Dynamics of a myoglobin mutant enzyme: 2D IR vibrational echo experiments and simulations

Myoglobin (Mb) double mutant T67R/S92D displays peroxidase enzymatic activity in contrast to the wild type protein. The CO adduct of T67R/S92D shows two CO absorption bands corresponding to the A(1) and A(3) substates. The equilibrium protein dynamics for the two distinct substates of the Mb double mutant are investigated by using two-dimensional infrared (2D IR) vibrational echo spectroscopy and molecular dynamics (MD) simulations. The time-dependent changes in the 2D IR vibrational echo line shapes for both of the substates are analyzed using the center line slope (CLS) method to obtain the frequency-frequency correlation function (FFCF). The results for the double mutant are compared to those from the wild type Mb. The experimentally determined FFCF is compared to the FFCF obtained from molecular dynamics simulations, thereby testing the capacity of a force field to determine the amplitudes and time scales of protein structural fluctuations on fast time scales. The results provide insights into the nature of the energy landscape around the free energy minimum of the folded protein structure.     

Ligand dynamics in myoglobin: calculation of infrared spectra for photodissociated NO.

We present molecular dynamics simulations of the photodissociated state of MbNO performed at 300 K using a fluctuating charge model for the nitric oxide (NO) ligand. After dissociation, NO is observed to remain mainly in the centre of the distal haem pocket, although some movement towards the primary docking site and the xenon-4 pocket can be seen. We calculate the NO infrared spectrum for the photodissociated ligand within the haem pocket and find a narrow peak in the range 1915-1922 cm(-1). The resulting blue shift of 1 to 8 cm(-1) compared to gas-phase NO is much smaller than the red shifts calculated and observed for carbon monoxide (CO) in Mb. A small splitting, due to NO in the xenon-4 pocket, is also observed. At lower temperatures, the spectra and conformational space explored by the ligand remain largely unchanged, but the electrostatic interactions with residue His64 become increasingly significant in determining the details of the ligand orientation within the distal haem pocket. The investigation of the effect of the L29F mutation reveals significant differences between the behaviour of NO and that of CO, and suggests a coupling between the ligand and the protein dynamics due to the different ligand dipole moments.     

On the formation, nature, stability and biological relevance of the primary reaction intermediates of myoglobins with hydrogen peroxide

The reaction between hydrogen peroxide and myoglobin (or haemoglobin) ferric haem is a two-electron redox process, yet the stable product is ferryl haem, retaining only one oxidizing equivalent. We have used SVD (singular value decomposition) and global spectroscopic analysis to examine the transient primary spectral intermediates in this reaction, which have been reported as either "compound 0"(ferric peroxide) or "compound I"(ferryl and porphyrin cation radical) types and which may precede the formation of ferrylmyoglobin. To test the hypothesis that the distal histidine facilitates ferryl formation we studied the myoglobin-like haemoglobin from the gastropod mollusc Aplysia fasciata, where this histidine is replaced by valine and its hydrogen bonding role is taken up by a non-homologous arginine. In this protein, consistent with the distal histidine hypothesis, a compound 0 intermediate is formed identified by an EPR spectrum typical of low spin ferric haem complexes. It is significantly more stable than any species seen with mammalian myoglobin. Thirdly, as ferryl haems and associated free radicals may play a role in disease, we have studied the action of myoglobin-peroxide mixtures towards external reductants. Even at a low pH, where ferrylmyoglobin is protonated and in its most reactive state, pre-incubation with reducing donors, including one-electron donors such as ferrocyanide, prior to peroxide addition renders both oxidizing equivalents available. The physiological antioxidant vitamin, ascorbate, is also able to trap both reactive species. Myoglobin can therefore act as a true ascorbate peroxidase. Ascorbate in vivo may be critical in controlling and preventing toxic side reactions of this and related haem proteins.     

