Abstract

   Inflammation is highlighted as an initial factor that helps orchestrate
   liver reconstitution. However, the precise mechanisms controlling
   inflammation during liver reconstitution have not been fully
   elucidated. In this study, a clear immune response is demonstrated
   during hepatic reconstitution. Inhibition of the hepatic inflammatory
   response retards liver regeneration. During this process, Ccl2 is
   primarily produced by type 1 innate lymphoid cells (ILC1s), and
   ILC1‐derived Ccl2 recruits peripheral ILC1s and regulatory T cells
   (Tregs) to the liver. Deletion of Ccl2 or Tregs exacerbates hepatic
   injury and inflammatory cytokine release, accelerating liver
   proliferation and regeneration. The adoption of Tregs and IL‐10
   injection reversed these effects on hepatocyte regenerative
   proliferation. Additionally, Treg‐derived IL‐10 can directly induce
   macrophage polarization from M1 to M2, which alleviated
   macrophage‐secreted IL‐6 and TNF‐α and balanced the intrahepatic
   inflammatory milieu during liver reconstitution. This study reveals the
   capacity of Tregs to modulate the intrahepatic inflammatory milieu and
   liver reconstitution through IL‐10‐mediated macrophage polarization,
   providing a potential opportunity to improve hepatic inflammation and
   maintain homeostasis.

   Keywords: inflammatory milieu, liver reconstitution, macrophage
   polarization, regulatory T cell, type 1 innate lymphoid cell
     __________________________________________________________________

   This work reveals a crosstalk between various immune cells in hepatic
   inflammatory milieu after liver injury and teases out the network to
   orchestrate the regenerative process, which provides a novel
   immune‐associated insights into liver reconstitution. This
   ILC1/Ccl2/Treg/IL‐10/M2 axis may be an effective immunotherapeutic
   strategy for liver diseases.

   graphic file with name ADVS-11-2403849-g001.jpg

1. Introduction

   The adult liver has a striking regenerative capacity following acute
   liver injury. Classic experiments in rodents have shown that after
   two‐thirds partial hepatectomy (PHx) or carbon tetrachloride (CCl[4])
   treatment, the liver mass can be reconstituted.^[ [52]^1 ^] Hepatic
   inflammation, initiated by endogenous molecules from dead or dying
   cells, is set in place to trigger hepatic repair and promote the
   re‐establishment of homeostasis. However, an inflammatory response that
   is too intense or fails is nearly always accompanied by aberrant
   reconstitution or causes irreversible damage to the liver.^[ [53]^2 ^]
   Accumulating evidence demonstrates that the inflammatory pathways,
   including TNF‐α and IL‐6 signaling pathways, are robustly activated by
   inflammatory mediators such as cytokines, chemokines, and complements,
   which stimulate quiescent hepatocytes to enter the cell cycle.^[ [54]^3
   ^] However, the communication between cytokines and immune cells, and
   the potential regulatory mechanisms to balance inflammation during
   liver reconstitution are largely unknown.

   The liver is abundant in various immune cells. To date, nearly all
   immune cell subsets have been reported to be involved in liver
   reconstitution.^[ [55]^4 ^] No single cell population or cytokine has
   enough credibility to fully understand the hepatic regenerative process
   without considering interactions between distinct immune cells.
   Macrophages have long been identified as key players in the response to
   acute liver injury,^[ [56]^5 ^] as they produce IL‐6 and participate in
   liver regeneration by promoting hepatocyte proliferation via
   phosphorylation of STAT3 and activation of the phosphoinositide‐3
   kinase (PI3K) pathway.^[ [57]^6 , [58]^7 ^] Recent evidence supports
   that acute liver injury results in the polarization of macrophages
   toward M1 (pro‐inflammatory) or M2 (anti‐inflammatory) phenotypes in
   response to extracellular stimuli.^[ [59]^8 ^] The balance between M1
   and M2 macrophages plays a critical role in the resolution of the
   hepatic inflammatory response. Therefore, the termination of hepatic
   remodeling is largely determined by macrophage polarization.

   Regulatory T cells (Tregs), a specialized T‐cell lineage, have a
   pivotal role in self‐tolerance and immune homeostasis. They contribute
   to the control and termination of the inflammatory response by
   producing anti‐inflammatory cytokines such as IL‐10, TGF‐β, and
   IL‐35.^[ [60]^9 ^] The X chromosome‐encoded transcription factor
   forkhead box P3 (Foxp3), identified as a lineage‐specific transcription
   factor of Tregs, is centrally involved in the development of Tregs and
   the maintenance of the Treg phenotype.^[ [61]^10 ^] Tregs may sense the
   initiation and progression of inflammation from the tissue
   microenvironment to adapt their suppressive functions.^[ [62]^11 ^]
   Accumulating evidence suggests that Tregs actively engage in tissue
   remodeling and repair in stressed or damaged organs. The enhanced
   frequency of intrahepatic Tregs has been reported in the injured liver
   to control liver inflammation, but whether they interact with other
   immune cell populations and their specific functions remains poorly
   understood.^[ [63]^12 , [64]^13 ^]

   The CCl[4]‐induced and two‐thirds PHx‐mediated mouse liver injury model
   was used to assess the interaction between different cell types in the
   injured liver. Here, we investigated the dynamics of cellular and
   molecular signatures in the injured liver regulated by type 1 innate
   lymphoid cells (ILC1s), Tregs, and macrophages, and how they together
   participate in controlling liver inflammation and modulating
   regenerative processes. We further identified the molecular mechanism
   by which Treg‐derived IL‐10 triggers M2 macrophage polarization during
   liver reconstitution.

2. Results

2.1. Immunosuppression Leads to Aberrant Hepatic Regeneration

   Two‐thirds PHx and CCl[4]‐induced hepatic injury mouse models were used
   to assess the liver immune microenvironment. To characterize the
   changes in immune cells following liver injury, we performed scRNA‐seq
   on CD45^+ cells isolated from CCl[4]‐treated liver. t‐distributed
   stochastic neighbor embedding (tSNE) was used to reveal nine different
   immune cell subsets (Figure [65]1A,B; Figure [66]S1A and Table [67]S1,
   Supporting Information). Indeed, liver injury induced an aberrant cell
   distribution (including upregulation of macrophages, innate lymphocyte,
   dendritic cell (cDC), and downregulation of T cells) and gene
   expression (Figure [68]1C,D; Figure [69]S1B and Table [70]S1,
   Supporting Information). Moreover, Go enrichment analysis indicated
   that the differentially expressed genes (DEGs) were associated with
   inflammatory responses and immune system processes (including TNF‐α and
   IL‐6 signaling) (Figure [71]1E; Table [72]S1, Supporting Information).

Figure 1.

   Figure 1
   [73]Open in a new tab

   DEX inhibits inflammatory response and regenerative proliferation after
   liver injury. A) tSNE projection visualizing the total (left) and each
   time point (right, control, 4 and 8 weeks) of intrahepatic immune cells
   after CCl[4] treatment (n = 3). B) Heatmap showing the marker genes
   among different cell subsets (n = 3). C) The cell count and fraction of
   indicated time point (n = 3). D) The expression of DEGs among ten major
   cell types (8 weeks versus control, n = 3). E) The enrichment score of
   DEGs in the inflammatory response and immune process (n = 3). F,G)
   Representative images of HE staining (F) and quantification of vacuolar
   degeneration (G) after saline or DEX treatment (n = 10, scale bar,
   20 µm). H) The serum level of ALT at the indicated time point after PHx
   (n = 5). (I–K) qPCR analysis of the relative mRNA expression of Il6
   (I), Tnfa (J), Ifng (K) (n = 5). L,M) Representative fluorescence
   images (L) and quantification (M) of Ki‐67‐positive cells (n = 5, scale
   bar, 20 µm). N) The ratio of liver/body weight (n = 8). O–Q)
   Representative images of western blotting (O) and quantification of the
   protein expression 24 (P) or 48 h (Q) after PHx (n = 5). Data represent
   three to five independent experiments. Data are shown as the mean + SEM
   along with individual data points and were compared using unpaired
   Student's t‐test (G, P,Q) or two‐way ANOVA followed by Bonferroni's
   multiple comparisons test (H–K, M,N). ^* p < 0.05, ^** p < 0.01, ^***
   p < 0.001, ns indicates p > 0.05.

   To determine whether the inflammatory response is involved in liver
   reconstitution, dexamethasone (DEX) was used to suppress the hepatic
   inflammatory response after PHx (Figure [74]S1C, Supporting
   Information). Notably, DEX significantly reduced vacuolar degeneration
   with lower serum alanine transaminase (ALT), Il6, Tnfa, and Ifng
   expression, suggesting alleviated hepatic inflammation
   (Figure [75]1F–K). Subsequently, livers with low inflammation showed
   decreased hepatocyte proliferation and liver‐to‐body weight ratio
   (Figure [76]1L–N). However, there was no significant change in cell
   apoptosis between saline‐ and DEX‐treated livers (Figure [77]S1D,
   Supporting Information). Western blotting analysis showed reduced
   expressions of Cyclin D1, Cdk2, p‐Akt, and p‐Stat3 in the injured
   livers from DEX‐treated mice (Figure [78]1O–Q), which confirms the view
   that hepatic inflammation plays a critical role in initiating liver
   regeneration.^[ [79]^14 ^]

2.2. Acute Liver Injury Enhances ILC1‐Derived Ccl2 Production

   Emerging evidence supports the notion that acute liver injury activates
   liver‐resident ILC1s, which promotes hepatocyte survival and protects
   the liver from damage.^[ [80]^15 , [81]^16 , [82]^17 , [83]^18 ^] To
   identify the role of ILC1 in liver injury, the profile of the ILC1
   transcriptome was analyzed using scRNA‐seq data (Figure [84]2A; Figure
   [85]S2A–C and Table [86]S2, Supporting Information). Notably,
   CCl[4]‐induced liver injury enhanced ILC1‐derived Ccl2 and Ccr2
   expression, although there was no significant change in the ILC1
   fraction (Figure [87]2B–D; Figure [88]S2D,E, Supporting Information).
   Similarly, 2/3 PHx induced an increase in the ILC1 population and Ccl2
   production (Figure [89]2E–H; Figure [90]S2F, Supporting Information).
   ELISA and qPCR results confirmed hepatic Ccl2 enhancement after PHx
   (Figure [91]2I,J). The deficiency of ILC1s eliminated the effect of
   Ccl2 augmentation in the injured liver, indicating that the elevated
   Ccl2 was mainly produced by ILC1s (Figure [92]2K). To exclude the
   impact of NK cells, Ccl2 expression was detected in NK cells. There was
   no significant difference in Ccl2 expression in NK cells between 0 and
   6 h after PHx (Figure [93]S2G,H, Supporting Information). Together,
   these results indicate that the increased Ccl2 primarily comes from
   ILC1s.

Figure 2.

   Figure 2
   [94]Open in a new tab

   Liver injury leads to ILC1‐derived Ccl2 production. A) The ILC was
   re‐clustering from Figure [95]1A. Three ILC subtypes were identified
   based on clustering after CCl[4] treatment (n = 3). B) The cell count
   and fraction of indicated time point after CCl[4] treatment (n = 3). (C
   and D) The specificity C) and intensity D) of Ccl2 expression in ILC
   subtypes (n = 3). E,F) Representative flow cytometry plots (E) and the
   statistical quantification (F) of hepatic ILC1s (CD49a^+ NK1.1^+) 6 h
   after PHx (n = 9). (G and H) Representative flow cytometry plots G) and
   percentage H) of the Ccl2^+ cells in ILC1s (CD49a^+ NK1.1^+, gated from
   Figure [96]2E) (n = 5). I) The level of Ccl2 measured by ELISA (n = 8).
   J) qPCR analysis of the relative mRNA expression of Ccl2 in the liver
   (n = 7). K) The level of Ccl2 measured by ELISA in WT, Rag2 ^−/− (lack
   T and B cells), Rag2 ^−/− γC ^−/− (lack T, B, ILC1s, and NK cells) mice
   (n = 5). L) 1×10^5 ILC1s (CD49a^+, NK1.1^+) were isolated from the
   regenerative liver 6 h after PHx. The western blotting of cGAS‐STING
   (p‐IRF3, IRF3, and STING) and Ccl2 protein expression (n = 3). M)
   Volcano plots of DEGs in ILC1s. The red dots indicate upregulated
   genes. The blue dots indicate downregulated genes (n = 4). Data
   represent three to five independent experiments. Data are shown as the
   mean + SEM along with individual data points and were compared using
   unpaired Student's t‐test (F,I,J) or Mann–Whitney test (H) or two‐way
   ANOVA followed by Bonferroni's multiple comparisons test (K). ^**
   p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

   Liver injury results in the production of massive amounts of
   danger‐associated molecular patterns (DAMPs) from dead cells, which
   activate the cGAS‐STING pathway.^[ [97]^19 , [98]^20 ^] Therefore, we
   isolated ILC1s from the regenerative liver. Interestingly, acute liver
   injury induced cGAS‐STING pathway activation with elevated Ccl2
   expression in ILC1s. These results indicate that cGAS‐STING activation
   may contribute to Ccl2 secretion in ILC1s (Figure [99]2L). To further
   explore ILC1s gene expression changes after PHx, we sorted ILC1s from
   the regenerated liver (6 h after PHx) and performed bulk RNA‐seq. Acute
   liver injury led to a significantly different profile of gene
   expression in ILC1s (Figure [100]2M; Figure [101]S3A and Table [102]S3,
   Supporting Information). A range of cytokine and cytokine receptor
   genes (Il17ra, Il13ra1, Ccl2, Ccl6, Ccl9, Ifngr1, Ccr1, Il1r1, Il1r2,
   Cxcr2, Il1b, Il18, Ccr2, Cxcl2, Ccr5) were elevated after injury
   (Figure [103]S3B and Table [104]S3, Supporting Information). The DEGs
   were also enriched in the “inflammatory response” and “cytokine
   production” gene sets (Figure [105]S3C,D and Table [106]S3, Supporting
   Information), including PI3K/AKT and IL‐6/STAT3 signaling (Figure
   [107]S3E,F and Table [108]S3, Supporting Information).

2.3. Acute Liver Injury Leads to Ccl2‐Mediated ILC1s and Tregs Recruitment

   Interestingly, few Ki‐67^+ ILC1s and Tregs were found in the
   regenerative liver, suggesting that the expansion of these cells is due
   to recruitment (Figure [109]S4A–D, Supporting Information). Previous
   studies also substantiate that the Ccl2/Ccr2 axis can recruit various
   immune cells (Tregs, monocytes, and NK cells) to specific organs.^[
   [110]^21 , [111]^22 ^] As noted, scRNA‐seq data revealed a clear
   co‐expression of Ccl2 and Ccr2 in ILC1s, indicating that chronic liver
   injury led to ILC1s recruitment via autocrine mechanism (Figure
   [112]3A,B). To determine whether the Ccl2/Ccr2 axis drives ILC1s homing
   to the injured liver, Ccl2 gene knockout (Ccl2 ^−/−) mice were used to
   assess Ccl2‐mediated recruitment. As expected, the deletion of Ccl2
   attenuated the recruitment of total and Ccr2^+ ILC1s in response to
   acute liver injury (Figure [113]3C,D). Additionally, decreased Ccr2^+
   and total Tregs were detected in Rag2 ^−/−γC^−/‐ and Ccl2 ^−/‐ livers
   after injury, suggesting that Ccl2 triggers Tregs homing (Figure
   [114]S4E–L, Supporting Information). To verify whether ILC1‐derived
   Ccl2 mediates Treg recruitment, we created Foxp3^+ cell‐specific Ccr2
   knockout (Ccr2 ^cKO) mice (Figure [115]S4M, Supporting Information).
   Although Ccr2 knockout did not influence the maturation of Tregs
   (Figure [116]S4N–Q, Supporting Information) or Tregs homing to the
   liver under steady state (Figure [117]S4R–U, Supporting Information),
   Ccr2 ^cKO mice exhibited lower Treg fraction in the injured liver
   (Figure [118]S4V,W, Supporting Information). Taken together, these
   results indicate that acute liver injury triggers the expansion of
   Tregs due to ILC1‐secreted Ccl2 recruitment.

Figure 3.

   Figure 3
   [119]Open in a new tab

   Deletion of Ccl2 results in deteriorated hepatic inflammation and
   enhanced hepatocyte proliferation in acutely injured livers. A) The
   scRNA‐seq data were analyzed from Figure [120]2A.The expression of Ccl2
   and Ccr2 in tSNE plot. Ccl2‐positive cell colored by red. Ccr2‐positive
   cell colored by green. Ccl2 and Ccr2 coexpression colored by yellow (n
   = 3). B) Correlation analysis between the Ccl2 and Ccr2 expression (n =
   3). C,D) Wildtype (WT) and Ccl2 ^−/− mice were subjected to 2/3 PHx.
   Representative flow cytometry plots (C) and the statistical
   quantification (D) of hepatic ILC1s (CD49a^+ NK1.1^+) and Ccr2^+ ILC1s
   6 h after PHx (n = 5, 6). (E and F) Representative images of HE
   staining E) and quantification F) of vacuolar degeneration 24 h after
   PHx (n = 5, scale bar, 20 µm). G) Serum levels of ALT (n = 5). H–J) The
   relative mRNA expression of Il6 (H), Tnfa (I) and Il10 (J) (n = 5).
   K,L) Representative fluorescence images (K) and quantification (L) of
   Ki‐67‐positive cells (n = 5, scale bar, 20 µm). M) The ratio of
   liver/body weight (n = 6). N–P) Representative images of western
   blotting (N) and quantification (O,P) of the relative protein
   expression (n = 5). Data represent three to five independent
   experiments. The Spearman's correlation analysis was used in panel B.
   Data are shown as the mean + SEM along with individual data points and
   were compared using unpaired Student's t‐test (D,F,O,P) or two‐way
   ANOVA followed by Bonferroni's multiple comparisons test (G–J,L,M). ^*
   p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001, ns indicates
   p > 0.05.

   Interestingly, we found significantly enhanced vacuolar degeneration
   accompanied by elevated serum ALT after PHx in Rag2 ^−/−γC^−/−, Ccl2
   ^−/‐ or Ccr2 ^cKO livers (Figure [121]3E–G; Figure [122]S5A–F,
   Supporting Information). Additionally, Ccl2 ^−/− or Rag2 ^−/−γC^−/−mice
   exhibited increased Il6 and Tnfa and decreased Il10 expression in the
   injured liver, demonstrating an aggravated inflammatory response during
   liver regeneration (Figure [123]3H–J; Figure [124]S5G,H, Supporting
   Information). Indeed, deficiency of the Ccr2/Ccl2 axis promoted
   hepatocyte regenerative proliferation, resulting in a notable increase
   in liver‐to‐body weight ratio (Figure [125]3K–M; Figure [126]S5I–L,
   Supporting Information). Notably, western blotting also revealed
   increased expression of Cyclin D1, Cdk2, p‐Akt, and p‐Stat3 in the
   injured livers of Ccl2 ^−/− mice (Figure [127]3N–P). These data suggest
   that Ccl2 modulates cell‐cycle progression and inflammatory signaling.

2.4. Tregs are Involved in Liver Regeneration and Development

   As mentioned above, the Ccr2/Ccl2 axis recruits peripheral Tregs to the
   injured liver. To explore the role of Tregs in hepatocyte
   proliferation, we used scurfy mice (Foxp3 ^−/y), Foxp3^DTR knock‐in
   mice, and CD25 neutralizing antibody (CD25 Ab)‐injected mice to
   determine whether Tregs affect hepatocyte proliferation during liver
   development or regeneration. Hepatocytes undergo rapid proliferation
   within the first month after birth. Notably, Treg deficiency aggravated
   hepatic vacuolar degeneration (Figure [128]S5M,N, Supporting
   Information) and increased hepatocyte proliferation (Figure [129]S5O,P,
   Supporting Information). We also observed an increase in the
   liver‐to‐body weight ratio independent of body weight changes (Figure
   [130]S5Q,R, Supporting Information). Moreover, we found a swollen liver
   with increased immune cell infiltration after 4 weeks of DT
   administration in Foxp3^DTR knock‐in mice (Figure [131]4A–C).
   Additionally, DT administration led to increased hepatocyte
   proliferation (Figure [132]4D,E). CD25 Ab administration also resulted
   in a decreased percentage of Tregs (Figure [133]4F–H), with similar
   changes observed during liver regeneration (Figure [134]4I–M). These
   observations further support that Tregs act as suppressors of hepatic
   inflammatory response both in physiological conditions and in liver
   injury.

Figure 4.

   Figure 4
   [135]Open in a new tab

   Treg balance hepatic inflammatory response to control liver
   regeneration and development. A) The liver tissue from Foxp3^DTR mice
   after 4 weeks of DT injection. Flow chart of DT administration. B,C)
   Representative images of histology (B) and quantification (C) of
   liver/body weight ratio (n = 5, gross scale bar, 2.5 mm; HE staining
   scale bar, 20 µm). D,E) Representative fluorescence images (D) and
   quantification (E) of Ki‐67‐positive cells (n = 5, scale bar, 20 µm).
   F) Flow chart of anti‐CD25 antibody administration. G,H) Representative
   flow cytometry plots (G) and the statistical quantification (H) of
   hepatic Tregs (CD4^+ Foxp3^+) (n = 5). I,J) Representative images of HE
   staining (I) and quantification (J) of vacuolar degeneration (n = 5,
   scale bar, 20 µm). K,L) Representative fluorescence images (K) and
   quantification (L) of Ki‐67‐positive cells (n = 5, scale bar, 20 µm).
   M) The ratio of liver/body weight (n = 6). N,O) The relative mRNA
   expression of Il6 (N) and Tnfa (O) (n = 5). P–R) Representative images
   of immunoblotting (P) and quantification (Q,R) of the relative protein
   expression (n = 5). S) The relative mRNA expression by qPCR 6 h after
   PHx (n = 5). T) The relative mRNA expression of Il10 (n = 5). U,V)
   Representative flow cytometry plots (U) and percentage (V) of IL‐10^+
   cells in Tregs (CD4^+ Foxp3^+, gated from Figure [136]4G) 6 h after PHx
   (n = 5). W) The T cells were re‐clustering from Figure [137]1A. Five T
   cell subtypes were identified based on clustering after CCl[4]
   treatment (n = 3). X) The cell count and fraction of indicated time
   point after CCl[4] treatment (n = 3). Y,Z) The specificity (Y) and
   intensity (Z) of Il10 expression in T cell subtypes (n = 3). Data
   represent three to five independent experiments. Data are shown as the
   mean + SEM along with individual data points and were compared using
   unpaired Student's t‐test (C,H,J, Q–S,V) or Mann–Whitney test (E) or
   two‐way ANOVA followed by Bonferroni's multiple comparisons test
   (L–O,T). ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001,
   ns indicates p > 0.05.

2.5. Tregs Balance Inflammation to Control Regenerative Progress

   Similarly, Tregs deficiency led to remarkably elevated expression
   levels of Il6 and Tnfa (Figure [138]4N,O). qPCR and Western blotting
   identified significantly activated AKT and STAT3 signaling in CD25
   Ab‐treated mouse livers, demonstrating enhanced inflammation and
   proliferation signals in Treg‐deficient livers (Figure [139]4P–S). We
   then investigated the potential mechanism by which Tregs regulate liver
   regeneration. Accumulating evidence supports the notion that Tregs play
   a pivotal role in controlling the inflammatory response via the
   production of anti‐inflammatory cytokines (IL‐10, TGF‐β, and IL‐35).^[
   [140]^10 , [141]^23 ^] Indeed, blocking CD25 significantly reduced Il10
   levels during liver regeneration, while CD25 Ab administration did not
   affect the expression of Tgfb and Il35 in the liver (Figure [142]4T;
   Figure [143]S5S,T, Supporting Information). Flow cytometry confirmed a
   decrease in the level of IL‐10 produced by Tregs (Figure [144]4U,V).
   Moreover, the gene profile of various T cell subsets (Naïve T, Memory
   T, Th17, CTL, Treg) was identified by scRNA‐seq data (Figure [145]4W;
   Figure [146]S6A–E and Tables [147]S1 and [148]S2, Supporting
   Information). These results also revealed an increased Treg fraction
   and Treg‐derived IL‐10 production during liver injury
   (Figure [149]4X–Z).

   To test whether Tregs could directly impact the proliferation of
   hepatocytes, we isolated splenic Tregs from Foxp3 ^YFP‐cre mice and
   co‐cultured the mouse hepatic cell line NCTC1469 with Tregs/IL‐10 in
   vitro (Figure [150]S7A, Supporting Information). Interestingly, CFSE
   analysis revealed that Tregs or IL‐10 did not directly affect
   hepatocyte proliferation (Figure [151]S7B, Supporting Information).
   Subsequently, we transfused Tregs and IL‐10 into Ccr2 ^cKO and CD25
   Ab‐treated mice according to the regimen (Figure [152]5A,B).
   Restoration of Tregs and IL‐10 prevented excessive vacuolar
   degeneration and liver injury during liver regeneration
   (Figure [153]5C–F), which reduced hepatocyte proliferation
   (Figure [154]5G–J) and the liver‐to‐body weight ratio
   (Figure [155]5K,L). Adoptive transfer of Tregs or IL‐10 treatment also
   alleviated liver injury (Figure [156]5M,N) and suppressed hepatic Il6
   and Tnfa expression (Figure [157]5O–R), indicating that Tregs could
   inhibit the inflammatory response and regenerative proliferation via
   IL‐10 production.

Figure 5.

   Figure 5
   [158]Open in a new tab

   Treg‐derived IL‐10 controls regenerative proliferation. A,B) Flow chart
   of Treg adoption (A) or anti‐CD25 antibody and IL‐10 administration
   (B). C–F) Representative images of HE staining (C,E) and quantification
   (D,F) of vacuolar degeneration 24 h after PHx (n = 5, scale bar,
   20 µm). G–J) Representative images (G,I) and quantification (H,J) of
   Ki‐67 positive cells 24 h after PHx (n = 6, 7, scale bar, 20 µm). K,L)
   The ratio of liver/body weight (n = 6–8). M,N) Serum levels of ALT (n =
   5). O–R) The relative mRNA expression of Tnfa (O,P) and Il6 (Q,R) (n =
   5). Data represent three independent experiments. Data are shown as the
   mean + SEM along with individual data points and were compared using
   one‐way ANOVA followed by Bonferroni's multiple comparisons test
   (D,F,H,J) or two‐way ANOVA followed by Bonferroni's multiple
   comparisons test (K–R). ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001,
   ^**** p < 0.0001, comparison between the 2 indicated groups, or
   comparison between the Ccr2 ^f/f + PBS and Ccr2 ^cKO + PBS groups, or
   comparison between the Control+PBS and CD25Ab + PBS groups; ^#
   p < 0.05, ^## p < 0.01, ^### p < 0.001, ^#### p < 0.0001, comparison
   between the Ccr2 ^cKO + PBS and Ccr2 ^cKO + Treg, or comparison between
   the CD25Ab + PBS and CD25Ab + IL‐10 groups.

2.6. Treg‐Derived IL‐10 Converts Macrophage Polarization from M1 to M2

   As mentioned above, Treg‐derived IL‐10 did not directly influence
   hepatocyte proliferation. Recent studies support the notion that IL‐10
   induces a dominant immunosuppressive M2 macrophage phenotype via
   polarization.^[ [159]^24 , [160]^25 ^] To explore whether Treg‐derived
   IL‐10 modulates proliferation by controlling macrophage polarization,
   macrophage subsets were analyzed using scRNA‐seq data (Figure [161]6A;
   Figure [162]S8A–C and Table [163]S2, Supporting Information). As
   expected, liver injury induced an expansion of M2 macrophages with
   elevated M1 macrophage‐derived Tnf levels (Figure [164]6B–D). In
   contrast, Ccl2 expression was rarely detected in both M1 and M2
   macrophages (Figure [165]S8D, Supporting Information). The DEGs of M1
   macrophages were enriched in the “inflammatory response” and “immune
   response” categories (Figure [166]6E; Table [167]S4, Supporting
   Information). Interestingly, Il10ra and Tnf were co‐expressed in M1
   macrophages, indicating that IL‐10 could interact with M1 macrophages
   to trigger M1 transformation during liver injury (Figure [168]S8E–H,
   Supporting Information). CellChat analysis also indicated closer
   interactions between ILC1, macrophages, and Tregs (Figure [169]S8I,
   Supporting Information).

Figure 6.

   Figure 6
   [170]Open in a new tab

   IL‐10 forces M2 macrophage polarization to regulate hepatocyte
   proliferation. A) The macrophages were re‐clustering from
   Figure [171]1A. Two macrophage subtypes were identified based on
   clustering after CCl[4] treatment (n = 3). B) The cell count and
   fraction of indicated time point after CCl[4] treatment (n = 3). C, D)
   The specificity (C) and intensity (D) of Tnf expression in T cell
   subtypes (n = 3). E) Go enrichment of DEGs in M1 (n = 3). F–I)
   Representative histogram plots (F and H) and quantification (G and I)
   of the fluorescence intensity of macrophage (CD11b^+)‐gated IL‐6, TNF‐α
   and CD206 24 h after PHx (n = 5). J–M) Macrophages were isolated and
   incubated with IL‐10 for 48 h. The relative mRNA expression by qPCR (n
   = 5). N–P) Representative images of western blotting (N) and
   quantification (O, P) of the relative protein expression (n = 5). Q)
   Flow chart of co‐culture procedure in vitro. R, S) NCTC1469 cells were
   labeled with CFSE and co‐cultured with macrophages (isolated from the
   injured liver) containing PBS/IL‐10 for 2 days or 4. Representative
   histogram plots (R) and quantification (S) of the proliferation ratio
   of NCTC1469 cells (n = 5). T) The expression levels of TNF‐α in the
   culture supernatants were measured by ELISA (n = 5). U, V)
   Representative images of immunoblotting (U) and quantification (V) of
   the relative protein expression in NCTC1469 cells (n = 5). Data
   represent three independent experiments. Data are shown as the mean +
   SEM along with individual data points and were compared using unpaired
   Student's t‐test (I, V) or one‐way ANOVA followed by Bonferroni's
   multiple comparisons test (G, J–M, and O–P) or two‐way ANOVA followed
   by Bonferroni's multiple comparisons test (S, T). ^* p < 0.05, ^**
   p < 0.01, ^*** p < 0.001, ^**** p < 0.0001.

   Subsequently, adoptive transfer of Tregs or IL‐10 injection induced M1
   to M2 polarization in total macrophages (CD11b^+), Kupffer cells
   (CD11b^+ F4/80^high), and monocyte‐derived macrophages (CD11b^+
   F4/80^low) (Figure [172]6F–I; Figure [173]S9A–C, Supporting
   Information), although no significant change in the total number of
   macrophages was observed after treatment (Figure [174]S9D–G, Supporting
   Information). Moreover, deficiency of Ccl2 and ILC1s resulted in
   elevated TNF‐α and reduced CD206 in macrophages, demonstrating that
   Ccl2‐induced Tregs convert macrophage polarization from M1 to M2
   (Figure [175]S9H–K, Supporting Information). To explore whether Ccl2
   regulates M1/M2 polarization, macrophages were isolated from the
   injured liver (6 h after PHx) and incubated with Ccl2 for 48 h.
   Interestingly, no difference in IL‐6, TNF‐α, or CD206 levels was found
   after treatment, suggesting that Ccl2's influence on M1/M2 polarization
   relies on Tregs (Figure [176]S9L,M, Supporting Information).
   Furthermore, co‐culture with macrophages and Tregs significantly
   induced M2 macrophage transformation (Figure [177]S9N,O, Supporting
   Information). Additionally, IL‐10 treatment inhibited the expression of
   Il6 and Tnfa and boosted the expression of Arg1 and Ym1 in a
   dose‐dependent manner (Figure [178]6J–M), indicating that Treg‐derived
   IL‐10 directly promotes M1 to M2 polarization in macrophages
   independently of direct interaction. IL‐10 treatment also inactivated
   both the MAPK and NF‐κB signaling pathways (important for macrophage
   polarization) in macrophages, as indicated by increased protein levels
   of p‐Erk1/2 and p‐NFκB (Figure [179]6N–P). NCTC1469 cells were
   co‐cultured with isolated macrophages and IL‐10 to assess the effect of
   macrophage polarization on hepatocyte proliferation (Figure [180]6Q).
   Co‐culture with macrophages and IL‐10 significantly impeded hepatocyte
   proliferation (Figure [181]6R,S), which was accompanied by decreased
   TNF‐α levels and reduced p‐Erk1/2 and p‐NFκB levels (Figure [182]6T–V).
   These findings indicate that by controlling M2 polarization, IL‐10
   modulates the hepatic immune milieu and liver regeneration.

3. Conclusion

   Immune activation and inflammatory surges are key features of various
   acute and chronic liver injuries. Unresolved inflammation acts as a
   detrimental factor in many liver diseases (hepatitis, cirrhosis, and
   cancer) by exacerbating liver damage.^[ [183]^1 , [184]^26 ^]
   Therefore, identifying the key factors that regulate the transition
   between healthy and pathogenic inflammation is critical for developing
   novel therapeutic strategies against inflammation‐related liver
   diseases. Several subsets of innate (including macrophages, NK, NKT,
   γδT, ILC, and cDC) and adaptive immune cells (including T and B cells)
   affect the hepatic immune microenvironment.^[ [185]^1 , [186]^4 ^]
   Previous studies have often focused on the function of specific cell
   types during liver regeneration. However, the interaction between
   distinct cell types to ensure an efficient and effective regeneration
   process remains unknown. Our work illustrates the network between
   immune cells that balances the inflammatory milieu during liver
   regeneration.

