Halofuginone Induces Matrix Metalloproteinases in Rat Hepatic ...

2 downloads 0 Views 511KB Size Report
Jan 3, 2006 - Halofuginone (HAL), a semisynthetic quinazolinone alkaloid origi- nally derived from the plant Dichroa febrifuga, has been used to prevent.
THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 22, pp. 15090 –15098, June 2, 2006 © 2006 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

Halofuginone Induces Matrix Metalloproteinases in Rat Hepatic Stellate Cells via Activation of p38 and NF␬B* Received for publication, January 3, 2006 Published, JBC Papers in Press, February 17, 2006, DOI 10.1074/jbc.M600030200

Yury Popov‡§1, Eleonora Patsenker‡2, Michael Bauer‡, Edith Niedobitek‡, Anja Schulze-Krebs‡, and Detlef Schuppan‡§3 From the ‡Department of Medicine I, University of Erlangen-Nuernberg, 91054 Erlangen, Germany and the §Division of Gastroenterology and Hepatology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts 02115 The semisynthetic plant alkaloid halofuginone (HAL) was reported to prevent and partly reverse experimental liver fibrosis. However, its mechanisms of action are poorly understood. We therefore aimed to determine the antifibrotic potential of HAL and to characterize involved signal transduction pathways in hepatic stellate cells (HSCs). Results were compared with its in vivo effects in a rat model of reversal of established liver fibrosis induced by thioacetamide. In vitro HAL inhibited HSC proliferation and migration dose dependently at submicromolar concentrations. HAL (200 nM) up-regulated matrix metalloproteinase (MMP)-3 and MMP-13 expression between 10- and 50-fold, resulting in a 2- to 3-fold increase of interstitial collagenase activity. Procollagen ␣1(I) and MMP-2 transcript levels were suppressed 2- to 3-fold, whereas expression of other profibrogenic mRNAs remained unaffected. p38 mitogen-activated protein kinase (p38 MAPK) and nuclear factor ␬B (NF␬B) pathways were activated by HAL, and specific inhibitors of p38 MAPK and NF␬B dose dependently inhibited MMP-13 induction. Treatment with HAL did not affect HSC viability, and observed effects were reversible after its removal. In vivo HAL upregulated MMP-3 and -13 mRNA expression 1.5- and 2-fold, respectively, in cirrhotic rats, whereas tissue inhibitor of metalloproteinase-1 was suppressed by 50%. In conclusion, submicromolar concentrations of HAL inhibit HSC proliferation and migration and up-regulate their expression of fibrolytic MMP-3 and -13 via activation of p38 MAPK and NF␬B. The remarkable induction of MMP-3 and -13 makes HAL a promising agent for antifibrotic combination therapies.

Liver cirrhosis as a consequence of many forms of chronic liver diseases is associated with a high morbidity and mortality. Treatment for advanced liver fibrosis and cirrhosis is unsatisfactory or inefficient. Thus, there is an urgent need for antifibrotic treatments that can prevent, halt, or even reverse advanced fibrosis. In recent years, significant progress has been made in our understanding of hepatic fibrosis as a dynamic process, characterized by an imbalance between collagen production and degradation, finally leading to distortion of normal hepatic

* This work was supported in part by the Deutsche Forschungsgemeinschaft (DFG, Grant 646/14-1), by the German Competence Network for Viral Hepatitis, and by the Interdisciplinary Center for Clinical Research of the University of ErlangenNuernberg (to D. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Recipient of Yamanouchi (2002) and Sheila Sherlock (2003) Fellowships of the European Association for the Study of the Liver. 2 Recipient of a scholarship of the DFG graduate college (GRK 750). 3 To whom correspondence should be addressed: Division of Gastroenterology and Hepatology, Beth Israel Deaconess MC, Harvard Medical School, Dana 501, 330 Brookline Ave, Boston, MA 02215. Tel.: 617-667-8377; Fax: 617-667-2767; E-mail: [email protected].

15090 JOURNAL OF BIOLOGICAL CHEMISTRY

architecture and loss of function (1, 2). Several studies suggest that human and experimental liver fibrosis may be reversible even at advanced stages, once the pathogenic trigger is eliminated (3– 6). In hepatic fibrosis the excessive extracellular matrix is produced by activated mesenchymal cells which resemble myofibroblasts and derive from quiescent hepatic stellate cells (HSCs)4 and periportal or perivenular fibroblasts (7–9). These myofibroblasts are the major target for antifibrotic therapies. Halofuginone (HAL), a semisynthetic quinazolinone alkaloid originally derived from the plant Dichroa febrifuga, has been used to prevent coccidiosis in poultry, when its antifibrotic properties were discovered accidentally, since skin tearing and loss of skin integrity were observed in broilers that received HAL in the diet (10). During the last decade HAL was investigated as an antifibrotic agent in various in vivo fibrosis models (reviewed in Ref. 11) and suggested to be a specific inhibitor of procollagen type I expression for various collagen-expressing cells (12– 15). More recently, HAL was shown to partly reverse established thioacetamide-induced rat liver fibrosis (16). However, inhibition of procollagen I synthesis alone can hardly explain the ability of HAL to reverse pre-established fibrosis in rats. So far, there exist no experimental studies that explain the mode of action of HAL in hepatic fibrosis and, especially, toward the hepatic fibrogenic effector cells, i.e. activated HSCs and myofibroblasts. Here we present an extensive analysis of the in vitro antifibrotic potential of HAL, study the signal transduction pathways triggered by the drug in HSCs/myofibroblasts, and provide in vivo evidence of its pro-fibrolytic mode of action in a rat model of fibrosis reversal. We show that HAL is a prominent activator of fibrolytic matrix metalloproteinase (MMP)-3 and -13 in HSCs, a hitherto unique property for an antifibrotic agent, which is mediated by activation of p38 mitogen-activated protein kinase (MAPK) and nuclear factor ␬B (NF␬B).

MATERIALS AND METHODS HSC Isolation and Culture—CFSC-2G (17), HSC line obtained from cirrhotic rat liver (kind gift from Dr. M. Rojkind, Washington, D. C.) and culture-activated primary rat HSCs were seeded onto 12-, 24-, or 96-well plastic plates at a density 40,000 cells per ml at 95% air/5% CO2 in a humidified atmosphere. Cells were maintained in Dulbecco’s modified Eagle’s medium (10% FCS, 1% penicillin/streptomycin) and used

4

The abbreviations used are: HSC, hepatic stellate cell; HAL, halofuginone; I␬B-␣, inhibitory ␬B-␣ protein; MF, myofibroblast-like cell; MMP, matrix metalloproteinase; NF␬B, nuclear factor ␬B; MAPK, mitogen-activated protein kinase; Erk, extracellular signalregulated kinase; PI3K, phosphatidylinositol 3-kinase; FCS, fetal calf serum; BrdUrd, bromodeoxyuridine; PDGF, platelet-derived growth factor; RT, reverse transcription; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; APMA, 4-aminophenyl mercuric acetate; TRITC, tetramethylrhodamine isothiocyanate; CREB, cAMP-response element-binding protein; TAA, thioacetamide; HYP, hydroxyproline; TIMP-1, tissue inhibitor of metalloproteinase-1; SAPK, stress-activated protein kinase; JNK, c-Jun N-terminal kinase.

