Treatment with sodium (S)-2-hydroxyglutarate prevents liver injury in an ischemia-reperfusion model in female Wistar rats

Synthesis of (S)-2HG

(S)-2HG was synthesized using a two-step methodology from L-glutamic acid, by using the optimized synthetic route reported by our laboratory (Cienfuegos-Pecina et al., 2020). Briefly, L-glutamic acid was subjected to a diazotization reaction in the presence of NaNO₂ and H₂SO₄ for 24 h. The reaction was stopped with NaCl and extracted three times with ethyl acetate, obtaining the 5-oxotetrahydrofuran-2-carboxylic acid, which was purified by column chromatography, using silica as the stationary phase and ethyl acetate as the mobile phase. This compound was hydrolyzed with NaOH, pH 10 for two hours, obtaining (S)-2HG, which was precipitated with anhydrous methanol.

The identity of the synthesis product was confirmed by nuclear magnetic resonance (NMR) spectroscopy and polarimetry. NMR data were acquired in a Bruker AVANCE III HD 400 MHz spectrometer (Bruker Corp., Billerica, MA, USA). A 50 mg sample of (S)-2HG was dissolved in double-distilled water and analyzed using sodium 3-(trimethylsilyl)propionate-2,2,3,3-d₄ (TSP) in D₂O (Sigma-Aldrich) as an internal standard. The ¹H-NMR and ¹³C-NMR spectra were acquired using the noesypr1d pulse sequence for water signal suppression and compared with those reported in the literature (Bal & Gryff-Keller, 2002; Cienfuegos-Pecina et al., 2020).

The specific optical rotation ([α]_D^20°C) of the synthetic (S)-2HG was measured in a PerkinElmer 341 Polarimeter (PerkinElmer, Waltham, MA, USA). Data were compared with previous reports in the literature (Cienfuegos-Pecina et al., 2020; Ritthausen, 1872).

Animals

Twenty-eight healthy female Wistar rats, weighing 238 ± 18 g, were used. Every single animal was considered an experimental unit. Rats were bred in-house, and they were kept under standard conditions of light and temperature (24 ± 3 °C, 12 h light-dark cycles), with access to commercial rat pellets (Nutrimix de México, Mexico) and water ad libitum. The animals were not genetically engineered, and no previous procedures were performed on the animals before the experiments. All the procedures were performed according to the specifications of the Mexican Official Norm NOM-062-ZOO-1999. This project was submitted to the Ethics and Research Committee of the School of Medicine, Universidad Autónoma de Nuevo León, and approved with the register number HI19-00002.

Experimental design

The sample size was decided based on the result of an a priori calculation using (1). (1)${\displaystyle n=\frac{{\left(Z\alpha +Z\beta \right)}^2\left({\sigma }_1^2+{\sigma }_2^2\right)}{{\left({\mu }_1-{\mu }_2\right)}^2}}$

Equation (1). Estimation of the sample size for comparison of means. Zα = Z-value for α; Zβ = Z-value for β; σ₁ = expected standard deviation of group 1; σ₂ = expected standard deviation of group 2; µ₁ = expected mean of group 1; µ₂ = expected mean of group 2.

As a reference, we considered a previously reported serum activity of ALT of 494 ± 84 U/L for rats subjected to the same IR-injury-induction protocol we used (Jiménez Pérez et al., 2016). We calculated the sample size expecting a 40% decreasing of the serum ALT activity after the treatment (with no change in the standard deviation) and considered a two-tail statistical significance (α) of 0.01 and a statistical power (1−β) of 95%, obtaining a total of 7 rats per group.

Rats were paired with a random number sequence obtained in R v. 4.01. The randomized animals were assigned to the following groups:

• Sham group (SH), n = 7: Rats were treated with double distilled water, p.o., twice per day, for two days. Then they were undergone to laparotomy with exposure of the liver hilum, without inducing IR injury. After 1 h and 20 min, rats were sacrificed by exsanguination, obtaining blood and tissue samples.

• Nontoxicity group (HGTox), n = 7: Rats were treated with double-distilled-water-dissolved (S)-2HG, at a dose of 25 mg/kg, in the same way as the SH group. After treatment, they underwent the same procedure as the SH group. A final n of 6 was considered for the data analysis since one of the rats suddenly died after the anesthesia was administered.

• IR group (IR), n = 7: Rats were treated in the same way as the SH group. Then they underwent laparotomy with an exposition of the liver hilum and induction of IR injury (20 min of ischemia, one hour of reperfusion).

