Synthesis and Cytotoxic Activity of 1,2,4-Triazolo-Linked Bis-Indolyl Conjugates as Dual Inhibitors of Tankyrase and PI3K

A series of new 1,2,4-triazolo-linked bis-indolyl conjugates (15a–r) were prepared by multistep synthesis and evaluated for their cytotoxic activity against various human cancer cell lines. It was observed that they were more susceptible to colon and breast cancer cells. Conjugates 15o (IC50 = 2.04 μM) and 15r (IC50 = 0.85 μM) illustrated promising cytotoxicity compared to 5-fluorouracil (5-FU, IC50 = 5.31 μM) against the HT-29 cell line. Interestingly, 15o and 15r induced cell cycle arrest at the G0/G1 phase and disrupted the mitochondrial membrane potential. Moreover, these conjugates led to apoptosis in HT-29 at 2 μM and 1 μM, respectively, and also enhanced the total ROS production as well as the mitochondrial-generated ROS. Immunofluorescence and Western blot assays revealed that these conjugates reduced the expression levels of the PI3K-P85, β-catenin, TAB-182, β-actin, AXIN-2, and NF-κB markers that are involved in the β-catenin pathway of colorectal cancer. The results of the in silico docking studies of 15r and 15o further support their dual inhibitory behaviour against PI3K and tankyrase. Interestingly, the conjugates have adequate ADME-toxicity parameters based on the calculated results of the molecular dynamic simulations, as we found that these inhibitors (15r) influenced the conformational flexibility of the 4OA7 and 3L54 proteins.


Introduction
Colorectal cancer (CRC) is the third leading malignancy and one of the major causes of cancer-related mortality in developed as well as developing countries [1,2]. Around the globe, nearly 1.4 million cases and about 694,000 deaths have been reported every year [3,4]. The early diagnosis of gene mutations may help to prevent disease progression [5].
Twenty-four mutated genes, including PI3K, AKT, mTOR, KRAS, APC, GSK3β, TANKS, etc. [6,7], are mainly responsible for colorectal cancer. Among them, PI3KCA and glycogen synthase kinase 3β (GSK3β) are the key elements of the β-catenin destruction complex associated with major cancer cases [8,9]. The PI3K/Akt and Wnt/β-catenin pathways were found to be responsible for the majority of colon cancers [10]. It is evident from the literature that there is a lack of PI3K and tankyrase dual inhibitors, and thus, there has been immense interest in the development of such dual inhibitors. Therefore, in view of the biological effects of 1,2,4-triazole and indole moieties for the inhibition of the tankyrase and PI3K/Akt pathways, some new 1,2,4-triazolo-linked bis- On the other hand, NVP-BKM120 (7), a quinazoline derivative is a promising PI3K/p-Akt inhibitor, as it decreases the cellular level of PI3K kinase [24]. Very few compounds, such as macrolactin A, were found to act as dual inhibitors of both the tankyrase and PI3K/Akt pathways [25] via reducing the nuclear β-catenin levels (tankyrase) and TCF/LEF transcriptional activity (in-vitro). The mouse model studies proved the dual inhibitory effect of SMA against tankyrase and PI3K/Akt [26]. Solberg and co-workers investigated the single and combined effects of the highly specific tankyrase 1/2 inhibitor G007-LK and the pan-class I (PI3K) inhibitor BKM120, and the results suggested that the combination of these two molecules successfully works as a dual inhibitor [11].
It is evident from the literature that there is a lack of PI3K and tankyrase dual inhibitors, and thus, there has been immense interest in the development of such dual inhibitors. Therefore, in view of the biological effects of 1,2,4-triazole and indole moieties for the inhibition of the tankyrase and PI3K/Akt pathways, some new 1,2,4-triazolo-linked bis-indolyl conjugates were synthesised and evaluated for their cytotoxic potential. Interestingly, these conjugates exhibited dual inhibiting activity for both tankyrase and PI3K. The rational design of the target molecules is depicted in Figure 2.

Structure-Activity Relationship (SAR)
Some SAR aspects were drawn from their cytotoxic activity based on (i) the nature of the N-substitution (aromatic/aliphatic) on the 1,2,4-triazolyl ring and (ii) the nature and position of the substituent (electron-donating/withdrawing) that is present in the aromatic ring, as shown in Figure 3. Moreover, amongst the aromatic and aliphatic N-substituted 1,2,4-triazole conjugates, aliphatic-substituted compounds demonstrated better activity. The nature and degree of the hydrophobicity of aliphatic-substituted compounds also influenced the activity remarkably. Meanwhile, N-propyl (15o) and N-cyclopropyl (15r)substituted conjugates were the most active compared to N-butyl, N-ethyl, N-isopropyl, N-cyclohexyl, N-allyl, and N-benzyl derivatives. The activity profiles of the aliphatic Nsubstituted compounds are depicted in anti-proliferative activity with IC50 values ranging between 2.64 and 8.1 μM against most of the cell lines tested. The site of the substitution also influenced the activity, as was observed in the case of 2-methoxy substitution (15k), which resulted in enhanced activity, followed by 3-methoxy (15l). Moderate activity was observed in the conjugate with the 4methoxy group substitution (15i). Among the halogenated compounds, the influence of the substituent on biological activity was observed as Br > Cl > F. Among all the compounds, it was observed that conjugates 15r and 15o showed the most promising cytotoxic activity, and thus, they were selected for detailed biological studies.

Cell Cycle Analysis
Among the tested compounds, conjugates 15r and 15o showed promising anti-proliferative activity against the HT-29 human cancer cell line with IC50 values of 0.85 μM and 2.04 μM, respectively, compared to the standard drug 5-FU (IC50 value of 5.31 μM). The effect of these compounds on the progression of cell cycle analysis on the HT-29 cancer cell line was studied. Their profiles in different cell cycle phases were examined using flow cytometry. The results displayed in Figure 4 show that the cell population increased in the G0/G1 phase, wherein the 15r and 15o were inhibited at 65.495% and 65.191%, respectively, at 1 μM and 2 μM concentrations, thus showing the most prominent effect. These conjugates provoked cell cycle arrest at the G0/G1 phase, thereby suggesting their ability to bind to DNA in addition to inhibiting replication in the HT-29 cancer cell line. Similarly, among the N-aryl substituted compounds, the nature and site of the substituent on the aromatic ring significantly influenced the biological activity. Among the N-aryl-substituted compounds, the molecules with methoxy substitution demonstrated enhanced activity when compared to the other substituents. Conjugate 15f, with 3,4,5trimethoxy benzene-substituted-1,2,4-triazole, also demonstrated an interesting profile of anti-proliferative activity with IC 50 values ranging between 2.64 and 8.1 µM against most of the cell lines tested. The site of the substitution also influenced the activity, as was observed in the case of 2-methoxy substitution (15k), which resulted in enhanced activity, followed by 3-methoxy (15l). Moderate activity was observed in the conjugate with the 4-methoxy group substitution (15i).
Among the halogenated compounds, the influence of the substituent on biological activity was observed as Br > Cl > F. Among all the compounds, it was observed that conjugates 15r and 15o showed the most promising cytotoxic activity, and thus, they were selected for detailed biological studies.

Cell Cycle Analysis
Among the tested compounds, conjugates 15r and 15o showed promising anti-proliferative activity against the HT-29 human cancer cell line with IC 50 values of 0.85 µM and 2.04 µM, respectively, compared to the standard drug 5-FU (IC 50 value of 5.31 µM). The effect of these compounds on the progression of cell cycle analysis on the HT-29 cancer cell line was studied. Their profiles in different cell cycle phases were examined using flow cytometry. The results displayed in Figure 4 show that the cell population increased in the G 0 /G 1 phase, wherein the 15r and 15o were inhibited at 65.495% and 65.191%, respectively, at 1 µM and 2 µM concentrations, thus showing the most prominent effect. These conjugates provoked cell cycle arrest at the G 0 /G 1 phase, thereby suggesting their ability to bind to DNA in addition to inhibiting replication in the HT-29 cancer cell line.

Mitochondrial Membrane Potential (MMP) Analysis
Conversion in the mitochondrial membrane potential is one of the early and main features of cells undergoing programmed cell death. Therefore, to assess whether 15r and 15o were engaged in damaging the integrity of the mitochondrial membrane, MMP analysis of these conjugates was performed using the JC-1 dye by flow cytometry in the HT-29 colon cancer cell line. The cells were treated with concentrations of 1 μM and 2 μM of the test compound, respectively. It was observed that there was a significant increase in the monomer formation, depicting the disruption of the mitochondrial membrane potential and a transition in the polarisation of the mitochondria by these conjugates. In contrast, the control cells had negative and intact MMP, Figure 5. Overall, the results show promising effects on the disruption of the mitochondrial potential, suggesting that the conjugates actively act on the mitochondria and reduce the Δψm loss.

Measurement of Cellular Apoptosis in HT-29 Cells
Targeting apoptosis in cancer treatment has been one of the most promising nonsurgical therapeutic strategies in curbing cancer. The loss of apoptosis control was shown to increase cancer cell survival and prolong proliferation, thus inducing tumour progression, angiogenesis stimulation, and cell proliferation deregulation. It was observed that 15r and 15o significantly increased the early and late apoptotic populations in HT-29 cells. The percentages of early and late apoptotic cell populations significantly increased upon treatment with these conjugates at doses of 1 μM and 2 μM for 24 h compared to the control group. (1 µM) in HT-29 cells showed significant G 0 /G 1 phase arrest as indicated by the increase in G 0 /G 1 phase cell population. (B) Bar graphs depict SubG 1 , G 0 /G 1 , S, and G 2 /M cell populations. Statistical significance was determined by two-way ANOVA followed by Tukey's post-hoc analysis, where *** p < 0.001 represents control vs. 1,2,4-triazolo-linked bis-indolyl (15o and 15r) conjugates in treatment groups and 'ns' represents not statistically significant in comparison to control.

Mitochondrial Membrane Potential (MMP) Analysis
Conversion in the mitochondrial membrane potential is one of the early and main features of cells undergoing programmed cell death. Therefore, to assess whether 15r and 15o were engaged in damaging the integrity of the mitochondrial membrane, MMP analysis of these conjugates was performed using the JC-1 dye by flow cytometry in the HT-29 colon cancer cell line. The cells were treated with concentrations of 1 µM and 2 µM of the test compound, respectively. It was observed that there was a significant increase in the monomer formation, depicting the disruption of the mitochondrial membrane potential and a transition in the polarisation of the mitochondria by these conjugates. In contrast, the control cells had negative and intact MMP, Figure 5. Overall, the results show promising effects on the disruption of the mitochondrial potential, suggesting that the conjugates actively act on the mitochondria and reduce the ∆ψm loss.

Measurement of Cellular Apoptosis in HT-29 Cells
Targeting apoptosis in cancer treatment has been one of the most promising nonsurgical therapeutic strategies in curbing cancer. The loss of apoptosis control was shown to increase cancer cell survival and prolong proliferation, thus inducing tumour progression, angiogenesis stimulation, and cell proliferation deregulation. It was observed that 15r and 15o significantly increased the early and late apoptotic populations in HT-29 cells. The percentages of early and late apoptotic cell populations significantly increased upon treatment with these conjugates at doses of 1 µM and 2 µM for 24 h compared to the control group.

Effect on ROS Production
Recently, ROS generation therapies have been a prominent target for treating cancer. Agents that induce oxidative stress by increasing ROS production and inhibiting antioxidant defences have received significant attention. Accumulated ROS have been shown to disrupt redox homeostasis, causing severe damage to the cancer cells. Conjugates 15r and 15o upregulated the ROS generation in HT-29 colon cancer cells, ultimately leading to cell death. They also increased mitochondrial ROS, superoxide levels, and total ROS generation, thus showing that it is one the most prominent and effective treatment methods for the treatment of cancer ( Figure 6).

Effect on ROS Production
Recently, ROS generation therapies have been a prominent target for treating cancer. Agents that induce oxidative stress by increasing ROS production and inhibiting antioxidant defences have received significant attention. Accumulated ROS have been shown to disrupt redox homeostasis, causing severe damage to the cancer cells. Conjugates 15r and 15o upregulated the ROS generation in HT-29 colon cancer cells, ultimately leading to cell death. They also increased mitochondrial ROS, superoxide levels, and total ROS generation, thus showing that it is one the most prominent and effective treatment methods for the treatment of cancer ( Figure 6).

Immunofluorescence Analysis of TAB-182 Levels and β-Catenin Levels
Several studies have reported that inhibiting the levels of β-catenin and TAB-182 (tankyrase) could be one of the promising strategies in the treatment of a variety of cancers. Typically, cancer cells have been found to express higher levels of β-catenin and TAB-182, which aids in enormous cell proliferation in cancers. Treatment with 15r and 15o at doses of 1 μM and 2 μM was found to effectively reduce the expression levels of these proteins in HT-29 colon cancer cells. Further, the treatment was found to inhibit cancer cell progression by inhibiting the nuclear translocation of β-catenin in HT-29 cells in a dose-dependent manner (Figure 7).

