Discovery and SAR Study of Quinoxaline–Arylfuran Derivatives as a New Class of Antitumor Agents

A novel class of quinoxaline–arylfuran derivatives were designed, synthesized, and preliminarily evaluated for their antiproliferative activities in vitro against several cancer cell lines and normal cells. The representative derivative QW12 exerts a potent antiproliferative effect against HeLa cells (IC50 value of 10.58 μM), through inducing apoptosis and triggering ROS generation and the accumulation of HeLa cells in vitro. Western blot analysis showed that QW12 inhibits STAT3 phosphorylation (Y705) in a dose-dependent manner. The BLI experiment directly demonstrated that QW12 binds to the STAT3 recombination protein with a KD value of 67.3 μM. Furthermore, molecular docking investigation showed that QW12 specifically occupies the pY+1 and pY-X subpocket of the SH2 domain, thus blocking the whole transmission signaling process. In general, these findings indicated that the study of new quinoxaline–aryfuran derivatives as inhibitors of STAT3 may lead to new therapeutic medical applications for cancer in the future.


Introduction
Today, cancer is among the leading causes of mortality and a major public health concern in the world. Approximately 19.3 million cases occurred in 2020 alone [1]. If current trends continue, new cancer cases will increase by 60% worldwide by 2040 and by more than 80% in low-and middle-income countries, where cancer diagnostic and treatment services are insufficient [1]. Although the increased understanding of tumorigenesis and progression has led to the development of numerous drugs available to treat various types of cancer, the marketing of anticancer drugs could not meet the demand of clinics until now [2]. Therefore, it is paramount to discover novel drugs for cancer therapies in order to minimize patient suffering and reduce the costs of expensive treatments.
Quinoxalines and their derivatives have been considered promising active compounds for the development of novel therapeutic agents, because of their broad biological activities, such as anticancer [3], anti-inflammatory [4], antifungal [5], antiproliferative [6], and antitubercular [7]. Many anticancer candidates with quinoxaline scaffolds (Figure 1) have been subjected to clinical trials, such as antineoplastic topoisomerase II inhibitors XK469 (1) and CQS (2) [8]. Compound 3, a synthesized derivative of acrylamide-quinoxaline, with moderate inhibitory activity on tumor cell lines derived from patients resistant to a firstgeneration EGFR inhibitor [9]. Compound 4 exhibited potent cytotoxicity against cancer Pharmaceutics 2022, 14, 2420 2 of 21 cells and significantly reduced tumor size in a dose-dependent manner [10]. Compound 5, a new quinoxaline-isoselenourea hybrid, showed good activity against five melanoma cell lines, including mutant B-RAFV600E and wild-type, with IC 50 values of 0.8-3.8 µM [11]. NVP-BSK805 (6), a new substituted quinoxaline derivative, exhibited inhibition against six human myeloma cell lines, with IC 50 values between 2.6 µM and 6.8 µM [12]. Compound 7 (Figure 2), a phenyloxy quinoxaline derivative, which potently suppresses proliferation against SET-2 cells in vitro and favorable rat pharmacokinetic properties in vivo [13]. This accumulated evidence indicated that quinoxaline was an important crucial heterocycle in the development of antitumor agents.
