Front-Cell-specific Expression of Membrane-Type 1 Matrix Metalloproteinase and Gelatinase A during Cohort Migration of Colon Carcinoma Cells Induced by Hepatocyte Growth Factor/Scatter Factor

Reagents.

Recombinant HGF/SF was purchased from Toyobo (Osaka, Japan), and TIMP-1 and -2 were from Fuji Chemical Industries (Takaoka, Japan).

Cell Cultures.

A highly metastatic subline to the liver (L-10) of the human colon adenocarcinoma cell line RCM-1 obtained by in vivo selection in nude mice was maintained in GM (for L-10), a 1:1 mixture of RPMI 1640 and Ham’s F-12 (Nissui Seiyaku, Tokyo, Japan) supplemented with 5% FCS (12) . Another human colon adenocarcinoma cell line, SW837, obtained from Dainihon Seiyaku (Osaka, Japan), was maintained in DMEM supplemented with 10% FCS (GM for SW837).

Cell Motility on a Lab-Tek Chamber Slide.

Motility assay was run as described (13) , with some modifications. Briefly, L-10 or SW837 cells were seeded into compartments of an 8-well Lab-Tek tissue culture chamber slide (Nunc, Naperville, IL; 1.2 × 10⁵ cells in 0.4 ml of GM) and allowed to attach for 40 h (1 day and overnight) in standard culture conditions. The cells formed interlinked, piled-up cell islands on the tissue culture glass substrate of the Lab-Tek chamber slide. The cells were then exposed to test medium (serum-free medium or GM with or without 20 ng/ml HGF/SF) for various periods up to 24 h. Stimulated cells migrated outward from the cell islands into interisland spaces as coherent cell sheets one-cell thick. Migration was quantified by counting the migrated cells in 10 randomly selected high-power fields (×400) after the cells were fixed with 3.7% formaldehyde in PBS and stained with hematoxylin. All experiments were performed in duplicate and repeated three times. Means and SEs of the mean were calculated, and statistical differences were analyzed using Student’s t test for nonpaired samples.

In experiments on gelatin-coated substratum, the glass substrate of each compartment of 8-well Lab-Tek chamber slides was coated with 200μ l of 1 μg/ml gelatin for 48 h at 4°C (7) . Diffuse coating was confirmed by protein staining. Using these coated chamber slides, cell migration was determined as described above.

In experiments with BB-94 (British Biotech Pharmaceuticals, Oxford, United Kingdom), 10 or 50 μm BB-94 was included in the gelatin solution (1 μg/ml, 200 ml) and used to coat the glass substrate of chamber slides. Coating was done for 48 h at 4°C. Cell migration was induced as described above, except for a 1-h preincuation of cells with 10 or 50 μm BB-94 prior to HGF/SF addition and presence of BB-94 during the whole assay period.

In experiments with TIMP-1 and -2 and the GelA PEX, cells were pretreated with TIMPs (10 μg/ml) or PEX-2 (25 or 50 μg/ml) for 1 h, followed by a 24-h incubation with HGF/SF (20 ng/ml) in the presence of TIMPs or GelA PEX.

Expression and in Vitro Folding of the GelA PEX Protein.

Expression vector GelA PEX/pET was constructed by cloning a PCR product of GelA PEX into protein expression vector pET-3a (Stratagene, La Jolla, CA). The PCR product corresponded to bases 1001–1939 (14) , and start and stop codons were added at its 5′ and 3′ ends, respectively. The nucleotide sequences of the GelA PEX product were confirmed by the dideoxy chain-termination method before cloning. The vector GelA PEX/pET was transfected into an Escherichia coli BL21(DE3)pLyS host (Stratagene), and expression and in vitro folding of the GelA PEX protein were done according to Huang et al. (15) . The obtained GelA PEX protein was concentrated using Diaflo ultrafiltration membranes (YM5; Amicon, Beverly, MA) and Centriprep 10 (Amicon) and was stored at −20°C until use.

Scattering Assays.

L-10 and SW837 cells were added to Lab-Tek tissue culture chamber slides at 5.0 × 10³ cells/well in the presence of 50 μg/ml anti-E-cadherin mAb (HECD-1; Takara) and allowed to grow for 7 days, after which loosely aggregated cell islands were obtained. The cells of these loose islands showed scattering in response to treatment with HGF/SF (20 ng/ml) for 24 h.

Immunofluorescent Staining.

After cell motility assays were done as described above, immunofluorescent staining of cells was performed as described (7) using mouse mAb against human MT1-MMP (114-6G6; Fuji Chemical Industries), MMP-2 (75-7F7; Fuji Chemical Industries), E-cadherin or vinculin (VG61110; Affinity Research Products, Nottingham, United Kingdom) as primary antibodies. For E-cadherin immunostaining, all of the solutions contained 1 mm CaCl₂.

