Genetic modification of spikelet arrangement in wheat increases grain number without significantly affecting grain weight

Crop yield is determined by the acquisition and allocation of photoassimilates in sink organs. Therefore, genetic modification of sink size is essential for understanding the complex signaling network regulating sink strength and source activities. Sink size in wheat depends on the number of spikelets per spike, floret/grain number per spikelet as well as the grain weight or dry matter accumulation. Hence, increasing spikelet number and improving sink size are targets for wheat breeding. The main objective of the present work was to genetically modify the wheat spike architecture, i.e., the sink size by introgressing the Miracle wheat' or the bh(t)-A1 allele into an elite durum wheat cv. Floradur. After four generations of backcrossing to the recurrent parent, Floradur (FL), we have successfully developed Near Isogenic Lines (NILs) with a modified spikelet arrangement thereby increasing spikelet and grain number per spike. Genotyping of bh(t)-A1 NILs using the Genotyping-By-Sequencing approach revealed that the size of the introgressed donor segments carrying bh(t)-A1 ranged from 2.3 to 38cM. The size of the shortest donor segment introgressed into bh(t)-A1 NILs was estimated to be 9.8mega base pairs (Mbp). Phenotypic analysis showed that FL-bh(t)-A1-NILs (BC3F2 and BC3F3) carry up to seven additional spikelets per spike, leading to up to 29% increase in spike dry weight at harvest (SDWh). The increased SDWh was accompanied by up to 23% more grains per spike. More interestingly, thousand kernel weight (TKW) did not show significant differences between FL-bh(t)-A1-NILs and Floradur, suggesting that besides increasing spikelet number, bh(t)-A1 could also be targeted for increasing grain yield in wheat. Our study suggests that the genetic modification of spikelet number in wheat can be an entry point for improving grain yield, most interestingly and also unexpectedly without the trade-off effects on TKW. Hence, FL-bh(t)-A1-NILs are not only essential for increasing grain number, but also for understanding the molecular and genetic mechanism of the source-sink interaction for a clearer picture of the complex signaling network regulating sink strength and source activities.