INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

D-GLUTAMATE enhances glutamine utilization and ammonium and alanine production in cultured kidney cells by blocking L-glutamate uptake and decreasing cellular L-glutamate levels (27). Because glutamine is a major fuel for most cells (14, 25) as well as a source of metabolically generated base in the kidneys (4), regulation of this pathway is of considerable interest. In this regard, L-glutamate has long been recognized as a competitive inhibitor of the phosphate-dependent glutaminase reaction (8, 11, 23), the first committed step in the intracellular pathway for glutamine utilization as a fuel. In view of the cultured kidney cell study, we wondered whether D-glutamate might also block L-glutamate uptake and activate the glutaminase pathway in vivo, thereby presenting a strategy for metabolic regulation amenable to therapeutic manipulation. Although D-glutamate is considered an "unnatural" amino acid, it is in fact the major D-amino acid for intestinal bacteria and when administered in vivo was readily converted to D-5-oxoproline (15, 19), a finding that raises the question as to how the D-glutamate enters cells. Indeed most (7, 24) but not all (13, 20) studies have concluded that the D-isomer is not transported by the acidic amino acid transporter. Recently, however, molecular cloning and functional expression of the excitatory amino acid transporters, EAAT-1 and EAAT-3 (2), have shown transport of the D-isomer by both of these subtypes, although at a lower affinity than that exhibited for the L-isomer; note that the D-isomer can apparently be transported at, or even above, the maximal rate of the L-isomer by EAAT-1. In line with transport of D-glutamate in vivo, Samarzija and Fromter (20) monitored renal tubular cell depolarization as an index of transport and observed that D-glutamate was more effective in depolarizing the cells than the L-glutamate at millimolar concentrations. Although it is unclear as to which cell surface was effecting the D-glutamate transport, these findings are consistent with D-glutamate uptake by the intact kidney and raise the possibility of competition with the L-isomer and consequently activation of the glutaminase pathway.

Transporter-mediated cotransport of acid might also modulate glutamine utilization apart from the direct glutamate effect on glutaminase activity. Indeed, expression of the cloned rabbit glutamate transporter, EAAC-1, in frog oocytes resulted in a fall in intracellular pH (9), suggesting that glutamate transport, be it the D- or L-isomer, might effect a cellular acidosis. Because cellular acidosis accelerates glutamate utilization and enhances ammoniagenesis (16, 18), D-glutamate's mode of action could be by blocking L-glutamate uptake and/or by delivering an acid load.

In light of the potential roles that glutamate transport plays in modulating in vivo glutamine metabolism, our purpose was twofold: first, to determine the extent of D-glutamate uptake by the intact kidney and the effect on uptake and cellular level of the L-isomer and, second, to demonstrate that renal glutaminase flux and ammoniagenesis are accelerated as a consequence. The results to follow show a surprisingly avid uptake of the D-glutamate at the antiluminal cell border leading to a fall in cellular L-glutamate and enhanced glutamine utilization and ammonium formation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

All experiments were performed on adult male Sprague-Dawley rats (Harlan, Indianapolis, IN) weighing between 350 and 450 g. Animals were housed one per cage and maintained on Purina Rat Chow. Food, but not water, was removed 16 h before an experiment. The animals were anesthetized with Inactin (120 mg/kg) and maintained at 37°C after insertion of a tracheal cannula and were prepared for renal extraction studies as previously described (5). After reaching equilibration, animals were either continuously infused for an additional 30 min with saline (time control) or infused with D-glutamate (neutralized with sodium bicarbonate) added to the saline vehicle and infused at the rate of 2.6 µmol/min after a 50-µmol D-glutamate prime injection. Urine was collected under oil from an indwelling bladder catheter over two consecutive 15-min periods; 0.3-ml blood samples were simultaneously drawn from the carotid artery and left renal vein at the end of each period with blood replacement from a donor littermate. Immediately after these, ~200-mg sections of the outer kidney cortex were cut and promptly homogenized in ice-cold 5% TCA for kidney cortex amino acid analysis (left kidney) and in 0.44 M sucrose, 10 mM MgCl₂, and 0.1 M Tris buffer (pH 7.2) using a polytron (half-speed for 30 s) for the ammoniagenesis from D-glutamate assay (right kidney).

