Epigenetic features are significantly associated with alternative splicing
© Zhou et al; licensee BioMed Central Ltd. 2012
Received: 15 October 2011
Accepted: 29 March 2012
Published: 29 March 2012
While alternative splicing (AS) contributes greatly to protein diversities, the relationship between various types of AS and epigenetic factors remains largely unknown.
In this study, we discover that a number of epigenetic features, including DNA methylation, nucleosome occupancy, specific histone modifications and protein features, are strongly associated with AS. To further enhance our understanding of the association between these features and AS, we cluster our investigated features based on their association patterns with each AS type into four groups, with H3K36me3, EGR1, GABP, SRF, SIN3A and RNA Pol II grouped together and showing strongest association with AS. In addition, we find that the AS types can be classified into two general classes, namely the exon skipping related process (ESRP), and the alternative splice site selection process (ASSP), based on their association levels with the epigenetic features.
Our analysis thus suggests that epigenetic features are likely to play important roles in regulating AS.
KeywordsAlternative splicing Epigenetics DNA methylation Nucleosome occupancy Histone modifications Transcription factors
Alterations in splice site choice, referred to alternative splicing (AS), can result in mRNA is forms coding for proteins with different chemical and biological activity [1–3]. AS events can be generally classified into the following categories: cassette exon (CE), exon skipping (ES), mutually exclusive exon (ME), alternative 5' splice site selection (A5SS), alternative 3' splice site selection (A3SS) and intron retention (IR) [1, 2]. Early estimates suggest that at least 60% of human genes may undergo AS . Now, with the application of next generation sequencing (NGS) technologies, it has been revealed that almost all genes (95%) in the human genome will undergo AS , highlighting its important roles in the complexity of eukaryotic gene expression. However, although much research work has been dedicated to the understanding of the AS phenomenon, the regulatory mechanism of AS is still not fully understood.
It has been widely accepted that chromatin state plays essential roles in regulating gene expression. DNA methylation, nucleosome occupancy and modifications of histone are involved in determining the chromatin state [6, 7], while some transcription factors (TFs) can bind to specific regulatory regions to interact with chromatin and regulate gene expression . All these factors can be considered as epigenetic features that regulate gene expression from a broad perspective . Though epigenetic signatures are mainly found to be enriched in promoters, it has become increasingly clear that they are also present in exon regions, indicating a potential link of epigenetic regulation to splicing [10–12]. For example, DNA methylation level shows distinctive differential patterns between intron and exon . Histone modifications such as H3K36me3, H3K79me1, H2BK5me1, H3K27me1, H3K27me2, and H3K27me3 were found to be related with exon expression .
Since epigenetic features have been implied to have a connection with splicing, naturally we wonder whether they may also be involved in AS. Indeed, nucleosome occupancy level was found to be lower in cassette exons than in constitutively spliced exons [14–17]. H3K36me3 and H3K9ac were found to be related to the exon skipping event of NCAM . The level of H3K36me3 was found to be different in mutually exclusive exons of FGFR2, TPM1, TPM2, and PKM2 between PNT2 and hMSC cell types. H3K36me3 was also found to be depleted in skipped exon in a genome-wide study across different species [17, 19], though this finding was still in an argument . Other histone methylations, such as H3K4me1, H3K4me3, H3K27me3 and H3K9me1, were found to be associated with the AS events of FGFR2 , while H3K4me3 was suggested to affect the AS events of CHD1 , and H3K9me3 was found to be associated with the multiple exon skipping of CD44 . A recent ChIP-chip assay also found that the levels of 13 histone modifications are different in between CE and constitutively spliced exon (CNE) . Despite the accumulating evidence suggesting a potential connection between epigenetics and AS, to date there have been no systematic studies on investigating the association of epigenetic features with different types of AS.
In this study, we have collected epigenetics data from six cell lines and carried out a comprehensive computational survey to investigate the association of different types of AS with epigenetic features, such as DNA methylation, nucleosome occupancy, histone modifications (including 10 histone methylations and 13 histone acetylations), and protein features (TFs) (including 9 TFs, CTCF and RNA Pol II). We find that a number of epigenetic features are strongly associated with AS. Based on their association pattern with AS events, these epigenetic features are grouped into four clusters, with the first cluster including H3K36me3, RNA Pol II, and a set of TFs and showing strongest association with all types of AS events. In addition, based on their association with the epigenetic features, the AS events are classified into two general classes: the exon skipping related process (ESRP) including ES and ME and the alternative splice site selection process (ASSP) including A3SS, A5SS and IR. Thus, our study highlights the potentially important roles of epigenetic regulation on AS, and provides list of epigenetic features strongly associated with AS, making it possible to design experiments to further investigate the molecular mechanisms of AS.
