis-configuration. Tubules were superfused (~12-15 min) with buffer containing no KG followed by a measurement of FL uptake using the superfusion buffer containing 1 µM FL plus KG. The cis-configuration tested the "immediate" effect of exogenous KG on FL uptake, including the potential inhibitory effect of KG as it competed with FL for binding to the extracellular face of the OA transporter (i.e., cis-inhibition).

Trans-configuration. Tubules were preloaded with KG by superfusing 12-15 min with buffer containing KG, and FL uptake was then measured using the superfusion buffer containing 1 µM FL and no KG. The trans-configuration tested the effect on FL uptake of an outwardly directed KG gradient (i.e., trans-stimulation). This configuration also eliminated any potential cis-inhibition of FL uptake by KG.

Cis-trans-configuration. Tubules were preloaded with KG by superfusing 12-15 min with buffer containing KG, and then FL uptake was measured using the superfusion buffer containing 1 µM FL and KG. The cis-trans configuration measured the simultaneous trans-stimulation and potential cis-inhibition effects on FL uptake produced by the continuous presence of exogenous KG (i.e., steady-state condition).

The influences of PAH, Li⁺, or methylsuccinate on the initial rate of FL uptake were tested using the superfusion buffer containing 1 µM FL plus the test substrate in the cis-configuration. The effect of sodium's absence on FL transport was determined by replacing sodium in the uptake medium with NMDG. The initial rates of FL uptake were first measured under the control condition (both superfusion buffer and uptake media contained ~150 mM sodium). Tubules were then superfused for 10 min with superfusion buffer, and FL uptake rates were measured using uptake medium containing FL and NMDG. When measuring the effects on FL transport resulting from various transmembrane configurations of KG or other experimental treatments (e.g., cis-inhibition), two to four measurements of FL uptake in the control condition and two to four measurements of FL uptake in the experimental condition were made per tubule.

Measuring Uptake of PAH Into Rabbit Proximal Tubules

Statistical Analysis

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Time Course of FL Accumulation and Precision of Initial Rate Measurements

View larger version (15K): [in this window] [in a new window]   Fig. 2.   A: fluorescence profile of an intact proximal renal tubule exposed first to a superfusion buffer containing 1 µM FL (FL influx) and then to a buffer containing no FL (efflux). B: initial rates of FL uptake into an intact proximal tubule measured using the brief-exposure protocol and a superfusion buffer containing 1 µM FL and 20 µM -ketoglutarate (KG) in the cis-trans- configuration. The 4 uptake rates shown are representative of 15 initial rates of FL uptake measured over a 3-h period.

When the initial rate of FL uptake was repeatedly measured in a single tubule, exposure to the uptake medium was limited to 30-35 s. Accumulated FL was then washed out of the tubule by superfusing the tubule for 10 min using an FL-free buffer. Figure 2B shows 4 representative measurements of uptake from an experiment where 15 replicates of FL uptake into a tubule were measured sequentially over a 3-h period using 20 µM KG in the cis-trans configuration and 1 µM FL. Typically, FL uptake rates were very reproducible when measured using this brief-exposure protocol and seldom declined by more than 15% over a 4-h period. If uptake rates in controls declined by more than 15%, then a mean uptake rate was calculated from rates measured in controls before and after the experimental treatment. The mean uptake rate was then compared with the rates of FL uptake measured under the experimental condition.

Interaction of FL with PAH Transport

View larger version (13K): [in this window] [in a new window]   Fig. 3.   A: effect of increasing FL concentrations on p-aminohippurate (PAH) uptake into single proximal S2 tubules. Initial rates of PAH uptake were based on 15-s incubations in buffer containing 5 µM [3H]PAH and concentrations of FL ranging from 0 to 1 mM. Each point is the mean ± SE of 3 separate experiments, each one of which involved measurement of uptake into at least 3 tubules at each test condition. Line was fitted to the data with the assumption of a competitive interaction between FL and PAH (10). B: effect of PAH on the uptake of FL into single rabbit renal proximal S2 tubules. Initial rates of FL uptake were first measured in controls, exposing tubules to uptake medium containing 1 µM FL. Rates of FL transport were then measured while exposing tubules to uptake medium containing 1 µM FL and 5 mM PAH. Height of each bar is mean ± SE of the rate of uptake (expressed in photometer units) determined in 3 separate experiments. * Decrease in FL uptake was significant (P < 0.05) as determined by a one-way, two-sample t-test.

