The role of forkhead box M1-methionine adenosyltransferase 2 A/2B axis in liver inflammation and fibrosis

Yang, Bing; Lu, Liqing; Xiong, Ting; Fan, Wei; Wang, Jiaohong; Barbier-Torres, Lucía; Chhimwal, Jyoti; Sinha, Sonal; Tsuchiya, Takashi; Mavila, Nirmala; Tomasi, Maria Lauda; Cao, DuoYao; Zhang, Jing; Peng, Hui; Mato, José M.; Liu, Ting; Yang, Xi; Kalinichenko, Vladimir V.; Ramani, Komal; Han, Jenny; Seki, Ekihiro; Yang, Heping; Lu, Shelly C.

doi:10.1038/s41467-024-52527-8

Download PDF

Article
Open access
Published: 27 September 2024

The role of forkhead box M1-methionine adenosyltransferase 2 A/2B axis in liver inflammation and fibrosis

Nature Communications volume 15, Article number: 8388 (2024) Cite this article

7308 Accesses
5 Citations
73 Altmetric
Metrics details

Subjects

Abstract

Methionine adenosyltransferase 2 A (MAT2A) and MAT2B are essential for hepatic stellate cells (HSCs) activation. Forkhead box M1 (FOXM1) transgenic mice develop liver inflammation and fibrosis. Here we examine if they crosstalk in male mice. We found FOXM1/MAT2A/2B are upregulated after bile duct ligation (BDL) and carbon tetrachloride (CCl₄) treatment in hepatocytes, HSCs and Kupffer cells (KCs). FDI-6, a FOXM1 inhibitor, attenuates the development and reverses the progression of CCl₄-induced fibrosis while lowering the expression of FOXM1/MAT2A/2B, which exert reciprocal positive regulation on each other transcriptionally. Knocking down any of them lowers HSCs and KCs activation. Deletion of FOXM1 in hepatocytes, HSCs, and KCs protects from BDL-mediated inflammation and fibrosis comparably. Interestingly, HSCs from Foxm1^Hep−/−, hepatocytes from Foxm1^HSC−/−, and HSCs and hepatocytes from Foxm1^KC−/− have lower FOXM1/MAT2A/2B after BDL. This may be partly due to transfer of extracellular vesicles between different cell types. Altogether, FOXM1/MAT2A/MAT2B axis drives liver inflammation and fibrosis.

Focal adhesion kinase promotes aerobic glycolysis in hepatic stellate cells via the cyclin D1/c-Myc/MCT-1 pathway to induce liver fibrosis

Article Open access 07 February 2025

Inhibition of intrahepatic monocyte recruitment by Cenicriviroc and extracellular matrix degradation by MMP1 synergistically attenuate liver inflammation and fibrogenesis in vivo

Article Open access 23 July 2024

MyD88 in hepatic stellate cells enhances liver fibrosis via promoting macrophage M1 polarization

Article Open access 28 April 2022

Introduction

Hepatic fibrosis is characterized by the accumulation of extracellular matrix (ECM) components such as collagens type I (COL1A1) and type III, which may further lead to cirrhosis and liver cancer¹. Activation of hepatic stellate cells (HSCs) is a crucial event in liver fibrosis. Forkhead box M1 (FOXM1) is a member of the forkhead box family with winged helix DNA binding domain that works as a crucial transcription factor in proliferation and oncogenesis^2,3. Recent studies have reported that FOXM1 could promote multiple organ fibrosis including liver fibrosis^4,5,6. Furthermore, Kurahashi et al.³ demonstrated that overexpression of FOXM1 in hepatocytes caused spontaneous liver injury, inflammation, fibrosis, and hepatocellular carcinoma (HCC). Forkhead domain inhibitory-6 (FDI-6), a small molecule that selectively targets the FOXM1 DNA-binding domain, could inhibit the transcription of FOXM1 target genes⁷. FDI-6 inhibited corneal inflammation and fibrosis after alkali burn in rats⁸ and attenuated subconjunctival fibrosis in a trabeculectomy rabbit model⁹. These studies suggest that targeting FOXM1 may be a potential therapeutic strategy in liver inflammation and fibrosis. However, this has not been examined.

Methionine adenosyltransferase (MAT) is an essential enzyme that catalyzes the biosynthesis of S-adenosylmethionine (SAMe)¹⁰. MAT2A encodes for the α2 catalytic subunit and MAT2B gene encodes the β regulatory subunit of the MATII isoenzyme¹⁰. MAT2A and MAT2B are expressed in extrahepatic tissues and non-parenchymal cells of the liver, such as HSCs and Kupffer cells (KCs)¹¹. MAT2A expression is induced in cultured activated HSCs, and in livers treated with carbon tetrachloride (CCl₄) or thioacetamide (TAA)¹². MAT2A and MAT2B affect HSCs activation through changes in SAMe levels and extracellular-regulated kinase and phosphoinositide 3-kinase signaling mechanisms¹⁰. In addition, MAT2B siRNA treatment reduced collagen level and inhibited HSCs activation in TAA-induced liver fibrosis in mice¹³. Wang et al.¹⁴ also found that knockdown of MAT2A alleviated CCl₄ and transforming growth factor-beta 1 (TGF-β1)-induced HSCs activation, whereas overexpression of MAT2A facilitated hepatic fibrosis. Here we investigated whether there is a crosstalk between FOXM1, MAT2A and MAT2B; to better understand the role of FOXM1 in a cell-type specific manner, we established Foxm1^Hep−/−, Foxm1^HSC−/−, and Foxm1^KC−/− mice and examined the role of the FOXM1/MAT2A/MAT2B axis in vitro and vivo.

In this work, we found that FOXM1, MAT2A and MAT2B expression are upregulated in hepatocytes, HSCs, and KCs after bile duct ligation (BDL) in mice. Treatment with FDI-6 significantly attenuates the development of liver injury and fibrosis but more remarkably, reverses liver fibrosis that was already established. We found FOXM1, MATα2 and MAT2β directly interact, and they positively regulate each other at the transcriptional level via FOX elements present in their promoters. Knocking down any of the three inhibits TGF-β1-induced HSCs activation as well as lipopolysaccharide (LPS)-induced macrophage activation. Interestingly, deleting FOXM1 from hepatocytes, HSCs or KCs all protect from liver injury and fibrosis induced by BDL comparably. This may be in part due to exchange of FOXM1, MATα2 and MAT2β in secreted extracellular vesicles (EVs) amongst the different cell types. Taken together, our results suggest targeting any of the components of the FOXM1/MAT2A/MAT2B axis may be a potential therapeutic strategy in liver fibrosis.

Results

FOXM1/MAT2A/MAT2B axis is induced in liver fibrosis

MATα2, MAT2β and FOXM1 are involved in liver fibrosis^3,13,14. Here we found the upregulation of FOXM1, MAT2A and MAT2B in hepatic cells and small bile ducts in human and murine fibrotic livers (Supplementary Fig. 1a–e). Hepatocytes express lower levels of Mat2a and Mat2b as compared to HSCs and KCs at baseline, but they are more induced following BDL and CCl₄ treatments (Supplementary Fig. 1e). Interestingly, Foxm1 was expressed at a lower level in HSCs than in hepatocytes and KCs but it was induced in all three cell types after BDL and CCl₄ treatment (Supplementary Fig. 1e). These proteins are also induced in primary biliary cholangitis (PBC) and primary sclerosing cholangitis (PSC) (Supplementary Fig. 2a–c). Furthermore, FOXM1 expression increased from day 1 to day 14 after bile duct ligation (BDL) while expression of MATα2 and MAT2β increased from day 2 to day 14 after BDL (Fig. 1a), suggesting FOXM1 may be upstream of MATα2 and MAT2β. Primary HSCs isolated from BDL mice at day 7 showed higher cytoplasmic and nuclear FOXM1, MATα2, and MAT2β staining compared with sham control (Fig. 1b). FDI-6 treatment reduced the mRNA and protein levels of FOXM1, MAT2A, and MAT2B in the day-5 culture activated HSCs and LX-2 cell line (Fig. 1c, d, densitometry is shown in Supplementary Fig. 3a, b). FDI-6 also reduced cytoplasmic and nuclear staining of FOXM1, MATα2, and MAT2β in LX-2 cells (Fig. 1e), and in BDL livers (Fig. 1f, densitometry is shown in Supplementary Fig. 3c, d). FDI-6 also lowered the proliferation (Fig. 1g) and migration (Fig. 1h) of LX-2 cells.

**Fig. 1: Expression of FOXM1/MAT2A/MAT2B during bile duct ligation (BDL) and after FDI-6 treatment.**

FDI-6 in the development and progression of liver fibrosis

To examine how inhibiting FOXM1 will impact on CCl₄-induced liver fibrosis, we divided the mice into two groups (prevention and treatment). In the prevention group, FDI-6 was administered from the start of CCl₄ treatment for three weeks. In the treatment group, FDI-6 was started after liver fibrosis was already established (after three weeks of CCl₄) and continued with CCl₄ for two additional weeks. In these two groups, immunohistochemistry (IHC) showed that FDI-6 treatment reduced the expression of COL1A1, F4/80, α-SMA, FOXM1, MATα2, and MAT2β induced by CCl₄ (Fig. 2a). FDI-6 also reduced the levels of alanine transaminase (ALT), aspartate transferase (AST), and hydroxyproline induced by CCl₄ (Fig. 2b–d). Furthermore, FDI-6 reduced the mRNA (Fig. 2e) and protein levels (Fig. 2f, densitometry is shown in Supplementary Fig. 3e) of FOXM1, MAT2A, MAT2B, α-SMA, F4/80, and COL1A1. Note that FDI-6 lowered fibrosis markers such as hydroxyproline and COL1A1 to below the levels seen with CCl₄ alone for three weeks, which suggests it can reverse already existing fibrosis. CCl₄ treatment also increased the interaction between FOXM1 and MATα2 and MAT2β, which was reduced by FDI-6 (Supplementary Fig. 3f, g).

