αvβ3 (LM609) and β5 (P1F6) blocking integrin antibodies 2627 were purchased from Chemicon. The Src 327 monoclonal antibody has been described previously 28. Src phospho-specific antibody, Y418, was purchased from Biosource. Antibodies to PKCα, PKCδ, PKCε, Rac1, and RhoA were obtained from BD Transduction Labs. PAK and phospho-PAK antibodies were purchased from Cell Signaling Technology. p190RhoGAP antibodies came from Santa Cruz. Alexafluor 488 goat anti-mouse and Alexafluor 546 phalloidin were from Molecular Probes and Hoechst 33258 was from Sigma.

The metastatic melanoma cell line C8161.9 29, obtained from Dr. Danny Welch, (University of Alabama, Birmingham, Alabama, USA) was grown in DMEM/F12 (Invitrogen) supplemented with 2 mM L-glutamine, 50 U/ml penicillin, 50 ug/ml streptomycin/ml, 0.1 mM NEAA, 1 mM NaPyruvate, and 5% fetal bovine serum (FBS). M14 melanoma cells (Dr. Han-Mo Koo, Van Andel Institute) were grown in RPMI-1640 (Invitrogen) supplemented with 2 mM L-Glutamine, Gentamicin, and 5% FBS. A375 metastatic melanoma cells were obtained from ATCC and grown in MEM Hanks (Invitrogen) supplemented with 50 U of penicillin, 50 ug/ml streptomycin, 0.1 mM NEAA, 1 mM NaPyruvate, and 10% FBS. HUVECS were obtained from Cascade Biololgicals and normal cultured primary melanocytes, NHEM, were purchased either from Clonetics or Cascade. Both cell lines were grown in media supplied by their respective companies. DF1 chicken fibroblast cell line 30 (Dr. Bart Williams, Van Andel Institute) was grown in DMEM (Invitrogen) supplemented with 2 mM L-Glutamine, 1 mM NaPyruvate, 50 U/ml penicillin, 50 ug/ml streptomycin, and 10% heat inactivated FBS. Stable C8161.9 and M14 cell lines expressing tVA, the chicken retrovirus receptor, were established as previously described 31.

Src inhibitors PD173955, PD180970, and SU6656, and PKC inhibitors bisindolylmaleimide and 12-O-tetradecanoylphorbol 13-acetate (TPA) were purchased from Calbiochem. Cells were treated overnight with 100 ng/ml TPA or for 30 minutes prior to use with 10 uM bisindolylmaleimide, 10 uM PD173955 or PD180970, or 2 uM SU6656.

One ug of pcDNA-RhoA-Q61L 32 was transiently cotransfected with 1 ug pCMV-GFP into C8161.9 cells with Lipofectamine Plus (Invitrogen) using the recommended protocol. 48 hours after transfection cells were plated on matrix for immunofluorescent staining.

pGST-Src and pGST-SrcK295M 33 were obtained from Dr. Sara Courtneidge (Burnham Institute, San Diego, California, USA) and subcloned via EcoRI into pENTR3c vector (Gateway, Invitrogen) and then recombined into the Gateway modified pDEST-RCAS retroviral vector 34. Full length rabbit PKCα and mouse PKCδ cDNAs from Dr. Shigeo Ohno (Yokohama City University, Japan) were subcloned via EcoRI into Bluescript cloning vector (Strategene) to generate pBS-PKCα and pBS-PKCδ. PKCα and PKCδ were released with NotI and ClaI and subcloned into pRCASBP-Y retroviral vector 34. RCAS-Src, RCAS-SrcK295M, RCAS-PKCα, and RCAS-PKCδ were transfected into DF1 chicken fibroblasts and retroviruses generated as previously described 30. C8161.9 cells or M14 cells expressing tVA were infected with virus (~1 × 10⁷ IU/100 mm plate) in the presence of 8 ug/ml Polybrene (Invitrogen). Seventy-two hours after infection, cells were selected in 7 ug/ml of puromycin (Invitrogen) or 400 ug/ml G418 (Invitrogen) for 2 weeks. Surviving cells were pooled and maintained in standard DMEM/F12 growth medium.

