Donald Bashford

Exploring protein native states and large‐scale conformational changes with a modified generalized born model

Implicit solvation models provide, for many applications, a reasonably accurate and computationally effective way to describe the electrostatics of aqueous solvation. Here, a popular analytical Generalized Born (GB) solvation model is modified to improve its accuracy in calculating the solvent polarization part of free energy changes in large‐scale conformational transitions, such as protein folding. In contrast to an earlier GB model (implemented in the AMBER‐6 program), the improved version does not overstabilize the native structures relative to the finite‐difference Poisson–Boltzmann continuum treatment. In addition to improving the energy balance between folded and unfolded conformers, the algorithm (available in the AMBER‐7 and NAB molecular modeling packages) is shown to perform well in more than 50 ns of native‐state molecular dynamics (MD) simulations of thioredoxin, protein‐A, and ubiquitin, as well as in a simulation of Barnase/Barstar complex formation. For thioredoxin, various combinations of input parameters have been explored, such as the underlying gas‐phase force fields and the atomic radii. The best performance is achieved with a previously proposed modification to the torsional potential in the Amber ff99 force field, which yields stable native trajectories for all of the tested proteins, with backbone root‐mean‐square deviations from the native structures being ∼1.5 Å after 6 ns of simulation time. The structure of Barnase/Barstar complex is regenerated, starting from an unbound state, to within 1.9 Å relative to the crystal structure of the complex. Proteins 2004;55:000–000.

Effective Born radii in the generalized Born approximation: The importance of being perfect

Generalized Born (GB) models provide, for many applications, an accurate and computationally facile estimate of the electrostatic contribution to aqueous solvation. The GB models involve two main types of approximations relative to the Poisson equation (PE) theory on which they are based. First, the self‐energy contributions of individual atoms are estimated and expressed as “effective Born radii.” Next, the atom‐pair contributions are estimated by an analytical function fGB that depends upon the effective Born radii and interatomic distance of the atom pairs. Here, the relative impacts of these approximations are investigated by calculating “perfect” effective Born radii from PE theory, and enquiring as to how well the atom‐pairwise energy terms from a GB model using these perfect radii in the standard fGB function duplicate the equivalent terms from PE theory. In tests on several biological macromolecules, the use of these perfect radii greatly increases the accuracy of the atom‐pair terms; that is, the standard form of fGB performs quite well. The remaining small error has a systematic and a random component. The latter cannot be removed without significantly increasing the complexity of the GB model, but an alternative choice of fGB can reduce the systematic part. A molecular dynamics simulation using a perfect‐radii GB model compares favorably with simulations using conventional GB, even though the radii remain fixed in the former. These results quantify, for the GB field, the importance of getting the effective Born radii right; indeed, with perfect radii, the GB model gives a very good approximation to the underlying PE theory for a variety of biomacromolecular types and conformations.

Identification and characterization of the first small molecule inhibitor of MDMX.

The p53 pathway is disrupted in virtually every human tumor. In ∼50% of human cancers, the p53 gene is mutated, and in the remaining cancers, the pathway is dysregulated by genetic lesions in other genes that modulate the p53 pathway. One common mechanism for inactivation of the p53 pathway in tumors that express wild-type p53 is increased expression of MDM2 or MDMX. MDM2 and MDMX bind p53 and inhibit its function by distinct nonredundant mechanisms. Small molecule inhibitors and small peptides have been developed that bind MDM2 in the p53-binding pocket and displace the p53 protein, leading to p53-mediated cell cycle exit and apoptosis. To date, peptide inhibitors of MDMX have been developed, but no small molecule inhibitors have been reported. We have developed biochemical and cell-based assays for high throughput screening of chemical libraries to identify MDMX inhibitors and identified the first MDMX inhibitor SJ-172550. This compound binds reversibly to MDMX and effectively kills retinoblastoma cells in which the expression of MDMX is amplified. The effect of SJ-172550 is additive when combined with an MDM2 inhibitor. Results from a series of biochemical and structural modeling studies suggest that SJ-172550 binds the p53-binding pocket of MDMX, thereby displacing p53. This lead compound is a useful chemical scaffold for further optimization of MDMX inhibitors that may eventually be used to treat pediatric cancers and various adult tumors that overexpress MDMX or have similar genetic lesions. When combined with selective MDM2 inhibitors, SJ-172550 may also be useful for treating tumors that express wild-type p53.

