Abstract Background & Aims Autophagy plays an important role in liver regeneration. However, most studies are limited to hepatocytes, and the function and mechanism of macrophage autophagy in liver regeneration remain unclear. This study investigated the role of the essential autophagy gene encoding autophagy-related 16-like 1 (ATG16L1), which regulates the macrophage phenotype in liver regeneration. Methods We generated FloxP-Atg16l1 (Atg16l1^FL/FL), Lyz2-Cre Atg16l1 knockout (KO) (Atg16l1^M-KO), and myeloid-specific Atg16l1-overexpression-knock-in (Atg16l1^OE) mice. These mice were subjected to 70% partial hepatectomy to demonstrate the role of ATG16L1 in macrophages during liver regeneration. Results ATG16L1 expression was significantly upregulated in macrophages during the early stages of liver regeneration. ATG16L1 deletion in macrophages substantially delayed liver regeneration in mice and caused a marked imbalance in Ly6C^hi and Ly6C^lo macrophage proportions in the liver. RNA-sequencing analysis revealed that, compared with macrophages isolated from Atg16l1^FL/FL mice, those from Atg16l1^M-KO mice exhibited significant downregulation of genes associated with oxidative phosphorylation and upregulation of proinflammatory gene expression. Mechanistically, ATG16L1 loss impaired mitophagy in macrophages, leading to the accumulation of mitochondrial damage and a metabolic shift that promoted proinflammatory macrophage polarization. ATG16L1 deficiency not only promoted macrophage mitochondrial (mt)DNA release and cyclic GMP-AMP synthase–stimulator of interferon genes (STING) activation, but also suppressed STING degradation. Sustained STING hyperactivation and subsequent increased release of downstream interferons further contributed to the inhibition of liver regeneration. Notably, pharmacological activation or genetic overexpression of ATG16L1 significantly enhanced liver regeneration in mice. Conclusions ATG16L1 has a pivotal role in liver regeneration by affecting the phenotype and function of macrophages. Thus, targeting ATG16L1 in macrophages could present a novel strategy for promoting liver regeneration. Impact and implications The autophagy-related gene ATG16L1 mediates mitophagy, facilitating the clearance of damaged mitochondria in macrophages following partial hepatectomy and maintaining a reparative macrophage phenotype. ATG16L1 deficiency triggers excessive STING activation and inhibits its degradation, thereby suppressing liver regeneration. Thus, targeting ATG16L1 in macrophages could represent a novel strategy to promote liver regeneration. Keywords: Proinflammatory, Mitophagy, mtDNA, STING, IFN, Oxidative phosphorylation Graphical abstract Image 1 [39]Open in a new tab Highlights * • ATG16L1 expression is increased after partial hepatectomy. * • Macrophage-specific ATG16L1 deletion impairs liver regeneration. * • ATG16L1 deletion remodels macrophage phenotypes by inhibiting mitophagy. * • Macrophage ATG16L1 deletion inhibits liver regeneration by promoting and sustaining cGAS–STING pathway activation. * • Targeted activation of ATG16L1 promotes liver regeneration. Introduction Liver regeneration can occur following various types of injury caused by drugs, toxins, and surgeries, but the extent of that regeneration directly affects patient prognosis.[40]^1 The strong regenerative ability of the liver is crucial for ensuring functional compensation and recovery in patients following liver resection.[41]^2 However, clinical cases of liver failure resulting from insufficient compensatory capacity of the remaining liver following major liver resection, such as hemihepatectomy or living-donor liver transplantation, are not uncommon.[42]^3 Factors such as hepatitis virus infection and liver cirrhosis can impair liver reserve function and its regenerative and reparative abilities. Therefore, in-depth research into the specific cellular and molecular mechanisms regulating liver regeneration and repair is essential for enhancing the regenerative capacity of liver and improving patient outcomes. Liver regeneration is an organized dynamic process primarily involving nonparenchymal cells in addition to the proliferation of liver parenchymal cells. Macrophages are not only crucial for maintaining liver homeostasis, but also have a role in liver pathology; thus, their role in liver regeneration and repair has gained increasing attention.[43]^4 For example, recent studies showed that macrophages are heterogeneous in both origin and function. Infiltrating macrophages involved in regulating regeneration and repair are often classified into two categories based on their Ly6C expression levels: classically activated proinflammatory macrophages (Ly6C^hi) and alternatively activated reparative macrophages (Ly6C^lo). Both types can have dual roles in either inhibiting or promoting liver regeneration and repair. Therefore, ensuring a coordinated balance between these two macrophage phenotypes is vital.[44]^5 Furthermore, Ly6C^hi macrophages can transition to Ly6C^lo macrophages during regeneration;[45]^6^,[46]^7 however, the specific mechanisms involved remain unclear. Liver regeneration requires sufficient energy and cellular materials for DNA replication and cell division. As a recycling mechanism that maintains homeostasis in eukaryotic cells, autophagy has an important role in providing energy and substrates during regeneration. A positive synergistic effect has been demonstrated between autophagy and hepatocyte proliferation.[47]^8 However, whether autophagy activation occurs in nonparenchymal cells, particularly macrophages, during regeneration and whether these cells rely on autophagic changes to perform their respective functions remain unclear. Autophagy related 16-like 1 (ATG16L1) is a key mediator in the lipidation of LC3 and the formation of autophagosomes during autophagy.[48]^9 It not only coordinates intestinal epithelial cell homeostasis, but also has an active role in lipid metabolism in hepatocytes. Previously, we reported that ATG16L1 deletion in macrophages exacerbated liver lipid accumulation and inflammation, ultimately promoting the progression of metabolic dysfunction-associated steatohepatitis.