ATG16L1 Regulates Reparative Function of Peritoneal Macrophages During Acute Drug-induced Liver Injury

Abstract

Background & Aims

Acute drug-induced liver injury (DILI) is a major cause of acute liver dysfunction and even liver failure. Peritoneal macrophages have been reported to invade into the injured liver for tissue repair. Herein, we aimed to investigate the role of autophagy-related 16 like 1 gene (ATG16L1) in regulating the reparative function of peritoneal macrophages during DILI caused by acetaminophen (APAP).

Methods

Myeloid ATG16L1 knockout (KO), overexpression (KI) or wild-type (WT) mice were challenged with a single dose of intraperitoneal APAP (300 mg/kg) injection. Intraperitoneal injection or depletion of peritoneal macrophages was conducted for the in vivo analysis. Co-culture of primary hepatocytes and peritoneal macrophages were applied for in vitro analysis.

Results

Peritoneal macrophages were able to rapidly invade into the liver in response to DILI. Peritoneal macrophage injection promoted, and peritoneal macrophage depletion impaired the resolution of inflammation and liver repair post DILI. DILI triggered ATG16L1 expression in intrahepatic accumulated peritoneal macrophages. Interestingly, compared with WT or KI peritoneal macrophages, KO peritoneal macrophages showed enhanced intrahepatic migration ability via Schlafen family member 5 (SLFN5)-CD44 signaling pathway, leading to less injury at early time of 24 hours post DILI in mice with KO peritoneal macrophage infusion. In addition, ATG16L1-mediated autophagy promoted phagocytosis and reparative phenotype of peritoneal macrophages by regulating reactive oxygen species (ROS)-Mer tyrosine kinase (MerTK) signaling. Moreover, peritoneal macrophage ATG16L1 promoted hepatocyte proliferation dependent on the interleukin (IL)-10–C-X-C motif chemokine receptor 2 (CXCR2) axis.

Conclusions

ATG16L1 activation enhanced peritoneal macrophage phagocytosis and reparative phenotype via autophagy-ROS-MerTK signaling and promoted IL-10-CXCR2-dependent hepatocyte proliferation during DILI. Peritoneal macrophage ATG16L1 might be a novel therapeutic target for DILI.

Graphical abstract

Summary.

This study reveals that autophagy-related 16 like 1 gene (ATG16L1) regulates the reparative role of peritoneal macrophages during acetaminophen -induced liver injury. By modulating migration, phagocytosis, and interleukin-10–C-X-C motif chemokine receptor 2-driven hepatocyte proliferation, ATG16L1 enhances inflammation resolution and tissue repair, identifying a promising therapeutic target for drug-induced liver injury.

Acute drug-induced liver injury (DILI) is a major cause of acute liver injury and even liver failure. Acetaminophen (APAP), a widely used antipyretic and analgesic particularly in Western countries, can lead to acute DILI and liver failure in a dose-dependent manner.^1,2 Hallmarks of DILI caused by APAP include extensive hepatocyte death and severe liver inflammation.³ However, the precise underlying mechanisms are not fully understood.

Early mechanistic studies revealed that APAP overdose induces mitochondrial oxidative stress and dysfunction,⁴ triggering hepatocyte necrosis, presenting an important rationale for the clinical use of N-acetylcysteine (NAC).⁵ However, besides mitochondrial oxidative stress, other cellular processes also contribute to APAP-induced acute liver injury, including sterile inflammation, endoplasmic reticulum (ER) stress, autophagy, and microcirculatory dysfunction.6, 7, 8, 9 Although NAC is an effective antidote when administered early, its clinical efficacy is limited in patients presenting late, when N-acetyl-p-benzoquinone imine (NAPQI)-induced glutathione depletion and hepatocellular damage are already advanced, or in cases of massive APAP overdose where the standard NAC regimen cannot fully restore glutathione levels. APAP-induced acute liver injury often progresses rapidly and is difficult to predict. The resolution of inflammation and the subsequent tissue regeneration are key determinants of clinical outcomes.¹⁰ Therefore, understanding the underlying mechanisms is critical to identifying new therapeutic targets and drugs.

Macrophages are the most abundant nonparenchymal cells in the liver, playing pivotal roles in immune defense, tissue repair, and metabolic regulation. Liver macrophages have diverse origins, primarily including liver resident Kupffer cells (KCs), monocyte-derived macrophages (MoMFs),¹¹ and peritoneal macrophages (PMs) that migrate into the liver.¹² Our previous studies demonstrated that KCs and MoMFs have important regulatory roles in liver inflammatory injury.^13,14 The peritoneal cavity, lined with the peritoneum, hosts various immune cells and serves as the first line of defense against inflammation, injury, and tumors. PMs belong to the family of tissue-resident macrophages and exert regulatory functions in multiple diseases.¹⁵ The transcription factor GATA6 is crucial for the specific gene expression, proliferation, and survival of PMs.^16,17 Recent studies have shown that PMs can rapidly migrate and infiltrate the liver following inflammatory injury, adopting an anti-inflammatory phenotype to promote repair.¹⁸ However, the precise regulatory mechanisms of PMs in inflammation resolution and liver regeneration during APAP-induced acute liver injury remain unclear.

Autophagy, a critical cellular stress response, regulates numerous physiological and pathological processes, including acute DILI.¹⁹ The protective mechanisms of autophagy in liver injury primarily involve the clearance of damaged organelles or debris, maintenance of genomic integrity, improved energy supply, and reduction of ER stress. Autophagy can also modulate macrophage immune responses, influencing inflammation and tissue repair.²⁰ Our previous research revealed that impaired autophagy in macrophages led to hyperactivation of the STING-NLRP3 pathway, exacerbating liver inflammation.^21,22 Autophagy-related 16 like 1 gene (ATG16L1) is a key autophagy regulator, and its role in modulating inflammatory diseases is well-established. A single nucleotide polymorphism (SNP) in the ATG16L1 gene increased the risk of Crohn's disease.²³ ATG16L1 deficiency triggered toll-like receptor 4 (TLR4) signaling activation in macrophages, elevating pro-inflammatory interleukin (IL)-1β and IL-18 secretion, thereby aggravating intestinal inflammation.²⁴ Our recent findings indicated that ATG16L1 deficiency inhibited lipid autophagy in macrophages, exacerbating metabolic dysfunction-associated steatohepatitis (MASH).²⁵ However, the role of ATG16L1 in regulating PM immune responses and its involvement in liver repair after APAP-induced liver injury remain largely unknown.

In the present study, we investigated the role of ATG16L1 in regulating peritoneal macrophages during liver regeneration following APAP-induced acute liver injury. We found that ATG16L1 affected the resolution of inflammation by modulating the intrahepatic migration and phagocytosis by peritoneal macrophages. Moreover, ATG16L1 promoted hepatocyte proliferation via IL-10/C-X-C motif chemokine receptor 2 (CXCR-2) signaling. Our findings provide new insights into the role of peritoneal macrophages in liver repair and offer a potential therapeutic target for clinical intervention.

Results

Intrahepatic Migration of PMs Promoted Liver Repair Post DILI

First, we evaluated whether PMs could accumulate into the liver in response to APAP-induced DILI. Liver tissues were collected 24 hours post DILI, and intrahepatic PMs were detected by immunostaining of F4-80 and GATA6. Indeed, F4-80 and GATA6 double-positive PMs were found in livers as shown in Figure 1A and B, indicating a rapid invasion of PMs in response to DILI.

