Transcriptional markers enable identification of rye-grass (Lolium sp.) plants with non-target-site-based resistance to herbicides inhibiting acetolactate-synthase.

Molecular detection of herbicide non-target-site-based resistance (NTSR) classically requires extensively validated NTSR genes. We assessed the feasibility of predicting NTSR phenotypes using expression data of NTSR transcriptional markers, i.e., transcripts which expression levels are statistically correlated to NTSR. Markers were sought by comparative RNA-Seq analysis of untreated NTSR or sensitive plants from four rye-grass populations followed by expression quantification in 299 individual plants with characterised sensitivity to two acetolactate-synthase-inhibiting herbicides. Multivariate analyses were implemented to predict NTSR using combined marker expression data. Nineteen markers (four cytochromes P450, four glutathione-S-transferases, three glycosyltransferases, two ABC transporters, two hydrolases, one aldolase, one peptidase, one transferase and one esterase) expressed significantly higher in NTSR plants were identified. Expression was highest in the most resistant plants. Some markers appeared co-regulated. Combined marker expression data enabled prediction of NTSR phenotypes in individual plants or of resistant plant frequencies in populations. Thus, NTSR detection based on transcriptional markers proved feasible. Accuracy can be improved by identifying additional markers, especially markers associated to NTSR regulation. Additionally, our data suggest that NTSR mechanisms emerged in different populations via redundant evolution, and that NTSR can evolve by selection for higher constitutive expression of whole herbicide-response pathways.