Skip Navigation


Human Molecular Genetics Advance Access originally published online on February 18, 2008
Human Molecular Genetics 2008 17(11):1631-1640; doi:10.1093/hmg/ddn051
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
17/11/1631    most recent
ddn051v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Okada, Y.
Right arrow Articles by Kiyosawa, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Okada, Y.
Right arrow Articles by Kiyosawa, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Comparative expression analysis uncovers novel features of endogenous antisense transcription

Yuki Okada1,2, Chiaki Tashiro4, Koji Numata5, Kazufumi Watanabe6, Hajime Nakaoka7, Naoyuki Yamamoto7, Kazue Okubo8, Rieko Ikeda5, Rintaro Saito1,3,*, Akio Kanai1,2,3, Kuniya Abe4,5, Masaru Tomita1,2,3 and Hidenori Kiyosawa4,5,*

1 Institute for Advanced Biosciences, Keio University, Tsuruoka 997-0017, Japan 2 Systems Biology Program, Graduate School of Media and Governance 3 Faculty of Environment and Information Studies, Keio University, Fujisawa 252-8520, Japan 4 Graduate School of Life and Environmental Sciences, University of Tsukuba, Ibaraki 305-0006, Japan 5 Technology and Development Team for Mammalian Cellular Dynamics, BioResource Center (BRC), RIKEN Tsukuba Institute, Ibaraki 305-0074, Japan 6 Custom Biotechnology Service Group, Hokkaido System Science Co. Ltd, 2-1 Shinkawa Nishi 2-1, Kita-ku, Sapporo 001-0932, Japan 7 C’s Lab Co. Ltd, Maruito Sapporo Bldg, 7F, Kita 2 Nishi 1, Kita-ku, Sapporo 060-0002, Japan 8 Genostaff Inc., Kawauchi Bldg. 6F 1-4-4, Nezu, Bunkyo-Ku, Tokyo 113-0031, Japan

* To whom correspondence should be addressed. Tel/Fax: +81 466475099; Email: rsaito{at}sfc.keio.ac.jp (R.S.); Tel/Fax: +81 298369199; Email: kiyosawa{at}rtc.riken.jp (H.K.)

Received December 7, 2007; Accepted February 13, 2008


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 CONTRIBUTIONS
 REFERENCES
 
Increasing numbers of sense–antisense transcripts (SATs), which are transcribed from the same chromosomal location but in opposite directions, have been identified in various eukaryotic species, but the biological meanings of most SATs remain unclear. To improve understanding of natural sense–antisense transcription, we performed comparative expression profiling of SATs conserved among humans and mice. Using custom oligo-arrays loaded with probes that represented SATs with both protein-coding and non-protein-coding transcripts, we showed that 33% of the 291 conserved SATs displayed identical expression patterns in the two species. Among these SATs, expressional balance inversion of sense–antisense genes was mostly observed in testis at a tissue-specific manner. Northern analyses of the individual conserved SAT loci revealed that: (i) a smeary hybridization pattern was present in mice, but not in humans, and (2) small RNAs (about 60 to 80 nt) were detected from the exon-overlapping regions of SAT loci. In addition, further analyses showed marked alteration of sense–antisense expression balance throughout spermatogenesis in testis. These results suggest that conserved SAT loci are rich in potential regulatory roles that will help us understand this new class of transcripts underlying the mammalian genome.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 CONTRIBUTIONS
 REFERENCES
 
The latest breakthroughs in the field of mammalian genomic and transcriptomic analysis have started to shed light on a new world of transcriptome complexity. Genomic regions that were previously considered to contain ‘junk DNA’ with no transcriptional activity are now known to be transcribed throughout the genome (1), yielding multitudes of new-class RNAs, including sense–antisense transcripts (SATs).

SATs are pairs of endogenously encoded transcripts that are transcribed from both strands of the DNA and are capable of forming sequences complementary to each other (2). SATs can be derived from both protein-coding genes and non-protein-coding genes, including exonic, intronic and intergenic sequences (3,4).

Increasing numbers of SATs have been identified in various eukaryotic organisms, including humans (3,5,6), mice (4,5,7), rats and chickens (8), flies (9), nematodes (8), rice (10), Arabidopsis (11,12) and yeast (13). These overlapping transcripts are estimated to cover from 5 to 29% of the whole transcript in animals, and from 7 to 30% in plants. Overall, the results of these studies confirm that SATs are phenomena that are frequently observable in the animal, plant and fungal kingdoms.

The finding that numerous SAT loci are harbored within the eukaryotic genome had led to further questions about the functional role of SATs. Studies performed in various organisms have suggested that SATs can participate in a broad range of regulatory events (14). For example, SAT coexpressed within a cell can lead to various kinds of antisense regulation (2), such as RNA masking (15), RNA editing (16–18) and RNA interference (RNAi) (19). Current reports have revealed that SATs have the potential to become RNAi targets in Caenorhabditis elegans (20), Schizosaccharomyces pombe (21) and Arabidopsis thaliana (22), indicating that mammalian SATs may also play a role in the RNAi pathway. SATs are also linked to monoallelic gene expression through mechanisms that include genomic imprinting (23), X-chromosome inactivation (24) and DNA methylation (25,26). However, quantities of experimentally validated pairs are still limited and vast numbers of unannotated SATs await the revelation of their cellular roles.

Cross-species comparative analysis of large-scale data on, for example, nucleotide sequences, genomic structure and gene expression is considered to be an effective approach to enriching our knowledge of the functionally important elements (27). Although several SAT comparative analyses that have used bioinformatics approaches have been published in the literature (5,28,29), no groups have taken large-scale expression profiling and experimental validation further to cover individual conserved SATs. Here, we implemented a comparative analysis of SAT pairs that especially encompassed human–mouse orthologous genes. We then performed microarray-based expression profiling and experimental validation of the selected SAT pairs. Through the results of expression analysis of mouse- and human-conserved SATs (cSATs), we demonstrate the novel characteristics of the transcription of evolutionarily conserved SAT loci at both the molecular and genome-wide levels, and we discuss the potential regulatory mechanisms involved.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 CONTRIBUTIONS
 REFERENCES
 
Microarray analysis of human- and mouse-conserved SATs
For the comprehensive SAT expression profile, custom-made oligo DNA (60-mer) chips targeting 1948 pairs of SATs in mouse and 1486 pairs in human that consisted of the full-length cDNA resources and NCBI RefSeq mRNA collection were used, including both pc (protein-coding) and npc (non-protein-coding) transcripts (Table 1). In humans, the proportion of SAT pairs containing npc genes (pc/npc and npc/npc) was smaller than in mice because a limited number of genes had been annotated as npc in the initial cDNA collection. Starting with these datasets, we found cSATs of which the genomic structure and nucleotide sequence were conserved (See Materials and Methods). By focusing on the conservation of exon-overlapping SATs, we hypothesized that our extraction would enrich the number of functionally important SATs. Therefore, we collected those that exhibited sense–antisense overlapping in exonic regions in both the human and mouse genomes. Finally, 291 pairs of SATs were selected as cSAT candidates and obtained for further comparative analysis (Supplementary Material, Fig. S1).


