Research ArticleKidney Disease

Nuclear receptor PXR targets AKR1B7 to protect mitochondrial metabolism and renal function in AKI

See allHide authors and affiliations

Science Translational Medicine  13 May 2020:
Vol. 12, Issue 543, eaay7591
DOI: 10.1126/scitranslmed.aay7591

A nuclear option for acute kidney injury

Acute kidney injury is a life-threatening health problem with many potential causes such as hypoxic injury, toxins, and sepsis. Patients can recover with supportive care, but they may have long-lasting kidney damage that increases the risk of chronic kidney disease, and there is no specific targeted treatment. Yu et al. discovered that nuclear receptor PXR is down-regulated in the setting of acute kidney injury, which contributes to the impairment of mitochondrial function. Conversely, interventions that restored PXR function in mouse models of acute kidney injury were effective at protecting the kidneys, suggesting the potential for a therapeutic approach targeting PXR.

Abstract

Acute kidney injury (AKI) is a worldwide public health problem with no specific and satisfactory therapies in clinic. The nuclear pregnane X receptor (PXR) is involved in the progression of multiple diseases, including metabolic diseases, atherosclerosis, hypertension, liver injury, etc. However, its role in kidney injury remains to be understood. In this study, we have investigated the role of PXR in AKI and underlying mechanism(s) involved in its function. PXR was robustly down-regulated and negatively correlated with renal dysfunction in human and animal kidneys with AKI. Silencing PXR in rats enhanced cisplatin-induced AKI and induced severe mitochondrial abnormalities, whereas activating PXR protected against AKI. Using luciferase reporter assays, genomic manipulation, and proteomics data analysis on the kidneys of PXR−/− rats, we determined that PXR targeted Aldo-keto reductase family 1, member B7 (AKR1B7) to improve mitochondrial function, thereby ameliorating AKI. We confirmed the protective role of PXR against kidney injury using genomic and pharmacologic approaches in an ischemia/reperfusion model of AKI. These findings demonstrate that disabling the PXR/AKR1B7/mitochondrial metabolism axis is an important factor that can contribute to AKI, whereas reestablishing this axis can be useful for treating AKI.

INTRODUCTION

Acute kidney injury (AKI), a life-threatening condition with a worldwide incidence of 13.5 million patients per year, not only contributes to about 1.7 million deaths every year but also predisposes patients to the development of chronic kidney disease (CKD) (1). The primary causes of AKI include ischemia, sepsis, and nephrotoxicity. Currently, no satisfactory therapies are available to treat established AKI, which greatly limits clinical efforts to improve the outcomes of patients with AKI and their kidneys (2, 3). Thus, new strategies that minimize AKI are urgently needed.

Normal kidney function depends on different cell types working in concert, which involves many energy-intensive processes (4). Renal proximal tubular epithelial cells are densely packed with mitochondria, whose function is to synthesize adenosine 5′-triphosphate (ATP) through electron transport and oxidative phosphorylation (OXPHOS) in conjunction with the oxidation of metabolites by the tricarboxylic acid cycle and catabolism of fatty acids by β-oxidation. Mitochondrial homeostasis is closely regulated by mitochondrial biogenesis, including the control of mitochondrial DNA (mtDNA) replication and mitochondrial dynamics such as mitochondrial fragmentation and mitophagy (autophagic clearance of damaged mitochondria) (5). Disruption of mitochondrial metabolism and mitochondrial integrity could cause the collapse of energy homeostasis, ultimately resulting in or promoting AKI (610). Considering the central role of mitochondria in energy metabolism, targeting mitochondrial metabolism may offer a promising strategy for alleviating AKI.

Nuclear receptors are ligand-activated transcriptional regulators of several key aspects of renal physiology and pathophysiology. As such, nuclear receptors control a large variety of metabolic processes, including kidney lipid metabolism, drug clearance, inflammation, fibrosis, cell differentiation, and oxidative stress. The pregnane X receptor (PXR) is a member of the nuclear receptor (NR) protein family (classified as Nr1i2) and is mainly expressed in organs with tubular epithelial structure, such as the kidney, liver, and gut (11, 12). In mammalian cells, PXR can be widely involved in the body’s material and energy metabolism by transcriptionally regulating the expression of downstream target genes (13, 14). In addition to its roles in the modulation of metabolic enzymes, PXR has recently emerged as a critical regulator of glycolipid metabolism, inflammation, and atherosclerosis (1517). Recent evidence has shown that dysregulation of the nuclear receptor PXR occurs in CKD, but no functional examination of its role in CKD pathology has been conducted (18). In addition, PXR activation by its ligands could ameliorate human diseases such as cholestatic liver diseases (19). However, there is no report demonstrating the involvement of PXR in AKI.

Before defining the role of PXR in AKI, we conducted an isobaric tags for relative and absolute quantitation (iTRAQ)–based quantitative proteomic analysis to assess differential protein expression in kidneys from wild-type and PXR−/− rats. Among the identified differentially expressed proteins, the expression of Aldo-keto reductase family 1, member B7 (AKR1B7) was markedly lower in kidneys of PXR−/− rats than in those of wild-type controls, suggesting that Akr1b7 might be a target gene of PXR. In agreement with this hypothesis, a previous study showed that AKR1B7 is a transcriptional target of PXR in the liver and intestine (20). AKR1B7 is a member of the AKR superfamily of genes encoding NAD(P)H (reduced form of nicotinamide adenine dinucleotide phosphate)–linked oxidoreductases. AKR1B7 is expressed in mouse kidney, steroidogenic tissues, and intestine and has been suggested to play an important role in the detoxification of lipid peroxidation by-products (21). A recent study showed that hepatic expression of AKR1B7 lowered hepatic triglyceride and cholesterol contents in db/db mice, indicating a function of AKR1B7 in lipid metabolism (22). In metabolic tissue, AKR1B7 was documented as a major regulator of white adipose tissue development through prostaglandin F2α (PGF2α)–dependent mechanisms (23). Hence, PXR-controlled AKR1B7 may play an important role in maintaining metabolic homeostasis and organ function in various contexts. However, this pathway has not been investigated in kidney disease.

In this study, we have used animal models, proximal tubular cells, and human renal biopsy samples to demonstrate the importance of PXR/AKR1B7 signaling in protecting against AKI and improving mitochondrial metabolism (OXPHOS and β-oxidation). These findings suggest a potential approach for treating AKI.

RESULTS

The down-regulation of PXR in kidneys was associated with AKI

First, to explore the association between PXR and human AKI, we performed immunohistochemistry for PXR in kidney biopsies from 20 patients with various forms of AKI resulting from ischemic, nephrotoxic, or combined insults. Para-carcinoma kidney tissues from six patients who underwent renal carcinoma resection were used as controls. The clinical features of the patients with AKI are listed in table S1. As shown in Fig. 1A, an apparent reduction in PXR expression was found in kidneys from patients with AKI relative to that in controls. Further analysis showed that the PXR expression in renal tubular cells was negatively correlated with the peak concentration of blood urea nitrogen (BUN) (r = −0.6377, P < 0.01) and serum creatinine (SCr) (r = −0.7819, P < 0.001) (Fig. 1B), indicating an association between PXR down-regulation and the pathogenesis of acute renal tubular injury. Next, we detected the renal expression of PXR in cisplatin-induced AKI. Consistent with the above data in patients with AKI, PXR expression in the kidneys of cisplatin-treated mice was markedly decreased in a time-dependent manner (Fig. 1, C to E). In cultured mouse renal tubular epithelial cells (RTECs), treatment with cisplatin reduced Pxr expression in a dose- and time-dependent manner (Fig. 1F).

Fig. 1 Regulation of PXR in the kidneys of patients with AKI and murine renal tubular cells treated with cisplatin in vivo and in vitro.

(A) Representative images of immunohistochemical staining of PXR in kidney biopsies from patients with AKI (n = 20) and para-carcinoma kidney tissues (n = 6). Scale bars, 50 μm. (B) Correlation between renal tubular expression of PXR and the peak concentrations of SCr and BUN in patients with AKI. Six random fields were taken from each kidney. IOD, integral optical density. (C) Representative images of immunohistochemical staining of PXR in the cisplatin (CP)–induced AKI mouse model (day 3 after CP injection; n = 6; scale bars, 50 μm). (D) Western blotting for PXR in the kidneys of mice with CP-induced AKI (days 0, 1, 2, and 3 after CP injection). GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (E) Quantitative real-time polymerase chain reaction (qRT-PCR) analysis for Pxr expression in the kidney of CP-induced AKI mice (days 0, 1, 2, and 3 after CP injection; n = 6). (F) qRT-PCR analysis of Pxr expression in RTECs exposed to CP (dose- and time-dependent; n = 3). Data are expressed as means ± SD. Statistically significant differences were determined by Student’s t test and one-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

PXR deficiency aggravated cisplatin-induced AKI

To elucidate the role of PXR in cisplatin-induced AKI, we subjected PXR-deficient (PXR−/−) rats to cisplatin treatment to induce AKI (Fig. 2A and fig. S1A). The SCr and BUN concentrations in cisplatin-treated PXR−/− rats were markedly higher than those in cisplatin-treated wild-type rats (Fig. 2B). Histological analysis showed that PXR−/− rats exhibited exacerbated tubular injury after cisplatin treatment as compared to wild-type rats (Fig. 2C). The expressions of BAX and cleaved caspase-3 and the number of TUNEL (terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick end labeling)–positive cells in the kidneys of cisplatin-treated PXR−/− rats were higher than those in the kidneys of cisplatin-treated wild-type rats (Fig. 2, D and E). Moreover, the quantitative real-time polymerase chain reaction (qRT-PCR) results showed that the mRNA expressions of the tubular injury markers kidney injury molecule 1 (Kim-1) and neutrophil gelatinase–associated lipocalin (Ngal), inflammatory factors (Il-1β, Il-6, and Tnf-α), and Bax in the kidneys of cisplatin-treated PXR−/− rats were higher than those in the kidneys of cisplatin-treated wild-type rats (Fig. 2, F and G). Furthermore, the number of cluster of differentiation 68 (CD68)–positive cells markedly increased in the kidneys of cisplatin-treated PXR−/− rats (fig. S1B).

Fig. 2 PXR deficiency aggravated CP-induced AKI.

(A) Identification of Nr1i2 gene knockout rats. (B) SCr and BUN concentrations were measured in wild-type (WT) and PXR−/− rats treated with CP. (C) Representative periodic acid–Schiff (PAS) staining images (scale bars, 50 μm) of the kidneys (left). Tubular injury scores in rats were analyzed (right). Six random fields were taken from each kidney. (D) Western blotting for BAX, caspase-3, and cleaved caspase-3 in the kidneys of CP-treated WT and PXR−/− rats. (E) Representative images and quantification of TUNEL (terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick end labeling) staining in the kidneys of CP-treated WT and PXR−/− rats. Scale bars, 10 μm; red arrows indicate the TUNEL-positive cells. Six random fields were taken from each kidney. (F) qRT-PCR analysis for renal Kim-1, Ngal, and Bax expressions. (G) qRT-PCR analysis for renal Il-1β, Il-6, and Tnf-α expression. (H) Representative electron micrographs of renal mitochondria (scale bars, 1 μm; red asterisks represent the damaged mitochondria). (I) Western blotting for LC3I, LC3II and P62 in the kidneys of CP-treated AKI rats. (J) qRT-PCR analysis for Atg3, Atg5, and Atg7 expressions. Data are expressed as means ± SD (n = 6). Statistically significant differences were determined by Student’s t test. #P < 0.05, ##P < 0.01, ###P < 0.001, *P < 0.05, **P < 0.01, and ***P < 0.001.

Timely removal of damaged mitochondria via autophagy (mitophagy) is critical for cellular homeostasis and function, thereby conferring protection against AKI (10). As shown in Fig. 2H, swollen mitochondria and disrupted cristae were evident in the kidneys of cisplatin-treated wild-type rats, and these morphological changes were exacerbated in cisplatin-treated PXR−/− rats. Moreover, the expressions of mitophagy-related genes (Atg3, Atg5, and Atg7) and proteins (LC3I and LC3II) in the kidneys of cisplatin-treated PXR−/− rats were lower than those in the kidneys of cisplatin-treated wild-type rats; the P62 protein (autophagy substrate marker) was increased in cisplatin-treated PXR−/− rats (Fig. 2, I and J). Furthermore, PXR deficiency aggravated the already reduced expression of mitochondrial genes in cisplatin-treated rats, consistent with the decreased expressions of Pgc-1α, Tfam, Sod2, and Nrf2 in the kidneys of cisplatin-treated rats (fig. S1, C and D). To evaluate whether aggravated AKI in PXR−/− rats is related to mitochondrial dysfunction, PXR−/− and wild-type rats were intraperitoneally administered MnTBAP [a cell-permeable mitochondrial superoxide dismutase (SOD) mimic] to antagonize mitochondrial oxidative stress before cisplatin treatment. As expected, MnTBAP markedly ameliorated cisplatin-induced kidney injury and mitochondrial oxidative stress in both wild-type and PXR−/− rats (fig. S2). Collectively, these results demonstrate that PXR inactivation impairs mitophagy and mitochondrial function that are important in AKI.

Activation of PXR by PCN attenuated cisplatin-induced AKI

To further examine the effects of PXR activation on cisplatin-induced AKI, mice were administered PCN [50 mg/kg per day, intraperitoneally (i.p.)] for 2 days before cisplatin treatment. In contrast with the findings in PXR−/− rats, the administration of PCN reduced the SCr and BUN concentrations in cisplatin-treated mice (Fig. 3A). Enzyme-linked immunosorbent assay (ELISA) showed that the increased circulating protein concentrations of interleukin-6 (IL-6) and tumor necrosis factor–α (TNF-α) in AKI mice were reduced upon the addition of PCN (Fig. 3B). Histological analysis showed that, in comparison to cisplatin treatment alone, PCN mitigated tubular injury along with a decrease in the release of Cytochrome c (Cyt-c) from mitochondria after cisplatin treatment (Fig. 3C). Moreover, PCN treatment reduced the number of TUNEL-positive cells and F4/80-positive cells (Fig. 3C) and ameliorated the changes in mitochondrial morphology caused by cisplatin (Fig. 3C). Consistent with these results, PCN treatment lowered the mRNA expression of tubular injury markers (Kim-1 and Ngal), apoptosis-associated proteins (Bax and Bcl-2), and inflammatory factors (Il-1β, Il-6, and Tnf-α) in the kidneys as compared to those only treated with cisplatin (Fig. 3, D to F). Pxr mRNA expression was also increased upon exposure to PCN (Fig. 3G).

Fig. 3 PCN attenuated CP-induced AKI.

