Vicinal Diaryl-Substituted Isoxazole and Pyrazole Derivatives with In Vitro Growth Inhibitory and In Vivo Antitumor Activity

The vicinal diaryl heterocyclic framework has been widely used for the development of compounds with significant bioactivities. In this study, a series of diaryl heterocycles were designed and synthesized based on an in-house diaryl isoxazole derivative (9), and most of the newly synthesized derivatives demonstrated moderate to good antiproliferative activities against a panel of hepatocellular carcinoma and breast cancer cells, exemplified with the diaryl isoxazole 11 and the diaryl pyrazole 85 with IC50 values in the range of 0.7–9.5 μM. Treatments with both 11 and 85 induced apoptosis in these tumor cells, and they displayed antitumor activity in vivo in the Mahlavu hepatocellular carcinoma and the MDA-MB-231 breast cancer xenograft models, indicating that these compounds could be considered as leads for further development of antitumor agents. Important structural features of this compound class for the antitumor activity have also been proposed, which warrant further exploration to guide the design of new and more potent diaryl heterocycles.


■ 1. INTRODUCTION
Vicinal diaryl-substituted heterocycles can be considered privileged scaffolds as they form the core structure of a number of compounds with diverse biological properties such as anticancer, antiviral, and anti-inflammatory activities. 1 Therefore, the use of this scaffold incorporating different central heterocycles could be pivotal in the drug development practice since they exist in the structures of various clinical therapeutics or compounds with therapeutic potential, which establishes their value as chemical tools in the field of medicinal chemistry (Figure 1). For example, the first COX-2 selective nonsteroidal anti-inflammatory drug celecoxib (1) was a member of vicinal diaryl pyrazole derivatives, 2 which stimulated the development of latter COX-2 inhibitors belonging to the vicinal diaryl heterocyclic class such as rofecoxib and valdecoxib. 3 Another vicinal diaryl pyrazole derivative developed as an antiobesity drug was rimonabant (2), 4 which continues to be an inspiration for further development of cannabinoid CB1 receptor antagonists, i.e., surinabant. 5 Oxaprozin (3) and licofelone (4) are antiinflammatory drugs for the treatment of osteo-and rheumatoid arthritis, also belonging to a vicinal diaryl oxazole and pyrrolizidine class, respectively. 6,7 Furthermore, a considerable number of vicinal diaryl heterocyclic motifs spread throughout the literature with various biological activities. 1 For example, a 4,5-diaryl isoxazole derivative BRP-187 (5) was reported to be a 5-lipoxygenaseactivating protein (FLAP) inhibitor with potent anti-inflammatory activity. 8,9 The 3,4-diarylpyrazole derivative mardepodect (6, PF-2545920) was discovered as a phosphodiesterase (PDE) 10A inhibitor, 10 which progressed through Phase-II clinical trials for Schizophrenia treatment, and was used as a model compound leading to a considerable number of follow-up PDE10A inhibitors having vicinal diaryl heterocyclic framework. 11 In addition, a good number of vicinal diaryl-substituted heterocycles were studied as anticancer therapeutics in which the vicinal diaryl scaffold was shown to be a simple pharmacophore for tubulin polymerization inhibitors mimicking the features of combretastatin A4 with potent anticancer activity as exemplified with combretatiophene (7). 12 In addition, 4,5-diarylisoxazole luminespib (8, was developed as an Hsp90 inhibitor for the treatment of several cancer types and reached Phase-II clinical trials. 13 Breast cancer (BC) is the most commonly occurring cancer in women and the most common cancer overall, while hepatocellular carcinoma (HCC) is the sixth most common and the second most lethal cancer. 14 A large number of treatment options including localized and/or systematic therapies are available for both cancer types. Despite the initial effectiveness in controlling tumor growth and prolonging patient survival, nearly all current treatments result in resistance to therapies. This necessitates the development of novel approaches as well as effective chemotherapeutic agents for cancer treatments. We have long been working with the vicinal diaryl heterocyclic skeleton toward novel antiinflammatory, antiplatelet, and anticancer agents. 8,9,15−20 In this context, inspired by the therapeutic potential of the vicinal diaryl motif around a central heterocycle, our compiled library of compounds belonging to this promising class was screened for their potential cytotoxic activity against an HCC cell line. Compound 9 with 3,4-diaryl isoxazole motif with synthetic accessibility for compound library generation was identified, although the cytotoxicity against HCC cells was negligible (IC 50 ≥ 20 μM) ( Figure 1, Table 1). However, encouraged by the anticancer potential of vicinal diaryl heterocyclic scaffold and the synthetic feasibility of compound 9 skeleton, a focused library of compounds around compound 9 was prepared by decorating the vicinal diaryl rings and by incorporating various heterocyclic rings as the central core, which resulted in new analogues with improved cytotoxicity against both HCC and BC cancer cell lines. Here, we report the biological assessment of a new set of compounds that we have developed against cell lines derived from BC and HCC, which would be valuable for the understanding of the main features of these new vicinal diaryl heterocyclic analogues as anticancer therapeutic leads.

