High threshold of β1 integrin inhibition required to block collagen I-induced membrane type-1 matrix metalloproteinase (MT1-MMP) activation of matrix metalloproteinase 2 (MMP-2)

MCF-7-MT1 cells: Stable transfection of MT1-MMP into MCF-7 cells previously transfected with β-galactosidase was previously described [32],[33]. The pcDNA-β1 vector was constructed by cloning human β1 integrin cDNA into pcDNA3.1. A codon-swapped mutant form of the human β1 integrin (β1-Mt) was generated by the QuickChange site-directed mutagenesis kit (Invitrogen, Melbourne, Australia). Two synthetic oligonucleotide primers were designed to generate 7 point mutations in the β1/6 integrin siRNA recognition site: TGCTGATATGGAAACTACTTATGATTATACACGACAGAAGGGAGT and ACTCCCTTCTGTCGTGTAT AATCATAAGTAGTTTCCATATCAGCA.

Three β1-integrin siRNAs (Pro-Oligo, Australia) were found active for β1 integrin knockdown: β1-integrin siRNA#4 (β1/4), AATTAGCATAACTTCAAATAA, β1-integrin siRNA#5 (β1/5), AATGGCTTAATTTGTGGAGGA, and β1-integrin siRNA#6 (β1/6), AAGCTTTTAATG ATAATTCAT.

Antibodies against the following proteins were purchased from Chemicon (Boronia, Australia): integrin β1 (AB2510), αV (L320), β1 (MAB19732) α3 (MAB19522) and MT1-MMP (AB815). Antibody toward α2 integrin (Ak7) was a kind gift from Dr. Michael Berndt (Monash University, Melbourne) and a pan-actin antibody was from Biosource (Camarillo, CA). Con A, TPA (12-O-tetradecanoylphorbol-13-acetate), vitronectin (VN), fibronectin (FN) and laminin-1 (LM) were from Sigma (Castle Hill, Australia). Recombinant full-length pro-MMP-2 expressed in a vaccinia virus system (rMMP-2) was a kind gift from Dr. Rafael Fridman (Wayne State University, Detroit, MI). Col I (Vitrogen 100) was obtained from Cohesion (Palo Alto, CA). Alexa Fluor 568-conjugated goat anti-mouse secondary antibody and Alexa Fluor 488-conjugated donkey anti-rabbit secondary antibody were obtained from Molecular Probes (Eugene, OR).

MCF-7-MT1 and MDA-MB-231 cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM; Life Technologies, Auckland, New Zealand) with 10% fetal bovine serum (FBS; JRH Biosciences, Lenexa, KS). SiRNA transfections were performed using Lipofectamine™ 2000 (Invitrogen, Life Technologies). Plasmid transfections were performed using Fugene6 reagent (Roche, Mannheim, Germany) with 2 µg of plasmid DNA. For co-transfection of human β1 integrin and β1 integrin siRNAs, MCF-7-MT1 and MDA-MB-231 cells were transfected with β1/6-integrin siRNA using Lipofectamine™ 2000 as described above. The siRNA-transfected cells were subsequently transfected with either wild-type or mutant β1 integrin as indicated using Fugene6 with 2 µg plasmid DNA. Recombinant full-length pro-MMP-2 was added at 100 ng/ml. At the appropriate time point, the conditioned media and cell lysates were collected.

