The general synthetic route of 1–32 is outlined in Scheme 1. We chose 3,4-dihydroxybenzoic acid methyl ester, 3,4-dihydroxybenzaldehyde, 3,4-dihydroxybenzoic acid, isovanillic acid, and caffeic acid as the starting materials. A series of aromatic ethers were synthesized through the Williamson ether synthesis, followed by selective mono-deallylation using 10% Pd/C or aminationa of the intermediate to provide part of target compounds [10]. Most of the compounds were easily obtained with excellent yields (60–95%). Compound 2, 3, 5, 8, 14, 19, 24, 25, 27–30 and 32 were new derivatives. All new compounds were fully characterized by HRMS and NMR. The purities (>95%) of all target compounds were checked by HPLC using a LC-2010c equipped with UV detector.

All compounds were evaluated for inhibitory activities on the pandemic influenza H1N1 virus [(A/Guangdong/03/2009 (H1N1)] neuraminidase in vitro, the NA inhibitors oseltamivir (Tamiflu) was used as positive control. The results (Figure 1) showed that most of the derivatives inhibited H1N1 virus neuraminidase in the range of 15–59% inhibition at the 50 μM concentration. Among them, compound 15 and 26 exhibited 50% or higher activities. In this screen, the model compound (Chart 1C) had no activity. The SAR analysis revealed that: (1) Compared with compounds 16–18, 20 and 25, compounds 9, 11–13, and 24 exhibited higher activities, it meant that substitution of 3- and 4-OH would reduce the activities; (2) compound 13 showed weaker activity than compound 15, and it indicated that OH group at C-4 position could improve the activity; (3) for compounds 9–13, it seems that allyl ether is better than isopentyl ether, propyl ether, propargyl ether and isopentene ether, which indicated that maybe the smaller group substituted in this series of compounds is helpful for the activities; (4) previous studies showed that carboxylic acid is an important group in maintaining activities. However, acid changed to aldehyde in 1-position of aromatic compounds was more favorable for the bioactivities in compounds 17, 21, 25 and 30 in our present study.

The effect of compound 15 was further investigated on the neuraminidase activity of three pandemic influenza H1N1 2009 virus strains [A/GuangdongSB/01/2009 (H1N1), A/Guangdong/03/2009 (H1N1) and A/Guangdong/05/2009 (H1N1)]. The results (Figure 2) revealed significant inhibitory effects of compound 15 on all three tested viral NA, with IC₅₀ values of 26.96 μM for A/GuangdongSB/01/2009 (H1N1), 27.73 μM for A/Guangdong/03/2009 (H1N1), and 25.13 μM for A/Guangdong/05/2009 (H1N1), respectively. At the same time, compound 15 showed low cytotoxity (>600 μM) to virus host cell—Madin-Darby canine kidney (MDCK) cells—and showed good selectivity index.

Molecular docking was performed to further assess the SAR of different substituents of the inhibitors by their interactions with the NA residues. The studies focused on the ether chains, the position of OH and the groups at C-1 position.

Compounds 9–13 with different ether chains in C-4 position produced little difference in activity. Introducing the hydrophobic groups in the small hydrophobic pocket formed by Trp178 and Ile222 as well as the side chain of Arg224 could enhance binding action. In Figure 3(a), all of the ether chains were near the pocket, but the allyl ether chain terminal displayed the best activity compared with other substituents. The triarginyl cluster of Arg118, Arg292, and Arg371 at the pocket is the predominant factor for orienting and stabilizing inhibitors. Compounds 17, 21, 25, 30 had different groups in C-4 position. Relative to 21, 25 and 30, compound 17 with an aldehyde group which made strong charge–charge interactions with the triarginyl pocket was more favorable for binding action (Figure 3(b)). From Figure 3(c), we found that the OH group of compound 15 was much closer to the nucleophilic oxygen of Asp151 compared to compound 13. Asp151 also showed significant favorable electrostatic interaction. The flexibility of the molecule during the docking calculation permitted turning the tail of the ether chains, thus the steric hindrance of the structure did not have a strong effect in binding with Arg156. Molecular docking studies further supported the aforementioned SAR analysis.

The most potent compound 15 was docked into the active sites of NA. Oseltamivir is an efficacious and commonly used neuraminidase inhibitor. The binding of compound 15 with NA was shown in Figure 3(d) and Figure 3(e). The activity of compound 15 could be explained by its interactions with the active residues of NA. The aldehyde group made strong charge–charge interactions with Arg118, Arg292, and Arg371. Asp151 exhibited favorable electrostatic interaction with the OH group at C-4 position. At the same time, the ether chains were suitable for the hydrophobic region formed by Trp178 and Ile222, as well as the side chain of Arg224. Compound 15 provided a binding mode with NA for this series of compounds from its geometric and electronic features.

