Functional and Structural Characterization of Pathogenicity of Human Arginine-Histidine Variants

Missense variants that change arginine-to-histidine and histidine-to-arginine (R>H; H>R) preserve positive charge yet alter pH-dependent behavior near neutrality, creating mutation type-specific but context-dependent effects on protein function, and thus could be pathogenic. To reveal the factors causing pathogenicity, we assembled high-confidence human R>H and H>R variants from ClinVar annotated as pathogenic or benign. It was found that in both cases, R>H/H>R, pathogenic variants are strongly enriched in cores and ordered regions, while benign variants were seen on surfaces and coils. Secondary structure analysis showed mutation type specificity; R>H pathogenic variants were enriched in helices, while H>R pathogenic variants were enriched in β-strands. Regarding the pH-optimum of activity, most R>H and H>R variants fell in physiological/near-physiological pH ranges, but R>H benign variants were more frequent in the neutral/physiological pH bin, whereas H>R pathogenic variants were overrepresented in the same neutral/physiological pH range. The last observation is consistent with histidine's pK_a being tunable near physiological range, while arginine's side chain introduces a permanent positive charge, and thus H>R substitution eliminates the wild-type pH-dependence. Functional protein analyses highlighted that pathogenic variants are overrepresented at binding/interface-heavy proteins (e.g., transcription factors) and selected enzymatic classes (e.g., oxidoreductases, ion channels, transporters, ligases). Interestingly, in the vast majority of cases, the proteins in our dataset had either R>H or H>R mutations, but not both present in the same protein. Proteins harboring both variant types, R>H and H>R, were very few, and typically they had both variants, either pathogenic or benign.