Synthesis and Biological Evaluation of NH2-Sulfonyl Oseltamivir Analogues as Influenza Neuraminidase Inhibitors

A series of NH2-sulfonyl oseltamivir analogues were designed, synthesized, and their inhibitory activities against neuraminidase from H5N1 subtype evaluated. The results indicated that the IC50 value of compound 4a, an oseltamivir analogue via methyl sulfonylation of C5-NH2, was 3.50 μM. Molecular docking simulations suggested that 4a retained most of the interactions formed by oseltamivir carboxylate moieties and formed an additional hydrogen bond with the methylsulfonyl group. Meanwhile, 4a showed high stability towards human liver microsomes. More importantly, 4a without basic moieties is not a zwitterion as reported on the general structure of neuraminidase inhibitors. This research will provide valuable reference for the research of new types of neuraminidase inhibitors.


Introduction
Influenza, an infectious disease caused by influenza viruses, has a high morbidity and mortality. In the 20th century, there were three pandemic outbreaks of influenza, for example, the "Spanish" influenza resulting in approximately 50 million deaths in 1918 and 1919 [1]. The pandemic nature of influenza presents a serious public health concern [2,3]. At present, there are two types of agents for influenza, adamantane-based M2 ion channel protein inhibitors and neuraminidase inhibitors (NAIs). The application of the former is limited due to adverse side effects and rapidly developed drug resistance. Thus the latter is widely used [4][5][6][7]. So far, four neuraminidase inhibitors, oseltamivir phosphate [6] and zanamivir hydrate [8] approved around the world, peramivir hydrate [9,10], and laninamivir octanoate [11] marketed only in a few countries, are available. Because of its convenient oral administration, oseltamivir has the largest market share [12,13].
Neuraminidase (NA) existing as a tetramer is responsible for removing sialic acid residues from receptors and facilitates virus release and spread to infect host cells [3,8,14]. The active site of NA is divided into five subsites (S1-S5) ( Figure 1). S1, composed of three basic amino acids (Arg118, Arg292, and Arg371), has affinity for the acidic moieties of the inhibitors to form a salt bridge beneficial for potent inhibitory activity. S2, formed by three acidic amino acids (Glu119, Asp151, and Glu227), has affinity for basic moieties to form a salt bridge [7,15]. Therefore, powerful inhibitors are generally zwitterions with poor lipophilicity and oseltamivir carboxylate (OC, the active ingredient of oseltamivir discovered compound D (Figure 2) which showed powerful inhibitory activities against wild-type NAs with IC50 values similar to those of OC and enhanced inhibitory activities against mutant NAs in the process of searching for inhibitors targeting the 150-cavity. What is more, compound D without basic moieties was not consistent with the traditional structure-activity relationships in that a basic moiety is essential for powerful activity [18]. In consideration that sulfonamido (R 1 SO2NHR 2 ) can act both as hydrogen bond donor and hydrogen bond acceptor, which may be beneficial for biological activity [19,20], we continue to search for NAIs without basic moieties. A series of oseltamivir analogues bearing sulfonamido were designed with the aim of the sulfonamido generating hydrogen bonds with the S2 and R (Figure 3) group interacting with the 150-cavity.  The 150-cavity close to S2, formed by residues of 147-152, offered new opportunities for the research of a new type of NAIs [7,14]. Guided by the 150-cavity, several oseltamivir derivatives substituted on the C5-amine have been reported. Xie and his colleagues revealed compounds A and B ( Figure 2) with IC 50 values of 1.9 nM and 2.1 nM against NAs from H5N1 (A/Chicken/China/1220/2012) respectively, which preliminarily showed that the 150-cavity was an advantage for enhancing inhibitory activity [17]. In 2010, Mohan et al. discovered compound C containing triazole with a K i value of 72 nM. Saturation transfer difference (STD) NMR spectroscopic studies suggested that compound C interacted with the 150-cavity successfully [12]. Our laboratory discovered compound D ( Figure 2) which showed powerful inhibitory activities against wild-type NAs with IC 50 values similar to those of OC and enhanced inhibitory activities against mutant NAs in the process of searching for inhibitors targeting the 150-cavity. What is more, compound D without basic moieties was not consistent with the traditional structure-activity relationships in that a basic moiety is essential for powerful activity [18]. In consideration that sulfonamido (R 1 SO 2 NHR 2 ) can act both as hydrogen bond donor and hydrogen bond acceptor, which may be beneficial for biological activity [19,20], we continue to search for NAIs without basic moieties. A series of oseltamivir analogues bearing sulfonamido were designed with the aim of the sulfonamido generating hydrogen bonds with the S2 and R ( Figure 3) group interacting with the 150-cavity. Glu227), has affinity for basic moieties to form a salt bridge [7,15]. Therefore, powerful inhibitors are generally zwitterions with poor lipophilicity and oseltamivir carboxylate (OC, the active ingredient of oseltamivir as a prodrug) is no exception [16]. The non-zwitterionic NAIs have potential advantages for several aspects such as lipophilicity.
The 150-cavity close to S2, formed by residues of 147-152, offered new opportunities for the research of a new type of NAIs [7,14]. Guided by the 150-cavity, several oseltamivir derivatives substituted on the C5-amine have been reported. Xie and his colleagues revealed compounds A and B (Figure 2) with IC50 values of 1.9 nM and 2.1 nM against NAs from H5N1 (A/Chicken/China/1220/2012) respectively, which preliminarily showed that the 150-cavity was an advantage for enhancing inhibitory activity [17]. In 2010, Mohan et al discovered compound C containing triazole with a Ki value of 72 nM. Saturation transfer difference (STD) NMR spectroscopic studies suggested that compound C interacted with the 150-cavity successfully [12]. Our laboratory discovered compound D ( Figure 2) which showed powerful inhibitory activities against wild-type NAs with IC50 values similar to those of OC and enhanced inhibitory activities against mutant NAs in the process of searching for inhibitors targeting the 150-cavity. What is more, compound D without basic moieties was not consistent with the traditional structure-activity relationships in that a basic moiety is essential for powerful activity [18]. In consideration that sulfonamido (R 1 SO2NHR 2 ) can act both as hydrogen bond donor and hydrogen bond acceptor, which may be beneficial for biological activity [19,20], we continue to search for NAIs without basic moieties. A series of oseltamivir analogues bearing sulfonamido were designed with the aim of the sulfonamido generating hydrogen bonds with the S2 and R ( Figure 3) group interacting with the 150-cavity.

