Deep digital gene expression profiling during early and late tension wood induction in Eucalyptus trees

Abstract

Genetic engineering of superior wood properties and exploiting natural genetic variation found within commercially important trees, such as Eucalyptus spp., promise to increase cellulose biomass production. It is therefore essential to understand the molecular genetics of wood formation. Digital Gene Expression (DGE) profiling is adept in not only assessing the expression level of genes transcriptome-wide, but also in characterising alternative splice forms of transcripts and identifying novel transcripts. Tension wood is a specialised type of wood which functions in the response to mechanical stress in trees and is formed on the upper side of a branch or a bent stem. The characteristics of tension wood differ from normal wood by increased cellulose and xyloglucan content and decreased lignin and xylan content. During tension wood formation, transcriptome-wide changes in the expression of genes involved in secondary cell wall formation underlie changes in cell wall composition. Most notably is an increase in fasciclin-like arabinogalactan protein (FLA) and xyloglucan endotransglucosylase (XTH) and a decrease in lignin biosynthesis gene expression. Differential expression patterns are shown by cellulose synthase (CesA) genes, which have been found to be either up- or down-regulated during tension wood formation. No previous study has profiled gene expression during early as well as late tension wood formation. The aim of this M.Sc study was to identify genes that are differentially expressed during early tension wood induction and late tension wood formation in the immature xylem tissues of Eucalyptus grandis x Eucalyptus urophylla hybrid trees. DGE profiling is a transcriptome-wide expression profiling technique based on ultra-high throughput second generation DNA sequencing technology. The processing, analysis and interpretation of DGE data has not yet been standardised. To address this problem, a case study was performed of DGE data mapping to seven well characterised Eucalyptus grandis CesA (EgCesA) genes. The DGE data processing guidelines developed based on this case study produced EgCesA expression profiles in normal wood that were comparable to the profiles of these genes determined with other technologies. A possible alternative splice variant occurring during tension wood formation was identified for the secondary cell wall gene EgCesA3. However future work is needed for the validation of this observation. Early tension wood induction and late tension wood formation was investigated by sampling differentiating xylem from ramets of a Eucalyptus grandis x Eucalyptus urophylla clone induced to form tension wood for 6 hours, 24 hours, 1 week, 2 weeks and 6 months. Up to 2,654 transcripts were found to be significantly differentially expressed during tension wood formation. FLA transcripts were the highest expressed transcripts and were, along with XTH genes, highly up-regulated in early and late tension wood formation. Genes differentially regulated during early tension wood formation reflected a general stress response and hormone signalling pathways. Late tension wood formation was marked by the differential regulation of secondary cell wall biosynthetic genes, which reflected the chemical composition of tension wood. Two secondary cell wall CesA genes were significantly up-regulated, while genes involved in lignin and xylan biosynthesis were significantly down-regulated. Observations suggest that the eucalypt trees used in this study formed tension wood to stabilise the bent stem, while apical dominance was transferred to new side branches which showed signs of extra secondary growth.