INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

TROGLITAZONE is a peroxisome proliferator-activated receptor- (PPAR-) agonist (11) that exhibits insulin-sensitizing activity and attenuates hyperinsulinemia and hyperglycemia in humans with type 2 diabetes mellitus (9). Apart from the recognized effects of troglitazone on glucose metabolism, we have shown troglitazone to inhibit glutamate transamination and alanine formation while increasing ammonium formation in rat mesangial cells (29). Because acidosis has a similar effect on glutamate metabolism (33), the question arises as to whether troglitazone induces a cellular acidosis and, if so, whether this could be an early and important parameter in subsequent troglitazone effects. Cell pH is an important signal for numerous cellular processes, including glycolysis (8) and glutamine metabolism (23, 32). Potential sites through which troglitazone might influence cellular acid-base balance are depicted in Fig. 1. Because intracellular pH (pH_i) is a balance between net metabolic acid production and net acid extrusion (1, 4), any effect of troglitazone on either of these two limbs of the balance could result in an alteration in cellular pH_i and hence cell function. In this regard, we (29) as well as others (12) have shown that troglitazone enhances glucose uptake, which is coupled to increased lactate formation and therefore enhanced acid production (8). Indeed, in our previous metabolic study (29), the media bicarbonate concentration was inversely related to lactate concentration, suggesting that metabolic acid production is largely, if not entirely, attributable to anaerobic glycolysis.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Potential sites for troglitazone (Tro) action in inducing cellular acidosis. Acid formation is depicted as balanced by acid extrusion systems in maintaining a constant cytosolic pH (pHi). Acid production from glycolysis leads to lactate (Lac) production and subsequent H+ generation. Acid extrusion systems are shown as the sodium/hydrogen ion exchanger and bicarbonate-activated acid extruder/base loader. Tro may act through either or both limbs of the balance to induce a decrease in pHi.

Mesangial cells express both a sodium/hydrogen ion exchanger as well as a bicarbonate-activated acid-extruding transporter (5, 14) as shown in Fig. 1. Because the exchanger activity is allosterically modulated by an increase in cytosolic hydrogen ion concentration (3), an increase in acid production and a fall in pH_i should upregulate the rate of acid extrusion. As a consequence, the decrease in pH_i would be far less than that predicted based solely on increased acid production. On the other hand, if troglitazone inhibits either one or both of the acid-extruding mechanisms while increasing metabolic acid production, then the resulting drop in pH_i is likely to be severe and physiologically significant, that is, to have effects on glutamine metabolism and cellular processes.

To test whether troglitazone would in fact induce a cellular acidosis, we loaded mesangial cells with the fluorescent pH indicator 2'7'-bis(2-carboxyethyl)-5,6-carboxyfluorescein (BCECF) and exposed them to troglitazone over a period sufficient to monitor both the above metabolic end products as well as pH_i. In addition, we assessed the ability of mesangial cells to extrude acid following an exogenous acid load. The results to follow show that troglitazone produces a profound cellular acidosis without increasing acid production but with a marked inhibition of the acid extrusion.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Glomerular mesangial cells were obtained from Sprague-Dawley rat kidney glomeruli isolated by differential sieving as previously described (28). Mesangial cells were grown to confluency in RPMI 1640 media (GIBCO, Life Technologies, Rockville, MD) containing 17% fetal calf serum. After confluency, the cells were harvested and seeded onto specially designed 30-mm chambers (Bioptechs, Biological Optical Technologies, Butler, PA) equipped with a heating element and cap port for O₂:CO₂ aeration. The chambers were placed uncapped inside a 60-mm covered tissue culture dish and incubated at 37°C and 5% CO₂ until confluent. Cells were then transferred to the stage of an Olympus IMT-2 microscope equipped with a heated stage insert (model TIS04201501; Kent Scientific). A glass pipette directed a stream of 5% CO₂-95% O₂ into the cap port after having been bubbled through distilled water. Fluorescent measurements were made at 37°C through an inverted epifluorescence microscope with a UV-F ×40 objective using a Photon Technology International (Brunswick, NJ) RM-D microspectrofluorometer outfitted for photometric ratio fluorescence studies. After an autofluorescence measurement made in RPMI media minus phenol red media (GIBCO BRL, Rockville, MD), the cells were loaded with the pH-sensitive fluorescent dye BCECF acetoxymethyl ester (BCECF-AM; 5 mM in DMSO stock, Molecular Probes, Eugene, OR) dissolved in RPMI media to 5 µM and added to the chambers for 25 min at 37°C. The chamber was then washed three times with RPMI, and fluorescence measurements were obtained with the cells in 0.8 ml RPMI. Light emitted from a 75-W xenon arc lamp alternately exposed cells to wavelengths of 490 and 440 nm. Excitation wavelengths were chopped (10 Hz) at 490 and 440 nm, and emissions from a minimum of three to four aggregated cells were monitored at 535 ± 25 nm using a low-pass optical filter. Instrument components and data acquisition and analysis were computer controlled using fluorescence software (FELIX). Changes in the emission ration (490/440 nm) were taken as an index of changes in pH_i. The recording periods at the various data-acquisition intervals were minimized to avoid BCECF photobleaching. Only preparations with 20-fold greater fluorescence intensity than that of the autofluorescence were used. The high K⁺/nigericin technique (31) was used to clamp pH_i to media standards of known pH (confirmed on a Corning 240 pH meter at 37°C after withdrawing the sample from the chamber) obtaining a pH calibration of the 490/440 signal ratio.

