Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Sep 4.
Published in final edited form as: Mol Cell. 2014 Aug 21;55(5):791–802. doi: 10.1016/j.molcel.2014.07.012

The long noncoding RNAs NEAT1 and MALAT1 bind active chromatin sites

Jason A West 1,2,4,6, Christopher P Davis 1,2,6, Hongjae Sunwoo 1,2, Matthew D Simon 1,2,5, Ruslan I Sadreyev 1,3, Peggy I Wang 1,3, Michael Y Tolstorukov 1, Robert E Kingston 1,2,7
PMCID: PMC4428586  NIHMSID: NIHMS616563  PMID: 25155612

Summary

Mechanistic roles for many lncRNAs are poorly understood in part because their direct interactions with genomic loci and proteins are difficult to assess. Using a method to purify endogenous RNAs and their associated factors, we mapped the genomic binding sites for two highly expressed human lncRNAs, NEAT1 and MALAT1. We show that NEAT1 and MALAT1 localize to hundreds of genomic sites in human cells, primarily over active genes. NEAT1 and MALAT1 exhibit colocalization to many of these loci, but display distinct gene body binding patterns at these sites, suggesting independent but complementary functions for these RNAs. We also identified numerous proteins enriched by both lncRNAs, supporting complementary binding and function, in addition to unique associated proteins. Transcriptional inhibition or stimulation alters localization of NEAT1 on active chromatin sites, implying that underlying DNA sequence does not target NEAT1 to chromatin and that localization responds to cues involved in the transcription process.

Introduction

Noncoding RNAs (ncRNAs) are critical to multiple cellular processes including the modulation of chromatin structure (Mattick, 2007). In addition to small regulatory RNAs and structural RNAs (reviewed in (Sabin et al., 2013; Wan et al., 2011)), long ncRNAs (lncRNAs) exert regulatory control on gene function through interaction with chromatin-associated proteins (reviewed in (Brockdorff, 2013; Mercer et al., 2009)). Examples of such RNAs include XIST, which mediates X-chromosome inactivation and interacts with Polycomb repressive complex 2 (PRC2) (Zhao et al., 2008), and HOTAIR, which acts as a scaffold for several chromatin modifying complexes (Tsai et al., 2010).

Recent efforts have annotated over 10,000 human lncRNAs (Derrien et al., 2012; Engstrom et al., 2006), with a majority of transcripts remaining functionally uncharacterized. Recently developed technologies, including Capture Hybridization Analysis of RNA Targets (CHART) (Simon et al., 2011), Chromatin Isolation by RNA Purification (ChIRP) (Chu et al., 2011), and RNA Affinity Purification (RAP) (Engreitz et al., 2013), allow mechanistic insights into lncRNA function by identifying putative trans genomic binding sites for endogenous RNAs. These procedures use biotinylated capture oligonucleotides (COs) that hybridize to an RNA of interest to isolate RNA-associated DNA or proteins. Determining the genomic binding of lncRNAs as well as the proteins associated with a given lncRNA using these techniques provides important information about lncRNA function. Cataloguing genomic binding sites can determine whether localization is genic and in cis versus trans as well as correlate lncRNA binding with sequence motifs and other known regulatory elements. RNA binding on chromatin may be targeted via direct interaction with nucleic acids or via specific protein binding partners. Thus, identifying proteins that interact with a given lncRNA can identify factors that might confer function and/or localization. We previously employed CHART to characterize the genomic binding sites for the roX2 RNA in Drosophila S2 cells and for the Xist lncRNA during X-chromosome inactivation in mouse cells (Simon et al., 2011; Simon et al., 2013). In this study, we apply CHART to the human lncRNAs NEAT1 (nuclear enriched abundant transcript 1) and MALAT1 (metastasis-associated lung adenocarcinoma transcript 1). NEAT1 and MALAT1 encode ∼3.7 knt and ∼8 knt single-exon lncRNAs that are conserved at least across placental and therian mammals, respectively (Hutchinson et al., 2007; Ulitsky et al., 2011). An additional 21.7 knt NEAT1 transcript (NEAT1_v2) has also been described. This isoform is transcribed from the same promoter, also localizes to paraspeckles, and may play a dominant role in NEAT1 function (Nakagawa et al., 2011; Sasaki et al., 2009; Sunwoo et al., 2009). While the NEAT1 and MALAT1 loci are adjacent to one another in the genome (approximately 53 kb between the 3′ end of NEAT1_v2 and the transcription start site of MALAT1 on chromosome 11), the RNAs appear to localize to distinct nuclear subdomains. MALAT1 localizes to nuclear speckles, which are nuclear bodies enriched for serine and arginine-rich (SR) splicing factors (Hutchinson et al., 2007). NEAT1 is required for the formation of paraspeckles (Clemson et al., 2009; Sasaki et al.; Sunwoo et al., 2009), which are nuclear bodies named for their close proximity to nuclear speckles (Fox et al., 2002). Paraspeckles are comprised of NEAT1 RNA and multiple proteins including many RNA splicing factors, and have been implicated in sequestering RNAs that respond to cellular stress (Naganuma et al., 2012). Nonetheless, paraspeckles appear to perform nonessential or redundant functions in mice (Nakagawa et al., 2011). While the localization of NEAT1 and MALAT1 has been explored at the scale of nuclear domains, it remains to be determined whether these RNAs interact with chromatin at specific sites within the nucleus.

Here, we identify the genomic binding sites for the human NEAT1 and MALAT1 lncRNAs in human cell lines using CHART-seq. We identify hundreds of trans sites for both NEAT1 and MALAT1 that almost entirely overlap active gene regions. Both RNA/DNA co-FISH and ChIP-seq of the paraspeckle component PSF support the identification of these trans sites. We also introduce the use of proteomics with CHART-enriched material (CHART-MS) to identify proteins that associate with NEAT1 and MALAT1 in vivo, identifying previously known and unknown proteins that associate with NEAT1 and MALAT1. We find that NEAT1 and MALAT1 lncRNAs co-enrich a number of trans genomic binding sites and protein factors, suggesting potential redundancy or cooperation between these two lncRNAs in regulating nuclear organization around nuclear bodies. These two lncRNAs display interesting differences, however. NEAT1 localizes to transcriptional start sites (TSS) and transcriptional termination sites (TTS), whereas MALAT1 primarily localizes across gene bodies and near the TTS, suggesting independent but complementary roles for these RNAs at co-bound sites. NEAT1 localization responds significantly to changes in transcription, indicating a contact with some component of the transcription process. CHART shows that these two lncRNAs interact with proteins resident in nuclear bodies and uniquely target active genic sequences, raising the possibility that NEAT1 and MALAT1 link active genes to nuclear subdomains.

Results

CHART enriches endogenous NEAT1 and MALAT1 RNA

We previously applied COs to study DNA-protein interactions at human telomeres (Déjardin and Kingston, 2009) and to determine the genomic binding sites of the Drosophila roX2 and the mouse Xist lncRNA (Simon et al., 2011; Simon et al., 2013). To identify genomic binding sites for the human lncRNAs NEAT1 and MALAT1 via CHART, we first designed DNA-based COs with a 3′ linker and biotin tag to directly hybridize endogenous RNA transcripts and facilitate RNA purification with its associated factors (both nucleic acids and proteins). To identify accessible RNA regions, we employed an RNase H mapping strategy across the MALAT1 and NEAT1 transcripts isolated from cross-linked nuclear extracts of the human breast adenocarcinoma cell line MCF-7 (see Extended Experimental Procedures). We combined this accessibility information with other criteria including sequence specificity and base content/melting temperatures to design COs (see Extended Experimental Procedures). We previously identified COs for NEAT1 and MALAT1 CHART, yet each condition required multiple COs (Simon et al., 2011). Here, we identify two independent COs targeting either NEAT1 or MALAT1. Each of these four COs generated enrichment levels between 1-10% of total input for RNA (Figure 1A, C), similar to CHART RNA enrichment observed for Xist in murine cells (Simon et al., 2013).