The second criterion, i.e., that transport of FL should be limited to the PAH transporter, was supported by the observation that 5 mM PAH in the uptake medium reduced the initial rate of FL uptake into single tubules by 95% (Fig. 3B), suggesting that peritubular transport of FL is effectively limited to the OA/DC exchanger. These results agree with the report that PAH inhibits FL transport into rabbit proximal S2 segments with a K_i of 141 µM, similar to the 110 µM value of K_t for peritubular PAH transport (5). Taken together, these results allow us to conclude that FL is comparable to PAH as a model substrate for the peritubular OA/DC exchanger.

Influence of "Physiological" Concentrations of Exogenous KG on Peritubular FL Transport

View larger version (16K): [in this window] [in a new window]   Fig. 4.   A: stimulation of uptake of 1 µM FL into intact proximal tubules resulting from exposure to 10 or 100 µM KG in the trans-configuration. Each bar is mean ± SE (n = 18; m = 16) of the rate of FL uptake compared with the average rate of uptake measured in KG-free buffer (controls). * Increase in the initial rate of FL uptake was significant (P < 0.05), compared with controls as determined by a one-way paired t-test. B: effect of the cis, trans, and cis-trans configurations of 10 µM KG on the rate of FL uptake into single renal proximal S2 tubule segments exposed to bicarbonate-buffered solutions. Tubules were exposed to each of the transmembrane configurations of a 10 µM concentration of KG (see text) prior to measurement of the initial rate of 1 µM FL uptake. Each bar is mean ± SE of uptake measured in 8-11 tubules (m = 9). Uptakes are expressed as percentage increase in transport over control uptake measured in absence of KG. * Significant increases in transport (P < 0.05; one-way paired sample t-test) over that measured under the control condition (not exposed acutely to and/or preloaded with KG).

Unlike the situation produced by the trans-configuration, transport of OAs in vivo occurs under conditions where exchangeable dicarboxylates are distributed at steady state across the peritubular membrane, i.e., a cis-trans configuration. At least one study has examined the effect of a steady-state "cis-trans" distribution of a low concentration of KG on the activity of the OA/DC exchanger. Pritchard (17) showed that simultaneous exposure of rat renal cortical slices to a solution containing 10 µM [³H]PAH and 10 µM KG causes a 30% increase in the steady-state tissue/medium ratio of PAH accumulation. Although these data suggest that a concentration and transmembrane configuration of KG found in vivo can stimulate accumulation of OAs, this interpretation is complicated by the fact that the tissue was incubated at room temperature in a nutrient-free, nominally bicarbonate-free, phosphate buffer. These conditions are likely to have influenced cellular metabolism and/or rates of OA transport. Thus we sought to clarify the extent to which exogenous KG influences peritubular OA transport under conditions similar to those to which proximal tubules are normally exposed in vivo.

We compared the effect of the three different KG exposure protocols, the cis-, trans-, and cis-trans-configurations, on the uptake of 1 µM FL at 37°C using an extracellular concentration of 10 µM KG and a bicarbonate buffer. Two observations were of particular interest. First, as shown in Fig. 4B, all three KG configurations significantly stimulated, by 54% to 76%, uptake of 1 µM FL, compared with the control. These data support the conclusion that exposure of tubules to exogenous KG under conditions that mimic those found in vivo (with respect to temperature, substrate concentration, and buffer composition) can significantly increase the rate of peritubular OA transport. Because of the variability in the degree of stimulation produced by tubules from different rabbits, a simple one-way repeated measures ANOVA did not indicate significant differences between the different transmembrane configurations of KG. However, we think it is also worth noting that of the five trans and cis-trans pairings, in four cases FL uptake measured under the cis-trans-configuration exceeded that measured under the trans-configuration (in the 5th case, the rates were virtually identical). This observation suggests that the continuous accumulation of exogenous KG produces a higher concentration of this exchangeable substrate at the cytoplasmic face of the OA/DC exchanger than can be sustained by diffusion from the elevated cytoplasmic pool of KG.

