Design, Synthesis and Bioactivity Evaluation of 4,6-Disubstituted Pyrido[3,2-d]pyrimidine Derivatives as Mnk and HDAC Inhibitors

Both HDACs and Mnks play important role in translating multiple oncogenic signaling pathways during oncogenesis. As HDAC and Mnk are highly expressed in a variety of tumors; thus simultaneous inhibit HDAC and Mnk can increase the inhibition of tumor cell proliferation and provide a new way of inhibiting tumor growth. Based on the previous work and the merge pharmacophore method; we designed and synthesized a series of 4,6-disubstituted pyrido[3,2-d]pyrimidine derivatives as HDAC and Mnk dual inhibitors. Among them; compound A12 displayed good HDAC and Mnk inhibitory activity. In vitro antiproliferative assay; compound A12 exhibited the best antiproliferative activity against human prostate cancer PC-3 cells. Docking study revealed that the pyrido[3,2-d]pyrimidine framework and hydroxamic acid motif of compound A12 were essential for maintaining the activity of HDAC and Mnk. These result indicated that A12 was a potent Mnk /HDAC inhibitor and will be further researched.


Introduction
Histone deacetylase (HDAC) plays an important role in regulating the acetylation level of histones or non-histone proteins, thus it can regulate cell proliferation, differentiation and apoptosis [1]. The abnormal changes in histone acetylation levels often induce the expression of oncogene. Thus, targeting HDAC could be a strategy to treat cancer. HDAC inhibitors can inhibit cell growth and differentiation by increasing histone and non-histone acetylation levels in tumor cells, and further promote apoptosis. So far, five HDAC inhibitors have been approved by the U.S. FDA and a large number of HDAC inhibitors have been reported. Although HDAC inhibitors show promising therapeutic effects in hematological cancers, but they show poor effect in solid tumors. For example, vorinostat has an unsatisfactory effect on breast cancer, colorectal cancer, non-small cell lung cancer and thyroid cancer [2]. Romidepsin, belisestat and pabistil are also not effective in solid tumors [3]. On the other hand, HDAC inhibitors have distinct side effects, which affect the patients' medication compliance, so simultaneously inhibiting HDAC and another target could show a higher therapeutic effect than only inhibiting HDAC.
Multitarget-directed ligands are agents that treat multifactorial disease by interacting with multiple target involved in pathogenesis [4]. This strategy can avoid many pitfalls, such as dose-limiting toxicities, drug−drug interactions and complex and unpredictable pharmacokinetic profiles. Dual/multi-target inhibitors based on HDAC and kinases have made significant progress in the last decade [5]. Among them, CUDC-101 [6] and CUDC-907 [7] were already in clinical research stage ( Figure 1). decade [5]. Among them, CUDC-101 [6] and CUDC-907 [7] were already in clinical research stage ( Figure 1). CUDC-101, which was obtained by combining vorinostat and erlotinib, could simultaneously inhibit HDAC, EGFR, HER2, whose IC50 value against HDAC, EGFR and HER2 were 4.4 nM, 2.4 nM and 15.7 nM, respectively. In multiple tumor cell lines, CUDC-101 has better antiproliferative effect than vorinostat, erlotinib and the combination of vorinostat and erlotinib. Currently, CUDC-101 is in phase I clinical trials. CUDC-907, which could simultaneously inhibit HDAC and PI3K, inhibited tumor growth in a dose-dependent manner in a Daudi non-Hodgkin lymphoma xenograft model. In vivo anti-tumor activity studies have shown that CUDC-907 can cause accumulation of Ac-H3 and a decrease in p-AKT levels. CUDC-907 is currently undergoing a phase II clinical trial for patients with lymphoma and advanced thyroid carcinoma. Although a large number of multiple inhibitors based on inhibit HDAC and other targets have been reported, only two compounds have been in clinical research. Therefore, it is necessary to develop multi-target inhibitors based on HDAC.
Mitogen-activated protein kinase interacting kinases (Mnk1 and Mnk2) belong to serine/threonine kinase and are the only known kinases that can phosphorylate eIF4E at Ser209 [8,9]. In humans, Mnks are presented in two isoforms: Mnk1 and Mnk2, and they share 80% sequence identity [10]. So it is feasible to simultaneously inhibite Mnk1 and Mnk2. Although the MnK signaling pathway is branched at multiple directions, but the Mnk double knock-out does not affect the normal growth of mice [11], which indicated that Mnks are dispensable in normal cells. In addition, Mnks can mediate the resistance of many clinical drugs [12,13], such as Imatinib and Cytarabine. Therefore, the combination of Mnk inhibitors or the design of dual-target inhibitors that simultaneously targeting Mnk and other targets may have better anti-tumor effects. In summary, inhibiting Mnks is expected to become a low-toxic and efficient anti-tumor solution.
