Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Feb 24.
Published in final edited form as: Mol Cell. 2012 Jan 19;45(4):459–469. doi: 10.1016/j.molcel.2011.12.022

Mediator Complex Regulates Alternative mRNA Processing via the Med23 Subunit

Yan Huang 1, Wencheng Li 2, Xiao Yao 1, Qi-jiang Lin 1, Jing-wen Yin 1, Yan Liang 1, Monika Heiner 1, Bin Tian 2, Jingyi Hui 1, Gang Wang 1,*
PMCID: PMC3288850  NIHMSID: NIHMS347083  PMID: 22264826

SUMMARY

Mediator complex is an integrative hub for transcriptional regulation. Here we show that Mediator regulates alternative mRNA processing via its Med23 subunit. Combining tandem affinity purification and mass spectrometry, we identified a number of mRNA processing factors that bind to a soluble recombinant Mediator subunit MED23 but not to several other Mediator components. One of these factors, hnRNP L, specifically interacts with MED23 in vitro and in vivo. Consistently, Mediator partially colocalizes with hnRNP L and the splicing machinery in the cell. Functionally Med23 regulates a subset of hnRNP L-targeted alternative splicing (AS) and alternative cleavage and polyadenylation (APA) events as shown by minigene reporters and exon array analysis. ChIP-seq analysis revealed that Med23 can regulate hnRNP L occupancy at their co-regulated genes. Taken together, these results demonstrate a crosstalk between Mediator and the splicing machinery, providing a molecular basis for coupling mRNA processing to transcription.

INTRODUCTION

Alternative splicing (AS) affects the majority of protein-coding genes in higher species, creating functional diversity for gene products (Sharp, 2005). Multiple cis elements and trans-acting factors contribute to the regulation of mRNA splicing (Black, 2003; Licatalosi and Darnell, 2010). Splicing regulators, such as members of the serine/arginine-rich (SR) protein family and the heterogeneous nuclear ribonucleoprotein (hnRNP) family, modulate alternative splicing events by binding to sequences near splice sites (Licatalosi and Darnell, 2010). SR and hnRNP proteins have also been implicated in the regulation of alternative cleavage and polyadenlyation (APA) (Licatalosi and Darnell, 2010), which could happen to more than half of the human genes (Tian et al., 2005). Pre-mRNA splicing, which often occurs co-transcriptionally, can be mechanistically coupled to transcription (Kornblihtt, 2006; Kornblihtt et al., 2004; Maniatis and Reed, 2002; Neugebauer, 2002). There are two models for coupling splicing to transcription, which are not mutually exclusive: The “recruitment coupling” refers to the association of splicing factors with the transcribing pol II machinery, whereas the “kinetic coupling” involves modulation of the pol II elongation rate (Munoz et al., 2010). Multiple factors have been shown to regulate the interplay between transcription and splicing, such as CTD of RNA polymerase II, transcription factors, elongation factors, and epigenetic regulators (Kornblihtt et al., 2004; Luco et al., 2010). However, the molecular mechanism of how transcription couples to splicing is still not fully understood.

Mediator is a transcriptional coactivator complex that conveys intra- or extra-cellular signals to the basal transcription machinery (Malik and Roeder, 2010; Taatjes, 2010). In addition to its role in initiation of transcription, Mediator has been implicated in stimulation of RNA synthesis at postrecruitment steps, such as isomerization of the preinitiation complex, promoter melting, promoter clearance, chromatin remodeling, and possibly, mRNA processing (Bjorklund and Gustafsson, 2005; Struhl, 2005; Wang et al., 2005).

In this study, we provide evidence that Mediator regulates alternative mRNA processing. Using affinity purification and mass spectrometry, we identified a number of pre-mRNA processing factors that specifically bind to the Mediator subunit MED23. We found that hnRNP L, one of the identified factors, interacts with MED23 in vitro and in vivo. Consistently, MED23 partially colocalizes with hnRNP L and splicing factors related to U1/U2 snRNPs. Functionally Med23 regulates a significant subset of hnRNP L-targeted AS and APA events and modulate hnRNP L occupancy at their promoters as shown by minigene reporter assays, exon array, and ChIP-seq analysis. Together, these findings establish a direct link between Mediator Complex and the splicing machinery, demonstrate an important function of Mediator in the regulation of alternative mRNA processing, and provide insights into the molecular mechanisms that couple pre-mRNA processing to transcription.

RESULTS

Characterization of soluble MED23 and its binding proteins

Mediator components generally lack predictable functional motifs and contain large intrinsically disordered regions (Taatjes, 2010; Toth-Petroczy et al., 2008). MED23, one of the components which interacts with MAPK-activated ELK1 and adenovirus E1A (Berk, 2005), is a highly hydrophobic protein and is difficult to be purified in a soluble form (Wang et al., 2001). We managed to overcome technical barriers in MED23 purification by using the Bac-to-Bac expression system and tandem affinity columns (Figure 1A). The purified soluble double-tagged human MED23 was shown to have the capability to interact with the phosphorylated but not unphosphorylated ELK1 (Wang et al., 2009), and with adenovirus E1A-CR3 but not with the nonfunctional, point mutant E1A H160Y (Figure 1B), suggesting that the recombinant MED23 likely adopts a functional conformation capable of binding to the right biological partners.

Figure 1. Identification of MED23 interacting proteins.

Figure 1

Open in a new tab

(A) Baculovirus expressed His-Flag-MED23 was purified by Ni-NTA beads, and analyzed by SDS-PAGE followed by coomassie blue staining. (B) Purified GST-E1A or its point mutant GST-H160Y was incubated with His-Flag-MED23 immobilized on the Ni-NTA beads at 4 °C overnight. After washing, the bound proteins were eluted by boiling and immunoblotted with the anti-GST (for E1A) and anti-Flag (for Med23) antibodies. Input proteins were shown by coomassie blue staining. (C) Scheme for affinity purification and MS/MS to identify MED23 binding proteins. (D) The protein profiles from the first elution (E1, after NTA affinity purification) and the second elution (E2, after M2 affinity purification) described in (C) were visualized by silver staining. (E) Gene-Ontology (GO) analysis of potential MED23-associated factors with or without RNase treatment. (F) Venn diagram of MED23-associated proteins identified by affinity columns and MS/MS (with or without RNase treatment during purification). (G) Venn diagrams of proteins identified by MS/MS following the affinity columns of MED23, in pair with that of MED24, MED16, MED15, and MED29, respectively.

