Wednesday, October 3, 2018 at 12:20pm
Abstract: Maize anther development starts with about 100 undifferentiated Layer1- and Layer-2 derived cells, and in the course of 9 days 50,000 cells form, organized into 4 lobes each with 5 distinct cell types and a central column of connective and vasculature tissues. Archesporial cells are the first cell type to be specified, in response to hypoxia, and they secrete MAC1 to program somatic specification. During somatic specification there is huge burst of 21-nt phasiRNAs regulated by the epidermal transcription factor OCL4. Loss of 21-nt phasiRNAs in ocl4 mutants results in male sterility. Later in development, at the start of meiosis, 24-nt phasiRNAs are generated in the tapetum under the control of the master transcription factor MS23. 24-phasiRNAs are required for tapetal redifferentiation as secretory cells under normal corn growing conditions. Both 21-nt and 24-nt phasiRNAs are anther-specific and seem to confer robustness to anther development. Under specific environmental conditions phasiRNAs are dispensable, and plants deficient in phasiRNAs are partially male fertile.
The key features of plant development are that the body plan is indefinite, with continual stem cell activity producing new organs, and that there is an alternation of generations in which the phenotypes of haploid cells are determined mainly by their genotype. These life cycle features allow somatic and gametic selection to operate more stringently than in complex animals with a fixed body plan and in animal gametes. Historically our primary focus has been the regulation of MuDR/Mu transposable elements in the context of the maize life cycle. The transposons switch from "cut and paste" to a net replicative mode of transposition in cells that have acquired pre-meiotic fate. To understand how MuDR/Mu exploit this cell fate specification event, we swtiched to studying cell fate specification in maize anthers to understand the basic biology of this organ.
Plants do not have a germ line. Instead, within each flower a small number of somatic cells must be programmed to adopt a pre-meiotic fate. On the male side, this cell fate specification event occurs in the anthers when pluripotent stem cells become archesporial cells. The anther lobes have just 5 cell types, including the cells that ultimately undergo meiosis. Using a panel of male sterile mutants, transcriptome profiling, proteomics, and genetic analysis we are defining how these archesporial and somatic cells acquire their fates, and then maintain them. We recently discovered that hypoxia, generating a signal mediated by the MSCA1 glutaredoxin, establishes which cells differentiate as pre-meiotic cells and then in turn program the somatic niche surrounding them using a secreted protein. Mobile secreted proteins play key roles in establishing cell fate and programming particular cell division patterns. MAC1 also inhibits archesporial cell division -- either directly or as a consequence of somatic differentiation -- until there is an entire column of such cells in each anther lobe; then the archesporial cells start transit amplifying divisions and a 5 days later start meiosis synchronously.
Using additional mutants and laser capture microdissection we are analyzing the steps in differentiation of individual cell types and investigating whether there are changes in DNA methylation. We are particularly interested in characteristics of the archesporial cells and the neighboring tapetum. Many male sterile mutants have defects in tapetal cell fate specification, commitment, or differentiation, later resulting in meiotic arrest. Our most intriguing finding about the archesporial cells is that as soon as they are specified they being making both the mRNA and proteins utilized in meiosis.
We have intriguing clues that a novel type of small RNA (phasiRNAs = phased small RNAs of 21 or 24 nucleotides) are critical for early steps in in anther development. PHAS loci are non repetitive, transcribed by RNA Pol II but do not encode proteins; the long non-coding transcript is processed into precisely the same 21 or 24 nt pieces by the binding of a 22 nt trigger molecule and the action of a specific Dicers (DCL4 for the 21 nt type and DCL5 for the 24 nt class). Only grass anthers produce the 24 nt phasiRNAs, and in maize they appear shortly before the start of meiosis. Based on current evidence, we hypothesize that epidermal cells make the 21 nt phasiRNAs and the tapetal cells adjacent to the meiotic cells make the 24 nt phasiRNAs. Genetic and molecular approaches are being used to discover the functions of these fascinating small molecules.