Renal Glutathione: Dual Roles as Antioxidant Protector and Bioactivation Promoter

Abstract

The tripeptide glutathione (GSH) possesses two key structural features, namely the nucleophilic sulfur and the γ-glutamyl isopeptide bond. The former allows GSH to serve as a critical antioxidant and anti-electrophile. The latter allows GSH to translocate throughout the systemic circulation without being degraded. The kidneys exhibit several unique processes for handling GSH. This includes the extraction of 80% of plasma GSH, in part by glomerular filtration but mostly by transport across the basolateral plasma membrane. Studies on the protective effect of exogenous GSH are summarized, showing the different inherent susceptibility of proximal tubular and distal tubular cells and the impact on pathological or disease states, including hypoxia, diabetic nephropathy, and compensatory renal growth associated with uninephrectomy. Studies on mitochondrial GSH transport show the coordination between the citric acid cycle and oxidative phosphorylation in generating driving forces for both plasma membrane and mitochondrial carriers. The strong protective effects of increasing expression and activity of these carriers against oxidants and mitochondrial toxicants are summarized. Although GSH plays a cytoprotective role in most situations, two distinct exceptions to this are presented. In contrast to expectations, overexpression of the mitochondrial 2-oxoglutarate carrier markedly increased cell death from exposure to the nephrotoxic chemotherapeutic drug cisplatin (CDDP). Another key example of GSH serving a bioactivation role in the kidneys, rather than a detoxification role, is the metabolism of halogenated alkenes such as trichloroethylene (TCE). Although considerable research has gone into this topic, unanswered questions and emerging topics remain and are discussed.

Graphical Abstract

1. Introduction

Ever since its discovery in 1888 by de Rey-Pailhade [1], glutathione (GSH) has drawn the attention of numerous researchers in diverse fields. Moreover, a PubMed search in January 2024 using just the term “glutathione” revealed more than 180,000 hits. Although de Rey-Pailhade [1] did not describe the full structure of GSH at that time, he rightly focused on the unique presence of sulfur in the molecule, calling it philothion, from the Greek words for love and sulfur. Of course, the other unique feature of the GSH molecule is the γ-glutamyl isopeptide bond, which makes GSH resistant to degradation by normal proteases. It was not until 1948 that Binkley and Nakamura [2] demonstrated that the first step in metabolism of GSH and various GSH S-conjugates was enzymatic cleavage of the γ-glutamyl bond. The enzyme catalyzing this unique reaction was then isolated and named γ-glutamyltransferase (GGT; also called γ-glutamyl transpeptidase) [3, 4].

The primary GGT, which in humans is encoded by GGT1, is expressed at its highest levels in the kidneys but is also found in several other epithelia, including the jejunum, biliary tract, and choroid plexus of the brain. The kidneys play a key role in turnover of GSH or in metabolism of GSH S-conjugates because they are quantitatively the predominant sites in the body for localization of GGT. However, there is an interesting species difference among mammals regarding the ratio of GGT activity in the kidneys to that in the liver [5]: While activity in rat kidneys is >800-fold higher than that in rat liver, that in human kidneys is only about 15-fold higher than that in human liver. This suggests that the liver plays a quantitatively more significant, although still secondary, role in the turnover of GSH and GSH S-conjugates in humans than in species such as the rat. Regardless of this species difference in the degree of predominance of the kidneys in γ-glutamyl bond cleavage activity, the kidneys are still the primary sites for GSH / GSH S-conjugate degradation in all mammals.

While the capacity for GSH synthesis exists in most tissues, it is also highest in the liver [6]. One of the major functions of this hepatic GSH is to serve as a nucleophile to bind electrophiles and other toxic, reactive species. Although mercapturates, the “terminal” products of the classic GSH conjugation pathway, have been known for almost 150 years [7], it was not until 1959 [8, 9] that GSH was identified as the source of the sulfur atom in mercapturates. As with GSH synthesis, the liver has the greatest capacity to synthesize GSH conjugates, with cytoplasmic GSH S-transferases (GSTs) accounting for as much as 5% of total liver cytoplasmic protein [10–13]. Along with high capacities for GSH and GSH conjugate synthesis, the liver also has a high capacity to efflux both GSH and its conjugates into bile or plasma [14, 15]. Accordingly, much of the hepatic GSH and GSH conjugates are further metabolized by other tissues, most prominently the small intestines (i.e., enterohepatic circulation) and the kidneys (renal-hepatic circulation).

While this prominent role for the kidneys in GSH turnover / GSH conjugate metabolism certainly aligns with the tissue distribution of GGT activity, analysis of total renal extraction of plasma GSH during a single pass of the blood through the renal circulation was observed to be much more than could be accounted for by the process of glomerular filtration [16–21]. This led to the conclusion that besides glomerular filtration and degradation in the tubular lumen by GGT, a basolateral plasma membrane transport mechanism had to exist to account for all the extraction of plasma GSH by the kidneys. Although other epithelial tissues were also subsequently shown to possess basolateral membrane transport activity for uptake of extracellular GSH, this is an additional key aspect of the renal handling of GSH and GSH conjugates. These key aspects of renal transport and metabolism processes are schematically illustrated in Figure 1.

Figure 1.

Scheme illustrating extraction of plasma glutathione (GSH) by the kidneys.

During a single pass through the renal circulation, the kidneys extract approximately 80% of the free GSH found in plasma. Glomerular filtration can only account for 30%, meaning that the other 50% is extracted by another mechanism that occurs by transport of plasma-derived GSH across the basolateral plasma membrane. Filtered or transported GSH in the tubular lumen of the proximal tubular cell is then degraded by γ-glutamyltransferase (GGT) and dipeptidase (DP) activity to the constituent amino acids. Percentages indicate the flux of GSH through each route as it is processed by the kidneys.

