Selection for seed size has uneven effects on specialized metabolite abundance in oat (Avena sativaL.)

Abstract

Plant breeding strategies to optimize metabolite profiles are necessary to develop health-promoting food crops. In oats (Avena sativa L.), seed metabolites are of interest for their antioxidant properties, yet have not been a direct target of selection in breeding. In a diverse oat germplasm panel spanning a century of breeding, we investigated the degree of variation of these specialized metabolites and how it has been molded by selection for other traits, like yield components. We also ask if these patterns of variation persist in modern breeding pools. Integrating genomic, transcriptomic, metabolomic and phenotypic analyses for three types of seed specialized metabolites – avenanthramides, avenacins, and avenacosides – we found reduced heritable genetic variation in modern germplasm compared to diverse germplasm, in part due to increased seed size associated with more intensive breeding. Specifically, we found that abundance of avenanthramides increases with seed size, but additional variation is attributable to expression of biosynthetic enzymes. In contrast, avenacoside abundance decreases with seed size and plant breeding intensity. In addition, these different specialized metabolites do not share large-effect loci. Overall, we show that increased seed size associated with intensive plant breeding has uneven effects on the oat seed metabolome, but variation also exists independently of seed size to use in plant breeding. This work broadly contributes to our understanding of how plant breeding has influenced plant traits and tradeoffs between traits (like growth and defense) and the genetic bases of these shifts.

Generalizable approaches for genomic prediction of metabolites in plants

“…This population is described by Brzozowski et al. (2021). Briefly, a panel of 235 inbred lines was evaluated in three U.S. Midwest production environments (Minnesota [MN], South Dakota [SD], and Wisconsin [WI]).…”

“…The metabolites were annotated by comparison to an in‐house spectral library RAMSearch (Broeckling et al., 2016) and MSFinder (Tsugawa et al., 2016), and details of measurement and annotation of this dataset are provided in Brzozowski et al. (2021). Campbell et al.…”

“…We used a validation germplasm panel to test the transferability of kernels between populations. This population is described by Brzozowski et al (2021). Briefly, a panel of 235 inbred lines was evaluated in three U.S. Midwest production environments (Minnesota [MN], South Dakota [SD],…”

“…Deregressed BLUPs of the metabolites for the discovery (diverse) germplasm panel are available in the supplementary material of Campbell et al (2021b). Genotype data is as used in Brzozowski et al (2021) and available online (https://datacommons.cyverse. org/browse/iplant/home/shared/GoreLab/dataFromPubs/ Brzozowski_OatMetabolome_2021).…”

“…Genotype data is as used in Brzozowski et al. (2021) and available online (https://datacommons.cyverse.org/browse/iplant/home/shared/GoreLab/dataFromPubs/Brzozowski_OatMetabolome_2021). Deregressed BLUPs and annotations of the metabolites for the validation germplasm panel and the R code for these analyses is available on a public repository (https://github.com/ljbrzozowski/OatMetaboliteGenomicPrediction).…”