Mammalian testis development and spermatogenesis play critical roles in male fertility and continuation of a species. Previous research into the molecular mechanisms of testis development and spermatogenesis has largely focused on the role of protein-coding genes and small non-coding RNAs, such as microRNAs and piRNAs. Recently, it has become apparent that large numbers of long (>200 nt) non-coding RNAs (lncRNAs) are transcribed from mammalian genomes and that lncRNAs perform important regulatory functions in various developmental processes. However, the expression of lncRNAs and their biological functions in post-natal testis development remain unknown. In this study, we employed microarray technology to examine lncRNA expression profiles of neonatal (6-day-old) and adult (8-week-old) mouse testes. We found that 8,265 lncRNAs were expressed above background levels during post-natal testis development, of which 3,025 were differentially expressed. Candidate lncRNAs were identified for further characterization by an integrated examination of genomic context, gene ontology (GO) enrichment of their associated protein-coding genes, promoter analysis for epigenetic modification, and evolutionary conservation of elements. Many lncRNAs overlapped or were adjacent to key transcription factors and other genes involved in spermatogenesis, such as Ovol1, Ovol2, Lhx1, Sox3, Sox9, Plzf, c-Kit, Wt1, Sycp2, Prm1 and Prm2. Most differentially expressed lncRNAs exhibited epigenetic modification marks similar to protein-coding genes and tend to be expressed in a tissue-specific manner. In addition, the majority of differentially expressed lncRNAs harbored evolutionary conserved elements. Taken together, our findings represent the first systematic investigation of lncRNA expression in the mammalian testis and provide a solid foundation for further research into the molecular mechanisms of lncRNAs function in mammalian testis development and spermatogenesis.
Citation: Sun J, Lin Y, Wu J (2013) Long Non-Coding RNA Expression Profiling of Mouse Testis during Postnatal Development. PLoS ONE 8(10): e75750. doi:10.1371/journal.pone.0075750
Editor: Wei Yan, University of Nevada School of Medicine, United States of America
Received: March 19, 2013; Accepted: August 19, 2013; Published: October 10, 2013
Copyright: © 2013 Sun et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by National Basic Research Program of China (grant numbers 2013CB967401 and 2010CB945001; http://www.most.gov.cn) and the National Nature Science Foundation of China (grant number 81121001; http://www.nsfc.gov.cn). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The mammalian testis is the site of spermatogenesis and testosterone production, so it plays a central role in the male reproductive system. Spermatogenesis is the primary biological process in the testis and produces mature haploid spermatozoa from diploid spermatogonia. This developmental process is complicated, and involves a series of cellular differentiation and cell biological events, including spermatogonial proliferation, meiosis of spermatocytes and morphological changes of round spermatids , . Elucidation of the molecular mechanisms underlying spermatogenesis is important for our understanding of the genetic regulation of normal male germ cell development. Importantly, this understanding can also direct strategies for clinical diagnosis and therapy of male infertility. Therefore, investigations into the molecular mechanisms of testis development and spermatogenesis are prominent in the field of reproductive biology. To date, these investigations have largely focused on the role of protein-coding genes and small non-coding RNAs, including microRNAs (miRNAs) and piwi-interacting RNAs (piRNAs). The unprecedented advances in high-throughput technologies, such as microarray screening and transcriptome sequencing, have delivered significant advances in the exploration of testis development and spermatogenesis. So far, gene expression profiling, proteome profiling, miRNA profiling, piRNA profiling during testis development or spermatogenesis in mouse have been investigated –.
Long non-coding RNAs (lncRNAs) represent a novel class of regulatory molecule, which are arbitrarily defined as transcripts of more than 200 nucleotides (nt) in length that lack significant open reading frames . Advances in genome-wide analyses of the mammalian transcriptome have revealed lncRNAs as a major class of transcript, that is pervasively transcribed . Most lncRNAs are transcribed by RNA polymerase II and possess a 5’ methyl cap as well as a polyadenylated tail, similar to protein-coding mRNAs . Numerous lncRNAs are expressed in a tissue/cell type-specific pattern or in a developmental stage-specific manner –. However, they exhibit very low sequence conservation compared to protein coding genes , prompting some to argue that lncRNAs may simply be “transcriptional noise” . Nevertheless, accumulating evidence indicates that lncRNAs are not the "dark matter" of the genome, but that they play significant regulatory roles in various biological processes, including X-inactivation, genomic imprinting, cell differentiation, cell apoptosis, stem cell pluripotency, brain development, retina development, nuclear trafficking, heat shock response, and genome rearrangement , . In addition, lncRNAs are also associated with a variety of diseases, such as cancer, neurological disorders, heart disease, and autoimmune disorders . Surprisingly, the mechanisms of action of lncRNAs are so diverse that they can modulate gene expression at multiple levels (e.g. transcriptional, post-transcriptional or epigenetic level). For example, lncRNAs can directly interact with a promoter or exon of a target gene, can regulate alternative splicing of pre-mRNA, and the expression of microRNAs and translation. LncRNAs can also modulate protein activity through binding to a target protein and can generate functional small RNAs as precursors. In addition, lncRNAs can mediate epigenetic changes by recruiting chromatin remodeling complexes to specific genomic loci, such as polycomb repressive complex 1 (PRC1), polycomb repressive complex 2 (PRC2), G9a complex, LSD1/CoREST/REST complex, TLS/CBP/p300 complex, and MLL/WDR5 complex , .
To date, very few studies on the roles of lncRNAs in mammalian testis development and spermatogenesis have been reported. Lee et al. identified a total of 50, 35 and 24 potential lncRNAs from type A spermatogonia, pachytene spermatocytes, and round spermatids, respectively, through searching SAGE data. Candidates were BLAST searched, mapped and RNA secondary structures were compared against various ncRNA databases. They found that levels of some lncRNAs decreased significantly following induction of differentiation by retinoic acid and the levels of several lncRNAs were decreased by more than a 1000-fold . A recent study revealed that an X-linked gene, Tsx, located within the X-inactivation center, is actually a new member of the lncRNA family and is abundantly expressed in meiotic germ cells. Tsx mutant male mice have smaller testes resulting from pachytene-specific apoptosis, indicating that Tsx performs general functions in male germ cell development . Another recent study found that a lncRNA, meiotic recombination hot spot locus (mrhl), negatively regulates Wnt signaling in mouse spermatogonial cells through its protein partner Ddx5/p68 .
However, our knowledge of the overall expression status of lncRNAs during post-natal development of the mammalian testis is still very limited. To understand roles of lncRNAs in mammalian testis development and spermatogenesis, we first utilized high throughput microarray screening to investigate lncRNA expression profiles of neonatal (6-day-old) and adult (8-week-old) mouse testes. By comparing the lncRNA expression profiles from two developmental stages, we identified differentially expressed lncRNAs. We further examined their genomic context, the gene ontology (GO) enrichment of their associated protein-coding genes, the epigenetic state of their promoter regions and the presence of evolutionary conserved elements. Our data indicate that lncRNAs are likely to play an important role in testis development and spermatogenesis and provide an important foundation for future research in this field.
