Introduction
Liver cancer is widely prevalent and ranks as the leading cause of cancer mortality in the digestive system worldwide.1 Hepatic fibrosis (HF), an essential pathological stage in chronic liver disease, is critical to the progression of hepatocellular carcinoma (HCC), developing from early inflammation to cirrhosis and then to liver cancer.2,3 Although significant advancements have been made in understanding the pathophysiological mechanism of liver fibrosis,4–7 there is currently no effective strategy to reverse the end stage of HF, apart from liver transplantation. Therefore, identifying new therapeutic targets and understanding their underlying mechanisms of action are critical to developing novel strategies to prevent or delay liver fibrogenesis.
Liver fibrosis is characterized by the activation and proliferation of hepatic stellate cells (HSCs) and subsequent extracellular matrix (ECM) accumulation due to persistent hepatocellular injury and inflammatory stimulation.8 Hepatocyte damage induced by various stimuli forms the initial stage of pathogenesis, contributing to inflammatory cell recruitment and cytokine secretion,9 activating the transition of HSCs into myofibroblasts that produce collagen fibers, resulting in excessive ECM deposition.3,5,9 However, most previous reports focus on the central role of HSCs and/or Kupffer cells in experimental HF models and human liver injuries.4,5,7,10 Few studies have examined hepatocytes, which are frequently damaged in chronic liver injuries, although they are not typically regarded as the primary effector cells responsible for HF progression.11,12 Hence, it is extremely important to explore the specific molecular interactions and underlying mechanisms in damaged liver cells for early HF prevention.
Mitochondrial dysfunction is a hallmark of liver fibrosis due to its frequent occurrence in various chronic liver diseases.13,14 Increasing evidence shows that mitochondrial disorders are associated with HF development primarily through oxidative stress damage to hepatocytes induced by excessive ROS.6,11–13 This emphasizes the need to pay more attention to mitochondrial homeostasis in hepatocytes during HF pathogenesis. However, there is limited characterization of the mechanisms protecting hepatocytes from mitochondrial damage in liver fibrogenesis.
The gene associated with retinoid-IFN-induced mortality 19 (GRIM19), also known as NDUFA13, a vital component of the mitochondrial respiratory chain (MRC) complex I,15 has been reported to be inhibited in numerous human cancers, including gastric cancer and HCC,15–17 indicating its potential role as a tumor suppressor. Our previous findings revealed a progressive downregulation of GRIM19 in chronic atrophic gastritis tissues, a precancerous lesion of gastric cancer, suggesting a crucial role of GRIM19 loss in inflammatory diseases and their premalignant progression.16–18 However, whether GRIM19 loss is involved in liver fibrogenesis remains largely unknown. This study aimed to investigate the potential relationship between GRIM19 loss and HF pathogenesis and to explore underlying mechanisms in vitro and in vivo.
Methods
Human liver disease spectrum tissue microarrays (TMA)
Liver TMA slides containing 104 samples, including normal liver tissues (n=16), hepatitis (n=24), hepatic cirrhosis (n=32), and HCC (n=32), were commercially purchased from US Biomax, Inc (Cat. LV20812, corresponding clinical information and ethical materials were included). All samples were correctly labeled clinically and pathologically. The TMA slide was stained using IHC staining and scored as previously described.16–18
Cells, tissue culture, and reagents
Murine hepatocyte cells AML12 (Procell, Wuhan, China) were maintained in DMEM/F12 medium supplemented with 10% FBS (Gibco, Gaithersburg, MD, USA), 1× insulin-transferrin-selenium media supplement (Procell, Wuhan, China), 40 ng/mL dexamethasone (Procell, Wuhan, China), streptomycin (100 ug/mL), and penicillin (100 U/mL) at 37°C in a humidified 5% CO2 atmosphere. The ROS scavenger N-acetylcysteine (NAC), NF-кB inhibitor PDTC, and Caspase1 inhibitor VX765 were obtained from AbMole Bioscience (Houston, TX, USA). The NLRP3 inhibitor MCC950 was from MCE (Houston, TX, USA). All chemical reagents were commercially obtained from Sigma-Aldrich (St Louis, MO, USA) unless otherwise indicated. All antibodies used in this study are provided in Supplementary Table 1.
GRIM19 gene-editing mediated by CRISPR/Cas9 lentivirus
CRISPR/Cas9-mediated recombinant lentivirus carrying specific small-guide RNA targeting GRIM19 at position 11656 (sgRNA-56), 11657 (sgRNA-57), and 11658 (sgRNA-58), as well as the corresponding negative control (NC), were commercially constructed with a GFP reporter gene by GeneChem Co (Shanghai, China). AML12 cells were infected by CRISPR/Cas9 recombinant lentivirus with a multiplicity of infection of 30–50, and GFP-positive cells were sorted by flow cytometry (FACS Aria II, BD) to establish single-cell clones following established protocols.17 Western blotting and immunofluorescence (IF) staining were performed to determine GRIM19 expression.