Biotin Labeling of Coated Gelatin (Double Staining with Phalloidin).

Gelatin-coating of Lab-Tek chamber slide substratum was done as described above. After rinsing with cold PBS twice, the gelatin was biotinylated by incubating in PBS containing water-soluble sulfo-N-hydroxysuccinimide-biotin (2 mg/ml; Pierce Chemical Co., Rockford, IL) for 2 h at 4°C, followed by washing with PBS three times and then rinsing in PBS overnight. L-10 cells (1.2 × 10⁵ cells/well) were added and allowed to attach for 40 h, and migration assays were run as described above. After the cells were fixed with 2% paraformaldehyde/0.1 m PB for 30 min at 4°C and permeabilized with 0.1% Triton X-100/PBS for another 30 min at 4°C, the biotinylated gelatin was visualized with FITC-conjugated streptavidin (Sigma Chemical Co.) for 2 h at room temperature. After rinsing with PBS four times, the cells were stained with rhodamine-phalloidin in PBS (Molecular Probes, Eugene, OR) for 20 min at room temperature.

To obtain footprints of migrating cell sheets, after cohort migration was induced on gelatin-coated substratum, cells were treated with 0.5 mm EDTA/PBS, followed by tapping of chamber slides by fingers. About 30% of cell islands with migrating cell sheets were removed by the treatment, and the remaining footprints were demonstrated by visualization of biotinylated gelatin matrix, as described above.

RT-PCR.

mRNA expression for MT1-MMP and MMP-2 in L-10 cells was analyzed using RT-PCR, as described (8) . Annealing was done at 60°C for MT1-MMP and 58°C for MMP-2. The sequences of the forward and reverse primers for MMP-2 are 5′-ACCTGGATGCCGTCGTGGAC-3′ (1800–1819) and 5′-TGTGGCAGCACCAGGGCAGC-3′ (2228–2247), respectively, and the PCR product size is 447 bp (14) . The sequences of the forward and reverse primers for MT1-MMP are 5′-CCCTATGCCTACATCCGTGA-3′ and 5′-TCCATCCATCACTTGGTTAT-3′, respectively, and are complementary to bases 598–617 and 1129–1148 (3) . The PCR product size is 550 bp.

Preparation of CM and Zymography.

After L-10 cells (5.0 × 10⁵ cells in 2 ml of GM) were added and cultured in 35-mm tissue culture dishes (Becton Dickinson, Lincoln Park, NJ) for 24 h, the cells were rinsed in serum-free media three times, then incubated in 1.0 ml of serum-free GM with or without HGF/SF (20 ng/ml) for another 24 h. CM harvested from these cultures was centrifuged twice (110 × g, 5 min; and 4000 × g, 20 min) to remove cellular debris, concentrated by 8-fold using centricon 10 (Amicon) and subjected to zymography using gelatin as a substrate, as described previously (16) .

Immunoblotting.

After L-10 cells were treated with HGF/SF (20 ng/ml) for 24 h, they were then lysed by boiling for 5 min in 2% SDS, 0.1 m dithiothreithol/1 m Tris-HCl (pH 6.8), followed by centrifugation at 11,000 × g for 5 min. These cell extracts and CM obtained as above were subjected to immunoblotting, as described (8) , using mouse mAb to human MMP-2 (75-7F7; Fuji Chemical Industries) or MT1-MMP (114-1F2; Fuji Chemical Industries) as primary antibodies.

Expression of MT1-MMP and MMP-2 in L-10 Cells.

We first examined whether MT1-MMP and MMP-2 were expressed in L-10 cells and also whether HGF/SF, which can induce cohort migration of L-10 cells, could affect the expressions. Analyses by RT-PCR and immunoblotting demonstrated the expression of MT1-MMP and MMP-2 in L-10 cells at mRNA and protein levels, respectively (Fig. 1A and B) ⇓ . On immunoblots, stimulation of MMP-2 production with HGF/SF treatment (20 ng/ml) was shown, whereas the MT1-MMP expression was not altered. Zymograpy of L-10 cell CM revealed the presence of both pro- (M_r 68,000) and active (M_r 62,000) forms of MMP-2 (Fig. 1C) ⇓ . The amounts of samples loaded for zymography were the same as those for immunoblotting, but gel lysis in zymography was so sensitive that a less amount of pro- and active MMP- 2 in L-10 cell CM without HGF/SF treatment could cause lysis of gelatin substrate like that by CM with HGF/SF treatment.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Expression of MT1-MMP and MMP-2 in L-10 cells. After cells were treated with or without HGF/SF (20 ng/ml), total RNA extracted from the cells was subjected to RT-PCR using primer pairs for MT1-MMP and MMP-2, which form 550-bp and 447-bp products, respectively (A). Total cell extracts were subjected to immunoblotting with anti-MT1-MMP mAb (B; MT1-MMP). The M_r 60,000 band corresponds to the active form of MT1-MMP, and the M_r 50,000 band corresponds to its processed form. Their conditioned media were subjected to immunoblot with anti-MMP-2 mAb (B; MMP-2) and gelatin zymography (C). Bands of M_r 68,000, M_r 62,000, and M_r 51,000 correspond to pro-, active, and further processed forms of MMP-2, respectively.

