Pregnane X receptor alleviates sepsis-induced liver injury through activation of yes-associated protein in mice

Abstract

The severity of sepsis is attributed to excessive inflammatory responses leading to liver injury. Pregnane X receptor (PXR), a nuclear receptor that controls xenobiotic and endobiotic metabolism, has been implicated in regulating inflammation and liver regeneration. This study aimed to investigate the role of PXR in sepsis-induced liver injury and the underlying mechanisms. Sepsis models were established in mice, the mice were administered the typical mouse PXR agonist PCN (100 mg·kg⁻¹·d⁻¹, i.p.) for 3 consecutive days in advance, then subjected to CLP operation or LPS administration 1 h after the last administration of PCN. The results showed that PCN pretreatment significantly increased the survival rate of septic mice, while the survival rate was reduced after the knockout of Pxr. In addition, PCN pretreatment effectively alleviated sepsis-induced liver injury. In Pxr knockout mice, liver injury was more severe, whereas the protective effects of PCN pretreatment were abolished. Mechanistically, PCN pretreatment significantly upregulated the expression of yes-associated protein (YAP) and its downstream targets and decreased the level of phosphorylated nuclear factor-κB (NF-κB). Moreover, liver-specific knockdown of Yap blocked the protective effects of PCN pretreatment against sepsis-induced liver injury and downregulated the phosphorylation level of NF-κB. In summary, this study demonstrated that PXR activation protects against sepsis-induced liver injury through activation of the YAP signaling pathway, providing a new strategy for the diagnosis and treatment of sepsis-induced liver injury.

Introduction

Sepsis is defined as a life-threatening multi-organ dysfunction caused by a dysregulated host response to infection [1, 2]. Sepsis affects approximately 50 million individuals worldwide each year, with a mortality rate of 20% to 30% [3]. Sepsis-related morbidity and mortality are associated with uncontrolled inflammation, coagulation, and hemodynamic compromise, which may lead to multiple organ system dysfunction and failure [2]. The liver plays a pivotal role in maintaining host defense mechanisms against pathogens and maintaining metabolic homeostasis [4]. Epidemiological studies reveal that 45.9% of sepsis patients develop liver injury, with hepatorenal failure significantly exacerbating sepsis severity and adversely impacting clinical outcomes [5, 6]. Unfortunately, current therapeutic approaches  for this life-threatening complication remain supportive rather than curative. Hence, investigating the pathogenesis of sepsis-induced multi-organ injury, particuly liver injury, is crucial for more effective prevention and treatment strategies against sepsis.

The pregnane X receptor (PXR, also known as NR1I2) is a ligand-activated transcription factor belonging to the nuclear receptor superfamily, primarily expressed in hepatocytes and intestinal epithelial cells [7]. PXR serves a crucial function in regulating endogenous and exogenous metabolism, including bile acids, lipids, steroids, glucose, and drugs, which is closely related to liver physiopathology [8, 9]. Previous research has investigated the positive role of PXR in modulating inflammatory diseases, including cholestatic liver disease, inflammatory liver disease, and inflammatory bowel disease [10–12]. For instance, Pxr knockdown exacerbated D-galactosamine (D-GalN)/lipopolysaccharide (LPS)-induced acute liver injury, whereas overexpression of Pxr protected against liver injury [11]. Our recent study found that the PXR agonist pregnenolone-16α-carbonitrile (PCN) could attenuate D-GalN/LPS-induced liver injury by regulating macrophage polarization [13]. Wang et al. further reported that indole-3-propionic acid attenuates sepsis-induced liver injury through PXR upregulation [14]. Nevertheless, a comprehensive comprehension of the contribution of PXR in the modulation of sepsis-induced liver injury is lacking, and the involved molecular signals remain poorly characterized.

Yes-associated protein (YAP) serves as a key downstream effector of the Hippo kinase cascade pathway, governing essential biological processes including cell proliferation, organ regeneration, inflammatory responses, and tumorigenesis [15–17]. Previous studies have found that YAP suppresses vascular inflammation via obstructing tumor necrosis factor receptor-associated factor 6-mediated NF-κB activation, thereby mitigating sepsis progression [18]. Meanwhile, YAP also inhibits the development of cecal ligation and puncture (CLP)-induced ferroptosis in hepatocytes, consequently reducing liver injury and mortality in septic mice [19]. Recently, we also verified that PXR directly interacts with YAP, where PXR activation enhances YAP-transcriptional enhancer factor domain family member (TEAD) binding and upregulates YAP downstream targets, ultimately facilitating physiological liver enlargement and regeneration [20, 21]. However, the precise role of PXR in mitigating sepsis-induced liver injury through the YAP signaling pathway remains to be elucidated.

Therefore, the current study intended to investigate the role of  PXR in sepsis-induced liver injury and investigate its underlying molecular mechanisms. The data revealed a promising new function of PXR in ameliorating liver injury and enhancing the survival of septic mice. Importantly, the YAP signaling pathway emerged as a critical mediator of PXR’s protective effects against sepsis-induced liver injury. Collectively, these results underscore the significance of PXR in sepsis pathophysiology and highlight its potential as a therapeutic target for managing hyperinflammatory conditions. This work provides critical mechanistic insights into the PXR-YAP axis, providing a strategic framework for developing targeted strategies to mitigate sepsis-induced organ dysfunction and improve clinical outcomes in septic patients.

