Abstract
Background:
Single cell transcriptome studies have recently advanced to the analysis of millions of cells in a single pipeline. This type of analysis requires expensive, high maintenance platforms, precluding its use in most laboratories. Therefore, there is a need to find innovative ways of using more accessible low throughput methods for clinical application. Understanding the gene expression of an individual cell has clinical applications ranging from improved diagnosis to more precise treatment. We developed a novel single cell transcriptome analysis method and adapted it for ultra-low input mRNA analysis to potentially improve cancer diagnosis.
Methods:
Our single cell transcriptome analysis method allowed us to determine if the detection of single prostatic cells in various backgrounds could identify novel quantitative PCR biomarkers that could be used in urine diagnosis of prostate cancer. We applied this method to detect single cells from the LNCaP and PC3 prostate cancer cell lines in populations of prostatic (LNCaP) and non- prostatic (HeLa) cells. We used up to 29 cells from the HeLa cell line as a non-prostatic background for detection of a single LNCaP and PC3 cells. We also used up to approximately 199 LNCaP cells as a prostatic background to detect a single PC3 cell.
Results:
Our method detected 23 gene isoforms that significantly distinguish (P < 0.05) a single PC3 cell from a background population of approximately 199 LNCaP cells. 161 gene isoforms significantly distinguished (P < 0.05) a single PC3 cell from a background population of 29 HeLa cells. 64 gene isoforms significantly distinguished (P < 0.05) a single LNCaP cell from a background population of 29 HeLa cells. The MAGED1 gene had the highest log fold change in PC3 cells compared to HeLa cells (log2 fold change = 9.3, P < 0.05) and significantly distinguished (P < 0.05) a single PC3 cell from a background of 29 HeLa cells. qPCR analysis confirmed the RNA-seq results for MAGED1 and PSA gene expression patterns in LNCaP and PC3 cells.
Conclusions:
Transcriptome wide single cell analysis identified several genes that indicate the presence of single PC3 and LNCaP cells in background populations of cells. This result supports a single cell sampling approach to highly sensitive diagnostics and prognostics for ultra-low input RNA samples. Our single cell analysis method compares positively with methods described in the literature with respect to genes detected per read.