Reaction of Aplysia limacina metmyoglobin with hydrogen peroxide

Myoglobin (Mb) from gastropod mollusc Aplysia limacina shows only 20% sequence homology to the 'prototype' sperm whale Mb but exhibits a typical Mb fold and can reversibly bind oxygen. An intriguing feature of aplysia Mb is that it lacks the distal histidine and displays a ligand stabilisation based on an arginine. Here we report the reaction of aplysia metMb with hydrogen peroxide studied by optical and electron paramagnetic resonance (EPR) spectroscopies. Two electron oxidation of the protein by H2O2 results in formation of two intermediates typical for this class of reactions, the oxoferryl haem state and a globin-bound free radical. An unusual characteristic of the aplysia Mb reaction is formation, prior to haem oxidation, of an optically distinct compound with an EPR spectrum typical of the low spin Fe3+ haem state. This compound is interpreted as the complex between H2O2 and the ferric haem state (Compound), formed prior to cleavage of the dioxygen bond. We conclude that H2O2 is singly deprotonated in Compound which can thus be notated as [Fe3+--OOH]. A new low spin ferric haem state has been observed over the period of Compound decay, and hypotheses have been formulated as to its identity and role. The location of the protein bound radical observed in aplysia Mb is discussed in light of the fact that the protein does not have any tyrosine residues, the most common site of free radical formation in the haem protein/peroxide systems. All intermediates of the reaction are kinetically characterised.     

Study on the drug resistance and the binding mode of HIV-1 integrase with LCA inhibitor

Human immunodeficiency virus type 1 (HIV-1) integrase (IN) is an essential enzyme in the lifecycle of this virus and also an important target for the study of anti-HIV drugs. The binding mode of the wild type IN core domain and its G140S mutant with L-Chicoric acid (LCA) inhibitor were investigated by using multiple conformation molecular docking and molecular dynamics (MD) simulation. Based on the binding modes, the drug resistance mechanism was explored for the G140S mutant of IN with LCA. The results indicate that the binding site of the G140S mutant of IN core domain with LCA is different from that of the core domain of the wild type IN, which leads to the partial loss of inhibition potency of LCA. The flexibility of the IN functional loop region and the interactions between Mg2 ion and the three key residues (i.e., D64, D116, E152) stimulate the biological operation of IN. The drug resistance also lies in several other important effects, such as the repulsion between LCA and E152 in the G140S mutant core domain, the weakening of K159 binding with LCA and Y143 pointing to the pocket of the G140S mutant. All of the above simulation results agree well with experimental data, which provide us with some helpful information for designing the drug of anti-HIV based on the structure of IN.     

Crystals in Minutes: Instant On‐Site Microcrystallisation of Various Flavours of the CYP102A1 (P450BM3) Haem Domain

Despite CYP102A1 (P450BM3) representing one of the most extensively researched metalloenzymes, crystallisation of its haem domain upon modification can be a challenge. Crystal structures are indispensable for the efficient structure‐based design of P450BM3 as a biocatalyst. The abietane diterpenoid derivative N‐abietoyl‐l ‐tryptophan (AbiATrp) is an outstanding crystallisation accelerator for the wild‐type P450BM3 haem domain, with visible crystals forming within 2 hours and diffracting to a near‐atomic resolution of 1.22 Å. Using these crystals as seeds in a cross‐microseeding approach, an assortment of P450BM3 haem domain crystal structures, containing previously uncrystallisable decoy molecules and diverse artificial metalloporphyrins binding various ligand molecules, as well as heavily tagged haem‐domain variants, could be determined. Some of the structures reported herein could be used as models of different stages of the P450BM3 catalytic cycle.     

Optical biosensing of nitric oxide using the metalloprotein cytochrome c'

The metalloprotein cytochrome c' was extracted and purified from the bacterium Paracoccus denitrificans in order to develop a specific biosensing system for nitric oxide (NO). The metalloprotein was encapsulated in a porous silicate sol-gel glass to enable spectroscopic changes in the haem centre as a function of NO ligation to be quantified using absorption measurements. Spectroscopic evidence suggested that, between 2 and 4 d after encapsulation, the cytochrome c' protein changed conformation in the locality of the haem moiety, possibly from a five to a six coordinate haem centre. Such conformational changes were also observed when the cytochrome c' was stored in solution, although over a 2-3 month period. The conformational changes occurring in the protein altered the spectral characteristics of the reduced, oxidised and nitrosyl complex of the cytochrome c' and appear to change the binding affinity of the protein towards NO. However, the encapsulated (reconformed) cytochrome c' was shown to retain its selectivity towards NO with good reproducibility (seven consecutive measurements of NO produced an intensity value with a relative standard deviation of 0.28%). An NO calibration curve, using the in situ release of NO from the donor diethylamine NONOate, was obtained for the encapsulated cytochrome c' with an approximate working range of 10-400 mumol l-1.     