Materials and methods

Animals

Male six to eight-week-old C57BL/6 J mice were acquired from GemPharmatech (Guangzhou, China) for all animal experiments. The Pxr^-/- mouse was gifted by Prof Fei Li of the Sichuan University of China. The mice were housed in a temperature-controlled colony room with ad libitum access to standard laboratory chow and water, under a 12-h light/dark cycle. All experimental protocols were approved by the local Animal Care and Use Committee of Southern Medical University, Guangzhou, China (Approval number: SMUL2023078).

Animal experiments

Sepsis models were established through CLP operation or LPS administration, following established protocols [22–24]. Briefly, mice were anesthetized with isoflurane (RWD, Shenzhen, China) before undergoing a 1–2 cm midline incision was made under aseptic conditions to expose the cecum. The distal portion of the cecum was ligated and punctured once with an 18-gauge needle (Zhejiang, China), allowing minimal fecal extrusion. Subsequently, the peritoneum of the mice was sutured, and fluid resuscitation was performed by injecting 1 mL of sterile saline (Beijing, China) subcutaneously. Finally, the mice were rewarmed by a heating pad until awakening. The LPS-induced sepsis model was induced by intraperitoneal injection of 20 mg/kg LPS (Lot#: L2630, Sigma-Aldrich, Germany). Furthermore, mice were anesthetized and sacrificed 12 h after CLP operation or LPS injection by intraperitoneal injection of 50 mg/kg pentobarbital sodium (Lot#: P3761, Sigma-Aldrich, Germany), with subsequent collection of blood and tissue samples. Survival rates were monitored continuously for 72 h after the CLP operation or LPS injection. To explore the function of PXR in septic mice, in the typical mouse PXR agonist PCN (100 mg·kg⁻¹·d⁻¹, dissolved in corn oil, Lot#: C3884, APExBIO, Texas, USA) was intraperitoneally administered for three consecutive days in advance, and induced CLP or LPS model 1 h after the last administration of PCN [21, 25]. The liver-specific Pxr and Yap knockdown was performed as described previously [15]. In brief, male six to eight-week-old C57BL/6 J mice were given an intravenous injection of Adeno-associated virus (AAV) Pxr-shRNA-EGFP and AAV control-EGFP (1.2 × 10¹² genome copies per mouse, Hanbio, China) to knockdown of Pxr, or intravenous injection of AAV Yap-shRNA-ZsGreen and AAV control-ZsGreen (1.3 × 10¹¹ genome copies per mouse, Hanbio, China) to knockdown of Yap, followed by an induced CLP or LPS model after 4 weeks. The sequences used are provided as follows: AAV Pxr-shRNA-EGFP (5′-GGAGGAAGATGGAGGTCTT-3′), AAV control-EGFP (5′-TTCTCCGAACGTGTCACGTAA-3′), AAV Yap-shRNA-ZsGreen (5′-GAAGCGCTGAGTTCCGAAATC-3′), AAV control-ZsGreen (5′-TTCTCCGAACGTGTCACGTAA-3′).

Biochemical analysis

The levels of serum aspartate transaminase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), blood urea nitrogen (BUN), creatinine (CR), creatine kinase isoenzymes (CK-MB) were assessed using biochemical kits (URIT, Guilin, China) following the manufacturer’s protocols and analyzed by URIT-8021A automatic chemistry analyzer (URIT, Guilin, China). The Elabscience ELISA kit (Wuhan, China) was used to measure serum and liver levels of cytokines interleukin 6 (IL-6) and tumor necrosis factor α (TNF-α).

Histopathological analysis

Liver, kidney, lung, and cardiac tissues obtained from septic mice were fixed in 4% paraformaldehyde (Biosharp, Anhui, China) for 48 h. Subsequently, 5 μm thick paraffin-embedded sections were prepared for hematoxylin and eosin (H&E, Biossci, Wuhan, China) and immunohistochemical staining analysis. At least 5 areas were randomly selected and statistics for each sample. Tissue injury severity was graded according to established criteria [24, 26]. The degree of severity of hepatic injury was scored according to parameters on a scale of 0 to 3, which included inflammatory infiltration, hemorrhage and destruction of the hepatic cord structure, and necrosis, with a cumulative maximum score of 12. Lung injury severity was scored on a scale of 0 to 3, assessing alveolar structure, alveolar wall thickness, hemorrhage, and inflammatory infiltrate, also with a maximum score of 12. Renal injury severity was evaluated based on tubular damage and glomerular atrophy on a scale of 0 to 5. Heart injury severity was assessed based on myocyte damage and inflammatory infiltration on a scale of 0 to 6.

Immunohistochemical analysis

Paraffin-embedded tissue sections underwent deparaffinization and rehydration, incubated with 3% H₂O₂ for 10 min, and blocked with goat serum for 30 min. Subsequently, the sections were incubated with F480 (Lot#: 70076, CST, MA, USA) or LY6G antibody (Lot#: ab238132, Abcam, UK) overnight at 4 °C. The next day, the sections were incubated with a streptavidin-conjugated horseradish peroxidase-labeled secondary antibody working solution for 30 min at 37 °C. After three washes with PBS (Biosharp, Anhui, China), the sections were stained with DAB (Lot#: DA1010, Solarbio, Beijing, China) and counterstained with hematoxylin. For each sample, at least five random fields of view were selected, and images were analyzed using ImageJ software (NIH, MD, USA) to quantify the area occupied by positive cells.