Overview of lncRNA and mRNA profiles in neonatal and adult mouse testes
To examine the lncRNA expression profiles of the mouse testis during post-natal development, we interrogated a commercial mouse lncRNA microarray (stringent version, Arraystar) with RNA isolated from neonatal (6-day-old, N) and adult (8-week-old, A) mouse testes. This microarray contains 14,724 lncRNA probes collected from RefSeq_NR, UCSC_knowngenes, NRED, Fantom 3.0, and 22,635 mRNA probes. The lncRNA probes involve all 20 pairs of chromosome and mitochondrial genome. The overview of lncRNA expression profiles is summarized in Table 1. We found that 56% of lncRNAs on the microarray (8,265 out of 14,724) exhibited expression above background (Table S1), and that 37% of these (3,025 out of 8,265) were significantly differentially expressed (absolute fold-change ≥5; P value ≤0.05) between neonatal and adult mouse testes (Table S2). By contrast, 82% of protein-coding mRNA transcripts on the microarray (18563 out of 22635) were expressed above background (Table S3), and 32% of these (5,964 out of 18,563) were significantly differentially expressed (Table S4). Similar to previous observations , , our data revealed that a smaller percentage of lncRNAs was detected above background compared to protein-coding genes. This result indicated that lncRNAs exhibited a greater temporal and spatial specificity than protein-coding genes, consistent with previous reports , . In addition, statistical analysis showed that lncRNAs expressed above background in mouse testis were widely scattered on all chromosomes (Figure 1, Table S5) and that the ratio (expressed probes/total probes) of lncRNAs expressed from each chromosome was very similar, except for the mitochondrial genome. We inferred that the high relative number of lncRNAs derived from the mitochondrial genome probably relates to the high abundance of mitochondrial lncRNAs in reproductive tissues, such as ovary and testis .
Chromosomes are on the X axis, and the distribution ratio is on the Y axis. Vertical bands show the ratio (expressed probes/total probes) of expressed lncRNAs derived from each chromosome. “chrM” represents mitochondrial genome.
Thousands of lncRNAs are differentially expressed in neonatal and adult mouse testes
According to the lncRNA expression profiles, 3,025 lncRNAs were differentially expressed (absolute fold-change ≥5; P value ≤0.05) between neonatal (N) and adult (A) mouse testes (Table S2). When we evaluated the expression levels of lncRNAs in paired samples (adult to neonatal ratio, A/N) by log fold-change, 1,062 lncRNAs were found to be significantly down-regulated and 1,963 lncRNAs were significantly up-regulated in the adult testis (Figure 2). The number of up-regulated lncRNAs was almost twice that of down-regulated lncRNAs. In contrast, with regard to the differentially expressed mRNAs (absolute fold-change ≥5), down-regulated mRNAs (3,736 out of 5,964) were more common than up-regulated mRNAs (2,228 out of 5964) (Table S4, Figure S1).
A hierarchical clustered heat map showing the log2 transformed expression values for differentially expressed lncRNAs (absolute fold-change≥5; P≤0.05) between neonatal (N) and adult (A) mouse testis. The intensity of the color scheme is calibrated to the log2 expression values such that red refers to higher transcript abundance and blue refers to lower transcript abundance. The bar code on the right represents the color scale of the log 2 values. Each column represents the data from one of three biological replicates of each sample.
In line with our expectations, some known haploid male germ cell-specific lncRNAs were up-regulated in adult testis, such as Aldoart2, Speer5-ps1 and Speer9-ps1 ,  (Table 2), and some well-known imprinted lncRNAs were down-regulated, including H19, Meg3, Airn (Table 3). Previous studies have shown that generating the methylation imprint on H19, a well-known paternally imprinted gene, is a continuous process during spermatogenesis. H19 methylation begins between E15.5 and E18.5, but only becomes fully methylated post-natally, by the pachytene spermatocyte stage , . It is noteworthy that our data show H19 to be the most significantly down-regulated lncRNA in the adult testis and this result reaffirms H19 imprinting in the testis as an incremental process.
In addition, we found that only the number of lncRNAs transcribed from the X chromosome is much larger in the down-regulated group than the up-regulated group. In contrast, the number of lncRNAs from all other chromosomes is larger in up-regulated group rather than down-regulated group (Figure S2). And predominant expression of X chromosomal lncRNAs early in spermatogenesis is positively correlated with the greater expression in spermatogonia of protein-encoding genes from the X chromosome.
qRT- PCR results are concordant with microarray data
To validate the microarray data, we investigated the expression patterns of eight randomly selected lncRNAs which identified to be differentially expressed on six time points of postnatal testis development involving 6-day-old, 12-day-old, 18-day-old, 24-day-old, 30-day-old and 8-week-old using quantitative real-time PCR (qRT-PCR). The results clearly showed that expression patterns of these eight selected lncRNAs were concordant with microarray data (Figure 3). Additionally, We found there was a high correlation (Spearman coefficient rho = 0.952, p<0.01, n = 8) between the microarray data and the qRT-PCR data (Table 4). These results demonstrated that the microarray results were reliable.
The vertical axis indicates relative expression levels of each lncRNA. The relative expression levels were assessed by Q-PCR and were normalized with GAPDH gene. Each result was the average of three independent biological replicates. The horizontal axis indicates the six time points of postnatal testis development from 6 days (d) to 8 weeks (w) after birth.
Genomic association of differentially expressed lncRNAs with protein-coding genes
Previous studies show that lncRNAs often originate from complex transcriptional loci, in which the lncRNAs are coordinately transcribed with their associated protein-coding transcripts , . A number of investigations have indicated that the exact nature of the genomic relationship between a lncRNA and its associated protein-coding gene usually has important functional consequences, often with the lncRNA regulating the expression of its protein-coding counterpart via epigenetic modification or transcriptional co-activation/repression –. To gain insight into the functional role of lncRNAs that are differentially expressed during testis development, we analyzed their genomic context based upon their orientation to local protein-coding genes. We categorized the relationship between lncRNAs and their associated protein-coding genes as exonic sense, intronic sense, exonic antisense, intronic antisense, bidirectional and intergenic according to our modified definition (see Materials and Methods). Of 3,025 differentially expressed lncRNAs, we identified 343 exonic sense, 495 exonic antisense, 433 intronic sense, 202 intronic antisense, 242 bidirectional and 1,310 intergenic lncRNAs (Figure 4, Table S6).
Exonic sense lncRNAs. Transcriptional profiling has shown that transcription of sense lncRNAs that overlap with exons of protein-coding genes is pervasive throughout the mammalian genome, and many of these lncRNAs can be considered non-coding transcript variants of protein-coding genes . Previous studies have demonstrated that exonic sense lncRNAs can regulate the expression of their associated protein-coding genes . Here, we found 343 lncRNAs were exonic sense. For example, AK011429, a down-regulated lncRNA, was identified as an exonic sense transcript to cyclin D2 (Ccnd2) (Figure 5A). The expression profile of AK011429 was positively correlated with that of Ccnd2. Ccnd2 is highly expressed in spermatogonia and plays an important role in spermatogonial stem cell (SSC) self-renewal .
Examples of exonic sense (A), exonic antisense (B), intronic sense (C), intronic antisense (D), bidirectional (E and F), and intergenic (G) lncRNAs are shown. Each panel illustrates the organization of the lncRNA (red) to associated protein-coding genes (blue). CpG islands are indicated by green strips. Arrows indicate the direction of transcription.
Exonic antisense lncRNAs. Exonic antisense transcripts are prevalent throughout the mammalian genome and are expressed at high levels in the testis , . Numerous studies have demonstrated that they can regulate the expression of their protein-coding counterparts via a range of mechanisms, including chromatin remodeling, alternative splicing, translational promotion, translational interference and promoter targeting . Within our data sets, we found that 495 of the differentially expressed lncRNAs were exonic antisense transcripts. For example, AK077193 was an exonic antisense transcript to synaptonemal complex protein 2 (Sycp2) (Figure 5B), a spermatocyte-specific gene required for synaptonemal complex assembly and chromosomal synapsis during male meiosis . Our data show that AK077193 was up-regulated in adult testis and there was a positive correlation between expression of AK077193 and Sycp2. Therefore, we speculated that AK077193 was likely to regulate Sycp2 expression.
