Hepatocyte-Specific Deletion of Yes-Associated Protein Improves Recovery From Acetaminophen-Induced Acute Liver Injury

Abstract

Overdose of acetaminophen (APAP) is the major cause of acute liver failure (ALF) in the Western world with very limited treatment options. Previous studies from our groups and others have shown that timely activation of liver regeneration is a critical determinant of transplant-free survival of APAP-induced ALF patients. Here, we report that hepatocyte-specific deletion of Yes-associated protein (Yap), the downstream mediator of the Hippo Kinase signaling pathway results in faster recovery from APAP-induced acute liver injury. Initial studies performed with male C57BL/6J mice showed a rapid activation of Yap and its target genes within first 24 h after APAP administration. Treatment of hepatocyte-specific Yap knockout (Yap-KO) mice with 300 mg/kg APAP resulted in equal initial liver injury but a significantly accelerated recovery in Yap-KO mice. The recovery was accompanied by significantly rapid hepatocyte proliferation supported by faster activation of Wnt/β-catenin pathway. Furthermore, Yap-KO mice had significantly earlier and higher pro-regenerative inflammatory response following APAP overdose. Global gene expression analysis indicated that Yap-KO mice had a robust activation of transcription factors involved in response to endoplasmic reticulum stress (XBP1) and maintaining hepatocyte differentiation (HNF4α). In conclusion, these data indicate that inhibition of Yap in hepatocytes results in rapid recovery from APAP overdose due to an earlier activation of liver regeneration.

Acetaminophen (APAP) is one of the most widely used antipyretic and analgesic agents in the world and is very effective when take as directed (Lee, 1995, 2004; Nourjah et al., 2006). However, overdose of APAP remains the leading cause of acute liver failure (ALF) in the Western world (Jaeschke et al., 2014; Lee, 2008, 2013). APAP-induced ALF is a triphasic process involving initiation of injury, progression of injury, and recovery and regeneration (Bhushan and Apte, 2019; Iorga et al., 2017; Jaeschke et al., 2014). Extensive studies, including from our laboratory, have demonstrated that stimulation of timely and robust liver regeneration is the critical factor in patient survival (Bhushan and Apte, 2019; Schmidt and Dalhoff, 2002, 2005). Extensive studies in the last decade have shown central role of the Hippo Kinase signaling pathway in regulation of liver size, hepatocyte proliferation, and liver cancer pathogenesis (Sylvester and Colnot, 2014; Yimlamai et al., 2015; Zhao et al., 2008). The downstream effector of Hippo Kinase pathway is a transcriptional coactivator called Yes-associated protein (Yap), which is regulated by a series of kinases including MST1/2 and Lats-1. When the Hippo Kinase pathway is on, Yap is phosphorylated at Ser127 by MST1/2-Lats signaling axis, undergoes cytoplasmic translocation and is degraded via 14-3-3-aided degradation. When the pathway is turned off, Yap phosphorylation is inhibited, Yap is active and undergoes nuclear accumulation, where it binds to transcription factor (TF) TEAD and initiates gene expression program. Yap-TEAD heterodimer regulates a diverse group of genes including cell proliferation genes, metabolic genes, and cell structure genes (Zhao et al., 2008).

Previous studies have demonstrated that Yap can induce hepatocyte proliferation and hepatocyte fate change (Camargo et al., 2007; Miyamura et al., 2017; Yimlamai et al., 2014). Role of Yap in pathogenesis of hepatocellular carcinoma is well-documented (Li et al., 2012; Sylvester and Colnot, 2014; Yimlamai et al., 2015). Ectopic expression of S127 mutated Yap results in the development of hepatoblastoma in mice (Min et al., 2019). Role of Yap in liver regeneration after partial hepatectomy has also been demonstrated (Oh et al., 2018). These data indicate that Yap promotes cell proliferation and liver regeneration. However, the exact role of Yap in regulation of liver regeneration after drug-induced liver injury, which is mechanistically different has not been investigated. Specifically, role of Yap in APAP-induced liver injury and recovery from APAP overdose has not been investigated. Here, we report interesting paradoxical role of Yap in regulation of recovery and regeneration following APAP-induced liver injury.