[49]^10 However, the specific role of macrophage ATG16L1 in liver regeneration following partial hepatectomy (PHx) remains unclear. In this study, we elucidated the molecular mechanisms by which ATG16L1 signaling regulates macrophage activation, hepatocyte proliferation, and intrahepatic inflammation during liver regenerative repair. We focused on the intrinsic links among autophagy, the reshaping of reparative macrophage phenotypes, and hepatocyte proliferation, thereby providing a theoretical foundation and potential therapeutic targets for the orderly regulation of liver regeneration and repair in clinical settings. Materials and methods Mice The wild-type (WT), FloxP-Atg16l1 (Atg16l1^FL/FL), Lyz2-Cre Atg16l1 knockout (KO) (Atg16l1^M-KO), and myeloid-specific Atg16l1-overexpression-knock-in (Atg16l1^OE) mice used in this study were obtained from GemPharmatech Co (Nanjing, Jiangsu, China)., Ltd. The FloxP-Tmem173 (Tmem173^FL/FL) and Lyz2-Cre Tmem173 KO (Tmem173^M-KO) mice were provided by Shanghai Model Organisms Center, Inc. (Shanghai, China) All mouse strains had a C57BL/6J genetic background, were housed under specific pathogen-free conditions, were maintained on a 12-h light/dark cycle, and had free access to water and standard chow. All animals were treated humanely, and all procedures were approved by the Institutional Animal Care and Use Committee (IACUC:2304026) of Nanjing Medical University. Methods All methods and materials used in the study are detailed in the supplementary data online. Data analysis All data are presented as the mean ± SEM of at least three biological replicates per group. Two-tailed student's t test was applied for comparisons of two groups and one-way analysis of variance (ANOVA) for comparisons of more than two groups. p <0.05 was considered statistically significant. Results Significant upregulation of ATG16L1 expression after PHx To investigate ATG16L1 expression during liver regeneration, we established a mouse model of 70% PHx. After PHx, we measured the expression of ATG16L1 at different time points. Notably, ATG16L1 protein and mRNA levels in the mouse liver peaked at 48 h post PHx before declining, eventually returning to baseline levels at 7 days post PHx ([50]Fig. 1A,B). To further confirm ATG16L1 expression in hepatic macrophages in the regenerative environment, we used western blot and quantitative (q)PCR. ATG16L1 expression in macrophages was significantly elevated after PHx, showing a synchronous pattern with the overall hepatic expression of ATG16L1 ([51]Fig. 1C,D). Dual immunofluorescence staining further confirmed that ATG16L1 expression was upregulated in both hepatocytes and macrophages following PHx ([52]Fig. 1E). In addition, the expression of autophagy-related genes, including ATG5, ATG12, and Beclin, was elevated after PHx ([53]Fig. S1). These results indicate that ATG16L1 expression is upregulated in liver tissue after PHx, with a synchronized expression trend between hepatocytes and hepatic macrophages. Peak ATG16L1 expression coincides with the peak period of hepatocyte proliferation.[54]^11 Moreover, ATG5, ATG12, and Beclin are also upregulated in macrophages. Fig. 1. [55]Fig. 1 [56]Open in a new tab ATG16L1 levels are significantly elevated in the mouse model after PHx. ATG16L1 protein (A) and mRNA (B) levels in mouse liver tissues at indicated time points post PHx (n = 4–6 per group). (C) Western blot analysis of ATG16L1 expression of liver macrophages at indicated time points post PHx (n = 4 per group). β-actin was used as a control. ∗∗∗p <0.001, by one-way ANOVA with Bonferroni post-test. (D) Relative expression levels of Atg16l1 mRNA in liver tissue, primary hepatocytes, and liver macrophages. ∗∗∗p <0.001, by an unpaired, 2-tailed Student’s t test. (E) Representative immunofluorescence co-staining images of ATG16L1 (red), F4/80 (green), and DAPI (blue) in mice and their quantification 36 h after PHx and in the sham group. Scale bar: 50 μm. ATG16L1, autophagy-related 1- like 1; PHx, partial hepatectomy. Macrophage-specific ATG16L1 deletion impairs liver regeneration To further elucidate the potential role of macrophage-specific ATG16L1 during liver regeneration, we generated Lyz2-Cre Atg16l1-KO mice and performed hepatectomy on Atg16l1^M-KO and Atg16l1^FL/FL mice. Before PHx, there were no differences between the two groups of mice in terms of average body weight, liver weight, or liver lipid accumulation ([57]Figs. S2A and B). However, compared with Atg16l1^FL/FL mice, Atg16l1^M-KO mice subjected to 90% PHx exhibited reduced tolerance to liver resection surgery and a significantly increased mortality rate ([58]Fig. 2A). In the 70% PHx model, the liver-to-body weight ratio was lower in Atg16l1^M-KO mice than in Atg16l1^FL/FL mice at 48 h post PHx. However, no significant difference was observed between the two groups at 7 days post PHx ([59]Fig. 2B), suggesting that the absence of ATG16L1 in macrophages leads to delayed recovery of liver mass. Furthermore, Atg16l1^M-KO mice exhibited a significantly delayed decrease in liver enzyme levels, with serum alanine aminotransferase and aspartate aminotransferase levels remaining significantly higher than those in Atg16l1^FL/FL mice between 12 and 72 h ([60]Fig. 2C). We evaluated tissue injury at different time points using H&E staining and found no differences in injury between Atg16l1^FL/FL and Atg16l1^M-KO mice. Overall injury levels were also low in this model. This could be closely associated with delayed liver regeneration in these mice ([61]Fig. 2D). Fig. 2. [62]Fig. 2 [63]Fig. 2 [64]Open in a new tab Macrophage-specific ATG16L1 deletion inhibits hepatocyte proliferation during liver regeneration. (A) Survival curve of Atg16l1^FL/FL and Atg16l1^M-KO mice in response to lethal 90% PHx (n = 14 per group). (B) Liver-to-body weight ratio of Atg16l1^FL/FL and Atg16l1^M-KO mice at indicated time points after 70% PHx (n = 6 per group). (C) Serum ALT and AST levels assayed by ELISA at indicated time points post 70% PHx of Atg16l1^FL/FL and Atg16l1^M-KO mice. (D) Representative H&E staining and quantification of necrotic zones at indicated time points after PHx in Atg16l1^FL/FL and Atg16l1^M-KO mice (n = 6 per group). (E) Immunohistochemical staining of liver tissue for Ki67 and PCNA and their quantification at indicated time post 70% PHx in Atg16l1^FL/FL and Atg16l1^M-KO mice (n = 6 per group). (F) Western blot analysis and corresponding quantification of PCNA protein expression levels in mice liver tissue at different time points post PHx (n = 4 per group). β-actin was used as a control. (G) RNA-sequencing analysis of liver tissues at 36 h post PHx in mice. Bar chart showing the top-30 significantly expressed genes based on GO enrichment. The cell cycle signaling pathways ranked top for both their p values and count numbers. (H) Quantification of CyclinA2, CyclinB1, CyclinD1, and CyclinE1 mRNA expression in Atg16l1^FL/FL and Atg16l1^M-KO mice at different time points post PHx (n = 6 per group). (I) Western blot analysis and corresponding quantification of Cyclin A and Cyclin D1 expression levels in mice liver tissue at different time points post PHx (n = 4 per group). β-actin was used as a control.∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, by two-way ANOVA with Bonferroni post-test. Scale bars: 50 μm (D), 100 μm (E). ALT, alanine aminotransferase; AST, aspartate aminotransferase; ATG16L1, autophagy-related 16-like 1; GO, Gene Ontology; KO, knockout; PCNA, proliferating cell nuclear antigen; PHx, partial hepatectomy. According to immunohistochemistry staining for Ki67 and proliferating cell nuclear antigen, as well as western blot results, the overall proliferation level in Atg16l1^M-KO mice was significantly lower than that in Atg16l1^FL/FL mice at 36–72 h post surgery ([65]Fig. 2E,F). Whole-liver sequencing further confirmed this finding because pathway enrichment analysis and qPCR results showed downregulation of the liver cell cycle pathway and significant reductions in the expression of the cell cycle markers Cyclin A2, Cyclin B1, Cyclin D1, and Cyclin E1, indicating a weakened G1–S phase transition in Atg16l1^M-KO mice. The western blot results showed similar changes ([66]Fig. 2G–I). We supplemented these findings with an acute CCl4-induced liver injury model. At 36–72 h after CCl4 injection, both Atg16l1^M-KO and Atg16l1^FL/FL mice exhibited significant liver damage, with the damage being more severe in Atg16l1^M-KO mice. Although liver injury in Atg16l1^FL/FL mice was fully repaired 120 h after CCl4 injection, Atg16l1^M-KO mice still displayed evident necrotic areas at 168 h post injection ([67]Fig. S2C). Consistent with the PHx model results, expression of the proliferation marker Ki67 was significantly suppressed in the livers of Atg16l1^M-KO mice ([68]Fig. S2D). Taken together, these results suggest that ATG16L1 deletion in macrophages delays hepatocyte proliferation after PHx. Macrophage-specific ATG16L1 deletion does not affect the initiation phase of liver regeneration after PHx Macrophages contribute to the entire process of liver regeneration.