Abstract 4856: Transcriptome sequencing of circulating tumor cells reveals their heterogeneity

Abstract

Abstract We have developed a novel sample preparation technology for complete purification of circulating tumor cells (CTCs). The target cells are magnetically labeled with anti-EpCAM particles and then immunostained with anti-EpCAM antibodies. CTCs have been selectively captured by applying absorption magnetic forces and desorbing fluidic shear force to the cells (MagSweeper technology). CTCs, defined as EpCAM+ CD45- cells are then picked individually for nucleic acid analysis. We have utilized this platform for isolation of CTCs from prostate cancer patients. From each case, 7.5ml blood sample is processed within 24hr of collection. The purified CTCs have been validated to be EpCAM+ CD45- by immunostaining and qPCR analysis (91%± 6% specificity). So far we have isolated CTCs from 14 out of 35 cases. The CTC counts isolated by MagSweeper have correlated well with the counts obtained from the same cases using FDA-cleared CellSearch system. We are also developing new genomic technologies for whole-genome expression profiling of single CTCs. The purified CTCs are lysed under control conditions, and then processed with our whole transcriptome amplification (WTA) protocol. The amplified cDNA are used for SBS sequencing library preparation. Our WTA protocol has been validated through mRNA-sequencing of cultured single prostate LNCaP and bladder T24 cells. We found that more than 66% of all genes are reproducibly sequenced and the expression profile of single LNCaP and T24 cells correlated well within each cell line (R⁁2=0.75 for LNCaP and R⁁2=0.87 for T24 cells). We have sequenced the whole transcriptome of 7 CTCs and one leukocyte cell isolated from blood sample of a metastatic prostate cancer patient. We found that 7/7 CTC cells express KLK3, AR and EpCAM, 6/7 express CD24 and AMACR, 5/7 express MYC and 0/7 express CD45. On the other hand, the WBC does not express any of these genes except CD45. At present we are analyzing the data for the presence of fusion genes. So far we have not detected the TMPRSS2-ERG fusion gene in any of the CTCs analyzed; however, the search for looking into other known fusion genes is currently under progress. In addition, we clustered CTCs with LNCaP and T24 single cells. We found that the 5/7 CTC cells cluster together, and are different from LNcap or T24 clusters. 2/7 CTCs show different expression profile, demonstrating expected heterogeneity between CTCs. We have introduced a new end-to-end integrated workflow for CTC genomic analysis. The workflow has been successfully tested in isolation and mRNA-seq profiling of CTCs from prostate cancer patient samples. This novel workflow will open up new opportunities in studying CTC biology. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 4856. doi:10.1158/1538-7445.AM2011-4856