Mode of cytotoxic action of nephrotoxic agents: oxidative stress and glutathione-dependent enzyme


Sensuke Konno, Department of Urology, New York Medical College, Munger Pavilion 4th Floor, Valhalla, NY 10595, USA.



To investigate the cytotoxic action of nephrotoxic agents using an in vitro renal cell model, focusing on the cellular oxidative status and a specific glutathione (GSH)-dependent enzyme, glyoxalase I (Gly-I).


Renal proximal tubular LLC-PK1 cells were exposed to mercuric chloride, glycerol, cisplatin, gentamicin and cyclosporin A, and cell number/viability were determined. Oxidative stress was assessed by lipid peroxidation (LPO) assay, and Gly-I activity was measured by enzymatic method on a spectrophotometer.


Both mercuric chloride (30 µm) and glycerol (2.5%) were highly toxic to LLC-PK1 cells, inducing >90% cell death within 24 h. The remaining agents led to slightly >50% growth inhibition at 72 h. The LPO levels at 3 h in cells exposed to mercuric chloride or glycerol were ≈2.5 times higher than that in controls. N-acetylcysteine (NAC), a potent antioxidant and precursor for GSH, almost completely (>95%) prevented renal cell death from mercuric chloride or glycerol. Gly-I activity was dependent on NAC and closely associated with cell viability. A ≈65% loss in Gly-I activity by mercuric chloride/glycerol led to >90% cell death, while restoring a basal activity of Gly-I with NAC was accompanied by complete cell viability.


The cytotoxic action of nephrotoxic agents appears to be triggered by oxidative stress, leading to Gly-I inactivation. As Gly-I plays a key role in cellular detoxification, its inactivation under oxidative stress probably becomes fatal to cells. However, cytoprotection provided with NAC is significant and might have implications in preventing renal cell injury mediated through nephrotoxic agents.


glutathione (reduced)


lipid peroxidation


glyoxalase I






buthionine sulphoximine


nitric oxide.


Acute renal cell injury or failure at the cellular level has been shown to be multifocal, resulting from loss of cellular polarization, intrinsic energy deficiency, calcium overload, release of toxic proteases and oxygen free radicals, derangement of the cell cytoskeleton, and vacuolar transformation of brush-border microvilli [1]. These events could seemingly lead to irreversible cellular injury, although the exact mechanism is not fully elucidated. We are particularly interested in acute renal injuries induced by nephrotoxic agents such as HgCl2, glycerol, cisplatin, gentamicin and cyclosporin A [2,3]. The cytotoxic mechanisms of these agents have not been yet delineated, but accumulating data suggest that oxidative stress, mediated through free radicals [3] could be commonly involved in such renal injuries.

In general, reactive oxygen metabolites, including superoxide anion, hydroxyl radical, hydrogen peroxide, hypochlorous acid and singlet oxygen, are generated by incomplete reduction of molecular oxygen during the aerobic metabolic process [4]. They can cause cellular damage and dysfunction by oxidizing nucleic acids, proteins and membrane lipids [5]. Indeed, they have been shown to affect the biological process critical to various renal disease states, e.g. ischaemic renal failure, acute nephrotoxic nephritis, complement-activated glomerular injury, and drug-induced nephrotoxicity [6,7]. Thus, it is plausible that increased oxidative stress exerted by those nephrotoxic agents might play a significant adverse role in the initiation and progression of renal injury.

Fortunately, antioxidants of differing chemical nature have been reported to have beneficial or protective effects on cellular damage associated with oxidative stress [8]. Among others, these include vitamins (ascorbic acid, α-tocopherol, or retinoids), α-lipoic acid, mannitol, uric acid, and glutathione (GSH, reduced). In particular, cellular GSH has been shown to provide cells with a potent protective effect against oxidative assault [9]. GSH not only physically scavenges free radicals but also acts as an essential cofactor required for activation of GSH-dependent enzymes [10], including glutathione peroxidase, glutathione S-transferase, and glyoxalase I (Gly-I). These enzymes are known to play a vital role in the cellular antioxidative and detoxifying systems against oxidative stress and/or cytotoxic metabolites and agents [10–12].