Western blot analysis

Total protein, cytoplasmic protein, and nuclear protein of the liver were extracted as previously described [15]. In brief, total liver proteins, and cytoplasmic/nuclear proteins were homogenized with protein RIPA lysate containing 1% PMSF (Lot#: K1020, APExBIO, Texas, USA), NE-PER Nuclear and Cytoplasmic Extraction Reagents (Lot#: 78835, Thermo Fisher Scientific, Massachusetts, USA), respectively. The protein concentration of the liver was determined by BCA kit (Lot#: 23225Thermo Fisher Scientific, Massachusetts, USA). Protein samples were separated by SDS-PAGE and transferred to a Polyvinylidene fluoride (PVDF) membrane (Merck Millipore, Germany), which was blocked with 5% skim milk (Biosharp, Anhui, China) at room temperature for 1 h, and then incubated with primary antibody at 4 °C overnight. PVDF membranes were washed 3 times with TBST [Tris-buffered saline (TBS) containing 0.1% Tween-20 (Biosharp, Anhui, China)] the next day and incubated with HRP-labeled secondary antibody (ABclonal, Wuhan, China) for 1 h at room temperature. The bands were detected by chemiluminescence (Merck Millipore, Germany) following three washes with TBST, and the images were analyzed by ImageJ software.

Quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) analysis

Total RNAs were isolated with TRIzol (Lot#:AG21101, Accurate Biology, Hunan, China), and cDNA was synthesized employing the Evo M-MLV Kit (Lot#:AG11706, Accurate Biology, Hunan, China). Then 20 ng cDNA was used for qRT–PCR through SYBR Green PCR Master Mix (Lot#:AG11701, Accurate Biology, Hunan, China). All primers used in research were listed in Tables 1 and 2.

Table 1.

Primers for qPCR.

Gene	Forward primer (5′-3′)	Reverse primer (5′-3′)
18S	CGATCCGAGGGCCTCACTA	AGTCCCTGCCCTTTGTACACA
Il-6	TAGTCCTTCCTACCCCAATTTCC	TTGGTCCTTAGCCACTCCTTC
Tnf-α	CCCTCACACTCAGATCATCTTCT	GCTACGACGTGGGCTACAG
Il-1β	CTGGTACATCAGCACCTCAC	AGAAACAGTCCAGCCCATAC
Il-10	CCAGGGAGATCCTTTGATGA	CATTCCCAGAGGAATTGCAT
Il-17	TCAGCGTGTCCAAACACTGAG	CGCCAAGGGAGTTAAAGACTT
Tgf-β	TGGAGCAACATGTGGAACTC	GTCAGCAGCCGGTTACCA
pxr	GATGGAGGTCTTCAAATCTGCC	CAGCCGACATTGCGTTTC
Cyp3a11	GGATGAGATCGATGAGGCTCTG	CAGGTATTCCATCTCCATCACAGT
Cyp2b10	AAGGAGAAGTCCAACCAGCA	CTCTGCAACATGGGGGTACT
Ugt1a1	GTCATCCAAAGACTCGGGCA	GACATTCAGGGTCACCCCAG
Yap	ACCCTCGTTTTGCCATGAAC	TGTGCTGGGATTGATATTCCGTA
Cyr61	TAAGGTCTGCGCTAAACAACTC	CAGATCCCTTTCAGAGCGGT
Ankrd1	GGAACAACGGAAAAGCGAGAA	GAAACCTCGGCACATCCACA
Ctgf	GGCCTCTTCTGCGATTTCG	GCAGCTTGACCCTTCTCGG

Table 2.

Primers for gene identification.

	Sequences (5'-3')
Pxr-F	CTGGTCATCACTGTTGCTGTACCA
Pxr-R1	GCAGCATAGGACAAGTTATTCTAGAG
Pxr-R2	CTAAAGCGCATGCTCCAGACTGC

Statistical analysis

Statistical analysis was performed using GraphPad Prism 8 software (GraphPad Software, San Diego, CA, USA). All data were shown as mean ± standard deviation (SD). The survival rates were analyzed by Kaplan-Meier curves. Data were analyzed using the 2-tailed unpaired Student’s t test between two different groups. Comparisons among three or more groups were analyzed using one-way ANOVA analysis. P-values less than 0.05 were considered significant statistically.

Results

Sepsis-induced liver injury severity negatively correlates with the expression of PXR downstream targets

To evaluate the potential function of PXR in sepsis, we sacrificed mice at 0, 6, 12, and 24 h after the establishment of the CLP model to examine liver injury and the expression of targets downstream of PXR at different time points. Notably, mice in the 0 h group underwent the CLP procedure, were immediately euthanized, and tissue samples were collected. Biochemical analysis revealed that serum levels of ALT and AST were highest at 12 h and slightly decreased at 24 h (Fig. 1a, b). Meanwhile, Il-6 mRNA levels peaked at 12 h, whereas Tnf-α levels were highest at 6 h and then decreased gradually (Fig. 1c, d). Furthermore, we found that the mRNA levels of Pxr were consistently downregulated at 0, 6, and 12 h, which was consistent with the decreased mRNA levels of PXR downstream targets cytochrome P450 family 3 subfamily A polypeptide 11 (Cyp3a11), Cyp2b10 and UDP-glucuronyltransferase 1A1 (Ugt1a1) (Fig. 1e–h). Normalized data subjected to Pearson correlation analysis demonstrated significant inverse relationships between liver injury markers (ALT, AST, Il-6) and PXR pathway components (Pxr, Cyp3a11, Ugt1a1), as visualized in the correlation heatmap (Fig. 1i, j). These results suggested that the severity of liver injury at different stages of sepsis negatively correlated with the downregulation of PXR downstream targets.