METHODS

Animals, Treatments, and Tissue Collection

Two- to 3 months old male C57Bl/6J mice, purchased from Jackson Laboratories (Bar Harbor, Maine), and YAP floxed/floxed (YAP^fl/fl) mice were generated at the Knockout Mouse Project (KOMP) repository. All animals were housed in Association for Assessment and Accreditation of Laboratory Animal Care-accredited facilities at the University of Kansas Medical Center (KUMC) under a standard 12 h light-dark cycle with access to chow and water ad libitum. All studies were approved by the Institutional Animal Care and Use Committee at KUMC. Hepatocyte-specific YAP knockout (Yap-KO) mice were generated by treating 2–3 months old male YAP^fl/fl mice with AAV8.TBG.PI.Cre.rBG virus (2.5 × 10^11 viral particles). YAP^fl/fl mice treated with AAV8.TBG.PI.eGFP.WPRE.bGH (Control) were used as controls. Complete Yap deletion within 1 week after cre administration was confirmed by Western blot analysis (Supplementary Figure 1A). Yap deletion did not affect liver to body weight ratio (Supplementary Figure 1B) and did not induce any discernable histological change (data not shown). Mice were fasted 12 h before APAP administration. APAP was dissolved in warm 0.9% saline, and mice were treated with either 300 mg/kg or 600 mg/kg APAP intraperitoneally. Food was given to the mice after 1 h of APAP treatment. Mice were sacrificed at 0, 3, 6, 12, 24, and 48 h after APAP administration to collect liver and blood samples. Serum samples were obtained by centrifuging blood at 5000 rpm for 10 min at 4°C and used for analysis of alanine aminotransferase (ALT) activity using commercially available kits (ThermoFisher Scientific, Pittsburg, Pennsylvania). The liver tissue was frozen in liquid N2 and stored at −80°C until it was used to isolate RNA and to prepare RIPA extracts.

Protein Isolation and Western Blot Analysis

Protein estimation and Western blot analysis were performed using RIPA extract prepared from frozen liver tissues as previously described (Bhushan et al., 2014).

Immunohistochemistry

Paraffin-embedded liver sections (4 μm thick) were used for immunohistochemical staining. Slides were stained with hematoxylin-eosin using an autostainer (model CV 5030, Leica Microsystems, Buffalo Grove, Illinois) as described previously (Bhushan et al., 2014). Paraffin sections were also stained with proliferating cell nuclear antigen (PCNA) to detect cell proliferation, as previously described (Bhushan et al., 2014) and YAP antibody (1:25 dilution).

Gene Array and Real-Time PCR

RNA was isolated from frozen liver tissues, according to the manufacturer’s protocol (Sigma, St. Louis, Missouri), and converted to cDNA, as previously described (Bhushan et al., 2014). Global gene expression analysis was conducted using Mouse Transcriptome 1.0 GeneChip array (Affymetrix) which has been rename Mouse Clariom D GeneChip array. The resulting global gene expression data were deposited in GEO database (GSE183119). Expression of specific genes was determined using SYBR Green technology on a real-time PCR system (model 7300, Applied Biosystems, Foster City, California).

APAP Protein Adducts and Glutathione Measurement

APAP protein adducts were measured as described before using HPLC-based method. Total glutathione (GSH) level was quantified as described before (Borude et al., 2018; McGill et al., 2012; McGreal et al., 2018).

Statistical Analysis

All results were expressed as mean ± SE. Comparison between 2 groups were performed with Student’s t-test. p < .05 was considered significant.

