LIX1-like protein drives hepatic stellate cell activation to promote liver fibrosis by regulation of chemokine mRNA stability

Dear Editor, Hepatic stellate cells (HSCs) play a key role in the fibrotic response, thus inactivating activated HSC could be a potential therapy for fibrosis. CCL20 expressed by HSCs and macrophages, may serve as a mediator of inflammation and fibrosis. LIX1L is a putative RNA-binding protein (RBP) that may play an important role in post-transcriptional gene regulation. However, the biological function of LIX1L in liver fibrosis remains unclear, we therefore aimed to characterize its functions in HSC activation and liver fibrosis. LIX1L expression was significantly upregulated in human cirrhotic liver (GSE25097), tumor tissue, and patients with fibrotic livers, and was positively correlated with liver fibrosis stage, level of inflammatory and α-SMA expression in cirrhotic liver (Supplementary Fig. s1a, b and Fig. 1a). Similarly, LIX1L expression was markedly increased in CCl4 or BDL induced mouse model of liver fibrosis (Fig. 1b). Next, Lix1l knockout (Lix1l) mice were generated and subjected to CCl4 and BDL treatment. H&E staining, Masson′s trichrome, and Sirius red staining, together with biochemical collagen quantification results showed decreased collagen deposition and pseudo lobular nodule formation in Lix1l mice. In addition, serum ALT and AST, proinflammation genes, and α-SMA protein expression were dramatically reduced in Lix1l mice compared to WT mice (Fig. 1c, d and Supplementary Fig. s2a, b). HSC and KC activation play important role in the process of liver fibrosis. The fibrogenic activity of HSC and hepatic macrophage infiltration as indicated by α-SMA and F4/80 expression was reduced in Lix1l mice (Fig. 1e). To determine the cellular distribution of LIX1L, we examined LIX1L expression in cultured primary HSCs, KCs, and hepatocytes. Primary cells isolated from healthy, CCl4 and BDL treated mice showed that LIX1L was more abundant in HSCs and KCs than in hepatocytes. Dual immunofluorescence staining of liver tissues displayed a co-localization of LIX1L with α-SMA-positive HSCs and F4/80-positive macrophages (Fig. 1f and Supplementary Fig. s3a, b). To confirm these results, we employed adeno-associated viral (AAV)-LIX1L-Flag to express LIX1L in hepatocytes of LIX1L knockout mice. Of note, LIX1L overexpression in hepatocytes failed to reverse the beneficial effect of LIX1L knockout on fibrosis (Supplementary Fig. s3c–f), indicating that LIX1L in HSC and KCs predominantly regulating liver fibrosis. HSC activation is key event in the development of liver fibrosis. We tested whether increased LIX1L regulated HSC activation and found that LIX1L levels gradually increased in primary HSCs during culture activation, with induction of α-SMA (Fig. 1g and Supplementary Fig. s4a). LIX1L deletion inhibited fibrogenic markers expression in primary HSCs and LX-2 cells and inhibited HSC activation in CCl4 challenged mice, while LIX1L overexpression further enhanced α-SMA and fibrogenic markers expression in LX-2 cells (Fig. 1h and Supplementary Fig. s4b–f). Together, these results suggest that LIX1L promotes liver fibrosis involving in regulation of HSC activation. To identify the mechanism(s) underlying LIX1L-induced HSC activation, we compared gene expression profiles of control and LIX1L knockdown LX-2 cells (Supplementary Fig. s6a). The results revealed that LIX1L knockdown reduced the expression of a wide spectrum of genes known to play critical roles in HSC activation and liver fibrosis (Fig. 1i). KEGG pathway analysis revealed that genes repressed by LIX1L knockdown were most enriched in the chemokine signaling pathway. Meanwhile, the expression of CCL20 and CCL2 was significantly reduced, among which CCL20 was reduced the most (Fig. 1j and Supplementary Fig. s6b, c). Ccl20 mRNA and CCL20 expression by Elisa assay were significantly reduced in LIX1L-deficient HSCs and KCs (Fig. 1k and Supplementary Fig. s6g, h), as well as the supernatant of cultured primary HSCs and KCs isolated from mice with BDL treatment (Fig. 1l), suggesting that LIX1L is required for CCL20 production in these cell types. Immunohistochemical analyses revealed that CCL20 staining was mostly in non-parenchymal cells, and reduced in Lix1l mice treated with CCl4 or BDL (Supplementary Fig. s6 i, j). In addition, CCL20 knockdown significantly reduced fibrotic gene expression in LX2 cells (Supplementary Fig. s6k). CCL20 could activate HSCs and reverse the inhibitory effect of LIX1L deletion on HSC activation (Supplementary Fig. s6l). Moreover, exogenous CCL20 could upregulate the expression of Lix1l in HSCs and KCs (Supplementary Fig. s6m). Given that endogenous CCL20 expression was decreased in Lix1l KCs compared to WT KCs. Therefore, in order to determine whether endogenous CCL20 affects the expression of LIX1L, we detected the expression of LIX1L in primary HSCs co-cultured with Lix1l KCs and WT KCs. We found that endogenous CCL20 could upregulate the expression of LIX1L (Supplementary Fig. s6n). Together, these data further suggest that LIX1L promotes HSC activation through the regulation of CCL20 signaling pathway. RNA immunoprecipitation assay indicated that CCL20 mRNA was enriched by LIX1L (Supplementary Fig. s7a). RNA pulldown assay showed that 3′ UTR region of CCL20 mRNA recruited LIX1L protein (Fig. 1m), suggesting that LIX1L may directly influence the expression of CCL20 mRNA at the post-transcriptional level. Indeed, Gene Set Enrichment Analysis (GSEA) identified LIX1L knockdown regulating RNA stability (Supplementary Fig. s7b), and LIX1L knockdown facilitated CCL20 mRNA degradation (Fig. 1n). Thus, LIX1L knockdown promoted CCL20 mRNA decay. The mRNAs of many chemokines and cytokines are targeted for rapid degradation through AU-rich elements (AREs) located in their 3′ UTR. We tested whether AREs are required for LIX1L binding, and found that these AREs in 3′ UTR of CCL20 mRNA are essential for

