Variants in the ethylmalonyl-CoA decarboxylase (ECHDC1) gene: a novel player in ethylmalonic aciduria?

Abstract

Ethylmalonic acid (EMA) is a major and potentially cytotoxic metabolite associated with short-chain acyl-CoA dehydrogenase (SCAD) deficiency, a condition whose status as a disease is uncertain. Unexplained high EMA is observed in some individuals with complex neurological symptoms, who carry the SCAD gene (ACADS) variants, c.625G>A and c.511C>T. The variants have a high allele frequency in the general population, but are significantly overrepresented in individuals with elevated EMA. This has led to the idea that these variants need to be associated with variants in other genes to cause hyperexcretion of ethylmalonic acid and possibly a diseased state. Ethylmalonyl-CoA decarboxylase (ECHDC1) has been described and characterized as an EMA metabolite repair enzyme, however, its clinical relevance has never been investigated. In this study, we sequenced the ECHDC1 gene (ECHDC1) in 82 individuals, who were reported with unexplained high EMA levels due to the presence of the common ACADS variants only. Three individuals with ACADS c.625G>A variants were found to be heterozygous for ECHDC1 loss-of-function variants. Knockdown experiments of ECHDC1, in healthy human cells with different ACADS c.625G>A genotypes, showed that ECHDC1 haploinsufficiency and homozygosity for the ACADS c.625G>A variant had a synergistic effect on cellular EMA excretion. This study reports the first cases of ECHDC1 gene defects in humans and suggests that ECHDC1 may be involved in elevated EMA excretion in only a small group of individuals with the common ACADS variants. However, a direct link between ECHDC1/ACADS deficiency, EMA and disease could not be proven.

Keywords: digenic inheritance; ethylmalonic aciduria (EMA); ethylmalonyl-CoA decarboxylase (ECHDC1); short-chain acyl-CoA dehydrogenase (SCAD); synergistic heterozygosity.

FIGURE 1

Location and predicted molecular consequences of ACADS and ECHDC1 variants identified in the three individuals with high EMA. A and B, Genomic location of the two common ACADS susceptibility variants and of the three ECHDC1 variants. The 5′ and 3′ UTR regions are displayed in orange. C, P3 carries an indel variant in the acceptor splice site of intron 2; 6nt are deleted (depicted in red) and 2nt are inserted (depicted in green). D, P2 carries a deletion of 4nt (depicted in red) next to a predicted branch point motif (blue box) in intron 5. E, P1 carries a c.389T>C missense variant in exon 4, which results in the replacement of p.Met130 with a Thr. Met130 is a highly conserved amino acid in numerous eukaryotic species

FIGURE 2

Decarboxylation activity of p.Met130Thr variant ECHDC1 compared to wild‐type ECHDC1. Both proteins were produced in E. coli as fusion proteins with an N‐terminal His‐tag. They were purified to homogeneity and incubated in the reaction mixture at the indicated protein concentrations in the presence of [¹⁴C]ethylmalonyl‐CoA (EMCoA). The reaction was stopped at the indicated times and the residual [¹⁴C]ethylmalonyl‐CoA was determined. Data represent means ± SD for 6 wild‐type (WT, 1.6 μg/mL), 4 p.Met130Thr (6.4 μg/mL) and 2 p.Met130Thr (1.6 μg/mL) determinations performed in 2 separate experiments. If not visible, the SD bar is smaller than the size of the symbol

FIGURE 3

ECHDC1 mRNA expression in fibroblasts from individual 2 and 3 compared to three controls. The error bars (±SD) represent triplicate RT‐qPCR measurements from each of 2 to 3 independent cell cultures

FIGURE 4

Molecular consequences of ECHDC1 c.221‐4_222delinsTA. A, The c.221‐4_222delinsTA mutation deletes intron 2 acceptor splice site, creating a weaker acceptor splice site 3nt downstream in exon 3, and activating a stronger cryptic acceptor splice site 44nt upstream in intron 2. The position of the acceptor‐splice‐site‐motif G (NAG/NN) is indicated. B, The strength of the two alternative acceptor splice sites, according to five different prediction programs, is shown; SSF: SpliceSiteFinder, MaxE: MaxEntScan, NNS: NNSplice GS: GeneSplicer, HSF: Human Splicing Finder (see Material and Methods). C, The PCR products obtained, using cDNA as template and primers located in exon 2 and 3, respectively, are shown (3% agarose gel). In P3, two bands were observed: a ~430 bp band and a ~380 bp band, similar in size to the wild‐type band. D, The activation of the upstream acceptor splice site, results in partial intron 2 retention (~430 bp band) and creation of a premature stop codon due to a shifted reading frame. The TA insertion is depicted in red. E, The PCR products obtained with primers, specific for the inserted intronic fragment, are displayed and shown to be unique for P3 cDNA (2% agarose gel). F, Sequence data from analysis of the ~380 bp band in panel (C). The creation of the downstream acceptor splice site results in 3 bp deletion (c.221_223del). A heterozygous 3 bp deletion is clearly visible in the sequencing data derived from P3, but not in C1. G, The 3 bp deletion causes the deletion of p.Gly74. The deleted nucleotides are enclosed in the red box. H, A multiple alignment analysis shows that p.Gly74 is conserved among Eukaryotic species

FIGURE 5

EMA Excretion in Fibroblasts with the Three Possible ACADS c.625G>A Genotypes, with or without ~50% ECHDC1 Knockdown and with or without Butyrate Stimulation. Fibroblasts were treated either with non‐targeting shRNA or with ECHDC1 targeting shRNA. Fibroblasts were treated for 24 hours either with 0 mM or with 5 mM butyrate. EMA excretion was measured for all cell lines and expressed as EMA/protein. Data from two experiments (c.[625G>A];[625G>A]) or three experiments (c.[=];[=] and c.[625G>A];[=]) are shown as the mean ± SD