To have an insight into the cytotoxic mechanism of nephrotoxic agents, we investigated their cellular effects on renal cells in vitro. Renal proximal tubular LLC-PK1 cells [13] were chosen for this study, because they retain many characteristics of the renal proximal tubule and have been widely used as an in vitro experimental model. Accordingly, we explored the mode of action of these toxic agents, focusing on their effects on cellular oxidative state and activity of a specific GSH-dependent enzyme. These studies might help to better understand the mechanisms of acute renal injuries induced by nephrotoxic agents. A potential usefulness of certain antioxidants in prevention and/or treatment of such renal injuries by regulating the cellular GSH level is also discussed.


Renal proximal tubular LLC-PK1 cells (American Type Culture Collection, Rockville, MD, USA) were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum, penicillin (100 units/mL), and streptomycin (100 µg/mL). For experiments, cells were seeded in six-well plates or T-75 flasks at an initial cell density of 1 × 105 cells/mL and treated with various nephrotoxic agents for specified times. Cell number and viability were then determined by the trypan blue-exclusion method and the AlamarBlue cell viability test (Biosource International, Camarillo, CA, USA), respectively. RPMI-1640, FBS, antibiotics, and all other agents were purchased from Sigma Chemical Company (St. Louis, MO) or Calbiochem (La Jolla, CA).

For the lipid peroxidation (LPO) assay we used an assay kit (Calbiochem), based on the formation of malondialdehyde (MDA), an end-product from peroxidation of polyunsaturated fatty acids in the plasma membrane. The detailed procedures were described in the vendor’s protocol, and the amount of MDA formed was determined from the MDA standards.

Gly-I activity was measured by the spectrophotometric method described by Ranganathan and Tew [14]. Cell lysates from control and agent-treated cells were prepared by freeze-thawing three times in liquid nitrogen. After preparing the reaction mixture (200 mm imidazole HCl, pH 7.0, 16 mm MgSO4, 7.9 mm methylglyoxal, 1 mm GSH), the reaction was started by adding cell lysates (40 µg). The increase in absorbance at 240 nm, due to a production of S-d-lactoylglutathione (E240 = 3.37 mm−1 cm−1), was measured with time on a spectrophotometer. The Gly-I activity was then expressed as units/mg protein, where one unit is defined to catalyse the formation of one µmole of S-d-lactoylglutathione/min under assay conditions.

For statistical analysis, all data are presented as the mean (sd) and the significance of differences between groups assessed using an unpaired Student’s t-test, with P < 0.05 considered to indicate significance.


To examine the cytotoxic effects of various nephrotoxic agents, LLC-PK1 cells were cultured with HgCl2 (30 µm), glycerol (2.5%), cisplatin (100 µm), gentamicin (5 mg/mL), and cyclosporin A (3 µm) for 72 h, and the cell number/viability determined. The concentrations used here were previously reported to have adverse effects on renal cells [1,2] and have also been confirmed to be effective in our pilot study. As a result, all five agents were capable of inducing a significant reduction in viability (P < 0.01) compared with untreated control cells. Particularly, both HgCl2 and glycerol were highly toxic, inducing >90% cell death in 24 h (Fig. 1). However, cisplatin, gentamicin and cyclosporin A were only moderately cytotoxic, leading to slightly >50% growth inhibition at 72 h (Fig. 1).

Figure 1.

Effects of nephrotoxic agents on renal cell viability. LLC-PK1 cells were cultured with HgCl2 (30 µm), glycerol (2.5%), cisplatin (100 µm), gentamicin (5 mg/mL) or cyclosporin A (3 µm). Cell viability was determined as the viable cell number at 72 h. Data are the mean (sd) from three separate experiments.

Some of the cells exposed to these agents had ‘cell blebbing’ (cytoplasmic vesicles), which was indicative of cells under oxidative stress (data not shown). Thus we assumed that oxidative stress (generation of free radicals) might be involved or play a key role in the early phase of growth inhibition and cell death induced by these nephrotoxins. Accordingly, cells were exposed to the five agents for 3 h and the LPO assay was used to assess possible membrane damage caused by oxidative stress [15]. The LPO levels were significantly (>50%; P < 0.05) higher in all agent-exposed cells than in controls, especially those exposed to HgCl2 or glycerol, at 2.5 times greater than in controls, indicating extensive plasma membrane damage (Fig. 2). Thus, these results suggest that gradual or rapid cell death caused by these agents appears to be principally due to oxidative stress, although the ‘severity’ might vary with the potency of the individual agent. Nevertheless, as both HgCl2 (30 µm) and glycerol (2.5%) had the optimum cytotoxic effect (>90% cell death) on LLC-PK1 cells, the remaining studies used these two highly effective agents.

