Combination of Neutron Scattering and Molecular Dynamics to Determine Internal Motions in Biomolecules

Abstract

The forms and frequencies of atomic dynamics on the pico- and nanosecond timescales are accessible experimentally using incoherent neutron scattering. Molecular dynamics simulations cover the same space and time domains and neutron scattering intensities can be calculated from the simulations for direct comparison with experiment. To illustrate the complementarity of neutron scattering and molecular dynamics we examine measured and simulation-derived elastic incoherent scattering profiles from myoglobin and from the crystalline alanine dipeptide. Elastic incoherent scattering gives information on the geometry of the volume accessible to the atoms in the samples. The simulation-derived dipeptide elastic scattering profiles are in reasonable accord with experiment, deviations being due to the sampling limitations in the simulations and experimental detector normalisation procedures. The simulated dynamics is decomposed, revealing characteristic profiles due to rotational diffusional and translational vibrational motions of the methyl groups. In myoglobin, for which the timescale of the simulation matches more closely that accessible to the experiment, good agreement is seen for the elastic incoherent structure factor. This indicates that the space sampled by the hydrogen atoms in the protein on the timescale <100 ps is well represented by the simulation. Part of the helix atom fluctuations can be described in terms of rigid helix motions.     

The asymmetric ATPase cycle of the thermosome: elucidation of the binding, hydrolysis and product-release steps

Using a combination of intrinsic fluorescence to report ATP-induced rearrangements, quenched-flow to measure ATP hydrolysis "on-enzyme" and optical methods to probe the kinetics of product release, we have begun to dissect the process of energy transduction in the thermosome, a type II chaperonin from Thermoplasma acidophilum. Stoichiometric measurements of ATP binding reveal the tight association of eight nucleotide molecules per hexa-decamer, implying the filling of only one ring owing to strong negative cooperativity. After binding, we show that these eight ATP molecules are hydrolysed over the next 50 s, after which hydrolysis slows down markedly during the establishment of the steady state in the ATPase reaction, demonstrating that the kinetic system is off-rate limited. Looking in more detail, this rapid first-turnover can be dissected into two phases; the first occurring with a half-time of 0.8 s, the second with a half-time of 14 s, possibly reflecting the differential behaviour of the four alpha and four beta subunits in a single thermosome ring. To investigate the post-hydrolytic events, we used two heat-stable enzyme-linked optical assays to measure the rate of evolution of ADP and of phosphate from the thermosome active site. Neither product showed a rapid dissociation phase prior to the establishment of the steady state, showing that both are released slowly at a rate that limits the cycle. These data highlight the importance of the highly populated thermosome/ADP/Pi complex in the molecular mechanism.     

Methods for structural characterization of prefibrillar intermediates and amyloid fibrils

Protein fibrillation is first and foremost a structural phenomenon. Adequate structural investigation of the central conformational individuals of the fibrillation process is however exceedingly difficult. This is due to the nature of the process, which may be described as a dynamically evolving equilibrium between a large number of structural species. These are furthermore of highly diverging sizes and present in very uneven amounts and timeframes. Different structural methods have different strengths and limitations. These, and in particular recent advances within solution analysis of the undisturbed equilibrium using small angle X-ray scattering, are reviewed here.     

Dynamical heterogeneity of specific amino acids in bacteriorhodopsin

Components of biological macromolecules, complexes and membranes are animated by motions occurring over a wide range of time and length scales, the synergy of which is at the basis of biological activity. Understanding biological function thus requires a detailed analysis of the underlying dynamical heterogeneity. Neutron scattering, using specific isotope labeling, and molecular dynamics simulations were combined in order to study the dynamics of specific amino acid types in bacteriorhodopsin within the purple membrane (PM) of Halobacterium salinarum. Motions of leucine, isoleucine and tyrosine residues on the pico- to nanosecond time scale were examined separately as a function of temperature from 20 to 300 K. The dynamics of the three residue types displayed different temperature dependence: isoleucine residues have larger displacements compared to the global PM above 120 K; leucine residues have displacements similar to that of PM in the entire temperature range studied; and tyrosine residues have displacements smaller than that of the average membrane in an intermediate temperature range. Experimental features were mostly well reproduced by molecular dynamics simulations performed at five temperatures, which allowed the dynamical characterisation of the amino acids under study as a function of local environment. The resulting dynamical map of bacteriorhodopsin revealed that movements of a specific residue are determined by both its environment and its residue type.     