Quinoxalines and their derivatives have been considered promising active compounds for the development of novel therapeutic agents, because of their broad biological activities, such as anticancer [3], anti-inflammatory [4], antifungal [5], antiproliferative [6], and antitubercular [7]. Many anticancer candidates with quinoxaline scaffolds (Figure 1) have been subjected to clinical trials, such as antineoplastic topoisomerase II inhibitors XK469 (1) and CQS (2) [8]. Compound 3, a synthesized derivative of acrylamide-quinoxaline, with moderate inhibitory activity on tumor cell lines derived from patients resistant to a first-generation EGFR inhibitor [9]. Compound 4 exhibited potent cytotoxicity against cancer cells and significantly reduced tumor size in a dose-dependent manner [10]. Compound 5, a new quinoxaline-isoselenourea hybrid, showed good activity against five melanoma cell lines, including mutant B-RAFV600E and wild-type, with IC50 values of 0.8-3.8 μΜ [11]. NVP-BSK805 (6), a new substituted quinoxaline derivative, exhibited inhibition against six human myeloma cell lines, with IC50 values between 2.6 µ M and 6.8 µ M [12]. Compound 7 (Figure 2), a phenyloxy quinoxaline derivative, which potently suppresses proliferation against SET-2 cells in vitro and favorable rat pharmacokinetic properties in vivo [13]. This accumulated evidence indicated that quinoxaline was an important crucial heterocycle in the development of antitumor agents. In addition to the quinoxaline moiety, the furan ring is always widely used in drug development for its broad bioactivities. For example, nifuroxazide exhibits potent antiproliferative activity against various melanoma cell lines by inhibiting STAT3, which could significantly inhibit tumor growth without obvious side effects in a model of A375bearing mice, by inducing apoptosis and reducing cell proliferation and metastasis [14][15][16]. Furthermore, certain furan derivatives with phenyloxy substituents, such as compounds 8 and 9 (Figure 2), have been obtained and exhibited high cytotoxicity against the human cancer cell lines MCF-7, TK-10, and UACC-62 [17]. It has been reported that phenyl-2-furan is a potential pharmacophore for antitumor proliferative activity with multiple mechanisms, such as inhibiting protein tyrosine phosphatase 1B [18] and inhibiting Pglycoprotein [19]. Thus, it is promising to use the phenyl-2-furan motif to discover anticancer agents. As an ongoing effort to identify effective anticancer agents with novel scaffolds, we combined quinoxaline and arylfuran motifs as the designed quinoxaline-arylfuran scaffold ( Figure 2). Recently, substituted hydrazide−hydrozones have attracted the researchers to develop their heterocyclic analogues as medicinal agents for their diverse biological activities, such as antibacterial [20], antitumor [21], and anti-inflammatory [22]. Thus, we chemically constructed a quinoxaline-arylfuran scaffold with a substituted acylhydrazone side chain ( Figure 2) and evaluated it for its anticancer properties against several cancer cell lines. The preliminary structure-activity relationship was studied, and then we performed cell cycle analysis and apoptosis studies on the most potent compound, QW12, in vitro. Furthermore, Western blot and BLI analysis revealed the mechanism of action of this compound. These results indicated that QW12 could be further investigated as an ideal lead compound for antitumor agents. In addition to the quinoxaline moiety, the furan ring is always widely used in drug development for its broad bioactivities. For example, nifuroxazide exhibits potent antiproliferative activity against various melanoma cell lines by inhibiting STAT3, which could significantly inhibit tumor growth without obvious side effects in a model of A375-bearing mice, by inducing apoptosis and reducing cell proliferation and metastasis [14][15][16]. Furthermore, certain furan derivatives with phenyloxy substituents, such as compounds 8 and 9 (Figure 2), have been obtained and exhibited high cytotoxicity against the human cancer data recorded were analyzed by means of the BLItz software and MATLAB to extrapolate the kinetic parameters. All association and dissociation curves were fitted by a single exponential function. Each acquisition was repeated twice to confirm reproducibility.