cDNA was prepared from 100 ng of total RNA isolated with the Qiagen RNeasy mini-column kit (Qiagen, GmbH, Germany) using Superscript II reverse transcriptase (Invitrogen) and random primers. Human ribosomal protein L32 mRNA was used as a housekeeping gene for normalization [34]. qRT-PCR was performed on an ABI Prism 5700 Sequence Detection System (Perkin-Elmer Applied Biosystems, Australia) with cDNA generated from an equivalent of 2 ng of RNA as previously described [24]. The primer sequences used were for L32: CAGGGTTCGTAGAAGATTCAAGGG and CTTGGAGGAAACATTGTGAGCGATC; β1 integrin: GCTGACCAGTCAGCAGCTTCATTT and CTCCAGAAGAAGCAGTAGCAGAGTTT; α2 integrin: GACCTATCCACTGCCACATGTGAAAAA and CCACAGAGGACCACATGTGAGAAAA; α3 integrin: CGCAGGTGGGCGCCTATTTT and GGCACCCCCTACTTCCTCTTT; αV integrin: GGCAGTGC CATAGCTCCTTT and CCCACTGCCCTTCAAGGGATTT; β1 integrin: GACTGATCAGTTCAGTTTGCTGT GTGTTT and CCCTGCTTGTATACATTCTCCACATGATTT. Reaction conditions were 95°C for 10 minutes followed by 50 cycles of 95°C for 15 s and 60°C for 1 minute. The difference in average cycle threshold (dCT) between L32 (housekeeping) and the gene of interest was determined from quadruplicate readings for each sample.

Cells were trypsinised and resuspended in FACS buffer (2% FCS in PBS, 0.02% sodium azide) at 1 10⁶ cells/ml, then incubated with 10 µg/ml antibodies to integrins β1, α2, α3, αV, and β1. Cells were then washed and incubated with FITC-conjugated goat anti-mouse or swine anti-rabbit IgG (Dako) and analyzed on a FACScan cytometer (Becton-Dickinson, Mountain View, CA).

Cells were plated on teflon printed glass slides (Electron Microscopy Sciences, Ft. Washington PA) and fixed with 3% paraformaldehyde/PBS then blocked with 5% BSA/0.25% Triton X-100. Cells were then washed in PBS and incubated with primary antibodies toward β1 integrin (AB2511) or MT1-MMP (AB815). Cells were washed and probed with a donkey anti-rabbit Alexa Fluor 488-conjugated antibody for β1 integrin or an Alexa Fluor 568 secondary antibody for MT1-MMP. Nuclei were then stained with propidium iodide (0.25 mg/ml), then the cells were washed and mounted using fluorescent mounting medium (Dako) and viewed by confocal microscopy (BioRad MRC 1024, Hemel Hempstead, UK).

Cells were lysed with 10 mM Tris HCl pH 7.6, 10 mM NaCl, 3 mM MgCl₂ and 1% Nonidet P-40 containing Protease Inhibitors (Roche), then cleared by centrifugation at 6,000 rpm for 10 minutes. Protein concentrations were determined with a BCA protein quantification assay (Pierce, Rockford IL). Equal amounts of lysates were electrophoresed with 10% SDS-PAGE under reducing conditions (100 mM DTT, Sigma). After electrophoresis, proteins were transferred to PVDF membranes and blocked with 5% skim milk. Blots were incubated with primary antibodies overnight and then probed with an HRP-conjugated secondary antibody (Pierce). Detection was performed using the ECL + Plus system (Pierce).

Cells were plated for near confluence at 37 C overnight. After each treatment, cells were washed twice with serum-free medium (SFM: unsupplemented DMEM), then replaced with fresh SFM. Recombinant full-length pro-MMP-2 (kindly, provided by Dr. Rafael Fridman, Wayne State University, Detroit, MI, USA) was added at 100 ng/ml concentration. At the appropriate time point, the conditioned medium was collected, transferred to a microcentrifuge tube and centrifuged at 6000 rpm for 10 minutes at 4°C. MMP-2 activation was analyzed by gelatin zymography using 5% polyacrylamide stacking gel and 10% polyacrylamide resolving gel (Bio-Rad Co., Richmond, VA) containing 1 mg/ml gelatin (BDH Laboratory Supplies, Poole, UK). Equal amounts of conditioned media were mixed with SDS sample buffer under non-reducing conditions. After electrophoresis, gels were washed with 50 mM Tris HCl pH 7.5, 5 mM CaCl₂ and 2.5% Triton X-100 and then incubated in 50 mM Tris HCl pH 7.5, 5 mM CaCl₂ at 37°C overnight. Gels were stained with 0.25% Coomassie Brilliant Blue (R-250) dye in 10% acetic acid and 10% isopropanol. Semiquantitative densitometry was performed using the Image J 1.46 program (NIH). Data is expressed as percent of area.