Figure 1.

Figure 1

Open in a new tab

CHART enriches the lncRNAs NEAT1 and MALAT1. (A and B) NEAT1 CHART enrichment of NEAT1 and MALAT1 RNA (panel A) and DNA (panel B) assessed by qPCR from MCF-7 cells. n=3 (C and D) MALAT1 CHART enrichment of NEAT1 and MALAT1 RNA (panel C) and DNA (panel D) assessed by qPCR from MCF-7 cells. n=3 (E) CHART-seq of DNA and PSF ChIP-seq over chromosome 11. Peaks extending beyond the y-axis are depicted with hatch marks and annotated with the actual peak height. (F) Co-enrichment of NEAT1 and MALAT1 RNA and PSF at the NEAT1 and MALAT1 genomic loci. All error bars in A-D represent s.e.m. See also Figure S1 and Table S1.

We determined whether DNA associated with NEAT1 and MALAT1 RNA could be enriched using single COs in the CHART protocol. DNA-enrichment was first measured at the gene loci for NEAT1 and MALAT1, as the site of transcription should be captured from cross-linked chromatin extracts via CHART (Simon et al., 2011). Between 0.1 to 1% of input for the genomic site of transcription was consistently enriched for NEAT1 or MALAT CHART (Figure 1B, D). Importantly, CHART-based DNA enrichment is RNA-mediated, as the use of sense COs (which cannot directly target the RNA but have sequence identity to the DNA locus) eliminates this signal.

To our surprise, we noticed robust RNA and DNA enrichment of MALAT1 from NEAT1 CHART, and vice versa (MALAT1 CHART enriches for NEAT1 RNA and DNA) (Figure 1A-D). This reciprocal localization is consistent with previous reports suggesting potential co-regulation of the NEAT1 and MALAT1 gene loci, including Malat1 knockout mice that demonstrate changes in Neat1 expression (Nakagawa et al., 2012; Zhang et al., 2012).

CHART-seq identifies trans genomic binding sites for NEAT1 and MALAT1

We used the COs characterized above to perform genome-wide mapping of NEAT1 and MALAT 1 trans binding sites using CHART-seq. We generated approximately 25-40 million mapped paired-end reads for each of two independent CO conditions for both NEAT1 and MALAT1. Two independent CO conditions enabled stronger trans site identification by reducing sequence-based off-target hybridization events from influencing downstream analysis (Chu et al., 2011). The two independent COs used for NEAT1 and MALAT1 generated similar CHART-seq results, exhibited by a strong Pearson correlation between the two CO conditions (r = 0.753 between NEAT1 CO1 and CO2; r = 0.756 between MALAT1 CO1 and CO2; Table S1).

The NEAT1 and MALAT1 RNAs bound at numerous sites along chromosome 11 (Figure 1E), including significant enrichment of their genomic site of transcription (Figure 1F, note change in y-axis scale). These results confirmed that NEAT1 CHART enriches the MALAT1 genomic locus, and reciprocally, MALAT1 CHART enriched the NEAT1 genomic locus. We note that these RNAs possess little sequence identity, the DNA region between these two gene loci was not significantly enriched, nor was enrichment observed at other nearby genes, including genes that produce highly expressed transcripts (Figure 1F), leading us to conclude that enrichment is specific.

To identify putative trans binding sites for NEAT1 and MALAT1 we analyzed sites of enrichment seen with both COs. We combined SPP-based (Kharchenko et al., 2008) and MACS-based (Zhang et al., 2008) peak calling approaches to identify 1251 putative NEAT1 binding sites and 670 MALAT1 binding sites (Figure 2A and Table S2). Broad regions of enrichment detected in our CHART-seq data had an average size of approximately 10 kb (Figure S1). DNA motif analysis using MEME does not support direct RNA-DNA hybridization at these trans sites as motifs identified within the CHART-enriched material lack sequence identity to the NEAT1 and MALAT1 transcript (Figure S2).

Figure 2.

Figure 2

Open in a new tab

NEAT1 and MALAT1 localize to actively-transcribed genomic loci. (A) Venn diagram depicting number of NEAT1-specific, MALAT1-specific, and co-enriched CHART-seq sites. (B) Overlap between gene bodies and CHART-seq binding sites. (C and D) Probability density distributions of expression levels (RPKM; reads per kb per million) of genes overlapping MALAT1 peaks (panel C), NEAT1 peaks (panel D), or both (panels C and D). (E-G) Probability density distributions of input-normalized coverage density for H3K4me3 (panel E), H3K36me3 (panel F), and H3K27me3 (panel G) over a 500 bp window around NEAT1 and MALAT1 peak maxima. Dotted lines in panels E-G clarify separation between peak maxima that are enriched (coverage density > 1) or depleted (coverage density < 1) for a particular modification. (H) Probability density distributions of NEAT1 and MALAT1 peak maxima positions along length-normalized gene bodies. (I) CHART-seq enrichment of NEAT1 and MALAT1 and ChIP-seq enrichment of PSF at the SAP18 genomic locus. See also Figure S2, Table S2, and Table S3.

Previously, it was suggested that paraspeckles do not interact with chromatin, as paraspeckles localized to DAPI-depleted subnuclear domains (Fox et al., 2002). However, NEAT1 has recently been linked to transcriptional regulation (Hirose et al., 2014; Imamura et al., 2014). Remarkably, the CHART-seq data revealed that NEAT1 and MALAT1 enriched regions almost entirely overlap gene bodies (Figure 2B, Table S3), suggesting specificity in the CHART-seq signal and organizational coupling of nuclear speckles and paraspeckles to active genes. Trans sites identified by CHART-seq for NEAT1 and MALAT1 conservatively identified 314 sites of co-enrichment (Figure 2A).

To characterize NEAT1 and MALAT1 interacting loci, we compared binding sites to previously reported datasets, including gene annotations, gene expression levels, and covalent histone marks (Consortium, 2004). We found that the sites of enrichment for NEAT1 and MALAT1 were associated with expressed genes (Figure 2C, D). Interestingly, while the putative binding sites of NEAT1 and MALAT1 were significantly comprised of expressed genes (Table S3), they did not show bias towards only the most abundant transcripts in MCF-7 cells, and a small proportion of genes overlapping NEAT1 and MALAT1 were not expressed (RPKM = 0, Figure 2C, D).

