Network-based systems biology has become an important tool for analysing high-throughput gene expression data and for gene function mining. Meiosis is a specialized type of cell division necessary for sexual reproduction in eukaryotes. The genetic mechanisms regulating meiotic progression in plants are still not fully understood. Our knowledge of the genes involved in meiosis in many crop species such as wheat is largely based on studies on model species. The latest advances of wheat genomics in particular the high-quality genome reference sequence[1] and a developmental gene expression atlas[2], together with the gene expression data collected from meiotic samples have provided the prerequisite resources for building a co-expression gene network to facilitate wheat meiotic studies.

In this study, we used the WGCNA package in R to build a meiotic gene co-expression network in wheat based on 130 wheat RNA-seq samples collected from a range of tissues including meiotic tissue (anthers at different meiotic stages). A set of 50,387 genes were expressed during meiosis (TPM  0.5 in one meiosis sample at least) and assigned to 66 modules according to their expression patterns. Three of the modules (modules 2, 28 and 41 containing 4940 genes, 544 genes and 313 genes, respectively) were significantly correlated with meiotic tissue samples (r > 0.5, FDR adjusted p < 0.001) but not with any other type of tissues. Gene Ontology (GO) term enrichment analysis showed that GO terms related to cell cycle, DNA replication, chromatin modifications and other processes occurring during meiosis were highly enriched (FDR adjusted p < 0.001) in the three modules. We also applied orthology informed approaches to evaluate the genes in the meiosis-related modules and found that wheat orthologs of meiosis genes were found in modules 2, 28 and 41. Module 2 in particular was significantly enriched possessing 166 meiosis orthologs (while the expected number of genes was 37).

The combination of co-expression network analysis in tandem with orthologue information will contribute to the discovery of new meiosis genes and greatly empowers reverse genetics approaches to validate the function of candidate genes. Ultimately this will lead to better understanding of the regulation of meiosis in wheat (and other polyploid plants) and subsequently improve wheat production. To our knowledge, this study represents the first meiotic co-expression gene network built in polyploids.

[1]International Wheat Genome Sequencing Consortium. (2018). Science. 361, eaar7191.

[2]Ramírez-González, et al. (2018). Science. 361, eaar6089.