Limits...
How modification of accessible lysines to phenylalanine modulates the structural and functional properties of horseradish peroxidase: a simulation study.

Navapour L, Mogharrab N, Amininasab M - PLoS ONE (2014)

Bottom Line: We also found a new persistent hydrogen bond between the protein moiety (F174) and the heme prosthetic group as well as two stitching hydrogen bonds between the connecting loops GH and F'F″ in mutated HRP.However, detailed analysis of functionally related structural properties and dynamical features suggests reduced reactivity of the enzyme toward its substrates.Results also demonstrated that the area and volume of the aromatic-substrate binding pocket are significantly decreased upon modifications.

View Article: PubMed Central - PubMed

Affiliation: Biophysics and Computational Biology Laboratory, Department of Biology, College of Sciences, Shiraz University, Shiraz, Iran.

ABSTRACT
Horseradish Peroxidase (HRP) is one of the most studied peroxidases and a great number of chemical modifications and genetic manipulations have been carried out on its surface accessible residues to improve its stability and catalytic efficiency necessary for biotechnological applications. Most of the stabilized derivatives of HRP reported to date have involved chemical or genetic modifications of three surface-exposed lysines (K174, K232 and K241). In this computational study, we altered these lysines to phenylalanine residues to model those chemical modifications or genetic manipulations in which these positively charged lysines are converted to aromatic hydrophobic residues. Simulation results implied that upon these substitutions, the protein structure becomes less flexible. Stability gains are likely to be achieved due to the increased number of stable hydrogen bonds, improved heme-protein interactions and more integrated proximal Ca2+ binding pocket. We also found a new persistent hydrogen bond between the protein moiety (F174) and the heme prosthetic group as well as two stitching hydrogen bonds between the connecting loops GH and F'F″ in mutated HRP. However, detailed analysis of functionally related structural properties and dynamical features suggests reduced reactivity of the enzyme toward its substrates. Molecular dynamics simulations showed that substitutions narrow the bottle neck entry of peroxide substrate access channel and reduce the surface accessibility of the distal histidine (H42) and heme prosthetic group to the peroxide and aromatic substrates, respectively. Results also demonstrated that the area and volume of the aromatic-substrate binding pocket are significantly decreased upon modifications. Moreover, the hydrophobic patch functioning as a binding site or trap for reducing aromatic substrates is shrunk in mutated enzyme. Together, the results of this simulation study could provide possible structural clues to explain those experimental observations in which the protein stability achieved concurrent with a decrease in enzyme activity, upon manipulation of charge/hydrophobicity balance at the protein surface.

Show MeSH

Related in: MedlinePlus

Time dependence of the backbone RMSD.(A) Backbone RMSD with respect to the starting structure during the entire course of 100 ns MD simulations. (B) Backbone RMSD relative to the structure of 70th ns (i.e. the starting structures of the analyzed time frames) during the last 30 ns MD simulations. Those of the n-HRP are reported in black and those of the p-HRP are shown in red.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4196758&req=5

pone-0109062-g002: Time dependence of the backbone RMSD.(A) Backbone RMSD with respect to the starting structure during the entire course of 100 ns MD simulations. (B) Backbone RMSD relative to the structure of 70th ns (i.e. the starting structures of the analyzed time frames) during the last 30 ns MD simulations. Those of the n-HRP are reported in black and those of the p-HRP are shown in red.

Mentions: Using the known x-ray crystallographic structure of HRP C (PDB: 1ATJ), two 3D molecular models of HRP C, n-HRP and p-HRP, were constructed differing in the residues 174, 232, and 241 (Fig. 1). Keeping the amino acid pattern of the original X-ray structure, the control model (n-HRP) was built with cationic lysines at these positions. In the next model (p-HRP) charged lysine residues were substituted by hydrophobic aromatic phenylalanine residues. The overall stability and structural relaxation of the enzymes were monitored by computing time evolution of the root mean square deviation (RMSD) of the backbone atoms along the simulations. The RMSD of n-HRP and p-HRP with respect to their starting structure were 2.46±0.47 and 3.03±0.52 Å, respectively. This indicates that the structure of mutated protein undergoes more conformational changes along the simulation. Comparison of the RMSD per residue profiles of the MD trajectories revealed that these changes are concentrated in the N terminal part of the protein structure (including residues 1–12), C terminal part of the loop D′E (residues 140–144) and the helix E (residues 145–153) (data not shown). For n-HRP model, the backbone RMSD as a function of time reaches a relative plateau after about 50 nanoseconds (ns) of simulation, but for p-HRP such a plateau is achieved after about 70 ns (Fig. 2A). Hence, to be statistically comparable, our analyses were focused on those trajectories obtained from the last 30 ns of simulations for both the native and mutated proteins. In order to check whether the sampling of conformational space is sufficient, the production MD period was divided into four 25 ns parts and the cosine content of the first 4 principal components was calculated for each sub-trajectory. For both n-HRP and p-HRP models, the best results of cosine content were obtained for the last quarter part of the 100 ns simulations. The values of the first 4 principal components for this time window were calculated to be 0.146, 0.459, 0.036, 0.110 for n-HRP and 0.197, 0.517, 0.010, 0.001 for p-HRP.The stability of the fluctuation of the potential energy was also examined by calculating the ratio between the variance and average of potential energy. This ratio for n-HRP and p-HRP was respectively about 0.00099 and 0.00092, thus showing that energy was conserved during the simulations and providing additional evidences indicating that the simulations and models were stabilized.