Intronic lncRNAs. A class of long non-coding RNAs transcribed from intronic regions of protein-coding genes, including from sense and antisense strands, was recently identified by the Verjovski-Almeida group, and they have demonstrated that intronic lncRNAs reside in a large portion of mammalian transcriptional units , . Previous studies indicated that intronic lncRNAs could regulate the expression of their host or neighboring genes via multiple mechanisms, including alternative splicing, microRNAs and RNA interference, transcriptional disruption and chromatin modification . In the current study, we identified 433 intronic sense and 202 intronic antisense transcripts from differentially expressed lncRNAs. For instance, AK080917, a down-regulated lncRNA in adult testis, was identified as an intronic sense transcript from the region between introns 4 and 5 of Zinc finger and BTB domain containing 16 (Zbtb16) (Figure 5C). Zbtb16 has a high expression level in spermatogonial stem cells and is a key gene required for their maintenance . A positive correlation between expression of AK080917 and Zbtb16 suggested that their function and/or regulation might be related. Another example is AK005744, which is specifically expressed in testis and up-regulated in adult testis. AK005744 is an intronic antisense transcript from the region between introns 6 and 7 of Spermatogenesis associated 17 (Spata17) (Figure 5D), a testis-specific gene involved in male germ cell apoptosis and strongly expressed in adult testis , . The positive correlation between expression profiles of AK005744 and Spata17 indicated that there might be some regulatory relationships between them.
Bidirectional lncRNAs. A major organizational theme within mouse and human transcriptomes, which applies to approximately 10% of known protein coding genes, is the prevalence of bidirectional transcripts. These are pairs of transcription initiation sites from two different transcripts that are in close proximity (<1000 bp) but in the opposite orientation , . Bidirectional genes usually share a common CpG island promoter , while bidirectional lncRNAs can promote or repress the expression of their neighboring protein-coding genes via epigenetic modification , . Among the differentially expressed lncRNAs, we found that 242 possessed a bidirectional pair. As an example, we identified AK016105 as a bidirectional pair to Piwi-like homolog 1 (Piwil1), a testis-specific gene specifically expressed in spermatocytes and spermatids. Piwil1 plays a central role during meiosis through piRNA-mediated repression of transposable elements and translation regulation . In addition, their common promoter region has a CpG island (Figure 5E). High expression levels of AK016105 in adult testis were concordant with the expression profile of Piwil1. Another example, AK160141, was identified as a bidirectional pair to Zinc finger protein 148 (Zfp148), which is strongly expressed in neonatal testis and is required for normal development of male germ cells . This bidirectional pair also shares a common CpG island promoter (Figure 5F). Our data showed that AK160141 was weakly expressed in neonatal testis. Therefore, it exhibited a discordant expression profile with Zfp148.
Intergenic lncRNAs. Genome-wide analysis in eukaryotes other than human and mouse has shown that long intergenic non-coding RNAs (lincRNAs) represent a large portion of the non-coding genes, for example in zebrafish, worm, and Arabidopsis , , , . Emerging evidence supports the view that lincRNAs play important roles in many fundamental biological processes, such as pluripotency and differentiation of embryonic stem cells, brain development and limb development, and are also involved in certain diseases, especially cancer , . Although, lincRNAs are poorly conserved across species , an increasing number of studies have demonstrated that they can modulate the expression of their neighboring protein-coding genes or other target genes scattered across the genome via directly recruiting histone-modifying enzymes to chromatin , . We identified 1,310 lincRNAs among the differentially expressed lncRNAs. In particular, a large number of lincRNAs were transcribed from HOX loci and were spatially expressed along developmental axes. These lincRNAs possess unique sequence motifs, and their expression can control the expression of neighboring HOX genes by affecting chromatin signature . For example, AK052984, a lincRNA highly expressed in adult testis, was transcribed from the intergenic region between Hoxc4 and Smug1. Hoxc4 is required for spermatogonial stem cell self-renewal and is highly expressed in neonatal testis . The expression of AK052984 was clearly negatively correlated with the expression of Hoxc4. Therefore, we speculated that AK052984 was likely to affect Hoxc4 expression.
Some differentially expressed lncRNAs overlap microRNAs
Small noncoding RNAs, such as microRNAs, can be processed from long primary ncRNAs . It is well-known that mature microRNAs can regulate the stability and fate of their target protein-coding RNAs via binding to the 3’-UTR, 5’-UTR or coding region . Recent studies show that lncRNAs can also be targeted by microRNAs . To explore the possibility that some lncRNAs that are differentially expressed during testis development might also act as primary transcripts for small RNAs, we systematically searched for genomic overlap between differentially expressed lncRNAs and known microRNAs. We found 20 lncRNAs overlapped annotated microRNAs (Table S7). For example, we detected that H19, a well-known paternally imprinted gene, was the most significantly down-regulated lncRNA in the adult testis and contained mmu-mir-675 (Figure 6A). A recent study has demonstrated that H19 can indeed be processed in vivo to give rise to the 23 nt-long miR-675 miRNA and that this ability is conserved in humans and mice . AK144366, a lncRNA down-regulated in neonatal testis, contains mmu-mir-202 within its second exon (Figure 6B). This direct overlap indicates that AK144366 is likely to be processed into and function via mmu-mir-202, and indeed mmu-mir-202 expression is up-regulated in neonatal testis and down-regulated in adult testis . Furthermore, the expression pattern of AK144366 was concordant with that of mir-202, according to Affymetrix Exon Tissues Track (hosted in the UCSC Genome Browser, http://genome.ucsc.edu/) (Figure 6C). This further indicated that AK144366 might be the primary precursor to mmu-mir-202.
Each panel illustrates the lncRNA (red) transcription initiation direction (indicated by an arrow) relative to an annotated microRNA (green). (A) H19 overlaps with mmu-mir-675. (B) AK144366 overlaps with mmu-mir-202. (C) Tissue expression pattern of AK144366, according to the Affymetrix Exon Tissues Track in the UCSC Genome Browser. Red indicates high level of expression.
LncRNA-associated protein-coding genes are more likely to function in transcription-related processes
As previously mentioned, lncRNAs are often coordinately transcribed with their associated protein-coding transcripts and can regulate the expression of their adjacent or overlapping protein-coding genes in multiple ways. To some extent, the function of lncRNAs may be reflected through their associated protein-coding genes. Therefore, gene ontology (GO) term enrichment analysis of associated protein-coding genes may provide insight into the function of lncRNAs. We submitted a list of 3,275 protein-coding genes associated with differentially expressed lncRNAs to the Database for Annotation, Visualization and Integrated Discovery (DAVID) for GO term enrichment analysis and 2,940 genes had functional annotation in DAVID. The most relevant GO terms were enriched in transcription-related processes and some key transcriptional regulator genes involved in spermatogenesis occurred in the “regulation of transcription” gene list, and including Ovol1, Ovol2, Lhx1, Sox3, Sox9, Plzf (Zbtb16), Tfam, Notch1 and Tnp1 (Table 5, Table S8). It is indicated that lncRNAs were likely to perform their function at transcriptional levels by regulating transcription-related genes. In contrast, GO terms related to testis development or spermatogenesis were poorly represented and enriched (Table 5). It is noteworthy that our results showed extremely similar trends to those observed in a previous investigation . Although there was little GO term enrichment for testis development and spermatogenesis-related genes, we still found that approximately 2.5% of genes (69 of 2,940) had associated GO terms relating to testis development or spermatogenesis, included some well-known spermatogenesis genes, such as Plzf (Zbtb16), Kit, Wt1, Notch1, Piwil1, Sycp2, Prm1 and Prm2 (Table S9).
