New Bioactive Fused Triazolothiadiazoles as Bcl-2-Targeted Anticancer Agents

A series of 3-(6-substituted phenyl-[1,2,4]-triazolo[3,4-b]-[1,3,4]-thiadiazol-3-yl)-1H-indoles (5a–l) were designed, synthesized and evaluated for anti-apoptotic Bcl-2-inhibitory activity. Synthesis of the target compounds was readily accomplished through a reaction of acyl hydrazide (1) with carbon disulfide in the presence of alcoholic potassium hydroxide to afford the corresponding intermediate potassium thiocarbamate salt (2), which underwent cyclization reaction in the presence of excess hydrazine hydrate to the corresponding triazole thiol (3). Further cyclisation reaction with substituted benzoyl chloride derivatives in the presence of phosphorous oxychloride afforded the final 6-phenyl-indol-3-yl [1,2,4]-triazolo[3,4-b]-[1,3,4]-thiadiazole compounds (5a–l). The novel series showed selective sub-micromolar IC50 growth-inhibitory activity against Bcl-2-expressing human cancer cell lines. The most potent 6-(2,4-dimethoxyphenyl) substituted analogue (5k) showed selective IC50 values of 0.31–0.7 µM against Bcl-2-expressing cell lines without inhibiting the Bcl-2-negative cell line (Jurkat). ELISA binding affinity assay (interruption of Bcl-2-Bim interaction) showed potent binding affinity for (5k) with an IC50 value of 0.32 µM. Moreover, it fulfils drug likeness criteria as a promising drug candidate.


Introduction
The anti-apoptotic Bcl-2 protein family influences oncogenesis signalling pathways through control of the intrinsic apoptotic process. The parent Bcl-2 protein is an apoptosis suppressor that is found to be frequently overexpressed in cancer cells [1], and the therapyresistant cancer phenotype is commonly induced by the overexpression of anti-apoptotic Bcl-2. Therefore, targeting the anti-apoptotic Bcl-2 protein with small molecules that mimic the interaction of BH3 domain proteins with anti-apoptotic Bcl-2 can selectively induce apoptosis and is a promising strategy for the development of novel chemotherapeutic agents with potent anticancer efficacy [2].
Indoles are nitrogen-based heterocyclic scaffolds that are found in most common FDAapproved anticancer drugs with the ability to sensitize cancer cells towards apoptosis [3]. Inspired by obatoclax as a clinically useful indole-based small molecule that inhibits the Bcl-2 protein [4], we previously designed and synthesized a series of indole-based compounds with oxadiazole [5][6][7] and triazole [8] scaffolds that show potent anticancer activity. Notably, the 1,3,4-thiadiazole scaffold was found to be an integral part of the anticancer agent moiety, with a toxophoric NCS moiety that may provide desirable pharmacophoric H-bond features alongside the lipophilic character of the sulphur atom [9]. Herein, we aimed to investigate the potential effect of incorporating a bioactive rigid [1,2,4]-triazolo [3,4-b]- [1,3,4]-thiadiazole scaffold within an indole-based structure with different substitutions (Scheme 1). These designed structures with enhanced physiochemical properties and fitting within the Bcl-2 binding site contribute to our continuous efforts to discover promising new potent therapeutics to combat cancer ( Figure 1).

Compound 5k and 5l Analogues Showed Potent and Selective Growth-Inhibitory Anticancer Activity on Bcl-2-Expressing Cell Lines
All the tested compounds (4, 5a-l) were found to inhibit the growth of the Bcl-2expressing human cancer cell lines (MDA-MB-231, HeLa and KG1a) with low micromolar IC 50 values, except 5c, which was inactive against the KG1A (Acute Myeloid Leukemia) cell line ( Table 1). The Bcl-2-inhibitory natural product gossypol was used as a positive control. In general, test compounds showed weak activity against the Bcl-2-negative T-cell lymphocytic leukemia cell line (Jurkat) except compounds 5a and 4. Compounds (5b, 5c, 5e, 5i, 5k and 5l) showed no activity against Jurkat cells, indicating that Bcl-2 could play a role in mediating the observed anticancer activity that later was confirmed through the ELISA binding assay. Considering the structure-activity relationships, the most active compound was 5k (containing a 2,4-dimethoxyphenyl group), followed by 5l (containing a 3,5-dimethoxyphenyl group). These observations reveal that the dimethoxyphenyl substitution was the most favourable for potent activity. In addition, the 4-methoxy 5b, 4-trifluoromethyl 5i and 4-bromo 5e substitutions showed moderately potent inhibitory activity against Bcl-2-expressing cell lines. The different phenyl substitution patterns affected the physiochemical properties, spatial arrangement and orientation within the binding active site. A selection of the most potent and moderate active and selective compounds (5b, 5e, 5i, 5k, 5l) were further investigated for their competitive Bcl-2 binding affinity (ELISA assay).

