De novo variants in EMC1 lead to neurodevelopmental delay and cerebellar degeneration and affect glial function in Drosophila

Abstract

Background: The endoplasmic reticulum (ER)-membrane protein complex (EMC) is a multi-protein transmembrane complex composed of 10 subunits that functions as a membrane-protein chaperone. Variants in EMC1 lead to neurodevelopmental delay and cerebellar degeneration. Multiple families with biallelic variants have been published, yet to date, only a single report of a monoallelic variant has been described, and functional evidence is sparse.

Methods: Exome sequencing was used to investigate the genetic cause underlying severe developmental delay in three unrelated children. EMC1 variants were modeled in Drosophila, using loss-of-function (LoF) and overexpression studies. Glial-specific and neuronal-specific assays were used to determine whether the dysfunction was specific to one cell type.

Results: Exome sequencing identified de novo variants in EMC1 in three individuals affected by global developmental delay, hypotonia, seizures, visual impairment and cerebellar atrophy. All variants were located at Pro582 or Pro584. Drosophila studies indicated that imbalance of EMC1-either overexpression or knockdown-results in pupal lethality and suggest that the tested homologous variants are LoF alleles. In addition, glia-specific gene dosage, overexpression or knockdown, of EMC1 led to lethality, whereas neuron-specific alterations were tolerated.

Discussion: We establish de novo monoallelic EMC1 variants as causative of a neurological disease trait by providing functional evidence in a Drosophila model. The identified variants failed to rescue the lethality of a null allele. Variations in dosage of the wild-type EMC1, specifically in glia, lead to pupal lethality, which we hypothesize results from the altered stoichiometry of the multi-subunit protein complex EMC.

Figure 1

Pedigree, EMC1 conservation and in silico protein modelling. (A) Pedigrees of the studied families. (B) The domain that includes the variants is highly conserved from fly to human. Biallelic variants are depicted above the protein diagram, and monoallelic variants are shown below the protein diagram. Variants reported here are shown in red. (C) Protein sequence alignment in multiple species confirms evolutionary conservation of Pro582 in Drosophila. (D) The β-propeller domain of the EMC1 of the Cryo-EM human ER membrane protein complex structure (PDB ID: 7ADO) is shown. The loop region is magnified and shown in the rectangular area, with the known mutated residues in the de novo cases labeled. Pro585 sits at the edge of the loop before the beginning of the β-strand.

Figure 2

EMC1 is a functional fly homolog of human EMC1. (A) Ubiquitous expression of EMC1 reference using the Da-GAL4 driver is toxic, but expression of the variants is less toxic. Numbers indicate the ratio of o/e flies. (B) Strong ubiquitous expression of EMC1 reference with act-GAL4 rescued the pupal lethality observed in EMC1^655G/EMC1^655G flies, whereas expression of EMC1 p.Pro506His or EMC1 p.Pro509His variants failed to rescue pupal lethality. Note that the expression of EMC1 p.Pro506His results in a few escapers, suggesting that EMC1 p. Pro506His is less pathogenic than EMC1 p.Pro509His in flies. Numbers indicate the ratio of o/e flies. (C) Schematic image showing the locus of EMC1^KG4 allele. (D) An expression pattern of EMC1 in adult CNS (Top) or larval CNS (Bottom). The data suggest that EMC1 is expressed in both neurons and glia.

Figure 3

Loss or gain of EMC1 in glia results in lethality. (A) Expression of EMC1 RNAi-1 or RNAi-2 by hs-Gal4 after 1 h heat-shock significantly reduced the level of EMC1 transcripts by 50% (RNAi-1) or 65% (RNAi-2). (B) Ubiquitous expression of EMC1 RNAi by Da-Gal4 or Act-Gal4 causes lethality, as does glial-specific expression of EMC1 RNAi (Repo > EMC1 RNAi-1 and 2). Neuronal-specific expression of EMC1 RNAi by Nsyb-Gal4 or elav-Gal4 does not cause lethality. Co-expression of human EMC1 reference rescued the lethality of Repo > EMC1 RNAi-1 and 2 (n = 9 crosses/genotype). (C) Co-expression of human EMC1 reference rescued the climbing defects of Repo > EMC1 RNAi-1 and 2 flies. (D) Glial overexpression of wild-type fly EMC1 is toxic, but expression of the variants is less toxic in flies. (E) Neuronal overexpression of either fly EMC1 wild-type or variants does not cause lethality. Numbers indicate the ratio of o/e flies. Statistical analyses are one-way ANOVA followed by a Tukey’s post hoc test. Results are mean ± s.e.m. ^***P < 0.001.