VOLUME 281 • NUMBER 22 • JUNE 2, 2006

Halofuginone Induces MMPs in Stellate Cells TABLE 1 Primers and probes used in real-time RT-PCR Target molecule

5ⴕ-Primer

Probe

3ⴕ-Primer

Procollagen ␣1(I) Procollagen ␣1(III) MMP-2 MMP-3 MMP-13 TIMP-1 TGF␤1 CTGF GAPDH

TCCGGCTCCTGCTCCTCTTA AATGGTGGCTTTCAGTTCAGCT CCGAGGACTATGACCGGGATAA CCGTTTCCATCTCTCTCAAGATGA GGAAGACCCTCTTCTTCTCA TCCTCTTGTTGCTATCATTGATAGCTT AGAAGTCACCCGCGTGCTAA ATCCCTGCGACCCACACAAG CCTGCCAAGTATGATGACATCAAGA

TTCTTGGCCATGCGTCAGGAGGG TGGAAAGAAGTCTGAGGAAGGCCAGCTG TCTGCCCCGAGACCGCTATGTCCA AGATGGTATTCAATCCCTCTATGGACCTCC TCTGGTTAGCATCATCATAACTCCACACGT TTCTGCAACTCGGACCTGGTTATAAGG ACCGCAACAACGCAATCTATGACAAAACCA CTCCCCCGCCAACCGCAAGAT TGGTGAAGCAGGCGGCCGAG

GTATGCAGCTGACTTCAGGGATGT TGTAATGTTCTGGGAGGCCC CTTGTTGCCCAGGAAAGTGAAG CAGAGAGTTAGATTTGGTGGGTACCA TCATAGACAGCATCTACTTTGTC CGCTGGTATAAGGTGGTCTCGAT TCCCGAATGTCTGACGTATTGA CAACTGCTTTGGAAGGACTCGC GTAGCCCAGGATGCCCTTTAGT

FIGURE 1. Effects of HAL on proliferation and migration of HSCs. Serum-stimulated DNA synthesis of CFSC-2G cells (A) and primary HSCs (B) (first passage and 7 days in culture) was measured by BrdUrd incorporation (6 – 8 parallel wells per experimental condition, % of controls stimulated with 10% fetal calf serum alone). C, cell counts of CFSC-2G cells treated with HAL in growth medium. D, inhibition of migration of liver myofibroblasts in the presence of HAL, as assessed by the scratch assay (% reduction of initial scratch width). E, reversibility of the antiproliferative effect of HAL. Cells seeded at 30,000/well in growth medium were exposed to 200 nM HAL from 24 to 48 h, after which the medium was replaced by growth medium without HAL. Cells grown in growth medium alone served as a control. All graphs are representative of ⱖ3 independent experiments and expressed as means ⫾ S.E.

for experiments at low (0.5% FCS) serum conditions after 12–24 h unless specified otherwise. HAL (kind gift of Dr. M. Pines, Rehovot, Israel) was added directly to the culture medium. All reagents were from Sigma (Taufkirchen, Germany) if not stated otherwise. Primary HSCs were isolated from male Wistar rats (Retired Breeders, 450 –500 g, Charles River, Sulzfeld, Germany) according to a previously published procedure (18). Briefly, the liver was perfused with 0.1% Pronase E and 0.025% type IV collagenase in Dulbecco’s modified Eagle’s medium for 10 –15 min, followed by digestion with 0.04% Pronase, 0.025% collagenase, and 0.002% DNase at 37 °C for 10 –30 min and by a two-step centrifugation through a 11 and 13% gradient of Nycodenz at 1,500 ⫻ g for 15 min. Cell viability was assessed by Trypan Blue exclusion and was routinely greater than 95–98%. Purity of HSC isolates was confirmed by their stellate shape, and cytoplasmic lipid-droplets showing greenish autofluorescence at 390 nm excitation. Contamination

JUNE 2, 2006 • VOLUME 281 • NUMBER 22

with Kupffer cells, as assessed by the ability to engulf 3-␮m latex beads, was ⬍3–5% after isolation and undetectable after the first passage. Cells were used freshly or after the first passage if not stated otherwise. Myofibroblast-like cells (MF) were obtained by outgrowth from primary HSC cultures (three to fifth passages). Cell Proliferation—DNA synthesis was evaluated measuring BrdUrd incorporation using an enzyme-linked immunosorbent assay (Roche Applied Science). CFSC-2G and culture-activated primary HSCs (5–7 days) were seeded at a density of 2,000 cells/well in 96-well plates in growth medium containing 10% FCS for 24 h. After 24 h of starvation in 0.25% FCS, cells were stimulated by 10% FCS for 24 h in the presence of increasing concentrations of HAL. BrdUrd was added during the last 4 h, and its incorporation was quantified using an anti-BrdUrd peroxidase-labeled antibody. All experiments were done with 6 – 8 wells in parallel and repeated at least three times. For

JOURNAL OF BIOLOGICAL CHEMISTRY

15091

Halofuginone Induces MMPs in Stellate Cells

FIGURE 2. HAL up-regulates expression of fibrolytic MMP-3 and -13. MMP-13 (A) and MMP-3 (B) mRNA in CFSC-2G cells treated with HAL at 200 nM. mRNA expression of MMP-3 (C) in primary early activated HSCs (first passage, 3 days in culture) of MMP-13 (D) and MMP-3 (E) in liver myofibroblasts (fifth passage), treated with HAL at 200 nM for 36 h. Transcript levels were quantified by real-time PCR. Data are representative of ⱖ3 independent experiments performed in triplicates (means ⫾ S.E., arbitrary units relative to GAPDH mRNA). HAL up-regulates expression of pro-MMP-13 and MMP-3 protein in CFSC-2G cells. F, Western blot using a rat-specific anti-MMP-13 antibody in conditioned media from CFSC-2G cells treated with HAL after 20-fold concentration by acetone precipitation. Rat keratinocyte pro-MMP-13 (62 kDa) served as positive control (first lane). G, Western blot from cell lysates using an anti human MMP-3 antibody detected a 52-kDa band corresponding to pro-MMP-3. ␤-Actin expression (lower band) served as loading control.

determination of cell numbers cells were seeded at 40,000 cells/well in 6-well plates and maintained in growth medium for 24 h followed by different periods of treatment in triplicates. After trypsinization cell numbers were determined with a Coulter counter (Coulter威 Z1, Coulter Electronics Ltd., Luton, UK). Cell Migration Assay—Cell migration was assessed by measuring the repopulation of a linear wound made in a confluent cell monolayer. Cells were grown in 24-well plates until confluency and starved in serum-free medium for 24 h. Thereafter a linear wound was generated in the monolayer by scraping a sterile 100-␮l plastic pipette tip perpendicular to three parallel drawn lines, and medium was changed to remove scraped cells. After 3 h, the distance between the cell fronts was measured with a micrometer, using the three lines as a reference. 10 ng/ml platelet-derived growth factor-BB (PDGF-BB) or 10% FCS and HAL at increasing concentrations were immediately added, and migration was assessed during the following 12–24 h. Three measurements per well were performed in three wells for each condition, and data are expressed as a reduction of initial scratch width. Real-time RT-PCR—Real-time RT-PCR was performed as described (19 –21). Briefly, total RNA was isolated from the cell lysates or liver homogenates using RNApure (PeqLab, Erlangen, Germany) and 0.5 ␮g of total RNA was reverse transcribed using Superscript II reverse tran-