■ RESULTS AND DISCUSSION
Chemistry. Based on the developable potential of compound 9, we developed additional analogues to establish structure−activity relationships (SAR). New 3, were synthesized according to the literature, 21 and synthesis procedures and schemes of the relevant intermediates are given in the Supporting Information. In brief, hydrogenation of 9 to remove the benzyl group produced the key intermediate 10, which was subsequently used to generate desired final compounds 11−42 through alkylation of the phenolic hydroxyl (Scheme 1). The similar procedures in Scheme 1 were successfully applied to the synthesis of 44−57 with modifications at the middle phenyl and 59 with amine linker, following the reaction conditions in Schemes 2 and 3. The synthesis of 4,5-diaryl-3-methylisoxazole analogue 60 (Table 3) was achieved as we previously reported. 8 The preparation of 4,5-diarylisoxazole derivative 63 was achieved starting with readily available starting materials as illustrated in Scheme 4. Compound 61 was obtained according to the corresponding procedures. 10,22 The obtained 4,5diarylisoxazole framework 62 was subsequently used to produce the desired 63 through first the hydrolysis of the methoxy and then the alkylation of the phenolic hydroxyl group (Scheme 4).
The preparation of the 1,5-diaryl-3-methylpyrazole derivative 69 carrying the 2-methylbenzyloxy arm on 5-aryl group is outlined in Scheme 7. The synthesis of the β-diketone 68 was achieved by Claisen−Schmidt condensation. 25 The regioselective synthesis of the 1,5-diaryl-3-methylpyrazole 69 was then conveniently achieved by condensation of the diketone 68 with the hydrochloride salt of the 4-chlorophenylhydrazine in methanol/triethylamine. 15 The central pyrazole series continued with the synthesis of 4,5-diaryl-1H-pyrazole 71 and 3,4-diaryl-1-methylpyrazole 72. Briefly, the in situ formed enaminone intermediate, obtained by the treatment of 70 with DMFDMA, 10 was cyclized by hydrazine or methylhydrazine to produce the desired pyrazole compounds 71 and 72, respectively (Scheme 8). For the synthesis of the vicinal diaryl 2-methylthiazole 75, hydrolysis of 73 to remove the methyl group furnished the intermediate 74, which was utilized to generate the desired 4,5-diarylthiazole compound 75 through the benzylation of the phenolic hydroxyl (Scheme 9). 1,5-Diarylpyrazole scaffolds 78 and 79 were produced according to previously developed procedures (Scheme 10). 26 Table 1. In Vitro Cell Growth Inhibitory Activity against Hepatocellular Carcinoma and Breast Cancer Cell Lines a a IC 50 values were determined at least with five different concentrations of the compounds from the cell growth inhibition percentages. Scheme 1. Reaction Conditions and Reagents: (i) Pd/C (10% w/w), HCl, EtOH, rt; (ii) Aryl/Alkyl Halide Derivative, K 2 CO 3 , MeCN, Δ; (iii) LiOH·H 2 O, THF:water, Δ 1,5-Diarylpyrazole analogues 85−89 bearing the benzyloxy fragment on the 5-aryl as opposed to that of 78 were obtained through the cyclization of enaminones 80−84 with the hydrochloride salt of the 4-chlorophenylhydrazine in ethanol (Scheme 11). In addition, analogous compounds where the ether bridge in 85 exchanged with amide 91 or amine 92 were accomplished starting from 90 and following the standard reaction steps outlined in Scheme 12.
