Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Biological or biochemical
Reexamination Certificate
2001-02-16
2004-12-14
Borin, Michael (Department: 1631)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Biological or biochemical
C702S027000
Reexamination Certificate
active
06832162
ABSTRACT:
TECHNICAL FIELD
The present invention relates generally to the field of molecular biology. More particularly, the invention relates to novel methods of predicting alpha helical segments, beta-sheets, and the tertiary structure of a polypeptide ab initio, provided with only the amino acid sequence of the polypeptide.
BACKGROUND
Proteins are essential molecules exhibiting complex structural and functional relationships. Biological functionality is defined by the native three-dimensional structure of the protein, which in turn depends on an intricate balance of molecular interactions. It is well known that many proteins fold spontaneously from random disordered states into compact (native) states of unique shape. However, the ability to explain the mechanisms that govern this transformation has not yet been realized. The goal in solving this protein folding problem is to understand this folding process and to predict the three dimensional structure of proteins from their one dimensional amino acid sequence.
Structure prediction of polypeptides and proteins from their amino acid sequences is regarded as a holy grail in the computational chemistry, molecular and structural biology communities. According to the thermodynamic hypothesis, the native structure of a protein in a given environment corresponds to the global minimum free energy of the system. Anfinsen,
Science
181, 223 (1973). Recent reviews assess the advances made in this area by researchers across several disciplines. Dill,
Prot Sci
8, 1166 (1999); Koehl and Levitt,
Nature Structural Biology
6, 108 (1999); Wales and Scheraga,
Science
285, 1368 (1999). In spite of pioneering contributions and decades of effort, the ab initio prediction of the folded structure of a protein remains a very challenging problem. The current approaches for the structure prediction of polypeptides can be classified as : (i) homology or comparative modeling methods; (ii) fold recognition or threading methods; (iii) ab initio methods that utilize knowledge-based information from structural databases (e.g., secondary and/or tertiary structure restraints); and (iv) ab initio methods without the aid of knowledge-based information.
Knowledge-based ab initio methods exploit information available from protein databases regarding secondary structure, introduce distance constraints, and extract similar fragments from multiple sequence alignments in an attempt to simplify the prediction of the folded three-dimensional protein structure. Contributions to the art include the work reported in Levitt,
J. Mol. Biol
. 170, 723 (1983); Hinds and Levitt,
J. Mol. Biol
. 243, 668 (1994); Ortizet al.,
Proc. Natl. Acad. Sci. USA
95, 1020 (1998a); Skolnick et al.,
J. Mol. Biol
. 265, 217 (1997); Simons et al.,
Proteins
34, 82 (1999a); Shortle et al.,
Proc. Natl. Acad. Sci. USA
95, 11158 (1998); Sun et al.,
Protein Engineering
8, 769 (1995); Monge et al.,
Proc. Natl. Acad. Sci. USA
91, 5027 (1994); Monge et al.,
J. Mol. Biology
247, 995 (1995); M. Standley et al.,
J Mol Bio
285, 1691 (1999).
Ab initio methods that are not guided by knowledge-based information represent the most challenging category. Important advances include, among others, Scheraga et al., J
Global Optimization
15, 235 (1999); Liwo et al.,
Proc. Natl. Acad. Sci. USA
96, 5482 (1999); Lee et al.,
Biopolymers
46, 103 (1998); Srinivasan and Rose,
PNAS
96, 14258 (1999); Yue and Dill,
Protein Science
5, 254 (1996); Dill et al.,
J. Computational Biology
4, 227 (1997). A recent assessment of the current status of both types of ab initio protein structure prediction approaches may be found in Orengo et al.,
Proteins
Suppl. 3, 149 (1999).
The above methods fail to predict accurately and reliably the tertiary structure of a polypeptide, as determined by lack of agreement with experimental observations of the tertiary structure. Thus, there is a need for more accurate and reliable methods of determining the tertiary structure of polypeptides ab initio.
SUMMARY OF THE INVENTION
The present invention provides the novel ASTRO-FOLD approach for the ab initio prediction of the three dimensional structures of proteins. The four stages of the approach are exemplified in FIG.
1
. The first stage involves the identification of helical segments. This aspect of the invention is accomplished by first partitioning the amino acid sequence into oligopeptides, for example, pentapeptides, such that consecutive pentapeptides possess an overlap of four aminoacids. Then, atomistic level modeling is performed using a selected force field. Many force field parameterizations exist for protein systems. In one aspect of the invention, the ECEPP/3 force field, which includes non-bonded, hydrogen-bonding, electrostatic, torsional and disulfide loop-closing terms is employed. Nemethy et al.,
J. Phys. Chem
. 96, 6472 (1992). The next steps involve generating an ensemble of low energy conformations, then calculating free energies that include entropic, cavity formation, polarization and ionization contributions for each pentapeptide, and finally the calculation of helix propensities for each residue using equilibrium occupational probabilities of helical clusters.
The partitioning of the amino acid sequence into overlapping oligopeptides is based on the idea that helix nucleation relies on local interactions and positioning within the overall sequence. The explicit consideration of local interactions through overlapping oligopeptides allows for detection of cases in which identical amino acid sequences adopt different conformations in different proteins. See, e.g., Minor and Kim,
Nature
380, 730 (1996). This is consistent with the observation that local interactions extending beyond the boundaries of the helical segment retain information regarding conformational preferences. See, e.g., Baldwin and Rose,
TIBS
24, 77 (1999). The partitioning pattern is generalizable and can be extended to heptapeptide, nonapeptide, and even longer systems, such as those equivalent in length to the longest known helical segments. See, e.g., Anfinsen and Scheraga,
Advances In Protein Chemistry
29, 205 (1975). Other methods have utilized partitioning schemes, but these only provide for discrete states of the central residue of non-overlapping peptides and have not considered the implications of free energy modeling. The overall methodology for the ab initio prediction of helical segments encompasses the following steps:
1. The overlapping pentapeptides are modeled as neutral peptides surrounded by a vacuum environment using the ECEPP/3 force field. An ensemble of low potential energy pentapeptide conformations, along with the global minimum potential energy conformation, is identified using a modification of the &agr;BB global optimization approach [Klepeis and Floudas,
J Chem Phys
110, 7491 (1999)] and the conformational space annealing approach [Lee et al.,
Biopolymers
46, 103 (1998)]. For the set of unique conformers, Z, determined by removing all duplicate and symmetric minima, including the clustering of any two minima that do not differ by more than 50 degrees for at least one dihedral angle, free energies (F
har
vac
) are calculated using the harmonic approximation for vibrational entropy. See Klepeis and Floudas,
J Chem Phys
110, 7491 (1999).
2. The energy for cavity formation in an aqueous environment is modeled using a solvent accessible surface area expression, F
cavity
=&ggr;A+b, wherein A is the surface area of the protein exposed to the solvent. This macroscopic free energy term is based on a fit to experimental free energies of the transfer of alkane molecules into water. The values for the &ggr; and b parameters are taken to be 0.005 kcal/molAA and 0.860 kcal/mol, respectively.
3. For the set of unique conformers, Z, the total free energy is calculated from the expression
F
total
F
vac
har
+F
cavity
+F
solv
+F
ionize
,
wherein F
sol
represents the difference in polarization energies caused by the transition from a vacuum to a solvated environment,
Floudas Christodoulos A.
Klepeis John L.
Borin Michael
The Trustees of Princeton University
Woodcock & Washburn LLP
LandOfFree
Methods of ab initio prediction of &agr; helices, &bgr;... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods of ab initio prediction of &agr; helices, &bgr;..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods of ab initio prediction of &agr; helices, &bgr;... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3310504