A Versatile and Sustainable Multicomponent Platform for the Synthesis of Protein Degraders: Proof-of-Concept Application to BRD4-Degrading PROTACs

The use of small molecules to induce targeted protein degradation is increasingly growing in the drug discovery landscape, and protein degraders have progressed rapidly through the pipelines. Despite the advances made so far, their synthesis still represents a significant burden and new approaches are highly demanded. Herein we report an unprecedented platform that leverages the modular nature of both multicomponent reactions and degraders to enable the preparation of highly decorated PROTACs. As a proof of principle, our platform has been applied to the preparation of potential BRD4-degrading PROTACs, resulting in the discovery of a set of degraders endowed with high degradation efficiency. Compared to the existing methods, our approach offers a versatile and cost-effective means to access libraries of protein degraders and increase the chance of identifying successful candidates.


■ INTRODUCTION
Protein degraders represent an attractive tool to control levels of disease-related proteins and are set to revolutionize drug discovery. 1,2 Among them, PROTACs (Proteolysis TArgeting Chimeras) have drawn attention as one of the most appealing approaches. 3 Being composed of a ligand for the protein of interest (POI) (referred to as warhead) and a recruiting moiety for the E3 ubiquitin ligase (referred to as anchor) connected through a suitable linker, PROTACs are able to hijack the ubiquitin−proteasome cascade and force the degradation of the POI (Figure 1). They effectively act as catalysts, being available for successive rounds of degradation upon completion of their cycle, thus minimizing the need for a continuous drug administration. 4 Around 100 different proteins have been targeted so far by applying the PROTAC technology, which also proved to be an effective modality for accessing the "undruggable genome". 5 However, a grand total of more than 1000 proteins might be considered as PROTACtable, offering opportunities for future efforts based on protein degraders. 6 Compared to the occupancy-driven approaches, PROTACs utilize an event-driven mechanism of action, which allows the knock down of disease-related POIs, offering an extraordinary strategy to overcome issues often experienced in classic drug discovery approaches, including resistance mechanisms (e.g., target protein overexpression and resistance mutations). 5,7−9 The great potential of the approach, which has grown exponentially in the past few years, has been confirmed by the first degraders that have reached clinical trials, including Arvinas degraders ARV-110, ARV-470, and ARV-766, 10,11 and DT2216 developed by Dialectic Therapeutic, whose structures have been recently disclosed ( Figure 2). 8,10 While the POIs targeted by PROTACs currently under clinical investigation show a high degree of heterogeneity, bromodomain and extraterminal domain (BET) proteins have emerged as a model POI in many pilot studies probing PROTACs. BET proteins include BRD2, BRD3, BRD4, and BRDT and recruit transcriptional complexes by binding to acetylated lysine residues on histones, thereby controlling genes involved in cellular proliferation and cell cycle progression. Because alterations in the regulation of activities by BET proteins, especially BRD4, 12 are associated with different inflammatory diseases and cancer, several BET degraders, including dBET1, 13 MZ1, 14 dBET21, 15,16 and ARV-771 17 ( Figure 3) bearing (+)-JQ1 18,19 as a warhead, have been developed, with the ultimate goal of finding more effective treatments compared to small-molecule inhibitors. 18−20 From a structural point of view, while a large variety of warheads have been explored to bind different POIs, independently from their mechanism of action, cereblon (CRBN) targeting ligands (i.e., immunomodulatory imide drug (IMiD)-based ligands, such as thalidomide and its structural analogues) 21 and Von Hippel-Lindau (VHL) targeting ligands (i.e., hydroxyproline-based molecules) are  the most commonly used anchors to recruit E3 ligases ( Figure  1). 4 In contrast with the limited room of maneuver on both warhead and anchor, the linker and the linkage points represent important sites of diversification and are the focus of several medicinal chemistry programs currently. Linkers of different length and chemical nature (i.e., linear aliphatic and polyethylene glycol (PEG) chains and extended glycol chains) have been investigated with the main goal to modulate the degradation efficiency induced by PROTACs. On the same note, diverse linkage points of the linker to both the anchor and the warhead have been explored so far and their chemical nature is mainly driven by the chemical feasibility of PROTACs synthesis rather than by a rational choice (i.e., amine, ether, amide moieties, Figure 1). The sites of attachment on the anchor and the warhead are critical and are usually selected by analyzing the solvent-exposed regions, to minimize the interference of the linker with the binding of the PROTAC to the E3 ligase and the POI, respectively ( Figure 1).
