243 GENE EXPRESSION PROFILING BY NEXT-GENERATION SEQUENCING OF cDNA SAMPLES FROM INNER CELL MASS AND TROPHECTODERM OF A SINGLE-DAY 12 PORCINE EMBRYO

Abstract

Traditional microarray approaches to gene expression profiling often require RNA or cDNA amplification when working with extremely small or valuable tissue samples.This process is generally viewed as being undesirable because there is potential for bias to be introduced during amplification. Very recently, the so-called next-generation sequencing technologies were adapted for use in global gene expression profiling. Herein we report our efforts to apply these sequencing technologies to assess relative transcript abundances in pre-implantation-stage porcine embryos, without additional nucleic acid amplification before sequencing. As a proof-of-principle experiment, we have isolated total RNA from the embryonic disc (inner cell mass; ICM) and a small piece of trophectoderm (TE) from a Day 12 in vivo-produced embryo, which were estimated to be composed of 500 to 1000 cells each. The RNA was reverse transcribed using oligo-dT priming followed by second-strand cDNA synthesis. The double-stranded cDNA was then randomly sheared by sonication, and 10 ng of double-stranded cDNA fragments was used for sample preparation before sequencing. Prepared cDNA fragments (at 7 picomolar concentrations) were submitted for sequencing using the Illumina/Solexa platform as recommended. The millions of short (36 bp) reads generated by Illumina sequencing for each sample were then aligned to the swine UniGene database from NCBI, allowing for zero or one mismatches. Relative transcript abundances between cell types were profiled by considering the read counts for a given UniGene member as a percentage of the total number of reads generated for each cell type. It was demonstrated that approximately 11 000 and 9000 UniGene members were represented by a normalized average of 5 or more short reads per lane (0.001% of the total) in the ICM and TE samples, respectively. As expected, pluripotency factors, chromatin remodeling components, and cell-cell communication molecules were overrepresented in the ICM sample as compared with the TE sample. Conversely, epithelial determinants, ion transporters, and components of the steroid biosynthesis pathways were more abundant in the TE sample than in the ICM sample. Relative abundances of representative transcripts in these samples were verified by quantitative RT-PCR. In conclusion, we demonstrate the utility of next-generation sequencing technologies for gene expression profiling using even minute tissue samples and show that such analyses are possible even in species without a sequenced genome.