Treatment in the current study implies the use of different matrixes for cell culture.
Growth protocol
Growth of cells on ECM. Gel matrices were made by layering either 0.8 ml of ice cold Matrigel (Becton-Dickinson, Bedford, MA), SISgel, or Type I collagen (BD Biosciences, San Jose, CA) onto polyethylene terephthalate membranes of 6-well cell culture inserts (Falcon, Becton-Dickinson Labware, Franklin Lakes, NJ), which were then allowed to gel at 37°C. SISgel was prepared either by solubilized small intestine submucosa provided to us by Cook Biotech (Lafayette, IN) or by material we prepared ourselves from decellularized small intestine submucosa using Cooks protocol (17). Briefly, ground SIS was partially digested with pepsin at pH 2.8 at 4°C for 5 days. This represents an optimization; too little digestion and the material forms clots instead of gelling, and excessive digestion produces material that will not gel. The pepsin was inactivated by raising the pH to 10 and incubating the gel at 4°C for 2 hours, followed by lowering the pH to 4 with 6 N HCl. The product was dialyzed against 10 mM HCl and sterilized with CHCl3, which was then dialyzed out against 10 mM HCl. To form a gel, the material was mixed with 1/10th volume of 10X phosphate-buffered saline and a small amount of Phenol Red to assist with pH adjustment. The pH was adjusted to 7.4 with sterile 1 M NaOH using a visual color standard that was checked against a pH meter. Collagen gels were prepared by mixing rat tail collagen, Type I with 1/10th volume of 10X PBS, adjusting the pH to 7.4 ± 0.1, and then layering 0.8 ml onto transwell inserts as described above. Cells were layered onto either Matrigel, SISgel or collagen at a concentration of 5 ×105 cells/200 µl media, and 2 ml of Dulbecco’s Modified Eagle Medium (DMEM, Life Technologies, Carlsbad, CA) containing 1% penicillin, streptomycin and 10% fetal calf serum were layered beneath the transwell supports in 6-well plates as described (14). Cultures were grown for 6 days with media changes every two days. Cultures were removed from the transwells, fixed in 1% formalin after overlaying with agarose or SISgel to prevent loss of cells on cutting of sections. Sections (5 µm) were stained with hematoxylin and eosin. The role of Integrin Linked Kinase (ILK) signaling was confirmed by adding the ILK inhibitor, QLT0267, (provided by S. Dedhar) to the culture medium of cells grown on Matrigel and SISgel at a concentration of 10 µM and determining the phenotype from stained sections. Cultured cells grown for 3 days were isolated from the matrices as follows. Cells grown on Matrigel were gently washed with Hank’s Balanced Salts Solution, HBSS, (Life Technologies, Carlsbad, CA), then coarsely minced gels were incubated with 1ml of dispsase (BD Biosciences, San Jose, CA) at 37°C until the gel dissolved. Cells were harvested from SISgel or collagen by washing with HBSS then incubating coarsely minced gels with 1ml of Collagenase IV, 450U/ml, (Life Technologies, Carlsbad, CA) at 37°C until the gel dissolved. Liberated cells were centrifuged to a pellet, washed twice with HBSS, then resuspended with DMEM. Cells grown on plastic were trypsinized and resuspended in DMEM. Growth of Cells Flank Xenograft Model. Cells resuspended in 100µl DMEM/flank were mixed with an equal volume of cold SISgel, Matrigel or collagen, type I. The mixture of gel and cells prepared as described in the previous section was immediately injected subcutaneously into either the right or left flank of a 5-week old nude mouse, nu/nu-nuBR, (Charles River Laboratories, Wilmington, MA). This protocol was reviewed by the Institutional Animal Care and Use Committee to ensure compliance with all standards of animal experimentation. Caliper measurements of tumor size were taken every week for the length of the study as were fluorescent images. Labeled cells were visualized with the Lightool’s LT-9900 system with the EGFP filter set of 470nm excitation filter and 515nm viewing filter and captured with a cooled CCD camera (Lightools, Encinitas, CA). The area and intensity of the tumor was measured using Adobe Photoshop by first selecting the tumor area, then counting pixels above a threshold selected to eliminate background, non-tumor areas. The integrated intensity was calculated by multiplying the average intensity of detected pixels by the number of pixels detected.
Extracted molecule
total RNA
Extraction protocol
Total RNA was isolated from J82 and JBV cells grown on Matrigel and SISgel and liberated as described above. Triplicate samples, each from a different six-well transwell, were prepared to use as biological replicates. The RNA was isolated using a filter method with the QuickGene-Mini80 device and reagents (FujiFilm Life Sciences marketed by AutoGen, Holliston, MA). http://www.autogen.com/assets/pdf/White%20Paper%20-%20QuickGene.pdf. RNA integrity was assessed using the Agilent (Agilent Technologies, Santa Clara, CA) microelectrophoresis system. The RIN was >9.0 for all the RNA preparations. The RNAs were converted to sense cDNA and hybridized to Affymetrix GeneChip Human Gene 1.0 ST Arrays (Affymetrix, Inc., according to the manufacturer's instructions. This array has 764,885 total 23-mer probes, which correspond to 28,869 annotated gene level probe sets and covers 99% of the sequences present in the RefSeq database.
Label
biotin
Label protocol
Isolated total RNA was used with GeneChip WT Expression kit (Applied Biosystems) to produce fragmented single-stranded DNA, which was labeled using terminal labeling kit (Affymetrix) to generate biotin-labeled sense targeted DNA probes.
Hybridization protocol
Fragmented and labeled DNA (540 ng) was hybridized following the manufacturer’s directions for 17h at 45C and rotating at 60rpm. Arrays were processed using Genechip Fluidics Station 450 according to recommended protocols (FS450_007) of double-staining and post-hybridization washes using GeneChip hybridization, wash and stain kit (Affymetrix #900720).
Scan protocol
Fluorescent images were captured using gene Array scanner 3000 and GCOS 1.2 software (Affymetrix).
Description
Gene expression from J82 cells grown on Matrigel RMA extracted, Log2
Data processing
Raw microarray data were processed using Affymetrix Expression Console version 1.1 (http://www.affymetrix.com/browse/level_seven_software_products_only.jsp?productId=131414&categoryId=35623#) to extract gene level intensities using the RMA method. The subsequent analysis was performed using the R statistical environment (http://www.R-project.org) and BioConductor packages. Quality control of the data included a comparison of boxplots of gene expression level among the conditions, and hierarchical clustering of condition-specific gene expression profiles to identify potential outliers. All data conform with the Minimal Information About a Microarray Experiment (MIAME) guidelines. Matrix-specific combined gene/protein expression profiles were compared using the limma (Smyth 2004) package. The Benjamini-Hochberg false discovery rate (FDR) correction for multiple testing was applied, along with 2-fold change cutoff. Gene set enrichment analysis (GSEA) method was used to analyze differentially expressed genes ranked by their fold change. Briefly, GSEA walks down the list of pre-ranked genes and calculates enrichment of gene groups in a given functional category. Functional categories containing genes consistently up- or downregulated are further assessed by a permutation method. GSEA assesses enriched categories in either phenotype by starting from either up- or downregulated genes. In this study, we considered enrichment of the genes in KEGG canonical pathways and all gene ontology categories. Further investigation of functional significance of differentially expressed genes/proteins was carried out by Ingenuity Pathway Analysis (IPA; Ingenuity Systems, Redwood City, CA).
