Amyloid precursor protein promotes MASH progression by upregulating death receptor 6–mediated hepatocyte apoptosis

Abstract

Metabolic dysfunction–associated steatohepatitis (MASH) is a complicated process that contributes to end-stage liver disease and, eventually, hepatocellular carcinoma. Hepatocyte apoptosis, a well-defined form of cell death in MASH, is considered the primary cause of liver inflammation and fibrosis. However, the mechanisms underlying the regulation of hepatocyte apoptosis in MASH remain largely unclear. We explored the proapoptotic effect of hepatocyte amyloid precursor protein (APP) in MASH. C57BL/6J mice were fed a Western diet plus sugar water, a high-fat high-fructose diet, or a methionine and choline deficiency diet to induce MASH. APP expression was analyzed in murine MASH specimens. App^−/− mice and mice with adeno-associated virus–mediated APP overexpression were established to study the role of APP in MASH. Palmitic acid was used to mimic lipotoxicity-induced MASH in AML12 cells. We identified a dramatic increase in APP expression in hepatocytes of patients with MASH and three different mouse models. Suppression of APP attenuated hepatic steatosis, inflammation, and fibrosis in MASH mice, whereas its restoration activated MASH pathogenesis. Furthermore, increased death receptor 6 (DR6) was observed in MASH mouse livers. Mechanistically, APP interacted with DR6, a tumor necrosis factor receptor, to facilitate DR6 expression and activation. Activated DR6 increased apoptosis in hepatocytes, which was associated with an increase in proapoptotic effectors (cleaved-caspase 3/7). Our results highlight the role of the APP–DR6 axis in hepatocyte apoptosis, inflammation activation, and fibrosis formation in murine MASH model, providing new insights into therapeutic strategies for MASH.

Metabolic dysfunction–associated steatotic liver disease (MASLD) is currently the most common chronic disease in clinical practice, affecting approximately 32.4% of the global population (1). In 25% of patients, MASLD progresses to metabolic dysfunction–associated steatohepatitis (MASH) (2, 3), and approximately 30% of MASH patients eventually develop liver fibrosis, liver cirrhosis, and even hepatocellular carcinoma (4). Although the "multiple hits" theory can partially explain the pathogenesis of MASH, there is still a lack of specific treatment methods for MASH (5). Hence, there is an urgent need to understand MASH pathogenesis and identify new therapeutic targets.

Hepatocyte apoptosis is a well-defined form of cell death in MASH (6). Apoptotic hepatocytes induce inflammatory cell infiltration and activation of hepatic stellate cells, further promote liver inflammation and fibrosis, and aggravate the progression of MASH (7). Previous literatures have shown that hepatocyte apoptosis is significantly increased in MASH patients and animals and positively correlated with disease severity, which is considered to be one of the pathological hallmarks of MASH (8, 9). Currently, the mechanisms of regulating hepatocyte apoptosis in MASH have not been fully understood, and elucidating this regulatory mechanism will provide important clues for the treatment of MASH.

Amyloid precursor protein (APP) belongs to the highly conserved APP superfamily and is encoded by human chromosome 21Q21.3 genes (10, 11). It is called APP because it can generate amyloid β peptides consisting of 40–42 amino acids after being cleaved by the amyloid pathway. APP is mainly expressed in the central nervous system, and the accumulation of amyloid β peptides generated by APP cleavage in the brain is the main cause of Alzheimer's disease (AD) (12). Although APP is highly expressed in the brain and regulates neuronal metabolism, it is also widely expressed in peripheral tissues, such as adipose tissue, myotubes, and liver (13, 14, 15). Many studies have reported a strong association between APP and metabolic diseases, including diabetes and obesity (13, 16). Zheng et al. (17) found that APP KO mice lost 15 to 20% of their body weight compared with control mice. An et al. (15) found that APP is highly expressed in the adipose tissue of obese mice, and overexpression of APP promoted the occurrence of obesity. Since obesity, diabetes, and MASH are all metabolic diseases, whether APP is involved in the progression of MASH has not been reported.