**Fig. 2: Prevention and treatment roles of FDI-6 in carbon tetrachloride (CCl₄)-induced liver fibrosis.**

Reciprocal positive regulation between FOXM1/MAT2A/MAT2B

To better understand how they crosstalk, we examined the relationship between FOXM1 and MAT2A/2B in LX-2 cells and primary cholangiocytes isolated from BDL livers. FOX binding sites are present in the human MAT2A (Supplementary Fig. 4a), MAT2B (Supplementary Fig. 4b), and FOXM1 promoter regions (Supplementary Fig. 5). Different fragments of these human promoter constructs were transfected in LX-2 cells and primary cholangiocytes and treatment with FDI-6 (5 μM) or FOXM1 siRNA significantly decreased MAT2A promoter activity with maximal effect seen with promoter fragment −271/+60 in LX-2 cells (Fig. 3a, left), MAT2B promoter activity with maximal effect seen with fragment −250/+3 (Fig. 3b, left), and FOXM1 promoter −1333/+107 fragment (Fig. 3c, left). Similar attenuation was seen in all three promoters in primary cholangiocytes after these treatments (Fig. 3a, right for Mat2a, Fig. 3b, right for Mat2b, Fig. 3c, right for Foxm1), suggesting important elements are within those regions of the MAT2A/MAT2B/FOXM1 promoters. Mutation at each of the FOX binding sites significantly attenuated the ability of FDI-6 and FOXM1 siRNA treatments to lower MAT2A/MAT2B/FOXM1 promoter activities (Fig. 3d–f). These results show that FOXM1 regulates MAT2A, MAT2B, and FOXM1 mainly via FOX binding sites.

**Fig. 3: FOXM1, MATα2 and MAT2β regulate each other positively at the transcriptional level.**

Since we found interaction between FOXM1, MATα2, and MAT2β in liver lysates (Supplementary Fig. 3f), we next examine if they interact at the two FOX binding sites of the FOXM1 promoter using chromatin immunoprecipitation (ChIP) and sequential ChIP (Seq-ChIP). We found that FOXM1 but not MATα2 or MAT2β can bind to the FOX region (Fig. 3g). However, in the presence of FOXM1, MATα2 and MAT2β co-occupy this region on Seq-ChIP (Fig. 3g). Furthermore, MAT2A, MAT2B, and FOXM1 overexpression (OV) led to an increase in FOXM1 and MATα2/2β binding and the opposite was true with MAT2A, MAT2B, and FOXM1 knockdown in LX-2 cells (Fig. 3g, h). FDI-6 treatment had the same results as FOXM1, MAT2A and MAT2B knockdown in LX-2 cells and cholangiocytes (Supplementary Fig. 6a, b). To confirm that these proteins were binding to the FOX element, we used electrophoretic mobility shift assay (EMSA), which confirmed that FOXM1, MAT2A and MAT2B siRNAs decreased and their overexpression increased protein binding to the FOX elements (Fig. 3i). Recombinant protein (r) of MATα2 or MAT2β cannot bind to the FOX element by itself. However, they can bind when combined with rFOXM1 as shown by supershift assay (Fig. 3j). Using co-immunoprecipitation (Co-IP) with purified recombinant proteins we found direct interaction between FOXM1 and MATα2 and MAT2β (Fig. 3k), and BDL increased their interaction in the liver (Fig. 3l). These results suggest these three proteins all interact at the FOX elements to activate the transcription of each other. Consistently, primary cholangiocytes isolated from BDL and sham control livers showed BDL increased cytoplasmic and nuclear FOXM1, MATα2 and MAT2β content (Supplementary Fig. 6c). Furthermore, FOXM1, MAT2A, and MAT2B overexpression increased each other’s protein expression and mRNA levels in LX-2 cells (Supplementary Fig. 7a), and the opposite was true with FOXM1, MAT2A, and MAT2B knockdown (Supplementary Fig. 7b). Consistent with the importance of FOXM1/MAT2A/MAT2B in HSCs activation, their siRNAs inhibited while overexpression raised migration of LX-2 cells (Supplementary Fig. 8). Lastly, FOXM1 (Supplementary Fig. 9a), MAT2A (Supplementary Fig. 9b) and MAT2B (Supplementary Fig. 9c) mRNA levels are all elevated in human liver cirrhosis (n = 31) and they positively correlate with each other (Supplementary Fig. 9d–f). Their high expressions in liver cirrhosis correlates with hepatitis B virus (HBV) infection (Supplementary Table 1).

Foxm1 ^Hep−/− mice have lower BDL-induced injury and fibrosis

Hepatocellular injury is a central mechanism of inflammation and disease progression in chronic liver diseases¹⁵. Hepatocytes, activated HSCs, and KCs cooperate in the establishment and resolution of liver fibrosis¹⁶. To explore how hepatocytes’ FOXM1 is involved in liver fibrosis, we generated Foxm1^Hep−/− mice (Supplementary Fig. 10a). Foxm1^Hep−/− mice subjected to BDL have lower CK19 (biliary/progenitor cell marker), α-SMA (a biomarker for myofibroblast differentiation), F4/80 (a marker of macrophages) and Sirius red staining (Fig. 4a, b), hydroxyproline (Fig. 4c), ALT (Fig. 4d), and AST (Fig. 4e) levels, lower F4/80 positive cells (Fig. 4f), CK19 and α-SMA area/total area (Fig. 4g) as compared to Flox controls after BDL. We next measured the expression levels of FOXM1, MATα2 and MAT2β in a cell-type specific manner from Foxm1^Hep−/− mice at baseline and after BDL. Foxm1 mRNA and protein are not detected in hepatocytes or cholangiocytes (Fig. 4h–j). Foxm1^Hep−/− mice have reduced protein and mRNA levels of Mat2a, and Mat2b in hepatocytes (Fig. 4h, i), cholangiocytes (Fig. 4h–j), and HSCs (Fig. 4h–k) after BDL but unchanged in KCs (Fig. 4h–l) as compared to Flox controls after BDL. See Supplementary Fig. 11a–d for a summary of densitometric changes of Fig. 4h.

**Fig. 4: Response of hepatocyte specific *Foxm1* knockout *(Foxm1*^Hep−/−) mice to BDL.**

Foxm1 ^HSC−/− mice have lower BDL-induced injury and fibrosis

Activated HSCs release ECM, which is considered a key event in liver fibrosis. To better understand the role of HSCs FOXM1 in liver fibrosis, we generated Foxm1^HSC−/− mice (Supplementary Fig. 10b). IHC results showed Foxm1^HSC−/− mice have lower α-SMA and Sirius Red staining (Fig. 5a, b) and hydroxyproline level (Fig. 5c) after BDL, but F4/80 and CK19 staining were similar to Flox controls after BDL (Fig. 5a). ALT (Fig. 5d) and AST (Fig. 5e) levels were lower, F4/80 positive number (Fig. 5f) and CK19 area/total area (Fig. 5g) were similar but α-SMA area/total area was lower than Flox controls after BDL (Fig. 5g). FOXM1 protein (Fig. 5h) and mRNA (Fig. 5k) in HSCs after BDL are undetectable from Foxm1^HSC−/−. Foxm1^HSC−/− mice have lower FOXM1, MATα2 and MAT2β protein (Fig. 5h) and mRNA (Fig. 5i) levels in hepatocytes as compared to Flox controls after BDL but cholangiocytes (Fig. 5h–j) and KCs (Fig. 5h–l) were similar. See Supplementary Fig. 11e–h for a summary of densitometric changes of Fig. 5h.

**Fig. 5: Response of hepatic stellate cell (HSC)-specific Foxm1 knockout (*Foxm1*^HSC−/−) mice to BDL.**

Foxm1 ^KC−/− mice have lower BDL-induced injury and fibrosis

KCs are resident macrophages that are localized within the lumen of the liver sinusoids, which play a vital role in liver inflammation¹⁶. In response to hepatocyte injury, KCs become activated and express cytokines and signaling molecules, such as tumor necrosis factor α (TNF-α), interleukin-1β (IL-1β), and interleukin-6 (IL-6)¹⁶. To examine the role of macrophage-derived FOXM1, we created Foxm1^KC−/− mice (Supplementary Fig. 10c). IHC results showed Foxm1^KC−/− mice have lower F4/80, Sirius red, and α-SMA staining after BDL but no change in CK19 staining (Fig. 6a, b, f, g). They have lower levels of hydroxyproline (Fig. 6c), ALT (Fig. 6d) and AST (Fig. 6e). FOXM1 protein (Fig. 6h) and mRNA (Fig. 6l) in KCs after BDL are undetectable in Foxm1^KC−/− mice. However, Foxm1^KC−/− mice have lower protein (Fig. 6h) and mRNA level of Foxm1, Mat2a, and Mat2b in hepatocytes (Fig. 6i) and HSCs (Fig. 6k) as compared to Flox controls after BDL, except for cholangiocytes (Fig. 6j), which were similar. See Supplementary Fig. 12a–d for a summary of densitometric changes of Fig. 6h.

**Fig. 6: Response of Kupffer cell (KC)-specific Foxm1 knockout (*Foxm1*^KC−/−) mice to BDL.**

FOXM1-MAT2A-MAT2B interplay drives fibrosis and inflammation

TGF-β1 promotes liver fibrosis by activating HSCs¹⁷. SMAD3 is the core transcription factor that mediates TGF-β signaling¹⁸. To see if MAT2A and MAT2B are important in FOXM1-driven fibrosis, we treated LX-2 cells with TGF-β1 (20 ng/ml) and found FOXM1, MATα2, MAT2β and p-SMAD3 levels were increased three hours after treatment, along with an increase in fibrosis markers such as α-SMA and COL1A1 (Fig. 7a, see Supplementary Fig. 12e for densitometric changes). Knocking down FOXM1, MAT2A, or MAT2B reduced protein expression of MATα2, MAT2β, SMAD3, p-SMAD3, α-SMA and COL1A1 (Fig. 7b, see Supplementary Fig. 12f for densitometric changes). To determine the effect of the FOXM1/MAT2A/MAT2B axis on TGF-β1-mediated HSC activation, we treated the LX-2 cells with MAT2A siRNA plus TGF-β1 or MAT2B siRNA plus TGF-β1. Knocking down either MAT2A or MAT2B attenuated TGF-β1-mediated induction of FOXM1, MATα2, MAT2β, SMAD3, p-SMAD3, α-SMA and COL1A1. FOXM1 overexpression further augmented TGF-β1-induced protein expression of MATα2, MAT2β, SMAD3, p-SMAD3, α-SMA and COL1A1, but not if MAT2A or MAT2B were silenced (Fig. 7c, see Supplementary Fig. 13a for densitometric changes). These results indicate the FOXM1/MAT2A/MAT2B axis is important in liver fibrosis induced by the TGF-β1 signaling pathway and MAT2A and MAT2B are required for FOXM1 to exert its pro-fibrogenic effect.

