ATF3-mediated transactivation of CXCL14 in HSCs during liver fibrosis

Abstract

Background and aims: Myofibroblasts, the primary producers of extracellular matrix, primarily originate from hepatic stellate cells (HSCs), and their activation plays a pivotal role in liver fibrosis. This study aimed to investigate the function of CXC motif ligand 14 (CXCL14) in the progression of liver fibrosis.

Approach and results: CXCL14 knockdown significantly reduced the extent of liver fibrosis. Using Ingenuity pathway analysis and qRT-PCR, activating transcription factor 3 (ATF3) was identified as a key upstream regulator of CXCL14 expression. Mechanistically, ATF3 was shown to bind to the CXCL14 promoter, promoting its transactivation by TGF-β in HSCs. Notably, both global CXCL14 deletion (CXCL14^-/-) and HSC/myofibroblast-specific CXCL14 knockdown significantly attenuated liver fibrosis in mice. RNA-seq comparisons between CXCL14^-/- and WT mice highlighted Jak2 as the most significantly downregulated gene, implicating its role in the antifibrotic effects of CXCL14 suppression on HSC inactivation. Moreover, Jak2 overexpression reversed the inhibition of liver fibrosis caused by CXCL14 knockdown in vivo.

Conclusions: These findings unveil a novel ATF3/CXCL14/Jak2 signalling axis in liver fibrosis, presenting potential therapeutic targets for the disease.

Keywords: CXCL14; hepatic stellate cell; liver fibrosis; transcriptional regulation.

FIGURE 1

Elevated CXCL14 during HSC activation. Primary HSCs were isolated and cultured in vitro for spontaneous activation. (A, B) Hepatocytes, KCs and primary HSCs were isolated from CCl₄‐ and BDL‐induced liver fibrosis mice, and CXCL14 expression was evaluated by qRT‐PCR and ELISA. (C) mRNA levels of α‐SMA, Col1a1 and CXCL14 in primary HSCs. (D) CXCL14 levels in primary HSCs detected by ELISA. (E, F) CXCL14 expression in LX‐2 cells measured by qRT‐PCR and ELISA following TGF‐β (2 ng/mL) or PDGF (10 ng/mL) treatment. (G) CXCL14 expression in liver tissues of healthy individuals and patients with cirrhosis, assessed by qRT‐PCR. (H) Correlation between CXCL14 and serum IV‐Collagen levels analysed by linear regression (n = 30). **p < .01, ***p < .001.

FIGURE 2

ATF3 mediates TGF‐β/PDGF‐induced CXCL14 transcription in HSCs. (A) IPA analysis. (B) mRNA levels of CXCL14 in LX‐2 cells transfected with specific siRNAs, followed by TGF‐β treatment. (C) Luciferase activity analysis of a CXCL14 promoter‐luciferase construct transfected into LX‐2 cells with increasing doses of ATF3. (D, E) Luciferase activity of various CXCL14 promoter‐luciferase constructs transfected into LX‐2 cells with or without ATF3, TGF‐β or PDGF. (F, G) ChIP analysis of LX‐2 cells treated with TGF‐β or PDGF. (H) mRNA levels of ATF3 in LX‐2 cells following TGF‐β or PDGF treatment. (I) Luciferase activity of WT and MT CXCL14 promoter‐luciferase constructs transfected into LX‐2 cells followed by TGF‐β or PDGF treatment. Each value represents the mean ± SD of three experiments. *p < .05, **p < .01, ***p < .001.

FIGURE 3

CXCL14 regulates HSC activation in vitro. Primary HSCs were isolated and cultured for spontaneous activation. (A) Immunofluorescence staining for α‐SMA (green) and Type I collagen (red) in LX‐2 cells transfected with indicated siRNAs and treated with TGF‐β. (B) mRNA levels of CXCL14, α‐SMA and Col1a1 in LX‐2 cells. (C) EdU assay in LX‐2 cells. (D) Immunofluorescence staining for α‐SMA (green) and CXCL14 (red) in LX‐2 cells and primary HSCs treated with rCXCL14 (50 ng/mL). (E) mRNA levels of α‐SMA and CXCL14 in LX‐2 cells and primary HSCs treated with rCXCL14 (50 ng/mL). (F) EdU assay in LX‐2 cells and primary HSCs. (G, H) Volcano plots and heatmap of primary HSCs treated with rCXCL14 (50 ng/mL), followed by RNA‐seq analysis. Each value represents the mean ± SD of three experiments. **p < .01, ***p < .001.