The second point apparent from the experiments summarized in Fig. 4B was that the increase in FL transport resulting from exposure to exogenous KG can occur quite rapidly. Under the cis-configuration, the 22-160% increase in FL transport we observed in 11 separate experiments occurred within the 5-7 s that represented the limit of temporal resolution of our present technique. The rapid onset of the cis-stimulation is evident in the data presented in Fig. 5. In this experiment, acute exposure of a tubule to 10 µM KG immediately increased the initial rate of FL uptake by 85% compared with the control condition. Simultaneous exposure of the tubule to 10 µM KG and 2 mM Li⁺ essentially eliminated the KG-mediated stimulation of FL transport. In five measurements from two separate tubules, 2 mM Li⁺ blocked 96% of the increase in FL uptake produced by 10 µM KG under the cis-configuration. Because Li⁺ is a specific inhibitor of the Na-DC cotransporter (1, 19, 31), we concluded that the increased transport produced by the cis-configuration was secondary to the uptake of exogenous KG. The increase in FL uptake presumably reflected an increased turnover of the OA/DC exchanger that was concomitant to an increase in the intracellular KG concentration, an increase that must have been at least proportional to the increase in FL transport. In other words, the stimulation of FL transport noted under the cis-configuration must have been secondary to an increase of 22% to 160% in the concentration of KG at the cytoplasmic face of the exchanger. The rapidity of the response necessitates that the increase in KG occur within a few seconds, a subject we address in the DISCUSSION.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Fluorescence record showing stimulation of FL uptake into a proximal tubule segment resulting from an acute cis-exposure to 1 µM FL plus 10 µM KG and blockage of that cis-stimulation by simultaneous exposure to 2 mM Li+. FL uptake is expressed in terms of photomultiplier output. Fluorescence output prior to t = 0 (period labeled "No FL") reflects autofluorescence of tubule. At time 0, superfusion buffer was switched to one containing 1 µM FL. In all measurements, there was a rapid increase in fluorescence (to ~450,000 photons/s) due to the addition of FL and a subsequent slower increase in fluorescence from the accumulation of FL in the tubule. Each point is mean ± SE of 3 determinations of FL uptake (where error bars are not evident, the error was smaller than the graphic representation of the mean). Solid lines, linear regressions determined from increases in fluorescence between t = 5 s and t = 25 s.

Effect of DC Transport on FL Transport in Absence of Exogenous KG

View larger version (13K): [in this window] [in a new window]   Fig. 6.   A: effect of sodium removal on the uptake of 1 µM FL into single renal proximal S2 tubule segments. Sodium in the uptake medium was replaced with N-methyl-D-glucamine (NMDG), whereas the control condition contained the normal concentration of sodium. Height of each bar is mean ± SE of uptake measured in 5 tubules (m = 4). Buffers used for these experiments did not contain KG. * Decrease in initial rate of FL uptake was significant (P < 0.05) compared with controls, as determined by using a one-way, two-sample t-test. B: effect of 2 mM Li+ and 1 mM methylsuccinate (MS) on 1 µM FL uptake into a single renal proximal S2 tubule segment. Height of each bar represents uptake as mean ± SE of 2-3 sequential measurements of FL uptake in presence of one or both of the test agents. Control bar is mean ± SE of 9 measurements of initial rate of uptake measured before and after each of the test conditions. All buffers used for this experiment did not contain KG. * Decrease in initial rate of FL uptake was significant (P < 0.05) compared with controls, as determined by ANOVA.