**Fig. 7: Roles of FOXM1, MAT2A and MAT2B in transforming growth factor (TGF)-β1 signaling in LX-2 cells and LPS effects in macrophages, and their presence in EVs.**

Macrophages have emerged as the central players in sustaining and amplifying chronic inflammation, which is the hallmark of liver fibrosis¹⁹. Lipopolysaccharide (LPS) stimulates RAW 264.7 macrophages to release pro-inflammatory activation, such as TNF-α and IL-6²⁰. We next examined the FOXM1/MAT2A/MAT2B axis in inflammation, first using RAW 246.7 cells, and then confirming the results in primary KCs. Interestingly, knocking down FOXM1, MAT2A or MAT2B reduced the protein expression of FOXM1, MATα2, MAT2β, TNF-α, and IL-6 at baseline and attenuated or completely blocked LPS-induced expression of these proteins (Fig. 7d, see Supplementary Fig. 13b for densitometric changes). Importantly, FOXM1 overexpression recapitulated the effects of LPS on TNF-α, and IL-6 but silencing MAT2A or MAT2B significantly attenuated FOXM1-mediated induction (Fig. 7e, see Supplementary Fig. 13c, d for densitometric changes). Silencing Foxm1, Mat2a or Mat2b attenuated LPS-induced TNF-α, and IL-6 release (Supplementary Fig. 13e). These results indicate the FOXM1/MAT2A/MAT2B axis also participates in inflammation.

Liver cells secrete EVs containing FOXM1/MATα2/MAT2β

To understand why HSCs from Foxm1^Hep−/−, hepatocytes from Foxm1^HSC−/− and HSCs and hepatocytes from Foxm1^KC−/− all have lower levels of FOXM1, MATα2 and MAT2β after BDL, we examined the possibility that these proteins are secreted in EVs and transferred to other cells. First, we isolated and cultured hepatocytes, HSCs, and KCs from WT and their respective KOs at day 14 of BDL. The following day EVs were isolated from culture medium and used to treat KO HSCs or hepatocytes with the following combinations: 1. HSCs from Foxm1^HSC−/− treated with EVs from hepatocytes of WT and Foxm1^Hep−/−, 2. hepatocytes from Foxm1^Hep−/− treated with EVs from HSCs of WT and Foxm1^HSC−/−, 3. HSCs from Foxm1^HSC−/− treated with EVs from KCs of WT and Foxm1^KC−/−, and 4. hepatocytes from Foxm1^Hep−/− treated with EVs from KCs of WT and Foxm1^KC−/−. We found cells treated with EVs from cell-type specific knockouts all resulted in lower expression of FOXM1, MATα2, and MAT2β as compared to cells treated with EVs from WT cells after BDL (Fig. 8a–d, see Supplementary Fig. 14a–d for densitometric changes). We also evaluated whether hepatocytes and HSCs can transfer EVs to KCs in vitro. Contrary to what we observed in vivo, the treatment of KCs with EVs from hepatocytes or HSCs from Flox mice after BDL showed an increase in the expression of FOXM1, MATα2 and MAT2β when compared to EVs from hepatocytes from Foxm1^Hep−/− or HSCs from Foxm1^HSC−/− after BDL (Fig. 8e, f, see Supplementary Fig. 14e, f for densitometric changes). After BDL, we also found that EVs isolated from hepatocytes of Foxm1^Hep−/−, HSCs of Foxm1^HSC−/−, and KCs of Foxm1^KC−/− have lower protein expression of FOXM1, MATα2, and MAT2β (Supplementary Fig. 14g) as compared to EVs from respective WT cells. Characterization of these EVs suggest they are most likely exosomes based on the size distribution (Supplementary Fig. 14h). These results suggest that FOXM1, MATα2 and MAT2β are released by hepatocytes, HSCs and KCs in EVs and they can influence the expression of these proteins in other liver cell types.

**Fig. 8: FOXM1, MATα2, MAT2β are secreted in EVs and exchanged between hepatocytes, HSCs, and KCs.**

Discussion

Liver fibrosis is associated with high morbidity and mortality, and liver transplantation is currently the most effective way to cure liver fibrosis²¹. Although FOXM1 is involved in multiple diseases²², such as pulmonary fibrosis, diabetes, vascular diseases, and psoriasis, earlier studies regarding FOXM1 mainly focused on its function in oncogenesis^{2,3,23,24,25,26}. In recent studies, Kurahashi et al.³ identified hepatocyte FOXM1 as a crucial driver of chronic liver inflammation as hepatocyte specific FOXM1 transgenic mice developed spontaneous liver inflammation, fibrosis, and HCC. However, the level of expression in a transgenic model is not physiologic so the role of endogenous FOXM1 in liver fibrogenesis is still not confirmed. FDI-6 could inhibit corneal inflammation, and subsequently attenuated subconjunctival fibrosis^8,9. However, how FOXM1 causes liver inflammation and fibrosis is uncertain and whether inhibiting FOXM1 can be effective in the treatment of existing liver fibrosis has not been investigated. We and others have shown MAT2A and MAT2B are essential for HSCs activation and liver fibrosis^10,12,13,14. In the current work, we examined if there is interplay between FOXM1 and MAT2A/MAT2B. We also investigated the contribution of cell-type-specific FOXM1 expression in liver inflammation and fibrosis.

Consistent with previous reports, the expression of FOXM1, MAT2A and MAT2B is induced in human and murine liver fibrosis. Following BDL, the cytoplasmic and nuclear content of all three proteins increased. MAT2A is best known as the gene that encodes the α2 catalytic subunit of MATII, which catalyzes the biosynthesis of SAMe in non-hepatic cells and non-parenchymal cells of the liver¹¹. We and others have shown that MATα2 is present in the nucleus where it can regulate gene expression via epigenetics and as a transcription factor^27,28. MAT2B is best known as the gene encoding for the regulatory subunit β that regulates the activity of MATII by lowering its K_m for methionine and K_i for SAMe¹⁰. MAT2A and MAT2B are often regulated in parallel, and both are overexpressed in many cancers¹⁰. MATα2 and MAT2β also stabilize each other²⁹. Here we found they also regulate each other positively at the transcriptional level and form a positive regulatory loop with FOXM1 so that knocking down any of them can be effective in stopping the progression of liver fibrosis. Consistently, FDI-6 treatment, which is a specific inhibitor of FOXM1, lowered the cytoplasmic and nuclear expression of FOXM1, MATα2 and MAT2β in activated HSCs, LX-2 cells, and cholangiocytes after BDL. Importantly, FDI-6 was effective in preventing as well as stopping the progression and regressing already established liver fibrosis that was induced by three weeks of CCl₄ treatment. These results suggest FDI-6 may be a potential therapy in liver fibrosis through inhibiting the FOXM1/MAT2A/2B axis.

FOXM1 activates transcription of many oncogenes, such as cell cycle-related genes, and interacts with proteins to enhance the oncogenesis via its winged helix DNA-binding domain^3,30. We also showed that FOXM1 activates NF-κB and forms a complex with NF-κB to bind to the FOX elements of the FOXM1 and MAT1A promoters to activate the former but repress the latter, thereby explaining why FOXM1 and MAT1A exert reciprocal negative regulation against each other². NF-κB subunit p65 was reported to trans-activate the MAT2A promoter^10,14, and we have shown that overexpression of MAT2A increased NF-κB reporter activity²⁸, suggesting NF-κB and MAT2A form a positive regulatory loop. In this work, we found that FOXM1, MAT2A and MAT2B exert reciprocal positive regulation on each other via FOX binding sites present in their promoters. MATα2 and MAT2β are not able to bind to the FOX element alone, but they were able to in the presence of FOXM1. This is because they can directly interact with FOXM1, as demonstrated by using recombinant proteins in Co-IP experiments. Interestingly, FDI-6 not only lowered the expression of the FOXM1, MAT2A and MAT2B, it also lowered their interaction. Human cirrhosis tissues also confirmed direct correlation between FOXM1, MAT2A, and MAT2B mRNA levels and upregulation of all three at the mRNA level. Interestingly, high levels of all three were noted in those with HBV infection. One plausible explanation is that the X protein of HBV was shown to increase MAT2A expression at the transcriptional level³¹. This could then further increase the expression of FOXM1 and MAT2B.

Hepatocytes damage underlies multiple acute and chronic liver diseases and contribute to disease progression^15,32, such as inflammation and liver fibrosis. HSCs activation forms the basis of liver fibrosis. KCs also enhance hepatic fibrosis by promoting activated HSCs survival in an NF-κB-dependent manner^16,33. In addition, IL-1 and TNF-α mediate the crosstalk between KCs and HSCs¹⁶. To better understand the role of FOXM1 in a cell-type specific manner, we constructed Foxm1^Hep−/−, Foxm1^HSC−/−, and Foxm1^KC−/− mice. We found that all three cell-type-specific knockouts were able to attenuate liver injury as measured by ALT and AST to comparable degrees and lowered liver fibrosis induced by BDL. Foxm1^Hep−/− mice also had less F4/80 number and ductular proliferation, which were unchanged in Foxm1^HSC−/− mice whereas Foxm1^KC−/− mice had less F4/80 number but no change in ductular proliferation. These results seem to suggest lowering F4/80 number or ductular proliferation were not the key determinants of protection in this model.

Unexpectedly we found deletion of Foxm1 in a cell-type-specific manner influenced the expression of FOXM1 in other cell types after BDL. Although we cannot rule out the possibility that there was leakiness in the Cre-lox system, our results suggest another explanation, namely exchange/transfer of FOXM1, MATα2 and MAT2β amongst different cell types via EVs. Consistently, HSCs and hepatocytes from respective knockouts had lower levels of FOXM1, MATα2 and MAT2β when treated with EVs from knockouts as compared to WTs after BDL. Indeed, EVs and their cargo transfer various cellular materials and signals (RNA, DNA, proteins, and lipids) between neighboring or distant cells, playing vital roles in the development of metabolic diseases and cancer^34,35,36,37. Transfer of EVs enriched in FOXM1, MATα2 and MAT2β between hepatocytes, HSCs, and KCs can maintain sustained activation of HSCs and KCs. Other studies have shown hepatocytes, HSCs, and KCs interaction via EVs^38,39. We cannot exclude the possibility that other components of the EVs are also involved to modulate the changes in the expression of these three proteins.