Immunolocalization of MT1-MMP and MMP-2 during Cohort Migration.

By using immunofluorescent staining techniques, we then investigated the MMP expression patterns during cohort migration of L-10 cells. Without HGF/SF treatment, L-10 cells formed piled-up cell islands with few cells migrated, and MT1-MMP immunostaining was detected as summation of that of each cells in the piled-up cell islands, with some accentuation along the cell borders (Fig. 2A ⇓ , a). When stimulated with HGF/SF (20 ng/ml, 24 h), L-10 cells migrated outward from the cell islands as coherent cell sheets one-cell thick (Fig. 2A ⇓ , c and e), as described previously (14) . In the migrating cell sheets, MT1-MMP was demonstrated only at their fronts, including the frontmost first to second rows of cells, with the following migrating cells being negative (Fig. 2A ⇓ , d and f). At a higher magnification, expression of MT1-MMP was observed predominantly in the leading edges of the frontmost cells, accompanied by some staining in perinuclear regions (Fig. 2A ⇓ , b). Negative control with nonimmune mouse IgG did not show any immunofluorescent staining (data not shown). MMP-2 was also demonstrated specifically in the frontmost cells of migrating cell sheets, especially at their leading edges (Fig. 2A ⇓ , g and h).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

A, immunolocalization of MT1-MMP and MMP-2 in L-10 cells showing cohort migration. c, e, g, and i, Normarski differential interference contrast images (NI). a, b, d, and f, immunofluorescent staining (IF) with anti-MT1-MMP mAb. h, IF with anti-MMP-2 mAb. j, IF with anti-E-cadherin mAb. Without HGF/SF treatment, L-10 cells form piled-up cell islands, where MT1-MMP immunofluorescent images show summation of membrane staining (a). When cohort migration was induced with HGF/SF treatment (20 ng/ml, 24 h; b–j), only the frontmost cells stain positive for MT1-MMP (b, d, and f) and MMP-2 (h), whereas the following cells in the migrating cell sheets are negative (∗). d and h, the remaining piled-up cell islands are indicated by arrows. b, distribution of MT1-MMP in the leading edges of the frontmost cells is shown at a high magnification. Nucleus boundaries are shown as broken lines. j, as control to demonstrate accessibility of antibodies to the inside cells within the migrating cell sheets, IF with anti-E-cadherin mAb is shown. B, expression of MT1-MMP and MMP-2 in scattering L-10 cells. Scattering of cells was induced with HGF/SF treatment in the presence of anti-E-cadherin antibody, as described in “Materials and Methods.” a and c, NI; b, IF with anti-MT1-MMP mAb; d, IF with anti-MMP-2 mAb. C, immunolocalization of MT1-MMP and MMP-2 in SW837 colon adenocarcinoma cells showing either cohort migration (a–d) or scattering (e–h). a, c, e, and g, NI; b and f, IF with anti-MT1-MMP mAb; d and h, IF with anti-MMP-2 mAb. a, nucleus boundaries are shown as circles. a and c, the position of the front line of the leading edges is indicated as a light green line. The cells were photographed at a magnification of ×50 (A, c and d), ×100 (A, a, and e–j; B, a–d; C, a–h), or ×250 (A, b).

To rule out the possibility that mAbs could not reach to the inside cells of the migrating cell sheets, we used other mAbs that should react with both front and inside (following) cells. E-cadherin was shown along the cell borders, even inside the migrating cell sheets (Fig. 2A ⇓ , i and j). Vinculin also successfully demonstrated focal adhesion plaques in both front and following cells of the migrating cell sheets (data not shown). These results clearly showed that mAbs can reach to not only frontmost cells but also following cells of the migrating cell sheets.

Expression of MT1-MMP and MMP-2 in Scattering L-10 Cells.

As another control, we induced scattering of L-10 cells by stimulating cell migration in the presence of anti-E-cadherin antibody and examined how MT1-MMP and MMP-2 were expressed in the scattering cells. Individual scattering L-10 cells expressed both MT1-MMP and MMP-2 in their leading edges and also within the cytoplasm, especially perinuclear regions (Fig. 2B ⇓ , a–d). The front cell-specific pattern of expression observed during cohort migration was lost.