Although these data are consistent with the idea that activity of the OA/DC exchanger is functionally linked to activity of the Na-DC cotransporter, it is difficult to dismiss the possibility that removal of Na⁺, even transiently, could have influenced cellular processes that exert indirect effects on peritubular OA transport. Consequently, we examined the effects of two comparatively selective inhibitors of the Na-DC cotransporter on the rate of FL transport. The first was exposure to 1 mM methylsuccinate, and the second was exposure to 2 mM Li⁺. Methylsuccinate is a high-affinity substrate (K_t = 10 µM) of the peritubular Na-DC cotransporter in rabbits (1) and a relatively weak inhibitor of the OA/DC exchanger (33). Thus a sufficiently large concentration of methylsuccinate in the uptake medium should prevent KG from entering proximal cells by the Na-DC cotransporter. A concentration of 1 mM methylsuccinate was selected, because in rabbit proximal tubules it reduces by more than 95% the stimulatory effect of exogenous KG on PAH uptake (3). As noted previously, Li⁺ is a relatively specific inhibitor of the peritubular Na-DC cotransporter (1, 19) that competes with Na⁺ for binding to the cotransporter and produces an activator-transporter complex with an extremely low affinity for dicarboxylates. Chatsudthipong and Dantzler (3) showed that 2 mM Li⁺ eliminates the stimulation of PAH transport into rabbit renal proximal tubules produced by exposure to 100 µM KG. The results presented in Fig. 5 confirmed that 2 mM Li⁺ blocks the accumulation of exogenous KG.

We tested the effects of Li⁺ and methylsuccinate on FL transport in separate experiments with tubules superfused with buffer containing no exogenous KG. Acute application of 2 mM Li⁺ or 1 mM methylsuccinate significantly (P < 0.05) reduced FL uptake as follows: Li⁺ by 18 ± 5.0% (n = 3, m = 3) and methylsuccinate by 29 ± 2.3% (n = 3, m = 3). Significantly, in neither case was the degree of inhibition of FL transport as substantial as that produced by complete removal of Na⁺ from the medium (Fig. 6A). If the effects of Li⁺ and methylsuccinate are through a common mechanism, i.e., inhibition of the Na-DC cotransporter, then their effects should not be additive. This hypothesis was tested in a separate experiment that used a single tubule, the results from which are summarized in Fig. 6B. Consistent with the observations reported above, Li⁺ and methylsuccinate each reduced the rates of FL uptake into the tubule by ~28%, inhibitions that were completely reversed by reexposing the tubule to the control buffer. When the tubule was then simultaneously exposed to 2 mM Li⁺ plus 1 mM methylsuccinate, the inhibition of FL uptake (33%) was not different from that produced by either Li⁺ or methylsuccinate alone (P > 0.05). Thus the effects of Li⁺ and methylsuccinate appear to be limited to inhibition of the Na-DC cotransporter. The consistent observation that selective inhibition of the Na-DC cotransporter in the absence of exogenous KG simultaneously reduced FL transport by ~25% suggests that 25% of peritubular OA/DC exchanger activity requires parallel activity of the Na-DC cotransporter.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The current observations support two conclusions. First, under conditions that mimic closely those found in vivo (i.e., substrate concentration, buffer composition, incubation temperature), exogenous KG (that is, KG in the extracellular medium, in contrast to endogenous, metabolically produced KG) significantly stimulates the peritubular uptake of FL, a model OA. Second, activity of the peritubular Na-DC cotransporter significantly enhances OA uptake, even in the absence of exogenous KG.

Prior to these observations, firm conclusions about these issues were not possible. Although it was apparent that peritubular OA transport involves mediated exchange with intracellular KG (18, 24) and that uptake of exogenous KG stimulated OA uptake, it was also evident that isolated renal tissue can actively accumulate OAs in dicarboxylate-free medium (e.g., Ref. 8). Thus it was not known whether uptake of exogenous KG would increase cytoplasmic KG to levels above that already present due to cellular metabolism. Also, previous studies frequently used nonbicarbonate buffers (14, 17, 20), nutrient-free media (11, 14, 17, 20), and nonphysiological incubation temperatures (17, 20). Any of these conditions is likely to have affected the rates of transport and/or cellular metabolism in ways that altered the influence that DC transport had on the transport of OAs. For example, in rabbit renal proximal cells, exposure to bicarbonate-free buffers decreases the pool size of cytosolic dicarboxylates in parallel with increases in the mitochondrial pool size of dicarboxylates (26). In intact renal proximal tubules from dog, low-bicarbonate conditions decrease cytosolic concentrations of KG without changing the concentration of mitochondrial KG (27). Bicarbonate deletion also results in a direct inhibition of mitochondrial respiration and a decrease in the cytosolic pool sizes of tricarboxylic acid cycle intermediates in rabbit proximal tubules (6). These observations all suggest that rates of OA transport measured in the absence of bicarbonate will reflect the influence of low cytoplasmic concentrations of KG. In that regard, it is relevant to note that exposure of intact rabbit proximal tubules to nominally bicarbonate-free (HEPES-containing) buffers reduces the rate of peritubular transport of PAH (3, 4) and FL (Welborn, unpublished results). Thus, despite previous observations showing that micromolar concentrations of KG can stimulate renal OA transport (17), it was difficult to assess the extent to which DC transport stimulates OA transport under physiological conditions.