Also, it is interesting that in vivo after BDL, hepatocyte-specific and HSC-specific Foxm1 deletion attenuated Foxm1/Mat2a/Mat2b induction in each other but they had no influence on their expression in KCs. This is in contrast to in vitro treatment of KCs with EVs from hepatocytes or HSCs after BDL that clearly showed an increase in the expression of FOXM1, MATα2 and MAT2β as compared to EVs from KO hepatocytes or HSCs. One consideration is the spatial relationship of these cells. For instance, hepatocytes and HSCs are in close contact with each other but KCs are in the sinusoids, separated from the other two cell types by sinusoidal endothelial cells. However, KC-specific Foxm1 KO mice had lower expression of Foxm1/Mat2a/Mat2b in hepatocytes and HSCs after BDL. One possibility is that Foxm1 deleted KCs are releasing less proinflammatory cytokines which are known to induce Foxm1 and Mat2a^2,10. Another observation is that In vivo the cholangiocytes do not appear to be influenced by the EVs released by the other three cell types. At present we are not sure of the underlying mechanisms, but one speculation is how the different cells receive their blood supply. Bile duct epithelial cells receive their blood supply from the hepatic artery⁴⁰. It is possible that hepatocytes release EVs mainly into sinusoidal blood under our experimental conditions.

Since FOXM1, MAT2A and MAT2B are positively regulating each other and their encoded proteins all physically interact, we resorted to testing how essential they are in fibrogenesis and inflammation using overexpression and knockdown. TGF-β1 pathway plays an essential role in the pathogenesis of liver fibrosis and it is known to activate NF-κB subunit p65, which contributes to TGF-β1-induced HSCs activation^10,14. Macrophage-derived TGF-β1 is the most potent known fibrogenic agonist¹⁶. LPS increases liver injury and inflammation via the release of inflammatory cytokines⁴¹, such as TNF-α and IL-6. We found FOXM1, MATα2 and MAT2β, α-SMA and COL1A1 were all increased at three hours and since TGF-β1 activates NF-κB, which activates both FOXM1² and MAT2A¹⁴, they may be induced simultaneously. However, after BDL, the induction in FOXM1 occurred slightly earlier, suggesting FOXM1 is induced first in vivo. Knocking down any of the three reduced fibrogenic markers and although FOXM1 overexpression can augment TGF-β1 signaling, this requires participation of MAT2A and MAT2B. A similar scenario also occurs with LPS-induced proinflammatory cytokines in macrophages, where knocking down any of the three attenuated/eliminated LPS’ effect and MAT2A and MAT2B are required for FOXM1 to exert its full pro-inflammatory effect. These results support the notion that the FOXM1-MAT2A-MAT2B axis is essential for liver fibrosis and inflammation.

Methods

The study protocol was approved by the Institutional Animal Care and Use Committee of CSMC (No. 8850).

Materials and reagents

See Supplementary Table 2 for a list of the reagents and sources.

Human samples

Microarrays for 16 healthy human liver (HHL), 24 cases of hepatitis and 28 cases of cirrhosis tissues were from US biolab Inc. (Cat# DLV20812A, Rockville, MD, https://usbiolab.com/index.php/tissue-array/product/liver/DLV20812a). Four cases of PSC and three cases of PBC human liver specimens obtained from liver biopsy from 2014–2019, and five healthy liver tissues obtained from surgical resection for patients suffering from intrahepatic ductal stones from 2018-2019 were obtained from the department of Pathology at the Xiangya Hospital Central South University, Changsha, Hunan province, China. The samples were embedded in paraffin and stored in the institutional biobank. PSC, PBC and healthy tissues were evaluated histologically to confirm presence or absence of cancer.

For mRNA assays of FOXM1, MAT2A and MAT2B in healthy human livers (NHL), ten case of healthy control samples were obtained from surgical resection of hepatic hemangioma. None of the healthy control individuals underwent preoperative chemotherapy, and liver histology demonstrated absence of both cirrhosis and malignancy. Liver tissues were immediately frozen in liquid nitrogen. Thirty-one patients with cirrhosis were diagnosed at Xiangya Hospital Central South University from 2016–2021. There was no bias in the selection of patients. The sex of the participants was self-reported, and neither sex nor gender was considered during participant selection. Liver biopsies were evaluated blindly by two expert hepatopathologists. Hepatitis B surface antigen (HBsAg) was measured with a commercial enzyme immunoassay (Shanghai Kehua Bio-engineering Co. Ltd Shanghai). Total RNA of NHL and cirrhosis was extracted from stored frozen liver specimens using QIAzol (Qiagen) reagent according to the manufacturer’s recommendations, and RNA quality and integrity were assessed using the RNA 6000 Nano assay on the Agilent 2100 Bioanalyzer. mRNA levels of FOXM1, MAT2A and MAT2B were measured by TaqMan. The research was conducted under both the Declarations of Helsinki and Istanbul, and the study protocol was approved by the IRB and the Medical Ethical Committee of Xiangya Hospital Central South University. All human materials were obtained with patients’ informed written consent.

Mice

Four-month-old male C57BL/6 littermates were used for this study. Male mice were used because they are more susceptible to liver fibrosis. Mice were housed with ad libitum access to food (diet: CAT#5053, Picolab, San Francisco, CA) and water and kept on a 12 h (8 a.m. to 8 p.m.) light-dark cycle in a temperature (72–77⁰F) and humidity (42%) controlled room at the animal facility of the Cedars-Sinai Medical Center (CSMC). Care of animals was within institutional animal care committee guidelines. Mice were euthanized using ketamine (100 mg/kg, intraperitoneally).

Generation of cell-type-specific Foxm1 knockout mice

Generation of Foxm1 ^Hep−/− mice

1.
Foxm1^fl/fl mice (Jackson Laboratory) were cross bred with Alb-Cre mice (Jackson Laboratory) transgenic mice. Foxm1^fl/flCre-negative littermates served as WT controls. Cell type-specific knockouts were confirmed by isolation of liver cells and messenger RNA (mRNA) expression. Twelve-week-old mice were used for the BDL experiments. All animal experiments were approved by the Institutional Review Board of the Cedars-Sinai Medical Center. Sequences of primers of forward 5’-TGGCTTCCCAGCAGTACAAATC-3’ and reverse 5’-TGCTTACAAAAGACACACTTGGACG’. The PCR reaction consisted of an initial denaturation at 94 °C for 3 min followed by 35 cycles at 95 °C for 30 s, annealing at 60 °C for one minute and extension at 72 °C for one minute using the 2XTaq red master (Apex, El Cajon, CA, Cat #: 42-138). PCR analysis of Foxm1^Hep−/−, Foxm1^Hep+/−and Foxm1^Hep+/+ depicted in Supplementary Fig. 10a.

Generation of Foxm1 ^HSC−/− mice

2.
Foxm1^fl/fl mice were crossed with Lrat-Cre transgenic mice (kindly provided by Dr. Robert Schwabe, Columbia University). Mice were back-crossed at least 10 generations onto the C57BL/6 background. Foxm1^fl/flCre-negative littermates served as WT controls. Sequences of primers of forward 5’-CCTTTCTTTGACCCCCTGCAG-3’ and reverse 5’-GACCGGCAAACGGACAGAAG’. The PCR reaction is same as Foxm1^Hep−/−. PCR analysis of Foxm1^HSC−/−, Foxm1^HSC+/−and Foxm1^HSC+/+ depicted Supplementary Fig. 10b.

Generation of Foxm1 ^KC−/− mice

3.
Foxm1^fl/fl mice were cross bred with Clec4f-Cre (Jackson Laboratory) transgenic mice. Foxm1^fl/flCre-negative littermates served as WT controls. Sequences of primers for mutant reverse 5’-ACACCGGCCTTATTCCAAG-3’, wild type reverse 5’-GAAAGACCCAAGGGAAGGAG-3’ and common 5’-CAAGAAGTCCACAGGGTGGT-3’. The PCR reaction is same as Foxm1^Hep−/−. PCR analysis of Foxm1^KC−/−, Foxm1^KC+/−and Foxm1^KC+/+ depicted in Supplementary Fig. 10c.

Genomic DNA from mouse ears was isolated and amplified by PCR for analysis of Foxm1^Hep−/−, Foxm1^HSC−/−, and Foxm1^KC−/− genotypes.

Carbon tetrachloride treatment and BDL

Mice were housed in a room maintained at a temperature of 22 °C and relative humidity of 50 ± 10% with alternating 12 h of light and darkness. Mice were acclimatized for one week prior to use and had free access to food and water during the entire experiment. In the CCl₄ groups, mice were intraperitoneally (i.p) injected with 0.3% CCl₄ (10 mL/kg, dissolved in corn oil). 64 mice at 8-weeks of age were randomly assigned to two groups: a prevention group and a treatment group. Prevention and treatment groups were divided into subgroups of Oil + DMSO, Oil + FDI-6, Oil + CCl₄ and CCl₄ + FDI-6. CCl₄ was given via i.p. injection three times/week for 3 weeks in prevention group and 5 weeks in the treatment group. FDI-6 was dissolved in 10% DMSO and diluted in 90% corn oil (final concentration: 2.5 mg/mL). FDI-6 treatment was given at 25 mg/kg/d via direct i.p. injection three times/week started at the same time as CCl₄ in the prevention group versus after 3 weeks of CCl₄ in the treatment group and continued with CCl₄ for 2 weeks.

For BDL surgery, 32 mice at 3-months of age from Foxm1^Hep−/−, Foxm1^HSC−/−, Foxm1^KC−/− or flox controls were fed chow ad libitum and housed at constant temperature (22 °C) with alternating 12 h of light and darkness. Following i.p. injection of ketamine (80 mg/kg) and xylazine (10 mg/kg), the common bile duct was exposed through a midline abdominal incision, ligated in two places with a silk thread and sectioned between the ligatures. Sham operated mice had anesthesia and laparotomy only. The serum and partial liver tissues were used for assays of ALT, AST, hydroxyproline and IHC. The hepatic cells were isolated for studies described below and used for western blots and mRNA determinations.

