Human antigen R contributes to hepatic stellate cell activation and liver fibrosis

Authors

  • Ashwin Woodhoo,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
    • Joint first authors.

  • Marta Iruarrizaga-Lejarreta,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
    • Joint first authors.

  • Naiara Beraza,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
  • Juan L García-Rodríguez,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
  • Nieves Embade,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
  • David Fernández-Ramos,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
  • Nuria Martínez-López,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
  • Virginia Gutiérrez-De Juan,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
  • Beatriz Arteta,

    1. School of Medicine and Dentistry, University of the Basque Country, Universidad del País Vasco/Euskal Herriko Unibertsitatea, Leioa, Spain
    Search for more papers by this author
  • Juan Caballeria,

    1. Liver Unit, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi i Sunyer, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Barcelona, Spain
    Search for more papers by this author
  • Shelly C Lu,

    1. Division of Gastrointestinal and Liver Diseases, University of Southern California Research Center for Liver Diseases, Southern California Research Center for Alcoholic and Pancreatic Diseases & Cirrhosis, Keck School of Medicine, University of Southern California, Los Angeles, CA
    Search for more papers by this author
  • José M Mato,

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
  • Marta Varela-Rey,

    Corresponding author
    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    • CIC bioGUNE, Technology Park of Bizkaia, 48160 Derio, Bizkaia, Spain
    Search for more papers by this author
    • fax: +34-944-061301

    • Joint senior authors.

  • María L Martínez-Chantar

    1. CIC bioGUNE, Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd), Technology Park of Bizkaia, Derio, Bizkaia, Spain
    Search for more papers by this author
    • Joint senior authors.


  • Potential conflict of interest: Nothing to report.

  • This work was supported by the National Institutes of Health (grant nos.: RO1AT-1576 and RO1AT-004896; to S.C.L., M.L.M-C., and J.M.M.), SAF 2011-29851 (to J.M.M.), ETORTEK-2011, Sanidad Gobierno Vasco 2008, Educación Gobierno Vasco 2011 (PI2011/29), and FIS (PI11/01588) (to M.L.M-C.), Sanidad Gobierno Vasco 2012 (to M.V.R.), FIS PS09/00094, Fundación Científica de la Asociación Española Contra el Cáncer (Cancer Infantil) (to A.W.), and the Program Ramón y Cajal, Ministry of Science and Innovation, Spain (to A.W. and N.B.). FIS PS09/02010 were awarded to N.B., and FIS PS09/01164 was awarded to J.C. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (Ciberehd) is funded by the Instituto de Salud Carlos III.

Abstract

RNA-binding proteins (RBPs) play a major role in the control of messenger RNA (mRNA) turnover and translation rates. We examined the role of the RBP, human antigen R (HuR), during cholestatic liver injury and hepatic stellate cell (HSC) activation. HuR silencing attenuated fibrosis development in vivo after BDL, reducing liver damage, oxidative stress, inflammation, and collagen and alpha smooth muscle actin (α-SMA) expression. HuR expression increased in activated HSCs from bile duct ligation mice and during HSC activation in vitro, and HuR silencing markedly reduced HSC activation. HuR regulated platelet-derived growth factor (PDGF)-induced proliferation and migration and controlled the expression of several mRNAs involved in these processes (e.g., Actin, matrix metalloproteinase 9, and cyclin D1 and B1). These functions of HuR were linked to its abundance and cytoplasmic localization, controlled by PDGF, by extracellular signal-regulated kinases (ERK) and phosphatidylinositol 3-kinase activation as well as ERK/LKB1 (liver kinase B1) activation, respectively. More important, we identified the tumor suppressor, LKB1, as a novel downstream target of PDGF-induced ERK activation in HSCs. HuR also controlled transforming growth factor beta (TGF-β)-induced profibrogenic actions by regulating the expression of TGF-β, α-SMA, and p21. This was likely the result of an increased cytoplasmic localization of HuR, controlled by TGF-β-induced p38 mitogen-activated protein kinase activation. Finally, we found that HuR and LKB1 (Ser428) levels were highly expressed in activated HSCs in human cirrhotic samples. Conclusion: Our results show that HuR is important for the pathogenesis of liver fibrosis development in the cholestatic injury model, for HSC activation, and for the response of activated HSC to PDGF and TGF-β. (HEPATOLOGY 2012;56:1870–1882)

