Brd4‐Brd2 isoform switching coordinates pluripotent exit and Smad2‐dependent lineage specification

Abstract Pluripotent stem cells (PSCs) hold great clinical potential, as they possess the capacity to differentiate into fully specialised tissues such as pancreas, liver, neurons and cardiac muscle. However, the molecular mechanisms that coordinate pluripotent exit with lineage specification remain poorly understood. To address this question, we perform a small molecule screen to systematically identify novel regulators of the Smad2 signalling network, a key determinant of PSC fate. We reveal an essential function for BET family bromodomain proteins in Smad2 activation, distinct from the role of Brd4 in pluripotency maintenance. Mechanistically, BET proteins specifically engage Nodal gene regulatory elements (NREs) to promote Nodal signalling and Smad2 developmental responses. In pluripotent cells, Brd2‐Brd4 occupy NREs, but only Brd4 is required for pluripotency gene expression. Brd4 downregulation facilitates pluripotent exit and drives enhanced Brd2 NRE occupancy, thereby unveiling a specific function for Brd2 in differentiative Nodal‐Smad2 signalling. Therefore, distinct BET functionalities and Brd4‐Brd2 isoform switching at NREs coordinate pluripotent exit with lineage specification.

A mESCs differentiating upon 2i release for 1.5 days were treated with 1 lM MLN4924 for a further 2.5 days. Klf4, Fgf5, Sox1 and Brachyury mRNA levels were determined by qRT-PCR and compared to DMSO control (grey shade). Data are presented as mean AE SD of technical replicates from two experiments. B mESCs differentiating upon 2i release for 1.5 days were treated with 1 lM BI-D1870 or SL0101 for a further 2.5 days and Brachyury mRNA level determined by qRT-PCR. Data presented as mean AE SD of technical replicates from two experiments; statistical significance was determined using two-tailed unpaired Student's t-test (ns = not significant and ***P < 0.001). C 2i mESCs were treated with 100 nM JQ1 or DMSO control for 24 h. Nanog and Lamin B1 protein levels were determined by immunoblotting. D mESCs differentiating upon 2i release for 1.5 days were treated with 100 nM JQ1(+) or the inactive stereoisomer JQ1(-), for a further 2.5 days. Brachyury and Lamin B1 levels were evaluated by immunoblotting. E mESCs differentiating upon 2i release for 1.5 days were treated with 100 nM JQ1 for a further 2.5 days, and mRNA expression of Mixl and Goosecoid was determined by qRT-PCR. Data are presented as mean AE SD of technical replicates from two experiments; statistical significance was determined using two-tailed unpaired Student's t-test (*P < 0.05, **P < 0.01). F mESCs differentiating for 1.5 days were treated with 100 nM JQ1(+), the inactive stereoisomer JQ1(-) or 3 lM SB505124 for a further 2.5 days. Phospho-Smad1, Smad1 and Lamin B1 levels were evaluated by immunoblotting.
Source data are available online for this figure. A mESCs differentiating upon 2i release were treated with 100 nM JQ1, 3 lM SB505124 or 1 lM LDN193189 for 1 h or 48 h. pSmad2 and Erk1/2 levels were determined by immunoblotting. B mESCs differentiating upon 2i release were treated with 100 nM JQ1 and stimulated with TGFb1, Activin A or BMP4 for 48 h. Mixl and Nodal mRNA levels were quantified by qRT-PCR. Data are presented as mean AE SD of technical replicates from a representative experiment; statistical significance was determined for each condition relative to control using two-tailed unpaired Student's t-test (ns = not significant, *P < 0.05 and **P < 0.01). Similar results were found in three independent experiments (n = 3). C PAI-1 luciferase U2OS cells were treated with vehicle control, SB505124 or JQ1 for 48 h, stimulated with TGFb1 and PAI-dependent luminescence determined. Data are presented as mean AE SD of technical triplicates from three experiments (n = 3).
Source data are available online for this figure. Low ex.
High ex. Figure EV3. BET activity is required for Smad2 signalling and Brachyury induction in hiPSCs.
A, B Control hiPSCs or hiPSCs induced to differentiate with CHIR99021 for 24 h were treated with JQ1 or SB505124. (A) Levels of Smad2 phosphorylation and Smad2 were evaluated by immunoblotting. (B) Total RNA was extracted and levels of Brachyury mRNA were quantified by qRT-PCR. Graph shows mean AE SD of technical replicates from a representative experiment; statistical significance was determined using two-tailed unpaired Student's t-test (**P < 0.01, ***P < 0.001). Similar results were found in three independent experiments (n = 3).
Source data are available online for this figure.
EMBO reports ª 2017 The Authors  with 100 nM JQ1 for a further 2.5 days, and levels of Nodal nascent mRNA and mature mRNA determined by qRT-PCR. Graph shows mean AE SD of technical replicates from two independent experiments; statistical significance was determined using two-tailed unpaired Student's t-test (**P < 0.01, ****P < 0.0001).  Figure EV5. Specific requirement for Brd2 in mesendoderm differentiation.
A 2i mESCs were transfected with control, Brd2, Brd3 or Brd4 siRNAs and differentiated in 2i release for 3 days. Sox1 mRNA level was determined by qRT-PCR. Box plots show technical replicates from three independent experiments. Boxes extend from the 25 th to the 75 th percentile, with the median represented as the line segment inside each box. The whiskers expand from the minimum to the maximum value. Statistical significance was determined for each condition relative to control using two-tailed unpaired Student's t-test. (****P < 0.0001, ns = not significant). B 2i mESCs were transfected with control, Brd2, Brd3 or Brd4 siRNAs and differentiated in 2i release for 3 days. Mixl, Lefty1 and Lefty2 mRNA levels were determined by qRT-PCR. Box plots show technical replicates from at least three independent experiments. Boxes extend from the 25 th to the 75 th percentile, with the median represented as the line segment inside each box. The whiskers expand from the minimum to the maximum value. Statistical significance was determined for each condition relative to control using two-tailed unpaired Student's t-test (ns = not significant, *P < 0.05, **P < 0.01, ***P < 0.001). C 2i mESCs were transfected with control, Brd2, Brd3 or Brd4 siRNAs and differentiated in 2i release for 3 days. Brd2, Brd3 and Brd4 protein levels were determined by immunoblotting and quantified using ImageLab software. Data are presented as mean AE SD of values obtained in two different experiments (n = 2). D 2i mESCs were transfected with control, Brd2 or Brd4 siRNAs and differentiated in 2i release for 3 days. Nodal and Brachyury mRNA levels were determined by qRT-PCR. Box plots show technical replicates from at least three independent experiments. Boxes extend from the 25 th to the 75 th percentile, with the median represented as the line segment inside each box. The whiskers expand from the minimum to the maximum value. Statistical significance was determined for each condition relative to control using two-tailed unpaired Student's t-test (ns = not significant, *P < 0.05, **P < 0.01, ***P < 0.001).