Figure 2.

LPO assays; after cells were exposed to the indicated nephrotoxic agents for 3 h, the LPO levels were assessed by measuring the amount of MDA formed. Data are the mean (sd) from three separate experiments.

To confirm that oxidative stress is the primary cause of cell death induced by HgCl2 or glycerol, we examined whether antioxidants such as N-acetylcysteine (NAC) [16], vitamin C or mannitol could provide cytoprotection against these agents, preventing cell death. Cells were exposed to HgCl2 (30 µm) or glycerol (2.5%) in the presence of NAC (500 µm), vitamin C (200 µm) or mannitol (20 mm) for 24 h. The results then showed that NAC almost completely (>95%) prevented HgCl2/glycerol-induced cell death, whereas vitamin C and mannitol gave little cytoprotection (Fig. 3A).

Figure 3.

(A) Effects of antioxidants on cytotoxicity induced by HgCl2 or glycerol; cells were cultured with HgCl2 (30 µm) or glycerol (2.5%) in the presence of NAC (500 µm), vitamin C (200 µm) or mannitol (20 mm) for 24 h. Cell viability was determined as the percentage of viable cells relative to the control. All data are the mean of three separate experiments. (B) The effects of dithiothreitol (DTT) on HgCl2/glycerol-induced cytotoxicity, as in A, but with 100, 300 or 500 µm of dithiothreitol. Cell viability at 24 h was determined as the percentage of viable cells. All data are the mean of three separate experiments.

Because only NAC was highly effective, it was plausible that the thiol-containing nature of NAC, which was absent in vitamin C and mannitol, could account for such a potent antioxidative effect. We then tested if one of the most common thiol compounds, dithiothreitol, might also provide cytoprotection as shown by NAC. Cells were cultured with HgCl2 (30 µm) or glycerol (2.5%) in the presence of 100, 300 or 500 µm of dithiothreitol; this agent provided no protective effects against HgCl2 or glycerol, in that the higher concentration (500 µm) led to a more severe cytotoxicity than the lower (100 µm) concentration (Fig. 3B). Thus, cytoprotection provided with NAC is unlikely to be a result of having a thiol group but to other properties capable of neutralizing or diminishing HgCl2/glycerol-induced cytotoxicity.

NAC is known to be a potent antioxidant, as shown above, as well as a cell-permeable precursor for GSH, which is a cofactor required for activation of GSH-dependent enzymes [10]. This implies that cellular GSH and its related enzymes might be important in protecting renal cells from oxidative assault. To test this possibility, we examined whether buthionine sulphoximine (BSO) [17], an inhibitor of GSH synthesis, might abolish the cytoprotection provided by NAC. Varying concentrations (10, 20 and 50 µm) of BSO were added to cell cultures containing HgCl2 (30 µm)/NAC (500 µm) or glycerol (2.5%)/NAC (500 µm). Cell viability at 24 h showed that NAC cytoprotection against HgCl2 or glycerol was significantly (P < 0.03) decreased by 50–60% with 20 µm BSO, and almost completely (>95%) diminished with 50 µm BSO (Fig. 4). Even 50 µm BSO alone (without HgCl2, glycerol or NAC) led to a ≈ 50% reduction in viability within 24 h (data not shown). Thus, the reversal by BSO of the cytoprotection provided by NAC confirms that the amount or availability of cellular GSH is crucial to cell viability.

Figure 4.

Effects of BSO on NAC-provided cytoprotection against HgCl2 or glycerol; 10, 20 or 50 µm of BSO were added to cells exposed to HgCl2 (30 µm)/NAC (500 µm) or glycerol (2.5%)/NAC (500 µm), and cell viability at 24 h was assessed as the percentage of viable cells. Data are the mean (sd) of three independent experiments.

