First- and second-shell metal binding residues in human proteins are disproportionately associated with disease-related SNPs.

Protein structure serves as a key determinant for revealing the molecular basis of human disease. Metal ions are among the most frequently bound heterogroups in proteins affecting structure and function. We analyzed the relationship between single nucleotide polymorphisms (SNPs) associated with human disease and metal binding sites in proteins on a database scale, using structural models and predictive tools. A match was identified for 586 disease-associated SNPs (dSNPs) located at 135 predicted metal binding sites and associated with 126 diverse diseases. For 104 diseases, a metal is known to bind at the predicted site in the homologue; for 22, the analysis gives a first indication for metal involvement in the disease. As second-shell residues play an important part in metal ion binding, our analysis included protein space up to 4.5 Å from metal binding sites. The ratio of disease-associated versus nondisease-associated SNPs (dSNP/ndSNP) for first-shell residues is 7.4 and for second-shell residues, 3.1. In addition, over 13% of all dSNPs were found to be associated with first- and second-shell residues, although these residues occupy only about 3% of protein space. These results show a disproportionate association of dSNPs and metal binding sites over a wide variety of diseases.