FIGURE 4

CXCL14 deficiency alleviates liver fibrosis in mice. Primary HSCs were isolated and cultured for spontaneous activation. (A) mRNA levels of α‐SMA and Col1a1 in primary HSCs isolated from WT and CXCL14^−/− mice. (B) Immunofluorescence staining for α‐SMA (green) and Type I collagen (red) in primary HSCs. (C) EdU assay. (D–G) WT and CXCL14 knockout mice were injected with CCl₄ for 6 weeks. (D, E) Levels of ALT, AST and Hyp. (F) mRNA levels of profibrogenic genes α‐SMA, Col1a1, Timp1 and Ctgf in liver tissues from WT and CXCL14^−/− mice after CCl₄ treatment. (G) HE, Masson, Sirius Red, F4/80 and Ly6G staining. Each value represents the mean ± SD of six experiments. ***p < .001.

FIGURE 5

HSC‐specific CXCL14 deletion mitigates liver fibrosis in mice Lenti‐shCXCL14 or Lenti‐shC were injected into C57/BL6 mice, followed by CCl₄ injections for 6 weeks. (A) Experimental design. (B, C) Levels of ALT, AST and Hyp. (D) mRNA levels of α‐SMA, Col1a1, Timp1 and Ctgf in liver tissues. (E) Protein levels of α‐SMA and Type I collagen in liver tissues. (F) HE, Masson, Sirius Red, F4/80 and Ly6G staining. Each value represents the mean ± SD of six experiments. ***p < .001; ns, not significant.

FIGURE 6

Jak2 is involved in CXCL14‐mediated HSC activation. Primary HSCs were isolated from WT and CXCL14^−/− mice, cultured for spontaneous activation, and analysed by RNA‐seq. (A) PCA plot. (B) Volcano plot. (C) GO analysis. (D) Heatmap of differentially expressed genes. (E, F) mRNA and protein levels of Jak2 in primary HSCs. (G) Heatmap of TFs. (H) mRNA levels of Runx1, Creb5, Myc, Gata3 and Tead2. (I) Runx1‐binding sites in the Jak2 promoter predicted by JASPAR (http://jaspar.genereg.net/). (J) Schematic of potential Runx1 binding sites in the Jak2 promoter. (K) ChIP analysis of Runx1 occupancy at the Jak2 promoter. (L) Luciferase reporter assays. Each value represents the mean ± SD of three experiments. **p < .01, ***p < .001.

FIGURE 7

CXCL14 promotes HSC activation and liver fibrosis via Jak2. (A, B) Primary HSCs were isolated from WT and CXCL14^−/− mice and transduced with adenovirus carrying a Jak2 expression vector (Ad‐Jak2) or control vector (Ad‐ctrl). (A) mRNA levels of α‐SMA and Col1a1 in primary HSCs. (B) Protein levels of α‐SMA, Type I collagen, and Jak2 in primary HSCs. (C–E) AAV8 carrying a Jak2 expression vector (AAV8‐Jak2) or control vector (AAV8‐ctrl) was injected into WT and CXCL14^−/− mice, followed by CCl₄ injection for 6 weeks. (C, D) Levels of ALT, AST and Hyp. (E) Masson and Sirius Red staining. Each value represents the mean ± SD of six experiments. ***p < .001.

FIGURE 8

Inhibition of CXCL14 ameliorates liver fibrosis in vivo. CCl₄‐induced liver fibrosis mice were treated with α‐CXCL14 (15 mg/kg) or isotype IgG (control) every 3 weeks for 12 weeks. (A) Experimental design. (B, C) Levels of ALT, AST and Hyp. (D) HE, Masson, Sirius Red, F4/80 and Ly6G staining. (E) mRNA levels of α‐SMA, Col1a1, Timp1 and Ctgf. Each value represents the mean ± SD of six experiments. **p < .01, ***p < .001.