An optimal combination of all different PROTAC structural elements is of pivotal importance in triggering the productive formation of the ternary E3 ligase−PROTAC−POI complex and the following proteasome cascade activation. 22,23 The rational design of effective PROTACs, which require cooperativity in assembling the ternary complex (TC) and establishing adventitious favorable interactions between the POI and the E3 ligase, is still a key challenge. However, progress in structural biology, X-ray crystallography, molecular modeling, and dynamics simulations has been recently made to rationalize the formation and stability of PROTAC-mediated TC and to understand the PROTAC mechanism of action. 19,24−26 Besides degradation efficiency, drug-like properties such as aqueous solubility, metabolic stability, and cell permeability represent additional challenges to face for achieving in vivo efficacy, with far more reason because PROTACs lie outside the rule-of-five space. 27−29 Hence, massive efforts have been devoted to rationally explore the role of the different structural elements in conferring favorable in vitro ADME properties. 30 Because it has become increasingly clear that the length and chemical nature of the linker influence both the bioactivity and physicochemical properties of PROTACs, a shift from synthetically tractable linear aliphatic and PEG-based chains to more sophisticated and rigid motifs (e.g., piperazines, piperidines) is currently ongoing. 27,31,32 Despite the advances made so far, the rational conversion of a ligand for the POI into an effective degrader in vivo remains rather empirical and based on a "trial and error" approach. 33 Furthermore, the lack of reliable and economically sustainable synthetic methods represents an important caveat that necessarily limits the number of accessible compounds. The assemble of PROTACs has proven to be far from straightforward, as it involves the asymmetric diversification of the two sides of the linker and their chemoselective reactions with the two protein binding motifs. This challenge is usually achieved through orthogonal conditions or protection/deprotection sequences. Linking strategies to couple the linker with the warhead and the anchor include amide bond formation, N-and O-alkylations, nucleophilic aromatic substitution (S N Ar), acylation, and Williamson ether synthesis, transformations that are known to suffer from poor reactivity, low chemoselectivity, and lack of atom economy ( Figure 1). 34 Thus, the preparation of PROTACs often requires low-yielding and cumbersome multistep synthetic routes, makes extensive use of protecting groups, and relies on highly functionalized and costly building blocks. 35 To overcome these limitations, advances aimed at simplifying the access to protein degraders and their screening have been reported 32 and include the use of orthogonally protected bifunctional linkers, 36 solid-phase synthesis, 37 click chemistry, 38 Staudinger ligation, 39 a direct-to-biology (D2B) approach, 40 and others. 41,42 Nevertheless, none of these strategies has established itself as the ideal method so far, due to existing limitations. For instance, among the recent methods reported to ease the preparation of PROTACs, the click chemistry platform allows the preparation of both CRBN-and VHL-recruiting degraders, but the introduction of the triazole ring is well-known for leading to poorly soluble products and for its reluctance to scale-up. 43 In an effort to streamline the discovery of successful PROTACs and expand the chemical space explored so far, we report a modular synthetic platform that capitalizes on the versatility of multicomponent reactions (MCRs) to assemble heterobifunctional protein degraders ( Figure 1). focused on Ugi 44,45 and Passerini 46 reactions, as these MCRs stand out for their efficiency, versatility, high atom economy, and simple and environmentally friendly experimental procedures, where the only byproduct, when present, is one water molecule per molecule of product. As a proof of concept, we decided to use the following synthons as a model system: five different isocyanides bearing chains of different length and the thalidomide-based CRBN-recruiting anchor (1−5), a carboxylic acid (6) bearing the warhead based on (+)-JQ1, 19 formaldehyde (7) as the carbonyl compound to minimize the interference of the linker with the binding to the POI, and three different commercially available amines (methylamine 8, tritylamine 9, and piperazine 10). Scheme 1 provides an overview of the synthesized PROTACs by using the aforementioned synthons and by exploiting the reactions as discussed in further detail below.

