Variable sensitivity to diethylene glycol poisoning is related to differences in the uptake transporter for the toxic metabolite diglycolic acid

Abstract

Introduction/Context:

Poisonings with diethylene glycol (DEG) are characterized by acute kidney injury (AKI) and by a peripheral neuropathy. In animal studies on the toxicities of DEG and its metabolite diglycolic acid (DGA), remarkable differences in susceptibility to AKI were observed in identically-dosed rats. In those studies, only about 60% showed AKI, yet all AKI rats showed marked DGA accumulation in tissues, while no DGA accumulated in rats without injury. DGA is taken into renal cells via sodium-dependent dicarboxylate transporters (NaDCs). When NaDC-1 is inhibited or knocked down in human kidney cells, DGA uptake and toxicity are reduced. We hypothesize that the variation in sensitivity to tissue DGA retention and to DEG/DGA toxicity is explained by differential expression of NaDC-1 in rat kidneys.

Methods:

Using kidney tissue from previous studies, we performed rt-PCR analysis of NaDC-1 mRNA. In those studies, Wistar-Han rats were either gavage dosed with DEG (6 g/kg every 12 h for 7 days) or with single doses of DGA (300 mg/kg). Kidney tissue was harvested after euthanasia and preserved in formalin. Tissue slices were homogenized and RNA was isolated using an RNAstorm FFPE RNA Isolation Kit. The expression of NaDC-1 mRNA was compared between groups that showed DGA accumulation and AKI, with those that showed no DGA accumulation or toxicity.

Results:

Significantly higher expression of NaDC-1 mRNA was present in kidneys of AKI rats with DGA accumulation compared to those in rats that had no DGA in kidneys and without AKI.

Discussion:

The likelihood of AKI after dosing of rats with DEG or DGA is linked with an enhanced ability to take up DGA into renal cells via the NaDC-1 transporter. The variability in DEG toxicity in humans, as reported in epidemiological studies, may also be linked with differences in tissue uptake of DGA.

Conclusions:

Animals with AKI after exposure to DEG or DGA had higher NaDC-1 expression and greater DGA accumulation in renal tissues than animals without AKI.

INTRODUCTION

Diethylene glycol (DEG) is a colorless organic solvent that is found in industrial lubricants and chafing fuel. It has also been mistakenly used in pharmaceutical formulations as a cheaper alternative to glycerin or has been an adulterant in the procured glycerin [1]. Ingestion of these adulterated pharmaceutical preparations has resulted in several epidemic poisonings, with multiple fatalities. The hallmark sign of DEG poisoning is renal failure or acute kidney injury (AKI), while other clinical manifestations include metabolic acidosis, mild to moderate hepatotoxicity, and a delayed peripheral neuropathy [2–4]. The kidney injury observed in many patients is characterized by remarkable necrosis of the proximal tubular epithelium [5]. DEG undergoes metabolism first by alcohol dehydrogenase, eventually yielding two primary metabolites, diglycolic acid (DGA) and 2-hydroxyethoxyacetic acid (2-HEAA). A study by Besenhofer et al. [6] showed that DEG toxicity is blocked when metabolism by alcohol dehydrogenase is inhibited in rats, suggesting that it is a metabolite of DEG that is responsible for the toxicity and not the parent compound. Moreover, several studies have shown that direct DGA administration both in-vitro and in-vivo mimic the toxicity found in DEG studies, suggesting that DGA is the metabolite responsible for the toxicity [7–9]. Furthermore, DGA accumulation of up to 100-fold is found in kidney tissue after DEG administration, compared to concentrations in the blood [10].

DGA is similar in structure to succinate and other Krebs cycle intermediates, which allows it to be taken up by the same transport mechanisms. Sodium-dependent dicarboxylate transporter 1 (NaDC-1), which is found on the apical membrane of the proximal tubule cells in the kidney, has been determined to be the main transporter responsible for the uptake and concentration of DGA in the kidney [11]. Pharmacologic inhibition or molecular knockdown of NaDC-1 activity decreases the cellular uptake and toxicity of DGA in human proximal tubule cells [11], suggesting the importance of this transporter in the accumulation of DGA in kidney tissue to toxic concentrations.

