Many proteins normally exist as oligomers, but the oligomer is often incomplete in the published PDB file. An educated guess about the complete oligomer, and the corresponding PDB file, is available at the Probable Quaternary Structures (PQS) site of the European Bioinformatics Institute.

In order for PQS to report a specific oligomer, that oligomer must occur in the crystal, or else the authors of the PDB file must specify the symmetry operations required to generate the oligomer (commonly done for virus capsids). PQS looks at the interchain contact patches in the crystal, and makes an automated judgement about whether they are likely to represent specific oligomer interfaces, or merely contacts that occur only in the crystal (Henrick & Thornton, 1998). Any contact less than 400 Ų per chain is deemed a For quaternary structures known from independent sources, the interfacial area per chain rarely falls below this value, and may be as high as 5,000 Ų. The mean area of specific contact between a protein antigen and an antibody is 800 Ų per chain (Janin, 1997). In order to be deemed a specific oligomeric contact by PQS, the contact must not only have an interfacial area larger than 400 Ų per chain, but must also satisfy an empirical weighted score with contributions from the loss of solvent accessible surface, gain in solvation free energy, the number of residues buried at the interface, and the number of salt bridges and disulfide bridges, if any.

There are numerous possibile relations between the asymmetric unit in the published PDB file, and specific oligomers. Here are some examples.

1. The published PDB file may contain a single chain, or multiple chains, but there is no specific oligomer, because all contacts in the crystal are crystal contacts only. Examples are 104l (lysozyme mutant) and 1b6b (serotonin N-acetyltransferase) : each of these PDB files contains a homodimer, but in each case, the interface is a crystal contact and there are no specific oligomers. By using the link to PQS available in the External Resources window (opened from the PE Site Map) you will learn that the homodimer is a crystal artifact, and you will be offered a PDB file containing only a

2. The published PDB file may happen to contain exactly the specific oligomer, as for the homodimers of 4mdh (malate dehydrogenase) or 1bdy (protein kinase C). Consulting PQS will reveal this.

3. The published PDB file may contain more than one copy of the specific oligomer. An example is 12e8 (Fab of immunoglobulin) which contains four chains representing two copies of the heterodimer. Consulting PQS will reveal this.

4. The published PDB file may contain only part of the specific oligomer. An example is 1hho (human oxyhemoglobin) which contains an alpha-beta heterodimer. This is only half of the specific tetramer. To get the complete tetramer, while viewing 1hho in PE, use the link to PQS available in the External Resources window (available from the PE Site Map).

   Another example is 5cev (arginase). The specific oligomer is a hexamer, but file 5cev contains two half-hexamers with a crystal contact interface between them.

5. A specific homodimer may appear in the PDB file as a crystal-contact dimer. That is, the actual oligomer may be a homodimer, and the published PDB file may contain a homodimer. However, the interface in the PDB-file dimer may be a crystal contact interface, rather than the specific homodimeric interface. An example is 1qrd (quinone reductase), where PQS suggests two plausible dimers, neither of which is the dimer in the published PDB file.

6. For truncated proteins, oligomers may occur that are impossible in the native protein. For example, 1bk5 (karyopherin alpha) is a truncated part of the natural chain, and forms a dimer that would be prevented by the full-length chain. Dimerization is dependent upon Y397. Mutation Y397D prevents this artifactual dimerization, leading to the monomer 1ee5.

7. The specific oligomer may be a complete virus capsid, such as for poliovirus (2plv), or simian virus 40 (1sva). (See also the Chime presentation SV40 Capsid Assembly.) In these cases, the authors typically specify the symmetry operations needed to construct the capsid in the PDB file header.

References

1. Dasgupta, S., G. H. Iyer, S. H. Bryant, C. E. Lawrence, and J. A. Bell. 1997. Extent and nature of contacts between protein molecules in crystal lattices and between subunits of protein oligomers. Proteins. 28:494-514.
2. Henrick, K., and J. M. Thornton. 1998. PQS: a protein quaternary structure file server. Trends. Biochem. Sci. 23:358-361.
3. Janin, J. 1997. Specific versus non-specific contacts in protein crystals [letter]. Nat. Struct. Biol. 4:973-974.
4. Ponstingl, H., K. Henrick, and J. M. Thornton. 2000. Discriminating between homodimeric and monomeric proteins in the crystalline state. Proteins. 41:47-57.