Targeting Endothelial Necroptosis Disrupts Profibrotic Endothelial–Hepatic Stellate Cells Crosstalk to Alleviate Liver Fibrosis in Nonalcoholic Steatohepatitis

Chronic liver diseases affect over a billion people worldwide and often lead to fibrosis. Nonalcoholic steatohepatitis (NASH), a disease paralleling a worldwide surge in metabolic syndromes, is characterized by liver fibrosis, and its pathogenesis remains largely unknown, with no effective treatment available. Necroptosis has been implicated in liver fibrosis pathogenesis. However, there is a lack of research on necroptosis specific to certain cell types, particularly the vascular system, in the context of liver fibrosis and NASH. Here, we employed a mouse model of NASH in combination with inducible gene knockout mice to investigate the role of endothelial necroptosis in NASH progression. We found that endothelial cell (EC)-specific knockout of mixed lineage kinase domain-like protein (MLKL), a critical executioner involved in the disruption of cell membranes during necroptosis, alleviated liver fibrosis in the mouse NASH model. Mechanistically, EC-specific deletion of Mlkl mitigated the activation of TGFβ/Smad 2/3 pathway, disrupting the pro-fibrotic crosstalk between endothelial cells and hepatic stellate cells (HSCs). Our findings highlight endothelial MLKL as a promising molecular target for developing therapeutic interventions for NASH.


Introduction
Non-alcoholic fatty liver disease (NAFLD) is a prevalent condition in the world [1] characterized by the accumulation of fat in the livers of individuals who do not consume substantial amounts of alcohol [2]. NAFLD is considered a significant public health concern due to its potential to progress to more severe liver damage, including non-alcoholic steatohepatitis (NASH), cirrhosis, and hepatocellular carcinoma (HCC) [3]. NASH is a more severe form of NAFLD characterized by ballooned hepatocytes along with steatosis and lobular inflammation, leading to liver fibrosis [4]. The incidence of non-alcoholic fatty liver disease (NAFLD) in China has been increasing year by year. According to a nationwide epidemiological survey, the prevalence of NAFLD among Chinese adults has exceeded 20%, and the rate is over 60% in the obese population [5,6]. It is estimated that approximately 20-30% of patients with NAFLD will progress to NASH [4]. The prevalence of NASH in the Chinese population ranges from 2.4 to 6.1% [7][8][9][10][11], which is close to the

Construction of Inducible Endothelial Cell (EC)-Specific Knockout of Mlkl Mice (Mlkl i∆EC/i∆EC )
Necroptosis has been implicated in liver fibrosis, yet there is a lack of research on necroptosis specific to certain cell types, particularly within the vascular system, in the context of liver fibrosis and NASH. To examine whether necroptosis of endothelial cells is involved in NASH and the fibrosis model, we generated a conditional Mlkl knockout mouse model using CRISPR/Cas9 technology. A gRNA targeting sequence of the mouse Mlkl gene following the PAM site was inserted into the Exon3 locus and introduced into mice ( Figure 1A). Genotyping assays were performed to distinguish homozygote (Mlkl f/f ), heterozygote (Mlkl f/+ ), and wild-type (WT) mice ( Figure 1B top). We also screened mice carrying VE-cad-Cre ERT2 to subsequently breed conditional knockout mice ( Figure 1B bottom). Mice carrying EC-specific promoter Cdh5/VE-cadherin-driven tamoxifen-responsive Cre (VE-cad-Cre ERT2 ) were crossed with Mlkl-floxed mice. The injection of tamoxifen into the resulting mouse offspring selectively induced deletion of Mlkl in vascular endothelial cells, generating EC-specific knockout of Mlkl in adult mice (Mlkl i∆EC/i∆EC ). Mice expressing Cdh5-PAC-Cre ERT2 alone served as the control group ( Figure 1C). After inducible deletion of Mlkl by tamoxifen injections, ECs were isolated from liver tissues. RT-qPCR showed that Mlkl relative mRNA levels in liver ECs of Mlkl i∆EC/i∆EC mice dropped to 25% compared with Mlkl +/+ mice ( Figure 1D). In parallel, we also examined knockout efficiency after tamoxifen injection, as assessed by the protein levels of MLKL ( Figure 1E). MLKL protein levels in liver ECs were reduced significantly. The result of quantification by image J was consistent with Western blot ( Figure 1F). These data indicate that construction of EC-specific knockout of Mlkl mice is successful.