Immunofluorescence Analysis of TAB-182 Levels and β-Catenin Levels
Several studies have reported that inhibiting the levels of β-catenin and TAB-182 (tankyrase) could be one of the promising strategies in the treatment of a variety of cancers. Typically, cancer cells have been found to express higher levels of β-catenin and TAB-182, which aids in enormous cell proliferation in cancers. Treatment with 15r and 15o at doses of 1 µM and 2 µM was found to effectively reduce the expression levels of these proteins in HT-29 colon cancer cells. Further, the treatment was found to inhibit cancer cell progression by inhibiting the nuclear translocation of β-catenin in HT-29 cells in a dose-dependent manner (Figure 7).

Immunofluorescence Analysis of NF-κB and PI3K-P85
NF-κB is an inflammatory marker that is expressed in high levels in cancer cells. The treatment of these conjugates (15r and 15o) was found to prominently reduce inflammatory markers, such as nuclear translocation and the expression levels of NF-κB, along with decreased expression of PI3K-P85 (phosphoinositide 3-kinase). These results suggest that these new conjugates were effective in the treatment of colon cancer and were found to curb the expression of markers that are responsible for cancer cell growth and proliferation. Similarly, F-actin staining and Phalloidin red staining showed prominent damage to the actin filaments in the HT-29 cells, further confirming the anti-proliferative activity of these conjugates (Figure 8). TAB-182 and β-catenin levels were markedly reduced when compared to the control group, as observed from bar graphs (C,D). Images were acquired using a confocal microscope at 63× and 2× magnification. Statistical significance was determined by one-way ANOVA followed by Tukey's post-hoc analysis, where * p < 0.05, ** p < 0.01, *** p < 0.001 represents control vs. 1,2,4triazolo-linked bis-indolyl conjugates (15o and 15r) in the treatment groups.

Immunofluorescence Analysis of NF-κB and PI3K-P85
NF-κB is an inflammatory marker that is expressed in high levels in cancer cells. The treatment of these conjugates (15r and 15o) was found to prominently reduce inflammatory markers, such as nuclear translocation and the expression levels of NF-κB, along with decreased expression of PI3K-P85 (phosphoinositide 3-kinase). These results suggest that these new conjugates were effective in the treatment of colon cancer and were found to curb the expression of markers that are responsible for cancer cell growth and proliferation. Similarly, F-actin staining and Phalloidin red staining showed prominent damage to the actin filaments in the HT-29 cells, further confirming the anti-proliferative activity of these conjugates (Figure 8).

Western Blotting
The effects of 15o and 15r on the expression levels of PI3K p85, β-actin, AXIN-2, TAB-182, and β-catenin were determined by Western blotting. The phosphorylation levels of the PI3K p85 molecule, TAB-182, and β-catenin were significantly inhibited at concentrations of 1μM and 2μM of these conjugates compared to the control. However, no change was observed in the expression levels of the PI3K p85 molecule, TAB-182. Furthermore, it was found that these conjugates reduced the expression levels of the cell proliferation marker β-catenin pathway, as shown in Figure 9. Hence, compounds 15o and 15r inhibited the key markers related to cell proliferation via the β-catenin pathway in colorectal cancer. Treatment with different 1,2,4-triazolo-linked bis-indolyl conjugates (15a-r) significantly reduced the expressions of NF-κB in the nucleus and PI3K p85 levels in HT-29 colon cancer cells. Images were acquired using a confocal microscope at 63× and 2× magnification. (C,D) Statistical significance was determined by one-way ANOVA followed by Tukey's post-hoc analysis, where *** p < 0.001 represents control vs. 1,2,4-triazolo-linked bis-indolyl conjugates (15o and 15r) in the treatment groups.

Western Blotting
The effects of 15o and 15r on the expression levels of PI3K p85, β-actin, AXIN-2, TAB-182, and β-catenin were determined by Western blotting. The phosphorylation levels of the PI3K p85 molecule, TAB-182, and β-catenin were significantly inhibited at concentrations of 1µM and 2µM of these conjugates compared to the control. However, no change was observed in the expression levels of the PI3K p85 molecule, TAB-182. Furthermore, it was found that these conjugates reduced the expression levels of the cell proliferation marker β-catenin pathway, as shown in Figure 9. Hence, compounds 15o and 15r inhibited the key markers related to cell proliferation via the β-catenin pathway in colorectal cancer.

