The oral SCC cell line, SCCKN (30) was grown in RD medium (a 1:1 mixture of RPMI-1640 and Dulbecco’s Modified Eagle’s Medium) supplemented with 5% fetal bovine serum (FBS) in 5% CO₂ at 37°C. The proliferation of the cells on ECM proteins was estimated as follows: The wells of 24-well tissue culture plates were incubated with 100 μg/ml type I collagen, type IV collagen, fibronectin, laminin or vitronectin (all from Sigma, St. Louis, MO, USA) overnight at 4°C. Poly-L-lysine (100 μg/ml; Sigma) was used as a non-integrin-dependent adhesion substrate. The cells (2×10⁴) suspended in RD containing 10 μg/ml bovine insulin, 5 μg/ml human transferrin, 0.5 mg/ml fatty acid-free bovine serum albumin (BSA), 10 μM 2-mercaptoethanol, 10 μM 2-aminoethanol and 10 nM sodium selenite (all from Sigma) (31) were seeded in each well of the culture plates and cultured in 5% CO₂ for 6 days at 37°C. The number of cells was measured by the Coulter counter (Beckman Coulter, Inc., Tokyo, Japan). The measurements were done in triplicate.

The open reading frame of human integrin αv was amplified by PCR from plasmid CDM8 containing human αv cDNA, which was provided by Dr Joseph C. Loftus (Mayo Clinic Arizona, Scottsdale, AZ, USA). The forward primer (5′-CGGAATTCTTCGGCGATGGCTTTTCCGC-3′) containing EcoRI site (underlined) and the reverse primer (5′-TCCCCCGGGTTAAGTTTCTGAGTTTCCTTCACCAT-3′) containing SmaI site (underlined) were used for the amplification. The PCR products were double-digested with EcoRI (New England Bio Labs, Ipswich, MA, USA) and SmaI (New England Bio Labs) and then ligated into pCI-neo Mammalian Expression Vector (Promega Corporation, Madison, WI, USA) digested with both EcoRI and SmaI. The inserted cDNA sequences were all verified by DNA sequence analysis. The resultant plasmid was termed as pCI/neo-αv, and pCI/neo without insert was used as negative control.

SCCKN cells were transfected with pCI/neo-αv or pCI/neo (5 μg per 60-mm dish) using TransFast transfection reagent (Promega Corporation) according to the manufacturer’s instruction. Selection was initiated 48 h after transfection by adding 600 μg/ml G418 (Geneticin; Invitrogen Life Technologies, San Diego, CA, USA) to the culture medium. The selection medium was changed every 4 days for 2 weeks until all non-transfected cells died. Resistant cell clones were isolated. Cell clones transfected with pCI/neo-αv or pCI/neo were termed as KNαv or KNmock, respectively.

To detect integrin αv and β8 proteins, SCCKN, KNmock and KNαv cells were lysed with lysis buffer [10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 1% protease inhibitor cocktail (Sigma)]. The samples containing 10 μg of total protein were electrophoresed on 10% SDS-polyacrylamide gels under reducing condition and transferred to polyvinylidene difluoride (PVDF) membrane filters (Millipore, Bedford, MA, USA). The filters were blocked in T-TBS (20 mM Tris-HCl, pH 7.5, 137 mM NaCl, 0.1% Tween-20) containing 5% skim milk for 1 h at room temperature and then incubated with rabbit anti-integrin αv polyclonal antibody (Santa Cruz Biotechnology, Inc., Dallas, TX, USA) or goat anti-integrin β8 polyclonal antibody (Santa Cruz Biotechnology, Inc.), followed with the incubation with horseradish peroxidase (HRP)-conjugated anti-rabbit IgG antibody (Cell Signaling Technology, Inc., Danvers, MA, USA) or HRP-conjugated anti-goat IgG antibody (KPL, Gaithersburg, MD, USA), respectively. Rabbit anti-β-actin polyclonal antibody (Cell Signaling Technology, Inc.) was used as internal control to confirm equal loading of total protein. Protein bands were visualized by enhanced chemiluminescence detection (ECL Plus System; GE Healthcare, Uppsala, Sweden).

The wells of 24-well culture plates were incubated with 100 μg/ml type I collagen, type IV collagen, fibronectin, laminin or vitronectin overnight at 4°C. Poly-L-lysine (100 μg/ml) was used as a non-integrin-dependent adhesion substrate. The wells were washed five times with phosphate-buffered saline (PBS) and incubated with in PBS containing 1% BSA for 1 h at 37°C to block non-specific binding. Subconfluent culture of cells was radiolabeled with 1 μCi/ml [methyl-³H]-thymidine (Perkin-Elmer, Waltham, MA, USA) for 24 h. The labeled cells (1×10⁵) suspended in RD medium containing 0.1% BSA were added to each well of the culture plates. After incubation for 30 min at 37 °C, the medium was aspirated, and the wells were gently rinsed twice with PBS. The cells adhering to the well were dissolved in 1 N NaOH, and the radioactivity was measured with a liquid scintillation counter (LSC903; Aloka, Co., Ltd., Tokyo, Japan). The adhesion capacity was determined relative to the radioactivity of seeded cells (1×10⁵) that was considered to be 100%. Each assay was performed in triplicates and repeated three times.

