Whole Transcriptome Profiling of Pure Cell Populations Isolated from Minute Prostate Needle Core Biopsies.
Joan Sweet, Natalie Stickle, Carl Virtanen, Lina Qi, Susan Done, Theodore Brown, Karen Hersey, Neil Fleshner, Neil Winegarden. University Health Network, Toronto, ON, Canada; Mt.Sinai Hospital, Toronto, ON, Canada
Background: Prostate cancer (PCa) research has focused on the malignant prostate epithelium. However, increasing evidence has shifted focus to the tumor stroma as key in the growth and metastasis of PCa. For a more complete understanding of the role of both epithelium and stroma in the development and progression of PCa, it is necessary to obtain pure cell populations through techniques such as laser capture microdissection (LCM). Typically however, in order to study cells at the whole transcriptome level, tens to hundreds of thousands of cells must be isolated to provide sufficient material. When utilising needle core biopsies, the ability to obtain sufficient material is limited and often mulitple cores are necessary leading to increased heterogeneity of the cell population. We present a novel method of RNA amplification that requires very few (less than 1000) cells and no RNA purification, yet allows robust full transcriptome profiling.
Design: Needle core biopsies were obtained intraoperatively at the time of prostatectomy to obtain tumor and non-tumorous samples from 5 cases. The control group consisted of 5 cases of benign prostate tissue taken in an identical manner at cystoprostatectomy for bladder cancer with no prostate cancer. The tissue was snap frozen in OCT medium. Sections were obtained in RNAse free conditions and stained with hematoxylin. LCM was performed using the Arcturus Pixell LCM system to separately collect epithelial cells and stroma for a total of 30 samples. For each sample, cells were lysed and mRNA was reverse transcribed into cDNA. Using a unique method of multi-stage PCR amplification, double-stranded product was produced in such a way as to maintain transcript abundance relationships. DNA product was dye labeled and hybridized to Agilent 44k whole genome microarrays.
Results: Data was filtered to show only probes that were in the upper 80th percentile of the distribution of intensities in 75% of any of the 1 of 6 above categories. Using a two-way ANOVA we were able to find 571 genes that differed in expression across all 3 tissue types between stroma and epithelium.
Conclusions: The profiling results reveal epithelial and stromal specific gene signatures that may be useful in revealing the complex nature of the interaction between these two tissue components in prostate cancer. In addition, the technique described here will have value for analysis of any tissue samples for which only small amounts are obtainable.
Wednesday, March 2, 2011 9:30 AM
Poster Session V # 217, Wednesday Morning