The final point that needs to be introduced is the function of GSH and how the kidneys can utilize GSH in some common but also some unique ways. As noted above, the nucleophilic sulfur atom of the cysteinyl residue of GSH can readily react with toxic electrophiles, either spontaneously or via GST-mediated catalysis, thereby preventing reaction of the electrophile with key functional groups on proteins, DNA, or lipids. In this manner, GSH conjugation is viewed as a critical component of cellular antioxidant defense. While this is indeed an accurate description for most chemicals metabolized by GSH-dependent pathways, for several classes of chemicals that undergo GSH conjugation, including some drugs and environmental contaminants, this reaction actually results in bioactivation. This occurs by the production of reactive, sulfur-containing intermediates, rather than detoxification to readily excreted mercapturates.

This review will, therefore, focus on several key GSH-dependent processes in the kidneys to illustrate both the cytoprotective and bioactivation roles that exist. Studies will primarily come from the author and colleagues over the past 40 or so years and will serve as examples to illustrate the diverse function of GSH in the kidneys. Where appropriate, to either support conclusions or illustrate additional areas for research, studies from others will be discussed. Although the antioxidant and nucleophilic properties of GSH are well established, the role of membrane transport processes across both the plasma and mitochondrial membranes provides some unique features that can be modulated to produce nephroprotection from toxicants or pathological conditions. Despite many of the studies that will be discussed being published 10 to as much as 30 years ago, the novelty of the presentation lies in providing novel insights into the complexity and sometimes contradictory functions of GSH in the kidneys.

The review will discuss the role of membrane transporters and renal enzymes in the GSH-dependent bioactivation of some classes of chemicals, using the cases of the chemotherapeutic agent cisplatin (CDDP) and the halogenated alkene and environmental contaminant trichloroethylene (TCE) as examples. The mechanism of action of nephrotoxic metabolites derived from these pathways includes interactions with renal mitochondria, which may provide a source of biomarkers for early exposures to these halogenated alkenes and other chemicals. The review will then highlight some open questions and areas of uncertainty that are areas worthy of future investigation.

2. Glutathione as a Cytoprotective Agent and Antioxidant.

Mammalian kidneys are composed of approximately one million or more nephrons, their basic structural unit. Each nephron in turn is comprised of several cell types, each with distinct morphologies, biochemical properties, and physiological functions [22]. The distinctive features of each nephron cell type include differences in redox regulation and cellular energetics (i.e., mitochondrial function) [23, 24]. In a series of studies on preparations of freshly isolated [25–32] and primary cultures of rat proximal tubular (rPT) and distal tubular (rDT) cells [33, 34], we investigated the inherent biochemical properties of two major nephron cell populations to assess their susceptibility to chemically induced injury.

The protective effects of improved cellular GSH status in rPT and rDT cells is illustrated in Figure 2. The summary scheme shows the mechanism of the protective effect of exogenous GSH in each cell population after incubation with the oxidant tert-butyl hydroperoxide (tBH). It is notable that exogenous GSH results in larger increases in intracellular GSH in rPT cells as compared to rDT cells. While the mechanism for these increases in rPT cells involves transport across the basolateral plasma membrane, the increases in rDT cells occur primarily by degradation of luminal GSH, uptake of the constituent amino acids, and intracellular resynthesis of GSH. These mechanisms were supported by the ability of preincubation of rPT cells with L-γ-glutamyl-L-glutamate (γ-GluGlu), which is a competitive inhibitor of renal basolateral GSH uptake [35], to increase intracellular concentrations of GSH. The interpretation is that the intracellular GSH increase was due to exchange of accumulated γ-GluGlu for extracellular GSH. In contrast, preincubation of rDT cells with γ-GluGlu eliminated the protective effect of GSH against tBH-induced cytotoxicity. Inclusion of either L-buthionine S,R-sulfoximine (BSO) or acivicin [L-(αS,5S)-α-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid; also known as AT-125], inhibitors of GSH synthesis [36] and GSH degradation by GGT [37], respectively, eliminated the protection by exogenous GSH of rDT cells from tBH-induced cytotoxicity. This is consistent with protection in rDT cells not being primarily associated with uptake of intact GSH as it is in rPT cells.

Figure 2.

Impact of tert-butyl hydroperoxide (tBH) on glutathione (GSH) status in rat proximal tubular (rPT) and rat distal tubular (rDT) cells.

Upon exposure of rPT cells to tBH, GSH is oxidized to glutathione disulfide (GSSG) as tBH is reduced to tert-butyl alcohol and water. Addition of L-γ-glutamyl-L-glutamate (γ-GluGlu), a competitive inhibitor of GSH transport across the basolateral plasma membrane (BLM) [process 1] alters the ability of exogenous GSH to protect rPT cells from tBH-induced toxicity. In rDT cells, in contrast, protection against tBH-induced toxicity by exogenous GSH primarily involves degradation of GSH in the tubular lumen by γ-glutamyltransferase (GGT) and dipeptidase [processes 1 and 2, respectively] and reuptake of the constituent amino acids [process 4] and resynthesis of intracellular GSH. Protection of rDT cells by exogenous GSH is inhibitor by the GGT inhibitor acivicin (L-(αS,5S)-α-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid) or the GSH synthesis inhibitor BSO (L-buthionine-S,R-sulfoximine).

Various pathological or chronic disease states affecting the kidneys are also associated with oxidative stress and include alterations in renal cellular GSH status. Studies in models of three different pathological or chronic disease states will be reviewed. These include hypoxia, diabetic nephropathy, and compensatory renal hypertrophy that occurs after uninephrectomy (NPX).