[69]^12 To determine whether the deletion of macrophage-specific ATG16L1 delays liver regeneration by affecting the initiation phase, we first examined the gene expression of the cell cycle markers Cyclin A2, Cyclin B1, Cyclin D1, and Cyclin E1, as well as the protein levels of proliferating cell nuclear antigen, in liver tissue. No significant differences were observed between Atg16l1^M-KO and Atg16l1^FL/FL mice ([70]Fig. 3A,B). Similarly, during the initiation phase, no notable difference was observed in the proportion of hepatic macrophages and neutrophils between the two groups ([71]Fig. 3C,D). Fig. 3. [72]Fig. 3 [73]Fig. 3 [74]Open in a new tab Macrophage-specific ATG16L1 deletion does not influence the initiation phase of liver regeneration after PHx. (A) Quantification of CyclinA2, CyclinB1, CyclinD1, and CyclinE1 mRNA expression in Atg16l1^FL/FL and Atg16l1^M-KO mice 0–6 h after PHx (n = 6 per group). (B) Western blot analysis and corresponding quantification of PCNA expression in Atg16l1^FL/FL and Atg16l1^M-KO mice 0–6 h after PHx (n = 4 per group). NS, no significance, by two-way ANOVA with Bonferroni post-test. (C) Representative FACS plots and quantification of MoMFs and KCs in Atg16l1^FL/FL and Atg16l1^M-KO mice liver tissues 3 and 6 h after PHx (n = 6 per group). (D) Representative FACS plots and neutrophil quantification in Atg16l1^FL/FL and Atg16l1^M-KO mice liver tissues 3 and 6 h after PHx (n = 6 per group). NS, no significance, by an unpaired, 2-tailed Student’s t test. (E) Quantification of IL-6 and TNF-α mRNA expression in Atg16l1^FL/FL and Atg16l1^M-KO mice at indicated time points after 70% PHx (n = 6 per group). (F) Serum IL-6 and TNF-α assayed by ELISA in Atg16l1^FL/FL and Atg16l1^M-KO mice 0, 3, and 6 h after 70% PHx (n = 6 per group). (G) Western blot analysis and corresponding quantification of STAT3 and p-STAT3 expression in Atg16l1^FL/FL and Atg16l1^M-KO mice. β-actin was used as a control. NS, no significance, by two-way ANOVA with Bonferroni post-test. ATG16L1, autophagy-related 16-like 1; KC, Kupffer cell; KO, knockout; MoMF, monocyte-derived macrophage; p-STAT3, phosphorylated signal transducer and activator of transcription 3; PCNA, proliferating cell nuclear antigen; PHx, partial hepatectomy; STAT3, signal transducer and activator of transcription 3; TNF, tumor necrosis factor. Macrophage-derived IL-6 and tumor necrosis factor (TNF) are crucial cytokines in the initiation phase of liver regeneration.[75]^13 Therefore, we compared the mRNA levels of IL-6 and TNF in liver tissue, as well as their serum concentrations during the early phase of regeneration. No significant differences were detected between the two groups ([76]Fig. 3E,F). In addition, signal transducer and activator of transcription 3 (STAT3), a downstream signaling mediator in response to IL-6 and a key regulator of cell proliferation,[77]^14 showed no difference in its active form (phosphorylated STAT3) between Atg16l1^M-KO and Atg16l1^FL/FL mice ([78]Fig. 3G). In summary, ATG16L1 deletion in macrophages does not inhibit liver regeneration by affecting the initiation phase. ATG16L1 deletion impairs liver regeneration by remodeling macrophage phenotypes Hepatic macrophages can interact with other immune cell subsets to indirectly influence liver regeneration.[79]^13^,[80]^15 Therefore, we hypothesized that the loss of ATG16L1 in macrophages might lead to altered crosstalk with immune cells, affecting liver regeneration. We assessed post-PHx immune cell infiltration in the liver in both Atg16l1^M-KO and Atg16l1^FL/FL mice. Flow cytometry and immunohistochemistry revealed a significant increase in neutrophil infiltration in Atg16l1^M-KO mice, with attenuated clearance of neutrophils during the regenerative phase ([81]Fig. 4A,B). Recently, a collaboration between neutrophils and macrophages in coordinating inflammation and tissue repair was reported.[82]^16 To investigate whether the inhibition of liver regeneration in ATG16L1-deficient macrophages is neutrophil dependent, we depleted neutrophils in vivo by administering Ly6G-specific antibodies and observed liver regeneration post PHx. Surprisingly, although neutrophil depletion delayed liver proliferation in both groups, hepatocyte proliferation at 36 h and 48 h post PHx remained significantly lower in Atg16l1^M-KO mice than in Atg16l1^FL/FL mice ([83]Fig. 4C–E). This suggests that inhibition of liver regeneration caused by ATG16L1 deficiency in macrophages is not neutrophil dependent. Next, we examined the infiltration of natural killer cells (NK1.1^hiCD3^−), CD4^+ T (CD3^+CD4^+), and CD8^+ T (CD3^+CD8^+) cells. Flow cytometry results showed no significant differences in cell proportions or absolute numbers between the two groups ([84]Fig. 4F). Fig. 4. [85]Fig. 4 [86]Open in a new tab Inhibition of liver regeneration by macrophage-specific ATG16L1 deletion is independent of other immune cells. (A) Representative FACS plots and neutrophil quantification of Atg16l1^FL/FL and Atg16l1^M-KO mice liver tissues at indicated time points after 70% PHx (n = 6 per group). (B) Representative immunohistochemical images and quantification of hepatic Ly6G^+ cells in Atg16l1^FL/FL and Atg16l1^M-KO mice at indicated time points after 70% PHx (n = 6 per group). (C) Mice were treated with the neutrophil-depleting anti-Ly6G monoclonal antibody or an isotype control (IgG) 3 h after PHx. (D) Representative liver sections 48 h after PHx and respective quantification of Ki67-positive hepatocytes in Atg16l1^FL/FL and Atg16l1^M-KO mice (n = 6 per group). (E) Liver-to-body weight ratio of Atg16l1^FL/FL and Atg16l1^M-KO mice at 48 h after 70% PHx (n = 6 group). ∗∗p <0.01, ∗∗∗p <0.001, by two-way ANOVA with Bonferroni post-test. (F) Representative FACS plots of CD4^+T, CD8^+T, and NK cells from Atg16l1^FL/FL and Atg16l1^M-KO mice 36 h after PHx and quantification of the proportion and numbers in CD45^+ cells (n = 6 per group). NS, no significance, by two-way ANOVA with Bonferroni post-test. Scale bars: 50 μm (B), 100 μm (D). ATG16L1, autophagy-related 16-like 1; KO, knockout; PHx, partial hepatectomy. These findings suggest that ATG16L1 influences liver regeneration by directly affecting macrophage phenotype transformation. Anti-F4/80 immunofluorescence staining revealed a significant increase in macrophage numbers