In an attempt to identify novel proteins that interact with MED23, we applied HeLa nuclear extract to a MED23-column for two constitutive rounds of affinity purification (Figure 1C), using Ni-agarose and anti-Flag M2 beads, followed by shotgun mass spectrometry (MS) (Figure 1D). Gene Ontology (GO) analysis of the MED23-binding candidates showed a significant enrichment of factors involved in mRNA processing (Figure 1E, no RNase). To address whether the interactions are mediated by RNA, we repeated the purification procedure with an RNase treatment step. GO analysis of the eluted proteins revealed again that mRNA processing factors ranked at the top (Figure 1E, RNase). A total of 55 proteins were identified in both experiments (Figure 1F), including over 10 proteins related to RNA processing (Table S1). In contrast, similar purification schemes applied to four other Mediator components, including MED15, MED24, MED16, and MED29, yielded protein sets without the enrichment of RNA processing factors. These protein sets only marginally overlapped with the MED23-binding set (Figure 1G). Most of the overlapping factors were abundant cytoskeleton or chaperone proteins, suggesting that binding of the mRNA processing factors to MED23 is specific. Western blotting further confirmed the association of the RNA processing factors and other identified proteins with MED23, not with MED29 (Figures S1A and S1B).

MED23 interacts with hnRNP L and other splicing factors in vivo and in vitro

Among the mRNA processing factors binding to MED23, hnRNP L was top ranked in the two independent purification procedures with or without RNase treatment. HnRNP L is a member of the RNA-binding protein family with the RNA recognition motif (RRM) and has been implicated in both AS (Hui et al., 2003; Hung et al., 2008; Rothrock et al., 2005) and APA (Guang et al., 2005; Hung et al., 2008). We performed a co-immunoprecipitation experiment by co-transfecting tagged Med23 and hnRNP L into 293T cells. HnRNP L-like protein (hnRNP LL), a homolog of hnRNP L, was used for comparison. Interestingly, MED23 interacted with hnRNP L, but not much with hnRNP LL in the co-immunoprecipitation experiment (Figure 2A). Consistently, GST-hnRNP L, but not GST-hnRNP LL, bound to the recombinant MED23 in vitro (Figure 2B). Furthermore, endogenous MED23 and hnRNP L were able to reciprocally co-immunoprecipitate with each other from HeLa nuclear extract (Figure 2C). We also analyzed the presence of hnRNP L in the purified Mediator fractions. Mediator fractions prepared through 4 columns (MED, Figure 2D) (Wang and Berk, 2002) or 2 columns (L-MED) (Wang et al., 2001) from HeLa nuclear extracts were immunoblotted for splicing factors. Only sub-stoichiometic amount of hnRNP L and SF3B1 can be detected within both Mediator preparations (Figure 2D), which is likely due to that the binding of hnRNP L and other splicing factors with the Mediator Complex could be dissociated during the multi-column purification procedure. This observation may also explain why the presence of hnRNP L within the purified Mediator fractions has not been shown so far. However, antibody to another Mediator subunit CDK8 can readily pull down Mediator components as well as SF3B1 and hnRNP L from the nuclear extracts of WT ES cells, but not of Med23−/− ES cells (Figure 2E), supporting that the association of hnRNP L with the Mediator Complex is Med23-dependent. Interestingly, hnRNP LL was detected in the anti-CDK8 precipitates from both the WT and Med23−/− ES nuclear extracts, suggesting its possible binding to Mediator in a Med23-independent manner (i.e., via another Mediator subunit). A domain mapping experiment with multiple deletions of Med23 and hnRNP L revealed that the N-terminal region of MED23 and the second classic RNA-recognition motif of hnRNP L (RRM2) are required for the Med23-hnRNP L interaction, and hnRNP L RRM1 seems to be also important for this interaction, but to a less extent (Figures 2F and 2G). Highly homologous RRM2 of hnRNP L and hnRNP LL were fused to Myc-tag and cotransfected each with Flag-Med23 respectively. MED23 bound to the hnRNP L RRM2 more strongly than to the hnRNP LL RRM2 (Figure S2A), explaining the MED23 binding preference to hnRNP L over to hnRNP LL. Because the N-terminal deletion mutants of MED23 could not be expressed, we were not able to further pinpoint the interaction domain. Additional co-immunoprecipitation experiments also verified that MED23 can interact with other splicing factors, such as DDX17 (an RNA helicase, also known as P72) (Figure S2B) and hnRNP H1 (Figure S2C), indicating multiple connections between Mediator and splicing factors in addition to hnRNP L.

Figure 2. HnRNP L specifically interacts with MED23 in vivo and in vitro.

Figure 2

Open in a new tab

(A) Co-IP of hnRNP L with MED23. Flag-Med23 expressing plasmid was cotransfected with Myc-hnRNP L or Myc-hnRNP LL into 293T cells. Whole cell extract was used for immunoprecipitation with the anti-Flag antibody, followed by western blotting using indicated antibodies. (B) Soluble His-Flag-MED23 was incubated at 4 °C overnight with immobilized GST-hnRNP L or GST-hnRNP LL. After washing, the bound proteins were eluted by boiling and immunoblotted with the indicated antibodies. (C) Reciprocal co-IP experiments with antibodies against endogenous hnRNP L and MED23 in HeLa nuclear extracts. (D) Mediator complex purified from HeLa nuclear extract was subjected to SDS-PAGE, then analyzed by silver staining and western blotting. MED: Mediator fractions prepared through 4 columns; L-MED: Mediator fractions prepared through 2 columns (see Supplemental Methods for details). (E) Nuclear extracts prepared from Med23+/+ (WT) or Med23−/− (KO) mouse embryonic stem cells were subjected to co-IP using CDK8 antibody immobilized on protein G beads. The immunoprecipitated proteins were detected with indicated antibodies by western blotting. (F) Deletion mutants of Flag-Med23 were co-immunoprecipitated with GST-hnRNP L using anti-Flag antibody. Asterisk indicates the IgG heavy chain. (G) Myc-tagged hnRNP L deletion mutants were co-immunoprecipitated with Flag-Med23 using the anti-Flag antibody. Asterisks indicate the IgG heavy and light chains.