(A) SCr and BUN concentrations were measured in CP-induced AKI mice with or without PCN administration. (B) Enzyme-linked immunosorbent assay (ELISA) showing the concentrations of circulating IL-6 and TNF-α. (C) Representative images and quantification for periodic acid–Schiff (PAS) staining (scale bars, 50 μm), TUNEL staining (scale bars, 10 μm; red arrows indicate the TUNEL-positive cells), and immunohistochemistry of Cyt-c and F4/80 (scale bars, 20 μm) in the kidneys and representative electron micrographs of renal mitochondria (scale bars, 1 μm; red asterisks denote the damaged mitochondria). Six random fields were taken from each kidney. TEM, transmission electron microscopy. (D) qRT-PCR analyses of renal Kim-1 and Ngal mRNA expressions. (E) qRT-PCR analyses for renal Bax and Bcl-2 mRNA expressions. (F) qRT-PCR analyses for renal Il-1β, Il-6, and Tnf-α mRNA expressions. (G) qRT-PCR analyses for renal Pxr mRNA expression. (H) SCr and BUN concentrations were measured in CP-treated PXR−/− and WT rats with or without PCN administration. NS, not significant. (I) Representative images for PAS staining (scale bars, 20 μm) in the kidneys (left). Tubular injury scores in rats were analyzed (right). Six random fields were taken from each kidney. (J) qRT-PCR analyses for renal Kim-1 and Ngal mRNA expressions. Data are expressed as means ± SD (n = 6). Statistically significant differences were determined by one-way ANOVA and two-way ANOVA. ###P < 0.001, ####P < 0.0001, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

To evaluate whether the protective effect of PCN against cisplatin-induced AKI involves a PXR-dependent mechanism, PXR−/− and wild-type rats were administered PCN (intraperitoneally) before cisplatin treatment. As expected, the protective effect of PCN was abolished in PXR−/− rats (Fig. 3, H to J), suggesting the involvement of PXR in the activity of PCN against cisplatin-induced AKI.

Overexpression of PXR in kidneys protected against cisplatin-induced AKI

To understand the role of PXR in cisplatin-induced AKI, we used the tail vein high-pressure injection method to introduce PXR plasmids into mice. As shown in Fig. 4 (A and B), the mRNA and protein expression of FLAG-tagged PXR markedly increased in the kidneys 36 hours after injection of the PXR plasmids. These results indicated that this gene delivery approach efficiently overexpresses PXR in the kidneys. These mice were administered cisplatin 36 hours after PXR plasmid delivery. Consistent with the results using the PXR agonist PCN, injecting the PXR plasmids markedly improved renal function in AKI mice compared to that in mice injected with scrambled plasmids, as evidenced by the lower concentrations of BUN and SCr (Fig. 4C), and reduced protein concentrations of circulating IL-6 and TNF-α, along with similar changes in the mRNA expression of Kim-1, Ngal, Il-1β, Il-6, and Tnf-α in the kidneys (Fig. 4, D to G). The up-regulated Bax was decreased, and reduced Bcl-2 got restored by overexpressing PXR, consistent with attenuated cellular apoptosis (Fig. 4, H and I). Moreover, the tubular injury scores, Cyt-c release from mitochondria, altered mitochondrial morphology, and the number of F4/80-positive cells were ameliorated upon overexpressing PXR in AKI mice (Fig. 4I). Collectively, these results demonstrate the benefits of PXR overexpression on cisplatin-induced AKI.

Fig. 4 Overexpression of PXR in kidneys protected against CP-induced AKI.

(A) Western blotting for the FLAG-tagged protein in the kidneys 36 hours after injecting PXR plasmids and scrambled plasmids. (B) qRT-PCR analysis for renal Pxr expression 36 hours after injection of PXR plasmids. (C) SCr and BUN concentrations were measured in CP-induced AKI mice injected with PXR or scrambled plasmids. (D) ELISA showing the concentrations of circulating IL-6 and TNF-α. (E) qRT-PCR analyses for renal Kim-1 mRNA expression. (F) qRT-PCR analyses for renal Ngal mRNA expression. (G) qRT-PCR analyses for renal Il-1β, Il-6, and Tnf-α mRNA expressions. (H) qRT-PCR analyses for renal Bax and Bcl-2 mRNA expressions. (I) Representative images and quantification of PAS staining (scale bars, 50 μm), TUNEL (scale bars, 10 μm; red arrows indicate the TUNEL-positive cells), and immunohistochemistry staining of Cyt-c and F4/80 (scale bars, 20 μm) in the kidneys and representative electron micrographs of renal mitochondria (scale bars, 1 μm; red asterisks denote the damaged mitochondria). Six random fields were taken from each kidney. Data are expressed as mean ± SD (n = 6). Statistically significant differences were determined by Student’s t test, one-way ANOVA, and two-way ANOVA. ####P < 0.0001, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. NC, negative control.

PCN treatment or PXR overexpression attenuated cisplatin-induced mitochondrial dysfunction in vitro

We further confirmed the direct effect of PXR activation on cisplatin-induced AKI using RTECs. Consistent with our in vivo results, PCN treatment prevented cisplatin-induced injury in RTECs and primary renal tubular cells, as shown by reduced cellular apoptosis; decreased expression of BAX, cleaved caspase-3, and Cyt-c; and restored expression of B-cell lymphoma-2 (BCL-2) (Fig. 5, A and B, and fig. S3A). Similarly, the inflammatory response induced by cisplatin was attenuated by PCN treatment (Fig. 5C).

Fig. 5 PCN treatment attenuated CP-induced apoptosis and mitochondrial dysfunction in vitro.

(A) Representative images for fluorescence-activated cell sorting (FACS) analysis after annexin V and propidium iodide (PI) staining. RTECs were pretreated with PCN and incubated with CP (5 μg/ml) for 24 hours. The percentages of apoptotic cells were quantified by FACS (n = 3). Ctrl, control; FITC, fluorescein isothiocyanate. (B) Western blotting showing BAX, BCL-2, cleaved caspase-3, and Cyt-c in CP-treated RTECs with or without PCN administration. (C) qRT-PCR analyses for Il-1β, Il-6, and Tnf-α mRNA expressions (n = 3). (D) FACS analysis for the accumulation of ROS in the mitochondria and MMP using the MitoSOX Red Mitochondrial Superoxide Indicator and JC-1 fluorescent probe in RTECs, respectively. qRT-PCR analysis for mtDNA copy number in RTECs. 18S was used as an internal control (n = 3). (E) Cellular ATP content was detected by a luciferase assay (n = 3). (F) Seahorse 96xf was used to detect basal OCR (representative of the basal mitochondrial OXPHOS activity) (n = 8). FCCP, carbonyl cyanide p-trifluoromethoxyphenylhydrazone. (G) Seahorse 96xf was used to detect spare respiratory capacity (n = 8). (H) qRT-PCR analysis for Pgc-1α, Tfam, Sod2, and Nrf2 mRNA expressions (n = 3). (I) qRT-PCR analysis for p62, Atg3, Atg5, and Atg7 mRNA expressions (n = 3). (J) qRT-PCR analysis for Pink1, Parkin, Mfn1, Mfn2, Opa1, Fundc1, and Nix mRNA expressions (n = 3). (K) RTECs were transfected with the pDsRed2-Mito plasmids, and red fluorescence was detected (n = 3; scale bars, 20 μm). (L) RTECs stably expressing mt-mKeima-cox8 were used to detect green and red fluorescence (n = 3; scale bars, 10 μm). DAPI, 4′,6-diamidino-2-phenylindole; GFP, green fluorescent protein; RFP, red fluorescent protein. (M) Representative images and quantification for the FACS analysis after annexin V and PI staining. RTECs were transfected with PXR siRNA followed by pretreatment with PCN. The percentages of apoptotic cells were quantified by FACS (n = 3). Data are presented as means ± SD. Statistically significant differences were determined by one-way ANOVA and two-way ANOVA. #P < 0.05, ##P < 0.01, ###P < 0.001, ####P < 0.0001, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. NC, negative control.