So far, no Mnk inhibitor has been approved for marketing. Only three Mnk inhibitors have entered clinical research ( Figure 1). BAY1143269 [14], which is a selective Mnk1 inhibitor and which structure has not been disclosed, was combined with docetaxel in a clinical phase I study for the treatment of non-small cell lung cancer, but that clinical study has now been terminated [15]. eFT508 is a selective Mnk inhibitor that has entered phase II clinical research for the treatment of solid tumors [16]. ETC-206 combined with dasatinib entered a clinical phase I study for the treatment of blast crisischronic myeloid leukemia [17]. We previously found that 4,6-disubstituted pyrido [3,2-d]pyrimidine derivatives as good Mnk inhibitors [18]. Among them, compound D06 showed good Mnk inhibitory activity, the IC50 against Mnk1 and Mnk2 were 1.1 nM and 26 nM, respectively.  Mnk is located downstream of the MAPK signaling pathway and thus can regulate cell survival and proliferation. What's more, HDAC and Mnk are highly expressed in a variety of tumors, and there are cases where both are overexpressed in a variety of cancers, such as colon cancer, prostate cancer and leukemia. In addition, HDAC and Mnk regulate the expression of some common oncoproteins, such as c-Myc, Cyclin D1 and VEGF ( Figure 2). Therefore, small molecule inhibitors that target both HDAC and Mnk can increase the inhibition of tumor cell proliferation and provide a new way of inhibiting tumor growth.
Molecules 2020, 25, x FOR PEER REVIEW 3 of 18 Mnk is located downstream of the MAPK signaling pathway and thus can regulate cell survival and proliferation. What's more, HDAC and Mnk are highly expressed in a variety of tumors, and there are cases where both are overexpressed in a variety of cancers, such as colon cancer, prostate cancer and leukemia. In addition, HDAC and Mnk regulate the expression of some common oncoproteins, such as c-Myc, Cyclin D1 and VEGF ( Figure 2). Therefore, small molecule inhibitors that target both HDAC and Mnk can increase the inhibition of tumor cell proliferation and provide a new way of inhibiting tumor growth. Molecular docking performed between D06 and Mnk2 (PDB code 2hw7) showed that the pyrido [3,2-d]pyrimidine core could imitate the structure of purine in ATP and made hydrogen bond interaction with Met162 in hinge. The 4-fluoroaniline group could occupy the gatekeeper hydrophobic region and made π-π stacking interaction with Phe159. The 6-position substituent at the core extended to solvent exposed region ( Figure 3). Based on the strategy of merge pharmacophore, we took the key fragment N-phenylpyrido [3,2d]pyrimidin-4-amine as the recognition cap of HDAC and introduce zinc-binding motif (ZBG) through suitable linker in the 6-position of the core, designed a series of small molecule compounds simultaneously targeting both Mnk and HDAC ( Figure 4). Molecular docking performed between D06 and Mnk2 (PDB code 2hw7) showed that the pyrido [3,2-d]pyrimidine core could imitate the structure of purine in ATP and made hydrogen bond interaction with Met162 in hinge. The 4-fluoroaniline group could occupy the gatekeeper hydrophobic region and made π-π stacking interaction with Phe159. The 6-position substituent at the core extended to solvent exposed region ( Figure 3). Mnk is located downstream of the MAPK signaling pathway and thus can regulate cell survival and proliferation. What's more, HDAC and Mnk are highly expressed in a variety of tumors, and there are cases where both are overexpressed in a variety of cancers, such as colon cancer, prostate cancer and leukemia. In addition, HDAC and Mnk regulate the expression of some common oncoproteins, such as c-Myc, Cyclin D1 and VEGF ( Figure 2). Therefore, small molecule inhibitors that target both HDAC and Mnk can increase the inhibition of tumor cell proliferation and provide a new way of inhibiting tumor growth. Molecular docking performed between D06 and Mnk2 (PDB code 2hw7) showed that the pyrido [3,2-d]pyrimidine core could imitate the structure of purine in ATP and made hydrogen bond interaction with Met162 in hinge. The 4-fluoroaniline group could occupy the gatekeeper hydrophobic region and made π-π stacking interaction with Phe159. The 6-position substituent at the core extended to solvent exposed region ( Figure 3). Based on the strategy of merge pharmacophore, we took the key fragment N-phenylpyrido [3,2d]pyrimidin-4-amine as the recognition cap of HDAC and introduce zinc-binding motif (ZBG) through suitable linker in the 6-position of the core, designed a series of small molecule compounds simultaneously targeting both Mnk and HDAC ( Figure 4). Based on the strategy of merge pharmacophore, we took the key fragment N-phenylpyrido [3,2-d]pyrimidin-4-amine as the recognition cap of HDAC and introduce zinc-binding motif (ZBG) through suitable linker in the 6-position of the core, designed a series of small molecule compounds simultaneously targeting both Mnk and HDAC ( Figure 4).