To further confirm the association of Mediator with the splicing machinery, we immunoprecipitated MED23 from HeLa nuclear extract and then analyzed the spliceosomal U1, U2, U4, U5, and U6 snRNAs in the precipitates by northern blot. Under a stringent condition, the antibody against MED23 pulled down U1 and U2 snRNAs, but not U4, U5, or U6 snRNAs. In the precipitates, other Mediator components such as MED16 and MED12 were also present, suggesting that Mediator likely functions as a whole complex to interact with U1 and U2 snRNAs (Figure 3A). Anti-CDK8 antibody can also pull down more U1 and U2 snRNAs from the HeLa nuclear extracts compared to the IgG precipitates (Figure S3A), and more U1 and U2 snRNAs from the nuclear extracts prepared from WT ES cells compared to that from the Med23−/− ES cells (Figure S3B). Consistent with this result, co-immunostaining revealed that MED23 partially colocalizes with hnRNP L (Figure 3B), as well as U1-70K and U2AF65, which are a component of the U1 snRNP and an auxiliary factor for U2 snRNP, respectively (Figures 3C and 3D). Collectively, these results suggest that Mediator may associate with the splicing machinery via the MED23 subunit, and Mediator complex could possibly play a role at an early step of spliceosome assembly, in which hnRNP L and U1/U2 are involved (Motta-Mena et al., 2010).

Figure 3. Association of MED23 with the splicing machinery.

Figure 3

Open in a new tab

(A) HeLa nuclear extract was immunoprecipitated with anti-MED23 or control IgG. Total RNAs were extracted and separated on a denaturing polyacrylamide gel containing 8 M urea and detected by northern blot analysis with individual U snRNA probes. The immunoprecipitated proteins were immunoblotted with the indicated antibodies. (B) HeLa cells were immunofluorescence-stained with antibodies against MED23 and hnRNP L. The scatter plot of FITC and RRX emission intensities are plotted on the X- and Y-axes, respectively, using Volocity Softwere. Coefficient for co-localization (Pearson correlation coefficient, r) was 0.44±0.17 (see Supplemental Methods for details). (C) HeLa cells were immunofluorescence-stained with antibodies against MED23 and U1-70K. Co-localization was analyzed as in (B). r = 0.59±0.06. (D) HeLa cells were immunofluorescence-stained with antibodies against MED23 and U2AF65. Co-localization was analyzed as in (B). r = 0.59±0.13.

Med23 regulates alternative splicing of some hnRNP L targets

The association of MED23 with splicing machinery and splicing factors suggests a regulatory role in AS. Previously, inclusion of the SLC2A2 exon 4 was shown to be repressed by hnRNP L via its binding to a CA-repeat intronic splicing silencer (Hui et al., 2005). We performed an in vivo splicing assay using the CMV-driven SLC2A2 minigene, which harbors either an intronic hnRNP L-targeted CA repeat sequence (wt) or a substitution mutant sequence (sub). As previously described, substitution of the CA repeat region with a non-related sequence activated exon 4 inclusion. Compared to its splicing pattern in Med23+/+ MEF cells, wt minigene resulted in an increased splicing of exon 4 in Med23−/− MEF cells. However, sub minigene had a comparable level of exon 4 inclusion in Med23+/+ and Med23−/− cells (Figure 4A), suggesting Med23 and hnRNP L have similar roles in regulating splicing of the SLC2A2 exon 4. To further examine whether Med23 and hnRNP L can function in the same “pathway”, we performed a double depletion experiment. Knocking down hnRNP L in Med23−/− cells did not further increase the inclusion rate of the SLC2A2 exon 4 compared to the effect of hnRNP L RNAi in wildtype cells (Figure S4A), suggesting that hnRNP L and Med23 may act in the same pathway in regulating this AS event. We also checked if Med23 regulates AS of SLC2A2 minigene in HeLa cells with viral-mediated Med23 knockdown. Again, Med23 RNAi led to the increased inclusion of the SLC2A2 exon 4 (Figure 4B). Interestingly, increasing Med23-overexpression also gradually reduced skipped exon 4 of SLC2A2 (Figure 4C). This is possibly due to the squelching effect that abundant MED23 monomer protein could inhibit the interaction between hnRNP L and Mediator, therefore resulting in a similar effect on AS to Med23-deficiency. Pre-mRNA splicing of eNOS was previously showed to be stimulated by hnRNP L (Hui et al., 2003). We found here that its splicing efficiency was greatly reduced in the Med23−/− cells (Figure 4D) and Med23 RNAi HeLa cells (Figure S4B). However, another hnRNP L-regulated AS event of RFXANK is Med23-independent (Figure 4E), suggesting that Med23-mediated splicing regulation can be gene-specific.

Figure 4. Med23 regulates alternative splicing.

Figure 4

Open in a new tab

(A) Schematic representation of SLC2A2 minigene driven by CMV promoter. The alternatively spliced exon 4 is indicated by a grey box. Wild type (wt) and substituted (sub) sequence elements are shown. Vertical lines within the introns mark the positions where intron sequences were deleted. Minigenes were transfected into the Med23+/+ (wild type) or Med23−/− MEF cells, and splicing pattern was examined by RT–PCR. Control transfection (mock) was carried out in the absence of DNA. Average inclusion rates and standard deviations are from three separate experiments. (B) CMV promoter-driven SLC2A2 minigenes were transfected into the si-Ctrl or si-Med23 HeLa cells, and splicing pattern was examined by RT–PCR. Average inclusion rates and standard deviations are from three separate experiments. (C) CMV promoter-driven SLC2A2 minigene and increasing dosage of Flag-Med23 were transfected into HeLa cells, and splicing pattern was examined by RT–PCR. The expressing of Flag-Med23 was detected by western blotting using Flag antibody. (D) Schematic representation of CMV-promoter eNOS construct carrying 32 CA repeats in the shortened intron 13. The eNOS construct was transiently transfected into Med23+/+ (wild type) or Med23−/− MEF cells, and splicing activity was determined by RT-PCR with exon 13/14-specific primers. The averages of splicing efficiencies with standard deviations are shown below the gel (n = 3). (E) Schematic of RFXANK minigene with the CMV promoter is represented. Wild type (wt) and substituted (sub) sequence elements are shown. Minigenes were transfected into the Med23+/+ (wild type) or Med23−/− MEF cells, and splicing pattern was examined by RT–PCR. Average inclusion rates and standard deviations are from three separate experiments. (F) Schematic of SLC2A2 minigene with the Egr1 promoter is represented. The Egr1 promoter-driven minigene was transfected into Med23+/+ (wild type) MEF, Med23−/− MEF, or Med23−/− MEF with re-expressing Med23. Splicing pattern was examined by RT–PCR. Average inclusion rates and standard deviations are from three separate experiments.

