Oxidative stress stimulates proliferation and invasiveness of hepatic stellate cells via a MMP2-mediated mechanism^†

Materials.

Culture media, trypsin, all restriction endonucleases, and DNA-modifying enzymes were obtained from GIBCO (Grand Island, NY). Stractan was obtained from Life Technologies (Milan, Italy). SB203580, PD98059, wortmannin, rapamycin, and MMP inhibitors (GM6001 and MMP-2 inhibitor I, OA-Hy) were obtained from Calbiochem (La Jolla, CA). Nitrocellulose membranes (Hybond) were obtained from Amersham (Milan, Italy). Pronase was obtained from Boehringer Mannheim (Monza, Italy). Antibodies against phosphorylated and diphosphorylated extracellular signal-regulated kinase (ERK1/2) and against 70-kd S6 kinase (p70^S6K) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies against MT1-MMP were obtained from Chemicon (Temecula, CA). Radioactive material was purchased from New England Nuclear (Boston, MA). All other reagents were obtained from Sigma Chemical Co. (Milan, Italy).

Isolation and Culture of HSC.

HSC were obtained from wedge sections of normal human liver that was unsuitable for transplantation after obtaining the approval of the appropriate local ethics committee. After combined 0.5% pronase/0.05% collagenase tissue digestion, human HSC were isolated by ultracentrifugation over four gradients (1.111/1.080/1.058/1.053) of stractan. They were characterized as previously described.19 The purity of HSC, estimated by autoflorescence of cells using UV-excited fluorescence microscopy, was beetwen 90% and 95%. Cells were cultured in Iscove's modified Dulbecco's Modified Eagle Medium supplemented with 20% fetal bovine serum, 2 mmol/L glutamine, 0.1 mmol/L nonessential amino acids, 1 mmol/L sodium pyruvate, 0.6 U/mL insulin, and 1% antibiotic-antifungal solution. Experiments were undertaken in triplicate using activated alpha smooth muscle–positive cells obtained between the first and third serial passage of cell lines obtained from five different patients.

Assay of DNA Synthesis and Cell Growth.

Cellular proliferation was evaluated by measuring DNA. The amount of [methyl-³H]thymidine ([³H]TdR) incorporated into trichloroacetic acid–precipitated material was determined, and cells were counted as previously described.19 Briefly, passage-activated HSC were seeded into 60-cm² flasks until the monolayers were 75% to 80% confluent. Cell cultures were rendered quiescent by incubation in serum-free and insulin-free (SFIF) Iscove's medium for 48 hours and were next incubated with a superoxide anion donor [xanthine (500 μmol/L) plus xanthine oxidase (20-2,000 μU/mL) (X/XO)] for 20 hours; the cells were then pulsed for 4 hours with 1.0 μCi/mL [³H]TdR (6.7 Ci/mmol). Cells were disolved in 0.2 N NaOH. Their radioactivity was then measured in a scintillation counter. Six determinations were made to generate each experimental value.

Counting of hematoxylin and eosin–stained cells was undertaken after 48 hours of incubation with X/XO; a computerized video image analysis system (Leica Quantimet Q500MC; Leica Cambridge Ltd., Cambridge, UK) was used. Experiments were undertaken in triplicate using three different cell preparations. Production of superoxide anion was measured as SOD-inhibitable reduction of cytochrome c.6 Some experiments were undertaken in the presence of antioxidants (superoxide dismutase polyethylene glycol 250 U/mL or vitamin E 100 μmol/L), and MMP inhibitors (GM6001, 15 μmol/L or OA-Hy, 100 μmol/L) or kinase inhibitors (SB203580 10 μmol/L, PD98059 30 μmol/L, wortmannin 100 nm/L, rapamycin 5 ng/mL). Preliminary experiments indicated that cellular toxicity of the different treatments used could be disregarded for the following reasons: (1) lack of Typtan blue staining of the cells in the presence of relevant concentrations of test substances during the experimental period; (2) absence of lactate dehydrogenase leakage from HSC into the culture medium using a cytotoxicity detection kit (Boehringer Mannhein, Mannhein, Germany).