The present observations were made using concentrations of KG that proximal tubules are exposed to in vivo (23). Exposing tubules to 10 µM KG significantly stimulated FL uptake under all conditions tested (Figs. 4 and 5). The condition most representative of the in vivo condition, i.e., cis-trans exposure of tubules to 10 µM KG in a bicarbonate buffer, increased the initial rate of FL uptake by 76% over that measured in the absence of exogenous KG (Fig. 4), indicating that uptake of exogenous KG can support 43% of total uptake of FL (i.e., 76%/176%).

The second conclusion supported by this study is that activity of the peritubular Na-DC cotransporter significantly enhances the renal uptake of OAs, even in the absence of exogenous KG. The tertiary active transport model for renal OA transport includes the recycling of KG across the membrane as the Na-DC cotransporter and the OA/DC exchanger operate in parallel (18, 22). However, there has been no evidence indicating the extent to which this functional interaction occurs under physiological conditions. Removal of Na⁺ from the external medium, which would eliminate activity of the Na-DC cotransporter, almost completely eliminated FL uptake (Fig. 6A), consistent with the effect of Na⁺ removal on OA transport reported in previous studies (12, 34). However, it is not apparent the degree to which this inhibition was influenced by the effect of Na⁺ removal on other cellular processes. The 25% inhibition of FL transport produced by a selective inhibition of the Na-DC cotransporter, although more modest than that produced by Na⁺ removal, is probably a more accurate indicator of the extent to which activity of Na-DC cotransporter results in the reaccumulation of KG lost from proximal cells through the OA/DC exchanger. This result also supports the contention that recycling of endogenous KG can support 25% of the peritubular transport of OAs.

The present measurements permit an estimate of the influence that Na-DC cotransport has on the rate of peritubular OA/DC exchange. Recycling of KG across the peritubular membrane supports ~25% of the "basal" activity of the OA/DC exchanger (Fig. 6B), i.e., the activity that is supported by cellular production of KG. Also, the peritubular membrane at steady state in vivo is exposed to 10 µM KG in the cis-trans- configuration, and under this configuration, KG increased the rate of FL transport by ~75% over the basal rate (Fig. 4B), presumably by increasing the cytoplasmic concentration of KG. These two observations suggest that Na-DC cotransporter activity is directly responsible for supporting ~57% [(75%+25%)/175%] of the total peritubular flux of OAs via the OA/DC exchanger. Figure 7 presents a model emphasizing the two modes of influence of the Na-DC cotransporter on the activity of the OA/DC exchanger.

View larger version (23K): [in this window] [in a new window]   Fig. 7.   Model of the organization of peritubular transporters involved in the "tertiary active" transport of FL into renal proximal tubule cells. Transporter 1 is the organic anion/dicarboxylate (OA/DC) exchanger, which accumulates extracellular FL through mediated exchange with intracellular KG. Active uptake of FL is supported by an outwardly directed KG gradient maintained by cellular metabolism and activity of the Na-DC cotransporter. Transporter 2 is the Na-DC cotransporter, which, through accumulation of exogenous KG and recycling of endogenous KG, elevates the concentration of KG at the cytoplasmic face of the OA/DC exchanger. Transporter 3 is the Na+-K+-ATPase, which maintains the inwardly directed Na+ electrochemical gradient that supports activity of the Na-DC cotransporter.

