Genetic dissection of secondary metabolite production during heartwood formation of Western redcedar (Thuja plicata)

Thuja plicata, commonly known as Western redcedar, is an economically important conifer species on the North-west coast of Canada and the USA. T. plicata lumber is valued for its natural durability. Paradoxically, logged trees are frequently culled because of extensive heartwood (HW) fungal rot, indicating genetic variation in rot resistance. Resistance to rot in the dead HW of trees is attributed primarily to secondary metabolites. T. plicata HW has high levels of fungistatic lignans and fungitoxic monoterpene-derived tropolones, primarily β-thujaplicin. However, it takes 15-20 years of growth before β-thujaplicin is produced at high levels, precluding breeding based on β-thujaplicin levels. If critical genes behind terpenoid biosynthesis were known, variants could be correlated with high levels of corresponding terpenoid, allowing selection for HW rot resistance based on genotype of seedlings after crosses. In a quest for relevant genes, we used RNA-seq to identify genes upregulated in the sap to heart wood transition zone in which secondary metabolites are produced. Among them, we found six putative terpene synthases (TpTSs) and three putative 2-oxoglutarate/Fe(II)-dependent dioxygenases (TpDOXs) that are candidates for synthesis of specific terpenes and for ring expansion of monoterpenes into tropolones, respectively. Enzyme assays using recombinant proteins and terpene substrates showed that one enzyme produced the monoterpene terpinolene and another terpinolene and 3-carene, both of which may be precursors to β-thujaplicin. One terpene synthase produced the sesquiterpene alcohol elemol, previously linked to termite resistance. In addition, we identified three monofunctional diterpene synthases, of which one produced the intermediate normal-copalyl diphosphate, an intermediate for diterpene biosynthesis, the second produced sandaracopimaradiene or syn-stemod-13(17)-ene depending on the substrate, and the third produced levopimaradiene. Enzyme assay of three TpDOXs with a potential substrate did not yield any β-thujaplicin. Taken together, we have identified genes and enzymes that for the first time provides detailed insights into terpene biosynthesis during heartwood formation in T. plicata. Lastly, we show that RNA-seq methodology and identified genes can be used also for field studies of biotic and abiotic stress in T. plicata silviculture.