Transgenic mice model and pharmacological intervention in vivo
GRIM19 conditional knockout (CKO) mice GRIM19flox/flox (GRIM19fl/fl) with the exon 3 of GRIM19 flanked by loxP (C57BL/6J X129/Sv background) were commercially obtained from Cyagen Bio Inc (Suzhou, China) as previously described.18–20 Alb-Cre transgenic mice [(B6.Cg-Tg(Alb-cre)21Mgn/J, JAX-003574)] were purchased from the Shanghai Biomodel Organism Center (Shanghai, China). Hepatocyte-specific GRIM19 knockout (KO) mice were acquired by crossing GRIM19fl/fl and Alb-Cre transgenic mice. DNAs from tail or liver tissues were subjected to PCR genotyping, while GRIM19 expression was confirmed by Western blotting and IF. GRIM19 heterozygous KO mice (GRIM19fl/−/Alb-Cre, named GRIM19fl/−) were used in this study, and GRIM19fl/fl mice were littermate controls. For experimental intervention in mice, GRIM19fl/− mice (eight to nine months) were randomly divided into two groups (n=8 mice/group) and administrated with either PBS control or MCC950 (10 mg/kg in PBS) via intraperitoneal (i.p.) injection three times per week for four consecutive weeks. Animals were kept, housed, and bred under specific pathogen-free conditions as previously described.16–20 All procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Chongqing Medical University (CHCMU-IACUC20210114023).
Histology and Immunohistochemistry (IHC)
Histology and IHC staining were performed as previously described.16–18 Liver tissues were fixed in 10% buffered formalin for 48 h, dehydrated in various ethanol concentrations, and embedded in paraffin for preparing histopathological analysis. Tissue sections (4 µm) underwent H&E staining and Masson’s trichrome staining (Solarbio, Beijing, China) in accordance with the manufacturer’s instructions. The fibrosis stage in the liver was evaluated based on the Ishak score. IHC staining was conducted utilizing Elivision plus Polymer HRP IHC Kit (Maixin, Fujian, China) and DAB Kit (ZSGB-Bio, Beijing, China). The number of positive cells and staining intensity were evaluated to determine the IHC score, as previously specified.16–18
Immunofluorescence (IF) staining
IF staining was performed on paraffin-embedded tissue sections (4 µm) or fresh frozen liver tissues (8 µm) as described previously.17–20 Briefly, normal goat serum (Solarbio, Beijing, China) was used to block tissue sections for 1 h at room temperature. Primary antibodies were incubated overnight at 4°C, followed by AF488, AF555, or AF647-conjugated secondary antibody (Bioss, Beijing, China) according to the manufacturer’s instructions at room temperature. 4′,6-diamidino-2-phenylindole (DAPI) was used to stain the nuclei. Images were acquired by the laser confocal microscope (A1R, Nikon, Japan), and mean fluorescent intensity (MFI) was quantified using NIS-Element 5.20 Software.
Intracellular ROS and mitochondrial ROS (mROS) measurements
Intracellular ROS and mROS in cells or fresh frozen tissue sections (8 µm) were measured using 10 µM Dihydroethidium (DHE) (Sigma, USA) or 5 µM MitoSOX Red (Thermo Fisher Scientific, Waltham, USA) by flow cytometry or IF staining as described previously.17,18 Indicated cells were resuspended, and MFI was measured by flow cytometry on a FACS Calibur flow cytometer (BD Bioscience), with data analyzed using FlowJo software (TreeStar, Ashland, OR). DAPI was used to stain the nuclei. Images were obtained using the laser confocal microscope (A1R, Nikon, Japan), and MFI was quantified through NIS-Element 5.20 Software.
Intracellular adenosine triphosphate (ATP) level assay
Intracellular ATP levels were measured using an ATP Assay Kit (Beyotime Biotech, Haimen, China) following the manufacturer’s instructions. AML12 cells (1.0×106) were homogenized and lysed in 200 µL of ice-cold lysis buffer, then centrifuged at 12,000 g for 5 m at 4°C. Relative Light Units were determined from the supernatant using a Synergy H1 microplate reader (Bio Tek, USA).
Intracellular reduced glutathione (GSH)/oxidized glutathione (GSSG) ratio and NADP+/NADPH ratio analysis
The intracellular NADP+/NADPH ratio and GSH/GSSG ratio were determined using the NADP+/NADPH Assay Kit (Beyotime, Haimen, China) and GSH/GSSG Ratio Detection Assay Kit (Beyotime, Haimen, China), respectively. The oxidative status in AML12 cells was assessed using a Synergy H1 microplate reader (Bio Tek, CA, USA) according to the manufacturer’s instructions as described previously.17,18
Western blotting
Protein extraction and western blotting were performed as described previously.18–20 Liver tissues or AML12 cells were subjected to total protein extraction. Cellular components were prepared using NE-PER nuclear and cytoplasmic extraction reagents (Thermo Fisher Scientific, Waltham, USA). Protein extracts were quantified using an enhanced BCA Protein Assay Kit (Beyotime, Haimen, China). Samples (20–30 µg) were resolved by SDS-PAGE and transferred to polyvinylidene fluoride membranes (0.22 µM) (Roche, Mannheim, Germany). QuickBlock™ Blocking Buffer (Beyotime, Haimen, China) was applied to block the polyvinylidene fluoride membranes, followed by incubation with the indicated primary antibody overnight at 4°C. HRP-conjugated secondary antibodies were then incubated for 1–2 h at room temperature after washing with TBST buffer. Visualization was performed using a WesternBright ECL kit (Advansta, Menlo Park, CA, USA) on a ChemiDocTMTouch Imaging System (Bio-Rad, Hercules, CA, USA). Total protein levels were normalized using β-actin as a loading control.
