Chemical Communications

Accepted Manuscript

This article can be cited before page numbers have been issued, to do this please use: R. Traquete, E. Henderson, S. Picaud, P. M. S. D. Cal, F. Sieglitz, T. Rodrigues, R. Oliveira, P. Filippakopoulos and G. J. L. Bernardes, Chem. Commun., 2019, DOI: 10.1039/C9CC05054J.

This is an Accepted Manuscript, which has been through the Royal Society of Chemistry peer review process and has been accepted for publication.
Accepted Manuscripts are published online shortly after acceptance, before technical editing, formatting and proof reading. Using this free service, authors can make their results available to the community, in citable form, before we publish the edited article. We will replace this Accepted Manuscript with the edited and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the Information for Authors.
Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal’s standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains.

Evaluation of linker length effects on a BET bromodomain probe

Rui Traquete,a Elizabeth Henderson,b Sarah Picaud,b Pedro M. S. D. Cal,a Florian Sieglitz,a Tiago Rodrigues,a Rudi Oliveira,c Panagis Filippakopoulos,b,* and Gonçalo J. L. Bernardesa,d,*
Fueled by the therapeutic potential of the epigenetic machinery, BET bromodomains have seen high interest as drug targets. Herein, we introduce different linkers to a BET bromodomain benzodiazepine ligand (I BET762) to gauge its implications in the development of hybrid drugs, imaging probes and small molecule drug conjugates. Biophysical studies confirmed minimal disruption to binding of the BRD4 cavity by the synthesized entities, which includes imaging probes. Target engagement was confirmed in a cellular context, but poor membrane diffusion was found despite efficient localization in the nuclei after membrane disruption. Our study highlights challenges and opportunities for the successful design of benzodiazepine- derived drug-delivery systems.
The epigenome is characterized by covalent post-translational modifications (PTMs) of histone proteins and DNA that indirectly affect transcription, DNA repair and replication.1, 2 Histone PTMs are dynamically controlled by chromatin-modifying enzymes in a highly-regulated manner. These regulations involve enzymes that apply (writers) or remove (erasers) the covalent modifications to specific aminoacids in histone proteins, as well as conserved protein interaction domains that recognize (read) and bind to those chemical modifications, which subsequently translates them in the context of chromatin reorganization and transcriptional control.3 Proteins of the bromo and extra-terminal (BET) family are among the readers of epigenetic PTMs. They recognize acetylated lysine (KAc) residues on histone tails by means of their evolutionary conserved bromodomain (BRD) modules that facilitate aggregation of transcriptional complexes to chromatin.4 Competitive displacement of BET proteins from chromatin with small-molecule inhibitors results in down-regulation of growth-promoting genes, such as the MYC oncogene, which leads to cell cycle and growth arrest, which in turn offers opportunities for therapeutic intervention in cancer.
Potent inhibitors of BET BDR proteins, such as (+)-JQ1 and I- BET762 (Compounds 1 and 6 in Fig. 1), have been reported.5, 6 The chemical scaffolds of these lead BET inhibitors have since been the subject of optimization with regards to BDR selectivity and activity.7

The excitement emerging from the preclinical success of several BET inhibitors, supported by potent antitumor activities in a wide range of genetic contexts, led to rapid clinical translation of these molecules. However, this translation has been hampered by toxicity, typically related to off-target effects.8 The targeted delivery of BET inhibitors and the development of hybrid constructs may circumvent the shortcommings of BET inhibitors. However, the feasibility of these approaches depends on the design of appropriate linkers and the tolerance of BET BDRs towards derivatized ligands. Here, we report on the design, synthesis and biological profiling of a focused set of I-BET762-derived entities that can be readily functionalized for either imaging and/or therapeutic purposes.

Fig. 1 Structures of BET bromodomain inhibitors.