■ RESULTS AND DISCUSSION
Preparation of Isocyanide-Based CRBN-Recruiting Anchors. While (+)-JQ1-based carboxylic acid (6), formaldehyde (7), and all the selected amines (8, 9, and 10) are commercially available, we applied a sustainable multistep protocol as depicted in Scheme 2 to access the desired thalidomide-based CRBN-recruiting isocyanides (1−5). The first step is the formylation of the proper amino alcohol (11−15) by using ethyl formate, followed by evaporation of the formylating agent. Further treatment of the N-formamide intermediate with tosyl chloride in the presence of triethylamine as base allowed us to obtain the corresponding linker (16−20) bearing both the isocyanide and the tosylate functionalities in 59−73% yield. According to a recent report by Meier et al. published in Green Chemistry, tosyl chloride represents an efficient and practical alternative to toxic and hazardous dehydrating agents such as phosphorus oxychloride in the preparation of aliphatic isocyanides. 47 This reagent proved to be suitable to our purpose, as it can simultaneously dehydrate the formamide group and convert the hydroxy function of 11−15 to an excellent leaving group. In the last step, an O-alkylation reaction takes place between derivatives 16−20 and 4hydroxythalidomide (21) in the presence of sodium bicarbonate to afford both linear aliphatic (1−3) and PEG (4 and 5) linkers in yields ranging from 46% to 85%. Notably, our optimized and mild experimental conditions avoided both the N-alkylation at the secondary nitrogen atom and the opening of the phthalimide/glutarimide ring of 21, two undesirable events that would result in the loss of E3 ligase recruitment. 48 Furthermore, all isolated isocyanides proved not to suffer from the unpleasant smell typical of these synthons, due to the high molecular weight and solid state at room temperature, and showed stability upon long-term storage. 49 Exploration of Different Amines To Probe the Warhead Linkage Point. With all isocyanide-base linkers in our hands, we turned our efforts into the variation of the amine component. We began with methylamine (8) as a model for conventional primary amines and PROTACs 22−26 (Scheme 1A) synthesized in 54−75% yields. To further reduce steric hindrance at the attachment point and minimally influence affinity of the warhead for its target protein, we then explored the use of tritylamine (9), that our laboratory had reported as an effective surrogate of ammonia, 50 especially when coupled with formaldehyde as the carbonyl component. The reaction was performed in methanol and, upon completion, the trityl group was cleaved by adding trifluoroacetic acid to give the desired PROTACs 27−31 in 55−82% yields (Scheme 1B). Finally, we carried out the split Ugi reaction, which we reported in 2006. 51 In this variant of the Ugi reaction a secondary amine is used, instead of a primary one, leading to the acylation of one nitrogen atom, while the other one is alkylated. This transformation particularly suited our purposes, as it allowed the one-pot incorporation of the piperazine ring, a privileged substructure of several PROTACs, including ARV-110 and ARV-471 ( Figure  2), 11,52 that helps to improve solubility and metabolic stability 27 and promotes the formation of stable TC. 53−55 Noteworthy is the use of piperazine as the amine component, which allows an increase in the sp 3 character and reduces the number of HBDs and HBAs that are associated with oral bioavailability improvement. 31 Piperazine-based PROTACs 32−36 were successfully synthesized in yields of 63−83% (Scheme 1C).
Due to the well-known configurational instability that affects thalidomide substructures, all PROTACs were obtained as a mixture of diastereomers, which include a racemate at the IMiD stereocenter and an enantiopure (+)-configuration at the JQ1 stereocenter. To explore the scalability of our synthetic approach, the preparation of isocyanide 3 was performed on a 6.00 mmol scale and the split Ugi reaction was scaled up to 1.50 mmol to afford 1 g of 34 in 83% yield.
Implementation of the Platform with the Passerini MCR. Very recently, Ciulli et al. have reported the bioisosteric replacement of an amide function with an ester at the linkage point between the warhead and the linker as a successful strategy to enhance cell permeability, without significantly affecting metabolic stability. 56 This report encouraged us to include the Passerini reaction 57 in our MCR platform. To this aim, we selected the isocyanide-bearing linker 2, which reacted with (+)-JQ1-based carboxylic acid 6 and formaldehyde 7 in the absence of the amine component in dichloromethane to afford PROTAC 37 in 60% yield (Scheme 1D).   In Vitro Biological Validation of PROTACs and Preliminary Assessment of Thermodynamic Aqueous Solubility and Metabolic Stability of Selected BRD4-Degrading PROTACs.