All reagents and solvents were of commercial quality. Melting points were determined on an X-4 micromelting point apparatus and were uncorrected. ¹H and ¹³CNMR data were recorded on a Varian Inova 400 MB NMR spectrometer operating at 400 and 125 MHz for ¹H and ¹³C, respectively. All chemical shifts are in ppm (δ) with respect to tetramethylsilane (TMS) as internal standard, and coupling constants (J) are in Hz. Mass spectra were obtained on DSQ (low resolution mass spectrometer) and MAT95XP (high resolution mass spectrometer) instruments.

The pandemic influenza H1N1 2009 virus Influenza A strains A/GuangdongSB/01/2009 (H1N1), A/Guangdong/03/2009 (H1N1) and A/Guangdong/05/2009 (H1N1) were kindly provided by Dr. Li at the Sun Yat-sen University Zhongshan School of Medicine (Guangzhou, Guangdong, China.) who isolated it during the human swine flu outbreak of 2009. Madin-Darby canine kidney (MDCK) cells were cultured in DMEM (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS) (Hyclone, Logan, UT, USA), 2 mM l-glutamine, 100 μg/mL streptomycin and 100 units/mL penicillin (Invitrogen, Carlsbad, CA, USA). The cultures were maintained at 37 °C in a humidified atmosphere of 5% CO₂. All viruses were amplified and titrated in MDCK cells and stored at −80 °C until use. Confluent MDCK monolayers in 6-well plates were washed three times with PBS and incubated with viruses at a multiplicity of infection (MOI) of 0.001 in DMEM for 2 h at 37 °C. After viral adsorption, media was removed, cells were washed three times with PBS and incubated with post-adsorption medium [DMEM, supplemented with 1% fetal bovine serum (FBS) (Hyclone, Logan, UT, USA), 2 mM l-glutamine, 100 μg/mL streptomycin and 100 units/mL penicillin (Invitrogen, Carlsbad, CA, USA), 25 mM HEPES buffer (pH 7.2), 0.1 mM non-essential amino acids, 1.0 mM sodium pyruvate, 2 mg/mL l-1-(tosylamido-2-phenyl) ethyl chloromethyl ketone (TPCK)-treated trypsin (Invitrogen, Carlsbad, CA, USA)].

Influenza virus neuraminidase (NA) sensitivity to the test compounds was evaluated by using an NA activity inhibition assay, based on the method of Maki Kiso et al. [11]. Briefly, 4-Methylumbelliferyl-N-acetylneuraminic acid (4-MUNANA, Sigma-Aldrich, St. Louis, MO, USA), at a final concentration of 0.15 mM, was used as the fluorescent substrate. The virus solution [(A/Guangdong/03/2009 (H1N1), ranging from 800 to 1200 fluorescence units of MU-NANA] was prepared in MES assay buffer [32.5 mM MES (pH 6.5), 4 mM CaCl₂, 0.1% NP-40, 0.3 mg/mL BSA], the test compounds were dissolved in MES assay buffer at various concentrations (1, 10, 20, 40, 50, 80 μM), and Tamiflu^® (Roche Laboratories Inc., Basel, Switzerland) was used as a positive control. Diluted virus was mixed with each test compound in 50 μL MES assay buffer, and the mixtures were incubated at 37 °C for 1 h with gentle agitation, then mixed with 50 μL 4-MUNANA buffer and incubated at 37 °C for another 1 h with gentle agitation. The reaction was terminated by adding 100 μL of 0.1 M NaOH in 80% ethanol (pH 10.0). Then the fluorescence was measured at an excitation wavelength of 360 nm and an emission wavelength of 465 nm, using a SpectraMax M5 multi-detection reader (Molecular Devices Corporation, Sunnyvale, CA, USA). Under the experimental conditions, a linear relationship between the enzyme amount and the fluorescence was confirmed. Each NA inhibition assay was done in triplicate. NA activity inhibition was determined using the following formula as described previously [11]: NA activity inhibition (%) = (1 − OD₄₆₅ of treated group/OD₄₆₅ of control group) × 100%. The half maximal inhibitory concentration (IC₅₀) was calculated by Bliss software [12] and the data, expressed as means ± SD, were analyzed by SPSS.

Surflex-Dock was applied to study molecular docking. Crystal structure of NA was retrieved from RCSB Protein Data Bank (PDB entry code: 2ht7) [13]. This is a particular structure with oseltamivir. The protein structure was utilized in subsequent docking experiments without energy minimization. All ligands and water molecules have been removed at first and the polar hydrogen atoms were added. Automatic docking was employed. Other parameters were established by default in software. All molecules were minimized using default parameter.