As a first step, we installed different moieties onto the eastern

region of 6. X-ray crystallography studies have shown that the

a. Instituto de Medicina Molecular João Lobo Antunes, Faculdade de Medicina,
Universidade de Lisboa, Av. Prof. Egas Moniz, 1649-028, Lisboa, Portugal. E-mail: [email protected]
b. Structural Genomics Consortium, Nuffield Department of Clinical Medicine, Oxford University, ORCRB, Roosevelt Drive, OX3 7DQ, UK. E-mail: [email protected]
c. Hovione Farmaciência SA, Sete Casas, 2674-506, Loures, Portugal.
d. Department of Chemistry, University of Cambridge, Lensfield Road CB2 1EW, Cambridge, UK. E-mail: [email protected]
Electronic Supplementary Information (ESI) available: Methods, supplementary data and figures. See DOI: 10.1039/x0xx00000x

amide moiety of 6 sits in a solvent-exposed region in BET BDRs, which motivated the exploitation of that region for the installation of chemical handles. Chemical syntheses for the required chemical matter were performed through established synthetic procedures (Fig. 2).9 In short, a propyl (8) or polyethylene glycol (9–11) side chains were installed through amide bond formation chemistry prior to tethering of fluorescein to afford compounds 14–16.
This journal is © The Royal Society of Chemistry 20xx J. Name., 2013, 00, 1-3 | 1


Fig. 2 Synthesis of I-BET762 derivatives equipped with different spacers and a fluorophore. HBTU = 2-(1h-benzotriazole-1-yl)- 1,1,3,3-tetramethyluronium hexafluorophosphate. DIPEA = N,N- diisopropylethylamine. THF = tetrahydrofuran. TFA= trifluoroacetic acid. NHS = N-hydroxysuccinimide. DMF = N,N-dimethylformamide.

First, by focusing exclusively on the influence of the spacer’s addition, we measured its in vitro affinity effect towards BET proteins. By using thermal shift assays we measured temperature shifts (ΔTm) for 11 against all BRD modules of the human BET family, which were comparable, albeit lower, to those observed for JQ1 and the parent chemotype I-BET762 (Fig. 3a. Table S1). Notably no binding outside the BET family was observed. This was further validated in isothermal titration calorimetry experiments against the first and second BRDs of BRD4 [BRD4(1) and BRD4(2), respectively], against which the compound exhibited dissociation constants (KD) values of 53 and 50 nM, respectively (VFieiwg.A3rtbic.leTOanblinlee S2), which are comparable to those repoDrtOeId: 10fo.1r0369./6CT9oCCf0u5r0th54eJr establish that the added linkers would not interfere with target binding, but would allow for the introduction of a fluorescent label, we determined the X-ray structure of BRD4(1) in complex with 11 at high resolution. The compound was well resolved in the density map and engaged the protein in the typical KAc-mode with the triazolo group sitting within the cavity and interacting with Asn140 through a hydrogen bond. The classic network of water molecules within the BRD cavity was retained (Fig. 3c and Table S4). Moreover, the protecting tert-butyloxycarbonyl (Boc) group was resolved in the density in two conformations, pointing away from the BRD cavity, and trapped between two molecules in the asymmetric unit (Fig. 3d).
Encouraged by our results, we wondered if our functionalized compounds could directly engage BET proteins in a cellular context. We used fluorescent recovery after photo-bleaching, a well- established technique for assessment of on-target engagement, previously used on BET proteins.5, 10 By using parental chemotype 6 as positive control we confirmed that a functionalized entity, such as 11, can engage full length BRD4 in cells and displace it from chromatin. Although the parental scaffold readily displayed rapid recovery (t1/2 = 7.4 seconds) after photo-bleaching, which suggests direct target engagement, 11 did not elicit a similar effect at 1 μΜ (1 h treatment. Fig. S1). However a 5 μM, overnight treatment resulted in increased recovery (t1/2 = 4.9 and 5.7 seconds for 6 and 11,
respectively. Fig. 4).

Fig. 3 In vitro validation of 11 binding. a. Thermal shift (ΔTm) data with BET and other BRDs; b. Isothermal titration calorimetry in solution binding with BRD4(1) and BRD4(2); c. Structural engagement of BRD4(1) by 11; d. Packing of 11 between two BRD4(1) molecules.