Ultrafast infrared spectroscopy reveals water-mediated coherent dynamics in an enzyme active site

Understanding the impact of fast dynamics upon the chemical processes occurring within the active sites of proteins and enzymes is a key challenge that continues to attract significant interest, though direct experimental insight in the solution phase remains sparse. Similar gaps in our knowledge exist in understanding the role played by water, either as a solvent or as a structural/dynamic component of the active site. In order to investigate further the potential biological roles of water, we have employed ultrafast multidimensional infrared spectroscopy experiments that directly probe the structural and vibrational dynamics of NO bound to the ferric haem of the catalase enzyme from Corynebacterium glutamicum in both H₂O and D₂O. Despite catalases having what is believed to be a solvent-inaccessible active site, an isotopic dependence of the spectral diffusion and vibrational lifetime parameters of the NO stretching vibration are observed, indicating that water molecules interact directly with the haem ligand. Furthermore, IR pump–probe data feature oscillations originating from the preparation of a coherent superposition of low-frequency vibrational modes in the active site of catalase that are coupled to the haem ligand stretching vibration. Comparisons with an exemplar of the closely-related peroxidase enzyme family shows that they too exhibit solvent-dependent active-site dynamics, supporting the presence of interactions between the haem ligand and water molecules in the active sites of both catalases and peroxidases that may be linked to proton transfer events leading to the formation of the ferryl intermediate Compound I. In addition, a strong, water-mediated, hydrogen bonding structure is suggested to occur in catalase that is not replicated in peroxidase; an observation that may shed light on the origins of the different functions of the two enzymes.     

Reduction of conformational mobility and aggregation in W60G β2‐microglobulin: assessment by 15N NMR relaxation

The amyloid pathology associated with long‐term haemodialysis is due to the deposition of β₂‐microglobulin, the non‐polymorphic light chain of class I major histocompatibility complex, that accumulates at bone joints into amyloid fibrils. Several lines of evidence show the relevance of the tryptophan residue at position 60 for the fibrillogenic transition of the protein. A comparative ¹⁵N NMR relaxation analysis is presented for wild‐type human β₂‐microglobulin and W60G β₂‐microglobulin, i.e. the mutant with a glycyne replacing the natural tryptophan residue at position 60. The experimental data, collected at 11.4 T and 310 K, were analyzed by means of the reduced spectral density approach. Molecular dynamics (MD) simulations and corresponding thermodynamic integration, together with hydrodynamic calculations were performed to support data interpretation. The analysis results for the mutant protein are consistent with a reduced aggregation with respect to the wild‐type counterpart, as a consequence of an increased conformational rigidity probed by either NMR relaxation and MD simulations. Although dynamics in solution is other than fibrillar competence, the assessed properties of the mutant protein can be related with its reduced ability of forming fibrils when seeded in 20% trifluoroethanol. Copyright © 2013 John Wiley & Sons, Ltd.     

Rationalizing perhydrolase activity of aryl-esterase and subtilisin Carlsberg mutants by molecular dynamics simulations of the second tetrahedral intermediate state

The perhydrolysis reaction in hydrolases is an important example of catalytic promiscuity and has many potential industrial applications. The mechanisms of perhydrolase activity of a subtilisin Carlsberg mutant and of an aryl-esterase mutant have been investigated using classical molecular dynamics simulations of the second tetrahedral intermediate (TI) state. The simulations demonstrated that hydrogen bonding between the second TI of the perhydrolysis reaction is possible in the mutants but not wild type. The stabilization by hydrogen bonds was specific for the perhydrolysis intermediate and either no hydrogen bonding or only weakened hydrogen bonding to the second TI state of the hydrolysis reaction was observed. Furthermore, a significant hindrance to the formation of the catalytically important hydrogen bond between His64 and Ser221 in the catalytic triad by competing hydrogen bonds was found for the subtilisin mutant but not wild type enzyme in case of the hydrolysis intermediate. The opposite was observed in case of the perhydrolysis intermediate. The result offers a qualitative explanation for the overall reduced hydrolysis activity of the subtilisin mutant. In addition, the simulations also explain qualitatively the perhydrolysis activity of the enzyme variants and may be helpful for designing enzyme mutants with further improved perhydrolysis activity.     