Consistent with localization over transcribed genes, NEAT1 and MALAT1 peaks were enriched for covalent histone modifications associated with euchromatin and active genes, including H3K4me3 (note NEAT1 enrichment, Figure 2E) and H3K36me3 (note MALAT1 enrichment, Figure 2F), but not H3K27me3, a marker of transcriptionally silent genes (Figure 2G). Interestingly, two populations of NEAT1 peak maxima were observed: one depleted of H3K4me3 and one enriched for this modification (Figure 2E). This dual population arose due to localization of NEAT1 at either the TSS or TTS, as NEAT1 peaks within 500 bp of a TSS were enriched for H3K4me3 and peaks located farther than 500 bp from a TSS were depleted for this mark (Figure S2). We observed MALAT1 peaks highly enriched for H3K36me3, which is associated with gene bodies, while MALAT1 peaks were not strongly enriched for the TSS-associated H3K4me3. We conclude that while NEAT1 and MALAT1 are enriched at many of the same genes, they have different sublocalizations at many of these sites. Indeed, metagene profiles as well as analysis of regions with maximal NEAT1 binding showed enrichment over both the TSS and the TTS of genes, while MALAT1 binding maxima were depleted at the TSS and enriched over the gene body and the TTS (Figure 2H, Table S3), as exemplified by the SAP18 locus (Figure 2I). NEAT1 and MALAT1 showed increased density at those genes co-enriched by both RNAs, as well as at transcriptionally active genes, while still maintaining distinct gene body binding patterns (Figure S2). Taken together, these analyses indicate that NEAT1 and MALAT1 bind to specific euchromatic, actively-transcribed regions, and the unique binding characteristics at regions co-bound by NEAT1 and MALAT1 suggest distinct roles for these RNAs.

NEAT1 and MALAT1 binding sites are distributed throughout the genome (Figure S3A), as seen for a representative autosome, chromosome 17, showing NEAT1 and MALAT1 binding across the chromosome (Figure 3A). Specific examples of trans genomic binding sites include a NEAT1-specific target SP3 (chromosome 2, Figure 3B), a NEAT1 and MALAT1 co-enriched locus HEXIM1 (chromosome 17, Figure 3C), and the MALAT1-specific CALR and RAD23A loci (chromosome 19, Figure 3D). Localization of NEAT1 and MALAT1 to selected trans genomic binding sites was confirmed by qPCR on CHART-enriched DNA (Figure 3E, F) and RNA (Figure S3B-C).

Figure 3.

Figure 3

Open in a new tab

NEAT1 and MALAT1 localize to genomic sites in trans. (A) NEAT1 and MALAT1 DNA CHART-seq and PSF ChIP-seq over chromosome 17. The HEXIM1 locus is boxed. (B-D) DNA CHART-seq and PSF ChIP-seq depicting the NEAT1-specific target SP3 (panel B), colocalization of NEAT1 and MALAT1 RNA at the HEXIM1 locus (panel C), and the MALAT1-specific target CALR/RAD23A (panel D). (E and F) qPCR validation of trans sites enriched by NEAT1 (panel E) and MALAT1 (panel F) DNA CHART in MCF-7 cells. Error bars represent s.e.m. n=3. See also Figure S3.

Molecular confirmation of NEAT1 binding sites

We used several approaches to confirm that NEAT1 associates with specific gene loci. We used FISH to image NEAT1 RNA and putative trans genomic sites and inspected cells for overlap in NEAT1 RNA and candidate DNA signal. We focused these efforts on karyotypically-normal BJ foreskin fibroblasts to avoid confounding results due to aneuploidy in MCF-7 cells. CHART qPCR of a set of candidate gene loci in BJ foreskin fibroblasts suggest that these cells possess similar NEAT1 and MALAT1 binding sites as those identified in MCF-7 (Figure S4A-D). RNA FISH in BJ cells confirmed the punctate, nuclear localization of NEAT1 RNA (Figure 4A). We performed RNA/DNA co-FISH and observed a strong correlation (p < 0.05; chi-squared test) between NEAT1 RNA and the MALAT1 genomic locus compared to a gene-poor region on chromosome 5 showing minimal NEAT1 localization in CHART-seq (Figure S4E), with roughly 90% of MALAT1 DNA loci colocalizing with NEAT1 RNA (Figure 4A and 4B). This result supports NEAT1 RNA binding at the MALAT1 genomic locus, however, these results may be due to the relatively low resolution of RNA/DNA co-FISH imaging, as NEAT1 is transcribed approximately 53 kb away from MALAT1.

Figure 4.

Figure 4

Open in a new tab

Validation of NEAT1 trans site localization by RNA/DNA co-FISH. (A) Representative images of NEAT1 RNA colocalization with MALAT1 and MAPK15 DNA, but not a gene-poor region, determined by RNA/DNA co-FISH. (B and C) Quantification of NEAT1 RNA colocalization with the MALAT1 locus (panel B) and each allele for trans genomic loci (panel C). “n” indicates number of DNA loci analyzed. Asterisks indicate a significant (p < 0.05) increase in NEAT1 colocalization with the particular DNA locus compared to the gene-poor region, determined by a chi-squared test. See also Figure S4.

To validate putative trans site binding of NEAT1, we tested for NEAT1 RNA colocalization using RNA/DNA co-FISH with several trans genomic sites that were highly enriched via CHART. We examined five genes with positive association with NEAT1, such as the MAPK15 locus (Figure 4A, Figure S4), RPS17 -- a highly expressed gene with low NEAT1 localization as measured by CHART-seq (Figure S4), and a gene-poor region. We found a statistically significant (p < 0.05, chi-squared test) increase in NEAT1 colocalization with four of the five sites compared to the gene-poor region (Figure 4C). The RPS17 locus also showed putative colocalization with NEAT1, despite low NEAT1 CHART-seq coverage density. This may reflect insufficient sequencing depth in our CHART-seq experiment or technical limits with interpretation of FISH, as RNA/DNA co-FISH has limited resolution. Here it is further limited by the high number of paraspeckles per cell, which increases the possibility of random overlap of RNA/DNA signal, and the lack of information on the dwell time of lncRNAs at any specific chromatin site (Clark et al., 2012; Mao et al., 2011). Thus, while RNA/DNA co-FISH supported the mapping of NEAT1 binding sites as determined by CHART, this analysis was not conclusive.

As another means of validation, we compared the genomic localization of NEAT1 to the binding sites of a protein known to interact with NEAT1 using ChIP-seq. We chose to examine PSF (Polypyrimidine-tract-binding protein-associated splicing factor), a paraspeckle component that associates with NEAT1 (Sasaki et al., 2009). Consistent with our CHART results, PSF localized to both the NEAT1 and MALAT1 gene loci (Figure 1F) and multiple NEAT1 and MALAT1 trans sites (Figure 3B-D and Figure S5A-B). PSF binding was observed at the TSSs of transcribed genes, while inactive genes lacked robust PSF signal (Figure 5A). PSF was enriched most strongly at sites where NEAT1 was bound either alone or in combination with MALAT1, whereas a weaker enrichment of PSF was seen at MALAT1-specific binding sites (Figure 5B; note that the location of MALAT1 and NEAT1 peaks differ within ‘shared’ sites so PSF overlap with those maxima are distinct.).

Figure 5.

Figure 5

Open in a new tab

NEAT1 and MALAT1 colocalize with the paraspeckle component, PSF. (A) Metagene profile of PSF ChIP-seq density over active (RPKM > 1) and inactive (RPKM <= 1) genes. (B) PSF ChIP-seq coverage centered over NEAT1-specific and MALAT1-specific peak maxima (“Unique”) and maxima of NEAT1 and MALAT1 peaks that overlap with MALAT1 and NEAT1 binding sites, respectively (“Shared”). See also Figure S5.