• (S)-2HG + IR group (HGIR), n = 7: A dose of 25 mg/kg of (S)-2HG, was administered to the rats in the same way as the HGTox group. Then, they underwent the same procedure as the IR group.

To minimize potential confounders, all rats were housed in the same room during the experiments, under the same conditions of light and temperature. In each surgery, an equal number of rats from each group underwent the surgical procedure. During the analysis of the biological samples, all the involved were blinded to the identity of the samples, except those who then performed the statistical analysis (Eduardo Cienfuegos-Pecina and Paula Cordero-Pérez).

Induction of liver injury

Ischemic liver injury was induced following the procedure reported by Jiménez Pérez et al. (2016). Rats were anesthetized with 100 mg/kg of ketamine (Anesket, PiSA Agropecuaria, S.A. de CV, Mexico), and 10 mg/kg of xylazine (Sedaject, Vedilab S.A. de CV, Mexico), i. p. After asepsis, an incision along the midline was performed, exposing the hepatic hilum, which was then occluded by using an atraumatic vascular clamp (Pringle maneuver) for 20 min. Following the ischemia, the clamp was withdrawn, and the rats were kept under general anesthesia for a reperfusion time of 1 h, after which rats were sacrificed by exsanguination by a puncture in the aorta. Criteria for early euthanasia was the failure of the anesthesia during the surgery, however, none of the animals needed the use of this protocol. Blood samples were centrifuged at 2,000 g for 12 min, then the serum was separated. The liver was resected, weighed, and samples of tissue were obtained and frozen at −80 °C until analysis.

Biochemical markers, oxidative stress biomarkers, and proinflammatory cytokines

To assess the magnitude of the induced liver injury, the serum activities of alanine aminotransferase (ALT), aspartate aminotransferase (AST), lactate dehydrogenase (LDH), and alkaline phosphatase (ALP), and the serum concentrations of total bilirubin and glucose were measured. The biochemical determinations were performed in an ILab Aries analyzer (Instrumentation Laboratory, SpA, Milan, Italy), using kinetic and end-point UV-Visible spectrophotometric methodologies.

Malondialdehyde (MDA) is one of the main products of arachidonic acid peroxidation, and it is a commonly used biomarker to assess oxidative stress (Ayala, Muñoz & Argüelles, 2014), alongside the activities of superoxide dismutase (SOD) and glutathione peroxidase (GPx). Since one of the key mechanisms involved in IR injury is the oxidative burst produced by the sudden activation of xanthine oxidase, we compared the tissue levels of MDA, SOD, and GPx among the study groups. To measure the concentration of MDA in liver tissue, 200 mg of liver tissue were mechanically homogenized in 1 mL of RIPA buffer in an ice bath, and then centrifuged at 1,600 g for 10 min at 4 °C. MDA concentration was measured spectrophotometrically in the supernatant, using the thiobarbituric acid colorimetric method with a thiobarbituric acid-reactive substances (TBARS) assay kit (Cayman Chemical Company, Ann Arbor, MI, USA). The product of this reaction was measured spectrophotometrically at a wavelength of 535 nm.

The total tissue activity of SOD was quantified by measuring the inhibition of the reduction of a tetrazolium salt to formazan by reactive oxygen species using the SOD Assay Kit (Cayman Chemical Company). Briefly, 200 mg of liver tissue were mechanically homogenized in one mL of 20 mM HEPES buffer, pH 7.2, containing 1 mM EDTA 210 mM mannitol, and 70 mM sucrose. Homogenization was done in an ice bath. Samples were then centrifuged for 15 min at 10,000 g, at 4 °C and the assay was performed using the supernatant, diluted in the buffer. 96-well microplates were read spectrophotometrically at a wavelength of 450 nm and total SOD activity was determined.

Tissue GPx activity was measured by quantifying the oxidation of NADPH to NADP⁺ using a GPx Assay Kit (Cayman Chemical Company). To perform the assay, 200 mg of liver tissue were mechanically homogenized in one mL of 50 mM Tris-HCl buffer, pH 7.5, containing 5 mM EDTA and 1 mM DTT. Homogenization was performed in an ice bath. Samples were centrifuged at 10.000 g for 15 min at 4 °C, and the supernatants were separated. The supernatants were diluted in the sample buffer provided by the manufacturer before the GPx determination. GPx activity was quantified by a UV kinetic method, measuring the change in the absorbance at a wavelength of 340 nm.