In this study, we found that hepatic APP expression was increased in murine MASH models. APP KO effectively ameliorated hepatic steatosis, inflammation, fibrosis, and apoptosis, while overexpression of APP aggravated the aforementioned phenotypes. Mechanistically, APP directly interacts with death receptor 6 (DR6) to facilitate DR6 expression and activation. Activated DR6 increased apoptosis in hepatocytes, which was associated with an increase in proapoptotic effectors (cleaved-caspase 3/7), subsequently promotes the activation of downstream proinflammatory and profibrotic-related genes. Our data demonstrated that APP promotes the progression of MASH by promoting DR6-induced hepatocyte apoptosis, which may serve as a potential therapeutic target for the MASH.

Results

APP expression is upregulated in patients and mouse models of MASH

First, we examined APP expression levels in two Gene Expression Omnibus datasets, GSE66676 and GSE48452, and found that the expression level of APP was significantly upregulated in MASH patients (Fig. S1). Then, we constructed three different MASH mouse models and cell models to detect changes in APP expression.

APP protein and mRNA were increased in the liver tissues of Western diet plus sugar water (WDSW)–induced MASH mice than in controls (Figs. 1A and S2A). Consistently, the mRNA and protein levels of APP were significantly increased in the livers of mice after high-fat high-fructose diet (HFHFrD) or methionine and choline deficiency diet treatment compared with controls (Figs. 1, B and C and S2, B, and C). In addition, APP mRNA and protein levels were increased substantially after exposure to palmitic acid (PA) for 24 h in AML12 cells (Figs. 1D and S2D). These results suggest a plausible role of APP in the development of MASH.

Figure 1.

APP is upregulated in MASH and affects lipid metabolism.A, Western blotting of APP in the livers from WDSW-fed and CDNW-fed mice (n = 4/group). B, Western blotting of APP in livers from HFHFrD- or SCD-fed mice (n = 4/group). C, Western blotting of APP in livers from MCD- or SCD-fed mice (n = 4/group). D, Western blotting of APP in AML12 cells treated with PA (n = 3/group). E, Western blot assay verified the knockdown efficiency of APP in AML12 cells (n = 3). F, Oil Red O staining of PA treated AML12 cells after knockdown of APP (n = 6). G, knockdown of APP on TG content of PA-treated AML12 cells (n = 6). H, Western blotting of APP in primary hepatocytes transfected with APP-overexpressing plasmids (n = 3). I, Oil Red O staining of PA-treated primary hepatocytes after APP overexpression (n = 6). J, TG content of PA-treated primary hepatocytes after APP overexpression (n = 4). Data represent the mean ± SEM. Significance was determined by two-tailed Student's t test in A–E, and H and one-way ANOVA in F, G, I, and J. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. APP, amyloid precursor protein; CDNW, chow diet and normal water; HFHFrD, high-fat high-fructose diet; MASH, metabolic dysfunction–associated steatohepatitis; MCD, methionine and choline–deficient diet; PA, palmitic acid; SCD, standard chow diet; TG, triglyceride; WDSW, Western diet and sugar water.

To investigate the effects of APP on lipid metabolism, we constructed siRNAs and overexpressing plasmids of APP (Fig. 1, E and H). We found that the knockdown or overexpression of APP had no significant effect on intracellular lipid deposition in the absence of PA stimulation. The knockdown of APP mitigated lipid accumulation after PA induction in AML12 cells (Fig. 1, F and G). Conversely, overexpression of APP aggravated lipid accumulation after PA induction in primary hepatocytes (Fig. 1, I and J). These data demonstrated that APP promoted fat accumulation in vitro.

APP KO alleviates WDSW-induced glucose metabolism disorder and hepatic steatosis in mice

To explore the role of APP in the development of MASH in vivo, we generated App^−/− mice (Fig. 2A). The APP protein was effectively depleted, as evidenced by immunoblotting (Fig. 2B). App^−/− mice and WT mice were fed with WDSW and chow diet normal water (CDNW) for 24 weeks to induce MASH model (Fig. 2C). App^−/− mice and WT mice fed with CDNW presented similar metabolic profiles (Fig. 2). After 24 weeks of WDSW administration, App^−/− mice exhibited lower body weights, liver weights, and liver weight-to-body weight ratios than WT mice (Fig. 2, D and E). Meanwhile, WDSW-fed App^−/− mice also had milder glucose intolerance and insulin resistance (IR) than WDSW-fed WT mice, as assessed by the glucose tolerance test and insulin tolerance test (Fig. 2, F and G). More importantly, App^−/− mice were characterized by decreased hepatic lipogenesis and ballooning compared with WT mice, as shown by H&E staining, Oil Red O staining, and triglyceride (TG) and total cholesterol (TC) assays (Fig. 2, H and I).