PKC isoform-specific siRNAs sequences (IDT) were as follows: PKCδ si1δ: 5'-UGGCGCCGUUCCUGCGCAU-3', PKCα si2α: 5'-CCUGCGAUAUGAACGUUCA-3', and si3α: 5'-AAGGCUUCCAGUGCCAAGA-3'. The scrambled siRNA control sequence was 5'-ACUACCGUUGUUAUAGGUG-3'. All sequences were blasted using BLASTN and contained no homology to any other known human genes. C8161.9 cells at 30–50% confluency in 100 mm tissue culture dishes were starved (DMEM/F12 + 0.1% FBS) 24 hours prior to transfection. Cells were transfected with 20 nM siRNAs with SiLentFect (Bio-Rad) as suggested by the manufacture. The cells were allowed to recover in growth media supplemented with 2% FBS for 24 hours. Cells were then placed in starvation media for an additional 48 hours before use in adhesion assays.

Serum starved cells were removed from tissue culture plates by trypsinization and rinsed with soybean trypsin inhibitor (Invitrogen). Cells resuspended in PBS + 0.1% BSA at 1 × 10⁶ to 1 ×10⁷ cells/ml were incubated with 5 ug/ml of αvβ3 or β5 integrin antibodies for 1 hour at 4°C. Cells were washed with PBS and incubated with FITC-conjugated secondary antibodies in PBS+0.1% BSA for 1 hour at 4°C. Cells were washed with PBS, fixed with 1% paraformaldehyde in PBS at 4°C for 20 minutes, washed in PBS, and analyzed by Fluorescent Activated Cell Sorting (FACS) with the FACSCalibur (Becton Dickinson) and CellQuest (Becton Dickinson).

For assays done on extracellular matrices, cells were prepared as previously described 35. Briefly, cells were growth factor-starved for 48 hours, trypsinized, treated with soybean trypsin inhibitor (Invitrogen), washed in PBS, and suspended in serum-free medium for 30–60 minutes. Cells were then plated on tissue culture plates or sterilized coverslips coated with 5 ug to 10 ug/ml vitronectin (Chemicon) and blocked with 1% BSA (Sigma). Suspension controls were maintained at 37°C. One to two hours after plating on matrix cells were either lysed for biochemical analysis or immunostained.

For biochemical analyses cells were lysed either in Triton-X (50 mM Tris pH 7.5, 100 mM NaCl, 0.5 mM EDTA, 1% TritonX-100, 50 mM NaF, 50 mM β-glycerophosphate, 5 mM sodium pyrophosphate, 1 mMNa₃VO₄, 1 mM PMSF, 100 U/ml aprotinin, 10 ug/ml pepstatin, and 10 ug/ml leupeptin) or RIPA (10 mM Tris pH 7.2, 158 mM NaCl, 1 mM EDTA, 0.1% SDS, 1% NaDOC, 1% Triton-X100, 1 mMNa₃VO₄, 1 mM PMSF, 100 U/ml aprotinin, 10 ug/ml pepstatin, and 10 ug/ml leupeptin) buffers. For immunoprecipitation, 500 to 1000 ug/ml protein was incubated with the appropriate antibodies for 3 hours at 4°C with either protein A- or protein G-conjugated agarose beads (Pierce) to capture the complexes. Washed immunoprecipitates were resuspended in 2× SDS sample buffer containing 10% β-mercaptoethanol. All samples were subjected to SDS-polyacrylamide gel electrophoresis, and transferred to a polyvinylidene difluoride membrane (PVDF). The PVDF membranes were blocked with 5% BSA in Tris-buffered saline containing 0.1% Tween 20 (TBST) for 2 hours, followed by 2 hour incubation with the appropriate primary antibodies in 5% BSA/TBST. After washing, blots were incubated with a horseradish peroxidase-conjugated secondary antibody for 1 hour in 5% BSA/TBST, visualized with a chemiluminescence reagent, and captured by a CCD camera in a Bio-Rad Chemi-Doc Imaging System and quantified using Quantity One software (Bio-Rad). Blots were stripped in low-pH 2% SDS at 65°C for 60 minutes, rinsed and reprobed for total levels of protein in the immunoprecipitates or cell lysates.