Molecular Docking
The molecular docking studies with QW12 were performed using Schrodinger based on the crystal structure of STAT3 downloaded from the Protein Data Bank (PDB:1BG1). Schrodinger software was used to prepare the macromolecule and ligand. The compound QW12 was docked using the Glide XP docking procedure. Gasteiger charges were assigned to the ligands by AutoDock Tools. The grid size was set to 105 Å × 75 Å × 68 Å, which is large enough to cover the entire active target site. After completing 10 million energy evaluations, the root-mean-square-deviation threshold was established as 1.5 Å, and all conformations of the ligands in the binding pocket of the macromolecule were clustered. The lowest energy clusters were identified, and the binding energy was evaluated. PyMOL (version 2.4.1) was used to create binding poses.

Synthesis
As shown in Scheme 1, these quinoxaline-arylfuran derivatives were synthesized according to the literatures [23,24]. Briefly, the synthesis begins with (2,5-dimethoxyphenyl) boronic acid turned into intermediates 11a-c, followed by treatment with NBS in DCM to obtain bromo derivatives 12a-c, which yield arylfuran derivatives 14a-c through treatment with 2-formylfuran-5-boronic acid. The intermediates 14a-c were reacted with hydrazide compounds to generate the corresponding hydrazone derivatives QW1-24. All synthesized quinoxalines compounds were tested for in vitro cytotoxicity against four cancer cell lines (HeLa, PC3, A549, and HCT116). The inhibitory effects of these compounds were evaluated using the CCK-8 assay after treatment with 20 μM of each compound. As shown in Table 1, most of the compounds exerted weak inhibitory activity against HeLa, PC3, A549, and HCT116 cell lines. Apparently, the quinoxaline scaffold is Scheme 1. Synthetic routes for compounds QW1-24. Regents and conditions: (a) Ar-Br, Pd(dppf)Cl 2 , Cs 2 CO 3 , dioxane, 90 • C; (b) NBS, DCM, rt; (c) Pd(dppf)Cl 2 , Cs 2 CO 3 , dioxane, 90 • C; (d) RCONHNH 2 , AcOH, EtOH. All synthesized quinoxalines compounds were tested for in vitro cytotoxicity against four cancer cell lines (HeLa, PC3, A549, and HCT116). The inhibitory effects of these compounds were evaluated using the CCK-8 assay after treatment with 20 µM of each compound. As shown in Table 1, most of the compounds exerted weak inhibitory activity against HeLa, PC3, A549, and HCT116 cell lines. Apparently, the quinoxaline scaffold is well tolerated, and most of the quinoxaline derivatives (QW1-16) exhibited higher activity than the naphthalene ring substitution derivatives (QW17-24). The species of phenyl substituents have a significant effect on the activity. The hydrazide side chain is p-cresol or pyridine, which showed higher inhibitory activity, such as in QW3-5 and QW11-13, indicating that the side chain with a hydrogen-bond acceptor or donor is more appropriate for promoting activity. This result was consistent with our molecular docking analysis in the docking-study section. By comparison, hydrophobic substitution at the phenyl ring gave derivatives with relatively low potency (QW7-8). In addition, large substituents on the phenyl ring exerted an unfavorable influence on the potency of the compounds (QW8 and QW16). Among these compounds, QW12 exhibited excellent activities against HeLa, PC3, A549, and HCT116 cancer cell lines, with inhibition rates of 76.35%, 52.55%, 50.78%, and 65.43%, respectively. The positive control we used was nifuroxazide, a STAT3 inhibitor, and its inhibition rates of HeLa, PC3, A549, and HCT116 were 80.38%, 75.29%, 91.23%, and 85.31%, respectively.                To further evaluate the anticancer effect of QW12, we evaluated its cytotoxicity against four human cancer cell lines: HeLa (Human cervical tumor), PC3 (Human prostate tumor), HCT116 (Human colorectal carcinomas), and A549 (Human lung carcinoma). To demonstrate whether QW12 would show the expected selectivity between normal cells vs. cancer cells, the cytotoxicity of QW12 against normal human hepatocytes L02 was also determined. As presented in Table 2, the IC 50 value of QW12 against A549 was 20.57 µM, while it showed potent cytotoxicity at low micromolar concentration (10.58-12.67 µM) against the other tumor cell lines. Furthermore, the IC 50 values of QW12 against PC3 and HCT116 were similar to those of nifuroxazide. Interestingly, compound QW12 showed higher safety against normal human hepatocyte L02 cells than nifuroxazide (>50 µM vs. 25.30 µM).

QW12 Inhibits the Proliferation and Migration of HeLa Cells
The wound-healing assay is a simple and cost-effective way to assess the invasiveness and migration of cells. In this study, a HeLa cell wound-healing assay under different treated conditions was carried out to observe the effect of QW12 on HeLa cell migration. As shown in Figure 3, the change in scratch closure indicated that cells treated with QW12 significantly prevented wound healing. After 48 h, the rate of scratch in closure with different concentrations of QW12 (10 and 40 µM) was 28.0% and 17.2%, respectively. All values in the QW12-treated groups were much lower than in the control group (55.7%). These data indicated that QW12 markedly inhibited HeLa cell would-healing in a doseand time-dependent manner, which means that QW12 can inhibit HeLa cell migration.  To evaluate the effect of QW12 on cell proliferation, we performed plate-cloning assays. In the plate-cloning experiment, cells treated with different concentrations of QW12 proliferated at a significantly lower rate than the control ( Figure 4B). Therefore, the results implied that QW12 could markedly inhibit HeLa cell proliferation. To evaluate the effect of QW12 on cell proliferation, we performed plate-cloning assays. In the plate-cloning experiment, cells treated with different concentrations of QW12 proliferated at a significantly lower rate than the control ( Figure 4B). Therefore, the results implied that QW12 could markedly inhibit HeLa cell proliferation.
HeLa cells. (B) Quantitative analysis of scratch closure changes. Values are presented in means±SD of three independent experiments., * p < 0.05, ** p < 0.01.
To evaluate the effect of QW12 on cell proliferation, we performed plate-cloning assays. In the plate-cloning experiment, cells treated with different concentrations of QW12 proliferated at a significantly lower rate than the control ( Figure 4B). Therefore, the results implied that QW12 could markedly inhibit HeLa cell proliferation.

