Functional and Structural Analysis of Influenza Virus Neuraminidase N3 Offers Further Insight into the Mechanisms of Oseltamivir Resistance

Recombinant NA production.

N3, 09N1, and 57N2 were prepared using the protocol from previous reports (11, 16, 25, 26). N3-H274Y, 09N1-H274Y, and 57N2-H274Y mutants were constructed by overlap PCR (Generay, China). The cDNA encoding amino acid residues 83 to 469 were inserted into the baculovirus transfer vector pFastBac1 (Invitrogen), with a GP67 signal peptide, a 6×His tag, a tetramerizing sequence, and a thrombin cleavage site at the N terminus. Recombinant baculovirus was prepared based on the manufacturer's protocol (Invitrogen). Hi5 suspension cultures were grown in Sf-900 II SFM serum-free medium (Gibco) (28°C and 120 rpm) and transfected with high-titer recombinant baculovirus. After the growth of the transfected Hi5 suspension cultures for 2 days, centrifuged medium were applied to a 5-ml HisTrap FF column (GE Health), which was washed with 20 mM imidazole, and then NA was eluted using 300 mM imidazole. After dialysis, thrombin digestion (Sigma, 3 U of NA/mg; overnight at 4°C), and gel filtration chromatography using a Superdex-200 10/300 GL column (GE Healthcare), NA fractions were analyzed by SDS-PAGE. High-purity NA fractions were pooled and concentrated using a membrane concentrator with a molecular mass cutoff of 10 kDa (Millipore). A buffer of 20 mM Tris-HCl–50 mM NaCl (pH 8.0) was used for gel filtration and protein concentration.

NA enzymatic inhibition assay.

A neuraminidase inhibition assay using MUNANA was performed as described by Potier et al. with modifications (25). Briefly, 10 μl of purified, recombinant N3, 09N1, and 57N2 and their H274Y mutant proteins (10 nM) were mixed with 10 μl of inhibitor separately and incubated for 30 min at room temperature. NA and inhibitors were carefully diluted in fresh phosphate-buffered saline. At least five concentrations of each inhibitor at an appropriate range were used for each repeat. After incubation, 30 μl of 166 μM MUNANA in 33 mM MES (morpholineethanesulfonic acid) and 4 mM CaCl₂ (pH 6.0) was added to the solution to start the reaction using a 12-tip pipette (Eppendorf). A positive and a negative control were included in each 12-well lane. After starting the reaction for each lane on the plate, the reaction mixture was immediately loaded on a SpectraMax M5 (Molecular Devices) where fluorescence was quantified over the course of 30 min at an excitation wavelength of 355 nm and an emission wavelength of 460 nm. Single time points were chosen where the positive control produced a relative fluorescence signal of approximately 1,000. At least two replicates were used for each NA-inhibitor pair, and the 50% inhibitory concentrations (IC₅₀s) for each inhibitor were calculated with sigmoidal fitting of the log inhibitor versus inhibition percentage using GraphPad Prism.

Crystallization and inhibitor soaking.

NA crystals were grown using the hanging-drop and sitting-drop vapor diffusion methods. Initial screening was performed using a commercial kit (Hampton Research). N3 crystals were obtained by mixing 1 μl of the concentrated protein at 10 mg/ml in 20 mM Tris (pH 8.0) and 50 mM NaCl with 0.1 M Bis-Tris propane (pH 9.0)–10% (vol/vol) Jeffamine ED-2001 (pH 7.0), the same concentrations used for 57N2 crystallization. N3-H274Y crystals were obtained in the reservoir solution of 1 M bicine (pH 8.5)–15% PEG1500. All crystals were first incubated in mother liquor containing 20 mM inhibitor and then flash-cooled at 100 K. Diffraction data were collected at KEK beamline Ne3A and SSRF beamline BL17U.

Data collection, processing, and structure solution.

Diffraction data were processed and scaled using HKL2000 (26). Data collection and processing statistics are summarized in Table 2. The structures of N3 and N3-H274Y were solved by molecular replacement using Phaser from the CCP4 program suite with the structure of N2 (Protein Data Bank [PDB] identification code [ID] 2BAT) as the search model (27, 28). Initial restrained rigid-body refinement and manual model building were performed using REFMAC5 (29) and COOT (30), respectively. Further rounds of refinement were performed using the phenix.refine program implemented in the PHENIX package with coordinate refinement, isotropic ADP refinement, and bulk solvent modeling (31). The stereochemical quality of the final model was assessed with the program PROCHECK (32).