Oxygen deprivation or hypoxia was modeled in suspensions of isolated rPT and rDT cells by incubating them in either 95% N₂/5% CO₂ (= 0% O₂), air (= 21% O₂), or 95% O₂/5% CO₂ [27]. rPT cells, but especially rDT cells, were most sensitive to oxidants or other agents that are known to modify thiols, when incubated in air and were least sensitive to these agents when incubated in 95% O₂/5% CO₂. Toxicants included the oxidant tBH and the thiol-alkylating quinone methyl vinyl ketone (MVK). While tBH reduces GSH by oxidation to GSSG, MVK reduces free GSH by alkylation. The conclusion was that redox status was most disturbed when cells were incubated in air because there is sufficient oxygen for oxidative processes to occur but insufficient oxygen for efficient detoxification. Little toxicity occurred in cells incubated without oxygen because oxidative processes could not occur. Finally, the 95% O2/5% CO2, which is the standard atmosphere for incubations of freshly isolated renal cells in suspension, produces an intermediate level of cellular injury because both oxidative and detoxification processes can occur with maximal catalytic efficiency.

Similar to the previously mentioned greater susceptibility of rDT cells to a thiol oxidant and an alkylating agent, these cells were also more sensitive than rPT cells to oxygen deprivation alone or addition of mitochondrial inhibitors [29]. Interestingly, this difference in sensitivity occurred despite similar extents of GSH oxidation or depletion occurring in the two cell types. This suggests that other processes besides or downstream of GSH oxidation or depletion ultimately determine cellular response to oxygen deprivation or oxidative stress.

Diabetic nephropathy, which is a common complication of both Type 1 and Type 2 diabetes, has been associated with oxidative and nitrosative stress [38–42]. To better understand the molecular effects of chronic hyperglycemia on the renal proximal tubule, male Sprague-Dawley rats were made diabetic by ip injections of streptozotocin and rPT cells were isolated from these diabetic rats and control rats [43]. Several indicators of oxidative stress, such as fluorescence of the reactive oxygen species (ROS) indicator 2,7-dichlorofluorescein, were found to be exacerbated in rPT cells from diabetic rats as compared to rPT cells from control rats. Most importantly from a toxicological point of view, incubation of diabetic rPT cells with either tBH or MVK resulted in greater increases in ROS and greater losses in cell viability than similarly incubated rPT cells from control or non-diabetic rats. Of note for the oxidative mechanism, preincubation of diabetic rPT cells with N-acetyl-L-cysteine (NAC) eliminated this geater sensitivity to oxidants. Surprisingly, although rPT cells from diabetic rats exhibited a similar decrease in ROS from NAC preincubation as cells from non-diabetic rats, the extent of protection from cytotoxicity was not as much in the cells from diabetic rats, consistent with an inherently greater sensitivity to oxidant-induced injury.

rPT cells were then isolated from rats who were diabetic for either one month or three months to assess redox status over a longer time period and to compare redox status with severity of diabetic nephropathy [44]. The most significant finding with respect to GSH status was that rPT cells from one-month diabetic rats exhibited adaptive increases in cellular concentrations of GSH that were accompanied by increases in both mitochondrial GSH transport and mitochondrial respiration. As anticipated, rPT cells from three-month diabetic rats, which exhibited more severe nephropathy, showed lower concentrations of GSH and diminished mitochondrial function.

An additional model of inherent oxidative stress associated with a chronic disease state is that of NPX and compensatory renal hypertrophy [45–47]. In the NPX model, which is a simulation of reduced nephron mass that occurs during aging or because of renal disease, the remnant kidney undergoes rapid and almost immediate changes to compensate for the lost renal function. For example, within 6 hours of removal of one kidney, the remnant kidney exhibits dramatic increases in processes such as glomerular filtration, sodium reabsorption, and mitochondrial respiration. The remnant kidney increases in size, mostly due to cellular hypertrophy (i.e., increased cell size) rather than cellular hyperplasia (i.e., increased cell number). Thus, the remnant tissue increases its level of work to compensate for the lost renal tubules. The increased workload results in an increased basal level of oxidants, primarily from the enhanced mitochondrial function. We investigated the influence of these changes on renal GSH status and found compensatory increases in renal GSH content in the cortex and outer stripe of the outer medulla and in isolated rPT cells; increases in activities were also observed in several GSH-dependent enzymes involved in redox homeostasis [48, 49].

In a similar approach to the studies described above on diabetic nephropathy, we isolated rPT cells from normal and NPX rats and studied their mitochondrial function and thiol redox status in both freshly isolated cells and primary cell cultures [50–54]. As with the diabetic nephropathy model, the rPT cells from the NPX rats showed compensatory changes in several processes involving cellular energetics and thiol redox status. Of toxicological significance, rPT cells from NPX rats were more susceptible to oxidant-induced injury despite any compensatory changes in GSH status.

Both the studies with rPT cells from rats with diabetic nephropathy and those from NPX rats illustrate the impact of changes in redox status, especially those involving the antioxidant properties of GSH. These pathological states also alter the response to oxidative stress and the susceptibility to various toxicants. As shown in Figure 3, both diabetic nephropathy and NPX with the ensuing compensatory renal growth result in increased mitochondrial activity and ROS due to increased workload. In the former case, the increased workload is due to the increased levels of glucose, which deliver an increasing amount of reducing equivalents to the mitochondria. In the latter case, the increased workload is due to the increased Na+ ion reabsorption and glomerular filtration that occur to compensate for the loss of functional nephrons. It is notable that both of these chronic disease processes converge on the mitochondria. In response to this, GSH redox status is altered such that GSH concentrations are initially elevated but as disease progresses, these concentrations decrease and result in increased susceptibility of the rPT cells to oxidant-induced injury.

