The Dynamics Of Gene Expression Profile During Postnatal Mouse Ovary Development Via Next Generation Sequencing Technology

Mammalian ovary has two basic functions:producing mature follicles and secreting sexual hormones necessary for female growth. Ovary development coordinates multiple cellular. molecular, and histogenetic events. However, certain essential cellular events for ovarv development such as formation of primordial follicles, initial recruitment of follicles and selection of dominant follicles, have'nt yet been well understood. A detailed study of ovary transcriptome on a genome-wide scale will be a great contribution for reproductive biology research. The transcriptome analysis of mammalian ovary based on DNA microarray technology revealed the genes involved in ovary development, as well as their expression levels. Thus, the outcome of microarray studies has been serving as basic information for gene expression pattern during ovary development. However, DNA microarray technology has several limitations:(1) the number of genes is fixed; (2) the sensitivity is limited by background hybridization; and (3) the inaccuracy of mRNA levels because of difference in hybridization among probes. With these limitations, DNA microarray studies can only identify genes at modest or high expression level and are not able to detect alternative splice transcripts.In recent years, a novel method based on next-generation sequencing technology termed RNA-seq is being exploited to analyze the dynamics of transcriptomes. The trancriptomic research on model organisms using RNA-seq indicates that RNA-seq can detect and quantify the dynamic of transcriptomes at a single nucleotide level on a genome-wide scale. Compared to microarray technology, RNA-seq is much more sensitive, reproducible and accurate. In addition, RNA-seq can be used to detect new alternatively spliced transcripts, which are usually not specifically detected in microarray experiments. Therefore, RNA-seq is more suitable for transcriptome study than DNA microarray. In this study, to help assess the full range of genes involved in postnatal ovary development, we use the next-generation sequencing technology (SOLiD) to carry out a comparative study on mouse ovary transcriptomes at three developmental stages:infant stage (7 days old), juvenile stage (4 weeks old), and adult stage (8 weeks old).We acquired 12,585,638,15,652,548 and 17,476,893 uniquely-mapped reads (50bp per read) from infant, juvenile, and adult ovary samples, respectively, and of the uniquely-mapped reads, 1987750,1919665 and 5128962 are mapped to exons. We subsequently identified 16,961,17,185 and 18,632 Ensemble-defined genes (≥2 uniquely mapped reads in exonic regions) expressed in the three samples. A majority of these expressed genes are mRNA genes but some are non-coding RNAs. The number of genes identified in our data indicates that RNA-seq detects 50% more genes than microarray. Further analysis indicates that about 6,692 genes in the mouse ovary have alternative transcripts, and 5002 alternative spliced variants found in our data are distinct from those defined by Ensemble-defined genes. We also found that some alternative transcripts expressed specially among three developmental stages.We used RPKM value to calculate the gene expression abundance and to normalize the gene expression profiles among the three stages. The gene expression profiles are clustered into two groups, one contains infant and juvenile ovaries, and the other contains adult ovary only. This result is consistent with the physiological characteristic of the three developmental stages. Previous studies indicated that the development of ovarian follicle can be differentiated as a two-phase process based on their dependence of gonadotropin:(1) the initial recruitment of the follicle from the primordial pool to preantral follicle, and (2) the cyclic recruitment of growing follicles involving gonadotropin-dependent stages of rapid growth from preantral to mature follicle. The infant and juvenile ovaries are composed of primordial, primary and small preantral follicles, whose development is gonadotropin-independent. In contrast, adult ovary contains a number of antral follicles for which the effect of gonadotropin is essential. Therefore, it's reasonable that the expression profile of juvenile ovary is more similar to that of infant ovary instead of adult ovary. Our cluster results supported the two-phase theory perfectly.Through comparative analysis of gene expression profiles among the three samples, we found the expression abundance of 454 genes changed significantly (P<0.005) between infant and juvenile ovaries, while the number of that between juvenile and adult ovaries was 4366. The number of significantly changed genes among the three stages supports the result of clustering analysis further. Compared to juvenile ovary,4100 differentially expressed genes are up-regulated in adult ovary. This phenomenon suggests that the gene expression abundance of ovary is much more enhanced after puberty. GO analysis showed that these up-regulated genes reflected a vast repertoire of genes involved in regulating steroid biosynthesis, energy metabolism, cell component, signal transduction, apoptosis. Moreover, we found that the transcriptional activity of genes associated with steroid synthesis such as Cyp11a1, Hsd17b1,Hsd3b1 were constantly up-regulated from infant to adult.Transcription factors (Tfs) play a key role in regulating gene expression. Our research indicates that 1535 transcription factors are involved in the development of mouse ovary, some of which are essential for female fertility such as Nr5a1, Emx2, Nobox, Foxl2 and Sox3. The function of most Tfs during ovary development is still unclear. Moreover, we find 561 Tfs with low expression abundance (RPKM<1) in mouse ovary, including some important genes for folliculogenesis:Figla, Lhx9 and Sox3. These lowly expressed Tfs that are undetectable by DNA microarray may give us an overall and in-depth understanding of ovary transcriptome. Our data also reveals a number of non-coding RNA genes involved in ovary development, whose exact function is yet to known. In the three samples, we identified 1100 non-coding RNA genes defined by Ensemble annotation. Those non-coding RNA genes are composed of snoRNA, snRNA, miRNA and miscRNA. We also found several known non-coding RNA genes (mir-199a, mir-let7a, mir-let7b, mir-709, mir-26a) with high expression level in mouse ovary. The expression level of certain non-coding RNA genes changed significantly among the three ovary samples, which indicated their predictable roles during ovary development.