Examining RNA Transcript Changes in Stem Cells Undergoing Differentiation, Using Short and Long Reads Sequencing Technologies

Differentiation of embryonic stem (ES) cells is among the most dynamic processes in biology. Mouse ES cells, derived from the inner cell mass of mouse blastocysts, are pluripotent cells that have the capacity to differentiate into cell types of all three primary germ layers. Retinoic acid (RA), the active form of vitamin A, induces differentiation in vitro of ES cells into neurons.

Towards the aim of identifying changes in alternative splicing in ES differentiation, we collected RNA before and after differentiation with RA, and sequenced it via Illumina short reads (RNA-Seq). We also used long-read sequencing methodology in order to directly generate individual reads that could span the full length of transcripts. Moreover, we sequenced both cDNA and direct RNA using Oxford Nanopore Technology (ONT). The direct RNA approach offers exciting opportunities, whereby individual poly-adenylated RNAs are sequenced directly, without the recoding and amplification biases inherent in other sequencing methodologies.

Our findings show that the correlation between the technologies is high (Spearman correlation minimum of 0.83) when analyzing gene expression. However, analysis on the genes exons expression level, results in a lower correlation between the Illumina and ONT sequencing technologies (Spearman correlation minimum of 0.5). Exploring the data, we find that many exons expression is detected in a platform-dependent fashion. Further, we demonstrate that ONT long reads gives us an enhanced ability to detect exons that are differentially expressed upon cell differentiation.