In this study, we explore whether epigenetic regulation is likely involved in AS by investigating the association of epigenetic features with AS. The epigenetic features include DNA methylation, nucleosome occupancy, histone modifications and some protein features. We focus on comparing the association level of each feature around the splice sites of between the alternatively spliced exon (ASE) and the constitutively spliced exon (CNE). Here, the types of ASE being investigated include only the alternative spliced internal exons, such as ME, ES, A3SS, A5SS and IR. ES only refers to the single exon skipping event. Details on the definition of internal exons and how to determine these ASE types can be found in Methods and Additional file 1. When evaluating the significance of the association, we divide the regions surrounding the acceptor and donor splice sites into the intronic and exonic regions each with a length of 100 bp (base pair), respectively, and compute a P-value in each region to indicate the significance of difference in between a type of ASE and CNE for a given feature. The difference is considered statistically significant if the P-value is smaller than 1e-2. In the following sections, unless otherwise specified, if a feature is said to be higher or lower in a given type of ASE, then it always refers to the results of comparison with the level of the feature in CNE; if no specific region, e.g., the exonic or intronic region, is mentioned, then it implies that the association is true for all the regions; in addition, if there are available data for a given feature in two or more cell lines, the results are always averaged across all available cell lines.
The association of DNA methylation and nucleosome occupancy with AS
Here, we investigate whether the level of genomic CpG dinucleotides (termed CG) and the percentage of methylated CpG dinucleotides (termed mCG) are different in between ASE and CNE. The mCG data is obtained from a study of DNA bisulfite sequencing (BS-seq) of H1 hESC and IMR90 cells . The nucleosome occupancy dataset is obtained from a study using high-throughput sequencing after MNase digestion in resting CD4+ cells .
For nucleosome occupancy, we find that its level is significantly lower in the exonic regions of both ME and ES, with the results of ME more significant than that of ES (Figure 1c). This finding is consistent with several early reports that ES has lower nucleosome occupancy [14–16]. The nucleosome occupancy level is also significantly lower in the intronic region of the donor splice site of IR. Both A5SS and A3SS do not have strong association with nucleosome occupancy. The overall association patterns of nucleosome occupancy with ASE are similar to that of mCG's, indicating a potential connection between them in AS.
The association of histone modifications with AS
We obtain histone modification data from several recently published ChIP-seq studies [10, 25, 26]. We select 10 histone methylations and 13 histone acetylations that have available data in at least two cell lines (see Additional file 4 for details about the histone modifications and the corresponding cell lines). The distribution of each histone modification is drawn by averaging the ChIP signal across all cell lines.
Histone acetylations are not as significantly associated as histone methylations with AS (Additional file 2). We only find that the level of H3K9ac and H3K27ac (Figure 2k-l) are higher in A3SS and A5SS, while the level of H3K14ac is slightly higher in the intronic region of ME (P = 0.0438, Additional file 5).
The association of protein features with AS
Besides the above-mentioned well known epigenetic features, other features, such as TFs, CTCF and RNA Pol II are also thought to be related with the modification of chromatin structures [27–30]. Despite that there have been no report on the association of these TFs with the splicing machinery, given their relationships with chromatin structure, it is reasonable to hypothesize that some of them may be involved in splicing either directly or indirectly, and therefore may likely have an impact with AS. For example, a recent study reported that the rate of RNA Pol II elongation might have an impact on AS , while CTCF was found in a recent study to be related to the alternative splicing of CD45 gene . However, to date the relationships of these TFs with AS have not been systematically investigated. Here, we analyze the association pattern of 9 TFs, CTCF and RNA Pol II that have ChIP-seq data [10, 25, 26] of more than 2 cells with AS.
