Identifying DNA-protein interactions and promoter function in the candidate apomixis APOLLO gene

Abstract

Among genes that control apomixis components, APOLLO is known for its strong linkage to apomeiosis. Previous studies on expression profiles of apomictic vs. sexual Boechera species revealed an ovule-specific expression for an apomictic allele which is characterized by a 20 bp Apo-insertion/Sex-deletion polymorphism conserved in the 5’UTR of APOLLO. In addition, there is allelic variation in Boechera species, as sexual individuals are homozygous for the Sex-alleles while apomicts are heterozygous for the Apo/Sex-alleles. I hypothesized that the apomixis-specific polymorphism (i.e., 20 bp Apo-insertion/Sex-deletion) in the 5′ untranslated region of Apo-allele (5’UTR; TGGCCCGTGAAGTTTATTCC) corresponds to specific transcription-factor binding sites (TF) which are absent in all Sex-alleles. To test this hypothesis a yeast one-hybrid assay was conducted by extracting RNAs of different tissues and creating corresponding cDNA libraries for both apomictic and sexual Boechera species. The libraries along with the bait sequences were sent to Hybrigenics (France) for the yeast one-hybrid assay experiment. The results of the yeast one-hybrid assay demonstrated novel transcription factors belonging to the APETALLA2 (AP2) family including ERF15, ERF107, ERF5 for apomictic Boechera spp., and ERF1B, ERF107 and GLABROUS1 enhancer-binding protein for sexual Boechera. AP2/ERF proteins are known to have important functions in the transcriptional regulation of a variety of biological processes such as response to environmental clues and flower development. These data are in line with the results of a previous study in which the expression of AP2 in sexual tetraploid pearl millet promoted both parthenogenesis and the production of haploid offspring. To learn about the regulation of APOLLO in planta, the function of the APOLLO promoter was studied in transgenic lines. Firstly, native APOLLO promoters, including 1 kb upstream of the transcription start site (TSS) of both Apo- and Sex-alleles were cloned into Arabidopsis using a cassete carrying a GUS reporter gene. Then, 2 kb upstream of the TSS of both Apo- and Sex-alleles were cloned into both Arabidopsis and Boechera followed by analyzing their GUS activity. No GUS activity occurred in plants transgenic for the 1 kb APOLLO promoter, neither for Apo nor Sex-alleles. Arabidopsis transgenic for 2kb Apo promoter showed tissue-specific GUS activity in anther and Boechera transgenic for the same construct in the stigma of different developmental stages, while the 2 kb Sex promoter did not show any GUS activity neither for Arabidopsis or Boechera transgenic lines. The data show that there are differences in GUS activity between Apo- vs. Sex-allele, though no GUS activity was observed in pre-meiotic ovules of Apo-allele transformants. This may mean that the 1 kb and 2 kb regions do not carry all needed regulatory elements and hence a longer promoter region can be tested for future study. Finally, five different synthetic ~2 kb promoters were made to test Apo- vs. Sex-specific promoter components. A comparison of various flower developmental stages in transgenic lines containing the different constructs with the 2 kb native transgenic lines revealed that changes to the APOLLO promoter causes shifts in tissue- and developmental-stage specificity. In addition, it was shown that the 20 bp apomixis-specific polymorphism along with other sequences of the 5’UTR gives rise to ovule-specific GUS activity only when adjacent to 2 kb of the Sex-allele. Together, these findings will serve as a foundation for understanding the complex regulation of the APOLLO gene in Boechera.