EC-Specific Knockout of Mlkl Alleviates Histopathological Phenotype Associated with NASH Progression
To explore the influence of necroptosis of ECs on liver fibrosis in NASH model, repetitive CCl 4 injection and a high fat and sugar diet (WD diet) were used to induce the NASH model ( Figure 2A). One dose of CCl 4 per week was injected into the enterocoelia of mice, and liver functions, steatosis, inflammation, liver fibrosis, collagen deposition, and index of liver fibrosis were examined after 12 weeks. Firstly, liver functions of the control and Mlkl i∆EC/i∆EC mice were assessed. We found that the liver functions of genetic knockout of Mlkl mice was improved. Specifically, the serum concentration of aspartate amino transferase (AST), alanine aminotransferase (ALT), and alkaline phosphatase (ALP) were decreased in Mlkl i∆EC/i∆EC mice compared with the control group ( Figure 2B). Steatosis is a major pathological feature of NASH [59]. Therefore, we performed Oil Red O staining and found that lipid droplet deposition in the livers of the control group was higher than that large surface area ( Figure 2C). The change in serum cholesterol (CHO) in Mlkl i∆EC/i∆EC mice was consistent with the above results ( Figure 2B). Next, we assessed the histopathology and collagen deposition ( Figure 2D, right and 2E) in the livers of the control and Mlkl i∆EC/i∆EC mice. Compared with the control group, which exhibited prominent histological features of NASH, including inflammation and ballooned hepatocytes (as indicated by black arrows), the livers of Mlkl i∆EC/i∆EC animals showed an improved phenotype ( Figure 2D left).    The excessive accumulation of extracellular matrix (ECM) proteins, including collagen (especially Collagen I and III), is a prominent hallmark of liver fibrosis. In addition, αsmooth muscle actin (α-SMA) serves as a marker of pathologic fibroblasts in fibrotic tissue. Therefore, to further assess the molecular characteristics of fibrosis, total RNA was isolated from the livers of Mlkl i∆EC/i∆EC and Mlkl +/+ NASH mice and RT-qPCR was performed. The mRNA expression levels of Col1a1, Col3al, and α-SMA (Acta2) were upregulated in Mlkl +/+ mice compared with Mlkl i∆EC/i∆EC mice ( Figure 3A). Meanwhile, the liver hydroxyproline (HYP) abundance, one of the main components of collagen, also significantly reduced in the NASH model with endothelial Mlkl deletion ( Figure 3B). Consistently, we also examined the fibrotic responses in the NASH model, as assessed by the protein levels of α-SMA, Desmin, and Collagen I ( Figure 3C,D). The protein levels of α-SMA, Desmin, and Collagen I were reduced in the genetic deletion of Mlkl mice. Analysis of immunostained liver sections obtained from individuals with NASH also suggested downregulation of Collagen I ( Figure 3E), α-SMA ( Figure 3F), and Desmin ( Figure 3G) deposition compared to the corresponding expression in control mice livers ( Figure 3H). Collectively, these findings suggest that the fibrotic response in Mlkl i∆EC/i∆EC mice is downregulated in the NASH model, indicating that MLKL plays an important role in the progression of liver fibrosis.