For time-curse of FOXM1, MAT2α and MAT2β during BDL, C57BL/6 mice (9–10 weeks) were anesthetized with isoflurane and subjected to midline laparotomy (~1 cm) to expose the common bile duct (CBD). Two knots were made on the CBD using a 5–0 silk suture and then, the CBD was cut between the two knots. The peritoneum was re-aligned, and the underneath muscle layers and skin were closed individually using a 5–0 silk suture. Mice. Briefly, mice were fed ad libitum with diet for 5 days and then divided into six groups of the day 0, 1, 2, 3, 7 and 14 (n = 6 per group, 36 total). Sham operation was on day 0. Sera and liver tissues were collected for various assays and cell isolation. Eight Flox male mice at three months of age were used for cholangiocyte isolation and promoter assays.

All procedure protocols, use, and the care of the animals were reviewed and approved by the Institutional Animal Care and Use Committee at Cedars-Sinai Medical Center and all experiments involving research animals were conducted in accordance with all relevant ethical regulations. All mice were housed under 12-h light/12-h dark cycle at an average temperature of 74 ⁰F and 40% humidity.

Isolation of different liver cell types

Foxm1^Hep−/−, Foxm1^HSC−/−, Foxm1^KC−/−, and their corresponding Flox are subjected to BDL for seven days. Isolations of primary HSCs from three-months old male knockout and wild-type animals were performed by pronase-collagenase perfusion followed by density gradient centrifugation. Briefly, after in situ perfusion of the liver with 1 mg/ml pronase (CAT#: 1145963001, Roche, Indianapolis, IN), followed by 0.35 mg/mL collagenase (CAT#: C9891-1G, Sigma), dispersed cell suspensions were layered on a discontinuous density gradient Optiprep solution (CAT#: 07820, Serumwerk, Bernburg) and washed with 1XHBSS (CAT#: 21-021-CV, CORNING). HSCs were collected from the gradient interface, and cell viability was verified by phase-contrast microscopy as well as trypan blue staining. The viability of all cell cultures used for the studies was >95%. For EVs isolation, the cells were plated at a density of 4,000,000/10 cm dish. The cells were kept in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS) and antibiotics in a humidified atmosphere with 5% CO₂ and 21% O₂ at 37 °C for 24 h and then changed to serum free DMEM media for 24 h. To isolate Kupffer cells, animals were performed by collagenase perfusion followed by low spin at 50 g for 5 min. The supernatants were centrifugated at 800 g for 10 min. The cell pellet was followed by density gradient centrifugation at a speed of 16,000 for 15 min. The cell fraction at middle layer was incubated with anti-F4/80 microbeads (Milenyl Biotec, Auburn, CA) antibody for 15 min in the dark in the refrigerator (2–8 °C). The cells were washed with 1–2 ml of buffer (a solution containing phosphate-buffered saline (PBS), pH 7.2, 0.5% bovine serum albumin (BSA) and 2 mM EDTA) and spun at 300 g for 10 min and aspirated supernatant completely. For EVs isolation, the cells were plated at a density of 5,000,000/10 cm dish. To isolate hepatocytes, animals were performed by collagenase perfusion followed by low spin at 50 g for 5 min. The cell pellet was washed by Hanks’ balanced salt solution (HBSS) twice. The cells were plated at a density of 3,000,000/10 cm dish. Cell culture condition for KCs and hepatocytes was the same as HSCs. Cell pellets were immediately stored at −80 °C for RNA or protein measurements.

Cell lines

LX-2 cells were kindly provided by Dr. Ekihiro Seki and the mouse macrophage cell line RAW 264.7 was purchased from American Type Cell Collection (ATCC, Manassas, VA). Both cell lines were cultured in DMEM containing 10% fetal bovine serum (FBS) and antibiotics (2 mM glutamine, 50 mM penicillin, and 50 mg/ml streptomycin sulfate).

Measurements of liver injury and fibrosis

Liver samples from CCl₄ with or without FDI-6 treatments with three kinds of Foxm1 cell specific knockout or Flox mice were homogenized in 6 N HCl, hydrolyzed overnight at 100 °C, and centrifuged at 10,000 × g for 3 min. Supernatants were transferred to a 96-well plate and analyzed following the Hydroxyproline assay kit protocol (MAR008-1KT, Sigma). Hydroxyproline content was measured using a CLARIO Star machine (absorbance at 560 nm) and normalized to liver weight. The ALT and AST levels in serum were determined using ALT kit (CAT# TR71121, ThermoFisher) and AST kit (CAT# TR70121, ThermoFisher), respectively. The collagen proportional area (CPA) was defined as the percentage of the area positive for Sirius red staining. Sirius Red area/total area was quantified according to the value of CPA⁴².

Real-time quantitative PCR (qPCR)

Total RNA was extracted using TRIzol reagent (ThermoFisher) or Quick-RNA miniPrep kit (Zymo Research, Irvine, CA) from liver tissues, hepatic cells of Foxm1^Hep−/−, Foxm1^HSC−/−, and Foxm1^KC−/−, or LX-2 (human activated HSCs) and RAW 264.7 (murine macrophages) cells. Human and mouse probes for MAT2A, MAT2B and FOXM1, and the PCR Supermix were purchased from ThermoFisher. Hypoxanthine phosphoribosyltransferase 1 (HPRT1) was included as a housekeeping gene. The thermal profile comprised of initial denaturation at 95 °C for 3 min followed by 40 cycles at 95 °C for 3 s and at 60 °C for 30 s. The cycle threshold (Ct value) of the target gene was normalized to that of the HPRT1 gene to obtain the delta Ct (DCt). Relative genes expression was calculated with 2^−ΔΔCT method and normalized to HPRT1 expression (ΔΔCt = ΔCt of target genes – ΔCt of HPRT1).

Western blot analysis

Western blot analysis was done with antibodies against FOXM1 (ab207298; CAT#: 13147-1-AP, Proteintech), MATα2 (55309-1-AP), MAT2β (BNP1-82797), α-SMA (ab5831), COL1A1 (ab270993), β-actin (ab8226), F4/80 (ab300421), SMAD3 (ab52903), tumor necrosis factor alpha (TNF-α, ab183218), tubulin (ab18251) and interleukin 6 (IL-6, ab259341). The membranes were incubated with secondary antibody (mouse anti-rabbit IgG, Cell Signaling, #7074, USA) for 40 min at room temperature. Membranes were stripped and re-probed with antibodies against β-actin or histone H3 as housekeeping control or CD9 as EVs control. A chemiluminescence system (Millipore Corporation, Billerica, MA, USA) was used for signal detection. All results were quantified by ImageJ software (versions 1.51 and 1.54j). Statistical significance was defined as p < 0.05.

Morphological and immunohistochemical analysis

All human and mouse liver tissues were embedded in paraffin, sectioned, and stained with hematoxylin & eosin (H&E), Sirius red, and F4/80. Five μM paraffin sections of the liver tissue samples were used for IHC staining. Slides were immunostained with antibodies to FOXM1, MATα2, MAT2β, COL1A1, EMR1 (F4/80), α-SMA, Keratin 19 (CK19, ab52625), and IgG according to an IHC detection kit (ab64264). All sections were counterstained with hematoxylin for 30 s and were observed under Digital Inverted Fluorescence Microscope (ThermoFisher, AMF4300). No staining occurred with normal mouse IgG. For every slide, 10 representative fields were captured randomly and analyzed by Image-Pro Plus v6.0 software.

Promoter analysis

The −1329/+60 PCR product representing the MAT2A 5’-flanking region was cloned in the sense orientation upstream of the luciferase coding sequence of the pGL3-basic vector (Promega)⁴³. To produce 5’-deletion constructs, MAT2A promoter was digested with NheI and Afl II to produce a −939 to /+60 deletion, NheI and EcoRV to produce a −571 to /+60 deletion, and NheI and Apa I (−47 to /+60 deletion. −271/ + 60 bp deletion was produced using forward primer 5’-CATCAAACAAGGAAGAGCAATCC-3’ (−250 to −271 bp relative to transcription start site) and reverse primer 5’-AGCTGCGGACAGCGTTCTACTCG-3’ ( + 39 to +60 bp relative to transcription start site). Mutagenesis of the MAT2A promoter FOX binding sites were performed as follows: 5’- CTATAAA-3’ (−16 to −22) was changed to 5’-CTATCAA-3’ using forward primer 5’- GGCGCTGCTCTATCAATACCGGGCC-3’ (−7 to −30 bp relative to transcription start site); 5’- CATCAAACA-3’ (−264 to −272) was changed to 5’- CATCCAACA -3’ using forward primer 5’- TTTCTCCCACATCCAACAAGGAAGAGC -3’, (−255 to −281). A −1319/ + 3 5′-flanking region of the human MAT2B was cloned into the SmaI site of pGL-3 basic vector⁴³. MAT2B promoter was subcloned with forward primers from −990 to –967, −713 to –688, and −250 to –225, and reverse primer was +3 to −20 to generate deletion constructs −990/+3, −713/+3, and −250/+3 MAT2B-LUC, respectively. All sequences are relative to the ATG start codon. Mutagenesis of the MAT2B promoter FOX binding site was performed as follows: 5’- CAAATA-3’ was changed to 5’-CAACTA-3’ (−140 to −145) using forward primer 5’-AATAAAAAGCACTCAACTAAAATCTCCGAAAC-3’ (−127 to −158) with QuikChange multisite-Directed mutagenesis kit (#200515-5) from Agilent Technologies (Carpinteria, CA). The mutant strand synthesis reaction consisted of an initial denaturation at 95 °C for 30 s followed by 16 cycles at 95 °C for 30 s, annealing at 55 °C for one minute and extension at 68 °C for one minute/kb of plasmid length using the PfuUltra HF DNA polymerase. DpnI digestion of the amplified PCR products and transformation of XL10-gold ultracompetent cells were done in accordance with their suggested protocol (Agilent Technologies). The human 1.4 kb FOXM1 promoter was purchased from GeneCopoeia (Rockville, MD). FOXM1 promoter was subcloned with forward primers from −1304 to –1282 and −123 to –102, and reverse primer was +89 to +107 to generate deletion constructs −1304/+107, and −123/+107 FOXM1-LUC, respectively. Mutagenesis of the FOX binding sites in the FOXM1 promoter were performed as follows: 5’-TGTTTG-3’ (−1022 to −1027) was changed to 5’-GGTTTG-3’ using forward primer 5’- GCAATAATTCAACATTGGTTTGTTTTGGAGAC-3’ (−1012 to −1043 bp relative to transcription start site), and 5’-TGTTTA-3’ (-732 to -737) was changed to 5’-TGTTGA-3’ using forward primer 5’-GCCCACATTTGTTGATTTGATTAAAATGTC-3’, (−717 to −751), with QuikChange multisite-Directed mutagenesis kit (Agilent Technologies, Inc. Santa Clara, CA). Plasmids were transfected into cells using Lipofectamine 2000 reagent (Invitrogen, #11668019, MA). Promoter constructs and pGL3/pLuc-MCS/PEZ/SV40 (1 μg) were then transfected into LX-2 and primary cholangiocytes using JetPRIME transfection reagent (Polyplus Transfection, #101000027, USA). Luciferase assays were performed by the Dual Luciferase Reporter Assay System (Promega, #E190, USA) 24 h after transfection.