Figure 2.

APP deficiency ameliorates WDSW diet–induced glucose metabolic disorders and hepatic steatosis.A, schematic diagram of App gene editing using CRISPR–Cas9 technology. B, relative protein levels of APP in the livers of mice of WT and App^−/− mice. C, experimental design. WT and App^−/− mice were fed a CDNW or WDSW for 24 weeks, starting at 8 weeks of age. The GTT was performed at week 22, and the ITT was performed at week 23. Twenty-four weeks later, the mice were harvested for analysis (WT CDNW, n = 6; App^−/− CDNW, n = 5; WT WDSW, n = 8; App^−/− WDSW, n = 8). D, body weight. E, liver weight and the ratio of liver weight to body weight. F, GTT and corresponding AUC of blood glucose level. G, ITT and corresponding AUC of blood glucose level. H, representative H&E staining and Oil Red O staining of the indicated groups and statistical analysis of Oil Red O staining. Scale bar represents 50 μm. I, liver TG and TC contents. J–K, representative Western blot image showing the expression levels of lipid synthesis and quantification of the indicated protein level normalized to that of GAPDH. Data represent the mean ± SEM. Statistical analysis was analyzed by one-way ANOVA in D–I and two-tailed Student's t test in K. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. APP, amyloid precursor protein; App^−/−, APP gene KO mice; AUC, area under curve; CDNW, chow diet and normal water; GTT, glucose tolerance test; ITT, insulin tolerance test; TC, total cholesterol; TG, triglyceride; WDSW, Western diet and sugar water.

Next, we detected the expression of key proteins involved in lipid metabolism and found that the expression of lipogenesis-related proteins, such as peroxisome proliferator–activated receptor gamma, sterol regulatory element-binding protein 1, fatty acid synthase, and acetyl-CoA carboxylase, was significantly lower in WDSW-fed App^−/− mice than in WT mice (Fig. 2, J and K), whereas there was no significant difference in lipid oxidation–related proteins between the two groups (e.g., peroxisome proliferator–activated receptor alpha, CPT1-α, and ACOX1) (Fig. S3A). These data showed that KO of APP inhibited WDSW-induced glucose metabolism disorder and hepatic steatosis in mice.

APP overexpression aggravates WDSW-induced glucose metabolic disorders and hepatic steatosis in mice

To identify the potential role of APP in the pathogenesis of MASH, C57BL/6 mice were given with adeno-associated virus (AAV)-negative control (NC) or AAV-APP through tail vein injection and then fed with WDSW or CDNW for 24 weeks (Fig. 3A). Successful overexpression of APP in the liver of C57BL/6J mice after 24 weeks of WDSW administration is shown in Fig. S4, A–C. No obvious change was observed between AAV-APP mice and AAV-NC mice fed a CDNW (Fig. 3). AAV-APP mice exhibited heavier body weights than AAV-NC mice after 24 weeks of WDSW feeding (Fig. 3B). Furthermore, overexpression of APP consistently aggravated the WDSW-induced increases in liver weights and liver weight-to-body weight ratios compared with these parameters in AAV-NC mice (Fig. 3C). These results indicated an increase in hepatic lipid accumulation in APP-overexpressing mice.

Figure 3.