View this table:
[in this window]
[in a new window]

  		Table 1. Classifications of SAT and cSAT pairs in mouse and human

 
We initially conducted a clustering analysis based on the expression ratios of sense and antisense gene pairs among cSATs. Microarray hybridization experiments were undertaken for four tissues and one fibroblast cell type (brain, heart, liver, testis and murine NIH3T3 or human HF19 cells) in both humans and mice. Target samples were prepared by oligo dT priming or a random priming method from each RNA sample (see Materials and Methods). Although there were some cSATs for which the expression ratios were comparatively similar among the tested samples in both species (Fig. 1Aa and Ba), others exhibited different patterns of expression balance (Fig. 1Ac and Bc), and others possessed completely reversed expression ratios in humans and mice (Fig. 1Ab and Bb). By the oligo dT priming method, 30% (87 pairs) of SATs had similar expression ratios among all tissues in both species; this value was 33% (96 pairs) by the random priming method. Notably, the number of similarly expressed cSATs was significantly higher than that of randomly paired pseudo-cSAT groups (P < 0.01; test for the mean) (Supplementary Material, Fig. S2). Since this outcome was observed only in exon-overlapping cSATs, and not in non-exon-overlapping conserved bidirectional gene pairs, our hypothesis—that conservation of not only nucleotide sequences but also exon-overlapping genomic structure may enrich functionally crucial features—is reasonable for detailed human–mouse comparative studies. Most of the pairs that exhibited similar expression patterns represented invariable intensity ratios between sense and antisense genes, but nine pairs showed expression ratio turnover specific to a single tissue by the oligo dT priming method; the number was six pairs by the random priming method. Notably, six of the nine pairs by the oligo dT priming method, followed by two of the six pairs by the random priming method, exhibited complete alteration of expression balance, specifically in the testis. In addition, 54 pairs showed completely reversed expression ratios between human and mouse by the oligo dT priming method, and 67 pairs by the random priming method (Fig. 1Ab and Bb). Remnant cSATs exhibited dissimilar expression patterns; there were 150 of these pairs by the oligo dT priming method and 128 pairs by the random priming method (Fig. 1Ac and Bc).


Figure 1
View larger version (62K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. Clustering analysis of the expression balance between the sense and antisense genes from cSATs in five types of sample. Heatmaps of the expression ratios of sense and antisense genes from cSAT pairs across four tissues and one fibroblast cell type (brain, heart, liver, testis and murine NIH3T3 cells or human HF19 cells) in the mouse (left columns) and human (right columns). (A) reflects the microarray data obtained by the oligo dT priming method, whereas (B) are data from the random priming method. When the sense gene expression is three times that of the antisense gene expression, the color shifts to red. In the opposite situation, the color is green. cSATs for which the expression ratios of the antisense and sense genes are comparatively similar among the tested samples in both species are grouped in (a). cSATs in which the two genes of the pair possess completely reversed expression ratios between human and mouse are shown in (b), and those that have neither of the above two patterns are shown in (c). Note that the direction of the human SAT genes are unified based on the published mouse genome sequence. Experimentally examined cSATs (cSAT-025, cSAT-042 and cSAT-177) are enclosed in yellow boxes. cSATs with the expression ratio turnover specific to a single tissue are marked in asterisk.

 
Although microarray experiments are powerful tools for overviewing the expression properties of genes in particular tissues and cell types, they are limited to detailed analysis of the transcriptional states of individual genes. To explore in detail the features of individual cSAT transcription by northern and in-situ hybridization analyses, we selected three cSAT pairs (cSAT-025, cSAT-042 and cSAT-177) that included non-protein-coding-like transcripts on one strand and similar genomic overlapping between the two species (Fig. 1).

Northern hybridization analysis of selected cSATs
We first carried out an expression analysis of the cSAT named cSAT-042. In the mouse, this cSAT contains a gene-encoding hypothetical serine-rich region containing protein (AK012036 [GenBank] ) in the plus strand and an unannotated EST (AK039347 [GenBank] ) in the minus strand (Fig. 2A). The corresponding cSAT in the human genome is composed of an unannotated EST (AK025621 [GenBank] ) in the plus strand and an unnamed protein coding gene (AK027898 [GenBank] ) in the minus strand (Fig. 2B). Both SATs are oriented in a head-to-head overlapping structure. The AK012036 [GenBank] gene encodes an ORF of 483 nt and has high sequence similarity to the AK027898 [GenBank] gene, suggesting that these two transcripts produce an orthologous protein (tblastx identity = 72.89%). On the other hand, the antisense-transcribed unit against these protein-coding genes presumably lacks protein-coding potential in both species. Microarray experiments detected expression signals at varying levels for AK012036 [GenBank] , AK039347 [GenBank] and AK027898 [GenBank] , but not for the non-protein-coding-like transcript AK025621 [GenBank] in humans (Supplementary Material, Fig. S3A–D).


Figure 2
View larger version (71K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2. Northern hybridization analysis of cSAT-042 in mouse and human. (A) cDNA mapping patterns and positions of AK012036 [GenBank] (protein-coding transcript, sense strand) and AK039347 [GenBank] (non-protein-coding transcript, antisense strand) in the mouse genome. The exons are shown by the solid bars. Direction of transcription is displayed on the last exon. (B) cDNA mapping patterns and positions of AK027898 [GenBank] (protein-coding transcript, antisense strand, possible ortholog of AK012036 [GenBank] ) and AK025621 [GenBank] (non-protein-coding transcript, sense strand) in the human genome. Genes that correspond to the mouse SAT gene ortholog are displayed by black bars. (C) Northern hybridization of AK012036 [GenBank] (sense) from total RNA and mRNA of (1) fibroblast (NIH3T3), (2) brain and (3) testis. Each lane was loaded with 10 µg of total RNA or 1 µg of mRNA. The band patterns for lane 2 (poly(A)+ RNA) were similar to those in lanes 1 and 3, but for lane 2 the band intensity on the exposed film was greater. (D) Northern hybridization of AK039347 [GenBank] (antisense). (E) Northern hybridization of AK027898 [GenBank] (antisense) from total RNA and mRNA of (1) fibroblast (HF19), (2) brain and (3) testis. Each lane was loaded with 10 µg of total RNA or 0.5 µg of mRNA. (F) Northern hybridization of AK025621 [GenBank] (sense). Thick black arrowheads indicate positions of the 28S and 18S ribosomal RNA. Light bands in the lower regions of the gels are indicated by narrow black arrows.

