Abstract
Spermatogenesis is a complex developmental process in which undifferentiated spermatogonia are differentiated into spermatocytes and spermatids through two rounds of meiotic division and finally giving rise to mature spermatozoa (sperm). These processes involve many testis- or male germ cell-specific gene products that undergo strict developmental regulations. As a result, identifying critical, regulatory genes controlling spermatogenesis provide the clues not only to the regulatory mechanism of spermatogenesis at the molecular level, but also to the identification of candidate genes for infertility or contraceptives development. Despite the biological importance in male germ cell development, the underlying mechanisms of stage-specific gene regulation and cellular transition during spermatogenesis remain largely elusive. Previous genomic studies on transcriptome profiling were largely limited to protein-coding genes. Importantly, protein-coding genes only account for a small percentage of transcriptome; the majority are noncoding transcripts that do not translate into proteins. Although small noncoding RNAs (ncRNAs) such as microRNAs, siRNAs, and Piwi-interacting RNAs are extensively investigated in male germ cell development, the role of long ncRNAs (lncRNAs), commonly defined as ncRNAs longer than 200 bp, is relatively unexplored. Herein, we summarize recent transcriptome studies on spermatogenesis and show examples that a subset of noncoding transcript population, known as lncRNAs, constitutes a novel regulatory target in spermatogenesis.
Introduction
Spermatogenesis refers to the continuous production of male gametes (i.e. sperms) from spermatogonial stem cells (SSCs) through a series of tightly regulated developmental programs involving mitosis, meiosis, and differentiation inside the testes. SSCs develop from gonocytes after birth and are able to self-renew and maintain a relatively constant population on the basement membrane of seminiferous tubules throughout reproductive life until the number declines due to aging factors. The microenvironment inside the seminiferous tubules where SSCs proliferate and differentiate represents the ‘niche’, contributed by Sertoli cells, the basement membrane, Leydig cells, and other interstitial cells. The ‘niche’ nourishes the SSC development and supports the whole process of spermatogenesis. A sub-population of SSCs commits differentiation pathway into type A spermatogonia, followed by the formation of type B spermatognia, preleptotene spermatocytes, leptotene spermatocytes, zygotene spermatocytes, pachytene spermatocytes, round spermatids, elongated spermatocytes, and finally spermatozoa (or sperms; Oatley & Brinster 2012; Fig. 1). Therefore, studying the developmental processes at various male germ cell stages during spermatogenesis allows a comprehensive understanding of the cellular and molecular regulations that underlie cellular proliferation and differentiation. Identification of key regulators or signaling pathways in these developmental programs will potentially be useful in generating novel hypotheses, designing diagnostic biomarkers, and therapeutic interventions in various reproductive health issues.
The complexity of mouse spermatogenesis involves different cell types, cellular stages at various developmental time-frames. Top layer: Diagrams showing the key developmental stages in mouse spermatogenesis - gonocytes (light pink), spermatogonia (dark pink), spermatocytes (purple), spermatids and sperm (cyan). Second layer: cell types development at each stage. Third layer: Developmental processes in spermatogenesis. Forth layer: Ploidity of each stage. Fifth layer: Postnatal day in which some of the representative cell types arisen in first wave of mouse spermatogenesis.
Citation: REPRODUCTION 147, 5; 10.1530/REP-13-0594
Noncoding RNAs: overlooked transcript species in genome biology
The central dogma of molecular biology suggests that proteins perform the most of structural, functional, and regulatory roles in various cellular functions. However, this protein-centric view has recently been challenged, as the number of protein-coding genes in higher organisms is very similar to lower ones, and this is not sufficient to account for their complexity. It was estimated that known genes only account for around 5–10% of the mammalian genome. However, recent large-scale transcriptome studies have suggested that the majority of genome is transcribed and outnumbers the current gene annotations. Importantly, a very large number of transcripts do not appear to encode for proteins. These transcript species are known as noncoding RNAs (ncRNAs). Previously, they were often considered as ‘junk’ in the genome, but emerging evidence has highlighted that ncRNAs can have a wide variety of functions. Prokaryotes have <25% noncoding DNA; simple eukaryotes have between 25 and 50% noncoding DNA; and more complex fungi, plants, and animals have more than 50%, rising to ∼98.5% noncoding DNA in humans (Amaral et al. 2008).
ncRNAs can be generally classified according to length. The short class refers to ncRNAs with <200 bp; anything longer than that is referred to as long ncRNA (lncRNA). Examples of short ncRNAs include rRNA, microRNA (miRNA), siRNA, and Piwi-interacting RNA (piRNA). Application of short RNAs has contributed significantly in biology and medicine in recent years. Short ncRNAs involved in spermatogenesis have been systematically reviewed previously (Bettegowda & Wilkinson 2010, Chuma & Nakano 2013, Yadav & Kotaja 2014) and have proved to be indispensible in various developmental processes in male germ cell development.
A novel regulatory layer in cellular development: lncRNAs
One of the major breakthroughs in recent high-throughput transcriptomic studies and genome-wide association studies is the recognition on multiple regulatory roles played by lncRNAs (Djebali et al. 2012). Many lncRNAs are transcribed by RNA polymerase II just as a mRNA without protein coding ability, or formed by processing other transcripts. While short RNA relies largely on partnership with protein, such as argonaut (Yadav & Kotaja 2014) and piwi proteins (Chuma & Nakano 2013), for regulatory functions, there is no specific group of protein regulator identified for universal processing of lncRNA transcripts.
Recently, a number of hallmark reports have revealed the versatile molecular mechanisms of regulation exerted by lncRNA in human or rodent embryonic stem cell (ESC) models, such as interacting with proteins and regulating their function (Hung et al. 2011, Ng et al. 2013), regulating the action of chromatin-modifying factors by directing interaction (Klattenhoff et al. 2013), inactivating X chromosome by direct binding (Ohhata et al. 2011), promoting organelle formation (Chen & Carmichael 2009), inhibiting antisense mRNA expression (Santoro et al. 2013), promoting antisense mRNA degradation (Kraus et al. 2013), regulating mRNA export from nucleus (Chen & Carmichael 2009), acting as competing endogenous RNAs (ceRNAs) to inhibit the function of miRNA (Cheng & Lin 2013), and acting as a decoy which binds DNA-binding proteins and prevents them from approaching target sites in genomic DNA (Kino et al. 2010, Redon et al. 2010). Given such a wide variety of regulatory roles, the exact molecular regulations conferred by lncRNAs remain to be elucidated. In brief, lncRNAs can be classified based on genomic locations in terms of transcript associated or nontranscript associated (intergenic) and the direction of transcription (sense vs antisense). Transcript-associated candidates can be further divided based on genomic features that are classified as promoter-, exon-, intron-, or UTR-associated lncRNAs. Based on recent studies, the molecular mechanisms of lncRNA regulation can mainly be summarized into three modes of regulations, including competitor, recruiter/activator, and precursor (Fig. 2). First, as a competitor, lncRNA can bind to DNA-binding proteins, such as transcription factors (Hung 2011, Ng 2013), and inhibit their attachment to target DNA. They can also bind complimentarily to some structures in DNA molecules and hinder protein binding to specific antisense mRNA (Kraus et al. 2013), or to miRNAs for post-transcriptional regulation (Zhang et al. 2010). Secondly, lncRNA can activate epigenetic modifiers or recruit them to specific target sites, thus enhancing DNA methylation (Berghoff et al. 2013) and histone modifications (Yap et al. 2010, Klattenhoff et al. 2013). lncRNAs are suggested to increase the proximity of those proteins to their reaction sites, thus enhancing the rate of reactions such as DNA methylation and histone modifications. Finally, lncRNAs can be processed by certain RNase such as Drosha and Dicers to form shorter RNAs, which in turn may carry out further post-transcriptional regulation–examples include H19 (Keniry et al. 2012) and HongrES2 (Ni et al. 2011).