For L-[¹⁴C]glutamate and L-[¹⁴C]glutamine unidirectional uptake studies, the carotid artery cannula was moved down to the junction with the aorta for direct infusion of L-[U-¹⁴C]glutamate (158 mCi/mmol, Sigma, St. Louis, MO) or L-[¹⁴C]glutamine (253 mCi/mmol, NEN, Boston, MA) into the aorta (5); [¹⁴C]glutamate was given as a priming dose, 4 µCi at 45 min followed immediately by the constant infusion at the rate of 1 µCi/min over 30 min; L-[¹⁴C]glutamine was infused at twice this rate. Arterial blood samples were drawn from the cannulated right femoral artery.

Experimental design. To discern an in vivo effect of D-glutamate on L-glutamate uptake into the kidney, two approaches were used. Removal of L-glutamate by the kidneys was determined from the arteriovenous glutamate concentration difference times renal plasma flow; this of course measures net removal, the summation of unidirectional uptake and simultaneous release. Renal plasma flow was estimated from creatinine clearance divided by the fraction of arterial creatinine removed by the kidneys as previously described (5). Arterial plasma and filtered glutamate loads were calculated from the renal plasma flow and glomerular filtration rate times arterial plasma concentration as previously described (5); fractional extraction was calculated as the net uptake divided by arterial load. Fractional reabsorption was calculated from the amount reabsorbed divided by the amount filtered. Net uptake of both glutamate and glutamine was corrected for excretory loss. To measure the unidirectional glutamate uptake, we used the L-[¹⁴C]glutamate tracer and fractional removal (arteriovenous L-[¹⁴C]glutamate difference divided by the arterial L-[¹⁴C]glutamate) times the arterial glutamate load to give unidirectional uptake; this of course only measures the unidirectional uptake of L-glutamate without the confounding simultaneous release. Tracer amounts of [¹⁴C]glutamate were infused to constant specific activity (arterial sp act 142 ± 14 and 153 ± 23 cpm/nmol at 15 and 30 min, respectively); arterial plasma glutamine specific activity was <10% of that for glutamate. We took advantage of the enzymatic determination of L-glutamate using glutamate dehydrogenase (27) to measure the effect of D-glutamate (determined as total glutamate analyzed by HPLC minus L-glutamate) on simultaneous L-glutamate uptake. Thus this design provided two independent measures of the D-glutamate effect on L-glutamate uptake: 1) chemical, expressed as net removal, and 2) isotopic, expressed as unidirectional uptake. As indexes of glutaminase flux, we used three independent measures (5): net glutamine removal, total ammonium production (combined renal venous + urine ammonium), and unidirectional glutamine uptake. Radiolabeled glutamine administered as a prime (4 µCi) followed by a constant infusion (2 µCi/min) via the aorta achieved a constant arterial plasma specific activity at 15 and 30 min (137 ± 25 and 149 ± 43 cpm/nmol at 15 and 30 min, respectively), with the arterial radiolabeled glutamine in equilibrium with the kidney cortex glutamine determined on the TCA extracts of outer cortex described above as judged by their similar specific activities measured at 30 min (149 ± 43 and 184 ± 43 cpm/nmol, respectively). We used the ratio of ammonium produced to glutamine utilized of 2 as an index for complete deamination and deamidation of the extracted glutamine during the D-glutamate infusion. Note that D-glutamate does not support ammonium production in rat kidney homogenates as measured by ammonium formed in the presence of 5 mM D-glutamate vs. that formed in the absence of exogenous substrates 150 mM KH₂PO₄ buffer (pH 7.2 alone; 1.89 ± 0.19 vs. 2.13 ± 0.04 nmol · min¹ · mg protein¹, respectively; n = 3).