Experimental design. To observe the pH_i response to troglitazone under the conditions previously used for mesangial cell metabolic studies (29), monolayers were initially studied in RPMI 1640 media. After a 0.5-h control period over which pH_i and lactate and alanine formation were monitored, the media was replaced with RPMI 1640 containing troglitazone, and the acid production and alanine formation associated with the change in pH_i were again determined. Comparisons were then made between control and troglitazone treatment differences in pH_i, acid production, and glutamine metabolism using the Student's t-test. Acid production was estimated from the change in media lactate content measured over the 0.5-h incubation period; change in alanine formation was taken as an index of inhibition of the transamination pathway as previously demonstrated (29, 33). The concentrations of lactate and alanine in the media were measured by enzymatic and HPLC methods as previously described (23) and expressed per milligram protein. Troglitazone concentrations of 5, 10, and 20 µM were used to construct a dose-response relationship with pH_i using ANOVA and a corrected Student's t-test (Bonferroni). To test for acid extrusion capability, the cells were incubated in Krebs-Henseleit (KH) media (containing in mM: 120 NaCl, 4.7 KCl, 1.9 CaCl₂, 1.2 MgSO₄, 1.2 KH₂PO₄, 24 NaHCO₃, and 10 D-glucose at pH 7.40) and then acid loaded with a 4-min exposure to KH media in which 20 mM NaCl had been replaced with 20 mM NH₄Cl. After returning the KH media, the pH_i response was monitored for 60 s initially after the 20 s required for temperature equilibration and again at 4 and 8 min. The recovery response was taken as the change in pH_i per time interval for the 8 min required to return to the preacid load pH_i. To test for the sodium/hydrogen ion exchanger response without the bicarbonate-activated component, cells were incubated in KH media in which sodium bicarbonate had been replaced with equimolar HEPES/N-methyl-D-glucamine (pH 7.40); after establishing the baseline pH_i, the cells were acid loaded with KH-HEPES containing 20 mM NH₄Cl for 4 min, and the recovery response was monitored. The effect of troglitazone on acid extrusion was then assessed by a second acid loading with 10 µM troglitazone added to the KH recovery media after the 4-min acid load; an equivalent fall in pH_i was used to ensure equal acid loading. Time control experiments for a repetitive acid load were also performed, establishing that the recovery rates were not different for the loading periods. Differences between the control preload pH_i and pH_i measured at 1, 4, and 8 min of recovery were analyzed using ANOVA and a corrected Student's t-test (Bonferroni) while differences within the recovery periods between control and troglitazone responses (pH_i/min) were determined using a paired Student's t-test. When appropriate, based on the a priori hypotheses presented in Fig. 1, a one-tailed t-table was consulted; otherwise a two-tailed t-test was used.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Figure 2 shows the pH_i profile of mesangial cells incubated in RPMI media for 30 min followed by an exchange for RPMI containing 10 µM troglitazone for a second 30-min incubation. Monitoring of the 490/440 ratio was begun within 20 s of exchanging the media (time for temperature equilibration at 37°C) and again at 15 and 30 min of incubation. Over the 30-min control period, pH_i remained stable with the 30-min pH_i value (7.52 ± 0.03; mean ± SD) taken as the average of the 1-, 15-, and 30-min readings. Adding RPMI containing 10 µM troglitazone resulted in a prompt fall in pH_i that continued to decline throughout the initial recording period, reaching a value of 6.87 at 15 min with a rise to 6.97 at 30 min. Nigericin-containing, high-potassium solutions were then utilized for the in situ calibration of the 490/440 signal ratio. The results from four additional experiments are shown in Fig. 3A. The pH_i for cells incubated in RPMI 1640 media for 30 min averaged 7.47 ± 0.04, whereas 10 µM troglitazone induced a decrease in pH_i to an average value of 6.95 ± 0.02 (P < 0.0001). Figure 3B shows that this cellular acidosis develops without a detectable increase in lactate production (17 ± 2 and 15 ± 3 nmol · min¹ · mg protein¹ for control and 10 µM troglitazone, respectively), indicating that increased metabolic acid production is not responsible for the cellular acidosis. However, as shown in Fig. 3C, troglitazone treatment reduced alanine production by 64% (325 ± 32 to 117 ± 32 pmol · min¹ · mg protein¹, P < 0.001) consistent with inhibition of the transamination pathway (29) and reduced glutamine metabolism (33).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Representative tracing of 490/440 nm signal ratio for 0.5 h control and 0.5 h during Tro incubation. Three readings of 60-s duration were taken at 1, 15, and 30 min after adding 1 ml RPMI media or 1 ml RPMI media containing 10 µM Tro added in 0.1 µl DMSO (controls received 0.1 µl DMSO). Nigericin clamped calibration curve was performed at the end of 1 h. Aeration with 5% CO2-95% O2 and constant 37°C were maintained throughout the 1-h incubation.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effect of 10 µM Tro on pHi (A), lactate formation (B), and alanine production (C). A: average pHi from 4 experiments measured over the 30-min control (Ctl) period vs. the 30-min Tro period. B: formation of lactate over the Ctl and Tro periods. C: alanine production over the Ctl and Tro periods. Measurements were made in the same experiments with results given as means ± SE. * Different from Ctl period (P < 0.01).