Boyden chambers inserts containing Matrigel in the upper chamber were purchased from Becton-Dickinson. The lower side of the porous membrane in the Boyden inserts was treated with 10 ug/mL vitronectin overnight at room temperature. This results in a vitronectin coating on the lower side of the membrane as well as some vitronectin seeping into the Matrigel above. After coating, the lower side of the membranes were washed with PBS and blocked with 1% BSA for 2–4 hours at 37°C while the Matrigel in the upper chamber was rehydrated with DMEM. Serum starved cells were placed in suspension in serum free medium. Inhibitors or blocking antibodies were added to suspension cells prior to plating in chambers and 1.5 × 10⁵ to 2 × 10⁵ cells were added to the top chamber in the absence of serum. Medium without serum was added to the lower chambers. All chambers were incubated in 5% CO₂ at 37°C for 72 hours. Chamber inserts were rinsed with PBS and then cells were stained and fixed with crystal violet/formalin mixture (Chemicon) for 20 minutes. Any cells remaining in the top chamber were removed with a cotton applicator tip. Cells having migrated to the bottom chamber and adhering to the lower side of the membrane were visualized with Nikon Eclipse TE300 Epi-flourescence Microscope using a 10× objective. Pictures were taken with the Hamamatzu Digital CCD camera using OpenLab imaging software (Improvision) and the RGB Colour Merge Automator (OpenLab). Five pictures of each well were taken and the cells in each image were counted. All assays were repeated at least 3 times.

Rho pull-down assays were carried out essentially as described 36 and the Rac pull-down protocol was adapted from del Pozo et al 37. GST-RBD (rhotekin binding domain) and GST-PBD (PAK binding domain) protein expression was induced with 0.5 mM Isopropyl-β-D-thiogalactopyranoside (IPTG) for 2 hr at 30°C in 4 liters of culture or for 3–4 hr at 37°C in 200 ml respectively. GST-RBD bacteria were resuspended in 40 ml cold lysis buffer (50 mM Tris, pH 7.5, 1% Triton X-100, 150 mM NaCl, 5 mM MgCl₂, 1 mM DTT, 10 ug/ml aprotinin, 10 μg/ml leupeptin and 1 mM PMSF). GST-PBD bacteria were resuspended in 10 ml cold lysis buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 10% glycerol, 2 mM EDTA, 10 ug/ml aprotinin, 10 ug/ml leupeptin and 1 mM PMSF). Suspensions were sonicated and clarified lysates were mixed with 0.6 ml glutathione beads (Sigma) and rotated at 4°C for 60 min. The GST-RBD beads were washed six times with 12 ml wash buffer (50 mM Tris, pH 7.5, 0.5% Triton X-100, 150 mM NaCl, 5 mM MgCl₂, 1 mM DTT, 1 ug/ml aprotinin, 1 ug/ml leupeptin and 0.1 mM PMSF) and GST-PBD beads were washed three times with 12 ml lysis buffer supplemented with 0.5% NP-40 then washed 3× with storage buffer (50 mM Tris pH7.5, 150 mM NaCl, 10% glycerol, 5 mM MgCl₂, 1 mM DTT). Beads were resuspended in 1 ml wash buffer with 10% glycerol and used immediately for pull-downs. Purified GST-RBD and GST-PBD proteins were quantified using a 12% SDS PAGE and BSA standards for comparison. These preparations typically yielded between 20–30 ug protein per 20 ul aliquot of bead suspension.

Following adhesion to matrix cells were placed on ice and washed twice with ice-cold Tris-buffered saline (50 mM Tris, pH7.4, 150 mM NaCl, 10 mM MgCl₂). For Rho assays cells were lysed in 500 ul of cold Rho lysis buffer (50 mM Tris, pH 7.2, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 500 mM NaCl, 10 mM MgCl₂, 10 μg/ml each of leupeptin and aprotinin, and 1 mM PMSF). For Rac assays, cells were lysed in 500 ul of cold Rac lysis buffer (50 mM Tris pH 7.4, 10 mM MgCl₂, 1% Triton-X, 10% glycerol, 500 mM NaCl, 0.1% DOC, 10 ug/ml leupeptin, 10 ug/ml aprotinin, 10 ug/ml pepstatin, 2 mM PMSF). 300–500 ug of lysates were combined with 20 ul of bead suspension and rotated at 4°C for 60 min. Bead suspension was washed four times with 600 μl ice cold buffer containing 50 mM Tris, pH 7.2, 1% Triton X-100, 150 mM NaCl, 10 mM MgCl₂, and 10 ug/ml each of leupeptin and aprotinin, and 0.1 mM PMSF. Samples were boiled 10 min in SDS sample buffer and separated on 12% Tris-Glycine SDS PAGE gels. Proteins were transferred to PVDF membrane for immunoblotting with RhoA or Rac antibodies.

