Computational analysis of human N-acetylgalactosamine-6-sulfate sulfatase enzyme

The heterogeneity of mutations in human N-acetylgalactosamine-6-sulfate sulfatase enzyme (GALNS) has limited the ability to establish a complete genotype–phenotype correlation for Morquio syndrome type A. To attempt the prediction of the disease severity, in this study we expanded the in silico evaluation of GALNS mutations by using several bioinformatics tools. The human GALNS sequence was compared with the one from other species and with other human sulfatases. Tertiary structure was modeled by using I-TASSER server and mutations were done by computational site-directed mutagenesis. Calcium was included within the model by YASARA. Affinity of wild type and mutated GALNS for their natural and artificial substrates was evaluated using AutoDock Vina and Molegro Virtual Docker. Human GALNS showed 198 amino acids highly conserved among GALNS sequences of the studied species, as well as 17 amino acids among human sulfatases. Forty-two out of 53 mutations in conserved amino acids were associated with severe phenotype. There was a clear correlation between minimization energy and phenotype, with severe phenotype mutations producing a higher increase in minimization energy. Molecular docking confirmed that GALNS has more affinity for natural substrates than for the artificial, and presence of calcium allowed a better modeling of substrate binding. In all the cases, mutations within the active site reduced the enzyme affinity for the substrates. An update in genotype–phenotype correlation for Morquio A from the computational biology approach is proposed, in which multiple parameters must be analyzed for prediction of disease severity. Future modeling should consider normalization of disease severity, since currently this is a subjective parameter.