Examples on molecular regulation of lncRNAs. The regulations can be summarized to three major categories: (A) competitor, (B) recruiter/activator, and (C) precursor.
Citation: REPRODUCTION 147, 5; 10.1530/REP-13-0594
Public resources for lncRNA prediction and analysis
The locations and configuration of lncRNA genes in the genome, known as annotation, are updated in various public databases regularly, providing the basis for lncRNA prediction. Currently, the majority of mammalian lncRNA annotation is based on humans and rodents. As listed in Table 1, some databases contain both coding RNA and ncRNA annotations, while some provided ready-to-use datasets of lncRNAs. One can download the lncRNA datasets in Browser Extensible Data (BED) or Gene Transfer Format (GTF) format to visualize them in genome browsers (e.g. UCSC genome browser: http://genome.ucsc.edu) or perform further analyses (for details, please visit: http://genome.ucsc.edu/FAQ/FAQformat.html). Furthermore, disease association and molecular interaction and function are also important issues in lncRNA studies. Some databases have taken the pioneer roles to systematically summarize relevant studies and to provide tools to predict potential interactions and functions of lncRNAs.
Online resources for lncRNA annotation and characterization.
| Database | Web addresses | Remarks |
|---|---|---|
| General lncRNA annotation | ||
| Reference Sequence Database (RefSeq), Pruitt et al. (2012) | http://www.ncbi.nlm.nih.gov/refseq/ | RefSeq provides RNA information on over 31 000 organisms. The related protein products, if any, are shown in corresponding RNA transcripts. Potential lncRNAs do not contain any coding sequence (CDS), and probably have accession prefix of NR or XR. Genomic information and expression evidence are provided for each transcript |
| Encyclopedia of DNA Elements (ENCODE) project, ENSEMBL, Derrien et al. (2012) | http://ensembl.org/index.html | RNA annotations of 78 species are available. Gene type of each transcript is shown. Long transcripts in 3′-overlapping ncRNA, antisense, lincRNA, polymorphic psuedogene, processed psuedogene, processed transcript, pseudogene, sense intronic, and sense overlapping may all contribute to lncRNAs. Gene information and expression evidence are provided for each transcript |
| USCS Genes, Meyer et al. (2013) | http://genome.ucsc.edu/ | UCSC genes is a summary of transcript annotations from other databases. lncRNAs found under this annotation are either annotated by RefSeq, or by GenBank plus one other source. The complete annotation can be directly visualized in UCSC genome browser or downloaded from UCSC table browsers for further analysis |
| ncRNA Expression Database (NRED), Dinger et al. (2009) | http://nred.matticklab.com/cgi-bin/ncrnadb.pl | NRED provides probe-centric expression information of mouse and human lncRNAs in over 50 tissues or stages. The data is based on results by custom ncRNA microarray experiments investigating six normal development backgrounds. Also, genomic information and evolutionary conservation are provided for each transcript |
| NONCODE, Bu et al. (2012) | http://www.noncode.org/ | NONCODE contains annotations of non-coding RNAs from eight species. lncRNAs can be easily selected by size |
| Functional RNA Database (fRNAdb), Kin et al. (2007) | http://www.ncrna.org/frnadb/ | fRNAdb contains noncoding RNA annotations of ten species. It provides information on transcription factors, associated diseases, sequence ontology, and gene ontology for some lncRNAs |
| Characterized lncRNA annotation | ||
| Long noncoding RNA Database (lncRNAdb), Amaral et al. (2011) | http://www.lncrnadb.org/ | This database provides expression, literature, and some other characterizations of 50 characterized lncRNAs in various species |
| The lncRNA and Disease Database (LncRNADisease), Chen et al. (2013) | http://cmbi.bjmu.edu.cn/lncrnadisease | This website provides both experimentally supported and predicted lncRNA-human disease relationships, based on hundreds of publications. It also shows relevant lncRNA–protein interaction, where applicable |
| Function and interaction prediction | ||
| Noncoding RNA Functional Annotation Server (ncFANs), Liao et al. (2011) | http://www.bioinfo.org/ncfans/ | This tool provides a platform for obtaining lncRNA expression information from human and mouse microarray data. It also predicts the functions of lncRNAs via four algorithms under user's statistics settings |
| ChIPBase, Yang et al. (2013) | http://deepbase.sysu.edu.cn/chipbase/ | ChIPBase provides expression profiles of human lncRNA and transcription factor binding maps of lncRNAs from six species. All the information provided is based on ChIP-Seq experiments |
| starBase, Yang et al. (2011) | http://starbase.sysu.edu.cn/ | This database provides information on lncRNA–miRNA and lncRNA–protein interactions. It also predicts ceRNA networks for lncRNA function prediction. All the information provided is based on CLIP–Seq studies |
| RNA–Protein Interaction Prediction (RPISeq), Muppirala et al. (2011) | http://pridb.gdcb.iastate.edu/RPISeq/ | This online tool provides a means to predict any possible interaction between lncRNA of known sequence and proteins |
lncRNAs in mammalian spermatogenesis
Transcriptome studies
Identification of functional lncRNA was technically challenging before the advent of high-throughput genome-wide expression assays. Only few lncRNAs, such as H19 (Pachnis et al. 1984), XIST (Brown et al. 1991), and KCNQ1OT1 (Lee et al. 1999), were discovered and characterized. Those discoveries were sporadic and confined to candidates with relatively high expression levels or close proximity to a cluster of known genes. Unbiased identification of noncoding transcripts in the genome is impossible without prior knowledge of genome sequence and high-throughput genomic analysis techniques. These limitations were eventually removed by the genome projects and large-scale expression studies on various male germ cell development models (Table 2). The common platforms for delineating the transcriptome landscape on spermatogenesis are summarized below.