Clustering of epigenetic features by its pattern in different AS types and clustering of AS types by its epigenetic pattern
Correcting the effect of ChIP-seq input and nucleosome occupancy
It has been reported that most histone modifications are related to the positioning of nucleosomes [14, 15]. In this study, nucleosome occupancy is found to have similar association patterns with AS to that of most modifications (Figure 1c), though its significance level is considerably lower. To investigate how significant nucleosome occupancy may contribute to the association of other features with AS, we perform the analysis by correcting for nucleosome occupancy. However, since nucleosome data is available only for CD4+ cell lines, among the significantly associated epigenetic features we only do the correction for H3K36me3 and RNA Pol II which have available data in that cell line. The results show that the association patterns of these two features remain the same as before, and the statistic tests are still significant, though the significance level of ES, ME and A5SS is lower than before (Figure 6b-c). Additionally, for all other epigenetic features with available data in CD4+ cell lines, the results show that after the correction for nucleosome occupancy, their association patterns are still similar to before (Additional file 12). Thus, we conclude that nucleosome alone is not enough to explain the association of other epigenetic features with AS.
Prior to our work, there have been a number of studies on how epigenetic features may affect splicing [20, 31, 33]. It has been proposed that specific histone modifications may help to recruit related TFs or splicing factors or to change the RNA Pol II elongation efficiency, which can affect splicing. As RNA Pol II elongation efficiency may play a role in affecting splicing, we reason that higher level of RNA Pol II around the splice sites may correspond to lower RNA Pol II elongation rate, which can help promote splicing efficiency. Based on our findings that the level of RNA Pol II is significantly higher in ASSP while significantly lower in ESRP, here we propose that lower RNA Pol II elongation rate may be necessary for selecting the alternative splice site, while higher RNA Pol II elongation rate can facilitate the exon skipping events. Specific histone methylations, such as H3K36me3, may help to recruit general TFs like EGR1, GABP, SRF, SIN3A, REST, TAF1 and USF1, which can help to retain RNA Pol II around the splice site and promote ASSP; while other histone methylations, such as H4K20me1, H3K4me1, H3K4me2, H3K9me1 and H3K79me2, may provide additional help for this process. In the absence of H3K36me3, the above-mentioned general TFs may not be recruited, making it easy for RNA Pol II to pass through the splice site, and resulting in ESRP events. Our model explains the general difference between the ASSP and ESRP events.
In this study, the transcriptional effects are not taken into consideration when analyzing the association of epigenetic features with different types of ASEs. This is because of the following reasons. Firstly, not all cell types investigated in this study have the available exon-array or RNA-seq data that were obtained under the same condition as that of the epigenetic features, making it difficult to reduce the transcriptional effects across different cell types. Secondly, even in cell line in which the exon-array or RNA-seq data were available, we find that there are usually significant variations of expression level in between the individual exons of a given gene, making it impossible to reduce the transcriptional effects on different types of exons. Thirdly, transcription level is a result of both epigenetic regulation and splicing, even with available data it may be difficult to distinguish the effects of epigenetic regulation from the effects of transcription level. Nevertheless, by focusing on comparing the distribution of epigenetic features in between a specific type of AS exons and the CNEs that are determined by comparing all Ensemble transcripts between each other, we are able to identify the statistically significant differences for some epigenetic features that are consistent across different cell lines. This highlights the likely difference of epigenetic regulation on potentially alternatively spliced exons from that on constitutively spliced exons.
Exon length may be another factor that affects the results of our analysis, because we do not separate exons by using different cutoff. However, the short exons (e.g., with a length shorter than 50 nt) and the long exons (e.g., with a length longer than 300 nt) only account for a small portion of all exons (Additional file 13), and therefore are unlikely to affect the results. In our analysis, we define the regions surrounding the splice site to be ± 100 nt to the splice site. When the regions are defined to be ± 50 nt to the splice site, the results remain almost the same (Additional file 2). In addition, Kolasinska-Zwierz et al. studied the alternative splicing of exons longer than 200 nt, which excluded most exons (the typical length of an exon is 147 nt), and found that the level of H3K36me3 is lower in ES compared to CNE , which is consistent with our findings, further suggesting that exon length is unlikely to affect our conclusions. On the other hand, it has been shown that most epigenetic features have completely different patterns in both TSS and PolyA than in gene body , while the levels of some histone modifications are increased from TSS to PolyA while the levels of some are decreased, hinting that the positional effects on the level of epigenetic features may cause some bias to our conclusions. In our analysis, we exclude AS types such as alternative first exon and alternative last exon and only consider exons that are not within 2000 bp to TSS and PolyA. To further clarify the possible positional effects, we group ASEs and CNEs based on their positions (2nd-4th exons), and repeat the analysis for the most significant epigenetic features (cluster 1 and cluster 2) on two cell lines (GM12878 and K562). The results suggest that their patterns at different positions exons remain the same as before (Additional file 14), indicating that the positional effects of the distribution of epigenetic features are unlikely to affect our conclusions. However, our analysis only focuses on the protein coding transcripts that are alternatively spliced, and the conclusion may not be directly applicable for AS of non-coding transcripts which were recently found to have different epigenetic signatures from protein coding transcripts [14, 16]. On the other hand, we focus only on the regions surrounding the splice sites, and do not consider the epigenetic features in the intronic regions far from the splice sites, which may also contribute to AS.