Inter-individual sensitivity has been an intriguing feature of animal studies of the toxicity of DEG or of DGA. Two studies by Jamison et al. [12, 13] showed that at most 60% of animals dosed with matching weight-based doses of DEG developed neurologic injury and AKI. Interestingly, the only animals that showed toxicity were those that showed marked DGA accumulation in the kidney (and brain), as well as markers of renal toxicity such as elevated BUN and creatinine. As such, the DGA accumulation was shown to be necessary for the AKI to occur [12, 13]. Similarly, Robinson et al [8] administered single doses of DGA to rats and observed that only 60% of the animals suffered from AKI even at the highest dose. There is indirect evidence that such variability in DEG toxicity may also occur in humans, as has been reported in multiple epidemiological studies [14–16]. We conducted this study to determine if differences in renal cell uptake transporter expression were distinct in subjects with AKI versus those without AKI. We hypothesize that elevations in NaDC-1 content in certain subjects can lead to the enhanced DGA accumulation in tissues and hence to toxicity when dosed with identical doses of DEG or DGA.

MATERIALS AND METHODS

Materials.

RNAstorm FFPE RNA Isolation Kit was obtained from CellData (Fremont, California). ActB (β-actin) and SLC13a2 (NaDC-1) primers were obtained from Integrated DNA Technologies (Coralville, Iowa). From the Robinson study [8], formalin fixed kidney tissue was obtained from rats that received 300 mg/kg of DGA. At that dose, 6 animals had evidence of AKI and showed DGA accumulation, while 4 did not. As described below, analysis was conducted with paired specimens, so there were 4 analyses. From the Jamison studies [12, 13], formalin fixed kidney tissue from the rats that had received 6 g/kg/12 h of DEG for up to 7 days was obtained. Of the 16 specimens at this dose, 9 animals exhibited AKI and DGA accumulation and 7 did not; hence there were 7 paired analyses.

Summarized Animal Protocols.

For the Robinson study [8], twenty-seven adult male Wistar rats were randomly placed in one of three treatment groups including 0 mg/kg control (n=8), which received water, 100 mg/kg DGA (n=9) and 300 mg/kg DGA (n=10). Fluids were administered via oral gavage at time 0. After 48 h of observation, animals were anesthetized using isoflurane induction followed by sodium pentobarbital (50 mg/kg, i.p.). Kidneys from all treatment groups were collected and one was cut in half, with that half being sliced (1 mm) and fixed in 10% neutral buffered formalin.

For the Jamison studies [12, 13], fifty-two adult male and sixteen adult female Wistar-Han rats were randomly placed into one of seven treatment groups (water control, 4, 5, or 6 g/kg DEG every 12 h or 4, 5, or 6 g/kg every 24 h) and administered their dose via oral gavage for up to 7 days. Rats were single-housed in metabolic chambers for urine collection and monitored at 12 h intervals for signs of morbidity (decreased urine output). Animals that showed a decrease in urine volume by at least 50% were not continued on their dose schedule and were euthanized. After rats were anesthetized (isoflurane induction followed by sodium pentobarbital, 50 mg/kg, IP), kidneys were collected and one was sliced (1 mm) and fixed in 10% neutral buffered formalin.

In these studies, animals were considered to have AKI when the BUN and the plasma creatinine concentrations were above the historical standard ranges for these parameters in rats (10-33 mg/dL and 0.5-2.2 mg/dL, respectively) and also statistically above the ranges measured in the simultaneously treated control rats (14-24 mg/dL and 0.2 – 0.7 mg/dL, respectively). Typical values for AKI rats were at least four times the upper amount of the control ranges. AKI was confirmed using histopathology of the kidneys post euthanasia.