   Type 1 innate lymphoid cells are tissue‐resident innate lymphocytes
   that regulate tissue inflammation and homeostasis.^[ [187]^15 ,
   [188]^16 , [189]^17 , [190]^18 ^] ILC1s rapidly produce specific
   cytokines in response to local inflammation.^[ [191]^27 ^] In the
   liver, ILC1s are the most predominant ILC subset, except for NK cells.
   However, the pathophysiological role of hepatic ILC1s remains
   incompletely explored. A recent study highlighted the critical role of
   liver ILC1s in tissue protection during acute liver injury.^[ [192]^15
   ^] Indeed, our results support the conclusion that acute liver injury
   induces ILC1s recruitment. Accumulating evidence indicates that
   Ly6c^high monocytes extravasate into the injured liver in a
   Ccl2/Ccr2‐dependent manner.^[ [193]^1 , [194]^28 ^] However, the origin
   of Ccl2 in the injured liver remains unknown. In this study, we
   demonstrate that ILC1s are primarily responsible for Ccl2 production in
   response to liver injury, which not only triggers monocyte recruitment
   but also drives ILC1s and Tregs homing to the liver.

   Tregs, which belong to a subset of CD4^+ T cells, regulate other immune
   cells in a dominant‐negative manner via the production of
   immunosuppressive cytokines such as IL‐10, TGF‐β, and IL‐35. The
   current study indicates Ccl2‐dependent Treg expansion in acutely
   injured livers, consistent with previous studies.^[ [195]^13 , [196]^29
   ^] As noted, Tregs exert protective effects against acute liver injury,
   thereby promoting liver regeneration. In line with findings from other
   studies, chronic inflammation results in an increased proportion of
   Tregs in the liver via the IL‐33/ST2 axis.^[ [197]^12 ^] Furthermore,
   we demonstrated that Treg‐derived IL‐10 plays an immunosuppressive role
   in maintaining hepatic inflammation, which regulates the liver
   regeneration process. This finding supports the notion that IL‐10
   exerts a negative role in modulating liver regeneration by inhibiting
   the inflammatory response and tempering hepatic IL‐6/STAT3 signaling
   activation.^[ [198]^30 ^]

   Macrophages are indispensable for the hepatic inflammatory milieu,
   fibrosis, and resolution, switching their phenotype from
   pro‐inflammatory to restorative. Additionally, macrophages are
   highlighted as key regulators that help orchestrate liver
   regeneration.^[ [199]^1 ^] It is believed that macrophage‐derived
   cytokines (TNF‐α and IL‐6) are essential for liver regeneration after
   PHx.^[ [200]^31 , [201]^32 ^] Two major classes of macrophages have
   been identified in the injured liver. The proportion of M1/M2 cells
   depends on their adaptive response to various stimuli. The
   heterogeneity or plasticity of macrophages exhibits different effects
   on liver regeneration.^[ [202]^1 ^] A recent study convincingly showed
   that macrophage polarization is implicated in inflammation resolution
   and liver regeneration. Ccl5 deficiency‐mediated M2 macrophage
   polarization accelerates liver repair by improving the inflammatory
   microenvironment.^[ [203]^8 ^] Here, we identified that IL‐10 shifts
   macrophage polarization from M1 to M2, which reshapes the hepatic
   inflammatory state and controls liver regeneration. These findings
   support the notion that the STAT3/IL‐10/IL‐6 axis is an important
   regulator of macrophage efferocytosis, survival, and phenotypic
   conversion,^[ [204]^33 ^] verifying that the balance of hepatic
   macrophage polarization is finely modulated by Treg recruitment and
   IL‐10 production.

   In summary, our study reveals a network of distinct immune cell
   interactions that regulate the inflammatory microenvironment during
   liver regeneration. ILC1‐derived Ccl2 drives ILC1s and Tregs homing to
   the injured liver via paracrine and autocrine mechanisms. This
   chemotactic effect and subsequent IL‐10 production play a key role in
   modulating hepatic inflammation and regeneration. Our work also
   illustrates the mechanism of macrophage polarization and clarifies the
   effect of macrophage phenotype on liver regeneration. This
   ILC1/Ccl2/Treg/IL‐10/M2 axis may provide new immune‐associated insights
   into liver injury and repair.

4. Experimental Section

Mice

   Rag2 ^−/− (Cat# 033562), Rag2 ^−/−γC^−/− (Cat# 014593), Ccl2 ^−/− (Cat#
   004434), Foxp3 ^YFP‐Cre (Cat# 016959), Scurfy mice (Foxp3 ^−/y, Cat#
   006775), Foxp3^DTR knock‐in mice (Cat# 016958) were purchased from The
   Jackson Laboratory (Bar Harbour, USA). Ccr2 ^cKO mice were generated by
   crossing Ccr2 ^loxP (Gempharmatech, Nanjing, China) mice with Foxp3
   ^YFP‐Cre mice. For Tregs ablation, 6‐week‐old Foxp3^DTR mice were
   intraperitoneally injected twice a week with 500 ng dose^−1 of
   diphtheria toxin (DT, List Labs, California, USA) for 3 weeks. All mice
   were housed and maintained under specific pathogen‐free (SPF)
   conditions. All animal experiments, housing, and feeding conditions
   were approved by the Institutional Ethical Commission for Animal
   Research at Xiamen University (XMULAC20190061).

Animal Treatment and Single‐Cell RNA Sequencing (scRNA‐seq)

   Mice (8‐ to 12 weeks old) were subjected to 2/3 partial hepatectomy
   (PHx) and carbon tetrachloride (CCl[4], Shanghai Acmec Biochemical
   Technology Co., Ltd, Shanghai, China)‐induced hepatic injury as
   described previously.^[ [205]^34 , [206]^35 ^] Mice were given an
   intraperitoneal injection of 2 mg kg^−1 dexamethasone (DEX, Shanghai
   Shyndec Pharmaceutical Co., Ltd, Shanghai, China) once a day for three
   days before PHx. CD25 neutralizing antibody (250 µg/dose, clone
   PC‐61.5, BioXcell, New Hampshire, USA) and IL‐10 (0.01 µg/g,
   Novoprotein, Suzhou, China) were injected intraperitoneally twice a
   week before PHx. The intrahepatic immune cells were isolated according
   to the previous study. Single CD45^+ cells were sorted from untreated,
   4‐week, and 8‐week CCl[4]‐treated liver samples by
   fluorescence‐activated cell sorting (FACS), followed by droplet‐based
   scRNA‐seq on a 10X Genomics Chromium chip (10X Genomics, Pleasanton,
   USA). The 10X Genomics Single Cell v2 kit was used for reverse
   transcription and library preparation. Subsequently, the libraries were
   sequenced on an Illumina HiSeq 4000 (Illumina, San Diego, USA).

Preprocessing and Integrating scRNA‐seq Data

   In this study, single‐cell RNA sequencing (scRNA‐seq) data were
   preprocessed and integrated using the well‐established Seurat R package
   (version 4.1.1). Initially, genes detected in fewer than three cells
   were filtered out to ensure data quality. Subsequently, low‐quality
   cells characterized by a limited number of measured genes (< 200 genes)
   were excluded and potential doublets exhibiting more than 6 000 genes.
   Additionally, cells with mitochondrial gene expression constituting
   more than 20% of the total gene counts were removed to further enhance
   data reliability. To integrate the single‐cell data derived from mice
   subjected to three distinct experimental conditions, the Seurat package
   was employed, selecting 2 000 features for cell integration. This
   approach resulted in a comprehensive dataset comprising 32256 cells in
   total, enabling to investigate cellular heterogeneity and expression
   patterns across different experimental groups.