With the aim of proving the utility of our platform and the compatibility of the substructures accessible by the MCR strategy with a proximity-based degradation, the ability of our PROTAC candidates to effectively act as BRD4 degraders was tested in the human breast cancer cell line MDA-MB-231, known to express detectable levels of this protein. To this aim, BRD4 protein levels were assessed by Western blot analysis after exposing cells to each of the selected compounds. dBET1 PROTAC was used as the benchmark reference for comparison of CRBN-recruiting compounds, while (+)-JQ1 was used as negative control of protein degradation. Alternatively, MZ1 was identified as reference and cis-MZ1 as negative control for VHLrecruiting PROTAC candidates 45−47. 13 To preliminarily identify which of these molecules showed effective BRD4 degradation, we ran time-course experiments (4 up to 24 h) using 1 μM concentration of each compound (Figures 4 and 5). As shown in Figure 4A−D, CRBN-recruiting compounds 27, 28, 29, and 34 led to a decreased expression of BRD4 at all the time-points tested. Also compounds 32 and 33 were able to induce BRD4 degradation, although the onset of this effect was delayed (after 8 h). Conversely, compound 30 caused a modest and transitory degradation of BRD4 (at 8 and 16 h of treatment), as well as compound 31 (at 16 h of treatment), while none of the VHL-recruiting compounds (45− 47), with the exception of MZ1, acted as a BRD4 degrader ( Figure 5).
According to time-course experiments, compounds 27, 28, 29, 32, 33, and 34 resulted in the most active degraders, albeit with different times of action. Therefore, these PROTACs were selected for further characterization of dose−response and proteasome-dependence of BRD4 degradation. To define whether their activity is dose-dependent, MDA-MB-231 cells were treated with increasing concentrations of each compound (from 100 nM up to 10 μM) for 8 h, because at this time-point they were found effective, as shown in Figure 4. By this approach, compounds 27, 28, and 32 showed dose-dependency, with an effect peaking between 1 and 10 μM concentrations. Conversely, compounds 29 and 33 displayed a narrower window of active concentrations because degradation was observed only at 1 and 3 μM concentrations ( Figure 7). Notably, 34 was the most potent compound, was able to cause maximal degradation of BRD4 at 300 nM, and showed the well-known "hook effect". 33 To assess the dependency of BRD4 degradation on the proteasome machinery, we studied the effect of our best compounds in the presence of the proteasome inhibitor bortezomib. As shown in Figure 6, cotreatment with 5 nM bortezomib was able, after 8 h, to abolish BRD4 degradation induced by compounds 27, 28, 29, 32, 33, and 34 (all used at the 1 μM concentration), demonstrating the specificity of their molecular activity.
Of note, compound 37, which was synthesized using the Passerini reaction and is characterized by a α-acyloxy amide instead of a bis-amide, caused degradation of BRD4, although this effect was evident only after 16 and 24 h of incubation at the 1 μM concentration ( Figure 8A). Similarly to 34, when dose−  response degradation was investigated after 8 h of treatment, the maximal effect was seen at 300 nM ( Figure 8B).
α-Tubulin was used as internal control for equal loading. Degradation activity is reported below each lane as % of BRD4immunoreactive bands densitometry relative to DMSO vehicle. Western blot results shown in Figure 8 are representative of three (n = 3) independent experiments. DC 50 (nM) and D max (%) of PROTACs 27−29, 32−34, and 37 were calculated and are reported in Table 1.
Having demonstrated that the introduction of additional HBAs and HBDs does not prevent cell membrane crossing and subsequent BRD4 degradation, we were curious to preliminarily assess the aqueous solubility of structures potentially accessible through our platform and, to this aim, we selected compounds 27−29 as representative of Ugi-like products with different lengths, 34 that exemplifies the split-Ugi products and 37 as the result of the Passerini reaction. Being aware of thalidomide hydrolytic degradation at physiological pH (about 30% degradation in 3 h), 27 we determined the thermodynamic solubility for the selected compounds for 2 h to limit the permanence in aqueous media and monitored hydrolysis by high performance liquid chromatography (HPLC) analysis. At physiological pH, except for 27, all tested PROTACs showed a poor solubility (<30 μM), which is comparable with the solubility of the majority of commercial PROTACs. 30 As expected, the Passerini product 37 that has one less HBD compared to 27−29 was the least soluble. Under acidic conditions, the solubility significantly increased, especially for compound 34 that displays an ionizable piperazine ring, further confirming the utility of introducing this substructure in PROTAC design (Table 2).