Chemistry
The synthesis route of A01 and A07 is described in Scheme 1. Starting from commercially available 2-cyano-3-nitro-6-chloropyridine, the intermediate 1 was obtained by reduction of the nitro group with stannous chloride dihydrate which was accompanied by hydrolysis of the nitrile.

Chemistry
The synthesis route of A01 and A07 is described in Scheme 1. Starting from commercially available 2-cyano-3-nitro-6-chloropyridine, the

Chemistry
The synthesis route of A01 and A07 is described in Scheme 1. Starting from commercially available 2-cyano-3-nitro-6-chloropyridine, the intermediate 1 was obtained by reduction of the nitro group with stannous chloride dihydrate which was accompanied by hydrolysis of the nitrile.  The intermediate 4 was obtained by nucleophilic substitution of 3 with 4-fluoroaniline in the presence of isopropanol, then a palladium catalyzed Suzuki-coupling to furnish the carboxylic acid 5. Subsequent treatment of the carboxylic acid with o-phenylenediamine was reacted in THF in the presence of HATU, DMAP and DIEPA to yield A01. Intermediate 5 was esterified and underwent ammonolysis with hydroxylamine aqueous solution under strong alkaline conditions to obtain A07.
The synthesis route of A02 and A08 is described in Scheme 2. The  The intermediate 4 was obtained by nucleophilic substitution of 3 with 4-fluoroaniline in the presence of isopropanol, then a palladium catalyzed Suzuki-coupling to furnish the carboxylic acid 5. Subsequent treatment of the carboxylic acid with o-phenylenediamine was reacted in THF in the presence of HATU, DMAP and DIEPA to yield A01. Intermediate 5 was esterified and underwent ammonolysis with hydroxylamine aqueous solution under strong alkaline conditions to obtain A07.
The synthesis route of A02 and A08 is described in Scheme 2. The   The synthesis route of target compounds A04-A06 and A10-A19 is shown in Scheme 4. Using 2-cyano-3-nitro-6-chloropyridine as starting material, nucleophilic substitution reactions with different amines afforded intermediates 12a-12c. The intermediates 15a-15c were obtained by cyclization and chlorination of 12a-12c. Intermediates 15a-15c underwent a C-N coupling reaction with different anilines in the presence of palladium catalyst to obtained intermediates 16a-16j. The target compounds A10-A19 were obtained by aminolysis reactions of intermediates 16a-16j. The intermediates 16a-16c afforded the target compounds A04-A06 by hydrolysis and condensation with o-phenylenediamine.
Molecules 2020, 25, x FOR PEER REVIEW 6 of 18 different amines afforded intermediates 12a-12c. The intermediates 15a-15c were obtained by cyclization and chlorination of 12a-12c. Intermediates 15a-15c underwent a C-N coupling reaction with different anilines in the presence of palladium catalyst to obtained intermediates 16a-16j. The target compounds A10-A19 were obtained by aminolysis reactions of intermediates 16a-16j. The intermediates 16a-16c afforded the target compounds A04-A06 by hydrolysis and condensation with o-phenylenediamine.