ECL images were captured by a CCD camera in a Bio-Rad Chemi-Doc Imaging System and the intensity of the activated RhoA and total RhoA ECL signals were quantified using Quantity One software (Bio-Rad). Extent of RhoA activation was normalized to total RhoA for each sample. Data were expressed as the fold in increase in RhoA activation compared to control or untreated samples.

Sterilized cover-slips (Fisher) were coated with 5–10 μg/ml vitronectin O/N at 4°C, blocked in PBS + 1% BSA for 1–2 hours at 37°C, and then washed with PBS prior to plating 2–3 × 10⁵ suspended cells in serum-free medium. After incubation at 37°C for 45 minutes cells were fixed with 4% paraformaldehyde in PBS at 4°C for 20 minutes. Cells were permeabilized with TBS (10 mM Tris, pH 8.0, 150 mM NaCl) + 0.5% TritonX-100 for 10 minutes at room temperature and blocked with 2% BSA/TBS-T for 20 minutes at room temperature. Cells were then incubated at room temperature with primary antibody and Phalloidin 546 in 2% BSA/TBS-T for 60 minutes and then incubated with Alexaflour 488 goat anti-mouse in 2% BSA in TBS-T for 60 minutes. Nuclei were stained with Hoechst for 5 minutes at room temperature. Cover-slips were mounted onto slides in 50% glycerol. Cells were visualized with Nikon Eclipse TE300 Epi-flourescence Microscope using a 60× oil immersion objective. Pictures were taken with the Hamamatzu Digital CCD camera using OpenLab imaging software (Improvision).

The primary ligand for αvβ3 is vitronectin (VN); however, another integrin, αvβ5, also binds VN. To determine which integrin is responsible for the biological effects mediated by adhesion to VN, αvβ3 and αvβ5 integrin expression was measured in the melanoma cells. The cell surface levels of αvβ3 and αvβ5 in cultured melanocytes and in a non-invasive cell line, M14 38, and a highly metastatic melanoma cell line, C8161.9 29, were quantified by FACS. Irrespective of their metastatic potential, both melanoma cell lines expressed similar levels of αvβ3, while normal cultured melanocytes and HUVECs (positive control) expressed over 2 times more αvβ3. On the other hand, the more highly metastatic melanoma cell line, C8161.9, expressed over 2× higher levels of αvβ5 than M14 (Figure 1A). To determine which integrin is required for the migratory and invasive properties of the melanoma cell lines, cells were treated with or without VN integrin-blocking antibodies 2627 or IgG and their ability to invade and migrate through vitronectin-enriched Matrigel to the lower chamber of a Boyden chamber in the absence of serum or growth factors was evaluated. M14 and normal melanocytes both failed to invade through VN/Matrigel, while invasion of the highly metastatic C8161.9 cells was dramatically diminished in the presence of αvβ3, but not αvβ5 blocking antibodies (Figure 1B). If VN is omitted C8161.9 cells are unable to invade. Thus, expression of αvβ3 is required for invasion of highly metastatic melanoma cells. However, αvβ3 expression alone is not sufficient, since αvβ3 is expressed on both M14 and normal cultured melanocytes and these cells were unable to invade. Therefore, additional signaling events must also be required. These signaling events must be intrinsic to the cells, since no exogenous growth factors, serum, or additional external stimuli other than extracellular matrix (ECM) proteins, were added to the invasion assays.

αvβ3 integrin is a strong activator of the nonreceptor tyrosine kinase Src 11, which is known to regulate cell migration and invasion 39. Therefore, to determine whether Src is required for the invasive properties of C8161.9 cells, the levels of Src protein and its activity were first measured by immunoblotting of cell extracts from normal melanocytes, M14, and C8161.9 cells after adhesion to VN. There were no significant differences in total levels of Src protein between these cell lines; however, the level of activated Src (as measured by phosphorylation at Y418) was dramatically increased in the highly metastatic C8161.9 cells (Figure 2A). A similar decrease in phosphorylation at Y529 was also observed (data not shown). Strikingly this increase in Src activity occurred independently of integrin engagement, as Src activity was high even in cells placed in suspension. Next, Src activity was inhibited in C8161.9 cells and αvβ3-dependent invasion was measured. Over expression of a dominant interfering form of Src (SrcK295M) 33 or treatment with two Src inhibitors dramatically reduced C8161.9 cell invasion (Figure 2B). Over expression of wild type Src caused a 50% increase in invasion. Normal cultured melanocytes and M14, both express αvβ3 integrin (Figure 1A); however, neither have high levels of active Src. Thus, constitutive activation of Src, in combination withαvβ3, is required for invasion by highly metastatic melanoma cells.