Fig. 1. Liver injury severity negatively correlates with the expression of PXR downstream target genes.

a, b The levels of serum ALT and AST (n = 5–6). c, d The mRNA levels of cytokines Il-6, and Tnf-α in livers (n = 4–5). e–h The mRNA levels of Pxr and downstream target genes Cyp3a11, Cyp2b10, and Ugt1a1 in livers (n = 4–5). i The variations in ALT, AST, Il-6, Tnf-α, Pxr, Cyp3a11, Cyp2b10, and Ugt1a1 levels at various times post-Z-Score normalization. j The heatmap of the correlation matrix displays the Pearson correlation coefficient values for all variables under study, with negative values represented in blue and positive values in red. The coefficient ranges from −1 to 1, where a value of −1 indicates a perfect negative linear relationship between variables. Data are presented as mean ± standard deviation (SD). The statistical significance is represented by: *P < 0.05, **P < 0.01, ***P < 0.001.

PXR activation ameliorates CLP-induced sepsis and liver injury

To further evaluate the protective role of PXR in sepsis-induced liver injury, we initially examined the effect of PXR activation on survival rates in septic mice (Fig. 2a). The findings revealed that the murine-specific PXR agonist PCN effectively triggered PXR activation, as evidenced by the upregulation of the downstream enzymes CYP3A11 and CYP2B10 (Fig. 2b, c, and Supplementary Fig. S5a). Specifically, CYP3A11 and CYP2B10 were downregulated in the CLP group and significantly upregulated after PXR activation. In addition, we also examined the protein expression of PXR, and there was no significant change in PXR expression between the Sham and PCN groups, while PXR expression was significantly down-regulated in the CLP group and was not significantly changed after PXR activation (Fig. 2b, and Supplementary Fig. S5b). Then, we found that PXR activation significantly increased the 72 h survival rate by 33% in CLP mice (Fig. 2d). Next, we sacrificed mice 12 h after CLP-induced sepsis and investigated the protective effect of PXR activation against multiple organ damage in sepsis (Fig. 2a). We found that serum ALT, AST, ALP, CR, BUN, and CK-MB were significantly increased in the CLP group (Fig. 2e, f, Supplementary Figs. S1a, and S2a–c). Meanwhile, ALT, AST, and ALP were significantly decreased after activation of PXR, while CR, BUN, and CK-MB tended to decrease without significant difference after activation of PXR (Fig. 2e, f, Supplementary Figs. S1a and S2a–c). Histological examination also showed that PXR activation diminished hepatic pathological changes and inflammatory cell infiltration in CLP mice, whereas there was no significant improvement in kidney, lung, and heart injury (Fig. 2g, i, and Supplementary Fig. S2d–g). Therefore, we next investigated the improvement of liver and systemic inflammation after PXR activation. Immunohistochemical staining showed that PXR activation reduced the hepatic infiltration of F480⁺ and LY6G⁺ cells in CLP mice (Fig. 2h, j). ELISA results indicated that serum and liver protein levels of IL-6 and TNF-α were upregulated in the CLP mice, while PXR activation significantly downregulated their protein levels (Fig. 2k–n), which was consistent with the decreased mRNA levels of liver Il-6, Tnf-α, Il-1β, Il-10, Il-17, and Tgf-β in CLP mice upon PXR activation (Fig. 2o–q, and Supplementary Fig. S1b–d). In addition, we evaluated the protective effect of PXR activation against septic liver injury during 24–72 h, which is considered to be the immunosuppressive phase of sepsis. We sacrificed and harvested samples 36 h after establishing the CLP-induced sepsis model. The results showed that activation of PXR significantly down-regulated serum ALT and AST. However, the levels of serum BUN, CR, and CK-MB tended to decrease, but there was no statistically significant difference (Supplementary Fig. S3). These results indicated that PXR activation effectively ameliorates liver injury but not kidney, lung, and heart injury in CLP-induced sepsis mice.

Fig. 2. PXR activation ameliorates CLP-induced liver injury.

a C57BL/6 mice were treated with PCN (100 mg·kg⁻¹·d⁻¹) for 3 days before the CLP treatment. b, c The protein expression of liver CYP3A11, CYP2B10, PXR (n = 3). d The survival rates of CLP mice (n = 6, 12). e, f The levels of serum ALT and AST (n = 5–6). g H&E staining of livers. h Immunohistochemical staining of liver F480, LY6G. i Liver injury score (n = 3). j The percentage of F480 and LY6G positive cells in livers (n = 3). k–n The protein levels of IL-6, TNF-α in serum and liver (n = 5–6). o–q The mRNA levels of cytokines Il-6, Tnf-α, Il-1β in livers (n = 5–6). Data are presented as mean ± SD. The survival rates of septic mice were analyzed using the Kaplan-Meier method with the log-rank test. The statistical significance is represented by: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