Bioactivity Studies and SAR. To deduce SARs, several specific areas were focused on compound 9. Hence, it was aimed to explore (i) the influence of differently positioned substituents on the benzyloxy arm and the replacement of the Scheme 2. Reaction Conditions and Reagents: (i) TEA, Diethyl Ether, 0°C; (ii) Phenyl/pyridyl Acetone Derivative, NaH, THF, 0°C phenyl ring of the benzyl group by heteroaryl moieties, (ii) the incorporation of different substituents at the isoxazole-4phenyl group, (iii) the modification of the middle phenyl, and (iv) the replacement of central isoxazole with isosteric heterocycles. Collectively, a total of seventy vicinal diaryl heterocyclic compounds were synthesized to screen their inhibitory effects on cancer cell proliferation against Huh7 and MCF7 cells by introducing different chemical functionalities with the aforementioned modifications (Tables 1−4). Initially, the substitution pattern at the benzyl functionality was scrutinized by comparison of the differently substituted analogues (11−42) as illustrated in Table 1. Methylation of the aromatic ring of the benzyl at 2-position (11) caused a sudden increase in the potency (IC 50 = 1.3 μM for Huh7 and 3.8 μM for MCF7) versus compound 9 (IC 50 ≥ 20 μM for both Huh7 and MCF7) with unsubstituted benzyl arm. However, relocation of the methyl group to 3-or 4-position (19 and 24) led to a significant activity loss (IC 50 ≥ 20 μM). Substitutions at 2-position other than methyl, i.e., electrondonating (12)(13)(14) and electron-withdrawing groups (15−18), were undesirable resulting in an activity decrease against both cell lines (IC 50 values of 8.5 to >20 μM). Seemingly, voluminous substituents than methyl at 2-position were not well tolerated and impaired the efficiency of the compounds. Benzyl analogues differently substituted at 3-(19-23) or 4position (24−29) were not chased further as they demonstrated a significant loss of inhibitory activity (IC 50 values of 12 to >20 μM) for both cell lines except for 20 with 3-methoxy, which showed comparable activity to 11 for the Huh7 cell line (IC 50 = 1.3 vs 3.6 μM). Compounds with dimethyl substitution pattern, i.e., 2,3-diMe (30) or 2,, suggested that additional methyl group other than 2-methyl was not well tolerated on the benzyl moiety.
Next, it was explored the impact of installing a heteroaromatic group in place of the phenyl ring of the benzyl functionality (32−42). When the 2-methylphenyl was replaced by 3-pyridinyl (32), 4-pyridinyl (33), or a voluminous 2quinonyl (34), these compounds showed decreased cytotoxic activity, while compound 35 with the 3-pyridine ring having a methyl at a topologically equivalent position as in 11 retained the potency with IC 50 values of 3.9 μM for Huh7 and 6.1 μM for MCF7. A similar analogue 36 with a 2-pyridine ring having a methyl at the same position partially restored the potency against Huh7 (IC 50 = 9.4 μM), while the activity loss toward MCF7 was more pronounced (IC 50 ≥ 20 μM). Of interest, several five-membered heteroaromatic counterparts with alkyl substitutions such as 3,5-dimethylisoxazole (39), 1,3-dimethylpyrazole (40), and 1-isopropylimidazole (41) were well tolerated for their inhibition potential against Huh7 with IC 50 values of 3.7, 0.9, and 2.5 μM, respectively, while these three compounds appeared to be less effective against MCF7 cells (IC 50 values of 12 to >20 μM). From this point of view, the 2methylbenzyl group remains the best choice for the 3,4-diaryl-3-methylisoxazole core, i.e., compound 11, for cytotoxic potency albeit several five-membered heteroaryl moieties exemplified with compounds 39−41 are also conceivable. Apparently, a consistent SAR for cytotoxic activity at this part was not accessible since small structural differences on the benzyl arm were not well tolerated.
