SUPPLEMENTARY METHODS

Cell Lines and Reagents. MDA-MB-435s cells were obtained from University of North Carolina Tissue Culture Facility (Chapel Hill, NC). Genetic analysis demonstrating that they are identical to melanoma M14 is in Supplementary Figure 7. WM melanoma cell lines were obtained from Dr. Meenhard Herlyn (Wistar Institute, Philadephia, PA), and STR analysis performed by the Herlyn Lab confirms their identity (data not shown). A375 was from Dr. Suyun Huang (M.D. Anderson Cancer Center, Houston, TX). Primary melanocytes, from Dr. John D'Orazio (University of Kentucky, Lexington, KY), were purchased from Cascade Biologics (Portland, OR). STAT3 cDNAs (pRc/CMV-STAT3C, pcDNA-STAT3; Addgene, Cambridge, MA) were donated by Dr. James Darnell. Vector and STAT3C-expressing cells were obtained by transfection, G418 selection (900mg/ml), and pooling expressing clones. Cells expressing GFP-tagged MMP-1 (Origene; Rockville, MD) were obtained by transfection/G418 selection, and flow sorting GFP-positive cells (University of Kentucky Flow Cytometry Facility).

The following antibodies were purchased commercially: c-Abl (K12; immunoprecipitation), Arg, PARP, a-tubulin, (Santa Cruz Biotechnology; Santa Cruz, CA); phospho-Crk/CrkL (Y221/Y207), caspase-3 (Cell Signaling Technology; Danvers, MA); c-Abl (8E9; western blotting), GAPDH, phospho-STAT3 (Y705), STAT3 (BD Biosciences; Chicago, IL), MMP-1 (Neomarkers/ThermoFisher; Pittsburgh, PA); MMP-3, MT1-MMP (Abcam; Cambridge, MA); and b-actin (Sigma, St. Louis, MO). STI571 (a.k.a. imatinib mesylate, Gleevec) and nilotinib were kind gifts of Novartis Pharmaceuticals (Basel, Switzerland). STI571 was dissolved in water at a concentration of 10mM and frozen in aliquots at -80oC, while nilotinib (10mM) was dissolved in DMSO and stored at 4oC in aliquots (in vitro studies). IGF-1 was obtained from Upstate Biotechnology (Charlottesville, VA); matrigel invasion chambers were from BD Biosciences (Chicago, IL); and recombinant, full-length MMP-1 (contains active and inactive forms) and active MMP-3 were from Abcam (Cambridge, MA). Luciferin D was purchased from Regeis (Chicago, IL). SCID-beige mice were from Harlan Laboratories (Indianapolis, IN). c-Abl and Arg constructs were described previously (Plattner et al 1999).

Transfection and RNAi: Cells were transfected with DNA or siRNA using Lipofectamine 2000 according to the manufacturer's instructions (Invitrogen, Carlsbad, CA). siRNAs (Applied Biosystems/Ambion; Carlsbad, CA) were utilized at the following concentrations: Abl #1 (1336; 20nM), Abl #2 (s866; 5nM); Arg #1 (1478, 20nM), Arg #2 (s872, 5nM); MMP-1-s8848 (10nM); MMP-3-s8853 (10nM), MT1-MMP-s8877 (10nM), STAT3-s743 (1nM), scrambled control (concentration equivalent to test siRNA). c-Abl and Arg siRNAs were transfected on two consecutive days to increase silencing efficiency.

GST-fusion proteins. GST and GST-Crk were described previously (Plattner et al 1999). pGEX2T-STAT3 was created by cloning STAT3 from pcDNA-STAT3 by digestion with XhoI, blunting with Klenow, and digesting with BamH1. The insert was cloned into pGEX2T by cutting the vector with EcoR1, blunting with Klenow, and digesting with BamH1. GST proteins were isolated as we described (Plattner et al 1999).

Kinase Assays: Assays were performed as described previously (Plattner et al 1999). Briefly, cells were serum-starved overnight, lysed in a Triton-X-100 based lysis buffer, c-Abl and Arg immunoprecipitated, washed with a series of stringent buffers, incubated in a kinase reaction with 32P-g-ATP, 1μM cold ATP and GST-Crk for 40’ at room temperature. Kinase reactions were run on SDS-PAGE gels, dried, exposed to film, and bands quantitated using a STORM phosphoimager and ImageQuant Software (Molecular Dynamics, GE Heathcare; Piscataway, NJK).

