Isoxazole‐Derived Amino Acids are Bromodomain‐Binding Acetyl‐Lysine Mimics: Incorporation into Histone H4 Peptides and Histone H3

Abstract A range of isoxazole‐containing amino acids was synthesized that displaced acetyl‐lysine‐containing peptides from the BAZ2A, BRD4(1), and BRD9 bromodomains. Three of these amino acids were incorporated into a histone H4‐mimicking peptide and their affinity for BRD4(1) was assessed. Affinities of the isoxazole‐containing peptides are comparable to those of a hyperacetylated histone H4‐mimicking cognate peptide, and demonstrated a dependence on the position at which the unnatural residue was incorporated. An isoxazole‐based alkylating agent was developed to selectively alkylate cysteine residues in situ. Selective monoalkylation of a histone H4‐mimicking peptide, containing a lysine to cysteine residue substitution (K12C), resulted in acetyl‐lysine mimic incorporation, with high affinity for the BRD4 bromodomain. The same technology was used to alkylate a K18C mutant of histone H3.

Deconvoluted masses (1.0 Da resolution, 10,000-20,000 Da range) from combined spectra across the protein region of LCMS chromatograms (around 4.3 min retention time). A H3K18C starting protein showing single species as β-ME adduct: calculated MW of reduced H3K18C 15213 Da + 76 Da for SC 2 H 4 OH. B Alkylation reaction after 1 h: reaction progression evident through observation of H3K18C (15214 Da) and H3K18C-DMI (15309 Da) protein species. C Alkylation reaction after 3 h and buffer exchange into H 2 O: complete monoalkylation of protein resulted in single protein species corresponding to H3K18C-DMI (15309 Da).

S15
Supporting Inhibitory values of unprotected amino acids given as average percentage inhibition (buffer blank -0% inhibition; 25% DMSO -100% inhibition) against each bromodomain, from measurements in triplicate (AlphaScreen). n.d. value not determined. Inhibition values highlighted as a heatmap from green (low inhibition) to red (high inhibition). The phenol derivative (2) bearing a 3,5-dimethylisoxazole moiety, with an IC 50 of 382 nM, [9] was included on each plate as a positive control, which in every case demonstrated 100% inhibition. Data for BAZ2A and BRD9 were obtained by Dr. Oleg Fedorov at the SGC. Several amino acids resulted in >50% inhibition at 250 μM, with significant inhibitory effect detected at lower concentrations of 25/50 μM. These findings represent an increase in binding affinity relative to that of KAc, which displayed negligible inhibition at 250 μM, and are the first unnatural free amino acids known to inhibit bromodomain recognition of an KAc-containing peptide. The data suggest that a three-atom linker (3-5 & 12-14) is long enough to allow the isoxazole to access the KAc pocket, and extended lengths are less well tolerated. (23.5% inhibition at 250 µM for three atom linker 12 against BRD4(1); 7.5% inhibition at 250 µM for four atom linker 10). Amide connections demonstrated a higher percentage of BRD4(1) inhibition than alkyl or ether, with a preferred orientation of the carbonyl adjacent to the isoxazole (31.2% inhibition at 250 µM for 5). The phenyl-linked isoxazole (8) showed the strongest inhibitory activity against BRD4(1) (>70% inhibition at 250 μM). Apparent selectivity over BRD9, might be due to increased rigidity, or the ability of the aromatic linker to interact with the hydrophobic region of the ZA channel and displace multiple, proximal KAc residues more effectively.

S16
Supporting  (20 mg/mL α-cyano-4-hydroxycinnamic adcid in 50% acetonitrile) and spotted onto an MTP 384 Massive target plate using the dried droplet method. The instrument was calibrated directly prior to data acquisition using mono-isotopic peptide masses with Peptide Calibration Standard II (Bruker Daltonics, Coventry, UK). Sample ionization was achieved with a N 2 laser (337 nm) at 35-50% laser energy and MS spectra were acquired by manual operation in reflectron mode: 2000-3000 data points were collected to make a total intensity of approximately 8×10 4 a.u. Alkylation sites were assigned unambiguously by MALDI-TOF/TOF mass spectrometry: MS/MS spectra were acquired by Laser-induced fragmentation (LIFT). [11] Predicted fragmentation masses were calculated using the Fragment Ion Calculator, as part of the Proteomics Toolkit from the Institute for Systems Biology, and compared manually to spectrum generated.  [9] ). For incubation steps, the plate was sealed, shaken for 10 sec at 600 rpm, and incubated at room temperature in the dark for 1 h.