Molecular Docking Studies
The immunofluorescence assay of the most potent conjugates suggests that this c of compounds inhibited the expression of tankyrase and PI3K levels. Therefore, it considered of interest to perform in silico docking studies for these newly synthesi conjugates against tankyrases and phosphoinositide 3-kinase (PI3K) proteins using Sch dinger software (Schrödinger's, LLC, New York, NY, USA) with PDB ID's 4OA7and 3L The docking scores for this series of compounds are illustrated in Table 2

Molecular Docking Studies
The immunofluorescence assay of the most potent conjugates suggests that this class of compounds inhibited the expression of tankyrase and PI3K levels. Therefore, it was considered of interest to perform in silico docking studies for these newly synthesised conjugates against tankyrases and phosphoinositide 3-kinase (PI3K) proteins using Schrödinger software (Schrödinger's, LLC, New York, NY, USA) with PDB ID's 4OA7and 3L54. The docking scores for this series of compounds are illustrated in Table 2. Conjugates 15o and 15r show better docking scores against both of the targets (tankyrase and PI3K) and are in correlation with the results of the anti-proliferative activity. The docking poses of the most potent molecules of 15o and 15r were investigated, and the 2D and 3D docking poses are shown in Figure 10 and Figure S5 (ESI).
In PDB ID 4OA7, both non-covalent hydrophobic and hydrophilic interactions were observed between tankyrase protein and these compounds. Conjugate 15r demonstrated hydrophilic interactions with TYR 1203 and ASP 1198 amino acid residues through hydrogen bonding with a bond length of 2.3 Å. In addition, 15r showed Pi-Pi stacking with HID 1201 amino acid residues. It also showed hydrophobic interactions with TYR 1213, LE 1212, ALA 1210, MET (Figure 10). Conjugates 15o and 15r were superimposed on both PI3K and tankyrase target binding groves. The phenyl group of the indole ring, triazole NH, and indole NH showed interactions with the ligand present in the protein shown in Figure 10. The docking scores of 15o and 15r were −9.614 kcal/mol and −9.223 kcal/mol, respectively, in tankyrase (PDB ID 4OA7), and −6.833 kcal/mol and −6.196 kcal/mol in PI3K (PDB ID 3L54).
The docking results of these compounds are in correlation with the results of the anti-proliferative activity. Conjugate 15r, with N-cyclopropyl substitution, demonstrated an enhanced anti-proliferative activity profile compared to conjugate 15o with N-propyl substitution, which can also be explicable by the difference in the orientation of their docking poses, as 15r exhibited more hydrophilic interactions with the tankyrase protein and more hydrophobic interactions with the PI3K protein because 1,2,4-triazole with N-substituted groups are more favourable for fitting at the hydrophobic groves. The superimpositions of the docking poses are shown in Figure 10.

Results of In Silico ADME Studies
Generally, a number of drug conjugates have poor ADME profiles, which contribute to their failure in clinical trials. Therefore, a molecule must have a favourable pharmacokinetic profile and biological activity in order to be deemed a potential therapeutic candidate. There are various in silico tools available for predicting the pharmacokinetics of a molecule. Herein, Lipinski's "Rule of Five" was used to evaluate the "drug-likeness" metrics. All parameters for the tested molecules were well within the prescribed range. The QikProp results suggest that the identified hits have drug-like physico-chemical properties, as depicted in Table 3.

Molecular Dynamic Simulation Studies
The docking complexes of the most active compound 15r, with the 4OA7 and 3L54 proteins, were subjected to molecular dynamics for a period of 10 nanoseconds by using Schrödinger's software. The study explored the possible key interactions between the ligand 15r with the 4OA7 and 3L54 proteins.
During the simulation, a frame was captured every 10 ps and saved into a trajectory. Overall, around 1000 frames were generated throughout the simulation exercise. The root mean square deviation (RMSD) for the protein and ligand was computed by aligning the structures generated during MD simulation in the trajectory with the initial frame. Figure 11A shows the RMSD for the ligand (15r)-protein (4OA7) complex, and it is quite evident that the complex was stable for the whole simulation period. However, slight drifts were observed at 5 ns and 7 ns. Later, it stabilised towards the end of the simulation at 1.5Å. Similarly, Figure 11B shows the RMSD for the 15r-3L54 complex, wherein a slight drift can be observed at 5-6 ns, with stabilisation at 0.9 Å. The root mean square fluctuation (RMSF) was used to analyse the conformational changes occurring along the protein side chain ( Figure 11C and 11D). Flexibility within the ranges of 0.5 to 3.5 Å and 0.5 to 3.7 Å can be interpreted from the RMSF data of the proteins 4OA7 and 3L54, respectively. The protein-ligand interactions (interaction of 15r with PI3K and tankyrase proteins) were also monitored throughout the simulation study, and an analysis report is depicted in Figure 11E and 11F. In both cases, hydrophilic and hydrophobic interactions were observed between the ligand and target proteins. Compound 15r showed a strong hydrogen bonding with TYR 1203 of 4A07 and as well as with TYR 867 of 3L54. The binding energy profiles of compound 15r with 4OA7 and 3L54 were ascribed from various components, such as the RMSD, radius of gyration (rGyr), intramolecular hydrogen bonds (intraHB), molecular surface area (MolSA), solvent accessible surface area (SASA), and polar surface area (PSA), as given by ESI.