Expression of MT1-MMP and MMP-2 in SW837 Cells.

To examine whether front cell-specific expression of MT1-MMP and MMP-2 is unique only to L-10 cells or can be seen in other cell lines, another human colon carcinoma cell line (SW837) was used for an immunolocalization study. This cell line also shows cohort migration in response to HGF/SF treatment (8) , and its expression of MT1-MMP and MMP-2 was demonstrated by RT-PCR (data not shown). When cohort migration was induced with HGF/SF treatment, MT1-MMP (Fig. 2C ⇓ , a and b) and MMP-2 (Fig. 2C ⇓ , c and d) were demonstrated predominantly in the frontmost cells or sometimes in the first and second rows of cells at the front of the migrating cell sheets. In contrast, when scattering was induced by treament with HGF/SF in the presence of anti-E-cadherin antibody, all scattering SW837 cells expressed MT1-MMP (Fig. 2C ⇓ , e and f) and MMP-2 (Fig. 2C ⇓ , g and h), without preferential localization in the front cells.

Degradation of Gelatin Matrix by Migrating Cells.

Using gelatin as a substrate, we studied whether MT1-MMP and MMP-2 expressed at the front of the migrating cell sheets cause ECM degradation. Gelatin labeled with biotin was visualized with avidin-conjugated FITC, and, at the same time, cytoplasmic actin filaments were demonstrated with rhodamine-phalloidin. At the sites of leading edges of the front cells of migrating cell sheets, limited gelatinolysis and consequent partial removal of gelatins occurred, causing radially arrayed lysis of gelatin matrix at the outermost portions (Fig. 3, a–d ⇓ ). Thin, also radially arrayed gelatin matrix was left in between the lysed portions. At the base of the radially arrayed gelatinolysis, an arcuate band-like gelatin matrix was also left beneath the leading edges.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Organized degradation of coated gelatin substrate by migrating L-10 cells. Cohort migration was induced with HGF/SF treatment (20 ng/ml, 20 h) on FITC-labeled gelatin matrix. a and c, rhodamine-phalloidin immunofluorescent images. b, d, and e, immunofluorescent images of FITC-labeled gelatin. b and d, dark areas caused by lysis of FITC-labeled gelatin is seen at the sites corresponding to leading edges (∗). The lysis of gelatin is caused in an organized manner: radially arrayed gelatinolysis (short arrows) is observed at the frontmost portion of the leading edges, with arcuate band-like gelatin matrix (long arrows) being left beneath the main part of the leading edges. Gelatinolysis caused by migrating cells was also demonstrated by removing cells with EDTA treatment and tapping the chamber slides with fingers after cohort migration had been induced, which we called footprints of cells (e). Radially arrayed gelatinolysis (arrows) is also shown at the frontmost portions. The cells were photographed at a magnification of ×100 (a–e).

We also tried to demonstrate migrating cell-induced gelatinolysis by removing cells with EDTA treatment and tapping the chamber slides with fingers after cohort migration had been induced, and we called these lysed portions left behind footprints of cells. The footprints also showed a similar organized way of degradation of gelatin matrix at the peripheral portions corresponding to leading edges of the frontmost migrating cells, as radially arrayed lytic marks (Fig. 3e) ⇓ .

An Essential Role of Gelatin Degradation in Cohort Migration.

We finally investigated whether the above organized degradation of gelatin matrix was necessary for cohort migration of carcinoma cells. BB-94, a synthetic inhibitor for MMPs, inhibited HGF/SF-induced cohort migration of L-10 cells in a dose-dependent manner: approximately 34% and 78% inhibition at 10 μm and 50 μm, respectively (Fig. 4) ⇓ . Both TIMP-1 and TIMP-2 also inhibited the migration, although TIMP-2 seemed to be more effective than TIMP-1: approximately 72% inhibition by TIMP-1 and 82% inhibition by TIMP-2. Furthermore, GelA PEX effectively inhibited the migration: approximately 60% inhibition at 25 μg/ml and 74% inhibition at 50 mg/ml. Taken together, these results indicate that MMP-2 activated on the cell surface by MT1-MMP plays a major role in cohort migration on gelatin matrix.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Effect of BB-94, TIMP-1 and -2, and GelA PEX on HGF/SF-induced cohort migration on gelatin-coated substratum. Cells were treated with BB-94 (10 or 50 μm), TIMP-1 or -2 (10μ g/ml), or GelA PEX (25 or 50 μg/ml) for 1 h, and then cohort migration was induced with HGF/SF treatment (20 ng/ml, 24 h) on gelatin-coated substratum in the presence of the above inhibitors. The control value was 53.9 ± 1.4 cells/high-power field. The values are means and SEs of the mean (n = 3). ∗, statistically significant compared with control, P < 0.01.