Hepatic fibrosis is the common consequence of chronic liver diseases (CLDs), such as viral and autoimmune hepatitis, alcohol consumption, biliary obstruction, and nonalcoholic fatty liver disease.1 Hepatic stellate cells (HSCs) are the major producers of collagen in the damaged liver.2 In healthy liver, HSCs have a quiescent phenotype, accumulating retinoids (i.e., vitamin A) and expressing markers characteristic of adipocytes.3 After continued liver damage, these quiescent HSCs are exposed to apoptotic hepatocytes, reactive oxygen species, as well as inflammatory and profibrogenic factors, and undergo a process of activation to a myofibroblastic phenotype. These activated HSCs increase proliferation and migration, acquire contractility and proinflammatory properties, and express myogenic markers, such as alpha smooth muscle actin (α-SMA) to become the major collagen type 1 alpha 1 (col1a1)-producing cells.4

In the liver, levels of many messenger RNAs (mRNAs) are regulated in response to fibrosis-inducing injuries.5 RNA-binding proteins (RBPs) can promote rapid spatiotemporal expression of proteins by binding to U- and AU-rich elements (AREs) in mRNAs.6 Human antigen R (HuR), a member of the Hu/Elav family, is a ubiquitously expressed RBP predominantly (>90%) localized in the nucleus of most unstimulated cells. In response to proliferative, stress, apoptotic, differentiation, senescence, inflammatory, and immune stimuli, HuR is exported to the cytoplasm, increasing the half-life and/or the rate of translation of target mRNAs.6 Several studies have shown that HuR has important functions in hepatocytes, including hepatocyte growth factor–induced hepatocyte proliferation,7 differentiation,8 and apoptosis9 as well as during hepatocyte malignant transformation.8, 10 Also, HuR expression is up-regulated in hepatocellular carcinoma (HCC) tissue, compared to healthy tissues,10 suggesting that it could represent a novel target for liver damage research.

The aims of the current work were to study the role of HuR in liver fibrosis and in HSC activation, and examine its role in controlling the functions of two principal mediators of HSC activation, platelet-derived growth factor (PDGF), and transforming growth factor beta (TGF-β).

Abbreviations

AKT, protein kinase B; α-SMA, alpha smooth muscle actin; AMPK, adenosine-monophosphate–activated protein kinase; AREs, AU-rich elements; BDL, bile duct ligation; BrdU, bromodeoxyuridine; CLDs, chronic liver diseases; CFSC-8B, cirrhotic liver fat-storing cells-8B; col1a1, collagen type I alpha 1; ERK, extracellular signal-regulated kinase; GFAP, glial fibrillary acidic protein; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; H&E, hematoxylin and eosin; HSCs, hepatic stellate cells; HuR, human antigen R; IHC, immunohistochemistry; IL, interleukin; iNOS, inducible nitric oxide synthase; LKB1, liver kinase B1; MAPK, mitogen-activated protein kinase; MCP-1, monocyte chemoattractant protein-1; MEK, MAPK/ERK kinase; MMP9, matrix metalloproteinase 9; mRNA, messenger RNA; NFκB, nuclear factor kappa-light-chain enhancer of activated B cells; PI3K, phosphatidylinositol 3-kinase; PDGF, platelet-derived growth factor; pLKB1, phosphorylated LKB1; qPCR, quantitative real-time polymerase chain reaction; RIP-qPCR, RNA immunoprecipitation of ribonucleotide complexes coupled to qPCR; RBP, RNA-binding protein; shRNA, short hairpin RNA; TGF-β, transforming growth factor beta; TNF-α, tumor necrosis factor alpha.

Materials and Methods

Reagents.

TGF-β and PDGF were purchased from PeproTech Inc. (Rocky Hill, NJ). SB203580 and BAY 11-7082 were purchased from from Calbiochem (San Diego, CA), U0126 was from Promega (Madison, WI), and LY-294002 was from Sigma-Aldrich (St. Louis, MO).

Human Samples.