Finally, we were interested in ruling out the possibility that additional amides and/or esters may represent potential soft spots with a negative impact on metabolic stability. To this aim, we evaluated the metabolic stability of the five selected PROTACs in mouse liver microsomes (MLM). Interestingly, metabolic biotransformation only occurred when the microsomal monooxygenase system was activated by NADPH, suggesting susceptibility to oxidative metabolism but not to hydrolysis. 27 Contrary to what happened in phosphate saline buffer, no significant degradation of thalidomide was observed in control incubations carried out in Tris-HCl (see Experimental Section). Concerning the active compounds of the Ugi-like products series (27−29), the stability increased depending on the linker length: indeed, the residual substrate percentage after 1 h incubation raised from 36% for derivative 27 to 88% for 29. Also PROTAC 37, synthesized by the Passerini reaction, proved to be highly stable in MLM, in accordance with previous findings on PROTACs displaying esters in plasma (Table 2). 56

■ CONCLUSIONS
In this study, we present an unprecedented MCR-based platform that streamlines the synthetic entry into the chemical space of protein degraders. By varying the amine component in the MCR, the platform allows extensive structure−activity and structure−property relationship studies around not only the linker and its length and composition but also the linkage point to the warhead that strongly influences the affinity of the PROTAC for the POI and might be responsible for metabolic instability. 27,59 When compared to the existing 2-CR methods, our approach stands out for being reliable, high yielding, versatile, protecting group-free, stereoconservative, and sustainable. While in this paper we have focused only on BRD4 degrading PROTACs, the platform might be used for targeting different POIs, provided that the warhead bearing a carboxylic group moiety is accessible. Our approach obviates the need to perform protection/ deprotection steps typically required in the synthesis of PROTACs, except for the use of tritylamine as a surrogate of ammonia. An additional feature of this strategy is given by the well-known stereoconservative nature of Ugi and Passerini reactions that is fundamental when chiral warheads and anchors are involved in the assembly of the protein degrader. 60,61 Finally, the process is characterized by a low environmental footprint. Except for the (+)-JQ1-based carboxylic acid, all the other substrates required for the preparation of both the linkers and the final degraders are inexpensive and easy to handle, in stark contrast with the cost associated with most of the synthons that are currently commercially available. 62 In this paper, we have applied the platform mainly to BRD4degrading PROTACs and, by using a model toolkit, we have demonstrated the compatibility of the substructures accessible by the MCRs and placed at the attachment point between the warhead and the linker with degradation activity, aqueous solubility, and in vitro metabolic stability, despite the presence of additional HBAs and HBDs and potential soft spots. The full characterization and development of BRD4-degrading PRO-TACs are beyond the scope of the present report; however, we have identified a cluster of CRBN-binding PROTACs (27,28,29,34, and 37) that cause a significant and proteasomaldependent reduction of BRD4 levels. At this stage, compound 34, displaying the privileged piperazine ring, and 37, endowed with an α-acyloxy amide, stand out for higher degrading activity, with DC 50 values of 60 and 62 nM, respectively, in MDA-MB-231 cells, and properties compatible with further preclinical development.
For the time being, we have only scraped the surface of the potential of the platform in the protein degraders field: the described variants have many applications and can be easily applied to other proteins for which ligands are available. Making protein degraders easily accessible to chemists at a reasonable cost must be a major objective in the time to come, and we envision that our method will help to achieve this goal. ■ EXPERIMENTAL SECTION Chemistry. Commercially available reagents and solvents were used as purchased without further purification. When needed, solvents were distilled and stored on molecular sieves. Reactions were monitored by thin layer chromatography (TLC) carried out on 5 cm × 20 cm silica gel plates with a layer thickness of 0.25 mm, using UV light as a visualizing agent. When necessary, TLC plates were visualized with aqueous KMnO 4 or aqueous acidic solution of cerium sulfate and ammonium molybdate reagent. Column chromatography was performed on flash  (Hz). Abbreviations used for multiplicity are as follows: s, singlet; d, doublet; t, triplet; q, quartet; quint, quintet; dd, doublet of doublets; dt, doublet of triplets; br, broad; m, multiplet. All PROTACs based on JQ1 structure were obtained as a mixture of diastereomers which include the racemic mixture at the IMiD stereocenter with enantiopure (+)-configuration at the JQ1 stereocenter.