In this study, we have uncovered a number of epigenetic features strongly associated with different types of AS. Some of the findings are consistent with other studies. For example, our finding on H3K36me3's association with ES is consistent with Kolasinska-Zwierz et al.'s , though their study involved less number of exons. A recent study performed by Dhami et al. found that 13 histone methylations are associated with the splicing of cassette exon  in which nine of them are investigated in this study. Since Dhami et al. did not report the p-value of the associations, we compare the tendency of the association of these epigenetic features in their study with ours. Except for H3K4me1, the tendency of all the other features is consistent with ours. However, because of the adjustment of multiple tests, four of them, H3K4me3, H3K27me3, H3K36me3 and H3K79me1, are significant in our study. The possible difference can be attributed to the following reasons: Dhami et al. used the ChIP-chip data, while we use the ChIP-seq data; they analyzed the AS events of only 268 well expressed genes, while we carry out a genome-wide study; they analyzed three cell lines, while we analyze six cell lines. It is worth noting that these recent studies are limited because their studies either focused on data from single cell line , or used ChIP-chip data which are considered to have lower quality compared to ChIP-seq data . In contrast, our study uses the ChIP-seq data from six cell lines to investigate the associations of not only histone methylations but also other epigenetic features, such as TFs, with AS, and discovers a number of epigenetic features significantly associated with AS, in which many of them are first reported, including EGR1, GABP, SRF, SIN3A, etc. These discovered significant epigenetic features will be useful to guide experimentalists to design specific experiments to further explore the mechanisms of epigenetic regulation on AS.
NCAM, Neural cell adhesion molecule; FGFR2, Fibroblast growth factor receptor 2; TPM1, Tropomyosin 1 (alpha); TPM2, Tropomyosin 2 (beta); PKM2, Pyruvate kinase muscle; CHD1, Chromodomain helicase DNA binding protein 1; CD44, CD44 molecule (Indian blood group)
CTCF, CCCTC-binding factor (zinc finger protein); EGR1, Early growth response 1; GABP, GA binding protein transcription factor; SIN3A, SIN3 homolog A transcription regulator; SRF, Serum response factor (c-fos serum response element-binding transcription factor); REST, RE1-silencing transcription factor (NRSF); TAF1, TAF1 RNA polymerase II TATA box binding protein (TBP)-associated factor; USF1, Upstream transcription factor 1; EP300, E1A binding protein p300 (P300); SPI1, Spleen focus forming virus (SFFV) proviral integration oncogene spi1 (PU.1)
Our results reveal that epigenetic features are strongly associated with AS, suggesting that epigenetic regulation may be involved in AS. By applying a clustering procedure, we identify four tight clusters of epigenetic features, with features in the first cluster showing strongest association with AS. The AS events are grouped to two classes: the exon skipping related process (ESRP) (including ME and ES) and the alternative splice site selection process (ASSP) (including A3SS, A5SS and IR) on the basis of their association patterns with epigenetic features, indicating that these two processes may involve different mechanisms of epigenetic regulation.
Annotation of alternative splicing (AS) events
All transcripts used to annotate genome-wide AS types are downloaded from Ensembl (build 59) annotations and only coding transcripts are selected . We identify AS events, such as ME, ES, A3SS, A5SS, IR only from internal exons that are not located within the region of 0-2000 bp to TSS or -2000-0 bp to the terminal site of a transcript. Because CE is complex, and include any AS events in which an exon is either skipped or included , e.g., both ME and ES are subtypes of CE, it is not analyzed in this study. For comparison, we identify CNE from the internal exons as well. The definitions of ASE follow a previous study . The procedures on identifying AS events and CNE involve the following steps (Additional file 1a). First, we map known transcripts to the reference genome sequences (hg19). Then, the exon-intron junction sites of the transcripts belonging to the same coding gene region are compared with each other. Next, the splicing events are first coded into binary codes and then converted to decimal code for convenience, see the Nagasaki's method  for details. Here, we employ a slightly modified version of the Nagasaki's method (see Additional file 1a), such that less information will get lost and complex situations can be avoided. Finally, we recognize the AS event according to the decimal code as shown in Additional file 1b. The numbers of CNE, ES, ME, A3SS, A5SS and IR we identify are 103,806, 2725, 1206, 2521, 1884 and 563, respectively. Then, we focus on the splice site of the above events for further analysis. Here, for A3SS and A5SS, we always choose the splice site corresponding to a longer exon in the analysis; for IR, the 3' acceptor corresponds to the 3' of the retained intron, while the 5' donor corresponds to the 5' of the retained intron; for ME, both exons are considered as independent counts, because it is impossible to know which exon is included and which one is not. The recognition script of AS is available online (Additional file 15).