Surgically resected liver tumor specimens from 16 patients with cirrhosis (hepatitis C virus [HCV], n = 7; alcoholic, n = 9) were examined. Informed consent to all clinical investigations, in accord with the principles outlined in the Declaration of Helsinki, was provided. The institutional review board of the Hospital Clínic de Barcelona (Barcelona, Spain) approved the protocol.

HSC Isolation.

Animals were maintained in the CIC bioGUNE (Derio, Spain) animal facility with appropriate approvals from the institutional review committee on animal use. HSCs were isolated from livers of male Sprague-Dawley rats, bile duct ligated (BDL) mice, and sham-operated mice, as previously described.11

BDL.

BDL was performed in 12-week-old mice by tying the common bile duct using a nonabsorbable filament. Mice (n = 8) were injected in the tail vein with 200 μL of a 0.75-μg/μL solution of HuR-specific short hairpin RNA (shRNA) (sense 5'-gatgcagagagagcaatca-3') or control shRNA (pSM2c; Open Biosystems, Lafayette, CO).

CCl4 Treatment.

Rats (n = 5) were treated with CCl4 diluted (1:1) in corn oil (0.5 μL of CCl4/g body weight) by intraperitoneal injection twice-weekly for 6 weeks. Control animals received vehicle alone (n = 5).

Viral Infection.

Cells were treated with short-hairpin lentiviral particles against HuR [CCGGCCCAC AAATGTTAGACCAATTCTCGAGAATTGGTCTAA CATTTGTGGGTTTTTG] or against LKB1 [CCGG CATCTACACTCAGGACTTCACCTCGAGGTGAA GTCCTGAGTGT-AGATGTTTTT] in the presence of hexadimethrine bromide (8 μg/mL). For control cells, HSCs were infected with pLKO.1 lentiviral vector (Sigma-Aldrich). After 24-hour transduction, cells were selected using puromycin (1.25 μg/mL).

Migration Assay.

Migration using the “scratch assay” was performed in liver kinase B1 (LKB1)- and HuR-silenced cells seeded onto poly-D-lysine–coated dishes, as previously described.13

RNA Isolation and Real-Time Polymerase Chain Reaction.

Quantitative real-time polymerase chain reaction (qPCR) was performed with primers described in Supporting Table 1.

RNA Immunoprecipitation of Ribonucleotide Complexes Coupled to qPCR.

Immunoprecipitation of endogenous RNA/protein complexes were performed as previously described.8, 10

Western Blotting Analysis.

Total proteins were extracted in radioimmunoprecipitation assay buffer. Cytoplasmic and nuclear lysates were prepared with the subcellular proteome extraction kit (Calbiochem). Immunoblotting analysis was performed with specific antibodies (Supporting Table 2).

Immunohistochemistry.

A detailed immunohistochemistry (IHC) protocol of paraffin-embedded sections is provided in the Supporting Materials.

Results

HuR Expression in HSCs From Human CLDs.

We found that activated HSCs (α-SMA+ cells) strongly expressed HuR in surgically resected liver samples from patients with alcoholic (Fig. 1A) and HCV cirrhosis (Fig. 1B). Similarly, activated HSCs expressed HuR in the nucleus of liver sections from two animal models of induced fibrosis—BDL mice (Fig. 1C) and rats treated with CCl45 (Fig. 1D)—suggesting that HuR could play a role during HSC activation.

Figure 1.

HuR in human cirrhosis, BDL mice, and CCl4-treated rats. IHC showing expression of HuR (DAB+ cells, brown, arrowheads) in activated HSCs (α-SMA+ cells, alkaline phosphatase, red) in liver sections from patients with (A) alcoholic and (B) HCV cirrhosis. Representative pictures of 9 alcoholic and 7 HCV samples are shown. (C) BDL mice 7 days after surgery (n = 5) (D) and CCl4-treated rats (n = 5).

HuR Silencing Attenuated Hepatic Fibrosis in BDL Mice.