Bioactivity Assay
To investigate the HDAC and Mnk inhibitory potency of the target compounds, the inhibition rate of Mnk1 and Mnk2 at 1 μM were determined using homogeneous time-resolved fluorescence assay (HTRF) and the IC50 value against HDAC was determined by fluorescence analysis. Compound D06, Staurosporine and Vorinostat were used as positive controls. The results are shown in Tables  1 and 2.
Based on the structure of compound D06, we first introduced a zinc ion binding group (ZBG) through a linker with different length and rigidity to obtain the compounds A01-A12. The results showed that when the ZBG was hydroxamic acid motif, the inhibitory against Mnks and HDACs were better than o-phenylenediamine (A04, A05, A06 vs A10, A11, A12). The linker's properties also have significant impact on the enzyme inhibitory activity of these compounds. When the compound linkers were rigid, the inhibitory activities of these compounds on HDACs were poor. When the linkers were replaced with a flexible alkyl chain, the potency of these compounds on Mnks and HDACs were increased significantly (A07-A10 vs A11, A12). The length of the linker also had important effect on the potency. Although when the linker was composed of a NH and five methylene units, the compound has better inhibitory activity against HDACs (A11, IC50 = 0.18 μM), but its inhibitory activity against Mnks was poor (Mnk1: 49%, Mnk2: 61%). Compared to compound A11, when the length of linker was a NH and six methylene groups, despite the fact the potency against HDACs declined (A12, IC50 = 0.78 μM), the inhibitory rate against Mnk was obviously improved (Mnk1: 54%, Mnk2: 70%), but the inhibitory activity of compounds A01-A12 against Mnk declined differently compared with D06.

