Identification of core, conditional and crosstalk components of tomato heat stress response using integrative transcriptomics and orthology

Heat stress significantly affects global agricultural yield and food security and as climate change is expected to increase the frequency and severity of heatwaves, this is a growing challenge. Tomato plants are prone to heat stress exposure both in the field and in greenhouses, making heat stress resilience a key trait for breeding. While the identification of heat-associated genes has been addressed in multiple individual studies, the quantitative integration of data from these studies holds potential for low-cost, high-value knowledge gain about the complex network of actors involved in heat stress response mechanisms. To address this challenge, we have compiled a comprehensive data resource containing both novel and publicly available RNA-seq data on tomato in heat stress spanning multiple tissues, genotypes, and levels and durations of stress exposure. We show that in each individual dataset the large majority of responses originates from an interaction between the stimulus and the specific experimental setup. Conversely, by intersecting differentially expressed genes across experiments, we identify a tomato-specific core response of only 57 genes encoding heat shock proteins, transcriptional regulators, enzymes, transporters and several uncharacterized proteins. 17 of these genes lie within previously identified genetic loci associated with heat tolerance traits. Applying the same approach to all publicly available RNA-seq data on drought and salt stress in tomato, we find large overlaps in the conditional parts of the stress responses but the robust and sustained core responses are mostly stress-specific. Finally, we show that the core responses to these stresses are enriched with evolutionarily ancient genes with orthologs across all domains of life and that the heat core response genes form identifiable co-evolving clusters within the Streptophyta. Our study exemplifies the importance and advantage of using FAIR public data to interpret results of new stress experiments, and provides tools to perform such analyses in a relatively short time.