15092 JOURNAL OF BIOLOGICAL CHEMISTRY

scriptase (Invitrogen) with 50 pmol of random hexamer and 100 pmol of oligo(dT) primers (Promega, Mannheim, Germany). Relative mRNA transcript levels were quantified on a LightCycler (Roche Applied Science) using the TaqMan technology. The housekeeping gene GAPDH was amplified in a parallel reaction for normalization. TaqMan probes and primer sets were designed using the Primer Express software (PerkinElmer Life Sciences) based on published sequences as summarized in Table 1. All TaqMan probes are positioned such as spanning exon-exon boundaries of corresponding genes to exclude co-amplification of genomic DNA. Sense and antisense primer (each at 0.5 ␮M) and 0.125 ␮M probe, labeled and phosphorylated at its 5⬘-end with 6-carboxyfluoresceine (6-FAMTM) and at the 3⬘-end with tetramethylrhodamine, were synthesized at MWG Biotech (Ebersberg, Germany) and validated using conventional RT-PCR and agarose chromatography (20). Relative mRNA transcript levels were expressed in arbitrary units as n-fold versus untreated controls (mean ⫾ S.E.) after normalization to GAPDH mRNA. MMP Protein Production—MMP protein production was semiquantified by Western blotting as described previously (22) in 20-fold concentrated conditioned media (for MMP-13) or in cell lysates (for MMP-3) of CFSC-2G cells treated with or without HAL for 0 –36 h. Pro-MMP-13 (60 kDa) from conditioned media of rat keratinocytes

VOLUME 281 • NUMBER 22 • JUNE 2, 2006

Halofuginone Induces MMPs in Stellate Cells

FIGURE 3. HAL down-regulates profibrogenic procollagen ␣1(I) and MMP-2 genes. Procollagen ␣1(I) (A and C) and MMP-2 (B and D) mRNA expression in CFSC-2G cells (A and B) and primary HSCs (C and D) (first passage, 3 days in culture) treated for 12 and/or 36 h with 200 nM HAL. Data are representative of three independent experiments performed in triplicates (means ⫾ S.E., arbitrary units relative to GAPDH).

(CCL-4 cells) served as positive control (MMP-13 antigen and antiMMP-13 antibody from Lab Vision, Westinghouse, CA; rabbit antihuman MMP-3 from AnaSpec, San Jose, CA). Conditioned media and cell lysates were standardized by equal cell numbers and volume of media (1 ⫻ 106cells/ml). Equal loading was confirmed by Ponceau S staining of membranes after protein transfer and/or concomitant blotting for ␤-actin. Interstitial Collagenase Activity Assay—Conditioned media were activated with 2 ␮M 4-aminophenyl mercuric acetate (APMA) at 37 °C for 5 h to measure total interstitial collagenase activity using degradation of native bovine biotinylated type I collagen as substrate (ECM710, Chemicon, Hampshire, UK). 100 ␮l of APMA-activated conditioned media was incubated with substrate for 2 h at 37 °C, and solubilized biotinylated fragments were transferred to a biotin-binding 96-well plate followed by streptavidin-peroxidase detection according to the manufacturer’s recommendations. APMA-activated human recombinant MMP-1 was used as a standard. Determination of NF␬B Activation—Cells were seeded in 75-cm2 flasks, grown until confluence, starved in serum-free medium for 24 h, and treated with HAL at 200 nM for various periods of time. After centrifugation in phosphate-buffered saline at 4 °C and 1200 rpm for 5 min, 2 ⫻ 106 cells were resuspended in ice-cold nuclear extraction buffer (5 mM HEPES-KOH, 0.75 mM MgCl2, 5 mM KCl, 0.25 mM dithiothreitol, 0.1 mM PMSF, pH7.9), incubated for 10 min, and vortexed for 1 min. Nuclei were collected at 1200 rpm for 30 s at 4 °C. The nuclear pellet was resuspended in 300 ␮l of cell lysis buffer (see MAPK activation assay below), boiled for 10 min under reducing conditions, and frozen at ⫺20 °C until use. The p65-subunit was detected in nuclear extracts by Western blotting (see MAPK activation assay) using an antip65 antibody (1:200, from Delta Biolabs). Inhibitory ␬B protein (I␬B-␣) degradation was assessed by Western blotting using anti-I␬B␣ antibody (1:1000, from Rockland Inc., Gilbertsville, PA) in CFSC-2G cell lysates 30 min obtained after addition of HAL. For immunofluorescence, subconfluent cells were starved in 0.25% FCS/Dulbecco’s modified Eagle’s medium for 24 h and treated with 200 nM HAL for 0.5– 6 h, washed with ice-cold phosphate buffer, and fixed in cold methanol for 10 min. p65 NF␬B was detected by incubating with the anti-p65 antibody (1:100) for

JUNE 2, 2006 • VOLUME 281 • NUMBER 22

FIGURE 4. HAL stimulates interstitial collagenolytic activity. Total interstitial collagenase activity in conditioned media from liver myofibroblasts, treated for 18 h with increasing concentrations of HAL (A), as determined by degradation of biotinylated native bovine collagen type I. Time course of collagenolytic activities of conditioned media from myofibroblast cultures, untreated (open bars) or treated (closed bars) with 200 nM HAL (B). Each figure is representative of two independent experiments performed in triplicates (means ⫾ S.E., expressed as nanograms of human recombinant MMP-1).

JOURNAL OF BIOLOGICAL CHEMISTRY

15093

Halofuginone Induces MMPs in Stellate Cells

FIGURE 5. Halofuginone activates p38 MAPK and NF␬B signaling pathways. Western blots showing phosphorylation of p38 and Erk1/2 (p42/44) in cells treated with HAL. CFSC-2G cells starved for 24 h were stimulated with PDGF-BB (10 ng/ml) for 10 min with or without pre-treatment with HAL (200 nM) for 6 h. A, cell lysates in parallel blots were probed with antibodies against phosphorylated and total Erk-1/2 and p38 MAPK (see “Materials and Methods”). B, Western blotting for p65 NF␬B nuclear translocation. Nuclear extracts were prepared from CFSC-2G cells treated for 0 –12 h with HAL at 200 nM and probed with anti-p65 NF␬B antibodies. C, subcellular localization of p65 as assessed by immunofluorescence in a cell monolayer of CFSC-2G cells demonstrates maximal nuclear translocation 6 h after addition of 200 nM HAL. Addition of MG132 (15 ␮g/ml) 30 min prior to HAL effectively abrogates p65 nuclear translocation. D, Western blotting of CFSC-2G lysates probed for I␬B-␣ 30 min after addition of 10 –200 nM HAL. The lower band is ␤-actin (loading control). Images are representative of at least three independent experiments.

