Dominant negative ADA2 mutations cause ADA2 deficiency in heterozygous carriers

Abstract

Human ADA2 deficiency (DADA2) is an inborn error of immunity with a broad clinical phenotype which encompasses vasculopathy including livedo racemosa and lacunar strokes, as well as hemato-immunological features. Diagnosis is based on the combination of decreased serum ADA2 activity and the identification of biallelic deleterious alleles in the ADA2 gene. DADA2 carriers harbor a single pathogenic variant in ADA2 and are mostly considered healthy and asymptomatic. However, some DADA2 carriers present a phenotype compatible with DADA2. Here, we report ten patients from seven kindreds presenting with a phenotype indicative of DADA2, in whom only a single pathogenic variant (p.G47R, p.G47V, p.R169Q, p.H424N) was identified. To test whether being heterozygote for specific variants could explain the patients' phenotype, we investigated the effect of the ADA2 missense variants p.G47A, p.G47R, p.G47V, p.G47W, p.R169Q, p.E328K, p.T360A, p.N370K, p.H424N and p.Y453C on ADA2 protein expression, secretion and enzymatic activity. Functional studies indicate that they exert a dominant negative effect on ADA2 enzymatic activity, dimerization and/or secretion. At the molecular level, heterozygosity for these variants mimics what is observed in DADA2. We conclude that humans with heterozygous dominant negative missense variants in ADA2 are at risk of DADA2.

Figure 1.. Pedigree analysis of 10 DADA2 carriers presenting with DADA2 clinical phenotype.

A. Pedigrees of seven kindreds showing familial segregation of ADA2 missense variants. M represents mutant. Individuals with unknown genotype are labeled ‘?’. Black filled symbols represent individuals with 2 pathogenic alleles, half-filled symbols represent individuals with 1 pathogenic allele ‘P’ represents individuals carrying 1 pathogenic allele with a DADA2 phenotype. B. Schematic representation of the functional domains of the ADA2 protein and the location of the ADA2 variants identified in our cohort (labeled ‘*’), in previous studies (labeled ‘~’) and in literature (labeled ‘§’).

Figure. 2. Secretion of ADA2 dimers in homozygous or heterozygous state on non-denaturating gel.

A. ADA2 dimer secretion of HEK293T cells transfected with WT and/or ADA2 variants. Cells and supernatant were collected 48h after transfection. B. Quantification of ADA2 secretion in Supernatant of transfected HEK293T cells with wild-type ADA2 or ADA2 variants in homozygous conditions. Bar graphs represent percentage of ADA2 protein secretion relative to wild-type ADA2 100%. C. Quantification of ADA2 secretion of co-transfected HEK293T cells of ADA2 variants together with wild-type ADA2 in heterozygous conditions. Bar graphs represent percentage of ADA2 secretion relative to wild-type ADA2 50%. experiments. A-C. Image shown represents 3 independent experiments.

Figure 3.. Adenosine deaminase activity of ADA2 variants in homozygous or heterozygous state

A. Adenosine deaminase activity in whole cell lysate of HEK293T transfected cells with WT and ADA2 variants in homozygous conditions. Bar graphs represent the percentage of enzymatic activity relative to wild-type ADA2 100%. B. Adenosine deaminase activity in supernatant of HEK293T transfected cells with WT and ADA2 variants in homozygous conditions. Bar graphs represent the percentage of enzymatic activity relative to wild-type ADA2 100%. C. Adenosine deaminase activity whole cell lysate of HEK293T transfected cells with WT and/or ADA2 variants in heterozygous conditions. Bar graphs represent the percentage of enzymatic activity relative to wild-type ADA2 50%.. D. Adenosine deaminase activity in supernatant of HEK293T transfected cells with WT and/or ADA2 variants. Bar graphs represent the percentage of enzymatic activity relative to wild-type ADA2 50%.. A-D. Data represents mean ± SD from 3 independent experiments.

Figure 4.. Molecular modeling of ADA2 variants G47A, G47R, G47V, G47W, R169Q, E328K, H424N and Y453C.

A. Comparative plots of the different mutations which displays the variability of B-factor values along the protein backbone compared to the wild target. Blue backbones represent lower fluctuations compared to the wild target, red backbones represent higher fluctuations. B. Zoom-in of residue R169 in the wild target protein. R169 is shown in magenta. Polar and aromatic interactions are shown as dashed lines. C. Zoom-in of residue E328 in the wild target protein. E328 is shown in magenta. Polar interactions are shown as dashed lines. D. Zoom-in of residue Y453 in the wild target protein. Y453 is shown in magenta. Polar interactions are shown as dashed lines.

Figure 5.. Serum ADA2 enzymatic activity of suspected DADA2 patients.

A. ADA2 enzyme activity (U/L) measured in serum samples of suspected DADA2 patients (n=10), adult healthy controls (n=35), pediatric healthy controls (n=17), healthy DADA2 carriers (n=19) and DADA2 patients (n=18). Each data point is plotted with mean ± SD. Statistical significance was assessed using the Mann-Whitney U test,*** p<0.0001. B. ADA2 protein secretion in Serum samples of pediatric controls, adult controls, DADA2 patients and cohort patients by Western blot. C-D. Quantification of ADA2 secretion in Serum samples of pediatric controls, adult controls, DADA2 patients and cohort patients. Bar graphs represent percentage of ADA2 protein secretion relative to Pediatric/adult controls. Each bar represents mean ± SD from 3 independent experiments.