Congenital disease SNPs target lineage specific structural elements in protein kinases

Abstract

The catalytic domain of protein kinases harbors a large number of disease-causing single nucleotide polymorphisms (SNPs) and common or neutral SNPs that are not known or hypothesized to be associated with any disease. Distinguishing these two types of polymorphisms is critical in accurately predicting the causative role of SNPs in both candidate gene and genome-wide association studies. In this study, we have analyzed the structural location of common and disease-associated SNPs in the catalytic domain of protein kinases and find that, although common SNPs are randomly distributed within the catalytic core, known disease SNPs consistently map to regulatory and substrate binding regions. In particular, a buried side-chain network that anchors the flexible activation loop to the catalytic core is frequently mutated in disease patients. This network was recently shown to be absent in distantly related eukaryotic-like kinases, which lack an exaggerated activation loop and, presumably, are not regulated by phosphorylation.

Fig. 1.

Kinase subdomains and SNP distribution. (A) The subdomains PKA (PDB entry 1ATP). Gray residues are intervening loops. Subdomains are numbered by roman numerals and color coded. (B) The distribution of kinase disease SNPs. Spheres denote residues with high disease SNP frequencies; red, eight SNPs; yellow, seven SNPs; orange, six SNPs; green, five SNPs; and blue, four SNPs. (C) Ratio of observed to expected SNPs per region. Roman numerals correspond to subdomains in A, where a denotes the intervening region between subdomains. Black bars, disease SNPs; gray bars, common SNPs. The image was created in part with Protein Workshop (54).

Fig. 2.

Distribution of disease and common SNPs in N-lobe subdomains. The distribution of disease and common SNPs and the degree of conservation per residue in subdomains I (A) and III–IV (B). Black bars, disease SNPs; gray bars, common SNPs. The character height is proportional to the degree of conservation. The number of common SNPs is adjusted for the difference in total common and disease SNPs occurring throughout the catalytic core. Arrow denotes disease hotspot G55.

Fig. 3.

Distribution of disease and common SNPs in subdomains VII and VIII. The distribution of disease and common SNPs and the degree of conservation per residue in subdomains VII (A) and VIII (B). Black bars, disease SNPs; gray bars, common SNPs. The character height is proportional to the degree of conservation. The number of common SNPs is adjusted for the difference in total common and disease SNPs occurring throughout the catalytic core. Arrow denotes disease hotspots R165 and E170.

Fig. 4.

Conserved core shared between EPKs and ELK and structural elements unique to EPK. Structural comparison of PKA and distantly related aminoglycoside kinase (APH). A conserved core shared between PKA and APH is shown (in transparent mode). The exaggerated activation segment, which is one of the distinguishing features of EPKs is shown in red. The C-terminal substrate binding regions of PKA and APH are omitted for clarity. These C-terminal regions are very different and likely contribute to substrate specificity. The buried side-chain network that anchors the flexible activation segment is also shown. This network appears to have coevolved with the activation segment as ELKs that lack an exaggerated activation segment also lack the side-chain network.

Fig. 5.

SNPs and allostery. The ePK conserved allosteric network of the C-terminal lobe. Red spheres, oxygen; blue spheres, nitrogen; dashed lines, hydrogen bonds. Zoom box shows the ePK conserved side-chain network.