Figure 3.

Scheme illustrating the progression of oxidative stress and mitochondrial dysfunction associated with diabetic nephropathy and uninephrectomy (NPX) and compensatory renal cellular hypertrophy.

Streptozotocin (STZ) was used to induce diabetes, which progresses to diabetic nephropathy (DN). Compensatory renal cellular hypertrophy (CRH) was induced by NPX. Both DN and NPK-CRH are associated with increased mitochondrial respiration, which then leads to increased production of reactive oxygen species (ROS). Compensatory changes in glutathione (GSH) status occur early in the pathological process and this causes altered susceptibility of the proximal tubular (PT) cells to toxicants.

Although GSH is found in plasma, it exists at micromolar concentrations that are much lower than those of other reductants such as albumin [55]. These relatively low concentrations raise the question of the importance of plasma GSH in maintaining redox balance. Maddipati and Marnett [56] and Takahashi et al. [57] in 1987 identified a distinct activity of glutathione peroxidase (GPX) in human plasma. Subsequent studies by Maser et al. [58] in mice and Cohen and colleagues [59–61] in human cells and tissues showed that the kidneys were the primary source of this unique extracellular GPX through secretion of the enzyme across the basolateral plasma membrane. Patients with chronic renal failure and on dialysis or those exposed to nephrotoxic drugs exhibited markedly diminished levels of plasma GPX than patients with normal renal function. Further, Cohen and colleagues [60] showed that exposure of rabbits to nephrotoxic agents known to disrupt proximal tubular function resulted in diminished levels of plasma GPX. This effect was further confirmed in pediatric patients exposed to nephrotoxicants.

While the precise role of plasma GPX remains unclear, Cohen and colleagues [60] proposed that monitoring of plasma GPX activity could have predictive value in assessing the function of transplanted kidneys or in evaluating the risk for patients from potential nephrotoxic drugs who might be especially susceptible to such injury. Further, in light of the continuous exposure of the kidneys to bloodborne oxidants, plasma GPX activity may protect the kidneys from oxidative damage. Additional studies are required to determine whether the micromolar concentrations of free GSH that are normally found in plasma are sufficient for this protective role or whether other molecules provide protection.

3. Interorgan Transport of Glutathione.

As noted above, most of the synthesis of GSH and GSH S-conjugates occurs in the liver. Because the liver is very efficient at transporting these products out of the tissue into either bile or plasma, these products enter the systemic circulation and reach other tissues, in particular the kidneys [14, 15]. Due to the process of glomerular filtration, the array of plasma membrane transporters on renal proximal tubular (PT) cells, and the exceptionally high activity of GGT on the renal PT cell brush-border plasma membrane, the kidneys play the key role in what has been termed interorgan metabolism of both GSH and GSH S-conjugates [16–21]. Plasma membrane transporters play a key role in determining this interorgan metabolism. While most synthesis of GSH and GSH S-conjugates occurs in the liver, this organ is very efficient at transporting these molecules out of the tissue into either bile or plasma by transport across the canalicular or sinusoidal plasma membranes, respectively. This initial transport step across hepatic plasma membranes is beyond the scope of this paper, as the focus here is on renal transport and metabolism; readers are referred to reviews that describe the roles of membrane carriers such as several of the multidrug resistance proteins (MRPs) and organic anion transporting polypeptide (OATP) carriers [62, 63].

As described in previous reviews [64–66], the dogma in glutathione biochemistry prior to the 1980s had been that GSH was not transported as the intact tripeptide into mammalian cells. The prominent role of the kidneys in removal of circulating GSH, much of it derived from the liver, is based on detailed analyses of plasma and tissue GSH concentrations [16–21]. A key question that was the subject of considerable controversy, however, was exactly how the kidneys so effectively extracted 80% of plasma GSH during a single pass through the renal circulation. Based on the process of glomerular filtration only accounting for 30% of the removal of plasma GSH, the obvious conclusion was that a process occurring on the basolateral plasma membrane must account for the remaining 50% of GSH that was extracted. While some investigators argued that GGT activity outside of the brush-border plasma membrane was responsible, by a mechanism involving GSH degradation and uptake of the constituent amino acids [67, 68], studies in basolateral plasma membrane vesicles from rat kidneys [35, 69, 70] and in rPT cells [30–32, 71] unambiguously demonstrated that both GSH and GSH S-conjugates are transported as the intact tripeptide or conjugate into the PT cell via multiple facilitated and active transport carriers.

The energetics and carriers that are putatively involved in the transport process are schematically summarized in Figure 4. Studies with selective transport inhibitors and assessments of ion and energy dependence support the involvement of potentially three carriers in the basolateral uptake of GSH and GSH S-conjugates, namely the sodium-dicarboxylate carrier 3 (NaC3; Slc13a3) and organic anion transporter 1 and 3 (OAT1/3; Slc22a6/8) [30–32, 35, 69, 70, 72]. The basolateral uptake process was partially inhibited by dimethylsuccinate (DMS), which is a selective inhibitor of NaC3, and by p-aminohippurate (PAH) and probenecid, which are broad-spectrum inhibitors of OATs. Cloning and expression of rat Oat3 cDNA in membrane vesicles and in a rat kidney cell line (NRK-52E cells) demonstrated the ability of OAT3 to transport GSH [73]. Transport was also significantly inhibited by incubations in a Na⁺-free transport medium [35, 69, 70], showing that the transport process was partially dependent on Na⁺ ions. Based on the inhibition by DMS and the significant Na⁺ ion dependence of transport, the NaC3, which couples uptake of dicarboxylates with 2 Na⁺ ions, was suggested as the likely Na⁺-dependent carrier responsible for GSH uptake. Note that the OATs are only indirectly Na⁺- or energy-dependent due to their function to exchange organic anions with 2-oxoglutarate (2-OG^2–), which is generated by the mitochondrial citric acid cycle. The relevance of the basolateral transporters in the protective functions of GSH in renal PT cells has been demonstrated by showing diminished protection by GSH against oxidants by co-incubation with transporter inhibitors [30–32].