APP overexpression exacerbates WDSW-induced glucose metabolic disorders and hepatic steatosis.A, experimental design. C57BL/6 mice were administered AAV-NC or AAV-APP through tail vein injection at 6 weeks of age. Two weeks later, mice were fed a CDNW or WDSW diet for 24 weeks. The GTT was performed at week 22, and the ITT was performed at week 23. Twenty-four weeks later, the mice were harvested for analysis (AAV-NC CDNW, n = 5; AAV-APP CDNW, n = 6; AAV-NC WDSW, n = 8; AAV-APP WDSW, n = 8). B, body weight. C, liver weights (left) and liver/body weight ratio (right). D, GTT and corresponding AUC of blood glucose levels in the four groups. E, ITT and corresponding AUC of blood glucose levels in the four groups. F, representative H&E staining and Oil Red O staining of the indicated groups and statistical analysis of Oil Red O staining. Scale bar represents 50 μm. G, Liver TG and TC contents. H–I, representative Western blot image showing the expression levels of lipid synthesis and quantification of the indicated protein level normalized to that of GAPDH. Data represent the mean ± SEM. Statistical analysis was analyzed by one-way ANOVA in B–G and two-tailed Student's t test in I. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. APP, amyloid precursor protein; AUC, area under curve; CDNW, chow diet and normal water; GTT, glucose tolerance test; ITT, insulin tolerance test; TC, total cholesterol; TG, triglyceride; WDSW, Western diet and sugar water.

Next, we evaluated the effect of hepatic APP on glucose metabolism. We found that WDSW-fed AAV-APP mice developed more serious glucose intolerance and IR than WDSW-fed AAV-NC mice, as assessed by glucose tolerance test and insulin tolerance test (Fig. 3, D and E). These parameters indicated that hepatic APP overexpression increased the susceptibility to IR under metabolic challenge.

Consistent with the aforementioned findings, WDSW-fed AAV-APP mice also exhibited remarkably higher intrahepatic lipid accumulation, as indicated by H&E and Oil Red O staining (Fig. 3F), as well as higher intrahepatic TG and TC contents (Fig. 3G). Accordingly, the protein levels of lipid synthesis were increased, whereas those of proteins related to fatty acid β-oxidation showed no significant change between the two groups (Figs. 3, H and I and S3, B, H and I). These data showed that hepatocyte-specific overexpression of APP significantly aggravated WDSW-induced glucose metabolic disorders and hepatic steatosis, further indicating the detrimental role of APP during WDSW-induced MASH progression.

APP KO attenuates WDSW-induced hepatic inflammation and fibrosis in mice

Considering that the pathogenesis of MASH is a complex process that includes metabolic disorders and uncontrolled chronic inflammation and fibrosis (18), we next investigated whether APP affects the liver inflammatory response and fibrosis. We found that serum alanine aminotransferase and aspartate aminotransferase levels of App^−/− mice were significantly lower than those of WT mice after WDSW challenge, indicating that knocking out APP could ameliorate liver injury in MASH mice (Fig. 4A). Sirius Red staining and immunohistochemical staining of F4/80 revealed markedly decreased collagen fiber deposition and inflammatory cell infiltration in the livers of WDSW-fed App^−/− mice compared with control mice (Fig. 4B). KO of APP also significantly dampened WDSW-induced activation of NF-κB signaling in the livers (Fig. 4C). Furthermore, the mRNA levels of inflammatory cytokines (tumor necrosis factor α, C-C chemokine 3 [Ccl3], C-C chemokine 5 [Ccl5], C-C chemokine 7 [Ccl7], and IL-6) were consistently decreased in App^−/− mice compared with the corresponding controls (Fig. 4D). Western blot result demonstrated that App^−/− mice fed a WDSW diet showed reduced alpha-smooth muscle actin protein expression levels compared with WT mice (Fig. 4E). Consistently, the expression levels of MASH-related profibrotic genes (collagen type I alpha 1 chain [Col1α1], collagen type III alpha 1 chain [Col3α1], actin alpha 2 [Acta2], tissue inhibitor of metalloprotease 1 [Timp1], and connective tissue growth factor [Ctgf]) were also significantly decreased in the App^−/− mouse (Fig. 4F). Recent literature has reported that Foxs1 serve as a biomarker for liver fibrosis (19). We also examined Foxs1 expression in liver samples from App^−/− mice and WT controls and found that KO of APP significantly reduced the mRNA levels of Foxs1 (Fig. 4F). These results showed that APP KO attenuates WDSW-induced hepatic inflammation and fibrosis in mice.

Figure 4.