Most differentially expressed lncRNAs exhibit epigenetic modification marks and are preferentially expressed in a tissue-specific manner
Epigenetic modifications, such as DNA methylation and histone modification, are important regulators in various biological processes, including testis development and spermatogenesis. A large number of protein-coding genes are regulated during testis development and spermatogenesis via epigenetic mechanisms . Similar to protein-coding genes, most lncRNAs are transcribed by RNA polymerase II and transcribed products have typical hallmarks of Pol II, such as a 5’ cap and poly(A) tail . Furthermore, many lncRNAs are expressed in specific cell and/or tissue-types and during specific developmental stages. These suggest that lncRNAs may be regulated epigenetically in a similar way to protein-coding genes. This is also supported by previous studies showing that lncRNAs promoters are subject to purifying selection, are on average more conserved than promoters of protein-coding genes, and are associated with chromatin marks , –. Therefore, to investigate whether differentially expressed lncRNAs are subject to epigenetic regulation in the adult testis, we carried out an analysis to identify promoters with high CpG content (HCG), histone H3 lysine 4 trimethylation (H3K4me3) and histone H3 lysine 27 trimethylation (H3K27me3) level.
In mammals, DNA methylation occurs at cytosines within the context of the CpG dinucleotide, which are frequently found in short genomic regions including gene promoters . Mammalian gene promoters can be classified into distinct categories based on their CpG dinucleotide content. HCG promoters are typically non-tissue-specific and frequently regulate housekeeping genes or genes with complex expression patterns and approximately 64% of protein-coding genes have an HCG promoter (HCP) . Therefore, we first identified the promoters of lncRNAs based on CpG content and found that only 23% of differentially expressed lncRNAs (689 out of 3,025) were associated with HCPs and that the CpG observed-to-expected ratio (CpG O/E) ranged from 0.61 to 1.43 (Table S10). This result indicated that a large proportion of differentially expressed lncRNAs were likely to function in a tissue-specific manner, as suggested by previous studies .
H3K4me3 and H3K27me3 regulate gene expression and therefore play key roles in multiple aspects of development . Previous studies indicate that H3K4me3 is generally associated with activation of transcription, while H3K27me3 is closely associated with transcriptional repression . To investigate whether H3K4me3 and H3K27me3 can affect the expression of differentially expressed lncRNAs, we assessed the levels of H3K4me3 and H3K27me3 on lncRNA promoters in adult mouse testis. We found that 73% of up-regulated lncRNAs (1,442 out of 1,963) possessed H3K4me3 marks on their promoters; in contrast, only 32% of down-regulated lncRNAs (338 out of 1062) have H3K4me3 marks (Table S10). The correlation between promoter H3K4me3 marks and lncRNA expression was consistent with that H3K4me3 involved in transcriptional activation. For H3K27me3, only 8% of down-regulated lncRNAs promoters (86 out of 1,062) were marked by H3K27me3, while 4% of up-regulated lncRNAs (71 out of 1,963) had H3k27me3-positive promoters (Table S10). Although a minority of differentially expressed lncRNAs had H3k27me3 marks on their promoters, these results were still consistent with the negative regulatory effect of H3k27me3. In addition, many promoters with H3K4me3 modification are simultaneously marked with H3K27me3 and these so-called “bivalent” domains mark developmentally-associated genes whose expression is thought to be “poised” for activation or repression during development . We found that only approximately 5% of differentially expressed lncRNAs (138 out of 3,025) had bivalent promoters. However, nearly 88% of H3K27me3-positive lncRNAs promoters (138 out of 157) were also marked by H3K4me3, which is similar to previous reports for protein-coding genes .
In addition, previous studies revealed that HCPs are almost always decorated with H3K4me3 and prefer to regulate ubiquitous housekeeping genes . Here, we found 94% of differentially expressed lncRNAs with a HCP were associated with H3K4me3 (649 out of 689), which further indicated that these lncRNAs were extremely likely to be ubiquitously expressed as housekeeping genes and that the majority of differentially expressed lncRNAs perform tissue-specific functions.
We concluded that the majority of lncRNAs differentially expressed during testis development were likely to be expressed in a tissue-specific pattern, and that they could be regulated by epigenetic modification, similar to protein-coding genes.
Most differentially expressed lncRNAs contain evolutionarily conserved elements
Over evolutionary time, the purifying selection of functional genomic elements results in the presence of sequences that exhibit high levels of conservation across multiple species, providing a useful indicator of function, for example, functional protein-coding sequence is often highly conserved. LncRNA sequences generally exhibit low conservation ; however, recent studies have revealed that a large number of lncRNAs contain a short highly conserved sequence in exons, especially in the exon-intron boundaries , . Therefore, we identified this conserved region (termed PhastCons elements) within differentially expressed lncRNAs that are strongly conserved across species (see Materials and Methods). We found that 70% of differentially expressed lncRNAs overlapped PhastCons elements (2,129 out of 3,025), 93% of which (1,986 out of 2,129) possessed ≥20 PhastCons bases (Table S11). Hence, the majority of lncRNAs differentially expressed during testis development exhibited evolutionary conservation of primary sequences. This result was consistent with previous reports .
In recent years, increasing evidence have demonstrated that lncRNAs play key roles in the development of various tissues, such as brain , retina , mammary gland , heart , , and inner ear . However, very few studies have been conducted on the potential roles of lncRNAs in mammalian testis development. Investigations into the molecular mechanisms of testis development and spermatogenesis, have mostly focused on protein-coding genes and small non-coding RNAs (miRNAs and piRNAs). Therefore, our understanding of lncRNA function in mammalian testis development is still extremely poor. For this reason, in the current study, we took advantage of the high throughput feature of microarrays to investigate lncRNAs expression profiles of neonatal and adult mouse testes. This is the first systematic expression profiling study of lncRNAs during post-natal development of the mammalian testis employing genome-wide techniques.
lncRNAs are generally expressed at lower levels than protein-coding genes and are more likely to display a tissue-specific pattern of expression , , . To compare the overall expression status of lncRNAs and mRNAs in mouse testis, we employed microarrays containing probes for both lncRNAs and mRNAs to simultaneously detect the expression of lncRNAs and mRNAs. Consistent with previous reports, we found that only 56% of lncRNAs were expressed above background in mouse testis, while this number reached 82% for protein-coding genes. This finding suggested that lncRNAs tended to be expressed at lower levels and in a testis-specific manner and that they might represent cryptic signals that mainly fulfill regulatory functions in the control of complex testis developmental processes and spermatogenesis. In addition, we found that nearly all mitochondrial genome-derived lncRNAs on the microarray could be detected above background. This finding indicated that mitochondrial genome-derived lncRNAs were relatively abundant in testis, and that lncRNAs might be key contributors to mitochondria-mediated regulation of spermatogenesis; mitochondria are known to play an important role in spermatogenesis, such as meiosis, quality control through apoptosis and sperm motility. We also found that the number of lncRNAs transcribed from the X chromosome is much larger in the down-regulated group than the up-regulated group. It is indiatded that X chromosome inactivation during spermatogenesis could affect the expression of lncRNA similar to protein-coding genes.
In this study, we identified 3,025 lncRNAs that were significantly differentially expressed between neonatal and adult mouse testis. The dynamic change of lncRNA expression during testis post-natal development further indicated that lncRNAs might play significant biological roles in testis development and spermatogenesis. Unlike protein-coding gene or microRNAs, the function of lncRNAs cannot currently be inferred from sequence or structure. Therefore, to date, most studies have predicted function via genomic association of lncRNAs with protein-coding genes because lncRNAs often regulate the expression of their overlapping or neighboring protein-coding genes . Therefore, we analyzed the genomic context of differentially expressed lncRNAs and classified them based on their genomic relationship with protein coding genes as exonic sense, intronic sense, exonic antisense, intronic antisense, bidirectional and intergenic. We found that all six categories of differentially expressed lncRNA could be examined. Strikingly, we found that nearly 43% of differentially expressed lncRNAs (1,310 out of 3,025) were long intergenic non-coding RNAs (lincRNAs). This result indicated that lincRNAs were more abundant in testis relative to other classes of lncRNAs that overlap with protein coding genes, and that they might be the major contributor to the regulatory roles mediated by lncRNAs. Interestingly, a recent study reported that 78% of all defined lincRNAs (4662 in total) exhibited tissue-specific expression patterns relative to protein-coding genes and almost a third of lincRNAs were specific to testis . Combined with our findings, this suggested that lincRNAs are abundant in testis and can be considered as a new class of RNA in the testis, like piRNA . To further define which biological processes lncRNAs may be involved in, we tested for GO enrichments in the set of protein-coding genes associated with lncRNAs in a genomic context. We found that protein-coding genes that overlap or are adjacent to lncRNAs were inclined to be enriched in transcription related processes. Notably, our result was consistent with previous reports on brain development and embryonic development , . Some lncRNAs may, therefore, function as “indirect regulators” of transcription by regulating protein-coding genes responsible for transcription.