Genomic and epigenomic data preparation
CpG dinucleotides are directly counted from the hg19 genome downloaded from UCSC genome center . The DNA methylation data of human H1 embryonic stem cell and IMR90 fetal lung fibroblasts cell are downloaded from SALK institute . As previous studies have suggested that the sense and antisense CpG methylation are coupled [11, 37], we consider cytosine methylation from both strands of CpG dinucleotides as one functional site. The nucleosome occupancy data of resting CD4+ cell are get from one publication . The ChIP-seq data of histone modifications and protein features from six cell lines are downloaded from various sources: the data of resting CD4+ cell are from several publications [10, 38]; the data of IMR90 cell are download from the NIH Roadmap Epigenomics ; the data of GM12878, K562, H1 hESC and Hep G2 cells are downloaded from UCSC genome center , which are generate by the Encode Project [25, 26] (see Additional file 4 for details about the available data of histone modifications and protein features from each cell). The data of histone modifications and protein features are all raw alignments of sequencing reads generated by ChIP-seq. To avoid the loss of information, all alignments are extended to 200 bp according to the direction of strand . If the ChIP-seq data are based on hg18, we use the liftOver tool downloaded from the UCSC genome browser  to convert the hg18 coordinates to hg19's. All processing scripts are written by Perl programming language, and can be obtained upon request.
We divide the regions surrounding each splice site (including both donor and acceptor splice sites) into the exonic and intronic regions with 100 bp length each. Then, we map the epigenetic features investigated in this study to those regions. To obtain the level of an epigenetic feature in a given region of a splice site, for CpG dinucleotides, we directly count the number of CpG dinucleotides in the region; for methylated CpG dinucleotides, we calculate the percentage of CpG dinucleotides that are methylated in the region; for all other epigenetic features, we calculate the total number of raw sequencing reads mapped the region. Next, to test the significance of the association of an epigenetic feature with ASE, we collect the level of the feature of all splice sites belonging to the ASE and CNE from all cell lines available, respectively, and employ a one-tailed t-test with Bonferroni adjustment to examine the difference between the mean level of the ASE and of CNE. When combining the data for a given epigenetic feature, its data from each splice site in a cell line are considered as an independent observation and are put together for all cell lines. Therefore, if the pattern is consistent among all cell lines, the p-value will be more significant. Conversely, even if an epigenetic feature may be significant in one cell line, with the above-described treatment, its p-value will be less significant if the data from more cell lines are combined. The correction of ChIP-seq input is done by subtracting the signal of each feature by the signals of ChIP-seq input, because many splice sites do not have ChIP-seq signals. The correction of nucleosome occupancy is done by dividing the signal of epigenetic features to that of nucleosome level, because we only consider the histone features with the evidence of existing nucleosomes.
To visualize the distribution of the level of the epigenetic features around the splice sites, all epigenetic features are first mapped to the -500-200 bp and the -200-500 bp region surrounding the acceptor and donor splice sites, respectively. Then, for a given type of ASE or CNE, at each nucleotide position, for CpG and methylated CpG dinucleotides, we compute the percentage of CpG and methylated CpG dinucleotides across all splice sites belonging to the ASE or CNE, followed by the application of sliding window of 147 bp, which is the typical length of an exon , to reduce noise; for all other epigenetic features, at each nucleotide position we compute the average number of raw reads overlapped with this position across all splice sites belonging to the ASE or CNE, without the use of sliding window because all raw reads have already been extended to 200 bp. When data are available for multiple cell lines for a given feature, the above signal levels are averaged across all cell lines. For the correction using both ChIP-seq input and nucleosome occupancy, the corrected plots are the result of divided raw reads to ChIP-seq input reads or nucleosome occupancy at each position. The heatmap illustrating the associations of all epigenetic features with all types of ASE is generated from a matrix of p-values obtained from the one-tailed t-test with Bonferroni adjustment. Here, the p-values are converted to -log (p-value) or log (p-value) if a feature has higher level in an ASE or in CNE, respectively. The heatmap is drawn using the "heatmap.2" function of R statistic language. K-means clustering is done by applying the "kmeans" function of R with 'center = 4'. The distance matrix used by K-means is converted from the heatmap matrix using 'dist' function of R with Euclidian distance.