Next, we determined the effect of PXR activation on mitochondrial reactive oxygen species (ROS) production in cisplatin-treated RTECs using the MitoSOX Red Mitochondrial Superoxide Indicator. As shown in Fig. 5D, PCN treatment reduced cisplatin-induced accumulation of mitochondrial ROS. Moreover, reduced mitochondrial membrane potential (MMP), copy number of mtDNA, and ATP synthesis were reversed in PCN-treated RTECs (Fig. 5, D and E). We then examined the oxygen consumption rate (OCR; which represents the basal mitochondrial OXPHOS activity) using a Seahorse analyzer; the decreased OCR and maximal respiratory capacity in cisplatin-treated RTECs were reversed upon exposure to PCN (Fig. 5, F and G). Furthermore, PCN treatment also reversed the dysregulated expression of mitochondrial genes in cisplatin-treated RTECs (fig. S3B). To better evaluate the effect of PXR on mitochondrial homeostasis, we examined some genes with known protective effects on mitochondria (2426) and found that PCN treatment increased the expressions of Pgc-1α, Tfam, Sod2, and Nrf2 (Fig. 5H).

Because timely mitophagy is critical for mitochondrial homeostasis, we investigated the status of mitophagy in our experimental setting. As expected, PCN activated mitophagy, as seen by the increased expressions of Atg3, Atg5, and Atg7 and decreased expression of p62 (Fig. 5I), along with the enhanced expressions of Pink1, Parkin, Fundc1, and Nix, which are the main regulators of mitophagy (Fig. 5J). Furthermore, the reductions in Mfn1, Mfn2, and Opa1 expression (indicative of mitochondrial fusion) were reversed upon exposure to PCN (Fig. 5J). To provide additional evidence for the role of PXR in mitophagy, we transfected RTECs with the pDsRed2-Mito plasmid that is designed to fluorescently label mitochondria. As shown in Fig. 5K, the intensity of red fluorescence decreased in cisplatin-treated RTECs and recovered upon PCN treatment. In addition, we established a stably expressing mt-mKeima-cox8 RTEC line, wherein the fluorescence emitted is converted by the Keima protein, thereby quantitatively reflecting mitophagic activity. As shown in Fig. 5L, the intensity of green fluorescence increased in the cisplatin-treated cells, whereas green fluorescence was replaced by red fluorescence in PCN-treated cells, suggesting an increase in mitophagy. To determine whether PCN protects against cisplatin-induced RTEC injury by activating PXR and rule out off-target effects, we silenced PXR using small interfering RNA (siRNA) and found that this silencing abrogated the inhibition of cisplatin-induced apoptosis and inflammatory cytokine production by PCN (Fig. 5M and fig. S3, C and D).

In accordance with the findings observed upon exposure to PCN, PXR overexpression also prevented cisplatin-induced injury in RTECs (fig. S4, A and B). Moreover, PXR overexpression ameliorated the cisplatin-induced inflammatory response and accumulation of mitochondrial ROS and reversed the cisplatin-induced alterations in MMP, mtDNA copy number, ATP production, and reduced OCR and maximal respiratory capacity (fig. S4, C to G). Furthermore, PXR overexpression also improved mitochondrial homeostasis and mitophagy as shown by the increased expression of Pgc-1α, Tfam, Sod2, Nrf2, Atg3, Atg5, Atg7, Pink1, Parkin, Fundc1, Nix, Mfn1, Mfn2, Opa1, and mitochondrial genes and decreased expression of p62 (fig. S4, H to K). When RTECs were transfected with the pDsRed2-Mito plasmid, the red fluorescence intensity was decreased in cisplatin-treated RTECs but recovered with overexpression of PXR (fig. S4L). Cisplatin-treated stably expressing mt-mKeima-cox8 RTECs exhibited increased green fluorescence, whereas green fluorescence was replaced by red fluorescence in PXR-overexpressing cells, indicating an increase in mitophagy (fig. S4M). We then investigated whether the effect of PXR on cisplatin-induced RTEC apoptosis was mediated by mitophagy. We inhibited mitophagy by treating RTECs with bafilomycin A1 (Baf-A1; an inhibitor of autophagosome-lysosome fusion) and found that Baf-A1 reversed the protective effect of PXR in preventing cisplatin-induced RTEC apoptosis (fig. S4N). In human renal epithelial cells (HK2) cultured in nonadherent Nunclon Sphera six-well plates, we also found a protective role of PXR against cisplatin-induced cell injury (fig. S5, A to D). Collectively, these results demonstrate that activation of PXR directly attenuates cisplatin-induced tubular cell injury and mitochondrial dysfunction, possibly by activating mitophagy and promoting mitochondrial fusion.

AKR1B7 was the transcriptional target of PXR in the kidney

We performed an iTRAQ-based quantitative proteomic analysis to assess differential protein expression in the kidneys of wild-type and PXR−/− rats. Compared to wild-type rats, PXR−/− rats exhibited 152 differentially expressed proteins (up-regulated ≥1.5-fold or down-regulated ≤0.67-fold, P < 0.05), including 47 down-regulated proteins and 105 up-regulated proteins (Fig. 6A). In the kidneys of PXR−/− rats, other nuclear receptors detected by this analysis did not change in response to PXR deficiency (fig. S6A). qRT-PCR showed that the mRNA expressions of some nuclear receptors also remained unchanged (fig. S6B). Among the differentially expressed proteins identified, AKR1B7 was markedly down-regulated in the kidneys of PXR−/− rats as compared to its expression in the kidneys of wild-type rats (Fig. 6B). Moreover, it has been previously shown that mouse AKR1B7 is a transcriptional target of PXR in the liver and intestine (20). Thus, we decided to examine whether AKR1B7 is a downstream transcriptional target of PXR in our experimental setting.

Fig. 6 Akr1b7 was the transcriptional target of PXR in kidneys and protected against CP-induced AKI in vivo.

