5-Hydroxyindole-Based EZH2 Inhibitors Assembled via TCCA-Catalyzed Condensation and Nenitzescu Reactions

5-Hydroxyindole derivatives have various demonstrated biological activities. Herein, we used 5-hydroxyindole as a synthetic starting point for structural alterations in a combinatorial process to synthesize 22 different compounds with EZH2 inhibitor pharmacophores. A series of 5-hydroxyindole-derived compounds were screened inhibitory activities against K562 cells. According to molecular modeling and in vitro biological activity assays, the preliminary structure-activity relationship was summarized. Compound L–04 improved both the H3K27Me3 reduction and antiproliferation parameters (IC50 = 52.6 μM). These findings revealed that compound L–04 is worthy of consideration as a lead compound to design more potent EZH2 inhibitors. During the preparation of compounds, we discovered that trichloroisocyanuric acid (TCCA) is a novel catalyst which demonstrates condensation-promoting effects. To gain insight into the reaction, in situ React IR technology was used to confirm the reactivity. Different amines were condensed in high yields with β-diketones or β-ketoesters in the presence of TCCA to afford the corresponding products in a short time (10~20 min), which displayed some advantages and provided an alternative condensation strategy.

Our group is devoted to the study and development of histone methyltransferase inhibitors. The enhancer of zeste homolog 2 (EZH2), known as the key catalyzed subunit of the polycomb repressor complex 2 (PRC2), can regulate trimethylated histone H3 lysine 27 (H3K27Me3) on chromatin, and subsequently silence tumor suppressor genes [22]. It is reported that an abnormally high expression of EZH2 was discovered in various malignant cells and other diverse biological processes [23]. In recent years, several EZH2 inhibitors (EZH2Is) have been the subject of clinical trials, summarized in Table 1 [24][25][26][27][28][29]. Table 1. Representative EZH2Is in clinical trials.

Phase I Acute myeloid and lymphocytic leukemia
Daiichi Sankyo PF-06821497 [28] Molecules 2020, 25, x FOR PEER REVIEW 2 of 23 chromatin, and subsequently silence tumor suppressor genes [22]. It is reported that an abnormally high expression of EZH2 was discovered in various malignant cells and other diverse biological processes [23]. In recent years, several EZH2 inhibitors (EZH2Is) have been the subject of clinical trials, summarized in Table 1 [24][25][26][27][28][29]. chromatin, and subsequently silence tumor suppressor genes [22]. It is reported that an abnormally high expression of EZH2 was discovered in various malignant cells and other diverse biological processes [23]. In recent years, several EZH2 inhibitors (EZH2Is) have been the subject of clinical trials, summarized in Table 1 [24][25][26][27][28][29]. In 2020, tazemetostat was the first agent in its class to receive approved by U.S. Food and Drug Administration (FDA) for the treatment of metastatic or locally advanced epithelioid sarcoma in adults and pediatric patients aged 16 years and older [25]. Tazemetostat is also being investigated as a chemotherapy agent for the treatment of several cancers, including follicular lymphoma, kidney cancer, diffuse large B-cell lymphoma, solid tumor, etc. [25]. These observations stimulated numerous EZH2 drug discovery programs and greatly encouraged us to develop more novel EZH2Is for the treatment of multiple cancer contexts.
Molecules 2020, 25, x FOR PEER REVIEW 3 of 23 In 2020, tazemetostat was the first agent in its class to receive approved by U.S. Food and Drug Administration (FDA) for the treatment of metastatic or locally advanced epithelioid sarcoma in adults and pediatric patients aged 16 years and older [25]. Tazemetostat is also being investigated as a chemotherapy agent for the treatment of several cancers, including follicular lymphoma, kidney cancer, diffuse large B-cell lymphoma, solid tumor, etc. [25]. These observations stimulated numerous EZH2 drug discovery programs and greatly encouraged us to develop more novel EZH2Is for the treatment of multiple cancer contexts. However, some agents' relatively modest potency and pharmacokinetic properties remain drawbacks for future studies as optimal EZH2Is; for example, GSK-126 is only administered by injection due to its poor pharmacokinetic properties [30]. Moreover, the single administration dosage of tazemetostat is up to 800 mg po bid [31]. Besides, the clinical therapeutic benefits of EZH2Is remain unsatisfactory and their applications are limited to only certain hematological malignancies. The abovementioned issues highlight the challenges that are associated with EZH2I therapy in the context of cancer. Therefore, the development of novel EZH2I structures could be a way to overcome these challenges. From the perspective of the structure of reported EZH2Is, while the initial structures have given invaluable insight into small molecule EZH2Is, a number of open questions remain. Most of the compounds take advantage of the structural characteristics of the pyridone moiety as an active pharmacophore that binds to EZH2 structures. This feature may serve as a boon for future inhibitor development.
Herein, based on the privileged 5-hydroxyindole scaffold and the main pyridone moiety pharmacophore, a novel series of 5-hydroxyindole-based EZH2Is has been designed and synthesized (Scheme 1), which then their antiproliferative effects against K562 cells in vitro and have been assessed and their ability to reduce cellular H3K27Me3 levels further characterized. Further a simple catalyst-1,3,5-trichloroisocyanuric acid (TCCA)-for promoting the condensation reactions of β-diketones or β-ketoesters and related amines during the preparation of compounds was accidentally discovered. Of note, TCCA has been previously reported as an inexpensive and relatively stable chlorination [32,33] and oxidation reagent [34,35] and our results therefore further expand the application scope of TCCA.  However, some agents' relatively modest potency and pharmacokinetic properties remain drawbacks for future studies as optimal EZH2Is; for example, GSK-126 is only administered by injection due to its poor pharmacokinetic properties [30]. Moreover, the single administration dosage of tazemetostat is up to 800 mg po bid [31]. Besides, the clinical therapeutic benefits of EZH2Is remain unsatisfactory and their applications are limited to only certain hematological malignancies. The abovementioned issues highlight the challenges that are associated with EZH2I therapy in the context of cancer. Therefore, the development of novel EZH2I structures could be a way to overcome these challenges. From the perspective of the structure of reported EZH2Is, while the initial structures have given invaluable insight into small molecule EZH2Is, a number of open questions remain. Most of the compounds take advantage of the structural characteristics of the pyridone moiety as an active pharmacophore that binds to EZH2 structures. This feature may serve as a boon for future inhibitor development.

Results and Discussion
Herein, based on the privileged 5-hydroxyindole scaffold and the main pyridone moiety pharmacophore, a novel series of 5-hydroxyindole-based EZH2Is has been designed and synthesized (Scheme 1), which then their antiproliferative effects against K562 cells in vitro and have been assessed and their ability to reduce cellular H3K27Me3 levels further characterized. Further a simple catalyst-1,3,5-trichloroisocyanuric acid (TCCA)-for promoting the condensation reactions of β-diketones or β-ketoesters and related amines during the preparation of compounds was accidentally discovered. Of note, TCCA has been previously reported as an inexpensive and relatively stable chlorination [32,33] and oxidation reagent [34,35] and our results therefore further expand the application scope of TCCA.

Condensation of Amines with β-Diketones or β-Ketoesters Catalyzed by TCCA
Generally, the condensation reactions of carbonyl compounds with amines are carried out via azeotropic removal of water due to their reversibility, which typically results in long reaction times, low yields and harsh reaction conditions, and these drawbacks limit their practical application. It was recently reported that some Lewis acids such as Zr(ClO 4 [36][37][38]. Herein, we systematically investigated the reaction conditions, including solvent, catalyst and molar ratio (seeing Supporting Information, Table S1). In this process, TCCA (at 2 mol%) could effectively accelerate the condensation reaction. The conversion of ethyl acetoacetate to the corresponding ethyl 3-(benzylamino)butyl-2-enoate was monitored by in situ React IR technology. As seen in Figure 2, less than 1 min after the addition of benzylamine to ethyl acetoacetate and TCCA in acetonitrile under ice bath conditions, a new sharp peak belonging to ethyl 3-(benzylamino)but-2-enoate appeared at 1647 cm −1 , and the intensity of the peak increased gradually as the reaction proceeded. Of note also is the disappearance of the 1752 cm −1 band in parallel to the appearance of the 1647 cm −1 band. We believed that the new infrared absorption at 1647 cm −1 is attributed to the formation of a C-N bonds. The conversion of ethyl acetoacetate was less than 50% without any catalyst (Figure 2A), whereas ethyl acetoacetate was completely converted in 15 minutes when the reaction was catalyzed by TCCA ( Figure 2B). Generally, the condensation reactions of carbonyl compounds with amines are carried out via azeotropic removal of water due to their reversibility, which typically results in long reaction times, low yields and harsh reaction conditions, and these drawbacks limit their practical application. It was recently reported that some Lewis acids such as Zr(ClO4)2 . 6H2O, InBr3, Bi(O2CCF3)3, etc. could catalyze this transformation under mild conditions [36][37][38]. Herein, we systematically investigated the reaction conditions, including solvent, catalyst and molar ratio (seeing Supporting Information, Table S1). In this process, TCCA (at 2 mol%) could effectively accelerate the condensation reaction. The conversion of ethyl acetoacetate to the corresponding ethyl 3-(benzylamino)butyl-2-enoate was monitored by in situ React IR technology. As seen in Figure 2, less than 1 min after the addition of benzylamine to ethyl acetoacetate and TCCA in acetonitrile under ice bath conditions, a new sharp peak belonging to ethyl 3-(benzylamino)but-2-enoate appeared at 1647 cm −1 , and the intensity of the peak increased gradually as the reaction proceeded. Of note also is the disappearance of the 1752 cm −1 band in parallel to the appearance of the 1647 cm −1 band. We believed that the new infrared absorption at 1647 cm −1 is attributed to the formation of a C-N bonds. The conversion of ethyl acetoacetate was less than 50% without any catalyst (Figure 2A), whereas ethyl acetoacetate was completely converted in 15 minutes when the reaction was catalyzed by TCCA ( Figure 2B). The reaction enjoys a wide substrate scope with respect to amine substrates, including primary aliphatic amines and aromatic amines ( Table 2). This facile method was successfully used for β-diketones or β-ketoesters, and excellent yield and short reaction times were observed in all cases. The condensation reaction between aliphatic amines and β-ketoesters catalyzed by TCCA required short reaction times of only 10-20 min (entries 1-9). A more sterically encumbered phenyl group at the β-position of the β-ketoesters, and bulky amine substrate also reacted without incident (entries 8,16). Since the aromatic amines have weaker nucleophilic activity, a long reaction time was also required (entries 3,7,8,[12][13][14][15]. Moreover, electron-withdrawing groups on the benzene ring were not beneficial to the reaction (entry 15); for example, nitro groups, resulted in longer reaction times and lower conversion rates. Having established a facile method to the synthesis of intermediates, we next turned our efforts toward the synthesis of 5-hydroxyindole-based EZH2Is. The reaction enjoys a wide substrate scope with respect to amine substrates, including primary aliphatic amines and aromatic amines ( Table 2). This facile method was successfully used for β-diketones or β-ketoesters, and excellent yield and short reaction times were observed in all cases. The condensation reaction between aliphatic amines and β-ketoesters catalyzed by TCCA required short reaction times of only 10-20 min (entries 1-9). A more sterically encumbered phenyl group at the β-position of the β-ketoesters, and bulky amine substrate also reacted without incident (entries 8,16). Since the aromatic amines have weaker nucleophilic activity, a long reaction time was also required (entries 3,7,8,[12][13][14][15]. Moreover, electron-withdrawing groups on the benzene ring were not beneficial to the reaction (entry 15); for example, nitro groups, resulted in longer reaction times and lower conversion rates. Having established a facile method to the synthesis of intermediates, we next turned our efforts toward the synthesis of 5-hydroxyindole-based EZH2Is.