Alternative spliced exon
Mutually exclusive exon
Alternative 5' splice site selection
Alternative 3' splice site selection
Exon skipping related process
Alternative splice site selection process
Transcription start site.
We thank B. Chang and Q. Wang for proofreading the manuscript. We thank K. Zhao's team, ENCODE and Roadmap Epigenomics Project for making their data publicly available. This work was supported by the National Basic Research Program of China (Grant No. 2012CB316505, Grant No. 2010CB529505); the National Natural Science Foundation of China (Grant No. 31071113, Grant No. 30971643); FDUROP, Fudan's Undergraduate Research Opportunity Program (to YZ).
- Black DL: Mechanisms of alternative pre-messenger RNA splicing. Annu Rev Biochem. 2003, 72: 291-336. 10.1146/annurev.biochem.72.121801.161720.View ArticlePubMedGoogle Scholar
- Nilsen TW, Graveley BR: Expansion of the eukaryotic proteome by alternative splicing. Nature. 2010, 463 (7280): 457-463. 10.1038/nature08909.PubMed CentralView ArticlePubMedGoogle Scholar
- Law JA, Jacobsen SE: Establishing, maintaining and modifying DNA methylation patterns in plants and animals. Nat Rev Genet. 2010, 11 (3): 204-220. 10.1038/nrg2719.PubMed CentralView ArticlePubMedGoogle Scholar
- Modrek B, Lee C: A genomic view of alternative splicing. Nat Genet. 2002, 30 (1): 13-19. 10.1038/ng0102-13.View ArticlePubMedGoogle Scholar
- Pan Q, Shai O, Lee LJ, Frey BJ, Blencowe BJ: Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing. Nat Genet. 2008, 40 (12): 1413-1415. 10.1038/ng.259.View ArticlePubMedGoogle Scholar
- Laird PW: Principles and challenges of genome-wide DNA methylation analysis. Nat Rev Genet. 2010, 11 (3): 191-203.View ArticlePubMedGoogle Scholar
- Kouzarides T: Chromatin modifications and their function. Cell. 2007, 128 (4): 693-705. 10.1016/j.cell.2007.02.005.View ArticlePubMedGoogle Scholar
- Goodrich JA, Tjian R: Unexpected roles for core promoter recognition factors in cell-type-specific transcription and gene regulation. Nat Rev Genet. 2010, 11 (8): 549-558.PubMed CentralPubMedGoogle Scholar
- Brinkman A, Stunnenberg H: Strategies for epigenome analysis. Epigenomics. 2009, 3-18.View ArticleGoogle Scholar
- Barski A, Cuddapah S, Cui K, Roh TY, Schones DE, Wang Z, Wei G, Chepelev I, Zhao K: High-resolution profiling of histone methylations in the human genome. Cell. 2007, 129 (4): 823-837. 10.1016/j.cell.2007.05.009.View ArticlePubMedGoogle Scholar
- Lister R, Pelizzola M, Dowen RH, Hawkins RD, Hon G, Tonti-Filippini J, Nery JR, Lee L, Ye Z, Ngo QM: Human DNA methylomes at base resolution show widespread epigenomic differences. Nature. 2009, 462 (7271): 315-322. 10.1038/nature08514.PubMed CentralView ArticlePubMedGoogle Scholar
- Kornblihtt AR, Schor IE, Allo M, Blencowe BJ: When chromatin meets splicing. Nat Struct Mol Biol. 2009, 16 (9): 902-903. 10.1038/nsmb0909-902.View ArticlePubMedGoogle Scholar
- Laurent L, Wong E, Li G, Huynh T, Tsirigos A, Ong CT, Low HM, Sung KWK, Rigoutsos I, Loring J: Dynamic changes in the human methylome during differentiation. Genome Res. 2010, 20 (3): 320-331. 10.1101/gr.101907.109.PubMed CentralView ArticlePubMedGoogle Scholar