To determine the correlation between PSF and lncRNA density over NEAT1 and MALAT1 peaks, we plotted CHART-seq and associated PSF ChIP-seq coverage density for each NEAT1 and MALAT1 peak. We obtained a positive correlation between binding of NEAT1 and PSF at sites where NEAT1 alone was bound (Figure S5C, Pearson's r = 0.49). As expected from the previous observations that PSF marks paraspeckles (Mao et al., 2011; Naganuma et al., 2012; Sasaki et al., 2009), we observed little correlation between PSF and MALAT1 binding over regions where MALAT1 was solely bound (Figure S5D, Pearson's r = 0.24). However, a substantial increase in correlation with PSF binding was seen when considering only the NEAT1 peaks that overlap with MALAT1 binding sites (Figure S5D, Pearson's r = 0.40), and similarly, the MALAT1 peaks that intersect with NEAT1 binding sites (Figure S5C, Pearson's r = 0.51). Similar correlations at shared sites were observed when comparing PSF density with either MALAT1 or NEAT1 density over NEAT1 or MALAT1 peaks, respectively (Figure S5E, F). We conclude from these data that NEAT1, MALAT1, and PSF co-occupy many sites in the genome (Mao et al., 2011; Naganuma et al., 2012; Sasaki et al., 2009).

NEAT1 localization responds to transcription

NEAT1 and MALAT1 localized to highly expressed genes (Figure 2C-D), raising the question as to whether NEAT1 and MALAT1 are responsive to the expression level of these target genes or whether they localize to their targets regardless of transcriptional status. To address this issue, we determined the effects of transcriptional inhibition or stimulation on NEAT1 and MALAT1 localization.

We treated cells with flavopiridol, a transcription elongation inhibitor that disrupts P-TEFb-mediated phosphorylation of serine 2 on the RNA polymerase II C-terminal domain (Pol II Ser2P) (Chao et al., 2000; Chao and Price, 2001). Cells treated with flavopiridol for 1 hour displayed decreased levels of Pol II Ser2P compared to the vehicle control, indicating global depletion of elongating Pol II (Figure S6A). Consistent with a reduction in transcription elongation, we observe decreased enrichment of NEAT1 and MALAT1 RNA within the NEAT1 and MALAT1 genomic loci and increased enrichment near the TSSs of these genes (Figure S6B-C). We observed alterations in NEAT1 and MALAT1 localization at trans genomic binding sites, such as the RPS24 locus, which showed a decrease in NEAT1 and MALAT1 binding after treatment with flavopiridol (Figure 6A, Figure S6D). This decrease in localization was not due to a reduction in NEAT1 and MALAT1 RNA, as steady state levels of NEAT1 and MALAT1 were not decreased upon flavopiridol treatment (Figure S6E). In contrast, treatment with flavopiridol also led to increased localization of NEAT1 at numerous regions, such as the promoter of the KMT2E gene (Figure 6B). To analyze these effects genome-wide, we identified regions enriched by NEAT1 and MALAT1 CHART for vehicle-treated and flavopiridol-treated cells. We selected peaks found in both biological replicates for each CO condition that were not identified in the corresponding sense CO CHART-seq (Table S2, Table S4). Peaks in the vehicle-treated samples show significant overlap with our previously defined set of high-confidence CHART-seq peaks (Figure S6F, Table S2), although they were fewer in number. We generated metagene profiles of NEAT1 and MALAT1 binding sites with respect to RefSeq annotated genes found within 3 kb of the CHART-seq peaks (Figure 6C-D, Figure S6G-H). NEAT1 and MALAT1 binding sites in vehicle-treated cells showed nearly uniform distribution across the gene body (Figure 6C-D). However, upon treatment with flavopiridol, NEAT1 localizes primarily to the TSS (Figure 6C), whereas MALAT1 shows a less striking change in localization (Figure 6D). We note that the NEAT1 binding profile for vehicle-treated cells does not precisely match the binding profile obtained from high-confidence NEAT1 binding sites (Figure 2H). This is likely due to differences in sample preparation and depth of coverage, as discussed in Extended Experimental Procedures. We conclude from these analyses that inhibition of transcription elongation leads to a distinct change in NEAT1 localization towards transcription initiation sites.

Figure 6.

Figure 6

Open in a new tab

NEAT1 localization is responsive to changes in transcription. (A and B) NEAT1 DNA CHART-seq over the RPS24 locus (panel A) and KMT2E locus (panel B) in cells treated with DMSO (“Vehicle”) or flavopiridol. (C-D) Probability density distributions of NEAT1 CO1 (panel C) or MALAT1 CO1 (panel D) peak midpoint positions along nearby length-normalized gene bodies in vehicle or flavopiridol treated MCF-7 cells. (E) NEAT1 DNA CHART-seq over the estrogen receptor-regulated gene GREB1 in cells treated with ethanol (“Vehicle”) or 17β-estradiol (“E2”). (F) Table describing number of CHART-seq peaks identified upon vehicle treatment and E2 treatment. (G-H) Scatterplots of transcript abundance in cells treated with vehicle or E2, as measured by GRO-seq (see Extended Experimental Procedures), for genes nearby vehicle-specific (black) or E2-specific (red) CHART-seq peaks for NEAT1 CO1 (panel G) or MALAT1 CO1 (panel H). Indicated p-values are calculated using a Wilcoxon signed rank test. See also Figure S6, Table S2, Table S4, and Table S5.

We addressed the converse situation, whether NEAT1 and MALAT1 localization is responsive to induction of previously repressed genes, by assessing the effects of estrogen signaling on NEAT1 and MALAT1 localization. Through the activity of estrogen receptors α and β (ERα and ERβ), treatment of ERα-positive cells with the ligand 17β-estradiol (E2) stimulates a well-characterized transcriptional response (reviewed in (Cheung and Kraus, 2010)). We treated ERα-positive MCF-7 cells with E2 for 40 minutes to stimulate widespread alterations in gene expression (Hah et al., 2011). After stimulation, we observed increased NEAT1 localization, but minimal increases in MALAT1 localization, at the GREB1 locus (Figure 6E, Figure S6I), a previously characterized ERα target that is transcriptionally upregulated upon E2 stimulation (Ghosh et al., 2000; Lin et al., 2004). To obtain a genome-wide view of the effects of E2 treatment on lncRNA localization, we identified regions enriched in both biological replicates for each CO condition that were also absent in the corresponding sense control CHART-seq (Figure 6F, Table S2). A greater number of regions are bound by NEAT1 and MALAT1 upon E2 stimulation, and many of these regions did not overlap regions identified in vehicle treated cells (Figure 6F, “E2-specific peaks”). Genes closest to these E2-specific peaks were significantly enriched for annotated ERα targets (Table S5). The transcriptional status of genes found near these E2-specific peaks was analyzed using previously generated global run-on sequencing (GRO-seq) data from MCF-7 cells treated with E2 (Hah et al., 2011). E2-specific NEAT1 targets identified in each CO condition show increased gene expression upon E2 treatment compared to vehicle treatment as measured by GRO-seq (Figure 6G, Figure S6J), while MALAT1 targets showed variable responses to E2 stimulation (Figure 6H, Figure S6K). These results show that NEAT1 localizes to genes upregulated during estrogen signaling and, together with the flavopiridol data, demonstrate that NEAT1 binding is responsive to the transcriptional status of genomic targets.