Collagen gels were made from Col I (Vitrogen 100, 3 mg/ml), neutralized with 1 M sodium hydroxide, pH 8.0 and then diluted in PBS to a final concentration of 2 mg/ml of Col I in 10 mM of sodium hydroxide. Cells were resuspended such that one part of the cell suspension was mixed with nine parts of the collagen solution, to give a final concentration of 4 10⁵ cell/ml. A total of 500 βl of cell-collagen mixture was added to 24 well plates and gels were allowed to polymerize for 60 min at 37°C. In order to release the gels, a yellow tip was inserted around and underneath the gels. The relaxed, floating gels were immersed with 1 ml medium and incubated at 37°C. The gel diameters were measured daily for 5 days using an inverted microscope. Collagen gel contraction is indicated as the decrease in gel area expressed as a percentage. Each experiment was performed in triplicate and with 3 independent experiments. Statistical analysis was performed by Student’s t-test.

Col I, fibronectin and vitronectin (0.5 ug/ml, 100 uL/well) were coated on 96-well plates at 37°C for 60 minutes and the wells were then washed and blocked in 3% BSA. Cells were detached from culture by Versene, resuspended at 2.5× 10⁵ cells/ml, and incubated at 37°C for 60 minutes to allow recovery of cell surface proteins. Cells (100 βl; 2.5× 10⁴ cells) were added to the wells and incubated at 37°C for 60 minutes. The adherent cells were stained with 0.5% Crystal Violet in 25% methanol for 5 minutes. Wells were then washed, dried and solubilization of the Crystal Violet was performed with 0.1 M sodium citrate containing 50% ethanol. Absorbance was measured at 540 nm (Power Wave X, Bio-Tek Instruments, Inc., Winooski, VT). Each experiment was performed in triplicate and 3 independent experiments were performed. Student’s t-test was used for statistical analysis.

β1 siRNAs β1/4, β1/5 and β1/6 were effective in knocking-down β1 integrin expression, however, β1/6 siRNA was more efficient than β1/4 and β1/5 in MCF-7-MT1 (Figure 1A) and MDA-MB-231 cells (data not shown). We hypothesized that β1 integrin knockdown would also affect the cell surface expression of the α subunit heterodimer partners. Integrin α1β1, α2β1, β10β1 and β11β1 are the major collagen-binding integrin receptors [35],[36]. MCF-7-MT1 cells had very low levels of β1, and no detectable levels of β10 or β11 integrin mRNA (data not shown). FACS analysis for α2, α3, and αV integrins showed that all three siRNAs also caused reduced cell surface levels of α2 and α3 integrin subunits. MCF-7-MT1 cells had low levels αV integrin, and no effect of the β1 siRNAs on αV integrin levels was observed (Figure 1B).

β1 integrin can bind multiple additional partners to form integrins in addition to the primary Col I receptors β1β1, α2β1, β10β1 and β11β1 [35],[36]. Integrins α3β1, α6β1, α6α4 and α7β1 are major LM receptors [37]. Integrins α5β1 and α8β1 are major FN receptors. Integrin αIIbβ3 is an alternate FN receptor, while αVα3 is specific for VN [36]. To examine the functional effects of inhibiting integrin β1 expression, adhesion analysis was performed with ECM substrates including VN, FN, and Col I after β1 integrin siRNA transfection. β1/6 integrin siRNA-treated cells showed significantly less adhesion (p <0.05) to Col I (Figure 2A). Decreased adhesion to FN was observed for β1/4, β1/5 and β1/6 siRNAs compared to the control siRNA cells (p < 0.05, 0.05 and 0.01 respectively), and again β1/6 integrin siRNA showed a stronger effect than β1/4 and β1/5 (data not shown). None of the β1 siRNA treated cells showed any reduction in adhesion to VN (ligand for αVα3), confirming that the β1 effects were specific (data not shown).