Because the therapeutic plasma concentration range for troglitazone is 5-15 µM (9), we studied the steady-state cell pH measured after 30-min incubation in the RPMI media containing 17% FCS in response to 5, 10, and 20 µM troglitazone [glitazones may bind to plasma proteins (9), and therefore studies in which they are excluded may not reflect the in vivo conditions]. As shown in Fig. 4, cells in RPMI plus 17% FCS have a higher pH_i than those in RPMI minus FCS (7.68 ± 0.06 vs. 7.47 ± 0.04, P < 0.05). Incubation for 30 min in 5, 10, and 20 µM troglitazone resulted in a decrease of 0.16 ± 0.04, 0.30 ± 0.07, and 0.43 ± 0.09 pH_i units, respectively (all P < 0.05 vs. control). These results show that troglitazone induces a dose-dependent cellular acidosis even under conditions in which plasma proteins are present

View larger version (29K): [in this window] [in a new window]   Fig. 4.   pHi response of mesangial cells incubated in RPMI media plus 17% FCS and control, 5, 10, and 20 µM of Tro for 30 min. Results are means ± SE obtained averages from triplicate readings made at 1, 15, and 30 min on 12 chambers.

To assess whether troglitazone inhibits either one, or both, of the potential acid extrusion systems depicted in Fig. 1, mesangial cells were incubated in either KH solution containing bicarbonate, KH + HCO at 24 mM (Fig. 5), or KH solution in which 24 mM HEPES (Fig. 6) was substituted for bicarbonate (KH + HEPES). The pH_i in cells maintained in the physiological bicarbonate buffer (7.40 ± 0.05, n = 5) was significantly higher (P < 0.05) than the pH_i in cells incubated in HEPES buffer (7.19 ± 0.06, n = 8) as can be seen by comparing Figs. 5 and 6 (control period); this is in accord with the presence of both the sodium/hydrogen ion exchanger and the bicarbonate-activated acid extruder as depicted in Fig. 1. The cells were then acid loaded with NH, producing a similar drop in pH_i (Figs. 5 and 6) for cells incubated in KH + HCO and KH + HEPES (6.96 ± 0.08 and 6.80 ± 0.08 pH_i/min for KH + HCO and for KH + HEPES, respectively). Thereafter, the pH_i for cells in KH + HCO returned to the resting pH_i faster (pH_i/min = 0.11 ± 0.02) than those in KH + HEPES (pH_i/min = 0.06 ± 0.03, Table 1, 1-4 min, P < 0.05), consistent with both acid-extruding mechanisms operating in the physiological buffer compared with the sodium/hydrogen ion exchanger alone in the HEPES buffer in agreement with Boyarski et al. (5). Troglitazone, 10 µM, markedly inhibited the acid extrusion response in the mesangial cells incubated in the bicarbonate buffer as shown in Table 1. Over the first 1-4 min, the rate of acid extrusion fell by 73% (pH_i/min = 0.03 ± 0.02 vs. 0.11 ± 0.02 for control, P < 0.001); note that the response to a second NH₄Cl load was not different from the first acid load, pH_i/min = 0.12 ± 0.03 (n = 3, data not shown). Over the 4- to 8-min period, the acid extrusion rate slowed in the control and reversed to a negative rate in the troglitazone-treated cells (0.02 ± 0.02 vs. 0.03 ± 0.03 pH_i/min, P < 0.002), consistent with acid production driving the pH_i (Fig. 1). For the overall 8-min recovery period, the response fell from 0.06 ± 0.02 to 0 ± 0.01 pH_i/min (P < 0.0001). In the HEPES-buffered media, which limits the response to the sodium/hydrogen exchanger alone, the response to 10 µM troglitazone was also to reduce acid extrusion but less than that occurring in the KH + HCO. Over the first 1-4 min, the recovery in the presence of troglitazone was 50% slower than control (pH_i/min = 0.03 ± 0.03 vs. 0.06 ± 0.03, P < 0.05); over the 4- to 8-min period, however, the response in troglitazone-treated cells reversed to acid loading as it did in the bicarbonate buffer (pH_i/min = 0.01 ± 0.01 vs. 0.03 ± 0.02). For the overall 8-min recovery period, the response fell from 0.04 ± 0.02 to 0.01 ± 0.01 pH_i/min (P < 0.002).