High-throughput studies with data available for lncRNA identification in male germ cell development.
| References | Platform | Sample | Remarks |
|---|---|---|---|
| McCarrey et al. (1999) | cDNA library | Whole testes, primitive type A, types A and B spermatogonia, preleptotene, leptotene, zygotene and pachytene spermatocytes, round spermatids, and Sertoli cells from mice and rats | |
| Wang et al. (2001) | cDNA library subtraction | Primitive type A from 6-day-old CD-1 mice, types A and B spermatogonia from 8-day-old CD-1 mice | |
| Liu et al. (2007) | EST library | Mouse type A spermatogonia, pachytene spermatocytes, round spermatids, and Sertoli cells | |
| Yao et al. (2004) | SAGE | Whole testes from 3- and 29-month-old BDF1 mice and 14-month-old SAMP1 mice | 61 929 SAGE tags for three libraries |
| Divina et al. (2005) | SAGE | Whole testes of adult mouse | 76 854 SAGE tags for two libraries |
| Wu et al. (2004), Chan et al. (2006) and Lee et al. (2009, 2012b) | SAGE | Type A spermatogonia from PND 6 mice, pachytene spermatocytes, and round spermatids from PND 60 mice | 333 470 SAGE tags for three libraries |
| Xiao et al. (2008) | Microarray (Genechip Mouse 430 v.2, Affymetrix) | Whole testes of mice aged 4, 9, 18, 35, 54 days, and 6 months | |
| Waldman Ben-Asher et al. (2010) | Microarray (MGU74 v.2, Affymetrix) | Whole testes of mice at PNDs 7, 10, 12, 14, and 17 | |
| Bao et al. (2013) | Microarray (Mouse LncRNA Array v2.0 8×60k, Arraystar) | Whole testes of mice at six time points from E12.5 to adult | |
| Sun et al. (2013) | Microarray (Mouse LncRNA Array v2.0 8×60k, Arraystar) | Whole testes from mouse at 6 days and 8 weeks | |
| Das et al. (2013) | Microarray, RNA-Seq | Fresh sperms and testes of adult horses | rRNA is naturally depleted in sperms |
| Lesch et al. (2013) | ChIP-Seq for epigenetic marks, RNA-Seq | Flow-sorted male germ cells in mouse embryo at E12.5–E14.5, pachytene spermatocytes and round spermatids were isolated from adult testes | Poly-adenylated transcripts for RNA-Seq |
| Schmid et al. (2013) | RNA-Seq | Whole testes of mice at PNDs 6 and 21 | Poly-adenylated transcripts only |
| Gong et al. (2013) | RNA-Seq | Whole testes of mouse at 6 days, 4 weeks, and 10 weeks | Poly-adenylated transcripts only |
| Gan et al. (2013) | RNA-Seq/5hmC-Seq | Isolated spermatogenic cells of eight stages, including primitive type A, types A and B spermatogonia, leptotene and pachytene spermatocytes, round and elongated spermatids, and spermatozoa. Sertoli cells were also included for reference | Poly-adenylated transcripts for RNA-Seq |
| Laiho et al. (2013) | RNA-Seq | Whole testes of mice at PNDs 7, 14, 17, 21, and 28 | rRNA removal with rRNA-specific probes |
cDNA and expressed sequence tag libraries
The era of high-throughput datasets began with the construction of cDNA or expressed sequence tag (EST) libraries. The idea of this technology was to clone all mature mRNAs in a sample with a pull-down target, most commonly a poly-A tail, into a collection of cloning vectors. Although at that time researchers usually focused on cloning coding genes, the libraries also captured mature lncRNA with poly-A tail, and thus is an evidence for lncRNA expression.
McCarrey et al. (1999) prepared 23 comprehensive testis-derived libraries at 2×105 plaque-forming units (PFU) for whole testes, peritubular cells, Sertoli cells, and nine representative spermatogenic cell types of several strains of mice and rats. The cDNA is unidirectionally cloned to the expression vector. After validation with the presence of some known genes, the library can be used for screening the expression of any transcripts at each cell source with sequence-specific probes, for subsequent sequencing or overexpressing the whole transcript. Wang et al. (2001) also applied the cDNA library subtraction approach to identify genes specifically expressed in spermatogonia. Owing to sensitivity, only 25 genes could be identified, half of which resided in sex chromosomes, while the other half was distributed evenly on autosomes. Although these cloned libraries theoretically contain all mRNA transcripts with poly-A tails, detecting the presence or abundance of a specific transcript still relies on low-throughput procedures using northern blotting and Sanger sequencing. Another limitation of this technology is its low detection sensitivity due to the requirement for abundantly expressed transcripts in the cloning procedure.
Other than full-length cDNA libraries, a relatively rapid approach could be achieved by EST library construction. EST libraries were also used to study different transcriptomes of different germ cell stages during spermatogenesis by means of 3′-processing sites. EST tags were a few hundred base pair long sequences from either end of transcripts in a cDNA library. It is particularly useful to detect RNA isoforms of the same gene, yet, with a relatively strong background noise and vector contamination issue. Liu et al. (2007) built EST libraries from primitive type A spermatogonia, pachytene spermatocytes, round spermatids, and Sertoli cells of mice, and focused on the 3′-UTR of the transcripts. They found that even for the same transcript, the 3′-UTR location and length varied at different stages in spermatogenesis.
An important part of current transcription evidence on lncRNAs came from cDNA or EST libraries. Back then, ncRNAs were not the focus and so no lncRNA candidates were reported. Another limitation of these libraries was their bias towards RNA transcripts with poly-A signals. Although poly-A is very useful to remove rRNAs and tRNAs, this method also screened off potential transcripts without poly-A features, leaving the lncRNA landscape incomplete.
Microarrays
Based on genomic information gathered from previous cDNA and EST libraries, specific cDNA or oligonucleotide probes can be designed for known RNA transcripts. This gave rise to the microarray platform, which is a hybridization-based technology to detect the presence of labeled cDNA targets on immobilized high-density array of synthetic DNA probes. The detection coverage of microarray completely depends on the content of probes, which can target exons, promoters, 3′-UTRs, or the full genes.
There were a number of studies using microarray to investigate male germ cell development. In early studies, custom-made microarrays were used to detect gene expression (Sha et al. 2002, Tanaka et al. 2002). Later, commercial microarrays such as GeneFilters GF400 (Pang et al. 2003), Affymetrix Genechips that covered mouse genome using U74 series (Su et al. 2002, Shima et al. 2004, Waldman Ben-Asher et al. 2010), 430 series (Xiao et al. 2008), and human genome U95 series (Su et al. 2002) were applied in different experimental setups for studying spermatogenesis.