Molecular dynamics study of a phospholipid monolayer at a water/triglyceride interface: towards lipid emulsion modelling

In order to mimic the surface of parenteral nutrition emulsion droplets, the first molecular dynamics simulation of a palmitoyloleoylphosphatidylcholine (POPC) monolayer at a water/triglyceride (trilinoleoylglycerol, LLL) interface was performed. Triglyceride influence was evaluated by comparing computed phospholipid properties to the ones in a similarly modelled hydrated POPC bilayer. As expected, polar head properties (molecular area, lipid hydration, headgroup orientation) were not affected by triglycerides. In contrast, slight differences were observed on phospholipid alkyl tail region (order parameter, diffusion, Van der Waals interactions). This first approach can reasonably be extended to a further more realistic multicomponent model of clinical nutrition emulsions.     

H-bonding networks of the distal residues and water molecules in the active site of Thermobifida fusca hemoglobin

The ferric form of truncated hemoglobin II from Thermobifida fusca (Tf-trHb) and its triple mutant WG8F-YB10F-YCD1F at neutral and alkaline pH, and in the presence of CN⁻ have been characterized by resonance Raman spectroscopy, electron paramagnetic resonance spectroscopy, and molecular dynamics simulations. Tf-trHb contains three polar residues in the distal site, namely TrpG8, TyrCD1 and TyrB10. Whereas TrpG8 can act as a potential hydrogen-bond donor, the tyrosines can act as donors or acceptors. Ligand binding in heme-containing proteins is determined by a number of factors, including the nature and conformation of the distal residues and their capability to stabilize the heme-bound ligand via hydrogen-bonding and electrostatic interactions. Since both the RR Fe–OH⁻ and Fe–CN⁻ frequencies are very sensitive to the distal environment, detailed information on structural variations has been obtained. The hydroxyl ligand binds only the WT protein giving rise to two different conformers. In form 1 the anion is stabilized by H-bonds with TrpG8, TyrCD1 and a water molecule, in turn H-bonded to TyrB10. In form 2, H-bonding with TyrCD1 is mediated by a water molecule. Unlike the OH⁻ ligand, CN⁻ binds both WT and the triple mutant giving rise to two forms with similar spectroscopic characteristics. The overall results clearly indicate that H-bonding interactions both with distal residues and water molecules are important structural determinants in the active site of Tf-trHb. This article is part of a Special Issue entitled: Oxygen Binding and Sensing Proteins.     

Molecular recognition of Cullin3 by KCTDs: Insights from experimental and computational investigations

Recent investigations have highlighted a key role of the proteins of the KCTD (K-potassium channel tetramerization domain containing proteins) family in several fundamental biological processes. Despite the growing importance of KCTDs, our current understanding of their biophysical and structural properties is very limited. Biochemical characterizations of these proteins have shown that most of them act as substrate adaptor in E3 ligases during protein ubiquitination. Here we present a characterization of the KCTD5-Cullin3 interactions which are mediated by the KCTD5 BTB domain. Isothermal titration calorimetry experiments reveal that KCTD5 avidly binds the Cullin3 (Cul3). The complex presents a 5:5 stoichiometry and a dissociation constant of 59 nM. Molecular modeling and molecular dynamics simulations clearly indicate that the two proteins form a stable (KCTD5–Cul3)₅ pinwheel-shaped heterodecamer in which two distinct KCTD5 subunits cooperate in the binding of each cullin chain. Molecular dynamics simulations indicate that different types of interactions contribute to the stability of the assembly. Interestingly, residues involved in Cul3 recognitions are conserved in the KCTD5 orthologs and paralogs implicated in important biological processes. These residues are also rather well preserved in most of the other KCTD proteins. By using molecular modeling techniques, the entire ubiquitination system including the E3 ligase, the E2 conjugating enzyme and ubiquitin was generated. The analysis of the molecular architecture of this complex machinery provides insights into the ubiquitination processes which involve E3 ligases with a high structural complexity.     