Recent studies have demonstrated that lncRNAs can be regulated by epigenetic modification of their promoter regions, such as DNA methylation and histone modification, in a manner similar to protein-coding genes. The dynamic lncRNA expression profiles during testis development prompted us to investigate promoter characteristics related to epigenetic modification, including whether promoters are high CpG content promoters (HCP), and determining levels of H3K4me3 and H3K27me3. We found that only 23% of differentially expressed lncRNAs were defined as HCPs. HCPs are generally associated with ubiquitously expressed “housekeeping” genes; therefore, we inferred that the majority of differentially expressed lncRNAs were likely to be expressed in a testis-specific manner. Our finding was consistent with previous observations showing that most lncRNAs usually exhibit tissue-specific expression patterns , . In addition, we found that there was also a strong correlation between histone modification, including H3K4me3 and H3K27me3, and the expression level of the associated lncRNAs, similar to that observed for protein-coding genes. For instance, in adult mouse testis, we found that almost 73% of up-regulated lncRNAs and only 32% of down-regulated lncRNAs were marked at their promoters by H3K4me3. This result was consistent with previous observations showing that the H3K4me3 mark is strongly correlated with gene activation. We also found that almost all lncRNAs with HCPs often exhibited the H3K4me3 mark simultaneously, very similar to previous reports for protein-coding genes. Clearly, this sub-class of promoters preferentially regulates ubiquitous housekeeping genes further indicated that the majority of differentially expressed lncRNAs might be specifically expressed in testis. It is noteworthy that we found a small percentage of differentially expressed lncRNAs marked by both H3K4me3 and H3K27me3 in adult testis. We speculate that these lncRNAs may be important developmental regulators because developmental regulator genes often carry bivalent histone modifications of both H3K4me3 and H3K27me3 .
In addition, we found that approximately half the differentially expressed lncRNAs were spliced according to annotation from the UCSC Mouse Genome Browser database (mm9), similar to most protein-coding genes (Table S12). This indicated that the majority of lncRNAs differentially expressed during testis development were transcribed by RNA polymerase II (Pol II) because pre-mRNA splicing is frequently coupled to transcription mediated by Pol II . Therefore, it was not surprising that lncRNAs could be regulated via similar epigenetic modifications to those that regulate protein-coding genes.
Most lncRNAs generally exhibit poor primary sequence conservation; however, recent investigations have found short, highly conserved regions in lncRNA primary sequences. Indeed, we found that nearly 70% of differentially expressed lncRNAs contain PhastCons conserved elements. An explanation for the low sequence conservation of lncRNA sequences may be that they do not require very much nucleotide sequence conservation to maintain their functionality. Previous studies have demonstrated that a large number of lncRNAs can bind histone modification complexes, such as polycomb repressive complex 2 (PRC2), then guide these complexes to specific sites and cause the silencing of target genes via histone modification . It is possible that similar secondary structures formed by lncRNAs with distinct sequence contribute to the specificity of interactions with the same protein partners. For example, Xist and HOTAIR, two well-known functionally annotated lncRNAs, bind to PRC2. Both have similar, short GC-rich stem-loop RNA motifs that are required for recruitment of PRC2 , . Therefore, to maintain normal function, these lncRNAs may only need to conserve short stretches of sequence that form similar secondary structures. In contrast, protein coding genes are under intense selection restraints due to the need to maintain correct amino acid coding and an open reading frame.
In summary, this is the first systematic study to examine the expression profiles of lncRNAs in mammalian testis on a genome-wide scale. The dynamic expression profiles and feature analyses, including genomic context, epigenetic modification of promoters and evolutionary conservation, indicate that lncRNAs may play important roles in post-natal development of the mammalian testis. Although due to technical restriction of microarray, it is not possible to profile entire transcriptome of lncRNAs in this study; our results still provide a solid foundation for the identification and characterization of key lncRNAs involved in testis development or spermatogenesis.
Materials and Methods
Twenty one neonatal (6-day-old) and six adult (8-week-old) male C57BL/6 mice were purchased from SLAC Laboratory Animal Co., Shanghai, China. Mice in each age group were divided into three groups to provide three biological replicates for microarray analysis. All procedures involving animals were approved by Institutional Animal Care and Use Committee (IACUC) at Shanghai Jiao Tong University, Shanghai, China [SYXK (Shanghai 2007-0025)], and were conducted in accordance with the National Research Council Guide for Care and Use of Laboratory Animals.
RNA isolation and labeling
RNA derived from mouse testes was purified using Trizol Reagent (Invitrogen) and treated with DNase I (Fermentas). The quantification and quality of RNA samples were assessed using a NanoDrop ND-1000 Spectrophotometer (Thermo Scientific). RNA integrity and genomic DNA contamination were examined by denaturing agarose gel electrophoresis. Double-stranded cDNA was synthetized from RNA using an oligo dT primer and a Superscript Double-Strand Synthesis Kit (Invitrogen, 11917-020). cDNA was labeled with Cy3 using the Quick Amp labeling kit (Agilent). Labeled cDNA quality assessment and quantification were performed using a NanoDrop-1000 Spectrophotometer (Thermo Scientific).
Microarray expression analysis
Labeled cDNA was hybridized to the Mouse Stringent LncRNA Microarray (Arraystar) using Agilent’s SureHyb Hybridization Chambers according to the One-Color Microarray-Based Gene Expression Analysis protocol (Agilent Technology). After hybridization and washing, slides were scanned with the Agilent DNA microarray scanner (G2505B) using the recommended settings. The resulting text files, extracted from Agilent Feature Extraction Software (version 10.5.1.1), were imported into the Agilent GeneSpring GX software (version 11.0) for quantile normalization and background correction. Probe level files and gene summary files were produced. Differentially expressed lncRNAs and mRNAs were identified through Absolute fold change and P values were calculated using Student's T-test (paired). An Absolute fold change of ≥5.0 and a P value of ≤0.05 were selected as thresholds for significant differential expression. Raw and processed microarray data have been deposited in the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) and are accessible through (GEO) series accession number, GSE43442.
Quantitative real-time PCR
Quantitative real-time PCR (qRT-PCR) was performed according to the manufacturer’s protocols (Roche). The qRT-PCR reactions were performed on a 96-well plate in triplicate. Each reaction consisted of a 25 µL mixture containing 12.5 µL 2×FastStart Universal SYBR Green Master (Roche), 0.3 µM forward and reverse primer and 40 ng cDNA. qRT-PCR amplifications were performed using the ABI PRISM 7500 Sequence Detection System (Applied Biosystems). Amplification conditions were as follows: 10 min at 95°C to activate the FastStart Taq DNA polymerase followed by 40 cycles of 15 s at 95°C and 30 s 60°C. A dissociation curve was drawn to ensure the validity of each specific PCR product. The qRT-PCR was repeated three times. The relative expression of genes was calculated based on the 2−ΔΔCt method using the mouse housekeeping GAPDH gene as an endogenous control . Differences in expression levels between two groups were evaluated using Student's t-test (paired). Primers used are shown in Table S12. The primers were designed using Primer Premier 5.0 and checked using Primer-BLAST searches to avoid cross-amplification. Amplification efficiency was evaluated via standard curve analysis.