ELISA Assay Showed Potent Competitive Binding Activity of 5k and 5l
The most active compounds from the cell line assay (5b, 5e, 5i, 5k, 5l) were further evaluated according to their ability to compete with the pro-apoptotic BH3 domain Bim peptide for binding to His-tagged Bcl-2 protein, as previously described [6,11]. The results are shown in Figure 2 and Table 2. The natural product (−)-gossypol [12] was used as a positive control. Compounds (5k and 5l) were found to be more potent than positive control gossypol in the binding assay (Table 2).  Compound 5k showed stable interactions within the Bcl-2 binding site (4AQ3). The triazole nitrogen atom (N 2 ) showed two H-bonds interacting with Arg-105; the indole nitrogen showed an H-bond interaction with Glu-95; and pi-pi stacking of the phenyl group with Phe-63 was also observed, along with hydrophobic interactions with Asp-70, Phe-63 and Try-67. These interactions provided an overall glide score of −5.8 kcal/mol ( Figure 3).

ADME Predictions Confirm the Drug-Like Properties of Compounds
The new series and the most potent candidate complied with the Lipinski rule of 5, including hydrogen bond donors ((HBD) < 5, hydrogen bond acceptors (HBA) < 10, calculated octanol-water partition coefficient (cLog P) < 5 and molecular mass (Mwt) < 500) with predicted high GI absorption, moderate solubility, non-permeability of BBB and good bioavailability ( Table 3). The compounds fulfil the criteria of drug likeness, indicating that they are promising drug candidates. The bioavailability radar pink area indicates the preferred properties range ( Figure 4).

R-Group Analysis
Structure-activity relationships (SAR) of the most potent and selective compounds (5b, 5e, 5i, 5k and 5l) were studied using the R-group analysis tool [13]. This analysis showed that the different structural substitution resulted in a variance in the biological activity and binding affinity. The different R-groups orientation showed dissimilar experimental binding activity with PIC 50 . The SAR heat map analysis showed different activities with different colours ranging from red to blue ( Figure 5A). The superior activity was indicated with the 2,4-dimethoxyphenyl-substituted moiety of 5k, followed by the 3,5-dimethoxyphenyl of 5l that showed that the orientation of the dimethoxy substitution enhanced the activity. The 4-methoxyphenyl and the trifluoromethylphenyl substitution showed almost similar activity. The 4-bromophenyl substitution showed the least activity. Moreover, the pharmacophoric R-QSAR analysis highlighted that the hydrogen bond acceptor (HBA) and negative ionic interactions can significantly increase activity ( Figure 5B). Therefore, the potent activity of 5k and 5l compounds is due to the dimethoxyphenyl substitution that has two oxygen-containing groups within the pharmacophoric HBA feature and negatively ionic feature, followed by one methoxy substitution. The trifluoromethylphenyl group of 5i is negatively ionic with a weak HBA feature, while 5e showed decreased activity with a weak HBA feature and less negatively ionic properties compared to the trifluoromethylphenyl substitution. Eventually, this information could be used for the future optimization of these compounds to significantly improve the anticancer activity.

