Wild type male C57BL/6 mice were purchased from Slyke jingda Biotechnology Company (CertificateSCXK 2016-0002; Hunan, China). The VDR knockout (VDR-KO), renal proximal tubular specific VDR overexpressing (VDR-OE) mice and littermates were constructed in cooperation with the Model Animal Research Center of Nanjing University. All experimental mice were fed under SPF conditions, and the experimental protocols were approved by the Laboratory Animal Ethics Committee of Central South University.

To induce AKI, 8-week-old male mice received intraperitoneal injections of either PBS (WT group) or 20 mg/kg LPS (O111:B4, L2630, Sigma Aldrich, LPS group) for one dose. To investigate the effect of VDR activation on LPS-induced AKI, mice were injected intraperitoneally with paricalcitol (an activated vitamin D analog, a present from professor Yan Chun Li, Chicago university, 0.2 μg/kg/day for wild-type mice (Du et al., 2019) and 0.1 μg/kg/day for VDR-OE mice (Jiang et al., 2021), respectively, 1 week before LPS injection) or the same volume of solvent. Finally, wild type male C57BL/6 mice were randomly divided into WT, WT + P, LPS group and LPS + P group; VDR-KO mice and their littermates were randomly divided into WT, KO, LPS, KO + LPS groups; and OE mice and their littermates were randomly divided into WT, OE, LPS, and OE + LPS groups. All mice were sacrificed 24 h after LPS administration. Their blood and kidneys were collected for subsequent experimental analysis.

Human proximal tubular epithelial cells (HK-2 cells) and their VDR knockout (VDR-KO) cell lines were used in this study, provided by the Institute of Nephrology, Central South University. HK-2 cells were transfected with VDR plasmid and blank plasmid using Lipofectamine 2000, and cultured in F12 (1:1) DMEM supplemented with 10% fetal bovine serum. Cells were seeded in six-well plates at a rate of 5 × 10⁴ cells/well. After 24 h of incubation at 37°C, 5% CO₂, the cell cultures were supplemented with paricalcitol (200 nM for HK-2cells, 100 nM for VDR-OE cells) for 24 h, and then treated with LPS (1 μg/mL) for a further 16–24 h to harvest cells for follow-up experiments. In some experiments, DCA (5 mM, 2 h pretreatment, 347795, Sigma Aldrich), Compound C (10 μM, 1h pretreatment, HY-13418, MedChemExpress) were used.

Blood and renal tissues were collected for biochemical analysis. BUN and creatinine levels in serum, and lactate levels, hexokinase activity in renal tissues were measured using the corresponding detection kits in accordance with the manufacturer’s instructions. All these ELISA kits purchased from Nanjing JianCheng Bioengineering Institute (Nanjing, China). In addition, lysate of HK-2 cells was also collected for detection of lactate levels (KTB1100, Abbkine).

Tissues were embedded and cut into 50–100 nm ultrathin sections by an ultramicrotome and a diamond knife. Then, they were double stained with 3% uranyl acetate and lead nitrate and examined with a Hitachi HT- 7700 electron microscope.

Briefly, isolated mouse kidney tissue was fixed immediately with formalin and embedded in paraffin, then, the tissue was cut into 5 μm thick sections for hematoxylin-eosin (H&E) staining, TUNEL fluorescent staining, F4/80 fluorescent staining, p-PDHA1 fluorescent staining and PDHA1 fluorescent staining. The intensities of p-PDHA1 and PDHA1 in the photos were detected by Image J software.

After extraction of tissue and cellular protein, protein was separated by SDS/PAGE and electro-transferred to PVDF membranes. The resulting membranes were blocked with 0.1% (w/v) BSA solution on a shaker for 1 h. Then, they were incubated with primary antibodies at 4°C overnight. The antibodies including: VDR (ab109234, Abcam), cleaved caspase-3 (ab49822, ab214430, Abcam), bcl2 (226593-1-AP, ProteinTech), PDHA1 (sc-377092, Santa Curz Biotechnology), p-PDHA1 (ab177461, Abcam), p-AMPK (2535, Cell Signaling Technology), AMPK (2532, Cell Signaling Technology), β-actin (20536-1-AP, ProteinTech), α-tublin (AF7010, Affinity). After that, the membrane was incubated with the fluorescent secondary antibody for 1 h before three times with TBST. Finally, the protein expression levels were visualized by the Image Studio software and band intensities were quantified with Image J gel analysis software.