RESULTS

Yap Activation Occurs During Initiation and Progression Phase of APAP-Induced ALI

We investigated Yap activation after a single acute overdose of APAP (300 mg/kg, ip) given to male C57Bl/6J mice using Western blotting, immunohistochemistry, and target gene expression. APAP treatment resulted in significant liver injury as measured by serum ALT levels, which peaked at 12 h after APAP administration and subsided by 48 h time point (Figure  1A). Cell proliferation measured by PCNA immunohistochemistry as an indicator of liver regeneration occurred from 12 to 72 h after APAP administration with a peak at 48 h (Figure  1B) as previously demonstrated. We performed Western blotting of total Yap and the inactive phosphorylated-Yap (Figure  1C) and then determined the ratio of Yap/phospho-Yap to estimate Yap activity (Figure  1D). Yap activity, i.e. the portion of Yap that was unphosphorylated, started increasing at 3 h time point, peaked at 6 h and decline by 24 h after APAP treatment. These data were further confirmed by expression of Yap target genes including AmotL2 (Figure  1E) and CTGF (Figure  1F), both of which increased significantly within 6 h and decline by 24–48 h after APAP treatment. Immunohistochemical staining of Yap showed that in normal mice Yap staining was limited to biliary epithelium (Figure  1G). However, with 3–6 h after APAP treatment, significant nuclear translocation of Yap was observed mainly in centrilobular and midzonal hepatocytes, a trend that continued till 24 h after APAP treatment. A significant decrease in nuclear Yap and a significant increase in cytoplasmic Yap staining in hepatocytes surrounding the necrotic areas were observed at 48 and 74 h after APAP treatment. By 96 h after APAP administration, Yap was localized only in the biliary epithelial cells.

Yap-KO Mice Exhibited Faster Recovery From APAP-Induced ALI as Compared to Controls

Next, we studied APAP-induced liver injury in control and hepatocyte-specific Yap-KO mice following a dose of 300 mg/kg APAP dose. Liver injury was measured using histopathological observation of H&E-stained liver sections (Figure  2A), serum ALT activity (Figure  2B), and necrosis scoring of H&E stained slides (Figure  2C). All injury parameters revealed an equal increase in liver injury in control and Yap-KO mice at 6 h and 12 h after APAP administration. However, liver injury rapidly declined in Yap-KO mice at 24 and 48 h after APAP treatment. By 48 h, most of the mice showed a complete recovery from APAP-induced liver necrosis and ALT levels returned to normal.

Hepatocyte-Specific Deletion of Yap Does Not Change Mechanisms of Initiation of APAP-Induced ALI

We investigated major mechanisms involved in initiation of APAP-induced liver injury after APAP overdose in Control and Yap-KO mice. Hepatocyte-specific Yap deletion neither affect hepatic CYP2E1 protein levels (Figure  3A) nor basal hepatic GSH levels (Figure  3B), 2 of the most important factors in initiation of APAP-induced liver injury.

Quantification of hepatic GSH content after APAP overdose showed equal GSH depletion in control and Yap-KO mice at 1 h after APAP administration (Figure  3C). Furthermore, an equal amount of APAP-protein adducts were observed in control and Yap-KO mice at 1 and 6 h post-APAP treatment (Figure  3D). Finally, Western blot analysis of total and phosphorylated (active) JNK, a crucial node in molecular signaling that drives APAP-induced liver injury, showed an equal activation of JNK in both control and Yap-KO mice at 1 and 12 h after APAP treatment (Figure  3E). JNK activation was significantly decreased in Yap-KO mice at 24 h as compared to control mice, consistent with decreased injury.

Faster Compensatory Proliferation in Yap-KO Mice After APAP Overdose

To determine whether the rapid decline in liver injury in Yap-KO mice correlates with changes in liver regeneration, we investigated cell proliferation and molecules that drive cell proliferation in the liver. Immunohistochemical staining (Figure  4A), quantification of PCNA positive cells (Figure  4B), and Western blot (Figure  4C) analysis of PCNA, the marker of cell proliferation, clearly demonstrated an earlier increase in cell proliferation in Yap-KO mice. A moderate increase in PCNA expression was observed in Yap-KO at 12 h after APAP treatment, which increased at 24 h and was sustained at 48 h. Control mice showed a robust PCNA activation as 48 h. The Yap-KO mice had an overall early and sustained increase in compensatory cell proliferation as compared to wild-type (WT) mice. Next, we determined expression of core cell cycle proteins including Cyclin D1, the G1/S cyclin, phosphorylated Rb (pRb), and CDK4, all of which showed a significantly earlier induction at 24 h after APAP administration (Figure  4B, lower panel). Control mice showed induction in cell cycle proteins only at 48 h after APAP administration. We further dissected the cell cycle progression at 24 and 48 h by conducting Western blot analysis on individual samples at 24 h (Figure  4D) and 48 h (Figure  4E) time points, which corroborated the data obtained using pooled samples. Control mice showed no cell cycle activation at 24 h but significant induction in expression of Cyclin D1, pRB, CDK4, and PCNA was observed in the Yap-KO mice. By 48 h, expression of these proteins was significantly higher in Control mice as compared to Yap-KO, which still showed some expression suggesting a sustained cell cycle progression.