(A) Volcano plot for differentially expressed proteins in the kidneys from WT and PXR−/− rats based on iTRAQ-based quantitative proteomic analysis (up-regulated ≥1.5-fold or down-regulated ≤0.67-fold, P < 0.05). (B) Heat map for differentially expressed mitochondrial-related proteins in the kidneys from WT and PXR−/− rats based on iTRAQ-based quantitative proteomic analysis. (C) Akr1b7 mRNA expression in the kidneys of WT and PXR−/− rats with or without CP (n = 6). (D) Akr1b7 mRNA expression in RTECs stably expressing PXR with or without CP (n = 3). (E) RTECs were cotransfected with a 2-kb-long pGL3-Akr1b7 reporter together with PXR vectors. Subsequently, transfected cells were treated with PCN for 24 hours before the luciferase assay (n = 5). DMSO, dimethyl sulfoxide. (F) In situ hybridization of AKR1B7 with a Cy3-labeled probe in RTECs (n = 3; scale bars, 10 μm). FISH, fluorescence in situ hybridization. (G) qRT-PCR analysis of Akr1b7 expression in the kidney of CP-induced AKI mice (days 0, 1, 2, and 3 after CP injection; n = 6). (H) Western blotting for FLAG-tagged proteins in the kidneys 36 hours after injection of the AKR1B7 plasmids and scrambled plasmids. (I) qRT-PCR analysis of renal Akr1b7 expression 36 hours after injecting the AKR1B7 plasmids (n = 4). (J) SCr and BUN concentrations were measured in CP-induced AKI mice injected with AKR1B7 or scrambled plasmids. (K) ELISA for circulating IL-6 and TNF-α. (L) qRT-PCR analyses for renal Kim-1 and Ngal mRNA expressions. (M) qRT-PCR analyses for renal Il-1β, Il-6, and Tnf-α mRNA expressions. (N) qRT-PCR analyses for renal Bax and Bcl-2 mRNA expressions. (O) Representative images and quantification for PAS (scale bars, 50 μm), TUNEL (scale bars, 10 μm; red arrows indicate the TUNEL-positive cells), and immunohistochemistry staining of Cyt-c and F4/80 (scale bars, 20 μm) in the kidneys and representative electron micrographs of renal mitochondria (scale bars, 1 μm; red asterisks denote the damaged mitochondria). Six random fields were taken from each kidney. Data are expressed as mean ± SD; NC, n = 6; NC + CP, n = 8; AKR1B7 + CP, n = 8. Statistically significant differences were determined by Student’s t test, one-way ANOVA, and two-way ANOVA. #P < 0.05, ##P < 0.01, ###P < 0.001, ####P < 0.0001, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. NC, negative control.

To further confirm that AKR1B7 is the transcriptional target of PXR in the kidney, we detected Akr1b7 mRNA expression in the kidneys of PXR−/− rats and cisplatin-treated RTECs stably expressing PXR. We found a reduction of Akr1b7 in PXR−/− rats in contrast to an up-regulation of Akr1b7 in cisplatin-treated PXR-overexpressing RTECs (Fig. 6, C and D). To further determine the effect of PXR on regulating Akr1b7 gene expression, we cotransfected RTECs with a 2-kb-long pGL3-Akr1b7 reporter gene and PXR vectors. Subsequently, transfected cells were treated with PCN for 24 hours before performing the luciferase assay. As shown in Fig. 6E, pGL3-Akr1b7 was activated by PCN in RTECs cotransfected with PXR vectors, whereas the mutation of Akr1b7 resulted in the loss of PXR effect.

Next, to investigate whether AKR1B7 is involved in cisplatin-induced AKI, we performed in situ hybridization of AKR1B7 using a Cy3-labeled probe in RTECs. As shown in Fig. 6F, the red fluorescence was primarily distributed in the cytoplasm, and the intensity of this fluorescence decreased in cisplatin-treated RTECs but recovered upon treatment with PCN. Consistent with the data for PXR, the expression of Akr1b7 in the kidneys of cisplatin-treated mice decreased in a time-dependent manner (Fig. 6G). In cultured mouse RTECs, cisplatin decreased the amounts of Akr1b7 in a dose- and time-dependent manner (fig. S7A).

AKR1B7 protected kidneys from cisplatin-induced AKI

To investigate the role of AKR1B7 in cisplatin-induced AKI, we used the tail vein high-pressure injection method to deliver AKR1B7 plasmids into mice. We determined the expression of AKR1B7 in kidney tissues 36 hours after plasmid delivery. As shown in Fig. 6 (H and I), the FLAG-tagged protein and mRNA expression of AKR1B7 in the kidney cortex markedly increased 36 hours after injection of the AKR1B7 plasmids. Then, mice were treated with cisplatin (intraperitoneally) after delivery of AKR1B7 plasmids for 36 hours. Consistent with the results of PXR modulation in animals, tail vein high-pressure injection of AKR1B7 plasmids markedly improved renal function in AKI mice compared to that in mice injected with scrambled plasmids; we observed lower concentrations of BUN, SCr, and the circulating proteins IL-6 and TNF-α (Fig. 6, J and K). Consistent with these results, the mRNA expression patterns of the tubular injury markers (Kim-1 and Ngal), the apoptosis-associated proteins (Bax and Bcl-2), and the inflammation factors (IL-1β, IL-6, and TNF-α) in the kidneys were reversed upon overexpressing AKR1B7 (Fig. 6, L to N). Compared to cisplatin treatment alone, overexpression of AKR1B7 with cisplatin treatment inhibited the release of mitochondrial Cyt-c, reduced the number of TUNEL-positive cells and F4/80-positive cells, and ameliorated mitochondrial abnormalities (Fig. 6O). Collectively, these results demonstrate the benefits of AKR1B7 in cisplatin-induced AKI.

AKR1B7 attenuated cisplatin-induced mitochondrial dysfunction

AKR1B7 was reported to play a role in modulating mitochondrial function (9). Our findings also indicated improved mitochondrial morphology in AKI mice overexpressing AKR1B7 (Fig. 6O). Thus, we investigated the effects of AKR1B7 on cell death and mitochondrial function in RTECs. Consistent with our in vivo results, AKR1B7 overexpression prevented cisplatin-induced cell injury—as evidenced by the decrease in cell apoptosis; the reduction of BAX, cleaved caspase-3, LC3I, LC3II, and Cyt-c; and the restoration of BCL-2—and inhibited the inflammatory response (Fig. 7, A to C). AKR1B7 overexpression reduced cisplatin-induced accumulation of mitochondrial ROS, accompanied by a recovery of the MMP, mtDNA copy number, and ATP production in RTECs (Fig. 7, D and E). Furthermore, the OCR and maximal respiratory capacity decreased in cisplatin-treated RTECs but improved after overexpressing AKR1B7 (Fig. 7, F and G). Moreover, AKR1B7 overexpression also reversed the reduced expression of the mitochondrial genes, Pgc-1α, Tfam, Sod2, and Nrf2, observed in RTECs treated with cisplatin (Fig. 7H and fig. S7B). In situ analysis and mitochondrial isolation showed that Akr1b7 mRNA was detectable in mitochondria (fig. S7, C and D); using RNA sequencing (RNA-seq), we found that AKR1B7 overexpression enhanced the expression of mitochondrial protective genes, including Ldhb, Enpp1, Eya2, and Lars2 (fig. S7E) (27, 28), suggesting a mitochondrial function of AKR1B7.

Fig. 7 AKR1B7 overexpression attenuated CP-induced mitochondrial dysfunction in vitro.

(A) Representative images for FACS analysis after annexin V and PI staining. AKR1B7 was overexpressed in RTECs followed by incubation with CP (5 μg/ml) for 24 hours. The percentages of apoptotic cells were quantified by FACS (n = 3). (B) Western blotting for BAX, BCL-2, LC3I, LC3II, and Cyt-c in CP-treated RTECs with or without AKR1B7 overexpression. (C) qRT-PCR analyses for Il-1β, Il-6, and Tnf-α mRNA expressions. (D) FACS analysis for the accumulation of ROS in the mitochondria and MMP using the MitoSOX Red Mitochondrial Superoxide Indicator and JC-1 fluorescent probe in RTECs, respectively. qRT-PCR analysis showed the copy number of the mtDNA in RTECs. 18S was used as an internal control (n = 3). (E) Cellular ATP content was determined by luciferase assay (n = 3). (F) Seahorse 96xf was used to detect basal OCR (n = 8). (G) Seahorse 96xf was used to detect spare respiratory capacity (n = 8). (H) qRT-PCR analysis for Pgc-1α, Tfam, Sod2, and Nrf2 mRNA expressions (n = 3). (I) qRT-PCR analysis for p62, Atg3, Atg5, and Atg7 mRNA expressions (n = 3). (J) qRT-PCR analysis for Pink1, Parkin, Mfn1, Mfn2, Opa1, Fundc1, and Nix mRNA expressions (n = 3). (K) RTECs were transfected with the pDsRed2-Mito plasmid followed by the detection of red fluorescence (n = 3; scale bars, 10 μm). Data are expressed as mean ± SD. Statistically significant differences were determined using one-way ANOVA and two-way ANOVA. #P < 0.05, ##P < 0.01, ###P < 0.001, ####P < 0.0001, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. NC, negative control.