Total RNA was extracted from the renal cortex and cell using the corresponding detection kits in accordance with the manufacturer’s instructions. cDNA was synthesized using a reverse transcription kit. Real-time quantitative PCR was performed using SYBR Green PCR Master Mix on a Roche Light Cycler 480 system. PCR primer sequences are shown in Table 1 (Chen et al., 2013; Li et al., 2022). The relative mRNA levels were calculated using the 2^−ΔΔCT formula.

Oxygen consumption rate (OCR) was measured using the SeahorseXF96 Extracellular Flux Analyzer (Seahorse Bioscience, North Billerica, MA, United States). HK-2 cells were seeded into 96-well cell culture plate at a density of 1.5 × 104 cells. When the cell confluence was about 90%, cells were washed twice with assay medium (49.5 mL basal medium, 500 μL sodium pyruvate and basal medium) and incubated in a non-CO₂ incubator for 40–60 min, OCR was measured. The working fluid concentration was as follows: oligomycin (1 μM), FCCP (2.5 μM), rotenone and antimycin A (1 μM).

All data were presented as means ± SD. Statistical comparisons were carried out using unpaired two-tailed Student’s t-test or one-way analysis of variance (ANOVA) as appropriate. Statistical significance was defined as p < 0.05.

As shown in Figure 1A, LPS-induced loss of renal function featured with elevated BUN and creatinine levels, and the levels were further increased in VDR-KO mice. HE staining revealed severe tubular dilation, cell shedding and brush-border disruption in the kidney cortex after LPS injection, and these pathological lesions were more serious in VDR-KO mice than in their WT littermates (Figure 1B). Additionally, immunofluorescence analysis of TUNEL and F4/80 (a marker of macrophage infiltration) in kidney sections showed that elevated tubular cell apoptosis and interstitial inflammatory infiltrate in LPS-treated mice were aggravated in the KO + LPS group (Figure 1C). In addition, compared with the LPS group, the KO + LPS group showed more robust caspase-3 (cleaved) activation, weaker bcl2 in western blot analysis and higher mRNA expression of IL-6 and MCP-1 by real-time RT‒PCR (Figures 1D, E). These results are consistent with those of our previous study (Du et al., 2019).

More importantly, we found that the lactate concentration and hexokinase activity in renal homogenate were increased after LPS injection at 24 h and further increased in VDR-KO mice (Figure 2A). Western blot and immunofluorescence analyses showed that the protein expression level of p-PDHA1/PDHA1 (a key catalytic enzyme that adjusts the tricarboxylic acid cycle and oxidative phosphorylation during glycolysis through phosphorylation and dephosphorylation) was increased in LPS-injected mice, and VDR-KO mice showed a further increase in the p-PDHA1/PDHA1 ratio in renal tissue (Figures 2B, C). Since mitochondria are the main site of aerobic oxidation, we observed morphological changes in mitochondria through transmission electron microscopy, and the results showed that LPS-treated kidney tissue had more swollen mitochondria and a reduced number of cristae with a more lamellar phenotype, and these characteristics were significantly exacerbated in VDR-KO mice (Figure 2D). These results confirm that VDR deficiency aggravated LPS-induced renal injury and glucose metabolism reprogramming.

To further confirm the role of VDR in glucose metabolism reprogramming, we constructed an LPS-induced AKI model in transgenic mice with renal proximal tubular-specific VDR-overexpressing (VDR-OE). As expected, compared with WT littermates treated with the same dose of LPS, renal function, kidney cortex pathological lesions, tubular cell apoptosis and interstitial inflammation were partially ameliorated in VDR-OE mice (Figure 3). Consistently, the increased lactate concentration and hexokinase activity in renal homogenate were lowered by overexpression of VDR (Figure 4A), and the ratio of renal p-PDHA1/PDHA1 expression was also significantly decreased in VDR-OE mice (Figures 4B, C). Electron microscopy showed that the swelling of mitochondria in VDR-OE mice treated with LPS was improved compared to their WT littermates treated with LPS (Figure 4D). These results suggest that VDR overexpression alleviated renal injury and glucose metabolism reprogramming in LPS-induced AKI.

