Mechanism of Action of Non-Synonymous Single Nucleotide Variations Associated with α-Carbonic Anhydrase II Deficiency

Abstract

Human carbonic anhydrase II (CA-II) is a Zinc (Zn2+) metalloenzyme responsible for maintenance of acid-base balance within the body through the reversible hydration of CO2 to produce protons (H+) and bicarbonate (BCT). Due to its importance, alterations to the amino acid sequence of the protein as a result of single nucleotide variations (nsSNVs) have detrimental effects on homeostasis. Six pathogenic CA-II nsSNVs, K18E, K18Q, H107Y, P236H, P236R and N252D were identified, and variant protein models calculated using homology modeling. The effect of each nsSNV was analyzed using motif analysis, molecular dynamics (MD) simulations, principal component (PCA) and dynamic residue network (DRN) analysis. Motif analysis identified 11 functionally important motifs in CA-II. RMSD data indicated subtle SNV effects, while PCA analysis revealed that the presence of BCT results in greater conformational sampling and free energy in proteins. DRN analysis showed variant allosteric effects, and the average betweenness centrality (BC) calculations identified Glu117 as the most important residue for communication in CA-II. The presence of BCT was associated with a reduction to Glu117 usage in all variants, suggesting implications for Zn2+ dissociation from the CA-II active site. In addition, reductions to Glu117 usage are associated with increases in the usage of the primary and secondary Zn2+ ligands; His94, His96, His119 and Asn243 highlighting potential compensatory mechanisms to maintain Zn2+ within the active site. Compared to traditional MD simulation investigation, DRN analysis provided greater insights into SNV mechanism of action, indicating its importance for the study of missense mutation effects in proteins and, in broader terms, precision medicine related research.

Understanding the Pyrimethamine Drug Resistance Mechanism via Combined Molecular Dynamics and Dynamic Residue Network Analysis

“…This points to distant effects on residue connectivity induced by drug-resistance mutations. Similar long-range effects due to mutations have been reported in previous studies [22].…”

“…Altogether, despite pairwise Pearson's correlation of ∆BC ( Figure S13) failing to group the different mutants based on respective degrees of resistance to pyrimethamine as discussed before [15] (WT: sensitive; SM: low resistant; DM1, DM2: moderately resistant; TM1, TM2: high resistant; and QM: highly resistant), these calculations provided useful clues related to catalytic efficiency. This is expected since DRN analysis highlights differences in intra-protein communication (which relates to protein function) when applied in comparative analysis between WT and mutant [22,37]. These findings further suggest that the pyrimethamine resistance mechanism is based on localized adjustments due to mutations within the Pf DHFR active site and may not correlate to its catalytic activity.…”

“…To be able to observe potential discrete conformational changes in protein systems and hence to understand the variation effect on them, RMSD distribution histograms of the WT and variant proteins both in ligand free and ligand bound were generated, as previously applied [22][23][24] (Figure 2A). The pyrimethamine-free WT model displayed nearly normal RMSD distribution while the bound system showed a bimodal curve, suggesting that the former predominantly sampled a single conformer while the latter sampled at least two conformations during simulation.…”

“…This indicates that there were only subtle differences in average BC values across the systems [37], which can be expected since drug-resistance mutations evolve in a conservative fashion such that protein fitness is not overly jeopardized. To further elucidate mutation-and pyrimethamine-induced differences on protein communication, we calculated the changes in BC (∆BC) values as differences between WT and mutants and as differences between ligand-free and corresponding ligand-bound systems, as demonstrated in previous studies [22,30].…”

Integrated Computational Approaches and Tools for Allosteric Drug Discovery

“…For example, network analysis is usually performed to determine the pathways that connect the mutation and the active site and aids in the study of allosteric communication [86,94,95]. A recent publication studied six validated non-synonymous single nucleotide variations (nsSNVs) to identify underlying mechanisms responsible for CA-II deficiencies resulting in the phenotype of osteopetrosis with renal tubular acidosis and cerebral calcification [96]. In this study, Sanyanga et al combined MD and DRN [97] analysis and showed that nsSNVs have indirect/allosteric effects, providing greater insights into SNV mechanism of action.…”