Résumé
Double haploid production is the most effective way to create true-breeding lines in a single generation. In Arabidopsis, haploid induction via mutation of the centromere-specific histone H3 (cenH3) has been shown when the mutant is outcrossed to the wild-type, and the wild-type genome remains in the haploid progeny. However, factors that affect haploid induction are still poorly understood. Here, we report that a mutant of the cenH3 assembly factor Kinetochore Null2 (KNL2) can be used as a haploid inducer when pollinated by the wild-type. We discovered that short-term temperature stress of the knl2 mutant increased the efficiency of haploid induction 10-fold. We also demonstrated that a point mutation in the CENPC-k motif of KNL2 is sufficient to generate haploid-inducing lines, suggesting that haploid-inducing lines in crops can be identified in a naturally occurring or chemically induced mutant population, avoiding the generic modification (GM) approach at any stage. Furthermore, a cenh3-4 mutant functioned as a haploid inducer in response to short-term heat stress, even though it did not induce haploids under standard conditions. Thus, we identified KNL2 as a new target gene for the generation of haploid-inducer lines and showed that exposure of centromeric protein mutants to high temperature strongly increases their haploid induction efficiency.