Analysis of the transcriptional program governing meiosis and gametogenesis in yeast and mammals

Abstract

During meiosis a competent diploid cell replicates its DNA once and then undergoes two consecutive divisions followed by haploid gamete differentiation. Important aspects of meiotic development that distinguish it from mitotic growth include a highly increased rate of recombination, formation of the synaptonemal complex that aligns the homologous chromosomes, as well as separation of the homologues and sister chromatids during meiosis I and II without an intervening S-phase. Budding yeast is an excellent model organism to study meiosis and gametogenesis and accordingly, to date it belongs to the best studied eukaryotic systems in this context. Knowledge coming from these studies has provided important insights into meiotic development in higher eukaryotes. This was possible because sporulation in yeast and spermatogenesis in higher eukaryotes are analogous developmental pathways that involve conserved genes. For budding yeast a huge amount of data from numerous genome-scale studies on gene expression and deletion phenotypes of meiotic development and sporulation are available. In contrast, mammalian gametogenesis has not been studied on a large-scale until recently. It was unclear if an expression profiling study using germ cells and testicular somatic control cells that underwent lengthy purification procedures would yield interpretable results. We have therefore carried out a pioneering expression profiling study of male germ cells from Rattus norvegicus using Affymetrix U34A and B GeneChips. This work resulted in the first comprehensive large-scale expression profiling analysis of mammalian male germ cells undergoing mitotic growth, meiosis and gametogenesis. We have identified 1268 differentially expressed genes in germ cells at different developmental stages, which were organized into four distinct expression clusters that reflect somatic, mitotic, meiotic and post-meiotic cell types. This included 293 yet uncharacterized transcripts whose expression pattern suggests that they are involved in spermatogenesis and fertility. A group of 121 transcripts were only expressed in meiotic (spermatocytes) and postmeiotic germ cells (round spermatids) but not in dividing germ cells (spermatogonia), Sertoli cells or two somatic control tissues (brain and skeletal muscle). Functional analysis reveals that most of the known genes in this group fulfill essential functions during meiosis, spermiogenesis (the process of sperm maturation) and fertility. Therefore it is highly possible that some of the �30 uncharacterized transcripts in this group also contribute to these processes. A web-accessible database (called reXbase, which was later on integrated into GermOnline) has been developed for our expression profiling study of mammalian male meiosis, which summarizes annotation information and shows a graphical display of expression profiles of every gene covered in our study. In the budding yeast Saccharomyces cerevisiae entry into meiosis and subsequent progression through sporulation and gametogenesis are driven by a highly regulated transcriptional program activated by signal pathways responding to nutritional and cell-type cues. Abf1p, which is a general transcription factor, has previously been demonstrated to participate in the induction of numerous mitotic as well as early and middle meiotic genes. In the current study we have addressed the question how Abf1p transcriptionally coordinates mitotic growth and meiotic development on a genome-wide level. Because ABF1 is an essential gene we used the temperature-sensitive allele abf1-1. A phenotypical analysis of mutant cells revealed that ABF1 plays an important role in cell separation during mitosis, meiotic development, and spore formation. In order to identify genes whose expression depends on Abf1p in growing and sporulating cells we have performed expression profiling experiments using Affymetrix S98 GeneChips comparing wild-type and abf1-1 mutant cells at both permissive and restrictive temperature. We have identified 504 genes whose normal expression depends on functional ABF1. By combining the expression profiling data with data from genome-wide DNA binding assays (ChIPCHIP) and in silico predictions of potential Abf1p-binding sites in the yeast genome, we were able to define direct target genes. Expression of these genes decreases in the absence of functional ABF1 and whose promotors are bound by Abf1p and/or contain a predicted binding site. Among 352 such bona fide direct target genes we found many involved in ribosome biogenesis, translation, vegetative growth and meiotic developement and therefore could account for the observed growth and sporulation defects of abf1-1 mutant cells. Furthermore, the fact that two members of the septin family (CDC3 and CDC10 ) were found to be direct target genes suggests a novel role for Abf1p in cytokinesis. This was further substantiated by the observation that chitin localization and septin ring formation are perturbed in abf1-1 mutant cells.