Identification of proteins interacting with NEAT1 and MALAT1

Based upon the mapping results, we were interested in determining whether these two RNAs might have overlapping specificity in protein binding. We adapted CHART to identify the full complement of proteins associated with RNAs in vivo by using mass spectrometry (CHART-MS). We first optimized NEAT1 and MALAT1 CHART enrichment for specific proteins known to associate with either paraspeckles (PSPC1 and PSF) or nuclear speckles (SRSF1) using immunoblotting of CHART-enriched extracts (Figure 7A; see also (Simon et al., 2011)). Enriched levels of these proteins were between 0.05% to 1% of the total level found in the input and were significantly enriched compared to histone H3. CHART-mediated protein enrichment is RNA-dependent, as sense COs fail to detectably enrich these protein factors. We next performed CHART-MS for NEAT1 and MALAT1 (Figure 7B, Table S6). We cross-referenced the proteins identified through CHART-MS with 140 proteins identified through proteomic analysis of nuclear speckles in mouse liver nuclei (Saitoh et al., 2004) and 40 paraspeckle components characterized through a fluorescence-based screen for proteins colocalizing with PSF (Naganuma et al., 2012). We restricted our analysis to proteins enriched at least ten-fold over input and observed numerous known nuclear speckle and paraspeckle components associated with MALAT1 and NEAT1 (Figure 7B). We observed a similar number of nuclear speckle and paraspeckle components associated with NEAT1 and MALAT1, and many of these proteins are associated with both RNAs (Figure 7C), consistent with the high degree of genomic colocalization seen for NEAT1 and MALAT1.

Figure 7.

Figure 7

Open in a new tab

CHART enriches proteins associated with NEAT1 and MALAT1 and reveals mechanistic insights into lncRNA function. (A) Western blot of NEAT1 and MALAT1 CHART-enriched material for the paraspeckle components PSPC1 and PSF, the nuclear speckle component SRSF1, and a chromatin-associated protein, histone H3. (B) Table indicating number of known nuclear speckle and paraspeckle proteins enriched ten-fold over input for the indicated CHART conditions. (C) Venn diagram depicting overlap between proteins associated with NEAT1 and MALAT1. (D) Western blot of NEAT1 and MALAT1 CHART-enriched material for PURA and ESRP2 protein. (E) Model of NEAT1 and MALAT1 binding to trans genomic sites. (F) Model depicting preferential localization of NEAT1 RNA at the TSS and TTS and MALAT1 RNA at the TTS. See also Table S6 and Figure S7.

We identified several proteins not previously associated with NEAT1 or MALAT1 (Figure S7). These include several proteins known to bind nucleic acids, such as PURA and PURB (reviewed in (White et al., 2009)) and RBM12B. Others have known roles in RNA processing, such as ESRP2 (Warzecha et al., 2009) and SAFB2 (Sergeant et al., 2007). Another class of proteins enriched by CHART-MS have roles in transcriptional regulation, such as NCOA5 (Sauvé et al., 2001). The identification of proteins involved in a variety of nuclear processes by CHART-MS is consistent with previous studies identifying splicing factors, RNA 3′-end processing factors, and transcriptional regulators localized to nuclear speckles (reviewed in (Spector and Lamond, 2011)). We confirmed the specificity of uncharacterized protein associations with NEAT1 and MALAT1 for two of the novel interacting proteins. Western blotting of CHART-enriched material confirmed NEAT1 and MALAT1 enrichment of PURA and ESRP2 at levels similar to known paraspeckle and nuclear speckle proteins (Figure 7D). We conclude that NEAT1 and MALAT1 interact with many overlapping proteins, although each also has unique interactions.

Discussion

Mechanistic studies of the functions of lncRNAs have been limited by the ability to assess direct targets for these molecules in vivo. We show here that NEAT1 and MALAT1 bind to hundreds of active genes and that transcriptional activity has a significant influence on NEAT1 localization. Thus, NEAT1 localization is responsive to transcriptional status, implying an ability of NEAT1 to take locational cues from either the proteins involved in transcription or from the newly transcribed RNA. A second finding from these studies is that NEAT1 and MALAT1 co-localize to many genes, although their binding patterns across these genes differ, with NEAT1 binding both the TSSs and TTSs of genes and MALAT1 enriching gene bodies and TTSs. The differential binding patterns of NEAT1 and MALAT1 over co-bound sites suggests independent but synergistic functional roles for these RNAs on co-occupied genes (Figure 7E-F).

While previous work has established the nuclear localization of MALAT1 and NEAT1 within nuclear speckles and paraspeckles, respectively, it has been assumed that paraspeckles do not significantly interact with the genome (Fox et al., 2002). Here, we show that NEAT1 and MALAT1 indeed associate with discrete genomic locations. MALAT1 has also been shown to interact with the Polycomb repressive complex 1 (PRC1) subunit CBX4 and may help to modulate growth-control gene localization in the nucleus, providing supporting evidence that MALAT1 is localized to chromatin (Yang et al., 2011). As NEAT1 and MALAT1 have relatively short half-lives, their high turnover rates may indicate these interactions are transient, consistent with the previously reported dynamic organization of paraspeckles (Clark et al., 2012; Mao et al., 2011). We found that NEAT1 was re-localized with treatments that altered transcription in a one-hour time frame, consistent with rapid and dynamic localization of this lncRNA.

We identified a significant number of gene loci co-enriched by both NEAT1 and MALAT1. Previous work has suggested crosstalk between nuclear speckles and paraspeckles. In addition to being frequently localized adjacent to one another in the nucleus (Fox et al., 2002), a number of canonical paraspeckle proteins, such as PSF and p54/nrb, are also found within nuclear speckles (Saitoh et al., 2004). Analysis of the three-dimensional chromatin interactions in human cells supports nuclear organization that places the MALAT1 genomic locus in close proximity to the NEAT1 locus (Jin et al., 2013). Isoforms of NEAT1 have also been shown to localize to nuclear speckles in HeLa cells upon treatment of cells with inhibitors of transcription elongation (Sunwoo et al., 2009). These prior studies are further supported by our CHART-seq and CHART-MS results showing substantial overlap between nuclear speckle and paraspeckle components associated with NEAT1 and MALAT1. Consistent with a potential co-regulatory role between NEAT1 and MALAT1, Malat1 knockout mice show variations in Neat1 RNA expression depending on the cell type examined (Nakagawa et al., 2012; Zhang et al., 2012). While functional defects in Malat1 and Neat1 knockout mice are subtle (Eißmann et al., 2012; Nakagawa et al., 2012; Nakagawa et al., 2011; Zhang et al., 2012), double knockout mice may reveal important and synergistic roles for these RNAs. In addition, CO design strategies that distinguish the various isoforms of these lncRNAs may reveal mechanistic insights for alternately processed forms of NEAT1 and MALAT1.

Some of the conclusions we have reached are based upon adapting the CHART protocol to identify RNA associated proteins using mass spectrometry (CHART-MS). CHART-MS represents a useful method to study lncRNAs by identifying proteins that may target lncRNAs to chromatin or serve independent functions. Alternative methods to probe interactions between proteins and RNAs include incubation of in vitro-transcribed RNA with nuclear extract or purified proteins, isolation of RNA through immunoprecipitation of a specific protein (Gilbert et al., 2004; Ule et al., 2003), and use of RNA aptamers, such as the MS2 tagging system (Zhou et al., 2002) or streptavidin-binding RNA aptamers (Srisawat and Engelke, 2001), to purify an RNA and associated proteins. Through CHART, we use biotinylated COs to isolate specific RNAs and associated proteins cross-linked in vivo without altering the endogenous RNA sequence or using cell-free extracts that might generate spurious interactions. We identified multiple proteins known to be present in either nuclear speckles or paraspeckles and implicated several proteins as uncharacterized components of these nuclear bodies. These data are consistent with a role for NEAT1 and MALAT1 in nuclear organization around nuclear bodies and establish CHART as a method to biochemically define DNA and proteins associated with lncRNAs. While NEAT1 and MALAT1 are highly expressed and therefore biochemically tractable targets for CHART-MS, further optimization of CHART-MS may be needed to identify proteins associated with low-abundance transcripts.