The animal protocols were approved by the Institutional Animal Care and Use Committee (Louisiana State University Health Sciences Center - Shreveport) and were in accordance with the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals.

Analysis of kidney tissue for mRNA content of NaDC-1.

Although there were some animals in the Jamison male study [12] that showed AKI with lower doses, we chose to use tissue only from the rats with the highest dose (6 g/kg/12h), because that was the only dose that produced AKI in the female study [13]. Hence, the results from all animals given the highest dose of 6 g/kg/12 h could be analyzed together. Tissues from both studies were stored in formalin at room temperature until processed. Formalin-fixed kidneys were sliced, and slices were homogenized. RNAstorm FFPE RNA Isolation Kit was used to isolate total RNA from the kidney homogenates because it is designed to efficiently extract RNA from formalin-fixed tissues. RNA was quantified and validated using a Nanodrop instrument. Quantitative Real Time-Polymerase Chain Reaction (qRT-PCR) was carried out on a Bio-Rad CFX96 Fast Real-Time PCR System using the iTaq Universal SYBR Green One-Step Kit. The fold change in NaDC-1 mRNA content was determined using the delta-delta Ct method, which uses the threshold cycles (Ct) of a housekeeping gene (β-actin) and a gene of interest (SLC13a2, which encodes NaDC-1) to determine the fold gene expression of a sample. The differences between the Ct of the housekeeping gene and the gene of interest were computed, then compared with a reference sample. The fold gene expression was calculated by 2^-(ΔΔCt) [17]. The reference samples were the animals that did not develop AKI, and the experimental samples were the animals that developed AKI. Samples from animals that had elevated BUN and creatinine concentrations, as defined above, along with renal DGA accumulation were considered to be AKI. Animals that did not have elevated markers of kidney damage and no DGA accumulation, but had received identical doses as the animals showing toxicity, were identified as animals without AKI. Since there were fewer animals without AKI in both DEG and DGA studies (because about 40% were without AKI), tissue from a rat without AKI was randomly paired with that from an AKI rat from the same study (DEG or DGA). These pairs were then analyzed simultaneously through the RNA extraction process and the RT-PCR analysis.

Statistics.

Comparisons between the AKI and without AKI groups were made using the Wilcoxon signed rank test. Data were plotted using box and whisker plots, where the whiskers denote the range of values, the top and bottom of the box represents the 75 and 25% respectively and the midline is the median value. All analyses were performed using GraphPad Prism 7.05 (La Jolla, Ca) for Windows. Tests were considered significant if p < 0.05.

RESULTS

DEG-treated animals.

Nine rats treated with DEG at 6 g/kg q12h for up to 7 days developed AKI, while 7 rats did not [12,13]. There was a 1.6-fold increase in NaDC-1 mRNA expression in the kidneys of those with AKI (p = 0.016) compared to those that without AKI (Figure 1).

Figure 1:

Fold change in mRNA of NaDC-1 in the kidney of rats treated with 6 g/kg DEG every 12 h for up to 7 days. n=7. Data analyzed by Wilcoxon signed-rank test. (p = 0.0156)

DGA-treated animals.

Six rats treated with DGA at 300 mg/kg developed AKI and four did not [8]. There was a two-fold, albeit non-significant increase in NaDC-1 mRNA expression in the kidney of rats with AKI (p = 0.12) (Figure 2). The increase in mRNA expression in the AKI group correlated with the increased accumulation of DGA in the kidney within that group.

Figure 2:

Fold change in mRNA of NaDC-1 in the kidney of rats treated with a single dose of 300 mg/kg DGA. n=4. Data analyzed by Wilcoxon signed-rank test. (p = 0.1250)

Combining animals irrespective of treatment resulted in a 1.7-fold increase in NaDC-1 mRNA expression (p = 0.001) in the kidneys of animals with AKI compared to those that did not develop AKI (Figure 3).

