Design, Synthesis, and Evaluation of Benzoheterocyclic-Containing Derivatives as Novel HDAC1 Inhibitors

In this study, the synthesis and biological evaluation of a variety of benzoheterocyclic-containing We designed and synthesized 14 novel benzamide HDAC1 inhibitors. By screening inhibitory activity of HDAC1 and cellular antiproliferative activity, we found that compound 3c possessed better inhibitory activities of HDAC1 and cellular antiproliferative activity superior to chidamide. Moreover, our data suggested that 3c did not possess signi ﬁ cant toxicity to primary human cells and the patch clamp hERG K þ ion channel. Compound 3c had good selectivity for HDAC1 over HDAC6 and HDAC8. The oral pharmacokinetic parameters of 3c were acceptable. Results from in vivo antitumor activity study showed that 3c exhibited potent oral antitumor activity in an HCT116 human colon carcino-ma mouse xenograft model. Further assessment of compound 3c is in and we will report in future.


Introduction
Histone deacetylases (HDACs) and histone acetyltransferases jointly regulate the acetylation levels of cellular histone proteins, thereby regulating the expression of genes. At present, 18 human HDACs have been identified and divided into four classes: Class I (HDAC1, HDAC2, HDAC3, and HDAC8), Class II (Zolinza; 2006), romidepsin (Istodax; 2009), belinostat (Beleodaq;2014), and panobinostat (Farydak;, have been approved by the U.S. Food and Drug Administration for cancer treatment. 5 However, most of the used HDACIs are of low selectivity, and may inhibit all or at least a few members of the HDAC family, leading to many side effects, low potency, or low stability during therapy. 6,7 Nowadays, a lot of benzamidebased HDACIs have entered clinical evaluation for the treatment of solid tumors and hematological malignancies, such as entinostat , chidamide (CS055), and mocetinostat (MGCD0103). 8,9 They can selectively and significantly inhibit HDACs 1-3. Chidamide was the first orally available benzamide class of HDACIs approved by the China Food and Drug Administration for the treatment of advanced peripheral T cell lymphoma. MS-275 is another benzamide-based HDACI that preferentially inhibits HDAC1 (IC 50 ¼ 510 nmol/L) over HDAC3 (IC 50 ¼ 1.7 μmol/L). It showed good anticancer efficacy in experiments. 10 Above all, benzamide derivatives are research hotspots for exploring HDACIs.
Although there are various structural features for HDACIs, most HDACIs can be widely described with a zinc-binding group (ZBG), a linker, and a cap group (CAP) as shown in ►Fig. 1. 11 The structural similarity of the existing benzamide HDA-CIs suggested that rational isosteric modification of the surface-recognition domain or ZBG is feasible. By analyzing the structure-activity relationship of HDACIs (MS-275 and chidamide), herein we reported the design, synthesis, and preliminary bioactivity evaluation of benzoheterocyclic-containing benzamide derivatives as HDACIs (►Fig. 1).