Figure 4.

Renal basolateral plasma membrane carriers mediating the uptake of glutathione (GSH) and GSH S-conjugates.

Three putative carriers have been either suggested or identified as contributing to the renal proximal tubular (PT) cell uptake of GSH and GSH S-conjugates. The sodium-dicarboxylate 3 (NaC3; SLC13A3) carrier has been proposed to function due to the partial dependence of transport on Na⁺ ions and the selective inhibition by dimethylsuccinate (DMS). Organic anion transporter 1 (OAT1; SLC22A6) may potentially function and OAT3 (SLC22A8) has been shown to function in transport. The function of both carriers in based on selective inhibition by both p-aminohippurate (PAH) and probenecid and for OAT3, function is also based on activity of the cDNA-expressed carrier. Also shown is the (Na⁺+K⁺)-ATPase, which functions to create the Na⁺-ion gradient. The function of this enzyme is maintained by mitochondrial oxidative phosphorylation whereas the supply of 2-oxoglutarate (2-OG^2–) is provided by the mitochondrial citric acid cycle.

4. Mitochondrial Glutathione Transport.

In terms of regulation of redox status in the renal PT cell, the mitochondria are critical sites largely because they are the primary organelles within the cell for oxygen consumption and ROS formation. Previous work in rat liver mitochondria showed that all detectable synthesis of GSH within the hepatocyte was catalyzed in the cytoplasm [73]. Several studies in rat hepatocytes and rat liver mitochondria [74–78] also demonstrated unique properties of the mitochondrial GSH – GSSG pool and supported the suggestion that this pool was separately regulated from the pool in the cytoplasm. Subsequent studies in renal proximal tubules [80] came to similar conclusions about the separate and unique regulation of the mitochondrial pool in the kidneys. It was further shown that similar to results in rat hepatocytes, GSH synthesis in rPT cells also appears to be restricted to the cytoplasm, with no detectable synthesis within the mitochondria [81].

Considering that the concentrations of free GSH in the mitochondrial matrix are similar to those in the cytoplasm, that synthesis appears to occur exclusively in the cytoplasm, and that the GSH molecule is a zwitterion, containing both positive and negative charges (2 negative charges on carboxylate groups, a thiol group that is partially deprotonated at physiological pH, and one positive charge on an amine group), it seemed likely to us that the mitochondrial pool of GSH is derived from transport from the cytoplasm via the function of specific carrier proteins (see [82–84] for reviews). Our approach to identifying potential carrier proteins involved in transport of GSH across the mitochondrial inner membrane was similar to that described above for the basolateral plasma membrane carriers. First, based on the charges on the GSH molecule (net charge of −1 or −2 at physiological pH), our starting point was to examine the potential role of the various known organic anion and amino acid transporters [85]. Use of selective inhibitors of carrier proteins and examination of substrate specificity in studies in isolated rat renal cortical mitochondria [80, 86] and in preparations of enriched and reconstituted mitochondrial inner membrane transporters [87] indicated that two anion carriers seemed to be responsible for renal mitochondrial uptake of GSH, namely the dicarboxylate carrier (DIC, Slc25a10) and the 2-oxoglutarate carrier (OGC, Slc25a11). The DIC catalyzes the electroneutral exchange of dicarboxylates, such as malate or 2-OG²− with inorganic phosphate whereas the OGC catalyzes electroneutral exchange of various dicarboxylates, including 2-OG^2–.

The pathways for mitochondrial GSH transport are thus dependent on generation of substrates by the citric acid cycle, which are in turn coupled to mitochondrial oxidative phosphorylation. Interestingly, citric acid cycle intermediates (especially 2-OG²−) and energy from mitochondrial oxidative phosphorylation to provide ATP and maintain the cellular transmembrane Na⁺ ion gradient are also integrated to maintain cellular and mitochondrial supplies of GSH. As illustrated in Figure 5, the function of the various carriers on the basolateral plasma membrane (BLM), brush-border plasma membrane (BBM) and mitochondrial inner membrane are either directly or indirectly dependent on gradients of Na⁺ ions or 2-OG²− and supply of ATP.

Figure 5.

Scheme showing integration between plasma membrane and mitochondrial glutathione (GSH) transport processes in renal proximal tubular cells with cellular energetics and ion gradients.

The scheme illustrates putative and confirmed GSH transporters at the renal basolateral plasma membrane (BLM), the renal brush-border plasma membrane (BBM) and the mitochondrial inner membrane. GSH is presumed to be transported with a net negative charge of –2 (GS^2–). At the BLM, three carriers are possible transporters for GSH: The organic anion transporter 1 and 3 (Oat1/3; Slc22a6/8), which exchanges GS^2– for 2-oxoglutarate (2-OG^2–); and the sodium-dicarboxylate carrier 3 (NaC3; Slc13a3), which cotransports either 2-OG^2– or GS^2– with 2 Na⁺ ions. At the BBM, GS^2– can efflux into the tubular lumen by exchange for an organic anion (OA^–) by Oatp1a1 (Slc21a1) or by the ATP-dependent multi-drug resistance protein 2 or 4 (Mrp2/4; Abcc2/4). At the mitochondrial inner membrane, GS^2– is transported into the mitochondrial matrix from the cytoplasm by either the dicarboxylate carrier (DIC; Slc25a10), which exchanges GS^2– or 2-OG^2- with inorganic phosphate (Pi^2–), or the 2-oxoglutarate carrier (OGC; Slc25a11), which exchanges GS^2– for 2-OG^2- or other dicarboxylates. The (Na⁺+K⁺)-ATPase on the BLM maintains the transmembrane Na⁺ and K⁺ ion gradients at the expense of cellular ATP. This scheme is based on one presented in [75] but was modified to focus on substrate charge and driving forces for transporters.