Flow cytometry
Flow cytometry was performed on a FACS Calibur flow cytometer (BD Bioscience), and the data were analyzed using FlowJo software (Tree Star, Ashland, OR) as described previously.17,18 GFP-positive AML12 cells were sorted after transfection.
Statistical analysis
Statistical analysis was performed using GraphPad Prism 8.0 (GraphPad, San Diego, CA) as previously described.17,18 All data were expressed as the mean±SD unless otherwise specified. An unpaired Student’s t-test (two groups) or a one-way analysis of variance (more than two groups) was applied to analyze statistical differences. A p-value of p<0.05 was considered statistically significant.
Results
GRIM19 is downregulated in human chronic liver fibrosis tissues
GRIM19 was identified as frequently reduced in various human cancers, including HCC.15–17,21 To investigate whether GRIM19 loss is associated with liver fibrogenesis, we evaluated GRIM19 protein levels in tissues from patients with chronic liver diseases using human liver disease spectrum TMAs containing normal liver, hepatitis, cirrhosis, and HCC tissues. As shown in Figure 1A, IHC staining revealed progressive downregulation of GRIM19 protein in hepatitis, cirrhosis, and HCC tissues compared with normal liver tissues. Furthermore, we observed that GRIM19 protein was predominantly distributed in the cytoplasm of normal liver tissues with dense dot-like or uniform patterns, whereas in hepatitis, cirrhosis, and HCC tissues, GRIM19 was clearly present perinuclearly or even in the nucleus with a dense dot distribution (Fig. 1B). These findings suggest that the downregulation of GRIM19 may correlate with the progression of chronic liver diseases.
Hepatocyte-specific GRIM19 heterozygous deletion is sufficient to trigger spontaneous liver fibrogenesis in mice
Given the downregulation of GRIM19 in chronic human liver diseases, we investigated whether GRIM19 loss is responsible for liver fibrogenesis. To this end, we generated a genetic hepatocyte-specific GRIM19 KO mice model using the Cre-loxP-based system by crossing transgenic GRIM19 CKO and Alb-Cre mice to obtain hepatocyte-specific heterozygous GRIM19 deletion mice (Supplementary Fig. 1). H&E staining demonstrated that hepatocyte-specific GRIM19 loss caused abnormal liver structure, sinusoidal destruction, and swollen hepatocytes compared with littermate controls (Fig. 2A). Interestingly, we did not observe obvious accumulation of extracellular collagen fibers in GRIM19-deleted liver tissues from four-week-old mice, as demonstrated by Masson staining (Fig. 2B). However, we observed a significant infiltration of CD45+ and MPO+ cells and a slight increase in F4/80+ cells (Fig. 2C). Therefore, these findings suggest that hepatocyte-specific GRIM19 ablation triggers spontaneous hepatitis at an early stage.
To explore whether GRIM19 loss-induced hepatitis could progress to HF, we extended our observation of hepatocytes-specific GRIM19 deletion mice to a period of two years. Similar to the results from four-week-old mice, H&E staining showed the presence of abnormal liver structures and damaged hepatocytes in two-year-old GRIM19-KO mice compared with littermate controls (Fig. 2D). However, as demonstrated in Figure 2E with Masson staining, we observed the accumulation of extracellular collagen fibers in GRIM19-deficient livers, indicating that liver fibrogenesis from hepatitis progressed with age. We also observed significant infiltrations of CD45+, MPO+, and F4/80+ cells (Fig. 2F). Tumor necrosis factor-α (TNFα) (Supplementary Fig. 2A, B) and TGFβ1 (Supplementary Fig. 2C, D) were significantly increased and co-localized with activated F4/80+ cells in GRIM19-deleted liver tissues, indicating that macrophage-derived TNFα and TGFβ1 play a vital role in GRIM19 loss-induced liver fibrogenesis. Therefore, our data suggest that heterozygous GRIM19 deletion induces persistent hepatitis and consequent HF with age.
Next, we further investigated HSC activation and the presence of fibrosis-associated proteins in liver tissues from two-year-old GRIM19-deficient mice. As shown in Figure 2G, GRIM19 ablation significantly increased both α-SMA and Desmin levels, indicating enhanced proliferation and activation of HSCs. Besides the fibrotic marker α-SMA, pro-fibrotic proteins Collagen I and TIMP1 demonstrated a marked increase (Fig. 2H, I), indicating an imbalance in ECM synthesis and degradation in GRIM19-deleted liver tissues. Moreover, we found that activated HSCs with normal GRIM19 expression (Supplementary Fig. 3A) were surrounded by higher levels of TGFβ1 and Collagen III (Supplementary Fig. 3B, C), indicating that macrophage-derived TGFβ1 may contribute to HSC activation. Together, these results suggest that hepatocyte-specific GRIM19 deletion is sufficient to trigger spontaneous liver fibrogenesis in mice.