Figure 1.

Intrahepatic migration of PMs promotes liver repair post-DILI. Liver tissues from APAP-treated mice were collected at 24 hours post DILI. F4/80 and GATA6 co-staining was analyzed by IHC (A) and IF (B). PMs were isolated and intraperitoneally injected into APAP-pretreated mice or depleted prior to APAP-induced DILI. Liver tissues and serum were collected at 24, 48, and 72 hours post DILI. Representative H&E (C), TUNEL (D), CD31 (E) staining, serum ALT (F) and AST (G), IL-1β (H–I), and IL-10 (J–K) levels were analyzed; furthermore, representative IF staining of CD11b (L) and Ly6G (M) was analyzed. n = 6/group. Data are shown as mean ± SEM. Each point represents an independent experiment. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

The function of PMs on regulating APAP-induced liver injury was then evaluated. PMs were isolated and then injected into the peritoneal cavity of mice after DILI. Interestingly, PM injection protected against APAP-induced liver injury at 24, 48, and 72 hours as shown by better preserved liver architecture (Figure 1C), less cell death (Figure 1D), enhanced revascularization (Figure 1E), and lower serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels (Figure 1F and G). Moreover, PM injection suppressed liver inflammation as indicated by decreased proinflammatory IL-1β (Figure 1H and I) and increased anti-inflammatory IL-10 (Figure 1J and K). In contrast, PM depletion aggravated liver injury and intrahepatic inflammation (Figure 1C–K). CD11b+ monocytes increased in the liver at 24 hours post-APAP and declined toward baseline by 48 to 72 hours. Compared with the corresponding control groups, the depletion of PMs resulted in an increase in hepatic monocyte numbers at 24 hours, whereas the infusion of PMs led to a reduction in hepatic monocytes (Figure 1L). Ly6G+ neutrophils infiltrated the liver at 24 hours and declined by 48 to 72 hours. Their numbers were further increased at 24 hours by PM depletion (Figure 1M). These findings confirmed that PMs could migrate into the liver and function to promote the resolution of inflammation and liver repair.

To determine whether depletion of PMs influences the early phase of liver injury, we evaluated earlier time points at 6 and 12 hours. Histological analysis revealed increased hepatic necrosis in clodronate liposome (CLL)-treated mice at 12 hours after APAP administration (Figure 2A). Consistently, serum ALT and AST levels were significantly higher in the APAP + CLL group compared with APAP alone (Figure 2B–C). Moreover, depletion of PMs elevated both protein and mRNA expression of the pro-inflammatory cytokine IL-1β (Figure 2D–E), while markedly reducing protein and transcript levels of the anti-inflammatory cytokine IL-10 (Figure 2F–G). To exclude the potential confounding effect of CLL-mediated depletion of hepatic KCs, we confirmed that low-dose intraperitoneal CLL administration did not alter the number of hepatic F4/80⁺Clec4f⁺ Kupffer cells at day 7, as determined by dual-color immunofluorescence (Figure 2H).

Figure 2.

Depletion of PMs exacerbates APAP-induced liver injury in the early phase. Mice with depleted PMs and control mice were treated with APAP using the standard protocol. Liver and serum samples were collected at 6 and 12 hours thereafter. Representative H&E (A) staining, serum ALT (B) and AST (C), IL-1β (D–E), and IL-10 (F–G) levels were analyzed. Following intraperitoneal injection of control or CLL (100 μL), livers from WT mice were collected at day 7 post-injection. Tissue sections underwent dual-color IF staining for F4/80 and Clec4f, with the proportion of F4/80⁺Clec4F⁺ cells quantified by confocal microscopy analysis (H). n = 6/group. Data are shown as mean ± SEM. Each point represents an independent experiment. ∗P < .05, ∗∗P < .01.

ATG16L1 Affected the Pro-repair Effects of PMs

Next, we investigated the underlying regulatory molecular signals that affect the pro-repair function of peritoneal macrophages. Recently we found that ATG16L1 regulated the inflammatory response of bone marrow-derived macrophages in a MASH model.²⁵ Therefore, we questioned whether ATG16L1 signaling was involved in modulating PM functions as well. Indeed, the expression of ATG16L1 was elevated in GATA6-positive (Figure 3A) or CD45.2-positive (Figure 3B and C) intrahepatic accumulated PMs after DILI. To further determine the essential role of ATG16L1, myeloid ATG16L1 deficiency (KO) or overexpression (KI) mice were generated, which were confirmed by Western blotting (WB) analysis (Figure 3D). PMs isolated from wild-type (WT), KO, or KI mice were injected into the peritoneal cavity of WT mice after APAP treatment, and then liver repair was measured. Interestingly, the transfusion of WT, KO, or KI PMs all promoted liver repair at 24, 48, and 72 hours post DILI (Figure 3E–M; group APAP vs APAP + WT-PMs, APAP + KO-PMs, or APAP + KI-PMs). Moreover, in contrast to KO PMs, transfusion of KI PMs showed better protective effects at 48 and 72 hours post DILI (Figure 3E–M; group APAP + KO-PMs vs APAP + KI-PMs, 48 and 72 hours). Unexpectedly, at 24 hours post DILI, mice with KO PM transfusion exhibited alleviated inflammatory liver injury as compared with that of KI PMs (Figure 3E–M; group APAP + KO-PMs vs APAP + KI-PMs, 24 hours). At 24 hours, adoptive transfer of ATG16L1-KO macrophages reduced monocyte numbers vs WT; whereas ATG16L1-KI transfer showed no difference. By 48 to 72 hours, monocyte counts converged (Figure 3N). In contrast, neutrophil counts were unchanged by transfer of any macrophage genotype (WT, KO, or KI) (Figure 3O). These results suggested that ATG16L1 signaling regulated the pro-repair effects of PMs.

Figure 3.

ATG16L1 influences the pro-repair effects of PMs. Liver tissues from APAP-treated mice were collected at 24 hours post DILI. Immunostaining for F4/80, GATA6, ATG16L1, and DAPI (A). CD45.2 PMs were injected intraperitoneally into CD45.1 mice that had been pretreated with APAP. After 24 hours, the intrahepatic NPCs were collected. To isolate the migrated CD45.2 macrophages specifically, they were purified from the liver sample using CD45.2-based magnetic sorting. Stable CD45.2 PMs were used as the control group. The samples were analyzed by flow cytometry (B) and WB for ATG16L1 (C); n = 3/group. WB analysis of ATG16L1 in peritoneal macrophages (D). PMs were injected into APAP-pretreated WT mice, and liver tissues/serum were collected at 0, 24, 48, and 72 hours post DILI. H&E (E), TUNEL (F), CD31 (G), ALT (H), AST (I), IL-1β (J–K), and IL-10 (L–M) analyses. Furthermore, representative IF staining of CD11b (N) and Ly6G (O) was evaluated. n = 6/group. Data are shown as mean ± SEM. Each point represents an independent experiment. ∗P < .05, ∗∗P < .01, ∗∗∗∗P < .0001.