Invasion Assays: Assays were conducted as previously described (Srinivasan and Plattner 2006) Briefly, cells were transfected with siRNAs, and serum-starved overnight. Cells (5X105 in 1% BSA migration medium) were placed in the top well of matrigel invasion chambers (BD Biosciences, Chicago, IL), chemoattractant (IGF-1, 10nM) was placed in the lower chamber in migration medium, and cells were allowed to invade/migrate for 48 hours at 37oC. Cells and matrigel on the upper surface of the membrane were removed, and cells on the undersurface were fixed, stained and all cells on the membranes were counted. Recombinant MMP-1 or MMP-3 (25ng/ml) was added to the top well of chambers for some experiments (Figure 5e,f), and nilotinib (0.5mM) was added to top and bottom chambers for other experiments (Figure 2b).

Tritiated Thymidine Assays: Cells, plated in 12-well dishes to obtain 50% confluence the day after plating, were treated with STI571 or nilotinib for 24h, labeled with tritiated thymidine (5 mCi) for 2h, and harvested as we previously described (Srinivasan et al 2008). For RNAi experiments, siRNA-transfected cells (1X 105) were replated in 12-well dishes 48h after the initial transfection, allowed to attach, serum-starved, labeled with tritiated thymidine, and harvested. Equivalent plating efficiencies were confirmed in replicate wells 8h following replating.

Western Blots: Immunoblotting was performed according to protocols from antibody manufacturers.

Semi-quantitative RT-PCR (Srinivasan and Plattner 2006): Total RNA was isolated from serum-starved cells transfected with siRNAs or treated with STI571 using an RNAeasy kit (Qiagen, Valencia, CA), and digested with DNase I to remove trace DNA (Applied Biosystems, Carlesbad, CA). Five micrograms of DNase-treated RNA was reverse transcribed using Superscript Reverse Transcriptase and random primers (Invitrogen, Carlesbad, CA). cDNA was subjected to PCR using specific primers (below) together with internal b-actin control primers (10mM), MgCl2 (1.5mM), dNTPs, and Taq DNA polymerase (Genscript, Piscataway, NJ). PCR cycling parameters involved 35 cycles of 95oC-1', 55oC-1', 72oC-1'. Aliquots were also taken at cycles 27 and 31 to check for linearity. Scanned photographs were quantified with ImageQuant (Molecular Dynamics, GE Healthcare, Piscataway, NJ) and specific bands normalized to b-actin internal control bands.

c-Abl forward primer, 5'CCTTCATCCCTCTCATATCAACC3',

c-Abl reverse primer, 5'TGGACCACTGCCTGCTGTCGC3'

Arg forward primer, 5'CATCCGTCCATCTGCTCAGAC3'

Arg reverse primer, 5'GGACAGTAGGTCAGCACATTC3'

MMP-1 forward primer, 5'AGCGTGTGACAGTAAGCTAAC3'

MMP-1 reverse primer, 5'TCCTCAGAAAGAGCAGCATCG3'

MMP-3 forward primer, 5'GTTCCGCCTGTCTCAAGATGA3'

MMP-3 reverse primer, 5'ATCCAGCTCGTACCTCATTTCC3'

MT1-MMP forward primer, 5'GCAGGCCGACATCATGATCTTC3'

MT1-MMP reverse primer, 5'TCCTCTCGTAGGCAGTGTTG3'

b-actin forward primer, 5'CCTTCCTGGGCATGGAGTCCT3',

b-actin reverse primer 5'ggagcaatgtctttgatcttc3'.

Paraffin-embedding of mouse lungs: Formalin-fixed lungs were dehydrated in 70% ethanol, 100% ethanol, 1:1 ethanol/xylene, followed by 100% xylene. Lungs were placed in melted paraplast X-tra:xylene (1:1) for 1hr at 54oC followed by room temperature incubation overnight. Paraplast (100%) was then added to the tissue for 1h at 54oC, incubated at room temperature for 2h, paraffin was melted, and mounted at room temperature. The mounts were stored at room temperature overnight followed by dessication for long-term storage, or sectioned on a ThermoShandon Finesse microtome (University of Kentucky Histology Facility).