Figure 3:

Fold change in mRNA of NaDC-1 in kidneys of rats treated with high doses of either DEG or DGA. n=11. * Indicates significant difference from the without AKI group (Wilcoxon signed-rank test, p = 0.0010).

DISCUSSION

A pooled analysis of samples from previous studies with rats treated with DEG [12,13] or DGA [8] showed that NaDC-1 mRNA expression was increased in rats with AKI compared with those without AKI. Previous animal studies of DEG or DGA exposure found a wide variability in toxic response, even within groups exposed to the same dose of toxicant. In a single-dose DGA study, only 60% of the animals in the high dose group (300 mg/kg) showed an accumulation of DGA in the kidney and development of severe kidney injury [8]. Similarly in two studies on the repeat-dose toxicity of DEG, slightly more than half of the animals at the highest dose of DEG developed symptoms of poisoning and showed brain and kidney tissue accumulation of DGA [12, 13]. A similar variability can also be seen in human epidemiological studies of DEG poisoning, although not experimentally examined per se. The sulfanilamide epidemic of 1937 in the US resulted in 260 poisonings, although 353 people ingested the DEG-contaminated preparation [14]. In the Panama epidemic, only 119 cases of AKI were identified from over a thousand people estimated to have been exposed to the DEG-contaminated cough syrup [15]. In a cohort study in the Haiti epidemic, only 32 subjects of the total 49 ingestions showed symptoms of poisoning [16]. This study in animals indicates that the ability of tissues to take up and retain DGA may be an important factor in the variability of toxic effects. It is possible that similar variability in sensitivity to toxicity in humans may also relate to differences in the ability of tissues to take up and retain DGA.

This variability suggests that the degree of toxicity in different subjects is dependent upon factors related to the handling of DEG, specifically steps that are involved in DGA homeostasis. The two main factors that would affect this toxicity would be metabolism of DEG into DGA and subsequent transport of DGA into target tissues. Toxicity was only identified in rats with an accumulation of DGA within its target tissues, suggesting that key steps involve differences in processing of DGA. Since the Robinson study observed the differences in DGA accumulation and in toxic response after dosing with DGA itself [8], it is more likely that the cause of the variability is related to the transport of DGA into the tissues, rather than its formation from DEG. For this reason, we focused on a possible difference in DGA transport between animals using tissue from three different studies – two involving DEG and the other DGA.

DGA is a 4-carbon dicarboxylic acid, similar in structure to the citric acid cycle intermediate, succinate. Succinate is taken into kidney cells by the SLC13 family of solute carriers known as the sodium dicarboxylate transporters (NaDC-1 and NaDC-3) [18]. NaDC-1 has broad substrate specificity, with preference for four carbon dicarboxylates, such as succinate and possibly DGA. Previous studies using human proximal tubule (HPT) cells in culture have shown that treatment with a small molecule inhibitor of NaDC-1 markedly reduced the uptakes of both succinate and DGA [11]. Also, the uptake of succinate and the toxicity of DGA were reduced in cells pre-treated with siRNA to knockdown NaDC-1 function. Lastly, DGA reduced the uptake of succinate by these cells. Parallel studies examining the role of NaDC-3 were not definitive for a role of NaDC-3 in DGA uptake. Taken together, these studies indicate that DGA is likely taken into kidney cells by the NaDC-1 transporter. Hence for this study, we have specifically focused on differences in NaDC-1 transporter expression.

In the repeated dose studies by Jamison et al [12,13], only 9 of the 16 animals exposed to the highest dose of DEG developed kidney injury after DEG-administration. Despite identical doses of DEG, some animals accumulated DGA in tissues and developed kidney injury, while others did not. Analysis by rt-PCR showed a nearly two-fold increase in mRNA for SLC-13a2, which encodes NaDC-1, in the kidneys of the animals that developed AKI compared to identically-dosed animals that did not show toxicity. The higher expression of NaDC-1 would suggest an increased ability to take up DGA, which correlates with the findings that only the animals that developed AKI had an accumulation of DGA within the kidney.

