Abstract
Drought alone or in combination with other stresses forms the major crop production constraint worldwide. Sorghum, one of the most important cereal crops is affected by drought alone or in combination with co-occurring stresses; notwithstanding, sorghum has evolved adaptive responses to combined stresses. Furthermore, an impressive number of sorghum genes have been investigated for drought tolerance. However, the molecular mechanism underling drought response remains poorly understood. We employed a systems biology approach to elucidate regulatory and broad functional features of these genes. Their interaction network would provide insight into understanding the molecular mechanisms of drought tolerance and underpinning signal pathways. Functional analysis was undertaken to determine significantly enriched genesets for pathways involved in drought tolerance. Analysis of distinct pathway cross-talk network was performed and drought-specific subnetwork was extracted. Investigation of various data sources such as gene expression, regulatory pathways, sorghumCyc, sorghum protein-protein interaction (PPI) and Gene Ontology (GO) revealed 14 major drought stress related hub genes (DSRhub genes). Significantly enriched genesets have shown association with various biological processes underlying drought-related responses. Key metabolic pathways were significantly enriched in the drought-related genes. Systematic analysis of pathways cross-talk and gene interaction network revealed major cross-talk pathway modules associated with drought tolerance. Further investigation of the major DSRhub genes revealed distinct regulatory genes such as ZEP, NCED, AAO, and MCSU and CYP707A1. These were involved in the regulation of ABA biosynthesis and signal transduction. Other protein families, namely, aldehyde and alcohol dehydrogenases, mitogene activated protein kinases (MAPKs), and Ribulose-1,5-biphosphate carboxylase (RuBisCO) were shown to be involved in the drought-related responses. This shows a diversity of complex functional features in sorghum to respond to various abiotic stresses. Finally, we constructed a drought-specific subnetwork, characterized by unique candidate genes that were associated with DSRhub genes. According to our knowledge, this is the first in sorghum drought investigation that introduces pathway and network-based candidate gene approach for analysis of drought tolerance. We provide novel information about pathways cross-talk and signaling networks used in further systems level analysis for understanding the molecular mechanism behind drought tolerance and can, therefore, be adapted to other model and non-model crops.