Bioactivity Assay
To investigate the HDAC and Mnk inhibitory potency of the target compounds, the inhibition rate of Mnk1 and Mnk2 at 1 µM were determined using homogeneous time-resolved fluorescence assay (HTRF) and the IC 50 value against HDAC was determined by fluorescence analysis. Compound D06, Staurosporine and Vorinostat were used as positive controls. The results are shown in Tables 1 and 2.
Based on the structure of compound D06, we first introduced a zinc ion binding group (ZBG) through a linker with different length and rigidity to obtain the compounds A01-A12. The results showed that when the ZBG was hydroxamic acid motif, the inhibitory against Mnks and HDACs were better than o-phenylenediamine (A04, A05, A06 vs. A10, A11, A12). The linker's properties also have significant impact on the enzyme inhibitory activity of these compounds. When the compound linkers were rigid, the inhibitory activities of these compounds on HDACs were poor. When the linkers were replaced with a flexible alkyl chain, the potency of these compounds on Mnks and HDACs were increased significantly (A07-A10 vs. A11, A12). The length of the linker also had important effect on the potency. Although when the linker was composed of a NH and five methylene units, the compound has better inhibitory activity against HDACs (A11, IC 50 = 0.18 µM), but its inhibitory activity against Mnks was poor (Mnk1: 49%, Mnk2: 61%). Compared to compound A11, when the length of linker was a NH and six methylene groups, despite the fact the potency against HDACs declined (A12, IC 50 = 0.78 µM), the inhibitory rate against Mnk was obviously improved (Mnk1: 54%, Mnk2: 70%), but the inhibitory activity of compounds A01-A12 against Mnk declined differently compared with D06. In order to improve the inhibitory activity of the compounds against HDAC while maintaining the Mnk activity, we fixed the linker as a flexible alkyl chain containing a NH unite and six methylene groups and introduced different substituents in the phenyl at the 4-position of the core to investigate the inhibitory activity of the compounds against Mnks and HDACs. The results are shown in Table  2, where it can be seen that when the 2-position or 3-position was substituted, the inhibitory activity of the compounds against Mnks declined (A13-A19 vs A12). This may be due to the increased steric hindrance when there are too many substituents on the 4-position of the benzene ring, and this may block the 4-substituted benzene ring from occupying the hydrophobic pocket. Meanwhile, the inhibitory activity against HDAC was changed differently. When the 2-position or 3-position was substituted, most compounds showed better HDAC inhibitory activity than A12, except for compounds A13 and A19. Especially when the substituent was an electron-withdrawing group, the inhibitory activity against HDAC increased significantly (A16, A17 vs A12). It was worth noting that when the substituents on the benzene ring were 2-methylcarbamoyl-4-fluoro, the compound A16 showed the best HDAC inhibitory activity among all the synthesized compounds, whose IC50 value against HDAC was 0.037 μM and was better than positive control Vorinostat. Unfortunately, the inhibitory activity of compound A16 on Mnk declined dramatically compared to compound A12.  In order to improve the inhibitory activity of the compounds against HDAC while maintaining the Mnk activity, we fixed the linker as a flexible alkyl chain containing a NH unite and six methylene groups and introduced different substituents in the phenyl at the 4-position of the core to investigate the inhibitory activity of the compounds against Mnks and HDACs. The results are shown in Table  2, where it can be seen that when the 2-position or 3-position was substituted, the inhibitory activity of the compounds against Mnks declined (A13-A19 vs A12). This may be due to the increased steric hindrance when there are too many substituents on the 4-position of the benzene ring, and this may block the 4-substituted benzene ring from occupying the hydrophobic pocket. Meanwhile, the inhibitory activity against HDAC was changed differently. When the 2-position or 3-position was substituted, most compounds showed better HDAC inhibitory activity than A12, except for compounds A13 and A19. Especially when the substituent was an electron-withdrawing group, the inhibitory activity against HDAC increased significantly (A16, A17 vs A12). It was worth noting that when the substituents on the benzene ring were 2-methylcarbamoyl-4-fluoro, the compound A16 showed the best HDAC inhibitory activity among all the synthesized compounds, whose IC50 value against HDAC was 0.037 μM and was better than positive control Vorinostat. Unfortunately, the inhibitory activity of compound A16 on Mnk declined dramatically compared to compound A12. In order to improve the inhibitory activity of the compounds against HDAC while maintaining the Mnk activity, we fixed the linker as a flexible alkyl chain containing a NH unite and six methylene groups and introduced different substituents in the phenyl at the 4-position of the core to investigate the inhibitory activity of the compounds against Mnks and HDACs. The results are shown in Table 2, where it can be seen that when the 2-position or 3-position was substituted, the inhibitory activity of the compounds against Mnks declined (A13-A19 vs. A12). This may be due to the increased steric hindrance when there are too many substituents on the 4-position of the benzene ring, and this may block the 4-substituted benzene ring from occupying the hydrophobic pocket. Meanwhile, the inhibitory activity against HDAC was changed differently. When the 2-position or 3-position was substituted, most compounds showed better HDAC inhibitory activity than A12, except for compounds A13 and A19. Especially when the substituent was an electron-withdrawing group, the inhibitory activity against HDAC increased significantly (A16, A17 vs. A12). It was worth noting that when the substituents on the benzene ring were 2-methylcarbamoyl-4-fluoro, the compound A16 showed the best HDAC inhibitory activity among all the synthesized compounds, whose IC 50 value against HDAC was 0.037 µM and was better than positive control Vorinostat. Unfortunately, the inhibitory activity of compound A16 on Mnk declined dramatically compared to compound A12.