To assess whether PM infusion or ATG16L1 modulation influences the severity of liver injury during the early phase, we evaluated hepatic damage at 18 hours after APAP challenge, corresponding to 6 hours following PM transfer in our model. Histological analysis revealed no significant differences in hepatic necrosis among mice receiving WT, ATG16L1-KO, or ATG16L1-KI PMs compared with APAP alone (Figure 4A). Consistently, serum ALT and AST levels showed no significant changes across groups (Figure 4B–C). Furthermore, protein and mRNA expression of the pro-inflammatory cytokine IL-1β (Figure 4D–E) and the anti-inflammatory cytokine IL-10 (Figure 4F–G) remained comparable regardless of PM infusion or ATG16L1 status. These findings indicate that PM transfer, irrespective of ATG16L1 expression, does not affect the severity of APAP-induced liver injury at this early time point.

Figure 4.

Neither infusion of PMs nor knockout or overexpression of ATG16L1 in them affects the severity of DILI within 6 hours. As per our protocol, liver tissues and serum were collected 18 hours after DILI was established in WT mice, which corresponded to 6 hours after the injection of PMs. Representative H&E (A) staining, serum ALT (B) and AST (C), IL-1β (D–E), and IL-10 (F–G) levels were analyzed. n = 6/group. Data are shown as mean ± SEM. Each point represents an independent experiment.

ATG16L1 Regulated Migration of PMs via SLFN5-CD44 Signaling

We further questioned whether the stronger protective role of KO PM transfusion at early time post DILI was due to the faster intrahepatic infiltration. WT, KO, or KI PMs (CD45.2-positive) were transperitoneal injected into the receipt mice (CD45.1-positive) with DILI, and then the intrahepatic CD45.2-positive macrophages were counted at 24 and 48 hours post DILI. Indeed, as compared with WT PMs, significantly more CD45.2-positive KO macrophages but less CD45.2-positive KI macrophages were detected at 24 hours post DILI (Figure 5A), whereas comparable numbers of CD45.2-positive WT, KO, and KI macrophages were found at 48 hours post DILI (Figure 5B). Meanwhile, WT, KO, and KI mice were subjected to APAP-induced DILI, and PMs that remained in the peritoneal cavity were isolated and counted as well. As shown in Figure 5C, at 24 hours post DILI, significantly less PMs were obtained from KO mice, whereas much more were obtained from KI mice. Comparable numbers of PMs were detected in all groups at 48 hours post DILI. These results indicated that ATG16L1 depletion promoted, but ATG16L1 overexpression suppressed the migration ability of PMs, leading to more intrahepatic infiltrated PMs and attenuated liver inflammatory injury at early time post DILI.

Figure 5.

ATG16L1 controls PM migration via SLFN5–CD44, with cells localized to HA-rich areas. PMs from CD45.2 mice were injected into APAP-pretreated CD45.1 mice. The intrahepatic NPCs were analyzed for CD45.2/F4/80 double-positive cells at 24 hours (A) and 48 hours (B) post DILI; n = 3/group. PMs remaining in the cavity were counted at 24 hours and 48 hours post DILI (C); n = 12/group. CD45.2 PMs were injected into CD45.1 APAP-induced mice pretreated with PTX, CD44 inhibitor, or HA, and NPCs were analyzed (D); n = 3/group. WT mice were subjected to APAP treatment, and liver tissues and serum were collected at 0, 24, 48, and 72 hours post DILI. (E) Fluorescence staining of HABP in liver tissues and (F) ELISA detection of HA in serum. PMs were isolated from GFP mice and intraperitoneally injected into WT mice without GFP expression followed by APAP treatment, and liver tissues were collected at 24 hours post DILI. (G) IF analysis of GFP, HABP, and DAPI in liver sections. WB (H) and qPCR (I) assessed CD44 expression in peritoneal macrophages. Transwell migration assays of PMs (J); n = 6/group. RNA-seq (K, L) and WB (M) identified key genes and proteins in PMs, with SLFN5/CD44 interactions confirmed via WB (N), qPCR (O) and co-immunoprecipitation (P); n = 3/group. Data are shown as mean ± SEM. Each point represents an independent experiment. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

We further determined the molecular mechanism of ATG16L1 in regulating PM migration. Galphai-protein coupled receptors, hyaluronate sodium (HA), and CD44 have been reported to regulate macrophage migration.¹⁸ Therefore, mice with DILI were treated with pertussis toxin (PTX), CD44 inhibitor, or HA prior to PM injection. Inhibition of CD44 or HA significantly suppressed intrahepatic PM migration, whereas PTX showed no markable effect (Figure 5D). Moreover, APAP treatment induced HA expression in liver tissues and blood (Figure 5E and F), and PMs were predominantly detected around HA-expressing areas (Figure 5G). ATG16L1 depletion upregulated and ATG16L1 overexpression downregulated protein and gene expression of CD44 (Figure 5H and I). Moreover, CD44 inhibition abrogated the role of ATG16L1 in regulating peritoneal macrophage migration in vitro at both 12 and 24 hours (Figure 5J). Mechanistically, RNA sequencing (RNA-seq) analysis showed that KO peritoneal macrophages had decreased gene expression of Schlafen family member 5 (SLFN5), OAS3, TRIM21, EIF2AK2, and DHX58 (Figure 5K and L) and significant decreased protein levels of SLFN5 (Figure 5M). Interestingly, SLFN5 suppression by small interfering RNA (siRNA) upregulated protein and gene expression of CD44 in WT PMs (Figure 5N and O). The immunoprecipitation analysis further confirmed the interaction between SLFN5 and CD44 (Figure 5P). Together, these findings suggested that ATG16L1 depletion promoted PM intrahepatic migration via SLFN5-CD44 signaling.

ATG16L1-mediated Autophagy Promoted PM Phagocytosis and IL-10 Expression Through Regulating ROS-MerTK Signaling

Macrophage phagocytosis has been implicated in regulating hepatic inflammation and repair. Interestingly, ATG16L1 depletion deceased, but ATG16L1 overexpression increased Mer tyrosine kinase (MerTK) expression in infiltrated PMs post DILI (Figure 6A and B). However, ATG16L1 showed no significant effects on both the protein and gene levels of AXL, TYRO3, and TIM4 (Figure 6A and C–E). In vitro, ATG16L1 deficiency impaired PM phagocytosis. In contrast, ATG16L1 overexpression enhanced PM phagocytosis, which was abrogated by MerTK inhibition by all-trans retinoic acid (ATRA) (Figure 6F). Moreover, ATG16L1 depletion suppressed but ATG16L1 overexpression increased IL-10 expression of PMs after in vitro cell phagocytosis, which was abrogated by MerTK inhibition (Figure 6G and H). MerTK inhibition by MerTK-siRNA effectively impaired the pro-repair effect of ATG16L1 KI peritoneal macrophage during DILI (Figure 6I).

Figure 6.

