MED16 dissociates from the Mediator complex and interacts with UBP1-TFCP2

(A) Co-IP experiment with antibodies against endogenous CDK8, MED1, and MED12 in 293T whole cell lysis. The resultant immunoprecipitates were washed by washing buffers containing 0%, 5%, and 10% of 1,6-hexanediol respectively. Input: 0.5% of the total lysate was loaded. (B) Co-IP experiment with antibodies against endogenous CDK8, MED1, and MED12 in HeLa nuclear extract. The resultant immunoprecipitates were washed by washing buffers containing 10% of 1,6-hexanediol or 10% 2,5-hexanediol. Input: 0.2% of the total HeLa nuclear extract was loaded. (C) Co-IP experiment with antibodies against endogenous MED1 in 293T whole cell lysis. The resultant immunoprecipitated was washed by washing buffers containing 150 mM, 300 mM, and 500 mM of NaCl respectively. Input: 0.2% of the total HeLa nuclear extract was loaded. (D) Gel filtration chromatography of HeLa nuclear extract. 500µl HeLa nuclear extract was applied to Superose 6 column then was run in Buffer D. Column fractions of 500µl were collected. The Void (void volumn) of Superose 6 is based on the volume of effluent required for the elution of blue dextran (molecular mass of ∼2000 kDa). The molecular weight of corresponding fractions was detected by protein standards. (E) Immunoblots of anti-MED16 immunoprecipitation in HeLa nuclear extract and gel filtration factions No.29 to No.31. The immunoprecipitated proteins were detected with indicated antibodies by western blotting. HeLa NE input: 0.2% of the total HeLa nuclear extract was loaded. (F) Immunoblots of anti-TFCP2 immunoprecipitation in HeLa nuclear extract and gel filtration factions No.29 to No.31. The immunoprecipitated proteins were detected with indicated antibodies by western blotting. HeLa NE input: 0.2% of the total HeLa nuclear extract was loaded.

MED16 αβ-domain interacts with UBP1 specific 36 amino acids sequence in Vivo and in Vitro.

(A) Schematic of Flag-MED16 deletion mutants. The WDR domains are shown as blue boxes. “1” represent full-length MED16 and “2” to “9” represent different truncations of MED16. All proteins were labeled with Flag-tag and expressed in 293T cell line as panel (B) indicated. (B) Domain mapping experiment of MED16. Flag-MED16 deletion mutants were coimmunoprecipitated with endogenous MED1 and TFCP2 using anti-Flag antibody. Asterisks indicate the IgG heavy and light chains. (C) Purified GST and GST-MED16 αβ-domain proteins were incubated with 6×His-tag TFCP2, UBP1, TFCP2L1, and YY1 at 4℃. After washing the interaction protein were eluted by boiling and immunoblotted with antibodies against GST-tag and His-tag. (D) Amino acids sequence analysis of UBP1 and TFCP2 protein, a 36 amino acids UBP1-specific sequence (UBP1-SP) as shown below. (E) Schematic of Flag-UBP1 deletion mutants. FL represents full-length UBP1, △S represents SAM motif truncated UBP1, and DBD represents UBP1 DNA binding domain. (F) Domain mapping experiment of UBP1. Flag-UBP1 deletion mutants were coimmunoprecipitated with endogenous MED1, MED23, MED16, and TFCP2. Asterisks indicate the IgG heavy and light chains.