A12-A19
As Mnks and HDACs are overexpressed in prostate cancer, the GI 50 values of the target compounds against PC-3 cell line were determined by an MTT assay. The results are presented in Table 3. The results showed that most of the target compounds showed moderate or excellent anti-proliferative effect on PC-3 cells. When the zinc-binding motif was hydroxamic acid, the anti-proliferative effect was better than o-phenylenediamine (except A03 and A04). The length of the flexible linker has a significant influence on the anti-proliferative activity: compounds containing a NH unite and 6 methylene groups were greatly better than compounds containing a NH unite and 5 methylene groups (A11 vs. A12).
The inhibitory activity of compounds on Mnks and HDACs was closely related to their anti-proliferative activity: when the compound only showed good inhibitory activity on Mnk or HDAC, the cell activity of the compound may be poor (A11, A14, A16, A17 vs. A12). When the substituent was 4-fluoroaniline, the compounds showed best anti-proliferative activity among the other substituents (A13-A19 vs. A12). The anti-proliferative activity of the most potent compound A12 (GI 50 = 2.91 µM) was better than D06 and equivalent to vorinostat, which should be further studied.

Docking Study
In order to better understand the binding mode of these compounds to Mnks and HDACs protein, we chose A12 to conduct a molecular docking study. The results are shown in Figure 5. As illustrated in Figure 5A,B, the docking result of A12 with Mnk2 is basically the same as D06: The N1 of pyrido [3,2-d]pyrimidine core can form conventional hydrogen bond interactions with Met162 in the hinge region. The 4-fluoroaniline can occupy the gatekeeper hydrophobic region and form a π-π stacking interaction with Phe159. The F atom can form a halogen interaction with Lys113 and Asp226 in the hydrophobic region. What's more, the NH and OH group of the hydroxamic acid motif can form hydrogen bond interactions with Glu92 in the solvent exposed region. This may be the reason why the compounds show better Mnk inhibitory when the hydroxamic acid is substituted. The possible binding mode of compound A12 with HDAC1 is exhibited in Figure 5C,D. Compound A12 binds to HDAC1 by embedding the flexible chain into the hydrophobic tunnel and anchoring the terminal hydroxamic acid group at the bottom of the pocket. Besides the coordination with the zinc ion, the hydroxamic acid group also forms hydrogen bond interactions with His131, His132 and Tyr297, respectively. In addition, the NH group of the linker can form hydrogen bond interactions with Glu92 in the hydrophobic tunnel. In the solvent exposed region, the pyridine ring of the pyrido [3,2-d]pyrimidine core and the 4-fluoroaniline part can form π-π stacking interactions with Phe200 and the F atom can form halogen interaction with Glu201. All in all, the predicted binding mode of compound A12 with Mnk2 and HDAC1 can further guide us to find more potent Mnk/HDAC dual inhibitors.
Molecules 2020, 25, x FOR PEER REVIEW 9 of 18 Phe200 and the F atom can form halogen interaction with Glu201. All in all, the predicted binding mode of compound A12 with Mnk2 and HDAC1 can further guide us to find more potent Mnk/HDAC dual inhibitors.
A B C D

General Information
Melting points were measured with an X-4 digital micro-melting point apparatus (Shanghai Jingke Co. Ltd, Shanghai, China ) and are uncorrected. 1 H-NMR spectra were measured with a ARX (600 MHz) spectrometer (Bruker, Karlsruhe, Germany) and 13 C-NMR with a Bruker AV (150 MHz) instrument. Chemical shifts are recorded in δ units using tetramethylsilane as the standard (NMR peak description: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad peak). Organic solutions were dried over anhydrous MgSO4 during workup. Column chromatography was carried out on a Combiflash Rf+ preparative liquid chromatograph (TELEDYNE ISCO, NE, USA). Silica gel 60 (200-300 mesh) and TLC were purchased from Qingdao Haiyang Chemical Co. Ltd. (Qingdao, China). All commercial reagents and solvents were used without further purification unless otherwise noted.

General Information
Melting points were measured with an X-4 digital micro-melting point apparatus (Shanghai Jingke Co. Ltd., Shanghai, China ) and are uncorrected. 1 H-NMR spectra were measured with a ARX (600 MHz) spectrometer (Bruker, Karlsruhe, Germany) and 13 C-NMR with a Bruker AV (150 MHz) instrument. Chemical shifts are recorded in δ units using tetramethylsilane as the standard (NMR peak description: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad peak). Organic solutions were dried over anhydrous MgSO 4 during workup. Column chromatography was carried out on a Combiflash Rf+ preparative liquid chromatograph (TELEDYNE ISCO, Lincoln, NE, USA). Silica gel 60 (200-300 mesh) and TLC were purchased from Qingdao Haiyang Chemical Co. Ltd. (Qingdao, China). All commercial reagents and solvents were used without further purification unless otherwise noted.