GRIM19 loss induces chronic liver injury by oxidative stress
To investigate the mechanism by which GRIM19 loss induces liver fibrogenesis in vitro, we used CRISPR/Cas9 sgRNA-mediated gene editing to knock down GRIM19 expression in the murine hepatocyte AML12 cell line (Supplementary Fig. 4A, B). Among three sgRNA targets, CRISPR/Cas9-mediated sgRNA-11658 (sgRNA-58) demonstrated the most effective inhibition of GRIM19 expression (Supplementary Fig. 4C). We then isolated stable GRIM19-deficient AML12 clones by FACS sorting (Supplementary Fig. 4D). Consequently, AML12 cells expressing sgRNA-58 (AML12-58 cells) or a negative control (NC) (AML12-NC cells) were used for subsequent analysis.
Next, we investigated the effect of GRIM19 loss on chronic liver injury, a critical initial event for liver fibrogenesis.3,9 As shown in Figure 3A, B, we observed that GRIM19 loss markedly increased intracellular ROS and mROS, as well as 8-OHdG, a marker of ROS-induced DNA damage18in vitro. Furthermore, GRIM19 loss significantly decreased ATP content (Fig. 3C) and the GSH/GSSG ratio (Fig. 3D, left panel), while increasing the NADP+/NADPH ratio (Fig. 3D, right panel) in AML12 cells, indicated that GRIM19 loss causes chronic hepatocyte injury via ROS-induced oxidative stress. In addition, we observed increased intracellular ROS (Fig. 3E), mROS production (Fig. 3F), and 8-OHdG levels (Fig. 3G) in GRIM19-deficient mice livers, indicating that GRIM19 plays an essential role in the homeostasis of oxidative stress in hepatocytes. Overall, these data suggest that GRIM19 loss exerts a crucial role in oxidative damage-induced hepatocyte injury.
GRIM19 loss induces ROS-dependent NF-κB activation in vitro and in vivo
NF-кB activation has been frequently involved in liver inflammation.22 Given the dual role of ROS in oxidative damage and NF-кB signaling, we examined the effect of GRIM19 loss on NF-кB activation. We found that the expression of both p65 and its phosphorylated form, p-p65, was significantly enhanced after GRIM19 knockdown in vitro (Fig. 4A). We also observed a clear increase in p-p65 levels, accompanied by a minor increase in p65 levels, in vivo (Fig. 4B). Furthermore, NF-кB-responsive targets including IL6, TNFα, VEGF, VCAM1, and ICAM1 were significantly increased both in vivo and in vitro (Fig. 4A, B). We discovered markedly altered NF-кB regulator proteins including p-IKKα/β, IKKα/β, p-IκB-α, and IκB-α (Fig. 4C), and also found that cytoplasmic and nuclear p65 levels were both increased after GRIM19 loss (Fig. 4D–F). In addition, NF-кB inhibitor PDTC significantly abrogated GRIM19 loss-induced p65 activation and downstream targets in AML12 cells (Fig. 4G). To further clarify the role of ROS in GRIM19 loss-induced NF-кB activation, NAC (a ROS scavenger) was applied to determine whether ROS suppression could inhibit GRIM19 loss-induced NF-кB activation. As shown in Figure 4H, I, NF-кB activation, its downstream targets, and the NF-кB regulator proteins normally promoted by GRIM19 loss were significantly attenuated after NAC intervention in AML12-58 cells. Hence, these results suggest that GRIM19 loss significantly activates NF-кB in a ROS-dependent manner.
GRIM19 loss-driven activation of the NLRP3 inflammasome is critical to liver fibrogenesis in mice
It has been shown that mROS facilitates liver fibrogenesis by activating the NLRP3 inflammasome,23,24 which directly promotes liver fibrosis by activating Kupffer cells (KCs) and HSCs in mice.25 Therefore, we investigated whether the NLRP3 inflammasome is involved in GRIM19 loss-driven liver fibrogenesis. Using hepatocyte-specific GRIM19 KO mice, we observed a significant increase in NLRP3, ASC, and Caspase1 protein expression (Fig. 5A). Moreover, IL1β and IL33, two critical NLRP3-induced proinflammatory cytokines, were markedly elevated in GRIM19-deficient liver tissues as shown by western blotting (Fig. 5A), indicating aberrant NLRP3 activation during liver fibrogenesis. Furthermore, dual IF staining revealed that the increased levels of NLRP3, IL33, and IL1β were mainly observed in hepatocytes (Fig. 5B) rather than HSCs (Supplementary Fig. 5) in KO mice. These findings indicate that hepatocyte-specific GRIM19 deletion can directly trigger abnormal activation of the NLRP3 inflammasome.