ATG16L1 promotes PM phagocytosis by activating MerTK. CD45.2 PMs were injected into APAP-pretreated CD45.1 mice. At 24 hours post DILI, the NPCs were obtained, and CD45.2-positive macrophages were sorted using CD45.2 magnetic beads. WB (A) and qPCR (B–E) analyzed MerTK, AXL, TYRO3, and TIM-4 expression. PMs stimulated with HA or HA+ATRA were labeled with CMFDA and co-cultured with pHrodo-labeled dead hepatocytes. (F) IF for pHrodo/CMFDA double-positive cells. PMs and supernatant were collected after in vitro cell phagocytosis. IL-10 levels (G) and gene expression (H). KI PMs pretreated with MerTK-siRNA or NS-siRNA were injected into DILI mice. (I) H&E staining of liver tissues at 0, 24, 48, and 72 hours post DILI. (n = 6/group). Data are shown as mean ± SEM. Each point represents an independent experiment. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

Interplay of autophagy and reactive oxygen species (ROS) has been implicated in regulating MerTK-mediated macrophage phagocytosis.²¹ Indeed, compared with PMs without DILI, infiltrated PMs isolated from livers post DILI showed increased activation of autophagy as indicated by enhanced LC3 staining and increased protein levels of LC3II/I ratio (Figure 7A and B). ATG16L1 depletion downregulated but ATG16L1 overexpression upregulated autophagy activation of PMs (Figure 7C). Autophagy inhibition suppressed PM phagocytosis (Figure 7D) and enhanced oxidative stress of WT and KI PMs (Figure 7E). ROS scavenge by NAC restored MerTK activation (Figure 7F) and phagocytosis by WT and KO PMs (Figure 7G). Functionally, mice with infusion of PMs with autophagy inhibitor pretreatment showed increased necrosis (Figure 8A), necrotic cells (Figure 8B), impaired angiogenesis (Figure 8C), higher levels of serum ALT and AST (Figure 8D and E), increased IL-1β (Figure 8F and G) but decreased IL-10 (Figure 8H and I) expression at 24, 48, and 72 hours post DILI. Therefore, ATG16L1-mediated autophagy activation promoted PM phagocytosis and IL-10 expression to facilitate liver repair post DILI.

Figure 7.

ATG16L1-mediated autophagy promotes PM phagocytosis via ROS-MerTK signaling. WT mice underwent APAP-induced DILI or control group without DILI. (A) IF for F4/80, GATA6, and LC3 in liver tissues at 24 hours post DILI. CD45.2 PMs were injected into APAP-treated CD45.1 mice, and intrahepatic CD45.2 PMs were sorted at 24 hours. (B) WB analysis of p62, and LC3-I/II in PMs isolated from control group and in intrahepatic CD45.2 PMs sorted from APAP group. WB analysis of p62, and LC3-I/II in PMs (C). CMFDA-labeled PMs pretreated with DC-LC3in-D5 co-incubated with pHrodo-labeled hepatocytes showed phagocytosis (D). ROS detection by DCFH staining (E). CMFDA-labeled peritoneal macrophages were pretreated with NAC and then co-cultured with pHrodo-labeled dead hepatocytes. MerTK expression (F) and phagocytosis (G) were evaluated. n = 6/group. Data are shown as mean ± SEM. Each point represents an independent experiment. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

Figure 8.

Inhibition of autophagy limits the pro-repair effects of PMs in APAP-induced DILI. PMs pretreated with DC-LC3in-D5 were intraperitoneally injected into APAP-treated recipient mice. Liver tissues and serum were collected at 0, 24, 48, and 72 hours post DILI. Representative H&E staining (A), TUNEL staining (B), and CD31 staining (C) of liver tissues were performed. Serum levels of ALT (D) and AST (E), IL-1β (F), and IL-10 (H) were measured, along with IL-1β (G) and IL-10 (I) expression in liver tissues. (n = 6/group). Data are presented as mean ± SEM. Each point represents an independent experiment. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001.

PM ATG16L1 Promoted Hepatocyte Proliferation via IL-10-CXCR2 Axis

Finally, we evaluated the direct role of PM ATG16L1 signaling in regulating hepatocyte proliferation. Indeed, treatment with PM injection promoted proliferating cell nuclear antigen (PCNA) and CXCR2 expression in livers post DILI, which was downregulated by ATG16L1 depletion and upregulated by ATG16L1 overexpression (Figure 9A–D). In vitro, WT, KO, or KI PMs stimulated by HA were, respectively, co-cultured with hepatocytes, and then PCNA and CXCR2 expression of hepatocytes was measured. As compared with WT controls, hepatocytes cocultured with KO PMs exhibited decreased PCNA and CXCR2 expression, whereas KI PMs increased PCNA and CXCR2 expression of hepatocytes (Figure 9E and F). Moreover, ATG16L1 depletion suppressed but ATG16L1 overexpression promoted IL-10 expression of PMs stimulated by HA. ROS scavenge further upregulated (Figure 9G and H) and autophagy inhibition downregulated (Figure 9I and J) IL-10 expression of PMs, respectively. IL-10 inhibition suppressed CXCR2 and PCNA expression of hepatocytes co-cultured with PMs (Figure 9K–M).

Figure 9.

ATG16L1 in PMs promotes hepatocyte proliferation via the IL-10-CXCR2 axis. PMs were injected into APAP-treated WT recipient mice. Liver tissues were collected at 24 and 48 hours post DILI. IHC for PCNA (A). WB (B) and qPCR (C, D) for CXCR2 and PCNA expression in liver tissues at 48 hours post DILI. Primary hepatocytes were co-cultured with PMs stimulated with HA. qPCR for CXCR2 and PCNA in hepatocytes (E, F). PMs were stimulated by HA combined with NAC or DC-LC3in-D5. IL-10 levels were assessed by qPCR and ELISA (G–J). IL-10 neutralizing antibody was added into the co-cultured primary hepatocytes and PMs stimulated with HA. Expression of CXCR2 and PCNA in hepatocytes (K–M). (n = 6/group). Data are presented as mean ± SEM. Each point represents an independent experiment. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

Discussion

Drug-induced acute liver injury remains one of the most challenging hepatic disorders, with APAP overdose as the leading cause.^26,27 Recent studies have emphasized the critical role of PMs in both resolving inflammation and promoting liver regeneration.^28,29 In this study, we investigated the role of ATG16L1 in regulating PM function during liver regeneration following APAP-induced DILI. We found that GATA6⁺ PMs rapidly migrate and infiltrate the liver after DILI, contributing to liver regeneration and repair, with ATG16L1 signaling playing a regulatory role in several aspects. Specifically, ATG16L1 deficiency promoted intrahepatic PM migration but inhibited phagocytosis, resulting in delayed inflammation resolution. Moreover, ATG16L1-deficient PMs exhibited reduced IL-10 secretion, leading to impaired hepatocyte proliferation. Thus, ATG16L1 in PMs represents a promising therapeutic target for enhancing liver repair post-DILI.

Macrophages play a dual role in APAP-induced liver injury, contributing to both hepatic damage and subsequent regeneration. After APAP overdose, macrophages are recruited to the liver, where they release pro-inflammatory cytokines such as tumor necrosis factor (TNF)-α and IL-1β, exacerbating hepatocyte necrosis and tissue injury. However, macrophages also facilitate liver regeneration by secreting growth factors like IL-6 and hepatocyte growth factor (HGF), promoting hepatocyte proliferation and tissue repair. PMs, tissue-resident macrophages found in the peritoneal cavity, serve as the first line of defense against pathological conditions, including peritoneal injuries, abdominal sepsis, or metastatic tumor growth.¹⁵ Although PMs have been shown to invade injured organs and promote tissue repair rapidly,^18,30 their precise role and underlying regulatory mechanisms in liver regeneration following APAP-induced DILI remain unclear.

