Genetic Load and Allelic Imbalance Estimated from Transcriptome Sequencing in Eucalyptus Hybrids

In addition to transcriptome-wide analysis of gene expression, next-generation RNA sequencing (RNA-seq) allows analysis of transcript encoded protein sequence variation underlying quantitative traits in plants. High impact amino acid polymorphisms predicted to cause loss of function (LoF) of protein-coding genes may contribute to the evolution of phenotypes. Such LoF variants are representative of high genetic load that is typically sheltered in highly heterozygous plants such as Eucalyptus trees. Using Illumina RNA-Seq analysis, we analysed the predicted effects of 332,789 SNPs and 16,286 indels on protein function encoded in the xylem transcriptomes of 158 trees from a single F2 interspecific backcross family of E. grandis x E. urophylla. A total of 1,461 LoF variants predicted to inactivate 1,285 genes were identified. Gene inactivation was often accompanied by evidence of regional decay of selective constraints manifesting as a 40% increase in accumulated auxiliary mutations compared to the genome-wide average. This increase in mutagenicity had the greatest impact ~150 bp upstream and downstream of the mutation, creating local “hot spots” of genetic variation. Similar hotspots were found in transcripts disrupted by indels. We further used heterozygous SNP loci to analyse the extent of allele specific expression (ASE) in this population. On average, ~7.9% of the heterozygous loci tested per individual showed evidence of allelic imbalance (P<0.001). 58% of genes exhibiting allelic imbalance had cis-eQTLs, consistent with the expectation for ASE. Interestingly, some genes exhibiting allelic imbalance and LoF variation have previously been implicated in wood formation, and thus warrant further investigation.