Possible role of ribosome biogenesis in the recovery from transient hepatic damage caused by ethylene glycol in rats

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Institute of Science Tokyo (ST) (approval No. A2023-080C) and conducted in accordance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines [17, 18]. Eight-week-old male Wistar rats (220–250 g; Oriental Yeast Co. Ltd., Tokyo, Japan) were housed under standardized conditions (12-h light/dark cycle, 25 °C) with free access to food and water. The rats were randomly assigned to either the control or the EG-treated group. Each animal received a single oral dose of either 0.9% NaCl (control) or 8 or 16 g/kg EG (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan). Samples were collected at 2 or 5 days after administration of EG (EG2 or EG5, respectively), after euthanasia with an overdose of sodium pentobarbital (40 mg/kg). Each group consisted of 6 male rats, with 3–5 animals per group used for each specific analysis. The 8 g/kg EG dose was selected based on a previous study investigating acute EG toxicity [19].

The Huh-7 human hepatoma cell line (RCB1366, RIKEN Bioresources Center, Tsukuba, Japan) was used as an in vitro model of EG-induced hepatotoxicity. Cells were maintained in Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum, 100 μg/mL streptomycin, and 100 U/mL penicillin. EG concentrations (0.3, 0.6, and 1.2 M) were selected based on a previous report [20], in which the IC50 of EG on the viability of KB (HeLa derivative) cells was estimated at 0.45 M. Cell viability was measured using the Cell Counting Kit-8 (Dojindo, Kumamoto, Japan).

Blood samples (anticoagulant: 1.5 mg/mL K₃-EDTA) were collected from the rats at 2 and 5 days after EG administration (8 g/kg). Plasma levels of albumin (ALB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), lactate dehydrogenase (LDH), and γ-glutamyl transpeptidase (γ-GTP) were measured according to the standard clinical chemistry methods (Oriental Yeast Co., Ltd., Tokyo, Japan). Blood samples collected prior to organ harvesting on days 2 and 5 were immediately analyzed for pH, bicarbonate (HCO₃⁻), and anion gap (AG) using the i-STAT®1 analyzer with EC8 + cartridges (Abbott Point of Care Inc., Princeton, NJ, USA).

Rat liver tissues were fixed in 4% paraformaldehyde, embedded in paraffin, and sliced into 2.5 µm sections. For histological analysis, the sections were stained with hematoxylin and eosin (H&E). Transmission electron microscopy (TEM) of rat liver was performed according to previously described methods [21, 22]. Briefly, approximately 1 mm³ of liver tissue was fixed in 2.5% glutaraldehyde, post-fixed in 1% osmium tetroxide, and embedded in epoxy resin (EPON). Ultrathin Sects. (80 nm) were stained with uranyl acetate and lead citrate and examined using a transmission electron microscope (HT7800, Hitachi High-Tech Co., Tokyo, Japan). Semi-thin Sects. (0.5 μm) of EPON-embedded tissue, prepared prior to ultrathin sectioning for TEM, were stained with toluidine blue and observed under both bright-field and dark-field microscopy. To visualize glycogen, periodic acid–Schiff (PAS) and periodic acid-methenamine silver (PAM) staining were performed after removing epoxy resin from EPON-embedded sections using a saturated KOH solution. Oil Red O staining of frozen sections was also conducted to visualize lipid droplets.

Total RNA from the livers of rats treated with or without EG was subjected to Clariom S array analysis (Thermo Fisher Scientific, Waltham, MA). Prior to analysis, RNA was purified using RNeasy Mini Kit (QIAGEN, Hilden, Germany), and RNA integrity was verified using a Bioanalyzer (Agilent, Santa Clara, CA). Transcriptome Analysis Console software (Thermo Fisher Scientific) was used for pathway enrichment analysis, and Metascape (https://metascape.org/gp/index.html#/main/step1) [23] was used for gene ontology (GO) enrichment analysis. Relative mRNA levels were measured using reverse transcription-quantitative PCR (RT-qPCR). Briefly, total RNA extracted from rat liver or Huh-7 cells using TRIzol reagent (Thermo Fisher Scientific) was used for cDNA synthesis. SuperScript II reverse transcriptase (Thermo Fisher Scientific) and oligo(dT)₁₅ primers were used to synthesize cDNA from total RNA. Gene-specific primers (Table S1), SYBR Green dye (Promega, Madison, WI), and the StepOne real-time PCR system (Thermo Fisher Scientific) were used for qPCR. Relative mRNA abundances were calculated using the comparative Ct method [24].