Normal mitochondrial function depends on mitophagy and proper mitochondrial dynamics (5, 29). We observed that overexpressing AKR1B7 increased the expression of Atg3, Atg5, Atg7, Pink1, Parkin, Fundc1, and Nix and decreased that of p62, suggesting activated mitophagy (Fig. 7, I and J). Moreover, the reduction in the expression of the mitochondrial fusion–related genes Mfn1, Mfn2, and Opa1 by cisplatin treatment was reversed by AKR1B7 overexpression (Fig. 7J). To provide additional evidence for the activation of mitophagy by AKR1B7, we transfected RTECs with the pDsRed2-Mito plasmids. As shown in Fig. 7K, the red fluorescence decreased upon cisplatin treatment but recovered in the AKR1B7-overexpressing RTECs. Next, we studied whether AKR1B7-mediated protection against cisplatin-induced RTEC apoptosis involved the activation of mitophagy. We inhibited mitophagy by treating RTECs with Baf-A1 and found that Baf-A1 largely reversed the protective effect of AKR1B7 on preventing cisplatin-induced RTEC apoptosis (fig. S7F). Collectively, these results demonstrate that AKR1B7 overexpression attenuates cisplatin-induced mitochondrial dysfunction by activating mitophagy and promoting mitochondrial fusion.

PXR/AKR1B7 activation attenuated cisplatin-induced lipid accumulation and improved mitochondrial fatty acid β-oxidation

Using iTRAQ-based quantitative proteomic analysis, we observed that PXR deficiency affected the metabolic pathways in the kidney the most (fig. S8, A and B). To further investigate this phenomenon, we performed orthogonal partial least squares discriminant analysis (OPLS-DA). The OPLS-DA score plots showed that the data from the PXR−/− and wild-type rats were completely separated (positive ion mode: R2X = 0.22, R2Y = 0.893, Q2 = 0.668) (fig. S9A), indicating metabolic dysregulation in the PXR−/− rats. The color-coded loading plots revealed differences in the metabolites between the PXR−/− and wild-type rats (fig. S9B). Rats with PXR deficiency showed a marked decrease in the concentrations of lactic acid, l-(+)-arginine, 5-aminopentanoic acid, glutamine, l-aspartate, threonine, l-(+)-arginine, proline, lauric acid, acetylenedicarboxylic acid, tyrosine, histidine, d-(+)-malic acid, etc. (fig. S9B), suggesting that PXR deficiency results in abnormal metabolism in the kidney.

Furthermore, we investigated the effect of PXR deficiency on lipid metabolism in the kidney using untargeted lipidomics analysis. As shown in fig. S9C from lipidomic analysis, the PXR−/− and wild-type rats were completely separated in the OPLS-DA score plots (positive ion mode: R2X = 0.229, R2Y = 0.826, Q2 = 0.568), indicating dysregulated lipid metabolism in the PXR−/− rats. Hierarchical cluster analysis of the metabolomics data from the PXR−/− and wild-type rats is shown in fig. S9D. The concentrations of TG(16:0/18:0/16:0), ubisemiquinone (coenzyme Q10), PE(22:1(13Z)/15:0), Cer(d18:1/24:1(15Z)), Pectachol, PC(22:6(4Z, 7Z, 10Z, 13Z, 16Z, 19Z)/16:1(9Z)), PE(P-18:1(9Z)/20:3(5Z, 8Z, 11Z)), and PE(P-18:1(9Z)/18:3(9Z, 12Z, 15Z)) were decreased in the kidneys of PXR−/− rats, which could weaken the antioxidant activity in the kidney (3032), suggesting that PXR deficiency increases oxidative stress resulting in aberrant metabolism of lipids and lipid complexes in the kidney.

Fatty acid oxidation is the main source of energy in RTECs and primarily occurs in the mitochondria and peroxisomes (33). Consistent with the lipidomics data, we observed that PXR deficiency markedly decreased renal Cpt-1α, Mcad, Hmgcs2, Pgc-1α, and Acox1 mRNA expressions, suggesting disrupted fatty acid β-oxidation in the kidneys of PXR−/− rats (fig. S9E). In addition, the Cpt-1α, Mcad, and Acox1 mRNA expressions in cisplatin-treated PXR−/− rats were markedly lower than those in cisplatin-treated wild-type rats (fig. S9F); moreover, accumulation of lipids in the kidneys of cisplatin-treated PXR−/− rats was higher than that in the kidneys of cisplatin-treated wild-type rats (fig. S9G). Consistent with these findings, PCN treatment or PXR overexpression markedly attenuated lipid accumulation in the kidneys of cisplatin-treated mice (fig. S9, H and I), in agreement with decreased fatty acid β-oxidation (fig. S9, J and K). To further determine whether PXR protects against cisplatin-induced mitochondrial metabolic damage by activating AKR1B7, we silenced AKR1B7 and found that AKR1B7 silencing abrogated the protective effects of PCN and PXR overexpression against cisplatin-induced impairment of mitochondrial oxidative capacity and fatty acid β-oxidation (fig. S10). Together, these results demonstrate that PXR/AKR1B7 activation ameliorates cisplatin-induced lipid accumulation possibly by improving mitochondrial metabolism and fatty acid oxidation.

PXR deficiency exacerbated I/R-induced AKI

Last, we investigated the role of PXR in ischemia/reperfusion (I/R)–induced AKI using PXR-deficient (PXR−/−) rats. Consistent with the results from the cisplatin-induced AKI model, the SCr and BUN concentrations, along with Kim-1 and Ngal mRNA expressions, were higher in the kidneys of PXR−/− rats after I/R challenge than those in the kidneys of wild-type rats subjected to I/R (Fig. 8, A and B). Moreover, compared to the wild-type rats, PXR−/− rats exhibited exacerbated tubular injury and an increased number of TUNEL-positive cells after I/R challenge (Fig. 8, C and D). Furthermore, PXR deficiency aggravated the decrease in the expression of the mitochondrial genes in the kidneys of PXR−/− rats after I/R challenge (fig. S11).

Fig. 8 PXR deficiency exacerbated I/R-induced AKI.

(A) SCr and BUN concentrations were measured in WT and PXR−/− rats with or without I/R. (B) qRT-PCR analyses for renal Kim-1 and Ngal mRNA expressions. (C) Representative images for PAS staining (scale bars, 50 μm) in the kidneys (left). Tubular injury scores in rats were analyzed (right). Six random fields were taken from each kidney. (D) Representative images and quantification for TUNEL staining (scale bars, 10 μm; red arrows indicate the TUNEL-positive cells) in the kidneys. Six random fields were taken from each kidney. (E) SCr and BUN concentrations were measured in I/R-induced AKI mice model with or without PCN administration. (F) qRT-PCR analyses for renal Kim-1 and Ngal mRNA expressions. (G) Representative images for PAS staining (scale bars, 50 μm) in the kidneys (left). Tubular injury scores in rats were analyzed (right). Six random fields were taken from each kidney. Data are expressed as mean ± SD (n = 6). Statistically significant differences were determined by Student’s t test and one-way ANOVA. #P < 0.05, ##P < 0.01, ####P < 0.0001, *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

To further examine the effects of PXR activation on I/R-induced AKI, mice were administered PCN (50 mg/kg per day, intraperitoneally) for 2 days before I/R. In contrast to the results in PXR−/− rats, the administration of PCN resulted in reduced SCr and BUN concentrations, along with reduced mRNA expressions of Kim-1 and Ngal and improved kidney histology in mice subjected to I/R (Fig. 8, E to G). Collectively, these results demonstrate that the beneficial effect of PXR activation can be extended to I/R-induced AKI, suggesting a potential therapeutic use of PXR in managing AKI.

