Residue Variations in Human N‑Acetyltransferase 2 Enzyme Alleles: From Rapid to Slow Acetylation

Understanding the effects of single-nucleotide variations on the three-dimensional structures of drug-metabolizing enzymes at the atomic level is not a well-studied field and deserves attention as it could provide insights into tackling both drug resistance and drug efficacy/toxicity issues. Here, we examined molecular dynamics simulations of the isoniazid-metabolizing enzyme arylamine N-acetyltransferase 2 (NAT2) and established a computational approach to decipher how NAT2 transitions from being a rapid acetylator to a slow acetylator in the presence of residue variations. After revealing slight differences between two rapid acetylators (the NAT2 reference allele and the legacy reference allele), we compared them to five variants with slow clinical phenotypes. In all slow acetylators, we observed a significant reduction in hydrogen bonding and dynamic residue network metric values in functional residues. For example, in the R64Q+K268R variant, the R64Q substitution causes a substantial reduction of hydrogen bonds, resulting in increased root-mean-square fluctuation and changes in residue networks. This is evidenced by the significantly decreased betweenness centrality values of catalytic D122 and putative isoniazid-binding residues S125 and F217, as well as the significantly decreased eigencentrality values of N72, D122, and the active site loop residue G124. Also, in this variant, the average distance between the putative isoniazid-binding residue F217 and catalytic residues H107 and D122 is significantly reduced. This structural shift may affect ligand binding affinity and catalysis. Overall, our findings offer new insights into how allosteric variations alter the NAT2 mechanism, leading to changes in drug exposure.