in the liver of Atg16l1^M-KO mice at 36 h post PHx ([87]Fig. 5A). We further evaluated the hepatic macrophage population using flow cytometry. The results showed a shift toward monocyte-derived macrophages in Atg16l1^M-KO mice, with a significant increase in the proportion of monocyte-derived macrophages and a relative decrease in Kupffer cells (KCs). Interestingly, although the proportion of KCs was significantly reduced, their absolute number showed no difference between the two groups ([88]Fig. 5B). Fig. 5. [89]Fig. 5 [90]Fig. 5 [91]Open in a new tab ATG16L1 deletion affects liver regeneration by remodeling macrophage phenotypes. (A) Representative IF co-staining images of F4/80 (green) and DAPI (blue) in Atg16l1^FL/FL and Atg16l1^M-KO mice and quantification 0 and 36 h after PHx. (B) Representative FACS plots and quantification of MoMFs and KCs in Atg16l1^FL/FL and Atg16l1^M-KO mice liver tissues 0 and 36 h after PHx (n = 6 per group). (C) Representative dual IF staining for iNOS (purple) and F4/80 (green) in Atg16l1^FL/FL and Atg16l1^M-KO mice liver sections. Representative dual IF staining for CD206 (purple) and F4/80 (green) in liver sections 36 h post PHx compared with sham controls (n = 6 per group). (D) Representative FACS plots and quantification of Ly6C^lo or Ly6C^hi hepatic macrophages 36 h after 70% PHx (n = 6 per group). (E) Schematic of in vitro co-culture of primary hepatocytes with CM from sorted macrophages in Atg16l1^FL/FL and Atg16l1^M-KO mice. (F) Representative IF staining images of Ki67 and quantification of Ki67-positive cells in Atg16l1^FL/FL and Atg16l1^M-KO macrophage CM-treated hepatocytes (n = 4 per group). (G) Western blot of PCNA expression in primary hepatocytes cultured for 48 h in DMEM and the supernatant of sorted macrophages from Atg16l1^FL/FL and Atg16l1^M-KO mice (n = 4–6 per group). Quantitative data are also shown. (H) ELISA analysis of IL-1β, TNF-α, IL-6, and HGF secreted by sorted macrophages from Atg16l1^FL/FL and Atg16l1^M-KO mice (n = 6 per group). (I) Volcano plot of all DEGs in Atg16l1^FL/FL and Atg16l1^M-KO liver macrophages. (J) RNA-sequencing analysis of hepatic macrophages from Atg16l1^FL/FL and Atg16l1^M-KO mice after 70% PHx (n = 3 per group). Heatmap showing proinflammatory gene expression profiles in Atg16l1^FL/FL and Atg16l1^M-KO mice. (K) Quantification of mRNA expression of genes detected in (J) in Atg16l1^FL/FL and Atg16l1^M-KO mice. ∗∗∗p <0.001, by an unpaired, 2-tailed Student’s t test. Scale bars: 50 μm (C,F), 100 μm (A). ATG16L1, autophagy-related 16-like 1; CM, Conditioned medium; DEG, differentially expressed gene; IF, immunofluorescence; KO, knockout; MoMF, monocyte-derived macrophage; KC, Kupffer cell; PCNA, proliferating cell nuclear antigen; PHx, partial hepatectomy; TNF, tumor necrosis factor. Dual immunofluorescence staining showed that, at 36 h post PHx, ‘M1-like’ macrophages (iNOS^+F4/80^+) were significantly more abundant in the livers of Atg16l1^M-KO mice than in those of Atg16l1^FL/FL mice, whereas the number of ‘M2-like’ macrophages (CD206^+F4/80^+) was markedly reduced ([92]Fig. 5C). In macrophage-mediated regeneration and repair studies, recruited reparative (M2-like) and proinflammatory (M1-like) macrophage subsets are often distinguished by Ly6C^lo and Ly6C^hi markers.[93]^17^,[94]^18 In the present study, ATG16L1 deletion significantly altered the ratio of Ly6C^lo to Ly6C^hi macrophages. At 36 h post PHx, the proportion of Ly6C^hi macrophages in the livers of Atg16l1^M-KO mice was significantly higher than that in Atg16l1^FL/FL mice, whereas the proportion of Ly6C^lo macrophages was significantly reduced ([95]Fig. 5D). To clarify the impact of this change on hepatocyte proliferation, we co-cultured primary hepatocytes from WT mice with conditioned media from the macrophages of either Atg16l1^M-KO or Atg16l1^FL/FL mice for 48 h. We observed a significant reduction in the number of proliferating hepatocytes, as indicated by Ki67 staining, and a marked decrease in proliferating cell nuclear antigen protein levels in the Atg16l1^M-KO group compared with those in the Atg16l1^FL/FL group. Analysis of cytokine levels in the conditioned media showed significantly higher levels of proinflammatory cytokines, such as IL-1β, IL-6, and TNF, in the Atg16l1^M-KO group, but a significantly lower level of hepatocyte growth factor (HGF), a key factor in promoting hepatocyte proliferation ([96]Fig. 5E–H). Subsequently, we analyzed the gene expression profile of hepatic macrophages 36 h post PHx. RNA-sequencing analysis revealed that ATG16L1 deletion led to significant upregulation of inflammation-related genes in macrophages, consistent with our earlier finding of a shift toward an ‘M1-like’ macrophage phenotype ([97]Fig. 5I–K). In conclusion, macrophage-specific ATG16L1 deletion disrupts the balance between proinflammatory and reparative macrophages following PHx, skewing the macrophage population toward a proinflammatory phenotype, which ultimately impairs liver regeneration. ATG16L1 deletion remodels macrophage phenotypes by inhibiting mitophagy To investigate the mechanism underlying macrophage phenotype transformation, we performed Kyoto Encyclopedia of Genes and Genomes pathway and gene enrichment analyses. The results showed that the oxidative phosphorylation pathway in macrophages from Atg16l1^M-KO mice had a significant p value, with related genes being significantly downregulated ([98]Fig. 6A). Macrophage metabolism is crucial for shaping their phenotype: ‘M1’-like macrophages rely on glycolysis, whereas ‘M2’-like macrophages depend on mitochondria-driven oxidative phosphorylation.[99]^19 Therefore, we introduced myeloid-specific ATG16L1 overexpression mice to further explore the effect of ATG16L1 on macrophage metabolism. Specifically, we used the Seahorse XFe96 analyzer to assess mitochondrial metabolism and glycolysis in hepatic macrophages from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice at 36 h post PHx. The results showed that ATG16L1 deletion significantly promoted glycolysis in macrophages, whereas oxidative phosphorylation was markedly decreased. Conversely, ATG16L1 overexpression enhanced mitochondrial metabolic activity and reduced glycolysis levels ([100]Fig. 6B,C). Our experimental findings confirmed the regulatory role of ATG16L1 in macrophage metabolic pathways. Furthermore, we measured mitochondrial metabolic activity and glycolysis levels in Ly6C^lo and Ly6C^hi macrophages in WT mice at 36 h post PHx. We found that Ly6C^lo macrophages preferred oxidative phosphorylation, whereas Ly6C^hi macrophages were more inclined toward glycolysis ([101]Figs. S3A and B). This further confirmed that ATG16L1 might alter the ratio of Ly6C^lo to Ly6C^hi macrophages in the liver after PHx by affecting macrophage metabolism. Fig. 6. [102]Fig. 6 [103]Fig. 6 [104]Open in a new tab ATG16L1 deletion mediates a macrophage metabolic shift by impairing mitophagy. (A) RNA-sequencing analysis of hepatic macrophages sorted by flow cytometry from mice 36 h after PHx. Bubble chart showing the top 20 significantly downregulated genes based on KEGG enrichment. The oxidative phosphorylation pathway in macrophages from Atg16l1^M-KO mice had a