N3 is a typical group 2 NA.

To investigate the structure of the second to last structure-unknown influenza A virus NA, N3 was crystallized, and its structure was solved at a resolution of 1.8 Å (Table 2). The overall structure of N3 is highly conserved relative with the other structure-solved influenza virus NAs with a typical “box-shaped” tetramer of four identical monomers, each containing six four-stranded antiparallel β-sheets in a propeller-like arrangement (Fig. 1A). Detailed analysis shows that N3 has no 150-cavity in its active site (Fig. 1B) like all other group 2 NA structures solved to date. The 150-loop (residues 147 to 152, based on N2 numbering) sequence of N3 is GTIKDR, which happens to be the same as 09N1 (PDB ID 3NSS). The 150-loop sequence of VN04N1 (PDB ID 2HTY), a typical group 1 NA, and that of N9 (PDB ID 7NN9), a typical group 2 NA, are GTVKDR and GTIHDR, respectively. Superposition of Cα atoms of the N3 150-loop with those of 09N1, VN04N1, and N9 gives average root mean square deviation (RMSD) values of 0.200, 2.259, and 1.404 Å, respectively.

Open in a separate window

Fig 1

Crystal structure of the N3 tetramer and analysis of its active site. (A) N3 tetramer in schematic representation, viewed from above the viral surface. (B) Molecular surface of the active site of N3 (yellow); N3 has no 150-cavity. (C) Molecular surface of the active site of H5N1 N1 (red) (PDB ID 2HTY); typical group 1 N1 has a 150-cavity. (D) The covalent complex formed between 3-fluoro-Neu5Ac and N3 Tyr406. The lack of density for the C-2 fluoride leaving group indicates that it has departed, allowing for the formation of the Tyr406-inhibitor covalent bond. (E) Laninamivir bound to the N3 active site with a typical binding mode. (F) Molecular surface of the active site of N2 (blue) (PDB ID 2BAT); group 2 N2 also has no 150-cavity.

We were also able to obtain a covalent complex between N3-Tyr406 and 3-fluoro-Neu5Ac by soaking N3 crystals with 2,3-difluoro-Neu5Ac. This demonstrates that N3 also operates via a nucleophilic mechanism, just like influenza virus N2, and NAs from other microorganisms (35–37). We also soaked N3 crystals with laninamivir (4-guanidino-7-methoxy-Neu5Ac2en) and solved the complex structure at a resolution of 1.7 Å. Like all other zanamivir and laninamivir complex structures, the laninamivir C4-guanidino is buried beneath the N3 150-loop, where it hydrogen bonds with E119, D151, W178, and E227 (Fig. 1C).

N3 contains three loops with unique conformations.

Upon close comparison of the N3 structure to the other structure-solved NAs, three special loops in N3 were discovered: the 270-loop (residues 267 to 276), the 380-loop (residues 380 to 390), and the 430-loop (residues 429 to 433) (Fig. 2).

Open in a separate window

Fig 2

Comparison of loops in NA from different influenza viruses. (A) Superposition of NA monomers with an emphasis on the 270-loop, 380-loop, and 430-loop. NAs are differentiated by color: 18N1, green; VN04N1, cyan; 09N1, red, N4, light magenta; N5, light pink; N8, salmon; Tokyo N2, white; N6, orange; N9, light blue; and N3, yellow. (B) Stereo view of the 270-loop. All group 1 members are on the left, while group 2 members are on the right. N3 is in the middle with a longer 270-loop. (C) Ribbon representation of the 380-loop. The 380-loop of N3 is also unique compared to all other NAs. (D) Stereo image of the 430-loop. N3 also has a relative flexible 430-loop.