DISCUSSION

Previous studies have demonstrated that nuclear receptor superfamily members, such as farnesoid X receptor (FXR), liver X receptor (LXR), and Nur77, attenuate acute renal injury by reducing inflammation and oxidative stress and improving renal lipid metabolism (3436). However, no research has been conducted to demonstrate the role of PXR in acute renal injury. A review of the literature indicated that in addition to its roles in the modulation of metabolic enzymes, PXR is also reported to be a critical regulator of glycolipid metabolism, inflammation, and atherosclerosis (37, 38). We also observed an abnormal metabolic status in PXR−/− rats’ kidneys by analyzing metabolomics and proteomics. Furthermore, we performed experiments using genetic and pharmacological approaches and provided insight into the role of PXR in the pathogenesis of AKI. Here, we reported the pathogenic effect of PXR deficiency on acute renal tubular injury and mitochondrial dysfunction in vivo and in vitro. Overexpression or activation of PXR potently inhibited mitochondrial damage and improved AKI. All these in vivo and in vitro experimental data suggested a protective role of PXR in regulating mitochondrial metabolism and mitochondrial homeostasis, helping protect the kidneys against acute injury.

By proteomics analysis, we found differential mitochondria-related protein expressions in the kidneys from PXR−/− rats, suggesting a role of PXR in regulating mitochondrial function. Next, we identified Akr1b7, which encodes an NAD(P)H-linked oxidoreductase, as a specific target of PXR in antagonizing renal tubular cell apoptosis and mitochondrial dysfunction. We also confirmed that AKR1B7 can be expressed in mouse kidneys. AKR1B7 has alcohol dehydrogenase (NADP+) and oxidoreductase activities, along with actions on cellular lipid metabolism and oxidation-reduction processes (20, 22). One of the major functions of AKR1B7 is to detoxify lipid peroxidation products (20). However, there had been no direct evidence that activating AKR1B7 could attenuate acute renal injury and mitochondrial dysfunction. Here, we observed that Akr1b7 was the transcriptional target of PXR in the kidney and confirmed this finding by proteomics, gene regulation experiments, and luciferase assays. Consistent with the results for PXR activation, AKR1B7 overexpression also markedly reduced cisplatin-induced renal tubular injury. Further research demonstrated that transcriptional regulation of Akr1b7 by PXR improved the biogenesis of mitochondria in renal tubular cells after acute insult. Considering that the known function of AKR1B7 is to detoxify lipid peroxidation products, it will be necessary to investigate whether PXR-controlled AKR1B7 contributes to alleviating the toxicity associated with cellular metabolism of lipids.

Given that renal tubular cells are rich in mitochondria, these cells are anticipated to be the main target of insults that cause AKI, including cisplatin nephrotoxicity (39). Mitochondria are the main target organelles in acute tubular cell damage, which is characterized by reduced MMP and OCR and impaired mitophagy and fatty acid β-oxidation (6). In agreement with the above concept, many studies have focused on improving the mitochondrial function to ameliorate AKI (10, 40, 41). Here, we found that a number of parameters associated with mitochondrial function, including the MMP, OCR, the expression of mitochondrial genes, Cyt-c oxidase activity, mitophagy, mitochondrial morphology, and mitochondrial fatty acid β-oxidation, were markedly altered after the modulation of PXR or AKR1B7 in vitro and in vivo, indicating that activation of the PXR/AKR1B7 signaling pathway can ameliorate cisplatin-induced mitochondrial dysfunction.

Mitochondrial dysfunction is not only a pathological phenomenon but also a pathogenic factor, resulting in oxidative stress, subsequent inflammation, and tubular cell injury (7, 42). Mitochondrial damage causes mtDNA leakage into the cytosol to activate cyclic guanosine 5′-monophosphate–adenosine 5′-monophosphate synthase (cGAS)–stimulator of interferon genes (STING) signaling, resulting in tubular inflammation and fibrotic response (43). We found that PXR/AKR1B7 signaling protected against both tubular cell inflammation and immune cell infiltration in AKI. However, the more detailed mechanism of PXR in ameliorating immune cell infiltration needs to be investigated in the future.

Besides the cisplatin model, we also extended the study to an I/R AKI model. Consistent with the findings from the animals with cisplatin AKI, deletion of PXR in rats also aggravated kidney injury, whereas the activation of PXR in mice protected kidney function against acute I/R injury. These results further suggested that PXR could serve as a common therapeutic target in various types of AKI.

A limitation of our study is that the roles of PXR in AKI improvement were mainly observed in mouse and cell models but not in humans. Although we did detect the down-regulation of PXR and its negative correlation with the degree of kidney injury in patients with AKI from ischemic or other nephrotoxic insults, the effectiveness and importance of PXR in protecting against human AKI remain to be further explored by a clinical trial of PXR agonist in the future.

In conclusion, this study demonstrated that the activation of the nuclear receptor PXR markedly attenuated AKI possibly by ameliorating mitochondrial dysfunction in ROS production, mitophagy, and lipid metabolism via transcriptional regulation of Akr1b7. These findings from the current research not only increase our understanding of the pathogenesis of AKI but also provide therapeutic potential by activation of the PXR/AKR1B7 pathway. Developing effective and safe agonists or activators of PXR and AKR1B7 and performing clinical trials of those agents in patients with AKI could offer a viable approach for AKI therapy.

MATERIALS AND METHODS

Study design

The objectives of this study were to examine the role of PXR in the pathogenesis of AKI and the underlying mechanisms. To achieve these objectives, we designed a study using human kidney tissues, AKI models in rats and mice, and in vitro models of renal tubular cell injury to determine the role of PXR and the potential mechanisms. For the human study, we enrolled 20 patients with AKI and acute tubular necrosis verified by renal biopsy from the Peking University First Hospital. AKI was diagnosed according to the Kidney Disease: Improving Global Outcomes (KDIGO) criteria (44). Para-carcinoma kidney tissues from six patients with no evidence of kidney injury were used as controls for the analysis of PXR expression in the kidney tissues. For animal studies, animals were chosen at random for the vehicle or treatment groups. A power analysis was used to calculate the necessary sample sizes of animal experiments to achieve reliable measurements. For in vitro studies, a minimum of three experimental replicates was performed. The numbers of replicates are presented in the figure legends. Human and animal kidney samples were carefully evaluated by two independent pathologists in a blinded manner. Statistical tests were chosen on the basis of nature of variables, assumption of data distribution, and effect size. Primary data are reported in data file S1.

Human specimens

Twenty patients from the Peking University First Hospital with AKI and acute tubular necrosis verified by renal biopsy were enrolled in the study. Clinical parameters, including age, sex, cause of AKI, and SCr concentrations at the peak and at the time of renal biopsy, were collected. AKI was diagnosed using the KDIGO criteria. Para-carcinoma kidney tissues from six patients who underwent renal carcinoma resection, with no evidence of kidney injury seen on routine pathological examination, were used as controls for the PXR assay. The protocol concerning the use of the patient samples in this study was approved by the Committee on Research Ethics of the Peking University First Hospital and by the Human Subjects Committee of the Nanjing Medical University. Informed consent was obtained from all participants.