30 min followed by TRITC-conjugated anti-rabbit IgG (1:200, Dako, Germany). Subcellular localization of p65 was assessed, and representative images were documented using a scanning confocal microscope (Carl Zeiss, Germany). To inhibit NF␬B nuclear translocation (23), the proteasome inhibitor MG132 (Rockland Inc.) was used. MAPK Activation Assays—Confluent, starved cells were stimulated with 10 ng/ml PDGF-BB for 10 min with or without pretreatment with HAL for 4 h. Cell extracts were prepared on ice using cell lysis buffer (50 mM Tris-HCl, 1% Tween 20, 0.25% SDS, 150 mM NaCl, 1 mM EGTA, 1 mM Na3VO4, 1 mM NaF) and CompleteTM protease inhibitor mixture (Roche Applied Science), boiled for 10 min under reducing conditions, and frozen at ⫺20 °C until use. To inhibit activation of individual kinases, several specific inhibitors were used: SB203580 (p38 MAPK), U0126 (Erk1/2), and LY294002 (PI3K) (all from LC Labs, Woburn, MA). Total and phosphorylated MAPKs, Akt-1, and phospho-CREB were visualized by Western blotting using respective antibodies recognizing phosphorylated and total kinases (from Cell Signaling Technology, Beverly, MA). 10 ␮g of cell lysates standardized by cell number was run on a 14% SDS-polyacrylamide gel, blotted onto nitrocellulose, and stained with 0.5% Ponceau S to assure equal protein loading and transfer. Membranes were blocked with 5% powdered milk in TBS-T (25 mM TrisHCl, pH 8.0, 144 mM NaCl, 0.1% Tween 20), incubated overnight at 4 °C with primary antibodies, washed, and incubated for 2 h with their corresponding peroxidase-conjugated secondary antibodies. Immunodetected proteins were visualized utilizing the enhanced chemiluminescence assay kit (Amersham Biosciences). Animal Experimentation— 48 male Wistar rats (300 –330 g) were purchased from Charles River (Sulzfeld, Germany). The protocol for animal experimentation was approved by the Government of Lower Franconia (permission number 621.2531.31-20/00). Cirrhosis was induced in 40 rats by intraperitoneal administration of thioacetamide

15094 JOURNAL OF BIOLOGICAL CHEMISTRY

(TAA) dissolved in saline, 200 mg/kg twice a week, for 12 weeks. Four rats died during induction and were not included in the study, and eight rats which received saline instead of TAA served as controls. After 12 weeks of TAA treatment, cirrhotic rats were randomly divided into three groups: PF (peak of fibrosis, n ⫽ 12), SR (spontaneous recovery, n ⫽ 12), and HAL (SR and treatment with HAL, n ⫽ 12). Group PF was sacrificed 5 days after the last TAA injection, and the HAL group received halofuginone (0.05 mg/kg), dissolved in 3% carboxymethylcellulose, by daily oral gavage for the following 8 weeks, while the SR group received equal amounts of vehicle for 8 weeks. Thereafter animals were sacrificed, and fibrosis parameters were evaluated. Total (milligrams/ liver) and relative (milligrams/g of liver) hydroxyproline (HYP) was determined biochemically from 250 mg of liver as described previously (19, 24). Relative mRNA expression was quantified by real-time RTPCR from total RNA, extracted from liver homogenates (⬃100 mg of tissue) as described above. Histological staining for connective tissue (Sirius Red) was performed in formalin-fixed, paraffin-embedded liver specimens from two different lobes of each animal according to the routine protocol of Department of Pathology, University of Erlangen. Statistical Analysis—Statistical analyses were performed using Microsoft EXCEL software. Data are expressed as means ⫾ S.E. The statistical significance of differences was evaluated using the unpaired, non-parametric Student’s t test.

RESULTS Halofuginone Inhibits Hepatic Stellate Cell Proliferation and Migration—HAL dose dependently inhibited serum-stimulated cell proliferation as measured by BrdUrd incorporation, with a maximal inhibition at 200 nM in both CFSC-2G and primary rat HSCs (Fig. 1, A and B). Inhibition of DNA synthesis was consistent with reduction in cell number (Fig. 1C). The effect of HAL on serum-induced migration of HSCs,

VOLUME 281 • NUMBER 22 • JUNE 2, 2006

Halofuginone Induces MMPs in Stellate Cells

FIGURE 6. p38 MAPK and NF␬B are causally involved in the induction of MMP-13 in HSCs. A, lack of effect of the Erk1/2 inhibitor U0126 (U, 5 ␮g/ml) and the PI3K inhibitor LY294002 (LY, 5 ␮g/ml), in contrast to complete abrogation of MMP-13 induction by SB203580 (SB, 5 ␮g/ml) and MG132 (MG, 15 ␮g/ml). Cells were treated with 200 nM HAL for 12 h 30 min after addition of inhibitors in 0.5% FCS Dulbecco’s modified Eagle’s medium. B, Western blots demonstrating efficient and specific inhibition of respective kinases by preincubation with inhibitors at doses as used in Fig. 6A. Cells were stimulated with 10 ng/ml PDGF-BB for 10 min and analyzed as described under “Materials and Methods.” Total p38 MAPK served as a loading control. C, inhibition of activation of p38 MAPK (upper band) and its downstream target phospho-CREB (middle band) by SB203580 in CFSC-2G cells in the presence of 0.5% FCS and 200 nM HAL. The phosphorylation state of kinases was evaluated after 10 min of stimulation with PDGF-BB (10 ng/ml) or fetal calf serum (0.5%). C, MMP-13 mRNA expression in CFSC-2G cells treated