Synthesis
The synthetic approaches to oseltamivir were investigated, and several approaches were found to be innovative or interesting [21][22][23][24], however, we followed our previous semi-synthetic procedure for oseltamivir (2) [18]. The synthetic route of the target compounds is depicted in Scheme 1. Oseltamivir was reacted with the corresponding sulfonyl chlorides to obtain 3a-3k [18,25,26]. The nitro fragment of 3i-3k was reduced to amino by iron powder to afford intermediates 5i-5k. Finally, 3a-3k and 5i-5k were treated with NaOH in aq. methanol to yield target compounds 4a-4k and 6i-6k via saponification. All compounds were confirmed by 1 H-NMR, 13 C-NMR, and HRMS (ESI).

Synthesis
The synthetic approaches to oseltamivir were investigated, and several approaches were found to be innovative or interesting [21][22][23][24], however, we followed our previous semi-synthetic procedure for oseltamivir (2) [18]. The synthetic route of the target compounds is depicted in Scheme 1. Oseltamivir was reacted with the corresponding sulfonyl chlorides to obtain 3a-3k [18,25,26]. The nitro fragment of 3i-3k was reduced to amino by iron powder to afford intermediates 5i-5k. Finally, 3a-3k and 5i-5k were treated with NaOH in aq. methanol to yield target compounds 4a-4k and 6i-6k via saponification. All compounds were confirmed by 1 H-NMR, 13 C-NMR, and HRMS (ESI).

Synthesis
The synthetic approaches to oseltamivir were investigated, and several approaches were found to be innovative or interesting [21][22][23][24], however, we followed our previous semi-synthetic procedure for oseltamivir (2) [18]. The synthetic route of the target compounds is depicted in Scheme 1. Oseltamivir was reacted with the corresponding sulfonyl chlorides to obtain 3a-3k [18,25,26]. The nitro fragment of 3i-3k was reduced to amino by iron powder to afford intermediates 5i-5k. Finally, 3a-3k and 5i-5k were treated with NaOH in aq. methanol to yield target compounds 4a-4k and 6i-6k via saponification. All compounds were confirmed by 1 H-NMR, 13 C-NMR, and HRMS (ESI).