We next examined whether the regulation of AS by Med23 is promoter-dependent. To this end, we fused the SLC2A2 minigene to the well-characterized Med23-controlled Egr1 promoter (Stevens et al., 2002; Wang et al., 2005). Consistent with previous studies, Egr1-driven expression of the reporter is decreased in Med23−/− cells compared to wild-type cells. Importantly, the exon 4 inclusion rate increased significantly in Med23−/− cells (Figure 4F), and this stimulation effect was partially reduced by re-expressing Med23 in Med23−/− cells. To further confirm if Med23 is important for recruiting hnRNP L to target genes, we compared the binding of hnRNP L to the Egr1 locus between Med23+/+ and Med23−/− cells by Chromatin Immumoprecipitation (ChIP) assays, and found that the association of hnRNP L is much stronger in Med23+/+ cells than in Med23−/− cells, at the Egr1 locus (Figure S4C). Collectively, these data suggest that Med23 may be important for both the hnRNP L recruitment and its function for some AS events.

Genome-wide identification of AS and APA events commonly targeted by Med23 and hnRNP L

To examine the effect of MED23-hnRNP L interaction on a genomic scale, we next performed an exon array experiment using HeLa cells expressing Med23 or hnRNP L siRNAs which were established by a virus-mediated siRNA technology (Figure 5A, left panel). Total RNA of these cell lines was processed and hybridized to the Affymetrix human exon array (see Supplemental Experimental Procedures). Notablly, si-Med23 did not change expression of hnRNP L and vice versa (Figure 5A, right panel).

Figure 5. Genome-wide identification of AS and APA events commonly targeted by Med23 and hnRNP L.

Figure 5

Open in a new tab

(A) Knock-down of Med23 or hnRNP L in HeLa cells by retrovirus-mediated siRNA. Western blot was performed to detect the knock-down efficiency (left). Log2 (ratio) in the exon array data was shown, indicating the mRNA expression level of Med23 and hnRNP L in knockdown samples (right). Standard deviations are from three separate experiments. (B) Scatter plot showing skipped exons in si-Med23 or si-hnRNP L cells as compared to si-Ctrl cells. Y and X axes are splicing index (SI, see Supplemental Experimental Procedures for detail). AS events regulated (P<0.05, T-test; and log2 (fold change) >0.5) in si-Med23 only are in black, those regulated in both si-Med23 and si-hnRNP L are in red, and other skipped exons are shown in grey. Pearson correlation coefficient (r) and P-value for the events shown in black and red are indicated. (C) The numbers of AS events regulated in the same or opposite directions between si-Med23 and si-hnRNP L. Ex, exclusion; in, inclusion. (D) As in (B) except that APA events are shown and Y and X axes are RUD values (see Supplemental Experimental Procedures for detail). (E) As in (C), except that APA events were analyzed. Le: 3′UTR lengthened (proximal polyA site less used); sh: 3′UTR shortened (proximal polyA site more used). (F) Occupancy of hnRNP L on the set of genes whose AS events are co-regulated by Med23 and hnRNP L in si-Ctrl (red) and si-Med23 (blue) Hela cells, determined by ChIP-seq analysis. All these genes were normalized to 3 kb for meta-gene, with 1 kb extended upstream from TSS, 1 kb downstream from TTS for analyzing average intensity in 50 bp bins. (G) Quantitation of hnRNP L occupancy on promoter (1 kb upstream from TSS) and gene body region (from TSS to TTS). hnRNP L average binding intensity in si-Med23 (blue) was normalized to si-Ctrl (red).

We first analyzed the exon array data for AS and APA events, and found that Med23 RNAi resulted in regulation of 1459 AS events and 279 APA events (P<0.05, T-test) (Figures S5A, S5B and Table S2). For the events highly regulated upon Med23 RNAi (P<0.05 and fold change >1.44), correlation coefficients of 0.58 and 0.66 were obtained for AS (skipped exons only) and APA, respectively, when Med23 RNAi was compared with hnRNP L RNAi (Figures 5B and 5D). Additionally, 199 AS and 35 APA events were commonly regulated by Med23 RNAi and hnRNP L RNAi (Figures S5A, S5B, and Table S2). Notably, the events consistently regulated significantly outnumbered those oppositely regulated (Figures 5C and 5E), suggesting these two proteins regulate common subsets of exons and polyadenylation (polyA) sites. The overlap between Med23-dependent and hnRNP L-dependent AS events was also statistically significant for other AS types, including alternative 5′ splice site usage (A5SS), alternative 3′ splice site usage (A3SS), intron retention (RI), and multiple cassette exon regulation (MCE) (Table S2). AS and APA co-dependence on these two factors was further validated by RT-PCR or qPCR assays (Figures S5C and S5D). These results indicate that Med23 and hnRNP L can globally target partially overlapping sets of AS and APA events.

To further address the question of whether Med23 affects the recruitment of hnRNP L, we performed a ChIP-seq experiment to compare the hnRNP L binding profile in si-Ctrl and si-Med23 HeLa cells. Interestingly, for the set of genes whose AS events are co-regulated by hnRNP L and Med23 as identified by exon array, Med23-depletion attenuated the hnRNP L occupancy at the promoter region, whereas it did not alter the overall hnRNP L binding intensity in the gene body (Figures 5F and 5G). Therefore, on the genomic scale, Med23 affects at least a subset of hnRNP L-regulated alternative mRNA processing events, likely by modulating hnRNP L occupancy at the promoter region.