General
All of the reagents, chemicals, and antibodies utilised in this study were procured from GLR, Merck (India), and Sigma Aldrich, as analytical reagent (AR) grades. Thinlayer chromatography (TLC) plates made of 0.25 mm silica gel were used to monitor the reactions. In the UV cabinet, UV light was employed for the visualisation spots in the reaction mixture. The 1 H and 13 C nuclear magnetic resonance (NMR) spectra were determined using the Bruker NMR (400 MHz and 100 MHz) spectrometer. Chemical shifts are expressed as ppm against the TMS internal reference, and the spectra were interpreted using TopSpin software. Using Agilent mass spectrometry, the mass spectra, including

General
All of the reagents, chemicals, and antibodies utilised in this study were procured from GLR, Merck (India), and Sigma Aldrich, as analytical reagent (AR) grades. Thin-layer chromatography (TLC) plates made of 0.25 mm silica gel were used to monitor the reactions. In the UV cabinet, UV light was employed for the visualisation spots in the reaction mixture. The 1 H and 13 C nuclear magnetic resonance (NMR) spectra were determined using the Bruker NMR (400 MHz and 100 MHz) spectrometer. Chemical shifts are expressed as ppm against the TMS internal reference, and the spectra were interpreted using TopSpin software. Using Agilent mass spectrometry, the mass spectra, including MS, were recorded using ESI-MS, whereas a Bruker ALPHA FT-IR spectrometer (Germany) was used to record the IR spectra. The automated melting point apparatus Buchi labortechnik AG 9230 was used to measure the melting points of all synthesised compounds (Switzerland). All the final conjugates prepared in this paper are new and were validated by spectral data analysis. All of the compounds synthesised were recrystallised in methanol and purified using column chromatography with basic alumina. Similarly, other reagents used for the in vitro studies included 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (>10T) reagent (Cat no: M2128), normal goat serum (Cat no: 5425), anti-rabbit IgG (Cat no: 7074S), anti-mouse IgG (Cat no: 7076S).

Chemistry
3.2.1. General Procedure for Synthesis of 1H-indole-3-carbohydrazide (10) To the ethanolic (absolute alcohol, 50 mL) solution of 10 g of ethyl 1H-indole-3carboxylate (9), 30 mL of hydrazine hydrate was added, and the reaction mixture was refluxed for 3 h. After the completion of the reaction, observed by TLC, the reaction mixture was slowly brought to ambient temperature and poured onto crushed ice. The white precipitate that formed was filtered off under vacuum to obtain a pure white solid; 98% yield; 1H NMR (500 MHz, DMSO-d6) δ (ppm): 11.55 (s, 1H), 9.19

Evaluation of Mitochondrial Membrane Potential
The JC-1 staining method was used to study the effects of 15r and 15o on the mitochondrial membrane potential (∆ψm) of HT-29 cells. It was assessed using JC-1 dye, a specific mitochondrial fluorescent probe. Normally ∆ψm JC-1 forms aggregates with high red fluorescence intensity, so a loss in the ∆ψm is indicated by a decrease in the red fluorescence and an increase in green fluorescence due to the shifting of the dye from the aggregate to the monomeric form. JC-1 dye was used at a concentration of 2 µM after the seeding and treatment of HT-29 cells in a 6-well plate following incubation for 30 min in an incubator. The red/green fluorescence ratio serves as an indicator of the ∆ψm loss [33].

Evaluation of Apoptosis by Annexin V/Propidium Iodide (PI)
Annexin V is a very sensitive dye used to detect cellular apoptosis, while PI detects late apoptotic or necrotic populations that are characterised by the loss of integrity of nuclear and plasma membranes. HT-29 cells were plated in a 6-well plate, and after attaining morphology and the desired cell density, treatment was given for 24 h. Then, the cells were harvested and washed in ice-cold PBS and resuspended in annexin-binding buffer. Annexin-V-FITC and PI were used to stain the cells for 5-15 min following cell analysis by flow cytometry [34].