To confirm the role of HuR in liver fibrosis, we silenced HuR in vivo in BDL mice. Thus, mice were injected in the tail vein with an HuR-specific or control shRNA at 0 hours as well as days 3 and 6 after BDL, then sacrificed 9 days after BDL. HuR silencing was confirmed by qPCR and western blotting in whole liver extracts (Supporting Fig. 1A,B) and, specifically, in HSCs by IHC (Supporting Fig. 1C). HuR silencing resulted in reduced histological liver damage, as observed by hematoxylin and eosin (H&E) staining (Fig. 2A) and decreased alanine aminotransferase and bilirubin serum levels (Supporting Fig. 1D,E). Notably, fibrosis development in these mice was significantly attenuated, as shown by reduced collagen deposition (Fig. 2B), α-SMA expression (Fig. 2C), and col1a1, α-SMA, and TGF-β mRNA levels (Fig. 2D).

Figure 2.

HuR silencing attenuated liver damage and hepatic fibrosis in BDL mice. Livers from BDL mice, injected with Sh control or Sh HuR plasmids, were extracted and the following analyses were performed: (A) H&E staining; (B) Sirius Red staining and quantification of positive areas; (C) α-SMA immunostaining and quantification of positive areas; and (D) qPCR of fibrogenic genes (*P < 0.05). Sh, short hairpin.

HuR silencing also led to reduced protein oxidation (Fig. 3A and Supporting Fig. 1F,G), proliferation (Fig. 3B), macrophage infiltration (Fig. 3C), and lower expression of genes involved in inflammation (iNOS [inducible nitric oxide synthase], IL-1α [interleukin-1α], and TNF-α [tumor necrosis factor alpha]) and infiltration (MCP-1 [monocyte chemoattractant protein-1], F4/80, ICAM-1 [intracellular adhesion molecule 1], MMP9 [matrix metalloproteinase 9], and Actin) (Fig. 3D). Altogether, our results suggest that HuR plays a crucial role in the pathogenesis of cholestatic liver injury.

Figure 3.

HuR silencing attenuated oxidative stress and hepatic inflammation in BDL mice. (A) Quantification of protein oxidation by western blotting, (B) analysis of proliferation by Ki67 IHC and quantification of positive cells, (C) analysis of macrophage infiltration determined by F4/80 IHC and quantification of positive areas, and (D) qPCR of inflammatory genes (*P < 0.05).

HuR Expression During HSC Activation.

The above data suggest that HuR could be regulating HSC activation and fibrosis development, either directly and/or indirectly, by a decrease in liver damage and inflammation. To characterize the effect of HuR in HSC activation only, we examined its expression in primary HSCs isolated from sham and BDL mice 9 days after surgery. HuR mRNA levels increased in HSCs isolated from BDL mice, correlating with HSC activation, as observed by the induction of α-SMA mRNA expression (Fig. 4A). Total, cytoplasmic, and nuclear HuR protein levels were also up-regulated (Fig. 4B).

Figure 4.

HuR is expressed in activated HSCs. (A) qPCR and (B) western blotting analysis in HSCs from sham and BDL mice 9 days after surgery. (C) qPCR and (D) western blotting analysis in culture-activated HSCs. (E) HuR and F-actin staining and (F) qPCR analysis of selected mRNAs after HuR silencing in culture-activated HSCs (7 days) (*P < 0.05).

Similarly, during in vitro activation of primary HSCs on a plastic surface,14 HuR mRNA levels also increased after 3, 5, and 7 days in culture, compared to quiescent HSCs (day 1), correlating with HSC activation, as observed by the induction of α-SMA, cyclin D1, and TGF-β as well as the down-regulation of GFAP (glial fibrillary acidic protein) expression (Fig. 4C). This up-regulation of HuR was confirmed by western blotting in 5-day cultured HSCs, compared to quiescent HSCs (Fig. 4D). HuR silencing in primary HSCs, as confirmed by immunocytochemsitry (Fig. 4E), induced morphological changes (F-actin immunostaining) (Fig. 4E), significantly reduced levels of activation (α-SMA, col1a1, and TGF-β) and proliferation markers (cyclin D1), and markedly increased expression of the quiescent marker, GFAP15 (Fig. 4F). Taken together, our data show that HuR could play a role during HSC activation.

Role of HuR in PDGF-Induced Migration and Proliferation.