The H93G myoglobin cavity mutant as a versatile template for modeling heme proteins: ferrous, ferric, and ferryl mixed-ligand complexes with imidazole in the cavity

One of the difficulties in preparing accurate ambient-temperature model complexes for heme proteins, particularly in the ferric state, has been the generation of mixed-ligand adducts: complexes with different ligands on either side of the heme. The difference in the accessibility of the two sides of the heme in the H93G cavity mutant of myoglobin (Mb) provides a potential general solution to this problem. To demonstrate the versatility of H93G Mb for the preparation of heme protein models, numerous mixed-ligand adducts of ferrous, ferric, and ferryl imidazole-ligated H93G (H93G(Im) Mb) have been prepared. The complexes have been characterized by electronic absorption and magnetic circular dichroism (MCD) spectroscopy in comparison to analogous derivatives of wild type Mb. The starting ferric H93G(Im) Mb state spectroscopically resembles wild-type ferric Mb as expected for a complex containing a single imidazole in the proximal cavity and water bound on the distal side. Addition of a sixth ligand to ferric H93G(Im) Mb, whether charge neutral (imidazole) or anionic (cyanide and azide), results in formation of six-coordinate low-spin complexes with MCD characteristics similar to those of parallel derivatives of wild-type ferric Mb. Reduction of ferric H93G(Im) Mb and subsequent exposure to either CO, NO, or O2 produces ferrous complexes (deoxy, CO, NO, and O2) that consistently exhibit MCD spectra similar to the analogous ferrous species of wild-type ferrous Mb. Most interestingly, reaction of ferric H93G(Im) Mb with H2O2 results in the formation of a stable high-valent oxoferryl complex with MCD characteristics that are essentially identical to those of oxoferryl wild-type Mb. The generation of such a wide array of mixed-ligand heme complexes demonstrates the efficacy of the H93G Mb cavity mutant as a template for the preparation of heme protein model complexes.     

Hydrogen‐bonding interactions in the active site of a low molecular weight protein‐tyrosine phosphatase

Molecular dynamics simulation of the Michaelis complex, phospho‐enzyme intermediate, and the wild‐type and C12S mutant have been carried out to examine hydrogen‐bonding interactions in the active site of the bovine low molecular weight protein‐tyrosine phosphatase (BPTP). It was found that the S_γ atom of the nucleophilic residue Cys‐12 is ideally located at a position opposite from the phenylphosphate dianion for an inline nucleophilic substitution reaction. In addition, electrostatic and hydrogen‐bonding interactions from the backbone amide groups of the phosphate‐binding loop strongly stabilize the thiolate anion, making Cys‐12 ionized in the active site. In the phospho‐enzyme intermediate, three water molecules are found to form strong hydrogen bonds with the phosphate group. In addition, another water molecule can be identified to form bridging hydrogen bonds between the phosphate group and Asp‐129, which may act as the nucleophile in the subsequent phosphate hydrolysis reaction, with Asp‐129 serving as a general base. The structural difference at the active site between the wild‐type and C12S mutant has been examined. It was found that the alkoxide anion is significantly shifted toward one side of the phosphate binding loop, away from the optimal position enjoyed by the thiolate anion of the wild‐type enzyme in an S_N2 process. This, coupled with the high pK_a value of an alcoholic residue, makes the C12S mutant catalytically inactive. These molecular dynamics simulations provided details of hydrogen bonding interactions in the active site of BPTP, and a structural basis for further studies using combined quantum mechanical and molecular mechanical potential to model the entire dephosphorylation reaction by BPTP. © 2000 John Wiley & Sons, Inc. J Comput Chem 21: 1192–1203, 2000     