 
Northern hybridization of AK012036 [GenBank] produced no specific band patterns—only a smear—in the case of total RNA, whereas several specific bands were detected with poly(A)+ RNAs (Fig. 2C). Smear patterns were also observed for AK039347 [GenBank] in both RNA types, which is the antisense transcribed gene (Fig. 2D). Northern hybridization of human AK027898 [GenBank] , a possible ortholog of mouse AK012036 [GenBank] , gave several specifically hybridized bands in poly(A)+ RNA samples; in the case of total RNA, a smear was detected below the 18S ribosomal RNA band. The human non-protein-coding-like transcript AK025621 [GenBank] had no smeary hybridization patterns, unlike in the mouse cSAT (Fig. 2F). While no significant expression was observed for AK025621 [GenBank] from the microarray data analysis (Supplementary Material, Fig. S3C–D), we detected a clear band in northern analysis. Although both expression analyses results seemed to contradict with the expression of AK025621 [GenBank] , it is necessary to keep in mind that the probe sequences used in the northern and microarray analyses were different. The microarray probes were constructed as one 60-mer probe per gene, whereas the probes used in the northern analysis used full-length or several hundreds nucleotides of the transcript. Thus, differences between the expression profiles may have occurred between the two analytical results because of differences in probe length and position. Overall, human cSAT-042 tended to have fewer smear patterns and the specific bands were also detected in the non-protein-coding-like transcripts. Interestingly, a similar phenomenon was observed with cSAT-025 (Supplementary Material, Fig. S4C–F) and cSAT-177 (Supplementary Material, Fig. S6C–F).

Gene-expression properties of SAT and non-SAT loci
As we had previously reported that the majority of mouse SATs tended to be poly(A)– (30), we hypothesized that the abundance of poly(A)+ and poly(A)– RNAs originating from mouse and human SAT loci might differ, since the cSATs exhibited northern smearing patterns that differed between the two species. We used the oligo dT and random priming methods to compare the expression profiles of SAT (1948 pairs in the mice and 1561 pairs in humans) and non-SAT (4223 in mice and 2271 in humans) genes obtained from five tissues and cell types. We expected that human SATs acquired by the random priming method would show results different from those of mice if the majority of the human SATs were abundant in poly(A)+ RNAs. As expected, on microarray analysis the average signal intensity for mouse SATs by the random priming method was higher than that by the oligo dT priming method (Fig. 3A). In contrast, for human SATs, the oligo dT priming method gave higher signals than did the random priming method (P < 0.05; Welch two sample t-test) (Fig. 3B). These trends were observed in non-SAT genes of mouse and human, implying that human SAT transcripts generally resemble non-SAT mouse genes rather than mouse SAT genes.


Figure 3
View larger version (25K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 3. Signal intensities for SAT and non-SAT candidates, averaged over all tissue samples used in the microarray analysis. Average signal intensities from all tissue samples from mice (A) and humans (B) are displayed. The light gray bars represent average signal intensities of candidates labeled by the oligo dT priming method and the dark gray bars represent those by the random priming method, for samples from five types of tissue or cell. Error bars indicate the standard error of the mean.

 
Small RNA molecules generated from SAT loci
During northern hybridization, we frequently detected bands on the lower part of the gels in both mice and humans; these were presumably small processed RNAs. (The detected bands are indicated by narrow black arrows in Fig. 2, Supplementary Material, Figs S4 and S6.) To further examine the transcriptional landscape originating from the sense–antisense overlapping region, we performed northern analysis with single-stranded oligo DNA probes derived from the exon-overlapping regions of cSAT-025 (Fig. 4A). cSAT-025 consists of transcript encoding a DENN-domain-containing protein (mouse: AK050033 [GenBank] , official symbol: Denn2d; human: NM_024901 [GenBank] , official symbol: DENN2D) and a probable non-protein-coding transcript (mouse: AK080971 [GenBank] , human: AK057181 [GenBank] ), located in a head-to-head overlapping manner (Supplementary Material, Fig. S4A and B). In all tested samples, northern hybridization of the murine AK050033 [GenBank] gene detected large amounts of small RNAs that originated from hybridization with the probe containing the 96 nt repeated sequence (SINE, B1) (Supplementary Material, Fig. S4C), because these bands faded during another trial of northern hybridization using probes that removed the corresponding repetitive region (data not shown).


Figure 4
View larger version (34K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 4. Northern hybridization of mouse cSAT-025 with a series of 30- to 40-mer single-stranded oligo DNA probes chosen from intron and exonic regions. (A) Extended mapping pattern. The thick horizontal lines with numbers underneath are the approximate positions of the 30- to 40-mer single-stranded oligo DNA probes. Two complement probes were designed for each position, corresponding to the sense and antisense strands. (B) Hybridization results for probes targeting AK050033 [GenBank] (represented as S), followed by its counterpart, AK080971 [GenBank] (represented as AS) in the six different numbered regions in (A). Each lane contained 10 µg of total RNA from the mouse fibroblast (NIH3T3). Bands indicate RNA fragments below 100 nt. (C) Northern hybridization of the S3 probe from total RNA of fibroblast, brain and testis. RNA was separated using 6% polyacrylamide–8 M urea gel and used for analysis. Each lane was loaded with 5 µg of total RNA. (D) Northern hybridization of the AS2 probe by a procedure similar to that described in (C). U1 gene expression was confirmed in both (C) and (D).

 
In the northern analysis with single-stranded oligo DNA probes, the probes at position no. 3 in the sense strand (S3) and position no. 2 in the antisense strand (AS2) detected small RNA molecules of less than 100 nt (Fig. 4B). In addition, weaker bands of a similar size were also detected by probes at positions no. 4 and 5 in the antisense strand (AS4 and AS5, respectively; Fig. 4B). More accurately, the lengths of these small RNA molecules were defined as around 80 nt (S3) and 60 nt (AS2), from the results of polyacrylamide gel electrophoresis (Fig. 4C and D). The results indicate that small RNA molecules of distinct size were processed from both strands of the SAT loci. Because the probe sequences were confirmed to be specific to the cSAT-025 loci, the small RNA molecules detected may not be repetitive elements of the mouse genome.