We next examined whether HuR activity controlled the functions of two principal mediators of HSC activation (i.e., PDGF and TGF-β). PDGF potently promotes HSC migration and proliferation during fibrosis.16 HuR silencing in primary HSCs isolated from BDL mice (Supporting Fig. 2A) significantly reduced their migratory rate, both basally (Supporting Fig. 2C) and after PDGF treatment (Supporting Fig. 2D), and decreased bromodeoxyuridine (BrdU) incorporation after PDGF stimulation (Supporting Fig. 2E).

HuR silencing in a cell line of activated HSCs (cirrhotic liver fat-storing cells-8B [CFSC-8B] cells)12 (Supporting Fig. 2B) also blocked PDGF-induced migration and proliferation (Fig. 5A–C). In CFSC-8B cells, HuR silencing prevented PDGF-induced increase in mRNA levels of genes regulating proliferation (cyclin D1 and B1), migration (MMP9 and Actin17), and infiltration (MCP-118) (Fig. 5D) as well as cyclin D1 protein (Supporting Fig. 2B). RNA immunoprecipitation of ribonucleaotide complexes coupled to qPCR (RIP-qPCR) analyses revealed a significantly increased binding of HuR to these mRNAs after PDGF stimuli (Fig. 5E).

Figure 5.

PDGF-induced responses are mediated by HuR. HuR silencing in CFSC-8B cells reduced PDGF-mediated (A and B) migration, as shown by slower rate of the gap closure at different time points, (C) proliferation, as shown by percentage of BrDU+ cells 24 hours after treatment, (D) up-regulation of specific mRNAs (Sh C, Sh Control; Sh H, Sh HuR). (E) RIP-qPCR showing binding of selected genes with HuR after PDGF stimulation (*P < 0.05). Sh, short hairpin.

These data demonstrate the importance of HuR in PDGF-mediated HSC proliferation and migration.

Role of Phosphatidylinositol 3-Kinase and Extracellular Signal-Related Kinase in PDGF-Induced HuR.

The abundance and subcellular localization of HuR are important determinants of its activity.19, 20 PDGF treatment increased the expression of HuR mRNA (Fig. 6A) and protein (Fig. 6B,C) levels in CFSC-8B cells as well as its cytoplasmic localization (Fig. 6D). Inhibition of both extracellular signal-related kinase (ERK) and phosphatidylinositol 3-kinase (PI3K) blocked PDGF-induced up-regulation of HuR mRNA and protein (Fig. 6A-C), thus controlling HuR abundance. Recently, it was reported that HuR transcription is controlled by nuclear factor kappa-light-chain enhancer of activated B cells (NFκB)/p65.21 We found that both ERK and PI3K induced nuclear translocation of the NFκB subunit (p65) in response to PDGF (Supporting Fig. 3A,C), and inhibition of this translocation by BAY 11-7802 treatment prevented PDGF-mediated up-regulation of HuR protein expression (Supporting Fig. 3B,D).

Figure 6.

Role of PI3K and ERK in PDGF-induced HuR. (A) qPCR and (B and C) western blotting analysis of total protein extracts showing that ERK inhibition using UO126 and PI3K inhibition using LY-294002 prevents PDGF-induced increase in HuR expression. (D and E) Western blotting showing that ERK inhibition, but not PI3K inhibition, prevents PDGF-induced HuR cytoplasmic translocation. (F) Western blotting showing V5 expression in cytoplasmic extracts of PDGF-treated CFSC-8B cells, transfected with plasmids containing wild-type or mutant HuR (*P < 0.05).

Conversely, we found that cytoplasmic localization of HuR was mediated by the ERK pathway only, and not by the PI3K pathway, unlike as described above (Fig. 6D,E and Supporting Fig. 3E). Post-translational modifications of HuR, such as phosphorylation, play an important role in its subcellular localization.19, 20 We performed mutagenesis of six serine and two threonine residues to the nonphosphorylable residue alanine of HuR protein. Mutation of serine residue 100 and threonine residues 293 or 295 prevented translocation to the cytosol of the mutant protein after PDGF treatment (Fig. 6F and Supporting Fig. 3F) without affecting nuclear levels (data not shown), suggesting that these phosphorylation sites are important for PDGF-induced HuR nucleocytoplasmic shuttling.

Role of Phosphorylated LKB1 in PDGF-Induced HuR.