X-ray and neutron small-angle scattering analysis of the complex formed by the Met receptor and the Listeria monocytogenes invasion protein InlB

The Listeria monocytogenes surface protein InlB binds to the extracellular domain of the human receptor tyrosine kinase Met, the product of the c-met proto-oncogene. InlB binding activates the Met receptor, leading to uptake of Listeria into normally nonphagocytic host cells. The N-terminal half of InlB (InlB₃₂₁) is sufficient for Met binding and activation. The complex between this Met-binding domain of InlB and various constructs of the Met ectodomain was characterized by size exclusion chromatography and dynamic light scattering, and structural models were built using small-angle X-ray scattering and small-angle neutron scattering. Although most receptor tyrosine kinase ligands induce receptor dimerization, InlB₃₂₁ consistently binds the Met ectodomain with a 1:1 stoichiometry. A construct comprising the Sema and PSI domains of Met, although sufficient to bind the physiological Met ligand hepatocyte growth factor/scatter factor, does not form a complex with InlB₃₂₁ in solution, highlighting the importance of Met Ig domains for InlB binding. Small-angle X-ray scattering and small-angle neutron scattering measurements of ligand and receptor, both free and in complex, reveal an elongated shape for the receptor. The four Ig domains form a bent, rather than a fully extended, conformation, and InlB₃₂₁ binds to Sema and the first Ig domain of Met, in agreement with the recent crystal structure of a smaller Met fragment in complex with InlB₃₂₁. These results call into question whether receptor dimerization is the basic underlying event in InlB₃₂₁-mediated Met activation and demonstrate differences in the mechanisms by which the physiological ligand hepatocyte growth factor/scatter factor and InlB₃₂₁ bind and activate the Met receptor.     

Open interface and large quaternary structure movements in 3D domain swapped proteins: insights from molecular dynamics simulations of the C-terminal swapped dimer of ribonuclease A

Bovine pancreatic ribonuclease (RNase A) forms two three-dimensional (3D) domain swapped dimers. Crystallographic investigations have revealed that these dimers display completely different quaternary structures: one dimer (N-dimer), which presents the swapping of the N-terminal helix, is characterized by a compact structure, whereas the other (C-dimer), which is stabilized by the exchange of the C-terminal end, shows a rather loose assembly of the two subunits. The dynamic properties of monomeric RNase A and of the N-dimer have been extensively characterized. Here, we report a molecular dynamics investigation carried out on the C-dimer. This computational experiment indicates that the quaternary structure of the C-dimer undergoes large fluctuations. These motions do not perturb the proper folding of the two subunits, which retain the dynamic properties of RNase A and the N-dimer. Indeed, the individual subunits of the C-dimer display the breathing motion of the beta-sheet structure, which is important for the enzymatic activity of pancreatic-like ribonucleases. In contrast to what has been observed for the N-dimer, the breathing motion of the two subunits of the C-dimer is not coupled. This finding suggests that the intersubunit communications in a 3D domain swapped dimer strongly rely on the extent of the interchain interface. Furthermore, the observation that the C-dimer is endowed with a high intrinsic flexibility holds interesting implications for the specific properties of 3D domain swapped dimers. Indeed, a survey of the quaternary structures of the other 3D domain swapped dimers shows that large variations are often observed when the structural determinations are conducted in different experimental conditions. The 3D domain swapping phenomenon coupled with the high flexibility of the quaternary structure may be relevant for protein-protein recognition, and in particular for the pathological aggregations.     

Structural characterization of intramolecular Hg(2+) transfer between flexibly linked domains of mercuric ion reductase