S3 seems to be located within an intronic region of the sense transcript AK050033 [GenBank] (Fig. 4A), but UCSC Genome Browser (31), on an mm8 assembly, showed the presence of an EST (Accession ID: BY708215 [GenBank] ) that was oriented in the sense strand and covered the entire region of the probe sequence. Therefore, a novel alternative isoform with an exonic region overlapping in S3 may exist at this locus, which may have originated specifically in the NIH3T3 cell line.

Dynamics of cSAT expression in spermatogenesis
The third cSAT, which we named cSAT-177, comprises a gene specifically expressed in the testis, glyceraldehyde-3-phosphate dehydrogenase spermatogenic (mouse: NM_008085 [GenBank] , official symbol: Gapdhs; human: NM_014364 [GenBank] , official symbol: GAPDHS) and a hypothetical protein-coding transcript (mouse: AK049951 [GenBank] ; human: AK092032 [GenBank] ) in a tail-to-tail overlapping structure (Supplementary Material, Fig. S6A and B). According to the cDNA mapping data, whereas murine AK049951 [GenBank] is formed in multiple exons, human AK092032 [GenBank] exists as a single exon. Microarray analysis exposed adequate expression of AK049951 [GenBank] , NM_008085 [GenBank] and NM_014364 [GenBank] , but not of AK092032 [GenBank] , which is the antisense counterpart of the glyceraldehyde-3-phosphate dehydrogenase spermatogenic gene in human (Supplementary Material, Fig. S7A–D). As in the case of cSAT-024, poly (A)– RNA seemed to be abundant only in the mouse AK049951 [GenBank] gene, as shown by the higher signal intensity by the random priming method than by the oligo dT priming method. (Supplementary Material, Fig. S7A and B).

As previously described, our microarray experiments indicated that most of the between-species reversals of sense-antisense expression ratios of cSATs in a single tissue occurred in the testis (Fig. 1A). We speculated that this trend might reflect the dynamic development underlying this tissue, i.e. spermatogenesis. To disclose the status of sense–antisense gene expression in the testis, we performed in-situ hybridization analysis of testis samples, targeting mouse cSAT-177, which showed a reversed expression balance of AK049951 [GenBank] and NM_008085 [GenBank] specific to the testis from the microarray data analysis (Supplementary Material, Fig. S7A and B). The AK049951 [GenBank] gene was detected around the outsides of the seminiferous tubules, in the germ cells that function in the early stages of spermatogenesis (Fig. 5A). In contrast, the NM_008085 [GenBank] gene was detected in the insides of the seminiferous tubules, in the germ cells that play a role in the late stages of spermatogenesis (Fig. 5B). This observation was confirmed by microarray experiments obtained from three fractionated testis samples (pachytene spermatocyte, round spermatid and elongated spermatid) (Fig. 5C). The data (random priming method) well reflected the in situ hybridization results; AK049951 [GenBank] and NM_008085 [GenBank] were reciprocally expressed in accordance with the germ cell differentiation from pre-sperm cells to mature sperm.


Figure 5
View larger version (83K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 5. In situ hybridization analysis for cSAT-177 in mouse testis. (A) In situ hybridization performed on AK049951. (B) In situ hybridization performed on NM_008085 [GenBank] . (C) Microarray analysis of testis fraction samples labeled by the random priming method. Y-axis is signal intensity.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 CONTRIBUTIONS
 REFERENCES
 
Constructed microarray platform for SAT expression in both mouse and human provides an important resource for comparative expression analysis of SATs. By using microarray datasets from mouse and human SATs, we successfully obtained cSATs with similar expression ratios between the two species. Interestingly, while exon-overlapping cSATs showed a higher probability of being similarly expressed than random shuffled pairs, non-exon-overlapping bidirectional gene pairs that were conserved in the two species did not show this phenomenon, suggesting that the expression of exon-overlapping cSAT pairs is mutually controlled, rather than being the result of mere transcriptional noise. One previous large-scale transcriptome analysis found that mammalian SATs had a higher probability of being coordinately expressed compared with random transcript pairs (4,5,7). However, because this and other similar analyses were performed with public expression data that lack non-protein-coding genes, the expressional properties of the genes involved in SATs were largely unknown. Some of the cSATs with similar expressional properties are formed in pc/npc pairs or nc/nc pairs, suggesting that the expression of non-protein-coding genes may be controlled for further antisense regulation.

Our further investigations of individual mouse and human cSATs also disclosed SAT characteristics that differed between the two species. Among the three randomly chosen cSAT pairs, all results correspond to those of previous studies (25,30,32,33), providing evidence that transcripts from SAT gene loci are abundant in poly(A)– RNAs (Supplementary Material, Figs S5 and S7, A and B) and are capable of exhibiting northern hybridization smearing despite the transcripts’ coding potential (Supplementary Material, Figs S4 and S6, C and D). However, northern analysis of the human transcriptome generally showed specific hybridized band patterns (Supplementary Material, Figs S4 and S6, E and F), followed by remarkably low expression in the case of the non-protein-coding transcripts of the cSAT-042 (AK025621 [GenBank] ) and cSAT-025 (AK057181 [GenBank] ) pairs and the antisense transcript of the cSAT-177 (AK092032 [GenBank] ) pair obtained from the microarray experiments (Supplementary Material, Figs S3, S5 and S7, C and D). It is not clear how the human tissue samples and total RNAs that we used were prepared, because RNA samples from the brain, heart, liver and testis were purchased from a commercial company. However, this specific pattern of expression is independent of whether or not the RNA samples were purchased, since no hybridization smearing was apparent in the testing of human cultured cell lines (HUC-Fm and HepG2) (data not shown). In addition, the mean signals of human SAT genes in microarray profiling were dissimilar to those of the mouse SAT genes (Fig. 3). The distinct nature of the expression patterns in humans and mice has also been argued by the systematic search and northern analysis of conserved non-protein-coding transcripts (34). Taken together, we suggest that the regulation and/or turnover of transcripts from many SAT loci occur in a species-specific manner in spite of their cross-species conservation.

