Functional characterisation of alpha-galactosidase a mutations as a basis for a new classification system in fabry disease

Abstract

Fabry disease (FD) is an X-linked hereditary defect of glycosphingolipid storage caused by mutations in the gene encoding the lysosomal hydrolase α-galactosidase A (GLA, α-gal A). To date, over 400 mutations causing amino acid substitutions have been described. Most of these mutations are related to the classical Fabry phenotype. Generally in lysosomal storage disorders a reliable genotype/phenotype correlation is difficult to achieve, especially in FD with its X-linked mode of inheritance. In order to predict the metabolic consequence of a given mutation, we combined in vitro enzyme activity with in vivo biomarker data. Furthermore, we used the pharmacological chaperone (PC) 1-deoxygalactonojirimycin (DGJ) as a tool to analyse the influence of individual mutations on subcellular organelle-trafficking and stability. We analysed a significant number of mutations and correlated the obtained properties to the clinical manifestation related to the mutation in order to improve our knowledge of the identity of functional relevant amino acids. Additionally, we illustrate the consequences of different mutations on plasma lyso-globotriaosylsphingosine (lyso-Gb3) accumulation in the patients' plasma, a biomarker proven to reflect the impaired substrate clearance caused by specific mutations. The established system enables us to provide information for the clinical relevance of PC therapy for a given mutant. Finally, in order to generate reliable predictions of mutant GLA defects we compared the different data sets to reveal the most coherent system to reflect the clinical situation.

Figure 1. Correlation analysis of α-Gal A level (semi-quantitative Western Blot) and activity.

A: No correlation between GLA level and residual activity for mutations possessing less than 6% residual activity (n = 76, Spearman correlation coefficient rs = 0.128, p = 0.272). This implies that the catalytic unit is affected by the mutation and thus high amount of enzyme cannot compensate for the loss of activity. B: For mutation possessing more than 6% residual GLA activity (n = 48), the in vitro enzyme activity and GLA levels correlate with each other, indicating that the catalytic core is still intact and mutation most likely affect protein trafficking (Spearman correlation coefficient rs = 0.866, p<0.001).

Figure 2. In vitro activity of specific GLA site mutations.

Note that mutations at position p.Asp264 almost always lead to a loss of GLA activity, while the same does not hold true for p.Arg118 and p.Ser126. Interestingly mutations in p.R118 do not lead to a loss of activity below 20% of WT and range from 20% to 80% while p.Ser126 can lose all activity with certain mutations and retains no more than 60% activity. This highlights the differential effects of the mutational site and amino acid change on α-Gal A activity. Given is the median activity of all mutations in each position (horizontal mark).

Figure 3. Lyso-Gb3 values for female and male Fabry patients compared to control.

The horizontal mark indicated the median. It is noteworthy that lyso-Gb3 levels in males are ∼10 times higher than in females. Each data point represents one patient. Indicated in pink are patients with the mutation p.S126G (8f/4m), in blue p.A143T (10f/8m) and in green p.D313Y (33f/24m) to illustrate that most found non-pathogenic mutations belong to either one or the other patient cohort. Other exceptions are: p.R118C, p.V316I, p.E418G (one male patient each) and p.A20P, p.D83N, p.I91T, p.S102L, p.R112C, p.R118C, p.D175E, p.G325S, p.A368T, p.T385A, p.W399*, c.1208delT, p.L415F, (one female patient each) and p.R252T (4×), p.N215S (3×). About 180 healthy probands were tested with no Fabry gene variation and had values of 0.9 ng/ml (95^th percentile calculation).