Rat liver tissues were immersed in lysis buffer [10 mM Tris–HCl (pH 8.0), 320 mM sucrose, 1 mM EDTA, 50 mM NaF, 2 mM Na₃VO₄, and protease inhibitor cocktail (Complete; Roche, Mannheim, Germany)] together with magnetic beads and disrupted using a bead crusher [3000 revolutions per minutes (rpm) for 10 s at room temperature, 3 cycles] (Multi-beads Shocker MB3000, Yasui Kikai, Osaka, Japan). The homogenates were centrifuged at 20,000 × g for 15 min to remove debris. Laemmli’s buffer was added to the supernatant [final composition is 0.06 M Tris–HCl (pH 6.8), 2% SDS, 3% 2-mercaptoethanol, 10% glycerol, and 0.005% BPB]. Huh-7 cells were collected in the Laemmli’s buffer, and lysates were obtained via sonication (Output, 40 kHz; Pulse Setting, continuous; Duration, 10 s; Temperature, room temperature) (SFX150HH, Branson, Brookfield, CT). The tissue as well as cell lysates were subjected to SDS-PAGE (15% gels for Gpx4, SLC7A11, GAPDH, phospho-eIFα, and eIFα; 10% gels for 4-hydroxy-2-nonenal) by the method of Laemmli [25], and transferred on to a PVDF membrane (Immobilon-P, Millipore, Billerica, MA) by the method of Towbin et al. [26]. Just before the use, PVDF membranes were activated through immersion in methanol for at least 20 min. For the analysis of 4-hydroxy-2-nonenal, lysates were subjected to SDS-PAGE in discontinuous (4–15%) gels (Mini-PROTEAN-TGX precast gel; Bio-Rad, Hercules, CA). The blots were blocked with 5% non-fat dry milk in Tris-buffered saline containing 0.05% Tween 20 for 30 min at room temperature (RT), then incubated with appropriate antibodies (Table S2) over night at 4 °C. The membranes were subsequently incubated with horseradish peroxidase-conjugated anti-IgG antibodies (Promega) for 30 min at RT and visualized using enhanced chemiluminescence (ECL) reagents (Western Lightning Plus, Revvity, Waltham, MA). Band intensities were quantified using Image Capture version 4 (ATTO, Tokyo, Japan).

We applied 8 and 16 g/kg EG orally to rats to assess the maximum sublethal toxicity, since the LD50 of EG was estimated at 6.1–13 g/kg when administered orally to rats [8]. Both rats (2 out of 2, 100%) died 1 day after the administration of 16 g/kg EG. In contrast, a total of 6 out of 8 (75%) rats survived 5 days after the administration of 8 g/kg EG. All the rats administered saline (5 out of 5) survived after the experiments. Based on these findings, 8 g/kg EG was used throughout this study. To examine EG-induced liver damage in animal models, livers were dissected from rats at 2 and 5 days after administration. Although no significant changes in body weight were observed, liver weights decreased by approximately one-half at 5 days post-EG administration (Table 1). Consistent with the decline in liver weights, which indicated liver damage, metabolic acidosis was observed in EG-administered rats, evidenced by a significant decline in blood pH, a tendency toward decreased HCO₃⁻ concentrations, and a significant elevation in serum AG (Table 1). Taken together, these findings indicate liver damage in EG-administered rats, as evidenced by organ weight loss and metabolic acidosis.

Serological markers were examined to evaluate liver injury in EG-administered rats. Significant increases in AST, ALT, and LDH were observed at 2 days after EG administration (EG2 group), followed by a significant decrease in ALT and declining trends in AST and LDH in the EG5 group compared with the EG2 group (Table 1), as shown schematically in Fig. 1. Although changes were modest, similar tendencies were also noted in ALB, ALP, and γ − GT, suggesting transient (Supplementary Fig. S1) liver injury that peaked at 2 days and recovered by 5 days after EG administration.

