Zinc proteomics and the annotation of the human zinc proteome

Abstract

Sequence databases can be searched for homologies of zinc coordination motifs with characteristic ligand signatures. Ensuing predictions that 3-10 % of the human genes encodes zinc proteins are most remarkable. But they seem conservative when considering that database mining cannot discover new signatures or coordination environments that employ nonsequential binding of ligands and sulfur-ligand bridges. Predictions also fall short for zinc/protein interactions at protein interfaces and for inhibitory zinc sites. Zinc ions transiently target proteins that are not known to be zinc proteins, adding a hitherto unrecognized dimension to the human zinc proteome. Predicted zinc sites need to be verified experimentally. The metal can be absent or sites may bind metal ions other than zinc because protein coordination environments do not have absolute specificity for zinc. The metaphor of the "galvanization of biology" continues to gain prominence in terms of the sheer number of approximately 3000 human zinc proteins and their annotation with new functions. Clearly, description of zinc proteomes cannot be pursued solely in silico and requires zinc proteomics, an integrated scientific approach. Progress hinges on a combination of bioinformatics, biology, and significantly, analytical and structural chemistry.