Early microarray platforms were biased towards the 3′-end of a transcript. A more comprehensive microarray platform, known as tiling microarray, provided an unbiased approach in analyzing the whole transcriptome. Unlike usual microarray, which covers only known transcripts from cDNA clones, probes in whole-genome tiling array cover the whole genome by evenly spaced probes along the genomic coordinates. Since the probes are designed in high-density and distributed evenly throughout the genome, all transcripts, both known and unknown, will be covered. We have demonstrated the use of the Affymetrix mouse tiling 1.1R microarray set to screen through transcriptomes of type A spermatogonia, pachytene spermatocytes, and round spermatids. In this study, rRNA was depleted using the biotinylated rRNA-specific probes approach, which can retain lncRNAs without poly-A tails for microarray analysis. Nearly a half of the transcripts we detected had not been characterized before, revealing a large hidden layer of molecular regulation (Lee et al. 2012a).
To date, microarray still offers an ideal platform to access the whole transcriptome in a relatively quick and economical approach compared with sequencing-based methods. Although microarray may not be able to distinguish different gene isoforms and orientation of the transcripts, it does pinpoint the regions of expression in gene expression study or enriched targets in immunoprecipitation (IP) setup. These regions are then subject to subsequent low-throughput validation assay using qPCR or ChIP–PCR. Another main advantage of using microarray data to identify lncRNAs is the availability of public data repositories such as Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) or ArrayExpress (www.ebi.ac.uk/arrayexpress). Researchers are able to access the original datasets for performing analysis tailored to their interest. Some of the Affymetrix microarray data can be reanalyzed with online programs such as ncFANs (Liao et al. 2011) to obtain the lncRNA expression data by a simple data retrieval procedure. Both databases also provide online tool options that allow expression analysis or target gene search without downloading the dataset.
Recently, microarrays with probes specified for lncRNA detection have become available commercially. These include Affymetrix Genechip 2.0 ST series and Arraystar Mouse Stringent LncRNA microarray, which are based on updated lncRNA genomic annotation resources such as RefSeq, ENCODE, UCSC Genes, NRED, NONCODE, and fRNAdb. These platforms allow direct measurement of multiple known lncRNA candidates. Wu's group identified 8265 lncRNAs expressed in testis from either 6-day- or 8-week-old mouse, where 3025 showed difference in expression (Sun et al. 2013). At the same time, Yan's group used a similar approach to investigate six time points in mouse testis development: embryonic day (E)12.5, E15.5, postnatal day 7 (PND 7), 14, 21, and adult (Bao et al. 2013). They identified over 1000 down- and over 1500 up-regulated lncRNAs at each time point when compared with the previous time point. Most lncRNAs displayed the highest expression after PND 14, suggesting that lncRNA expression is most active in the meiosis and spermiogenesis phases. Also, both studies detected a significant portion of differentially expressed lncRNAs overlapping with small RNAs, suggesting possible precursor or regulator relationships. Furthermore, both studies observed a close association between lncRNAs and epigenetic regulation. Wu's group found lncRNA expression is correlated with epigenetic modification marks H3K4me3 and H3K27me3 (Sun et al. 2013). Yan's group further showed by cross-link RNA IP experiment that 11 lncRNAs interact with one or multiple factors of chromatin-remodeling complexes, such as EZH2 and LSD1 (Bao et al. 2013). Taken together, these data suggest that epigenetic regulation by lncRNAs may be involved in chromatin remodeling during spermatogenesis.
Serial analysis of gene expression
The serial analysis of gene expression (SAGE) is a quantitative sequencing technique to assess the abundance of individual transcript in cells before the advent of next-generation sequencing (NGS). Compared with microarrays, it is an open system and the results can be reanalyzed with updated genomic annotations. It also provides transcript orientation information and allows scalable library construction.
In a typical SAGE protocol (Velculescu et al. 1995), all transcripts with poly-A tails are pulled-down and the abundance of a 14-bp tag sequence starting from the 3′-most restriction site of a tagging enzyme, typically BsmFI, which recognize a tagging sequence of CATG, are determined. After mapping the tag sequences with sequencing mapping data such as SAGEmap (http://www.ncbi.nlm.nih.gov/SAGE/), the expression level of each gene can be calculated based on the number of unique tags present in the library. This technique allowed researchers to obtain the expression information of all poly-adenylated, coding, and noncoding transcripts.
Mori's team (Yao et al. 2004) and Forejt's team (Divina et al. 2005), each generated a SAGE library of mouse testis to study reproductive aging and chromosome clustering of genes, while Chan's group (Wu et al. 2004) built a cell-type specific SAGE database on type A spermatogonia, pachytene spermatocytes, and round spermatids of mouse. These data provided unbiased transcriptome information for coding and noncoding gene expression analyses. Non-annotated transcripts with poly-A signals will also be detected by this sensitive method. However, again, this method is weak in distinguishing isoforms of genes. Moreover, the resolution of SAGE is limited by the small number of base of a tag, which often led to multiple alignments for the same SAGE tag.
RNASeq
With the availability of NGS, transcriptome can be analyzed by a technique known as RNASeq. It provides advantages on speed, sensitivity, resolution, and coverage compared with the Sanger sequencing method. Driven by the falling cost of NGS, RNAseq has become a popular approach in transcriptome analysis and in lncRNA identification. Thousands of lncRNAs were identified in recent human and mouse genome projects (Derrien et al. 2012). Each run of RNASeq generates billions of reads, with typical lengths, of 30–400 bp, depending on the technology used. The number of reads mapped to a location directly reflects the expression level and splicing status of transcripts, facilitating inter-sample comparison.
There are a number of recent RNAseq studies on testis and sperm development, aiming to investigate the global transcriptomic dynamics in male germ cell development. Laiho et al. (2013) used RNA from whole testis on PNDs 7, 14, 17, 21, and 28 for analyzing transcriptomic changes in the first wave of spermatogenesis. By comparing two adjacent time points, they discovered 17–46 down-regulated and 16–605 up-regulated lncRNAs in each transition. However, the functions of those lncRNAs were completely unknown. With a similar approach, Gong et al. (2013) performed RNASeq on testis samples of 6-day-, 4-week-, and 10-week-old mice. Their main emphasis was on global protein-coding genes, revealing that the expressions of many stage-specific marker genes and genes for specific developmental processes, such as self-renewal and differentiation, largely coincided with the RNASeq or microarray results.
Schmid et al. (2013) focused on the splicing landscape during the meiotic phase of mouse spermatogenesis, by comparing the transcriptome of testis from PNDs 6 to 21, when the first-round meiosis begins. They identified 150 splicing switches which affected the isoform formation of protein-coding genes, and a number of exons highly expressed only during meiosis, but not at earlier stage or in other tissues. Although not reported, it is reasonable to expect that lncRNAs may also be regulated by similar alternative splicing mechanisms at a specific stage, and this would be an interesting topic for further study.