View larger version (27K): [in this window] [in a new window]   Fig. 5.   pHi measured from cells incubated in Krebs-Henseleit media (solid bar) following acid loading in control media and media containing 10 µM Tro. Cells were loaded with acid by exposure to 20 mM NH4Cl (open bar) for 4 min after which this media was replaced with control or Tro-containing media and the pHi was measured for 30 s at 1, 4, and 8 min (striped bars). See Experimental design for details. Results are means ± SE from 5 chambers. * Different from control pHi at 4 and 8 min (P < 0.01).

View larger version (30K): [in this window] [in a new window]   Fig. 6.   pHi measured from cells incubated in Krebs-Henseleit minus bicarbonate media (substituted with HEPES) with the protocol as described for Fig. 5. See Experimental design for details. Results are means ± SE from 8 chambers. * Difference between control and Tro media (P < 0.01).

The presence of 2 mM glutamine added to the HEPES media accelerated the recovery response to the acid load as shown in Fig. 7. After the first pulse in media with D-glucose as the only substrate, HEPES with 2 mM L-glutamine was used in the subsequent recovery periods. As can be seen, the recovery was approximately twice as fast measured over the first 4 min (0.09 ± 0.02 vs. 0.5 ± 0.01 pH_i/min, n = 4, P < 0.05) in the presence of glutamine in agreement with the previous study (26). However, with 10 µM troglitazone added to the HEPES + 2 mM L-glutamine media, the recovery response was markedly reduced (0.01 ± 0.01 pH_i/min, P < 0.05). These results show that glutamine accelerates the sodium/hydrogen exchanger response to an acid load and that troglitazone inhibits this glutamine-enhanced acid extrusion as well.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Glutamine acceleration of the recovery response to acid loading and the effect of Tro to inhibit the response in the presence of glutamine. A representative 490/440 signal tracing showing mesangial cells responding to repetitive acid loads (NH) for 4 min in HEPES, HEPES + 2 mM L-glutamine (Gln), and HEPES + 2 mM L-Gln + 10 µM Tro.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Our purpose was to determine if troglitazone would produce a cellular acidosis in rat mesangial cells and, if so, whether this effect could be attributed to either an increase in acid production or a decrease in acid extrusion or to both mechanisms. To estimate pH_i, we utilized the fluorescent dye BCECF; the excitation of BCECF at 490 nm is directly related to the ambient pH, while excitation at its isosbestic excitation wavelength, 440 nm, provides an index of the BCECF present inside the cell (24). The ratios obtained under high potassium and nigericin "clamp" conditions were directly related to pH_i, consistent with these ratios providing an estimate of the pH_i under the conditions of our experimental protocol (data not shown). The BCECF intracellular distribution is cytosolic (5, 24), which is the cellular compartment most relevant to modulation of functional glutaminase and resulting glutamate metabolism (18, 27). Because we utilized each monolayer as its own control, our results are reflective of the change in pH_i rather than a "true" physiological pH_i, which may or may not be determined by this technique (24). Nevertheless, the BCECF-estimated control pH_i values in KH media approximated those previously obtained from mesangial cells incubated in a physiological saline media containing bicarbonate (5). Similarly, we observed a significantly lower pH_i in cells incubated in HEPES-buffered KH media, in agreement with the HEPES-buffered saline in the previous studies (5, 14). The higher pH estimate in bicarbonate-containing media can be attributed to the operation of both the sodium/hydrogen ion exchanger as well as the bicarbonate-activated acid extruder (5). The higher pH_i in RPMI media may reflect the presence of 2 mM L-glutamine, which enhances the sodium/hydrogen ion exchanger activity associated with glutamate metabolism via the transamination pathway in vivo and in an intestinal cell line (2, 26) and appears to play a similar role in acid extrusion from mesangial cells (Fig. 7). Interestingly, this action of glutamine can be blocked by eliminating the transamination pathway (26), an effect on glutamine metabolism also exerted by troglitazone (29, 33).