We have assumed that the stimulation of FL transport produced by exposing tubules to exogenous KG was the result of an increase in the intracellular concentration of KG and its effect on turnover of the OA/DC exchanger. This assumption is supported by many observations including the mediated exchange of PAH for KG and structurally related dicarboxylates in isolated basolateral membrane vesicles (19, 24) and the direct effect that intracellular KG concentration has on the rate of peritubular PAH transport in renal cortical slices (17). It is also unlikely that a stimulation of cell metabolism and consequent changes in cell ATP content was responsible for the increase in FL transport caused by exposure to exogenous KG. First, exposing tubules to 100 µM KG under conditions similar to those used in the present study has no effect on cell ATP content (2). Moreover, active uptake of OAs has been shown to occur under conditions that should have drastically curtailed cellular metabolism, suggesting that cell ATP content has no direct effect on peritubular OA transport (16). Therefore, an increase in OA uptake following exposure to KG is probably secondary to an increase in the concentration of KG at the cytoplasmic face of the OA/DC exchanger. As noted earlier, the rapidity with which exposure to KG stimulates FL transport (e.g., Fig. 5) necessitates that equally rapid and proportionately large increases occur in the cytoplasmic concentration of KG.

It appears relevant to discuss the implications of such changes in the light of what is known about the cytoplasmic pool of KG and rates of OA transport. The cytoplasmic KG concentration in renal proximal cells is ~200 µM (17), which is near the apparent K_t for KG interaction with the OA/DC exchanger (17). Consequently, a 54% increase in the activity of the OA/DC exchanger resulting from exposure to 10 µM KG in the cis-configuration (Fig. 4B) would require, at least, a 54% increase in the concentration of KG at the cytoplasmic face of the exchanger; i.e., an increase of 108 µmol/liter cell water. Although the kinetics of peritubular KG uptake have, to our knowledge, not been measured, Ullrich and colleagues (7) have measured the maximal rate of Na⁺-methylsuccinate cotransport in intact rat proximal tubules (7): 0.5 pmol · cm¹ · s¹ or ~1,400 µmol · liter cell water¹ · min¹. The K_i of KG's interaction with this process is 80 µM (31), from which we can infer an uptake rate of ~17.5 µmol/liter cell water from an external KG concentration of 10 µM during the 5-7 s required to resolve the cis-stimulation of FL transport (Fig. 5). While acknowledging that the rate of peritubular KG transport may be substantially greater in rabbit proximal tubules, it still seems unlikely that the accumulation of KG would be sufficiently rapid to increase the cytoplasmic concentration by more than 100 µM within a few seconds.

Alternatively, uptake of exogenous KG could elevate the local KG concentration at the peritubular face of the OA/DC exchanger. Consistent with this idea is the observation that rates of FL transport were typically higher when measured under the cis-trans configuration than under the trans-configuration (Fig. 4B), despite the fact that both configurations should have included the same intracellular concentration of KG. In other words, it appears that continuous uptake of exogenous KG supported a higher rate of transport than that arising from simply loading the entire cell with KG. This observation could be explained if Na-DC cotransport increased the local concentration of KG near the cytoplasmic face of the OA-DC exchanger. It would, however, be premature to conclude that the Na-DC cotransporter could support a standing gradient of KG concentration within the cell. Similar ideas have been suggested and rejected with respect to the potential influence of transport processes within the microvillus brush border of some epithelial cells (9). Diffusion is quite fast over distances of a few microns, making it difficult for a gradient to be sustained between two points within a cell. For membrane transport to be effective in producing a local concentration substantially larger than that in the bulk cytoplasm, some rather severe constraints must be imposed on the system involving both the rate of transport (it must be large) and the morphology of the space into which solute is transported (it must be restricted). The complex morphology of the basolateral aspect of proximal tubule cells (35) offers some interesting possibilities with respect to this idea, but much more information on transport rates and cellular morphometry will be required before conclusions can be drawn about the subcellular mechanisms of solute coupling between the peritubular transport of KG and OAs.

In conclusion, under physiological conditions, accumulation of exogenous KG into renal proximal tubules occurs at rates sufficient to significantly increase the peritubular uptake of the organic anion FL, probably due to an increase in the intracellular KG concentration. Activity of the Na-DC cotransporter also supports the peritubular OA/DC exchange by recycling cellular KG. As a result of these separate influences on activity of the OA/DC exchanger, we conclude that the Na-DC cotransporter directly supports ~60% of the renal transport of OAs.