To discover the potential regulation of a novel epigenetic modification in spermatogenesis, Gan et al. (2013) performed a combined analysis on RNASeq and 5-hydroxymethylated cytosine (5hmC) profiling by chemical labeling and sequencing, showing that 5hmC status in different genomic regions changes continuously during eight selected stages of mouse spermatogenesis. They concluded some coding genes accommodated with abruptly up-regulated 5hmC signal were highly expressed in the late stages, suggesting that 5hmC may be involved in gene activation. This mechanism is also expected in the regulation of lncRNAs.
The RNASeq studies not only provided high-resolution transcriptome data, but also led to novel insights for lncRNAs identification in spermatogenesis. These high-throughput data are available for public access. Therefore, the data can be reanalyzed and mapped to various reference annotation databases such as RefSeq, Ensembl, Noncode, Fantom, and fRNAdb for lncRNAs prediction.
Given the abundance of expression evidence, the results from different studies can be combined to improve the prediction confidence and compliment the limitations of a specific platform. For instance, stage- and process-specific lncRNA transcripts could be retrieved from multiple expression platforms by applying signal cutoffs and Venn diagram approach. This approach is also essential to allow de novo lncRNA prediction when the IncRNA annotation is not known. The completeness of a lncRNA transcript could be predicted with supporting regulatory signals, such as presence of RNA polymerase II, CpG island, CAGE tags, ChIP-Seq mapping data on various transcription factors in the promoter region, and polyA signals in the 3′-UTR.
Functional lncRNAs in spermatogenesis
Despite many potential lncRNAs identified in male germ cell development, only a few have been functionally annotated and characterized. Some were identified by the high-throughout approach described above, while others were identified by the de novo annotation approach. The expression patterns of these novel lncRNA candidates in spermatogenesis are illustrated (Fig. 3). Reviewing the discoveries and regulatory roles of these candidates may bring insights for future lncRNA identification.
Functionally characterized lncRNAs expressed in spermatogenesis. The expression pattern is indicated alongside with the specific male germ cell stage.
Citation: REPRODUCTION 147, 5; 10.1530/REP-13-0594
Meiotic recombination hot spot locus (Mrhl)
Meiotic recombination hot spot locus (Mrhl) RNA was identified when Nishant et al. (2004) were focusing on a 17.2 kb fragment of a novel meiotic recombination hot spots in mouse chromosome 8. Mrhl is a 2.4-kb mono-exonic lncRNA located in the nucleus. Subsequent study suggested that Mrhl RNA regulates spermatogenesis potentially by two molecular mechanisms. The first mechanism was the cleavage of Mrhl by Drosha to form an 80-nucleotide RNA intermediate (Ganesan & Rao 2008). Evidence indicated that both RNAs were localized in nucleoli of the GC1 spermatogonia cell line, suggesting their possible interactions with chromatin (Ganesan & Rao 2008). Besides, Mrhl could inhibit the Wnt signaling pathway through interacting with p68 (Arun et al. 2012). However, the functional effects on spermatogenesis of this lncRNA remain elusive, and may not be easily deciphered without using a knockout animal.
HongrES2
HongrES2 is a 1588-bp lncRNA and expressed specifically in the cauda region of epididymis (Ni et al. 2011). It was detected at a constant level in rats aged from 30 days, when the first wave of spermatogenesis had just completed, to 450 days. This expression pattern perfectly coincided with its reported function in regulating sperm maturation.
The transcript is preferably localized in nucleus, where it is processed to a 23-bp miRNA-like small RNA, named as mil-HongrES2, which inhibits the expression of an epididymis-specific protein CES7 and hence its cholesterol esterase activity. Moreover, overexpression of mil-HongrES2 results in retarded sperm capacitation as indicated by the global tyrosine phosphorylation level (Ni et al. 2011), suggesting that an endogenous low level of the lncRNA expression is a gatekeeper for normal sperm maturation process in epididymis.
Testis-specific X-linked (Tsx)
Testis-specific X-linked (Tsx) was thought to be a protein-coding gene, but has recently been shown to be a lncRNA (Anguera et al. 2011). The same group also found that Tsx is specifically expressed in pachytene spermatocytes, but not in spermatogonia or round spermatids, suggesting a regulatory role in germline meiotic division. TUNEL assay analysis showed that Tsx knockout mice had an abnormally high proportion of pachytene spermatocytes that had undergone apoptosis, suggesting that these lncRNAs are somehow critical for progression of meiosis. Detailed mechanism remains open for investigation.
Dmrt1-related gene (Dmr)
Dmrt1-related gene (Dmr) is another interesting example of a testis-specific functional lncRNA. It was discovered unintentionally when Zhang et al. (2010) tried to clone the Dmrt1 gene, and found a novel RNA isoform of unexpected size and sequence on the 3′-half. Genetic mapping showed this mysterious 3′-half was actually coded by a gene, named therefore Dmr, in chromosome 5, while Dmrt1 is a gene in chromosome 19. In other words, Dmr is able to form a trans-splicing RNA isoform with Dmrt1. The formation of this chimeric mRNA disrupted the coding region of Dmrt1 and replaced the 3′-UTR of Dmrt1, both of which resulted in down-regulation of normal Dmrt1 protein. Since Dmrt1 protein is an essential transcription factor to promote spermatogonial development by up-regulating Sohlh1 and also prevents premature meiosis in spermatogonia by repressing Stra8 (Ottolenghi et al. 2000, Agbor et al. 2013), Drm suppression on Dmrt1 is also possibly involved in the switching between mitosis and meiosis of the development germ cell.
Spga-lncRNAs
To demonstrate the idea of integrating high-throughput data on lncRNA identification, we used previously published SAGE data (Chan et al. 2006, Lee et al. 2006, 2009) and whole-genome tiling microarray data (Lee et al. 2012a). Both platforms were previously used to study the transcriptomes of three critical germ cell stages, namely type A spermatogonia, pachytene spermatocytes, and round spermatids, where the initial focus was on coding genes with differential expression.
With this approach, we identified 50, 35, and 24 single-exonic lncRNAs with annotation specifically expressed in type A spermatogonia, pachytene spermatocyte, and round spermatids respectively (Lee et al. 2012b). Further validated by qPCR, northern blot, and RACEs, two spermatogonia-specific lncRNA candidates, known as Spga-lncRNA1 and 2, demonstrated significant differentiation inhibitory effects in in vitro model, suggesting they may be important in maintaining stemness of spermatogonia.