PXR activation ameliorates LPS-induced sepsis and liver injury

We also confirmed the curative function of PXR in sepsis-induced liver injury using an LPS-induced sepsis model (Fig. 3a). We similarly found that the PXR downstream target protein CYP3A11 and CYP2B10 was downregulated in the LPS group, while significantly upregulated after PXR activation (Fig. 3b, c, and Supplementary Fig. S5c). There was also no significant change in PXR expression between the Vehicle and PCN groups, while PXR expression was significantly down-regulated in the LPS group and was not significantly changed after PXR activation (Fig. 3b, and Supplementary Fig. S5d). Then, we found that PXR activation significantly increased the 72 h survival rate by 20% in LPS mice (Fig. 3d). ALT and AST were significantly decreased after PXR activation in LPS mice, while CR, BUN, and CK-MB tended to decrease without significant differences (Fig. 3e, f, and Supplementary Fig. S4a–c). Histological examination confirmed that PXR activation diminished hepatic pathological changes and inflammatory cell infiltration in LPS mice, whereas there was no significant improvement in kidney, lung, and heart injury (Fig. 3g, h, and Supplementary Fig. S4d–g). Real-time PCR results showed that PXR activation downregulated the mRNA levels of liver inflammatory factors Il-6, Tnf-α, Il-1β, Il-10, Il-17, and Tgf-β in LPS mice (Fig. 3i–n). Immunohistochemical staining also showed that PXR activation reduced hepatic infiltration of F480⁺ and LY6G⁺ cells in LPS mice, and this was in agreement with CLP mice (Fig. 3o, p). The above results indicated that PXR activation effectively ameliorates liver injury but not kidney, lung, and heart injury in LPS-induced septic mice.

Fig. 3. PXR activation ameliorates LPS-induced liver injury.

a C57BL/6 mice were treated with PCN (100 mg·kg⁻¹·d⁻¹) for 3 days before the LPS treatment. b, c The protein expression of liver CYP3A11, CYP2B10, PXR (n = 3). d The survival rates of LPS mice (n = 6, 12). e, f The levels of serum ALT and AST (n = 5–6). g H&E staining of livers. h Liver injury score (n = 3). i–n The mRNA levels of cytokines Il-6, Tnf-α, Il-1β, Il-10, Il-17, and Tgf-β in livers (n = 5–6). o Immunohistochemical staining of liver F480, LY6G. p The percentage of F480 and LY6G positive cells in livers (n = 3). Data are presented as mean ± SD. The survival rates of septic mice were analyzed using the Kaplan-Meier method with the log-rank test. The statistical significance is represented by: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

The loss of Pxr aggravates sepsis-induced liver injury and abolishes the protective effect of PCN

The above results demonstrated that activation of PXR effectively ameliorated sepsis-induced liver injury, and we next used Pxr knockout mice to further validate the protective effect of PXR and its agonist PCN in septic mice (Fig. 4a, and Supplementary Fig. S7a). The mRNA levels of Pxr in Pxr^+/+ and Pxr^-/- mice demonstrated that Pxr was successfully knockout (Supplementary Fig. S6a–c). The results indicated that the survival time was decreased after the loss of Pxr in CLP mice (Fig. 4b). We similarly found that serum ALT and AST were higher in Pxr^-/- mice compared to Pxr^+/+ mice in the sepsis model and were not obviously improved by PXR activation (Fig. 4c, d, and Supplementary Fig. S7b, c). H&E staining indicated that liver injury was more severe after the loss of Pxr in the sepsis model, while PXR activation showed no pathological improvement (Fig. 4e, g and Supplementary Fig. S7d, f). Meanwhile, immunohistochemical results showed that the liver of Pxr^-/- mice had more F480⁺, LY6G⁺ cell infiltration than Pxr^+/+ mice in the sepsis model, whereas there was no significant change after PXR activation (Fig. 4f, h and Supplementary Fig. S7e, g). We also evaluated inflammatory factor levels in serum and liver tissues. Our results suggested that protein levels of IL-6 and TNF-α in serum and liver were more highly expressed in Pxr^-/- mice than Pxr^+/+ mice in the sepsis model (Fig. 4i, j), which was consistent with the upregulated mRNA levels of liver Il-6, Tnf-α, Il-1β in Pxr^-/- mice (Fig. 4k, and Supplementary Fig. S7h). Moreover, PXR activation failed to reduce inflammatory factor levels in the serum and liver of Pxr^-/- mice (Fig. 4i, j).

Fig. 4. Pxr knockdown aggravates the CLP-induced liver injury.

a Pxr^+/+ and Pxr^-/- mice were treated with PCN (100 mg·kg⁻¹·d⁻¹) for 3 days before the CLP treatment. b The survival rates of Pxr^+/+ and Pxr^-/- mice. c, d The levels of serum ALT and AST (n = 5–6). e H&E staining of livers. f Immunohistochemical staining of liver F480, LY6G. g Liver injury score (n = 3). h The percentage of F480 and LY6G positive cells in livers (n = 3). i, j The protein levels of IL-6, and TNF-α in serum and livers (n = 5–6). k The mRNA levels of cytokines Il-6, Tnf-α, Il-1β in livers (n = 5–6). Data are presented as mean ± SD. The survival rates of septic mice were analyzed using the Kaplan-Meier method with the log-rank test. The statistical significance is represented by: *P < 0.05, **P < 0.01.