Analyses. Plasma, blood, and urine concentrations of creatinine, ammonium, glutamine, and glutamate were analyzed as described (5). Glutamine and glutamate were determined in TCA extracts of plasma and kidney by HPLC after precolumn derivatization with o-phthalaldehyde (OPA). Radiolabeled glutamate and glutamine OPA derivatives (retention times 7.3 and 11.4 min, respectively) were collected under their respective peaks and detected by liquid scintillation spectrometry as described (5, 16). Recovery of stock radiolabeled L-[¹⁴C]glutamate and L-[¹⁴C]glutamine OPA derivatives added to the column and isolated under the respective peaks was 95 ± 6 and 87 ± 5%, respectively (n = 3). Homoserine (1 mM) was routinely added to the TCA extract as an internal standard. To measure L-glutamate concentration in the D-glutamate infusion studies, the previously described fluorometric-enzymatic assay was employed (27). The D-glutamate concentration was taken as the difference between total glutamate determined by HPLC and the L-glutamate determined by enzymatic assay. Recovery of L-glutamate from an equimolar (0.125 mM) mixture of L- and D-glutamate was 103 ± 1% (n = 3); although D-glutamate (Sigma) is listed as 99% pure, some 6 ± 2% of the D-glutamate reacted as the L-isomer in this assay, thus making the actual recovery closer to 97%. The HPLC-determined values for total glutamate in the absence of infused D-glutamate were assumed to represent L-glutamate, although there was in fact a 14 ± 4% discrepancy between the HPLC and enzymatic analysis of arterial plasma glutamate, suggesting the presence of the D-isomer. In contrast, L-glutamate standards gave excellent agreement between the two methods (HPLC = 102 ± 5% of enzymatic assay, n = 6, at 1 mM L-glutamate). Nevertheless, we assumed that, in the time-control group, glutamate represented solely the L-isomer and was therefore measured by HPLC.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Figure 1 shows the arterial glutamate load delivered compared with that extracted by both kidneys for the control and D-glutamate-infused groups (Table 1 presents these and other pertinent data). Infusing D-glutamate at the rate of 2.6 µmol/min resulted in an arterial plasma D-glutamate concentration of 242 ± 25 nmol/ml vs. 142 ± 9 nmol L-glutamate/ml for time control (Table 1) and a D-glutamate load to the kidney some 2.4-fold greater than time controls (2,581 ± 217 nmol D-glutamate/min vs. 1,075 ± 131 nmol L-glutamate/min, respectively) due to rise in both the arterial plasma D-glutamate concentration and renal plasma flow with D-glutamate infusion (Table 1). Although the control group's kidneys removed a respectable 33% of their arterial L-glutamate load, surprisingly those infused with D-glutamate removed an even greater percentage, 73%, of the arterial D-glutamate load (Table 1). This highly efficient extraction rivals that observed for substances, such as p-aminohippurate (PAH), which are taken up from the peritubular capillaries. Infusing D-glutamate depressed the simultaneously measured L-glutamate removal from 33 to 24% of that delivered (355 ± 38 vs. 224 ± 53 nmol/min for time controls and D-glutamate, respectively, P < 0.05; Table 1). This decrease was associated with a fall in the arterial L-glutamate concentration (142 ± 9 to 83 ± 12 nmol/ml, P < 0.05; Table 1) as well as a reduction in the arteriovenous L-glutamate concentration difference (48 ± 5 to 22 ± 6 nmol/ml, P < 0.05; Table 1). Thus the fall in extraction was associated with a lowered arterial plasma L-glutamate concentration while the arterial load was maintained (Fig. 1) by the enhanced renal plasma flow (Table 1). When the glutamate removed from the plasma was compared with the amount filtered at the glomerulus, the results become even more revealing. With the D-glutamate, the kidneys removed an amount of D-glutamate from the blood that exceeded the amount filtered by 132% (Fig. 1; 1,882 ± 227 vs. 813 ± 93 nmol/min for removal and filtration, respectively; Table 1); in the time control, L-glutamate removed exceeded the amount filtered by only 6% (355 ± 38 vs. 334 ± 35 nmol/min, respectively; Table 1). These results are consistent with D-glutamate being transported into the kidneys predominantly across the antiluminal cell surface (blood side), whereas clear-cut antiluminal L-glutamate uptake is not apparent.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Renal glutamate (Glu) handling in control and D-Glu-infused groups. Arterial plasma load delivered to kidneys and that filtered across glomerulus were calculated from renal plasma flow and glomerular filtration rate (GFR) times arterial plasma concentration, whereas amount removed was determined from arterial minus renal venous Glu concentration difference times renal plasma flow. Results are means ± SE; see Table 1 for additional pertinent data. * Difference in L-Glu removal between time control and D-Glu infused is significant at P < 0.05.