The cytoprotective effect of mitochondrial GSH transporters was demonstrated in studies conducted in a normal rat kidney proximal tubular cell line, NRK-52E cells. These cells exhibit numerous properties of rPT cells and are useful to investigate redox processes relevant to mechanisms of toxicant-induced renal injury [88]. Overexpression of the DIC in NRK-52E cells significantly protected these cells from apoptosis and necrosis induced by the oxidant tBH or the nephrotoxicant and mitochondrial toxicant S-(1,2-dichlorovinyl)-L-cysteine (DCVC) [89]. Similarly, overexpression of the wild-type OGC protected the NRK-52E cells from both tBH and DCVC whereas overexpression of a double-cysteine mutant OGC that has markedly reduced transport activity failed to protect the cells [90]. In all cases, the extent of protection or lack of protection from oxidant or toxicant induced injury correlated with mitochondrial GSH transporter expression and activity and mitochondrial GSH concentration.

A final example of modulation of mitochondrial GSH transport expression being used to improve cellular function was a study conducted in primary cultures of rPT cells derived from rats that had undergone NPX and compensatory renal growth [91]. In this study, the primary cultures of rPT cells from NPX rats exhibited a phenotype characterized by an elevated level of oxidative stress and increased mitochondrial function and increased susceptibility to oxidants, as shown previously [51–54]. Overexpression of either the DIC or OGC resulted in a conversion of the cellular phenotype to normal, demonstrating the protective effect of increased mitochondrial GSH in a chronic pathological state.

In contrast to the above studies in which GSH was shown to have its “classic” antioxidative and protective effect, investigation of the impact of overexpression of the mitochondrial GSH transporter OGC in NRK-52E cells on chemically induced cytotoxicity was extended to a clinically used drug, cisplatin (CDDP), that is known to have nephrotoxicity as a dose-limiting side effect [92]. CDDP is considered by the World Health Organization as one of the 100 essential drugs and is used in chemotherapy [93–95]. Despite this, nephrotoxicity is a well-known side effect of its use in chemotherapy [96–100]. Numerous studies have shown the involvement of oxidative stress in CDDP-induced renal cell injury and the ability of thiol antioxidants, including NAC and GSH, to protect against the toxicity [101–104]. Moreover, the mitochondria in PT cells have been identified as sensitive and early targets for CDDP and mitochondrial dysfunction is a key mechanism in the renal cellular injury caused by CDDP [112–121].

Based on the above information that oxidative stress and mitochondrial dysfunction are critical components in CDDP-induced nephrotoxicity and that antioxidants protect, we investigated the impact of increasing mitochondrial GSH by overexpression of the OGC in NRK-52E cells, with the expectation that this would significantly protect those cells against CDDP [92]. Despite that expectation, NRK-52E cells overexpressing the OGC exhibited markedly increased cell death from exposure to CDDP (10–100 μM), both in the form of necrosis (lactate dehydrogenase release) and apoptosis. Although further study is required to confirm the mechanism of this increased injury, our hypothesis was that the increased mitochondrial GSH pool, which arose from OGC overexpression, resulted in enhanced accumulation of CDDP within the mitochondria of these cells.

5. Glutathione Conjugation as a Bioactivation Mechanism.

As noted above, conjugation of numerous electrophiles with GSH is a major protective function of GSH. GST-catalyzed formation of some GSH conjugates, however, results in bioactivation rather than detoxification [122]. A major class of chemicals that undergo bioactivation via GSH conjugation include halogenated alkenes such as the environmental contaminant and human carcinogen trichloroethylene (TCE). Besides haloalkenes, GSH conjugates of hydroquinones and isothiocyanates are examples of other diverse classes of chemicals that cause nephrotoxicity [123–125]. The present review will focus on the bioactivation of TCE as an example. Reasons include the wealth of data available on this chemical and that it illustrates very clearly several important principles that are characteristic of GSH conjugate nephrotoxicity.

Numerous reviews and toxicity assessments of TCE by various national and international organizations have been published in the past decade or so [e.g., 126–131]. TCE undergoes metabolism by either cytochrome P450 (CYP)-dependent oxidation or GST-catalyzed GSH conjugation [129]. The kidneys are one target organ for TCE, and it is metabolites derived from the GSH conjugation pathway that have been the ones that are associated with these effects [126–131]. A detailed description of the numerous studies on TCE bioactivation and the various human health risk assessments are beyond the scope of this review. Readers are referred to the aforementioned reviews and assessments for more information. The key focus here is to cite this as an example of GSH conjugation as a bioactivation mechanism.

Figure 6 broadly summarizes the overall pathway for TCE metabolism by GSH conjugation that leads to nephrotoxicity or nephrocarcinogenesis. Several steps of the pathway were confirmed in several experimental models, ranging from isolated renal cells from rats or humans to in vivo studies in rats or mice, by demonstrating that inhibition of specific steps results in protection against the adverse kidney effects.

Figure 6.

Bioactivation of trichloroethylene (TCE) via glutathione (GSH) conjugation.