To assess the inflammatory response after the induction of IR injury, we measured the tissue concentrations of the proinflammatory cytokines interleukin 1β (IL-1β), interleukin 6 (IL-6), and tumor necrosis factor (TNF) using a commercial sandwich enzyme-linked immunosorbent assay (ELISA) (R&D Systems, Minneapolis, MN, USA). The same samples from the GPx determination were used, undiluted. ELISA protocols were performed following the manufacturer’s instructions. Cytokine concentrations were determined spectrophotometrically, at a wavelength of 450 nm.

All measurements in tissue were normalized to the protein concentration in the homogenates, measured using Bradford’s method. Bradford’s reagent was prepared as previously reported in the literature (Kruger, 2009). MDA concentration is reported as mmol/mg of protein, SOD activity is reported as IU/mg of protein, GPx activity is reported as nmol/min/mg of protein, and proinflammatory cytokines are reported as pg/mg of protein.

Histological evaluation

Samples of liver tissue were fixed in phosphate-buffered 10% formalin solution pH 7.4 and then embedded in paraffin blocks, which were cut using a microtome at a thickness of 4 µm. The tissue sections were deparaffinized and processed using the standard histological technique. Hematoxylin-eosin (H&E) stained sections were blindly evaluated under the microscope, assessing the severity of necrosis, cytoplasmic vacuolization, and sinusoidal congestion by using the damage scale reported by Susuki et al. (1993): 0 = no evident injury; 1 = single-cell necrosis, minimal congestion or vacuolization; 2 = necrosis <30%, mild congestion or vacuolization; 3 = necrosis <60%, moderate congestion or vacuolization; and 4 = necrosis >60%, severe congestion or vacuolization.

Quantitative RT-PCR

We used RT-qPCR to measure the expression of Hmox1, Vegfa, and Pdk1, the genes coding for heme-oxygenase 1, vascular endothelial growth factor A, and pyruvate dehydrogenase kinase 1, whose expression is directly regulated through the HIF-1 pathway (Bernhardt et al., 2009; Bujaldon et al., 2019; Forsythe et al., 1996; Zhang et al., 2018). We performed a total RNA extraction from 100 mg of liver tissue using TRIzol reagent (Invitrogen, Thermo Fisher Scientific, Carlsbad, CA, USA) according to the manufacturer’s specifications. The RNA was quantified using a Microdrop Multiskan GO (Thermo Fisher Scientific, Carlsbad, CA, USA) and stored at −80 °C until analysis.

We performed all the RT-qPCRs using the GoTaq 1-Step kit (Promega Corporation, Madison, WI, USA). The gene coding for β-actin (Actb) was used as the housekeeping gene. The following primers were used: Hmox1 forward 5′-GCCTGCTAGCCTGGTTCAAGA-3′, Hmox1 reverse 5′-GAGTGTGAGGACCCATCGCA-3′, Vegfa forward: 5′-CCGTCCTGTGTGCCCCTAAT-3′, Vegfa reverse: 5′-AAACAAATGCTTTCTCCGCT-3′, Pdk1 forward: 5′-GATTGCCCATATCACGCCTCT-3′, Pdk1 reverse: 5′-CTCGTGGTTGGT TCTGTAATGC-3′, Actb forward 5′-CCCTGGCTCCTAGCACCAT-3′, and Actb reverse 5′-GATAGAGCCACCAATCCACACA-3′. We performed every reaction according to the manufacturer’s specifications, using 200 ng of RNA for each reaction and the primers at a concentration of 100 nM, to complete a final volume of 20 µL. The following reaction conditions were used: one cycle of reverse transcription at 37 °C for 15 min, one cycle of reverse transcription inactivation and Go Taq DNA Polymerase activation at 95 °C for 10 min, and 40 cycles of denaturation at 95 °C for 10 s and annealing/extension at 60 °C for 30 s. Fold changes of gene expression from the SH group were calculated using the 2^−ΔΔCt method.

Statistical analysis

Data were analyzed using a Shapiro–Wilk normality test, followed by either a one-way ANOVA test with a Tukey post hoc test or a Kruskal-Wallis test with a Dunn post hoc test. For the gene expression assessment, the −ΔΔC_T values were analyzed. All the statistical analyses were performed in R v. 4.0.1, using the packages tidyverse, cowplot, and PMCMRplus. The full datasets are supplied in Supplemental Informations 1–4, while the code used for the statistical analysis is supplied in Supplemental Information 5. The results are expressed as mean ± standard deviation or median (interquartile range), depending on their distribution. Differences between groups are considered significant at p < 0.05.