inflammation. Indeed, LIX1L up-regulation was observed in HSCs and KCs isolated 48 from mice treated with CCl4 and BDL, but not in primary hepatocytes. Together, these 49 findings support the conclusion that LIX1L knockout prevents progression of liver 50 fibrosis if applied onset of chronic liver injury. LIX1L may be a potential therapeutic 51 target for liver fibrosis.

52
One critical question is how LIX1L promotes liver fibrosis. HSCs activation is a 53 central driver of fibrosis in experimental models and human liver injury 2 . LIX1L 54 deficiency impairs HSC activation in vitro and in vivo. However, LIX1L knockdown 55 did not affect apoptosis or proliferation of LX-2 cells (Supplementary Fig. s4 g,h).

58
Further pathway analysis showed that LIX1L knockdown highly affected the 59 chemokine signaling pathway, and CCL20 is the most decreased chemokine. CCL20 60 could activate HSCs and is involved in chronic liver inflammation and fibrosis 3,4 . 61 During the development of liver fibrosis, HSCs, KCs, hepatocytes, LSECs and other 92 cells are involved in the regulation of the occurrence and development of fibrosis. To 93 fully elucidate the role of LIX1L in liver fibrosis is will be necessary to characterize 94 and identify its cell-type-specific functions. HSCs are the source and target of 95 chemokines 7-9 , and the interaction of HSCs with proinflammatory cells such as KC is a 96 crucial event in HSC activation and fibrosis, where chemokines and their receptors are 97 likely to serve as important contributors to this interaction 10 . Thus, cell-type specific 98 expression of LIX1L may assist in determining the outcome of liver disease. LIX1L is 99 highly expressed in activated HSCs and KCs. CCL20, mainly expressed by 100 macrophages and HSCs, exerts proinflammatory and profibrogenic effects in liver 101 pathological processes 11,12 . Our present data provide compelling evidence that LIX1L 102 regulates CCL20 expression in KCs and HSCs. Furthermore, we identified a paracrine 103 effect of LIX1L-mediated CCL20 secretion produced by KCs on HSC activation. In 104 this study, in the in vitro primary co-culture experiment, we found that the expression It was consistent with in vitro experiment, the expression of fibrogenic genes in Lix1l -

114
/-BM→Lix1l -/was lower than Lix1l -/-BM→ Lix1l -/-, as well as than Lix1l -/- BM→WT,115 demonstrating that LIX1L signaling in liver resident cells strongly contributes to the 116 pathogenesis of liver fibrosis. The results of Lix1l chimeric mice indicated that LIX1L 117 in liver resident cells is a major contributor to liver fibrosis. Since KCs are heterogenic 118 cell population from the BM and liver resident, we conclude that ablation of LIX1L 119 signaling in liver resident HSCs and KCs was critical for inhibition of liver fibrosis. 120 Therefore, targeting LIX1L in liver resident cells may serve as a target for the treatment 121 of liver fibrosis in the future.   Liver specimens were fixed in 10% neutral buffered formalin, embedded in paraffin and cut into 4μm sections. Next, the specimens were deparaffinized, hydrated and 156 stained by standard methods. To examine hepatic morphology and assess liver fibrosis, 157 H&E, Sirius Red and Masson's trichrome staining were performed, respectively. Tissue 158 sections were immunostained for α-SMA (Abcam, ab7817), F4/80 (Abcam, ab6640), 159 and CCL20 (Abcam, ab9829), the sections were scanned, and the images were then

306
The results are presented as mean ± SEM. Statistical differences between two groups 307 were analyzed by the unpaired Student's t test with a two-tailed distribution.

308
Differences between multiple groups of data were analyzed by one-way ANOVA with Cytokine Growth Factor Rev 14, 409-426 (2003  Hepatocyte apoptosis (left) and proliferation (right) in WT mice and Lix1l -/mice 387 following CCL4 for 48, 72h or BDL for 3 days treatment (n=8 per group).