Functional Analysis of Candidate Apomixis Genes

Although sexual reproduction is ubiquitous in the plant kingdom, asexual reproduction through seeds or apomixis predominates in some angiosperm species. Apomixis has been attracting and challenging scientists for decades due to its potential impact on crop improvement and study of plant development and evolution. Unfortunately, not many major crops have wild relatives which can be used for wide hybridization (Koltunow et al. 1995), and furthermore, the introgression of apomixis from apomictic relatives to sexual crops has proved to be slow and largely unsuccessful (Ozias-Akins and vanDijk 2007).

Another alternative is to introduce one or more genes controlling apomixis through transformation.  Candidate genes for apomixis have been isolated through comparative study of differential gene expression between ovules from sexual and apomictic genotypes or direct sequencing of BAC clones.  In our lab, BAC libraries have been constructed with Cenchrus ciliaris and with a polyhapoid apomictic F1 of a tetraploid pearl millet (Pennisetum glaucum) × P. squamulatum. BAC libraries were subjected to screening with apomixis-linked markers. The cytogenetic mapping using fluorescence in situ hybridization with apomixis-linked BACs revealed that one region from a single chromosome of P. squamulatum is sufficient for inheritance of the apomixis trait.  This region has been previously designated as Apospory Specific Genomic Region (ASGR).  Partial sequencing of ASGR-linked BACs revealed some candidate apomixis genes such as ASGR-BBML (Conner et al. 2008).

ASGR-BBML was characterized by RNA interference and over-expression in Arabidopsis and RNAi in pearl millet × P. squamulatum hybrids.  RNAi is an excellent tool to study gene function because of its specificity and efficiency in knocking down expression of redundant genes. In order to knock downASGR-BBML using RNAi, transgenic pearl millet was crossed with the apomictic parent carrying the target gene and hybrids were analyzed (see figure at right).  We generated two F1 plants in which ASGR-BBMLexpression was significantly reduced. Embryo initiation was reduced and embryo development was significantly delayed in these two F1 plants compared with the control plant.

This work was supported by National Science Foundation award # 0115911.

Peggy Ozias-Akins | The University of Georgia