To clarify the crucial role of NLRP3 activation in GRIM19 loss-induced liver fibrogenesis, we applied MCC950, an NLRP3 inhibitor, through intraperitoneal administration to analyze whether NLRP3 inhibition could alleviate the fibrotic liver phenotype in GRIM19 KO mice (Supplementary Fig. 6A). We found that MCC950 treatment significantly decreased NLRP3 inflammasome activation and IL1β and IL33 levels in GRIM19 KO mice (Fig. 5C–D). Furthermore, we also observed a decrease in GRIM19 loss-induced CD45+, MPO+ cells, and M2-type macrophages after MCC950 administration (Supplementary Fig. 6B, C, Supplementary Fig. 7). Interestingly, NLRP3 inhibition also attenuated GRIM19 loss-induced liver fibrosis, with reduced collagen deposition around the hepatic central vein (Fig. 5E) and decreased expression of α-SMA, fibrotic molecules Collagen I, and TIMP1 (Fig. 5F, G). These results suggest that NLRP3 activation is critical to GRIM19 loss-induced liver fibrogenesis in mice.
GRIM19 loss triggers aberrant NLRP3/IL33 activation through ROS/NF-κB signaling
Considering the increase in IL33 induced by NLRP3 in GRIM19 KO mice, we attempted to explore the regulatory pathway involved in NLRP3/IL33 activation, which may play a pivotal role in HF pathogenesis.26 Consistent with our observations in liver tissues in vivo, we found considerably elevated levels of IL1β and IL33 in GRIM19-deficient AML12 cells (Fig. 6A, B). Therefore, we applied specific inhibitors to determine the exact roles of ROS and NF-кB in NLRP3/IL33 activation. We found that both NAC (ROS scavenger) and PDTC (NF-кB inhibitor) markedly diminished GRIM19 loss-driven NLRP3 inflammasome activation, as well as IL1β and IL33 expression (Fig. 6C, D). We also used MCC950 and VX765 (an inhibitor of Caspase1) to further investigate the role of the NLRP3 complex in the processing of IL1β and IL33. As shown in Figure 6E, F, both MCC950 and VX765 treatment abrogated GRIM19 loss-driven IL1β and IL33 levels, indicating that the NLRP3 inflammasome exerts a pivotal role in GRIM19 loss-induced IL1β and IL33 expression in AML12 cells. Thus, these data indicate that GRIM19 loss contributes to abnormal NLRP3/IL33 activation through the ROS/NF-кB pathway.
Discussion
Deepening our understanding of the potential molecules involved in liver fibrogenesis would not only enhance our comprehension of HF pathogenesis but also identify promising approaches to treat or prevent liver fibrogenesis. Our previous study revealed a progressive decrease in GRIM19 protein during the progression of chronic atrophic gastritis, indicating a potential role for GRIM19 in digestive tract-related inflammation.16–18 Although our preliminary data showed that GRIM19 inactivation promoted spontaneous hepatitis and liver fibrosis in mice,19,20 the underlying mechanism remained largely unknown. Thus, it is crucial to understand whether and how GRIM19 loss contributes to liver fibrogenesis, particularly at the onset of hepatocyte injury, which is an initial event in chronic liver disease. Using human liver disease spectrum TMAs, we identified a progressive decrease in GRIM19 in chronic liver disease tissues ranging from hepatitis to cirrhosis to HCC. Following these findings, we generated a GRIM19 KO mice model and GRIM19-deficient AML12 cells, revealing that hepatocyte-specific GRIM19 knockout triggered spontaneous hepatitis followed by HF, caused by NLRP3/IL33 activation through ROS/NF-κB signaling. This study is the first to systematically elucidate the functional role of GRIM19 in liver fibrogenesis, offering a prospective diagnostic marker and therapeutic target for HF treatment.
As the most abundant cells in the liver, hepatocytes play a vital role in maintaining internal homeostasis and liver regeneration.27 Although HSC activation has been well characterized during liver fibrogenesis in response to various inflammatory signals,28–30 hepatocytes are considered the main target of most irritants, such as oxygen radicals, in various chronic liver diseases.31 Thus, our investigation focused on potential targets responsible for hepatocyte damage observed during liver fibrogenesis. Mitochondria have recently received significant attention for their association with liver fibrogenesis. Oxidative stress induced by mitochondrial structural or functional disorders has been defined as one of the main causative agents of liver fibrosis.11,13 mROS, a product of mitochondrial MRC I, significantly contributes to oxidative damage.32,33 Mitochondrial GRIM19, a key subunit of mitochondrial MRC I, is considered essential for its normal functioning,34 especially in maintaining mitochondrial activity.35,36 Here, we have shown that GRIM19 loss led to oxidative DNA damage by inducing abnormal intracellular ROS and mROS in vitro and in vivo, resulting in decreased ATP content and GSH/GSSG ratios, as well as increased NADP+/NADPH ratios in hepatocytes. These findings are further supported by a recent study on cardiac-specific GRIM19 heterozygous deletion, where an ROS leak from complex I increased cytoplasm-localized H2O2.34 Therefore, our data suggest that GRIM19 loss in hepatocytes is responsible for oxidative stress-induced liver injury, leading to liver fibrosis.