Bild vergrößern

Microscopic examination of livers from EG-administered rats was conducted to identify histological features of EG-induced hepatotoxicity (Supplementary Fig. S2).

In the kidney from the EG2 group, congestion and large vacuole-like structures were observed in the cytoplasm compared with the control group (Supplementary Fig. S2A, b and e). Although nuclear morphology showed no apparent differences (Supplementary Fig. S2A, b and e), an increased number of cells contained multiple nucleoli per nucleus (Supplementary Fig. S2B).

In the kidney from the EG5 group, lesser congestion and vacuole-like structures were observed in the cytoplasm compared with the EG5 group (Supplementary Fig. S2A, c and f). As shown in Supplementary Fig. S2, the occurrence of nuclei with multiple nucleoli was significantly increased compared with that in controls (Supplementary Fig. S2B).

To further examine the degenerative changes observed in EG-administered rat livers, electron microscopy was performed.

In the kidney from the EG2 group, the most notable finding was that hepatocyte cytoplasm contained regions of low electron density, suggestive of fluid accumulation in localized cytoplasmic areas (Fig. 2, b, e, and h). Closer inspection of these regions revealed 10–30 nm granule-like structures, consistent in size with glycogen (Fig. 2, h). Although mitochondria and ER appeared displaced to peripheral areas of the cytoplasm, no apparent mitochondrial degeneration was observed (Fig. 2, e). In contrast, the proportion of rough ER relative to total ER volume increased compared with that in controls (Fig. 2, e).

Bild vergrößern

In the kidney from the EG5 group, restoration of these observations (fluid accumulation as well as increased proportion of rough ER observed in the EG2 group) to basal levels was observed (Fig. 2, c, f, and i).

To further evaluate the low electron density region observed in the liver of the EG2 group, TEM, PAS, PAM, and Oil Red O staining were performed.

In the kidney from the EG2 group, both PAS and PAM staining showed typical glycogen staining patterns in the hepatic cytoplasm (Supplementary Fig. S3A, b and e), confirming that the granules observed by TEM (Fig. 2, h) were glycogen granules. Oil Red O staining did not stain the hepatocytes, excluding the possibility that the area contained lipids (Supplementary Fig. S3A, h). In addition, we observed crystals in the kidney from EG2 group (Supplementary Fig. S3B).

In the kidney from the EG5 group, restoration of these observations (accumulation of glycogens as well as crystals observed in the EG2 group) to basal levels was observed (Supplementary Fig. S3A, c and f).

Toluidine blue staining, which does not stain glycogen, was applied to liver sections from EG-administered rats and examined under bright-field (Supplementary Fig. S4, upper row) and dark-field (Supplementary Fig. S4, lower row) microscopy.

In the EG2 group, large areas of cytoplasm surrounding the nuclei were negative under bright-field microscopy.

In the EG5 group, the toluidine blue-negative cytoplasm showed bright birefringent granules under dark-field microscopy, suggesting the presence of oxalate (Supplementary Fig. S4, c and f).

To further investigate the molecular mechanisms underlying transient degeneration in EG-administered rat liver, transcriptome analysis using DNA microarray was performed on liver samples from rats with and without EG administration. The top 20 most upregulated genes in the EG2 and EG5 groups are listed in Tables 2 and 3, respectively.

In the EG2 group, both the most upregulated gene (SLC3A1) [27] and the fifth most upregulated gene (SLC7A13) [28, 29] were transporters involved in cystine uptake (Table 2). Pathway enrichment analysis showed that the “Burn wound healing” pathway, which includes genes related to apoptosis, inflammation, and collagens, ranked first (Fig. 3).