Collagen gel contraction was clearly observed with MDA-MB-231 cells and partial contraction was seen with MCF-7-MT1 cells. Collagen gel contraction appeared completely inhibited by β1/6 siRNA compared to β1/4 and β1/5 at Days 1, 3 and 5 in MDA-MB-231 cells (Figure 2C). All of the β1 integrin siRNA treated cells showed less collagen contraction than the siRNA control at Day 5 (Figure 2C). However, only β1/6 siRNA-treated cells showed significantly (p < 0.01) reduced ability to initiate collagen gel contraction on the first and all subsequent days (Figure 2D), while β1/4 and β1/5 integrin siRNAs initially failed contraction on the first day, but resumed contraction on the second and subsequent days. Consistent with the above results, only β1/6 integrin siRNA caused a significant reduction in collagen gel contraction (16.9%, p <0.001) compared to β1/4 and β1/5 (Figure 2D).

In MCF-7-MT1 cells, Col I caused activation of MMP-2 despite MT1-MMP being under the control of a heterologous promoter. Furthermore, when these cells were treated with cycloheximide to block new protein synthesis, some reduction in MMP-2 activation was seen, but the MMP-2 activational response to Col I was still observed [32]. Thus, MCF-7-MT1 cells provide an ideal model for studying non-transcriptional regulation of Col I-induced MMP-2 activation. Again, integrin expression appeared completely abrogated by β1/6 siRNA compared to β1/4 and β1/5 at 72, 96 and 120 hours in MCF-MT1 cells (Figure 3A). All three β1 integrin siRNAs caused reduced MMP-2 activation. Both β1/4 and β1/5 integrin siRNAs reduced Col I-induced MMP-2 activation at 24 hours and showed sustained reduction until 48 hours after knockdown, after which Col I-induced MMP-2 activation was not evident (i.e. between 72 and 120 hours after knockdown). On the other hand, reduced MMP-2 activation with β1/6 integrin siRNA was seen at 72 hours after knockdown, and continued to decrease, again showing that β1/6 was more efficient than β1/4 and β1/5 in both MCF-7-MT1 (Figure 3B) and MDA-MB 231 cells (data not shown)

Col I-stimulated MMP-2 activation functions principally via MT1-MMP in fibroblasts, as no MMP-2 activation occurs in MT1-MMP-deficient mouse fibroblasts in response to Col I [38],[39]. Since we found that abrogation of β1 integrin expression was able to suppress Col I-stimulated MMP-2 activation, the influence of β1 integrin on MT1-MMP activity was examined. Interestingly, β1/5 and β1/6 integrin siRNAs caused noticeably decreased MT1-MMP levels, however, the β1/6 integrin siRNA showed the strongest decrease, which corresponded to the lowest β1 integrin level compared to controls (siRNA control) (Figure 3C and D) and the highest suppression of MMP-2 activation.

Cell surface β1 integrin and MT1-MMP were also investigated by immunofluorescence. β1 integrin expressing cells were better spread, whereas all β1/5 and β1/5 knockdown cells displayed rounder and less-spread morphology, with β1/6 integrin siRNA being most pronounced (Figure 3E). Cell surface β1 integrin localized to the cell periphery while MT1-MMP was more diffusely distributed across the cell but concentrated on the cell surface, and the levels of MT1-MMP appeared to correlate with those of β1 integrin. Reduced cell surface β1 integrin and MT1-MMP was only seen for β1/6 integrin siRNA-treated cells, compared to β1/4 and β1/5, irrespective of Col I treatment (Figure 3F). Additionally, β1 integrin and MT1-MMP levels examined by immunofluorescence corresponded to Western blot determination (Figure 3C, D). Thus, β1/6 integrin siRNA showed the most efficient β1 integrin knockdown, which coincided with the strongest reduction of MT1-MMP.