Discussion
The indole-based Bcl-2 inhibitor obatoclax mesylate was used as the rational basis for the design and optimization of novel Bcl-2 inhibitors, presented in this study. Formerly, we reported the efficacy of the indole-based scaffold as potent and selective inhibitor of Bcl-2 protein [5][6][7][8]. The hybridization of indole-based structure with different heterocyclic moieties gained significant importance in the discovery of anticancer drugs. These hybrid structures showed unique physiochemical and pharmacokinetic properties. Further, hybridization of two pharmacophore scaffolds could reduce drug resistance and side effects [14]. The lipophilic property of the heterocyclic moiety enhanced its absorption and cellular uptake. The privileged heterocyclic moieties such as oxadiazole [14], triazole [15] and thiazole [11,16] possess a critical role in anticancer drug development.
We previously prepared a series of indole-based oxadiazoles with different aromatic groups [5][6][7][8]. This led to the identification of a potent indole-based oxadiazole amine analogue with 3-chloro-phenyl substitution. The 3-chloro analogue showed a significantly potent IC 50 value against Bcl-2-expressing human cancer cell lines, with a similar ELISA binding activity to gossypol as positive control [5]. Further structural optimization with different connecting groups aimed to improve the activity. SAR study revealed that direct connection of oxadiazole to the para trifluoromethyl phenyl substitution boosted the activity by two-fold more than the positive control gossypol using a Bcl-2 ELISA binding affinity assay [6].
While the indole-based triazole series were found less active compared to the corresponding oxadiazoles [8], herein, we investigated the potential role of novel indole-based compounds with rigid fused triazolo-thiazole scaffolds and different 6-position aromatic substitutions on the activity. This analysis shows that the different structural substitution resulted in a variance in the biological activity and binding affinity. The screening led to the identification of (5b, 5e, 5i, 5k and 5l) compounds as novel, selective and potent candidates. They showed a potent activity against Bcl-2-expressing human cancer cell lines, with limited inhibitory activity against a non-Bcl-2-expressing cancer cell line.
A competitive Bcl-2 ELISA assay recognizes compound 5k (3-[6-(2,4-dimethoxyphenyl)--triazolo[3,4-b]-[1,3,4]-thiadiazol-3-yl]-1H-indole) as a potent Bcl-2 inhibitor, with two-fold lower IC 50 value compared to (−)-gossypol positive control. The analysis highlights that the hydrogen bond acceptor (HBA) and negative ionic interactions significantly enhanced the activity. Hence, the potent activity of 5k and 5l could be attributed to the presence of two oxygen pharmacophoric HBA groups and the negatively ionic properties, followed by one HBA of 5b. The 5i compound has a negatively ionic fluorine atom with a weak HBA feature, while 5e compound showed decreased activity with a weak HBA feature and less negatively ionic properties.
The study presented here is focused on early design, synthesis and SAR of novel triazolothiadiazole-based heterocycles as Bcl2-inhibitory anticancer hit compounds. Whilst we provide an early indication that the most active new compounds bind to Bcl-2 (with greater potency than the positive control (−)-gossypol), we are planning to further expand our study to substantiate the wider anticancer potential and mechanistic features of these active compounds such as 5k and 5l. This further work will include screening within a wider panel of cancer cell lines, including normal cell line controls.
A future aim for study will be to examine the mode of cytostatic/cytotoxic activity of the novel inhibitors alongside their effects on cell cycle; apoptosis; autophagy, given previous reports of G 0 /G 1 arrest; and apoptosis induction for gossypol and obatoclax [17][18][19].
Finally, we note that overexpression of Bcl-2 in Bcl-2-negative cell lines may induce sensitivity to our new inhibitors. Similarly, knockdown of Bcl-2 in Bcl-2-overexpressing cell lines may induce resistance to the inhibitors. These future experiments will help to further clarify the role of Bcl-2 in mediating anticancer effects, as has been the case for related clinical agents such as venetoclax. [20][21][22][23]

Materials and Methods
All computational work was carried on the Maestro graphical user interface of Schrödinger Suite 12.7 available at (www.schrodinger.com). Assay kit was purchased from (Thermo Scientific, Rockford, IL, USA). All chemical and reagent was purchased from (Sigma-Aldrich, Gillingham, UK).

Chemistry
Chemicals and solvents employed in chemical synthesis were of analytical grade and, when necessary, were purified and dried by standard methods. Reactions were monitored by thin-layer chromatography (TLC) using pre-coated silica gel plates (Kieselgel 60 F254, BDH), and spots were visualized under UV light (254 nm). Melting points were determined using a Gallenkamp melting point apparatus and are uncorrected. Column chromatography was performed with Merck silica gel 60 (40-60 µM). Both 1H NMR and 13C NMR spectra were recorded on a Bruker Avance 500 MHz spectrometer. Chemical shifts were expressed in parts per million (ppm) relative to tetramethylsilane. Coupling constant (J) values were represented in hertz (Hz), and the signals were designated as follows: s, singlet; d, doublet; t, triplet; m, multiplet. Mass spectroscopic data were obtained through electrospray ionization (ESI) mass spectrum (Bruker MicroTOF instrument). Elemental analysis (% CHN) was run by combustion analysis through an outsourced service (Medac Ltd., Surrey, UK).

Cell Viability-CellTiter-Blue ® Assay
Human acute myeloid leukemia KG1a and acute T-cell lymphocytic leukemia Jurkat cells were cultured and maintained as previously described [5][6][7][8]. For each experiment, 15,000 cells in 0.1 mL of medium were seeded into a solid-black fluorescence 96-well plate and incubated for 24 h. The cells were treated with different concentrations of the test compounds (from 0.00001 to 100 µM) over a 10-fold dilution series and incubated for 24 h. CellTiter-Blue ® solution was added, followed by reading of the fluorescence according to our previous protocol [5][6][7][8]. IC 50 values were obtained by using GraphPad Prism 5 software (San Diego, CA, USA). For each concentration, six independent experiments were carried out.