JUNE 2, 2006 • VOLUME 281 • NUMBER 22

as measured in an in vitro wounding assay, was completely blocked in the presence of 100 nM HAL, with significant inhibition already at 10 nM (Fig. 1D). PDGF-BB-induced migration was inhibited similarly (not shown). Cell morphology, as visualized by phase-contrast microscopy, and cell viability, as assessed by Trypan Blue exclusion, were not affected by HAL. Moreover, the antiproliferative effect of 200 nM HAL added for 24 h was readily reversible after removal of the drug (Fig. 1E). Halofuginone Affects Expression of Fibrolytic Genes in HSCs—In CFSC-2G cells 200 nM HAL strongly up-regulated MMP-3 and MMP-13 mRNA expression between 10- and 50-fold (Fig. 2, A and B). MMP-13 mRNA was maximal at 12 h after addition and slowly declined to basal levels after 36 h (Fig. 2A), whereas MMP-3 mRNA peaked at 36 h (Fig. 2B). In freshly isolated, early culture-activated HSCs (ⱕ14 days) MMP-13 mRNA was almost undetectable in both untreated and HALtreated cells (data not shown), whereas HAL induced MMP-3 mRNA 6-fold at 36 h (Fig. 2C), similar to its effect in CFSC-2G cells. Myofibroblasts obtained by outgrowth from long term cultured primary HSCs were found to be positive for MMP-13 mRNA. Here, HAL up-regulated MMP-3 and -13 mRNAs 12- and 19-fold, respectively, after 36 h of treatment (Fig. 2, D and E). Western blotting of concentrated CFSC-2G conditioned media and cell lysates confirmed an increase in (pro-) MMP-3 and -13 protein expression, respectively, upon treatment with HAL (Fig. 2, F and G). In CFSC cells and culture-activated HSCs, HAL down-regulated procollagen ␣1(I) and MMP-2 transcripts 2- and 3-fold, respectively, within 24 and 36 h (Fig. 3, A–D). Similar results were obtained for myofibroblasts (not shown). These findings confirm previous observations by others in avian fibroblasts and a rat HSC line for procollagen ␣1(I) (12, 16), and in bladder carcinoma cells for MMP-2 (25). Procollagen ␣1(III) expression was suppressed similarly as procollagen ␣1(I) mRNA, whereas transforming growth factor ␤1, CTGF (connective tissue growth factor), and TIMP-1 and -2 transcript levels remained unaffected in both CFSC-2G cells and primary HSCs (data not shown). Halofuginone Stimulates Collagenolytic Activity of HSCs—Total (APMA-activated) collagenolytic activity was measured in conditioned media of myofibroblasts, obtained by outgrowth from primary HSCs. HAL enhanced degradation of native collagen type I by hepatic myofibroblasts in a time- and dose-dependent manner, with a 2- to 3-fold increase by 200 nM HAL after 12–18 h (Fig. 4, A and B). Intrinsic (without APMA activation) collagenolytic activity was below the detection limit of our system (not shown). Halofuginone Acts via Activation of p38 MAPK and NF␬B—To explore the mechanisms that lead to MMP-3 and MMP-13 induction by HAL, we studied activation of several MAPKs and NF␬B. Incubation of cells with HAL alone without subsequent stimulation by PDGF-BB did not modulate phosphorylation of kinases as compared with untreated controls (not shown). However, stimulation of the cells with 10 ng of PDGF for 10 min after preincubation with 200 nM HAL for 2 h induced a strong increase of p38 MAPK phosphorylation, without a change of extracellular signal-regulated kinase 1/2 (Erk1/2) phosphorylation as compared with cells, stimulated with PDGF-BB alone (Fig. 5A). Activation of SAPK/JNK or PI3K/Akt was unaffected by HAL (not shown). In addition, HAL caused a time-dependent increase of nuclear p65-NF␬B compared with untreated controls as detected by semiquantitative Western blot (Fig. 5B). Anti-p65 immunohistochemistry on CFSC-2G cells exposed to 200 nM HAL confirmed nuclear translocation of p65 with HAL at 200 nM for 12 h in presence of 0.5% FCS after a 30-min pre-treatment with increasing concentrations of the specific p38 MAPK and NF␬B-inhibitors, SB203580 (SB) and MG132 (MG). mRNA data are representative of three independent experiments performed in triplicates (means ⫾ S.E., arbitrary units relative to GAPDH).

JOURNAL OF BIOLOGICAL CHEMISTRY

15095

Halofuginone Induces MMPs in Stellate Cells TABLE 2 Changes in liver/spleen weight and hepatic collagen content in rats with TAA-induced cirrhosis, and its spontaneous recovery versus HAL treatment for 8 weeks Liver and spleen were weighed at the end of experiment, collagen deposition was determined biochemically as relative (micrograms/g) HYP content in 250 mg of liver samples from two different lobes, and total HYP (milligrams/whole liver) was calculated based on individual liver weight and the relative HYP value of its representative sample. Con: non-cirrhotic control group (n ⫽ 8); PF: cirrhotic rats at the peak of TAA-induced fibrosis (n ⫽ 12); SR: spontaneous resolution, after an additional 8 weeks of vehicle treatment (n ⫽ 12), HAL: spontaneous resolution, after an additional treatment with 0.05 mg/kg HAL for 8 weeks (n ⫽ 12). Data are expressed as means ⫾ S.D. Liver weight, g Spleen weight, g Relative HYP, ␮g/g Total liver HYP, mg a b

Con

PF

SR

HAL

16.49 ⫾ 1.60 0.82 ⫾ 0.07 298.32 ⫾ 36.9 4.91 ⫾ 0.66

17.25 ⫾ 2.26 1.84 ⫾ 0.50 1139.33 ⫾ 387.66 19.25 ⫾ 5.03

20.38 ⫾ 3.70a 1.72 ⫾ 0.60 949.65 ⫾ 168.35 19.58 ⫾ 5.79

17.76 ⫾ 3.03 1.65 ⫾ 0.44 817.07 ⫾ 117.22b 14.45 ⫾ 2.89b

p ⬍ 0.05 as compared to the peak of fibrosis (PF) group. p ⬍ 0.05 as compared to the spontaneous resolution (SR) group.

FIGURE 7. Histological assessment of fibrosis in rat liver sections (Sirius Red staining). Representative liver sections stained for connective tissue of the non-cirrhotic control group (A, Con, n ⫽ 8); rats at the peak of TAA-induced fibrosis (B, PF, n ⫽ 12); rats with “spontaneous resolution” after an additional 8 weeks of treatment with vehicle alone (C, SR, n ⫽ 12); after an additional treatment with 0.05 mg/kg HAL daily for 8 weeks (D, HAL, n ⫽ 12).

already at 1 h and peaking at 6 h. p65 nuclear translocation was abrogated by pre-treatment with the proteasome inhibitor MG132 (Fig. 5C). This was preceded by degradation of I␬B-␣ at 30 min (Fig. 6D). To answer the question whether these pathways are indeed causally involved in the observed up-regulation of fibrolytic MMP expression by HAL, cells were preincubated with specific inhibitors of several kinases. Confirming their lack of activation by HAL, inhibition of Erk1/2 or PI3K by their specific inhibitors U0126 and LY294002 did not affect MMP-13 mRNA expression, which was used as read-out for inhibition, whereas inhibitors of p38 MAPK (SB203580) phosphorylation and of NF␬B nuclear translocation (proteasome inhibitor MG132), blocked induction of MMP-13 mRNA expression (Fig. 6A). Pharmacological inhibition of kinases and NF␬B was efficient and specific (Figs. 5C and 6B). Thus, SB203580 inhibited activation of p38 and one of its downstream targets (CREB) under the same experimental conditions as used for the MMP-13 read-out experiment (Fig. 6C). Moreover, HAL-induced MMP-13 expression was inhibited dose dependently by both SB203580 and MG132 (Fig. 6D), clearly indicating that these two signaling pathways are upstream of collagenase induction by HAL. HAL Stimulates Fibrolysis in Cirrhotic Liver via Induction of Fibrolytic mRNAs—To investigate whether MMP induction by HAL in vitro is also operative in vivo, cirrhosis was induced in rats with TAA for 12 weeks, followed by HAL treatment for another 8 weeks. TAA for 12 weeks caused advanced liver fibrosis, with a 3.8- and 3.9-fold increase in relative and total hepatic HYP, respectively (Table 2), and the histological picture of micronodular cirrhosis (Fig. 7B). After 8 weeks of spontaneous recovery there was a trend toward a decreased relative hepatic HYP content, obviously due to liver regeneration as it was accompanied