To confirm that the reduced MMP-2 activation after β1 integrin knockdown was due to reduced MT1-MMP, MCF-7-MT1 cells were transfected with β1/4, β1/5 and β1/6 integrin siRNAs, and subsequently transfected with MT1-MMP. Forced MT1-MMP expression was confirmed (Figure 3G), and the expected suppression of Col I-induced MMP-2 activation by β1/6 integrin siRNA treatment was rescued (Figure 3H). These results confirm a role for β1 integrin in regulating MT1-MMP to facilitate MMP-2 activation in response to Col I stimulation.

There are several factors that may influence the efficiency of RNAi in mammalian systems including the choice of the target site of degradation, the transfection method and the turnover rate of the protein. In addition, the mRNA degradation rate is an important factor influencing target gene knockdown by siRNA technology [40],[41]. This effect is dependent on the dose response to siRNA concentration. We examined whether β1/4, β1/5 and β1/6 siRNAs had dose-dependent effects on the abrogation of β1 integrin mRNA expression, and whether this was relative to the reduction seen in MMP-2 activation. We hypothesized that the β1/6 siRNA showed a quicker mRNA degradation rate than either β1/4 or β1/5 siRNA. Serial concentrations of β1/6 siRNA (2.5, 5, 10, 15 and 20 µM) were tested. The data showed that at the lowest concentration (2.5 µM), β1/6 siRNA was able to knockdown β1 integrin expression to an extent comparable to the highest concentration (20 µM) of β1/4 or β1/5 siRNA (Figure 4A). The β1 integrin expression was below the threshold of detection 72 hours after transfection with 2.5 and 5 µM of β1/6 siRNA, and this was also seen with β1/4 and β1/5 siRNA at 20 µM. MMP-2 activation reductions were observed only at 20 µM of β1/6 integrin siRNA (Figure 4C), was which relative to reduction of MT1-MMP protein level at this concentration (Figure 4B). These results indicate that the reduction of Col I-induced MMP-2 activation seen only with β1/6 may require the more rapid and complete knock down of β1 integrin seen with this siRNA. A threshold of β1 integrin knockdown that cannot be surpassed by β1/4 or β1/5 siRNAs may exist for the MMP-2 activation effect.

Wild-type β1 integrin transfection was successful at increasing β1 integrin expression at all time points, however, strong inductions were observed at 12 and 24 hours after transfection. Consistently, cells over-expressing β1 integrin showed higher MT1-MMP expression than vector control cells at 24, 48 and 72 hours (Figure 5A). The MT1-MMP levels in the vector control cells resembled that seen in β1 integrin knockdown cells without any further transfection (Figure 3C).

Co-transfection of β1 integrin codon-swapped mutant and β1/6 integrin siRNA failed to knockdown β1 integrin, whereas co-transfection of wild-type β1 integrin and β1/6 integrin siRNA (β1/6 integrin siRNA alone) successfully knocked down β1 integrin (Figure 5A). Co-transfection of β1 integrin codon-swapped mutant and β1/6 integrin siRNA rescued the otherwise inhibited MMP-2 activation in both MCF-7-MT1 cells (Figure 5B) and MDA-MB-231 cells (data not shown). These results indicated that the stronger effect of β1/6 integrin siRNA was specific to β1 integrin.

Stimulators of MMP-2 activation that do not function through β1 integrins, such as Con A and TPA, would be expected to remain unaffected by β1 integrin knockdown. As expected, MMP-2 activation by Con A or TPA was not inhibited by β1/6 integrin siRNA treatment (Figure 5C).