Synthetic Methods
Compound names are those generated by ChemBioDraw™ (CambridgeSoft) following IUPAC nomenclature.
Reagents and solvents used, unless otherwise stated, were of commercially available reagent grade quality and were used without further purification. Where appropriate and if not stated otherwise, all non-aqueous reactions were carried out in a flame dried flask under an inert atmosphere of nitrogen or argon. Anhydrous solvent was purchased from SigmaAldrich UK in SureSeal™ bottles, or dried according to the procedure outlined by Pangborn et al., [13] and used without purification unless otherwise indicated. Solvents for use in organometallic reactions were degassed by three freeze-thaw cycles under vacuum and were stored under an argon atmosphere over 3 Å molecular sieves. Flash column chromatography was carried out either on Merck silica gel 60 (240-400 mesh), eluting with solvents as supplied under a positive pressure of compressed air, or on a Biotage SP1 system using KP-Sil™ cartridges.
Melting points were determined using a Kofler hot stage microscope and are uncorrected.
Specific optical rotations were measured using a Perkin-Elmer 241 or 341 polarimeter with a water-jacketed 1 dm path-length cell maintained at 20 °C. The light source was maintained at 589 nm. The concentration (c) is expressed in g/100 mL and specific rotations are denoted [α] D with implied units of 10 −1 deg cm 2 g −1 .
Infrared spectra were obtained as a thin film on sodium chloride discs or from neat samples using a diamond ATR module. The spectra were recorded on a Bruker Tensor 27 spectrometer and a representative number of absorption maxima are reported in wavenumbers (cm −1 ). The intensity of each signal is indicated by: (w) weak; (m) medium; (s) strong; (br) broad. 1 H NMR spectra were recorded on Bruker DPX400, AVII400 or AVIII400 (400 MHz) and Bruker DRX500 or AVII500 with cryoprobe (126 MHz) spectrometers using deuterochloroform (unless indicated otherwise) as a reference for internal deuterium lock. The chemical shift data for each signal are given as δH in units of parts per million (ppm) relative to tetramethylsilane (TMS) where δH (TMS) = 0.00 ppm. The multiplicity of each signal is indicated by: s (singlet); br s (broad singlet); d (doublet); t (triplet); q (quartet); dd (doublet of doublets); dq (doublet of quartets); tt (triplet of triplets); qd (quartet of doublets); ddd (doublet of doublet of doublets); or m (multiplet); app (apparent). The number of protons (n) for a given resonance signal is indicated by nH. Coupling constants (J) are expressed in Hz and are recorded to the nearest 0.1 Hz. Identical proton coupling constants (J) are averaged in each spectrum and reported to the nearest 0.1 Hz. The coupling constants are determined using Bruker TopSpin software.
13 C NMR spectra were recorded on Bruker AVII400 or AVIII400 (101 MHz) and Bruker DRX500 or AVII500 with cryoprobe (126 MHz) spectrometers using broadband proton decoupling and an internal deuterium lock. The chemical shift data for each signal are given as δC in units of parts per million (ppm) relative to tetramethylsilane (TMS) where δC (TMS) = 0.0 ppm. Where appropriate, coupling constants (J) are expressed in Hz and are recorded to the nearest 0.1 Hz. 1 H and 13 C spectra were assigned using 2D NMR experiments including COSY, HSQC and HMBC.
Mass spectra were acquired on either a Micromass LCT Premier spectrometer, Agilent 6120 Quadrapole spectrometer or Bruker MicroTOF spectrometer using electrospray ionization, operating in positive or negative mode, from solutions of methanol. m/z values are reported in Daltons and followed by their percentage abundance in parentheses.
Elemental analyses were obtained by the microanalysis service of the London Metropolitan University, UK.
High-Performance Liquid Chromatography was carried out using a PerkinElmer Flexar system with a Binary LC pump (flow rate 0.6 mL/min) and UV/VIS LC detector. For determination of compound purity a Dionex Acclaim 120 column (C18, 5 µm, 120 Å, 4.6 × 150 mm) was used with the method described below ( ; samples were injected in CH 3 OH. Chromera software was used to determine purity and enantiomeric excess from relative peak areas of UV/VIS absorbance at 254 nm. Step length ( (20 mL). The organic layer was dried (MgSO 4 ), filtered and concentrated in vacuo to give crude products which were subsequently purified.