Unsupervised Clustering and Cell Identification

   The conventional workflow in the Seurat R package was employed to
   perform dimensionality reduction and unsupervised clustering on each
   scRNA‐seq dataset. The optimal cluster resolution value was determined
   using the cluster R package (version 0.5.0). To identify cell types and
   subpopulations, nine cell types were initially distinguished based on
   the expression of canonical marker genes. Specifically, B cells were
   identified by positive expression of Cd79a and Cd79b, while T cells
   were recognized by high expression of Cd3d and Cd3e. Natural killer
   (NK) cells were marked by Ncr1, Nkg7, and Klrk1 expression, and
   classical dendritic cells (cDCs) were differentiated by Xcr1 and Sirpa
   expression. Neutrophils were characterized by Ly6g and Csf3r
   expression, macrophages by Cd68 and Csf1r, mast cells by Kit and Cpa3,
   plasma cells by Jchain and Igkc, and innate lymphoid cells (ILCs) by
   Ly6a and Thy1. Non‐immune cells were identified by high expression of
   Pimreg and Pclaf. To further delineate T cell and ILC subpopulations,
   the same conventional pipeline was applied after subsetting all T cells
   and ILCs separately. T cells were classified into CD4 T cells (Cd4,
   Icos), CD8 T cells (Cd8a, Gzmk), naïve T cells (Lef1, Tcf7), memory T
   cells (Il7r, Ms4a6b), and regulatory T cells (Tregs; Cd5, Foxp3,
   Ilzf2). Regarding ILCs, ILC1, ILC2, and ILC3 were distinguished based
   on canonical markers. ILC1 expressed Ncr1, Gzma, and Plek, ILC2
   expressed Gata3, Rora and Nr4a2, and ILC3 expressed Th17‐associated
   markers, such as Il17a, Il17f, and Rorc. Moreover, macrophages were
   classified into M1 macrophages (Tnf, Cd86, Il1b) and M2 macrophages
   (Bst2, Il4ra).

Cell–Cell Communication Analysis

   The intercellular communication between ILCs, T cell, and macrophages
   was analyzed using the CellChat R package (version 1.6.1).

Differential Expression Gene Detection

   Differentially expressed genes (DEGs) among each cell cluster between
   groups were identified using the “FindMarkers” function from the Seurat
   R package with the Wilcoxon rank‐sum test algorithm. Genes with a q
   value < 0.05, log2|Fold Change| > 0.25, and min. pct > 0.1 were
   considered differentially expressed.

Pathway Enrichment Analysis

   To investigate the biological significance of the DEGs identified in
   the previous step, a Gene Ontology (GO) pathway enrichment analysis was
   performed, focusing on overrepresented immunity‐related and
   inflammatory‐associated biological processes within the context of the
   DEGs. The clusterProfiler R package (version 4.2.2) was employed to
   conduct the enrichment analysis, utilizing a hypergeometric test to
   calculate the statistical significance of the overrepresented GO terms.

Bulk RNA Sequencing (RNA‐seq) and Quantitative PCR (qPCR)

   Total RNA was isolated using TRIeasy LS Total RNA Extraction Reagent
   (Yeasen Biotechnology, Shanghai, China). cDNA was synthesized by
   chloroform extraction and isopropanol precipitation as described
   previously.^[ [207]^36 ^] A Transcriptor First Strand cDNA Synthesis
   Kit (Yeasen Biotechnology) was used for cDNA synthesis and
   strand‐specific library preparation. The libraries were sequenced on an
   Illumina NovaSeq 6 000 instrument (Illumina, San Diego, USA) using the
   PE150 strategy. R package DESeq2 was used to compare the DEGs. Genes
   with a q value < 0.05 and log2|Fold Change| > 1 were deemed
   differentially expressed. Gene Ontology (GO) and Kyoto Encyclopedia of
   Genes and Genomes (KEGG) analyses were performed using the R package
   ClusterProfiler.

   To detect mRNA expression, qPCR was performed using SYBR Green PCR mix
   (Yeasen Biotechnology) and conducted with the qTOWER3G touch (Analytik
   Jena, Jena, Germany). Relative gene expression levels of mRNA were
   normalized against Gapdh and calculated using the 2^−ΔΔCt method.
   Primer sequences are shown in Table [208]S5 (Supporting Information).

Immune Cell Isolation and Flow Cytometry

   Isolation of intrahepatic immune cells was conducted as described in
   the previous study. Briefly, the excised liver tissue was minced into
   small pieces and gently triturated through a 75 µm gauge stainless
   steel mesh. The suspension was centrifuged at 850 × g for 30 min with
   37.5% Percoll (Solarbio Life Sciences, Beijing, China) in Hank's
   balanced salt solution (HBSS, Solarbio Life Sciences). The single
   suspension was then stained with fluorescently conjugated antibodies
   against PE‐CD3 (17A2), APC‐CD3 (17A2), PerCP‐CD45 (30‐F11), APC‐CD49a
   (clone: HMα1), APC‐Ccr2 (SA203G11), FITC‐NK1.1 (S17016D), PE‐Ccl2
   (2H5), PE‐Foxp3 (clone: 150D), FITC‐CD4 (clone: GK1.5), FITC‐F4/80
   (BM8), PE‐CD11b (M1/70), APC‐IL‐6 (MP5‐20F3), APC‐Ki‐67 (16A8),
   PerCP/Cyanine5.5‐lL‐10 (JES5‐16E3), PE‐TNF‐α (MP6‐XT22) and PE‐CD206
   (C068C2) from BioLegend. The cells were fixed and permeabilized using
   the Foxp3/Transcription Factor Staining Buffer Set (eBioscience,
   Shanghai, China) for cytokine detection. Briefly, the cells were
   treated with Fixation/Permeabilization Solution for 30 min at 4 °C,
   avoiding light, followed by washing with Permeabilization Buffer.
   Subsequently, the cytokine antibodies were used for staining at 4 °C
   for 45 min. The stained cells were examined using a BD Aria III machine
   and analyzed with FlowJo software (version 10.6.2, BD, Ashland, USA).