Invasion Assay.

The ability of cells to invade through a Matrigel-coated filter was measured in Boyden chambers. Polyvinylpyrrolydone-free polycarbonate filters (pore size 8 μm) were coated with basement membrane Matrigel (200 μg/filter) (Collaborative Research Inc., Bedford, MA), as described in the standard protocol.20 Confluent cells were starved of serum for 48 hours. After washing and trypsinization, they were placed in the upper chambers at a density of 1 × 10⁵ in serum-free medium with or without X/XO. Dulbecco's Modified Eagle Medium containing 0.1% fetal bovine serum was placed in the lower compartments of the Boyder chambers. Chambers were incubated for 12 hours in 5% CO₂ at 37° C. At the end of the incubation, the cells on the upper surface were completely removed by wiping with a cotton swab. Filters were fixed in methanol and stained with hematoxylin and eosin. Cells from various areas of the lower surface were counted using a computerized video image analyzing system (Leica Quantimet Q500MC). Each assay was undertaken in triplicate.

RNA Extraction and RNase Protection Assay.

Total RNA was extracted from cultured human HSC using the guanidinium-phenol-chloroform method. Cellular levels of MMP-1, MMP-2, TIMP-1, TIMP-2, and MT1-MMP messenger RNA (mRNA) were determined using a previously described ribonuclease (RNase) protection assay.21

“Anti-sense” (complementary to mRNA) RNA probes were obtained by run-off transcription of pGEM1 (Promega, Madison, WI) plasmids that contained the following complementary DNA fragments: the 730-bp SstI-EcoRI complementary DNA fragment of pCllase1, which codes for human MMP-1 (ATCC 57685) and the 1,300-bp EcoRI-BglII fragment of pK121 (kindly provided by Dr. K. Tryggvason), which codes for human gelatinase A (MMP-2). For TIMP-1, TIMP-2, and MT-MMP, probes were generated using oligo(dT)-primed reverse transcription of total cellular RNA prepared from human placenta and HT144 melanoma cells as previously described.12, 21 A 110-bp complementary DNA fragment of the GAPDH gene, subcloned into the HindIII site of pGEM1, was used as a control. T7 or SP6 RNA polymerase (Promega, Milan, Italy) was used to run off the anti-sense [³²P]-labeled RNA probe. The newly synthesized RNA probes (specific activity: approx. 1 × 10⁹ cpm/μg) were hybridized overnight with RNA samples obtained from HSC. Protected fragments were separated by electrophoresis through a 7.0-mol/L urea/6% acrylamide gel and were then exposed to Kodak X-omat film at −80°C for 72 hours.

Gelatinase Activity Assay.

Aliquots (20 μg/protein/lane) of cell culture medium were electrophoresed in 7.5% SDS polyacrylamide gels containing 0.1% gelatin (Sigma Chemical Co., Milan, Italy). After electrophoresis, sodium dodecyl sulfate was removed by soaking the gels for 20 minutes each in buffers containing 50 mmol/L Tris (pH 7.6), 10 mmol/L CaCl, and 2.5% Triton X-100, followed by the same buffer containing 1% Triton. Digestion was allowed to occur at 37°C for 24 hours. Gels were then stained with Coomassie brilliant blue and destained until clear bands became evident.

ELISA.

Immunoreactive levels of MMP-2 in conditioned medium were measured using a Biotrak Human MMP-2 ELISA kit (Amersham Biosciences, Arlington Heights, IL), according to the manufacturer's instructions. MMP-2 levels could be measured in the range 1.5 to 24 ng/mL; the sensitivity of the assay was 0.37 ng/mL.

TIMP-2 levels in conditioned media were measured using a human TIMP-2 immunoassay kit (Chemicon International, Temecula, CA), according to the manufacturer's instructions. TIMP-2 could be measured in the range of 20 to 320 ng/mL; the sensitivity of the assay was 20 ng/mL. Experiments were undertaken in triplicate.

Western Blotting.

