Sugarcane yellow leaf (SCYL), caused by the sugarcane yellow leaf virus (SCYLV) is a major disease affecting sugarcane, a leading sugar and energy crop. Despite damages caused by SCYLV, the genetic base of resistance to this virus remains largely unknown. Several methodologies have arisen to identify molecular markers associated with SCYLV resistance, which are crucial for marker-assisted selection and understanding response mechanisms to this virus. We investigated the genetic base of SCYLV resistance using dominant and codominant markers and genotypes of interest for sugarcane breeding. A sugarcane panel inoculated with SCYLV was analyzed for SCYL symptoms, and viral titer was estimated by RT-qPCR. This panel was genotyped with 662 dominant markers and 70,888 SNPs and indels with allele proportion information. We used polyploid-adapted genome-wide association analyses and machine-learning algorithms coupled with feature selection methods to establish marker-trait associations. While each approach identified unique marker sets associated with phenotypes, convergences were observed between them and demonstrated their complementarity. Lastly, we annotated these markers, identifying genes encoding emblematic participants in virus resistance mechanisms and previously unreported candidates involved in viral responses. Our approach could accelerate sugarcane breeding targeting SCYLV resistance and facilitate studies on biological processes leading to this trait.

Bru1 is currently the major gene conferring brown rust resistance in sugarcane, and diagnostic markers are available. A survey for the presence of this gene was conducted on 391 genotypes including Brazilian cultivars, clones and basic germplasm. The efficiency of these markers for identifying resistant cultivars and artificially inoculated basic germplasm was also evaluated. The Bru1 frequency among cultivars (73.5%) suggests this gene is the prevalent source of brown rust resistance in Brazilian sugarcane breeding programmes. Most of the cultivars known to be resistant were positive for Bru1, although other genes for resistance could be present in lines not having Bru1. Only 17.8% of the basic germplasm accessions were positive for the Bru1 gene, and a low correlation between Bru1 diagnostic markers and brown rust severity was observed for basic germplasm accessions. Overall, Bru1 diagnostic markers proved to be efficient identifying resistant cultivars and clones and have potential to be in screening brown rust resistance in Brazilian breeding programmes.

No abstract