15096 JOURNAL OF BIOLOGICAL CHEMISTRY

by an increase in liver mass. However, total hepatic HYP content remained unchanged indicating no spontaneous fibrolysis during 8 weeks (Table 2), which was also confirmed by connective tissue stainings in liver specimens showing no reduction in fibrillar collagen content (Fig. 7C). Treatment with 0.05 mg/kg HAL for 8 weeks after TAA withdrawal resulted in a significant reduction of relative and total liver HYP (Table 2). Histologically this was accompanied by thinning of fibrotic septa, which was a typical finding in a group receiving HAL during recovery (Fig. 7D). At the peak of TAA-induced cirrhosis, procollagen ␣1(I), TIMP-1, and MMP-13 mRNA were up-regulated 13-, 5-, and 3,3-fold, respectively. After 8 weeks of spontaneous recovery, procollagen ␣1(I) mRNA declined by 77%, whereas TIMP-1 and MMP-13 were down-regulated by ⬃40%. In the HAL-treated group, as compared with the vehiclealone treated group, procollagen ␣1(I) expression remained unaffected (Fig. 8, A and B), whereas MMP-3 and -13 mRNA were up-regulated 1.4- and 2-fold, respectively, and TIMP-1 was down-regulated 2-fold (Fig. 8, C and D).

DISCUSSION Our study explored the multiple effects of halofuginone (HAL) on activated hepatic stellate cells (HSCs). HAL suppressed HSC proliferation and migration, which was accompanied by profibrolytic modulation of extracellular matrix-related genes. In particular HAL induced a pronounced up-regulation of MMP-3 and -13 transcripts, which resulted in a 2- to 3-fold increased total interstitial collagenolytic activity. Of note, MMP-3 and interstitial collagenases are implicated to play a key role in fibrolysis and potential reversion of liver fibrosis (26–28). Furthermore, the time course of inducible interstitial collagenolytic activity coincides with the expression of MMP-13, supporting an important role of MMP-13 as hepatic interstitial collagenase. This effect was accompanied by down-regulation of procollagens I and III and MMP-2 mRNA. MMP-2 is suspected to favor fibrogenesis rather than fibrolysis due to degradation of collagen IV (29), which serves as scaffold for differentiation and quiescence-inducing basement membranes (30, 31). Furthermore, MMP-2 is up-regulated by the predominant profibrogenic cytokine transforming growth factor-␤1, in contrast to interstitial collagenases and stromelysins (e.g. MMP-3 and -13), which are down-regulated by this cytokine. In this line, MMP-2 is implicated in promoting HSC migration and proliferation (32, 33), activities that are causally linked to fibrogenesis, and it is up-regulated 2- to 3-fold in experimental and human liver fibrosis in vivo (34, 35). In our TAA-induced cirrhosis reversal model HAL significantly reduced hepatic collagen content; however, fibrosis resolution was less marked as compared with the previous study by Bruck et al. (16) who used a similar model. This may be explained by the higher hepatic HYP content in our model, suggesting more advanced and thus less reversible fibrosis, leading to a less potent antifibrotic efficacy of

VOLUME 281 • NUMBER 22 • JUNE 2, 2006

Halofuginone Induces MMPs in Stellate Cells

FIGURE 8. Halofuginone alters the hepatic mRNA expression profile toward a profibrolytic pattern in vivo. Hepatic procollagen ␣1(I) (A), TIMP-1 (B), MMP-13 (C), and MMP-3 (D) mRNA expression as quantified by real-time RT-PCR in total liver RNA. Con: non-cirrhotic control group (n ⫽ 8); PF: cirrhotic rats at peak of TAA-induced fibrosis (n ⫽ 12); SR: spontaneous resolution, vehicle for 8 weeks (n ⫽ 12); HAL: spontaneous resolution and 0.05 mg/kg HAL for 8 weeks (n ⫽ 12). Data are expressed as means ⫾ S.D. and in arbitrary units relative to GAPDH mRNA. *, p ⬍ 0.05 as compared with the PF group; ⫹, p ⬍ 0.05 as compared with the SR group.

HAL. Nonetheless, our animal data provide first in vivo evidence that HAL significantly up-regulates MMP-3 and -13, whereas it downregulates the ubiquitous MMP inhibitor TIMP-1. The low procollagen ␣1(I) expression observed after 8 weeks of spontaneous recovery suggests that degradation of the pre-existing excess of extracellular collagen is a rate-limiting factor in our model of TAA-induced rat liver fibrosis reversal rather than inhibition of de novo procollagen expression. Thus our in vitro as well as in vivo data show that fibrolytic MMP induction is the most striking effect of HAL as compared with its relatively moderate inhibitory effect on procollagen I gene expression described previously (12, 16). In the same vein two recent studies reported that adenoviral gene delivery of interstitial collagenases to fibrotic rat liver resulted in significant fibrosis reversal (36, 37). However, this treatment modality is still far from clinical reality due to safety concerns. In contrast, use of an oral drug that is capable of stimulating fibrolysis and displays a reasonable clinical safety profile (38) might provide an attractive alternative for long term antifibrotic treatment with no or little adverse effects. Obviously, the main targets of the profibrolytic action of HAL in the liver are HSCs and myofibroblasts. Because expression of TIMP-1, which was down-regulated by HAL in cirrhotic liver in vivo, was not affected in HSCs in vitro, HAL might inhibit TIMP-1 expression by cells other than HSCs, such as hepatocytes and endothelial cells (39). The observed effects of HAL on HSCs proliferation, migration, and fibrosis-related gene expression in vitro were found in the same concentration range (10 –200 nM), suggesting a common mechanism for the antifibrotic mechanisms induced by HAL. Therefore, we investigated if and how far HAL modulates central signal transduction pathways, which are known to be of importance in HSC activation, especially those involving Erk1/2, p38 MAPK, PI3K/Akt, SAPK/JNK, and NF␬B (40 – 42). Using phosphorylation and NF␬B nuclear translocation assays we could identify two pathways modified by HAL treatment, namely activation of p38 MAPK and nuclear translocation of p65 NF␬B. Although the role of p38 and NF␬B in HSCs activation are controversial and far from clear (23, 43– 47), two prior studies suggested the potential antifibrogenic role of these pathways. Thus, inhibition of p38 increased HSC proliferation (46) and p38 MAPK mediated down-regulation of procol-