Why expect troglitazone to elicit a cellular acidosis? Our previous studies showed that troglitazone markedly inhibited glutamate transamination (29, 33) and shifted the glutamate formed in the cytosol and utilized by the alanine aminotransferase pathway (20, 23) into the glutamate dehydrogenase pathway localized within the mitochondria (29, 33). If the transport of glutamate across the inner mitochondrial membrane is accelerated by a fall in the cytosolic pH (30), then one possible explanation for the these findings would be the development of an intracellular (cytosolic) acidosis following exposure to troglitazone. To test this hypothesis, we incubated mesangial cells for successive 0.5 h in RPMI media and RPMI media containing troglitazone concentration corresponding to the conditions observed in the previous metabolic study. Under these conditions, troglitazone produced a dose-dependent reduction in the cellular pH (Figs. 3A and 4), demonstrating that troglitazone does indeed induce a spontaneous cellular acidosis. Because the functional glutaminase is active in the cytoplasmic compartment (19, 27), the glutamate formed would either undergo transamination, forming alanine, or be transported into the mitochondrial matrix and undergo deamination, forming ammonium. Consequently, the fall in alanine formation (Fig. 3C) and the increase in ammonium formation as previously shown (29, 33) support and, in turn, may be explained by the drop in pH_i actually observed (Figs. 2, 3A, and 4).

Does the cellular acidosis result from increased acid production or decreased acid extrusion? We (29) and others (12) had shown that troglitazone increases glucose uptake and lactate production, suggesting that an increase in metabolic acid production could contribute to the troglitazone-induced cellular acidosis. By measuring simultaneously both the lactate production and the cell pH, a comparison of the metabolic acid production over 30 min to the fall in cell pH was obtained. These results showed that a marked cellular acidosis develops without an increase in lactic acid production (Fig. 3, A and B). Of course, we cannot rule out an early increase in acid production that subsequently becomes inhibited by the sharper fall in pH_i (Fig. 2); however, this drop in pH_i maintained at least over 30 min would be unusual if it were driven by increased acid production alone because of the activation of the acid extrusion systems by a similar fall in pH_i induced by an exogenous acid load (Fig. 5). Indeed, the lowest spontaneous pH_i observed with troglitazone (Fig. 3A) and in the absence of an exogenous acid load was comparable to the transient low pH_i observed immediately after NH loading (Figs. 5 and 6). On the other hand, troglitazone virtually eliminated the extrusion of an exogenous acid load (Table 1) so that this impaired acid extrusion combined with the endogenous acid production suffices to account for the severe spontaneous cellular acidosis actually observed (Fig. 2).