Recent studies have shown that PDGF induces LKB1 (Ser428) phosphorylation by ERK-induced activation in a cell-type–dependent manner.22 Here, using the CFSC-8B cell line, we found that PDGF-induced LKB1 phosphorylation was blocked by the MAPK/ERK kinase (MEK) inhibitor, U0126 (Fig. 6D and Supporting Fig. 3E). No regulation by the PI3K inhibitor, LY-294002, was observed (Fig 6E and Supporting Fig. 3E). LKB1 silencing did not affect PDGF-induced ERK and protein kinase B (AKT) phosphorylation (Supporting Fig. 4A), showing that LKB1 is a downstream kinase of ERK. Importantly, LKB1 knockdown (Supporting Fig. 4A) prevented HuR cytoplasmic localization (Fig. 7A and Supporting Fig. 4B) and blocked PDGF-induced cyclin D1 protein expression (Supporting Fig. 4C,D) as well as MMP9, actin, MCP-1, cyclin D1, and cyclin B1 mRNA expression (Fig. 7B). Finally, basal and PDGF-induced HSC migration (Fig. 7C) and PDGF-induced proliferation (Fig. 7D) were both reduced after LKB1 silencing.

Figure 7.

LKB1 is a downstream target of ERK in PDGF-treated cells. LKB1 silencing in CFSC-8B cells reduces PDGF-induced (A) HuR translocation, (B) up-regulation of specific mRNAs, (C) migration using the scratch assay, and (D) proliferation, as measured by BrDU incorporation. IHC showing expression of pLKB1 (DAB+ cells, brown, arrowheads) in activated HSCs (α-SMA+ cells, alkaline phosphatase, red) in liver sections from patients with alcoholic (E) and HCV cirrhosis (F). Representative pictures of 9 alcoholic and 7 HCV samples are shown (*P < 0.05).

It is known that LKB1 phosphorylates and regulates adenosine-monophosphate–activated protein kinase (AMPK), and recent studies have shown that activation of AMPK in HSCs leads to the reduction of induced proliferation and migration of HSCs.23, 24 Here, however, we show that in activated HSCs (CFSC-8B), PDGF induced phosphorylated LKB1 (pLKB1) without affecting phosphorylated AMPK levels (Supporting Fig. 5A), and that AMPK silencing did not affect PDGF-induced HuR cytosolic translocation (Supporting Fig. 5B). Altogether, our results suggest that in activated HSCs, AMPK does not mediate LKB1-induced HuR translocation in response to PDGF.

In primary HSCs isolated from BDL mice, PDGF-induced HuR cytosolic localization was also accompanied by LKB1 phosphorylation (Supporting Fig. 3G), and LKB1 silencing (Supporting Fig. 6A) also reduced migration both basally and after PDGF treatment (Supporting Fig. 6B,C) and inhibited PDGF-induced proliferation (Supporting Fig. 6D). Finally, we found strong LKB1 phosphorylation in activated HSCs (α-SMA+ cells) from BDL mice and CCl4-treated rats (Supporting Fig. 6E) and, more important, in human cirrhotic samples (Fig. 7E,F). In summary, our data suggest that LKB1 activation by ERK could be mediating PDGF-induced proliferation and migration in HSCs by regulating, at least partly, cytoplasmic localization of HuR and expression of its critical target genes.

Role of HuR in TGF-β-Induced Profibrogenic Response.

TGF-β is another major mediator of liver fibrogenesis.25 HuR silencing in the CFSC-8B cell line markedly reduced up-regulation of col1a1, α-SMA, and TGF-β mRNA after TGF-β treatment (Fig. 8A). RIP-qPCR analysis showed that α-SMA and TGF-β, but not col1a1, were bound to HuR in TGF-β-stimulated cells (Fig. 8A).

Figure 8.

TGF-β-induced fibrogenesis is mediated by HuR. (A) qPCR showing reduced expression of profibrogenic mRNAs after HuR silencing and RIP-qPCR showing binding of selected genes with HuR after TGF-β stimulation in CFSC-8B cells. (B) HuR silencing prevents the antiproliferative effects of TGF-β, as measured by BrdU incorporation. (C) qPCR showing reduced p21 expression after HuR silencing and RIP-qPCR showing its binding to HuR after TGF-β stimulation. Western blotting showing TGF-β-induced (D) HuR translocation and (E) p38 MAPK activation. (F) Western blotting showing that p38 MAPK inhibition prevents TGF-β-induced HuR translocation (*P < 0.05).