significant p value. (B) OCR of macrophages sorted from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice 36 h post PHx assessed by a Seahorse assay before and after the sequential addition of oligomycin, FCCP, rotenone, and antimycin (n = 6 per group). (C) ECAR assessed by a Seahorse assay of macrophages sorted from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice at 36 h post PHx (n = 6 per group). ∗∗p <0.01,∗∗∗p <0.001, by one-way ANOVA with Bonferroni post-test. Quantification of mitochondrial membrane potential (Δψm) (D) and mtROS levels (E) of liver F4/80 and CD11b double-positive macrophages from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice 36 h post PHx (n = 6 per group). (F) Representative DCFDA fluorescence images for mtROS levels of sorted liver macrophages in Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice 0 and 36 h post PHx (n = 6 per group). (G) Quantification of mitochondrion GSH/GSSG levels in liver macrophages sorted from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice at 0 and 36 h post PHx (n = 4 per group). (H) Representative electron microscopy images of Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE liver macrophages 36 h after PHx (n = 4 per group). Mitochondria are marked by red arrowheads, and autophagosomes are marked by yellow arrowheads. (I) Representative confocal microscopy images of liver macrophages 0 and 36 h post PHx stained with TOM20 (red) and dsDNA (green). Quantification of relative numbers of cytosolic mtDNA in macrophages from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice (n = 6 per group). qPCR analysis of the mtDNA/nDNA ratio (J) and representative Western blot images of TOM20, TIM23, VDAC1, P62, LC3b, and β-ACTIN levels in Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE liver macrophages 36 h post PHx (n = 4–6 per group). (L) Western blot analysis of TOM20, P62, LC3B, and β-actin expression of sorted macrophages in Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice with or without leupeptin 36 h post PHx compared with sham controls (n = 4–6 per group). ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, by one-way ANOVA with Bonferroni post-test. Scale bars: 5 μm (I), 20 μm (F), 500 μm (H). ATG16L1, autophagy-related 16-like 1; ECAR, extracellular acidification rate; GSH, glutathione; GSSG, glutathione oxidized; KEGG, Kyoto Encyclopedia of Genes and Genomes; KO, knockout; Ly6C^hi, classically activated pro-inflammatory macrophages; Ly6C^lo, alternatively activated reparative macrophages; OCR, oxygen consumption rate; mtDNA, mitochondrial DNA; nDNA, nuclear DNA; ROS, reactive oxygen species; PHx, partial hepatectomy; WT, wild type. Cellular metabolism depends on the structure and function of mitochondria.[105]^20 Therefore, we examined mitochondrial function in hepatic macrophages at 36 h post PHx. In ATG16L1-deficient macrophages, we observed a reduction in mitochondrial membrane potential, an increase in mitochondrial reactive oxygen species (ROS) production, decreased glutathione (which represents the antioxidant level in mitochondria), and significantly elevated overall levels of cellular ROS. Conversely, ATG16L1 overexpression significantly improved mitochondrial function and reduced oxidative stress within the mitochondria ([106]Fig. 6D–G). Cellular electron microscopy results showed reduced mitochondrial electron density and edema in macrophages after liver resection. In the Atg16l1^M-KO group, mitochondrial damage was more severe, with mitochondrial fragmentation and loss of intramitochondrial cristae, along with a significant reduction in the number of autophagosomes. By contrast, the Atg16l1^OE group exhibited better overall mitochondrial morphology and a significantly increased number of autophagosomes compared with the Atg16l1^FL/FL group ([107]Fig. 6H). Given that ATG16L1 is a key gene in the autophagy process,[108]^9^,[109]^21 we hypothesized that ATG16L1 deletion might inhibit mitophagy in hepatic macrophages after PHx, leading to the accumulation of mitochondrial damage. Therefore, we conducted mitophagy assays. Double immunofluorescence staining for TOM20 and dsDNA revealed mtDNA leakage and accumulation in the cytoplasm of hepatic macrophages in both Atg16l1^FL/FL and Atg16l1^M-KO mice at 36 h post PHx, with more considerable mtDNA accumulation in Atg16l1^M-KO macrophage cytoplasm. By contrast, mtDNA leakage was significantly reduced in Atg16l1^OE macrophage cytoplasm ([110]Fig. 6I). We subsequently used qPCR to measure mtDNA/nuclear (n)DNA copy numbers and performed western blotting to detect several mitochondrial proteins, including TOM20, TIM23, and VDAC1, as well as the autophagy-related proteins P62 and LC3b. The results showed that ATG16L1 deletion significantly inhibited the conversion of LC3b I to LC3b II, thereby blocking the occurrence of mitophagy. By contrast, ATG16L1 overexpression significantly increased the mitophagy flux in macrophages ([111]Fig. 6J,K). To determine autophagic flux, we compared mice treated with leupeptin, a lysosomal inhibitor, with those without leupeptin treatment. Following PHx, the levels of LC3-II in hepatic macrophages of Atg16l1^FL/FL and Atg16l1^OE mice were significantly elevated, and this increase was further augmented in the presence of leupeptin. However, this was not observed in the hepatic macrophages of Atg16l1^M-KO mice, indicating that ATG16L1 deletion impairs the lipidation process of LC3b-I, thereby inhibiting autophagy. Furthermore, changes in mitochondrial protein levels demonstrated that mitophagy was also suppressed ([112]Fig. 6L). In summary, these results indicate that ATG16L1 can mitigate mitochondrial damage during liver regeneration by regulating mitophagy, thereby maintaining metabolic activity in macrophages and ultimately controlling the macrophage phenotype. Macrophage ATG16L1 deletion inhibits liver regeneration by promoting and sustaining cGAS–STING pathway activation Given that ATG16L1 deletion induces macrophage metabolic phenotype shifts through mitochondrial damage, we hypothesized that other pathways may also inhibit liver regeneration. A previous study reported that ATG16L1 deletion significantly enhances stimulator of interferon genes (STING) activity in intestinal epithelial cells.[113]^22 Our earlier work revealed that defects of mitochondrial autophagy in macrophage senescence cause mtDNA leakage and activate the intracellular STING pathway.[114]^23 Furthermore, interferon (IFN) can inhibit liver regeneration.[115]^24 Therefore, we hypothesized that, after liver resection, STING in macrophages might act as a crucial mediator inhibiting liver regeneration in the context of ATG16L1 deficiency. We identified 18 genes involved in the IFN-I response from our sequencing data and validated their upregulation using qPCR. Interaction analysis using the STRING database indicated that the cyclic GMP-AMP synthase (cGAS)–STING pathway may be implicated in the response to ATG16L1 deletion ([116]Fig. 7A–C).