How modification of accessible lysines to phenylalanine modulates the structural and functional properties of horseradish peroxidase: a simulation study.

Navapour L, Mogharrab N, Amininasab M - PLoS ONE (2014)

Time dependence of the backbone RMSD.(A) Backbone RMSD with respect to the starting structure during the entire course of 100 ns MD simulations. (B) Backbone RMSD relative to the structure of 70th ns (i.e. the starting structures of the analyzed time frames) during the last 30 ns MD simulations. Those of the n-HRP are reported in black and those of the p-HRP are shown in red.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4196758&req=5

pone-0109062-g002: Time dependence of the backbone RMSD.(A) Backbone RMSD with respect to the starting structure during the entire course of 100 ns MD simulations. (B) Backbone RMSD relative to the structure of 70th ns (i.e. the starting structures of the analyzed time frames) during the last 30 ns MD simulations. Those of the n-HRP are reported in black and those of the p-HRP are shown in red.
Mentions: Using the known x-ray crystallographic structure of HRP C (PDB: 1ATJ), two 3D molecular models of HRP C, n-HRP and p-HRP, were constructed differing in the residues 174, 232, and 241 (Fig. 1). Keeping the amino acid pattern of the original X-ray structure, the control model (n-HRP) was built with cationic lysines at these positions. In the next model (p-HRP) charged lysine residues were substituted by hydrophobic aromatic phenylalanine residues. The overall stability and structural relaxation of the enzymes were monitored by computing time evolution of the root mean square deviation (RMSD) of the backbone atoms along the simulations. The RMSD of n-HRP and p-HRP with respect to their starting structure were 2.46±0.47 and 3.03±0.52 Å, respectively. This indicates that the structure of mutated protein undergoes more conformational changes along the simulation. Comparison of the RMSD per residue profiles of the MD trajectories revealed that these changes are concentrated in the N terminal part of the protein structure (including residues 1–12), C terminal part of the loop D′E (residues 140–144) and the helix E (residues 145–153) (data not shown). For n-HRP model, the backbone RMSD as a function of time reaches a relative plateau after about 50 nanoseconds (ns) of simulation, but for p-HRP such a plateau is achieved after about 70 ns (Fig. 2A). Hence, to be statistically comparable, our analyses were focused on those trajectories obtained from the last 30 ns of simulations for both the native and mutated proteins. In order to check whether the sampling of conformational space is sufficient, the production MD period was divided into four 25 ns parts and the cosine content of the first 4 principal components was calculated for each sub-trajectory. For both n-HRP and p-HRP models, the best results of cosine content were obtained for the last quarter part of the 100 ns simulations. The values of the first 4 principal components for this time window were calculated to be 0.146, 0.459, 0.036, 0.110 for n-HRP and 0.197, 0.517, 0.010, 0.001 for p-HRP.The stability of the fluctuation of the potential energy was also examined by calculating the ratio between the variance and average of potential energy. This ratio for n-HRP and p-HRP was respectively about 0.00099 and 0.00092, thus showing that energy was conserved during the simulations and providing additional evidences indicating that the simulations and models were stabilized.

Bottom Line: We also found a new persistent hydrogen bond between the protein moiety (F174) and the heme prosthetic group as well as two stitching hydrogen bonds between the connecting loops GH and F'F″ in mutated HRP.However, detailed analysis of functionally related structural properties and dynamical features suggests reduced reactivity of the enzyme toward its substrates.Results also demonstrated that the area and volume of the aromatic-substrate binding pocket are significantly decreased upon modifications.

View Article: PubMed Central - PubMed

Affiliation: Biophysics and Computational Biology Laboratory, Department of Biology, College of Sciences, Shiraz University, Shiraz, Iran.

ABSTRACT
Horseradish Peroxidase (HRP) is one of the most studied peroxidases and a great number of chemical modifications and genetic manipulations have been carried out on its surface accessible residues to improve its stability and catalytic efficiency necessary for biotechnological applications. Most of the stabilized derivatives of HRP reported to date have involved chemical or genetic modifications of three surface-exposed lysines (K174, K232 and K241). In this computational study, we altered these lysines to phenylalanine residues to model those chemical modifications or genetic manipulations in which these positively charged lysines are converted to aromatic hydrophobic residues. Simulation results implied that upon these substitutions, the protein structure becomes less flexible. Stability gains are likely to be achieved due to the increased number of stable hydrogen bonds, improved heme-protein interactions and more integrated proximal Ca2+ binding pocket. We also found a new persistent hydrogen bond between the protein moiety (F174) and the heme prosthetic group as well as two stitching hydrogen bonds between the connecting loops GH and F'F″ in mutated HRP. However, detailed analysis of functionally related structural properties and dynamical features suggests reduced reactivity of the enzyme toward its substrates. Molecular dynamics simulations showed that substitutions narrow the bottle neck entry of peroxide substrate access channel and reduce the surface accessibility of the distal histidine (H42) and heme prosthetic group to the peroxide and aromatic substrates, respectively. Results also demonstrated that the area and volume of the aromatic-substrate binding pocket are significantly decreased upon modifications. Moreover, the hydrophobic patch functioning as a binding site or trap for reducing aromatic substrates is shrunk in mutated enzyme. Together, the results of this simulation study could provide possible structural clues to explain those experimental observations in which the protein stability achieved concurrent with a decrease in enzyme activity, upon manipulation of charge/hydrophobicity balance at the protein surface.

Show MeSH
Related in: MedlinePlus