EC-Specific Knockout of Mlkl Downregulates the Liver Fibrosis Indexes after Repeated CCl 4 Injury
To examine whether the effects of necroptosis of ECs in the NASH model can be generalized to another liver injury model, we used repeated CCl 4 injection to investigate the contribution of vascular maladaptation to liver fibrosis and the underlying mechanisms ( Figure 4A). According to our previous research, both control mice and Mlkl i∆EC/i∆EC mice were treated with 40% CCl 4 every 3 days to establish a chronic liver fibrosis model [60]. After 3 weeks, mice were sacrificed and liver tissues were harvested to evaluate the fibrotic responses. Since lymphatic vessel endothelial receptor-1 (Lyve-1) is the marker of ECs, endothelial damage after injury was tested by immunostaining of the ECs markers Lyve-1 ( Figure 4B). The staining results suggested a liver vascular maladaptation in this mouse liver injury model. We found that, in the Mlkl i∆EC/i∆EC group, liver function parameters of mice, including ALT, ALP, and total bilirubin (TBIL), were lower than those in the control group ( Figure 4C). As explicated in Figure 4D, there were large contiguous fibrotic masses in the H&E staining liver sections in the control group. Furthermore, the liver architecture was severely damaged, and part of it was not preserved. In the Sirius red staining results, collagen fibers were dyed bright yellow or red. It showed that the mean staining intensity of collagen deposition is about 50% of the microscopic field. However, after knockout of Mlkl in mouse liver ECs, the damage of the liver tissue was reduced. There was no doubt that knockout of Mlkl in liver ECs could reduce collagen fibers deposition compared to the control group, and the liver sections of the Mlkl i∆EC/i∆EC group tended to be normal.
Next, we found that in the CCl 4 -induced fibrosis model, the mRNA expression of Col1a1, Col3a1, and α-SMA increase, while these elevations were abrogated in the Mlkl i∆EC/i∆EC group, as shown by RT-qPCR ( Figure 5A). Similarly, we also found that the level of hydroxyproline was increased after liver injury but reduced in the Mlkl i∆EC/i∆EC group. In parallel, we examined fibrotic responses after injection of CCl 4 , the protein levels of α-SMA and Collagen I were downregulated in the Mlkl i∆EC/i∆EC group, compared to the control group ( Figure 5C,D). Immunofluorescence results, as quantified in Figure 5E-G, verified that knockout of Mlkl in mouse liver ECs restrained the upregulation of Collagen I and Desmin induced by CCl 4 injection. Based on these findings, we postulated that necroptosis of ECs might contribute to impaired liver repair and fibrosis after injury. Therefore, targeting of endothelial MLKL during chronic liver injury could potentially alleviate fibrosis.    Next, we found that in the CCl4-induced fibrosis model, the mRNA expression of Col1a1, Col3a1, and α-SMA increase, while these elevations were abrogated in the control group ( Figure 5C,D). Immunofluorescence results, as quantified in Figure 5 verified that knockout of Mlkl in mouse liver ECs restrained the upregulation of Coll I and Desmin induced by CCl4 injection. Based on these findings, we postulated necroptosis of ECs might contribute to impaired liver repair and fibrosis after in Therefore, targeting of endothelial MLKL during chronic liver injury could potent alleviate fibrosis.

EC-Specific Knockout of Mlkl Reduces the Pro-Fibrotic TGFβ/Smad 2/3 Signaling Axis between Endothelial Cells and HSCs
Based on the significant attenuation of liver fibrosis in NASH mice with EC-specific knockout of Mlkl, we sought to define the molecular mechanisms underlying the antifibrotic effect of Mlkl i∆EC/i∆EC mice. The TGFβ/Smad2/3 pathway is a major profibrogenic signal for HSCs activation and induction of the fibrotic response [61]. Primary endothelial cells were isolated from livers of indicated mice, and the RNA extracting from harvested ECs in Figure 6A was used for RT-qPCR. We found that the expression of profibrotic gene TGFβ was decreased in the NASH model of Mlkl i∆EC/i∆EC mice ( Figure 6B). To determine the effects of endothelial MLKL on the TGFβ/Smad 2/3 signaling in HSCs, we examined the protein levels of the phosphorylation of Smad2/3 (p-Smad2/3) in corresponding experimental groups. Immunoblotting showed that p-Smad2/3 levels were reduced in the genetic deletion of Mlkl mice ( Figure 6C). We also performed double immunofluorescence staining of p-Smad2/3 with α-SMA (an activated HSC marker) [48]. In the Mlkl +/+ NASH mice, we observed strong colocalization of p-Smad2/3 and α-SMA (white arrow in Figure 6D). However, the proportion of colocalization was dramatically reduced in the genetic deletion of Mlkl mice ( Figure 6D), indicating that necroptosis of endothelial cells might regulate the fibrosis response via TGFβ/Smad 2/3 signaling axis. Collectively, these findings indicate that EC-specific deletion of Mlkl effectively alleviates liver fibrosis in NASH by suppressing the activation of the TGFβ/Smad 2/3 pathway and disrupting the pro-fibrotic crosstalk between endothelial cells and HSCs.

