Normally, plants are diploid (they can sometimes be tetraploid, but leave that aside for now), meaning they have two of each type of chromosome, one from the male parent and one from the female. Scientists and crop farmers prefer plants that ‘breed true’: plants that produce offspring with identical traits to those found in the parents. This can be tricky to achieve with diploid plants. On the other hand, a haploid plant with only one set of chromosomes will breed true by definition. They only have one version of each gene. Although haploid plants have been created in the past, the methods used were expensive and only worked some of the time.
Ravi and Chan were studying a particular histone called CENH3. Histones are the proteins that package loose strands of DNA into tightly packed chromosomes. CENH3 is located at the centromere, the region of the chromosome where homologous chromosomes pair up and where the chromosome is pulled into daughter cells as the cell divides.
Ravi damaged and fluorescently tagged CENH3 in Arabidopsis thaliana (frequent readers of this blog will have seen this plant before), and bred the mutants with wild-type (normal) plants. He expected to see plants containing both a gene for normal CENH3 and a gene for mutant CENH3. Instead, the offspring only had the normal gene (no fluorescence). It turned out that the resultant plants were all haploid. Because the centromeres could not line up properly, half the chromosomes were eliminated in a process called ‘genome elimination’. This technique of creating haploids by breeding wild-type plants with CENH3 deficient plants should be applicable to most plant species.
Of course, haploid plants cannot produce germ cells, so the plants must double their chromosomes before reproducing. This is accomplished by ‘meiotic non-reduction’ in which the chromosomes replicate as during cell division, but the cell itself never actually divides.
As an aside, CENH3 is found in animals as well as plants, but unlike most essential genes, which tend to be highly conserved, CENH3 shows great variability. Ravi and Chan speculate that differences in centromeres may prevent different species from interbreeding.