Higher and Faster Inflammatory Response in Yap-KO Mice After APAP-Induced ALI

To determine the molecular mechanisms of faster liver regeneration after APAP-induced acute liver injury (ALI) in Yap-KO mice, we performed global gene expression analysis using Clariom D Mouse Transcriptomics Array. At 24 h, after APAP treatment 249 genes were upregulated and 317 were gene downregulated 2-fold higher in the Yap-KO mice as compared to Control mice. The gene array data revealed a significant induction in serum amyloid A (SAA) mRNA in Yap-KO livers both before (0 h) and after (24 h) APAP treatment. The significantly higher mRNA levels of SAA1 and SAA2 after APAP treatment were confirmed using qPCR analysis (Figure  5A). We quantified SAA serum levels using an ELISA-based assay, which showed significantly higher levels in Yap-KO mice at 24 and 48 h after APAP treatment (Figure  5B). We stained liver sections for F4/80, a marker of macrophages (Figure  5C and E) and Ly6G, a marker for neutrophils (Figure  5D and F). Our data indicate that Yap-KO livers had significantly higher numbers of macrophages and neutrophils in the liver at 24 h after APAP overdose. These data suggested a heightened and faster inflammatory response in the Yap-KO mice.

Yap-KO Mice Exhibit Faster β-Catenin Activation as Compared to Control Mice After APAP Overdose

Previous studies from our and other laboratories have shown that canonical Wnt/β-catenin signaling plays a critical role in liver regeneration after APAP overdose (Apte et al., 2009; Bhushan et al., 2014). We investigated the status of canonical Wnt pathway activation in control and Yap-KO mice. Western blot analysis (Figure  6A and B) for activated β-catenin, the dephosphorylated active form of the downstream effector of the Wnt/β-catenin pathway indicated showed a faster induction in Yap-KO mice as early as 12 h after APAP treatment, which continued till 48 h. Control mice also showed an increase in active β-catenin at 24 and 48 h after APAP overdose, but it was significantly lower than Yap-KO mice. We determined mRNA expression of several canonical Wnt target genes using qPCR at 24 h after APAP administration. The data revealed a substantial activation of several Wnt target genes including Regucalcin, Cyclin D1, Axin 2, Lect2, and others in Yap-Ko mice (Figure  6C–E).

Upstream Regulator Analysis Reveals Additional Mechanisms Including Maintenance of Hepatic Differentiation and Lower endoplasmic reticulum Stress

The gene array data were used to conduct Upstream Regulator Analysis (URA) using the Ingenuity Pathway Analysis (IPA, Table  1). This analysis predicted XBP1, a transcription factor (YF) that alleviates endoplasmic reticulum (ER) stress, HNF4α, the master regulator of hepatocyte differentiation, as the 2 most activated TFs in the Yap-KO mice after APAP treatment. Additionally, activation HNF1α, a well-characterized target gene of HNF4α, and SREBP2, an associated TF involved in regulation of lipid metabolism, was also predicted in Yap-KO mice after APAP overdose (Supplementary Figure 2A) and hepatic differentiation (Supplementary Figure 2B). Furthermore, IPA analysis also predicted downregulation of several TFs including CREB1, an activator of gluconeogenesis; Myc, an oncogene which shows an inverse relationship with HNF4α expression; HIF1α, a TF responsive to hypoxia and NUPR1, a TF involved in stress-associated regulation of signal transduction. Some of the top gene expression changes that dictated the URA results are shown in Supplementary Figure 2. Overall, these data revealed that Yap-KO mice initiate a transcriptional program associated with maintenance of hepatocyte differentiation and decreased cellular stress.