Discussion
The pathogenesis of NASH is a complex process and its cellular and molecular mechanisms remain unclear, which hinders the development of therapeutic strategies for NASH [62][63][64]. In the present study, we showed that endothelial necroptosis contributes to NASH progression and blockage of endothelial necroptosis alleviates liver fibrosis in NASH. NASH development and progression have been linked to different forms of cell death, including necroptosis, ferroptosis, pyroptosis, and apoptosis [65][66][67][68]. In addition to necroptosis of endothelial cells in our study, there is emerging evidence that necroptosis in other cell types may play a role in NASH pathogenesis and modulation of necroptosis in other cell types has been shown to alleviate liver fibrosis and NASH progression in experimental models [69,70]. A potential therapeutic approach for biliary atresia (BA) might be to target macrophage necroptosis, which plays an important role in BA liver fibrosis pathogenesis [71]. It is reported that necroptotic hepatocytes can trigger the recruitment of immune cells, leading to inflammation and contributing to NASH development and progression. Clearance of necroptotic hepatocytes by liver macrophages is a potential strategy to mitigate hepatic stellate activation and dampen liver fibrosis in NASH [72][73][74][75]. Hepatic stellate cell necroptosis induced by curcumol might provide a therapeutic strategy for the treatment of liver fibrosis [76]. Overall, these findings suggest that targeting necroptosis in various cell types may represent an effective therapeutic strategy to treat liver fibrosis and dampen NASH progression.
Our study suggests that endothelial necroptosis contributes to liver sinusoidal endothelial cell loss and dysfunction, thereby resulting in NASH development and progression. Capillaries are lined with endothelial cells, which act as a barrier between the circulatory system and tissues [77,78]. As part of blood vessels, endothelial cells participate in delivering blood and nutrients. Vascular endothelial cells can also establish vascular niches by deploying various angiocrine factors such as TGFβ, hepatocyte growth factor (HGF), and Wnt signaling molecules to guide a variety of physiological processes, including liver regeneration, fibrosis, and NASH progression [41,57,79]. Therefore, it is possible that as a result of necroptosis, endothelial cells become lost and dysfunctional, leading to maladaptive vascular niches causing liver fibrosis in NASH.
Liver sinusoidal endothelial cells (LSECs) can influence the occurrence and development of NASH by regulating adjacent hepatic stellate cells (HSCs) and immune cells [38,39,42,80]. Deleve et al. have shown that LSECs can prevent HSC activation and fibrosis in physiological conditions [81]. However, in pathological conditions, LSECs promote the activation of HSCs and macrophages by secreting pro-fibrotic and pro-inflammatory factors, leading to liver fibrosis [82][83][84]. Activated Kupffer cells promote LSEC capillarization, thereby regulating the hepatic microenvironment by directly secreting ECM and pro-fibrotic factors, promoting the progression of steatosis to NASH [44,[85][86][87][88][89][90]. In this study, we found that the necroptosis of endothelial cells promotes TGFβ expression and activates HSCs, thereby promoting liver fibrosis in NASH. TGFβ is a vital profibrogenic cytokine for HSCs activation, which is the dominant event in hepatic fibrogenesis. Mechanically, TGFβ binds to its receptor (TβRI, TβRII) and leads to the phosphorylation of Smad2/3 [61]. The p-Smad2/3 is required for their nuclear translocation and transcriptional regulation of their fibrotic target genes (Figure 7) [61]. Our results indicate that necroptosis of endothelial cells might produce a pro-fibrotic endothelial signal involving TGFβ which activates Smad2/3 signaling in HSCs. Indeed, EC-specific deletion of Mlkl blocked necroptosis and reduced the activation of the TGFβ/Smad 2/3 pathway, thereby blocking a pro-fibrotic communication between endothelial cells and HSCs (Figure 7). Targeting of the communication between endothelial cells and HSCs could potentially facilitate therapy for liver fibrosis and NASH. Guo et al. have shown that Breviscapine, a natural flavonoid prescription drug isolated from the traditional Chinese herb Erigeron breviscapus, alleviates NASH through reducing lipid accumulation, pro-fibrotic factors, and inflammation in hepatocytes [91]. LSECs and hepatocytes are the two most abundant cell types in the liver, with LSECs accounting for 15%-20% of the total cells in livers [92]. Dysfunctional hepatocytes and endothelial cells may primarily regulate the occurrence and development of NASH through synergistic effects. Therefore, developing approaches that target dysfunctional hepatocytes and endothelial cells in a coordinated manner to reduce pro-fibrotic factors and inhibit inflammation and HSC activation may be an effective therapeutic strategy for NASH.
LSECs and hepatocytes are the two most abundant cell types in the liver, with LSECs accounting for 15%-20% of the total cells in livers [92]. Dysfunctional hepatocytes and endothelial cells may primarily regulate the occurrence and development of NASH through synergistic effects. Therefore, developing approaches that target dysfunctional hepatocytes and endothelial cells in a coordinated manner to reduce pro-fibrotic factors and inhibit inflammation and HSC activation may be an effective therapeutic strategy for NASH. Currently, there is no approved drug for the treatment of NASH, and it is considered an area of highly unmet medical need [93,94]. The global NASH drug market is expected to exceed USD 10 billion by 2025 [95]. Several drugs targeting specific molecular pathway are currently in clinical trials to mitigate NASH [93,[96][97][98]. MLKL represents a critical factor in the process of necroptosis [18,99]. In the present study, we found that EC-specific knockout of MLKL effectively alleviates the progression of NASH. Building upon this insight, researchers can design inhibitors that specifically target MLKL in endothelial cells, or develop specific adeno-associated viruses (AAV) to knock down MLKL in endothelial cells. This approach achieves the objective of specifically blocking endothelial necroptosis, repairing the vascular microenvironment, and mitigating liver fibrosis in NASH. Currently, there is no approved drug for the treatment of NASH, and it is considered an area of highly unmet medical need [93,94]. The global NASH drug market is expected to exceed USD 10 billion by 2025 [95]. Several drugs targeting specific molecular pathway are currently in clinical trials to mitigate NASH [93,[96][97][98]. MLKL represents a critical factor in the process of necroptosis [18,99]. In the present study, we found that EC-specific knockout of MLKL effectively alleviates the progression of NASH. Building upon this insight, researchers can design inhibitors that specifically target MLKL in endothelial cells, or develop specific adeno-associated viruses (AAV) to knock down MLKL in endothelial cells. This approach achieves the objective of specifically blocking endothelial necroptosis, repairing the vascular microenvironment, and mitigating liver fibrosis in NASH.