Because of the potential for generating natural dsRNA, the relationship between RNAi and SATs has been of great interest to researchers. In fission yeast, transcripts derived from the centromeric repeat sequences are used to form dsRNA to activate RNAi for gene regulation (21). Thus, as we detected possible processed small RNAs from the probe containing a repetitive region (Supplementary Material, Fig. S4C), it may be possible that the repeat sequence was somehow processed for generating small RNAs. We also detected small RNAs of approximately 80 nt and 60 nt during northern analysis of the cSAT-025 pairs derived from the regions capable of producing dsRNA (Fig. 4B–D). We implemented similar northern analyses of cSAT-042 pairs; they also displayed multiple-sized band patterns, including small RNA bands, ranging from about 0.3 to 4 kb in length, at the exonic overlapping regions (data not shown). Additionally, northern analysis performed with fractionated cytoplasmic and nuclear RNAs revealed that these small RNA molecules were exclusively detectable in the cytoplasm but not in the nucleus (data not shown). We also have previously observed small RNAs associated with the SAT loci, specifically in the mouse (30), suggesting that the presence of small RNAs originating from the SAT loci may be a common phenomenon. Interestingly, a recent systematic genome-wide study highlighted large amounts of steady-state short RNAs (<200 nt) enriched in the mammalian genome (35). Short RNAs were reported to be enriched in the synthetically conserved sequences of humans and mice, with more frequent appearances among genes than in the intergenic regions, suggesting that the small RNAs originating from the cSAT loci may be related to these unannotated RNAs. In our preliminary result, small RNA that was observed in Fig. 4D was also detected on the northern blot with the Dicer-deficient mouse ES cells (36) suggesting that these size of small RNAs (approximately 60 to 80 nt) may be generated through the Dicer-independent pathway.

The RNA expression pattern resulting from in situ hybridization of the cSAT-177 pair in the mouse testis reflects the fact that the dynamic reactions of spermatogenesis are occurring continuously in this tissue. Interestingly, the microarray data obtained from the testicular fraction revealed a reversal of the SAT expression ratio as the differentiation of sperm cells proceeded. Because there may be a need for rapid gene regulation during spermatogenesis, regulation within the RNA level may be more beneficial than in the protein level, since it can save more time and energy within the cell. In total, in our dataset, 16 mouse cSAT pairs from oligo dT priming and 33 pairs from random priming displayed reversed expression ratios during spermatogenesis (Supplementary Material, Tables S2 and S3). This observation of an inverted expression pattern among sense and antisense genes is reminiscent of two models of antisense gene regulation, namely transcriptional interference and RNAi, where one gene regulates another antisense counterpart (14,22). Although further detailed analyses are required to characterize their detailed molecular function, our data suggests that antisense-mediated gene regulation might occur in spermatogenesis in the case of cSAT-177 and many other cSATs.

The overall expression profiles of cSATs appear to represent various characteristics in terms of the biological functions that the transcripts may possess. Further functional analyses are needed to give us an understanding of these biological functions, but the evolutionarily conserved SATs are positively enriched with new features that will help to reveal the underlying functions of these transcripts in the mammalian genome.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
SUPPLEMENTARY MATERIAL
 FUNDING
 CONTRIBUTIONS
 REFERENCES
 
Procedures for mapping cDNA sequences to the corresponding genomes
Data from the mapping of all FANTOM2 (http://fantom.gsc.riken.jp/) and NCBI RefSeq (http://www.ncbi.nih.gov/RefSeq) sequences to the mouse genome (MGSCv3 (37)), including list of SAT pairs, were collected from the results of our earlier research (4).

Human full-length cDNA resources were downloaded from H-invitational (http://www.jbirc.aist.go.jp) and NCBI RefSeq (http://www.ncbi.nih.gov/RefSeq). Human genome sequences (hg17) were obtained from the UCSC Genome Bioinformatics site (http://hgdownload.cse.ucsc.edu/downloads.html).

cDNA and mRNA sequences were scanned against human genome sequences by using the BLAT (38) and SIM4 (39) programs. A BLAT search was initially conducted to align the whole cDNA and mRNA sequences onto the whole-genome sequence. When the same cDNAs or mRNAs were aligned to multiple areas of the genome, those that were mapped to the longest region on the genome sequence were selected for analysis. The SIM4 program was used to further realign the matched sequences in accordance with their exon–intron splicing junctions with the corresponding genome sequences. During this process, the intron boundary consensus sequence (GT/AG) for every aligned transcript with multiple exons was examined, selecting only cDNAs or mRNAs that possessed the GT/AG consensus. Regions aligned with at least 90% identity, along with their transcript sequence coverage, were defined as the origins of the transcripts’ corresponding genomic regions.

Genome-wide screening of human SAT pairs
In accordance with their genomic positions, all transcripts that were successfully aligned previously to the same genomic loci were compiled into sets of individual transcriptional clusters on the basis of sense and antisense strands. Using this information, we screened out those that overlapped with another transcription cluster positioned on the opposite side of the genome strand. We classified each pair of overlapping transcription clusters observed into exon-overlapping sense–antisense pairs and non-exon-overlapping bidirectional pairs, in which the two transcripts did not overlap but were derived from the same genomic locus. From each group of bidirectional pairs extracted, we selected pairs of representative transcripts with the longest transcription length. In total, 2519 human gene pairs were extracted as bidirectional gene pairs; they were further classified into 1486 exon-overlapping pairs and 1033 non-exon-overlapping pairs.

Custom oligo-array data in mice and humans
The mouse DNA microarray experiment protocols used in our previous research were followed (30). The RNA samples used for the mouse oligo-array experiments came from the brain, NIH3T3 cells (fibroblast cell line), liver, heart and testis. RNA from mouse tissue (C57BL/6J, 8 to 10 weeks, male and female mixed) and mouse fibroblast line NIH3T3 was isolated by using Trizol reagent (Invitrogen). The mouse custom oligo DNA microarray chip (microarray format: 11K) contained 3896 probes that represented genes in 1948 exon-overlapping SAT pairs and had been loaded on our previous 8.5 K microarray platform (30); 1114 probes for genes in 557 non-exon-overlapping bidirectional gene pairs; 4223 probes that corresponded to non-SAT genes; and other probes that were unrelated to the SAT analysis. The total mean signal on the mouse chip in each hybridization experiment was adjusted with SL10 cells (fibroblast cell line), which were not used for any other purpose in this study.