Paricalcitol (pari), an active vitamin D analog, was used in our study to illustrate the role of vitamin D in LPS-induced AKI. Our results show that pari treatment ameliorated renal insufficiency and pathological damage induced by LPS in C57 mice (Figures 5A, B). Additionally, elevated tubular cell apoptosis and interstitial inflammatory infiltrate in LPS-treated mice were inhibited in the LPS + P group (Figures 5C–E), which is consistent with our previous study (Du et al., 2019). Moreover, the increased lactate accumulation, hexokinase activity and p-PDHA1/PDHA1 ratio in LPS-treated mice were relieved by pari treatment (Figures 6A–C). The mitochondrial damage induced by LPS injection was significantly attenuated by pari treatment (Figure 6D). These results confirm that paricalcitol alleviated glucose metabolism reprogramming in LPS-induced acute kidney injury.

We constructed VDR knockout and VDR overexpression HK-2 cell lines to verify the effect of VDR on LPS-induced glucose metabolism reprogramming in vitro. Our results show that LPS induced glucose metabolism reprogramming, including a decreased oxygen consumption rate (OCR) and increased lactate levels, which was more serious in VDR-KO cells (Figures 7A, B). On the contrary, overexpression of VDR significantly attenuated LPS induced reprogramming of glucose metabolism featured with restored OCR and decreased lactate levels (Figures 8A, B). Furthermore, compared with the LPS group, the KO + LPS group had a higher p-PDHA1/PDHA1 ratio (Figure 7C), while the OE + LPS group had a lower p-PDHA1/PDHA1 ratio (Figure 8C). The expression of caspase-3 (cleaved) and bcl2 by western blot (Figures 7C, 8C) and the mRNA expression of IL-6 and MCP-1 (by PCR) (Figures 7D, 8D) showed that VDR knockout could promote cell apoptosis and inflammation in HK-2 cells, while VDR overexpression could improve these alterations induced by LPS. These results suggests that VDR alleviates LPS-induced glucose metabolism reprogramming and cell injury in renal tubular cells.

The above results indicate that VD-VDR can reduce the level of phosphorylated PDHA1 (p-PDHA1) but has no evident effect on the total protein level of PDHA1 in LPS-induced glucose metabolism reprogramming. Thus, to confirm that VDR can play a protective role in glucose metabolism reprogramming and renal injury in LPS-induced AKI by regulating PDHA1, DCA (a p-PDHA1 inhibitor) was used to evaluate LPS-induced tubular cell injury in vitro. In HK2 cells, the decreased OCR levels and increased cellular lactate accumulation induced by LPS were protected by paricalcitol or the p-PDHA1 inhibitor DCA, respectively. Importantly, when treated with both DCA and pari, the protective effect on the glucose metabolism reprogramming of HK-2 cells was no better than that of DCA alone (Figures 9A, B), and the same phenomenon also appeared in the impact on the ratio of p-PDHA1/PDHA1 expression (Figure 9C). Interestingly, the anti-apoptotic and anti-inflammatory effects of the DCA and pari combination in LPS-induced HK2 cells were comparable to those of pari alone (Figures 9C, D). These data confirm that VDR activation can alleviate LPS-induced glucose metabolism reprogramming by inhibiting the phosphorylation of PDHA1.

We further investigated how VD-VDR inhibits the phosphorylation of PDHA1. Since our previous research confirmed that VD-VDR can activate AMPK in diabetic nephropathy (Li et al., 2022), we detected the level of protein expression of p-AMPK in LPS-induced AKI mice and LPS-treated HK-2 cells. As Figures 10, 11 show, p-AMPK levels were increased in LPS-induced AKI mice and LPS-treated HK-2 cells, and paricalcitol or VDR overexpression further promoted the expression of p-AMPK (Figures 10B, C, 11B). In addition, the increased level of p-AMPK was weakened in VDR-KO mice and HK-2 cells (Figures 10A, 11A).

It has been reported that the glycolysis shift induced by LPS is related to AMP-activated protein kinase (AMPK) (Tan et al., 2021). Therefore, we examined the protein expression of p-AMPK/AMPK in LPS-treated HK2 cells and observed glycolytic metabolism in HK2 cells treated with Compound C, an AMPK inhibitor. As our data show, Compound C and LPS induced the metabolic state switch from oxidative phosphorylation (lower OCR levels) to glycolysis (higher lactic acid levels), while the effects on alleviating glucose metabolism reprogramming by pari were greatly weakened after Compound C treatment (Figures 11C, D). Similar results were observed in the effect of Compound C on the expression of p-PDHA1/PDHA1 and caspase-3 (cleaved) (Figure 11E). Real-time PCR analysis indicated that Compound C could induce increased expression of IL-6 mRNA, and the reduction in IL-6 by paricalcitol was also largely abolished in the presence of Compound C, but it had no influence on MCP-1 mRNA (Figure 11F).