An object-oriented programming suite for electrostatic effects in biological molecules An experience report on the MEAD project

We have developed a set of object-oriented classes and programs in C++ that implement molecular electrostatic models that can be described by the term, Macroscopic Electrostatics with Atomic Detail (MEAD). In the course of developing the MEAD suite, we have shifted from a class hierarchy rooted in atoms and molecules, to a system in which the top-level classes are the electrostatic potential and the entities that determine the potential in the equations of electrostatics: the charge distribution, the dielectric environment and the electrolyte environment. Atoms and molecules are then seen as objects giving rise to, or occurring as subclasses of, charge distributions, dielectric environments, etc. This shift in focus from the physical objects (molecules) to the more abstract objects that appear in the underlying physics has facilitated the development of alternative approximation schemes and numerical methods through subclassing. It also provides a natural way of writing high level programs in terms of potentials and distributions. Some of the newer elements of C++, such as templates and RTTI, have proven useful to solve multi-method and default method problems. MEAD is distributed as free software.

Catalysis and sulfa drug resistance in dihydropteroate synthase.

Sulfas Crystal View The sulfonamide antibiotics (sulfa drugs) have been used to treat infections for over 70 years; however, emerging resistance has eroded their clinical utility. Sulfa drugs target dihydropteroate synthase, a key enzyme in the bacterial folate pathway. By performing the reaction in the crystalline form of the enzyme, Yun et al. (p. 1110) have characterized the key structural intermediates. In combining structural data with theoretical and mutagenesis studies, they propose a detailed mechanism for dihydropteroate synthase catalysis. By resolving this structure with a sulfa drug bound to the enzyme, they showed how inhibition occurred and indicated how resistance could emerge. Structures of a target enzyme in the bacteria that cause anthrax and bubonic plague may lead to effective drugs. The sulfonamide antibiotics inhibit dihydropteroate synthase (DHPS), a key enzyme in the folate pathway of bacteria and primitive eukaryotes. However, resistance mutations have severely compromised the usefulness of these drugs. We report structural, computational, and mutagenesis studies on the catalytic and resistance mechanisms of DHPS. By performing the enzyme-catalyzed reaction in crystalline DHPS, we have structurally characterized key intermediates along the reaction pathway. Results support an SN1 reaction mechanism via formation of a novel cationic pterin intermediate. We also show that two conserved loops generate a substructure during catalysis that creates a specific binding pocket for p-aminobenzoic acid, one of the two DHPS substrates. This substructure, together with the pterin-binding pocket, explains the roles of the conserved active-site residues and reveals how sulfonamide resistance arises.

Proton Affinity Changes Driving Unidirectional Proton Transport in the Bacteriorhodopsin Photocycle

Bacteriorhodopsin is the smallest autonomous light-driven proton pump. Proposals as to how it achieves the directionality of its trans-membrane proton transport fall into two categories: accessibility-switch models in which proton transfer pathways in different parts of the molecule are opened and closed during the photocycle, and affinity-switch models, which focus on changes in proton affinity of groups along the transport chain during the photocycle. Using newly available structural data, and adapting current methods of protein protonation-state prediction to the non-equilibrium case, we have calculated the relative free energies of protonation microstates of groups on the transport chain during key conformational states of the photocycle. Proton flow is modeled using accessibility limitations that do not change during the photocycle. The results show that changes in affinity (microstate energy) calculable from the structural models are sufficient to drive unidirectional proton transport without invoking an accessibility switch. Modeling studies for the N state relative to late M suggest that small structural re-arrangements in the cytoplasmic side may be enough to produce the crucial affinity change of Asp96 during N that allows it to participate in the reprotonation of the Schiff base from the cytoplasmic side. Methodologically, the work represents a conceptual advance compared to the usual calculations of pK(a) using macroscopic electrostatic models. We operate with collective states of protonation involving all key groups, rather than the individual-group pK(a) values traditionally used. When combined with state-to-state transition rules based on accessibility considerations, a model for non-equilibrium proton flow is obtained. Such methods should also be applicable to other active proton-transport systems.

Structural details, pathways, and energetics of unfolding apomyoglobin.

