Plants have evolved a unique life strategy with alternating generations and continuous postembryonic development. These specialized features have important implications for the development of reproductive cells (gametes) as well as for seed formation. The life cycle of flowering plants consists of two phases: the diploid phase encompasses the visible vegetative organs (roots, stems, leaves, and floral organs), and the haploid phase is represented by the gametes found deep within floral reproductive organs. Unlike animals, in which meiotic products differentiate directly into gametes, the haploid precursors of flowering plants undergo several division cycles before they differentiate into functional reproductive cells.
The formation of plant gametes presents an opportunity for natural selection on the haploid genome, an evolutionary driving force that might be the origin of genetic and epigenetic mechanisms, such those ensuring tight regulation of reproductive development. These mechanisms include maternal effects that act during embryo development, genomic imprinting, and the nonequivalency of parental genome activity during early seed formation. Despite this selective pressure, numerous examples of developmental alternatives suggest flexible regulatory control of the mechanisms leading to gamete formation and embryo development. Many species have developed reproductive strategies that allow them to generate viable seeds without the fusion of sperm and egg by means of apomixis. This method of asexual reproduction culminates in the formation of clonal seeds and has great potential for crop improvement.
By elucidating epigenetic mechanisms that are crucial for early seed formation, our group is trying to determine how haploid-derived cells in the ovule of Arabidopsis thaliana acquire their reproductive identity. We have used a combination of enhancer-detection tagging and RNA interference (RNAi)–induced posttranscriptional silencing to determine the role of several genes that are crucial for female gametogenesis. We demonstrated that AtAGP18, a signaling gene encoding a classic arabinogalactan protein, is essential for initiating gamete formation. In plants lacking AtAGP18 activity, the female haploid precursor fails to enlarge and mitotically divide, indicating that AtAGP18 is essential for the diploid-to-haploid transition. These results revealed the molecular nature of signaling proteins that, since the early 1990s, were thought to be involved in gamete specification.
We have also characterized the function of chromatin-remodeling factors that act during ovule formation. CHR11 encodes an ISWI-like protein that is abundantly expressed during female gametogenesis and embryogenesis in Arabidopsis. To determine the function of CHR11, we specifically degraded endogenous CHR11 mRNA by RNAi. Plants that constitutively lacked CHR11 activity in the diploid phase showed reduced plant height and embryos with limited cell expansion. In contrast, plants in which CHR11 was specifically silenced at the onset of female gametogenesis had normal height and embryo size but defective haploid nuclear proliferation, demonstrating the functional versatility of the SWI2/SNF2 chromatin-remodeling factors during both phases of the life cycle.
We have also focused on small noncoding RNAs that act as regulators of eukaryotic gene expression. Using a bioinformatic approach, we identified two microRNAs (miRNAs) (miR854 and miR855) that regulate the expression of OLIGOURIDYLATE-BINDING PROTEIN 1b (UBP1b) by imperfectly binding to its 3′ untranslated region at multiple sites, a mechanism previously found only in animals. UBP1b encodes a member of a heterogeneous nuclear RNA-binding protein family (hnRNPs) involved in pre-mRNA maturation and splicing in both animals and plants. We showed that miR854 is abundantly expressed in ovules. We also identified miRNAs homologous to miR854 and their conserved hnRNP targets in invertebrates and mammals that included worms, mice, chimpanzees, and humans. The identification of this miRNA family suggests that these miRNAs may have an ancient origin as regulators of basal transcriptional mechanisms.
We recently generated a large-scale collection of short mRNA-derived and noncoding RNA-derived massively parallel signature sequencing (MPSS) tags. Our first MPSS signature collection is derived from wild-type ovules containing cellularized female gametes before pollination. Our second MPSS collection is derived from homozygous sporocyteless/nozzle (spl/nzz) ovules that do not produce female gametes. In collaboration with the group of Blake Meyers (University of Delaware), we compared these signatures with the annotation of the Arabidopsis genomic sequence. In our two libraries, the comparison yielded 28,293 (wild type) and 23,891 (spo/nzz) distinct signatures with sense expression and between 2,942 and 3,343 distinct signatures with antisense expression. An additional 1,200 to 1,900 signatures map to intergenic regions of the genome. The transcriptional abundance of reference genes such as BELL and INOconfirms that our two collections are ovule specific, with no contamination from other floral tissues. A preliminary comparison of signature abundance shows that the activity of many genes is over-represented in wild-type ovules. Transcripts that are significantly represented in the wild type but not in the spo/nzz collection are likely to represent genes that are expressed in female gametes but not in somatic cells of the ovule, whereas transcripts that are overrepresented in spo/nzz ovules might represent genes that are downregulated by female gametogenesis. A whole-mount in situ hybridization procedure implemented by our group allows rapid confirmation of gene expression for many of these differentially regulated genes. Using this information, we are investigating the overall importance of parent-of-origin effects during early seed development.
It is becoming clear that the genetic control of female gametogenesis and seed formation is directed by epigenetic mechanisms that are crucial for the developmental events that distinguish sexual from asexual embryo formation in the ovule. Understanding asexual embryo formation should lead to an understanding of the cellular decisions that plants use to maintain a remarkable developmental plasticity while retaining strict control over growth. We expect that our results will contribute substantial knowledge about the fundamental mechanisms that cause apomixis in a plant ovule, enabling attempts at its induction in cultivated crops.
Last updated March 2007