DISCUSSION

Extensive studies in last decade have revealed central role of Yap and Hippo Kinase pathway in liver pathobiology (Driskill and Pan, 2021; Yimlamai et al., 2015; Zhao et al., 2008). Although Yap activation has been implicated mainly in liver cancer pathogenesis, its role in liver development, postnatal growth, and regeneration have also been recognized (Li et al., 2012; Septer et al., 2012; Yi et al., 2018). In general, studies show that Yap activation is associated with increased cell proliferation and change in hepatocyte fate, both of which can promote liver carcinogenesis. Our studies are the first to demonstrate that deletion of Yap can promote regeneration and recovery from APAP overdose. Our initial investigation with 300 mg/kg dose of APAP in male C57BL/6J mice raised the question whether Yap activation observed during the first 24 h after APAP treatment is related to development of liver injury or stimulation of liver regeneration. Immunohistochemistry analysis of Yap showed cytoplasmic localization of Yap at 24 h, 48 h, and 72 h after APAP, time points where hepatocytes proliferate, liver regenerates, and recovers from the acute injury. Supported by Yap target gene expression, these data hinted that Yap inhibition may be required for faster hepatocyte proliferation after APAP overdose. This possibility was further supported by significantly faster recovery of hepatocyte-specific Yap-KO mice.

Hepatocyte-specific deletion of Yap did not affect events critical for APAP-induced hepatocyte death including the metabolic activation and protein binding (Ramachandran and Jaeschke, 2019). Yap deletion in hepatocytes did not affect baseline values of CYP2E1 protein and hepatic GSH levels. In addition, the bioactivation of APAP to NAPQI was not affected as indicated by similar GSH depletion at 1 h after APAP and presence of similar amount of protein adducts during the early injury phase. Furthermore, downstream signaling pathways such as JNK activation were also not affected in the Yap-deficient mice. This was consistent with the finding that both control and Yap-KO mice had similar liver injury up to first 12 h after APAP overdose. However, the Yap-KO mice recovered significantly faster than WT mice from APAP-induced ALI. This was due to faster initiation of cell proliferation, which increased significantly earlier than the Control mice (12 h vs 48 h). At 48 h time point, the WT mice showed moderately higher cell proliferation than Yap-KO mice. However, the Yap-KO mice had significantly earlier and sustained proliferation as compared to WT mice despite similar increase in liver injury. Previous studies have demonstrated that early and sustained stimulation of liver regeneration is critical for rapid recovery following drug-induced acute liver (Apte et al., 2002; Bhushan et al., 2013, 2014). These data indicate that deletion of Yap did not impact the injury phase but resulted in rapid recovery from APAP overdose due to faster liver regeneration.

Further studies revealed that faster cell proliferation in Yap-KO mice was secondary to 2 major events. First, we observed a significantly higher and faster activation of β-catenin in the Yap-KO livers starting at 12 h after APAP overdose and continuing till 48 h in Yap-KO as compared to starting at 24 h and 48 h after APAP treatment in Control mice. This observation is consistent with our previous reports that β-catenin activation is critical for stimulation of liver regeneration after an APAP overdose (Apte et al., 2009; Bhushan et al., 2014). Signaling between Yap and β-catenin is context dependent. Although mutated versions of Yap and β-catenin can conspire to induce liver tumors (Sylvester and Colnot, 2014; Yimlamai et al., 2015), specifically hepatoblastoma (Tao et al., 2014), other studies have shown that Yap can inhibit β-catenin in normal hepatocytes and dampen the canonical Wnt signaling response (Kim et al., 2017a,b; Park et al., 2015). We observed significantly faster and sustained induction in activated (unphosphorylated) β-catenin protein and activation of several canonical Wnt/β-catenin target genes in the Yap-KO mice after APAP overdose. The exact nature of signaling interaction between Yap and β-catenin following APAP overdose is not clear. It is possible that phosphorylated Yap can bind to Disheveled and inhibit canonical Wnt signaling (Imajo et al., 2012). Alternatively, Yap could bind directly to β-catenin, sequestering it and reducing its nuclear translocation. Further studies are needed to identify these mechanisms.