Conclusion and perspectives
Spermatogenesis is a complex and highly regulated biological process. Based on our current view in ncRNA biology, we suspect that the current understanding of lncRNA regulations in spermatogenesis is very incomplete. Although advances in experimental methods and public lncRNA annotation resources allow rapid discovery of potential lncRNA candidates, the functional annotation of lncRNAs in spermatogenesis remains incomplete. The generation of various transcriptome data will definitely provide better expression evidence for lncRNA prediction and annotation. While the use of public lncRNA annotation could be a straightforward approach for rapid lncRNA identification in spermatogenesis, there is still a subset of transcripts that are not covered by the public annotations. One possibility is that these transcripts may be cell type or process specific. Therefore, applying de novo lncRNA prediction would be a better way to identify and characterize these important novel candidates. We anticipate lncRNAs will be an attractive direction in male germ cell research, and more functional lncRNAs will be characterized when more high-throughput data, refined bioinformatics pipelines, and regulatory evidence from ENCODE, ontology databases, and protein–RNA binding models become available.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the review.
Funding
This work was supported by funds from the Hong Kong Research Grant Council (grant numbers 469711 and 468312) and CUHK Direct Grants (MD10783 and MD11449), Faculty of Medicine, The Chinese University of Hong Kong.
References
Agbor VA, Tao S, Lei N & Heckert LL 2013 A Wt1–Dmrt1 transgene restores DMRT1 to sertoli cells of Dmrt1(−/−) testes: a novel model of DMRT1-deficient germ cells. Biology of Reproduction 88 51. (doi:10.1095/biolreprod.112.103135)
Amaral PP, Dinger ME, Mercer TR & Mattick JS 2008 The eukaryotic genome as an RNA machine. Science 319 1787–1789. (doi:10.1126/science.1155472)
Amaral PP, Clark MB, Gascoigne DK, Dinger ME & Mattick JS 2011 lncRNAdb: a reference database for long noncoding RNAs. Nucleic Acids Research 39 D146–D151. (doi:10.1093/nar/gkq1138)
Anguera MC, Ma W, Clift D, Namekawa S, Kelleher RJ & Lee JT 2011 Tsx produces a long noncoding RNA and has general functions in the germline, stem cells, and brain. PLoS Genetics 7 e1002248. (doi:10.1371/journal.pgen.1002248)
Arun G, Akhade VS, Donakonda S & Rao MR 2012 mrhl RNA, a long noncoding RNA, negatively regulates Wnt signaling through its protein partner Ddx5/p68 in mouse spermatogonial cells. Molecular and Cellular Biology 32 3140–3152. (doi:10.1128/MCB.00006-12)
Bao J, Wu J, Schuster AS, Hennig GW & Yan W 2013 Expression profiling reveals developmentally regulated lncRNA repertoire in the mouse male germline. Biology of Reproduction 89 107. (doi:10.1095/biolreprod.113.113308)
Berghoff EG, Clark MF, Chen S, Cajigas I, Leib DE & Kohtz JD 2013 Evf2 (Dlx6as) lncRNA regulates ultraconserved enhancer methylation and the differential transcriptional control of adjacent genes. Development 140 4407–4416. (doi:10.1242/dev.099390)
Bettegowda A & Wilkinson MF 2010 Transcription and post-transcriptional regulation of spermatogenesis. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 365 1637–1651. (doi:10.1098/rstb.2009.0196)
Brown CJ, Ballabio A, Rupert JL, Lafreniere RG, Grompe M, Tonlorenzi R & Willard HF 1991 A gene from the region of the human X inactivation centre is expressed exclusively from the inactive X chromosome. Nature 349 38–44. (doi:10.1038/349038a0)
Bu D, Yu K, Sun S, Xie C, Skogerbo G, Miao R, Xiao H, Liao Q, Luo H & Zhao G et al. 2012 NONCODE v3.0: integrative annotation of long noncoding RNAs. Nucleic Acids Research 40 D210–D215. (doi:10.1093/nar/gkr1175)
Chan WY, Lee TL, Wu SM, Ruszczyk L, Alba D, Baxendale V & Rennert OM 2006 Transcriptome analyses of male germ cells with serial analysis of gene expression (SAGE). Molecular and Cellular Endocrinology 250 8–19. (doi:10.1016/j.mce.2005.12.018)
Chen LL & Carmichael GG 2009 Altered nuclear retention of mRNAs containing inverted repeats in human embryonic stem cells: functional role of a nuclear noncoding RNA. Molecular Cell 35 467–478. (doi:10.1016/j.molcel.2009.06.027)
Chen G, Wang Z, Wang D, Qiu C, Liu M, Chen X, Zhang Q, Yan G & Cui Q 2013 LncRNADisease: a database for long-non-coding RNA-associated diseases. Nucleic Acids Research 41 D983–D986. (doi:10.1093/nar/gks1099)
Cheng EC & Lin H 2013 Repressing the repressor: a lincRNA as a microRNA sponge in embryonic stem cell self-renewal. Developmental Cell 25 1–2. (doi:10.1016/j.devcel.2013.03.020)
Chuma S & Nakano T 2013 piRNA and spermatogenesis in mice. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 368 20110338. (doi:10.1098/rstb.2011.0338)
Das PJ, McCarthy F, Vishnoi M, Paria N, Gresham C, Li G, Kachroo P, Sudderth AK, Teague S & Love CC et al. 2013 Stallion sperm transcriptome comprises functionally coherent coding and regulatory RNAs as revealed by microarray analysis and RNA-seq. PLoS ONE 8 e56535. (doi:10.1371/journal.pone.0056535)
Derrien T, Johnson R, Bussotti G, Tanzer A, Djebali S, Tilgner H, Guernec G, Martin D, Merkel A & Knowles DG et al. 2012 The GENCODE v7 catalog of human long noncoding RNAs: analysis of their gene structure, evolution, and expression. Genome Research 22 1775–1789. (doi:10.1101/gr.132159.111)
Dinger ME, Pang KC, Mercer TR, Crowe ML, Grimmond SM & Mattick JS 2009 NRED: a database of long noncoding RNA expression. Nucleic Acids Research 37 D122–D126. (doi:10.1093/nar/gkn617)