Deficiency of APP alleviates WDSW-induced hepatic inflammation and fibrosis.A, serum ALT (left) and AST levels (right). B, representative images and quantitative analysis of Sirius Red staining and F4/80 staining of mouse liver tissues from the indicated groups. Scale bar represents 50 μm. C, representative Western blot image showing the expression levels of PP65 and P65. D, relative mRNA levels of genes related to inflammation normalized to Gapdh. E, representative Western blot image showing the expression levels of α-SMA. F, relative mRNA levels of genes related to fibrosis normalized to Gapdh. Data represent the mean ± SEM. Statistical analysis was performed using one-way ANOVA in A and B, and Student's t test in C–F. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. α-SMA, α-smooth muscle actin; ALT, alanine aminotransferase; APP, amyloid precursor protein; App^−/−, APP gene KO mice; AST, aspartate aminotransferase; CDNW, chow diet and normal water; PP65, phosphorylation of p65; WDSW, Western diet and sugar water.

Overexpression of APP exacerbates WDSW-induced hepatic inflammation and fibrosis in mice

APP KO attenuates WDSW-induced hepatic inflammation and fibrosis, prompting us to address whether overexpression of APP has the opposite effects. We found that serum alanine aminotransferase and aspartate aminotransferase of AAV-APP mice were significantly higher than those in the AAV-NC group, indicating that overexpression of hepatic APP aggravates liver injury in MASH mice (Fig. 5A). Compared with AAV-NC mice, AAV-APP mice exhibited more hepatic fiber deposition and higher intrahepatic infiltration of F4/80-positive macrophages (Fig. 5B). The consistently upregulated phosphorylation of p65 in AAV-APP mice compared with NC mice indicated the activation of NF-κB–mediated inflammatory signaling in response to WDSW feeding (Fig. 5C).

Figure 5.

Overexpression of APP aggravates hepatic inflammation and fibrosis in the WDSW-induced MASH model.A, serum ALT (left) and AST levels (right). B, representative images and quantitative analysis of Sirius Red staining and F4/80 staining of mouse liver tissues from the indicated groups. Scale bar represents 50 μm. C, representative Western blot image showing the expression levels of PP65 and P65. D, relative mRNA levels of genes related to inflammation normalized to Gapdh. E, representative Western blot image showing the expression levels of α-SMA. F, relative mRNA levels of genes related to fibrosis normalized to Gapdh. Data represent the mean ± SEM. Statistical analysis was performed using one-way ANOVA in A and B and student's t test in C–F. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. APP, amyloid precursor protein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CDNW, chow diet and normal water; MASH, metabolic dysfunction–associated steatohepatitis; PP65, phosphorylation of p65; α-SMA, α-smooth muscle actin; WDSW, Western diet and sugar water.

In addition, the expression of proinflammatory genes, such as C-C chemokine 2 (Ccl2), Ccl3, Ccl5, Ccl7, and C-X-C motif chemokine ligand 2 (Cxcl2), was also significantly increased in the livers of AAV-APP mice (Fig. 5D). AAV-APP mice also exhibited more pronounced hepatic fibrosis than did AAV-NC mice after WDSW feeding, as evidenced by alpha-smooth muscle actin protein expression and profibrotic gene expression (e.g., Col1α1, Col3α1, Acta2, Timp1, Foxs1) (Fig. 5, E and F). Collectively, these results indicate that in contrast to the improvement of hepatic inflammation and fibrosis in App^−/− mice, hepatocyte-specific overexpression of APP caused the opposite effects. The aforementioned data indicated that APP promotes MASH progression by worsening hepatic steatosis, inflammation, and fibrosis during metabolic challenges in vivo.

APP promotes HFHFrD-induced hepatic steatosis and inflammation in mice

We further verified the important role of APP in the HFHFrD-induced MASH model (Fig. S5A). We found that App^−/− mice exhibited lighter body weights, liver weights, and liver weight-to-body weight ratios than WT mice after HFHFrD feeding (Fig. S5, B and C). HFHFrD-fed App^−/− mice also showed alleviated glucose intolerance and IR (Fig. S5, D and E). Meanwhile, lipid accumulation was alleviated by knocking out APP, as shown in H&E and Oil Red O, TG and TC assays (Fig. S5F). Consistently, the expression of sterol regulatory element-binding protein 1, which stimulates fatty acid synthesis, was decreased, whereas the levels of proteins that promote fatty acid oxidation were not significantly altered in App^−/− mice compared with WT mice after HFHFrD feeding (Fig. S5G).