Neuraminidase Enzyme Inhibitory Assay
Compounds 4a-4k and 6i-6k were evaluated as inhibitors of the NA from the H5N1 subtype of influenza A. The inhibition rates at 10 µM and 100 µM are shown in Table 1 [18]. The relatively good compounds such as compound 4a, 4h, 4i, and 6i were further screened to give IC 50 values. The inhibitory activities varied from the sulfonyl fragments substituted on the amine of OC. Compounds 4a-4e showed high to weak inhibitory activities. The length of the sulfonyl moieties played a crucial role in the inhibitory activities. The increased length of the substituents led to decreased inhibitions against NA as suggested by 4a-4e. Compound 4a, possessing the shortest substituent, exhibited the most powerful neuraminidase inhibitory activity with an IC 50 value of 3.50 µM (Table 2). Meanwhile, 4f as a fluoro-substituted congener of 4a did not exhibit good inhibitory activity, indicating fluorine substitution was harmful for interaction with amino acids near or belonging to S2. Compounds 4g-4k and 6i-6k bearing aromatic rings exhibited moderate to weak inhibitory activities. Compound 4g containing 4-acetylamido phenyl exerted little inhibition even at 100 µM. Compared with 4f, the inhibitory activity of compound 4h bearing trifluoromethyl was enhanced, and the IC 50 value was 12.00 µM. Among the compounds 4i-4k and 6i-6k, ortho-substituted 4i and 6i showed more potent inhibitory activities than the meta-substituted and para-substituted counterparts. Inhibitory activities decreased in the order of ortho-, meta-, and para-position. Compared to the nitro, the amino group was more beneficial for inhibitory activity.  Compound 4a exerted the most powerful inhibitory activity. The inhibition of 4a was weaker than that of OC, but 4a without basic moieties is not a zwitterion, which is not consistent with traditional structure-activity relationships of NAIs.

Molecular Docking Model Analysis
As shown in Figure 4B, due to the added methylsulfonyl group the OC fragment of compound 4a did not overlap well with OC, so the interactions formed by the OC moiety with NA were discounted to a certain extent. It is a pity that on account of the short length of the methylsulfonyl compound 4a could not access the 150 cavity as expected. The oxygen of the sulfonyl of compound 4a generated a new hydrogen bond with Arg152 ( Figure 4A). The new interactions generated by methylsulfonyl were not enough to make up for the lost or decreased ones formed by the C5 amine of the OC fragment with S2, which explained the weaker inhibitory activity of 4a compared to that of OC. discounted to a certain extent. It is a pity that on account of the short length of the methylsulfonyl compound 4a could not access the 150 cavity as expected. The oxygen of the sulfonyl of compound 4a generated a new hydrogen bond with Arg152 ( Figure 4A). The new interactions generated by methylsulfonyl were not enough to make up for the lost or decreased ones formed by the C5 amine of the OC fragment with S2, which explained the weaker inhibitory activity of 4a compared to that of OC.

Metabolic Stability in Human Liver Microsomes In Vitro
Compound 4a with the most powerful inhibition was selected to test the metabolic stability in the presence of human liver microsomes in vitro and the control compounds of testosterone, diclofenac, and propafenone validated the assay [27][28][29][30]. The results are shown in Table 3. After incubating with human liver microsomes for 1 h, almost of all of 4a (101.6%) was detected and the T1/2 (half time) was more than 145 min, which indicated that 4a showed high stability towards human liver microsomes in vitro.

Metabolic Stability in Human Liver Microsomes In Vitro
Compound 4a with the most powerful inhibition was selected to test the metabolic stability in the presence of human liver microsomes in vitro and the control compounds of testosterone, diclofenac, and propafenone validated the assay [27][28][29][30]. The results are shown in Table 3. After incubating with human liver microsomes for 1 h, almost of all of 4a (101.6%) was detected and the T 1/2 (half time) was more than 145 min, which indicated that 4a showed high stability towards human liver microsomes in vitro.

