Prediction of phenotypes and tetrahydrobiopterin-responsiveness in phenylketonuria using data from the genotypes and locus-specific databases. N. Blau1, S. Wettstein2, W. W. Yue3, J. Underhaug3, B. D. Marsden4, A. Martinez4, A. Honegger5, B. Perez6 1) Div Metabolic Disorders, Univ Children's Hosp, Heidelberg, Germany; 2) Div Metabolism, Univ Childrens Hospital, Zürich, Switzerland; 3) Structural Genomics Consortium, Univ Oxford, Oxford, UK; 4) Dept of Biomedicine, Univ Bergen, Bergen, Norway; 5) Univ Zürich, Dept Biochemistry, Zürich, Switzerland; 6) Dept Mol Biology, CSICUAM, Univ Autonoma, Madrid, Spain.
Background: Management of phenylketonuria (PKU) patients depends on the individual phenotype. The variability in the metabolic phenotypes in PKU is caused by different mutations within the PAH gene and thus residual phenylalanine hydroxylase (PAH) activity. In addition, it has been shown that genotypes are useful in predicting cofactor tetrahydrobiopterin (BH4; sapropterin) responsiveness in PKU. Objectives: To analyze data from available PKU-associated databases (locus-specific PAHvdb and genotypes BIOPKU) and to establish algorithms for genotype-phenotype correlation and BH4-responsiveness prediction. Methods: First, the relative frequencies of mutations, genotypes, affected gene regions and protein domains were calculated. Subsequently, PAH mutations and genotypes were scored using data from FoldX (protein damage algorithm that uses an empirical force fields), SIFT (protein function prediction based on the degree of conservation of AA residues), Polyphen2 (impact of an AA substitution on the protein structure and function), SNPs3D (molecular functional effects of non-synonymous SNPs based on structure and sequence analysis), and Rosetta ddG (impact of a sequence change on a protein's stability) prediction tools. The 3D atomic environment of each mutation was visualized using the interactive iSee concept. The PAHvdb database (833 variations; www.biopku.org) and BIOPKU database (4181 PKU patients with full genotype; www.biopku.org) were used. Results/Discussion: Amongst the 4181 patients (15,1% HPA, 24,4% mild PKU, 41,3% classic PKU, 19,2% no information) we observed 463 different mutations. The most frequently affected sites were exon 7 (22,9%) and intron 10 (9,4%), and the most affected PAH region was the catalytic domain (60,8%), followed by the regulatory (14,0%) and tetramerization domain (4,9%). c.1066-11G>A /c.1066-11G>A was the most frequent genotype (3,3%). BH4-responsiveness data were available from 2128 patients (44,4% responders). Using genotype scoring both the phenotype and BH4-responsiveness was estimated, offering a robust method for patients characterization and management.
You may contact the first author (during and after the meeting) at