N3 has a longer, unique 270-loop than other structure-solved NAs due to an additional lysine residue, K271′ (Fig. 2B and Fig. 3A). The main-chain carbonyl at residue 273 of all structure known group 1 members hydrogen bonds to Gln249 and Tyr252, and the 273 main-chain nitrogen hydrogen bonds with residue 270 (Fig. 3B). In contrast, residues 272 and 273 of typical group 2 NA form hydrogen bonds with residues 314 and 316; thus, the 270-loop of group 2 NA is pulled toward the opposite direction relative to group 1 members (Fig. 3C). Unexpectedly, Gln273 in N3 forms hydrogen bonds with Tyr252 and K270, which is only observed in group 1 NA and yet also has group 2-specific interaction with Y316 (Fig. 3D). K271′ of N3 forms hydrogen bonds with the carboxyl oxygen of T314 (Fig. 3D and Table 3). Therefore, N3 possesses characteristics of both group 1 and group 2 NA, as well as unique characteristics, regarding the 270-loop.

Open in a separate window

Fig 3

Comparison of the 270-loop in different NAs. (A) Superimposed 270-loops of VN04N1 (orange), N2 (blue), and N3 (yellow). The 270-loop of N1 is much closer to the 240-loop. (B) Y273 in VN04N1 forms hydrogen bonds with Y252, Q249, and A270, which produces a tight turn in the 270-loop. (C) Q273 and A272 in N2 hydrogen bond to Y316 and S314, respectively, thus pulling the 270-loop in N2 to turn in the opposite direction of the N1 loop. (D) Q273 in N3 hydrogen bonds with Y252, K270, Y316, and I272 and K271′ hydrogen bonds with T314.

The 380-loop of N3 is also unique relative to other structure known NAs (Fig. 2C and Fig. 4). Detailed analysis indicates that all of the known influenza virus A NA structures, including N3, have hydrogen bonds between residues in the 380-loop and residues in the 320-loop (residues 315 to 320) and 330-loop (Fig. 4 and Table 4). Relative to N3, all other structure known NAs have additional hydrogen bonds with the 320- and 330-loops (Fig. 4, Table 4, underlined text). This lack of hydrogen bonds with the N3 380-loop is likely a major factor for the unique 380-loop conformation (Fig. 4).

Open in a separate window

Fig 4

Comparison of the N3 380-loop in with distinct NA structures. (A) Location of the 380-loop, 320-loop, and 330-loop in VN04N1 (orange), N2 (blue), and N3 (yellow). The distance between the 380-loop and 320-loop in N3 is the greatest. (B to D) Residues 319 and 320 form hydrogen bonds with the 380-loop in N1, N2, and N3. The 380-loop in N1 and N2 forms more hydrogen bonds with the 320-loop and adjacent residues relative to N3. Residues that take part in hydrogen bonds between and among the 380-loop, 320-loop, and 330-loop are labeled. Residues that are marked with an asterisk (*) indicate that they only take part in hydrogen bonds via their main chain oxygen or nitrogen.

In addition, N3 also possesses a unique 430-loop where the N3 residue K433 interacts with two water molecules, whereas all other NA form hydrogen bonds between E433 and G429 (Fig. 2D and Fig. 5). This results in a 430-loop in N3 that is farther away from the 150-loop; however, P431 remains in the same relative position.

Open in a separate window

Fig 5

N3 contains a special 430-loop. K433 in N3 (yellow) coordinates with two water molecules, while E433 in all other structure-known NAs form a hydrogen bond with G429.

This study was supported by the National 973 Project (grant 2011CB504703) and the National Natural Science Foundation of China (NSFC; grant 81021003), and G.F.G. is a leading principal investigator of the NSFC Innovative Research Group. C.J.V. is a recipient of the Chinese Academy of Sciences Young International Scientist Research Fellowship (2011Y2SA01) and the NSFC Research Fund for Young International Scientists (31150110147).

The assistance of the staff at Shanghai Synchrotron Radiation Facility (SSRF-Beamline 17U) is acknowledged. We thank Mingyang Wang, Yanfang Zhang, and Yue Liu for help with protein preparation and discussion. We also thank Liu Di and Li Wei at the Network Information Center, Institute of Microbiology, Chinese Academy of Sciences, for help with sequence analysis.

G.F.G. conceived and supervised the project. G.F.G. and Q.L. designed the project. Q.L. carried out the protein expression, purification, and crystallization. J.Q. solved the structures. Q.L. and C.J.V. carried out the functional analysis. H.K., K.T., H.O., Y.S., and Y.S. provided the covalent inhibitor. All of the authors analyzed the data. Q.L., C.J.V., and G.F.G. wrote the paper.