We propose, based upon these studies, that nuclear bodies may be organized around the genes or other genomic loci that they regulate. Previous studies have described dynamic interactions between nuclear speckles and gene bodies, where speckles may serve as a concentrated reservoir of splicing factors that shuttle to transcribed genes (Misteli et al., 1997; Zeng et al., 1997). Speckles frequently localize within the vicinity of actively transcribed genes undergoing co-transcriptional splicing (Huang and Spector, 1996; Smith et al., 1999). Here, we show that perturbing transcription, either through inhibition or stimulation, alters the binding of NEAT1. As NEAT1 localization responds to these changes in transcription, NEAT1 interactions with chromatin likely depend upon contact with some component of the transcription process and not due to DNA sequence alone.

Our CHART data demonstrate NEAT1 and MALAT1 binding at active genes. NEAT1 and MALAT1 could be playing a structural role in the organization of nuclear bodies at highly transcribed loci, as NEAT1 has been shown to organize paraspeckle formation around its site of transcription (Clemson et al., 2009; Mao et al., 2011). One possibility, which builds upon previous theories concerning lncRNAs such as Xist and HOTAIR (Engreitz et al., 2013; Simon et al., 2013; Tsai et al., 2010), is that NEAT1 and MALAT1 serve as scaffolds that bind proteins that also interact with components of nuclear speckles and paraspeckles and proteins that interact directly with the RNA and/or DNA at specific transcribed loci. Thus these lncRNAs could bridge chromosomal locations to these nuclear structures.

This nuclear organization might have particular disease relevance, as brain tissue demonstrating frontotemporal lobar degeneration (FTLD) have been found to contain high transcript levels for both NEAT1 and MALAT1 in TDP43-positive disease-associated inclusions when compared to healthy tissue (Tollervey et al., 2011). MALAT1 was also identified as a prognostic marker of non-small cell lung cancer metastasis (Ji et al., 2003) and has been further characterized as a potential regulator of metastasis-associated genes (Eißmann et al., 2012). Further exploration of the effects of aberrant expression of lncRNAs, such as NEAT1 and MALAT1, will yield critical insight into the function of lncRNAs, their mechanisms of targeting and function, and their potential effects on gene expression and nuclear organization in human disease.

Experimental Procedures

CHART

Cross-linked and sonicated nuclear extract was incubated with capture oligonucleotides and hybridized overnight. Hybridized material was captured with magnetic streptavidin resin (Invitrogen). Bound material was washed and eluted with RNase H (New England Biolabs) as previously described (Simon et al., 2011). A detailed description of nuclear extract preparation, capture oligonucleotide design, and purification and analysis of eluted nucleic acids and proteins is described in Extended Experimental Procedures.

Chromatin immunoprecipitation

Chromatin isolated from 108 cross-linked MCF-7 cells was immunoprecipitated overnight using Protein A magnetic beads (New England Biolabs) resuspended in RIPA buffer and conjugated with PSF antisera (sc-101137, Santa Cruz Biotechnology) or mouse IgG antisera (sc-2025, Santa Cruz Biotechnology). Beads were washed twice with RIPA buffer supplemented with 100 mM urea and then washed four times with RIPA buffer. Bound material was eluted using Proteinase K (Ambion), and heated to reverse cross-links. Purified DNA was analyzed using qPCR or sequencing as detailed in Extended Experimental Procedures.

RNA and DNA FISH

BJ cells grown on a glass coverslip and fixed in 4% formaldehyde in PBS. Cells were washed with PBS, permeabilized with Triton X-100, dehydrated through a series of ethanol washes, and hybridized overnight with a digoxigenin-labeled NEAT1 probe. RNA FISH signal was detected by incubating with FITC-labeled anti-digoxygenin antibody (Roche). Cells were fixed again in 4% formaldehyde. Cellular RNAs were removed by RNase A (Life Technologies). Chromosomal DNA was denatured and samples were dehydrated with ethanol. DNA FISH was performed with Cy3 labeled-BAC probes. Nuclei were counter-stained with Hoechst 33342 (Life Technologies). Imaging was performed on Nikon 90i microscope equipped with a 60X/1.4 N.A. VC objective lens, Orca ER camera (Hamamatsu) and Volocity software (Perkin Elmer). All probes were prepared by nick translation using DNA polymerase I (Promega), DNase I (Promega), Digoxigenin-dUTP (Roche), or Cy3-dUTP (Enzo Life Sciences). BACs used for this study are listed in Extended Experimental Procedures. To assess colocalization of NEAT1 RNA and trans loci DNA signal, only 2n cells were included for analysis and each allele within a cell was considered an independent data point. Colocalization was quantified using CellProfiler (Kamentsky et al., 2011). Statistical significance of the difference between NEAT1 localization at a trans site and the gene-poor region was assessed using a chi-squared test.

See Extended Experimental Procedures for additional details on all protocols and methods used.

Supplementary Material

01
02
03
04

Highlights.

  • High resolution genomic binding sites for the human lncRNAs NEAT1 and MALAT1

  • NEAT1 and MALAT1 exhibit distinct colocalized binding at active chromatin

  • Transcription impacts NEAT1 localization patterns

  • CHART-MS identifies proteins associated with endogenous RNAs in vivo

Acknowledgments

We thank J. Engreitz, M. Guttman, and the Kingston lab for helpful discussions, A. Plys for critical reading of the manuscript, S. Bowman for library construction optimization, D. Levens, S. Diederichs, T. Gutschner, F. Rigo, and ISIS Pharmaceuticals for reagents not directly used in this study, R. Tomaino for advice on proteomic analysis, and U. Kim and the MGH Sequencing Core. J.A.W. and R.E.K. were supported by the National Institute of General Medical Sciences of the NIH (F32GM093491 to J.A.W.; R01GM043901 and R37GM048405 to R.E.K.). C.P.D. was supported by a National Science Foundation graduate fellowship. M.D.S. was supported by the Helen Hay Whitney Foundation.

Footnotes

Author Contributions: J.A.W., C.P.D., M.D.S., and R.E.K. designed the project. H.S. performed RNA FISH, and J.A.W. and C.P.D. conducted all other experiments. C.P.D., R.I.S., P.I.W., and M.Y.T. performed bioinformatic analyses. J.A.W., C.P.D., and R.E.K. wrote the manuscript.

