Abstract:
Gray leaf spot (GLS) disease in maize, caused by the fungus Cercospora zeina, is a
threat to maize production globally. Understanding the molecular basis for quantitative
resistance to GLS is therefore important for food security. We developed a de novo
assembly pipeline to identify candidate maize resistance genes. Near-isogenic maize
lines with and without a QTL for GLS resistance on chromosome 10 from inbred
CML444 were produced in the inbred B73 background. The B73-QTL line showed a
20% reduction in GLS disease symptoms compared to B73 in the field (p = 0.01). B73-
QTL leaf samples from this field experiment conducted under GLS disease pressure
were RNA sequenced. The reads that did not map to the B73 or C. zeina genomes
were expected to contain novel defense genes and were de novo assembled. A total
of 141 protein-coding sequences with B73-like or plant annotations were identified
from the B73-QTL plants exposed to C. zeina. To determine whether candidate gene
expression was induced by C. zeina, the RNAseq reads from C. zeina-challenged
and control leaves were mapped to a master assembly of all of the B73-QTL reads,
and differential gene expression analysis was conducted. Combining results from both
bioinformatics approaches led to the identification of a likely candidate gene, which was
a novel allele of a lectin receptor-like kinase named L-RLK-CML that (i) was induced
by C. zeina, (ii) was positioned in the QTL region, and (iii) had functional domains for
pathogen perception and defense signal transduction. The 817AA L-RLK-CML protein
had 53 amino acid differences from its 818AA counterpart in B73. A second “B73-
like” allele of L-RLK was expressed at a low level in B73-QTL. Gene copy-specific
RT-qPCR confirmed that the l-rlk-cml transcript was the major product induced fourfold
by C. zeina. Several other expressed defense-related candidates were identified, including a wall-associated kinase, two glutathione s-transferases, a chitinase, a glucan beta-glucosidase, a plasmodesmata callose-binding protein, several other receptorlike
kinases, and components of calcium signaling, vesicular trafficking, and ethylene
biosynthesis. This work presents a bioinformatics protocol for gene discovery from de
novo assembled transcriptomes and identifies candidate quantitative resistance genes.