After having established that the installed linker does not perturb binding to BET BDRs, we explored how our fluorescent analogues behave in vitro (14–16. Fig. 2). We evaluated the activity of these chemotypes against BRD2(1), BRD3(1), BRD4(2) and BRD4(1) in AlphaScreen™ assays (Fig. 5). All compounds inhibited the target proteins despite the presence of a bulky substituent, such as fluorescein. The inhibitory potency data for BRD4(1) and

BRD4(2) proteins were the most similar. PEGylated probes 15 and 16 displayed a marked decrease in potency against BRD4(1) relative to 6 (Fig. 5) with an IC50 value of 5.2 μM (12-fold decrease) and 2.8 μM (7-fold decrease), respectively. This observation may be explained with an entropic penalty that results from the installation of flexible side chains in 6.11 Interestingly, compound 14 that comprises a simple propyl chain as spacer inhibited BRD4(1) with

2 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Fig. 4 Cellular on-target engagement of 11. a. Fluorescent recovery after photo-bleaching of GFP-FL-BRD4 in U2-OS cells after overnight treatment with 5 μM of 11. b. Quantitative comparison of time to half-maximal fluorescence recovery for BRD4 fluorescent recovery after photo-bleaching shown in a. Data represent means ± SEM (n= 20) and are annotated with p values obtained from a two-tailed t test (**** p < 0.0001).

high potency (IC50 = 0.85 μM). Analysis of the inhibitory potency of compounds across the different proteins showed no correlation to spacer composition. While 14 was 16-times more potent (IC50 = 0.24 μM) than 6 (IC50 = 3.9 μM) against BRD2(1), the same compound was unable to induce competitive peptide displacement in BRD3(1) at concentrations as high as 100 μM.

Fig. 5. Binding affinity of fluorescent-labelled probes towards the BET-proteins BRD2(1), BRD3(1), BRD4(1) and BRD4(2).

To further validate our fluorescent-based probes in a cellular system, we probed the intracellular localization of the fluorescein- labelled derivatives. LNCaP cells have been shown to be responsive to the inhibition of BET BDRs and particularly to 6.12, 13 We therefore used LNCaP cells as an in vitro model system for this exploratory assay. In a preliminary flow-cytometry experiment the rate of compound internalization was measured as the total number of cells that incorporated fluorescein. After 2 h treatment with 10 μM of 16, only 0.8% of the total cells were positive for fluorescein. When the dose was increased to 50 μM, 36.6% of the population of cells was fluorescently labelled. However, at these high concentrations treatment with the fluorescein-linker that does not have a BET- targeting ligand (Fig. 6a) also resulted in 12.6% of fluorescein- positive cells (Fig. S2). This data suggests that most of the probe internalization at 50 μM was non-specific. Non-functionalized 6 (with no fluorescein) was used for normalization to account for possible effects in cells induced by compound treatment.
As an orthogonal approach, imaging experiments with confocal microscopy confirmed our previous findings. LNCaP cells were incubated in the presence of 10 μM of 16 and 17 for

2 h before images were captured. The amount ofVieinwtAertricnleaOlinzleinde dye was minimal in 16-treated cells, yet gDreOaI:t1e0r.1t0h3a9n/Cf9oCrCc0o5n0t5r4oJl 17 without the BET BDR ligand (Fig. S3). It has been reported that fluorescein has a high rate of fluorescence quenching and low photo-stability, which could explain the poor signal observed.14 To investigate this further, the cellular membrane was disrupted with Triton X-100 detergent prior to incubation with conjugates 16 and 17. In these conditions all cells treated with 16 showed a pronounced fluorescent signal, whereas in samples treated with 17 no staining was observed (Fig. S3). In addition, the fluorescent signal of 16 was substantially higher in the nuclei in which BET-BDR proteins are localized, which suggests efficient targeting in cells. Our results fully corroborate previous findings where intracellular targets were revealed by employing BRD-directed optical probes,15–17 despite their impaired permeability.
The same trend was observed when incubating cells with the fluorescent conjugate 14, which has only a 3-atom spacer. Fluorescein derivative bearing no BET-targeting ligand (depicted in Fig. 6) showed no internalization nor nuclei co- localization, whereas the fluorescein-labelled 14 showed internalization in live LNCaP cells with an intact cellular membrane, and nuclear co-localization as expected (Fig. 6), although with a modest fluorescent intensity. Single-time point confocal images were taken after 1 and 4 h incubation with 10 μM doses and no differences were observed.
Fig. 6 Confocal microscopy analysis on the internalization rate of 17 (a–d) and 14 (e–h) in live cells after 1 h incubation with compounds (40 x magnification; blue: DAPI; Green: Fluorescein; Grey: Transmitted light).