To further confirm whether PCN ameliorated sepsis-induced liver injury is dependent on the hepatic PXR, we generated liver-specific Pxr knockdown mice via AAV8-mediated shRNA delivery (Fig. 5a). EGFP fluorescence demonstrated high infective efficiency of AAV vectors in mice livers (Fig. 5b). Moreover, the mRNA levels of Pxr were significantly decreased in AAV Pxr-shRNA mice than in AAV control mice (Fig. 5c), which indicated a successful hepatic Pxr-specific knockdown. Subsequently, we found that serum ALT and AST were higher in AAV Pxr-shRNA mice compared to AAV control mice during sepsis (Fig. 5d, e). Moreover, ALT and AST were decreased in AAV control mice with sepsis after PXR activation, whereas there was no significant improvement in AAV Pxr-shRNA mice (Fig. 5d, e). Histological analysis indicated that sepsis-induced liver injury, F480⁺ and LY6G⁺ inflammatory cell infiltration were more severe in AAV Pxr-shRNA mice than AAV control mice (Fig. 5f–j). Meanwhile, PXR activation in AAV Pxr-shRNA mice did not ameliorate sepsis-induced liver injury and inflammatory cell infiltration (Fig. 5f–j). We also evaluated inflammatory factor levels in liver tissues between AAV control and AAV Pxr-shRNA mice. The results showed that the mRNA levels of liver Il-6, Tnf-α, and Il-1β were higher in AAV Pxr-shRNA mice compared to AAV control mice during sepsis, and PXR activation in AAV Pxr-shRNA mice did not decrease the mRNA levels of liver Il-6, Tnf-α, Il-1β (Fig. 5k–m). These results indicated that the loss of hepatic Pxr aggravates sepsis-induced liver injury, and PCN ameliorated sepsis-induced liver injury in a PXR-dependent manner.

Fig. 5. Liver-specific knockdown of Pxr attenuates the effect of PXR activation on CLP-induced liver injury.

a AAV control and AAV Pxr-shRNA mice were treated with PCN (100 mg·kg⁻¹·d⁻¹) for 3 days before the CLP treatment. b Fluorescence images of EGFP in AAV control and AAV Pxr-shRNA mice. c The mRNA level of liver Pxr (n = 3). d, e The levels of serum ALT and AST (n = 5). f H&E staining of livers. g Immunohistochemical staining of liver F480, LY6G. h Liver injury score (n = 3). i, j The percentage of F480 and LY6G positive cells in livers (n = 3). k–m The mRNA levels of cytokines Il-6, Tnf-α, Il-1β in livers (n = 5). Data are presented as mean ± SD. The statistical significance is represented by: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

PXR regulates and activates the yes-associated protein signaling pathway in sepsis liver

Our previous results found that PXR directly bind to YAP, activates YAP signaling, and promotes the expression of YAP and its downstream targets [21]. Moreover, Wang et al. showed that liver injury was more severe after liver-specific knockdown of Yap in sepsis [19]. Therefore, we next investigated whether the YAP signaling pathway was involved in the PXR-triggered protection against sepsis-induced liver injury. The results indicated that the mRNA levels of Yap and its downstream targets ankyrin repeat domain 1 (Ankrd1), cysteine-rich angiogenic inducer 61 (Cyr61), and connective tissue growth factor (Ctgf) were downregulated in CLP mice and significantly upregulated after PXR activation (Fig. 6a–d). The protein expression of total-YAP, nuclear-YAP, and its downstream targets CYR61, ANKRD1, and CTGF was decreased in the liver of CLP and LPS mice and obviously increased after PXR activation, while the p-YAP level was reduced after PXR activation (Fig. 6e–j). Lastly, we detected the effect of Pxr knockout on YAP protein expression. The expression of hepatic YAP proteins did not show significant change between Pxr^+/+ and Pxr^-/- mice. (Supplementary Fig. S8a, b). Interestingly, Pxr knockdown significantly downregulated liver YAP protein expression in CLP mice, with no obvious change after PXR activation (Supplementary Fig. S8c, d). These results indicated a regulatory role for PXR on the YAP signaling pathway in sepsis-induced liver injury.

Fig. 6. PXR activation promotes liver Yes-associated protein and its downstream target protein expression in sepsis mice.

a–d The mRNA levels of Yap and downstream target genes Cyr61, Ankrd1, and Ctgf in the liver of the CLP model (n = 5–6). e Liver protein levels of total-YAP, nuclear-YAP, p-YAP, and downstream target genes CYR61, ANKRD1, CTGF in CLP model. f, g The statistics of liver total-YAP, nuclear-YAP, p-YAP, CYR61, ANKRD1 and CTGF protein levels in CLP model. h Liver protein levels of total-YAP, nuclear-YAP, p-YAP, and downstream target genes CYR61, ANKRD1, CTGF in LPS model. i, j The statistics of liver total-YAP, nuclear-YAP, p-YAP, CYR61, ANKRD1, and CTGF protein levels in the LPS model. Data are presented as mean ± SD. The statistical significance is represented by: *P < 0.05, **P < 0.01, ***P < 0.001.