Synthesis of 5-Hydroxyindole-Based EZH2Is
With the established protocol in hand, a series of 5-hydroxyindoles were obtained, as depicted in Scheme 2. Initially, compounds 3a-h were reacted with benzoquinone or naphthalene to yield 5-hydroxyindoles 4a-g via an anhydrous ZnCl 2 -catalyzed Nenitzescu indole synthesis. The fragments 6a-e were prepared from 4a by alkylation and ammonolysis. Then, esters 6a-e were hydrolyzed and acidified to afford carboxylic acids 7a-e. Meanwhile, 5-hydroxyindole 4a was transformed via alkylation with brominated alkanes in the presence of NaH into ethers 6f-h, respectively, which were hydrolyzed and acidified to afford carboxylic acids 7f-h. On the other hand, carboxylic acids 7i-o were obtained from 4a-g by direct hydrolysis and acidification. Finally, carboxylic acids 7a-o were assembled with pyridone derivatives in the presence of EDCI and HOBt to furnish the target compounds L-01~L-22. The synthesis of the pyridone derivatives is described in the Supporting Information.

Synthesis of 5-hydroxyindole-based EZH2Is
With the established protocol in hand, a series of 5-hydroxyindoles were obtained, as depicted in Scheme 2. Initially, compounds 3a-h were reacted with benzoquinone or naphthalene to yield 5-hydroxyindoles 4a-g via an anhydrous ZnCl2-catalyzed Nenitzescu indole synthesis. The fragments 6a-e were prepared from 4a by alkylation and ammonolysis. Then, esters 6a-e were hydrolyzed and acidified to afford carboxylic acids 7a-e. Meanwhile, 5-hydroxyindole 4a was transformed via alkylation with brominated alkanes in the presence of NaH into ethers 6f-h, respectively, which were hydrolyzed and acidified to afford carboxylic acids 7f-h. On the other hand, carboxylic acids 7i-o were obtained from 4a-g by direct hydrolysis and acidification. Finally, carboxylic acids 7a-o were assembled with pyridone derivatives in the presence of EDCI and HOBt to furnish the target compounds L-01~L-22. The synthesis of the pyridone derivatives is described in the Supporting Information. Scheme 2. Synthetic route to 5-hydroxyindole-based EZH2Is.

In vitro Bioactivity Assay and SAR
Following the synthesis, the antiproliferative effects of the 5-hydroxyindole-based EZH2Is on K562 cells were then evaluated by a cell counting kit-8 (CCK-8) assay using tazemetostat as positive control. The results showed that most of the target compounds exhibited a similar antiproliferative effect compared with the control (Table 3). Different lengths of aliphatic chains were introduced to Scheme 2. Synthetic route to 5-hydroxyindole-based EZH2Is.

In Vitro Bioactivity Assay and SAR
Following the synthesis, the antiproliferative effects of the 5-hydroxyindole-based EZH2Is on K562 cells were then evaluated by a cell counting kit-8 (CCK-8) assay using tazemetostat as positive control. The results showed that most of the target compounds exhibited a similar antiproliferative effect compared with the control (Table 3). Different lengths of aliphatic chains were introduced to evaluate the appropriate distance between the indole scafford and the basic functional group. The results showed that the length of chain has no obvious effect on the inhibitory activity; for instance, compounds L-01 (containing a two-carbon linker) and L-03 (containing a three-carbon linker) showed comparable IC 50 values. Additionally, it appeared that the terminal amine was essential for maintaining the activity. The hydroxy group at the C5 position of an indole substituted with an amine side chain was beneficial for the antiproliferative effect; for example, L-01 and L-06 had IC 50 values of 83.8 µM and above 100 µM, respectively. Furthermore, a hydrophobic piperidine group (compounds L-04 and L-05) had a beneficial effect on the IC 50 value compared with that of hydrophilic morpholine moiety (compounds L-01 and L-03).
As a refinement, some 5-hydroxylindoles bearing different substituents at N1 were synthesized and investigated. Thus, all other N-substituted compounds (L-10, L-12~L-15), containing phenethyl (L-10), cyclohexyl (L- 14) or (isopropyl, L-15) produced stronger inhibitory activities than benzyl (L-09) and phenyl (L-12). Particularly, it was noteworthy that the tricyclic compound L-11 showed inhibition activity with an IC 50 value of 55.2 µM. Among the different aliphatic alkanes (methyl, ethyl, propyl, isopropyl) on the pyridone ring, the results suggested a similar biochemical potency. Compounds L-20 and L-22 (with cyclohexane on the pyridone ring ) and L-18 (with a cyclopentane on pyridone ring) showed less inhibition activities, while compounds L-19 and L-21 showed medium antiproliferative effect, with IC 50 values of 68 µM and 52 µM, respectively. To correlate the K562 cell inhibition results with EZH2, we assessed the cellular H3K27Me3 levels in K562 cells after being treated with compounds L-01~L-04 and tazemetostat (Figure 3). Western blot data showed that L-01~L-04 and tazemetostat significantly reduced cellular H3K27Me3 levels in K562 cells at concentrations of 5 µM and 10 µM in a dose-dependent manner. Thus, our results provide clear evidence that these compounds could inhibit the EZH2 protein at low molarity, which is consistent with the inhibition of K562 cell growth.