Animal studies

To evaluate the effect of the PXR agonist sodium PCN (Cayman Chemical) on cisplatin-induced AKI, 8-week-old male C57BL/6 mice were assigned to three groups (control, n = 6; cisplatin-treated mice, n = 6; cisplatin plus PCN-treated mice, n = 6). Mice were pretreated with PCN (50 mg/kg per day) by intraperitoneal injection for 2 days, followed by a single intraperitoneal injection of cisplatin (20 mg/kg; Sigma-Aldrich). PCN was continuously administered for the next 3 days, and then the animals were euthanized. To confirm the effect of the PXR gene on cisplatin-induced AKI, 8-week-old male wild-type and PXR−/− Sprague-Dawley (SD) rats, provided by the Model Animal Research Center of the Nanjing University, were divided into four groups (vehicle-treated wild-type rats, n = 6; vehicle-treated PXR−/− rats, n = 6; cisplatin-treated wild-type rats, n = 6; cisplatin-treated PXR−/− rats, n = 6). Rats were intraperitoneally injected with cisplatin (7.5 mg/kg, single injection). Three days later, rats were sacrificed after sodium pentobarbital injection (30 mg/kg, intraperitoneally). To confirm the effect of PXR in another AKI model, induced by I/R, 8-week-old male SD rats were divided into four groups (sham wild-type rats, n = 6; sham PXR−/− rats, n = 6; I/R wild-type rats, n = 6; I/R PXR−/− rats, n = 6). I/R was induced by clamping bilateral renal arteries for 35 min. Rats in the sham group were subjected to the same procedure except the vascular clamping. After I/R for 24 hours, rats were euthanized after pentobarbital sodium injection (30 mg/kg, intraperitoneally). To further validate the roles of PXR and AKR1B7 in cisplatin-induced AKI, 60 μg of PXR plasmids (pLVX-mCherry-c1-PXR-Flag), AKR1B7 (pLenti-EF1a-AKR1B7-3Flag-CMV-GFP-Puro), and negative control plasmids (pLVX-mCherry-c1 or pLenti-EF1a-CMV-GFP-Puro) were administered to mice within 10 s via tail vein injection. After 36 hours, cisplatin was administered as described above. In a MnTBAP treatment study, wild-type and PXR−/− rats were pretreated with MnTBAP at a dose of 10 mg/kg by intraperitoneal injection, and then the rats received cisplatin treatment 24 hours later as described above. After the animals were euthanized, the blood was collected, and the isolated serum was stored at −80°C. Kidney tissues for histological analyses were fixed in 4% paraformaldehyde. The remaining kidney tissues were stored at −80°C for mRNA and protein analyses. All animal procedures were approved by the Nanjing Medical University Institutional Animal Care and Use Committee (registration number: IACUC-1809017).

Statistical analysis

The raw data were converted to ABF format by Reifycs Abf (analysis base file) Converter. Alignment was performed using MS-DIAL version 2.76 (45) based on the m/z (mass/charge ratio) values and the retention times of the ion signals. Identified metabolites from both electrospray ionization (ESI) and ESI+ modes were merged and imported into MetaboAnalyst 3.0 (46) for multivariate analysis. OPLS-DA was used to identify potential metabolic biomarkers by comparison between wild-type and PXR−/− rats. The biomarkers were filtered and confirmed by combining the variable importance (VIP) values (VIP > 1), t test results (P < 0.05), and fold change values (FC > 2). The quality of the model fit can be explained by the R2 and Q2 values. R2 represents the variance explained in the model and indicates the quality of the fit. Q2 represents the variance in the data, indicating the model’s predictability. Hierarchical clustering of signature metabolites altered in PXR−/− rats compared to wild-type rats was performed in MetaboAnalyst 2.0. The color intensity correlates with the degrees of increase (pink) and decrease (blue) relative to the mean metabolite ratio. Other experimental data are expressed as the mean ± SD of triplicate experiments performed in parallel unless otherwise indicated. Statistically significant differences were determined by analysis of variance (ANOVA), followed by a Bonferroni multiple comparisons test or Student’s t test using GraphPad Prism 6 software. A value of P < 0.05 was considered significant.

SUPPLEMENTARY MATERIALS

stm.sciencemag.org/cgi/content/full/12/543/eaay7591/DC1

Materials and Methods

Fig. S1. PXR deficiency aggravated cisplatin-induced tubular inflammation and mitochondrial damage.

Fig. S2. MnTBAP therapy attenuated cisplatin-induced AKI in wild-type and PXR−/− rats.

Fig. S3. PCN treatment ameliorated cisplatin-induced apoptosis, inflammation, and the dysregulation of mitochondrial genes in vitro.

Fig. S4. PXR overexpression attenuated cisplatin-induced apoptosis and mitochondrial dysfunction in vitro.

Fig. S5. PCN treatment or PXR overexpression attenuated cisplatin-induced apoptotic response in three-dimensional cultured HK2 cells.

Fig. S6. Expressions of other nuclear receptors remained unchanged in PXR−/− rats.

Fig. S7. AKR1B7 overexpression prevented cisplatin-induced mitochondrial injury in vitro.

Fig. S8. PXR deficiency changed mitochondrial metabolic pathways.

Fig. S9. PXR/AKR1B7 activation attenuated cisplatin-induced lipid accumulation and improved mitochondrial fatty acid β-oxidation.

Fig. S10. The protective effect of PXR on mitochondrial oxidative capacity and fatty acid β-oxidation was diminished by silencing AKR1B7.

Fig. S11. PXR deficiency aggravated I/R-induced mitochondrial damage.

Table S1. Clinical data of patients with AKI.

Data file S1. Primary data.

Data file S2. Proteomics analysis of all differential proteins between wild-type and PXR−/− groups.

Data file S3. Untargeted lipidomics analysis of all differential lipids between wild-type and PXR−/− groups.

Data file S4. Untargeted metabolomics analysis of all differential metabolites between wild-type and PXR−/− groups.

Data file S5. RNA-seq analysis of all differential transcriptomes between NC and AKR1B7-overexpressed RTECs.

Data file S6. Primers used for PCR amplification.

References (4754)

REFERENCES AND NOTES

Acknowledgments: We thank the Peking University First Hospital for provision of tissue samples and clinical data. Funding: This work was supported by grants from the National Natural Science Foundation of China (81325004, 81830020, 81530023, 91742205, and 81625004 to A.Z.; 81700604 to X.Y.; 81570616 to Y.Z.; and 81670647 and 81873599 to Z.J.), the National Key Research and Development Program (2016YFC0906103 to A.Z.), and the Natural Science Foundation of Jiangsu Province (BK20141079 to S.H., BL2014007 to A.Z., and BK20170148 to X.Y.). Author contributions: X.Y., Z.J., Y.Z., and A.Z. designed the study. X.Y., X.M., M.X., S.L., M.B., and Y.Z. performed the experiments. X.Y., X.M., M.X., R.Y., and Q.L. analyzed the data. X.Y., L.Y., S.H., Z.J., and A.Z. interpreted the results. X.Y., Z.J., and A.Z. wrote the manuscript. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data associated with this study are present in the paper or the Supplementary Materials.

Stay Connected to Science Translational Medicine

Navigate This Article