Procedure for the Synthesis of A01 and A07
3-Amino-6-chloropicolinamide (1) 2-Cyano-3-nitro-6-chloropyridine (2.0 g, 10.9 mmol), stannous chloride dihydrate (10.0 g, 44.3 mmol) and ethanol (20 mL) were added to a 50 mL pear-shaped flask and stirred at 78 • C for 2 h. After the reaction was complete, the solution was cooled to room temperature and the ethanol was evaporated to dryness. The residue was dissolved in ethyl acetate and the mixture was adjusted to pH 8 with a 2 mol/L aqueous sodium hydroxide solution, filtered and collected the organic layer, and the aqueous layer was washed twice with ethyl acetate. The organic layer was combined and washed twice with saturated sodium chloride, dried over anhydrous sodium sulfate, filtered and concentrated to obtain intermediate 1 as a yellow solid (3.0 g, yield 80.40%). (2) Intermediate 1 (2.5 g, 14.6 mmol) and triethyl orthoformate (100 mL, 60 mmol) were added to a 250 mL pear-shaped flask, and reacted at 146 • C for 2 h. After reaction completion was indicated by TLC, the mixture was cooled to room temperature, filtered and the solid dried to obtain intermediate 2 as a yellow solid (2.25 g, yield 85.14%).

4,6-Dichloropyrido[3,2-d]pyrimidine (3)
To a 100 mL pear-shaped flask intermediate 3 (1.81 g, 10 mmol), phosphorus oxychloride(36 mL) and 10 drops of N,N-dimethylaniline were added, followed by stirring for 1 h at 106 • C. The reaction was monitored by TLC. Upon completion the reaction mixture was cooled and evaporated. The obtained solid was redissolved with dichloromethane and the pH of the solution was adjusted to 8 with aqueous sodium hydrogen carbonate. The organic layer and the aqueous layer were separated, and the aqueous layer was extracted twice with dichloromethane. The organic layers were combined and dried over sodium sulfate, concentrated and purified by column chromatography on silica (petroleum ether:ethyl acetate = 100:3) to give intermediate 3 as a white solid (1.2 g, yield 60.61%).

4-(4-((4-Fluorophenyl)amino)pyrido[3,2-d]pyrimidin-6-yl)-N-hydroxybenzamide (A07)
Intermediate 6 (0.1 g, 0.27 mmol) was placed in a 250 mL pear-shaped flask and dissolved with 120 mL of methanol. The mixture was stirred for 5 min at 0 • C and then the pH was adjusted to 11 by adding 2 M NaOH aqueous solution. Then 5 mL aqueous hydroxylamine solution (50%) was added and the reaction was allowed to warm to room temperature. After the reaction was complete, as monitored by TLC, the mixture was partially evaporated and the pH of the residue was adjusted to 6 by adding 10% aqueous hydrochloric acid solution. The crude A07 was obtained by filtration and then purified by column chromatography on silica (dichloromethane:methanol = 15:1) to afford the target compound A07 as a light yellow solid (0.06 g, yield 59.23%). Mp: 249.0-251.