Thus we assumed that a specific GSH-dependent enzyme, e.g. Gly-I [18], might play a key role in NAC cytoprotection. Gly-I was than assayed in cells exposed for 6 h to HgCl2 (30 µm), glycerol (2.5%), NAC (500 µm), BSO (50 µm) or their combinations; the results are summarized in Table 1. Gly-I activity in NAC-exposed cells was 40% higher than that in control cells, while cells exposed to HgCl2 or glycerol lost ≈65% of Gly-I activity. However, such Gly-I inactivation by HgCl2/glycerol was fully prevented or reversed by NAC, but this reactivation of Gly-I with NAC was again completely reversed by BSO, resulting in >80% loss in Gly-I activity. BSO alone was also capable of inactivating Gly-I by ≈60% (concomitant with a ≈50% reduction in viability in 24 h). Thus, these results suggest that Gly-I appears to be critical in NAC-mediated cytoprotection against HgCl2/glycerol assault.

Table 1.  Effects of HgCl2, glycerol, NAC or BSO on Gly-I activity, as the mean of three experiments
ConditionsGly-I activity, µmol/mg protein at 6 h (% of control)
Control0.64 (100)
+ HgCl2 (30 µm)0.21 (33)
+ glycerol (2.5%)0.23 (36)
+ NAC (500 µm)0.89 (140)
+ HgCl2 (30 µm) + NAC (500 µm)0.71 (111)
+ glycerol (2.5%) + NAC (500 µm)0.75 (117)
+ HgCl2 (30 µm) + NAC (500 µm) + BSO (50 µm)0.10 (16)
+ glycerol (2.5%) + NAC (500 µm) + BSO (50 µm)0.11 (17)
+ BSO (50 µm)0.27 (42)


Although various nephrotoxic agents including HgCl2, glycerol, cisplatin, gentamicin and cyclosporin A are known to have cytotoxic effects on the renal tubular cells [2,3], the exact mechanism of such cytotoxicity remains elusive. To better understand the possible mode of action of these toxic agents, we first examined their cytotoxic effects on renal LLC-PK1 cells in vitro. Both HgCl2 and glycerol caused extensive cell death (>90%) within 24 h, while other agents required ≈72 h to gradually induce cell death. These studies showed that all agents tested are capable of killing cells, despite the differences in their cytotoxic potencies. To test the possibility that such irreversible cellular damage resulting in cell death could be due to oxidative stress, the integrity of the plasma membrane was then examined by LPO assays. Significantly increased levels of LPO, determined by formation of MDA [15], were consistently detected in cells exposed to these agents, indicating apparent membrane perturbation or damage. Particularly, the 2.5 times greater LPO levels in HgCl2- or glycerol-exposed cells indicated severe membrane disintegration, which might in part account for the rapid cell death induced by these agents. In addition, DNA analysis using agarose-gel electrophoresis further showed the smeared and degraded DNA bands from cells exposed to HgCl2 or glycerol (data not shown), which presumably resulted from oxidative stress. Thus, these nephrotoxins appear to cause oxidative stress that is pivotal in renal cell death, involving plasma membrane disintegration (at the cellular level) and DNA fragmentation (at the nuclear level).

To further verify and prevent such oxidative stress-induced cell death, three common antioxidants (vitamin C, mannitol and NAC) were tested for their antioxidative effects. Only NAC was able to prevent cell death induced by HgCl2 or glycerol, restoring >95% cell viability, while neither vitamin C nor mannitol had any protective effects. Thus, NAC appears to be a more potent antioxidant (than vitamin C or mannitol), physically and effectively scavenging harmful free radicals generated by HgCl2 or glycerol. Moreover, this NAC-specific cytoprotection suggests that cellular GSH and its related enzymes made the significant contribution. The importance of GSH was first examined using BSO as an inhibitor of cellular GSH synthesis [17]; BSO (50 µm) alone was capable of inducing a ≈50% reduction in cell viability and abrogated NAC cytoprotection (against HgCl2 and glycerol) by >95% (Fig. 4). This shows that BSO can counteract NAC, presumably depriving the cellular GSH, leading to cell death. Thus, this finding supports the substantial importance of GSH in lethal oxidative stress exerted by HgCl2 or glycerol.

