Abstract 1930: New technology of RNA profiling at single cell level leads to the discovery of tumor heterogeneity in grade 1 endometrial cancer

Background: Due to high heterogeneity, a robust single-cell level profiling method is in great demand to analyze genomic and transcriptomic changes during cancer progression. However, existing amplification methods exhibit variable amplification efficiencies depending on the nature of templates and experimental conditions, which can generate significantly skewed representation of sequence fragments in the final profiling. This issue is much severer in RNA profiling due to lack of proper normalization controls. Methods: Tissue blocks from grade 1 endometrial cancer patients, as well as normal individuals were obtained from the Tumor Bank at The University of Texas MD Anderson Cancer center. Single glands were microdissected from either normal or tumor tissue blocks. To accurately profile the transcriptome of each gland, we introduced modified polyT primers that harbor a random 10-nucleotide (10N) stretch, a definitive 7-nucleotide (7X) sample barcode, and a universal primer binding site. As a result, each newly synthesized cDNA contained a unique 10N stretch as its identity tag, a definitive 7X stretch as its sample barcode, and a universal primer binding site for massive amplification. The amplified cDNAs were subject to library construction and Illumina Hiseq 2000 sequencing. 7X sample barcodes were used for multiplexing, and 10N identity tags were used to calculate the expression level. Using this novel technology, the final transcriptome profiling faithfully revealed the mRNA expression level of each gene in spite of massive amplification bias. We further performed two-sample t-test to assess the differential gene expression between normal and tumor tissues on a gene-by-gene basis. We used the False-Discovery-Rate (FDR) to adjust the original p values using the Benjamini-Hochberg method. Results: We identified 2991 candidate genes as differentially expressed between normal and grade 1 tumor endometrial glands, with FDR adjusted p value Conclusions: This novel RNA profiling technology provided reasonable sensitivity and reliability with a small number of cells. The identified differential degree of heterogeneity in early endometrial carcinomas can be evaluated as a prognostic factor and facilitate personalized therapeutic decisions. Our profiling technique can also be broadly applied in other clinical fields that demand a higher resolution for cancer diagnosis and prognosis. Citation Format: Zhu Zhu, Samuel C. Mok, Xiaoping Su, Ying Yuan, Karen H. Lu. New technology of RNA profiling at single cell level leads to the discovery of tumor heterogeneity in grade 1 endometrial cancer. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1930. doi:10.1158/1538-7445.AM2015-1930