- Email: [email protected]
Structural transitions in neutral and charged proteins in Vacuo
ty of molecular behavior. Computer simulations of thermodynamic and kinetic behavior can be used to characterize the nature of this selectivity and aid in the design of drugs, novel antibodies, or enzymes. New computer architecture as available at national supercomputer facilities such as the SDSC can help in this effort. One example of the ways in which computer simulations can be used to understand properties of biological molecules is Dr. McCammon's work on acetylcholinesterase. Molecular dynamics of acetylcholinesterase show a stability of the negative electrostatic potential around the active sites that may contribute to the speed with which the enzyme works by steering the positively charged substrate to the appropriate binding site. The award for best paper was given to Izik Pe'er (Tel Aviv University, Israel) for "Spectrum Alignment: Efficient Resequencing by Hybridization." This method is universally applicable to standard microarray chips and removes the requirement for the redesign of chips for each individual DNA sequence to be analyzed. Mr. Pe'er's method uses spectrum data and DNA homology information to computationally construct a target sequence. The method allows for insertions, deletions, hybridization errors, and the use of a profile or Hidden Markov model. Tod Klingler (Prospect Genomics, Inc., San Francisco, CA, USA) showed a database of more than 230,000 modeled 3D protein structures generated by Andrej ~ali (Rockefeller University, New York, NY, USA) using his Modeler and ModPipe programs. The Modeler program uses protein homology modelling, also called comparative modelling, to produce theoretical structures based on sequence homology to experimentally-determined protein structures. Prospect Genomics has exclusively licensed the database from Rockefeller University. A copy of the proceedings is available from AAAI Press, 445 Burgess Drive, Menlo Park, CA 94025. The next ISMB meeting will be held July 21-25, 2001, in Copenhagen Denmark.
Links Conference website: http://ismb2OOO.sdsc.edu J. Andrew McCammon: http://mccammon.ucsd.edu Harold Scheraga: http://www.chem.cornell.edu/ department/Faculty/Scheraga/scheraga, html San Diego Supercomputer Center: http://www.sdsc.edu/
COMPANY PROFILE GeneFormatics Inc. ow that the genome projects have put masses of genes N in the hands of researchers, drug discovery and agbio companies are frantically hunting for targets. GeneFormatics Inc. plans to be a leader in the search for targets by predicting protein structure and function. Company founders are Jacquelyn Fetrow (formerly of The Scripps Research Institute), Adam Godzik (Burnham Institute, La Jolla, CA), and Jeffrey Skolnick and Andrzej Kolinski (Donald Danforth Plant Science Center, St. Louis, MO).
548
. J. Mol. Graphics Mod., 2000, Vol. 18, August-October
GeneFormatics' proprietary algorithms use protein structural information to predict function. The method can predict a protein's function starting from just a protein sequence, using a function identification algorithm with a predicted structure. Prediction is done using a threading algorithm, which aligns the sequences to the best matching structure in a structural database and extends sequence analysis beyond local sequence identity. After a structural model is obtained, the active site is identified using a "fuzzy functional form" (FFF), a three-dimensional descriptor of the protein's active s~te. The method is rapid and can be applied to genome sequence databases. The founders have published a demonstration of the method' in which all the open reading frames of the Escherichia coli genome were screened for the thiol-disulfide oxidoreductase activity of the glutaredoxin/thioredoxin protein family. They were able to identify sequences known to have the activity and to identify other proteins with the activity that were not previously known. The FFF method can distinguish proteins with similar active sites from proteins with similar topological folds. The company claims that using the FFF method they can identify 10 to 30 percent more hits and fewer false positives than a sequence motif analysis. GeneFormatics offers structure prediction and protein functional annotation services for pharmaceutical or biotechnology companies, analyzing client's proprietary sequence data. In the near term, GeneFormatics plans to sell subscriptions to its own database of protein structures and functions. In the long term, GeneFormatics will build its own intellectual property, based on structures and protein functions it is able to predict. For more information, contact GeneFormatics, 5830 Oberlin Drive, San Diego, CA 92130 USA. Tel. 1-858-450-3331. http://www.geneformatics.com
Reference I. Fetrow, J.S., Godzik, A., Skolnick, J.J. Functional analysis of the Escherichia coli genome using the sequence-to-structureto-function paradigm: identification of proteins exhibiting the glutaredoxin/thioredoxin disulfide oxidoreductase activity. J. Mol. Biol. 1998, 282:703-71 I.
MEETING ABSTRACTS
Protein Flexibility and Folding Traverse City, Michigan, USA August 13-17, 2000
Structural Transitions in Neutral and Charged Proteins in Vacuo Gustavo A. Arteca, Departement de Chimie et Biochimie, Laurentian University, Ramsey Lake Road, Sudbury, Ontario P3E 2C6, Canada In vacuo proteins provide the simplest laboratory to study
the specific roles of sequence, initial configuration, and environment on protein folding. From the experimental viewpoint, these systems are now beginning to be characterized at low resolution. Molecular dynamics (MD) simulations, in combination with tools for protein shape analysis, can complement these experiments and provide further insights on folding-unfolding transitions. Here, we illustrate some aspects of this issue by using an ensemble of MD trajectories for hen egg-white lysozyme. In lysozyme ions, unfolding can be triggered by Coulombic repulsion. In neutral lysozyme, unfolding can be induced by centrifugal forces and by weakening the monomer-monomer interaction. In both cases, the resulting unfolded transients act as initial configurations for relaxation dynamics. The trajectories are analyzed in terms of two descriptors of backbone shape: the anisotropy and the complexity of chain entanglements. This approach allows us to quantify separately the degree of polymer collapse and the evolution of largescale folding features. Our results allow us to address these questions: 1. Under which conditions can we trigger unfolding in neutral or charged proteins in vacuo?
suggestive of an external principle underlying protein evolution, which, in turn, is shown to be possibly associated with the emergence of secondary structures.
Applications of NMR for the Characterization of Protein Dynamics and Folding Clay Bracken, Department of Biochemistry, Weill Medical College of Cornell University, New York, NY 10021 USA
Protein flexibility is increasingly being recognized for its significance in the folding and function of proteins. A variety of high resolution NMR techniques can now be applied to characterize protein flexibility at the individual residue level to elucidate dynamic motions, and conformational changes in proteins. Using a combination of ~SN relaxation experiments in conjunction with the temperature dependence of NMR chemical shifts, we can identify specific locations for the formation of secondary structure in disordered regions of proteins. The backbone entropic cost of structure formation can be estimated by measuring changes in NMR derived order parameters. Additionally, NMR relaxation experiments can be employed for estimation of millisecond timescale protein folding rates under equilibrium conditions. Recent examples of NMR dynamics analysis will be presented focusing on assessing the dynamics of helical proteins undergoing helix-coil interactions.
2. Is there a common pattern of structural transitions associated with the unfolding of species in various charged states? 3. How does refolding proceed once the unfolding conditions are switched off? 4. How does the relaxation of an unfolded protein compare with the nonspecific collapse of a homopolymer? 5. Is there evidence of a well-defined folding process in vacuo, starting from the unfolded state ensemble?
Molecular Simulations and Acid.Induced Protein Unfolding David A. Case, Department of Molecular Biology, The Scripps Research Instit,tte, La Jolla, CA 92037 USA
Emergence of Helical Structures from a Variational Principle Jayanth R. Banavar.l Amos Maritan, Cristian Micheletti, Flavio Seno, and Antonio Trovato, JPenn State University
A challenging question in biochemistry, physics, and even geometry is the origin of highly regular motifs such as orh e l i c e s in the folded state of biopolymers and proteins. We formulate a dynamical variational principle for selection in conformation space based on the requirement that the backbone of the native state of biologically viable polymers be rapidly accessible from the denatured state. The variational principle is shown to result in the emergence of helical order in compact structures. A novel approach, validated by an analysis of barnase and chymotrypsin inhibitor, is introduced to elucidate the role played by the geometry of the protein backbone in steering the folding to the correct native state. It is found that native state structures of proteins, in comparison with compact artificial backbones, have associated with them an exceedingly large number of conformations with a given amount of structural overlap with them; moreover the density of overlapping conformations, at a fixed value of the overlap, of unrelated proteins of the same length are nearly equal. These results are
It is often useful in computer simulations to use a simple description of solvation effects, instead of explicitly representing the individual solvent molecules. Continuum dielectric models often work well in describing the thermodynamic aspects of aqueous solvation, and approximations to such models that avoid the need to solve the Poisson equation are attractive because of their computational efficiency. I will discuss on approach, the generalized Born model, which is simple and fast enough to be used for molecular dynamics simulations of proteins and nucleic acids. Strengths and weaknesses will be discussed, both for fidelity to the underlying continuum model, and for the ability to replace explicit consideration of solvent molecules in macromolecular simulations. The focus will be on versions of the generalized Born model that have a pairwise analytical form, and therefore fit most naturally into conventional molecular mechanics calculations. These approaches are especially useful for exploring charge changes, such as those arising from protonation of side chains at low pH. I will discuss the use of both explicit solvent simulations and generalized Born theories to study conformational changes as a function of pH. Preliminary results for the acid-induced unfolding of apo-myoglobir~ will be presented.
J. Mol. Graphics Mod., 2000, Vol. 18, August-October
549
Links Conference website: http://ismb2OOO.sdsc.edu J. Andrew McCammon: http://mccammon.ucsd.edu Harold Scheraga: http://www.chem.cornell.edu/ department/Faculty/Scheraga/scheraga, html San Diego Supercomputer Center: http://www.sdsc.edu/
COMPANY PROFILE GeneFormatics Inc. ow that the genome projects have put masses of genes N in the hands of researchers, drug discovery and agbio companies are frantically hunting for targets. GeneFormatics Inc. plans to be a leader in the search for targets by predicting protein structure and function. Company founders are Jacquelyn Fetrow (formerly of The Scripps Research Institute), Adam Godzik (Burnham Institute, La Jolla, CA), and Jeffrey Skolnick and Andrzej Kolinski (Donald Danforth Plant Science Center, St. Louis, MO).
548
. J. Mol. Graphics Mod., 2000, Vol. 18, August-October
GeneFormatics' proprietary algorithms use protein structural information to predict function. The method can predict a protein's function starting from just a protein sequence, using a function identification algorithm with a predicted structure. Prediction is done using a threading algorithm, which aligns the sequences to the best matching structure in a structural database and extends sequence analysis beyond local sequence identity. After a structural model is obtained, the active site is identified using a "fuzzy functional form" (FFF), a three-dimensional descriptor of the protein's active s~te. The method is rapid and can be applied to genome sequence databases. The founders have published a demonstration of the method' in which all the open reading frames of the Escherichia coli genome were screened for the thiol-disulfide oxidoreductase activity of the glutaredoxin/thioredoxin protein family. They were able to identify sequences known to have the activity and to identify other proteins with the activity that were not previously known. The FFF method can distinguish proteins with similar active sites from proteins with similar topological folds. The company claims that using the FFF method they can identify 10 to 30 percent more hits and fewer false positives than a sequence motif analysis. GeneFormatics offers structure prediction and protein functional annotation services for pharmaceutical or biotechnology companies, analyzing client's proprietary sequence data. In the near term, GeneFormatics plans to sell subscriptions to its own database of protein structures and functions. In the long term, GeneFormatics will build its own intellectual property, based on structures and protein functions it is able to predict. For more information, contact GeneFormatics, 5830 Oberlin Drive, San Diego, CA 92130 USA. Tel. 1-858-450-3331. http://www.geneformatics.com
Reference I. Fetrow, J.S., Godzik, A., Skolnick, J.J. Functional analysis of the Escherichia coli genome using the sequence-to-structureto-function paradigm: identification of proteins exhibiting the glutaredoxin/thioredoxin disulfide oxidoreductase activity. J. Mol. Biol. 1998, 282:703-71 I.
MEETING ABSTRACTS
Protein Flexibility and Folding Traverse City, Michigan, USA August 13-17, 2000
Structural Transitions in Neutral and Charged Proteins in Vacuo Gustavo A. Arteca, Departement de Chimie et Biochimie, Laurentian University, Ramsey Lake Road, Sudbury, Ontario P3E 2C6, Canada In vacuo proteins provide the simplest laboratory to study
the specific roles of sequence, initial configuration, and environment on protein folding. From the experimental viewpoint, these systems are now beginning to be characterized at low resolution. Molecular dynamics (MD) simulations, in combination with tools for protein shape analysis, can complement these experiments and provide further insights on folding-unfolding transitions. Here, we illustrate some aspects of this issue by using an ensemble of MD trajectories for hen egg-white lysozyme. In lysozyme ions, unfolding can be triggered by Coulombic repulsion. In neutral lysozyme, unfolding can be induced by centrifugal forces and by weakening the monomer-monomer interaction. In both cases, the resulting unfolded transients act as initial configurations for relaxation dynamics. The trajectories are analyzed in terms of two descriptors of backbone shape: the anisotropy and the complexity of chain entanglements. This approach allows us to quantify separately the degree of polymer collapse and the evolution of largescale folding features. Our results allow us to address these questions: 1. Under which conditions can we trigger unfolding in neutral or charged proteins in vacuo?
suggestive of an external principle underlying protein evolution, which, in turn, is shown to be possibly associated with the emergence of secondary structures.
Applications of NMR for the Characterization of Protein Dynamics and Folding Clay Bracken, Department of Biochemistry, Weill Medical College of Cornell University, New York, NY 10021 USA
Protein flexibility is increasingly being recognized for its significance in the folding and function of proteins. A variety of high resolution NMR techniques can now be applied to characterize protein flexibility at the individual residue level to elucidate dynamic motions, and conformational changes in proteins. Using a combination of ~SN relaxation experiments in conjunction with the temperature dependence of NMR chemical shifts, we can identify specific locations for the formation of secondary structure in disordered regions of proteins. The backbone entropic cost of structure formation can be estimated by measuring changes in NMR derived order parameters. Additionally, NMR relaxation experiments can be employed for estimation of millisecond timescale protein folding rates under equilibrium conditions. Recent examples of NMR dynamics analysis will be presented focusing on assessing the dynamics of helical proteins undergoing helix-coil interactions.
2. Is there a common pattern of structural transitions associated with the unfolding of species in various charged states? 3. How does refolding proceed once the unfolding conditions are switched off? 4. How does the relaxation of an unfolded protein compare with the nonspecific collapse of a homopolymer? 5. Is there evidence of a well-defined folding process in vacuo, starting from the unfolded state ensemble?
Molecular Simulations and Acid.Induced Protein Unfolding David A. Case, Department of Molecular Biology, The Scripps Research Instit,tte, La Jolla, CA 92037 USA
Emergence of Helical Structures from a Variational Principle Jayanth R. Banavar.l Amos Maritan, Cristian Micheletti, Flavio Seno, and Antonio Trovato, JPenn State University
A challenging question in biochemistry, physics, and even geometry is the origin of highly regular motifs such as orh e l i c e s in the folded state of biopolymers and proteins. We formulate a dynamical variational principle for selection in conformation space based on the requirement that the backbone of the native state of biologically viable polymers be rapidly accessible from the denatured state. The variational principle is shown to result in the emergence of helical order in compact structures. A novel approach, validated by an analysis of barnase and chymotrypsin inhibitor, is introduced to elucidate the role played by the geometry of the protein backbone in steering the folding to the correct native state. It is found that native state structures of proteins, in comparison with compact artificial backbones, have associated with them an exceedingly large number of conformations with a given amount of structural overlap with them; moreover the density of overlapping conformations, at a fixed value of the overlap, of unrelated proteins of the same length are nearly equal. These results are
It is often useful in computer simulations to use a simple description of solvation effects, instead of explicitly representing the individual solvent molecules. Continuum dielectric models often work well in describing the thermodynamic aspects of aqueous solvation, and approximations to such models that avoid the need to solve the Poisson equation are attractive because of their computational efficiency. I will discuss on approach, the generalized Born model, which is simple and fast enough to be used for molecular dynamics simulations of proteins and nucleic acids. Strengths and weaknesses will be discussed, both for fidelity to the underlying continuum model, and for the ability to replace explicit consideration of solvent molecules in macromolecular simulations. The focus will be on versions of the generalized Born model that have a pairwise analytical form, and therefore fit most naturally into conventional molecular mechanics calculations. These approaches are especially useful for exploring charge changes, such as those arising from protonation of side chains at low pH. I will discuss the use of both explicit solvent simulations and generalized Born theories to study conformational changes as a function of pH. Preliminary results for the acid-induced unfolding of apo-myoglobir~ will be presented.
J. Mol. Graphics Mod., 2000, Vol. 18, August-October
549