Mice
C57BL/6J mice were obtained from GemPharmatech Co., Ltd. (Nanjing, China) and C57BL/6JSmoc-Mlkl em1(flox)Smoc mice were acquired from Shanghai Model Organisms Co., Ltd. (Shanghai, China). All mice were maintained in a pathogen-free facility at the Experimental Animal Center of West China Second Hospital, and were fed freely on a standard 12-h light-dark cycle. Animal experiments were approved by the Experimental Animal Ethics Committee of West China Second Hospital of Sichuan University. All experimental mice were randomly allocated to each group.
The inducible Mlkl knockout mice with loxP-flox fragments were generated through homologous recombination of the fertilized egg using CRISPR/Cas9 technology. Specifically, the flox region of the coding sequence (CDS) region of Mlkl was replaced, and the recombinant site was on chromosome 8 (ENSMUSG00000012519.15). According to the Cre-loxP system, tamoxifen-induced activation of Cre recombinase can effectively delete the sequence between two loxP sites, achieving specific knockout of the Mlkl gene. To achieve endothelial cell-specific gene knockout, mice carrying endothelial cell (EC)-specific promoter Cdh5/VE-cadherin-driven tamoxifen-responsive Cre (VE-cad-CreERT2) were crossed with Mlkl-floxed mice. Mice were injected intraperitoneally with 2% tamoxifen (Sigma-Aldrich, Burlington, MA, USA; cat: T5648) dissolved in corn oil (Sigma), which were treated once a day for three days, and then treated again at intervals of three days, for a total of nine times. The dosage is 100 mg/kg body weight (calculated by mouse weight), as previously described [60,100].