Procedure for the Synthesis of A02 and A08
(E)-3-(4-Boronophenyl)acrylic acid (7) A mixture of (4-acetylphenyl)boronic acid(0.10 g, 0.67 mmol), malonate (0.21 g, 0.20 mmol) and pyridine (0.053 g, 0.67 mmol) in dry toluene (10 mL) was stirred at reflux for 2 h. After the reaction was complete as indicated by TLC, the mixture was cooled to room temperature and poured into water. The pH of the mixture was adjust to 5 by adding 10% aqueous hydrochloric acid solution and then it was filtered to give intermediate 7 as a white solid (0.16 g, yield 80.73%). 50 mmol) and cesium carbonate (0.16 g, 5 mmol) in DMF (20 mL) was stirred at 110 • C under a N 2 atmosphere. After the reaction was complete, as indicated by TLC monitoring, the mixture was cooled to room temperature and the DMF was evaporated to dryness. The residue was redissolved with dichloromethane and water and brine were used to wash the mixture. The organic layer was then dried over sodium sulfate, concentrated and purified by column chromatography on silicagel (petroluem ether:ethyl acetate = 5:1) to gave the intermediate 10 as a yellow solid (0.54 g, yield 53.04%). (11) Intermediate 10 was placed in a 50 mL pear-shaped flask and dissolved with 10 mL of DMF. Then 10 mL 10% aqueous sodium hydroxide solution was added to the mixture that was then stirred at 60 • C for 4 h. After the reaction was complete as monitored by TLC, the mixture was poured to 100 mL of water and the pH of the mixture was adjusted to 5 using 10% aqueous hydrochloric acid solution and then filtered to gave the intermediate 11 as a light yellow solid (0.092 g, yield 95.3%).

General Procedure for the Synthesis of A04-A06
General Procedure for the Synthesis of 12a-12c A mixture of 2-cyano-3-nitro-6-chloropyridine (1.0 equivalent), the required amine (1.5 equivalents) and N,N-dimethylaniline (3 equivalents) in 10 mL of DMF was stirred at 60 • C for 2 h. After the reaction was complete as monitored by TLC, the mixture was cooled to room temperature, then poured into 100 mL water, stirred for 10 min and filtered to gave intermediates 12a-12c.

General Procedure for the Synthesis of 13a-13c
Intermediates 13a-13c were prepared from 12a-12c using the same reaction conditions described above for making intermediate 1.

General Procedure for the Synthesis of 14a-14c
Intermediates 14a-14c were prepared from 13a-13c using the same reaction conditions described above for making intermediate 2.

General Procedure for the Synthesis of 15a-15c
Intermediates 15a-15c were prepared from 14a-14c using the same reaction conditions described above for making intermediate 3.
General Procedure for the Synthesis of 17a-17c Intermediates 17a-17c were prepared from 16a-16c using the same reaction conditions described above for making intermediate 11.
vorinostat for 10 min. Fluorescence intensity was measured using a Varioskan Flash Station (Thermo Scientific, MA, USA) at excitation and emission wavelengths of 355 (or 360 for the HeLa NuEx) and 460 nm, respectively. The IC 50 values were extracted by curve fitting the dose/response slopes.

Docking Study
Discovery Studio 3.0 (BIOVIA, Waltham, MA, USA) was used to perform in silico docking. The Xray crystal structure of Mnk and HDAC1 in complex with the ligand was retrieved from Protein Data Bank (PDB code 2HW7 and 1C3S). All calculations and manipulations were performed with CDOCKER modules in the Discovery Studio software package. All water molecules were removed and hydrogen was added. By applying the default parameters, the best docking result was selected to analysis based on the favorable binding affinity rank in kcal/mol.

Conclusions
We report herein a series of 4,6-disubstituted pyrido [3,2-d]pyrimidine derivatives as Mnk/HDAC dual inhibitors. The structure-activity relationships of these compounds was summarised. The vitro enzymatic assay result showed that most of the synthesized compounds showed good potency against Mnks and HDACs. Among the compounds, compound A12 showed the best inhibitory against Mnks and also good potency on HDAC, whose Mnk1 and Mnk2 inhibitory activity with the inhibition rate of 54% and 70% at the concentration of 1 µM and the HDAC inhibitory activity with the IC 50 value of 0.78 µM. Most of the compounds showed preferable anti-proliferation activity against human prostate cancer PC-3 cell lines with the GI 50 value of 2.91 µM-18.26 µM. The most potent compound A12 was selected to conduct molecular docking studies. Docking results showed that the hydroxamic acid motif could form hydrogen bond interactions with both Mnk and HDAC. The pyrido [3,2-d]pyrimidine core and the 4-substituted aniline group were also essential for forming key interactions between the compounds and targets. A series of novel Mnk/HDAC inhibitors bearing pyrido[3,2-d]pyrimidine scaffold were provided and the further research was still needed.