Transcriptional control by Med23 and hnRNP L

Since the transcription and splicing are tightly-coupled processes, we also examined if hnRNP L may also regulate transcription in a Med23-related manner. Gene expression profiling using the array data revealed that Med23 and hnRNP L regulated a largely overlapping set of target genes as indicated by the heatmap (Figure 6A) and Venn diagram (Figure S6A). Scatter plot analysis indicated a positive correlation of gene expression changes between si-Med23 and si-hnRNP L treatments, suggesting a functional commonality in transcriptional control between these two proteins (Figure 6B). Some of the co-regulated genes were validated by real-time quantitative PCR (qPCR) (Figure S6B). Notably, the Med23-controlled Egr1 had attenuated expression as a result of hnRNP L depletion (Figure S6B), and knockdown of hnRNP L or Med23 in mouse embryonic stem cells (ESCs) both led to reduced Egr1 expression and decreased Pol II binding (Figures 6C, 6D, 6E, 6F, 6G, and 6H), suggesting that hnRNP L can also regulate Med23-controlled transcription.

Figure 6. Transcriptional control by Med23 and hnRNP L.

Figure 6

Open in a new tab

(A) Heatmap showing gene expression changes in si-Med23 or si-hnRNP L cells. Log2(ratio) for si-Med23 or si-hnRNP L vs. si-Ctrl is represented by color using the scheme shown at the bottom. Only genes regulated in si-Med23 and/or si-hnRNP L are shown. (B) Scatter plot showing regulation of gene expression in si-Med23 or si-hnRNP L cells as compared to si-Ctrl cells. Y and X axes are log2(ratio). Genes regulated (P<0.05, T-test; and fold change > 1.3) in si-Med23 only are in black, those regulated in both si-Med23 and si-hnRNP L are in red, and other genes are in grey. Pearson correlation coefficient (r) and its P-value for genes shown in black and red are indicated. Experimentally validated genes (FigureS6B) are circled. (C) Western blot detection of RNAi efficiency in Med23 or hnRNP L knock-down ES cells. (D) Scheme shows primers used for real time PCR of Egr1 gene locus. (E–J) Egr1 mRNAs were detected by real time PCR in si-L (E) or si-23 (F) HeLa cells. Standard deviations are from three separate experiments. Serum stimulated ES cells were subjected to ChIP procedure using anti-pol II antibody (G and H) or anti-H3K36me3 antibody (I and J) in si-L or si-23 HeLa cells. (K) Model for the role of Mediator in coupling transcription and alternative splicing. The transcriptional Mediator cofactor complex communicates with the splicing machinery via interaction with splicing regulator.

A previous study found that hnRNP L is a regulator for KMT3a/Set2, a histone methyltransferase required for H3K36me3 in vivo (Yuan et al., 2009). We therefore performed the ChIP assay for H3K36me3, and observed that knocking down either hnRNP L or Med23 decreased the levels of H3K36me3 at the Egr1 coding region (Figures 6I and 6J). We have also performed a ChIP-seq experiment in control and Med23 knockdown HeLa cells using an antibody against H3K36 trimethylation. For the set of genes whose AS events are co-regulated by Med23 and hnRNP L as identified by exon array, Med23 knockdown resulted in decreasing enrichment level of H3K36 trimethylation over the gene coding region (Figure S6C), suggesting a connection between Mediator and epigenetic regulation. Considering recent reports on epigenetic connection with Mediator Complex (Ding et al., 2008) and alternative splicing (Luco et al., 2010), it is possible that Mediator Complex may function to integrate the processes of transcription, splicing, and epigenetic dynamics, at least in part through interaction with hnRNP L.

DISCUSSION

Extensive studies have established the functional versatility of the Mediator complex in sensing, integrating, and processing diverse signaling pathways by physically interacting with an array of transcription factors (Malik and Roeder, 2010). In addition, Mediator has been shown to physically or functionally interact with many preinitiation complex components, including Pol II, TFIIB, TFIID, TFIIE, and TFIIH (Taatjes, 2010). Recent studies have demonstrated an interaction between Mediator and the cohesin complex to loop the enhancer and promoter regions for gene activation (Kagey et al., 2010). Mediator has also been shown to interact with several cofactors or their complexes (Ebmeier and Taatjes, 2010; Johnson et al., 2002; Khorosjutina et al., 2010; Wallberg et al., 2003); interestingly, most of them have been shown to play roles in mRNA processing, such as Brm (Batsche et al., 2006), CHD (Sims et al., 2007), SRC (Auboeuf et al., 2002), PGC-1 (Monsalve et al., 2000), and TFIID (Dantonel et al., 1997). Here we identified interactions between Mediator and mRNA processing factors, which links the Mediator complex with the splicing machinery (Figure 6K). Therefore, in addition to its recruitment function for the preinitiation complex formation, the Mediator complex also plays a role in alternative mRNA processing.

Our data appear to be consistent with the “recruitment” model involved in the coupling of splicing to transcription (Munoz et al., 2010), in the sense that Mediator may help to recruit splicing regulators at the promoter, or interact with splicing regulators associated with the nascent mRNA. However, it is also possible that Mediator modulates some splicing events by altering the kinetics of transcription. In fact, two recent reports revealed a role of Mediator in regulation of Pol II elongation (Donner et al., 2010; Takahashi et al., 2011).

We showed that MED23 interacts with hnRNP L in vitro and in vivo, providing a link between Mediator and the splicing machinery. In depth analyses using mini-gene reporters and genome-wide exon array data indicate hnRNP L and Med23 have significant functional overlaps. Remarkably, Med23-MO (morpholino antisense oligonucleotide) or hnRNP L-MO injected zebrafish embryos displayed similar severe abnormalities in development, such as shortened body, delayed yolk sac absorption, curved tail, and slower heart beat (Figures S6D and S6E). Therefore, multiple genetic and biochemical lines of evidence suggest that hnRNP L and Med23 are intimately related.

On the other hand, it is also noticeable that Med23 regulates some, but not all hnRNP L targets. This could be due to the complexity of hnRNP L’s mode of action in mRNA processing which could be different when it is loaded to the genes via Mediator vs. recruited co-transcriptionally. Its context-dependent regulation has been noted before (Motta-Mena et al., 2010). In addition, our MS/MS data indicate that Mediator interacts with other mRNA processing factors, such as hnRNP H1, DDX17, and polyadenylation factor CPSF7 (Figure S2 and Table S1), which may explain why Med23 affects many hnRNP L-independent AS and APA events. In this vein, it would be interesting to address what determines a Med23 target in the future.