Our findings suggest that the NLRP3 inflammasome plays a crucial role in GRIM19 loss-induced liver fibrosis in vivo. Over the past several years, significant advances have been made in understanding the critical role of the NLRP3 inflammasome in HF.23–25 Activation of the NLRP3 inflammasome can directly trigger chronic inflammation,25 whereas NLRP3 inhibition can improve liver inflammation pathology and modulate the most critical outcome of liver fibrosis.24 Importantly, the activated NLRP3 inflammasome can also activate resident KCs, which produce the inflammatory cytokine TNFα and profibrotic factor TGFβ1, resulting in HSC activation and liver fibrogenesis.4,5,7,23–25 This is consistent with our observations in the two-year-old mice livers. The assembly of the inflammasome complex with ASC and caspase1 is mainly mediated by NLRP3 and facilitates the cleavage of proinflammatory cytokines such as IL1β and IL33.37 mROS, as a secondary messenger in inflammatory signaling, plays a key role in the activation of the NLRP3 inflammasome through the NF-кB pathway.38–40 We have found that GRIM19 loss induces chronic liver injury through ROS-mediated oxidative damage, resulting in aberrant NF-кB activation via an IKK/IкB partner. We have further revealed that GRIM19 loss induces abnormal NLRP3/IL33 activation through ROS/NF-кB signaling in vivo and in vitro, while pharmacological NLRP3 inhibition can attenuate GRIM19 loss-induced hepatitis and liver fibrosis in mice. Thus, our findings suggest that GRIM19 loss-triggered NLRP3/IL33 activation is responsible for HF pathogenesis, providing mechanistic insights into how GRIM19 loss drives liver fibrogenesis.
Our findings demonstrate NLRP3-dependent IL33 secretion and maturation during GRIM19 loss-induced liver fibrosis, offering promising prospects for clinical applications in the diagnosis and treatment of liver fibrosis. IL33, a member of the IL1 cytokine family, is considered an alarm hormone involved in immune regulation and the inflammatory response.41 Clinical studies have revealed a significant accumulation of IL33 levels in the plasma of patients with chronic hepatitis, as well as in biopsy liver tissues of patients with HF and cirrhosis.42,43 Additionally, increased IL33 levels can be detected in CCl4-induced liver fibrosis in mice, while IL33 knockout significantly inhibited CCl4 or cholestasis-induced HF progression.43,44 Furthermore, IL33 derived from injured liver cells was shown to induce innate immune cells to produce TH2-type cytokines, promoting the polarization of M2 macrophages and further activating HSCs by releasing fibrotic factors such as TGFβ1.45 Our study reveals that NLRP3 inhibition not only decreased GRIM19 loss-induced IL33 activation but also abrogated GRIM19 loss-driven liver fibrosis in vivo. Furthermore, we found that NLRP3 inhibition reduced GRIM19 loss-induced M2 macrophage aggregation, suggesting a crucial role for NLRP3/IL33 in GRIM19 loss-induced M2 macrophages during liver fibrosis. A recent study also identified NLRP3 as a critical transcription factor for IL33 in the epithelial cells of atopic dermatitis,37 further indicating a crucial role for NLRP3-dependent IL33 activation in GRIM19 loss-driven liver fibrogenesis. Thus, our findings demonstrate NLRP3-dependent IL33 secretion and maturation during GRIM19 loss-induced liver fibrosis, providing a potential diagnostic target for clinical diagnosis of liver fibrosis and offering a promising therapeutic target for clinical treatment.
Our study presents a novel HF mouse model in which hepatocyte-specific GRIM19 ablation can trigger spontaneous liver fibrosis, offering potential strategies for studying clinical interventions for liver fibrosis. Generally, genetic models rarely develop spontaneous liver fibrosis due to genetic manipulation and require restimulation to induce disease, except for MDR2-deficient mice, which develop biliary fibrosis.46 Our study revealed that the genetic knockdown of GRIM19 was sufficient to trigger spontaneous chronic liver fibrosis, providing potential clinical applications for preclinical drug assessment against liver fibrosis or therapeutic intervention of liver fibrosis through GRIM19 gene transfer. Intriguingly, our most recent study revealed a significant decrease in GRIM19 protein levels in CCl4-treated fibrotic liver tissues, while adeno-associated virus 8 (AAV8)-mediated GRIM19 overexpression alleviated CCl4-induced liver fibrogenesis by inhibiting NLRP3 activation,47 indicating an essential role for GRIM19 activation in chemical-induced liver fibrosis. However, it remains unclear whether GRIM19 loss is also involved in diet-induced liver fibrosis, such as methionine- and choline-deficient, choline-deficient L-amino acid-defined, or high-fat high-sugar dietary animal models of nonalcoholic steatohepatitis, which are considered more suitable models to mimic pathophysiological progression than chemical CCl4-induced liver fibrosis models.48 Thus, further efforts are needed to investigate the role of GRIM19 loss in diet-related chronic liver diseases.
Our study has some limitations. First, although genomic hypermethylation was associated with GRIM19 downregulation,17 the diverse and complex stimuli of liver injury make it challenging to understand the mechanism of GRIM19 loss in hepatocytes. Second, more long-term observations are needed to determine whether GRIM19 loss-driven liver fibrogenesis can progress into spontaneous HCC. Finally, we noted that hepatocyte-specific heterozygous GRIM19-deleted mice presented a lower birthrate compared to either GRIM19 CKO or Alb-Cre mice (data not shown), indicating an essential role for GRIM19 in liver development in mice. Similar results can also be found in a recent report that complete GRIM19 depletion was lethal for mouse embryos.21,34 Thus, a tamoxifen-induced specific GRIM19 knockout in hepatocytes should be performed using Alb-Cre-ERT2 mice in the future. Mechanistically, previous reports showed a critical role of GRIM19 loss in autophagy in colorectal cancer and adenomyosis;49,50 however, our unpublished data shows that neither apoptosis nor autophagy are involved in GRIM19 loss-induced liver fibrogenesis, suggesting that an unidentified mechanism contributes to GRIM19 loss-induced liver fibrogenesis.