The sequences of 60-mer DNA specific to the human sense and antisense genes were chosen by Agilent Technologies, and the Agilent custom oligo DNA microarray chips were manufactured by using this information. RNA was labeled and hybridized by using Fluorescent Direct Label Kits (oligo dT primed labeling) (Agilent Technologies), in accordance with the manufacturer’s protocols. For labeling with random nanomers, we used the CyScribe First-Strand cDNA Labeling Kit (Amersham). The RNA samples used for microarray experiments came from human brain, heart, liver, testis and HF19 cells (fibroblast cell line). The total brain, heart and testis RNAs used in the array experiments were purchased from Ambion. Total RNAs were isolated from the fibroblast cells by using Trizol reagent (Invitrogen). The same total RNA samples were labeled with single color Cy3, hybridized to the oligo DNA on the chip and dye-normalized; the processed signals were obtained by using Feature Extraction software (Agilent Technologies). The custom oligo DNA microarray chip (microarray format: 11 K) contained 2793 probes for 1486 pairs of exon-overlapping SATs; 1816 probes for genes in 1033 non-exon-overlapping bidirectional gene pairs; 2271 probes that represented non-SAT genes; and other probes unrelated to the SAT analysis. The data on all genes on the chip were used to enable the Feature Extraction software to produce the processed signals. For further analysis, the Cy3-labeled processed signal was used as the processed signal from the expression of a particular gene. The total mean signal on the human chip in each hybridization experiment was adjusted to that from the brain sample (oligo dT priming), so that the relative differences in gene expression could be compared among cell lines and tissues. The raw data from the array experiments were deposited in the Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (NCBI) under the accession number GSE9581 [NCBI GEO] (http://www.ncbi.nlm.nih.gov/geo/). The genomic structure of each SATs and the expression data can be viewed from antisense viewer which is a java applet downloadable from the following url (http://www.brc.riken.jp/archives/Kiyosawa/HMG08/).

Identification of cSATs
We used the following criteria to identify the cSAT pairs. First, we applied the mapping of 1948 pairs of mouse SATs to the human genome (hg17) by using BLAT. The alignment of those SAT genes with less than 80% sequence identity, along with those that were not marked as having the best alignment score, was discarded. Next, both SAT genes that were aligned to a certain human SAT locus were selected for further screening. In addition, since 21% (832/3896) of the mouse SAT dataset consisted of non-protein-coding transcripts, which were less well conserved than the coding sequences (4), we set up the rule that even if only one of the mouse SAT genes were successfully aligned to a human SAT locus, the other gene in the pair would also be eligible for further screening. Finally, the remaining SATs that passed the screening were further classified by their exon-overlapping pattern conservation, and those with conserved exon-overlap structures were defined as conserved SATs. The total numbers of cSATs and the procedures used are illustrated in Supplementary Material, Fig. S1. Lists of cSATs are provided in Supplementary Material, Table S1.

Microarray data analysis of cSATs
Expression signals for each cSAT pair obtained from mouse and human microarray data from five tissues or cells (brain, heart, liver, testis and fibroblast) were further categorized on the basis of the similarity of sense and antisense gene expression ratios among the two species (Fig. 1). Since the mouse and human fibroblasts originated from different tissues (mouse: NIH3T3, human: HF19), we used the expression data from the four remaining tissues to determine the expression ratio similarity. The expression ratios of cSATs were clustered by using CLUSTER (40), and the results were further visualized by using TREEVIEW (41). Expression ratio similarity was judged from the coloring of the clustering results. For example, if the mouse and human cSATs both displayed a red coloring in a certain tissue, we determined the pairs to be similarly expressed candidates.

Northern hybridization
RNA from mouse tissue (C57BL/6J, 8 to 10 weeks, male and female mixed), the mouse fibroblast line NIH3T3 and the human fibroblast line HF19 was isolated by using Trizol reagent (Invitrogen). Human total and poly(A)+ RNA was purchased from Ambion. Northern analysis was performed as previously described (30). Loading of equal amounts of RNA samples was confirmed by visualization of ethidium bromide-stained RNA in the gel. For hybridization (32), P-CTP-labeled strand-specific cRNA was prepared from the full-length mouse cDNA clones (purchased from K.K. DNAFORM, Japan). In the case of the human probes, cDNA fragments were cloned to pGEM-T Easy Vector (Promega), and strand-specific cRNA was prepared for hybridization (Supplementary Material, Table S4). The sequences of the synthetic oligo 30–40 mer DNA probes used for the sense gene AK050033 [GenBank] and antisense gene AK080971 [GenBank] (Fig. 4) are shown in Supplementary Material, Table S5.

In situ hybridization
Probes that corresponded to the mouse cSAT-177 genes were amplified by PCR using the following primers: cSAT-177 sense gene (AK049951 [GenBank] ), 5'-CCACCCCTTCTCCACTACTCT-3' (nt –593 to –847) and 5'-GGAACACCCAGCGTCACTTTG-3' (reverse complement to nt –847 to –593); cSAT-177 antisense gene (NM_008085 [GenBank] ), 5'-TGGAGAAGGGCATTAGGGTGG-3' (nt –374 to –641) and 5'-GAAATATGTGCCGAAGCTGCC-3' (reverse complement to nt –641 to –374). Probes (cRNA) labeled with digoxigenin (DIG) were hybridized to the testis tissue sections, and this procedure was followed by treatment with anti-DIG alkaline phosphatase-conjugated antibodies and visualization by BCIP/NBT.

Fractionation of testicular cells
Fractionation of germ cells on the basis of the three stages of spermatogenesis in the mouse (C57BL/6J) testis was performed by previous methods (42,43).


   SUPPLEMENTARY MATERIAL
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
FUNDING
 CONTRIBUTIONS
 REFERENCES
 
Supplementary Material is available at HMG Online


   FUNDING
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
CONTRIBUTIONS
 REFERENCES
 
This research was supported in part by grants from the Non-coding RNA Project of the New Energy and Industrial Technology Development Organization (NEDO) of Japan and by a Grant-in-Aid for Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.


   CONTRIBUTIONS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 CONTRIBUTIONS
REFERENCES
 
Y.O. analyzed the data and wrote the paper with editing from K.N., R.S., A.K., K.A., M.T. and H.K. H.K. designed the experiments and C.T., K.W., K.O., R.I., and H.K. performed the experiments. H.N. developed the antisense viewer under the supervision of N.Y. All authors discussed the results and commented on the manuscript.


   ACKNOWLEDGEMENTS
 
We thank Dr Yuko Osada and Mr Shinya Murata (Keio University, Japan) for their stimulating discussions and support. We are also grateful to Prof. Shinichi Kashiwabara (University of Tsukuba, Japan) for fractionation of the testis germ cells, and to members of GENOSTAFF Inc. for the in situ hybridization data. We also thank Kouichi Tatsuguchi and Yukiaki Kikuta for organizing the antisense RNA viewer project at C’s Lab Co. Ltd, and the staffs at Hokkaido System Science Co. Ltd, for organizing the hybridization process. Dicer-deficient murine ES cells were kindly provided from Dr Elizabeth P. Murchison and Dr Gregory J. Hannon of Cold Spring Harbor Laboratory. We acknowledge Mr Naoto Kaneko for technical support of the Dicer-null northern hybridization analysis.