In summary, Mediator can act as a key master coordinator, orchestrating diverse factors and machineries at multiple stages in the regulation of gene expression and production of transcript isoforms. Considering the Mediator Complex contains multiple subunits and has a malleable structure with a huge surface, it is conceivable that Mediator functions as a platform for diverse dynamic protein interactions involved in multiple stages of the transcriptional process. Temporal, spatial, and gene-specific features of these dynamic interactions involving the Mediator Complex remain challenging for future investigations.

EXPERIMENTAL PROCEDURES

Cell Culture

Med23+/+ and Med23−/− MEF cells were isolated from 9.5-day embryos and self-immortalized using standard procedures (Balamotis et al., 2009; Wang et al., 2009). Immortalized MEF cells, 293T, and HeLa cells were maintained in Dulbecco’s modified Eagle’s medium containing 10% FBS (Hyclone). ES cells were cultured in ES medium containing leukemia inhibitory factor (LIF). All the cells were maintained inside the incubator with 5% CO2.

Retrovirus Infection

Establishing stable cell lines to knock-down or overexpress a gene of interest were based on the manufacturer’s recommendation (Clontech). More details were described previously (Wang et al., 2009). The targeted sequence was determined using the Whitehead Institute siRNA designing tool and verified by BLAST searches to ensure specificity. See Supplemental Experimental Procedures for oligonucleotide sequences. Retroviruses were generated following the cotransfection of recombinant pSiren-RetroQ plasmids with pCL10A1 helper plasmid into 293T cells using Lipofectamine 2000 (Invitrogen). Tissue culture supernatants containing retroviruses were harvested 48 hr later and passed though a 0.45 μm filter. Cells were plated into 6-well plates before retroviral infection. Virus-containing supernatants were supplemented with 20 μg/ml polybrene and added to the cells for a centrifugation at 2500 rpm at 30 °C for 1.5 h. 24 hour after spin infection, MEFs were selected with 50 μg/ml puromycin (Sigma-Aldrich); HeLa and HepG2 cells were selected with 2 μg/ml puromycin.

Soluble Recombinant MED23 Purification, MS/MS and Gene-Ontology Analysis

See Supplemental Experimental Procedures.

In Vitro Binding and Co-immunoprecipitation (Co-IP) Assays

See Supplemental Experimental Procedures.

Purification of Mediator Complex

Mediator fractions purified from two columns and four columns were described previously (Wang and Berk, 2002; Wang et al., 2001). See Supplemental Experimental Procedures for more detail.

Northern Blot

RNA was separated by 8% Urea denaturing gel, transferred onto nylon membrane (Hybond N, GE Healthcare) and hybridized to U1, U2, U4, U5 and U6 digoxigenin-labeled DNA probes (prepared by PCR DIG-labeling Mix kit from Roche) and then detected by CSPD (Roche). See Supplemental Experimental Procedures for probe sequences.

Immunofluorescence

Cells were washed by PBS twice, fixed in 4% paraformaldehyde, permeabilized with 0.2% Triton X-100 and incubated for 1 hour in blocking buffer (2% BSA in PBS). Primary antibodies were diluted in blocking buffer and applied overnight at 4 °C. After three washes in PBS, secondary antibodies (form Jackson Laboratory) were diluted in blocking buffer and applied for 1 hour at room temperature. After three washes by PBS, nuclei were stained with DAPI (1:4000) for 4 min. Antibodies used for immunostaining were as follows: MED23 (BD Biosciences, mouse), hnRNP L (Santa Cruz, rabbit), SNRNP70 (Abcam, rabbit), U2AF65 (Abcam, rabbit). Individual images were analyzed for co-localization using Volocity Software (PerkinElmer, Waltham, MA, USA). See Supplemental Experimental Procedures for statistics analysis of the co-localization data.

Real-Time PCR (qRT-PCR) and RT-PCR

Total RNA was isolated from cells using TRIZOL (Invitrogen). The first-strand cDNA was generated using MMLV transcriptase (Promega), and real-time PCR was performed in triplicate in an Eppendorf Mastercycler. Endogenous mRNA values were normalized to the level of GAPDH mRNA. For endogenous APA validation, qRT-PCR were performed using two pairs of primers F2R2 and F1R1 (See Figure S5D), and the relative signal for primer set F2R2/F1R1 ratios was calculated.

In Vivo Splicing Assay

For in vivo splicing assay, MEF or HeLa cells (in 6 well plate, 90% confluency) were transfected with 1 μg minigene with Lipofectamine 2000. After 24 hour, total RNA was harvested using TRIZOL. 4 μg RNA was treated by RNase-free DNase (Roche) 30 min in 37 °C, purified by Phenol-Chloroform extraction, dissolved with DEPC water, and subjected to reverse transcription using random primer or vector specific primer BGH-rev, and performed the PCR with specific primer sets. Alternative splicing products were analyzed by 2% agarose gel and quantitated by software Quantity One (Bio-Rad). For eNOS CA32 splicing assay, 2 μg eNOS plasmid was transfected using standard calcium phosphate method. After 6 hours, splicing of eNOS mRNA was analyzed by RT-PCR. See Supplemental Experimental Procedures for primer sequences.

Chromatin Immumoprecipitation (ChIP), ChIP-seq and Data Analysis

ChIP assays were performed as described previously (Wang et al., 2005); however, the immunoprecipitated DNA was quantified using real-time PCR. All values were normalized to the input. See Supplemental Experimental Procedures for primer sequences. For ChIP-Seq, the ChIP’ ed DNA concentration was measured by Qubit Fluorometer (Invitrogen). DNA was purified using ChIP-seq Sample Prep Kit (Illumina) and subjected to 76 bases of sequencing on Genome analyzer IIx (Illumina). See Supplemental Experimental Procedures for more detail.

Affymetrix Exon Array and Data Analysis

HeLa cells were expressed with three different Med23 or hnRNP L siRNAs which were established by a virus-mediated siRNA technology. RNAs were extracted by TRIZOL, purified by QIAGEN RNAeasy kit and subjected to Affymetrix GeneChip Human Exon 1.0 ST Array. Raw data were normalized by the RMA method in the Affymetrix Power Tools (APT) program and probe sets with detection above background (DABG) P value < 0.05 in at least one sample group were used for further analysis. See Supplemental Experimental Procedures for more detail.