LncRNA genomic context analysis
We determined the genomic context of lncRNAs in relation to protein-coding genes according to a protocol that was updated for this study (Figure S3) . In summary, exonic sense and antisense lncRNAs were defined where the corresponding transcript was mapped to the positive and opposite strand, respectively, of a RefSeq-annotated exon [includes 5′-untranslated region (UTR), coding exon, and 3′-UTR]. Intronic sense and antisense lncRNAs were defined where the corresponding transcript was mapped to the positive and opposite strand, respectively, of an intron of a protein-coding gene. Bidirectional lncRNAs were defined where the corresponding transcript was oriented head-head to a protein-coding gene at a distance of <1000 bp. Intergenic lncRNAs were defined where the corresponding transcript was located within an intergenic region and no overlapping or bidirectional coding transcripts were nearby.
Gene Ontology (GO) term analysis of lncRNAs associated with protein-coding genes
Information on Gene Ontology functions of differentially expressed lncRNAs associated with protein-coding genes was obtained from the Database for Annotation, Visualization and Integrated Discovery v6.7 (DAVID) (http://david.abcc.ncifcrf.gov/) . Statistically over-represented GO terms in the biological process and molecular function categories were obtained by applying a Fisher's exact p-value cutoff <0.05, and correcting for multiple testing with the Benjamini false discovery rate. The mouse genome was used as the reference set.
LncRNA promoter analysis
High CpG content promoters (HCPs) were identified as described previously . Briefly, transcripts in a 500 bp interval, within −0.5 to +2 kb of the transcription start site (TSS), with a GC fraction≥0.55 and a CpG observed-to-expected ratio (CpG O/E) ≥0.6 were classified as HCPs. The CpG O/E was calculated as described previously . H3K4me3 and H3K27me3 status of promoter regions (−0.5 kb to +2 kb of the TSS) was obtained from publicly available H3K4me3 (UCSC accession: wgEncodeEM002489) and H3K27me3 (UCSC accession: wgEncodeEM002723) ChIP-Seq data previously generated from testes of 8-week-old mice, respectively. The peaks score corresponding to a promoter region was used to evaluate the level of H3K4me3 and H3K27me3 within the promoter.
Evolutionary conservation analysis of lncRNAs
Conservation of each LncRNA was determined by intersecting its sequence with those of genome-wide PhastCons elements. The number of total bases annotated as PhastCons elements was used to evaluate the evolutionary conservation. The PhastCons program uses genome alignments across 30 vertebrate species (30-way) to identify conserved genomic regions based upon a phylogenetic hidden Markov model .
Hierarchical clustering of mRNAs differentially expressed in neonatal and adult mouse testis. A hierarchical clustered heat map showing the log2 transformed expression values for differentially expressed lncRNAs (absolute fold-change ≥5; P≤0.05) between neonatal (N) and adult (A) mouse testes. The intensity of the color scheme is calibrated to the log2 expression values such that red refers to higher transcript abundance and blue refers to lower transcript abundance. The bar code on the right represents the color scale of the log 2 values. Each column represents the data from one of three biological replicates of each sample.
Relative chromosomal distribution of up and down-regulated lncRNAs. Chromosomes are on the X axis, and the distribution ratio is on the Y axis. Vertical bands show the ratio (up or down-regulated probe number/total probes number) of up and down-regulated lncRNAs derived from each chromosome.
Genomic organization of lncRNAs and their associated protein-coding genes. Schematic diagram illustrating the six categories of genomic association of lncRNAs (orange) with protein-coding genes (blue). Transcription initiation direction is indicated by an arrow (black).
LncRNAs expressed in mouse testis above background.
LncRNAs differentially expressed in neonatal and adult mouse testis.
mRNAs expressed in mouse testis above background.
mRNAs differentially expressed in neonatal and adult mouse testis.
Chromosome distribution of expressed lncRNAs.
Annotation for genomic context of differentially expressed lncRNAs.
Differentially expressed lncRNAs overlapping with annotated microRNAs.
List of lncRNAs associated protein-coding genes that enriched in transcription-related GO terms.
List of lncRNAs associated protein-coding genes that enriched in testis development or spermatogenesis-related GO terms.
Epigentic status in promoter of differentially expressed lncRNAs in adult testis.
Evolutionarily conserved elements in differentially expressed lncRNAs.
List of primers used in the validation of microarray results by Q-PCR.
Conceived and designed the experiments: JW. Performed the experiments: JS. Analyzed the data: JS. Contributed reagents/materials/analysis tools: YL. Wrote the paper: JS.
- 1. Hecht NB (1998) Molecular mechanisms of male germ cell differentiation. Bioessays 20: 555–561.
- 2. Dym M (1994) Spermatogonial stem cells of the testis. Proc Natl Acad Sci U S A 91: 11287–11289.
- 3. Sha J, Zhou Z, Li J, Yin L, Yang H, et al. (2002) Identification of testis development and spermatogenesis-related genes in human and mouse testes using cDNA arrays. Mol Hum Reprod 8: 511–517.
- 4. Pang AL, Taylor HC, Johnson W, Alexander S, Chen Y, et al. (2003) Identification of differentially expressed genes in mouse spermatogenesis. J Androl 24: 899–911.
- 5. Guo X, Shen J, Xia Z, Zhang R, Zhang P, et al. (2010) Proteomic analysis of proteins involved in spermiogenesis in mouse. J Proteome Res 9: 1246–1256.
- 6. Yan N, Lu Y, Sun H, Tao D, Zhang S, et al. (2007) A microarray for microRNA profiling in mouse testis tissues. Reproduction 134: 73–79.
- 7. Gan H, Lin X, Zhang Z, Zhang W, Liao S, et al. (2011) piRNA profiling during specific stages of mouse spermatogenesis. RNA 17: 1191–1203.
- 8. Kapranov P, Cheng J, Dike S, Nix DA, Duttagupta R, et al. (2007) RNA maps reveal new RNA classes and a possible function for pervasive transcription. Science 316: 1484–1488.
- 9. Carninci P, Kasukawa T, Katayama S, Gough J, Frith MC, et al. (2005) The transcriptional landscape of the mammalian genome. Science 309: 1559–1563.
- 10. Guttman M, Amit I, Garber M, French C, Lin MF, et al. (2009) Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals. Nature 458: 223–227.
- 11. Pauli A, Valen E, Lin MF, Garber M, Vastenhouw NL, et al. (2012) Systematic identification of long noncoding RNAs expressed during zebrafish embryogenesis. Genome Res 22: 577–591.
- 12. Pang KC, Dinger ME, Mercer TR, Malquori L, Grimmond SM, et al. (2009) Genome-wide identification of long noncoding RNAs in CD8+ T cells. J Immunol 182: 7738–7748.
- 13. Mercer TR, Dinger ME, Sunkin SM, Mehler MF, Mattick JS (2008) Specific expression of long noncoding RNAs in the mouse brain. Proc Natl Acad Sci U S A 105: 716–721.
- 14. Pang KC, Frith MC, Mattick JS (2006) Rapid evolution of noncoding RNAs: lack of conservation does not mean lack of function. Trends Genet 22: 1–5.
- 15. Struhl K (2007) Transcriptional noise and the fidelity of initiation by RNA polymerase II. Nat Struct Mol Biol 14: 103–105.
- 16. Mattick JS (2011) Long noncoding RNAs in cell and developmental biology. Semin Cell Dev Biol 22: 327.
- 17. Mattick JS (2011) The central role of RNA in human development and cognition. FEBS Lett 585: 1600–1616.
- 18. Wapinski O, Chang HY (2011) Long noncoding RNAs and human disease. Trends Cell Biol 21: 354–361.
- 19. Wang KC, Chang HY (2011) Molecular mechanisms of long noncoding RNAs. Mol Cell 43: 904–914.