The purity of selected compounds was determined by HPLC using a Shimadzu HPLC system (Shimadzu, Kyoto, Japan) equipped with a Kinetex C18 (150 × 4.6 mm, 5 μm d.p., Phenomenex Torrance, CA) and using 0.2% formic acid in water and 0.2% formic acid in acetonitrile as eluents (for further details, see Supporting Information, SI). The purity of all tested compounds is 95% or higher.
General Procedure A for the Synthesis of Isocyanides 16−20. Amino alcohol (7.50 mmol, 1 equiv) 1−5 and ethyl formate (5.00 mL) were heated under reflux for 6 h. Then the volatile was removed in vacuo. The corresponding formamide (7.50 mmol, 1 equiv) was solubilized in dry CH 2 Cl 2 (14.0 mL) under nitrogen and, after adding TEA (45.0 mmol, 6 equiv), the reaction mixture was cooled to 0°C. Then TsCl (22.5 mmol, 3 equiv) was added, and the reaction was stirred at room temperature for 5 h. The mixture was quenched with a saturated aqueous Na 2 CO 3 solution and stirred at 0°C for 30 min. Water was added, and the aqueous phase was extracted with CH 2 Cl 2 (×3). The combined organic layers were dried over sodium sulfate and evaporated. The crude product was purified by column chromatography using the eluent indicated below.
3-Isocyanopropyl 4-Methylbenzenesulfonate (16). The title compound was synthesized following the general procedure A. The crude material was purified by column chromatography using PE/ EtOAc 95:5 as eluent, affording compound 16 (1.31 g, 73%) as a yellow oil. 1 (17). The title compound was synthesized following the general procedure A. The crude material was purified by column chromatography using PE/EtOAc 95:5 as eluent, affording compound 17 (1.23 g, 65%) as a yellow oil. 1  5-Isocyanopentyl 4-Methylbenzenesulfonate (18). The title compound was synthesized following the general procedure A. The crude material was purified by column chromatography using PE/ EtOAc 9:1 as eluent, affording compound 18 (1.14 g, 57%) as a yellow oil. 1 (19). The title compound was synthesized following the general procedure A. The crude material was purified by column chromatography using PE/ EtOAc 9:1 as eluent, affording compound 19 (1.31 g, 65%) as a yellow oil. 1 (20). The title compound was synthesized following the general procedure A. The crude material was purified by column chromatography using PE/EtOAc 6:4 as eluent, affording compound 20 (1.67 g, 71%) as a yellow oil. 1  General Procedure B for the Synthesis of Thalidomide-Bearing Isocyanides 1−5. To a solution of thalidomide derivative 21 (2.20 mmol, 1 equiv) in dry DMF (6.00 mL) was added NaHCO 3 (3.30 mmol, 1.5 equiv) under nitrogen, and the reaction mixture was heated at 65°C. After 15 min, a solution of isocyanide 16−20 (2.64 mmol, 1.2 equiv) in dry DMF (500 μL) was added dropwise and the resulting mixture was stirred at 65°C overnight. The reaction mixture was diluted with CH 2 Cl 2 and washed with water (×3). The organic layer was dried over sodium sulfate and the volatile solvent was removed in vacuo. The crude product was purified by column chromatography using the eluent indicated below. (1). The title compound was synthesized following the general procedure B, starting from isocyanide 16. The crude material was purified by column chromatography using PE/EtOAc 6:4 as eluent, affording compound 1 (616 mg, 82%) as a yellow solid. 1

Journal of Medicinal Chemistry pubs.acs.org/jmc
Article amino)-2-oxoethyl)-N-methylacetamide (26). The title compound was synthesized following the general procedure C, starting from thalidomide-bearing isocyanide 5. The crude material was purified by column chromatography using CH 3 Figure S5 for NMR spectra of compound 26.

Synthesis of Isocyanide-Based Library of CRBN-Recruiting Anchor 3 on a Multigram Scale.