Supplementary Material

01

  1. Mediator MED23 specifically binds to a number of mRNA processing factors

  2. MED23 specifically interacts with hnRNP L in vitro and in vivo

  3. MED23 controls a subset of hnRNP L-targeted alternative mRNA processing events

  4. MED23 regulates hnRNP L occupancy at their co-regulated genes

Acknowledgments

We thank Guanzhen Yang and Dongming Hou for technical assistance, Dr. Arnie Berk for training and reagents, Drs. Jun Yan, Xin Fu, and Yuting Liu for advice on statistical analysis, Dr. Rong Zeng at SIBS Proteomics Core Facility for MS/MS support, and Drs. Dangsheng Li, Hong Cheng, and Michael Mathews (UMDNJ) for helpful suggestions. This work was supported in part by grants from China MOST (2009CB941100 & 2011CB510104), CAS (XDA01010401), and CNSF (30770452 & 81030047) to G.W., and NIH (R01GM084089) to B.T. G.W. and J.H. are scholars of the “Hundred Talent Program”.

Footnotes

ACCESSION NUMBERS

Exon array and ChIP-Seq data in this study are available for download from Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo). GEO accession ID for the exon array data is GSE33771, and for the ChIP-seq data is GSE33887.

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. Auboeuf D, Honig A, Berget SM, O’Malley BW. Coordinate regulation of transcription and splicing by steroid receptor coregulators. Science. 2002;298:416–419. doi: 10.1126/science.1073734. [DOI] [PubMed] [Google Scholar]
  2. Balamotis MA, Pennella MA, Stevens JL, Wasylyk B, Belmont AS, Berk AJ. Complexity in transcription control at the activation domain-Mediator interface. Sci Signal. 2009;2:ra20. doi: 10.1126/scisignal.1164302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Batsche E, Yaniv M, Muchardt C. The human SWI/SNF subunit Brm is a regulator of alternative splicing. Nat Struct Mol Biol. 2006;13:22–29. doi: 10.1038/nsmb1030. [DOI] [PubMed] [Google Scholar]
  4. Berk AJ. Recent lessons in gene expression, cell cycle control, and cell biology from adenovirus. Oncogene. 2005;24:7673–7685. doi: 10.1038/sj.onc.1209040. [DOI] [PubMed] [Google Scholar]
  5. Bjorklund S, Gustafsson CM. The yeast Mediator complex and its regulation. Trends Biochem Sci. 2005;30:240–244. doi: 10.1016/j.tibs.2005.03.008. [DOI] [PubMed] [Google Scholar]
  6. Black DL. Mechanisms of alternative pre-messenger RNA splicing. Annu Rev Biochem. 2003;72:291–336. doi: 10.1146/annurev.biochem.72.121801.161720. [DOI] [PubMed] [Google Scholar]
  7. Dantonel JC, Murthy KG, Manley JL, Tora L. Transcription factor TFIID recruits factor CPSF for formation of 3′ end of mRNA. Nature. 1997;389:399–402. doi: 10.1038/38763. [DOI] [PubMed] [Google Scholar]
  8. Ding N, Zhou H, Esteve PO, Chin HG, Kim S, Xu X, Joseph SM, Friez MJ, Schwartz CE, Pradhan S, Boyer TG. Mediator links epigenetic silencing of neuronal gene expression with x-linked mental retardation. Mol Cell. 2008;31:347–359. doi: 10.1016/j.molcel.2008.05.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Donner AJ, Ebmeier CC, Taatjes DJ, Espinosa JM. CDK8 is a positive regulator of transcriptional elongation within the serum response network. Nat Struct Mol Biol. 2010;17:194–201. doi: 10.1038/nsmb.1752. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ebmeier CC, Taatjes DJ. Activator-Mediator binding regulates Mediator-cofactor interactions. Proc Natl Acad Sci U S A. 2010;107:11283–11288. doi: 10.1073/pnas.0914215107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Guang S, Felthauser AM, Mertz JE. Binding of hnRNP L to the pre-mRNA processing enhancer of the herpes simplex virus thymidine kinase gene enhances both polyadenylation and nucleocytoplasmic export of intronless mRNAs. Mol Cell Biol. 2005;25:6303–6313. doi: 10.1128/MCB.25.15.6303-6313.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hui J, Hung LH, Heiner M, Schreiner S, Neumuller N, Reither G, Haas SA, Bindereif A. Intronic CA-repeat and CA-rich elements: a new class of regulators of mammalian alternative splicing. EMBO J. 2005;24:1988–1998. doi: 10.1038/sj.emboj.7600677. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hui J, Stangl K, Lane WS, Bindereif A. HnRNP L stimulates splicing of the eNOS gene by binding to variable-length CA repeats. Nat Struct Biol. 2003;10:33–37. doi: 10.1038/nsb875. [DOI] [PubMed] [Google Scholar]
  14. Hung LH, Heiner M, Hui J, Schreiner S, Benes V, Bindereif A. Diverse roles of hnRNP L in mammalian mRNA processing: a combined microarray and RNAi analysis. RNA. 2008;14:284–296. doi: 10.1261/rna.725208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Johnson KM, Wang J, Smallwood A, Arayata C, Carey M. TFIID and human mediator coactivator complexes assemble cooperatively on promoter DNA. Genes Dev. 2002;16:1852–1863. doi: 10.1101/gad.995702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kagey MH, Newman JJ, Bilodeau S, Zhan Y, Orlando DA, van Berkum NL, Ebmeier CC, Goossens J, Rahl PB, Levine SS, et al. Mediator and cohesin connect gene expression and chromatin architecture. Nature. 2010;467:430–435. doi: 10.1038/nature09380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Khorosjutina O, Wanrooij PH, Walfridsson J, Szilagyi Z, Zhu X, Baraznenok V, Ekwall K, Gustafsson CM. A chromatin-remodeling protein is a component of fission yeast mediator. J Biol Chem. 2010;285:29729–29737. doi: 10.1074/jbc.M110.153858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kornblihtt AR. Chromatin, transcript elongation and alternative splicing. Nat Struct Mol Biol. 2006;13:5–7. doi: 10.1038/nsmb0106-5. [DOI] [PubMed] [Google Scholar]