Conflict of Interest statement. Authors declare no conflicts of interest.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 CONTRIBUTIONS
 REFERENCES
 

  1. Birney E., Stamatoyannopoulos J.A., Dutta A., Guigo R., Gingeras T.R., Margulies E.H., Weng Z., Snyder M., Dermitzakis E.T., Thurman R.E., et al. Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project. Nature (2007) 447:799–816.[CrossRef][Medline]

  2. Lavorgna G., Dahary D., Lehner B., Sorek R., Sanderson C.M., Casari G. In search of antisense. Trends Biochem. Sci. (2004) 29:88–94.[CrossRef][Web of Science][Medline]

  3. Chen J., Sun M., Kent W.J., Huang X., Xie H., Wang W., Zhou G., Shi R.Z., Rowley J.D. Over 20% of human transcripts might form sense-antisense pairs. Nucleic Acids Res. (2004) 32:4812–4820.[Abstract/Free Full Text]

  4. Kiyosawa H., Yamanaka I., Osato N., Kondo S., Hayashizaki Y., RIKEN GER Group GSL Members. Antisense transcripts with FANTOM2 clone set and their implications for gene regulation. Genome Res. (2003) 13:1324–1334.[Abstract/Free Full Text]

  5. Engström P.G., Suzuki H., Ninomiya N., Akalin A., Sessa L., Lavorgna G., Brozzi A., Luzi L., Tan S.L., Yang L., et al. Complex Loci in human and mouse genomes. PLoS Genet. (2006) 2:e47.[CrossRef][Medline]

  6. Yelin R., Dahary D., Sorek R., Levanon E.Y., Goldstein O., Shoshan A., Diber A., Biton S., Tamir Y., Khosravi R., et al. Widespread occurrence of antisense transcription in the human genome. Nat. Biotechnol. (2003) 21:379–386.[CrossRef][Web of Science][Medline]

  7. Katayama S., Tomaru Y., Kasukawa T., Waki K., Nakanishi M., Nakamura M., Nishida H., Yap C.C., Suzuki M., Kawai J., et al. Antisense transcription in the mammalian transcriptome. Science (2005) 309:1564–1566.[Abstract/Free Full Text]

  8. Sun M., Hurst L.D., Carmichael G.G., Chen J. Evidence for variation in abundance of antisense transcripts between multicellular animals but no relationship between antisense transcription and organismic complexity. Genome Res. (2006) 16:922–933.[Abstract/Free Full Text]

  9. Misra S., Crosby M.A., Mungall C.J., Matthews B.B., Campbell K.S., Hradecky P., Huang Y., Kaminker J.S., Millburn G.H., Prochnik S.E., et al. Annotation of the Drosophila melanogaster euchromatic genome: a systematic review. Genome Biol. (2002) 3. RESEARCH0083.

  10. Osato N., Yamada H., Satoh K., Ooka H., Yamamoto M., Suzuki K., Kawai J., Carninci P., Ohtomo Y., Murakami K., et al. Antisense transcripts with rice full-length cDNAs. Genome Biol. (2003) 5:R5.[CrossRef][Medline]

  11. Wang X.J., Gaasterland T., Chua N.H. Genome-wide prediction and identification of cis-natural antisense transcripts in Arabidopsis thaliana. Genome Biol. (2005) 6:R30.[CrossRef][Medline]

  12. Yamada K., Lim J., Dale J.M., Chen H., Shinn P., Palm C.J., Southwick A.M., Wu H.C., Kim C., Nguyen M., et al. Empirical analysis of transcriptional activity in the Arabidopsis genome. Science (2003) 302:842–846.[Abstract/Free Full Text]

  13. David L., Huber W., Granovskaia M., Toedling J., Palm C.J., Bofkin L., Jones T., Davis R.W., Steinmetz L.M. A high-resolution map of transcription in the yeast genome. Proc. Natl Acad. Sci. USA (2006) 103:5320–5325.[Abstract/Free Full Text]

  14. Lapidot M., Pilpel Y. Genome-wide natural antisense transcription: coupling its regulation to its different regulatory mechanisms. EMBO Rep. (2006) 7:1216–1222.[CrossRef][Web of Science][Medline]

  15. Hastings M.L., Ingle H.A., Lazar M.A., Munroe S.H. Post-transcriptional regulation of thyroid hormone receptor expression by cis-acting sequences and a naturally occurring antisense RNA. J. Biol. Chem. (2000) 275:11507–11513.[Abstract/Free Full Text]

  16. Kawahara Y., Nishikura K. Extensive adenosine-to-inosine editing detected in Alu repeats of antisense RNAs reveals scarcity of sense-antisense duplex formation. FEBS Lett. (2006) 580:2301–2305.[CrossRef][Web of Science][Medline]

  17. Neeman Y., Dahary D., Levanon E.Y., Sorek R., Eisenberg E. Is there any sense in antisense editing? Trends Genet. (2005) 21:544–547.[CrossRef][Web of Science][Medline]

  18. Peters N.T., Rohrbach J.A., Zalewski B.A., Byrkett C.M., Vaughn J.C. RNA editing and regulation of Drosophila 4f-rnp expression by sas-10 antisense readthrough mRNA transcripts. RNA (2003) 9:698–710.[Abstract/Free Full Text]

  19. Brantl S. Antisense-RNA regulation and RNA interference. Biochim. Biophys. Acta (2002) 1575:15–25.[Medline]

  20. Sijen T., Plasterk R.H. Transposon silencing in the Caenorhabditis elegans germ line by natural RNAi. Nature (2003) 426:310–314.[CrossRef][Medline]

  21. Verdel A., Jia S., Gerber S., Sugiyama T., Gygi S., Grewal S.I., Moazed D. RNAi-mediated targeting of heterochromatin by the RITS complex. Science (2004) 303:672–676.[Abstract/Free Full Text]

  22. Borsani O., Zhu J., Verslues P.E., Sunkar R., Zhu J.K. Endogenous siRNAs derived from a pair of natural cis-antisense transcripts regulate salt tolerance in Arabidopsis. Cell (2005) 123:1279–1291.[CrossRef][Web of Science][Medline]