In HSCs, TGF-β also plays a major role in inhibiting proliferation in HSCs.26 TGF-β treatment decreased levels of the cell-cycle activators, cyclin D1 and B1, while increasing levels of the cell-cycle inhibitor, p21 (Supporting Fig. 7A,B). HuR knockdown abrogated the antiproliferative effects of TGF-β in primary HSCs from BDL mice (Supporting Fig. 7C) and in the CFSC-8B cell line (Fig. 8B). This antiproliferative effect of TGF-β was likely the result of reduced p21 levels (Fig. 8C). RIP-qPCR showed that TGF-β treatment induced an increased binding of HuR to p21 while reducing the interaction of cyclin D1 and B1 mRNA with HuR (Fig. 8C).

TGF-β treatment did not regulate HuR at mRNA and protein levels, unlike PDGF (Supporting Fig. 7D,E). However, TGF-β induced increased cytoplasmic localization of HuR, both in primary HSCs (Supporting Fig. 3G) and in the CFSC-8B cell line (Fig. 8D and Supporting Fig. 7F). This translocation is unlikely to be mediated by ERK, AKT, or LKB1, because TGF-β did not activate any of these kinases (Fig. 8E). However, TGF-β activated p38 MAPK (Fig. 8E), and inhibition of this pathway prevented TGF-β-induced HuR translocation (Fig. 8F). TGF-β did not affect phosphorylation at any of the eight residues that we previously tested for PDGF-induced translocation (data not shown), suggesting that TGF-β and PDGF mediate HuR translocation by different post-translational modifications.

In summary, we found that the profibrogenic and antiproliferative actions of TGF-β could be controlled by HuR-mediated regulation of critical genes.

Discussion

Liver fibrosis and cirrhosis result from the majority of chronic liver insults and represent a difficult clinical challenge. Recent studies have shown that HuR regulates angiotensin II–induced kidney fibrosis27 and ventricular remodeling after myocardial infarction.28 However, HuR functions during liver fibrosis development are unknown. Several studies have shown that HuR regulates the expression of several mRNAs encoding proinflammatory cytokines (e.g., TNF-α, IL-6, TGF-β, and interferon-gamma), proinflammatory mediators (e.g., iNOS), and chemoattractant factors (e.g., MCP-1).29 Most of these factors are involved in the pathogenesis of liver fibrosis.4 Here, we show that HuR silencing in a cholestactic liver injury model (i.e., BDL) reduces the expression of several of these genes, leading to decreased liver damage, oxidative stress, inflammation, macrophage infiltration, and liver fibrosis development. This suggests that HuR silencing would have a beneficial effect after cholestactic liver injury.

More important, our study also shows that HuR regulates HSC activation, which likely results in the reduced fibrosis observed in vivo after HuR silencing. HSC activation is highly regulated, with hundreds of genes up- and down-regulated.5 Modulation of mRNA stability and translation rates plays an important role in the regulation of gene expression during liver fibrosis development and hepatic stellate activation.1 Here, we show that HSC activation in vitro and in vivo after BDL is accompanied by an increase in HuR. HuR silencing significantly reduces the expression of HSC activation markers. Importantly, we observed that HuR mediates the response of two of the principal mediators of HSC activation (PDGF and TGF-β).30, 31 These data, together with the finding that HSC from human samples of hepatic cirrhosis expressed HuR, suggest that HuR has a significant role in fibrosis development after liver injury by controlling HSC activation itself, in addition to liver damage and inflammation.

