Unknown,Transcriptomics,Genomics,Proteomics

Dataset Information

Suppression and Activation of the Malignant Phenotype by Extracellular Matrix in Xenograft Models of Bladder Cancer: A Model for Tumor Cell Dormancy

ABSTRACT: A major problem in cancer research is the lack of a tractable model for delayed metastasis. Herein we show that cancer cells suppressed by SISgel, a gel-forming normal ECM material derived from Small Intestine Submucosa, in flank xenografts show properties of suppression and re-activation that are very similar to normal delayed metastasis and suggest they can serve as a novel model for developing therapeutics to target micrometastases or suppressed cancer cells. Co-injection with SISgel suppressed the malignant phenotype of highly invasive J82 and T24 bladder cancer cells and highly metastatic JB-V cells in flank xenografts. Cells could remain viable up to 120 days without forming tumors and appeared much more highly differentiated and less atypical than tumors from cells co-injected with Matrigel. In 40% of SISgel xenografts, growth resumed in the malignant phenotype after a period of suppression or dormancy for at least 30 days and was more likely with implantation of 3 million cells or more cells. Ordinary Type I collagen did not suppress malignant growth, and tumors developed about as well with collagen as with Matrigel. A clear signal in gene expression over different cell lines was not seen, but in contrast, Reverse Phase Protein Analysis of 250 proteins across 4 cell lines identified a clear signal at the protein level involving Integrin Linked Kinase (ILK) signaling that was confirmed by an ILK inhibitor. We suggest that cancer cells suppressed on SISgel could serve as a model for dormancy and re-awakening to develop therapeutic targets for micrometastases. Earlier we demonstrated that the phenotype of bladder cancer cells was radically different in 3-dimensional organotypic culture when grown on a normal extracellular matrix preparation (SISgel) as compared to that observed on a cancer-modulated permissive extracellular matrix preparation (Matrigel). SISgel is a gel-forming material derived from acellular small intestine submucosa, whereas Matrigel is a basement membrane preparation obtained from a mouse sarcoma. On Matrigel the bladder cancer cells recapitulated the phenotype reported for the original tumor; in sharp contrast, most of the malignant properties were lost when the cells were grown on SISgel. Cell lines derived from papillomas formed a layered structure reminiscent of normal urothelium, whereas cell lines derived from higher grade tumors formed a noninvasive layer of cells. These findings suggested that growth of cancer cells on normal ECM could provide a model to investigate the phenomenon of suppression of malignancy by normal ECM in metastasis and recurrence. In this study we explored whether the phenotypic suppression seen in organotypic culture of bladder cancer cells on SISgel also is observed in vivo. Positive findings support the use of SISgel as a model for investigations of the dormant or suppressed tumor cell phenotype and of mechanisms by which the normal ECM exerts an inhibitory influence on tumorigenesis and metastasis. The findings strongly suggest that interactions of cancer cells with normal ECM play an important role in recurrence and metastasis and further suggest that targeting suppressed cells could represent a heretofore unexploited point of vulnerability in cancer therapy.

ORGANISM(S): Homo sapiens

SUBMITTER: Mikhail Dozmorov

PROVIDER: E-GEOD-35947 | biostudies-arrayexpress |

REPOSITORIES: biostudies-arrayexpress

ACCESS DATA

Similar Datasets

Project description:This study was undertaken to identify how gene expression of the RT4 bladder papilloma cells respond to extracellular matrics (ECM) Matrigel. Gene expression profile of RT4 cells grown on Plastic was compared with expression profiles of RT4 cells grown on Matriges over 12 hours-9 days timecourse. Total 5881 hypervariable (HV) genes were found to vary significantly over the time points, indicating a response to matrix change. Correlational clustering showed these HV genes fell into distinct families, predominantly genes which became highly expressed on ECM. Another cluster show genes expressed at 12 hours after grown on Matrigel and third cluster show genes whose expression decreased below noise level on Matrigel. This indicate that processes responsible for cell adaptation to ECM happen within first 12 hours and the second cluster form last response of those active processes. Over 1-9 days of cell grows on Matrigel those genes are constantly expressed. Cell signaling and cellular growth and proliferation accounted for major functions for those genes. G-protein signaling, immune response and NF-kB signaling accounted for most dramatically affected canonical pathways. Genes sharing similar ontologies also share several overrepresented transcription regulatory elements confirmed by independent experiment on Panomics platform.Thus, rapid response for ECM orchestrated by several transcription factors was observed. Identified canonical pathways indicate possible targets for manipulating cell phenotype on ECM. Keywords: extracellular matrix, gene expression, RT4, Matrigel, microarray, time course Three dimensional gel cultures with Matrigel were performed by layering 0.8 mls of ice cold Matrigel onto polyethylene terephthalate membranes of 6-well cell culture inserts (Falcon, Becton-Dickinson Labware, Franklin Lakes, NJ). Gels were solidified at 37Âº C. RT4 cells were trypsinized, brought up on 20 mls McCoys media with 10% Fetal Calf Serum (FCS). Cells were centrifuged then brought up in McCoys media with 10% FCS to a concentration of 500,000 cells/250Âµl. 250Âµl added on top of each solidified matrigel disc and 2mls of McCoys with 10% FCS layered underneath each culture insert. Cells were maintained for 10 days and media replenished underneath the insert twice a week. Cells were harvested at 12 hrs, 24hrs, 2 days, 3 days, 4days, 6 days, 7 days, 8 days, and 9 days of growth in culture using Matrisperse (Falcon, Becton-Dickinson Labware, Franklin Lakes, NJ) and RNA collected. Data were normalized as was described previously in Dozmorov I et al., Nucleic Acids Res. 2004 Oct 28;32(19). In general, intensities were converted to expression and normalized to low-expressed genes, which provide possible technical noise level. The arrays are then adjusted to each other by robust linear regression and this Log10 transformed data presented for each sample. Group of low-expressed genes is selected by normal distribution fitting. This group provides internal standard of measurement noise (ISMN). Genes above 5SD from mean of ISMN were selected as expressed

Dataset Information

Suppression and Activation of the Malignant Phenotype by Extracellular Matrix in Xenograft Models of Bladder Cancer: A Model for Tumor Cell Dormancy

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure