Mathematics and Molecular NeurobiologyNathan A. Baker, Kaihsu Tai, Richard Henchman, David Sept, Adrian Elcock, Michael Holst and J. Andrew McCammonIn "Computational Methods for Macromolecular Modeling: Challenges and Applications," Tamar Schlick, Hin Hark Gan, Eds., Springer-Verlag, Vol. 24, pp. 31-60 (2002)
Advances in mathematics and computer technology, together with advances
in structural biology, are opening the way to detailed modeling of
biology at the molecular and cellular levels. One objective of such
studies is the development of a more complete understanding of
biological systems, including the emergence of behavior at the cellular
level from that at the molecular level. Another objective is the
development of more sophisticated models for structure-aided discovery
of new pharmaceuticals.
Binding of Aminoglycoside Antibiotics to the Small Ribosomal Subunit: A Continuum Electrostatics InvestigationChiansan Ma, Nathan A. Baker, Simpson Joseph and J. Andrew McCammonJournal of the American Chemical Society, Vol. 124, Issue 7, pp. 1438-1442 (2002) [PubMed 11841313]The binding of paromomycin and similar antibiotics to the small (30S)
ribosomal subunit has been studied using continuum electrostatics
methods. Crystallographic information from a complex of paromomycin with
the 30S subunit was used as a framework to develop structures of similar
antibiotics in the same ribosomal binding site. Total binding energies
were calculated from electrostatic properties obtained by by solution of
the Poisson-Boltzmann equation combined with a surface area-dependent
apolar term. These computed results showed good correlation with
experimental data. Additionally, calculation of the ribosomal
electrostatic potential in the paromomycin binding site provided insight
into the electrostatic mechanisms for aminoglycoside binding and clues
for the rational design of more effective antibiotics.
Molecular Dynamics of AcetylcholinesteraseTongye Shen, Kaihsu Tai, Richard H. Henchman and J. Andrew McCammonAccounts of Chemical Research, Vol. 35, Issue 6, pp. 332-340 (2002) [PubMed 12069617]Molecular dynamics simulations are leading to a deeper understanding of
the activity of the enzyme acetylcholinesterase. Simulations have shown
how breathing motions in the enzyme facilitate the displacement of
substrate from the surface of the enzyme to the buried active site. The
most recent work points to the complex and spatially extensive nature of
such motions, and suggests possible modes of regulation of the activity
of the enzyme.
Extracting Hydration Sites around Proteins from Explicit Water SimulationsRichard H. Henchman and J. Andrew McCammonJournal of Computational Chemistry, Vol. 23, Issue 9, pp. 861-869 (2002) [PubMed 11984847]Two new methods are assessed for determining the location of hydration
sites around proteins from computer simulation. Current methods extract
hydration sites from peaks in the water density constructed in the
protein frame. However, the dynamic nature of the water molecules, the
nearby protein residues and the protein reference frame as a whole tend
to smear out the water density, making it more difficult to resolve
sites. Two techniques are introduced to better resolve the water
density. The first is to construct the water density from the time
averaged position of each water molecule in the protein frame while the
water remains within a given distance of this averaged position. The
second technique is to construct the water density from the time
averaged position of each water in the reference frame only of the
nearby residues. Criteria for determining hydration sites from the water
density are examined. Both techniques are found to significantly improve
the detail in the water density and the number of hydration sites
detected.
Enlarging the Landscape: Editorial OverviewJ. Andrew McCammon and Peter G. WolynesCurrent Opinion in Structural Biology, Vol. 12, Issue 2, pp. 143-145 (2002)
Molecular theory and simulation are vibrant and essential elements of
modern biology. Indeed, theory and simulation contribute to our
understanding of structure-activity correlations with ever-increasing
detail and scope. The detail is illustrated by recent work on the
evolution of the excited electronic states in the green fluorescent
protein and other macromolecules endowed with chromophores. And the
range is illustrated by simulations of the activity of such
supramolecular structures as ribosomes and the cytoskeleton.
Entropy Loss of Hydroxyl Groups of Balanol upon Binding to Protein Kinase AGergely Gidofalvi, Chung F. Wong and J. Andrew McCammonJournal of Chemical Education, Vol. 79, No. 9, pp. 1122-1126 (2002)
This article describes a short project for an undergraduate to learn
several techniques for computer-aided drug design. The project involves
estimating the loss of the rotational entropy of the hydroxyl groups of
balanol upon its binding to the enzyme protein kinase A (PKA), as the
entropy loss can significantly influence PKA-balanol binding affinity.
This work employs semi-empirical quantum mechanical techniques for
estimating the potential energy curves for the rotation of the hydroxyl
groups of balanol in vacuum and in PKA, and solves the Poisson equation
to correct the potential energy curves for hydration effects.
Statistical mechanical principles were then applied to estimate the
desired entropy loss from the potential energy curves. The analysis
focuses on examining the importance of hydration effects on influencing
the rotational preference of the hydroxyl groups and the significance of
the rotational entropy in determining binding affinity.
Properties of Water Molecules in the Active Site Gorge of Acetylcholinesterase from Computer SimulationRichard H. Henchman, Kaihsu Tai, Tongye Shen and J. Andrew McCammonBiophysical Journal, Vol. 82, No. 5, pp. 2671-2682 (2002) [PubMed 11964254]A 10-ns trajectory from a molecular dynamics simulation is used to
examine the structure and dynamics of water in the active site gorge of
acetylcholinesterase to determine what influence water may have on its
function. While the confining nature of the deep active site gorge slows
down and structures water significantly compared to bulk water, water in
the gorge is found to display a number of properties that may aid ligand
entry and binding. These properties include fluctuations in the
population of gorge waters, moderate disorder and mobility of water in
the middle and entrance to the gorge, reduced water hydrogen-bonding
ability, and transient cavities in the gorge.
Mechanism of Acetylcholinesterase Inhibition by Fasciculin: A 5-ns Molecular Dynamics SimulationKaihsu Tai, Tongye Shen, Richard H. Henchman, Yves Bourne, Pascale Marchot and J. Andrew McCammonJournal of the American Chemical Society, Vol. 124, No. 21, pp. 6153-6161 (2002) [PubMed 12022850]Our previous molecular dynamics simulation (10 ns) of mouse
acetylcholinesterase (EC 3.1.1.7) revealed complex fluctuations of the
enzyme active site gorge. Now we report a 5 ns simulation of
acetylcholinesterase complexed with fasciculin 2. Fasciculin 2 binds to
the gorge entrance of acetylcholinesterase with excellent
complementarity and many polar and hydrophobic interactions. In this
simulation of the protein-protein complex, where fasciculin 2 appears to
sterically block access of ligands to the gorge, again we observe a
two-peaked probability distribution of the gorge width. When fasciculin
is present, the gorge width distribution is altered such that the gorge
is more likely to be narrow. Moreover, there are large increases in the
opening of the alternative passages, namely the side door (near Thr 75)
and the back door (near Tyr 449). Finally, the catalytic triad
arrangement in the acetylcholinesterase active site is disrupted with
fasciculin bound. These data support that, in addition to the steric
obstruction seen in the crystal structure, fasciculin may inhibit
acetylcholinesterase by combined allosteric and dynamical means.
Additional data from these simulations can be found at
http://mccammon.ucsd.edu/" target="_blank"
class="ref">http://mccammon.ucsd.edu/.
Computational Drug Design Accommodating Receptor Flexibility: The Relaxed Complex SchemeJung-Hsin Lin, Alexander L. Perryman, Julie R. Schames and J. Andrew McCammonJournal of the American Chemical Society, Vol. 124, No. 20, pp. 5632-5633 (2002) [PubMed 12010024]A novel computational methodology for drug design that accommodates
receptor flexibility is described. This "relaxed-complex" method
recognizes that ligand may bind to conformations that occur only rarely
in the dynamics of the receptor. We have shown that the ligand-enzyme
binding modes are very sensitive to the enzyme conformations, and our
approach is capable of finding the best ligand-enzyme complexes. This
new method serves as the computational analog of the experimental "SAR
by NMR" and "tether" methods, which permit a building block approach for
constructing a very potent drug.
Bridging the Implicit and Explicit Solvent Approaches for Membrane ElectrostaticsJung-Hsin Lin, Nathan A. Baker and J. Andrew McCammonBiophysical Journal, Vol. 83, No. 3, pp. 1374-1379 (2002) [PubMed 12202363]Conformations of a zwitterionic bilayer were sampled from a molecular
dynamics (MD) simulation and their electrostatic properties analyzed by
solution of the Poisson equation. These traditionally implicit
electrostatic calculations were performed in the presence of varying
amounts of explicit solvent to assess the magnitude of error introduced
by a uniform dielectric description of water surrounding the bilayer. It
was observed that, while membrane dipole potential calculations in the
presence of explicit water were significantly different than wholly
implicit solvent calculations, the calculated dipole potential converged
to a reasonable value when four or more hydration layers were included
explicitly.
Molecular Dynamics Simulations of Macromolecules: A PerspectiveMartin Karplus and J. Andrew McCammonNature Structural Biology, Vol. 9, pp. 646-652 (2002) [PubMed 12198485]It is now twenty-five years since the first molecular dynamics
simulation of a macromolecule of biological interest was published. The
simulation concerned the bovine pancreatic trypsin inhibitor (BPTI),
which has served as the "hydrogen molecule" of protein dynamics because
of its small size, high stability and relatively accurate x-ray
structure, available in 1975; interestingly, its physiological functions
remain unknown. Although this simulation was done in vacuum with a crude
molecular mechanism potential and lasted for only 9.2 ps, the results
were instrumental in replacing our view of proteins as relatively rigid
structures (In 1981, Sir D.L. Phillips commented: "Brass models of DNA
and a variety of proteins dominated the scene and much of the
thinking".) with the realization that they were dynamic systems, whose
internal motions play a functional role. Of course, there were already
experimental data, such as the hydrogen exchange experiments of
Linderstrom-Lang and his coworkers pointing in this direction and the
Feynman Lectures on Physics, published 14 years earlier contained the
prescient sentence, "Certainly no subject or field is making more
progress on so many fronts at the present moment than biology, and if we
were to name the most powerful assumption of all, which leads one on and
on in an attempt to understand life, it is that
all things are made
of atoms italics in original, and that
everything that living
things do can be understood in terms of the jigglings and wigglings of
atoms. italics added".
Unfolding Proteins under External Forces: A Solvable Model under the Self-consistent Pair Contact Probability ApproximationTongye Shen, Lawrence S. Canino and J. Andrew McCammonPhysical Review Letters, Vol. 89, article 068103, 4 pages (2002) [PubMed 12190614]We extend a model of Micheletti, et. al.
http://dx.doi.org/10.1103/PhysRevLett.87.088102" target="_blank"
class="ref">Phys. Rev. Lett, 87, 088102 used to study protein
conformations in equilibrium to the case in which there is an external
force field. Under the self-consistent pair contact probability
approximation, we show that this residue-level resolution model can
still be solved when a constant, external force is used to pull at the
termini of the protein. We implement the algorithm using heterogeneous
contact parameters and study the force-induced unfolding of a helical
segment from the protein GART and of the beta-stranded Ig and Fn-3
domains from the protein titin. The results are qualitatively consistent
with the results from more expensive, atomistic dynamics simulation.
Despite the mean-field-like approach, we observed a sharp and
cooperative unfolding transition within this model.
Studying Enzyme Binding Specificity in Acetylcholinesterase using a Combined Molecular Dynamics and Multiple Docking ApproachJeremy Kua, Yingkai Zhang and J. Andrew McCammonJournal of the American Chemical Society, Vol. 124, No. 28, pp. 8260-8267 (2002) [PubMed 12105904]A combined molecular dynamics simulation and multiple ligand docking
approach is applied to study the binding specificity of
acetylcholinesterase (AChE) with its natural substrate acetylcholine
(ACh), a family of substrate analogs and choline. Calculated docking
energies are well correlated to experimental
kcat/KM values, as well as to experimental
binding affinities of a related series of TMTFA inhibitors. The
"esteratic" and "anionic" subsites are found to act together to achieve
substrate binding specificity. We find that the presence of ACh in the
active site of AChE not only stabilizes the setup of the catalytic
triad, but also tightens both subsites to achieve better binding. The
docking energy gained from this induced fit is 0.7 kcal/mol for ACh. For
the binding of the substrate tail group to the anionic subsite, both the
size and the positive charge of the tail group are important. The
removal of the positive charge leads to a weaker binding of 1.1 kcal/mol
loss in docking energy. Substituting each tail methyl group with
hydrogen results in both an incremental loss in docking energy, and also
a decrease in the percentage of structures docked in the active site
correctly set up for catalysis.
Thalassiolins A-C: New Marine-derived Inhibitors of HIV cDNA IntegraseDavid C. Rowley, Mark S.T. Hansen, Denise Rhodes, Christoph A. Sotriffer, Haihong Ni, J. Andrew McCammon, Frederic D. Bushman and William FenicalBioorganic & Medicinal Chemistry, Vol. 10, Issue 11, pp. 3619-3625 (2002) [PubMed 12213478]Replication of human immunodeficiency virus (HIV) requires the
integration of viral cDNA into the host genome, a process mediated by
the viral enzyme integrase. We describe a new series of HIV integrase
inhibitors named thalassiolins A-C (1-3) isolated from the Caribbean sea
grass
Thalassia testudinum. The thalassiolins are distinguished
from other flavones previously studied by the substitution of a sulfated
beta-D-glucose at the 7-position, a substituent that imparts increased
potency against integrase in biochemical assays as well as antiviral
activity against HIV in cell culture. Thalassiolin A(1), the most active
of these molecules, displays
in vitro inhibition of the integrase
catalyzed strand transfer reaction at 0.4 mu-M and an antiviral
IC
50 of 30 mu-M. Molecular modeling studies indicate a
favorable binding mode is possible at the catalytic core domain of HIV-1
integrase.
Structural and Dynamical Properties of Water Around AcetylcholinesteraseRichard H. Henchman and J. Andrew McCammonProtein Science, Vol. 11, Issue 9, pp. 2080-2090 (2002) [PubMed 12192064]Structural and dynamic properties of water molecules around
acetylcholinesterase are examined from a 10 ns molecular dynamics
simulation to help understand how the protein alters water properties.
Water structure is broken down into hydration sites constructed from the
water density less than 3.6 Å from the protein surface. These
sites are characterized according to occupancy, number of water
neighbors, hydrogen bonds, dipole moment and residence time. The site
description provides a convenient means to describe the extend and
localization of these properties. Determining the network of paths that
waters follow from site to site and measuring the rate of flow of waters
from the sites to the bulk make it possible to quantitatively study the
time scales and paths that water molecules follow as they move around
the protein.
Role of the Catalytic Triad and Oxyanion Hole in Acetylcholinesterase Catalysis: An ab inito QM/MM StudyYingkai Zhang, J. Kua and J.A. McCammonJournal of the American Chemical Society, Vol. 124, No. 35, pp. 10572-10577 (2002) [PubMed 12197759]The initial step of the acylation reaction catalyzed by
acetylcholinesterase (AChE) has been studied by a combined ab inito
quantum mechanical/molecular mechanical (QM/MM) approach. The reaction
proceeds through the nucleophilic addition of the Ser203 O to the
carbonyl C of acetylcholine, and the reaction is facilitated by
simultaneous proton transfer from Ser203 to His447. The calculated
potential energy barrier at the MP2(6-31+G*) QM/MM level is 10.5
kcal/mol, consistent with the experimental reaction rate. The third
residue of the catalytic triad, Glu334, is found to be essential in
stabilizing the transition state through electrostatic interactions. The
oxyanion hole, formed by peptidic NH groups from Gly121, Gly122, and
Ala204, is also found to play an important role in catalysis. Our
calculations indicate that, in the AChE-ACh Michaelis complex, only two
hydrogen bonds are formed between the carbonyl oxygen of ACh and the
peptidic NH groups of Gly121 and Gly122. As the reaction proceeds, the
distance between the carbonyl oxygen of ACh and NH group of Ala204
becomes smaller, and the third hydrogen bond is formed both in the
transition state and in the tetrahedral intermediate.
AutoDocking Dinucleotides to the HIV-1 Integrase Core Domain: Exploring Possible Binding Sites for Viral and Genomic DNAAlexander L. Perryman and J. Andrew McCammonJournal of Medicinal Chemistry, Vol. 45, No. 26, pp. 5624-5627 (2002) [PubMed 12477345]To understand the binding of both viral and human DNA to HIV-1
Integrase, fully flexible dinucleotides were docked onto the core domain
of Integrase. AutoDocking did identify sites on Integrase where
favorable interactions with nucleotides can occur, and those sites were
in agreement with recently published protein fingerprinting data. By
analyzing the phosphates of the docked dinucleotides, we developed a
model indicating where the viral cDNA and human DNA bind to the
Integrase core domain.
Changes in Flexibility Upon Binding: Application of the Self-Consistent Pair Contact Probability Method to Protein-Protein InteractionsLawrence S. Canino, Tongye Shen and J. Andrew McCammonJournal of Chemical Physics, Vol. 117, Issue 21, pp. 9927-9933 (2002)
We extend the self-consistent pair contact probability (SCPCP) method to
the evaluation of the partition function for a protein complex at
thermodynamic equilibrium. Specifically, we adapt the method for
multi-chain models and introduce a parameterization for amino
acid-specific pairwise interactions. This method is simular to the
Gaussian network model (GNM) but allows for the adjusting of the
strengths of native state contacts. The method is first validated on a
high resolution x-ray crystal structure of bovine Pancreatic
Phospholipase A2 comparing calculated B-factors with reported values. We
then examine binding-induced changes in flexibility in protein-protein
complexes, comparing computed results with those obtained from x-ray
crystal structures and molecular dynamics simulations. In particular, we
focus on the mouse acetylcholinesterase:fasciculin II and the human
α-thrombin:thrombomodulin complexes.