MED16 is required for UBP1 activation of gene transcription

UBP1 reporter assay (A) with co-transfection of UBP1, MED16, and their mutants in various cell lines. In HeLa cells (B), 300 ng of UBP1 DNA binding domain (DBD), SAM motif-deleted mutant (ΔS), or full-length UBP1 (FL) plasmid was transfected with the reporter. (C) In MED16 knockdown and control HeLa lines generated by shRNA, 300 ng of UBP1 plasmid or control vector was co-transfected with the UBP1 reporter. Knockdown efficiency was confirmed by qPCR and Western blotting (D). (E) In wild-type and MED16 knockout mouse pancreatic cancer HT cells, the UBP1 reporter was co-transfected with 300 ng of mouse UBP1 plasmid or vehicle control. Knockout efficiency was validated by Western blotting (F). (G) Effect of increasing MED16 doses (0.3 μg, 0.5 μg, 1 μg) or αβ-domain truncated MED16 on the UBP1 reporter. (H) Effects of MED16, UBP1, and TFCP2 combinations on UBP1 reporter activity. Firefly luciferase activity was normalized to Renilla luciferase activity. The luciferase expression levels between groups were compared using two-tail unpaired Student’s t-test, The data are presented as the means ± SDs from n = 3 independent experiments. *: p < 0.05, **: p-value<0.01 and “ns”: no significant.

UBP1 and MED16 co-regulated transcriptome analysis

(A) The number of genes repressed or activated by UBP1 overexpression in the sh-NC and shMED16 HeLa cell line (FC > |1.5|, p ≤ 0.05). (B) Venn diagram of UBP1-activated genes in the sh-NC HeLa cell line and the shMED16 HeLa cell line (FC > |1.5|, p ≤ 0.05). 198 genes which were activated by UBP1 overexpression but abolished by sh-MED16 were defined as UBP1 and MED16 co-regulated genes. (C) Heatmap showing the RNA-seq analysis of 217 UBP1-activated genes in sh-MED16 and sh-NC cell line treated with or without UBP1 overexpression. UBP1-activated genes were defined in sh-NC cell line under UBP1 overexpression at fold change>|1.5|, p-value≤0.05. (D) GO analysis of 198 UBP1 and MED16 co-regulated genes. The top 10 biological process items were presented. (E) Volcano map analysis of differentially expressed genes in sh-NC HeLa cell line with or without UBP1 overexpression. Genes failed to be activated by UBP1 in the sh-MED16 HeLa cell line we’re indicated by "×". (F) UBP1 motif occurrence analysis of SFTPA1, SFTPA2, SFTPB and SFTPC. Motif occurrence were showed within the promoter region of the genes (-1000 bp to 200 bp relative to the TSS).

MED16 cooperates with UBP1 to inhibit HIV-1 reporter expression

(A) Schematic illustration of the HIV-1 reporter, with the HIV-1 core promoter and +1 to +60 TAR RNA sequence cloned upstream of the luciferase coding sequence (CDS). (B) The HIV-1 reporter was co-transfected with 1.5 μg of MED16, UBP1, TFCP2, YY1, or Flag-tag control plasmids in 293T cells. (C and D) Increasing amounts (0.7 μg, 1 μg, 1.5 μg) of MED16 or αβ-domain-deleted MED16 were transfected with the HIV-1 reporter. (E) The HIV-1 reporter was co-transfected with 1.5 μg of MED23, MED24, or Flag-tag control plasmids as additional controls. (F) Comparison of HIV-1 reporter luciferase activity between sh-NC and sh-MED16 293T cell lines. Firefly luciferase activity was normalized to Renilla luciferase activity. The luciferase expression levels between groups were compared using two-tail unpaired Student’s t-test, The data are presented as the means ± SDs from n = 3 independent experiment. * p < 0.05, ** p-value<0.01 and “ns”: no significant.

UBP1 binding site determines the UBP1-MED16 inhibition in HIV-1 transcription

Schematic showing a chimera reporter (A) driven by the Chicken β-ACTIN promoter, followed by an HIV-1 UBP1 binding site that also serves as the HIV-1 transcription start site and another chimera reporter (B) driven by the Chicken β-ACTIN promoter, a UBP1 binding site modify from HIV-1 TSS placed upstream of it. 1 μg of different truncation of MED16 or UBP1 plasmid was co-transfected with the reporter. The Luciferase activity comparison between two chimera reporters were shown in (C). “cβa-UBP1 bd” indicate the reporter that panel A described, and “UBP1 bd-cβa” indicate the reporter that panel B described. Firefly luciferase activity was normalized to Renilla luciferase activity. The luciferase expression levels between groups were compared using two-tail unpaired Student’s t-test, The data are presented as the means ± SDs from n = 3 independent experiment. * p < 0.05, ** p-value<0.01 and “ns”: no significant. UBP1 motif enrichment analysis of top 50 upregulated or top 50 downregulated transcripts induced by UBP1 were shown in (D). The transcripts changes levels are shown on the left as a heatmap, and average profile plots of UBP1 motif density are shown on the right. UBP1 motif occurrence analysis of UBP1 downregulated gene KIFC1 and GTF2H4 were shown in (E). Motif occurrence were showed within the promoter region of the genes (-500 bp to 200 bp relative to the TSS).