Molecular Docking
We used molecular modeling to determine a binding mode of some representative inhibitors (L-04, L-19, L-22 and tazemetostat) in the binding site of EZH2 (PDB ID: 4W2R). Initially, we docked L-04 and tazemetostat to their respective binding conformations. As indicated in Figure 4A, L-04 could maintain similar binding patterns compared with tazemetostat. The main pyridone moiety pharmacophore could form a hydrogen bond with Trp 521 . The carbonyl oxygen of the amide linker warhead engages EZH2 via tight hydrogen bonds with the catalytic tyrosine residue Tyr 111 . The indole scaffold and phenyl ring form face-to-face contacts with Tyr 111 and Tyr 558 . The C5 position of the indole substituted with the amine side chain projects into the solvent region, where the protonated piperidine did not form additional hydrogen bonds ( Figures 4B and 4C).

Molecular Docking
We used molecular modeling to determine a binding mode of some representative inhibitors (L-04, L-19, L-22 and tazemetostat) in the binding site of EZH2 (PDB ID: 4W2R). Initially, we docked L-04 and tazemetostat to their respective binding conformations. As indicated in Figure

Molecular Docking
We used molecular modeling to determine a binding mode of some representative inhibitors (L-04, L-19, L-22 and tazemetostat) in the binding site of EZH2 (PDB ID: 4W2R). Initially, we docked L-04 and tazemetostat to their respective binding conformations. As indicated in Figure 4A  Therefore, the substituents at the C5 position of indole could be modified to improve the pharmacokinetic properties; for example, compared with unsubstituted compound L-09, L-04 bearing the basic chain showed lower binding energy. Small substituents such as C5 or C6-methyl or ethyl groups on the pyridone ring were well compatible, while big substituents such as 5 or 6-membered rings lost the hydrogen bond with Trp 521 , resulting in an invalid interaction mode with the EZH2 protein ( Figure 4D). From the docking score perspective, L-18 and L-20 containing 5 or 6-membered ring substituents on the pyridone ring resulted in a higher binding energy. In all, our data suggests that L-04 is an ideal template to develop improved EZH2Is using molecular modeling to rationally design new agents to treat multiple cancers.

General Information
All reagents and starting materials were obtained from commercial sources and used as received.

General Procedure for the Syntheses of Ethyl Enoates 3a-p
TCCA (714/34.8 mg, 31/0.15 mmol) was added to ethyl acetoacetate (20/1 g, 154/7.68 mmol) or tert-butyl 3-oxobutanoate (1.2 g, 7.59 mmol) or acetylacetone (0.77 g, 7.69 mmol) or ethyl benzoylacetate (1.48 g, 7.71 mmol) in acetonitrile (20 mL) under ice bath cooling, and the appropriate amine (0.23/0.012 mol) was added dropwise. The mixture was stirred under ice bath conditions and allowed to naturally warm up to room temperature for the indicated time. After the reaction was complete, the mixture was poured into water, extracted with dichloromethane (3 × 50 mL), washed with water (3 × 50 mL) and brine (50 mL). The organic solution was dried over magnesium sulfate, filtered, evaporated under reduced pressure and used in the following reaction without any further purification.     26 g, 9.25 mmol) was added to a stirred suspension of p-benzoquinone (10 g, 92.5 mmol) or naphthoquinone (14.6 g, 92.5 mmol) in dry DCM (50 mL). After heating at reflux, a solution of crotonamine (92.5 mmol) in dry DCM (50 mL) was added into the mixture over 40 min and stirred at reflux for a further 45 min. The mixture was cooled to room temperature and held at 4 • C for 30 min to allow product precipitation. The solid was filtered and washed with DCM, water and acetonitrile to give a corresponding products 4a-g.