- 20. Rinn JL, Chang HY (2012) Genome regulation by long noncoding RNAs. Annu Rev Biochem 81: 145–166.
- 21. Lee T-L, Cheung A-H, Rennert O, Chan W-Y (2011) RNA Expression in Male Germ Cells During Spermatogenesis (Male Germ Cell Transcriptome). In: Zini A, Agarwal A, editors. Sperm Chromatin. New York, NY: Springer. pp. 107–121.
- 22. Anguera MC, Ma W, Clift D, Namekawa S, Kelleher RJ 3rd, et al. (2011) Tsx produces a long noncoding RNA and has general functions in the germline, stem cells, and brain. PLoS Genet 7: e1002248.
- 23. 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. Mol Cell Biol 32: 3140–3152.
- 24. Dinger ME, Amaral PP, Mercer TR, Pang KC, Bruce SJ, et al. (2008) Long noncoding RNAs in mouse embryonic stem cell pluripotency and differentiation. Genome Res 18: 1433–1445.
- 25. Mercer TR, Qureshi IA, Gokhan S, Dinger ME, Li G, et al. (2010) Long noncoding RNAs in neuronal-glial fate specification and oligodendrocyte lineage maturation. BMC Neurosci 11: 14.
- 26. Ulitsky I, Shkumatava A, Jan CH, Sive H, Bartel DP (2011) Conserved function of lincRNAs in vertebrate embryonic development despite rapid sequence evolution. Cell 147: 1537–1550.
- 27. Rackham O, Shearwood AM, Mercer TR, Davies SM, Mattick JS, et al. (2011) Long noncoding RNAs are generated from the mitochondrial genome and regulated by nuclear-encoded proteins. RNA 17: 2085–2093.
- 28. Spiess AN, Walther N, Muller N, Balvers M, Hansis C, et al. (2003) SPEER—a new family of testis-specific genes from the mouse. Biol Reprod 68: 2044–2054.
- 29. Vemuganti SA, Bell TA, Scarlett CO, Parker CE, de Villena FP, et al. (2007) Three male germline-specific aldolase A isozymes are generated by alternative splicing and retrotransposition. Dev Biol 309: 18–31.
- 30. Davis TL, Trasler JM, Moss SB, Yang GJ, Bartolomei MS (1999) Acquisition of the H19 methylation imprint occurs differentially on the parental alleles during spermatogenesis. Genomics 58: 18–28.
- 31. Davis TL, Yang GJ, McCarrey JR, Bartolomei MS (2000) The H19 methylation imprint is erased and re-established differentially on the parental alleles during male germ cell development. Hum Mol Genet 9: 2885–2894.
- 32. Wang PJ, McCarrey JR, Yang F, Page DC (2001) An abundance of X-linked genes expressed in spermatogonia. Nat Genet 27: 422–426.
- 33. Kapranov P, Drenkow J, Cheng J, Long J, Helt G, et al. (2005) Examples of the complex architecture of the human transcriptome revealed by RACE and high-density tiling arrays. Genome Res 15: 987–997.
- 34. Engstrom PG, Suzuki H, Ninomiya N, Akalin A, Sessa L, et al. (2006) Complex Loci in human and mouse genomes. PLoS Genet 2: e47.
- 35. Rinn JL, Kertesz M, Wang JK, Squazzo SL, Xu X, et al. (2007) Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 129: 1311–1323.
- 36. Yu W, Gius D, Onyango P, Muldoon-Jacobs K, Karp J, et al. (2008) Epigenetic silencing of tumour suppressor gene p15 by its antisense RNA. Nature 451: 202–206.
- 37. Kotake Y, Nakagawa T, Kitagawa K, Suzuki S, Liu N, et al. (2011) Long non-coding RNA ANRIL is required for the PRC2 recruitment to and silencing of p15(INK4B) tumor suppressor gene. Oncogene 30: 1956–1962.
- 38. Wang X, Arai S, Song X, Reichart D, Du K, et al. (2008) Induced ncRNAs allosterically modify RNA-binding proteins in cis to inhibit transcription. Nature 454: 126–130.
- 39. Lee J, Kanatsu-Shinohara M, Morimoto H, Kazuki Y, Takashima S, et al. (2009) Genetic reconstruction of mouse spermatogonial stem cell self-renewal in vitro by Ras-cyclin D2 activation. Cell Stem Cell 5: 76–86.
- 40. Rosok O, Sioud M (2004) Systematic identification of sense-antisense transcripts in mammalian cells. Nat Biotechnol 22: 104–108.
- 41. Okada Y, Tashiro C, Numata K, Watanabe K, Nakaoka H, et al. (2008) Comparative expression analysis uncovers novel features of endogenous antisense transcription. Hum Mol Genet 17: 1631–1640.
- 42. Faghihi MA, Wahlestedt C (2009) Regulatory roles of natural antisense transcripts. Nature Reviews Molecular Cell Biology 10: 637–643.
- 43. Yang F, De La Fuente R, Leu NA, Baumann C, McLaughlin KJ, et al. (2006) Mouse SYCP2 is required for synaptonemal complex assembly and chromosomal synapsis during male meiosis. J Cell Biol 173: 497–507.
- 44. Reis EM, Louro R, Nakaya HI, Verjovski-Almeida S (2005) As antisense RNA gets intronic. OMICS 9: 2–12.
- 45. Louro R, El-Jundi T, Nakaya HI, Reis EM, Verjovski-Almeida S (2008) Conserved tissue expression signatures of intronic noncoding RNAs transcribed from human and mouse loci. Genomics 92: 18–25.
- 46. Louro R, Smirnova AS, Verjovski-Almeida S (2009) Long intronic noncoding RNA transcription: expression noise or expression choice? Genomics 93: 291–298.
- 47. Costoya JA, Hobbs RM, Barna M, Cattoretti G, Manova K, et al. (2004) Essential role of Plzf in maintenance of spermatogonial stem cells. Nat Genet 36: 653–659.
- 48. Deng Y, Nie DS, Wang J, Tan XJ, Nie ZY, et al. (2005) Molecular cloning of MSRG-11 gene related to apoptosis of mouse spermatogenic cells. Acta Biochim Biophys Sin (Shanghai) 37: 159–166.
- 49. Nie DS, Liu Y, Juan H, Yang X (2013) Overexpression of human SPATA17 protein induces germ cell apoptosis in transgenic male mice. Mol Biol Rep 40: 1905–1910.
- 50. Trinklein ND, Aldred SF, Hartman SJ, Schroeder DI, Otillar RP, et al. (2004) An abundance of bidirectional promoters in the human genome. Genome Res 14: 62–66.
- 51. Sigova AA, Mullen AC, Molinie B, Gupta S, Orlando DA, et al. (2013) Divergent transcription of long noncoding RNA/mRNA gene pairs in embryonic stem cells. Proc Natl Acad Sci U S A 110: 2876–2881.
- 52. Kalitsis P, Saffery R (2009) Inherent promoter bidirectionality facilitates maintenance of sequence integrity and transcription of parasitic DNA in mammalian genomes. BMC Genomics 10: 498.
- 53. Morris KV, Santoso S, Turner AM, Pastori C, Hawkins PG (2008) Bidirectional transcription directs both transcriptional gene activation and suppression in human cells. PLoS Genet 4: e1000258.
- 54. Lee JY, Khan AA, Min H, Wang X, Kim MH (2012) Identification and characterization of a noncoding RNA at the mouse Pcna locus. Mol Cells 33: 111–116.
- 55. Deng W, Lin H (2002) miwi, a murine homolog of piwi, encodes a cytoplasmic protein essential for spermatogenesis. Dev Cell 2: 819–830.