MED16-UBP1 complex prohibited PIC formation of HIV-1 transcription

(A) Schematic of biotin-labeled immobilized template assay. (B) Immobilized template assay in wild-type, Med16 knockout and Med16 overexpression HT cell line. The binding proteins in individual steps were collected for immunoblotting. Beads: the streptavidin beads incubated with NE as a negative control; PIC: the pre-initiation step of transcription; Trx.: transcription initiation from the PIC step by adding NTPs. (C) qPCR quantification of generated RNA from different NE. The mRNA expression levels were compared between groups using two-tail unpaired Student’s t-test, The values are presented as the means ± SDs from n = 3 independent experiment. ** p-value<0.01 and “ns”: no significant.

MED16:UBP1 complex prohibited PIC formation of HIV-1 transcription

(A and B) Stable overexpression of MED16 (MED16 OE) in the J-Lat 10.6 cell line using MSCV retrovirus. Control: Control J-Lat 10.6 cell line generated via infection with a vehicle retrovirus. (C) Fluorescence microscopy images of MED16 OE and Control J-Lat 10.6 cells treated with or without 5 μM JQ1 for 48 hours. Percentage of GFP-positive cells measured by FACS (D), and mean fluorescence intensity (MFI) quantified using the BioTek Synergy 2 system (E). The values are presented as the means ± SDs from n = 3 independent experiment. ** p-value<0.01

Venn diagram of MED1, MED12, CDK8 and MED16 interacting proteins identified by IP-MS.

MED16 specific interacting proteins were highlight on the right.

Identifying the MED16-controled regulatory elements at the HIV-1 promoter

293T cells were transfected with 300 ng of either Flag control, MED16 overexpression, or UBP1 overexpression plasmid, along with the indicated reporter plasmids. (A) Cells were co-transfected with an HIV-1 reporter plasmid, then treated with or without 60 nM Panobinostat for 16 hours. Deacetylation inhibition was validated by western blot analysis (shown below). (B) Co-transfection with a TAR RNA mutant HIV-1 reporter and 50 ng of Tat overexpression plasmid, to assess the mutation’s effects. (C) Co-transfection with a TAR RNA mutant HIV-1 reporter to evaluate regulation by Flag control, MED16, or UBP1 overexpression. (D) Co-transfection with a 4×GC box reporter plasmid to examine MED16’s impact on GC box activity. (E) Co-transfection with an AP-4 binding site mutant HIV-1 reporter to assess MED16’s effect on the AP-4 site. The data are presented as the means ± SDs from n = 3 independent experiments; “ns”: no significant.

Effect of MED16 truncations co-transfected with HIV-1 reporter

(A) Diagram of the HIV-1 promoter, with UBP1 binding consensus sequences highlighted in red. (B) 293T cells were transfected with 1 μg of MED16 truncation plasmids (Δ3WD, Δ6WD, and Δ9WD, corresponding to constructs 2, 3, and 4 in Fig. 2A) along with the HIV-1 reporter plasmid shown in (A). (C) 293T cells were transfected with increasing amounts (0.3 μg, 0.5 μg, and 1.0 μg) of MED16 αβ-domain peptide overexpression plasmid together with the HIV-1 reporter. The data are presented as the means ± SDs from n = 3 independent experiment. * p-value<0.05, ** p-value<0.01 and “ns”: no significant.