General Procedure for Ammonolysis. Syntheses of 6a-e
To a solution of 5a (5 g, 12.0 mmol) or 5b (5. 16 g, 12.0 mmol) in acetonitrile (20 mL), KI (4.0 g, 24.0 mmol) were added, and the reaction mixture was heated to reflux for 30 min. After cooling to r.t., the amine (36 mmol) and anhydrous K 2 CO 3 (6.6 g, 48.0 mmol) were added, and the mixture was heated to reflux for the indicated time while monitoring the disappearance of starting material by TLC. The cooled mixture was evaporated under reduced pressure, and water was added to the residue. After stirring for 10 min, the solid suspension was filtered, washed with water, and dried to afford target products 6a-e.

General
Procedure for the Syntheses of O-Substituted 5-Hydroxylindoles 6f-h 5-Hydroxylindole 4a (0.8 g, 2.59 mmol) was added to a stirred suspension of NaH (60% dispersion in mineral oil, 0.20 g, 517 mmol) in DMF (5 mL) under ice bath cooling. After 20 min, a haloalkane (3.88 mmol) was added dropwise, and the mixture was allowed to react for 6 h. The mixture was poured into water, and the resulting precipitate was filtered off, washed with water, and dried to afford the intermediates 6f-h, which were used in the following reactions without any further purification.

General Hydrolysis Procedure: Syntheses of Compounds 7a-o
Carboxylate compound 4a-g or 6a-h (2.0 mmol) were suspended in a mixture of EtOH (10 mL) and 60% NaOH aqueous (30 mmol), and refluxed for 3 h. The ethanol was removed under vacuum, and the mixture was cooled and acidified. The solid was filtered and dried to afford intermediates 7a-o.

General Procedure for the Syntheses of the Target Compounds L-01~L-22
To a solution of indole derivatives (1 mmol), EDCI (0.29 g, 1.5 mmol), HOBt (0.20 g, 1.5 mmol) in dry dichloromethane (10 mL) Et 3 N (0.76 g, 5 mmol) was added and the mixture was stirred at room temperature for 30 min. Then, the corresponding pyridone derivatives (1 mmol) was added, and the resulting mixture was reacted at room temperature for indicated time while monitoring by TLC. When the reaction was complete the reaction mixture was poured into water, extracted with dichloromethane (15 × 3 mL), and washed with water (3 × 15 mL) and brine (15 mL). The organic solution was dried with anhydrous sodium sulfate, filtered and evaporated under reduced pressure. The solid residue was purified by column chromatography on silica gel.

React IR Experiment
The ReactIR 15 DiComp probe was inserted in a 50 mL, three-neck flask equipped with a magnetic stirrer. IR spectra were obtained every 15 s. Data collection began at the start of the experiment. The solution of ethyl acetoacetate (1.0 g, 7.68 mmol) and TCCA (34.8 mg, 0.15 mmol) in MeCN (5 mL) was cooled to 0 • C, and benzylamine (0.82 g, 7.68 mmol) was added dropwise. The resulting solution was stirring under ice bath and naturally elevated to at room temperature for indicated time. After the reaction was completed, the workup was consistent with the general procedure 3.1.1.