Then, it was investigated if the 4-chloro substituent on the phenyl ring at C(4)-isoxazole was replaceable with different atoms or groups, while keeping the 2-methylbenzyl unit due to its good impact on the activity (compounds 44−55 in Table  2). Moving the 4-chloro in 11 to 2-position (44) or replacing it with a smaller fluoro (45) or larger methoxy (46) group resulted in a decreased potency against both cell lines (IC 50 = 9.9 to >20 μM). The amino replacement of 4-chloro (51) further reduced the potency for both cell types, similar to 45 and 46. On the other hand, substitution of the 4-chloro with a methyl group (47), which is about the same size as a chlorine atom, was pertinent and regained the bioactivity in both cell lines (IC 50 = 1.8 and 4.7 μM, respectively), whereas with a linear and less voluminous nitrile group in this position (48), the efficiency again dropped, especially against MCF7 cells (IC 50 ≥ 20 μM). Introducing a nitrogen atom to the 4methylphenyl ring in 47 also caused an activity loss against MCF7 as exemplified with 54. Finally, compounds with substitutions at 3-position (49-50) or with 2,4-dihalogen substitutions (52−53) as well as with 4-pyridyl (55) were also significantly less effective indicating that the phenyl ring of C4isoxazole may require substitution at 4-position with a similar steric size such as methyl or chlorine ( Table 2). The next goal was to explore the substituent effect on the middle phenyl group, and this was briefly examined by fluoro substitution at 2-(56) and 3-(57) positions (Table 2). Although the 2-F substituted 56 preserved the potency against Huh7 and MCF7 cells with IC 50 of 2.25 and 9.53 μM, respectively, 3-F substitution in the compound 57 hampered the potency against both cell lines, implying that 2-substitution on the middle phenyl ring may be allowable for retaining or even improving the activity. Lastly, our efforts at the benzyloxyphenyl component included an isosteric exchange of the ether bridge with an amino linker (59), which was found tolerable for the Huh7 potency (IC 50 = 4 μM) with a concomitant loss of the activity against the MCF7 cell line (IC 50 ≥ 20 μM) ( Table 2).
After the biological confidence was rationalized with 3,4diarylisoxazole derivatives with respect to C3-benzyloxyphenyl and C4-phenyl pendants, the next area of interest was to explore the central isoxazole with a series of common fivemembered heterocycles to broaden the SAR investigation. To this end, eleven heterocyclic congeners with different heteroatom counts and orders were synthesized to probe the effect on the cytotoxic activity against HCC (Huh7 and Mahlavu) and BC (MCF7) cells and to search for the optimal central core for further SAR studies ( Table 3). The position exchange of nitrogen and oxygen atoms in 11 to produce 60 resulted in a significant drop of activity against MCF7 cells (IC 50 ≈ 19 μM), while the potency toward Huh7 cells was still preserved (IC 50 = 3.8 μM). Furthermore, removal of 3-methyl of isoxazole in 60 to afford 63 appeared to diminish the potency against Mahlavu and MCF7 cells (IC 50 ≥ 20 μM) but improved the cytotoxic activity for Huh7 (IC 50 = 1.5 μM). Based on this result, the methyl in 11 was replaced with a polar amino group to further explore the hydrophobic effect and procured 65 with comparable potency to parent 11 against all cell lines with IC 50 = 2.0-3.9 μM. Interestingly, switching the sites of C3-benzyloxyphenyl unit and C4-phenyl in 11 to afford 67 appeared to diminish the cytotoxic activity for all three cell lines (IC 50 ≥ 20 μM). Meanwhile, other combinations of heteroatoms such as a group of pyrazoles as well as a thiazole core were also evaluated for their potential to replace the isoxazole core in 11. While thiazole 75 and pyrazole 78 did not produce the desired activity enhancement, other pyrazole derivatives with different orders of nitrogens and methyl substitutions (69, 71, 72, 79, and 85) maintained adequate cytotoxicity against all cell lines with IC 50 values in the range of 0.7 to 14.4 μM. Gratifyingly, the reversal of benzyloxyphenyl and 4-chlorophenyl components in 78 to produce 85 displayed a potency boost for Huh7 and MCF7 with IC 50 values