Elevated expression of PKC has been shown to be associated with metastatic melanoma 2122. Highly metastatic melanoma cells may also require PKC for αvβ3-mediated invasion. The two highly invasive cell lines, C8161.9 and A375, expressed high levels of both PKCα and PKCδ compared to the less invasive melanoma cell line M14 and normal cultured melanocytes (Figure 2C). PKC activity was inhibited either by treatment with the PKC-specific inhibitor bisindolylmaleimide, or down regulation of PKC expression by long term treatment with TPA. Under these conditions, αvβ3-mediated invasion of C8161.9 cells was inhibited 5–10 fold (Figure 2D). Thus, PKC activity is required for invasion of C8161.9 cells.

Adhesion of normal cultured melanocytes or M14 melanoma cells to VN induced the formation of multiple stress fibers and large focal adhesion complexes from which the stress fibers emanated (Figure 3A). However, adhesion of C8161.9 cells to VN failed to induce stress fiber formation and the number and size of focal adhesions were dramatically reduced. Since RhoA has been shown to regulate the formation of stress fibers and focal adhesions, RhoA signaling may be impaired in C8161.9 cells. To test this C8161.9 cells were transiently cotransfected with a constitutively active form of RhoA and GFP. Only the three cells expressing green GFP (i.e. transfected with and expressing constitutively active RhoA Q61L) displayed prominent stress fibers (Figure 3B). The surrounding cells (asterisks) not expressing GFP and RhoA Q61L do not have prominent stress fibers. Src has been shown to be a negative regulator of RhoA activity 40. Therefore, it is possible that the high levels of Src activity in C8161.9 cells may be inhibiting RhoA activity. Indeed, inhibition of Src with the specific inhibitor SU6656 or over expression of the K295M Src dominant interfering mutant was sufficient to induce stress fibers and increase focal adhesion formation (Figure 3C). Thus, the aberrant stress fiber and focal adhesion formation in C8161.9 cells plated on VN is likely due to the constitutively high level of active Src, which effectively down regulates RhoA activity.

To further verify that Src is regulating RhoA activity, RhoA activity was measured in C8161.9 cells plated on VN using GST-RBD pull-down assays. Inhibition of Src activity with the Src-specific inhibitor SU6656 resulted in a 2.1-fold increase in RhoA activation after plating on VN (Figure 4A). Similarly, expression of the dominant interfering mutant of Src (SrcK295M) increased RhoA activity 2.5-fold compared to control cells plated on VN (Figure 4B). Conversely, over expression of wild type Src decreased RhoA activation 1.4-fold. The Rho-specific GAP, p190RhoGAP, is known to be activated by adhesion to matrix and its activity is regulated by Src-dependent tyrosine phosphorylation 414243. Activation of p190RhoGAP by active Src in C8161.9 cells could be responsible for the inhibition of RhoA activity following adhesion to matrix. Treatment of C8161.9 cells with the Src inhibitor SU6656 or expression of mutant Src blocked VN-induced tyrosine phosphorylation of p190RhoGAP (Figure 4C, D). All together these data demonstrate that αvβ3-mediated adhesion of highly metastatic melanoma cells to its extracellular matrix substrate VN results in Src-dependent phosphorylation and up-regulation of p190RhoGAP activity, a Src-dependent decrease in RhoA activity and stress fiber and focal adhesion formation, and a Src-dependent increase in αvβ3-mediated invasion. Thus, the elevated Src activity in the highly invasive metastatic melanoma cells is required for effective αvβ3-mediated invasion and functions in part to limit RhoA activity, which may favor a more migratory phenotype.