This scheme summarizes how TCE is metabolized by GSH conjugation, first to S-(1,2-dichlorovinyl)glutathione (DCVG) and then to the cysteine conjugate S-(1,2-dichlorovinyl)-L-cysteine (DCVC). Studies in vivo in rats and mice and in various in vitro preparations from rat, mouse and human kidney have validated the function of each step in the different pathways for bioactivation that lead to either acute kidney injury (typically high-dose, short-term exposures) or kidney cancer (typically lower-dose, long-term exposures). Thus, the function of organic anion transporters (OATs) in the renal cellular uptake of both DCVG and DCVC was shown by protection from pretreatment with probenecid. Similarly, the requirement for metabolism of DCVG by γ-glutamyltransferase (GGT) was demonstrated by showing protection by pretreatment with acivicin. The penultimate, cytotoxic and carcinogenic agent is DCVC. The requirement for metabolism by either the cysteine conjugate β-lyase (CCBL) or the flavin-containing monooxygenase (FMO) was shown by protection by pretreatment with either aminooxyacetic acid (AOAA) or methimazole, inhibitors of CCBL and FMO, respectively.

Despite the clear consensus that TCE metabolism by the GSH conjugation pathway ultimately results in reactive metabolites that are nephrotoxic and nephrocarcinogenic, there are several unanswered questions related to this pathway. One key question that was recently resolved related to the quantitative relevance of TCE metabolism by CYP vs. GST. This was addressed by development of new methodology to simultaneously quantify and validate formation of several of the key metabolites derived by the two pathways and detailed analyses of the large database on TCE metabolism [132, 133].

Related to this issue of what are correct absolute amounts for rates of GSH conjugation is the question of how interindividual variations in metabolism in humans, such as those due to genetic polymorphisms or developmental differences, may contribute to rates of metabolism and the consequent nephrotoxicity. Little direct information is available that addresses this question. However, there are some recent animal studies and physiologically-based pharmacokinetic (PBPK) modeling studies that have been conducted to try and address this issue. Both categories of studies used multiple strains of mice or humanized mice to try and mimic some of the variability that exists in the human population. Chiu et al. [134] provided data on both CYP-dependent oxidation and GSH conjugation of TCE in 16 inbred and 1 hybrid mouse strains. Application of these data to PBPK models, particularly at low concentrations that mimic environmental exposures, allowed the authors to simulate some of the variability that exists in the human population. In a similar approach, Rusyn and colleagues [135, 136] used either different mouse strains, Cyp2e1 knockouts, or humanized mice. Their results highlighted the close relationship between metabolism in both sexes and across species and toxicity. Other work [137, 138] showed that the distinct effects of TCE on liver and kidneys are species-, strain-, and sex-dependent and are correlated with differences in patterns or rates of metabolism. How these strain-dependent differences can be related to differences in the human population, however, remain unclear.

A final example involving species-dependent variability in TCE metabolism involves a study by Commandeur and colleagues [139], who demonstrated large differences in the rates of formation and distribution among regioisomers for the GSH and cysteine conjugates of TCE in rats and humans. The toxicological significance of this involves the proportions by which metabolism results in either the 1,2-cis, 1,2-trans, or 2,2-isomers of both the GSH and cysteine conjugates of TCE [S-(1,2-dichlorovinyl)glutathione (DCVG) and DCVC, respectively]. Whereas the 1,2-cis- and 1,2-trans conjugates were readily converted to reactive species that formed crosslinks with a model nucleophile, the 2,2-isomers could not generate reactive intermediates, could not alkylate the model nucleophile, and were not toxic. Comparisons between rat and human liver showed that there is a significant species difference: While rat liver cytoplasm primarily formed 1,2-cis-DCVG, human liver cytoplasm primarily formed the relatively inactive 2,2-DCVG. However, discrepancies between rates of DCVG formation published by Commandeur et al [139] and those published by Rusyn and colleagues [135–137] still provide some uncertainties about the accuracy of predictions for humans.

Another question that requires additional studies relates to the relative importance of the cysteine conjugate β-lyase (CCBL) vs. the flavin-containing monooxygenase (FMO) in the bioactivation of DCVC to the ultimate reactive and toxic metabolite. Some data on metabolism and cytotoxicity in PT cells from both rat [140–142] and human [143–146] kidneys suggest that the CCBL has a greater role in DCVC bioactivation than the FMO in rat kidney whereas the reverse is the case for the FMO, namely that FMO appears to play a greater role than the CCBL in DCVC bioactivation in human kidney. This is based on the effectiveness of metabolism inhibitors in protecting cells (aminooxyacetic acid (AOAA) as an inhibitor of the CCBL and methimazole as an inhibitor of the FMO).

A final open question relating to TCE bioactivation by the GSH conjugation pathway is how environmental exposures to this chemical can be detected before significant and irreversible renal injury occurs. The kidneys have a significant capacity to repair damage by deleting injured cells and regenerating new cells that undergo proliferation and redifferentiation. The field of biomarker discovery has become a major focus of research in recent years, especially focusing on the kidneys [147]. With this in mind, we hypothesized that exposures to low, environmentally-relevant concentrations of TCE could result in subtle changes in the PT cell that could be detected in the extracellular medium in an in vitro model or in urine in either an in vivo model or in exposed humans [148, 149]. Thus, as cells are injured, they release certain contents, namely proteins and metabolites, that can be recovered in the extracellular space. Moreover, as noted earlier, renal mitochondria are early and sensitive targets for the key TCE metabolite DCVC. Accordingly, we further hypothesized that mitochondria would be the source of a high proportion of the detected proteins or metabolites. Research is ongoing to identify these biomarkers to better detect exposure of kidneys prior to significant toxicity.