- Andersson R, Enroth S, Rada-Iglesias A, Wadelius C, Komorowski J: Nucleosomes are well positioned in exons and carry characteristic histone modifications. Genome Res. 2009, 19 (10): 1732-1741. 10.1101/gr.092353.109.PubMed CentralView ArticlePubMedGoogle Scholar
- Spies N, Nielsen CB, Padgett RA, Burge CB: Biased Chromatin Signatures around Polyadenylation Sites and Exons. Mol Cell. 2009, 36 (2): 245-254. 10.1016/j.molcel.2009.10.008.PubMed CentralView ArticlePubMedGoogle Scholar
- Tilgner H, Nikolaou C, Althammer S, Sammeth M, Beato M, Valcarcel J, Guigo R: Nucleosome positioning as a determinant of exon recognition. Nat Struct Mol Biol. 2009, 16 (9): 996-U124. 10.1038/nsmb.1658.View ArticlePubMedGoogle Scholar
- Schwartz S, Meshorer E, Ast G: Chromatin organization marks exon-intron structure. Nat Struct Mol Biol. 2009, 16 (9): 990-995. 10.1038/nsmb.1659.View ArticlePubMedGoogle Scholar
- Schor IE, Rascovan N, Pelisch F, Allo M, Kornblihtt AR: Neuronal cell depolarization induces intragenic chromatin modifications affecting NCAM alternative splicing. Proc Natl Acad Sci USA. 2009, 106 (11): 4325-4330. 10.1073/pnas.0810666106.PubMed CentralView ArticlePubMedGoogle Scholar
- Kolasinska-Zwierz P, Down T, Latorre I, Liu T, Liu XS, Ahringer J: Differential chromatin marking of introns and expressed exons by H3K36me3. Nat Genet. 2009, 41 (3): 376-381. 10.1038/ng.322.PubMed CentralView ArticlePubMedGoogle Scholar
- Luco RF, Pan Q, Tominaga K, Blencowe BJ, Pereira-Smith OM, Misteli T: Regulation of alternative splicing by histone modifications. Science. 2010, 327 (5968): 996-1000. 10.1126/science.1184208.PubMed CentralView ArticlePubMedGoogle Scholar
- Sims RJ, Millhouse S, Chen CF, Lewis BA, Erdjument-Bromage H, Tempst P, Manley JL, Reinberg D: Recognition of trimethylated histone h3 lysine 4 facilitates the recruitment of transcription postinitiation factors and pre-mRNA splicing. Mol Cell. 2007, 28 (4): 665-676. 10.1016/j.molcel.2007.11.010.PubMed CentralView ArticlePubMedGoogle Scholar
- Saint-Andre V, Batsche E, Rachez C, Muchardt C: Histone H3 lysine 9 trimethylation and HP1 gamma favor inclusion of alternative exons. Nat Struct Mol Biol. 2011, 18 (3): 337-U127. 10.1038/nsmb.1995.View ArticlePubMedGoogle Scholar
- Dhami P, Saffrey P, Bruce AW, Dillon SC, Chiang K, Bonhoure N, Koch CM, Bye J, James K, Foad NS: Complex Exon-Intron Marking by Histone Modifications Is Not Determined Solely by Nucleosome Distribution. Plos One. 2010, 5 (8):
- Schones DE, Cui KR, Cuddapah S, Roh TY, Barski A, Wang ZB, Wei G, Zhao KJ: Dynamic regulation of nucleosome positioning in the human genome. Cell. 2008, 132 (5): 887-898. 10.1016/j.cell.2008.02.022.View ArticlePubMedGoogle Scholar
- Birney E, Stamatoyannopoulos JA, Dutta A, Guigo R, Gingeras TR, Margulies EH, Weng ZP, Snyder M, Dermitzakis ET, Stamatoyannopoulos JA: Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project. Nature. 2007, 447 (7146): 799-816. 10.1038/nature05874.View ArticlePubMedGoogle Scholar
- Celniker SE, Dillon LAL, Gerstein MB, Gunsalus KC, Henikoff S, Karpen GH, Kellis M, Lai EC, Lieb JD, MacAlpine DM: Unlocking the secrets of the genome. Nature. 2009, 459 (7249): 927-930. 10.1038/459927a.PubMed CentralView ArticlePubMedGoogle Scholar