In the study by Robinson et al [8], high dose animals exposed to 300 mg/kg DGA showed variable kidney toxicity, with four showing anuria and increases in BUN and creatinine, three having histopathological changes in the kidney and oliguria, and three showing little toxicity. Analysis of these kidneys by rt-PCR showed a two-fold increase in mRNA for SLC-13a2 in the animals that showed AKI. This increase was not statistically significant, but only 4 pairs of animals were available to be compared. When all animals, whether treated with DEG or DGA, were compared to their respective controls, NaDC-1 expression was twice as high in animals with marked DGA accumulation and AKI (Figure 3). This increase in NaDC-1 is consistent with an increase in DGA accumulation in the kidneys.

This study adds to the therapeutic implications as NaDC-1 being a new molecular target for the treatment of DEG toxicity. Assuming the amount of NaDC-1 is critical for controlling uptake of DGA into tissues, a small molecule inhibitor of DGA uptake via NaDC-1 would be a valuable adjunct in the treatment of DEG poisoning. Current therapy centers on supportive care such as bicarbonate to combat the acidosis and the use of extracorporeal measures to remove DEG and its metabolites (such removal is implied, although it has not really been shown experimentally) and to fix acid-base abnormalities [1]. It is also likely that inhibition of alcohol dehydrogenase, either with ethanol or fomepizole, should be therapeutic. Animal studies using fomepizole have confirmed its efficacy in blocking DEG toxicity [6], but convincing human data are lacking. Another issue with fomepizole is that it would have to be given relatively early in the poisoning in order to block the formation of the toxic metabolites. Because of difficulties in diagnosing DEG poisoning, particularly since most cases involve mass poisonings of unknown origin, it is often not possible to treat with fomepizole in human cases. Because animal studies with DEG have shown that DGA accumulates slowly in the blood, peaking at 24 h and remaining elevated through 48 h [10], there is a longer therapeutic window where inhibiting DGA uptake from blood into tissues would still be useful. By targeting DGA accumulation in the kidney, a therapy could be developed to limit target organ toxicity at later stages of poisonings. Expression of NaDC-1 in the kidney has been shown to be increased by metabolic acidosis [19], such that bicarbonate treatment to reverse the acidosis might have another benefit in that it might lower NaDC-1 expression and thus lessen DGA uptake into tissues. The advantage of such an approach is that it is essentially available today, whereas a small molecule inhibitor would probably take years to develop. However, a small molecule inhibitor of NaDC-1 would act more quickly because it would inhibit NaDC-1 directly, while bicarbonate therapy would take time to reduce NaDC-1 expression.

Study limitations

The main limitation of this study is that it measured the mRNA expression of NaDC-1 in kidney tissue instead of protein amounts. The main factor in choosing to measure mRNA expressions as a surrogate marker of the protein of interest is that previous attempts to quantitate NaDC-1 protein content using Western blotting have not produced usable blots, likely due to insufficiently selective NaDC-1 antibodies.

Another limitation would be the relatively small number of animal tissues examined. For this study, we needed to pair samples in order to use the delta delta Ct method of quantitation. As such, we needed to use one AKI kidney with one non-AKI kidney. In the existing studies, about 60% of the rats had AKI, which thus limited the number of pairs examined to the number of non-AKI animals.

CONCLUSION

This study focused on the differences in transport between identically-dosed animals that showed variability in toxicity to both DEG and its metabolite DGA. The animals that exhibited AKI showed an increased mRNA expression for NaDC-1, which could explain the increased accumulation of DGA. Since DGA accumulation in the kidney was necessary for the toxicity to occur, this suggests that the variability of toxicity seen in these studies could be attributed to differential amounts of NaDC-1 in the kidney.