Cells were scraped and homogenized in ice-cold buffer (pH 7.4) that consisted of 50 mmol/L Tris (pH 7.4), 150 mmol/L KCl, 1% Triton X-100, 1 mmol/L ethylenediaminotetra-acetic acid, 5 mmol/L N-ethylmaleimide, and 0.2 mmol/L phenyl-methyl-sulfonyl-fluoride. Cell extracts (50 μg/lane) were separated via 12% gel electrophoresis and electroblotted onto nitrocellulose filters. Nonspecific binding sites were blocked by incubating nitrocellulose sheets for 1 hour in phosphate-buffered saline containing 5% low-fat dry milk. Blots were developed using an enhanced chemiluminescence detection system (ECL plus kit; Pharmacia Bioscience, Arlington Heights, IL).

Statistical Analysis.

Results are expressed as the mean ± SEM. Group means were compared via ANOVA followed by the Student-Newman-Keuls test, if the former was significant. A P value of less than .05 was considered significant.

Effect of X/XO on HSC proliferation and Invasiveness of Activated HSC.

To evaluate the effects of X/XO on HSC proliferation, cells were starved of serum and insulin for 48 hours and then incubated for an additional 24 hours with different concentrations of the superoxide anion donor. Application of the [³H]TdR incorporation assay indicated that X/XO significantly stimulated DNA synthesis in HSC (Fig. 1A). The increase in [³H]TdR incorporation was associated with a significant dose-dependent increase in the number of cells after incubation for 48 hours (Fig. 1B). The effect of oxidative stress on the ability of HSC to invade through reconstituted basement membranes was assessed using invasion chambers coated with Matrigel as already described. Exposure to X/XO induced a dose-dependent stimulation of the invasiveness of HSC (Fig. 1C). Concentrations of XO higher than 200 μU/mL did not induce any further significant increase of either proliferation or invasiveness.

X/XO Effects Are Associated With Increased MMP-2 Expression.

Gelatin zymography of culture media of unstimulated HSC revealed a gelatinolytic band of 72 kD that represented latent MMP-2 (Fig. 2A). After addition of X/XO to HSC and culture for 24 hours, MMP-2 levels were increased. In addition, less intense bands of 66 to 64 kD were present in zymograms of supernatants collected after 3 and 6 hours of incubation. These bands corresponded to partially or completely activated forms of MMP-2.

The increase of MMP-2 synthesis, suggested by gelatin zymography, was confirmed by a marked upregulation of MMP-2 mRNA expression 3 to 12 hours after X/XO treatment (Fig. 2B). In contrast, MMP-1 mRNA levels were not affected by X/XO treatment (data not shown). Furthermore, immunoreactive MMP-2, measured via ELISA, was markedly increased in the medium of cultured HSC that had been stimulated by X/XO for 24 hours, and the increase was dose-dependent (Fig. 2C).

Incubation with antioxidants, such as vitamin E and superoxide dismutase polyethylene glycol, abolished increased MMP-2 gene expression and secretion induced by X/XO (Fig. 3A–B) and impaired X/XO-induced proliferation and migration of activated HSC (Fig. 3C).

To confirm the role of MMP-2 in the regulation of proliferation and invasiveness of HSC, we added two chemically different MMP inhibitors. Both the broad spectrum MMP inhibitor, GM6001,22 and the specific MMP-2 inhibitor, 1 (OA-Hy),23 completely abrogated the effects of X/XO treatment on the growth and motility of HSC (Fig. 3D).

Effect of X/XO on Expression of TIMP-2 and MT1-MMP by HSC.

To examine the influence of oxidative stress on the expression of components of the MMP-2 activation complex, we assessed TIMP-2 and MT1-MMP expression in HSC treated with X/XO.

Figure 4A and C indicate a significant time-dependent upregulation of steady-state levels of TIMP-2 and MT1-MMP mRNA in X/XO-treated HSC. A similar induction of their protein products was found, as shown by the results of applying an ELISA assay for TIMP-2 and Western blotting for MT1-MMP. X/XO treatment induced a dose-dependent increase in TIMP-2 secretion (Fig. 4B) and the expression of both latent (65-kD) and activated (63-kD) forms of MT1-MMP; expression of the 63-kD form was greater than that of the 65-kD form (Fig. 4D). These findings suggest that oxidative stress can induce MT1-MMP activation that is required for the mediation of cell surface pro–MMP-2 activation. Pretreatment with vitamin E and superoxide dismutase polyethylene glycol abolished the increase of TIMP-2 and MT1-MMP mRNA levels induced by X/XO (Fig. 4E).