Evaluation of Total Reactive Oxygen Species (ROS)
HT-29 cells were plated in a 6-well plate, and after subsequent treatment, the plate was incubated with CM-H2DCFDA (2 ,7 -dichlorodihydrofluorescein diacetate), Sigma Aldrich, at a concentration of 5 µM for 20 min at 37 • C. After the incubation of the dye, the cells were washed with PBS followed by trypsinisation with 0.25% trypsin. Then, the percentage of intracellular ROS generation was measured using flow cytometry (Attune NXT, Thermo Fisher Scientific, Waltham, MA, USA), and 10,000 events were acquired for each treatment group [35].

Evaluation of Mitochondrial ROS
MitoSOX™ Red reagent was used to measure superoxide levels in live cells. Mi-toSOX™ Red is a novel fluorogenic dye that explicitly targets the mitochondrial membrane, and upon oxidation, produces red fluorescence. HT-29 cells were plated in a 6-well plate and post-treatment-incubated with 5 µM MitoSOX™ Red for 1 h at 37 • C in an incubator. Following incubation, the cells were washed and trypsinised. A total of 10,000 events were run in flow cytometry (Attune NXT, Thermo Fisher Scientific), and the mean fluorescence intensity was measured [36].

Immunocytochemistry (ICC)
HT-29 cells were plated on poly D-lysine-coated coverslips in a 6-well culture plate at a density of 2 × 10 6 cells per well. Upon treatment with test compounds, the cells were washed with PBS, fixed with 4% paraformaldehyde, and then permeabilised with 0.2% Triton X-100. The cells were blocked with 5% normal goat serum (NGS), washed, and incubated overnight with primary antibodies: β-catenin (1:200 dilution), TAB-182 (1:200 dilution), NF-κB (1:800 dilution), PI3K-P85 (1:200 dilution) at 4 • C overnight incubation. Next, the cells were washed with PBS and then incubated with secondary antibodies, namely, Alexa Flour TM 488 goat anti-rabbit IgG (H+L) ActinGreen and Phalloidin Red, which were added to the cells for 15 min following PBS washing and mounting. Subsequently, the nuclei were stained with Vectashield mounting medium for fluorescence with 4 ,6-dia-midino-2-phenylindole (DAPI) (Cat no. H-1200, Vector Laboratories, Burlingame, CA, USA). Negative control slides were prepared by the exclusion of the primary antibody. The slides were kept in a cool place until observation under oil emersion at 63× magnification with a confocal microscope [37].

Molecular Docking
All the synthesised compounds were docked against PI3K (PDB ID: 4OA7), and tankyrase (PDB ID: 3L54) protein targets using Schrödinger software. The protein data bank (PDB) was used to obtain the protein structures of the PI3K [39] and tankyrase [40]. (https: //www.rcsb.org, accessed on 1 October 2022), with a resolution of 2.301 Å, based on the best resonation, R (free) value, and the number of residues resolved. Initially, Schrödinger's protein preparation wizard panel was used to analyse the ligand. (Schrödinger's, LLC, New York, NY, USA). Proteins were present in tetrameric as well as monomeric forms in the workspace. Then, the tetrameric forms were changed into monomeric forms of the protein to later use the OPLS3 2015 force field for protein-energy minimisation in protein preparation. Then, crystallised water molecules were removed and the ligand was retained as such [41]. After protein preparation, a grid was generated around the ligand using receptor grid generation of the glide molecule and applying all the standard glide tool parameters.
Ligand preparation: All eighteen molecules were drawn in ChemDraw, along with the standard, and converted into SDF files. After entering all entities, the ligand preparation command was applied. Ligand docking occurred in this process in the receptor grid, and prepared ligands were subjected to glide XP docking using the standard protocol. After some time, the docking poses and docking scores were obtained and are shown in Figure 10 and Table 2.

ADMET Property Prediction
The QikProp module of Schrödinger software was used to predict the pharmacokinetic properties, such as the molecular weight (<500 Daltons), predicted octanol/water partition coefficient (QlogP o/w ), percentage of human oral absorption, H-bond donors (<5), and Hbond acceptors (<10), which were examined for their compliance with Lipinski's rule of five. The rule explains how molecular characteristics affect the medication pharmacokinetics in the human body, such as absorption, distribution, metabolism, and excretion (ADME). There are various in silico tools available for predicting the pharmacokinetics of a molecule. QikProp module version 5.4 of Maestro was used to calculate the molecular descriptor and predict the ADMET profile of the synthesised compounds [42].