Results and Discussion
The preparation of target compounds 3a-3n is shown in Scheme 1. The intermediates 2a-2 g could be obtained by condensation of 1a-1 g and 4-aminomethyl-benzoic acid under the conditions of O-benzotriazol-1-yl-tetramethyluronium hexafluorophosphate (HBTU). Then intermediates 2a-2 g coupled with HBTU and further reacted with phenylenediamine or 4-fluorophenylenediamine to obtain the target compounds 3a-3n. All synthetic compounds 3a-3n are confirmed by 1 H NMR (nuclear magnetic resonance) and ESI-MS (electrospray ionization mass spectrometry).
The inhibitory effects of compounds 3a-3n against HDAC1 enzyme subtype were determined by using a fluorescence-based assay as described before. 12 Data are presented as IC 50 values in ►Table 1. The results showed that all the compounds showed potent HDAC1 inhibitory activity, but not significantly superior to chidamide except for compound 3c. Compound 3c exhibited a much smaller IC 50 value (0.64 μmol/L) than that of chidamide (IC 50 ¼ 1.28 μmol/L) (►Table 1). As such, compound 3c was used for the following study. As presented in ►Table 1, analogs containing the 2amino-4-fluorophenyl group in the ZBG positions, such as 3b, 3d, 3f, 3h, 3j, 3k, and 3m, were typically less potent than the derivatives with 2-aminophenyl substitution at the same position (3a, 3c, 3e, 3 g, 3i, 3l, and 3n). The introduction of more N atoms in the CAP region was not conducive to HDAC1 inhibitory activity of the compound (such as 3a versus 3c), suggesting that the electrostatic properties of the substitutions on the ''CAP" group may play a critical role to contribute HDAC1 inhibitory activities.
To further understand the interaction between 3c and HDAC1, molecular docking studies were performed to evaluate the possible binding modes of 3c with the active site of HDAC1 (PDB entry: 4BKX) using Syble/FlexX module. 13 The docking results showed that compound 3c could well insert into HDAC1 active sites (►Fig. 2). The benzamide group of compound 3c could chelate the Zn 2þ very well (1.96 Å to the nitrogen atom in aniline) in a bidentate fashion. Besides, four amino groups formed four hydrogen-bonding interactions with ASP-99 and GLY-149, respectively. In general, the result of molecular docking supported a tight interaction between compound 3c and HDAC1.
MTT assay was further performed to assess the antiproliferative activity of 3c in human cancer cell lines (PC-3, HT-29, HCT-116, SK-BR-3, Jurkat E6-1, A549, Colo205, and MCF-7) and human fetal lung fibroblast normal cell line . 14 ►Table 2 demonstrates that compound 3c presented better antiproliferative activity in most tested cancer cell lines with IC 50 values obviously superior to chidamide, yet, a weak inhibitory activity against human normal cell MRC-5 cell lines (IC 50 > 10 μmol/L), suggesting that compound 3c inhibits Table 1 The chemical structures and HDAC1 inhibitory activities of target compounds 3a-3n proliferation of cancer cells with desirable selectivity over human normal MRC-5 cell lines.
The selective inhibitory effect of compound 3c for HDAC isoforms, including HDAC1 (Class I), HDAC8 (Class I), and HDAC6 (Class IIb), was also assessed (►Table 3). We found that 3c showed potent inhibitory activity toward HDAC1 and HDAC2 (IC 50 ¼ 2.1 μmol/L) while weak activity against HDAC6 (IC 50 > 10 μmol/L) and HDAC8 (IC 50 > 10 μmol/L). Further, we investigated the cardiac safety of compound 3c through a patch-clamp hERG K þ channel screening. The results showed that 3c was inactive in our hERG binding assay (IC 50 > 30 μmol/L), demonstrating that the potential risk cardiotoxicity of 3c was relatively low. The acute toxicity of compound 3c was tested in ICR mice by oral administration. The result suggested that administration of 3c at or below 1,500 mg/kg (po [oral]) may be safe for mice.
The pharmacokinetic study of 3c was evaluated using Sprague-Dawley (SD) rats after single ig (intragastric) at 10 mg/kg and iv (intravenous) at 2 mg/kg. The C max of 3c was 5,260 μg/L after iv dosing for 0.05 hours. After ig dosing of 3c, it was well cleared (CL ¼ 6.26 L/h/kg) in rats and the terminal phase half-life of 3c was 3.23 hours. Meanwhile, 3c was well distributed (V z ¼ 29.10 L/kg) and the oral bioavailability was moderate (F ¼ 10.35%) in rats (►Table 4). Therefore, research focus on enhancing the oral bioavailability of 3c through exploring the novel formulation was further studied in our laboratory. Unfortunately, data are unavailable at present.
The antitumor activity of compound 3c was further tested in a HCT-116 mouse xenograft model at daily intragastric doses of 45, 80, and 150 mg/kg for 28 days (►Fig. 3). It is apparent that 3c inhibited tumor growth in vivo. Under the dose of 45, 80, and 150 mg/kg, the ratio of tumor volume in treated versus control mice (T/C) was 73.22, 71.01, and 43.43%, respectively, suggesting the obvious antitumor activity of compound 3c in vivo.

Conclusion
We designed and synthesized 14 novel benzamide HDAC1 inhibitors. By screening inhibitory activity of HDAC1 and cellular antiproliferative activity, we found that compound 3c possessed better inhibitory activities of HDAC1 and cellular antiproliferative activity superior to chidamide. Moreover, our data suggested that 3c did not possess significant toxicity to primary human cells and the patch clamp hERG K þ ion channel. Compound 3c had good selectivity for HDAC1 over HDAC6 and HDAC8. The oral pharmacokinetic parameters of 3c were acceptable. Results from in vivo antitumor activity study showed that 3c exhibited potent oral antitumor activity in an HCT116 human colon carcinoma mouse xenograft model. Further assessment of compound 3c is in progress and we will report in the future.

Experimental Section
Chemicals and Instruments

General Synthetic Procedure of 3a-3n
To a suspension of HBTU (3.03 g, 8 mmol) in CH 3 CN (5 mL) was added the corresponding 1a-1 g (8.0 mmol) in CH 3 CN (5 mL), and the mixture was stirred for 1.5 hours at room temperature (r.t.). The resulting solution was added to a solution of NaOH (0.32 g, 8 mmol) and 4-(aminomethyl)benzoic acid (1.21 g, 8 mmol) in water (100 mL). After stirring for 4 hours at r.t., the solution was acidized with HCl (pH 5) to precipitate a white solid which was collected by filtration, washed with water (100 mL) and methanol (50 mL), respectively, and dried to give the benzoic acid derivatives 2a-2 g without further purification. To a solution of the corresponding benzoic acid derivatives 2a-2 g (1 equiv.) in DMF was added 1,2-phenylenediamine or 4-fluorophenylenediamine (1 equiv.), HBTU (1.2 equiv.), and TEA (4 equiv.) dropwise. The reaction mixture was stirred at r.t. for 6 hours, after which TLC analysis indicated the reaction was completed. The solution was added with water (20 equiv.) to precipitate a white solid. The mixture was poured into water and stirred for 30 minutes. Insoluble material was filtered. The solid was collected and dried to be recrystallized in aqueous ethanol to give target compounds 3a-3n. N- (4-((2-aminophenyl)