Chemistry
All of the raw materials and solvents were purchased from commercial suppliers. Melting points were determined in slides on a WRX-4 Micro melting point apparatus (Yice). High resolution mass spectra (HRMS) were recorded on an Agilent 6530 ultrahigh definition (UHD) accurate mass Q-TOF MS by ESI mode. The 1H-nuclear magnetic resonance (NMR) and 13 C-NMR spectra were recorded on a Bruker ARX 600 MHz using tetramethylsilane as the internal standard. The reaction process was monitored by thin-layer chromatography (TLC) on silica gel GF254. Plates were visualized using UV light (254 nm). The purity of target compound was determined by HPLC. The detailed method is as follows: a Shimadzu (Kyoto, Japan) HPLC; Column: DIAMONSIL ® C18, 250 mm × 4.6mm, 5 µm; mobile phase (isocratic elution): 50% acetonitrile (0.1% trifluoroacetic acid) for method A; 35% acetonitrile (0.1% trifluoroacetic acid) for method B; 25% acetonitrile (0.1% trifluoroacetic acid) for method C. Flow rate: 1 mL/min; Detector: UV (254 nm).
General Procedure for the Preparation of Compounds 4a-4k and 6i-6k Oseltamivir (2, 312.4 mg, 1.0 mmol), TEA (208 µL, 1.5 mmol) and 10 mL CH 2 Cl 2 were charged in a 50 mL round bottom flask, then the corresponding sulfonyl chloride (1.2 mmol) was added dropwise. The mixture was stirred at 0 • C until oseltamivir was completely consumed, as indicated by TLC analysis. The organic layer was washed successively with 1N HCl aqueous solution, saturated aq.Na 2 CO 3, and brine. The organic layer was concentrated in vacuo. The crude product was purified by column chromatography to obtain one of the intermediates (3a-3k).
One of intermediates 3a-3k (1.0 mmol), 1N NaOH aqueous solution (2.5 mmol), methanol (15 mL) and deionized water (V (methanol): V (water) = 5:1) were added to a round bottom flask. The mixture was stirred at room temperature. Then the methanol was evaporated in vacuo and the residual solution was acidified with 1N HCl aqueous solution to pH 1 to 2. The precipitate was separated and filtered. Finally, one of the title compounds (4a-4k) was obtained.

Metabolic Stability Assay
The human liver microsomes were purchased from BD (San Jose, USA). NADPH was purchased from Sigma-Aldrich (Munich, Germany). Two parallel assays with and without NADPH regenerating system were determined.
The compound 4a and control compounds (testosterone, diclofenac, or propafenone) were prepared by dilution of reaction buffer, and the final concentration of human liver microsomes was 0.5 mg/mL. Incubation was carried out in a thermostat at 37 • C and started by the addition of the appropriate compound. The samples were taken at 0, 5, 10, 20, 30, and 60 min. The reaction was terminated by the addition of stop solution. After collection, samples were centrifuged (20 min, 4000 rpm), then the centrifuged supernatant was directly analyzed by using LC-MS/MS analysis. Metabolic half-time (T 1/2 ) was calculated by using the equation of first order kinetics.

Conclusions
In summary, a series of oseltamivir analogues bearing the sulfonamido group were designed based on our previous work on potent neuraminidase inhibitors without basic moieties and with the opportunities offered by the 150-cavity. Among these analogues, compound 4a showed the most potent inhibition against NA from H5N1 subtype with an IC 50 value of 3500 nM. Molecular docking simulations revealed that 4a retained most of the interactions formed by the OC fragment and generated a new hydrogen bond. Due to the short length of the methylsulfonyl group, 4a did not get access to the 150-cavity as expected. Meanwhile, 4a exhibited high metabolic stability against human liver microsomes in vitro. What is more, compound 4a without basic moieties is not consistent with traditional inhibitors as zwitterions. Thus this study has enriched the structure types of NAIs and may provide valuable reference for the discovery of new types of NAIs. Funding: This work was financially supported by the Foundation of Shenyang Science and Technology Bureau (NO. F13-196-9-00).

Acknowledgments:
We gratefully acknowledge the financial support from the Foundation of Shenyang Science and Technology Bureau (NO. F13-196-9-00).

Conflicts of Interest:
The authors declare no conflicts of interest.