Multiple hepatic lineages, including monocytes and neutrophils, shape APAP injury and repair.^31,32 Although our study centers on PMs, we recognize bidirectional crosstalk with these populations. Because PM transfer occurred 12 hours after APAP, we quantified CD11b+ monocytes and Ly6G+ neutrophils by immunofluorescence under PM depletion/transfer and ATG16L1 manipulation.³³ CD11b+ monocytes rose at 24 hours and declined by 48 to 72 hours; PM depletion further increased 24-hour monocytes, whereas transfer of ATG16L1-KO PMs reduced them relative to WT, with convergence by 48 to 72 hours. Ly6G+ neutrophils peaked at 24 hours and normalized thereafter; PM depletion increased 24-hour neutrophils, but PM transfer (WT/KO/KI) had minimal effect, likely because neutrophil influx peaks at 12 hours.³⁴ Together, these data suggest PM abundance and ATG16L1-dependent PM function tune early monocyte recruitment, with more limited effects on neutrophils at the time points examined.^33,35, 36, 37

The ability of macrophages to migrate to sites of liver injury is crucial for their roles in both exacerbating and resolving tissue damage. PMs rapidly adhere to damaged tissue and migrate into the injury site. GATA6⁺ PMs are recruited into the liver within 1 hour of injury by crossing the mesothelium.¹⁸ ATP released from necrotic hepatocytes is sensed by PMs via PX27, whereas the binding of CD44 and hyaluronan at the damaged site triggers further PM infiltration into the liver parenchyma. Depletion of PMs using CLLs impaired tissue repair after liver and intestinal injuries.^18,30 Interestingly, we found that ATG16L1 deficiency promotes PM migration via the SLFN5/CD44 signaling pathway, whereas no significant changes in CD44 expression were observed in ATG16L1-deficient dendritic cells.³⁸ Consistent with these findings, our study also demonstrated that PM depletion exacerbated liver injury and intrahepatic inflammation, highlighting PM modulation as a potential therapeutic strategy for mitigating liver damage and promoting regeneration.

Monocyte-derived macrophages dominate the early (8–24 hour) APAP phase, and CCR2 blockade limits their infiltration and injury.³⁹ The magnitude of our PM-directed manipulations was modest, which we attribute to pool size and timing: the peritoneal cavity contains only 2–4 × 10⁶ GATA6⁺ LPMs; transferred PMs made up about 20% of hepatic macrophages in our migration assays; and we delivered 2 × 10⁶ PMs at 12 hours—more likely to influence resolution than the initial metabolic injury. Even so, PMs can shape monocyte/neutrophil recruitment, access the liver via capsular “nonvascular” routes¹⁸ and display strong phagocytic capacity,^40,41 supporting roles in later repair. Our data therefore clarify how ATG16L1 programs PM function within this multicellular orchestration.

In our model, APAP was administered intraperitoneally, a validated and commonly applied route.^42,43 Pharmacokinetics show rapid absorption and systemic distribution within 180 minutes,⁴⁴ with APAP-CYS peaking at 2 hours⁴⁵ and low plasma APAP by 12 hours.⁴⁶ To minimize direct APAP effects on PMs, we scheduled PM transfer 12 hours post-APAP, beyond the absorption phase. Nevertheless, we agree that residual local exposure may still affect PM function.

To study the early effects of PMs on APAP-induced injury, we depleted PMs and observed aggravated liver injury at 12 hours post-APAP, with increased necrosis, ALT/AST, and IL-1β, and reduced IL-10 (Figure 2). At 18 hours (6 hours after PM transfer), however, necrosis and cytokine levels were comparable between PM-injected and controls (Figure 4), likely due to limited engraftment and function within that short window. These findings suggest PMs can influence early APAP injury, but timing and kinetics are critical. Although our in vivo transfer and hepatocyte co-culture assays minimized confounding by early injury, we cannot fully exclude indirect effects on regeneration. Future time-controlled and cell-specific ATG16L1 knockouts (eg, DTR-based) will provide more definitive insights.

Autophagy, a crucial cellular process for maintaining homeostasis, significantly influences macrophage function during liver injury and repair. Impaired autophagy leads to increased oxidative mitochondrial damage and a self-mtDNA-mediated pro-inflammatory response in macrophages, aggravating liver injury.^21,22 In macrophages, miR223 deficiency has been shown to increase ATG16L1 expression, thereby ameliorating central nervous system inflammation.⁴⁷ Similarly, ATG16L1 deficiency resulted in elevated ROS production, reduced microbial clearance, and worsened colitis.⁴⁸ In our study, we found that macrophage autophagy, particularly through ATG16L1, was essential for controlling the severity of APAP-induced liver injury.

Human data underscore the clinical relevance of ATG16L1. The T300A variant marks a Crohn’s disease subtype with Paneth-cell endoplasmic-reticulum stress and bacterial persistence, implying altered antimicrobial function.⁴⁹ In a Guangxi cohort, rs4663402 and rs4663396 showed no association with antineutrophil cytoplasmic antibody-associated vasculitis, although rs4663396 CT+TT correlated with arthralgia.⁵⁰ In a phase III trial of advanced microsatellite-stable colorectal cancer, higher ATG16L1 expression predicted poorer immunotherapy outcomes in KRAS-mutant tumors.⁵¹ Collectively, these observations position ATG16L1 as a potential biomarker and therapeutic node across inflammatory, autoimmune, and neoplastic disease.

The timely and effective clearance of dead cells and cellular debris is crucial for maintaining tissue homeostasis. The phagocytosis of necrotic cells by macrophages is key to inflammation resolution.⁵² We previously found that defective phagocytosis in aged macrophages promoted inflammatory liver injury.⁵³ This phagocytic process involves a well-orchestrated signaling network, including specialized phagocytic receptors, bridging molecules, and “find-me” and “eat-me” signals. MerTK, a family of receptor tyrosine kinases (TAM) receptor, is involved in phagocytosis by recognizing phosphatidylserine (PS) on the surface of necrotic cells.^54,55 MerTK can be cleaved by A disintegrin and metalloproteinase 17 (ADAM17), and its cleavage product, soluble MER, competes for GAS6, inhibiting phagocytosis. ROS/ADAM17-mediated downregulation of MerTK impairs macrophage phagocytosis.⁵³ In our study, we found that ATG16L1 promotes MerTK activation by limiting ROS/ADAM17 signaling.

Hepatocyte proliferation and liver regeneration are driven by complex mitogens and signaling pathways.⁵⁶ Complete mitogens like HGF and EGFR ligands can induce liver enlargement and hepatocyte replication in serum-free, chemically defined media. However, the absence of auxiliary mitogens such as TNF, IL-6, and noradrenaline may delay liver regeneration. CXCR2, a chemokine receptor expressed by hepatocytes, neutrophils, and macrophages, plays a controversial role in APAP-induced liver injury. Both CXCR2 antagonists and agonists have been reported to protect against liver injury and enhance regeneration.^57,58 A recent study showed that IL-10 released by KCs induces CXCR2 expression on surviving hepatocytes, with CXCR2hi hepatocytes expressing pro-regenerative genes to initiate liver repair.⁵⁹ IL-10, a key anti-inflammatory cytokine, modulates the hepatic microenvironment during both acute and chronic liver injuries.⁶⁰ We previously found that macrophage-derived IL-10 suppressed inflammatory liver injury.¹³ In this study, we further demonstrated that ATG16L1-deficient macrophages produce lower levels of IL-10, resulting in reduced CXCR2 expression on hepatocytes, impaired hepatocyte proliferation, and delayed tissue repair, indicating a cross-talk between hepatocytes and PMs during APAP-induced DILI.