HuR regulates PDGF-induced proliferation and migration, controlling the expression of several genes involved in these processes. PDGF binding to its receptor leads to the sequential activation of RAF photo-oncogene serine/threonine-protein kinase, MEK, and ERK1/2. ERK signaling is involved in PDGF-stimulated mitogenesis, migration, and chemotaxis. PI3K also mediates PDGF-induced proliferation, migration, and chemotaxis, at least in part, through ERK-independent pathways.30 Here, we demonstrated that ERK1/2, but not PI3K, regulates the cytoplasmic translocation of HuR. PDGF also induces LKB1 (Ser428) phosphorylation through ERK activation.22 LKB1 has been classically described as a tumor suppressor,32 but seems to have the opposite role in the liver, controlling HuR nucleocytoplasmic shuttling and proliferation in HGF-stimulated hepatocytes and during apoptosis in hepatoma cell lines.8, 9 Here, we also identified LKB1 as a downstream target of ERK1/2 in PDGF-stimulated HSCs, and silencing LKB1 significantly reduced PDGF-induced migration and proliferation. These functions of LKB1 are possibly mediated by HuR activity, because LKB1 regulates the nucleocytoplasmic shuttling of HuR and both regulate the expression of a common set of mRNAs. It is known that LKB1 phosphorylates and regulates AMPK; however, we observed that PDGF-induced HuR cytosolic localization was independent of AMPK activity. This observation is in agreement with previous work describing that AMPK exerts antiproliferative properties in HSCs,23, 24 as well as with studies in melanoma cells, which show that LKB1 can be active without affecting AMPK activity.22 Previous studies have shown that PI3K and ERK are activated in HSCs in vivo after liver injury.33, 34 Here, we found that, similarly, LKB1 (Ser428) phosphorylation is also expressed in vivo in activated HSCs in two animal models of hepatic fibrosis (i.e., BDL and CCl4) and, importantly, in patients with cirrhosis.

TGF-β1 is another major mediator of liver fibrogenesis.35 We found that TGF-β1 treatment increased the binding of HuR to several target mRNAs, such as α-SMA and TGF-β, and that HuR silencing significantly reduced their expression. Increasing evidence supports a mechanism by which autocrine production of TGF-β is required to maintain the pathogenic myofibroblast phenotype in several cell types.36 We found that col1a1 was significantly reduced after HuR silencing, likely the result of reduced TGF-β autocrine secretion, rather than by regulation of its stability and translation, because we did not find increased binding of col1a1 to HuR. TGF-β1 is also an important negative regulator of proliferation in activated HSCs.25 Our results showed that TGF-β increased the stabilization or translation of p21 mRNA, increasing its binding to HuR. Conversely, we observed a markedly reduced association between HuR and cyclin D1 and cyclin B1 mRNAs in response to TGF-β. The TGF-β-induced decrease in proliferation was abrogated by HuR silencing, suggesting that HuR is an important mediator of the antiproliferative effects of TGF-β.

This role of HuR in TGF-β-treated cells is in sharp contrast to its effects in PDGF-treated cells, where we showed that HuR positively regulated HSC proliferation. Although PDGF activates the ERK/LKB1-signalling pathway to promote HuR translocation, TGF-β induced HuR translocation through p38 MAPK activation. In addition, TGF-β does not phosphorylate the same residues of HuR protein that control its cytoplasmic translocation, induced by PDGF. Thus, it is possible that the specific post-translational modification of HuR induced by the two signals could determine its binding to different mRNA targets. Similarly, PDGF and TGF-β have contrasting roles in regulating the levels of HuR. PDGF, through ERK- and PI3K-mediated activation of NFκB, is sufficient to increase HuR transcription. This is in agreement with other studies, which show that NFκB activity is regulated by cytokines in activated HSCs,11 and that p65 binds to the HuR promoter in gastric tumor cells.21

HuR has been implicated in several biological events, such as carcinogenesis, cell proliferation, differentiation, and inflammation.29 However, both low and high levels of HuR have been correlated with good prognosis in cancer, making careful designs of interventions to modulate HuR functions necessary. These generate the need to study the advantages or disadvantages of HuR silencing in different pathologies, as well as the identification of its specific mediators.29 Here, we have demonstrated that HuR silencing has pleiotropic and beneficial functions during cholestactic liver injury and HSC activation. Importantly, we find that HuR levels in human cirrhotic samples strongly correlate with the degree of HSC activation, suggesting that it could be a valuable therapeutic target for treatment of liver fibrosis and, possibly, its progression to HCC in humans.

Ancillary