Accession Numbers: All datasets generated for this study are available under GEO accession number GSE58444.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Brockdorff N. Noncoding RNA and Polycomb recruitment. RNA. 2013;19:429–442. doi: 10.1261/rna.037598.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Chao SH, Fujinaga K, Marion JE, Taube R, Sausville EA, Senderowicz AM, Peterlin BM, Price DH. Flavopiridol inhibits P-TEFb and blocks HIV-1 replication. J Biol Chem. 2000;275:28345–28348. doi: 10.1074/jbc.C000446200. [DOI] [PubMed] [Google Scholar]
  3. Chao SH, Price DH. Flavopiridol inactivates P-TEFb and blocks most RNA polymerase II transcription in vivo. J Biol Chem. 2001;276:31793–31799. doi: 10.1074/jbc.M102306200. [DOI] [PubMed] [Google Scholar]
  4. Cheung E, Kraus WL. Genomic analyses of hormone signaling and gene regulation. Annu Rev Physiol. 2010;72:191–218. doi: 10.1146/annurev-physiol-021909-135840. [DOI] [PubMed] [Google Scholar]
  5. Chu C, Qu K, Zhong F, Artandi S, Chang H. Genomic maps of long noncoding RNA occupancy reveal principles of RNA-chromatin interactions. Mol Cell. 2011;44:667–678. doi: 10.1016/j.molcel.2011.08.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Clark M, Johnston R, Inostroza-Ponta M, Fox A, Fortini E, Moscato P, Dinger M, Mattick J. Genome-wide analysis of long noncoding RNA stability. Genome Res. 2012;22:885–898. doi: 10.1101/gr.131037.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Clemson C, Hutchinson J, Sara S, Ensminger A, Fox A, Chess A, Lawrence J. An architectural role for a nuclear noncoding RNA: NEAT1 RNA is essential for the structure of paraspeckles. Mol Cell. 2009;33:717–726. doi: 10.1016/j.molcel.2009.01.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Consortium, ENCODE Project. The ENCODE (ENCyclopedia Of DNA Elements) Project. Science. 2004;306:636–640. doi: 10.1126/science.1105136. [DOI] [PubMed] [Google Scholar]
  9. Consortium ENCODE Project. A user's guide to the encyclopedia of DNA elements (ENCODE) PLoS Biol. 2011;9:e1001046. doi: 10.1371/journal.pbio.1001046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Déjardin J, Kingston R. Purification of proteins associated with specific genomic Loci. Cell. 2009;136:175–186. doi: 10.1016/j.cell.2008.11.045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Derrien T, Johnson R, Bussotti G, Tanzer A, Djebali S, Tilgner H, Guernec G, Martin D, Merkel A, Knowles D, et al. The GENCODE v7 catalog of human long noncoding RNAs: analysis of their gene structure, evolution, and expression. Genome Res. 2012;22:1775–1789. doi: 10.1101/gr.132159.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Eißmann M, Gutschner T, Hämmerle M, Günther S, Caudron-Herger M, Groß M, Schirmacher P, Rippe K, Braun T, Zörnig M, et al. Loss of the abundant nuclear non-coding RNA MALAT1 is compatible with life and development. RNA Biol. 2012;9:1076–1087. doi: 10.4161/rna.21089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Engreitz JM, Pandya-Jones A, McDonel P, Shishkin A, Sirokman K, Surka C, Kadri S, Xing J, Goren A, Lander ES, et al. The Xist lncRNA exploits three-dimensional genome architecture to spread across the X chromosome. Science. 2013;341:1237973. doi: 10.1126/science.1237973. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Engstrom PG, Suzuki H, Ninomiya N, Akalin A, Sessa L, Lavorgna G, Brozzi A, Luzi L, Tan SL, Yang L, et al. Complex Loci in human and mouse genomes. PLoS Genet. 2006;2:e47. doi: 10.1371/journal.pgen.0020047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fox A, Lam Y, Leung A, Lyon C, Andersen J, Mann M, Lamond A. Paraspeckles: a novel nuclear domain. Curr Biol. 2002;12:13–25. doi: 10.1016/s0960-9822(01)00632-7. [DOI] [PubMed] [Google Scholar]
  16. Ghosh MG, Thompson DA, Weigel RJ. PDZK1 and GREB1 are estrogen-regulated genes expressed in hormone-responsive breast cancer. Cancer Res. 2000;60:6367–6375. [PubMed] [Google Scholar]
  17. Gilbert C, Kristjuhan A, Winkler GS, Svejstrup JQ. Elongator interactions with nascent mRNA revealed by RNA immunoprecipitation. Mol Cell. 2004;14:457–464. doi: 10.1016/s1097-2765(04)00239-4. [DOI] [PubMed] [Google Scholar]
  18. Hah N, Danko CG, Core L, Waterfall JJ, Siepel A, Lis JT, Kraus WL. A rapid, extensive, and transient transcriptional response to estrogen signaling in breast cancer cells. Cell. 2011;145:622–634. doi: 10.1016/j.cell.2011.03.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hirose T, Virnicchi G, Tanigawa A, Naganuma T, Li R, Kimura H, Yokoi T, Nakagawa S, Benard M, Fox AH, et al. NEAT1 long noncoding RNA regulates transcription via protein sequestration within subnuclear bodies. Mol Biol Cell. 2014;25:169–183. doi: 10.1091/mbc.E13-09-0558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Huang S, Spector D. Intron-dependent recruitment of pre-mRNA splicing factors to sites of transcription. J Cell Biol. 1996;133:719–732. doi: 10.1083/jcb.133.4.719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hutchinson J, Ensminger A, Clemson C, Lynch C, Lawrence J, Chess A. A screen for nuclear transcripts identifies two linked noncoding RNAs associated with SC35 splicing domains. BMC Genomics. 2007;8:39. doi: 10.1186/1471-2164-8-39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Imamura K, Imamachi N, Akizuki G, Kumakura M, Kawaguchi A, Nagata K, Kato A, Kawaguchi Y, Sato H, Yoneda M, et al. Long noncoding RNA NEAT1-dependent SFPQ relocation from promoter region to paraspeckle mediates IL8 expression upon immune stimuli. Mol Cell. 2014;53:393–406. doi: 10.1016/j.molcel.2014.01.009. [DOI] [PubMed] [Google Scholar]
  23. Ji P, Diederichs S, Wang W, Boing S, Metzger R, Schneider PM, Tidow N, Brandt B, Buerger H, Bulk E, et al. MALAT-1, a novel noncoding RNA, and thymosin beta4 predict metastasis and survival in early-stage non-small cell lung cancer. Oncogene. 2003;22:8031–8041. doi: 10.1038/sj.onc.1206928. [DOI] [PubMed] [Google Scholar]
  24. Jin F, Li Y, Dixon JR, Selvaraj S, Ye Z, Lee AY, Yen CA, Schmitt AD, Espinoza CA, Ren B. A high-resolution map of the three-dimensional chromatin interactome in human cells. Nature. 2013;503:290–294. doi: 10.1038/nature12644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kamentsky L, Jones T, Fraser A, Bray MA, Logan D, Madden K, Ljosa V, Rueden C, Eliceiri K, Carpenter A. Improved structure, function and compatibility for CellProfiler: modular high-throughput image analysis software. Bioinformatics. 2011;27:1179–1180. doi: 10.1093/bioinformatics/btr095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kharchenko P, Tolstorukov M, Park P. Design and analysis of ChIP-seq experiments for DNA-binding proteins. Nat Biotech. 2008;26:1351–1359. doi: 10.1038/nbt.1508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lin CY, Strom A, Vega VB, Kong SL, Yeo AL, Thomsen JS, Chan WC, Doray B, Bangarusamy DK, Ramasamy A, et al. Discovery of estrogen receptor alpha target genes and response elements in breast tumor cells. Genome Biol. 2004;5:R66. doi: 10.1186/gb-2004-5-9-r66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Mao Y, Zhang B, Spector D. Biogenesis and function of nuclear bodies. Trends Genetics. 2011;27:295–306. doi: 10.1016/j.tig.2011.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Mattick JS. A new paradigm for developmental biology. J Exp Biol. 2007;210:1526–1547. doi: 10.1242/jeb.005017. [DOI] [PubMed] [Google Scholar]