- Yu JX, de Belle I, Liang HG, Adamson ED: Coactivating factors p300 and CBP are transcriptionally crossregulated by Egr1 in prostate cells, leading to divergent responses. Mol Cell. 2004, 15 (1): 83-94. 10.1016/j.molcel.2004.06.030.View ArticlePubMedGoogle Scholar
- Ravel-Chapuis A, Vandromme M, Thomas JL, Schaeffer L: Postsynaptic chromatin is under neural control at the neuromuscular junction. EMBO J. 2007, 26 (4): 1117-1128. 10.1038/sj.emboj.7601572.PubMed CentralView ArticlePubMedGoogle Scholar
- Schlesinger J, Schueler M, Grunert M, Fischer JJ, Zhang Q, Krueger T, Lange M, Toonjes M, Dunkel I, Sperling SR: The Cardiac Transcription Network Modulated by Gata4, Mef2a, Nkx2.5, Srf, Histone Modifications, and MicroRNAs. Plos Genetics. 2011, 7 (2):
- Garriga-Canut M, Schoenike B, Qazi R, Bergendahl K, Daley TJ, Pfender RM, Morrison JF, Ockuly J, Stafstrom C, Sutula T: 2-Deoxy-D-glucose reduces epilepsy progression by NRSF-CtBP-dependent metabolic regulation of chromatin structure. Nat Neurosci. 2006, 9 (11): 1382-1387. 10.1038/nn1791.View ArticlePubMedGoogle Scholar
- Ip JY, Schmidt D, Pan Q, Ramani AK, Fraser AG, Odom DT, Blencowe BJ: Global impact of RNA polymerase II elongation inhibition on alternative splicing regulation. Genome Res. 2010, 21 (3): 390-401.View ArticlePubMedGoogle Scholar
- Shukla S, Kavak E, Gregory M, Imashimizu M, Shutinoski B, Kashlev M, Oberdoerffer P, Sandberg R, Oberdoerffer S: CTCF-promoted RNA polymerase II pausing links DNA methylation to splicing. Nature. 2011, 479 (7371): 74-U99. 10.1038/nature10442.View ArticlePubMedGoogle Scholar
- Luco RF, Misteli T: More than a splicing code: integrating the role of RNA, chromatin and non-coding RNA in alternative splicing regulation. Curr Opin Genet Dev. 2011Google Scholar
- Flicek P, Amode MR, Barrell D, Beal K, Brent S, Chen Y, Clapham P, Coates G, Fairley S, Fitzgerald S: Ensembl 2011. Nucleic Acids Res. 2011, 39: D800-D806. 10.1093/nar/gkq1064.PubMed CentralView ArticlePubMedGoogle Scholar
- Nagasaki H, Arita M, Nishizawa T, Suwa M, Gotoh O: Automated classification of alternative splicing and transcriptional initiation and construction of visual database of classified patterns. Bioinformatics. 2006, 22 (10): 1211-1216. 10.1093/bioinformatics/btl067.View ArticlePubMedGoogle Scholar
- Fujita PA, Rhead B, Zweig AS, Hinrichs AS, Karolchik D, Cline MS, Goldman M, Barber GP, Clawson H, Coelho A: The UCSC Genome Browser database: update 2011. Nucleic Acids Res. 2011, 39: D876-D882. 10.1093/nar/gkq963.PubMed CentralView ArticlePubMedGoogle Scholar
- Jia D, Jurkowska RZ, Zhang X, Jeltsch A, Cheng XD: Structure of Dnmt3a bound to Dnmt3L suggests a model for de novo DNA methylation. Nature. 2007, 449 (7159): 248-U213. 10.1038/nature06146.PubMed CentralView ArticlePubMedGoogle Scholar
- Wang ZB, Zang CZ, Rosenfeld JA, Schones DE, Barski A, Cuddapah S, Cui KR, Roh TY, Peng WQ, Zhang MQ: Combinatorial patterns of histone acetylations and methylations in the human genome. Nat Genet. 2008, 40 (7): 897-903. 10.1038/ng.154.PubMed CentralView ArticlePubMedGoogle Scholar
- Bernstein BE, Stamatoyannopoulos JA, Costello JF, Ren B, Milosavljevic A, Meissner A, Kellis M, Marra MA, Beaudet AL, Ecker JR: The NIH Roadmap Epigenomics Mapping Consortium. Nat Biotechnol. 2010, 28 (10): 1045-1048. 10.1038/nbt1010-1045.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.