Brachypodium genome reveals conserved gene networks governing seed dormancy in temperate cereals

Seed dormancy, the inability of a viable seed to germinate even when conditions are otherwise favourable, is one of the most economically important traits in cereal farming. Now researchers at KU Leuven have carried out a wide-ranging comparative genomic analysis of the genes that control this trait, using the small grass Brachypodium distachyon as a stand-in for temperate cereals such as wheat, barley and oat, and comparing it with the dicot model plant Arabidopsis thaliana and the tropical cereal rice.

The study, led by Renqiang Li and Koen Geuten of the Department of Biology at the Leuven Plant Institute, curated genes with experimentally proven roles in seed dormancy and germination and traced their evolutionary relationships across the species. The team searched protein databases including Phytozome, UniProt and Ensembl Plants, using BLASTP searches with an E-value threshold of 1e-20 followed by reciprocal BLAST against Arabidopsis proteins to confirm homology. For gene families, they built profile Hidden Markov Models with the HMMER suite and performed maximum likelihood and Bayesian phylogenetic analyses using PhyML and MrBayes, with bootstrap support summarised over 100 replicates.

The central conclusion is that the Brachypodium genome encodes conserved components of the major pathways that regulate seed dormancy induction, maintenance and release. These include seed covering-mediated dormancy imposition, the metabolism and signalling of several plant hormones, and the perception and integration of environmental signals such as temperature, light and nutrient availability. Abscisic acid, or ABA, sits at the heart of the network: ABA synthesised by the zygote establishes dormancy during seed maturation, and both its metabolism and signalling are fundamental to how dormancy is set, maintained and broken.

The phylogenetic analysis identified homologs of key ABA synthesis and catabolism genes in Brachypodium and the temperate cereals. The rate-limiting biosynthetic enzyme 9-cis-epoxycarotenoid dioxygenase, NCED, belongs to a family in which Arabidopsis NCED6 and NCED9 show seed-specific expression and play essential roles in setting dormancy. Brachypodium carries two close counterparts, BdNCED1 and BdNCED2, which share the highest sequence similarity with rice genes proven to act in grain dormancy and germination. Their expression levels correlated with Brachypodium seed dormancy status under imbibition conditions, and a gene expression time course during grain development supported the pattern. On the catabolic side, the Arabidopsis CYP707A genes encode ABA 8′-hydroxylases, and their cereal counterparts, ABA8’OH-1 and ABA8’OH-2, have been consistently associated with grain dormancy in wheat, barley and Brachypodium through gene expression analysis. In bread wheat, a mutant deficient in an ABA 8′-hydroxylase gene showed stronger germination inhibition under field conditions and has been deployed for pre-harvest sprouting resistance.

ABA signalling itself is built from a trio of component families: the PYR/PYL/RCAR receptors, type 2C protein phosphatases known as PP2Cs, and SNF1-related protein kinase 2 proteins, or SnRK2s. The analysis showed that in group 1 and group 3 ABA receptors, Brachypodium and rice homologs cluster with their corresponding Arabidopsis proteins, while in group 2 four rice homologs sit apart, suggesting possible divergence between rice and Brachypodium. Among the PP2Cs, several wheat homologs including TaPP2C-a5, TaPP2C-a6 and TaPP2C-a10 have been shown to regulate seed dormancy and germination through interaction with core ABA signalling components and with wheat DOG1L proteins. Both Arabidopsis, rice and Brachypodium genomes encode ten SnRK2 members, and germination phenotypes have been reported for the rice SAPK8, SAPK9 and SAPK10 proteins as well as the barley and wheat PKABA1 kinase, indicating functional conservation of this subgroup in dormancy and germination control.

Beyond ABA, the study mapped the gibberellin, auxin, ethylene, brassinosteroid and jasmonate pathways, which coordinate with or antagonise ABA to modulate dormancy. Gibberellin metabolism is shaped by GA20ox and GA3ox biosynthetic enzymes and GA2ox catabolic enzymes; the Brachypodium genome encodes twelve GA2ox members compared with nine in Arabidopsis, and an extra GA2ox8 inferred as a paralog from a possible tandem duplication. OsGA20ox2 in rice has been identified as the candidate gene for a major seed dormancy locus, qSD1-2, which regulates endosperm-imposed dormancy. DELLA proteins, the germination repressors of GA signalling, cluster with grass-specific SLR1-LIKE groups, and functional characterisation of BdSLR1 and BdSLRL1 has already revealed conserved roles in GA-mediated growth regulation, although whether they act in Brachypodium seed dormancy requires further experimental work.