Altogether, the cellular data of 14 and 16 excluded the possibility for fluorescein photo quenching. Moreover, the data demonstrated a lack of correlation between spacer length and the rate of cellular internalization, in agreement with the biochemical assays for inhibitory potency. To understand why fluorescent probes were effective only when cellular membranes were disrupted, we employed a parallel membrane permeability assay to study compound diffusion across the cell membrane. The Pe values (10–6 cm/sec) for derivatised compounds 16 (0.22), 15 (0.36), 14 (0.4), and 17 (0.20) were close to the detection limit in the acceptor compartment (Table S5). This result indicated that the permeability of these compounds was comparable to the detection limit of low permeable standard atenolol. Conversely, parent molecule 6 was highly permeable. In summary, our data demonstrated that


diffusion across cellular membranes of fluorescein-derivatives

3. P. Filippakopoulos and S. Knapp, Nat. Rev. Drug DisVcioewv.A, r2ti0cl1e 4O,n1lin3e,

is strongly impaired in compounds 14–16 337–356.

Given the limited efficacy of the fluorescein-based probes to enter cells by diffusion, a new strategy was envisaged with nitrobenzofurazan (NBD) derivatives instead of fluorescein. In the absence of an extracellular receptor aiding in internalization, the use of a smaller and more neutral dye like NBD could be beneficial. To this end, 4-chloro-7- nitrobenzofurazan was modified with 2-bromoethylamine and subsequently reacted with 13 after a deprotection step to remove Boc to generate 18 (see ESI). Internalization of compound 18 in LNCaP cells was then evaluated. Confocal images were taken after 1, 2 and 4 h incubation and compared with 16 that bears a fluorescein derivative. The retrieved images indicated that the fluorescent signal was modest, which suggests poor internalization of the NBD-labelled probes (Fig. S4). Moreover, no improvements in the fluorescence intensity were observed relative to 16. The use of NBD in place of fluorescein was therefore an unsuccessful strategy to enhance dye diffusion and accumulation in cells. Altogether, these results highlight the absence of a receptor mediating the internalization process.
By considering the biophysical and cell-based experiments presented herein, one can determine the appropriateness of the site of conjugation in compound 6. The systematic exploration presented agrees with structure activity relationship studies present in the literature.11, 18 Our findings also expose challenges and opportunities for forthcoming drug delivery systems based on BET protein inhibitors. Although these may be suitable per se as ligands or payloads for targeted constructs, the required derivatization and spacer length can strongly disrupt the physicochemical properties needed for cellular membrane permeation. These features are key if intracellular ligand recognition is responsible for the selectivity of the construct. In a broader context, our data highlights the importance of informed design and selection of linkers and payloads, which is often neglected.
We thank the Hovione Farmaciência (PhD studentship to R.T.) Royal Society (URF\R\180019 to G.J.L.B.), DFG (SI 2117/1-
1 to F.S.), FCT Portugal (IF/00624/2015 to G.J.L.B., CEECIND/04518/2017 to P.M.S.D.C. and CEECIND/00887/2017 to T.R., and 02/SAICT/2017, Grant 28333 to T.R.) and the Molibresib Medical Research Council (MRC grant MR/N010051/1 to P.F.) for funding. The authors thank Dr Vikki Cantrill for her help with the preparation and editing of this manuscript.