However, L-glutamate transport into the kidney cells across the antiluminal cell surface can in fact be demonstrated from the unidirectional uptake as shown in Fig. 2 and Table 2. From the fractional extraction of L-[U-¹⁴C]glutamate and the arterial L-glutamate load, one obtains the unidirectional transport of L-glutamate into the kidney (see MATERIALS AND METHODS, Experimental design for details); this effectively eliminates the confounding factor of the simultaneous release of "cold" L-glutamate, which is included when net amounts removed are determined (arteriovenous concentration difference times renal plasma flow, see MATERIALS AND METHODS, Experimental design). As depicted in Fig. 2 and in detail in Table 2, the unidirectional L-glutamate uptake into the kidney was nearly double the amount that was filtered across the glomerulus (544 ± 12 vs. 283 ± 46 nmol/min, P < 0.01) and much higher than that detected by measuring net removal (544 ± 12 vs. 297 ± 10 nmol/min, P < 0.05; Table 2). Therefore almost half of the L-glutamate normally delivered to the kidneys is taken up at the antiluminal cell surface. With D-glutamate infusion, the unidirectional L-glutamate uptake is reduced 55% (544 ± 12 to 245 ± 66 nmol/min, P < 0.02) to a value that is not different from the amount filtered at the glomerulus (Fig. 2). D-Glutamate infusion did not increase L-glutamate excretion as shown in Table 1, whereas luminal L-glutamate uptake was more avid than the D-glutamate uptake (99.7 ± 0.1 vs. 97.7 ± 0.7% fractional reabsorption, respectively) in contrast to the fractional extraction (25 ± 5 vs. 73 ± 9% fractional extraction, respectively, Table 1). These findings are consistent with the studies showing that the D-isomer does not compete with L-glutamate for transport into the cell at the luminal cell surface (6, 24).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Unidirectional Glu uptake compared with amount filtered in control and D-Glu-infused groups. L-[U-14C]Glu was infused to a constant specific activity and uptake determined as described in MATERIALS AND METHODS; filtered Glu was determined from arterial Glu concentration times GFR. Results are means ± SE from n = 4; see Table 2 for pertinent data. * Difference between D-Glu infused and control uptake significant at P < 0.05.

The fall in arterial L-glutamate concentration during D-glutamate infusion (Table 1) raises the possibility of an effect of D-glutamate on some extrarenal site, and this might cause the renal uptake to fall. Therefore L-glutamate was also infused, raising the L-isomer to concentrations well above the time-control level. As shown in Fig. 3, when the arterial plasma ratio of D- to L-glutamate increases despite an elevated L-glutamate plasma concentration, the L-glutamate uptake falls. At the lowest D- to L-glutamate ratio, 0.2 (56 and 231 nmol/ml for D- and L-glutamate, respectively), 30 and 34% of their arterial loads were extracted; at a ratio of 0.4 (110 and 273 nmol/ml, respectively), L-glutamate extraction dropped to 21 as D-glutamate rose to 44%; at the ratio of 1.2 (198 and 158 nmol/ml, respectively), L-glutamate uptake was pretty nearly abolished with 89% of the D-glutamate extracted. Note that neither L- nor D-glutamate excretion rates contributed significantly to the uptake rates (excreted glutamate being <1% of their respective uptake rates). Therefore these results demonstrate that the effect of D-glutamate on L-glutamate uptake cannot be attributed to the spontaneous fall in arterial L-glutamate concentration but instead points to a competitive interaction between these isomers in the kidney and specifically at the antiluminal surface.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Uptake of Glu isomers at different arterial concentration ratios for D- and L-Glu. Relative D-, L-Glu concentrations were (in nmol/ml) 56/231, 100/273, 186/232, and 198/158, giving arterial plasma ratios 0.2, 0.4, 0.8 and 1.25, respectively. Results are from 4 rats.