Analysis of a Critical Residue Determining Herbicide Efficiency Sensitivity in Carboxyltransferase Domain of Acetyl-CoA Carboxylase from Poaceae by Homology Modeling and Free Energy Simulation

Carboxyltransferase domain(CT) of acetyl-coenzyme A carboxylase(ACCase, EC 6.4.1.2) from a family of Poaceae is an important target of commercial herbicide APPs for controlling grass weed growth. As the abuse of APPs herbicides, the resistant ACCase due to the mutation of a single residue(Ile→Leu), which is lo-cated in CT active site, is emergent in many populations and species of Poaceae. So it is urgent to understand the re-sistant mecha-nism so as to design new effect herbicides. Herein lies the complex of CT dimmer from Lolium rigi-dum and herbicide haloxyfop successfully constructed for wild type enzyme and Ile/Leu mutant, respectively, pro-viding a basis for explaining the resistance from microscopic structure. Moreover, the binding free energy difference between wild type and mutant enzymes was predicted in good agreement with the known observation, and the various contributions to it were analyzed, by Molecular mechanics-Poisson-Boltzmann surface area(MM-PBSA) method. The results indicate the van der Waals interaction difference between the protein and inhibitor, –22.94 kJ/mol of CT wild type lower than that of mutant, is the major reason for resistance. Structure analysis further suggests that van der Waals interaction difference is originated from the steric hindrance between the side chain of mutated residue Leu and the chiral methyl group of haloxyfop. All these findings enhance the understanding of resistant mechanism of ACCase to herbicide by Ile/Leu mutation and provide an important clue for the rational design of high effective herbicides.     

Structural Interpretation of Metastable States in Myoglobin–NO

Nitric oxide binding and unbinding from myoglobin (Mb) is central to the function of the protein. By using reactive molecular dynamics (MD) simulations, the dynamics following NO dissociation were characterized in both time and space. Ligand rebinding can be described by two processes on the 10 ps and 100 ps timescale, which agrees with recent optical and X‐ray absorption experiments. Explicitly including the iron out‐of‐plane (Fe‐oop) coordinate is essential for a meaningful interpretation of the data. The proposed existence of an “Fe‐oop/NO‐bound” state is confirmed and assigned to NO at a distance of approximately 3 Å away from the iron atom. However, calculated XANES spectra suggest that it is diffcult to distinguish between NO close to the heme‐Fe and positions further away in the primary site. Another elusive state, with Fe?ON coordination, was not observed experimentally because it is masked by the energetically more favorable but dissociative ⁴A state in this region, which makes the Fe?ON local minimum unobservable in wild‐type Mb. However, suitable active‐site mutations may stabilize this state.     

Primary Reaction Dynamics of Proteorhodopsin Mutant D97N Observed by Femtosecond Infrared and Visible Spectroscopy†

Femtosecond time‐resolved spectroscopy in the visible and IR range was utilized to study the primary reaction dynamics of the proteorhodopsin (PR) D97N mutant in comparison with wild type PR at different pH values. The analysis of the data obtained in the mid‐IR closely resembles the results for wild type PR. The observation of the first ground state intermediate K is initially obscured by a complex reaction scheme of vibrational relaxation and heating effects, but its spectral signature clearly emerges at long delay times. In the visible range, a biexponential decay of the excited state within 30 ps and the formation of the K photoproduct is observed. The decay time constants derived for the D97N mutant in D₂O are slightly larger than in H₂O due to H/D exchange. This kinetic isotope effect is even less pronounced than for wild type PR at pH 6. These results support the current notion of a pH dependent hydrogen bonding network in the retinal binding pocket of PR and a weaker interaction between the retinal Schiff base and the counter ion complex compared to bacteriorhodopsin.     

