Protein tertiary structure
||This article has multiple issues. Please help improve it or discuss these issues on the talk page.
In biochemistry and molecular biology, the tertiary structure of a protein or any other macromolecule is defined by its three-dimensional structure through its the atomic coordinates. Tertiary structure is formed by the packing of protein secondary structure elements into compact globular units called protein domains. Whole proteins can comprise one or several such domains, and its tertiary structure can refer to each individual domain as well as to the complete configuration of the whole protein, provided it contains a single, contiguous polypeptide chain backbone. Proteins that are formed by the assembly of separate, folded polypeptide chains give rise to quaternary structure.
Thorough understanding of the tertiary structure of proteins and their determinants represents a long-standing issue in biochemistry. The first predicted structure of globular protein was the cyclol model of Dorothy Wrinch, but this was quickly discounted as being inconsistent with experimental data. Modern methods are sometimes able to predict the tertiary structure de novo to within 5 Å for small proteins (<120 residues) and under favorable conditions, e.g., confident secondary structure predictions.
Determinants of tertiary structure 
In globular proteins, tertiary interactions are frequently stabilized by the sequestration of hydrophobic amino acid residues in the protein core, from which water is excluded, and by the consequent enrichment of charged or hydrophilic residues on the protein's water-exposed surface. In secreted proteins that do not spend time in the cytoplasm, disulfide bonds between cysteine residue helps to maintain the protein's tertiary structure. A variety of common and stable tertiary structures appear in a large number of proteins that are unrelated in both function and evolution - for example, many proteins are shaped like a TIM barrel, named for the enzyme triosephosphateisomerase. Another common structure is the highly stable dimeric coiled coil structure composed of 2-7 alpha helices. Proteins are classified by the folds they represent in databases like SCOP and CATH.
Stability of native states 
The most typical conformation of a protein in its cellular environment is generally referred to as the native state or native conformation. It is commonly assumed that this state is also the most thermodynamically stable conformation attainable for any given primary structure. This is a reasonable first approximation but the claim assumes that the reaction is not under kinetic control - that is, the time required for the protein to attain its native conformation before being translated is small.
In the cell, a variety of protein chaperones assist a newly synthesized polypeptide in attaining its native conformation. Some of these proteins are highly specific in their function, such as protein disulfide isomerase. Others are very general and can be of assistance to most globular proteins - the prokaryotic GroEL/GroES system and the homologous eukaryotic Heat shock proteins Hsp60/Hsp10 system fall into this category.
Some proteins explicitly take advantage of the fact that they can become kinetically trapped in a relatively high-energy conformation due to folding kinetics. Influenza hemagglutinin, for example, is synthesized as a single polypeptide chain that acts as a kinetic trap. The "mature" activated protein is proteolytically cleaved to form two polypeptide chains that are trapped in a high-energy conformation. Upon encountering a drop in pH, the protein undergoes an energetically favorable conformational rearrangement that enables it to penetrate a host cell membrane.
Relationship to primary structure 
Tertiary structure is considered to be largely determined by the protein's primary structure - the sequence of amino acids of which the protein is composed of. Efforts to predict the tertiary structure from the primary structure are known generally as protein structure prediction. However, the environment in which a protein is synthesized and allowed to be folded are significant determinants of its final shape and are usually not directly taken into account by current prediction methods. Most of such methods do rely on comparisons between the sequences to be predicted and sequences of known structure in the Protein Data Bank. Thus, they account for the environment indirectly, assuming the target and template sequences share similar cellular contexts.
Proteins, due to the precise conformations they fold into, are nature's original nanomachines. Developing an inexpensive and practical way of designing and targeting proteins would revolutionize medicine and would have incredibly far-reaching implications. The significance of such a discovery cannot be overstated.
The majority of protein structures known to date have been solved with the experimental technique of X-ray crystallography, which typically provides data of high resolution but provides no time-dependent information on the protein's conformational flexibility. A second most-common way of solving protein structures uses NMR. It provides somewhat lower-resolution data in general and is limited to relatively small proteins, but can provide time-dependent information about the motion of a protein in the solution. Dual polarisation interferometry is a time resolved analytical method for determining the overall conformation and conformational changes in the surface captured proteins providing complementary information to these high resolution methods. More is known about the tertiary structural features of soluble globular proteins than about membrane proteins because the latter class is extremely difficult to study using these methods.