Conclusions

In conclusion, our study highlights the critical role of ATG16L1 in regulating APAP-induced acute liver injury and regeneration by modulating PM migration, phagocytosis, and cytokine production. Targeting ATG16L1 and related pathways may offer promising therapeutic strategies for enhancing liver regeneration while minimizing liver injury in the context of APAP-induced hepatotoxicity.

Materials and Methods

Animals and Strains

Male C57BL/6J WT recipients (6–8 weeks old) were used to establish the APAP–induced acute liver injury model. Donor PMs were obtained from age-matched (6–8 weeks old) WT, FloxP-Atg16L1 (Atg16L1fl/fl, WT), Lyz2-Cre Atg16L1-knockout (Atg16L1ΔMφ, KO), and myeloid-specific Atg16L1-overexpressing-knockin (Atg16L1OE, KI) male mice; all strains were on a C57BL/6J background. For tracking experiments, CD45.1 recipients (C57BL/6JGpt-Ptprcem1Cin(p.K302E) /Gpt) received CD45.2 donor cells, and a Cre-dependent EGFP reporter line (C57BL/6JGpt-Rosa26-CAG-LSL-EGFP-3xHA-WPRE-polyA) was used to verify recombination where indicated. Unless otherwise stated, all donor mice used in this study were CD45.2 (C57BL/6J). Mice were maintained under specific pathogen-free conditions (12-hour light/12-hour dark) with ad libitum access to water and chow; all procedures were approved by the Institutional Animal Care and Use Committee of Nanjing Medical University (NMU08-092). Genotyping was performed by polymerase chain reaction (PCR); primer sequences are provided in Table 1.

Table 1.

The Primers Used to Genotype the Macrophage-specific Transgenic Overexpression or Knockout Mice

Genotype	Primer	Sequence 5′-3′
ATG16L1ΔMф	F1	CTTTCTGGTCACCTTAGTGAAACCC
	R1	CAGTGGCAAATCTTGAGCCTTTATG
	F2	CTTTCTGGTCACCTTAGTGAAACCC
	R2	AAGCAGGCAGAAGGGTCTTTGAG
	F3	CCCAGAAATGCCAGATTACG
	R3	CTTGGGCTGCCAGAATTTCTC
	F4	CTTGGGCTGCCAGAATTTCTC
	R4	TTACAGTCGGCCAGGCTGAC
ATG16L1OE	F1	TGGCTGCCATGAACAAAGGTTG
	R1	GACCCACTATTCTGCACCACTCATTAG
	F2	CAGCAAAACCTGGCTGTGGATC
	R2	ATGAGCCACCATGTGGGTGTC
	F3	AGTGCTGAAGTCCATAGATCGG
	R3	GTCACTCACTGCTCCCCTGT
	F4	AGTGCTGAAGTCCATAGATCGG
	R4	CTGATTCTCCTCATCACCAGG

APAP Injury and Adoptive Transfer of PMs

At time = 0 hours, WT recipients received APAP (300 mg/kg, intraperitoneal, HY-66005, MCE) freshly dissolved in warm phosphate-buffered saline (PBS); control mice received equal-volume PBS. Unless otherwise specified, mice were fasted before APAP. Mice were euthanized at 24, 48, or 72 hours after APAP for sample collection. At time = 12 hours after APAP, donor PMs were isolated by peritoneal lavage from the indicated donor mice (WT, Atg16l1 ΔMφ, or Atg16l1 OE), counted, and assessed for viability; 2 × 10⁶ cells were resuspended in 200 μL PBS and injected intraperitoneally into APAP-treated recipients; vehicle controls received PBS. In congenic experiments, CD45.1 recipients received CD45.2 donor macrophages prepared as above to enable donor–recipient discrimination by flow cytometry. For macrophage depletion studies, mice were administered CLL (0.1 mL, intraperitoneally CLL, CLD-8914, Encapsula NanoSciences) or control liposomes 7 days before APAP.¹⁸ Thus, the timeline is APAP at time = 0 hours, adoptive transfer at time = 12 hours, and endpoint collections at time = 24/48/72 hours (ie, 12/36/60 hours post-transfer).

Hepatocellular Function Assay

Blood samples were centrifuged to obtain serum, and the levels of ALT and AST were measured using an automatic chemical analyzer from Olympus.

Extraction, Culture, and Counting of Mouse PMs

Mice were euthanized by cervical dislocation following orbital blood collection. The peritoneal cavity was flushed with 10 mL of ice-cold PBS, and the lavage fluid was collected. The lavage fluid was centrifuged at 1000 rpm for 5 minutes, and the supernatant was discarded. The cells were resuspended in 2 mL of Dulbecco’s Modified Eagle Medium (DMEM) (including 10% fetal bovine serum [FBS], 10 mM HEPES) and plated. After 1 hour, adherent cells were digested and counted.

In certain vitro experiments, we used ATRA (R2625, Sigma Aldrich, 10 μM, 3 h), NAC (HY-B0215, MCE, 8 μM, 12 h), DC-LC3in-D5 (HY-141882, MCE, 3 μM,12h), and/or HA (EHBY-3000, FishReag, 3000 kDa, 1 mM,12h) to stimulate PMs.

Isolation of Liver Cells

The liver was perfused in situ through the portal vein using calcium- and magnesium-free Hanks’ Balanced Salt Solution (HBSS) supplemented with 2% heat-inactivated FBS, followed by perfusion with 0.27% collagenase IV (Sigma-Aldrich). After perfusion, the liver was dissected and processed through a 70-μm nylon mesh cell strainer (BD Biosciences). The liver cells containing hepatocytes and nonparenchymal cells (NPCs) were resuspended in 20 mL of DMEM with 10% FBS.

To separate the cells, the suspension was centrifuged at 50 g for 2 minutes, and the supernatant containing NPCs was collected. This process was repeated 3 times. The collected NPCs were then isolated by centrifugation at 800 g for 5 minutes and resuspended in DMEM. The cell suspension was subsequently used for flow cytometry or magnetic bead sorting.

For primary hepatocytes, the cell suspension was centrifuged at 50 g for 2 minutes. The pellet was resuspended in 20 mL of 40% cold Percoll solution (P1644, Sigma) and centrifuged at 150 g for 7 minutes. The hepatocytes were then washed once with DMEM containing 10% FBS and plated onto collagen type I-coated plates.¹³

Sorting of CD45.2-positive Cells From the Livers of CD45.1 Mice

After in situ perfusion and liver digestion, NPCs were isolated, and liver-infiltrated PMs were sorted using a kind of MagniSort Mouse CD45.2 Positive Selection Kit (8802-6849-74, Invitrogen) according to the manufacturer’s protocols, which is designed to enable magnetic separation of CD45.2+ cells by positive selection. Briefly, it is optimized to isolate CD45.2+ cells from mouse liver using biotinylated anti-mouse CD45.2 antibody and streptavidin-coated magnetic beads. CD45.2+ cells bind to the antibody and subsequently to the magnetic beads. By placing them in a magnetic rack, non-target cells can be separated from CD45.2+ cells by decantation.