General Procedure 2 for the deprotection of aspartic acid derived amino acids
To a solution of the corresponding Boc-, O-Bn protected amino acid (1.0 eq.) in anhydrous CH 2 Cl 2 (0.07 mM) at -10 °C, BBr 3 (5.0 eq.) was added. The reaction mixture was left to stir at -10 °C for 1 h and then for a further duration at rt as indicated. The solution was quenched with H 2 O (10-20 mL) and washed with EtOAc (5 × 10-20 mL). Purification by ion exchange chromatography (Dowex 50WX8, 100-200 mesh), washing with H 2 O then eluting with 1 M aq. NH 4 OH, followed by lyophilization, yielded unprotected isoxazole-containing amino acids.

3,5-Dimethyl-4-(trifluoroboranyl)isoxazole, potassium salt (28)
Following the procedure of Molander et al., [14] to a suspension of 3,5-dimethylisoxazol-4ylboronic acid 27 (2.50 g, 17.7 mmol, 1.0 eq) in MeOH (5 mL) at 0 °C was added KHF 2 (4.16 g, 53.2 mmol, 3.0 eq). H 2 O (11.8 mL) was then added dropwise. The solution was allowed to warm to rt and stirred for 10 min, then concentrated and dried overnight in vacuo. The crude solid was purified by Soxhlet extraction (16 h) with acetone (100 mL). The collected solvent was concentrated in vacuo, and the residues were redissolved in the minimum amount of boiling acetone (400 mL). The product was precipitated by the addition of Et 2 O (600 mL) and collected by filtration. The filtrate was concentrated in vacuo, redissolved in acetone (100 mL) and further product was precipitated by the addition of Et 2 O (200 mL) and collected by filtration.
The filtrate was again concentrated in vacuo, redissolved in acetone (30 mL) and further product was precipitated by the addition of Et 2 O (60 mL) and collected by filtration. The combined solids were dried in vacuo to give 28 (3.02 g, 84%) as a powdery colorless solid. mp >275 °C (from acetone; lit. >200 °C); 1

Method A
To a solution of 47 (2.00 g, 12.9 mmol, 1.0 eq) in anhydrous EtOH (25 mL) at 0 °C under an argon atmosphere was added NaBH 4 (585 mg, 15.5 mmol, 1.2 eq). The reaction was allowed to warm to rt and stirred for 7 h, after which time TLC analysis indicated complete consumption of 47. The reaction was quenched with H 2 O (20 mL) and aq. HCl (1 M, 10 mL), and concentrated in vacuo to remove EtOH. The residues were extracted with EtOAc (3 × 30 mL), and the combined organic layers were washed with brine (100 mL), dried (MgSO 4 ), filtered and concentrated in vacuo. Purification via silica gel chromatography (gradient elution 60 to 70% Et 2 O in petroleum ether) yielded 48 as a pale yellow oil (989 mg, 68%).

5-(Bromomethyl)-3-methylisoxazole (49)
To a solution of 48 (793 mg, 7.01 mmol, 1.0 eq) in anhydrous CH 2 Cl 2 (25 mL) at 0 °C under an argon atmosphere was added PPh 3 Br 2 (3.55 g). The reaction was stirred in the dark for 1.5 h at 0 °C then for 1.5 h at rt, after which time TLC analysis indicated complete consumption of 48. Most CH 2 Cl 2 was removed in vacuo, then hexane (10 mL