Transcription Regulation and Plant Diversity

Abstract

Transcript abundance measured for any gene on a microarray can be considered as a quantitative trait. If the transcription profile of a sufficient number of individuals from a segregating progeny is generated by microarrays, it allows mapping of genomic regions regulating variation in transcript abundance using traditional methods of QTL analysis. We generated transcript level profiles of wood forming tissue (differentiating xylem) collected from 91 individuals from a E. grandis x E. globulus F1 hybrid x E. grandis backcross population, using microarrays containing 2608 cDNAs. Least-square means estimates of transcript abundance were generated for each individual and cDNA, and mapped as QTLs in two single-tree linkage maps (hybrid paternal and E. grandis maternal) using composite interval mapping. QTLs were identified for 811 genes in the Eucalyptus hybrid map, displaying in most cases a simple genetic architecture, with a single QTL controlling up to 70% of the transcript level variation. A more complex genetic architecture was detected in one third of the genes, where up to five QTLs could be identified for three genes. QTL hotspots were identified in both maps, typically for genes encoding several enzymes of specific metabolic pathways, suggesting coordinated genetic regulation. Transcript level QTLs were co-localized to QTLs detected previously in this family for wood quality and growth traits and candidate genes were identified by the analysis of correlation between gene expression and phenotypic variation.