Divina P, Vlcek C, Strnad P, Paces V & Forejt J 2005 Global transcriptome analysis of the C57BL/6J mouse testis by SAGE: evidence for nonrandom gene order. BMC Genomics 6 29. (doi:10.1186/1471-2164-6-29)
Djebali S, Davis CA, Merkel A, Dobin A, Lassmann T, Mortazavi A, Tanzer A, Lagarde J, Lin W & Schlesinger F et al. 2012 Landscape of transcription in human cells. Nature 489 101–108. (doi:10.1038/nature11233)
Gan H, Wen L, Liao S, Lin X, Ma T, Liu J, Song C, Wang M, He C & Han C et al. 2013 Dynamics of 5-hydroxymethylcytosine during mouse spermatogenesis. Nature Communications 4 1995. (doi:10.1038/ncomms2995)
Ganesan G & Rao SM 2008 A novel noncoding RNA processed by Drosha is restricted to nucleus in mouse. RNA 14 1399–1410. (doi:10.1261/rna.838308)
Gong W, Pan L, Lin Q, Zhou Y, Xin C, Yu X, Cui P, Hu S & Yu J 2013 Transcriptome profiling of the developing postnatal mouse testis using next-generation sequencing. Science China. Life Sciences 56 1–12. (doi:10.1007/s11427-012-4411-y)
Hung T, Wang Y, Lin MF, Koegel AK, Kotake Y, Grant GD, Horlings HM, Shah N, Umbricht C & Wang P et al. 2011 Extensive and coordinated transcription of noncoding RNAs within cell-cycle promoters. Nature Genetics 43 621–629. (doi:10.1038/ng.848)
Keniry A, Oxley D, Monnier P, Kyba M, Dandolo L, Smits G & Reik W 2012 The H19 lincRNA is a developmental reservoir of miR-675 that suppresses growth and Igf1r. Nature Cell Biology 14 659–665. (doi:10.1038/ncb2521)
Kin T, Yamada K, Terai G, Okida H, Yoshinari Y, Ono Y, Kojima A, Kimura Y, Komori T & Asai K 2007 fRNAdb: a platform for mining/annotating functional RNA candidates from non-coding RNA sequences. Nucleic Acids Research 35 D145–D148. (doi:10.1093/nar/gkl837)
Kino T, Hurt DE, Ichijo T, Nader N & Chrousos GP 2010 Noncoding RNA gas5 is a growth arrest- and starvation-associated repressor of the glucocorticoid receptor. Science Signaling 3 ra8. (doi:10.1126/scisignal.2000568)
Klattenhoff CA, Scheuermann JC, Surface LE, Bradley RK, Fields PA, Steinhauser ML, Ding H, Butty VL, Torrey L & Haas S et al. 2013 Braveheart, a long noncoding RNA required for cardiovascular lineage commitment. Cell 152 570–583. (doi:10.1016/j.cell.2013.01.003)
Kraus P, Sivakamasundari V, Lim SL, Xing X, Lipovich L & Lufkin T 2013 Making sense of Dlx1 antisense RNA. Developmental Biology 376 224–235. (doi:10.1016/j.ydbio.2013.01.035)
Laiho A, Kotaja N, Gyenesei A & Sironen A 2013 Transcriptome profiling of the murine testis during the first wave of spermatogenesis. PLoS ONE 8 e61558. (doi:10.1371/journal.pone.0061558)
Lee MP, DeBaun MR, Mitsuya K, Galonek HL, Brandenburg S, Oshimura M & Feinberg AP 1999 Loss of imprinting of a paternally expressed transcript, with antisense orientation to KVLQT1, occurs frequently in Beckwith–Wiedemann syndrome and is independent of insulin-like growth factor II imprinting. PNAS 96 5203–5208. (doi:10.1073/pnas.96.9.5203)
Lee TL, Alba D, Baxendale V, Rennert OM & Chan WY 2006 Application of transcriptional and biological network analyses in mouse germ-cell transcriptomes. Genomics 88 18–33. (doi:10.1016/j.ygeno.2006.03.008)
Lee TL, Cheung HH, Claus J, Sastry C, Singh S, Vu L, Rennert O & Chan WY 2009 GermSAGE: a comprehensive SAGE database for transcript discovery on male germ cell development. Nucleic Acids Research 37 D891–D897. (doi:10.1093/nar/gkn644)
Lee TL, Rennert OM & Chan WY 2012a Revealing the transcriptome landscape of mouse spermatogonial cells by tiling microarray. Methods in Molecular Biology 825 75–92. (doi:10.1007/978-1-61779-436-0_7)
Lee TL, Xiao A & Rennert OM 2012b Identification of novel long noncoding RNA transcripts in male germ cells. Methods in Molecular Biology 825 105–114. (doi:10.1007/978-1-61779-436-0_9)
Lesch BJ, Dokshin GA, Young RA, McCarrey JR & Page DC 2013 A set of genes critical to development is epigenetically poised in mouse germ cells from fetal stages through completion of meiosis. PNAS 110 16061–16066. (doi:10.1073/pnas.1315204110)
Liao Q, Xiao H, Bu D, Xie C, Miao R, Luo H, Zhao G, Yu K, Zhao H & Skogerbo G et al. 2011 ncFANs: a web server for functional annotation of long non-coding RNAs. Nucleic Acids Research 39 W118–W124. (doi:10.1093/nar/gkr432)
Liu D, Brockman JM, Dass B, Hutchins LN, Singh P, McCarrey JR, MacDonald CC & Graber JH 2007 Systematic variation in mRNA 3′-processing signals during mouse spermatogenesis. Nucleic Acids Research 35 234–246. (doi:10.1093/nar/gkl919)
McCarrey JR, O'Brien DA & Skinner MK 1999 Construction and preliminary characterization of a series of mouse and rat testis cDNA libraries. Journal of Andrology 20 635–639. (doi:10.1002/j.1939-4640.1999.tb02565.x)
Meyer LR, Zweig AS, Hinrichs AS, Karolchik D, Kuhn RM, Wong M, Sloan CA, Rosenbloom KR, Roe G & Rhead B et al. 2013 The UCSC Genome Browser database: extensions and updates 2013. Nucleic Acids Research 41 D64–D69. (doi:10.1093/nar/gks1048)
Muppirala UK, Honavar VG & Dobbs D 2011 Predicting RNA–protein interactions using only sequence information. BMC Bioinformatics 12 489. (doi:10.1186/1471-2105-12-489)
Ng SY, Bogu GK, Soh BS & Stanton LW 2013 The long noncoding RNA RMST interacts with SOX2 to regulate neurogenesis. Molecular Cell 51 349–359. (doi:10.1016/j.molcel.2013.07.017)
Ni MJ, Hu ZH, Liu Q, Liu MF, Lu MH, Zhang JS, Zhang L & Zhang YL 2011 Identification and characterization of a novel non-coding RNA involved in sperm maturation. PLoS ONE 6 e26053. (doi:10.1371/journal.pone.0026053)
