Differential expression analysis of cold tolerance adaptation in D. buzzatii by RNA-seq de novo approach

MOREYRA NICOLÁS; MENSCH JULIÁN; HURTADO JUAN; HASSON ESTEBAN

Bahía Blanca

Congreso; VI Argentinean Conference on Bioinformatics and Computational Biology; 2015

BackgroundManytraits and biological processes can be affected by climatic changes and geneticvariation has a major role in the ecological adaptation to thermal changes duringdevelopment steps. Adaptation to these kinds of environments appears toinvolve a set of traits related to reproduction, stress resistance andmetabolic processes. We have performed a RNA-seq analysis in Drosophila buzzatii to investigateprofiles of gene expression changes during female reproductive diapause. Materialsand methodsThe study was conducted by explosingsets of females to three different conditions of temperatures. Two conditionswere controls with little differences between each other (named J25 and C) andthe third condition implied the measure of cold tolerance (named D ? diapausestate). We run all treatments and replicates simultaneously.We used the Trinity package softwareto generate a de novo RNA-seq assembly from ~100 paired-end Illumina reads. Next,we mapped reads and Trinity transcripts to a reference genome (obtained fromFlybase) guided by the Trinity protocol [3]. To analyze expression levels ofthe Trinity-reconstructed transcripts, we aligned the RNA-seq reads dataagainst the contigs (Trinity transcripts), and then estimated the number ofRNA-seq fragments that had mapped to each contig. Taking into account the skewgenerated for the different transcripts proportions in each sample, we appliedthe TMM normalization method to generate the normalized FPKM expression valuesseparately for each sample and therefore to get the differences in RNAcomposition. Thereby, we extracted transcripts that were at least 6.25-folddifferentially expressed at a significance of 0.001(FDR) in any of the samplecomparisons. A transcripts cluster dendrogram was created based on expressionprofiles. From these, we selected only two clusters involving overexpressed andunderexpressed genes in diapause relative to other treatments. ResultsWe found 87 transcriptdifferentially expressed into two clusters. All transcripts wereblasted against genome assembly, annotated genes and non-redundant databases(Flybase and NCBI). The most relevant transcripts exhibited high identity tochorion protein coding genes of D. melanogaster, Vitelline membranecysteine-rich domains, Heat Shock Protein (HSP) coding genes and criticalrespiratory chain genes of the mitochondrial genome (D. mojavensis).ConclusionsAll CDS were consistent with theinitial hypothesis and therefore, for the near future, we propose a validationby the Real-Time PCR (qPCR) method. In addition, examine the data set to findoverexpressed genes across conditions will allow us to determine which regionsof the transcriptome preferentially participate depending on the environmentaltemperature.