  19. Kornblihtt AR, de la Mata M, Fededa JP, Munoz MJ, Nogues G. Multiple links between transcription and splicing. RNA. 2004;10:1489–1498. doi: 10.1261/rna.7100104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Licatalosi DD, Darnell RB. RNA processing and its regulation: global insights into biological networks. Nat Rev Genet. 2010;11:75–87. doi: 10.1038/nrg2673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Luco RF, Pan Q, Tominaga K, Blencowe BJ, Pereira-Smith OM, Misteli T. Regulation of alternative splicing by histone modifications. Science. 2010;327:996–1000. doi: 10.1126/science.1184208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Malik S, Roeder RG. The metazoan Mediator co-activator complex as an integrative hub for transcriptional regulation. Nat Rev Genet. 2010;11:761–772. doi: 10.1038/nrg2901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Maniatis T, Reed R. An extensive network of coupling among gene expression machines. Nature. 2002;416:499–506. doi: 10.1038/416499a. [DOI] [PubMed] [Google Scholar]
  24. Monsalve M, Wu Z, Adelmant G, Puigserver P, Fan M, Spiegelman BM. Direct coupling of transcription and mRNA processing through the thermogenic coactivator PGC-1. Mol Cell. 2000;6:307–316. doi: 10.1016/s1097-2765(00)00031-9. [DOI] [PubMed] [Google Scholar]
  25. Motta-Mena LB, Heyd F, Lynch KW. Context-dependent regulatory mechanism of the splicing factor hnRNP L. Mol Cell. 2010;37:223–234. doi: 10.1016/j.molcel.2009.12.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Munoz MJ, de la Mata M, Kornblihtt AR. The carboxy terminal domain of RNA polymerase II and alternative splicing. Trends Biochem Sci. 2010;35:497–504. doi: 10.1016/j.tibs.2010.03.010. [DOI] [PubMed] [Google Scholar]
  27. Neugebauer KM. On the importance of being co-transcriptional. J Cell Sci. 2002;115:3865–3871. doi: 10.1242/jcs.00073. [DOI] [PubMed] [Google Scholar]
  28. Rothrock CR, House AE, Lynch KW. HnRNP L represses exon splicing via a regulated exonic splicing silencer. EMBO J. 2005;24:2792–2802. doi: 10.1038/sj.emboj.7600745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Sharp PA. The discovery of split genes and RNA splicing. Trends Biochem Sci. 2005;30:279–281. doi: 10.1016/j.tibs.2005.04.002. [DOI] [PubMed] [Google Scholar]
  30. Sims RJ, 3rd, 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;28:665–676. doi: 10.1016/j.molcel.2007.11.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Stevens JL, Cantin GT, Wang G, Shevchenko A, Shevchenko A, Berk AJ. Transcription control by E1A and MAP kinase pathway via Sur2 Mediator subunit. Science. 2002;296:755–758. doi: 10.1126/science.1068943. [DOI] [PubMed] [Google Scholar]
  32. Struhl K. Transcriptional activation: Mediator can act after preinitiation complex formation. Mol Cell. 2005;17:752–754. doi: 10.1016/j.molcel.2005.03.002. [DOI] [PubMed] [Google Scholar]
  33. Taatjes DJ. The human Mediator complex: a versatile, genome-wide regulator of transcription. Trends Biochem Sci. 2010;35:315–322. doi: 10.1016/j.tibs.2010.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Takahashi H, Parmely TJ, Sato S, Tomomori-Sato C, Banks CA, Kong SE, Szutorisz H, Swanson SK, Martin-Brown S, Washburn MP, et al. Human Mediator subunit MED26 functions as a docking site for transcription elongation factors. Cell. 2011;146:92–104. doi: 10.1016/j.cell.2011.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Tian B, Hu J, Zhang H, Lutz CS. A large-scale analysis of mRNA polyadenylation of human and mouse genes. Nucleic Acids Res. 2005;33:201–212. doi: 10.1093/nar/gki158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Toth-Petroczy A, Oldfield CJ, Simon I, Takagi Y, Dunker AK, Uversky VN, Fuxreiter M. Malleable machines in transcription regulation: the Mediator complex. PLoS Comput Biol. 2008;4:e1000243. doi: 10.1371/journal.pcbi.1000243. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Wallberg AE, Yamamura S, Malik S, Spiegelman BM, Roeder RG. Coordination of p300-mediated chromatin remodeling and TRAP/mediator function through coactivator PGC-1alpha. Mol Cell. 2003;12:1137–1149. doi: 10.1016/s1097-2765(03)00391-5. [DOI] [PubMed] [Google Scholar]
  38. Wang G, Balamotis MA, Stevens JL, Yamaguchi Y, Handa H, Berk AJ. Mediator requirement for both recruitment and postrecruitment steps in transcription initiation. Mol Cell. 2005;17:683–694. doi: 10.1016/j.molcel.2005.02.010. [DOI] [PubMed] [Google Scholar]
  39. Wang G, Berk AJ. In vivo association of adenovirus large E1A protein with the human Mediator complex in adenovirus-infected and -transformed cells. J Virol. 2002;76:9186–9193. doi: 10.1128/JVI.76.18.9186-9193.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Wang G, Cantin GT, Stevens JL, Berk AJ. Characterization of Mediator complexes from HeLa cell nuclear extract. Mol Cell Biol. 2001;21:4604–4613. doi: 10.1128/MCB.21.14.4604-4613.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Wang W, Huang L, Huang Y, Yin JW, Berk AJ, Friedman JM, Wang G. Mediator MED23 links insulin signaling to the adipogenesis transcription cascade. Dev Cell. 2009;16:764–771. doi: 10.1016/j.devcel.2009.04.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Yuan W, Xie J, Long C, Erdjument-Bromage H, Ding X, Zheng Y, Tempst P, Chen S, Zhu B, Reinberg D. Heterogeneous nuclear ribonucleoprotein L Is a subunit of human KMT3a/Set2 complex required for H3 Lys-36 trimethylation activity in vivo. J Biol Chem. 2009;284:15701–15707. doi: 10.1074/jbc.M808431200. [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
NIHMS347083-supplement-01.doc (1.3MB, doc)

RESOURCES