ERK1/2 and p70^S6K Activation Is Involved in X/XO-Mediated Induction of MMP-2.

To assess the mechanism of action of X/XO in the regulation of MMP-2 expression, we undertook experiments to assess the intracellular signaling involved. To assess the role of the PI3K and ERK1/2 signal transduction pathways in the regulation of MMP-2 expression,24, 25 we determined the effects of specific kinase inhibitors on the expression of MMP-2 induced by X/XO. MMP-2 mRNA expression (Fig. 5A) and protein secretion (Fig. 5B) induced by X/XO were blocked by wortmannin, a PI3K inhibitor, and by rapamycin, a mTOR/ p70^S6K inhibitor. Similarly, PD98058, a MEK inhibitor, significantly abrogated the effect of X/XO treatment on MMP-2 expression, whereas SB203580, a p38 inhibitor, had a negligible effect on this system.

It has been shown that PI3K and ERK, both of which can activate p70^S6K, play a key role in regulating the proliferation and migration of HSC via external stimuli.26-29 Western blot analysis was undertaken to evaluate the effects of X/XO on p70^S6K and ERK1/2 activation. p70^S6K activation is associated with increased phosphorylation of serine and threonine residues that results in reduced mobility of the protein on SDS-PAGE. The effects of X/XO on ERK1/2 phosphorylation were determined using antibodies that specifically recognize the active tyrosine-phosphorylated forms of ERK1 and ERK2.

X/XO induced p70^S6K phosphorylation; it caused an appreciable decrease in the mobility pattern of the 70-kD and 85-kD protein triplets relative to untreated controls (Fig. 6A, upper panel). This change in gel mobility pattern became evident 10 minutes after X/XO treatment; it was sustained at 30 minutes, but was lost by 60 minutes (Fig. 6B). Also, X/XO induced an increase in ERK1/2 phosphorylation after incubation for 10 minutes; the increase was less at 30 minutes and was no longer present at 60 minutes (Fig. 6A, middle panel; Fig. 6B). The equivalent loading of protein in each lane was confirmed by probing membranes with anti-ERK1/2 antibodies (Fig 6A, lower panel).

Effect of Intracellular Pathway Inhibitors on X/XO-Induced Proliferation and Invasiveness of HSC.

The signal transduction pathways involved in the regulation of MMP-2 expression induced by X/XO, induced us to evaluate the effects of different kinase inhibitors on the number of cells and chemoinvasion of cells after X/X0 treatment. Cellular proliferation and chemoinvasion of cells induced by X/XO were significantly reduced by addition of the MEK inhibitor, PD98059 (Fig. 7A–B). Similarly, the addition of either the PI3K inhibitor, wortmannin, or the mTOR/p70^S6K inhibitor, rapamycin, reduced X/XO-induced proliferation and invasiveness in HSC. No similar effect occurred when cells were treated with the p38 inhibitor, SB203580.

The effects of various kinase inhibitors on p70^S6K and ERK1/2 activation were also evaluated (see Fig. 7). It has been shown that p70^S6K can be a downstream component of PI3K pathways, which may also contribute to ERK1/2 activation in HSC.29 We tested the effect of the PI3K inhibitor, wortmannin, on p70^S6K and ERK1/2 phosphorylation in HSC incubated with X/XO. Wortmannin completely blocked X/XO-induced p70^S6K phosphorylation, but it did not affect ERK1/2 activation (Fig. 7C). We also studied the effect of the addition of the MEK inhibitor PD98059 on control and X/XO-treated HSC. The increase in p70^S6K and ERK1/2 phosphorylation induced by X/XO was completely abrogated by PD98059 (Fig. 7D). In contrast, rapamycin did not have any effect on ERK phosphorylation, but it completely abrogated phosphorylation of its specific substrate, p70^S6K (Fig. 7E).