Nishant KT, Ravishankar H & Rao MR 2004 Characterization of a mouse recombination hot spot locus encoding a novel non-protein-coding RNA. Molecular and Cellular Biology 24 5620–5634. (doi:10.1128/MCB.24.12.5620-5634.2004)
Oatley JM & Brinster RL 2012 The germline stem cell niche unit in mammalian testes. Physiological Reviews 92 577–595. (doi:10.1152/physrev.00025.2011)
Ohhata T, Senner CE, Hemberger M & Wutz A 2011 Lineage-specific function of the noncoding Tsix RNA for Xist repression and Xi reactivation in mice. Genes and Development 25 1702–1715. (doi:10.1101/gad.16997911)
Ottolenghi C, Veitia R, Barbieri M, Fellous M & McElreavey K 2000 The human doublesex-related gene, DMRT2, is homologous to a gene involved in somitogenesis and encodes a potential bicistronic transcript. Genomics 64 179–186. (doi:10.1006/geno.2000.6120)
Pachnis V, Belayew A & Tilghman SM 1984 Locus unlinked to α-fetoprotein under the control of the murine raf and Rif genes. PNAS 81 5523–5527. (doi:10.1073/pnas.81.17.5523)
Pang AL, Taylor HC, Johnson W, Alexander S, Chen Y, Su YA, Li X, Ravindranath N, Dym M & Rennert OM et al. 2003 Identification of differentially expressed genes in mouse spermatogenesis. Journal of Andrology 24 899–911. (doi:10.1002/j.1939-4640.2003.tb03142.x)
Pruitt KD, Tatusova T, Brown GR & Maglott DR 2012 NCBI Reference Sequences (RefSeq): current status, new features and genome annotation policy. Nucleic Acids Research 40 D130–D135. (doi:10.1093/nar/gkr1079)
Redon S, Reichenbach P & Lingner J 2010 The non-coding RNA TERRA is a natural ligand and direct inhibitor of human telomerase. Nucleic Acids Research 38 5797–5806. (doi:10.1093/nar/gkq296)
Santoro F, Mayer D, Klement RM, Warczok KE, Stukalov A, Barlow DP & Pauler FM 2013 Imprinted Igf2r silencing depends on continuous Airn lncRNA expression and is not restricted to a developmental window. Development 140 1184–1195. (doi:10.1242/dev.088849)
Schmid R, Grellscheid SN, Ehrmann I, Dalgliesh C, Danilenko M, Paronetto MP, Pedrotti S, Grellscheid D, Dixon RJ & Sette C et al. 2013 The splicing landscape is globally reprogrammed during male meiosis. Nucleic Acids Research 41 10170–10184. (doi:10.1093/nar/gkt811)
Sha J, Zhou Z, Li J, Yin L, Yang H, Hu G, Luo M, Chan HC & Zhou K 2002 Identification of testis development and spermatogenesis-related genes in human and mouse testes using cDNA arrays. Molecular Human Reproduction 8 511–517. (doi:10.1093/molehr/8.6.511)
Shima JE, McLean DJ, McCarrey JR & Griswold MD 2004 The murine testicular transcriptome: characterizing gene expression in the testis during the progression of spermatogenesis. Biology of Reproduction 71 319–330. (doi:10.1095/biolreprod.103.026880)
Su AI, Cooke MP, Ching KA, Hakak Y, Walker JR, Wiltshire T, Orth AP, Vega RG, Sapinoso LM & Moqrich A et al. 2002 Large-scale analysis of the human and mouse transcriptomes. PNAS 99 4465–4470. (doi:10.1073/pnas.012025199)
Sun J, Lin Y & Wu J 2013 Long non-coding RNA expression profiling of mouse testis during postnatal development. PLoS ONE 8 e75750. (doi:10.1371/journal.pone.0075750)
Tanaka K, Tamura H, Tanaka H, Katoh M, Futamata Y, Seki N, Nishimune Y & Hara T 2002 Spermatogonia-dependent expression of testicular genes in mice. Developmental Biology 246 466–479. (doi:10.1006/dbio.2002.0671)
Velculescu VE, Zhang L, Vogelstein B & Kinzler KW 1995 Serial analysis of gene expression. Science 270 484–487. (doi:10.1126/science.270.5235.484)
Waldman Ben-Asher H, Shahar I, Yitzchak A, Mehr R & Don J 2010 Expression and chromosomal organization of mouse meiotic genes. Molecular Reproduction and Development 77 241–248. (doi:10.1002/mrd.21139)
Wang PJ, McCarrey JR, Yang F & Page DC 2001 An abundance of X-linked genes expressed in spermatogonia. Nature Genetics 27 422–426. (doi:10.1038/86927)
Wu SM, Baxendale V, Chen Y, Pang AL, Stitely T, Munson PJ, Leung MY, Ravindranath N, Dym M & Rennert OM et al. 2004 Analysis of mouse germ-cell transcriptome at different stages of spermatogenesis by SAGE: biological significance. Genomics 84 971–981. (doi:10.1016/j.ygeno.2004.08.018)
Xiao P, Tang A, Yu Z, Gui Y & Cai Z 2008 Gene expression profile of 2058 spermatogenesis-related genes in mice. Biological & Pharmaceutical Bulletin 31 201–206. (doi:10.1248/bpb.31.201)
Yadav RP & Kotaja N 2014 Small RNAs in spermatogenesis. Molecular and Cellular Endocrinology 382 498–508. (doi:10.1016/j.mce.2013.04.015)
Yang JH, Li JH, Shao P, Zhou H, Chen YQ & Qu LH 2011 starBase: a database for exploring microRNA-mRNA interaction maps from Argonaute CLIP-Seq and Degradome-Seq data. Nucleic Acids Research 39 D202–D209. (doi:10.1093/nar/gkq1056)
Yang JH, Li JH, Jiang S, Zhou H & Qu LH 2013 ChIPBase: a database for decoding the transcriptional regulation of long non-coding RNA and microRNA genes from ChIP-Seq data. Nucleic Acids Research 41 D177–D187. (doi:10.1093/nar/gks1060)
Yao J, Chiba T, Sakai J, Hirose K, Yamamoto M, Hada A, Kuramoto K, Higuchi K & Mori M 2004 Mouse testis transcriptome revealed using serial analysis of gene expression. Mammalian Genome 15 433–451. (doi:10.1007/s00335-004-2347-7)
Yap KL, Li S, Munoz-Cabello AM, Raguz S, Zeng L, Mujtaba S, Gil J, Walsh MJ & Zhou MM 2010 Molecular interplay of the noncoding RNA ANRIL and methylated histone H3 lysine 27 by polycomb CBX7 in transcriptional silencing of INK4a. Molecular Cell 38 662–674. (doi:10.1016/j.molcel.2010.03.021)
Zhang L, Lu H, Xin D, Cheng H & Zhou R 2010 A novel ncRNA gene from mouse chromosome 5 trans-splices with Dmrt1 on chromosome 19. Biochemical and Biophysical Research Communications 400 696–700. (doi:10.1016/j.bbrc.2010.08.130)