Bild vergrößern

In the EG5 group, three of the five most upregulated genes (Sult2a6, Sult2a1, and Sult2a2) were associated with sulfate metabolism and conjugation, which enhance water solubility of xenobiotics for excretion [30, 31], whereas the most upregulated gene (Akr1b7) was involved in bile acid generation [32, 33] and the fourth most upregulated gene encoded a glutathione transferase (GST3π) (RGD1307603, Table 3). Pathway enrichment analysis showed that the “Burn wound healing” pathway ranked second in the EG5 groups (Fig. 4). Notably, the “cytoplasmic ribosomal proteins” pathway was ranked first in the EG5 group (Fig. 4). GO enrichment analysis also identified “Ribosome” as the top category in the EG5 group, suggesting increased ribosome synthesis following EG exposure (Supplementary Fig. S5).

Bild vergrößern

Collectively, these findings indicate that cysteine transport may be upregulated as early as 2 days after EG administration. By 5 days, xenobiotic detoxification, which is consisted of chemical modifications of xenobiotics including sulfate conjugation, appeared to be induced, along with increased ribosomal biogenesis, which should be increased during proliferation that may support hepatocyte recovery from EG-induced damage (Figs. 1 and 2).

The upregulation of SLC3A1 and SLC7A13 peaked in the EG2 group and returned to basal levels in the EG5 group (−1.37 for SLC3A1 and 1.01 for SLC7A13). In contrast, another cystine transporter, SLC7A11, was downregulated throughout EG administration (−2.07 in EG2 and −3.33 in EG5). Downregulation of this transporter, along with decreased glutathione peroxidase 4 (Gpx4), has been shown to result in ferroptosis, a form of regulated necrosis characterized by an iron-dependent increase in lipid peroxidation and reduced cellular capacity to eliminate toxic lipid peroxides [34, 35]. Gpx4 is an only Gpx that can reduce lipid peroxide and SLC7A11 is required for elevation of cellular glutathione levels through providing cysteine, a precursor of glutathione. We thus examined SLC7A11 and Gpx4 expression by immunoblotting. As shown in Fig. 5A, SLC7A11 levels showed a transient but significant decrease followed by partial recovery. In contrast, Gpx4 levels showed a time-dependent decline, reaching a significant reduction in the EG5 group compared with the control group (Fig. 5A).

Bild vergrößern

We next investigated whether the increased number of nucleoli (Supplementary Fig. S2B) and ribosomal gene expression (Fig. 4 and supplementary Fig. S5) were associated with increased protein synthesis. Immunoblot analysis of eukaryotic initiation factor 2α (eIF2α), whose phosphorylation correlates with inhibition of protein synthesis [36], showed decreased phosphorylation in response to EG administration, indicating increased translation (Fig. 5B). Thus, the increase in nucleolar number might lead to an increase of protein synthesis. To further evaluate the status of hepatic EG metabolism (Fig. 5C), we quantified mRNA levels of enzymes catalyzing major steps of EG metabolism in hepatocytes. As shown in Fig. 5D, both alcohol dehydrogenase (Adh1) and hydroxy acid oxidase (HAO1) were transiently upregulated during EG administration. Given that EG itself is relatively less toxic and its metabolites play major roles in toxicity, these findings are consistent with the transient hepatic damage observed after EG administration (Fig. 1). Finally, we examined 4-hydroxy-2-nonenal (4HNE)-modified protein levels, which reflect lipid peroxidation due to a decrease of SLC7A11-Gpx4 axis [37, 38], and confirmed accumulation of 4HNE-modifed proteins (Fig. 5E), indicative of lipid peroxidation.

To further confirm the negative effect of EG metabolism on the SLC7A11-Gpx4 axis in hepatocytes, EG was administered to Huh-7 human hepatoma cells at concentrations (0.3, 0.6, and 1.2 M) previously reported to produce approximately 50% inhibition of KB cell viability after 3 days [20]. As shown in Fig. 6A and B, cell viability decreased in a concentration-dependent manner. Significant reductions in SLC7A11 and Gpx4 levels were also observed in 1.2 M EG group compared with those in the control group, confirming that EG exerts a negative effect on the SLC7A11-Gpx4 axis in hepatic cells (Fig. 6C). Evaluation of EG metabolism, along with solute carrier family 26 member 1 (SLC26A1), which is considered the main transporter for oxalate extrusion [39], showed induction of EG metabolism and oxalate extrusion following EG administration (Fig. 6D).

Bild vergrößern