Fluorescence Assay of HDAC Inhibition Activities
In vitro HDAC inhibition assays were performed as previously described. 13 In brief, 10 μL of enzyme solution (HeLa cell nuclear extract, HDAC1, HDAC2, HDAC6, or HDAC8, obtained from BPS Bioscience, San Diego, California, United States) was mixed with different concentrations of the tested compound (50 μL). The mixture was incubated at 37°C for 5 minutes, followed by adding 40 mL of the fluorogenic substrate. The substrate Boc-Lys(Acetyl)-AMC (Bidepharm, Shanghai, China) was used for assaying HDAC1, HDAC2 and HDAC6; and boc-lys(triflouroacetyl)-AMC (Bidepharm, Shanghai, People's Republic of China) was used for assaying HDAC8. After incubation at 37°C for 30 minutes, the mixture was quenched by 100 μL of the developer containing trypsin and trichostatin A, and incubated for another 20 minutes. Fluorescence intensity was measured using SpectraMax M5 (Molecular Devices, LLC., San Jose, California, United States) at excitation wavelengths of 390 nm and emission wavelengths of 460 nm. The inhibition ratios were calculated from the fluorescence intensity readout of tested wells relative to those of control wells, and the IC 50 values were calculated using the Prism nonlinear curve fitting method (allowed to float and fitted as a parameter).

Compound on hERG Activity
Whole-cell recordings were performed using automated Qpatch (Sophion, Biolin Scientific, Stockholm, Sweden). Cells were voltage clamped at a holding potential of À80 mV. The hERG current was activated by depolarizing at þ20 mV for 5 seconds, after which the voltage was taken back to À50 mV for 5 seconds to remove the inactivation and observe the deactivating hERG tail current. The voltage stimulation was applied per 15 seconds. Compound solutions were administrated from low to high concentration with 2 minutes for each concentration, and 10 μmol/L cisapride was applied at the end of perfusion of compound solution. Each concentration was tested on at least three parallels. The degree of hERG channel inhibition was determined by the following equation: Inhibition (%) ¼ (1 À variation in the current before and after addition of a test substance/variation in the current before and after addition of a medium) Â 100.

In Vivo Bioavailability Study (Pharmacokinetic Parameter)
Male SD rats (n ¼ 3) were purchased form SLRC laboratory Animal Inc., Shanghai, People's Republic of China, and used in the pharmacokinetic parameter studies of compound 3c. Compound 3c was dissolved and vortexed in 5% DMSO, 10% Tween 80, and 75% physiological saline for a concentration of 0.2 and 1 mg/mL. Rats were housed in a room with controlled temperature and humidity and allowed free access to food and water. The rats were split into iv group (2 mg/kg) and ig group (10 mg/kg) before starting treatment (24 rats in each group). At indicated time points, blood was collected from rats at various time points up to 12 hours. The concentrations of compounds in plasma were determined by LC/MS/MS (Shimadzu LC-30AD, Kyoto Japan). A solution of 0.05 N HCl in saline was used as the vehicle.

In Vivo Antitumor Activity Assay
The antitumor effect of compound 3c was assessed in a mouse xenograft HCT116 tumor model. Female BALB/C mice (6 weeks old) were purchased form SLRC laboratory Animal Inc. and housed and maintained under specificpathogen free conditions. Animal procedures were performed according to institutional ethical guidelines of animal care. The HCT-116 cells (5 Â 10 7 /mL) in logarithmic growth phase were suspended with Geltrex. The cell suspension (120 μL) was injected subcutaneously into the right flank of mice with a 1 mL syringe. The tumor in mice was observed regularly until reaching to 100-300 mm 3 . Then the tumorbearing mice (n ¼ 40) were randomly divided into four groups: vehicle control group (n ¼ 16) and three test substance groups (n ¼ 8). Vehicle groups were given vehicle alone. The treatment groups were given 3c (ig) once a day for 21 days at a dose of 45, 80, and 150 mg/kg, respectively .
The tumor diameter was measured twice a week, and the tumor volume was calculated as the following: V ¼ [length (mm) Â width 2 (mm 2 )]/2.

Ethics Statement
The present study was approved by the animal ethics committee and abides by the relevant agreements of China State Institute of Pharmaceutical Industry, Shanghai, People's Republic of China.

Funding
This work was financially supported by the National Science and Technology Major Project (Grant No. 2018ZX09711002-002-009), the National Natural Science Foundation of China (Grant No. 81703358), and the Science and Technology Commission of Shanghai Municipality (Grant Nos. 22ZR1460300, 18QB1404200, and 21S11908000).