Grain number is one important yield component of wheat’s grain yield and measurable either as per unit area, per plant, or per spike. A further increase in the number of grains per plant can be achieved by modified spike architecture with enhanced reproductive sink capacity, and thus, producing more potential grain sites. The wheat and barley spike is characterized by sessile spikelets directly borne on the main axis, forming a branchless spike. However, wheat and barley mutants that display non-canonical spike-branching have been characterized in our laboratory of which one tetraploid wheat mutant, ‘Miracle Wheat’ (i.e. branched head, bh^t), produced significantly more grains per spike that consequently led to higher spike yield. Moreover, we positionally cloned and identified mutant alleles of two other spike architecture genes, compositum 1 (com1) and six-rowed spike 4 (vrs4) in barley, that resulted in branch formation. Therefore, the three above-mentioned genes—Com1, Bh, Vrs4—represent exciting targets for engineering novel spike architecture in hexaploid wheat to potentially produce more spikelets and subsequently more grains per spike. During the OSIRIS project, we therefore will apply genome engineering technology derived from the bacterial CRISPR/CAS immune system, to knock-out these spike-branching genes in hexaploid wheat. Consequently, single, double, or triple mutants of the homoeo-alleles will be identified via latest NGS methodology of amplicon sequencing. Mutant plants will be phenotypically characterized for altered spike architecture under greenhouse and field-like conditions to identify the most promising wheat mutants with improved yield potential or sink capacity. Moreover, independently and complementary to the first approach described above, we will explore the existing natural genetic variation for spike-related yield components in a diverse bread wheat germplasm pool following a Genome Wide Association Scan (GWAS) and mapping approach. Our GWAS approach will include detection of association signals affecting grain number per spike, grain size and specifically grain weight, thereby considering the known trade-off effect between increased grain number and seed size/weight. Ultimately, we will perform strategic crossings between genotypes carrying resource-efficient and/or large-seeded related QTLs (identified through GWAS) and novel sink-elevated, branched-spike genotypes (obtained through our knock-out approach via site-directed mutagenesis).

During OSIRIS we specifically aim to:

• Employ RNA-guided CAS endonucleases for three spike architecture genes of hexaploid wheat.
• Generate, select and phenotype homozygous mutants with altered spike architecture and evaluate their yield potential.
• Explore the natural genetic variation for grain number and more specifically grain size/weight in a diverse wheat germplasm using GWAS.
• Conduct strategic crossings between resource-efficient and/or large-seeded genotypes and novel sink-elevated, branched-spike genotypes.