[117]^25 Immunofluorescence staining of liver tissue showed a significant increase in phosphorylated (p)-STING^+ macrophages in the liver of Atg16l1^M-KO mice at 36 h post PHx, whereas p-STING expression in Atg16l1^OE macrophages was significantly reduced. Western blot analysis further confirmed that ATG16L1 deletion significantly upregulated the cGAS–STING signaling pathway in hepatic macrophages, whereas ATG16L1 overexpression markedly inhibited this pathway ([118]Fig. 7D,E). Fig. 7. [119]Fig. 7 [120]Fig. 7 [121]Open in a new tab Macrophage ATG16L1 deletion inhibits liver regeneration by promoting and sustaining cGAS–STING pathway activation. (A) Heatmap showing the clustering of 18 upregulated genes affecting the IFN-I response. (B) STRING-based network analysis of genes detected in (A), revealing the strong contribution of an IFN-I-related signature. (C) Quantification of mRNA expression of genes detected in (A) in Atg16l1^FL/FL and Atg16l1^M-KO liver macrophages. ∗∗∗p <0.001, by an unpaired, 2-tailed Student’s t test. (D) Representative IF co-staining images of p-STING (green), F4/80 (red), and DAPI (blue) in Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice 0 and 36 h after PHx. (E) Representative Western blot images of cGAS, p-STING, p-TBK1, TBK1, p-IRF3, IRF3, and β-ACTIN levels in Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE liver macrophages 36 h post PHx compared with sham controls (n = 4–6 per group). STING was silenced by AAV vectors. (F) Representative liver sections 36 and 48 h after PHx and respective quantification of Ki67-positive hepatocytes in Atg16l1^FL/FL+AAV-CON, Atg16l1^M-KO+AAV-CON, Atg16l1^M-KO+AAV-shTmem173 mice. Quantification of liver-to-body weight ratio 36 and 48 h after 70% PHx (G) and serum IL-1β, TNF-α, IL-6, IFN-α, and IFN-β levels assayed by ELISA 36 h post PHx (H). (I) Representative western blot images of p-TBK1, TBK1, p-IRF3, IRF3, and β-ACTIN levels in liver macrophages 36 h post PHx compared with sham controls (n = 4–6 per group). (J) Representative IF staining images of Ki67 and quantification of Ki67-positive cells in Atg16l1^FL/FL+AAV-CON, Atg16l1^M-KO+AAV-CON, Atg16l1^M-KO+AAV-shTmem173 macrophage CM-treated hepatocytes. (K) ELISA analysis of cytokines secreted by sorted macrophages. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, by one-way ANOVA with Bonferroni post-test. Scale bars: 50 μm (D,J), 100 μm (F). AAV, adeno-associated virus; ATG16L1, autophagy-related 16-like 1; cGAS, cyclic GMP-AMP synthase; CM, conditioned medium; IF, immunofluorescence; IFN, interferon; IRF3, interferon regulatory factor 3; p-, phosphorylated; PHx, partial hepatectomy; STING, stimulator of interferon genes; TBK1, TANK-binding kinase 1; TNF, tumor necrosis factor. To further explore the role of STING pathway activation in liver regeneration after ATG16L1 deletion, we silenced STING using adeno-associated virus (AAV) vectors and analyzed liver regeneration post resection. The results showed partial recovery of liver regeneration and reduced serum inflammatory cytokine levels, particularly IFN, in STING-silenced Atg16l1^M-KO mice compared with Atg16l1^M-KO mice. Western blot analysis also revealed the downregulation of STING downstream signals, including p-TANK-binding kinase 1 (TBK1) and p-IFN regulatory factor 3 (IRF3), in hepatic macrophages ([122]Fig. 7F–H). When primary hepatocytes from WT mice were cultured with conditioned media from the three groups of macrophages for 48 h, STING silencing in Atg16l1^M-KO macrophages had a similar effect on hepatocyte proliferation to that observed in vivo ([123]Fig. 7J,K). Next, we assessed mitochondrial metabolic activity and glycolysis levels in hepatic macrophages from the three groups of mice after PHx. Notably, STING silencing did not reverse the metabolic pathways in Atg16l1^M-KO macrophages ([124]Figs. S4A and B), which might be because of irreversible mitochondrial damage caused by ATG16L1 deficiency. Notably, when we performed PHx in Tmem173^FL/FL and Tmem173^M-KO mice, liver regeneration was enhanced in Tmem173^M-KO mice compared with Tmem173^FL/FL mice ([125]Figs. S4C and D). These findings indicate that STING acts as a key mediator in the inhibition of liver regeneration following ATG16L1 deletion in macrophages. ATG16L1 deficiency in BJ1 fibroblasts affects the degradation of STING pathway proteins.[126]^26 Thus, we hypothesized that a similar phenomenon may occur in macrophages. We treated bone marrow-derived macrophages from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice with the STING agonist DMXAA and examined the expression of p-STING, STING, p-TBK1, and TBK1 at different time points. Immunofluorescence and western blot results demonstrated that ATG16L1 deficiency inhibited the degradation of STING signaling proteins, promoting sustained activation of the pathway, whereas STING degradation was accelerated in ATG16L1-overexpressing macrophages ([127]Figs. S5A and B). However, the precise mechanism by which ATG16L1 influences STING degradation requires further investigation. In summary, our findings suggest that macrophage ATG16L1 deficiency inhibits liver regeneration by promoting and sustaining STING pathway activation. Targeted activation of ATG16L1 promotes liver regeneration To confirm the role of macrophage ATG16L1 activation in liver regeneration, we extracted bone marrow-derived macrophages from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice and implanted them into the spleens of WT mice undergoing PHx. At 36 h post PHx, we analyzed liver-to-body weight ratios and performed Ki67 immunohistochemical staining of liver tissue. The transfer of Atg16l1^OE macrophages significantly promoted liver regeneration in WT mice, whereas the transfer of Atg16l1^M-KO macrophages had the opposite effect ([128]Fig. 8A–D). In addition to cell therapy, we explored the effects of pharmacologically induced ATG16L1 activation on liver regeneration. After treating Atg16l1^FL/FL mice with the ATG16L1 enhancer peretinoin, we observed significant enhancement of liver regeneration ([129]Fig. 8E,F). In conclusion, the targeted activation of ATG16L1 can effectively promote liver regeneration. Fig. 8. [130]Fig. 8 [131]Open in a new tab Targeted activation of ATG16L1 promotes liver regeneration. (A) Strategy of BMDM infusion into mice after PHx. (B) PKH26 Dye-labeled BMDMs (red) were detected in F4/80^+ macrophages (green) under a fluorescence microscope 36 h following cell infusion. BMDMs for infusion were sorted from Atg16l1^FL/FL, Atg16l1^M-KO, and Atg16l1^OE mice. (C) Representative liver sections 36 h after PHx and cell infusion and respective quantification of Ki67-positive hepatocytes (n = 6 per group). (D) Liver-to-body weight ratio of mice implanted with BMDMs 36 h after 70% PHx (n = 6 per group). ∗p <0.05, ∗∗∗p <0.001, by one-way ANOVA with Bonferroni post-test. (E) The ATG16L1 enhancer peretinoin was used immediately after 70% PHx. Representative immunohistochemical staining images of Ki67 and quantification of Ki67-positive cells in liver sections after PHx in Atg16l1^FL/FL mice. (F) Liver-to-body weight ratio of Atg16l1^FL/FL mice 36 h after 70% PHx. ∗∗∗p <0.001, by an unpaired, 2-tailed Student’s t test. Scale bars: 20 μm (B), 100 μm (C,E), ATG16L1, autophagy-related 16-like 1; BMDM, bone marrow-derived macrophage; KO, knockout; PHx, partial hepatectomy. Discussion Given that liver regeneration is a complex and dynamic process, its orderly progression requires the coordination and involvement of immune cells.