Stimulation of Primary Hepatocytes

After obtaining primary hepatocytes and culturing them in DMEM (including 10% FBS) for 3 hours, the medium was changed and stimulated with APAP (10 mmol/L for 6 hours) to induce cell death.

In Vitro Phagocytosis Assay

PMs were plated in confocal dishes at a concentration of 2 × 10⁵ cells/mL, and the required stimulation (HA combined with or without ATRA/DC-LC3in-D5/NAC) was added. The macrophages were washed twice and incubated with 1 μM CMFDA-green dye (40721ES72, Yeasen Biotechnology) for 30 minutes.

Hepatocytes were treated with APAP to induce cell death and then labeled with 20 ng/mL pHrodo-red (P36600, Thermo Fisher Scientific) for 30 minutes.

The labeled dead hepatocytes and CMFDA-green-labeled PMs were co-incubated at a 1:1 ratio for 45 minutes. The fluorescence intensity of pHrodo-red and the percentage of cells double-positive for pHrodo-red and CMFDA-green fluorescence were analyzed.

PM Isolated and Supernatant Collection After In Vitro Phagocytosis

PMs were seeded in 6-well plates at a density of 5 × 10⁵ cells per well. After 1 hour, the medium was replaced, and the required experimental stimuli (HA or HA combined with ATRA) were applied, then, dead hepatocytes were added to the macrophages at a 1:1 ratio and co-incubated for 45 minutes. The cells were then washed with PBS, harvested, and centrifuged at 50 g for 2 minutes (repeated 3 times) to collect the macrophage-containing supernatant. We resuspended the separated macrophages by centrifuging at 800 g for 5 minutes to form a pellet. Then, we resuspended the pellet in DMEM (containing 10% FBS and 10 mM HEPES), and plated the cells for 15 minutes. After this, we changed the medium and incubated the cells in a 37°C incubator for 12 hours. At the end of the incubation, both the supernatant and cells were collected, and IL-10 levels in the supernatant, as well as IL-10 mRNA expression in the macrophages, were measured.

Establishment of the PM/Hepatocyte Co-culture System

For the co-culture system, primary mouse hepatocytes were seeded in the lower chamber of 6-well plates at a density of 2 × 10⁵ cells per well. A transwell insert with 0.4-μm pores containing 1 × 10⁶ PMs stimulated by HA was placed above the hepatocytes. The co-culture was maintained in DMEM (including 10% FBS, 10 mM HEPES) for 12 hours.

In certain experiments, IL-10 neutralizing antibody (50 ng/mL, JES5-2A5, MCE) or an isotype control antibody (50 ng/mL, Rat IgG1 kappa, 14-4301-82, Thermo Fisher Scientific) was added to the culture medium to assess the role of IL-10 in the system.

Terminal Deoxynucleotidyl Transferase dUTP Nick-end Labeling Assay

Sections of paraffin-embedded hepatic tissue were washed with PBS, fixed with 4% polyformaldehyde, and then incubated with a terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) reaction mixture (Roche). Briefly, the slides were rinsed twice with PBS. Next, 50 μL of the TUNEL reaction mixture was added to the samples, which were incubated for 60 minutes at 37°C in a humidified atmosphere in the dark. The slides were then rinsed 3 times with PBS. The samples were analyzed in a drop of PBS under a fluorescence microscope, using an excitation wavelength of 450 to 500 nm and detection in the range of 515 to 565 nm (green).

Immunohistochemical and Immunofluorescence Staining

Hematoxylin and eosin (H&E) staining was used in liver sections to observe inflammation and tissue damage. For immunohistochemical (IHC) and immunofluorescence (IF) staining, the sections were dehydrated and antigens were retrieved by incubation overnight at 4°C with the following primary antibodies: rat anti-mouse F4/80 (1:100, ab11101, Abcam), mouse anti-mouse GATA6 (1:200, 67943-1-Ig, proteintech), rabbit anti-mouse ATG16L1 (1:200, ab188642, Abcam), rabbit anti-mouse platelet endothelial cell adhesion molecule-1 (CD31) (1:100, ab222783, Abcam), hyaluronan-binding protein 1/complement component 1, q subcomponent binding protein (HABP1/C1QBP) (HY-P76185, MCE), Mouse (His) (1:100, ab213204, Abcam), rabbit anti-mouse LC3 (1:200, ab192890, Abcam), rat anti-mouse CD11b (1:50, ab8878, Abcam), rat anti-mouse Ly6G (1:500, 88876, CST), rat anti-mouse Clec4f (MAB2784; 1/1000, R&D Systems), and rabbit anti-mouse PCNA (1:200, ab29, Abcam) antibodies. The primary antibodies were detected with goat anti-mouse AF488 (1:1000, A-11001, Invitrogen), donkey anti-rat Cy3 (1:500, 712-165-153, Jackson), and goat anti-rabbit Cy5 (1:1000, ab6564, Abcam) conjugated secondary antibodies. Positive cells were counted blindly at 10 high-power fields (HPFs) per section.

Measurement of ROS Levels

ROS activity in PMs were measured by an ROS assay kit (S0033S, Beyotime) according to the instructions. The collected macrophages were seeded in a 96-well plate and incubated with 2',7'-dichlorodihydrofluorescein (DCFH). Fluorescence was detected by a confocal microscope system at 488 nm excitation and 525 nm emission.

Flow Cytometry

The NPCs, including PMs recruited to the liver, were obtained as previously described. After the cells had been stained with a fixable viability dye (eBioscience), they were incubated with Fc block (BD Biosciences) and the following primary antibodies for 30 minutes: CD45.2 (109828, BioLegend), Fixable Viability Stain 780 (565388, BD), F4/80 (111603, BioLegend). For staining ATG16L1(ab233796, Abcam), cells were fixed and permeabilized before staining. The cells were then analyzed using a BD FACS Canto II flow cytometer. Antibodies used for flow cytometry are provided in Table 2.

Table 2.

Antibodies Used for Flow Cytometry

Antigen	Conjugation	Dilution	Reference	Provider
CD45.2	percp/cy5.5	1:400	109828	Biolegend
FVS780	APC-Cy7	1:1000	107017	BD
F4/80	PE	1:400	111603	Biolegend
ATG16L1	V450	1:300	ab233796	Abcam

SiRNA Transfection

Slfn-5 (sc-153590, Santa Cruz Biotechnology, Inc), MerTK (155331, Thermo Fisher Scientific) and negative control siRNA (NS-siRNA) (SI03650318, Qiagen) transfection was performed according to the manufacturer’s protocol. Each well received 10 pmol siRNA mixed with 250 μL serum-free DMEM medium. Then, 5 μL diluted Lipofectamine 3000 reagent (Thermo Fisher Scientific, L3000015) was added. Following a 5-minute incubation, the complexes were added to the cells and incubated for 12 hours. After transfection, the PM medium was replaced for subsequent experiments.