Genotyping
To obtain DNA for genotyping, 1-2 mm tissues were removed from the toe or tail tip of mice and mixed with 100 µL tissue lysis solution, which contained 1 x MGB buffer, 10% Triton, 1% β-Mercaptoethanol, premixed with protease K (20µL protease K/mL lysis solution; Sigma-Aldrich; cat: 39450-01-6). The mixture was incubated at 56 • C overnight. After denaturation, the samples were stored at 4 • C as a DNA template for genotyping. For the detection of Mlkl and Cre, 2× Taq Plus MasterMix with dye (Cwbio, Beijing, China; cat: CW2849) was mixed with the related primers, and polymerase chain reactions (PCR) were performed. Finally, the PCR products were subjected to agarose gel electrophoresis to determine the genotype, as previously described [41,60].

NASH Model
NASH model was induced by a Western diet (WD) and chemical injury [101]. After the mice were treated with tamoxifen for three weeks, they were fed with a high-fat diet containing 21.1% fat, 41% sucrose, and 1.25% cholesterol, in a weight ratio, along with a high sugar aqueous solution diet (23.1 g/L fructose and 18.9 g/L glucose) for 12 weeks. In addition, 20% carbon tetrachloride (CCl 4 ) in corn oil was injected intraperitoneally once a week for 12 weeks, with a dosage of 2.5 mL/kg body weight (calculated by mouse weight). After 12 weeks of treatment, serum was collected from the fundus venous plexus, and the mice were sacrificed and perfused with 0.9% saline through the mice left ventricle. The liver lobes were then harvested for further testing, as previously described [63,102].

Chemically Induced Liver Fibrosis Model
Mice were fed with a standard diet for at least three weeks after the induction of the tamoxifen treatment. Mice were then injected intraperitoneally with 40% carbon tetrachloride/corn oil solution every three days in a total of three times [41]. Control mice were treated with the same dose of corn oil for the same duration. After three times of treatment, mouse serum was obtained through the fundus venous plexus, and the mice were sacrificed and perfused with 0.9% saline through the left ventricle. The liver lobes were harvested for further testing, as previously described [60].

Blood Biochemistry
Blood samples were collected and kept at room temperature for 30 min. After centrifugation at 3000 g for 20 min at 4 • C, the supernatant was aspirated, and serum was isolated and stored at −80 • C for biochemical detection. The serum levels of aspartate transaminase (AST), alanine transaminase (ALT), ALP (alkaline phosphatase), total bilirubin (TBIL), and cholesterol (CHO) were measured by the automated biochemical analyzer (Roche, Beijing, China;) for animals, as previously described [94].

Histochemical Staining
Liver tissues were harvested for histological analysis flowing a previously described protocol [41,57]. In brief, after sacrificing the mice, a piece of liver from each mouse was fixed in 4% paraformaldehyde (PFA) overnight at 4 • C. The liver tissues were then dehydrated, embedded in paraffin wax, and sectioned into 5 µm slices. These slices were stained with hematoxylin and eosin (H&E) and Sirius red for morphological assessment. Frozen liver tissue sections (10µm) were stained with Oil Red O to evaluate histopathological changes under microscopy (3D Pannoramic MIDI, Budapest, Hungary).

Immunofluorescence
Fresh liver samples were embedded in Tissue-Tek ® O.C.T. Compound and stored at −80 • C. Liver cryosections (8 µm in thickness) were fixed in 4% paraformaldehyde; all sections were permeabilized in 0.3% Triton X-100 for 20 min and then blocked with 2% BSA for 10 min. The sections were incubated with primary antibodies against Collagen I (Abcam, Waltham, MA, USA; cat: ab34710), α-SMA (Abcam; cat: ab7817), Desmin (Abcam; cat: ab15200) and Lyve1 (Fitzgerald, Dublin, Ireland; cat: 70R-LR003) overnight at 4 • C. Corresponding fluorescent secondary antibodies conjugated with Alexa Fluor 488 (1:500; Jackson ImmunoResearch, West Grove, PA, USA) were used to stain the sections for 1 h at room temperature in the dark, after which the primary antibodies on the slides were washed away. The nuclei were stained with DAPI (Thermo Fisher Scientific, Waltham, MA, USA) for 5 min. The slides were imaged with a Zeiss LSM 900 microscope and analyzed with ImageJ software (v1.8.0, Institutes of Health (NIH) Bethesda, MD, USA), as previously described [102].