  23. Kiyosawa H., Abe K. Speculations on the role of natural antisense transcripts in mammalian X chromosome evolution. Cytogenet. Genome Res. (2002) 99:151–156.[CrossRef][Web of Science][Medline]

  24. Lee J.T., Davidow L.S., Warshawsky D. Tsix, a gene antisense to Xist at the X-inactivation centre. Nat. Genet. (1999) 21:400–404.[CrossRef][Web of Science][Medline]

  25. Imamura T., Yamamoto S., Ohgane J., Hattori N., Tanaka S., Shiota K. Non-coding RNA directed DNA demethylation of Sphk1 CpG island. Biochem. Biophys. Res. Commun. (2004) 322:593–600.[CrossRef][Web of Science][Medline]

  26. Tufarelli C., Stanley J.A., Garrick D., Sharpe J.A., Ayyub H., Wood W.G., Higgs D.R. Transcription of antisense RNA leading to gene silencing and methylation as a novel cause of human genetic disease. Nat. Genet. (2003) 34:157–165.[CrossRef][Web of Science][Medline]

  27. Boffelli D., Nobrega M.A., Rubin E.M. Comparative genomics at the vertebrate extremes. Nat. Rev. Genet. (2004) 5:456–465.[CrossRef][Web of Science][Medline]

  28. Numata K., Okada Y., Saito R., Kiyosawa H., Kanai A., Tomita M. Comparative analysis of cis-encoded antisense RNAs in eukaryotes. Gene (2007) 392:134–141.[CrossRef][Web of Science][Medline]

  29. Veeramachaneni V., Makalowski W., Galdzicki M., Sood R., Makalowska I. Mammalian overlapping genes: the comparative perspective. Genome Res. (2004) 14:280–286.[Abstract/Free Full Text]

  30. Kiyosawa H., Mise N., Iwase S., Hayashizaki Y., Abe K. Disclosing hidden transcripts: mouse natural sense-antisense transcripts tend to be poly(A) negative and nuclear localized. Genome Res. (2005) 15:463–474.[Abstract/Free Full Text]

  31. Karolchik D., Baertsch R., Diekhans M., Furey T.S., Hinrichs A., Lu Y.T., Roskin K.M., Schwartz M., Sugnet C.W., Thomas D.J., et al. The UCSC genome browser database. Nucleic Acids Res. (2003) 31:51–54.[Abstract/Free Full Text]

  32. Furuno M., Pang K.C., Ninomiya N., Fukuda S., Frith M.C., Bult C., Kai C., Kawai J., Carninci P., Hayashizaki Y., et al. Clusters of internally primed transcripts reveal novel long noncoding RNAs. PLoS Genet. (2006) 2:e37.[CrossRef][Medline]

  33. Williamson C.M., Turner M.D., Ball S.T., Nottingham W.T., Glenister P., Fray M., Tymowska-Lalanne Z., Plagge A., Powles-Glover N., Kelsey G., et al. Identification of an imprinting control region affecting the expression of all transcripts in the Gnas cluster. Nat. Genet. (2006) 38:350–355.[CrossRef][Web of Science][Medline]

  34. Babak T., Blencowe B.J., Hughes T.R. A systematic search for new mammalian noncoding RNAs indicates little conserved intergenic transcription. BMC Genomics (2005) 6:104.[CrossRef][Medline]

  35. Kapranov P., Cheng J., Dike S., Nix D.A., Duttagupta R., Willingham A.T., Stadler P.F., Hertel J., Hackermuller J., Hofacker I.L., et al. RNA maps reveal new RNA classes and a possible function for pervasive transcription. Science (2007) 316:1484–1488.[Abstract/Free Full Text]

  36. Murchison E.P., Partridge J.F., Tam O.H., Cheloufi S., Hannon G.J. Characterization of Dicer-deficient murine embryonic stem cells. Proc. Natl Acad. Sci. USA (2005) 102:12135–12140.[Abstract/Free Full Text]

  37. Waterston R.H., Lindblad-Toh K., Birney E., Rogers J., Abril J.F., Agarwal P., Agarwala R., Ainscough R., Alexandersson M., An P., et al. Initial sequencing and comparative analysis of the mouse genome. Nature (2002) 420:520–562.[CrossRef][Medline]

  38. Kent W.J. BLAT—the BLAST-like alignment tool. Genome Res. (2002) 12:656–664.[Abstract/Free Full Text]

  39. Florea L., Hartzell G., Zhang Z., Rubin G.M., Miller W. A computer program for aligning a cDNA sequence with a genomic DNA sequence. Genome Res. (1998) 8:967–974.[Abstract/Free Full Text]

  40. Eisen M.B., Spellman P.T., Brown P.O., Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc. Natl Acad. Sci. USA (1998) 95:14863–14868.[Abstract/Free Full Text]

  41. Saldanha A.J. Java Treeview—extensible visualization of microarray data. Bioinformatics (2004) 20:3246–3248.[Abstract/Free Full Text]

  42. Bellve A.R., Cavicchia J.C., Millette C.F., O’Brien D.A., Bhatnagar Y.M., Dym M. Spermatogenic cells of the prepuberal mouse. Isolation and morphological characterization. J. Cell Biol. (1977) 74:68–85.[Abstract/Free Full Text]

  43. Romrell L.J., Bellve A.R., Fawcett D.W. Separation of mouse spermatogenic cells by sedimentation velocity. A morphological characterization. Dev. Biol. (1976) 49:119–131.[CrossRef][Web of Science][Medline]

  44. Okazaki Y., Furuno M., Kasukawa T., Adachi J., Bono H., Kondo S., Nikaido I., Osato N., Saito R., Suzuki H., et al. Analysis of the mouse transcriptome based on functional annotation of 60,770 full-length cDNAs. Nature (2002) 420:563–573.[CrossRef][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Nucleic Acids ResHome page
M. Carlile, D. Swan, K. Jackson, K. Preston-Fayers, B. Ballester, P. Flicek, and A. Werner
Strand selective generation of endo-siRNAs from the Na/phosphate transporter gene Slc34a1 in murine tissues
Nucleic Acids Res., April 1, 2009; 37(7): 2274 - 2282.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
S. Shpiz, D. Kwon, Y. Rozovsky, and A. Kalmykova
rasiRNA pathway controls antisense expression of Drosophila telomeric retrotransposons in the nucleus
Nucleic Acids Res., January 1, 2009; 37(1): 268 - 278.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
17/11/1631    most recent
ddn051v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (5)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Okada, Y.
Right arrow Articles by Kiyosawa, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Okada, Y.
Right arrow Articles by Kiyosawa, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?