In the initial stage of wheat spike development, the inflorescence meristem (IM) generates lateral meristems that acquire spikelet meristem (SM) identity. The SMs then generate lateral meristems that acquire floret meristem (FM) identity, and the FMs generate florets and floral organs. The timing of these meristem identity transitions affects the number of spikelets per spike and florets per spikelet, which determines the maximum number of grains that can be formed in a spike and impacts grain yield potential.

We demonstrate that wheat MADS-box genes VRN1 and FUL2 play critical and redundant roles in the determination of SM identity. Combined loss-of-function mutations of these two genes (vrn1ful2-null) were sufficient to revert the lateral SMs into vegetative meristems. In these mutants, the spikelets were replaced by leafy shoots and the IM growth became indeterminate. Based on the latter result, we hypothesized that mutations in VRN1 and/or FUL2 or in genes that regulate their expression would affect the number of spikelets per spike. Single vrn1-null and ful2-null mutants showed increased number of spikelets per spike, likely due to a delayed formation of the terminal spikelet. In field experiments the ful2null mutant showed a 30% increase in the average number of grains per spike.

We also found that mutations in two AP2 paralogs play critical and redundant roles in the determination of FM identity. Loss-of-function mutations of these two genes were sufficient to transform most FMs into sterile glume-like lemmas. Mutations in one of these two AP2 genes (ap2-5) increased the number of florets per spikelet.

Finally, we identified several genes that affect the number of spikelets per spike and that likely operate upstream of VRN1/FUL2. Mutants and natural variation in ELF3, PPD1, FT1 and FT2, affect spikelet number per spike. These genes are part of the wheat photoperiod pathway, which is known to converge in the transcriptional regulation of VRN1. WAPO1, a wheat ortholog of the rice ABERRANT PANICLE ORGANIZATION 1, also affects the number of spikelets per spike but seems to operate by an alternative mechanism. We will present a general model that integrates the effects of the different genes on the determination of the number of spikelets per spike and florets per spikelet. Our results suggest that a better understanding of the basic mechanisms underlying wheat spike development can inform future strategies to improve grain yield in wheat.