Kinetic Analysis and Structural Interpretation of Competitive Ligand Binding for NO Dioxygenation in Truncated Hemoglobin N

The conversion of nitric oxide (NO) into nitrate (NO₃^?) by dioxygenation protects cells from lethal NO. Starting from NO‐bound heme, the first step in converting NO into benign NO₃^? is the ligand exchange reaction FeNO+O₂→FeO₂+NO, which is still poorly understood at a molecular level. For wild‐type (WT) truncated hemoglobin N (trHbN) and its Y33A mutant, the calculated barriers for the exchange reaction differ by 1.5 kcal mol^?1, compared with 1.7 kcal mol^?1 from experiment. It is directly confirmed that the ligand exchange reaction is rate‐limiting in trHbN and that entropic contributions account for 75 % of the difference between the WT and the mutant. Residues Tyr 33, Phe 46, Val 80, His 81, and Gln 82 surrounding the active site are expected to control the reaction path. By comparison with electronic structure calculations, the transition state separating the two ligand‐bound states was assigned to a ²A state.     

Functional assessment of “in vivo” and “in silico” mutations in the guanine binding site of RNase T1: A DFT study

Experimentally, the functional assessment of amino acid side chains in proteins is carried out by comparing parameters such as binding constants for the wild‐type protein and a mutant protein in which the considered side chain is deleted. In the present study, we apply a density functional theory (DFT) methodology to obtain changes in binding energy upon mutations in the enzyme ribonuclease T₁. Mutant structures were either taken directly from crystallographic data (“in vivo”) allowing for conformational changes upon mutation, or derived from the wild‐type (“in silico”). Excluding entropic contributions, the computed interaction energy changes upon mutation in vivo correlate qualitatively well with experimental binding free energy changes. In contrast, the in silico approach does not perform as well, especially for residues that contribute largely to binding. Subsequently, we assessed the applicability of the in vivo approach by analyzing the functional cooperativity between pairs of side chains. © 2004 Wiley Periodicals, Inc. Int J Quantum Chem, 2004     

Toward quantitative simulations of carbon monoxide escape pathways in myoglobin

Straightforward molecular dynamics trajectories have been computed to explore the diffusion of carbon monoxide through myoglobin. The classical equations of motion were integrated for 2 ns and the resulting pathways analyzed. Two types of runs were examined. Type i: Myoglobin and a ligand embedded in a periodic box with 9996 water molecules; the water molecules are rigid but the bonds of the protein are flexible. Type ii: Myoglobin with a solvation shell (153 water molecules) in which all bond lengths are fixed. In trajectories of type i, the diffusing ligand visits a significant part of the protein matrix and was not constrained to the proximity of the heme pocket before escaping. The maximum time of the trajectories was 2 ns. It was shorter if the ligand escaped earlier. Two ligands (from a total of 88) escape to the solvent from nonclassical gates (non-E-helix gates). In trajectories of type ii, the overall fluctuations of the protein are smaller and the ligand explores significantly smaller internal space. The escape rate from type ii trajectories (11 of 400) is comparable to type i and is not dramatically different from experiment (1 of 100). Interestingly, the two simulations with comparable rates sampled different pathways. In trajectories of type ii, we observe escapes from the classical gate (His 64) and from the Xe4 cavity. Further studies (that are underway) are required to define the escape pathways and the overall rate.     

PHOTOCONTROL OF ANTHOCYANIN SYNTHESIS IN TOMATO SEEDLINGS: A GENETIC APPROACH*

The photocontrol of anthocyanin synthesis in dark-grown seedlings of tomato (Lycopersicon esculentum Mill.) has been studied in an aurea (au) mutant which is deficient in the labile type of phytochrome, a high pigment (hp) mutant which has the wild-type level of phytochrome and the double mutant au/hp , as well as the wild type. The hp mutant demonstrates phytochrome control of anthocyanin synthesis in response to a single red light (RL) pulse, whereas there is no measurable response in the wild type and au mutant. After pretreatment with 12 h blue light (BL) the phytochrome regulation of anthocyanin synthesis is 10-fold higher in the hp mutant than in the wild type, whilst no anthocyanin is detectable in the au mutant, thus suggesting that it is the labile pool of phytochrome which regulates anthocyanin synthesis. The au/hp double mutant exhibits a small (3% of that in the hp mutant) RL/far-red light (FR)-reversible regulation of anthocyanin synthesis following a BL pretreatment. It is proposed that the hp mutant is hypersensitive to the FR-absorbing form of phytochrome (P_fr) and that this (hypersensitivity) establishes response to the low level of P_fl. (below detection limits in phytochrome assays) in the au/hp double mutant.