Conclusion
Our research demonstrates that mitochondrial GRIM19 loss in hepatocytes triggers chronic liver fibrogenesis through NLRP3/IL33 activation via ROS/NF-κB signaling (Supplementary Fig. 8). This finding not only identifies a causal relationship between GRIM19 loss and liver fibrosis but also offers prospective therapeutic approaches for the treatment of liver fibrogenesis.
Supporting information
Supplementary Table 1
Antibodies used for Western blot, IHC, IF, and FACS in this study.
(DOCX)
Supplementary Fig. 1
Generation of hepatocyte-specific GRIM19 (gene associated with retinoid-IFN-induced mortality 19) knockout mice.
(A) The schematic diagram for generating hepatocyte-specific GRIM19 knockout (KO) mice. GRIM19 conditional knockout (CKO) mice (GRIM19fl/fl) were commercially constructed by targeting embryonic stem cells with a pair of loxP sites flanking exon 3 of the GRIM19 gene. Hepatocyte-specific heterozygous GRIM19 KO (GRIM19fl/−/Alb-Cre, named GRIM19fl/−) mice were generated by intercrossing GRIM19fl/fl and Alb-Cre transgenic mice. (B) DNA from tail and liver tissues were subjected to PCR genotyping for GRIM19 Flox and Cre alleles (Cre allele: Wild type=351bp, Mutant=150bp; Flox allele: Wild type=302bp; Mutant=382bp). (C) Western blotting was used to detect GRIM19 levels in the liver tissues of control and GRIM19fl/− mice. (D) Dual immunofluorescence (IF) staining for CK18 and GRIM19 expression in control and GRIM19fl/− mice. β-actin was used as a loading control. DAPI was used to stain the nuclei. Mean fluorescent intensity (MFI) was used to quantify the protein expression after IF staining. Data are presented as mean±SD of three independent experiments. Representative images are shown. Scale bars: 200 µm. ***p<0.001 between the indicated two groups determined by unpaired student’s t-test. DAPI, 4′,6-diamidino-2-phenylindole.
(TIF)
Supplementary Fig. 2
GRIM19 (gene associated with retinoid-IFN-induced mortality 19) loss increases macrophage-derived TNFα and TGFβ1 in liver tissues.
(A) Immunohistochemistry (IHC) staining for TNFα in the liver tissues from control and GRIM19fl/− mice with two years of age. (B) Dual immunofluorescence (IF) staining for F4/80 and TNFα expression in the liver tissues from control and GRIM19fl/− mice with two years of age. (C) IHC staining for TGFβ1 in the liver tissues from control and GRIM19fl/− mice with two years of age. (D) Dual IF staining for F4/80 and TGFβ1 expression in the liver tissues from control and GRIM19fl/− mice with two years of age. DAPI was used to stain the nuclei. Mean fluorescent intensity (MFI) was used to quantify the protein expression after IF staining. Data are presented as mean±SD of three independent experiments. Representative images are shown. Scale bars: (main) 200 µm; (inset) 50 µm. ***p<0.001 between the indicated two groups determined by unpaired student’s t-test. DAPI, 4′,6-diamidino-2-phenylindole.
(TIF)
Supplementary Fig. 3
Hepatocyte-specific GRIM19 (gene associated with retinoid-IFN-induced mortality 19) deletion promotes TGFβ1 and Collagen III expression in HSCs (hepatic stellate cells).
Dual immunofluorescence (IF) staining was used to detect co-expression of Desmin and GRIM19 (A), α-SMA and TGFβ1 (B), α-SMA and Collagen III (C) in the liver tissues from control and GRIM19fl/− mice with 2 years of age. DAPI was used to stain the nuclei. Mean fluorescent intensity (MFI) was used to quantify the protein expression. Data are presented as mean±SD of three independent experiments. Representative images are shown. Scale bars: 50 µm. ***p<0.001 between the indicated two groups determined by unpaired student’s t-test. α-SMA, alpha-smooth muscle actin; DAPI, 4′,6-diamidino-2-phenylindole.
(TIF)
Supplementary Fig. 4
CRISPR/Cas9 lentivirus-mediated GRIM19 (gene associated with retinoid-IFN-induced mortality 19) gene editing in AML12 cells.
Clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein-9 (Cas9) lentivirus-mediated sgRNA-58 target was used to knockout the GRIM19 gene in AML12 cells. (A) Schematic diagram of three sgRNA sites (11656, 11657, and 11658) targeting the GRIM19 gene. (B) CRISPR/Cas9 lentivirus vector and the corresponding GRIM19 sgRNA targets. (C) GRIM19 expression was detected by western blotting in AML12 cells. (D) GFP-positive AML12-58 cells were analyzed by flow cytometry after cell expansion. Negative control (NC) was used as a control. β-actin was used as a loading control. Three independent experiments were performed. Representative images are shown. AML-12, alpha mouse liver 12.