Isolation of extracellular vesicles (EVs) and characterization

Since FOXM1, MATα2 and MAT2β expression increased dramatically in hepatocytes, HSCs, and KCs after BDL, we isolated EVs from these cells after BDL and used them to treat other liver cell types. We used commercial kits (Thermo Fisher cat# 4484450, 4478360; SBI cat# EQ806A-1) for EVs isolation. Briefly, serum-free media was collected and centrifuged at 2000 g for 30 min to remove the cell debris. Then we further concentrated the 150 ml supernatant media to 1 ml with MilliporeSigma UFC701008 Centricon Plus-70 Centrifugal Filter (100 kd)⁴⁴. The concentrated medium was then processed with the SBI kit according to the provided protocol. The concentration and size distribution of EVs were assessed using Nanoparticle Tracking Analysis (NTA) via NanoSight NS300. Validation of EVs was performed through western blot analysis employing EVs markers⁴⁵.

We used 10 μg of EVs to treat hepatocytes, HSCs, and KCs. HSCs from Foxm1^HSC−/− were treated with EVs from hepatocytes of Foxm1^Hep−/− and Flox. Hepatocytes from Foxm1^Hep−/− were treated with EVs from HSCs of Foxm1^HSC−/− and Flox. Hepatocytes from Foxm1^Hep−/− and HSCs from Foxm1^HSC−/− were treated with EVs from Flox and Foxm1^KC−/−. Expression of FOXM1, MATα2, and MAT2β were measured by western blot in HSCs, hepatocytes, and KCs.

ChIP and sequential-ChIP (Seq-ChIP)

LX-2 cells were crosslinked with formaldehyde to bind the proteins to the DNA. The fragmentation of DNA was made by cell sonication (shearing of DNA into small fragments ranging from 50 to 500 bp). Immunoprecipitation (IP) using FOXM1, MATα2, and MAT2β and lgG antibodies enriches the fragments bound by the protein. Then, crosslink reversal frees the DNA fragments for further processing. ChIP and Seq-ChIP were performed using a EpiTect ChIP OneDay kit (Qiagen, Germantown, MD) in accordance with the kit instructions. For the FOX element of the human FOXM1 promoter, DNA immunoprecipitation (IP) was performed using antibodies against MATα2, MAT2β and FOXM1. lgG was used for control. This was followed by a second round of IP using antibodies against MATα2, MAT2β and FOXM1. The purified DNA was detected by PCR analysis. PCR primers for promoter regions containing FOX binding sites were FOXM1 - forward 5’- CATTTGTTTGTTTTGGAGACGGTGTC-3’ (−1006 to −1031) and reverse 5’-GGAAGAGGGGCACAGACATTTTAATC-3’ (−703 to −728) (GenBank® accession no. NM-001243088). All PCR products were electrophoresed on 2% agarose gel. The PCR protocol involved initial denaturation at 94 °C for 3 min and 25 cycles at 94 °C for 30 s, followed by annealing and extension at 65 °C for 90 s using the Advantage GC 2 PCR kit (Clontech, Mountain View, CA).

For ChIP-qPCR assays, chromatin was prepared from LX-2 cells treated with scrambled control, siRNAs of MAT2A, MAT2B and FOXM1, overexpression of MAT2A, MAT2B and FOXM1, or empty vector. Chromatin was enzymatically sheared and immunoprecipitated with anti-MATα2, anti-MAT2β and anti-FOXM1 using Pierce™ Magnetic ChIP Kit (Cat#: 26157, ThermoFisher Scientific) in accordance with the kit’s instructions. Rabbit IgG was used as a mock antibody for negative control. Immunoprecipitated chromatin was subjected to real-time qPCR with the SYBR Premix DimerEraser (RR091A, Takara). Two ChIP-qPCR primer pairs (forward primer: 5’- CATTTGTTTGTTTTGGAGACGGTGTC-3’ (−1006 to −1031) and reverse primer: 5’- ATAAACAAATGTGGGCTGGGCATAG-3’ (−730 to −755)) overlap two FOX binding sites of the human FOXM1 promoter region. Samples were run in triplicates, and data from MATα2, MAT2β and FOXM1 IP, and control IP were presented as enrichment relative to input DNA. ChIP-qPCR was repeated triplicates to confirm the reproducibility of results. The quality of chromatin enzymatically sheared was assessed using agarose gel electrophoresis. The percent input is used to normalize ChIP-qPCR data. 1% of starting chromatin is used as input. ChIP-qPCR data is normalized for both background levels and input chromatin going into the ChIP. Pierce™ Magnetic ChIP Kit (Cat#: 26157).

Electrophoretic mobility shift assay (EMSA)

In 10 cm dishes, LX-2 cells were transfected with scrambled control, siRNA (si) against FOXM1, MAT2A, or MAT2B, empty vector (EV), FOXM1 OV, MAT2A OV, or MAT2B OV for 48 h. LX-2 cells were grown to 70–80% and after removing the medium, cells were washed twice with ice-cold PBS and scraped into a 15 mL conical tube and centrifuged at 1000 rpm for 5 min. The nuclear proteins from LX-2 cells were isolated using the Nuclear-Extraction kit (Abcam, ab113473). Double FOX binding sites (5’-ACATTTGTTTGTTTTGACATTTGTTTATTTGA-3’) were used for gel shift and supper shift probe. EMSA assays were done by following the EMSA Kit (ThermoFisher Scientific, E33075). The shift bands were detected from the gel images with three biological replicates. The shifted bands were compared with that of the non-shifted band in each lane^46,47.

MTT and migration assays

For MTT assays, LX-2 Cells were cultured at 1 × 10⁶ cells per ml in a transparent plate and rinsed twice with PBS. 50 µL serum-free medium and 50 µL MTT reagent was added according to manuscript (ab211091, USA). After 3 h of incubation at 37 °C, 150 µL of MTT solvent was added to each well. Plate was wrapped in foil and shaken for 15 min on an orbital shaker. Data was read at OD = 590 nm. For migration assays, LX-2 cells (1 × 10⁵) were seeded into transwell inserts (8-μm pore size, BD Falcon). After 24 h of culture for the migration assays, the wells were stained with crystal violet (Beyotime, C0121). Migrated cells number was quantified using the ImageJ software.

Immunofluorescence staining

Primary HSCs and LX-2 cells were plated on 6-well plates containing coverslips. Cells were fixed with 4% paraformaldehyde for 30 min at room temperature and ice-cold methanol for 15 min. Cells were then blocked by incubation in PBS with 0.05% TWEEN® 20 (PBST, Sigma, P3563) containing 10% goat serum and 3% BSA for 45 min. Samples were incubated with FOXM1, MATα2, and MAT2β antibodies at 250-time dilution (1:250) in PBS for 1 h at 37 °C and incubated with fluorescent secondary antibodies (Invitrogen, Waltham, MA) for 3 h. Negative control was used with IgG antibody (Abcam, ab171870; Rabbit). VECTASHIELD mounting medium (Sigma, #DUO82040) containing DAPI was used to mount slides. EVOS™ FL and EVOS™ FL color Imaging systems (ThermoFisher Scientific, AMF4300) were used to capture images. Photoshop (version 23.5.2) overlay method was used to contrast and overlay images.

FOXM1-MATα2-MAT2β interactions

40 μL of A/G beads (Cat#: SC-2003, Santa Cruz Biotechnology), 500 μL IP buffer (150 mM NaCl, 50 mM Tris-HCl PH7.6, 1 mM EDTA PH7.0, 1 mM EGTA7.4, 25 mM NaF, 1% NP-40), and 2 μg of MATα2, MAT2β or FOXM1 antibody were mixed and rotated at 4 °C for one hour. The beads were washed 6 times with 500 μL IP buffer. For recombinant proteins (FOXM1, MATα2, and MAT2β) interactions, after the final wash with 500 μL of IP buffer above, 1 μg of recombinant protein was added. For the extracted proteins, 500 μL of IP buffer and 500 μg lysates were needed. 500 μL of IP buffer and 0.5 μL of rabbit IgG were mixed as control. All the reaction tubes were rotated at 4 °C overnight. Then the beads were washed 6 times with 500 μL IP buffer. After the last wash buffer was removed and beads were boiled in 2× SDS sample buffer for 10 min at 95 °C, and proteins were separated on 10% SDS-PAGE followed by Immunoblot analysis for MATα2, MAT2β and FOXM1.

Enzyme-linked immunosorbent assay (ELISA)

Raw 264.7 cells were cultivated in high-glucose DMEM containing 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37 °C in a humidified incubator containing 5% CO₂. After overnight culture in a 96-well plate (2 × 10⁴ cells/well, 200 μL medium/well), cells were transfected with siRNAs of scrambled, FOXM1, MAT2A and MAT2B for 8 h and treated with LPS (Escherichia coli, Sigma-Aldrich L4130 0111: B4) at 1 μg/ml for an additional 16 h. The culture supernatant from each well was collected at the end of scheduled experiments and used to measure TNF-α (KE10002) and IL-6 (KE10091, proteintech, Rosemont, IL, USA) concentration by ELISA according to the manufacturer’s instructions, and the absorbance was measured at 450 nm and 630 nm, respectively, using a microplate reader.