of <1 μM, while still maintaining a decent activity against Mahlavu cells (IC 50 = 3.7 μM) compared with the parent 11. Based on the promising potential of 85 as an anticancer agent, the SAR around the middle phenyl and the ether linker was briefly examined, while preserving the 2-methylphenyl and 4chlorophenyl units, which had been found optimal for the potency ( Table 4). The introduction of a nitrogen atom into the middle phenyl in 85 resulted in isosteric pyridine mimics 27 86 and 87 in which the 2-pyridyl regioisomer (86) exhibited higher cytotoxicity toward Huh7 and Mahlavu cells (IC 50 = 2.0 and 7.1 μM, respectively), although this was accompanied by a reduction in the potency against MCF7 cells (IC 50 ≥ 20 μM). The bioisosterism between the azine C-N and the aryl C-F bond in compounds 88 and 89 was also examined. 28 The activity results indicated an efficient bioisosterism of the C-F bond with the pyridine N atom in this context because the replacement of the pyridine N atom with the C-F moiety led to improved potency in 88 (IC 50 = 1.6−6.4 μM). This was also in good correlation with its isoxazole congener 56 with the 2-F substituted middle phenyl group (Table 2). Lastly, the role of ether linker was studied in 91 and 92. As seen, while the amide linker was not tolerated for any cell line, the amine replacement of the ether oxygen (92) was tolerable for Huh7 and Mahlavu cells (IC 50 = 3.6 and 4.5 μM, respectively) with a diminished activity against MCF7 cells (Table 4), again in good agreement with its isoxazole counterpart 59 (Table 2). According to the SAR results, compound 11 with 3,4diarylisoxazole and compound 85 with 1,5-diarylpyrazole frameworks were selected for further analysis, and unambiguous structural elucidation of 11 and 85 using the single-crystal X-ray diffraction method was accomplished showing the accurate rearrangement of aromatic rings and atoms in 3Dshape ( Figure S1). 29,30 Next, judged by the cellular activity of 11 and 85, both compounds in addition to Huh7, Mahlavu, and MCF7 cell lines were further screened against a panel of the hepatocellular carcinoma (HepG2, SNU475, Hep3B, FOCUS, Hep40, and PLC-PRF-5) and breast cancer (MDA-MB-231, MDA-MB-468, SKBR3, and ZR-75) cell lines along with the nontumorigenic immortalized breast epithelial cells MCF10A ( Table 5). The results demonstrated that both compounds are endowed with potent antiproliferative activity against all cancer cells with IC 50 values in the range of 1.3−9.5 μM for 11 and 0.77−7.8 μM for 85 for hepatocellular carcinoma and breast cancer cell lines, while found less toxic to the MCF10A immortalized normal breast epithelial cells (Table 5).
Moreover, in vitro single-dose anticancer screening of 11 and 85 at 10 μM was performed utilizing the stable panel of 60 cell lines comprising 9 different cancer types at the National Cancer Institute (NCI) under the Developmental Therapeutics Program (Table S1). 31 Compounds 11 and 85 demonstrated variations in sensitivity and selectivity against individual cell lines in the panel, as illustrated in Table S1. Collectively, both compounds exhibited similar antiproliferative activity against 22 cancer cells with growth inhibition (GI) values in the range of 45−100% at 10 μM. The NCI panel results revealed that both compounds displayed a good preference for leukemia such as CCRF-CEM, HL-60, K-562, MOLT-4, and SR, and colon cancer including HCT-116, HCT-15, HT-29, KM-12, and SW-620 cell lines, in addition to hepatocellular carcinoma and breast cancer cells of this work.
Based on the encouraging potency of 85 in cellular assays against hepatocellular carcinoma and breast cancer as well as in the NCI-60 panel, we decided to evaluate in vitro ADME and pharmacokinetic (PK) properties in mice ( Figure S2).  Metabolic stability of 85 in human microsomes was moderate (55% remaining after 45 min incubation). The compound is highly lipophilic with logD 7.4 of 6.64, sparingly soluble at the physiological pH, and highly plasma protein-bound (99.9%). The capacity of 85 to inhibit human CYP isoforms 2C9, 2D6, and 3A4 was insignificant (7.3, 22.2, and 0% at 10 μM, respectively) implying a safe window for clinical drug interactions.
