RESEARCH ARTICLE observation by Mitsubishi Pha Methods).We have also identified a route to synthesize each enantiome ous re this nall human bromodomains were n svs nd triazola s within the b nding po the the toas BRD4(D):Protein Data Bank code2 ned Table )withT igand,JQ1 (Fig.la 10.1 BRD4(I to (1)all gnly se ctive, ne ster -JQ1 showe ished struc Within a f mily of proteins orrelation bet en DSF AT referedtoasjQlandac 020406080700120140 :8 0品品。 d12 ERD42 -8 nts a least twith an Rof The grey.(-)-JQ1 c e into buffer (A)an smonitorigg B()or of an BBP are showr 1o6INATUREIVoL46sI2i/B0DEceSI2nPbhsLnd.Arosmenad observation byMitsubishi Pharmaceuticals that simple thienodiazepines possessed binding activityfor BRD4 (ref. 20). Previous researchfrom this group indicates that these compounds emerged from anti-inflammatory phenotypic studies, such as inhibition of CD28 co-stimulation as a means of treating autoimmune diseases21,22. A rich literature has estab￾lished the synthetic accessibility and favourable pharmacological prop￾erties of this privileged class of drug-like small molecules23. Indeed, the core scaffold described appears in FDA-approved substances such as alprazolam and triazolam. Inferring structure–activity relationships also derived from molecular modelling of candidate ligands within the binding pocket of the apo crystal structure of the first bromodomain of BRD4 (hereafter referred to as BRD4(1); Protein Data Bank code 2OSS), we designed a prototype ligand, JQ1 (Fig. 1a). JQ1 is a novel thieno-triazolo-1,4-diazepine, possessing an appended, bulky t-butyl ester functional group at C6 in order to (1) allow for additional pendant group diversity, as needed, and (2) to mitigate binding to the central benzodiazepine receptor as predicted by published structure–activity relationships23. We first estab￾lished a high-yielding, seven-step synthetic route to access racemic JQ1 (hereafter referred to as JQ1) and derivatives (scheme 1 in Supplementary Methods).We have also identified a route to synthesize each enantiomer, (1)-JQ1 and (2)-JQ1 (scheme 2 in Supplementary Methods). To establish a biochemical platform for comprehensive selectivity screening, all human bromodomains were subcloned into bacterial expression vectors. Testing of an average of 15 expression constructs per bromodomain resulted in the identification of 37 expression systems that yielded soluble protein suitablefor specificity screening and covered all bromodomain subfamilies (Supplementary Table 1). Because the specific substrates of most bromodomains are unknown, a general bind￾ing assay based on differential scanning fluorimetry (DSF) was imple￾mented24. Binding of (1)-JQ1 significantly increased the thermal stability of all bromodomains of the BET family (Fig. 1b and Sup￾plementary Table 2) with DTm obs values between 4.2 uC (BRDT(1)) and 10.1 uC (BRD4(1)). No significant stability shifts were detected for bromodomains outside the BET family, indicating that this ligand is highly selective. In contrast, the stereoisomer (2)-JQ1 showed no significant interaction with any bromodomain present in our panel. Within a family of proteins a linear correlation between DSFDTm obs values and binding constants has been observed, with temperature shifts larger than 6 uC corresponding to compounds with nanomolar log[conc.(μM)] Inhibition (%) –10 –8 –6 –4 0 50 100 BRD4(1) (+)-JQ1 BRD4(2) (+)-JQ1 CREBBP (+)-JQ1 BRD4(1) (–)-JQ1 40 60 80 100 120 140 160 180 200 4 6 8 10 12 ΔT o m bs (°C) Kd (nM) 0 20 40 60 80 100 120 140 –0.8 –0.7 –0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0.0 Time (min) ΔP (μcal s–1) 0.0 0.5 1.0 1.5 2.0 Molar ratio –8 –6 –4 –2 kcal mol–1 of injectant 0 A B C (–)-JQ1 A: blank (+)-JQ1 B: (–)-JQ1 C: (+)-JQ1 >9 °C 7–9 °C 4–7 °C 1–4 °C 0–1 °C <0 °C CREBBP EP300 BAZ1A BRD8(1) BRD8(2) BRDT(1) BRD4(1) BRD2(1) BRD3(1) BRDT(2) BRD3(2) BRD2(2) BRD4(2) BRPF1 BRPF3 BRD9 BRD7 KIAA1240 ATAD2 WDR9(2) PHIP(2) BRWD3(2) BAZ1B BAZ2B BAZ2A ZMYND11 PRKCBP1 TAF1(1) TAF1L(1) TAF1(2) TAF1L(2) CECR2 FALZ GCN5L2 PCAF WDR9(1) BRWD3(1) PHIP(1) ASH1L PB1(3) PB1(6) PB1(1) PB1(4) PB1(5) PB1(2) SMARCA2 SMARCA4 MLL SP110 SP100 SP140 TRIM28 TIF1 TRIM66 TRIM33 LOC93349 BRD1 (+)-JQ1 (–)-JQ1 Cl N N O O S N N Cl N N O O S N N a c b d e Figure 1 | Structure and selectivity of JQ1. a, Structure of the two JQ1 stereoisomers. The stereocentre at C6 is indicated by an asterisk. b, Assessment of inhibitor selectivity using differential scanning fluorimetry (DSF). Shown are averaged temperature shifts (DTm obs) in degrees Celsius upon binding of 10 mM (1)-JQ1. The temperature shifts are represented by spheres as indicated in the inset. Screened bromodomains are labelled and proteins not included in the selectivity panel are shown in grey. (2)-JQ1 did not reveal any significant temperature shifts to any of the screened bromodomains. c, Isothermal titration calorimetry (ITC). Differential power (DP) data time course of raw injection heats are shown for a blank titration of BRD4(1) into buffer (A), and reverse titrations using the inactive isomer (2)-JQ1 (B) and the active isomer (1)-JQ1 (C). The inset shows normalized binding enthalpies corrected for the heat of dilution as a function of binding site saturation (symbols as indicated in the inset). Solid lines represent a nonlinear least squares fit using a single-site binding model. d, Thermal shifts (DTm obs) show good correlation to dissociation constants (Kd) determined by ITC for the BET bromodomains. The dotted red line represents a least squares fit with an R of 96%. The DTm obs data represent the mean 6 s.d. (n 5 3). Error for ITC data was based on deviations to least squares fit described in c. e, Competitive displacement of a histone peptide from human bromodomains is exhibited by JQ1 using a bead￾based proximity assay. Alpha-screen titrations monitoring the displacement of a tetra-acetylated histone H4 peptide by JQ1 isomers using the bromodomains BRD4(1), BRD4(2) or of an acetylated H3 peptide using CREBBP are shown. RESEARCH ARTICLE 1068 | NATURE | VOL 468 | 23/30 DECEMBER 2010 ©2010 Macmillan Publishers Limited. All rights reserved