(S)-2HG was successfully synthesized from L-glutamic acid

(S)-2HG was produced in a 29.78% yield after alkaline hydrolysis, and it was obtained as a beige powder, highly hygroscopic, and soluble in water but insoluble in organic solvents such as methanol, ethanol, acetone, and ethyl acetate. The ¹H-NMR spectrum showed the following signals: 1.82 ppm (1H, m); 1.96 ppm (1H, m); 2.22 ppm (2H, m), and 4.00 ppm (1H, dd, J’ = 7.6 Hz, J” = 4.0 Hz) (Fig. S1). In the ¹³C-NMR spectrum, the following signals were observed: 33.91, 36.38, 75.01, 184.10, and 185.71 ppm (Fig. S2). Both spectra were consistent with the reported in the literature for this compound (Bal & Gryff-Keller, 2002; Cienfuegos-Pecina et al., 2020). A value of [α]_D 20° = 8.4° cm³g⁻¹dm⁻¹ was observed, and it was consistent with previously reported data (Cienfuegos-Pecina et al., 2020; Ritthausen, 1872).

(S)-2HG is hepatoprotective, but not hepatotoxic at the tested dose

To evaluate whether (S)-2HG produces an acute hepatotoxic effect, we compared the levels of liver injury biomarkers between the SH and HGTox groups. We did not observe a significant difference between these groups in any of the assessed biochemical markers (Fig. 1).

To assess whether (S)-2HG had a hepatoprotective effect, we compared the levels of the liver injury biomarkers among the SH, IR, and HGIR groups. Neither the treatment with (S)-2HG nor the IR injury induction affected the liver weight of the animals (Fig. S3). The induction of IR injury significantly increased the serum activities of ALT, AST, and LDH compared to the SH group, while the treatment with (S)-2HG produced a significant decrease in the serum activities of these enzymes compared to the IR group (Fig. 1). A significant decrease in serum glucose concentration was observed in both, the IR and HGIR groups compared to the SH group. No significant differences were observed among groups in the serum activity of ALP and the serum concentration of total bilirubin (Fig. 1).

The experimental model of total liver IR injury did not affect the levels of oxidative stress biomarkers in liver tissue

No significant differences were observed in the levels of oxidative stress biomarkers between the IR and SH groups (Table 1). Their levels were not affected by the administration of (S)-2HG in the HGTox group. We observed a decrease in the tissue activity of GPx in the HGIR group compared with the SH group (Table 1).

(S)-2HG modulates the concentration of proinflammatory cytokines

Our model of acute liver injury produced a significant increase in the tissue concentration of IL-1β in the IR group compared to the SH group. Treatment with (S)-2HG prevented this increase, as we observed a concentration of IL-1β in the HGIR group significantly lower than the concentration in the IR group (Fig. 2).

The induction of liver IR injury did not affect the tissue concentrations of IL-6 and TNF, however, a significant decrease in the tissue concentration of these cytokines was observed in the HGIR group, compared with both IR and SH groups (Fig. 2). The administration of (S)-2HG to the HGTox group did not cause any alteration in the levels of the assessed cytokines compared to the SH group (Fig. 2).

(S)-2HG produced a trend to decrease the severity of the histological liver injury, but it was not significative

The administration of (S)-2HG to the HGTox group did not produce a significant change in the scores of histological liver injury compared to the SH group (Table 2, Fig. 3). On the other hand, our model of IR injury induced a significant degree of necrosis and sinusoidal congestion in the IR group compared to the SH group, without producing a significant increase in the degree of cytoplasmic vacuolization (Table 2, Fig. 3). The histopathological evaluation of the IR group showed cellular eosinophilia and nuclear hyperchromasia. Cellular necrosis was severe at zone 3 of the liver acinus, with disruption of the central vein endothelial border (Fig. 3). The administration of (S)-2HG tended to decrease the severity of necrosis and sinusoidal congestion compared to the IR group, although this decrease was not statistically significant (Table 2, Fig. 3).

The treatment with (S)-2HG decreased the expression of Vegfa and Pdk1 in liver tissue, without affecting the expression of Hmox1

Contrarily as we expected, the treatment with (S)-2HG did not increase the expression of Hmox1, Vegfa, or Pdk1 in liver tissue. The expression of Hmox1 in liver tissue was not affected by the administration of (S)-2HG or by the induction of IR injury (Fig. 4). On the other hand, we observed a significant decrease in the expression of both, Vegfa and Pdk1 in the rats of the HGTox group compared to the SH group (Fig. 4). The induction of IR also decreased the expression of these genes compared to the SH group (Fig. 4). These observations suggest that the hepatoprotective effect of (S)-2HG is not mediated by the HIF-1 metabolic axis, since we did not observe the genetic footprint of the HIF-1α stabilization, indicating the involvement of a different, currently unidentified mechanism.