Abstract
BACKGROUND: We examined the effects of the R325W mutation on the three-dimensional (3D) structure of the beta-cell-specific Zn2+ (zinc) transporter ZnT-8. METHODS: A model of the C-terminal domain of the human ZnT-8 protein was generated by homology modeling based on the known crystal structure of the Escherichia coli (E. coli) zinc transporter YiiP at 3.8 A resolution. RESULTS: The homodimer ZnT-8 protein structure exists as a Y-shaped architecture with Arg325 located at the ultimate bottom of this motif at approximately 13.5 A from the transmembrane domain juncture. The C-terminal domain sequences of the human ZnT-8 protein and the E. coli zinc transporter YiiP share 12.3% identical and 39.5% homologous residues resulting in an overall homology of 51.8%. Validation statistics of the homology model showed a reasonable quality of the model. The C-terminal domain exhibited an alphabetabetaalphabeta fold with Arg325 as the penultimate N-terminal residue of the alpha2-helix. The side chains of both Arg325 and Trp325 point away from the interface with the other monomer, whereas the epsilon-NH3+ group of Arg325 is predicted to form an ionic interaction with the beta-COO- group of Asp326 as well as Asp295. An amino acid alignment of the beta2-alpha2 C-terminal loop domain revealed a variety of neutral amino acids at position 325 of different ZnT-8 proteins. CONCLUSIONS: Our validated homology models predict that both Arg325 and Trp325, amino acids with a helix-forming behavior, and penultimate N-terminal residues in the alpha2-helix of the C-terminal domain, are shielded by the planar surface of the three cytoplasmic beta-strands and hence unable to affect the sensing capacity of the C-terminal domain. Moreover, the amino acid residue at position 325 is too far removed from the docking and transporter parts of ZnT-8 to affect their local protein conformations. These data indicate that the inherited R325W abnormality in SLC30A8 may be tolerated and results in adequate zinc transfer to the correct sites in the pancreatic islet cells and are consistent with the observation that the SLC30A8 gene variant R325W has a low predicted value for future type 2 diabetes at population-based level.