QW12 Induces Intracellular ROS Production
Reactive oxygen species (ROS) are generated as by-products of normal aerobic metabolism or as second messengers in various signal transduction pathways in response to oxidative stress, which can elicit a wide spectrum of biological responses, such as macromolecular damage and cell death [26][27][28]. Substantial evidence suggests that the generation of ROS is part of the mechanism by which many anticancer agents kill tumor cells [29][30][31]. It is reported that nifuroxazide could induce apoptosis through ROS accumulation [15]. Although the 5-nitro group is believed to be the main responsible mechanism for the generation of ROS for nifuroxazide, there is evidence that ROS generation can also be independent of the presence of the nitro group [32]. Since the side chain of QW12 is similar to nifuroxazide, we were motivated to figure out whether QW12 could also trigger ROS production in HeLa cells. As illustrated in Figure 5A,B, compound QW12 stimulates ROS accumulation in a dose-dependent manner in HeLa cells. To further confirm the correlation between QW12-induced intracellular ROS accumulation and antiproliferation

QW12 Induces Intracellular ROS Production
Reactive oxygen species (ROS) are generated as by-products of normal aerobic metabolism or as second messengers in various signal transduction pathways in response to oxidative stress, which can elicit a wide spectrum of biological responses, such as macromolecular damage and cell death [25][26][27]. Substantial evidence suggests that the generation of ROS is part of the mechanism by which many anticancer agents kill tumor cells [28][29][30]. It is reported that nifuroxazide could induce apoptosis through ROS accumulation [15]. Although the 5-nitro group is believed to be the main responsible mechanism for the generation of ROS for nifuroxazide, there is evidence that ROS generation can also be independent of the presence of the nitro group [31]. Since the side chain of QW12 is similar to nifuroxazide, we were motivated to figure out whether QW12 could also trigger ROS production in HeLa cells. As illustrated in Figure 5A,B, compound QW12 stimulates ROS accumulation in a dose-dependent manner in HeLa cells. To further confirm the correlation between QW12-induced intracellular ROS accumulation and antiproliferation activity, HeLa cells were pretreated with scavenger N-acetylcysteine (NAC) for 1 h, then QW12 was added, and cell viability was examined. The results showed that the NAC pretreatment of cells blocked the accumulation of ROS induced by QW12 ( Figure 5D,E), and NAC significantly blocked the antiproliferative effect of QW12 ( Figure 5C), suggesting that the elevation of ROS levels is a critical event in the proliferation inhibition induced by QW12. activity, HeLa cells were pretreated with scavenger N-acetylcysteine (NAC) for 1 h, then QW12 was added, and cell viability was examined. The results showed that the NAC pretreatment of cells blocked the accumulation of ROS induced by QW12 ( Figure 5D, E), and NAC significantly blocked the antiproliferative effect of QW12 ( Figure 5C), suggesting that the elevation of ROS levels is a critical event in the proliferation inhibition induced by QW12.
Previous studies have reported that ROS could inhibit the PI3K/AKT pathway in various tumor cells treated with antitumor drugs [33,34]. Therefore, we attempted to explore the effect of QW12 on this pathway in HeLa cells. As shown in Figure 5F, treatment with QW12 significantly decreased phosphorylated AKT levels (T308), while total AKT was not affected. Furthermore, the inhibitory effect of QW12 on AKT phosphorylation could be reversed by NAC treatment. Thus, these results confirm that QW12 could block the PI3K/AKT pathways in a ROS-dependent fashion.   Previous studies have reported that ROS could inhibit the PI3K/AKT pathway in various tumor cells treated with antitumor drugs [32,33]. Therefore, we attempted to explore the effect of QW12 on this pathway in HeLa cells. As shown in Figure 5F, treatment with QW12 significantly decreased phosphorylated AKT levels (T308), while total AKT was not affected. Furthermore, the inhibitory effect of QW12 on AKT phosphorylation could be reversed by NAC treatment. Thus, these results confirm that QW12 could block the PI3K/AKT pathways in a ROS-dependent fashion.