6. Summary and Conclusions.

This review highlighted several of the diverse functions of GSH in the renal proximal tubule. Whereas the kidneys use GSH in the “classic” manner as an antioxidant and nucleophile to protect against many forms of chemically induced injury, several novel functions and processes exist in the kidneys. All of the functions of GSH can be associated with its two key structural features: The nucleophilic sulfur on the cysteinyl residue and the γ-glutamyl isopeptide bond. In the case of the sulfur atom, this is responsible for the antioxidant and anti-electrophilic roles of GSH. The γ-glutamyl isopeptide bond is what makes the GSH molecule resistant to normal proteases and enables it and GSH conjugates to translocate through the systemic circulation, primarily from the liver to the intestines or liver to the kidneys, as the intact tripeptide or intact conjugate. The novelty of this review lies in the presentation of a collection of studies over many years that highlight the diverse processes in and responses of the kidneys to modulation of GSH status. These processes include metabolism reactions and membrane transport pathways and responses include both protection against cellular injury by GSH and exacerbation of cellular injury in a GSH-dependent manner.

Several studies were summarized that demonstrated the antioxidant and cytoprotective roles of GSH in the renal proximal tubule and other nephron cell types. Included were studies on freshly isolated cells from rat kidneys showing how inherent differences in antioxidant capacity and enzyme activities in different nephron regions contribute to susceptibility to oxidant injury. Studies were also described showing how the kidneys uniquely handle exogenous GSH or GSH conjugates, involving both glomerular filtration and degradation in the tubular lumen and basolateral plasma membrane transporters. Studies on the antioxidant and cytoprotective functions of GSH in three different pathological or disease states were also described. This included studies modeling oxygen deprivation (i.e., hypoxia), diabetic nephropathy, and NPX and compensatory renal growth. Additional studies in the diabetic nephropathy model showed how early after induction of diabetes, several processes relating to GSH transport and antioxidant function are upregulated in a compensatory manner whereas later in the progression of the disease, many of these processes are downregulated.

Mitochondrial handling of GSH demonstrated the key role of GSH as the predominant non-protein thiol in the subcellular organelle in which most of the oxygen consumption and most of the ROS generation occurs. Studies were summarized that identified the primary carriers in renal cortical mitochondria responsible for the mitochondrial GSH pool, namely the DIC and OGC. Modulation of the expression and activity of these carriers were demonstrated to protect renal cells from oxidants but, surprisingly, to enhance cytotoxicity of the drug CDDP.

Perhaps the renal function of GSH that is most divergent from the classical view of GSH as an antioxidant and anti-electrophile is that of GSH conjugation serving to bioactivate certain classes of chemicals, as exemplified by the halogenated solvent TCE. Studies were briefly described that highlight the role of plasma membrane transport processes and specific enzymatic steps in generating reactive intermediates from TCE that can result in either acute kidney injury from relatively high-dose, short-term exposures or kidney cancer from relatively low-dose but long-term exposures. Some key, recent controversies and unanswered questions regarding this bioactivation process were also described. Finally, using the mechanism of action, which involves targeting mitochondria in the early phase of exposure to TCE, ongoing work is described to identify potential biomarkers of this early exposure so that renal repair and regeneration can be promoted.

It should be clear from this review that the myriad of roles for GSH in regulating kidney function and response of the kidneys to toxicant exposures or pathological or chronic disease states have been much studied over many years. Despite this, several questions remain that can serve as the focus of future research. Although modulation of intracellular GSH concentrations by targeting GSH synthesis or transport have been demonstrated, their translation to humans is unclear. With respect to GSH-dependent bioactivation of drugs or environmental chemicals, information on sex-, strain-, and species-dependent differences are only partially characterized. Moreover, information about interindividual differences in humans in metabolism and transport is lacking or minimal. Further, how specific differences correlate with incidence or risk for nephrotoxicity are unclear. Differences in other key steps in the bioactivation of nephrotoxicants among species or individuals are also lacking. Obtaining such information can improve knowledge of how these pathways lead to nephrotoxicity and can allow customization of models based on individual characteristics.

List of Abbreviations:

2-OG²‒

2-oxoglutarate

Acivicin or Aci

L-(αS,5S)-α-amino-3-chloro-4,5-dihydro-5-isoxazoleacetic acid

AOAA

aminooxyacetic acid

BBM

brush-border plasma membrane

BLM

basolateral plasma membrane

BSO

L-buthionine-S,R-sulfoximine

CCBL

cysteine conjugate β-lyase

CDDP

cisplatin

CYP

cytochrome P450

DCVC

S-(1,2-dichlorovinyl)-L-cysteine

DCVG

S-(1,2-dichlorovinyl)glutathione

DIC

dicarboxylate carrier

DMS

dimethylsuccinate

FMO

flavin-containing monooxygenase

γ-GluGlu

L-γ-glutamyl-L-glutamate

GGT

γ-glutamyltransferase

GPX

glutathione peroxidase

GSH

glutathione

GSSG

glutathione disulfide

GST

glutathione S-transferase

MRP

multidrug resistance protein

MVK

methyl vinyl ketone

NAC

N-acetyl-L-cysteine

NaC3

sodium-dicarboxylate carrier 3

NPX

uninephrectomy

OAT

organic anion transporter

OATP

organic anion transporting polypeptide

OGC

oxoglutarate carrier

PAH

p-aminohippurate

PBPK

physiologically-based pharmacokinetic

PT

proximal tubular

rDT

rat distal tubular

rPT

rat proximal tubular

ROS

reactive oxygen species

tBH

tert-butyl hydroperoxide

TCE

trichloroethylene