- 56. Takeuchi A, Mishina Y, Miyaishi O, Kojima E, Hasegawa T, et al. (2003) Heterozygosity with respect to Zfp148 causes complete loss of fetal germ cells during mouse embryogenesis. Nat Genet 33: 172–176.
- 57. Nam JW, Bartel DP (2012) Long noncoding RNAs in C. elegans. Genome Res 22: 2529–2540.
- 58. Liu J, Jung C, Xu J, Wang H, Deng S, et al. (2012) Genome-Wide Analysis Uncovers Regulation of Long Intergenic Noncoding RNAs in Arabidopsis. Plant Cell 24: 4333–4345.
- 59. Cabili MN, Trapnell C, Goff L, Koziol M, Tazon-Vega B, et al. (2011) Integrative annotation of human large intergenic noncoding RNAs reveals global properties and specific subclasses. Genes Dev 25: 1915–1927.
- 60. Tsai MC, Spitale RC, Chang HY (2011) Long intergenic noncoding RNAs: new links in cancer progression. Cancer Res 71: 3–7.
- 61. Wang KC, Yang YW, Liu B, Sanyal A, Corces-Zimmerman R, et al. (2011) A long noncoding RNA maintains active chromatin to coordinate homeotic gene expression. Nature 472: 120–124.
- 62. Schmidt JA, Avarbock MR, Tobias JW, Brinster RL (2009) Identification of glial cell line-derived neurotrophic factor-regulated genes important for spermatogonial stem cell self-renewal in the rat. Biol Reprod 81: 56–66.
- 63. He S, Su H, Liu C, Skogerbo G, He H, et al. (2008) MicroRNA-encoding long non-coding RNAs. BMC Genomics 9: 236.
- 64. Fabian MR, Sonenberg N, Filipowicz W (2010) Regulation of mRNA translation and stability by microRNAs. Annu Rev Biochem 79: 351–379.
- 65. Hansen TB, Wiklund ED, Bramsen JB, Villadsen SB, Statham AL, et al. (2011) miRNA-dependent gene silencing involving Ago2-mediated cleavage of a circular antisense RNA. EMBO J 30: 4414–4422.
- 66. Cai X, Cullen BR (2007) The imprinted H19 noncoding RNA is a primary microRNA precursor. RNA 13: 313–316.
- 67. Ro S, Park C, Sanders KM, McCarrey JR, Yan W (2007) Cloning and expression profiling of testis-expressed microRNAs. Dev Biol 311: 592–602.
- 68. Zamudio NM, Chong S, O'Bryan MK (2008) Epigenetic regulation in male germ cells. Reproduction 136: 131–146.
- 69. Wu SC, Kallin EM, Zhang Y (2010) Role of H3K27 methylation in the regulation of lncRNA expression. Cell Res 20: 1109–1116.
- 70. Sandoval J, Heyn H, Moran S, Serra-Musach J, Pujana MA, et al. (2011) Validation of a DNA methylation microarray for 450,000 CpG sites in the human genome. Epigenetics 6: 692–702.
- 71. Ponjavic J, Ponting CP, Lunter G (2007) Functionality or transcriptional noise? Evidence for selection within long noncoding RNAs. Genome Res 17: 556–565.
- 72. Bird A (1992) The essentials of DNA methylation. Cell 70: 5–8.
- 73. Saxonov S, Berg P, Brutlag DL (2006) A genome-wide analysis of CpG dinucleotides in the human genome distinguishes two distinct classes of promoters. Proc Natl Acad Sci U S A 103: 1412–1417.
- 74. Nottke A, Colaiacovo MP, Shi Y (2009) Developmental roles of the histone lysine demethylases. Development 136: 879–889.
- 75. Martin C, Zhang Y (2005) The diverse functions of histone lysine methylation. Nat Rev Mol Cell Biol 6: 838–849.
- 76. Mikkelsen TS, Ku M, Jaffe DB, Issac B, Lieberman E, et al. (2007) Genome-wide maps of chromatin state in pluripotent and lineage-committed cells. Nature 448: 553–560.
- 77. Bernstein BE, Mikkelsen TS, Xie X, Kamal M, Huebert DJ, et al. (2006) A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125: 315–326.
- 78. Marques AC, Ponting CP (2009) Catalogues of mammalian long noncoding RNAs: modest conservation and incompleteness. Genome Biol 10: R124.
- 79. Chodroff RA, Goodstadt L, Sirey TM, Oliver PL, Davies KE, et al. (2010) Long noncoding RNA genes: conservation of sequence and brain expression among diverse amniotes. Genome Biol 11: R72.
- 80. Meola N, Pizzo M, Alfano G, Surace EM, Banfi S (2012) The long noncoding RNA Vax2os1 controls the cell cycle progression of photoreceptor progenitors in the mouse retina. RNA 18: 111–123.
- 81. Askarian-Amiri ME, Crawford J, French JD, Smart CE, Smith MA, et al. (2011) SNORD-host RNA Zfas1 is a regulator of mammary development and a potential marker for breast cancer. RNA 17: 878–891.
- 82. Schonrock N, Harvey RP, Mattick JS (2012) Long Noncoding RNAs in Cardiac Development and Pathophysiology. Circ Res 111: 1349–1362.
- 83. Klattenhoff CA, Scheuermann JC, Surface LE, Bradley RK, Fields PA, et al. (2013) Braveheart, a long noncoding RNA required for cardiovascular lineage commitment. Cell 152: 570–583.
- 84. Roberts KA, Abraira VE, Tucker AF, Goodrich LV, Andrews NC (2012) Mutation of Rubie, a novel long non-coding RNA located upstream of Bmp4, causes vestibular malformation in mice. PLoS One 7: e29495.
- 85. Kutter C, Watt S, Stefflova K, Wilson MD, Goncalves A, et al. (2012) Rapid turnover of long noncoding RNAs and the evolution of gene expression. PLoS Genet 8: e1002841.
- 86. Derrien T, Johnson R, Bussotti G, Tanzer A, Djebali S, et al. (2012) The GENCODE v7 catalog of human long noncoding RNAs: analysis of their gene structure, evolution, and expression. Genome Res 22: 1775–1789.
- 87. Mercer TR, Dinger ME, Mattick JS (2009) Long non-coding RNAs: insights into functions. Nat Rev Genet 10: 155–159.
- 88. Girard A, Sachidanandam R, Hannon GJ, Carmell MA (2006) A germline-specific class of small RNAs binds mammalian Piwi proteins. Nature 442: 199–202.
- 89. Ponjavic J, Oliver PL, Lunter G, Ponting CP (2009) Genomic and transcriptional co-localization of protein-coding and long non-coding RNA pairs in the developing brain. PLoS Genet 5: e1000617.
- 90. Munoz MJ, de la Mata M, Kornblihtt AR (2010) The carboxy terminal domain of RNA polymerase II and alternative splicing. Trends Biochem Sci 35: 497–504.
- 91. Zhao J, Ohsumi TK, Kung JT, Ogawa Y, Grau DJ, et al. (2010) Genome-wide identification of polycomb-associated RNAs by RIP-seq. Mol Cell 40: 939–953.
- 92. Zhao J, Sun BK, Erwin JA, Song JJ, Lee JT (2008) Polycomb proteins targeted by a short repeat RNA to the mouse X chromosome. Science 322: 750–756.
- 93. Tsai MC, Manor O, Wan Y, Mosammaparast N, Wang JK, et al. (2010) Long noncoding RNA as modular scaffold of histone modification complexes. Science 329: 689–693.
- 94. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25: 402–408.
- 95. Huang da W, Sherman BT, Lempicki RA (2009) Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 4: 44–57.
- 96. Gardiner-Garden M, Frommer M (1987) CpG islands in vertebrate genomes. J Mol Biol 196: 261–282.
- 97. Siepel A, Bejerano G, Pedersen JS, Hinrichs AS, Hou M, et al. (2005) Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res 15: 1034–1050.