Effect of BET Missense Mutations on Bromodomain Function, Inhibitor Binding and Stability

Abstract

Lysine acetylation is an important epigenetic mark regulating gene transcription and chromatin structure. Acetylated lysine residues are specifically recognized by bromodomains, small protein interaction modules that read these modification in a sequence and acetylation dependent way regulating the recruitment of transcriptional regulators and chromatin remodelling enzymes to acetylated sites in chromatin. Recent studies revealed that bromodomains are highly druggable protein interaction domains resulting in the development of a large number of bromodomain inhibitors. BET bromodomain inhibitors received a lot of attention in the oncology field resulting in the rapid translation of early BET bromodomain inhibitors into clinical studies. Here we investigated the effects of mutations present as polymorphism or found in cancer on BET bromodomain function and stability and the influence of these mutants on inhibitor binding. We found that most BET missense mutations localize to peripheral residues in the two terminal helices. Crystal structures showed that the three dimensional structure is not compromised by these mutations but mutations located in close proximity to the acetyl-lysine binding site modulate acetyl-lysine and inhibitor binding. Most mutations affect significantly protein stability and tertiary structure in solution, suggesting new interactions and an alternative network of protein-protein interconnection as a consequence of single amino acid substitution. To our knowledge this is the first report studying the effect of mutations on bromodomain function and inhibitor binding.

Fig 1. Alignment of BET bromodomain mutants.

(A) Secondary structure elements are shown at the top of the sequence alignment. Mutated residues are highlighted in red and studied mutations are listed. The conserved asparagine (N391in BRD3(2) numbering) is highlighted in yellow. The green dots represent the residues involved in binding with inhibitor JQ1 (PDB ID: 3ONI, 3S92, 3MXF). The residues underlined in blue are involved in PFI-1binding (PDB ID: 4E96). (B) Location of the mutations. Shown are the first (left) and second bromodomain of BRD2. The mutated residues are highlighted in ball and stick and the position of Cα atoms are shown as a sphere. The main structural elements are labelled.

Fig 2. Structure of BET mutants and tertiary structure of mutants in solution.

(A) Superimposition of wild type BRD2(1) shown as a ribbon diagram with the mutants BRD2(1) R100L and D161Y shown as protein worm in green and magenta, respectively. The mutated residues are shown in ball and stick representation and main structural elements are labelled. (B) Details of interactions formed by R100 in the wild type compared to the mutated residue. (C) Detailed view of BRD2(1) wild type and D161Y. (D) Superimposition of BRD2(2) shown as a ribbon diagram and the mutant BRD2(2) Q443H shown as protein worm in blue. (E) Comparison of the near UV CD spectra of wild type BRD2(1) and all generated mutants. (F) Comparison of the near UV CD spectra of wild type BRD4(1) and the mutants BRD4(1) A89V. (G) Comparison of the near UV CD spectra of wild type BRD2(2) and the mutants BRD2(2) R419W and Q443H. (H) Comparison of the near UV CD spectra of wild type BRD3(2) and the mutants BRD3(2) H395R. (I) Comparison of the near UV CD spectra of wild type BRD4(2) and the mutants BRD4(2) A420D. Near-UV CD spectra were recorded at 20°C in a 1.0-cm quartz cuvette in 20 mM Tris/HCl, pH 7.5 containing 0.20 M NaCl and 2.00 mM DTT, as described in Materials and Methods.

Fig 3. Far-UV CD spectra of wild type bromodomains and mutants.

Far-UV CD spectra were recorded at 20°C in a 0.1-cm quartz cuvette in 20 mM Tris/HCl, pH 7.5 containing 0.20 M NaCl and 0.40 mM DTT, as described in Materials and Methods. Wild type spectra are shown as black solid lines and mutants are coloured as indicated in the figure.

Fig 4. Binding of BET bromodomain mutants to acetylated histone peptides and inhibitors.

(A) BLI study showing the interaction of some mutants with acetylated histone peptides. Shown is the maximum response at a protein concentration of 20 μM after subtraction of non-acetylated reference peptides using a colour code as indicated in the figure capture. (B) Structure of BRD3(2) mutant H395R. The mutated residues are highlighted. A nitrate molecule, present in the crystallization solution occupied the acetyl-lysine binding site in BRD3(2) H395R. The conserved asparagine (N391) formed canonical hydrogen bonds with the nitrated ion. (C) Superimposition of the wild type JQ1 complex with BRD3(2) H395R. (D) ITC experiments measuring the interaction of the panBET inhibitor JQ1 with wild type BRD3(2) (black curve) and BRD3(2) H395R (blue curve). Shown are raw titration heats (top panel) as well as normalized binding enthalpies as a function of ligand/protein ratio (lower panel). The best fit to a single binding site model is shown as solid lines. (E) ITC experiments showing the binding of the pan-BET inhibitor PFI-1 with BRD3(2) and the H395R mutant. A significant reduction in binding enthalpy and binding affinity is evident. Data on all ITC titrations are summarized in S2 Table.