The enzyme mercuric ion reductase MerA is the central component of bacterial mercury resistance encoded by the mer operon. Many MerA proteins possess metallochaperone-like N-terminal domains (NmerA) that can transfer Hg²⁺ to the catalytic core domain (Core) for reduction to Hg⁰. These domains are tethered to the homodimeric Core by ∼ 30-residue linkers that are susceptible to proteolysis, the latter of which has prevented characterization of the interactions of NmerA and the Core in the full-length protein. Here, we report purification of homogeneous full-length MerA from the Tn21 mer operon using a fusion protein construct and combine small-angle X-ray scattering and small-angle neutron scattering with molecular dynamics simulation to characterize the structures of full-length wild-type and mutant MerA proteins that mimic the system before and during handoff of Hg²⁺ from NmerA to the Core. The radii of gyration, distance distribution functions, and Kratky plots derived from the small-angle X-ray scattering data are consistent with full-length MerA adopting elongated conformations as a result of flexibility in the linkers to the NmerA domains. The scattering profiles are best reproduced using an ensemble of linker conformations. This flexible attachment of NmerA may facilitate fast and efficient removal of Hg²⁺ from diverse protein substrates. Using a specific mutant of MerA allowed the formation of a metal-mediated interaction between NmerA and the Core and the determination of the position and relative orientation of NmerA to the Core during Hg²⁺ handoff.     

Molecular dynamics simulation of highly charged proteins: comparison of the particle-particle particle-mesh and reaction field methods for the calculation of electrostatic interactions

Molecular dynamics (MD) simulations of the activation domain of porcine procarboxypeptidase B (ADBp) were performed to examine the effect of using the particle-particle particle-mesh (P3M) or the reaction field (RF) method for calculating electrostatic interactions in simulations of highly charged proteins. Several structural, thermodynamic, and dynamic observables were derived from the MD trajectories, including estimated entropies and solvation free energies and essential dynamics (ED). The P3M method leads to slightly higher atomic positional fluctuations and deviations from the crystallographic structure, along with somewhat lower values of the total energy and solvation free energy. However, the ED analysis of the system leads to nearly identical results for both simulations. Because of the strong similarity between the results, both methods appear well suited for the simulation of highly charged globular proteins in explicit solvent. However, the lower computational demand of the RF method in the present implementation represents a clear advantage over the P3M method.     

All-Atom Structural Models for Complexes of Insulin-Like Growth Factors IGF1 and IGF2 with Their Cognate Receptor

Type 1 insulin-like growth factor receptor (IGF1R) is a membrane-spanning glycoprotein of the insulin receptor family that has been implicated in a variety of cancers. The key questions related to molecular mechanisms governing ligand recognition by IGF1R remain unanswered, partly due to the lack of testable structural models of apo or ligand-bound receptor complexes. Using a homology model of the IGF1R ectodomain IGF1RΔβ, we present the first experimentally consistent all-atom structural models of IGF1/IGF1RΔβ and IGF2/IGF1RΔβ complexes. Our explicit-solvent molecular dynamics (MD) simulation of apo-IGF1RΔβ shows that it displays asymmetric flexibility mechanisms that result in one of two binding pockets accessible to growth factors IGF1 and IGF2, as demonstrated via an MD-assisted Monte Carlo docking procedure. Our MD-generated ensemble of structures of apo and IGF1-bound IGF1RΔβ agrees reasonably well with published small-angle X-ray scattering data. We observe simultaneous contacts of each growth factor with sites 1 and 2 of IGF1R, suggesting cross-linking of receptor subunits. Our models provide direct evidence in favor of suggested electrostatic complementarity between the C-domain (IGF1) and the cysteine-rich domain (IGF1R). Our IGF1/IGF1RΔβ model provides structural bases for the observation that a single IGF1 molecule binds to IGF1RΔβ at low concentrations in small-angle X-ray scattering studies. We also suggest new possible structural bases for differences in the affinities of insulin, IGF1, and IGF2 for their noncognate receptors.     

Hydroxyl groups in the betabeta sandwich of metallo-beta-lactamases favor enzyme activity: Tyr218 and Ser262 pull down the lid