(TIF)
Supplementary Fig. 5
GRIM19 (gene associated with retinoid-IFN-induced mortality 19) loss induces aberrant NLRP3/ IL33 activation in liver tissues.
Dual immunofluorescence (IF) staining was used to detect co-expression of α-SMA & NLRP3 (A) and α-SMA & IL33 (B) in control and GRIM19fl/− mice liver. DAPI was used to stain the nuclei. Mean fluorescent intensity (MFI) was used to quantify the protein expression. Representative images are shown. Data are expressed as mean±SD. Scale bars: 50 µm. ***p<0.001 between the indicated two groups determined by unpaired student’s t-test. α-SMA, alpha-smooth muscle actin.
(TIF)
Supplementary Fig. 6
Pharmacological NLRP3 inhibition attenuates GRIM19 (gene associated with retinoid-IFN-induced mortality 19) loss-induced hepatitis phenotypes.
(A) Schematic diagram for pharmacological intervention of NLRP3 inhibitor MCC950 in hepatocytes-specific GRIM19 deficient mice. GRIM19fl/− mice were treated with PBS or MCC950 (10 mg/kg in PBS) three times per week (n=8 mice/group) for four consecutive weeks, by intraperitoneal (i.p.) injection. (B) Hematoxylin and eosin (H&E) staining was performed to analyze histological changes in the liver. (C) Immunohistochemistry (IHC) staining for immune cell markers CD45, F4/80, and MPO in liver tissues of PBS or MCC950- treated GRIM19fl/− mice. Positive cells per field were analyzed, respectively. Representative images are shown. Data are expressed as mean±SD. Scale bars: (main) 200 µm; (inset) 50 µm. ***p<0.001 between the indicated groups determined by unpaired student’s t-test. MPO, myeloperoxidase; PBS, phosphate-buffered saline.
(TIF)
Supplementary Fig. 7
MCC950 treatment suppresses GRIM19 (gene associated with retinoid-IFN-induced mortality 19) loss-induced M2 macrophage polarization.
(A) Co-expression of F4/80 & CD86 and F4/80 & CD163 was detected by dual immunofluorescence (IF) staining in control and GRIM19fl/− mice liver. (B) Co-expression of F4/80 & CD86 and F4/80 & CD163 was detected by dual IF staining in liver tissues from MCC950-treated GRIM19fl/− mice. DAPI was used to stain the nuclei. F4/80+CD86+ and F4/80+CD163+ cells were counted for M1 and M2 macrophages, and the proportions were shown, respectively. Mean fluorescent intensity (MFI) was used to quantify the protein expression. Representative images are shown. Data are expressed as mean±SD. Scale bars: 50 µm. *p<0.05, **p<0.01, ***p<0.001 between the indicated two groups determined by unpaired student’s t-test. phosphate-buffered saline.
(TIF)
Supplementary Fig. 8
Schematic diagram of mitochondrial GRIM19 loss-induced liver fibrosis by ROS/NF-κB/NLRP3/IL33 pathway.
(A) GRIM19 maintains homeostasis in hepatocytes in normal liver tissues. GRIM19 protects hepatocytes against disturbed reactive oxygen species (ROS) and mitochondrial homeostasis, thus keeping the extracellular matrix (ECM) synthesis and degradation balance. (B) GRIM19 loss is involved in liver fibrogenesis. Decrease of GRIM19 expression induces ROS-mediated oxidative damage and triggers aberrant NLRP3/IL33 activation via ROS/NF-кB signals, resulting in the activation of resident Kupffer cells (KCs) and hepatic stellate cells (HSCs), thereby contributing to the imbalance of ECM synthesis and degradation.
(TIF)
Declarations
Acknowledgement
We thank Qixi Huang (Clayton High School, Clayton, Missouri, United States) for her writing assistance on the early version of this manuscript.
Ethical statement
All procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Chongqing Medical University (CHCMU-IACUC20210114023). All animals received humane care.
Data sharing statement
The data supporting the findings of this study are available in the supplementary material and the remaining data are available upon reasonable request from the corresponding author.
Funding
This study was partially supported by the National Nature Science Foundation of China [No.32171119; No.32371173]; the general basic research project from the Ministry of Education Key Laboratory of Child Development and Disorders [GBRP-202116]; the Nature Science Foundation of Chongqing Science and Technology Bureau [CSTB2022NSCQ-MSX0838]; the Science and Technology Research Program of Chongqing Municipal Education Commission [KJZD-K202100401]; and the Future Medical Youth Innovation Team Support Project of Chongqing Medical University [W0175].
Conflict of interest
The authors have no conflict of interests related to this publication.
Authors’ contributions
Conception and design (XX, JF, YH), experiments and data analysis (XX, JF, XW, XZ, XH, YL, ZF, LZ, DH), article drafting (XX, JF, YH), critical article revision for important intellectual content (XX, MY, TL, LB, YH). All authors participated in the discussions and approved the manuscript.