To a solution of thalidomide derivative 21 (1.65 g, 6.00 mmol, 1 equiv) in dry DMF (25.5 mL) was added NaHCO 3 (756 mg, 9.00 mmol, 1.5 equiv) under nitrogen, and the reaction mixture was heated at 65°C. After 15 min, a solution of isocyanide 18 (1.92 g, 7.20 mmol, 1.2 equiv) in dry DMF (1.50 mL) was added dropwise and the resulting mixture was stirred at 65°C overnight. The reaction mixture was diluted with CH 2 Cl 2 and washed with water (×3). The organic layer was dried over sodium sulfate, and the volatile solvent was removed in vacuo. The crude product was purified by silica gel column chromatography using PE/EtOAc 5:5 as eluent, affording compound 3 (1.88 g, 85%) as a white solid.
Cell Culturing and Western Blot Analysis. The human breast adenocarcinoma MDA-MB-231 cell line was cultured in DMEM (Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS), 1% glutamine, and 1% penicillin/streptomycin. For Western blot analysis, 15 × 10 5 cells/well were plated into 12-well plates containing DMEM complete growth medium, to be treated with the synthesized compounds, at the indicated concentrations, for different time lengths. After treatments, cells were scraped and lysed to obtain protein extract using RIPA lysis buffer supplemented with protease (PIC, Merck-Millipore) and phosphatase inhibitors (NaF 1 M, Na 3 VO 4 1 M, PMSF 100 nM, Sigma-Aldrich). Cells lysates were centrifuged at 13 000 rpm for 10 min at 4°C to remove membranes and insoluble fractions. Protein quantification of the supernatant fraction was performed with Braford protein assay (Sigma-Aldrich). Thirty milligram aliquots of protein extracts were used for 12% SDS-PAGE. After transfer into nitrocellulose membranes, BRD4 and tubulin protein levels were evaluated by immunoblotting using a rabbit polyclonal anti-BRD4 (Bethyl Laboratories) and a mouse monoclonal α-tubulin (Sigma-Aldrich) IgG as primary antibodies and an antirabbit or antimouse HRP-conjugated secondary antibody (Bio-Rad Laboratories). Protein levels were detected by chemiluminescent conversion of the HRP substrate LumiGLO (Thermo Fisher Scientific) using the ChemiDoc imaging system (Bio-Rad). Band densitometry was assessed, normalized to α-tubulin immunoreactive bands, and reported as percentage of the 0.3% DMSO control lane.
Thermodynamic Solubility. Saturated solutions were prepared by dissolving the tested compounds in 0.1 M phosphate-buffered saline (pH = 7.4) or 0.01 N HCl solutions and sonicated for 30 s. After equilibration on an orbital shaker at 25°C for 2 h, the solutions were centrifuged at 13 000 rpm for 5 min and filtered through a 0.2 μm regenerated cellulose (RC) syringe filter. A 400 μL aliquot of each filtrate was diluted with 100 μL of acetonitrile and analyzed by LC-UV. When the measured peak area was out of the calibration range (5−100 μM), the sample was further diluted in water−acetonitrile 1:1 mixture.
Metabolic Stability. Mouse liver microsomes (MLM) (pooled male mouse CD-1, protein concentration: 20 mg/mL) were purchased from Corning B.V. Life Sciences (Amsterdam, The Netherlands). The standard incubation mixture (250 μL final volume) was carried out in a 50 mM Tris (tris[hydroxymethyl]aminomethane) buffer (pH 7.4) containing 150 mM KCl, 1.5 mM, 3.3 mM MgCl 2 , 1.3 mM NADPNa 2 , 3.3 mM glucose 6-phosphate, 0.4 units/mL glucose 6-phosphate dehydrogenase, acetonitrile as cosolvent (1% of total volume), and the substrate (50 μM). After pre-equilibration of the mixture, an appropriate volume of MLM suspension was added to give a final protein concentration of 1.0 mg/mL. The mixture was shaken for 60 min at 37°C. Control incubations were carried out without the presence of an NADPH-regenerating system or microsomes. Each incubation was stopped by the addition of 250 μL of ice-cold acetonitrile, vortexed, and centrifuged at 13 000 rpm for 5 min. The supernatants were analyzed by LC-UV.

* sı Supporting Information
The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jmedchem.2c01218. 1 H and 13 C NMR spectra of final compounds; LC-UV methods for determination of purity of selected active compounds, aqueous solubility, and metabolic stability; HPLC chromatograms (PDF) Molecular