QW12 Induces the Apoptosis of HeLa Cells
Hoechst 33324 dyes can stain concentrated nuclei, thus distinguishing apoptotic cells from normal cells. Therefore, we used Hoechst 33,342 to stain the nucleus of HeLa cells to analyze the inhibitory effect of QW12 on cell apoptosis. As shown in Figure 6A, significant apoptotic morphological characteristics were observed in the nucleus of cells treated with QW12 compared to the control group, while the nucleus of control cells was rectangular in shape without any apoptotic bodies.
Annexin V-FITC is used to detect the early stage of apoptosis. PI can pass through the cell membrane to stain the cells in the middle and late stages of apoptosis. Therefore, when Annexin V-FITC is used in combination with PI, cells in different stages of apoptosis can be distinguished. As shown in Figure 6B, QW12-treated HeLa cells clustered mostly in the upper right quadrant. Cell apoptosis rates treated with QW12 at concentrations of 10 µM and 40 µM are 13.9% and 34.7%, respectively. They were much higher than in the control group (3.5%) ( Figure 6C). These results suggested that QW12 has great potential application in inducing HeLa cell apoptosis.
To investigate whether the expression of pro-or anti-apoptotic proteins in the Bcl-2 family undergoes changes during QW12 treatment, we determined the expression of Bax and Bcl-2. Bax expression increased slightly, while Bcl-2 expression decreased significantly compared to the control group ( Figure 6D). Therefore, Western blot analysis ( Figure 6E) revealed a significant downregulation of the Bcl-2/Bax expression ratio in the 40 µM (0.66 ± 0.01) group compared to the control group (1.00) ( Figure 6E).

QW12 Blocks HeLa Cells in Phase S
To investigate the effect of QW12 on the distribution of the cell cycle, a flow cytometric analysis was performed. HeLa cells were treated with 0, 10, and 40 µM of QW12 for 24 h and then subjected to a flow cytometric analysis after DNA staining. As shown in Figure 7A, the untreated cells exhibited the expected pattern for continuously growing cells, while the cells treated with QW12 progressively increased during the S phase of the cell cycle at 10 and 40 µM. For example, the HeLa cell population gradually increased from 23.19% at 10 µM to 51.88% at 40 µM in the S phase ( Figure 7B). Simultaneously, the percentage of cells in phases G0/G1 and G2/M decreased significantly, indicating that QW12 mainly arrests the cell cycle in phase S. h and then subjected to a flow cytometric analysis after DNA staining. As shown in Figure  7A, the untreated cells exhibited the expected pattern for continuously growing cells, while the cells treated with QW12 progressively increased during the S phase of the cell cycle at 10 and 40 μM. For example, the HeLa cell population gradually increased from 23.19% at 10 μM to 51.88% at 40 μM in the S phase ( Figure 7B). Simultaneously, the percentage of cells in phases G0/G1 and G2/M decreased significantly, indicating that QW12 mainly arrests the cell cycle in phase S.