Previous studies have demonstrated that the relative differences in RhoA and Rac activity in cells reflects the net migratory phenotype 44. Cells with high Rac, but low RhoA, are more migratory, while cells with high RhoA, but low Rac, are less migratory. Given that C8161.9 cells have low RhoA activity and their adhesion and actin structures resemble those seen in cells with high Rac activity, elevated Rac activity may account in part for the increased invasiveness of C8161.9 cells. The total levels of Rac protein and Rac activity were measured and compared between C8161.9 and M14 cells. Using cell lysates containing comparable levels of Rac expression and normalizing for total Rac levels, 2.5–3-fold more Rac activity is present in C8161.9 cells than M14 cells (Figure 5A).

One possibility is that the elevated Src activity in C8161.9 cells not only inhibits RhoA, but also increases Rac activity. To test this hypothesis, Src activity was inhibited in C8161.9 cells and the effects on Rac activity monitored. Surprisingly, inhibition of Src activity with the dominant interfering Src mutant (Figure 5B) or treatment with Src inhibitors (not shown) had no effect on Rac activity. Downstream signaling to the Rac effector PAK was similarly unaffected by Src inhibition (data not shown). Thus the elevated Rac activity in C8161.9 cells is not due to elevated Src activity.

PKC activity was inhibited in C8161.9 cells, either by long term treatment with TPA to inhibit PKC expression (Figure 5C) or with the catalytic inhibitor bisindolylmaleimide, and the effect on Rac activation after plating on VN was analyzed. Inhibition of PKC resulted in reduced Rac activity (Figure 5D) and reduced activation of the Rac effector PAK (Figure 5E). To determine which PKC isoform, PKCα or PKCδ, was responsible for regulating Rac activity an siRNA approach was used. Transient transfection of siRNAs specific to PKCα or PKCδ specifically inhibited their respective expression, while the scrambled siRNA had no effect on either PKC isoform (Figure 5C). Inhibition of expression was maximal at 72 hours and remained sustained for up to at least 144 hours (data not shown). None of the siRNAs had any effect on PKCε. Inhibition of PKCα, but not PKCδ, expression in C8161.9 cells by PKC siRNAs dramatically reduced Rac activity (Figure 5D). Loss of PKCα, but not PKCδ, also reduced VN-induced PAK activity (Figure 5E). Thus, PKCα is required for activation of Rac and its downstream effector PAK in C8161.9 cells.

Interestingly, Rac activity was often elevated in suspension cells (Figure 5B, D), while PAK activity was not (Figure 5E) suggesting an apparent disconnect between these two molecules. However, previous studies have demonstrated that the ability of Rac to signal to PAK is dependent on cell adhesion 37. The ability of Rac to activate PAK is dependent on membrane translocation of Rac, which requires an integrin-dependent adhesion event. Thus, even though Rac is sometimes highly active in suspended C8161.9 cells, it cannot complete the circuit to signal to PAK in the absence of integrin engagement.

Inhibition of Src was sufficient to induce stress fibers and increase focal adhesion formation in C8161.9 cells adherent to VN (see Figure 3C). Inhibition of PKCα expression by siRNA increased focal adhesion formation, but only partially restored stress fibers – mostly around the nucleus (Figure 6A). Inhibition of PKCδ primarily increased stress fibers, with some increase in the size, but not number of focal adhesions. However, neither response was as strong as seen with inhibition of Src (see Figure 3C). Thus PKCα appears to more strongly influence focal adhesions, while PKCδ primarily targets stress fibers.

To delineate which PKC isoform is required for αvβ3-mediated melanoma invasion, C8161.9 cells were transfected with scrambled or PKC isoform-specific siRNAs to inhibit PKC expression (Figure 6B) and tested for their ability to invade VN-enriched Matrigel. Loss of PKCδ resulted in a slight (20%), but consistently reproducible decrease in invasion (Figure 6C). Loss of PKCα resulted in a 2-fold decrease in Matrigel invasion (Figure 6C).

Src, PKCα, or PKCδ was over expressed in M14 cells to determine if Src and PKC are sufficient to convert αvβ3-expressing M14 cells to an invasive phenotype (Figure 6D). Over expression of PKCα or PKCδ alone was not sufficient to induce M14 cell invasion of VN-enriched Matrigel (Figure 6E). However, over expression of Src alone or both PKCα and PKCδ increased M14 invasiveness 1.6- and 2-fold respectively.