The HFHFrD-induced mouse model is often accompanied by liver inflammation, which plays an important role in the development of MASH (20, 21). In response to 12 weeks of HFHFrD feeding, App^−/− mice also displayed mild inflammation in the liver, as evidenced by F4/80-positive inflammatory cell staining, phosphorylation of p65 protein expression, and proinflammatory gene expression in the liver (Fig. S5, F, H, and I). Furthermore, we found that AAV8-mediated hepatocyte-specific overexpression of APP aggravated HFHFrD-induced metabolic disorders, hepatic steatosis, and inflammation in MASH mice (Fig. S6, A–J). These results confirmed that APP could promote MASH progression in HFHFrD-induced MASH mice.

APP directly interacts with DR6 to regulate MASH

Previous studies revealed that APP could bind with DR6 to regulate neural inflammatory effects and trigger neuron death in AD (22). The interaction between APP and DR6 also promotes the activation of death receptor (23). Activation of DR6 can induce apoptosis, and the occurrence of apoptosis is closely related to MASH (8, 24, 25). Herein, we explored whether APP promotes MASH progression through the DR6-induced apoptosis. First, based on bioinformatics analysis, the STRING database (https://string-db.org/) predicted that DR6 interacts with APP and Caspase3 (Fig. 6A). Then, coimmunoprecipitation results indicated the interaction between APP and DR6 in human embryonic kidney 293T cells and AML12 cells (Fig. 6, B and C). Furthermore, immunofluorescence staining indicated a considerable degree of colocalization of APP and DR6 in primary hepatocytes (Fig. 6D). We also observed that the hepatic expression of DR6 was dramatically upregulated at both the mRNA and protein levels in MASH model mice (Fig. 6, E and F). In the livers of WDSW-induced MASH models, we found that KO of APP further downregulated the protein expression of DR6, cleaved caspase 3, and cleaved caspase 7 (Fig. 6G). In contrast, overexpression of APP significantly upregulated DR6 protein expression in MASH mouse livers (Fig. 6H). These findings showed that APP directly interacts with and regulates DR6 in MASH livers.

Figure 6.

APP directly interacts with DR6 to regulate MASH.A, the STRING database predicted that APP interacts with DR6 and Caspase3. B and C, IP assays were performed in HEK293T cells (B) and AML12 cells (C). D, representative immunofluorescence images of APP and DR6 in murine primary hepatocytes (630×). Na⁺/K⁺ ATPase is a cell membrane marker. E, Western blotting of DR6 in the livers from WDSW-fed and CDNW-fed mice (n = 4/group), HFHFrD- or SCD-fed mice (n = 4/group). F, relative mRNA levels of Dr6 in liver tissues after 24 weeks of WDSW (n = 10) or CDNW (n = 10); 12 weeks of HFHFrD (n = 5) or SCD (n = 9). G and H, the protein levels of DR6, active caspase 3, and caspase 7 were detected in the livers of WDSW-treated WT or App^−/− mice (G) and WDSW-treated AAV-NC or AAV-APP mice (H). Data represent the mean ± SEM. Statistical analysis was performed using Student's t test. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001. APP, amyloid precursor protein; CDNW, chow diet and normal water; DR6, death receptor 6; HFHFrD, high-fat high-fructose diet; MASH, metabolic dysfunction–associated steatohepatitis; SCD, standard chow diet; WDSW, Western diet and sugar water.

DR6 is indispensable for APP-mediated hepatocyte lipid accumulation, apoptosis, and inflammation

To further explore whether DR6 has a profound effect on APP-regulated hepatic lipid accumulation, apoptosis, and inflammation upon metabolic challenge, we conducted the following experiments in mouse AML12 cells. First, we treated AML12 cells with different concentrations of PA and found that the protein expression levels of APP and DR6 were significantly increased in a concentration-dependent manner (Fig. 7A). Subsequently, we knocked down APP in AML12 cells and found that it could significantly reduce PA (200 μM)-induced lipid accumulation and hepatic apoptosis, which was aggravated when DR6 was overexpressed (Fig. 7B). In addition, several key proteins related to lipid synthesis and apoptosis were found to exhibit altered expression. In APP-knockdown AML12 cells, the expression of acetyl-CoA carboxylase and cleaved caspase-3 proteins was significantly downregulated, while the changes caused by knockdown of APP were enhanced by overexpression of DR6 (Fig. 7C). In addition, quantitative PCR results showed that knockdown of APP significantly alleviated PA (200 μM)-induced inflammation, whereas overexpression of DR6 aggravated the aforementioned phenomenon (Fig. 7D). These findings suggested that DR6 is indispensable for APP-mediated hepatocyte lipid accumulation, inflammation, and apoptosis.