QW12 Inhibits STAT3 Y705 Phosphorylation in Cell-Based Assays
The activation of STAT3 is tightly regulated during various physiological processes, and the aberrant and persistent activation of STAT3 has been found in various types of cancer [34][35][36]. Nifuroxazide, which has a furan-hydrazone core similar to QW12, inhibited STAT3 [14][15][16]. Western blot analysis determined whether QW12 could also regulate the STAT3 signaling pathway in HeLa cells. HeLa cells were treated with QW12 for 12 h, and total cell lysates were prepared; Y705-phosphorylated STAT3 proteins were detected using a specific antibody. As shown in Figure 8, QW12 inhibited Tyr705 phosphorylation in a dose-dependent manner, and the total STAT3 level also decreased at these concentrations. Therefore, these results indicated that QW12 effectively inhibit STAT3 phosphorylation. The activation of STAT3 is tightly regulated during various physiological processes, and the aberrant and persistent activation of STAT3 has been found in various types of cancer [34][35][36]. Nifuroxazide, which has a furan-hydrazone core similar to QW12, inhibited STAT3 [14][15][16]. Western blot analysis determined whether QW12 could also regulate the STAT3 signaling pathway in HeLa cells. HeLa cells were treated with QW12 for 12 h, and total cell lysates were prepared; Y705-phosphorylated STAT3 proteins were detected using a specific antibody. As shown in Figure 8, QW12 inhibited Tyr705 phosphorylation in a dose-dependent manner, and the total STAT3 level also decreased at these concentrations. Therefore, these results indicated that QW12 effectively inhibit STAT3 phosphorylation. h and then subjected to a flow cytometric analysis after DNA staining. As shown in Figure  7A, the untreated cells exhibited the expected pattern for continuously growing cells, while the cells treated with QW12 progressively increased during the S phase of the cell cycle at 10 and 40 μM. For example, the HeLa cell population gradually increased from 23.19% at 10 μM to 51.88% at 40 μM in the S phase ( Figure 7B). Simultaneously, the percentage of cells in phases G0/G1 and G2/M decreased significantly, indicating that QW12 mainly arrests the cell cycle in phase S. The activation of STAT3 is tightly regulated during various physiological processes, and the aberrant and persistent activation of STAT3 has been found in various types of cancer [35][36][37]. Nifuroxazide, which has a furan-hydrazone core similar to QW12, inhibited STAT3 [14][15][16]. Western blot analysis determined whether QW12 could also regulate the STAT3 signaling pathway in HeLa cells. HeLa cells were treated with QW12 for 12 h, and total cell lysates were prepared; Y705-phosphorylated STAT3 proteins were detected using a specific antibody. As shown in Figure 8, QW12 inhibited Tyr705 phosphorylation in a dose-dependent manner, and the total STAT3 level also decreased at these concentrations. Therefore, these results indicated that QW12 effectively inhibit STAT3 phosphorylation.

Kinetic Affinity of QW12 versus STAT3
To confirm whether QW12 is a direct STAT3 inhibitor, a BLI (biolayer interferometry) analysis was performed. Briefly, recombinant STAT3 with a His tag was immobilized on SSA biosensors, and the dissociation constants were determined by measuring the binding to serial dilutions of QW12 at concentrations ranging from 5.2 to 333 µM ( Figure 9A). The signal was collected, and the software calculated the kinetic affinity, which showed that the steady-state curve reaches saturation ( Figure 9B). The binding affinity (KD) was calculated as 67.3 µM. The results indicated that compound QW12 has a moderate binding affinity to the STAT3 protein.

Kinetic Affinity of QW12 Versus STAT3
To confirm whether QW12 is a direct STAT3 inhibitor, a BLI (biolayer interferometry) analysis was performed. Briefly, recombinant STAT3 with a His tag was immobilized on SSA biosensors, and the dissociation constants were determined by measuring the binding to serial dilutions of QW12 at concentrations ranging from 5.2 to 333 μM ( Figure 9A). The signal was collected, and the software calculated the kinetic affinity, which showed that the steady-state curve reaches saturation ( Figure 9B). The binding affinity (KD) was calculated as 67.3 μM. The results indicated that compound QW12 has a moderate binding affinity to the STAT3 protein.

Compound QW12 Docking Study with the STAT3 SH2 Domain
To account for the potent STAT3 phosphorylation-inhibition activities of QW12 in HeLa cells, a molecular docking study was carried out to evaluate potential interactions with the STAT3 SH2 domain (PDB: 1BG1). As shown in Figure 10, the hydroxyl group in the phenyl ring of the side-chain hydrazone formed two hydrogen bonds with residues Arg595 (3.0 Å ) and Lys591 (2.7 Å ), and the carbonyl oxygen atom of the hydrazone moiety also interacted with residue Lys591 (3.3 Å ) through hydrogen bonding. The N atom in the hydrazone group could interact with Ser636 (3.3 Å ) by hydrogen bonding. The oxygen atom of the methoxy group immediately next to the furan ring formed a hydrogen bond with Gln635 (3.4 Å ). On the basis of in silico results, it would be beneficial to introduce a side chain on the methoxyphenyl ring to occupy the pY subpocket, thereby increasing the selectivity and binding affinity. Moreover, a suitable conformation of QW12 contributed a lot to the tight binding. The predicted binding energy was −56.6 kcal/mol. These docking results explained that the hydrazone moiety and the substitution of p-hydroxyl may be necessary to target the STAT3 SH2 domain to some extent.