Cell Growth Inhibition Assay
The human chronic myeloid leukemia cell line K562 was purchased from the American Type Culture Collection (ATCC, Perry Pkwy, Gaithersburg, USA). The cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 1% penicillin and streptomycin at 37 • C in a humid atmosphere containing 5% CO 2 in air. In the cell growth experiments, cells were seeded with 5000 cells/well density in the culture medium containing 100 µL diluent series of test compounds (100, 10, 1, 0.1 µM) in 96 well cell culture plates. After treatment for 48 hours, cell growth was measured by WST-8 assay based on lactate dehydrogenase using the Spectramax Paradox Multimode detection platform. The CCK-8 reagent was added to each well of 10 µL, and the cells were incubated for another 1-4 hours and read at 450 nm. The readings were standardized relative to the cells treated by the carrier, and IC 50 was calculated by nonlinear regression analysis using SPSS statistics 20.0 software (IBM, Chicago, Michigan, USA).

Western Blot Analysis
The cells were treated with test compounds (10 and 5 µM), and the extract was prepared by adding radio immunoprecipitation assay (RIPA) lysis buffer to the cells under ice bath. The protein concentrations were determined by the bicinchoninic acid (BCA) assay. Protein samples were separated by sodium lauryl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene fluoride (PVDF) membrane. Cell membranes were blocked with 5% milk for 1 hour at room temperature and overnight at 4 • C with designated primary antibodies. After washing with tris buffered saline Tween (TBST) three times for 30 min, incubated with horseradish peroxidase-labeled secondary antibody for 1 h at room temperature, and then washing with TBST three times for 30 min to enhance chemiluminescence. The band intensity in Figure 3 was quantified by the Odyssey software (Li-Cor, Lincoln, Nebraska, USA) and histone H3 was used as a housekeeping protein for normalization.

Molecular Docking
We processed EZH2 protein (PDB code: 4W2R) using the protein preparation wizard in Schrödinger suite [40][41][42]. First, the complex's protonation states were adjusted for consistency with pH 7.4. Second, all hydrogen atoms and the total structure for the protein and small molecules were evaluated for energy accuracy using an all-atom force field, OPLS_2005; restrained minimization; and heavy atom convergence to a 0.3 Å RMSD. Third, the above prepared EZH2 protein and co-crystallized molecule were used to generate a receptor grid file. The position of co-crystallized molecule was used to determine the active site location (x = 30.57, y = 18.29, and z = 54.62) and size (inner box = 10 Å × 10 Å × 10 Å; outer box = 20 Å × 20 Å × 20 Å). The grid was generated using the OPLS_2005 force field. The glide docking was then carried out between the prepared protein and molecules. The docking result was exported and analyzed using Discovery Studio 2019 and PyMol.

Conclusions
In summary, we have synthesized a series of 5-hydroxyindole scaffold compounds bearing as main pharmacophore a pyridone moiety. According to molecular modeling and in vitro biological activity assays, the preliminary structure-activity relationships were determined and summarized. Compound L-04 improved both the H3K27Me3 reduction and antiproliferation parameters (IC 50 = 52.6 µM) on K562 cells. It bears mentioning that L-21 had comparable inhibitory effect (IC 50 = 51.8 µM) to L-04; however, molecular docking indicated that L-21 (with the N1 position substituted with an isopropyl group) had reduced binding energy due to the absence of π-π interactions created by the benzyl group. In all, we believe that L-04 should remain a potential candidate to design more EZH2Is, and future in vivo studies using patient-derived tumor xenograft models are required to further assess the utility of this compound.
During the preparation of compounds, we discovered a novel catalytic activity of TCCA in condensation reactions. Different amines were condensed with β-diketones or β-ketoesters in the presence of TCCA to afford the corresponding product in short times with high yields, which displayed some significant advantages and provided an alternative condensation strategy. In addition, we established a model for L-04 binding to EZH2 using molecular docking, which will be utilized to direct future analogue design and synthesis. These findings further support the therapeutic potential of EZH2Is as anti-cancer agents.
Funding: This research was funded by the National Natural Science Foundation of China (81973365), and Liaoning Revitalization Talents Program (XLYC1908031, XLYC1902008). The APC was funded by Liaoning Revitalization Talents Program (XLYC1908031, XLYC1902008).

Acknowledgments:
The authors are indebted to Testing Center of School of Pharmaceutical Engineering for providing the HPLC-MS data for this study.

Conflicts of Interest:
The authors declare no conflict of interest.