Statistical analysis

All sample data were presented as mean ± standard error of the mean (SEM). Sample data statistical analysis was used Student t-test and one-way ANOVA, followed by Fisher’s test for multiple comparisons. Chi-square test was used to analyze mRNA levels of FOXM1, MAT2A or MAT2B and clinicopathological features in liver cirrhosis. Pearson correlation was used to measure the strength of the linear relationship between FOXM1 and MAT2A, FOXM1 and MAT2B, MAT2A and MAT2B in human liver cirrhosis tissues. All experiments were repeated at least three times, and p < 0.05 was considered statistical significance. Excel (16.54) and Graphpad Prism 9.0.0 were used for calculations.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data supporting the findings described in this manuscript are available in the article and in the Supplementary Information and from the corresponding author upon request. Source data are provided with this paper.

References

Friedman, S. L. Liver fibrosis – from bench to bedside. J. Hepatol. 38, S38–S53 (2003).
Article PubMed Google Scholar
Li, Y. et al. Reciprocal regulation between forkhead box M1/NF-κB and methionine adenosyltransferase 1A drives liver cancer. Hepatology 72, 1682–1700 (2020).
Article CAS PubMed Google Scholar
Kurahashi, T. et al. Forkhead box M1 transcription factor drives liver inflammation linking to hepatocarcinogenesis in mice. Cell. Mol. Gastroenterol. Hepatol. 9, 425–446 (2020).
Article PubMed Google Scholar
Sun, W. et al. ALKBH5 promotes lung fibroblast activation and silica-induced pulmonary fibrosis through miR-320a-3p and FOXM1. Cell. Mol. Biol. Lett. 27, 26 (2022).
Article CAS PubMed PubMed Central Google Scholar
Filliol, A. & Schwabe, R. F. FoxM1 induces CCl2 secretion from hepatocytes triggering hepatic inflammation, injury, fibrosis, and liver cancer. Cell. Mol. Gastroenterol. Hepatol. 9, 555–556 (2020).
Article PubMed PubMed Central Google Scholar
Wang, Y., Zhou, Q., Tang, R., Huang, Y. & He, T. FoxM1 inhibition ameliorates renal interstitial fibrosis by decreasing extracellular matrix and epithelial-mesenchymal transition. J. Pharmacol. Sci. 143, 281–289 (2020).
Article CAS PubMed Google Scholar
Gormally, M. V. et al. Suppression of the FOXM1 transcriptional programme via novel small molecule inhibition. Nat. Commun. 5, 5165 (2014).
Article ADS CAS PubMed Google Scholar
Lan, C. et al. Forkhead domain inhibitor-6 suppresses corneal neovascularization and subsequent fibrosis after alkali burn in rats. Invest. Ophthalmol. Vis. Sci. 63, 14 (2022).
PubMed PubMed Central Google Scholar
Lan, C. et al. Forkhead domain inhibitory-6 attenuates subconjunctival fibrosis in rabbit model with trabeculectomy. Exp. Eye. Res. 210, 108725 (2021).
Article CAS PubMed Google Scholar
Murray, B., Barbier-Torres, L., Fan, W., Mato, J. M. & Lu, S. C. Methionine adenosyltransferases in liver cancer. World J. Gastroenterol. 25, 4300–4319 (2019).
Article CAS PubMed PubMed Central Google Scholar
Mato, J. M. & Lu, S. C. Role of S-adenosyl-L-methionine in liver health and injury. Hepatology 45, 1306–1312 (2007).
Article CAS PubMed Google Scholar
Ramani, K., Donoyan, S., Tomasi, M. L. & Park, S. Role of methionine adenosyltransferase α2 and β phosphorylation and stabilization in human hepatic stellate cell trans-differentiation. J. Cell. Physiol. 230, 1075–1085 (2015).
Article CAS PubMed PubMed Central Google Scholar
Hu, X. & Zhou, Y. Curcumin reduces methionine adenosyltransferase 2B expression by interrupting phosphorylation of p38 MAPK in hepatic stellate cells. Eur. J. Pharmacol. 886, 173424 (2020).
Article CAS PubMed Google Scholar
Wang, K. et al. TGF-β1/p65/MAT2A pathway regulates liver fibrogenesis via intracellular SAM. EBioMedicine 42, 458–469 (2019).
Article PubMed PubMed Central Google Scholar
Luedde, T., Kaplowitz, N. & Schwabe, R. F. Cell death and cell death responses in liver disease: mechanisms and clinical relevance. Gastroenterology 147, 765–783 (2014).
Article CAS PubMed Google Scholar
Koyama, Y. & Brenner, D. A. Liver inflammation and fibrosis. J. Clin. Invest 127, 55–64 (2017).
Article PubMed PubMed Central Google Scholar
Dewidar, B., Meyer, C., Dooley, S. & Meindl-Beinker, A. N. TGF-β in hepatic stellate cell activation and liver fibrogenesis-updated 2019. Cells 8, 1419 (2019).
Article CAS PubMed PubMed Central Google Scholar
Zehender, A. et al. TGFβ promotes fibrosis by MYST1-dependent epigenetic regulation of autophagy. Nat. Commun. 12, 4404 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Binatti, E., Gerussi, A., Barisani, D. & Invernizzi, P. The role of macrophages in liver fibrosis: new therapeutic opportunities. Int. J. Mol. Sci. 23, 6649 (2022).
Article CAS PubMed PubMed Central Google Scholar
Yamaguchi, M., Hashimoto, K., Jijiwa, M. & Murata, T. The inflammatory macrophages repress the growth of bone metastatic human prostate cancer cells via TNF-α and IL-6 signaling: involvement of cell signaling regulator regucalcin. Cell. Signal. 107, 110663 (2023).
Article CAS PubMed Google Scholar
Rho, H., Terry, A. R., Chronis, C. & Hay, N. Hexokinase 2-mediated gene expression via histone lactylation is required for hepatic stellate cell activation and liver fibrosis. Cell. Metab. 35, 1406–1423 (2023).
Article CAS PubMed Google Scholar
Zhang, Z., Li, M., Sun, T., Zhang, Z. & Liu, C. FOXM1: functional roles of FOXM1 in non-malignant diseases. Biomolecules 13, 857 (2023).
Article CAS PubMed PubMed Central Google Scholar
Hu, G. et al. FOXM1 promotes hepatocellular carcinoma progression by regulating KIF4A expression. J. Exp. Clin. Cancer Res. 38, 188 (2019).
Article PubMed PubMed Central Google Scholar
Khan, M. A., Khan, P., Ahmad, A., Fatima, M. & Nasser, M. W. FOXM1: a small fox that makes more tracks for cancer progression and metastasis. Semin. Cancer Biol. 92, 1–15 (2023).
Article CAS PubMed PubMed Central Google Scholar
Liang, S. K. et al. FOXM1 is required for small cell lung cancer tumorigenesis and associated with poor clinical prognosis. Oncogene 40, 4847–4858 (2021).
Article CAS PubMed Google Scholar
Wang, K. et al. Peptide-based PROTAC degrader of FOXM1 suppresses cancer and decreases GLUT1 and PD-L1 expression. J. Exp. Clin. Cancer Res. 41, 289 (2022).
Article CAS PubMed PubMed Central Google Scholar
Tomasi, M. L. et al. Methionine adenosyltransferase α2 sumoylation positively regulate Bcl-2 expression in human colon and liver cancer cells. Oncotarget 6, 37706–37723 (2015).
Article PubMed PubMed Central Google Scholar
Liu, T. et al. Mechanisms of MAFG dysregulation in cholestatic liver injury and development of liver cancer. Gastroenterology 155, 557–571 (2018).
Article CAS PubMed Google Scholar
Tomasi, M. L. et al. S-adenosylmethionine and methylthioadenosine inhibit cancer metastasis by targeting microRNA 34a/b-methionine adenosyltransferase 2A/2B axis. Oncotarget 8, 78851–78869 (2017).
Article PubMed PubMed Central Google Scholar
Gartel, A. L. FOXM1 in cancer: interactions and vulnerabilities. Cancer Res. 77, 3135–3139 (2017).
Article CAS PubMed PubMed Central Google Scholar
Liu, Q. et al. The X protein of hepatitis B virus inhibits apoptosis in hepatoma cells through enhancing the methionine adenosyltransferase 2A gene expression and reducing S-adenosylmethionine production. J. Biol. Chem. 286, 17168–17180 (2011).
Article CAS PubMed PubMed Central Google Scholar
Wree, A., Mehal, W. Z. & Feldstein, A. E. Targeting cell death and sterile inflammation loop for the treatment of nonalcoholic steatohepatitis. Semin. Liver. Dis. 36, 27–36 (2016).
Article CAS PubMed PubMed Central Google Scholar
Pradere, J. P. et al. Hepatic macrophages but not dendritic cells contribute to liver fibrosis by promoting the survival of activated hepatic stellate cells in mice. Hepatology 58, 1461–1473 (2013).
Article CAS PubMed Google Scholar
Kalluri, R. & LeBleu, V. S. The biology, function, and biomedical applications of exosomes. Science 367, eaau6977 (2020).
Article CAS PubMed PubMed Central Google Scholar
Luo, Z. et al. CRIg(+) Macrophages prevent gut microbial DNA-containing extracellular vesicle-induced tissue inflammation and insulin resistance. Gastroenterology 160, 863–874 (2021).
Article CAS PubMed Google Scholar
Isaac, R., Reis, F. C. G., Ying, W. & Olefsky, J. M. Exosomes as mediators of intercellular crosstalk in metabolism. Cell. Metab. 33, 1744–1762 (2021).
Article CAS PubMed PubMed Central Google Scholar
Mathieu, M., Martin-Jaular, L., Lavieu, G. & Théry, C. Specificities of secretion and uptake of exosomes and other extracellular vesicles for cell-to-cell communication. Nat. Cell. Biol. 21, 9–17 (2019).
Article CAS PubMed Google Scholar
Cho, Y. E., Song, B. J., Akbar, M. & Baek, M. C. Extracellular vesicles as potential biomarkers for alcohol- and drug-induced liver injury and their therapeutic applications. Pharmacol. Ther. 187, 180–194 (2018).
Article CAS PubMed PubMed Central Google Scholar
Hernández, A. et al. Chemical hypoxia induces pro-inflammatory signals in fat-laden hepatocytes and contributes to cellular crosstalk with Kupffer cells through extracellular vesicles. Biochim. Biophys. Acta Mol. Basis. Dis. 1866, 165753 (2020).
Article PubMed Google Scholar
Tabibian, J. H., Masyuk, A. I., Masyuk, T. V., O’Hara, S. P. & LaRusso, N. F. Physiology of cholangiocytes. Compr. Physiol. 3, 541–565 (2013).
Article PubMed Google Scholar
Perea, L. et al. Pentraxin-3 modulates lipopolysaccharide-induced inflammatory response and attenuates liver injury. Hepatology 66, 953–968 (2017).
Article CAS PubMed Google Scholar
Tsochatzis, E. et al. Collagen proportionate area is superior to other histological methods for sub-classifying cirrhosis and determining prognosis. J. Hepatol. 60, 948–954 (2014).
Article CAS PubMed Google Scholar
Yang, H. et al. Expression pattern, regulation, and functions of methionine adenosyltransferase 2beta splicing variants in hepatoma cells. Gastroenterology 134, 281–291 (2008).
Article CAS PubMed Google Scholar
Chhoy, P., Brown, C. W., Amante, J. J. & Mercurio, A. M. Protocol for the separation of extracellular vesicles by ultracentrifugation from in vitro cell culture models. Star. Protoc. 2, 100303 (2021).
Article CAS PubMed PubMed Central Google Scholar
Liu, J. et al. Extracellular vesicles-encapsulated let-7i shed from bone mesenchymal stem cells suppress lung cancer via KDM3A/DCLK1/FXYD3 axis. J. Cell. Mol. Med. 25, 1911–1926 (2021).
Article CAS PubMed Google Scholar
Yang, B. et al. CSNK2A1-mediated MAX phosphorylation upregulates HMGB1 and IL-6 expression in cholangiocarcinoma progression. Hepatol. Commun. 7, e00144 (2023).
Article PubMed PubMed Central Google Scholar
Lu, L. et al. Deregulated 14-3-3ζ and methionine adenosyltransferase α1 interplay promotes liver cancer tumorigenesis in mice and humans. Oncogene 40, 5866–5879 (2021).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

