Abstract
Purpose :
Human tyrosinase (Tyr) is a membrane glycoprotein that is the rate-limiting enzyme for the production of melanin pigment in melanosomes. Tyr mutations are involved in the genetic disease oculocutaneous albinism type 1 (OCA1), which is described by the complete (Type A) or partial (Type B) absence of pigment in the skin, hair, and eyes related to Tyr enzymatic activity. Recently, we expressed truncated Tyr and studied the effect of OCA1B mutant variants on protein function. The human tyrosinase crystal structure is currently unknown. Our attempts to crystalize the Tyr protein were unsuccessful, further intensifying our purpose for creating an atomic model.
Methods :
The glycosylated Tyr intra-melanosomal domain and OCA1 mutant structures (T373K, P406L, R402Q, R422Q, R422W) were built by homology modeling using non-human tyrosinase and hemocyanin crystal structures as templates. The models were further refined with molecular dynamics simulations and analyzed to ensure accuracy. Finally, the Gibbs free energy changes and stabilizing energies were calculated for the wild-type and mutants using the computer algorithm FoldX.
Results :
Our results confirmed that the OCA1A mutation, T373K, is associated with a severe structural change that causes protein misfolding (ΔΔG = 3.89kcal/mol). In addition, we calculated free energy changes between the wild type and OCA1B mutants, compared these results with experimental free energy changes from our unfolding/refolding experiments, and demonstrated a strong correlation (Pearson’s r = 0.98). The unfolding effect of 4 OCA1B mutants has a lesser structural effect compared to that of the OCA1A mutant, showing a less significant decrease of protein stability (0.14 – 1.03kcal/mol). Mutation P406L had the most significant change, suggesting about 80% of misfolded protein. R402Q, a genetic polymorphism, also demonstrated a decrease of protein stability indicating clinical importance.
Conclusions :
In conclusion, we built an atomic model of the human tyrosinase catalytic domain and confirmed its accuracy by comparing biochemically-determined effects of clinical mutations. We have used this model to predict phenotype, further helping to clarify the genotype to phenotype relationship in OCA1 patients. In the future, our model will be useful for in silico analysis to understand the role of chemical compounds in recovering enzymatic activity of Tyr mutant variants.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.