Liver-specific knockdown of Yap attenuates the effect of PXR activation on sepsis-induced liver injury

To confirm whether PXR ameliorates sepsis-induced liver injury by activating the YAP signaling pathway, AAV Yap-shRNA mice were employed in both CLP and LPS-induced sepsis models (Fig. 7a, and Supplementary Fig. S9a). ZsGreen fluorescence demonstrated high hepatic transduction efficiency of AAV vectors (Fig. 7b, and Supplementary Fig. S9b). Hepatic protein expression of YAP was obviously down-regulated in AAV Yap-shRNA mice (Fig. 7c, d, and Supplementary Fig. S9c, d). The mRNA levels of Yap and their downstream targets (Cyr61, Ankrd1, and Ctgf) were also significantly decreased in AAV Yap-shRNA mice (Supplementary Fig. S10a, c). Furthermore, the mRNA levels of Yap in the ileum, kidney, spleen, lung, and heart tissues were not significantly changed between AAV control and AAV Yap-shRNA mice (Supplementary Fig. S10b, d). These results indicated a successful hepatic Yap-specific knockdown. We also verified liver PXR was successfully activated in AAV control and AAV Yap-shRNA mice, as revealed by CYP3A11 protein expression (Supplementary Fig. S11a–d). Subsequently, we found that ALT and AST were decreased in AAV control mice with sepsis after PXR activation, whereas there was no significant improvement in AAV Yap-shRNA mice (Fig. 7e, f, and Supplementary Fig. S9e, f). Histological analysis indicated that following PXR activation in the sepsis mice, liver injury and inflammatory cell infiltration were more pronounced in AAV Yap-shRNA mice than in AAV control mice (Fig. 7g–j, and Supplementary Fig. S9g–k). We also evaluated inflammatory factor levels in serum and liver tissues between AAV control and AAV Yap-shRNA mice. The results showed that, following PXR activation in the sepsis model, protein levels of serum and liver IL-6, TNF-α were more higher in AAV Yap-shRNA mice than in AAV control mice (Fig. 7k, l), which was consistent with the higher mRNA levels of liver Il-6, Tnf-α, Il-1β in AAV Yap-shRNA mice (Fig. 7m, and Supplementary Fig. S9l–n). Previous studies demonstrated that NF-κB activation contributes to the hyperinflammatory response during sepsis, and both PXR and YAP are associated with the regulation of NF-κB activation [18, 27]. Therefore, we next investigated whether PXR affected NF-κB activation through the YAP signaling pathway. The results showed that PXR activation significantly decreased the level of phosphorylated p65 in septic mice (Fig. 7n, and Supplementary Fig. S11e), whereas the level of phosphorylation of p65 was obviously increased when hepatic Yap was specifically knockdown (Fig. 7o, and Supplementary Fig. S11f). This result suggested that PXR probably inhibited NF-κB activation through the YAP signaling pathway and further down-regulated hepatic inflammatory factor expression. Overall, these results indicated that PXR ameliorates sepsis-induced liver injury by activating the YAP signaling pathway and inhibiting NF-κB activation (Fig. 8).

Fig. 7. Liver-specific knockdown of Yap attenuates the effect of PXR activation on CLP-induced liver injury.

a AAV control and AAV Yap-shRNA mice were treated with PCN (100 mg·kg⁻¹·d⁻¹) for 3 days before the CLP treatment. b Fluorescence images of ZsGreen in AAV control and AAV Yap-shRNA mice. c, d The protein level of liver YAP (n = 3). e, f The levels of serum ALT and AST (n = 5–6). g H&E staining of livers. h Immunohistochemical staining of liver F480, LY6G. i Liver injury score (n = 3). j The percentage of F480 and LY6G positive cells in livers (n = 3). k, l The protein levels of IL-6, and TNF-α in serum and livers (n = 5–6). m The mRNA levels of cytokines Il-6, Tnf-α, Il-1β in livers (n = 5–6). n The protein expression of liver p-p65, p65 in CLP mice. o The protein expression of liver p-p65, p65 in AAV control and AAV Yap-shRNA mice. Data are presented as mean ± SD. The statistical significance is represented by: *P < 0.05, **P < 0.01.

Fig. 8. Pregnane X receptor alleviates sepsis-induced liver injury through activation of yes-associated protein in mice.

Upon PXR activation by the agonist PCN, PXR binds and promotes the expression of total YAP and nuclear YAP, while simultaneously inhibiting YAP phosphorylation. This interaction enhances the transcriptional activity of YAP by facilitating its nuclear translocation and upregulating its downstream targets, including CYR61, ANKRD1, and CTGF. Consequently, PXR activation suppresses hepatic inflammation, alleviates sepsis-induced liver injury, and improves survival outcomes in septic mice.

Discussion

Sepsis-induced liver failure represents a critical complication that exacerbates disease severity and impairs clinical outcomes [4]. While the therapeutic potential of PXR activation in mitigating this condition has been underexplored. Our current study indicated that PXR activation alleviated sepsis-induced liver injury and enhanced the survival of septic mice. Conversely, Pxr knockdown accelerated the death of septic mice and exacerbated liver injury. Furthermore, mechanistic analysis revealed that PXR exerts protective effects by activating the YAP signaling pathway while suppressing NF-κB activation. These findings collectively establish PXR as a key regulator of sepsis-induced liver injury, functioning through coordinated tissue repair promotion and inflammatory cascade inhibition.