Migration Assay

In vivo

PMs from CD45.2 mice were infused into CD45.1 mice with APAP-induced liver injury. Subsequently, intrahepatic NPCs were extracted, and the proportion of CD45.2/F4/80 double-positive cells was assessed using flow cytometry. In specific experiments, mice with APAP-induced liver injury were pretreated with a Gαi receptor PM inhibitor (PTX, Sigma-Aldrich, P7208, 15 μg/kg), a CD44 inhibitor (Angstrom6, HY-P2230, MCE, 100 mg/kg), or HA (5 units/kg, Sigma Aldrich, H3884).

In vitro

A Transwell chamber with an 8-μm pore size was employed for the experiment. The upper chamber housed 4 × 10⁴ PMs, whereas the lower chamber (a 24-well plate) contained 4 × 10⁴ hepatocytes that had been stimulated with APAP for 6 hours, along with DMEM supplemented with 10% FBS and 10 mM HEPES. A CD44 inhibitor (Angstrom6, MedChemExpress, 10 nM) was introduced to designated experimental groups.

Western Blot Analysis

Liver tissue or cellular proteins were extracted using ice-cold lysis buffer (50 mM Tris, 150 mM Nacl, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS) and 1% Triton-100). Proteins (20 μg/sample) were separated by 10% SDS-polyacrylamide gel electrophoresis (PAGE) and transferred to polyvinylidene fluoride (PVDF) nitrocellulose membranes. Rabbit anti-mouse ATG16L1 (1:1000, ab187671, Abcam), rabbit anti-mouse CD44 (1:1000, ab243894, Abcam), rabbit anti-mouse OAS3 (1:1000, 21915-1-AP, proteintech), rabbit anti-mouse SLFN5 (1:1000, AF15102, AIFang biological), rabbit anti-mouse DHX58 (1:1000, 11355-1-AP, proteintech), rabbit anti-mouse EIF2AK2 (1:5000, 18244-1-AP, proteintech), rabbit anti-mouse TRIM21 (1:5000, 12108-1-AP, proteintech), rabbit anti-mouse MerTK (1:1000, 27900-1-AP, proteintech), rabbit anti-mouse Axl (1:1000, ab215205, Abcam), rabbit anti-mouse TYRO3 (1:1000, 28513-1-AP, proteintech), rabbit anti-mouse TIM4 (1:1000, ab47637, Abcam), rabbit anti-mouse CXCR2 (1:1000, 20634-1-AP, proteintech), rabbit anti-mouse LC3 (1:1000, ab192890, Abcam), rabbit anti-mouse p62 (1:1000, ab109012, Abcam), rabbit anti-mouse PCNA (1:1000, ab29, Abcam), and mouse anti-mouse β-actin (1:1000, #3700S, Cell Signaling Technology) were used.

Quantitative Real-time-PCR

Total RNA was purified from cells using TRIzol reagent (Invitrogen, 15596026) and reverse transcribe using a Transcriptor First-Strand cDNA Synthesis kit (Roche) according to the manufacturer’s instructions. Quantitative real-time PCR was performed using SYBR green (Roche) on a StepOnePlus Real-Time PCR System (Applied Biosystems). Quantitative PCR was repeated 3 times for each sample. The expression levels of target genes were normalized to that of hypoxanthine phosphoribosyltransferase. The primers used in our study are listed in Table 3.

Table 3.

Primers Used for Quantitative Polymerase Chain Reaction

Murine gene	5′-3′ primer sequence or assay ID
Il1b	Fw TGGACCTTCCAGGATGAGGACA Rev GTTCATCTCGGAGCCTGTAGTG
Il10	Fw CGGGAAGACAATAACTGCACCC Rev CGGTTAGCAGTATGTTGTCCAGC
CD44	Fw CGGAACCACAGCCTCCTTTCAA Rev TGCCATCCGTTCTGAAACCACG
Mertk	Fw ATCATCCTCGGCTGCTTCTGTG Rev ACGACCAGTTGGGAATCCTCCT
Axl	Fw GGTGTTTGAGCCAACCGTGGAA Rev GCCACCTTATGCCGATCTACCA
Tyro3	Fw GAGGATGTCCTCATTCCAGAGC Rev CACTGCCACTTTCACGAAGGAG
Tim4	Fw AGAATGTGCGCTTGGAGCTGAG Rev GGTTGGGAGAACAGATGTGGTC
Cxcr2	Fw CTCTATTCTGCCAGATGCTGTCC Rev ACAAGGCTCAGCAGAGTCACCA
Pcna	Fw CAAGTGGAGAGCTTGGCAATGG Rev GCAAACGTTAGGTGAACAGGCTC
Actb	Fw CATTGCTGACAGGATGCAGAAGG Rev TGCTGGAAGGTGGACAGTGAGG

Enzyme-linked Immunosorbent Assay

The levels of IL-1β, IL-10 in serum or cell supernatants were measured using enzyme-linked immunosorbent assay (ELISA) kits according to the manufacturer’s protocols (Invitrogen).

RNA-seq

PMs from Atg16L1^fl/fl and Atg16L1^ΔMφ mice were extracted and then cultured in DMEM (including 10% FBS, 10 mM HEPES) for 1 hour, followed by RNA-seq (n = 3 per group). Total RNA was extracted from cells using TRIzol reagent (Invitrogen). The purity and fragment length of RNA were evaluated using the NanoDrop spectrophotometer (Peqlab) and Agilent 2100 Bioanalyzer (Agilent Technologies), respectively. Subsequently, cDNA libraries were constructed for each pooled RNA sample using the NEBNext Ultra RNA Library Prep Kit for Illumina (NEB) according to the manufacturer’s instructions. Pathway analysis, based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, was conducted to identify pathways significantly associated with differentially expressed genes (DEGs).

Co-immunoprecipitation

Co-immunoprecipitations were performed with SureBeads Starter Kit Protein A (BioRad) according to the manufacturer’s protocol. The cell lysates and magnetic beads were incubated overnight at 4°C on a vertical shaker, followed by two washes. Then, 500 μL of 1 × kinase buffer was added to a centrifuge tube, and the precipitate was resuspended in 40 μL of 1× kinase buffer after washing. Next, 200 μM ATP was added for substrate binding reaction, and the samples were incubated at 37°C for 30 minutes. Subsequently, 30 μL of 2 × SDS loading buffer containing mercaptoethanol was added to terminate the reaction, and the samples were centrifuged for 30 seconds. Finally, the samples were heated in a boiling water bath at 95 to 100°C for 10 minutes and stored at −20°C. Antibodies used in WB included anti-CD44 antibody (Abcam, ab243894), anti-SLFN5 antibody (Aifang, AF15102), and IPKine HRP Mouse Anti-Rabbit IgG LCS (Abbkine, A25022).

Statistical Analyses

All data are expressed as the mean ± standard error of the mean (SEM) from at least 3 biological replicates per group. Comparisons between 2 groups were conducted using a 2-tailed Student’s t-test, whereas 1-way analysis of variance (ANOVA) was used for comparisons among more than 2 groups. These analyses were performed using GraphPad Prism Version 10.0 software (GraphPad Software). Statistical significance was defined as P ≤ .05.