Protein folding is often difficult to characterize experimentally because of the transience of intermediate states, and the complexity of the protein-solvent system. Atomistic simulations, which could provide more detailed information, have had to employ highly simplified models or high temperatures, to cope with the long time scales of unfolding; direct simulation of folding is even more problematic. We report a fully atomistic simulation of the acid-induced unfolding of apomyoglobin in which the protonation of acidic side-chains to simulate low pH is sufficient to induce unfolding at room temperature with no added biasing forces or other unusual conditions; and the trajectory is validated by comparison to experimental characterization of intermediate states. Novel insights provided by their analysis include: characterization of a dry swollen globule state forming a barrier to initial unfolding or final folding; observation of cooperativity in secondary and tertiary structure formation and its explanation in terms of dielectric environments; and structural details of the intermediate and the completely unfolded states. These insights involve time scales and levels of structural detail that are presently beyond the range of experiment, but come within reach through the simulation methods described here. An implicit solvation model is used to analyze the energetics of protein folding at various pH and ionic strength values, and a reasonable estimate of folding free energy is obtained. Electrostatic interactions are found to disfavor folding.

Implicit Solvent Electrostatics in Biomolecular Simulation

We give an overview of how implicit solvent models are currently used in protein simulations. The emphasis is on numerical algorithms and approximations: since even folded proteins sample many distinct configurations, it is of considerable importance to be both accurate and efficient in estimating the energetic consequences of this dynamical behavior. Particular attention is paid to calculations of pH-dependent behavior, as a paradigm for the analysis of electrostatic interactions in complex systems.

Multiple-site ligand binding to flexible macromolecules: Separation of global and local conformational change and an iterative mobile clustering approach

This article concerns the calculation of equilibria of ligand binding to multiple sites in macromolecules in the presence of conformational flexibility and conformation‐dependent interaction among the sites. A formulation of this problem is presented in which global conformational changes are distinguished from conformational changes that are confined to “locally flexible regions.” The formalism is quite general in that ligands of different types, multivalent binding sites, tautomeric binding sites, and sites that bind more than one type of ligand can be accommodated. Strictly speaking, the separation of the conformational problem into global and local parts does not impose any loss of generality, although in practice it is necessary to restrict the number of global and local conformers. Because of the combinatorics of binding and conformational states, the computational complexity of a problem having only local conformational flexibility grows exponentially with the number of sites and the number of locally flexible regions. An iterative mobile clustering method for cutting off this exponential growth and obtaining approximate solutions with low computational cost is presented and tested. In this method, a binding site is selected, and a “cluster” of strongly interacting sites is set up around it; within the cluster, the binding and conformational states are fully enumerated, whereas the influences of sites outside the cluster on the sites inside are treated by a mean field approximation. The procedure then moves to the next site around which another (possibly overlapping) cluster is formed and the calculation is repeated. The procedure iterates through the list of sites in this way, using the results of previous iterations for the mean‐field terms of current iterations until a convergence criterion is met. The method is tested on a large set of randomly generated problems of varying size, whose geometries are chosen to have protein‐like statistical properties. It is found that the method is accurate and rapid with the computational cost scaling linearly to quadratically with the number of sites, except for a minority of cases in which large clusters occur by chance. The new method is more accurate than a Monte Carlo method, and may be faster or slower depending on the clustering criteria and details of the macromolecule. ©1999 John Wiley & Sons, Inc. J Comput Chem 20: 1091–1111, 1999

HLA-DRB1*07:01 is associated with a higher risk of asparaginase allergies

Asparaginase is a therapeutic enzyme used to treat leukemia and lymphoma, with immune responses resulting in suboptimal drug exposure and a greater risk of relapse. To elucidate whether there is a genetic component to the mechanism of asparaginase-induced immune responses, we imputed human leukocyte antigen (HLA) alleles in patients of European ancestry enrolled on leukemia trials at St. Jude Childrens Research Hospital (n = 541) and the Childrens Oncology Group (n = 1329). We identified a higher incidence of hypersensitivity and anti-asparaginase antibodies in patients with HLA-DRB1*07:01 alleles (P = 7.5 × 10(-5), odds ratio [OR] = 1.64; P = 1.4 × 10(-5), OR = 2.92, respectively). Structural analysis revealed that high-risk amino acids were located within the binding pocket of the HLA protein, possibly affecting the interaction between asparaginase epitopes and the HLA-DRB1 protein. Using a sequence-based consensus approach, we predicted the binding affinity of HLA-DRB1 alleles for asparaginase epitopes, and patients whose HLA genetics predicted high-affinity binding had more allergy (P = 3.3 × 10(-4), OR = 1.38). Our results suggest a mechanism of allergy whereby HLA-DRB1 alleles that confer high-affinity binding to asparaginase epitopes lead to a higher frequency of reactions. These trials were registered at www.clinicaltrials.gov as NCT00137111, NCT00549848, NCT00005603, and NCT00075725.