Five hundred-microliters of 0.21% type I collagen gel solution (Koken, Co., Ltd, Tokyo, Japan) in RD neutralized with reconstitution buffer (0.05 N NaOH, 2.2% NaHCO₃, 0.2 M HEPES) was pipetted into each well of 24-well culture plate and gelled as a basal layer with incubation for 1 h at 37°C. Thereafter, 500 μl type collagen solution containing cells (2.5×10⁴) was poured onto the basal layer and gelled, and received 1 ml RD containing 5% FBS. The cells in the gels were cultured for 12 days at 37°C. The colonies that formed in the gel were fixed with phosphate-buffered 10% formalin. Sections were prepared and stained with hematoxylin and eosin.

SCCKN, KNmock and KNαv cells suspended in RD containing 2% BSA were seeded on culture dishes coated with 100 μg/ml type I collagen. At various incubation times, the cells were washed with ice-cold PBS containing 1 mM sodium vanadate and lysed with Laemmli sample buffer (62.5 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 1% β-mercaptoethanol) supplemented with protease inhibitor cocktail and 1 mM sodium vanadate. The samples were separated on 10% SDS-polyacrylamide gels and transferred onto PVDF membrane filters. The immunoblot analysis was performed using rabbit anti-phospho-focal adhesion kinase (FAK) monoclonal antibody, rabbit anti-phospho-mitogen-activated protein kinase kinase 1/2 (MEK1/2) monoclonal antibody and rabbit anti-phospho-extracellular signal-regulated kinase 1/2 (ERK1/2) monoclonal antibody. To detect total FAK, MEK1/2 and ERK1/2 proteins, rabbit anti-FAK monoclonal antibody, rabbit anti-MEK1/2 monoclonal antibody and rabbit anti-ERK1/2 monoclonal antibody were used, respectively. After incubation with primary antibodies, the membranes were incubated with HRP-conjugated secondary antibody, and protein bands were detected using an enhanced chemiluminescence reagent. All antibodies used for the phosphorylation assay were purchased from Cell Signaling Technology, Inc.

To downregulate integrin β8, a morpholino antisense oligonucleotide specific for integrin β8 obtained from GeneTools (Philomath, OR, USA) was used. The sequence of the antisense oligonucleotide is as follows: 5′-AAGCCAGGGCCGAGCCGCACATAAT-3′. A standard control morpholino oligonucleotide (5-CCTCTTACCTCAGTTACAATTTATA-3) was used as a negative control. Delivery of the oligonucleotides into the cells was performed according to the GeneTools protocol. Briefly, 80–100% confluent SCCKN, KNmock or KNαv cells were treated with 10 μM of the morpholino antisense oligonucleotide or the standard control oligonucleotide, and 6 μM of Endo-Porter reagent (GeneTools). After 24 h, the cells were used for the subsequent experiments.

Total cytoplasmic RNA of SCCKN, KNmock or KNαv cells in confluent cultures was isolated with using TRIzol reagent (Life Technologies, Rockville, MD, USA) according to manufacture’s instruction. Total RNA (20 μg) obtained from the cells were separated on a 1% agarose gels containing 2.2 M formaldehyde and transferred directly from the gel to a nylon membrane (Hybond-N+; GE Healthcare) in 10X SSC (1X SSC is 0.15 M NaCl plus 1.5 mM sodium citrate) overnight. After transfer, RNA was UV cross-linked (120,000 μJ of UV), and the membrane was prehybridized with Rapid-hyb buffer (GE Healthcare) for 15 min at 65°C. The specific probe for integrin β8 was obtained by reverse transcriptase-PCR as follows: First-strand cDNA was synthesized from total RNA of SCCKN cells with ReverTra Ace (Toyobo, Osaka, Japan) and PCR amplification was performed using forward primer (5′-GATCAGACGTCTCATCTCGC-3′) and reverse primer (5′-CTCTTCCACTGCACACTTGG-3′). The PCR products (961 bp) were subcloned into pGEM-T Easy Vector (Promega Corporation), and the inserted cDNA sequences were verified by DNA sequence analysis. The plasmid was digested with EcoRI, and the insert was gel-purified and radiolabeled with [α-³²P]-dCTP (Perkin-Elmer) using Rediprime II DNA Labeling System (GE Healthcare). Hybridization was carried out for 2 h at 65°C in Rapid-hyb buffer containing 1×10⁶ cpm/ml probe. The membrane was washed with 2X SSC/0.1% SDS for 20 min at room temperature and washed twice with 0.5X SSC/0.1% SDS for 15 min at 65°C. Hybridization signals were detected by the BAS 2000 image analyzer (Fujifilm, Tokyo, Japan). Equivalent loading of ribosomal RNA was confirmed by methylene blue staining.

A small amount of integrin αv protein was observed in SCCKN and KNmock cells. In contrast, integrin αv transfectants in KNαv cells showed a large amount of integrin αv protein (Fig. 1A).