Molecular Dynamic Simulation
Molecular dynamic (MD) simulations were carried out using Schrödinger's software. MD simulation determines the physical movements of molecules and atoms in the proteinligand molecular docking complex [43,44]. The MD simulation was performed on the docked complex of compound 15r with tankyrase1 and PI3K gamma protein (PDB ID: 4OA7 and 3L54) [45,46]. Before performing the MD simulation, the protein was critically vetted for any missing residues using the structure check wizard tool. Ligand energies were minimised using the OPLS3e force field [46]. Later, the 'Desmond' program was opened to start the system building, and the protein structure was solvated using the PI3TP water model.
The protein was placed in the centre of the orthorhombic box using a minimised volume, where the distance between any atom of the solute and the edge of the solvent box was at least 10Å. Counter-ions were added to neutralise the system, and the salt ion concentration was set to 0.15M based on the psychological strength. After adding all the parameters, the process was started using a standard protocol for system building (Desmond version). Following the energy minimisation, a Nose-Hoover chain thermostat was used to carry out the NPT equilibration at 310 K to maintain a constant temperature. Additionally, a Martyn-Tobias-Klein barostat was employed to maintain a pressure of 1 bar [47]. MD simulations were carried out with the default settings of the normal calculation method, with the following parameters: the simulation length was 10 ns, and the solvent model was an explicit method. The periodic boundary conditions were taken into consideration while performing the MD simulations to avoid edge effects. The energy of the protein-ligand complex was minimised to 0.25 kcal/mol. After the MD simulation was completed, the trajectory was examined for RMSD and RMSF plots as well as protein-ligand contacts.

Conclusions
In conclusion, a series of eighteen 1,2,4-triazolo-linked bis-indolyl conjugates (15a-r) were prepared in a multistep synthetic methodology. All the synthesised compounds were screened for their cytotoxic activity against a panel of nine different human cancer cell lines. Among them, conjugates 15a, 15b, 15d, 15i, 15l, 15f, 15h, 15k, and 15m illustrated good cytotoxic activity against different cancer cell lines. Conjugates 15o and 15r displayed the most promising activity with IC 50 values of 2.04 µM and 0.85 µM, respectively, against colorectal adenocarcinoma (HT-29). These two conjugates were found to induce cell cycle arrest at the G 0 /G 1 phase and also enhanced cellular ROS production with increased superoxide levels. They demonstrated prominent effects in the destruction of the mitochondrial membrane potential of cancer cells. They also significantly increased the early and late apoptotic cell populations, as evidenced by annexin/PI staining. Immunofluorescence assays in the latter experiments revealed crucial changes in the expression levels of proteins responsible for cancer cell growth and proliferation. Interestingly, 15r and 15o significantly reduced the expression levels of TAB-182, PI3K-P85 and also inhibited the nuclear translocation of NF-κB and β-catenin. The Western blot results show that compounds 15o and 15r inhibited the expression levels of PI3K p85, β-actin, AXIN-2, TAB-182, and β-catenin proteins, as these markers are involved via the β-catenin pathway in colorectal cancer. The above experiments revealed that these conjugates possess promising cytotoxic activity and have immense potential to become possible leads for the treatment of colon cancer.
The results of the in silico studies show that the conjugates 15o and 15r demonstrated different hydrophobic/hydrophilic interactions with the target proteins with good docking scores of −9.614 kcal/mol and −9.223 kcal/mol, respectively, in tankyrase (PDB ID 4OA7), and −6.833kcal/mol and −6.196 kcal/mol in PI3K (PDB ID 3L54). Interestingly, all the synthesised conjugates were within the ADME parameters, and thus, they seem to be druggable in nature. Moreover, the molecular dynamic simulation studies demonstrated that the RMSD value of 15r and the protein (4OA7 and 3L540) complexes did not exceed 2.0 Å, and the relative mean square fluctuation (RMSF) plot fluctuations of the amino acid residues were less than 5.0 Å, thereby indicating the stability of the protein conformation. This was followed by MM-GBSA calculation to identify the interaction pattern and strength of the interaction. The combination of MD simulation and experimental techniques is an a priori means to solve biological problems and provides an in-depth understanding of the relationship between protein structure and function. Therefore, based on the immunofluorescence assay, Western blot assay, and docking results, it appears that conjugates 15r and 15o act as dual inhibitors of tankyrase and PI3K in colorectal cancer. Therefore, this class of conjugates could serve as an excellent template for the discovery and development of tankyrase and PI3K dual inhibitors.