It would be useful to know how HgCl2 or glycerol generate free radicals, but the underlying mechanisms have not yet been fully elucidated, although several studies, including ours, showed such nephrotoxin-induced oxidative stress as a physiological effect. It was reported that HgCl2 stimulates the generation of a specific free radical, hydrogen peroxide, in LLC-PK1 cells [19], and that such HgCl2-induced H2O2 generation was associated with increased oxygen consumption, due to the alteration of mitochondrial transmembrane potential [19]. Another in vitro study [20] also suggested that mercuric ions (Hg2+) might facilitate mitochondrial H2O2 generation in kidney cortical cells under conditions of impaired respiratory-chain electron transport. Thus, these studies imply that mercury might primarily target or have some direct effects on mitochondria, increasing H2O2 production that could further contribute to the increased cellular production of H2O2.

Similarly, glycerol is known to induce acute renal failure, involving decreased renal blood flow and intrarenal haemodynamic changes. In particular, nitric oxide (NO) is believed to be important in regulating renal haemodynamics and function [21]. Although NO could have a beneficial role it can also lead to the formation of the peroxynitrite anion, a harmful free radical, which is capable of inducing renal injury [22]. In addition, it was shown that glycerol-induced myolysis and haemolysis leads to the release of iron, that can promote free radical formation, specifically the hydroxyl radical [23]. Compared to these in vivo studies, only a few have been reported of in vitro experiments to explore the potential mechanism of glycerol-induced oxidative stress. It was reported that cisplatin, an anti-neoplastic agent, caused increased iron release in LLC-PK1 cells, resulting in increased hydroxyl radical formation [24]. Thus, it is possible that glycerol might facilitate/stimulate NO production (leading to peroxynitrite formation) and/or hydroxyl radical formation (via iron release) in vitro. Further studies are required to define the in vitro mechanism of glycerol-induced oxidative stress.

Apart from HgCl2/glycerol-induced detrimental oxidative stress, it is possible that certain GSH-dependent enzyme(s) could have been affected by the reduced or deprived GSH level, contributing to cell death. Gly-I requires GSH for its activation [18], and measuring Gly-I activity allowed us to indirectly assess the cellular GSH level. Compared to the basal Gly-I activity in control cells, the activity was increased by 40% with NAC at 6 h, while it was decreased by ≈ 60% with BSO (Table 1). Moreover, NAC-activated Gly-I in the cells exposed to HgCl2 or glycerol was again inactivated by >80% with BSO, resulting in >95% cell death. These findings thus indicate that an increase or decrease in cellular GSH (by NAC or BSO) is directly associated with Gly-I activity and cell viability. Although only Gly-I was tested here as a representative of GSH-dependent enzymes, the possible involvement of two other such enzymes, glutathione peroxidase and glutathione S-transferase [10], in cytoprotection against toxic agents cannot be excluded. More studies are required to fully elucidate the exact biochemical processes leading to cell death. Nevertheless, these results are highly suggestive that cellular GSH and its related enzyme(s) might play a critical defensive and protective role against nephrotoxic agents. For a better and clearer understanding of the present study, the effects of HgCl2 or glycerol on oxidative stress (lipid peroxidation), Gly-I activity and cell viability are summarized in Table 2.

Table 2.  Effects of HgCl2 or glycerol on LPO (as MDA formed) level, Gly-I activity and cell viability
ConditionsMDA, µm at 3 hGly-I, µmol/mg at 6 hCell viability (%) at 24 h
+ HgCl2 (30 µm)1.340.21<10
+ GLC (2.5%)1.300.23<10

In conclusion, the cytotoxic action of many nephrotoxic agents appears to be mediated through oxidative stress, accompanied by depletion of cellular GSH, leading to inactivation of Gly-I. However, such cytotoxic cell death was markedly prevented by NAC providing complete restoration of Gly-I activity. Therefore, renal cell injury due to nephrotoxins could be effectively prevented and/or treated with certain antioxidants (e.g. NAC) capable of up-regulating the cellular GSH level concomitantly with activation of GSH-dependent enzyme(s). Potential clinical studies are warranted.


None declared. Source of funding: departmental.