Conflicts of interest
R.T. and G.J.L.B. are co-inventors on a patent application.
Notes and references
1. P. Tessarz and T. Kouzarides, Nat. Rev. Cell Biol., 2014, 15, 703– 708.
2. P. A. Jones, J.-P. J. Issa and S. Baylin, Nat. Rev. Genet., 2016, 17, 630–641.

4. Y. Zhao, C.-Y. Yang and S. Wang, J. Med. Chem., 2013, 56, 7498–
5. E. Nicodeme, K. L. Jeffrey, U. Schaefer, S. Beinke, S. Dewell, C.-w. Chung, R. Chandwani, I. Marazzi, P. Wilson, H. Coste, J. White, J. Kirilovsky, C. M. Rice, J. M. Lora, R. K. Prinjha, K. Lee and A. Tarakhovsky, Nature, 2010, 468, 1119–1123.
6. K. J. Falkenberg and R. W. Johnstone, Nat. Rev. Drug Discov., 2014, 13, 673–691.
7. M. Pérez-Salvia and M. Esteller, Epigenetics, 2017, 12, 323–339.
8. P. Filippakopoulos, J. Clin. Oncol., 2018, 36, 3040–3042.
9. P. Filippakopoulos, J. Qi, S. Picaud, Y. Shen, W. B. Smith, O. Fedorov, E. M. Morse, T. Keates, T. T. Hickman, I. Felletar, M. Philpott, S. Munro, M. R. McKeown, Y. Wang, A. L. Christie, N. West, M. J. Cameron, B. Schwartz, T. D. Heightman, N. La Thangue, C. A. French, O. Wiest, A. L. Kung, S. Knapp and J. E. Bradner, Nature, 2010, 468, 1067–1073.
10. C.-e. A. Chang, W. Chen and M. K. Gilson, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 1534–1539.
11. O. Mirguet, R. Gosmini, J. Toum, C. A. Clément, M. Barnathan, J.-
M. Brusq, J. E. Mordaunt, R. M. Grimes, M. Crowe, O. Pineau, M. Ajakane, A. Daugan, P. Jeffrey, L. Cutler, A. C. Haynes, N. N. Smithers, C.-w. Chung, P. Bamborough, I. J. Uings, A. Lewis, J. Witherington, N. Parr, R. K. Prinjha and E. Nicodème, J. Med. Chem., 2013, 56, 7501–7515.
12. I. A. Asangani, V. L. Dommeti, X. Wang, R. Malik, M. Cieslik, R. Yang, J. Escara-Wilke, K. Wilder-Romans, S. Dhanireddy, C. Engelke, M. K. Iyer, X. Jing, Y.-M. Wu, X. Cao, Z. S. Qin, S. Wang,
F. Y. Feng and A. M. Chinnaiyan, Nature, 2014, 510, 278–282.
13. A. Wyce, Y. Degenhardt, Y. Bai, B. Le, S. Korenchuk, M. C. Crouthame, C. F. McHugh, R. Vessella, C. L. Creasy, P. J. Tummino and O. Barbash, Oncotarget, 2013, 12, 2419–2429.
14. O. Hegener, R. Jordan and H. Häberlein, J. Med. Chem., 2004, 47, 3600–3605.
15. S. Pan, S.-Y. Jang, D. Wang, S. S. Liew, Z. Li, J.-S. Lee and S. Q. Yao,
Angew. Chem. Int. Ed., 2017, 56, 11816–11821.
16. Z. Li, D. Wang, L. Li, S. Pan, Z. Na, C. Y. J. Tan and S. Q. Yao, J. Am. Chem. Soc., 2014, 136, 9990–9998.
17. X. Li, Y. Wu, G. Tian, Y. Jiang, Z. Liu, X. Meng, X. Bao, L. Feng, H. Sun, H. Deng and X. D. Li, J. Am. Chem. Soc., 2019, DOI: 10.1021/jacs.9b02738.
18. Steven G. Smith, R. Sanchez and M.-M. Zhou, Chem. Biol., 2014,
21, 573–583.

4 | J. Name., 2012, 00, 1-3 This journal is © The Royal Society of Chemistry 20xx

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>