JUNE 2, 2006 • VOLUME 281 • NUMBER 22

lagen I induced by tumor necrosis factor-␣ (48), while transient transfection of HSCs using an NF␬B p65 expression plasmid together with a collagen reporter gene demonstrated a strong inhibitory effect on transcription of the procollagen ␣1(I) gene promoter (43). In general, up to now there existed no evidence of MMP induction driven by these pathways in HSCs, whereas several studies performed in fibroblasts showed that activated p38 and NF␬B are involved in MMP-1, -3, and -13 induction (49 –51). Moreover, NF␬B was demonstrated as a downstream effector of p38 MAPK when fibroblasts were induced to express MMP-1 in three-dimensional collagen lattices (52). Interestingly, it was shown that MMP-13 expression stimulated by interleukin-1 can be abolished by p38 MAPK inhibition in the hepatic myofibroblast cell line MG2 (53), and a recent study demonstrated that in the rat HSC line CFSC-2G procollagen ␣1(I) and MMP-13 expression were modulated reciprocally by tumor necrosis factor-␣ and transforming growth factor-␤1 (22). These observations are in accordance with our results that showed a clearly antifibrogenic and profibrolytic modulation of HSC proliferation, migration, and gene expression by HAL. Thus we were able to inhibit HAL-dependent MMP-13 induction by specific inhibitors of p38 and NF␬B, but not of Erk1/2 or PI3K, confirming that the mode of action of HAL in HSCs is indeed dependent on these two signaling pathways. Although we did not perform experiments investigating the hierarchy of HAL-dependent activation of p38 and NF␬B, it seems very likely that p38 may also be involved in HSC proliferation and migration inhibition induced by HAL, whereas NF␬B is acting further downstream of p38 similar to fibroblasts. So far, little is known about the putative intra- or extracellular receptors or activities of HAL. Their characterization is important, because to our knowledge HAL is the first proven agent with a strong fibrolytic effect documented in vivo and in vitro. Obviously, its mode of action is different from that of previously described plant-derived potential antifibrotics, which predominantly act via their antioxidative properties. Such antioxidants are often criticized as to their true antifibrotic effect, because they have usually been tested in fibrosis models that depend on free radical formation (e.g. fibrosis due to chronic CCl4 administration), which often cannot be reproduced in models with no or less oxidative

JOURNAL OF BIOLOGICAL CHEMISTRY

15097

Halofuginone Induces MMPs in Stellate Cells stress, such as rat secondary biliary cirrhosis (54) or in a model of fibrosis reversion, as was used in the present study. Importantly, HAL was active in vitro at very low concentrations, between 10 and 200 nM, indicating that together with emerging in vivo safety data (38) the substance could be given to patients at a very low dose. Thus, phase I clinical trials may become feasible in the near future. In conclusion, our data show that HAL induces a clear profibrolytic phenotype of HSCs and myofibroblasts at submicromolar concentrations, leading to significant fibrolysis in cirrhotic rat liver. In addition, we identified p38 MAPK and NF␬B as important antifibrogenic signal transducers activated by HAL that induce collagenolytic activity in HSCs. Our findings support that HAL is a promising agent for the treatment of liver fibrosis, the antifibrotic effect of which could be potentiated further in combination with other drugs that predominantly target profibrogenic pathways. Acknowledgments—We thank Prof. V. U. Buko (Institute of Biochemistry, Grodno, Belarus) for his continuous support, Prof. G. Niedobitek (Institute of Pathology, Erlangen, Germany) for kind help in histological assessment of the samples and access to confocal microscope, and Dr. C. Lohwasser for valuable suggestions in developing MAPK assays.

REFERENCES 1. McCrudden, R., and Iredale, J. P. (2000) Histol. Histopathol. 15, 1159 –1168 2. Kossakowska, A. E., Edwards, D. R., Lee, S. S., Urbanski, L. S., Stabbler, A. L., Zhang, C. L., Phillips, B. W., Zhang, Y., and Urbanski, S. J. (1998) Am. J. Pathol. 153, 1895–1902 3. Dufour, J. F., DeLellis, R., and Kaplan, M. M. (1997) Ann. Intern. Med. 127, 981–985 4. Shiffman, M. L., Hofmann, C. M., Contos, M. J., Luketic, V. A., Sanyal, A. J., Sterling, R. K., Ferreira-Gonzalez, A., Mills, A. S., and Garret, C. (1999) Gastroenterology 117, 1164 –1172 5. Hammel, P., Couvelard, A., O’Toole, D., Ratouis, A., Sauvanet, A., Flejou, J. F., Degott, C., Belghiti, J., Bernades, P., Valla, D., Ruszniewski, P., and Levy, P. (2001) N. Engl. J. Med. 344, 418 – 423 6. Dienstag, J. L., Goldin, R. D., Heathcote, E. J., Hann, H. W., Woessner, M., Stephenson, S. L., Gardner, S., Gray, D. F., and Schiff, E. R. (2003) Gastroenterology 124, 105–117 7. Reeves, H. L., and Friedman, S. L. (2002) Front. Biosci. 7, d808 – d826 8. Knittel, T., Kobold, D., Piscaglia, F., Saile, B., Neubauer, K., Mehde, M., Timpl, R., and Ramadori, G. (1999) Histochem. Cell Biol. 112, 387– 401 9. Cassiman, D., Libbrecht, L., Desmet, V., Denef, C., and Roskams, T. (2002) J. Hepatol. 36, 200 –209 10. Granot, I., Bartov, I., Plavnik, I., Wax, E., Hurwitz, S., and Pines, M. (1991) Poult. Sci. 70, 1559 –1563 11. Pines, M., and Nagler, A. (1998) Gen. Pharmacol. 30, 445– 450 12. Granot, I., Halevy, O., Hurwitz, S., and Pines, M. (1993) Biochim. Biophys. Acta 1156, 107–112 13. Halevy, O., Nagler, A., Levi-Schaffer, F., Genina, O., and Pines, M. (1996) Biochem. Pharmacol. 52, 1057–1063 14. McGaha, T. L., Kodera, T., Spiera, H., Stan, A. C., Pines, M., and Bona, C. A. (2002) Arthritis Rheum. 46, 2748 –2761 15. Nagler, A., Miao, H. Q. A., H., Pines, M., Genina, O., and Vlodavsky, I. (1997) Arterioscler. Thromb. Vasc. Biol. 17, 194 –202 16. Bruck, R., Genina, O., Aeed, H., Alexiev, R., Nagler, A., Avni, Y., and Pines, M. (2001) Hepatology 33, 379 –386 17. Greenwel, P., Rubin, J., Schwartz, M., Hertzberg, E. L., and Rojkind, M. (1993) Lab. Invest. 69, 210 –216 18. Niki, T., Rombouts, K., De Bleser, P., De Smet, K., Rogiers, V., Schuppan, D., Yoshida, M., Gabbiani, G., and Geerts, A. (1999) Hepatology 29, 858 – 867 19. Popov, Y., Patsenker, E., Fickert, P., Trauner, M., and Schuppan, D. (2005) J. Hepatol. 43, 1045–1054 20. Benten, D., Kumaran, V., Joseph, B., Schattenberg, J., Popov, Y., Schuppan, D., and Gupta, S. (2005) Hepatology 42, 1072–1081 21. Schulze-Krebs, A., Preimel, D., Popov, Y., Bartenschlager, R., Lohmann, V., Pinzani,