Enzyme-Linked Immunosorbent Assay (ELISA)
ELISA assay was performed to evaluate the ability of the compounds to compete with Bim protein and inhibit its interaction with Bcl-2 protein. [5][6][7][8]. The addition of His-tagged Bcl-2 protein to a biotinylated Bim-coated plate according to our previously published protocol was followed by the addition of anti-His antibody (Abcam, Cambridge, UK) containing horseradish peroxidase enzyme (Qiagen, Crawley, UK), o-phenylenediamine (Sigma-Aldrich, UK) and hydrogen peroxide. The optical density was determined using a plate reader at wavelength of 450 nm. The experiments were carried out on three separate occasions, including both negative and positive controls. A plot of log nM concentration for each inhibitor against the percentage reduction in the Bim peptide was created, using non-linear regression curve analysis (GraphPad Prism 5, San Diego, CA, USA) and a software used to generate IC 50 for the Bcl-2 inhibition activity.

Computational Design Protein Preparation and Molecular Docking
The 3D crystal structure of human Bcl-2 (PDB ID: 4AQ3) protein was downloaded from the protein data bank (https://www.rcsb.org/ obtained on September 2021). The proteins were prepared and refined using the Protein Preparation Wizard Maestro to ensure structural correctness [19]. Crystallographic water molecules beyond 5 Å were removed.
All the missing hydrogen atoms were added at pH 7.3 for appropriate ionization, and the tautomerization state of amino acid residues and proper bond order were assigned. Next, the refining of protein structures was performed, and the water molecules with <3 hydrogen bonds to non-waters were deleted. Finally, the energy minimization was done using OPLS4 to relieve steric clashes [20]. The ligand 2D structure were converted to 3D structure using LigPrep (Schrödinger) [21]. Hydrogen atoms were added, and the salt ions were removed. The subsequent energy minimization of each structure was carried out using OPLS4 force field [20]. The ligand in the crystal structure of Bcl-2 protein was used for grid generation. A grid box was generated at the centroid of the active site for docking studies, and the active site was defined around the ligand crystal structure. Molecular docking was performed within the catalytic pocket site of the proteins using standard precision (SP) mode of Grid using Glide [22,23] without applying any constraints. The prepared ligands were docked against grid generated bcl-2 (PDB: 4AQ3) in SP flexible mode [24].

Pharmacokinetics and Drug Likeness Filter and Target Prediction
ADME properties of target compounds were predicted using Swiss-ADME server [25], using compounds in SMILES format [26]. Two different filters, Lipinski's rule of five and Ghose filter, were used to evaluate the drug-like properties.

R-Group Analysis
The R-group mapping analysis was performed in the Schrödinger suite. First, the input LigPrep structure with the associated IC 50 value was converted to PIC 50 values. The maximum common core was defined with Combi-Glide bond labelling and alignment of structure for fingerprint similarity of side-chain to minimize the number of attached R-groups. The heat map analysis displayed the effect of different functional group position with different colour ranges that reflect its pIC 50 activity. A QSAR model was generated based on pharmacophoric features such as hydrogen bond donor (D), acceptor (A), hydrophobic group (H), negatively ionizable (N), positively ionizable (P) and aromatic ring (R).

Conclusions
This study rationalized the design, synthesis and evaluation of novel series of (6substituted phenyl-[1,2,4]-triazolo-[3,4-b]-[1,3,4]-thiadiazol-3-yl)-1H-indoles (5a-l) as novel inhibitors of Bcl-2 protein. The various aromatic substitutions and spatial orientations of the (5a-l) compounds within the Bcl-2 binding pockets afforded variable activity, while the most potent compound with sub-micromolar IC50 values was (5k), containing a (2,4dimethoxy-phenyl) substitution. Moreover, molecular docking studies showed stable interactions of the (5k) compound within the BH3 domain of Bcl-2. This was confirmed with ELISA assay, which displayed a potent inhibitory IC50 value of 0.3 µM, representing a twofold greater potency than gossypol as positive control. The predicted ADME properties showed that the compounds fulfil the criteria of drug likeness and preferable properties range. This indicates that they are promising drug candidates for further development as a lead compound against a Bcl-2 anti-apoptotic target. Institutional Review Board Statement: Not applicable.

Data Availability Statement:
The authors confirm that the data supporting the finding are available within the article.