Stanford University's Folding@home project is a distributed computing research effort which uses its approximately 5 petaFLOPS (~10 x86 petaFLOPS) of computing power to attempt to model the tertiary and quaternary structures of proteins, as well as other aspects of how and why proteins fold into inordinately complex and varied shapes. No currently existing algorithm is yet able to consistently predict a proteins' tertiary or quaternary structure given only its primary structure. Learning how to accurately predict the tertiary and quaternary structure of any protein given only its amino acid sequence and the pertinent cellular conditions would be a monumental achievement. The calculations performed by the algorithms are constantly evolving, increasing in complexity and nuance, and involve an enormous number of variables. These techniques are superficially comparable to weather models that show hurricane storm tracks; each of several algorithms independently models a complex system somewhat different from each of its sister weather algorithms. The average of all the algorithms' output is taken to be the most likely "storm track". The shape of proteins can be elucidated through a somewhat similar process.
Researchers are also interested in proteins that can fold into more than one stable configuration. Protein aggregation diseases such as Alzheimer's Disease and Huntington's Disease as well as prion diseases such as Mad Cow disease can be better understood by constructing (and reconstructing) disease models. The most common way of doing this is by developing a way of inducing the desired disease state in test animals, for example by administering MPTP to give the animals Parkinson's disease, or knocking out a gene essential for the prevention of certain tumors from the animals' genomes. The Folding@home project and other projects similar to it now allow for the modelling of such disease states. Perhaps more importantly, full human proteins encoded by full human genes can be used without any of the ethical problems that arise in studying living human beings. They are quickly becoming indispensable tools among researchers from a broad variety of disciplines.
See also 
- Protein structure
- Folding (chemistry)
- Quaternary structure
- Structural biology
- Protein contact map
- Proteopedia, a collaborative encyclopedia of proteins and other molecules.
- IUPAC, Compendium of Chemical Terminology, 2nd ed. (the "Gold Book") (1997). Online corrected version: (2006–) "tertiary structure".
- Branden, Carl & Tooze, John (1991, 1999). Introduction to Protein Structure. New York: Garland Publising, Inc.
- Kyte, Jack (1995). Structure in Protein Chemistry. New York: Garland Publishing, Inc. ISBN 0-81531701-8.
- Whisstock J, Bottomley S (2006). "Molecular gymnastics: serpin structure, folding and misfolding". Curr Opin Struct Biol 16 (6): 761–8. doi:10.1016/j.sbi.2006.10.005. PMID 17079131.
- Gettins P (2002). "Serpin structure, mechanism, and function". Chem Rev 102 (12): 4751–804. doi:10.1021/cr010170. PMID 12475206.
- Whisstock JC, Skinner R, Carrell RW, Lesk AM (2000). "Conformational changes in serpins: I. The native and cleaved conformations of alpha(1)-antitrypsin". J Mol Biol. 296 (2): 685–99. doi:10.1006/jmbi.1999.3520. PMID 10669617.
- "Folding@home". Retrieved 2010-12-18.
- "Folding@home - FAQ". Retrieved 2010-12-18.
- "Folding@home - Science".
- "NHC Model Overview". Retrieved 2010-12-18.
- Schober A (October 2004). "Classic toxin-induced animal models of Parkinson's disease: 6-OHDA and MPTP". Cell Tissue Res. 318 (1): 215–24. doi:10.1007/s00441-004-0938-y. PMID 15503155.
- "Tp53 Knockout Rat". Cancer. Retrieved 2010-12-18.
- "Feature - What is Folding and Why Does it Matter?". Retrieved December 18, 2010.
- Protein Data Bank
- Display, analyse and superimpose protein 3D structures
- Visualize, analyze and compare multiple protein/DNA structures and their sequences simultaneously.
- Display, analyse and superimpose protein 3D structures
- WWW-based course teaching elementary protein bioinformatics
- Critical Assessment of Structure Prediction (CASP)
- Structural Classification of Proteins (SCOP)
- CATH Protein Structure Classification
- DALI/FSSP software and database of superposed protein structures
- TOPOFIT-DB Invariant Structural Cores between proteins
- PDBWiki — PDBWiki Home Page - a website for community annotation of PDB structures.