Figure 7.

Overexpression of DR6 exacerbates APP deletion ameliorated MASH in vitro.A, relative protein levels of APP and DR6 in PA-induced AML12 cells. B, representative images and quantitative analysis of Oil Red O staining and TUNEL staining after knockdown of APP and overexpression of DR6 in PA (200 μM)-treated AML12 cells. Scale bar represents 50 μm. C, Western blot analysis of the protein expression of APP, DR6, ACC, Cle-caspase3, and Caspase3 in PA (200 μM)-treated AML12 cells followed by knockdown of APP and further overexpression of DR6. D, inflammatory genes in PA (200 μM)-treated AML12 cells after knockdown of APP and overexpression of DR6. Data represent the mean ± SEM. Statistical analysis was performed using one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. ACC, acetyl-CoA carboxylase; APP, amyloid precursor protein; DR6, death receptor 6; MASH, metabolic dysfunction–associated steatohepatitis; PA, palmitic acid.

Discussion

MASH is widely accepted as a progressive liver disease that can lead to cirrhosis and hepatocellular carcinoma. In this study, we investigated whether APP is involved in disease progression and whether it could be an additional therapeutic target in MASH. We found that APP expression was significantly increased in the livers of mice with diet-induced MASH. KO of APP alleviated diet-induced liver steatosis, inflammation, fibrosis, and apoptosis, whereas overexpression of APP had the opposite effects. Mechanistically, APP directly interacts with DR6 to promote hepatocyte apoptosis and subsequently promotes the activation of downstream proinflammatory and profibrotic genes. We integrated each of these findings in a diagram that depicts an innovative regulatory mechanism of APP in MASH development (Fig. 8).

Figure 8.

Summary of the pathological mechanisms. APP promotes the progression of MASH by interacting with DR6 and inducing apoptosis. APP, amyloid precursor protein; DR6, death receptor 6; MASH, metabolic dysfunction–associated steatohepatitis.

Although APP is mainly expressed in the central nervous system, a large number of studies have shown that APP is also widely expressed in peripheral tissues and participates in the metabolism of peripheral tissues (11). As early as 2008, Lee et al. (13) found that the expression of APP was significantly upregulated in the adipocytes of obese patients. Later studies found that APP deficiency can resist the occurrence of high-fat diet–induced obesity (16). In 2019, An et al. (15) found that APP is highly expressed in the adipose tissue of obese mice, and overexpression of APP can promote the occurrence of obesity. At the same time, they found that adipo-APP transgenic mice developed hepatic steatosis, with elevated hepatic TG levels (15). These results all indirectly indicate that APP may be involved in the development of MASLD. However, there is still a lack of direct evidence to support this conclusion. In our study, we demonstrated for the first time that APP expression was upregulated in the livers of MASH mice and played an important role in regulating MASH.

DR6, also known as TNFRSF21, is a member of the tumor necrosis factor receptor superfamily (TNFRSF) (26, 27). Abnormal expression and activation of DR6 contributes to various diseases, including AD (22), immune diseases (28, 29), and tumors (30, 31), but its involvement in the development of MASH has not been reported. Currently, the only identified ligand for DR6 is APP (23, 32). These results initially suggested that the interaction between APP and DR6 requires the production of extracellular segments through the amyloid cleavage pathway. However, subsequent works further demonstrated that APP could still interact with DR6 without the cleavage of β-site amyloid precursor protein cleaving enzyme-1 (BACE1) and then activate DR6 (23, 33, 34). In this study, we found that the expression levels of APP and DR6 were significantly increased (Figs. 1 and 6, E and F), whereas BACE1 expression was significantly decreased in different MASH mouse models (Fig. S7A). In addition, in PA-treated AML12 cells, we also found that APP and DR6 were upregulated (Fig. 7A), whereas BACE1 was downregulated (Fig. S7B). Through coimmunoprecipitation and immunolocalization, we found a significant interaction between APP and DR6. These results suggest that the interaction between APP and DR6 is BACE1 independent in MASH progression.