The D-glutamate-induced reduction in L-glutamate uptake results in a fall in cortical L-glutamate content, as shown in Fig. 4. With D-glutamate infusion, cortical L-glutamate content decreased 44% (37 ± 6 to 17 ± 3 nmol/mg protein, P < 0.01) with the appearance of D-glutamate (29 ± 2 nmol/mg protein). Based on some 7 µl cell water/mg protein, these values translate to 5.3, 2.4, and 4.1 mM glutamate, respectively. The decrease in L-glutamate is well within the range of modulation of the glutaminase's inhibition constant (K_i) of 5 mM for L-glutamate (23). Also shown are the cortical glutamine contents for the 2 groups; compared with the control, D-glutamate infusion reduced glutamine 27%, resulting in a reduction in the L-glutamate to L-glutamine ratio (3.7 ± 0.9 to 1.9 ± 0.4, P < 0.05) consistent with deinhibition of the glutaminase flux.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Renal cortical Glu and glutamine (Gln) content in sections of cortex obtained at end of time control and D-Glu infusions (see MATERIALS AND METHODS). Results are means ± SE from n = 13 time controls and 8 D-Glu-infused rat kidneys per group, respectively. * Difference significant from time controls at P < 0.05.

The effects of D-glutamate on glutamine utilization and ammonium production are presented in Fig. 5 and Tables 3 and 4. In the control groups, net glutamine uptake was small and much less than the total ammonium production rate (Fig. 5, Table 3), indicating that other ammoniagenic precursors are supporting the control ammonium production rate, as previously noted (4, 5). With D-glutamate, there is a sevenfold elevation in glutamine removal (170 ± 89 to 1,311 ± 219 nmol/min, P < 0.01), whereas ammonium production rises to a ratio (2,433 ± 326 / 1,311 ± 219 nmol/min) consistent with complete deamidation and deamination of the consumed glutamine. Note that although glutamine excretion increased with D-glutamate (2.5 ± 0.3 to 4.1 ± 0.7 nmol/min, P < 0.05), this would not contribute significantly to the increased net uptake. To confirm the accelerated glutaminase flux, unidirectional glutamine uptake rates were measured from the fractional extraction of [¹⁴C]glutamine. As shown in Table 4, glutamine's unidirectional glutamine uptake rate increased nearly threefold (393 ± 121 to 1,168 ± 161 nmol/min, P < 0.05) with D-glutamate and L-glutamate plasma concentrations of 269 ± 33 and 51 ± 9 nmol/ml. At these arterial D- and L-glutamate concentrations, glutamate net uptake decreased 42% (314 ± 45 to 182 ± 27 nmol/min, P < 0.05), with a large uptake of D-glutamate (1,746 ± 516 nmol/min), in agreement with the results obtained in the previous experiments and presented in Table 1. The infusion of D-glutamate tended to reduce glutamine production (377 ± 185 to 153 ± 90 nmol/min, P = 0.10) so that the net glutamine extraction would increase by 224 nmol/min. Nevertheless the accelerated rate of glutamine breakdown (373 to 1,168 or 775 nmol/min) was a far greater contributor to the enhanced net uptake (998  13, or 985 nmol/min) than the 224 nmol/min gained from the fall in glutamine synthesis.

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Renal Gln consumed and ammonium (NH+4) released by time-control and D-Glu-infused groups measured after 30 min of infusion. Gln consumed represents net uptake determined from arteriovenous concentration difference times renal plasma flow over 15- to 30-min period. NH+4 released is sum of that released into renal vein and urine (see MATERIALS AND METHODS for details). Results are means ± SE; see Table 3 for pertinent data. * Differences from time controls significant at P < 0.05.

D-Glutamate uptake was associated with a fall in urinary pH (6.56 ± 0.08 to 6.07 ± 0.06 units, P < 0.01). Because filtered bicarbonate would have increased in conjunction with the rise in glomerular filtration, the increased urinary acidification is probably due to the movement of acid into tubule cells in association with the D-glutamate and subsequent secretion of the proton.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The goals of this study were twofold: 1) to demonstrate that D-glutamate could in fact be extracted by the functioning kidney and 2) to confirm that blocking L-glutamate uptake and reducing the cellular L-glutamate would accelerate the glutaminase flux (Fig. 6).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   D-Glu transporter asymmetry and putative effects on renal Gln metabolism. D-Glu blocks antiluminal L-Glu uptake, lowering this glutaminase inhibitor and thereby activating Gln hydrolysis (1) and NH+4 production (2 and 3). D-Glu transport-associated tubule acid load (H+) would further enhance NH+4 formation from Gln's amino nitrogen. Relatively greater uptake of D-isomer at antiluminal cell surface (renal vein) as opposed to L-Glu and vice versa for luminal cell surface (urine) are consistent with Glu transporter asymmetry in functioning organ.