The analysis also highlighted major dormancy regulators whose homology patterns differ strikingly between dicots and cereals. DELAY OF GERMINATION 1, DOG1, is a major regulator of seed dormancy in Arabidopsis, and the phylogeny shows distinct clustering of Arabidopsis and grass DOG1 genes, with AtDOGL4 falling inside the grass clades. In rice, OsDOG1L3 has been proposed as the candidate gene for a seed dormancy QTL, and ectopic expression studies suggest conservation of Triticeae DOG1L genes in dormancy regulation. The authors also traced the key residues needed for the interaction between DOG1 and the PP2C phosphatase AHG1, finding that the grass DOGL3 and DOGL4 clades retain the same proline residue that is critical for this interaction in Arabidopsis. Other major genes, including MFT, MKK3, Sdr4 and the alanine aminotransferase gene Qsd1, show high homology between cereal species while lacking similar roles in Arabidopsis in some cases. Qsd1, the causal gene of a critical quantitative trait locus controlling barley seed dormancy, has a confirmed wheat homolog, and its mechanism has remained elusive since cloning, partly because of the difficulty of working with large crop genomes.

The comparative approach also revealed clear signatures of gene duplication and gene loss since the split of the grass and dicot lineages. Duplication could explain the additional grass proteins seen in the LEC2, GA2ox, AUX3, AUX4, ETR4, ERS1 and AFP clades, as well as expanded families involved in epigenetic and environmental regulation such as FIE1, REF6, VIL3, ROS1, RBOHB and CRY1. Conversely, no homologs were found in grasses for several Arabidopsis proteins, including VAL1 and VRN2 in all grass species analysed, VIL1 in wheat and barley, and AUX2, AUX5, RBOHD, GA20ox5 and DOG1L4 in rice. The authors caution that these absences may be confined to the specific Brachypodium accession examined, Bd21, rather than extending across the wider grass lineage. Epigenetic regulators also came into view: the maternally expressed rice FIE1 and EMF2b genes have been implicated in seed development and dormancy, and the Brachypodium counterparts showed stage-specific expression during grain development that suggests possible roles awaiting experimental confirmation.

Environmental integration emerged as a recurring theme. Blue light inhibits germination in temperate cereals including Brachypodium, wheat and barley, an opposite effect to its role in Arabidopsis, and work in barley has traced this inhibition to the cryptochrome CRY1 acting through CYP707A. No evidence of red and far-red light involvement in grain germination has been reported in domesticated cereals, but observations in Brachypodium and weedy rice hint at possible phytochrome participation. Flowering time genes such as FLC and FT family members, including MFT, also connect temperature sensing to dormancy, with MFT homologs contributing to dormancy and germination control in rice, wheat and barley. Reactive oxygen species and nitrogen signals, mediated by RBOH oxidases and NIN-like proteins respectively, round out the environmental inputs that converge on the central hormonal pathway.

The authors are explicit about the limits of the work. The study is a hypothesis-generating comparative genomics framework, and functional prediction based on sequence homology alone can lead to skewed conclusions. The duplication and loss events described are restricted to the Bd21 accession and may differ in other Brachypodium accessions showing seed dormancy variation. Some gene families, including ARF, PRC2 and DML, showed limited bootstrap support, and results derived from public RNA-seq datasets might reflect annotation errors rather than biology. The team proposes specific follow-up experiments, including yeast two-hybrid tests of the DOG1-PP2C interaction predicted for wheat homologs, phenotypic analysis of Brachypodium mutants lacking identified dormancy gene homologs, qRT-PCR validation of key expression patterns, and complementation of Arabidopsis mutants with grass homologs. If those experiments succeed, Brachypodium, with its small genome and close relationship to wheat and barley, could become the practical route to identifying dormancy genes that protect temperate cereal harvests from pre-harvest sprouting.

Subject of Research: Comparative genomics of seed dormancy genes in Brachypodium distachyon and temperate cereals

Article Title: Comparative genomics of seed dormancy using Brachypodium distachyon as a model for temperate cereals

Article References: Li, R., & Geuten, K. (2026). Comparative genomics of seed dormancy using Brachypodium distachyon as a model for temperate cereals. Discover Plants, 3(1), Article 463. https://doi.org/10.1007/s44372-026-00914-4

Image Credits: AI Generated

DOI: 10.1007/s44372-026-00914-4

Keywords: seed dormancy, Brachypodium distachyon, comparative genomics, temperate cereals, abscisic acid, pre-harvest sprouting, phylogenetic analysis, gene duplication, gene loss, phytohormones, wheat, barley