15098 JOURNAL OF BIOLOGICAL CHEMISTRY

M., and Schuppan, D. (2005) Gastroenterology 129, 246 –258 22. Schaefer, B., Rivas-Estilla, A. M., Meraz-Cruz, N., Reyes-Romero, M. A., HernandezNazara, Z. H., Dominguez-Rosales, J. A., Schuppan, D., Greenwel, P., and Rojkind, M. (2003) Am. J. Pathol. 162, 1771–1780 23. Hellerbrand, C., Jobin, C., Iimuro, Y., Licato, L., Sartor, R. B., and Brenner, D. A. (1998) Hepatology 27, 1285–1295 24. Cho, J. J., Hocher, B., Herbst, H., Jia, J. D., Ruehl, M., Hahn, E. G., Riecken, E. O., and Schuppan, D. (2000) Gastroenterology 118, 1169 –1178 25. Elkin, M., Reich, R., Nagler, A., Aingorn, E., Pines, M., de-Groot, N., Hochberg, A., and Vlodavsky, I. (1999) Clin. Cancer Res. 5, 1982–1988 26. Watanabe, T., Niioka, M., Hozawa, S., Kameyama, K., Hayashi, T., Arai, M., Ishikawa, A., Maruyama, K., and Okazaki, I. (2000) J. Hepatol. 33, 224 –235 27. Lee, H. S., Huang, G. T., Miau, L. H., Chiou, L. L., Chen, C. H., and Sheu, J. C. (2001) Hepatogastroenterology 48, 1114 –1117 28. Knittel, T., Mehde, M., Grundmann, A., Saile, B., Scharf, J. G., and Ramadori, G. (2000) Histochem. Cell Biol. 113, 443– 453 29. Benyon, R. C., and Arthur, M. J. (2001) Semin. Liver Dis. 21, 373–384 30. Friedman, S. L., Roll, F. J., Boyles, J., Arenson, D. M., and Bissell, D. M. (1989) J. Biol. Chem. 264, 10756 –10762 31. Gaca, M. D., Zhou, X., Issa, R., Kiriella, K., Iredale, J. P., and Benyon, R. C. (2003) Matrix Biol. 22, 229 –239 32. Benyon, R. C., Hovell, C. J., Da Gaca, M., Jones, E. H., Iredale, J. P., and Arthur, M. J. (1999) Hepatology 30, 977–986 33. Olaso, E., Ikeda, K., Eng, F. J., Xu, L., Wang, L. H., Lin, H. C., and Friedman, S. L. (2001) J. Clin. Invest. 108, 1369 –1378 34. Milani, S., Herbst, H., Schuppan, D., Grappone, C., Pellegrini, G., Pinzani, M., Casini, A., Calabro, A., Ciancio, G., and Stefanini, F. (1994) Am. J. Pathol. 144, 528 –537 35. Takahara, T., Furui, K., Yata, Y., Jin, B., Zhang, L. P., Nambu, S., Sato, H., Seiki, M., and Watanabe, A. (1997) Hepatology 26, 1521–1529 36. Iimuro, Y., Nishio, T., Morimoto, T., Nitta, T., Stefanovic, B., Choi, S. K., Brenner, D. A., and Yamaoka, Y. (2003) Gastroenterology 124, 445– 458 37. Siller-Lopez, F., Sandoval, A., Salgado, S., Salazar, A., Bueno, M., Garcia, J., Vera, J., Galvez, J., Hernandez, I., Ramos, M., Aguilar-Cordova, E., and Armendariz-Borunda, J. (2004) Gastroenterology 126, 1122–1133; discussion 1949 38. Pines, M., Snyder, D., Yarkoni, S., and Nagler, A. (2003) Biol. Blood Marrow Transplant 9, 417– 425 39. Roeb, E., Purucker, E., Breuer, B., Nguyen, H., Heinrich, P. C., Rose-John, S., and Matern, S. (1997) J. Hepatol. 27, 535–544 40. Hellerbrand, C., Jobin, C., Licato, L. L., Sartor, R. B., and Brenner, D. A. (1998) Am. J. Physiol. 275, G269 –G278 41. Reif, S., Lang, A., Lindquist, J. N., Yata, Y., Gabele, E., Scanga, A., Brenner, D. A., and Rippe, R. A. (2003) J. Biol. Chem. 278, 8083– 8090 42. Pinzani, M., Marra, F., and Carloni, V. (1998) Liver 18, 2–13 43. Rippe, R. A., Schrum, L. W., Stefanovic, B., Solis-Herruzo, J. A., and Brenner, D. A. (1999) DNA Cell Biol. 18, 751–761 44. Elsharkawy, A. M., Wright, M. C., Hay, R. T., Arthur, M. J., Hughes, T., Bahr, M. J., Degitz, K., and Mann, D. A. (1999) Hepatology 30, 761–769 45. Reeves, H. L., Dack, C. L., Peak, M., Burt, A. D., and Day, C. P. (2000) J. Hepatol. 32, 465– 472 46. Schnabl, B., Bradham, C. A., Bennett, B. L., Manning, A. M., Stefanovic, B., and Brenner, D. A. (2001) Hepatology 34, 953–963 47. Furukawa, F., Matsuzaki, K., Mori, S., Tahashi, Y., Yoshida, K., Sugano, Y., Yamagata, H., Matsushita, M., Seki, T., Inagaki, Y., Nishizawa, M., Fujisawa, J., and Inoue, K. (2003) Hepatology 38, 879 – 889 48. Varela-Rey, M., Montiel-Duarte, C., Oses-Prieto, J. A., Lopez-Zabalza, M. J., Jaffrezou, J. P., Rojkind, M., and Iraburu, M. J. (2002) FEBS Lett. 528, 133–138 49. Reunanen, N., Li, S. P., Ahonen, M., Foschi, M., Han, J., and Kahari, V. M. (2002) J. Biol. Chem. 277, 32360 –32368 50. Westermarck, J., Holmstrom, T., Ahonen, M., Eriksson, J. E., and Kahari, V. M. (1998) Matrix Biol. 17, 547–557 51. Ravanti, L., Heino, J., Lopez-Otin, C., and Kahari, V. M. (1999) J. Biol. Chem. 274, 2446 –2455 52. Xu, J., Clark, R. A., and Parks, W. C. (2001) Biochem. J. 355, 437– 447 53. Lee, H. S., Miau, L. H., Chen, C. H., Chiou, L. L., Huang, G. T., Yang, P. M., and Sheu, J. C. (2003) J. Biomed. Sci. 10, 757–765 54. Schuppan, D., Popov, Y., Patsenker, E., and Bauer, M. (2003) in Disease Progression and Carcinogenesis in the Gastrointestinal Tract (Galle, P., Gerken, G., Schmidt, W. E., Wiedenmann, B., eds) pp. 169 –79, Falk Symposium 132, Kluwer, Dordrecht, Netherlands

VOLUME 281 • NUMBER 22 • JUNE 2, 2006