To examine the effect of integrin αv on cell adhesion to ECM proteins, SCCKN, KNmock and KNαv cells were seeded on various ECM protein-coated wells and incubated for 30 min. Only 5 or 10% of SCCKN and KNmock cells adhered to any ECM protein. In contrast, over 35% of KNαv cells adhered to type I collagen, type IV collagen, fibronectin and vitronectin after 30-min incubation (Fig. 1B).

To examine the participation of integrin αv in the proliferation of SCC cells, SCCKN, KNmock and KNαv cells were grown on various ECM proteins in the absence of serum for 6 days. Transfection with integrin αv cDNA led to a marked increase in cell proliferation on type I collagen. The number of KNαv cells grown on type I collagen was about 3-fold compared to the number of SCCKN and KNmock cells on type I collagen. In contrast, other ECM proteins exhibited no significant effect on the proliferation of KNαv cells (Fig. 2A).

The behavior of SCCKN, KNmock and KNαv cells was examined by three-dimensional culture using type I collagen gel. The cells embedded in type I collagen gel were cultured for 12 days. SCCKN and KNmock cells formed small and spherical colonies in the gel. In contrast, KNαv cells formed dilated colonies with irregular margins, and some cells migrated into the surrounding collagen gel, suggesting that transfection with integrin αv cDNA led to the enhancement of invasiveness of SCC cells (Fig. 2B).

Type I collagen enhanced the proliferation of KNαv cells, and KNαv cells exhibited the enhanced invasiveness into type I collagen gel compared to SCCKN and KNmock cells. Several studies have shown that the MEK/ERK signaling pathway via integrin αv regulates cell proliferation (32–34). To clarify the participation of the MEK/ERK signaling pathway in type I collagen-induced proliferation and invasion of KNαv cells, SCCKN, KNmock and KNαv cells were detached and replated onto type I collagen, and the phosphorylation of FAK, MEK1/2 and ERK1/2 in the cells was investigated after cultivation for various periods. The phosphorylation of FAK, MEK1/2 and ERK1/2 in KNαv cells was observed at 2, 6 h and 6 h after replating on type I collagen, respectively. In contrast, the phosphorylation of FAK, MEK1/2 and ERK1/2 in SCCKN and KNmock cells was observed at 12, 24 h and 24 h after replating, respectively (Fig. 3).

To examine the effect of the suppression of integrin β8 on the adhesion of SCC cells to ECM proteins, the cells were transfected with a morpholino antisense oligonucleotide targeting integrin β8 subunit (Fig. 4A) or a control oligonucleotide (Fig. 4B). The adhesion of SCCKN and KNmock cells to ECM proteins were not greatly affected by the suppression of integrin β8. In contrast, the suppression of integrin β8 by the antisense oligonucleotide led to the remarkable decrease in the attachment of KNαv cells to type I collagen compare to the attachment of KNαv cells transfected with the control oligonucleotide. However, the suppression of integrin β8 did not reduce the adhesion of KNαv cells to any ECM protein except type I collagen.

We examined the participation of integrin αvβ8 in type I collagen-induced proliferation of KNαv cells. The suppression of integrin β8 by a morpholino antisense oligonucleotide targeting integrin β8 led to remarkable decrease in the proliferation of KNαv cells cultured on type I collagen. However, transfection with the antisense oligonucleotide did not have strong effect on the proliferation of SCCKN and KNmock cells on type I collagen (Fig. 5A).

We next examined the effect of the suppression of integrin β8 on the morphology of colonies of KNαv cells in type I collagen gels. KNαv cells transfected with the control oligonucleotide formed dilated colonies with irregular margin, and some cells migrated into the surrounding collagen gel. In contrast, the colonies of KNαv cells transfected with the antisense oligonucleotide were small and spherical colonies (Fig. 5B).

The phosphorylation of FAK, MEK1/2 and ERK1/2 in KNαv cells transfected with the control oligonucleotide was observed at 2, 6 h and 6 h after replating on type I collagen, respectively. In contrast, the phosphorylation of FAK, MEK1/2 and ERK1/2 in KNαv cells transfected with the antisense oligonucleotide was observed at 12, 24 h and 24 h after replating, respectively (Fig. 5C).

The proliferation of KNαv cells on type I collagen was reduced by the suppression of integrin β8. Moreover, type I collagen-stimulated phosphorylation of FAK, MEK1/2 and ERK1/2 in KNαv cells was also inhibited by the suppression of integrin β8. These findings suggest that the interaction of integrin αvβ8 with type I collagen might activate FAK and the MEK/ERK signaling pathway and induce the proliferation of SCC cells. Therefore, the expression of integrin β8 protein and mRNA in SCCKN, KNmock and KNαv cells was examined by western blotting and northern blotting, respectively. The expression of integrin β8 protein in KNαv cells was enhanced compared to SCCKN and KNmock cells. In contrast, there was no remarkable difference in the expression of integrin β8 mRNA between SCCKN, KNmock and KNαv cells (Fig. 6).