Abstract

Creating true-breeding homozygotes (e.g. recombinant inbred lines or RILs) from a heterozygous F1 typically involves many generations of inbreeding. To accelerate this process, plant breeders produce haploid plants from a heterozygous parent, then convert them into fertile diploids that are homozygous for every locus in the genome. Arabidopsis thaliana haploids can now be made through a simple genetic cross. When a cenh3 GFP -tailswap mutant with altered centromeres is crossed to wild type, mutant chromosomes are lost after fertilization. Up to 50% of viable progeny are haploids produced by complete genome elimination, and we have introduced dominant markers into cenh3 GFP -tailswap to facilitate their selection. Haploid Arabidopsis plants convert into fertile diploids spontaneously. Each haploid yields >50 fertile diploid seeds through random chromosome segregation during meiosis. Haploid genetics has many applications: 1) New RIL sets can be made in only two generations. 2) Multiple mutant construction: it is feasible to homozygose 8 unlinked mutations in a single generation. 3) Gametophyte lethal mutations can be studied in a haploid plant. 4) Any nuclear genome can be combined with the cytoplasmic genomes of choice. 5) Tetraploid Arabidopsis can be converted into diploids to facilitate genetic manipulations. Lastly, we are using the principle of centromere-mediated genome elimination to engineer clonal reproduction (synthetic apomixis) in Arabidopsis. Crossing a mutant with diploid gametes (spo11 rec8 osd1, or MiMe) to a mutant with altered centromeres yielded up to 34% clonal progeny with the same heterozygous genotype as their MiMe parent. Thus, clonal reproduction in an Arabidopsis cross can be created by manipulating four conserved genes. This result raises hope that apomixis can eventually be engineered in crops, allowing vigorous hybrids to be propagated through seed.