  30. Mercer TR, Dinger ME, Mattick JS. Long non-coding RNAs: insights into functions. Nat Rev Genet. 2009;10:155–159. doi: 10.1038/nrg2521. [DOI] [PubMed] [Google Scholar]
  31. Misteli T, Cáceres J, Spector D. The dynamics of a pre-mRNA splicing factor in living cells. Nature. 1997;387:523–527. doi: 10.1038/387523a0. [DOI] [PubMed] [Google Scholar]
  32. Naganuma T, Nakagawa S, Tanigawa A, Sasaki Y, Goshima N, Hirose T. Alternative 3′-end processing of long noncoding RNA initiates construction of nuclear paraspeckles. EMBO J. 2012;31:4020–4034. doi: 10.1038/emboj.2012.251. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Nakagawa S, Ip J, Shioi G, Tripathi V, Zong X, Hirose T, Prasanth K. Malat1 is not an essential component of nuclear speckles in mice. RNA. 2012;18:1487–1499. doi: 10.1261/rna.033217.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Nakagawa S, Naganuma T, Shioi G, Hirose T. Paraspeckles are subpopulation-specific nuclear bodies that are not essential in mice. J Cell Biol. 2011;193:31–39. doi: 10.1083/jcb.201011110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Sabin LR, Delas MJ, Hannon GJ. Dogma derailed: the many influences of RNA on the genome. Mol Cell. 2013;49:783–794. doi: 10.1016/j.molcel.2013.02.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Saitoh N, Spahr C, Patterson S, Bubulya P, Neuwald A, Spector D. Proteomic analysis of interchromatin granule clusters. Mol Biol Cell. 2004;15:3876–3890. doi: 10.1091/mbc.E04-03-0253. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sasaki YT, Ideue T, Sano M, Mituyama T, Hirose T. MENepsilon/beta noncoding RNAs are essential for structural integrity of nuclear paraspeckles. Proc Nat Acad Sci USA. 2009;106:2525–2530. doi: 10.1073/pnas.0807899106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sauvé F, McBroom L, Gallant J, Moraitis A, Labrie F, Giguère V. CIA, a novel estrogen receptor coactivator with a bifunctional nuclear receptor interacting determinant. Mol Cell Biol. 2001;21:343–353. doi: 10.1128/MCB.21.1.343-353.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sergeant K, Bourgeois C, Dalgliesh C, Venables J, Stevenin J, Elliott D. Alternative RNA splicing complexes containing the scaffold attachment factor SAFB2. J Cell Sci. 2007;120:309–319. doi: 10.1242/jcs.03344. [DOI] [PubMed] [Google Scholar]
  40. Simon M, Wang C, Kharchenko P, West J, Chapman B, Alekseyenko A, Borowsky M, Kuroda M, Kingston R. The genomic binding sites of a noncoding RNA. Proc Nat Acad Sci USA. 2011;108:20497–20502. doi: 10.1073/pnas.1113536108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Simon MD, Pinter SF, Fang R, Sarma K, Rutenberg-Schoenberg M, Bowman SK, Kesner BA, Maier VK, Kingston RE, Lee JT. High-resolution Xist binding maps reveal two-step spreading during X-chromosome inactivation. Nature. 2013;504:465–469. doi: 10.1038/nature12719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Smith K, Moen P, Wydner K, Coleman J, Lawrence J. Processing of endogenous pre-mRNAs in association with SC-35 domains is gene specific. J Cell Biol. 1999;144:617–629. doi: 10.1083/jcb.144.4.617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Spector DL, Lamond AI. Nuclear speckles. Cold Spring Harb Perspect Biol. 2011;3 doi: 10.1101/cshperspect.a000646. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Srisawat C, Engelke DR. Streptavidin aptamers: affinity tags for the study of RNAs and ribonucleoproteins. RNA. 2001;7:632–641. doi: 10.1017/s135583820100245x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Sunwoo H, Dinger ME, Wilusz JE, Amaral PP, Mattick JS, Spector DL. MEN epsilon/beta nuclear-retained non-coding RNAs are up-regulated upon muscle differentiation and are essential components of paraspeckles. Genome Res. 2009;19:347–359. doi: 10.1101/gr.087775.108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Tollervey J, Curk T, Rogelj B, Briese M, Cereda M, Kayikci M, König J, Hortobágyi T, Nishimura A, Zupunski V, et al. Characterizing the RNA targets and position-dependent splicing regulation by TDP-43. Nat Neuro. 2011;14:452–458. doi: 10.1038/nn.2778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Tsai MC, Manor O, Wan Y, Mosammaparast N, Wang J, Lan F, Shi Y, Segal E, Chang H. Long noncoding RNA as modular scaffold of histone modification complexes. Science. 2010;329:689–693. doi: 10.1126/science.1192002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Ule J, Jensen KB, Ruggiu M, Mele A, Ule A, Darnell RB. CLIP identifies Nova-regulated RNA networks in the brain. Science. 2003;302:1212–1215. doi: 10.1126/science.1090095. [DOI] [PubMed] [Google Scholar]
  49. Ulitsky I, Shkumatava A, Jan CH, Sive H, Bartel DP. Conserved function of lincRNAs in vertebrate embryonic development despite rapid sequence evolution. Cell. 2011;147:1537–1550. doi: 10.1016/j.cell.2011.11.055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Wan Y, Kertesz M, Spitale RC, Segal E, Chang HY. Understanding the transcriptome through RNA structure. Nat Rev Genet. 2011;12:641–655. doi: 10.1038/nrg3049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Warzecha C, Sato T, Nabet B, Hogenesch J, Carstens R. ESRP1 and ESRP2 are epithelial cell-type-specific regulators of FGFR2 splicing. Mol Cell. 2009;33:591–601. doi: 10.1016/j.molcel.2009.01.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. White MK, Johnson EM, Khalili K. Multiple roles for Puralpha in cellular and viral regulation. Cell Cycle. 2009;8:414–420. doi: 10.4161/cc.8.3.7585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Yang L, Lin C, Liu W, Zhang J, Ohgi K, Grinstein J, Dorrestein P, Rosenfeld M. ncRNA- and Pc2 methylation-dependent gene relocation between nuclear structures mediates gene activation programs. Cell. 2011;147:773–788. doi: 10.1016/j.cell.2011.08.054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Zeng C, Kim E, Warren S, Berget S. Dynamic relocation of transcription and splicing factors dependent upon transcriptional activity. EMBO J. 1997;16:1401–1412. doi: 10.1093/emboj/16.6.1401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zhang B, Arun G, Mao Y, Lazar Z, Hung G, Bhattacharjee G, Xiao X, Booth C, Wu J, Zhang C, et al. The lncRNA Malat1 is dispensable for mouse development but its transcription plays a cis-regulatory role in the adult. Cell Reports. 2012;2:111–123. doi: 10.1016/j.celrep.2012.06.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhang Y, Liu T, Meyer C, Eeckhoute J, Johnson D, Bernstein B, Nusbaum C, Myers R, Brown M, Li W, et al. Model-based analysis of ChIP-Seq (MACS) Genome Biol. 2008;9 doi: 10.1186/gb-2008-9-9-r137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhao J, Sun B, Erwin J, Song JJ, Lee J. Polycomb proteins targeted by a short repeat RNA to the mouse X chromosome. Science. 2008;322:750–756. doi: 10.1126/science.1163045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Zhou Z, Sim J, Griffith J, Reed R. Purification and electron microscopic visualization of functional human spliceosomes. Proc Nat Acad Sci USA. 2002;99:12203–12207. doi: 10.1073/pnas.182427099. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS616563-supplement-01.pdf (6.8MB, pdf)
02
NIHMS616563-supplement-02.xlsx (108.3KB, xlsx)
03
NIHMS616563-supplement-03.xlsx (16KB, xlsx)
04
NIHMS616563-supplement-04.xlsx (314.6KB, xlsx)

RESOURCES