Our current study found that the severity of sepsis-induced liver injury was dynamically changed at different time points, as revealed by serum ALT, AST, and liver inflammatory factor Il-6, Tnf-α expression at different time points. Interestingly, the levels of Pxr and its downstream targets at different time points of sepsis were negatively correlated with the severity of liver injury, indicating that PXR might play a vital role in sepsis-induced liver injury and its activation may serve as protection against this process. Moreover, several clinically approved drugs—including rifampicin and mifepristone—are known PXR agonists [28]. The lipid-lowering drug atorvastatin has also been identified as an agonist of PXR [29]. Clinical evidence indicates that patients who were on atorvastatin therapy before being hospitalized for sepsis and continued its use during hospitalization experienced significantly lower mortality rates compared to those who had never used statins [30]. Furthermore, in a CLP model, the administration of atorvastatin 6 h post-induction significantly increased the survival rate of mice [31]. These collective findings posit PXR activation as a plausible mechanism underlying atorvastatin's therapeutic benefits in sepsis management.

The tissue distribution of PXR parallels its function. PXR exhibits considerable expression in the liver and intestinal tissues, with minimal expression in other organ tissues [7]. Consistently, we found that PXR activation ameliorated sepsis-induced liver injury, but not kidney, lung, and heart injury. Specifically, serum ALT, AST, and ALP were significantly downregulated after activation of PXR, while CR, BUN, and CK-MB tended to decrease without significance. Histological examination also showed that PXR activation diminished pathological changes and inflammatory cell infiltration in the liver of septic mice, whereas there was no significant improvement in kidney, lung, and heart injury. These findings suggest that the remarkable improvement in septic liver injury following PXR activation may be due to the organ-specific expression of PXR in liver tissue. Notably, while PXR activation improved survival and systemic inflammation in septic mice, its clinical utility may be constrained in patients with concurrent severe multiorgan dysfunction. Others also reported that indole-3-propionic acid attenuates sepsis-induced liver injury and upregulates the mRNA level of Pxr [14]. Hence, we further focused on the effect of PXR activation in sepsis-induced liver injury and verified it in Pxr knockout and hepatic knockdown mice. The results showed that PXR activation obviously reduced the level of liver inflammation and attenuated liver injury in septic mice. In addition, Pxr knockout or hepatic-specific knockdown accelerated the death of septic mice and exacerbated liver injury, whereas PXR activation did not significantly ameliorate sepsis-induced liver injury in Pxr^-/- and AAV Pxr-shRNA mice. In short, we innovatively demonstrate that PXR activation attenuates sepsis-induced liver injury, the loss of Pxr aggravates sepsis-induced liver injury and PCN alleviates sepsis-induced liver injury in a PXR-dependent manner.

YAP is a transcriptional coactivator downstream of the Hippo signaling pathway that promotes cell proliferation and inhibits apoptosis, thereby regulating tissue growth and organ size [17]. Recent studies reported that YAP has a protective role in sepsis and inflammatory disease. For instance, Wang et al. indicated that YAP protects against septic liver injury via ferroptosis resistance [19]. The current study found that the expressions of total-YAP, nuclear-YAP, and downstream targets CYR61, ANKRD1, and CTGF were downregulated in the liver of septic mice and obviously upregulated after PXR activation, whereas the p-YAP level was downregulated after PXR activation. We further found that the expression of hepatic YAP proteins did not show a significant change between Pxr^+/+ and Pxr^-/- mice. Interestingly, Pxr knockout significantly downregulated liver YAP protein expression in CLP mice, and there was no obvious change after PCN treatment. These results hint that YAP plays a vital role in PCN ameliorating sepsis-induced liver injury. Moreover, we demonstrated that PXR activation ameliorates sepsis-induced liver injury and is dependent on the YAP signaling pathway by specifically knocking down Yap in hepatocytes. We further found that PXR inhibited NF-κB activation through interaction with YAP in the liver, suggesting that PXR may inhibit IL-6 and TNF-α levels via the NF-κB pathway in sepsis.

Previous studies have demonstrated that other members of the nuclear receptor family, including PPAR-α and FXR, also played a protective role in sepsis-induced liver injury [32, 33], underscoring the broader therapeutic potential of nuclear receptors in sepsis management. However, it is worth noting that PXR is involved in the disposition of drugs by regulating the drug-metabolizing enzymes and transporters such as CYP3A4, CYP2B6, UGT1A1, OATP2, and MRP3. Therefore, the patient’s clinical concurrent medications should be carefully considered when using PXR agonists to protect against liver injury in sepsis patients [34]. Consequently, clinicians must carefully evaluate potential drug-drug interactions when considering PXR agonists for liver protection in septic patients, particularly those receiving concurrent medications.

Our current study revealed a novel function of PXR in ameliorating liver injury and enhancing the survival of septic mice. Meanwhile, we found that PXR ameliorates sepsis-induced liver injury by activating the YAP signaling pathway and inhibiting NF-κB activation, which might provide a novel strategy for the treatment of sepsis-induced liver injury.

Author contributions

HCB, JHF, and XY conceived and designed the project. CHW, SH, DL, XWJ, HOY, GFB, PW, FTL, and WHZ performed the experiments. HCB, CHW, and JHF wrote and revised the manuscript.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41401-025-01552-4.