Breeding Low Cyanide Sorghum Varieties through Metabolomic and Transcriptomic Analysis

Introduction. Sorghum (Sorghum bicolor(L.) Moench) is an important cereal crop for human and animal feed. However, sorghum contains cyanogenic glucoside which is degraded into toxic cyanide causing intoxication when it takes over a certain amount. Consequently, selecting individual lines having low cyanide content is crucial in sorghum breeding. Our long-term goal is to develop low cyanide sorghum varieties to prevent intoxication, which would be very useful for the nutrition of humans and animals.

Materials and Methods. Sorghum seeds of various lines were obtained from National Agrobiodiversity Center in Rural Development Administration (reference line BTx623, three local varieties), the University of Georgia (14 wild relatives), and Wild plant resources seed bank in Korea University (one wild species). Sorghum plants were grown in plastic greenhouse under natural condition. Samples were individually harvested at seedling stage, vegetative phase (right before heading), and reproductive phase (ripening). Using collected samples, we plan to analyze metabolites by HPLC and to perform RNA-sequencing using RNAs isolated from the samples.

Results and Discussions. Sorghum seeds of 19 different lines were obtained. After growing, leaf samples were harvested at different developmental stages. Prior to metabolome analysis using HPLC system, experimental condition was set to verify three metabolites, dhurrin, p-hydroxybenzaldehyde (HB), and p-hydroxyphenylacetic acid (HPA), simultaneously. Standard compounds were separated well by specific experimental condition. Under this condition, three metabolites were quantified in each line by HPLC system. We observed that the metabolite contents were changed differently by developmental stages and lines. We clustered these results into five groups as patterns of their contents by developmental stages. Almost lines showed that HB content was very high at seedling stage and decreased rapidly at vegetative phase. Interestingly, the patterns of dhurrin content were very different among clusters. However, HPA content was maintained low levels by developmental stages in almost lines. The results would demonstrate how dhurrin and alternative degradation pathways are different in each line. Furthermore, we plan to perform high-throughput transcriptome analysis by RNA-sequencing using the same samples used for metabolome analysis. We expect that expression patterns of dhurrin biosynthesis- and degradation-related genes may be different in each line. Through metabolome and transcriptome analysis for all different lines, we would reveal relationship between metabolite contents and expression patterns of dhurrin biosynthesis- and degradation-related genes. These results may provide important information on the promotion of low cyanide sorghum varieties.

Acknowledgements. This work was carried out with the support of "Cooperative Research Program for Agriculture Science & Technology Development (Project No. PJ011797)" Rural Development Administration, Republic of Korea.