An autosomal-dominant childhood-onset disorder associated with pathogenic variants in VCP

Abstract

Valosin-containing protein (VCP) is an AAA+ ATPase that plays critical roles in multiple ubiquitin-dependent cellular processes. Dominant pathogenic variants in VCP are associated with adult-onset multisystem proteinopathy (MSP), which manifests as myopathy, bone disease, dementia, and/or motor neuron disease. Through GeneMatcher, we identified 13 unrelated individuals who harbor heterozygous VCP variants (12 de novo and 1 inherited) associated with a childhood-onset disorder characterized by developmental delay, intellectual disability, hypotonia, and macrocephaly. Trio exome sequencing or a multigene panel identified nine missense variants, two in-frame deletions, one frameshift, and one splicing variant. We performed in vitro functional studies and in silico modeling to investigate the impact of these variants on protein function. In contrast to MSP variants, most missense variants had decreased ATPase activity, and one caused hyperactivation. Other variants were predicted to cause haploinsufficiency, suggesting a loss-of-function mechanism. This cohort expands the spectrum of VCP-related disease to include neurodevelopmental disease presenting in childhood.

Figure 1

Variants in VCP in this cohort versus published cases (A) Displayed is the canonical VCP protein and the N, D1, and D2 domains. The theoretical protein products for probands 1 and 3 are shown above; the novel sequence is in light blue, and both products are predicted to undergo nonsense-mediated mRNA decay before protein production. In-frame deletions are noted by yellow circles, and missense mutations in our cohort are denoted by orange squares, followed by the proband ID in parentheses. Below the bar, 44 previously described missense mutation sites are noted in blue (classic MSP with features of IBMPFD1), red (any other phenotype), or purple (classic MSP and other phenotype overlap).^,, (B) Cryo-EM structures with a single subunit of VCP; the N, D1, and D2 domains are in white, blue, and purple, respectively. On the left, marked in red are several classic MSP-associated variants, clustered in the N-D1 interface. On the right are the sites of variants described in this cohort, which are spread across the protein.

Figure 2

ATPase activity of VCP variants (A) Immunoblots of lysates from U2OS cells expressing an empty vector or expression plasmids with VCP variants and a C-terminal Myc tag with an anti-VCP antibody (top), anti-Myc antibody (middle), or anti-GAPDH as the loading control (bottom). Variants include classic MSP-associated variants p.Ala232Glu and p.Arg155His and variants from all probands except 1 and 3, whose variants are predicted to undergo nonsense-mediated decay. On the right, data were run separately for p.Lys251Asn (proband 4) and p.Gly271Asp (proband 7). Note that a small doublet is present with anti-VCP antibody, demonstrating the endogenous VCP (lower band) and the recombinant myc-tagged VCP (upper band). The p.Arg625Pro variant fails to be produced. (B) Longer exposure, using the anti-myc antibody, of a similar set of lysates shows a faint band for the p.Arg625Pro variant, as well as a high-molecular-weight smear and multiple degradation products for the variants. (C) Recombinant purified VCP was obtained from bacteria, and in vitro ATPase activity was assessed by a standard colorimetric assay and normalized to that of wild-type VCP. A comparison of these variants against classic MSP mutations can be found in Figure S3. n = 2–4 biological replicates with two technical replicates each. Error bars display the SD. p values from one-way ANOVA with correction for multiple comparisons with the wild type are denoted by ^∗p < 0.05 and ^∗∗∗p < 0.001.

Figure 3

Facial features and head circumference (A) Photos of probands (from left to right): 2, 4, 8, 9, 10, 11, and 13. There was no recognizable facial gestalt for this syndrome, although some probands share features such as a broad forehead and down-slanting palpebral fissures (in probands 4, 8, 11, and 13). (B) Macrocephaly is defined as a head circumference > 2 SD above the mean. The number of SD (Z score) was plotted for the head circumference of each proband at birth (filled circle or square) and at last evaluation (open circle or square). Proband 4 (gray circle) had relative macrocephaly, as defined by a standardized head circumference > 2 SD above standardized height. Probands 3 and 9 (gray squares) were never macrocephalic. Proband 10 was evaluated at the 98^th percentile (Z + 1.97), and proband 12 was evaluated at Z + 1.93, which we consider within the margin of measuring error for macrocephaly.