Isolation of Endothelial Cells from Liver Tissues
On the last day of NASH and CCl 4 -induced liver fibrosis modeling, mice were sacrificed and perfused with 0.9% saline through the left ventricle. Perfused liver tissues were minced and dissociated with digestive enzymes, including Collagenase I (Gibco, Billings, MT, USA; cat: 17100-017), DNase I (Sigma-Aldrich; cat: D4527), at 37 • C for 25 min. The digested cell suspension containing liver sinusoidal endothelial cells (LSECs) were incubated with Dynabeads magnetic beads (Invitrogen, Waltham, MA, USA; cat: 11035) coated with rat anti-mouse CD31 (BD Bioscience, San Jose, CA, USA; cat: 553370). Negative depletion with anti-CD45 (BD Bioscience; cat: 553076) Dynabeads were performed to exclude hematopoietic cells. The cells collected after all treatments were CD31 + CD45 − . The collected LSECs were stored at −80 • C for further experiments, as previously described [103].

RNA Extraction and Real-Time Quantitative PCR (qPCR)
Total RNA was extracted from the liver tissues or isolated liver endothelial cells using TRIzol TM (Invitrogen), following the manufacturer's instructions. RNA reverse transcription was performed using the PrimeScript TM RT reagent Kit (Takara, San Jose, CA, USA; RR047A-1), according to the manufacturer's instructions. Relative RNA expressions were detected using SYBR Green Master Mix (Vazyme, Nanjing, China; cat: Q712-02/03) on a CFX96 Real-Time PCR System (Bio-Rad, Hercules, CA, USA), as previously described [60].

Western Blotting
The liver tissue and the collected LSECs were lysed with RIPA buffer containing 1% PMSF and loaded into the wells of the SDS-PAGE gel. After transferring the gel containing proteins onto PVDF membranes, the immunoblots were incubated with the primary antibodies against GAPDH (Servicebio, Wuhan, China; GB11002), MLKL (Huabio; cat: ET1601-25), Smad2/3 (Cell Signaling Technology, Danvers, MA, USA; cat: 8685), and p-Smad2/3 (Cell Signaling Technology; cat: 8828) for more than 8 h at 4 • C. The corresponding HRP-linked secondary antibody was then added and incubated for 1 h. The PVDF membrane containing target protein was laid on the exposure plate for the detection of protein. ImageJ software was used for image analysis as previously described [103].

Hydroxyproline (HYP) Analysis
Hydroxyproline levels were detected using a HYP assay kit (Solarbio, Beijing, China; cat: BC0255). Liver tissues were cut into pieces and digested in the extraction solution at approximately 100 • C for 2-6 h, until the tissues were totally hydrolyzed. The PH of the mixture was adjusted to 10 using 10 M NaOH. The adjusted mixture was then centrifuged at 16,000 rpm for 30 min, and the supernatant was collected for testing. The OD value of each well was measured at 560 nm with a SpectraMax ABS PLUS reader. The HYP content was calculated as previously described [94].

Statistical Analysis
All data were expressed as mean ± SEM. Statistical analysis was performed using Student's t-test for comparisons between two groups or one-way ANOVA for experiments with more than two groups, using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). A p value < 0.05 was considered statistically significant.

Conclusions
In summary, our findings indicate that the occurrence of endothelial necroptosis during the progression of NASH exacerbates the pathological process. Notably, EC-specific ablation of MLKL effectively attenuates liver fibrosis in NASH by blocking necroptosis and mitigating the TGFβ/Smad 2/3 signaling axis between endothelial cells and HSCs. Our study contributes to a better understanding of the underlying mechanisms involved in NASH progression and provides novel therapeutic approaches for this disease.