The PK parameters after intravenous (iv) administration of 85 constitute a moderate volume of distribution, low total plasma clearance, and a moderate iv half-life of 3.85 h resulting in mice with low oral bioavailability ( Table 6). Although PK parameters are nonoptimal, this may be counterbalanced by the potent in vitro antiproliferative activity spectrum of 85 and is not considered limiting for further progression of 85 to in vivo antitumor efficacy studies. Characterization of Cell Death Mechanism Induced by 11 and 85. To control the cell death mechanism triggered by 11 and 85, Huh7 and Mahlavu hepatocellular carcinoma cells and MDA-MB-231 and MCF-7 breast cancer cells were cultured according to their cell growth rate and were treated with both 11 and 85 for 48 h. The apoptotic morphological changes were observed in both breast cancer cells and hepatocellular carcinoma cells upon treatment by nuclear staining with Hoechst compared to the control group ( Figure  3A). The apoptotic cell populations were further examined with Annexin V staining using flow cytometry. Compared to the control group, the percentage of apoptotic populations in both breast cancer cell lines and hepatocellular carcinoma cell lines treated with 11 or 85 was increased after 24 h ( Figure  3B). For further investigation of apoptosis activation through 11 and 85 treatments, apoptosis-associated PARP protein levels were assessed using western blotting. Except for 11 treated MCF7 and Huh7 cells, both 11 and 85 compounds caused the increase in PARP cleavage in both breast cancer cells (MCF7 and MDA-MB-231) and hepatocellular carcinoma cells (Mahlavu) (Figure 3C). These results further supported the increased cytotoxic effects of compounds on breast cancer and hepatocellular carcinoma cancer cells. We next investigated in vivo antitumor efficacy of 11 and 85 on nude mice tumor xenografts.

Real-Time Cellular
In Vivo Antitumor Effects of 11 and 85 in Mice Xenograft Models. The antitumor effects of 11 and 85 in the hepatocellular carcinoma (Mahlavu cells) and breast (MDA-MB-231 cells) xenograft models were assessed twice a week by oral administration of 11 and 85 at 40 mg/kg for 4 weeks. Both compounds conferred a sustained antitumor efficiency (Figures 4 and S3). In the Mahlavu xenografts, mice administered with compounds 11 and 85 had a significant reduction in tumor volume following 4 weeks of treatment, i.e., 85 and 40% reductions in tumor volumes, respectively. Moreover, for MDA-MB-231 xenografts, mice treated with both compounds resulted in about a 50% decrease in tumor volumes as compared to the control group ( Figure 4). In all studies, the administered dose was well tolerated and neither significant bodyweight loss nor toxic effects or mortality were observed.

■ CONCLUSIONS
We synthesized a series of vicinal diaryl isoxazole and pyrazole derivatives as putative anticancer agents and checked their growth inhibitory activity against human hepatocellular carcinoma and breast cancer cell lines. Most of the derivatives represented moderate cytotoxicity in the selected cancer cells. In general, the substitution type and pattern on the benzyloxyphenyl group linked to the central heterocycle had a significant effect on the potency and the selectivity of the compounds in the tested cancer cell lines. Following a comprehensive evaluation of twelve central heterocyclics comprising a combination of different heteroatom positions and numbers, we observed that the type of the central heterocycle, i.e., isoxazole and pyrazole, was also crucial for the observed anticancer potency. Subsequently, two analogues, the 3,4-diarylisoxazole derivative 11 and 1,5-diarylpyrazole 85, stand out as developable anticancer compounds based on their significant in vitro antiproliferative activities toward the 13 hepatocellular and breast cancer cell lines with IC 50 values in the range of 0.77 to 9.53 μM. We also demonstrated that both compounds displayed dose-and time-dependent growth inhibition through the RT-CES system, which was also correlated to initial SRB screening results. Further analysis showed that both compounds induced apoptotic cell death in    hepatocellular and breast tumor xenografts with inhibition rates between 40% and 85%, and with insignificant effect on the mice bodyweight.
In conclusion, our results revealed that these diaryl-isoxazole and -pyrazole derivatives exemplified with 11 and 85 show promise as leads for further development of improved anticancer compounds against hepatocellular carcinoma and breast cancers. Hence, further investigation of novel analogues that could be integral to the current SAR in this study would be of value to identify novel diaryl heterocycles with strong anticancer and druglike properties. Elucidation of detailed molecular mechanisms associated with 11 and 85 such as direct profiling of compounds using NanoString analysis as well as target fishing experiments with biotin-labeled conjugates to identify potential molecular targets are under investigation and will be reported in due time.

-M e t h y l -4 -( 3 -fl u o r o m e t h y l p h e n y l )
Statistical Analysis. All data were obtained from three independent experiments and standard deviation (S.D) values were accessed. All experiments except western blotting were done two times with n ≥ 3 biological replicates. One-way ANOVA and two-way ANOVA were applied using GraphPad (Prism) for statistical analysis. Results were shown as follows: ns: not significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. ■ ASSOCIATED CONTENT
Chemical procedures and experimental data of intermediates, NCI-60 cancer cell panel results for compounds 11 and 85, X-ray data, PK and in vivo test data, and copies of 1 H and 13