Although we suspected some D-glutamate extraction by the kidneys, the avidity and quantitative extent were truly remarkable. Indeed, almost 75% of the D-glutamate delivered to these kidneys was removed in a single pass, an uptake efficiency that is far greater than that seen with the L-isomer or, for that matter, greater than that for L-glutamine, a major renal fuel, and that approaches that of PAH, which is taken up from the peritubular capillaries. This large extraction was all the more surprising in that reabsorption of D-glutamate does not occur in the proximal tubule (24, 26), but D-glutamate transport does occur in the loop segment by an ill-defined system that accepts both L- and D-isomers and yet shows no competitive inhibition (6). Thus the displacement from the uptake process of the L-isomer by D-glutamate as shown in vivo (Fig. 3) as well as in vitro (27) is not likely to reflect competition with transport of L-glutamate at the luminal surface. We favor the antiluminal cell surface as the site at which D-glutamate effects the reduction in L-glutamate uptake, for the following reasons: first, significant L-glutamate transport into the kidney cells occurs across this surface, at least 50% of the L-isomer extracted entering from the blood side (see Fig. 2), and second, D-glutamate infusion decreases the L-glutamate extraction to the amount filtered that is not subject to competitive inhibition by the D-isomer. By blocking L-glutamate uptake at the antiluminal cell surface, D-glutamate could effect the measured fall in cortical L-glutamate content (Fig. 4) without inhibiting luminal L-glutamate uptake. This of course implies that uptake across both cell surfaces is important to maintain cellular glutamate concentration as proposed (20, 21).

Virtually all in vitro studies of the acidic amino acid transporter activity have reported that D-glutamate exhibits much less reactivity with the transporter than does the L-isomer of glutamate. Paradoxically, the present in vivo study suggests that whatever transporter activity is responsible for D-glutamate uptake, its reactivity toward D-glutamate in the functioning kidney is at least as great as that for the L-glutamate. Although the antiluminal transporter is undefined, the apparent competitive interaction with L-glutamate (Fig. 3) suggests that D-glutamate uptake may indeed reflect the activity of some acidic amino acid transporter. If so, the physiological conditions imposed may play a greater role in the transporter's specificity characteristics than we previously suspected (but see Refs. 13, 20). Besides basolateral glutamate transporters, organic ion transporters, i.e., for PAH or lactate, are also expressed in this membrane which could contribute to both L- and D-glutamate uptake. At present, studies are being undertaken to identify and further characterize the transporter subtype involved, i.e., EAAT-1, EAAT-2, and so forth (2). Although D-glutamate is generally considered an unnatural amino acid, it is probably generated daily in considerable amounts from intestinal bacterial flora (where it is a major component of cell walls) and found in a number of tissues, at least in the rat (10). The present study shows a surprisingly avid removal of the exogenously infused D-glutamate by the kidneys (Fig. 1). Indeed, this rate of removal (~2 µmol/min) accounts for the major fraction of the D-glutamate infused (2.6 µmol/min), whereas the rate of D-glutamate excretion is in fact insignificant (0.02 µmol/min), raising the question as to its metabolic fate. Although we did not attempt to follow its quantitative conversion to D-5-oxoproline as previously demonstrated (15, 19), we did confirm that D-glutamate is not a source of ammonium, since in neither the previous study (27) nor the present study was ammonium production increased above that produced in the absence of exogenous substrate. This would also be expected from Ratner's much earlier study (19), showing that when ¹⁵N-labeled D-glutamate was administered to rats it was recovered in the urine as D-5-oxoproline with virtually an identical ¹⁵N enrichment. According to Sekura et al. (22), this conversion of D-glutamate to D-5-oxoproline catalyzed by -glutamylcyclotransferase is the major, if not only, metabolic pathway participating in its disposal. As pointed out (22), rapid metabolic conversion prevents intracellular accumulation of the D-isomer, which is a potent inhibitor of glutathione synthesis. Indeed, -glutamylcysteine synthase exhibits a greater affinity for D-glutamate than for the L-isomer, effectively blocking glutathione synthesis (22). Consequently any buildup of D-glutamate could have effects on a host of physiological processes dependent on glutathione and glutamine homeostasis. In the present experiments endogenous D-glutamate appears to be present in rat plasma, consistent with reports of D-glutamate present in tissues (10). If so and given a K_i of 0.8 mM for D-glutamate in -glutamylcysteine synthesis (22), the potential for impaired glutathione synthesis certainly exists. It is noteworthy that we have observed an apparent increase in circulating endogenous D-glutamate in aged (24 mo) as opposed to young rats (6 mo) that appears related to diminished renal function in the former. From these findings and those of the present study, one might predict an impaired glutathione synthesis and fall in this important antioxidant with the aging process, as has in fact been reported (12).