[132]^1^,[133]^27 Although previous studies have thoroughly described the positive relationship between autophagy and liver regeneration, most have been limited to hepatocytes.[134]^8^,[135]^28 Therefore, our understanding of the role of autophagy in immune cells during liver regeneration remains limited. The results of this study reveal the regulatory role and mechanism of the autophagy-related gene ATG16L1 in macrophages during liver regeneration after PHx. We demonstrated that the conditional deletion of ATG16L1 in macrophages after PHx hinders mitophagy, leading to the accumulation of mitochondrial damage and forcing a metabolic and functional shift in macrophages from a reparative to a proinflammatory phenotype. Furthermore, ATG16L1 deficiency triggers activation of the STING signaling pathway in macrophages and inhibits its degradation, thereby reshaping the macrophage phenotype and ultimately suppressing liver regeneration. Both cell therapy and pharmacological studies suggest that the targeted activation of ATG16L1 represents a novel strategy for promoting liver regeneration. Effective liver regeneration is critical for a favorable clinical outcome in various liver diseases. Small-for-size syndrome is a major cause of postoperative liver failure and even death in patients undergoing liver resection or transplantation.[136]^3 In the present study, both Atg16l1^FL/FL and Atg16l1^M-KO mice showed mortality in a 90% hepatectomy model. Notably, most of these mice died within 48 h post hepatectomy. Thus, early mortality in these mice may have resulted from the remnant liver volume being insufficient to provide adequate compensatory function. This would have led to significant declines in metabolic, detoxification, and digestive capacities, ultimately resulting in early death. However, the overall survival rate of Atg16l1^M-KO mice was significantly lower than that of Atg16l1^FL/FL mice, suggesting that differences in macrophage-mediated regeneration capacity also affect mortality. Therefore, in the 90% hepatectomy mouse model, both liver failure caused by extensive resection and aberrant macrophage-mediated regeneration regulation are likely key contributors to early mortality. Liver regeneration is also influenced by injury factors, including the type, duration, and severity of the injury.[137]^29 However, compared with the regeneration models induced by toxicity or drugs, liver resection is a low-injury model. In the present study, we assessed liver injury at various time points following 70% hepatectomy. We found that although alanine aminotransferase and aspartate aminotransferase levels were elevated in Atg16l1^M-KO mice compared with those in the control group, liver histopathology with H&E staining showed no significant differences between the groups, indicating a relatively mild liver injury. Thus, the difference in injury caused by myeloid Atg16l1 deficiency might not significantly affect liver regeneration. In addition, the macrophage transfusion and co-culture experiments of macrophage-conditioned medium with hepatocytes further confirmed that ATG16L1 signaling in macrophages had a crucial role in liver regeneration. Together, the delayed regeneration in Atg16l1^M-KO mice might be predominantly the result of the role of macrophages during the regenerative repair phase rather than of differences in injury. As the primary immune cell population in the liver, macrophages have a crucial role in maintaining liver homeostasis and in liver diseases.[138]^30^,[139]^31 Owing to their heterogeneity under different conditions, macrophages exhibit dual roles in disease progression. Historically, macrophages have been classified into proinflammatory M1 or anti-inflammatory M2 phenotypes; however, this classification does not apply to all scenarios.[140]^32 Increasing attention has been given to the crucial impact of the balance between Ly6C^hi and Ly6C^lo macrophages in liver diseases.[141]^33 During inflammatory liver injury, Ly6C^hi macrophages promote damage progression by secreting inflammatory mediators, such as IL-1β, IL-6, and TNF-α, whereas alternatively activated Ly6C^lo macrophages produce anti-inflammatory mediators, such as IL-4 and IL-10, promoting tissue repair and regeneration.[142]^34 During liver regeneration and repair, a transition from Ly6C^hi to Ly6C^lo macrophages occurs.[143]^35 Liver-resident macrophages (KCs) undergo cell death in various liver disease conditions. In our previous study, we observed that KCs underwent early cell death in an acute liver ischemia–reperfusion injury model, with infiltration of peripheral macrophages into the liver, both leading to dynamic changes in the proportion of hepatic macrophages.[144]^36 However, this study did not analyze the impact of macrophage death on liver regeneration. We quantified the absolute number of KCs and found no difference between Atg16l1^FL/FL and Atg16l1^M-KO mice. Therefore, the differences in macrophage proportions might be mainly the result of a significant increase in monocyte-derived macrophages following ATG16L1 deletion. The role of autophagy in KC death remains controversial. In a chronic intermittent hypoxia model, autophagy inhibition exacerbated KC apoptosis.[145]^37 Conversely, autophagy inhibition significantly increased KC survival and reduced nanoparticle-induced liver damage.[146]^38 Therefore, autophagy could have a dual role in KC death, depending on the context. Whether the liver regeneration model induces KC death and whether ATG16L1 influences KC death remains to be studied further. Macrophage phenotype is closely linked to autophagy, with studies showing that autophagy-deficient macrophages trigger stronger inflammatory responses,[147]^39^,[148]^40 and that the M2 polarization of macrophages heavily depends on autophagy.[149]^41 ATG16L1, a key autophagy gene, regulates both canonical and non-canonical autophagy by influencing LC3 lipidation.[150]^42 Therefore, the loss of ATG16L1 in macrophages often leads to excessive inflammation.[151]^43 Our previous metabolic dysfunction-associated steatohepatitis model demonstrated that deletion of ATG16L1 in macrophages activated a proinflammatory phenotype, exacerbating disease progression.[152]^10 Similarly, in the present study, Atg16l1^M-KO mice exhibited a tendency toward proinflammatory macrophages at 36 h post PHx, with a significant decrease in reparative macrophages. This was accompanied by elevated levels of inflammatory cytokines, such as IL-1β, IL-6, and TNF-α, as well as a marked reduction in the macrophage-secreted HGF, resulting in suppressed liver regeneration. Apart from ATG16L1, we also detected that autophagy genes, such as ATG5 and ATG12, in macrophages exhibited upregulated expression following liver resection. The involvement of hepatocyte ATG5 in liver regeneration has been demonstrated, but further examination is warranted to elucidate its potential influence on macrophage phenotype and function during liver regenerative. HGF secretion levels vary among macrophage populations, with Ly6C^lo macrophages secreting higher levels of HGF compared with Ly6C^hi macrophages.