Sorghum bicolor (L) Moench is a cultivated and self-pollinated species used traditionally for human consumption, livestock feed and forage production. Recently, sorghum has received attention as a bioenergy crop because of its high water use efficiency, yield biomass potential and biomass composition. To develop superior sorghum inbred lines for bioenergy production, it would be beneficial to understand the genetic mechanisms controlling plant architecture traits. Sorghum presents a remarkable diversity in the species bicolor that can be exploited to identify new genes associated with those traits of interest. Brassinosteroids (BR) are steroid hormones that control different aspects of plant growth and development. Brassinosteroid biosynthesis and signaling pathways have been studied in model species such as Arabidopsis, rice and maize, but not in sorghum.

In this study, candidate gene association mapping and a diverse sorghum collection of 315 accessions were used to assess marker-trait associations between BR biosynthesis and signaling genes and eight plant architecture traits. Phenotypic data for plant height, leaf angle, stem circumference, panicle exertion, panicle length, number of tillers, number of internodes and flowering time was collected in three locations in Iowa in 2010 and 2012.

Using 702 genome-wide single nucleotide polymorphisms (SNPs), population structure was determined as five subpopulations that corresponded to three major races Caudatum, Durra and Kafir and the two intermediate races Guinea/bicolor and Guinea Caudatum (west Africa). Coefficient of co-ancestry was also estimated to account for the degree of ancestor relatedness.

A total of 82 SNPs markers out of 263 SNPs present in 19 BR candidate genes were significantly associated with seven phenotypes of interest, confirming the genetic complexity of the traits. Nineteen markers were found associated with more than one phenotype, showing possible pleiotropic effects of BR candidate genes on plant architecture phenotypes. Differences in one or few genes are expected to have an effect on more than one trait. Moreover, the effects of associated markers over phenotypes were consistent with phenotypic correlations, which provide more support to our findings.

This study constitutes the first association analysis for plant architecture traits with BR in sorghum. It contributes to understand the effect of BR candidate genes on sorghum plant architecture and generates molecular tools that can be used to improve sorghum lines for biofuel production.