As mentioned earlier, the interaction between APP and DR6 can trigger apoptosis of neuronal cell bodies (24). Based on the aforementioned clues, we searched for the relationship between APP, DR6, and Caspase3, the key executive molecule of apoptosis, in STRING database, and the results suggested that DR6 interacts with APP and Caspase3. Caspase 3 activation and hepatocyte apoptosis are also prominent features of MASH animal models, as well as human MASH, and have been shown to correlate with disease severity (8, 25, 35). In this study, we found that KO of APP significantly reduced the expression of DR6 and the apoptosis executors, caspase 3 and 7, in MASH mice. However, the expression of DR6 and cleaved caspase 3 and 7 increased significantly after APP overexpression. Based on this, we hypothesized that APP interacts with DR6 to promote apoptosis in the progression of MASH.

Although we have defined the role and mechanism of APP in the progression of MASH, there are still some limitations in our study. First, we used systemic APP KO mice in this study, and it is inevitable to avoid the potential influence of APP in other organs. However, we found that, compared with that in the control group, APP overexpression in the liver significantly aggravated the MASH phenotype. The effects of APP on hepatic steatosis were replicated in primary hepatocytes and AML12 cells. These results indicate that APP in hepatocytes is necessary for the MASH process and that targeting APP may provide important clues for MASH prevention. Future studies using liver-specific APP KO mice could address this issue directly. Similar to the results of Zheng et al. we found that the food intake of App^−/− mice was lower than that of WT mice (Fig. S8A). Admittedly, this may be part of the reason why App^−/− mice have milder hepatic steatosis. However, whether APP KO reduces MASH by inhibiting appetite or hepatocyte apoptosis still needs to be concluded by comparing liver steatosis in liver-specific KO mice with that in brain-specific KO mice. In this study, we found that liver-specific overexpression of APP could aggravate MASH, and there was no significant difference in food intake between AAV-NC and AAV-APP groups (Fig. S8B). In addition, in AML12 cells, we found that knockdown of APP reduced hepatic lipidosis and hepatocyte apoptosis. Based on the aforementioned results, we can conclude that hepatocyte APP promotes MASH by interacting with DR6 to induce hepatocyte apoptosis. Second, in this study, although we used a variety of different MASH models to demonstrate the role of APP, clinical data were lacking to analyze APP expression in liver samples from MASH patients and to further analyze the correlation between APP expression and disease severity.

In summary, findings from our current study have provided new insights that APP exerts an accelerated role in diet-induced MASH. Mechanically, APP directly binds with and activates DR6. Activated DR6 increased apoptosis in hepatocytes, which was associated with an increase in proapoptotic effectors (cleaved-caspase 3/7) (Fig. 8). To this end, specific targeting of APP, the APP–DR6 axis, should be considered as a promising strategy in the therapeutics of hepatic apoptosis and MASH progression.

Experimental procedures

Generation of genetically modified mice

APP gene KO mice (App^−/−) were generated by GemPharmatech Co, Ltd using the CRISPR/CRISPR-associated protein 9 (Cas9)–based genome editing system. The schematic diagram is shown in Figure 2A. Detailed explanations are provided in the Supporting Information.

Animals and treatments

All animal experiments were approved by the Animal Care and Use Committee of the First Affiliated Hospital, Zhejiang University School of Medicine (reference number: 2022-8). Male C57BL/6J (6–8 weeks old) mice were purchased from Zhejiang Academy of Medical Sciences. Detailed explanations of all animal treatments are provided in the Supporting Information.

Statistical analyses

All data were analyzed using the two-tailed Student's t test (two-group comparisons) or one-way ANOVA with post hoc t tests (multiple-group comparisons). The data are presented as the mean ± SEM. p < 0.05 (2-tailed) was considered significant. GraphPad Prism, version 9.0 (GraphPad Software, Inc) was used for statistical analysis.

Data availability

Data are available on request from the corresponding author.

Supporting information

This article contains supporting information.

Conflict of interest

The authors declare that they have no conflicts of interest with the contents of this article.

Reviewed by members of the JBC Editorial Board. Edited by Ursula Jakob