There is no doubt that D-glutamate infusion results in a large increase in the glutaminase flux and ammonium release (see Fig. 6). There are three lines of evidence that support this: first, the sevenfold increase in net glutamine removal (Fig. 6, 1); second, the large increase in ammonium production (Fig. 6, 2 and 3); and third, the threefold increase in unidirectional glutamine uptake (Table 4). In addition to the activated glutaminase flux, a decrease in glutamine synthesis also contributes to the D-glutamate-enhanced net glutamine uptake, although clearly less significantly than the former. Although -glutamyltransferase (-GT) is present at both the apical and basal poles of proximal tubules and converts extracellular glutamine to glutamate and ammonium, D-glutamate, unlike PAH, does not activate this enzyme (26). Nevertheless the net glutamate release observed in Fig. 3 may reflect -GT-catalyzed extracellular glutamine hydrolysis and the effect of D-glutamate to block subsequent L-glutamate uptake. Mechanistically, activation of the intracellular glutaminase flux is likely to result from the reduction in L-glutamate uptake and lowered glutamate concentration (8), since cortical glutamate content dropped nearly 60% (Fig. 4). Activation of the normally suppressed glutaminase is also consistent with the fall in cortical glutamine content.

D-Glutamate may also affect ammoniagenesis by accelerating glutamate flux through glutamate dehydrogenase as a consequence of reducing the cellular pH. Because glutamate transport is associated with a fall in cellular pH (9), the pH-sensitive glutamate dehydrogenase flux would be expected to increase (18). In support of this, there is an accelerated glutamate deamination that keeps pace with glutaminase flux, since the ammonium produced-to-glutamine consumed ratio is pretty nearly 2. Note that the glutaminase flux, on the other hand, is if anything decreased by a moderate fall in cellular pH (18), so that the reduced inhibitor concentration (L-glutamate) constitutes the major driving force for accelerated glutaminase flux. In addition to the increased glutamate dehydrogenase flux, there is additional support for D-glutamate uptake delivering an acid load, namely, the drop in urinary pH. Because this occurs at a time when filtered bicarbonate is increased, a fall in urine pH indicates enhanced tubular H⁺ secretion, and this appears to be associated with D-glutamate uptake rather than the increase in glutamine uptake because the proton contribution from glutamine enters the tubule lumen as ammonium (17) and therefore would not contribute to the reabsorption of filtered bicarbonate. If supported by further experimentation, these observations provide the basis for a dual role for antiluminal D-glutamate uptake in urinary acidification (directly) and ammoniagenesis (via glutamine metabolism).

Besides effecting urinary acidification, D-glutamate uptake might also contribute indirectly to the enhanced renal plasma flow actually observed. If the apical Na⁺/H⁺ exchanger activity present in the proximal and loop segments were accelerated and led to enhance Na⁺-HCO₃ reabsorption (1), then the reduced solute load to the load-sensing macula densa might effect a relaxing of the afferent arteriole "tone," resulting in increased renal plasma flow including an elevated filtration rate. Whether this or other potential mechanisms, i.e., local vasoactive factors, account for the increase in flow remains to be determined.

Finally, these studies point to a potentially important role for glutamate transport in metabolic and perhaps functional regulation as the result of delivering an acid load at the organ sites participating in interorgan glutamate fluxes (3). Note that both the interorgan fluxes (30) as well as organ-specific transporters are under the influence of growth factors (29) and glucocorticoid hormone (5). Obviously the integrative role of these two fluxes, interorgan and cellular transport, conveys considerable importance for both acid-base and nitrogen homeostasis.