The second significant mechanism is a robust and faster inflammatory response in Yap-KO mice following APAP overdose. Gene array data, which was confirmed by qPCR analysis, showed a significant increase in SAA (SAA1 and SAA2) in Yap-KO livers. SAA are acute phase proteins and are known to initiate pro-regenerative cytokine signaling networks (De Buck et al., 2016; Schrodl et al., 2016; Urieli-Shoval et al., 2000; Ye and Sun, 2015). Interestingly, we observed relatively moderate increase in serum SAA as compared to the massive increase in gene expression of SAA. This difference may be because of lag in protein synthesis, differences in local liver mRNA expression and systemic protein expression and possibly due to some anti-inflammatory signaling that may have limited translation of mRNA. Nevertheless, we observed significantly higher number of F4/80⁺ macrophages and Ly6G⁺ neutrophils in the livers of Yap-KO mice following APAP overdose. The role of inflammation in APAP-induced hepatotoxicity has been debated with both pro-injury and pro-recovery arguments (Jaeschke and Ramachandran, 2020; Woolbright and Jaeschke, 2018). Our data show that in mice with hepatocyte deletion of Yap, which have equal initial liver injury and faster recovery due to rapid liver regeneration, exhibit a significantly higher proinflammatory response following APAP overdose. Furthermore, promoter analysis of SAA1 and SAA2 genes shows several TEAD4 binding sites suggesting that Yap along with its common DNA binding partner TEAD could inhibit expression of these genes (data not shown). These data indicate that inhibition of hepatocyte Yap results in a pro-regenerative inflammatory response after APAP overdose.

The gene array data further revealed the mechanisms that drive faster recovery from APAP-induced hepatotoxicity following Yap deletion. Upstream regulator analysis indicated increased activity of transcription factors XBP1 and HNF4α. XBP1 is a transcription factor involved in ER stress sensing and its prompt induction is essential in resolving unfolded protein response (Malhi and Kaufman, 2011). Gene expression data indicate that Yap-KO mice have a higher XBP1 activation suggesting a faster response to UPR, which should aid in faster recovery. Similarly, IPA analysis indicated that Yap-KO mice had higher activation of HNF4α and its target gene HNF1α, both transcription factors critical in maintaining hepatocyte differentiation (Walesky and Apte, 2015; Watt et al., 2003). This is an interesting observation because several studies have shown that Yap activation in hepatocytes can induce a dedifferentiation and progenitor cell like phenotype. It is known that Yap overexpression results in hepatocyte fate change and recent studies demonstrated that this dampens the regenerative response of the hepatocytes (Bou Saleh et al., 2021; Camargo et al., 2007; Yimlamai et al., 2014). Our data indicate that Yap activation observed during initial time points following APAP overdose may result in a general dedifferentiation, which is prevented in the Yap-KO mice. This should result in better maintenance of hepatocyte function in the liver, which will help in faster recovery. Upstream regulator analysis also showed inhibition of Myc, a multipotent transcription factor involved in metabolism and cell proliferation. The observation that Yap-KO mice have higher HNF4α and lower Myc activity is consistent with our previous studies which showed that HNF4α and Myc have an inverse activation status (Walesky et al., 2013).

Taken together, our data indicate that hepatocyte-specific deletion of Yap does not affect the injury phase after an APAP overdose but results in faster recovery from APAP-induced ALI. This is because of the rapid cell proliferation and liver regeneration in the Yap-KO mice secondary to faster β-catenin activation and pro-regenerative inflammatory response. Our studies have revealed a novel signaling crosstalk between Yap, β-catenin and HNF4α and highlighted Yap as a novel therapeutic target in APAP-induced ALF.

SUPPLEMENTARY DATA

Supplementary data are available at Toxicological Sciences online.

DECLARATION OF CONFLICTING INTERESTS

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

FUNDING

This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases (grant/award number: R01 DK98414 and R56 DK112768).

REFERENCES

Author notes