Cell Culture

   Splenic Tregs (CD4^+YFP^+) from Foxp3 ^YFP‐cre mice were sorted and
   cultured in RPMI supplemented with 10% FBS (Yeasen, Shanghai, China).
   The isolated Tregs were incubated with 1000 IU IL‐2 (Novoprotein),
   100 ng mL^−1 PMA (Sigma, Shanghai, China), and 500 ng mL^−1 ionomycin
   (Sigma) together for 72 h. For Treg adoption, a total of 1×10^7 Tregs
   were intravenously infused into each mouse 2 days before PHx. Mouse
   NCTC1469 normal liver cells (American Type Culture Collection, ATCC)
   were cultured in DMEM (Solarbio Life Sciences) supplemented with 10%
   FBS (Cegrogen Biotech, Eupen, Belgium). To detect cell proliferation,
   NCTC1469 cells were labeled with 0.2 µM CFSE (Invitrogen, Shanghai,
   China) and analysed by flow cytometry. All cells were cultured at 37 °C
   in a humidified atmosphere with 5% CO[2].

   For co‐culture transwell experiments, 1 × 10^5 macrophages (CD11b^+,
   isolated from PHx‐6 h liver) were seeded in the lower chamber and
   co‐cultured with 1 × 10^5 Tregs (CD4^+YFP^+, isolated from the spleen)
   in the upper chamber of the transwell for 48 h. For polarization
   analysis, 1 × 10^5 macrophages (CD11b^+, isolated from mouse liver)
   were seeded in DMEM high glucose medium (Solarbio Life Sciences) with
   10% FBS. The indicated concentrations of IL‐10 and Ccl2 (Novoprotein,
   Shanghai, China) were used to stimulate the cells for 6 h.

ELISA Analysis and Serum Alanine Aminotransferase (ALT) Detection

   Hepatic Ccl2, IL‐6, and TNF‐α levels were quantified using a mouse
   Quantikine ELISA kit (RD Systems, Minneapolis, USA) according to the
   manufacturer's instructions. The optical density (OD) of the samples
   was measured with a Varioskan Flash (Thermo Fisher Scientific,
   Shanghai, China) microplate reader. For ALT detection, peripheral blood
   samples were analyzed using the UniCel DxC 800 Synchron Clinical System
   (Beckman Coulter GmbH, Brea, USA).

Western Blotting

   Liver samples were collected and sonicated in RIPA buffer (Solarbio
   Life Sciences) supplemented with a protease inhibitor cocktail (Bimake,
   Shanghai, China). The total protein concentration was determined using
   a BCA Protein Assay Kit (Solarbio Life Sciences). Twenty micrograms of
   protein were separated on an SDS‐polyacrylamide gel and transferred to
   PVDF membranes (LABLEAD, Beijing, China), followed by blocking with
   skimmed milk (OXOID, Hants, UK). Membranes were probed with primary
   antibodies and incubated with HRP‐conjugated secondary antibodies.
   Chemiluminescence detection systems (GE Healthcare Life Sciences,
   Pittsburgh, USA) were used to visualize the protein bands. The
   antibodies used were as follows: Cyclin D1 (Cell Signaling Technology,
   Cat#55506, 1:1000), Cdk2 (Cell Signaling Technology, Cat#18048,
   1:1000), p‐Akt (S473, Solarbio, Cat# K009354P, 1:1000), Akt (Solarbio,
   Cat# K106557P, 1:1000), p‐Stat3 (S727, Solarbio, K010036P, 1:1000),
   Stat3 (Solarbio, Cat# K000124M, 1:1000), Phospho‐IRF3 (S396, Huabio,
   HA722772, 1:1000), IRF3 (Servicebio, GB11368, 1:1000), STING
   (Servicebio, [209]GB111415, 1:1000), Ccl2 (Servicebio, GB11199,
   1:1000), Phospho‐NF‐κB p65 (S468, Proteintech, Cat# 82335, 1:1500),
   NF‐κB p65 (Proteintech, Cat# 80979, 1:2000), Phospho‐Erk1/2
   (Thr202/Tyr204, Proteintech, Cat# 80031, 1:2000), Erk1/2 (Proteintech,
   Cat# 66192, 1:2000), Tubulin (Cell Signaling Technology, Cat#5666,
   1:2000).

Histology and Immunohistochemistry

   For histological analysis, livers were fixed in 4% paraformaldehyde
   (Macklin Inc., Shanghai, China) and embedded in paraffin.
   Paraffin‐embedded tissue was sliced and stained with
   haematoxylin‐eosin. Subsequently, the sections were subjected to a
   standard process as described previously for immunohistochemistry and
   immunofluorescence assays.^[ [210]^34 ^] Images were acquired using a
   DM2700 microscope (Leica). The primary antibodies used for
   immunostaining included Albumin (Huabio, ET1702‐55, 1:1000), Ki‐67
   (Servicebio, Cat# [211]GB111141, 1:500), and CD11b (Servicebio, Cat#
   GB11058, 1:500). A TUNEL Cell Apoptosis Detection Kit (Servicebio,
   Wuhan, China) was used to detect apoptotic cells.

Statistical Analysis

   Statistical analyses were performed using GraphPad Prism Version 9
   (GraphPad). Depending on the normality (Shapiro‐Wilk test) and
   homogeneity of variance (Brown‐Forsythe test) of the data, various
   statistical tests were employed, including the unpaired Student's
   t‐test for normally distributed data between two groups, the
   Mann–Whitney U test for non‐normally distributed data between two
   groups, one‐way ANOVA with Bonferroni's post hoc test for comparisons
   among three or more groups with normally distributed data, and two‐way
   ANOVA with Bonferroni's post hoc test for analyzing the interaction
   between two independent variables. When the data did not meet the
   assumptions required for ANOVA, the Kruskal–Wallis test was used as a
   non‐parametric alternative. In addition, the Spearman's correlation
   analysis was performed to determine the possible correlation. Wherever
   possible, the analyses were performed in a blind manner to minimize
   bias. All statistical tests were conducted in a two‐sided manner with a
   significance level set at p < 0.05. Group data were expressed as mean +
   SEM. The number of animals used (n) in every experiment are provided in
   figure legends. Detailed statistics are provided in Table [212]S6
   (Supporting Information).

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   R.W., Q.L., and Q.Z. contributed equally to this work. K.W. and P.Y.
   conceived and designed the study; R.W., Q.L., Q.Z., S.Z., Y.L., B.L.,
   Q.F., X.B., Y.M., X.M., and B.C. performed the experiments; N.W. and
   P.Y. analyzed the sequencing data; Y.Z. and K.W. wrote the manuscript.

Supporting information

   Supporting Information
   [213]ADVS-11-2403849-s003.docx^ (6.3MB, docx)

   Supplemental Table 1
   [214]ADVS-11-2403849-s001.xlsx^ (4.2MB, xlsx)

   Supplemental Table 2
   [215]ADVS-11-2403849-s005.xlsx^ (12KB, xlsx)

   Supplemental Table 3
   [216]ADVS-11-2403849-s007.xlsx^ (5.5MB, xlsx)

   Supplemental Table 4
   [217]ADVS-11-2403849-s006.xlsx^ (897.3KB, xlsx)

   Supplemental Table 5
   [218]ADVS-11-2403849-s002.xlsx^ (9.3KB, xlsx)

   Supplemental Table 6
   [219]ADVS-11-2403849-s004.xlsx^ (47.4KB, xlsx)

Acknowledgements