[153]^17 We found that autophagy can influence HGF secretion by driving macrophage phenotype switching; following ATG16L1 KO, liver macrophages shifted toward the Ly6C^hi phenotype, and HGF secretion was significantly reduced. In addition to determining macrophage phenotypes, autophagy itself may affect HGF expression. Autophagy deficiency in tumor cells can block HGF expression by inhibiting DNMT1 degradation, suggesting that HGF secretion is dependent upon autophagy.[154]^44 However, given that autophagy often induces broad intracellular changes, multiple signaling pathways might be involved in the regulation of HGF expression by ATG16L1, warranting further investigation. In this study, RNA-sequencing analysis revealed that macrophages with ATG16L1 deletion had downregulated oxidative phosphorylation pathways. It is widely accepted that macrophage phenotype changes are dependent on their metabolic reprogramming. As the core of cellular metabolism, mitochondria directly influence macrophage polarization.[155]^45 PHx can induce mitochondrial damage within cells, a phenomenon observed in hepatocytes.[156]^8^,[157]^46 However, whether PHx induces mitochondrial damage in macrophages in the regenerative environment has not been fully studied. By assessing mitochondrial function and macrophage metabolism in the liver, we found that PHx led to mitochondrial damage in macrophages, which was exacerbated by ATG16L1 deficiency, shifting their metabolism toward glycolysis. Conversely, ATG16L1 overexpression significantly improved mitochondrial function and metabolic activity. Autophagy, particularly mitophagy, has a crucial role in regulating mitochondrial quantity and maintaining normal mitochondrial function, and is closely associated with the development of various systemic diseases.[158]^47 Mitophagy defects in macrophages, which lead to mitochondrial damage and metabolic alterations, drive the proinflammatory transformation of macrophages.[159]^48 Our earlier research revealed that diminished mitophagy in senescent macrophages exacerbates sterile liver inflammation.[160]^23 However, whether mitophagy in macrophages has a protective role in mitochondria during liver regeneration has not previously been elucidated. Through further examination of mitophagy flux in macrophages, we found that ATG16L1 deletion impairs mitophagy by affecting LC3b lipidation, whereas ATG16L1 overexpression in liver macrophages enhances mitophagy, reducing mitochondrial damage. Thus, we demonstrated that ATG16L1 drives mitophagy and protects macrophage mitochondria in the liver regeneration environment. However, as a key molecule in autophagy, ATG16L1 influences other autophagy-related processes, but whether these changes also have a role in liver regeneration requires further investigation. The cGAS–STING signaling pathway has garnered attention because of its role in antimicrobial and antitumor responses. In recent years, research identified STING as a key therapeutic target for inflammatory liver diseases, where inhibition of STING signaling effectively limits liver inflammatory damage. IFN can inhibit liver regeneration, but STING, as a key molecule in IFN activation, has not yet been fully explored in the context of liver regeneration.[161]^24 The cytosolic release of mtDNA in macrophages can act as an intrinsic signal to activate STING in various disease models. Following PHx, we observed that mtDNA leakage in macrophages triggered STING activation, which was more pronounced in the absence of ATG16L1. Inhibiting STING signaling partially restored liver regeneration in ATG16L1-deficient mice, although the recovery was incomplete. This may be because STING inhibition cannot fully reverse mitochondrial damage in macrophages. Beyond STING activation, crosstalk exists between autophagy and STING degradation. For example, in BJ1 fibroblasts, ATG16L1 deletion affects STING pathway protein degradation.[162]^26 In our model, we also found that ATG16L1 deficiency inhibits the degradation of STING in macrophages. In summary, our study highlights the crucial regulatory role of macrophage ATG16L1 signaling in liver regeneration. ATG16L1 deficiency drives metabolic reprogramming and proinflammatory polarization in macrophages by inhibiting mitophagy and simultaneously sustains excessive STING activation by both activating the pathway and inhibiting STING degradation, resulting in elevated levels of IFN and other inflammatory cytokines, thereby suppressing liver regeneration. Pharmacological activation or genetic overexpression of ATG16L1 in macrophages significantly promotes liver regeneration and repair. These findings suggest that ATG16L1 acts as a key switch for the reprogramming of macrophages into a reparative phenotype, and that the targeted activation of ATG16L1 represents a promising new therapeutic approach for enhancing liver regeneration in clinical settings. Abbreviations AAV, adeno-associated virus; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ATG16L1, autophagy related 16-like 1 gene; BMDM, bone marrow-derived macrophage; cGAS, cyclic GMP-AMP synthase; CM, conditioned medium; DEG, differentially expressed gene; ECAR, extracellular acidification rate; GO, Gene Ontology; GSH, glutathione; HGF, hepatocyte growth factor; IF, immunofluorescence; IFN, interferon; IRF3, interferon regulatory factor 3; KC, Kupffer cell; KEGG, Kyoto Encyclopedia of Genes and Genomes; KO, knockout; Ly6C^hi, classically activated proinflammatory macrophages; Ly6C^lo, alternatively activated reparative macrophages; MoMF, monocyte-derived macrophage; mtDNA, mitochondrial DNA; mtROS, mitochondrial reactive oxygen species; nDNA, nuclear DNA; NK, natural killer; OCR, oxygen consumption rate; p-, phosphorylated; PCNA, proliferating cell nuclear antigen; PHx, partial hepatectomy; qPCR, quantitative PCR; ROS, reactive oxygen species; STAT3, signal transducer and activator of transcription 3; STING, stimulator of interferon genes; TBK1, TANK-binding kinase; TNF, tumor necrosis factor; WT, wild type. Financial support This study was supported by the National Natural Science Foundation of China (82370668, 82071798), Natural Science Foundation of Jiangsu Province (BK20240053), Jiangsu Science and Education Capacity Enhancement Project (ZDXYS202201, CXZX202203), and Postgraduate Research & Practice Innovation Program of Jiangsu Province (KYCX24_2030). Authors’ contributions XZ: study design, conducting experiments, and drafting the manuscript. YB, QZ: conducting experiments, and supplementing the manuscript.YG: acquisition of data, analysis, and interpretation of data. QB, FL: critical revision of the manuscript for important intellectual content. ZZ: technical support. ZR, HZ: study concept and design, analysis, and interpretation of data, drafting the manuscript, obtaining funding, and study supervision. Data availability statement Data are available on reasonable request to the corresponding authors. Conflicts of interest The authors declare no potential conflicts of interest with respect to the research, authorship, and publication of this article. Please refer to the accompanying ICMJE disclosure forms for further details. Acknowledgements