Which acid extrusion system is inhibited by troglitazone? Mesangial cells utilize both the sodium/hydrogen ion exchanger and the bicarbonate-activated acid extruder for transporting acid out of mesangial cells (5, 6, 14) as depicted in Fig. 1. Therefore, we designed experiments to include both the bicarbonate-activated system as well as the sodium/hydrogen ion exchanger. Under physiological conditions, both systems should engage to handle an acid load, either that metabolically generated or as an exogenous NH load. In contrast, only the sodium/hydrogen ion exchanger would operate in the HEPES-buffered media. Therefore, we deployed media with and without bicarbonate to discern which of these two systems was the major site of inhibition. We challenged these cells with a standard 20 mM NH load (4, 5) for 4 min and monitored the acid extrusion response in the two different buffer systems. Although the acid load was identical under both conditions, the acid extrusion response was nearly twice as fast in the bicarbonate as in the HEPES-buffered media (Table 1) as expected from the presence of the dual systems depicted in Fig. 1. Given these two potential sites of inhibition, one might expect a slowing of the response in the bicarbonate media and the elimination of the response in the HEPES-buffered media if the sodium/hydrogen ion exchanger(s) was the sole site of inhibition. Alternately, one of these systems might compensate in response to a lower pH_i so that acid extrusion would only momentarily be slowed. Surprisingly, troglitazone markedly inhibited the acid recovery response in both conditions (Table 1). In fact, over the first 4 min when the recovery response was inhibited by 73% in the bicarbonate buffer, the response in the HEPES was reduced only 49%; after 4 min, the recovery response was virtually eliminated in both the bicarbonate and HEPES-buffered media. These findings point to a greater sensitivity of the bicarbonate-activated acid extrusion system than the sodium/hydrogen ion exchanger to troglitazone's action (10). Nevertheless, inhibition of both the acid-extruding systems appears to underlie the profound decrease in the spontaneous pH_i that accompanies troglitazone exposure. The sodium/hydrogen ion exchanger isoform(s) inhibited remain to be determined, but studies in the proximal tubule-like oppossum kidney-derived OK cell line, which expresses only the apical membrane NHE3 isoform (15), would be of some interest. In the 16-h metabolic studies in which alanine and ammonium formation from glutamine's amino nitrogen are inversely related, a troglitazone-induced fall in pH_i could readily explain these events. Whether the spontaneous cellular acidosis induced by troglitazone is maintained or acts as an early signal for subsequent adaptive responses (34) remains to be determined.

Is there any potential clinical significance of the cellular acidosis induced by troglitazone? We know that in the Zucker obese fatty rat model of type II diabetes mellitus that these animals develop a glomerulosclerosis that is indistinguishable from that observed in human type II diabetes (21) and in other forms of renal fibrosis occurring in diabetes and hypertension (13), as well as end-stage renal disease (ESRD) (19). In both the animal models of type II diabetes and ESRD, the chronic administration of troglitazone halts the mesangial expansion and as a functional correlate reverses the proteinuria (Fig. 8A and Refs. 19 and 21); associated with the halting of the mesangial and tubular interstitial matrix expansion is an enhanced ammonium excretion (Fig. 8B) consonant with a troglitazone-induced cellular acidosis and shift in glutamine metabolism from supporting protein synthesis to supporting ammoniagenesis (26) in mesangial and tubular cells (33). These results point to a cellular acidosis as a suppressing factor in inhibiting collagen type 1 (28) and laminin (33) production in the kidneys and potentially preventing renal fibrosis (13) in other forms of renal disease. Although the reduced matrix expansion is a potentially important beneficial effect of troglitazone, the excretion of the cation NH in place of excreted Na⁺ could lead to Na⁺ retention and the development of edema in subjects receiving chronic treatment of glitazones, as reported (9). If so, cellular acidosis as opposed to a systemic metabolic acidosis may underlie both the beneficial as well as the harmful side effects of this class of antidiabetic agents.

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Tro reverses proteinuria associated with increased ammonium excretion in experimental type II diabetes mellitus. Zucker diabetic fatty rats (untreated) and Zucker lean rats (controls) were compared with Zucker diabetic fatty rats given Tro (12 mg/g of rat chow) for 6 mo with bladder urine obtained at the time of death and assayed for creatinine, protein, and ammonium as described (16, 26). Results are means ± SE from 12 animals per group. * P < 0.001 vs. control lean group. A: protein excretion factored by creatinine excretion was increased above control in the untreated diabetic animals and reduced to control level in diabetic animals treated with Tro. B: ammonium excretion factored by creatinine excretion by the same animals with ammonium excretion increased above the untreated diabetic animals and control groups in the Tro-treated group.