This work was financially supported by NIH grant P01CA233452 (S.C. Lu, H. Yang, E. Seki), Plan Nacional of I + D SAF2017-88041-R (J.M. Mato), National Natural Science Foundation of China 82070632 (Liu T.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author information

These authors contributed equally: Bing Yang, Liqing Lu, Ting Xiong.

Authors and Affiliations

Karsh Division of Gastroenterology and Hepatology, Department of Medicine, Cedars-Sinai Medical Center, LA, Los Angeles, CA, 90048, USA
Bing Yang, Liqing Lu, Ting Xiong, Wei Fan, Jiaohong Wang, Lucía Barbier-Torres, Jyoti Chhimwal, Sonal Sinha, Takashi Tsuchiya, Nirmala Mavila, Maria Lauda Tomasi, Jing Zhang, Hui Peng, Komal Ramani, Jenny Han, Ekihiro Seki, Heping Yang & Shelly C. Lu
Department of Geriatric Endocrinology and Metabolism, Guangxi Key Laboratory of Precision Medicine in Cardio-cerebrovascular Diseases Control and Prevention, Guangxi Clinical Research Center for Cardio-cerebrovascular Diseases, The First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi, 530021, China
Bing Yang & Xi Yang
Key Laboratory of Cancer Proteomics of Chinese Ministry of Health, Xiangya Hospital, Central South University, Changsha, Hunan, 410008, China
Liqing Lu
Department of Pharmacy, The Third Hospital of Changsha, Changsha, Hunan, 410015, China
Ting Xiong
Department of Biomedical Sciences, CSMC LA, Los Angeles, CA, 90048, USA
DuoYao Cao
Department of Oncology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
Jing Zhang
CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology, Park of Bizkaia, 48120, Derio, Bizkaia, Spain
José M. Mato
Department of Gastroenterology, Xiangya Hospital, Key Laboratory of Cancer proteomics of Chinese Ministry of Health, Xiangya Hospital, Central South University, Changsha, Hunan, 410008, China
Ting Liu
Phoenix Children’s Research Institute, Department of Child Health, University of Arizona College of Medicine, Phoenix, AZ, 85004, USA
Vladimir V. Kalinichenko
Division of Neonatology, Phoenix Children’s Hospital, Phoenix, AZ, 85016, USA
Vladimir V. Kalinichenko
Department of Society and Genetics, UCLA LA, Los Angeles, CA, 92620, USA
Jenny Han

Authors

Bing Yang
View author publications
Search author on:PubMed Google Scholar
Liqing Lu
View author publications
Search author on:PubMed Google Scholar
Ting Xiong
View author publications
Search author on:PubMed Google Scholar
Wei Fan
View author publications
Search author on:PubMed Google Scholar
Jiaohong Wang
View author publications
Search author on:PubMed Google Scholar
Lucía Barbier-Torres
View author publications
Search author on:PubMed Google Scholar
Jyoti Chhimwal
View author publications
Search author on:PubMed Google Scholar
Sonal Sinha
View author publications
Search author on:PubMed Google Scholar
Takashi Tsuchiya
View author publications
Search author on:PubMed Google Scholar
Nirmala Mavila
View author publications
Search author on:PubMed Google Scholar
Maria Lauda Tomasi
View author publications
Search author on:PubMed Google Scholar
DuoYao Cao
View author publications
Search author on:PubMed Google Scholar
Jing Zhang
View author publications
Search author on:PubMed Google Scholar
Hui Peng
View author publications
Search author on:PubMed Google Scholar
José M. Mato
View author publications
Search author on:PubMed Google Scholar
Ting Liu
View author publications
Search author on:PubMed Google Scholar
Xi Yang
View author publications
Search author on:PubMed Google Scholar
Vladimir V. Kalinichenko
View author publications
Search author on:PubMed Google Scholar
Komal Ramani
View author publications
Search author on:PubMed Google Scholar
Jenny Han
View author publications
Search author on:PubMed Google Scholar
Ekihiro Seki
View author publications
Search author on:PubMed Google Scholar
Heping Yang
View author publications
Search author on:PubMed Google Scholar
Shelly C. Lu
View author publications
Search author on:PubMed Google Scholar

Contributions

B.Y., L.L., T.X., H.Y., J.H., W.F., T.T., D.C., J.Z. and H. P. performed the experiments and analyzed the data. B.Y., L.L., T.X. and H.Y. reviewed the literature, drafted the manuscript. L.L. and T.L. collected the clinic sample and data. J. W. performed cell culture, analysis, and interpretation. N. M. and M.L.T. bred the mice. L.B.-T., J.M.M., X.Y., J.C., S.S. and K.R. provided critical reading and editing of the manuscript. V.K. provided the Foxm1 flox mice. E.S. provided the L-rat-Cre mice and critical reading of the manuscript. H.Y. supervised the experiments. S.C.L. obtained funding, conceptualized the experiments, provided overall direction, and edited the manuscript. All authors have read and approved the final manuscript.

Corresponding authors

Correspondence to Heping Yang or Shelly C. Lu.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Chang Liu, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Source data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Yang, B., Lu, L., Xiong, T. et al. The role of forkhead box M1-methionine adenosyltransferase 2 A/2B axis in liver inflammation and fibrosis. Nat Commun 15, 8388 (2024). https://doi.org/10.1038/s41467-024-52527-8

Download citation

Received: 22 September 2023
Accepted: 09 September 2024
Published: 27 September 2024
DOI: https://doi.org/10.1038/s41467-024-52527-8

Subjects

Abstract

Similar content being viewed by others

Focal adhesion kinase promotes aerobic glycolysis in hepatic stellate cells via the cyclin D1/c-Myc/MCT-1 pathway to induce liver fibrosis

Inhibition of intrahepatic monocyte recruitment by Cenicriviroc and extracellular matrix degradation by MMP1 synergistically attenuate liver inflammation and fibrogenesis in vivo

MyD88 in hepatic stellate cells enhances liver fibrosis via promoting macrophage M1 polarization

Introduction

Results

FOXM1/MAT2A/MAT2B axis is induced in liver fibrosis

FDI-6 in the development and progression of liver fibrosis

Reciprocal positive regulation between FOXM1/MAT2A/MAT2B

Foxm1 Hep−/− mice have lower BDL-induced injury and fibrosis

Foxm1 HSC−/− mice have lower BDL-induced injury and fibrosis

Foxm1 KC−/− mice have lower BDL-induced injury and fibrosis

FOXM1-MAT2A-MAT2B interplay drives fibrosis and inflammation

Liver cells secrete EVs containing FOXM1/MATα2/MAT2β

Discussion

Methods

Materials and reagents

Human samples

Mice

Generation of cell-type-specific Foxm1 knockout mice

Generation of Foxm1 Hep−/− mice

Generation of Foxm1 HSC−/− mice

Generation of Foxm1 KC−/− mice

Carbon tetrachloride treatment and BDL

Isolation of different liver cell types

Cell lines

Measurements of liver injury and fibrosis

Real-time quantitative PCR (qPCR)

Western blot analysis

Morphological and immunohistochemical analysis

Promoter analysis

Isolation of extracellular vesicles (EVs) and characterization

ChIP and sequential-ChIP (Seq-ChIP)

Electrophoretic mobility shift assay (EMSA)

MTT and migration assays

Immunofluorescence staining

FOXM1-MATα2-MAT2β interactions

Enzyme-linked immunosorbent assay (ELISA)

Statistical analysis

Reporting summary

Data availability

References

Acknowledgements

Author information

Authors and Affiliations

Contributions

Corresponding authors

Ethics declarations

Competing interests

Peer review

Peer review information

Additional information

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Source data

Source data

Rights and permissions

About this article

Cite this article

Share this article

Search

Quick links

Foxm1 ^Hep−/− mice have lower BDL-induced injury and fibrosis

Foxm1 ^HSC−/− mice have lower BDL-induced injury and fibrosis

Foxm1 ^KC−/− mice have lower BDL-induced injury and fibrosis

Generation of Foxm1 ^Hep−/− mice

Generation of Foxm1 ^HSC−/− mice

Generation of Foxm1 ^KC−/− mice