Exploring the mechanism of zanamivir resistance in a neuraminidase mutant: a molecular dynamics study

Abstract

It is critical to understand the molecular basis of the drug resistance of influenza viruses to efficiently treat this infectious disease. Recently, H1N1 strains of influenza A carrying a mutation of Q136K in neuraminidase were found. The new strain showed a strong Zanamivir neutralization effect. In this study, normal molecular dynamics simulations and metadynamics simulations were employed to explore the mechanism of Zanamivir resistance. The wild-type neuraminidase contained a 3(10) helix before the 150 loop, and there was interaction between the 150 and 430 loops. However, the helix and the interaction between the two loops were disturbed in the mutant protein due to interaction between K136 and nearby residues. Hydrogen-bond network analysis showed weakened interaction between the Zanamivir drug and E276/D151 on account of the electrostatic interaction between K136 and D151. Metadynamics simulations showed that the free energy landscape was different in the mutant than in the wild-type neuraminidase. Conformation with the global minimum of free energy for the mutant protein was different from the wild-type conformation. While the drug fit completely into the active site of the wild-type neuraminidase, it did not match the active site of the mutant variant. This study indicates that the altered hydrogen-bond network and the deformation of the 150 loop are the key factors in development of Zanamivir resistance. Furthermore, the Q136K mutation has a variable effect on conformation of different N1 variants, with conformation of the 1918 N1 variant being more profoundly affected than that of the other N1 variants studied in this paper. This observation warrants further experimental investigation.

Figure 1. Interaction between active site residues and zanamivir plotted by ligplot from PDBsum website , .

ZMR is shown as ribbon with purple color, and active site residues are colored in brown. Carbon, nitrogen and oxygen atom is shown as black, blue and red balls. Green dashed lines represent hydrogen bond between ZMR and active site residues. Non-bonded contacts are shown as dark red “eyelashes”.

Figure 2. Root mean square deviation (RMSD) and fluctuation (RMSF) in two systems of two H1N1 viruses.

Root mean square deviation (RMSD) of heavy atoms of WT (Figure 2A) and Q136K (Figure 2B) in 1918 N1. Figure 2D and 2E show heavy atoms RMSD of WT and Q136K of 2009 N1. Black, red, green color lines represent three times simulation of apo form structures. Figure 2C and 2F show Root mean square fluctuation (RMSF) of each residue during apo form simulations for 1918 N1 and 2009 N1. Black and red lines represent WT and Q136K respectively.

Figure 3. Secondary structure propensity of residues in two systems of two H1N1 viruses.

Secondary structure propensity of residues in WT and Q136K of 1918 N1 (Figure 3A) and 2009 N1 (Figure 3B) respectively. The secondary structures were assessed by DSSP package.

Figure 4. Structure comparison between WT and Q136K in two H1N1 viruses.

Snapshot of WT (green) and Q136K (grey) of NA at t = 30 ns (Figure 4A for 1918 N1, 4C for 2009 N1). Representative structures of WT (green) and Q136K (grey) show the interaction between the 150 loop and the 430 loop (Figure 4B for 1918 N1, 4D for 2009 N1). The representative structure is the center of the largest cluster obtained in the whole simulation. Structures were superimposed by Pymol.

Figure 5. Probability of hydrogen bond formation between T148 and T439.

Probability of hydrogen bond formation between T148 and T439 in WT (black) and Q136K (red) of time evolution in 1918 N1 (Figure 5A) and 2009 N1 (Figure 5B). The probability was analyzed every 1 ns during simulation, and the dots here represent average value of the 3 trajectories.

Figure 6. Hydrogen bond formation between ZMR and binding sites.

Number of hydrogen bond formed between ZMR and active site residues in 1918 N1 (Figure 6A) and 2009 N1 (Figure 6B). Black and red lines represent WT and Q136K respectively.

Figure 7. Distance distribution and conformation comparison between WT and Q136K.

Minimal distance distribution between side chain of D151 and Q/K136. Lines in black, red and green color represent the distance distribution of the separate holo WT simulations, and lines in blue, magenta, cyan color in Figure 7A and 7C indicate Q136K simulations. The snapshot of WT and Q136K of holo form simulations at t = 30 ns, showing the interaction between ZMR and the active site residues. Structures in green and pink color indicate WT and Q136K in 1918 N1 (Figure 7B) and 2009 N1 (Figure 7D) respectively. Structures were superimposed by Pymol.

Figure 8. Free energy landscape and representing conformation of WT and Q136K in 1918 N1system.

Free energy landscape of WT (Figure 8A) and Q136K (Figure 8B), here CV1 is the distance between ZMR and 3 carbon α atoms of residues located at the bottom of binding pocket. CV2 represents the radius of gyration of selected group of heavy atoms located at the outer ring of binding pocket. A1 and A2 show the representative structures of the 1 and 2 local minimum of Figure 8A, and B1 and B2 show the representative structures of the 1 and 2 local minimum of Figure 8B. The 150 loop and the 430 loop is shown in orange color, amino acids that interact with ZMR is shown in green color. Free energy landscape was plot with OriginPro 8, and structure figure was plot using Pymol.

Figure 9. Free energy landscape of WT and Q136K in 2009 N1 system.

Free energy landscape of WT (Figure 9A) and Q136K (Figure 9B), CV1 is the distance between ZMR and 3 carbon α atoms of residues locate at the bottom of binding pocket. CV2 represents the radius of gyration of selected group of heavy atoms located at the outer ring of binding pocket.