Compound QW12 Docking Study with the STAT3 SH2 Domain
To account for the potent STAT3 phosphorylation-inhibition activities of QW12 in HeLa cells, a molecular docking study was carried out to evaluate potential interactions with the STAT3 SH2 domain (PDB: 1BG1). As shown in Figure 10, the hydroxyl group in the phenyl ring of the side-chain hydrazone formed two hydrogen bonds with residues Arg595 (3.0 Å) and Lys591 (2.7 Å), and the carbonyl oxygen atom of the hydrazone moiety also interacted with residue Lys591 (3.3 Å) through hydrogen bonding. The N atom in the hydrazone group could interact with Ser636 (3.3 Å) by hydrogen bonding. The oxygen atom of the methoxy group immediately next to the furan ring formed a hydrogen bond with Gln635 (3.4 Å). On the basis of in silico results, it would be beneficial to introduce a side chain on the methoxyphenyl ring to occupy the pY subpocket, thereby increasing the selectivity and binding affinity. Moreover, a suitable conformation of QW12 contributed a lot to the tight binding. The predicted binding energy was −56.6 kcal/mol. These docking results explained that the hydrazone moiety and the substitution of p-hydroxyl may be necessary to target the STAT3 SH2 domain to some extent.
atom of the methoxy group immediately next to the furan ring formed a hydrogen bond with Gln635 (3.4 Å). On the basis of in silico results, it would be beneficial to introduce a side chain on the methoxyphenyl ring to occupy the pY subpocket, thereby increasing the selectivity and binding affinity. Moreover, a suitable conformation of QW12 contributed a lot to the tight binding. The predicted binding energy was −56.6 kcal/mol. These docking results explained that the hydrazone moiety and the substitution of p-hydroxyl may be necessary to target the STAT3 SH2 domain to some extent.

Conclusions
In this study, we designed and synthesized a series of quinoxaline-furan derivatives and evaluated their antiproliferative effects in vitro. Among these compounds, QW12 was found to be the most potent hybrid against five cancer cell lines, especially inhibiting HeLa cells with an IC50 value of 10.58 μM. QW12 was also shown to inhibit the tumor migration and invasion of HeLa cells. Furthermore, it increased the production and accumulation of ROS in HeLa cells, which accounts for its partial antiproliferation effect through inducing cell apoptosis. The Hoechst 33342 and annexin V-FITC/PI staining experiments proved that QW12 induced the apoptosis of HeLa cells. The protein expression of anti-apoptotic and pro-apoptotic proteins Bcl-2 and Bax was also affected by QW12 in a concentration-dependent manner. Western blot analysis indicated that QW12 inhibited STAT3 phosphorylation levels. The kinetic affinity assay confirmed that QW12 could directly bind to STAT3 with a KD value of 67.3 μM. The docking study implied that the compound QW12 bound to the cavities pY+1 and pY-X of the STAT3 SH2 domain. In summary, all of these data provided a structural reference for the development of a novel scaffold and candidate for tumor treatment.

Conclusions
In this study, we designed and synthesized a series of quinoxaline-furan derivatives and evaluated their antiproliferative effects in vitro. Among these compounds, QW12 was found to be the most potent hybrid against five cancer cell lines, especially inhibiting HeLa cells with an IC 50 value of 10.58 µM. QW12 was also shown to inhibit the tumor migration and invasion of HeLa cells. Furthermore, it increased the production and accumulation of ROS in HeLa cells, which accounts for its partial antiproliferation effect through inducing cell apoptosis. The Hoechst 33342 and annexin V-FITC/PI staining experiments proved that QW12 induced the apoptosis of HeLa cells. The protein expression of anti-apoptotic and pro-apoptotic proteins Bcl-2 and Bax was also affected by QW12 in a concentration-dependent manner. Western blot analysis indicated that QW12 inhibited STAT3 phosphorylation levels. The kinetic affinity assay confirmed that QW12 could directly bind to STAT3 with a KD value of 67.3 µM. The docking study implied that the compound QW12 bound to the cavities pY+1 and pY-X of the STAT3 SH2 domain. In summary, all of these data provided a structural reference for the development of a novel scaffold and candidate for tumor treatment.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.