Metallo-beta-lactamases (MBLs) efficiently hydrolyze and thereby inactivate various beta-lactam antibiotics in clinical use. Their potential to evolve into more efficient enzymes threatens public health. Recently, we have identified the designed F218Y mutant of IMP-1 as an enzyme with superior catalytic efficiency compared to the wild-type. Thus, it may be found in clinical isolates in the future. In an effort to elucidate the molecular mechanisms involved in enhanced activity, we carried out molecular dynamics simulations of ten MBL variants in complex with a cefotaxime intermediate. The stability of these near-transition state enzyme-substrate intermediate complexes was modeled and compared to the experimental catalytic efficiencies k(cat)/K(M). For each of the ten complexes ten independent simulations were performed. In each simulation the temperature was gradually increased and determined upon breakdown of the complex. Rankings based on the experimental catalytic efficiencies and the data from computer simulations were in good agreement. From trajectory analysis of stable simulations, the combination of Tyr218 and Ser262 was found to lead to an altered hydrogen bonding network, which translates into a closing down movement of a beta-hairpin loop covering the active site. These observations may explain the significantly decreased K(M) and increased k(cat)/K(M) values of this variant toward all substrates recently tested in experiment. Previously, we have discovered that mutations G262S (yielding IMP-1) and G262A in IMP-6 stabilize the Zn(II) ligand His263 and thus the enzyme-substrate intermediate complex through a domino effect, which enhances conversion of drugs like ceftazidime, penicillins, and imipenem. Together, the domino effect and the altered beta-hairpin loop conformation explain how IMP-6 can evolve through mutations G262S and F218Y into an enzyme with up to one order of magnitude increased catalytic efficiencies toward these important antibiotics. Furthermore, the previously proposed binding of a third zinc ion close to the active site of IMP-6 mutant S121G was corroborated by our simulations.     

The C2B Domain Is the Primary Ca Sensor in DOC2B: A Structural and Functional Analysis

DOC2B (double-C₂ domain) protein is thought to be a high-affinity Ca^2 + sensor for spontaneous and asynchronous neurotransmitter release. To elucidate the molecular features underlying its physiological role, we determined the crystal structures of its isolated C2A and C2B domains and examined their Ca^2 +-binding properties. We further characterized the solution structure of the tandem domains (C2AB) using small-angle X-ray scattering. In parallel, we tested structure–function correlates with live cell imaging tools. We found that, despite striking structural similarity, C2B binds Ca^2 + with considerably higher affinity than C2A. The C2AB solution structure is best modeled as two domains with a highly flexible orientation and no difference in the presence or absence of Ca^2 +. In addition, kinetic studies of C2AB demonstrate that, in the presence of unilamellar vesicles, Ca^2 + binding is stabilized, as reflected by the ~ 10-fold slower rate of Ca^2 + dissociation than in the absence of vesicles. In cells, isolated C2B translocates to the plasma membrane (PM) with an EC₅₀ of 400 nM while the C2A does not translocate at submicromolar Ca^2 + concentrations, supporting the biochemical observations. Nevertheless, C2AB translocates to the PM with an ~ 2-fold lower EC₅₀ and to a greater extent than C2B. Our results, together with previous studies, reveal that the C2B is the primary Ca^2 + sensing unit in DOC2B, whereas C2A enhances the interaction of C2AB with the PM.     

Revisiting the bilayer structures of fluid phase phosphatidylglycerol lipids: Accounting for exchangeable hydrogens

We recently published two papers detailing the structures of fluid phase phosphatidylglycerol (PG) lipid bilayers (Ku?erka et al., 2012 J. Phys. Chem. B 116: 232–239; Pan et al., 2012 Biochim. Biophys. Acta Biomembr. 1818: 2135–2148), which were determined using the scattering density profile model. This hybrid experimental/computational technique utilizes molecular dynamics simulations to parse a lipid bilayer into components whose volume probabilities follow simple analytical functional forms. Given the appropriate scattering densities, these volume probabilities are then translated into neutron scattering length density (NSLD) and electron density (ED) profiles, which are used to jointly refine experimentally obtained small angle neutron and X-ray scattering data. However, accurate NSLD and ED profiles can only be obtained if the bilayer's chemical composition is known. Specifically, in the case of neutron scattering, the lipid's exchangeable hydrogens with aqueous D₂O must be accounted for, as they can have a measureable effect on the resultant lipid bilayer structures. This was not done in our above-mentioned papers. Here we report on the molecular structures of PG lipid bilayers by appropriately taking into account the exchangeable hydrogens. Analysis indicates that the temperature-averaged PG lipid areas decrease by 1.5 to 3.8 Å², depending on the lipid's acyl chain length and unsaturation, compared to PG areas when hydrogen exchange was not taken into account.