Bi-allelic genetic variants in the translational GTPases GTPBP1 and GTPBP2 cause a distinct identical neurodevelopmental syndrome

Abstract

The homologous genes GTPBP1 and GTPBP2 encode GTP-binding proteins 1 and 2, which are involved in ribosomal homeostasis. Pathogenic variants in GTPBP2 were recently shown to be an ultra-rare cause of neurodegenerative or neurodevelopmental disorders (NDDs). Until now, no human phenotype has been linked to GTPBP1. Here, we describe individuals carrying bi-allelic GTPBP1 variants that display an identical phenotype with GTPBP2 and characterize the overall spectrum of GTP-binding protein (1/2)-related disorders. In this study, 20 individuals from 16 families with distinct NDDs and syndromic facial features were investigated by whole-exome (WES) or whole-genome (WGS) sequencing. To assess the functional impact of the identified genetic variants, semi-quantitative PCR, western blot, and ribosome profiling assays were performed in fibroblasts from affected individuals. We also investigated the effect of reducing expression of CG2017, an ortholog of human GTPBP1/2, in the fruit fly Drosophila melanogaster. Individuals with bi-allelic GTPBP1 or GTPBP2 variants presented with microcephaly, profound neurodevelopmental impairment, pathognomonic craniofacial features, and ectodermal defects. Abnormal vision and/or hearing, progressive spasticity, choreoathetoid movements, refractory epilepsy, and brain atrophy were part of the core phenotype of this syndrome. Cell line studies identified a loss-of-function (LoF) impact of the disease-associated variants but no significant abnormalities on ribosome profiling. Reduced expression of CG2017 isoforms was associated with locomotor impairment in Drosophila. In conclusion, bi-allelic GTPBP1 and GTPBP2 LoF variants cause an identical, distinct neurodevelopmental syndrome. Mutant CG2017 knockout flies display motor impairment, highlighting the conserved role for GTP-binding proteins in CNS development across species.

Keywords: GREND syndrome; GTPBP1; GTPBP2; NBIA; animal models; ectodermal disorders; neurodegeneration; neurodevelopmental disorders; ribosome stalling; ribosomopathies.

Graphical abstract

Figure 1

Genetic summary of the reported individuals with homozygous GTPBP1 and GTPBP2 variants (A) Pedigrees of the sixteen families described. Square, male; circle, female; black filled symbol, affected individual; white symbols, unaffected individuals. Double lines indicate consanguinity. (B) Protein multiple sequence alignment in GTPBP1 and GTPBP2 orthologs shows high conservation of the residues involved by the non-synonymous variants (highlighted in yellow) and almost complete conservation of the nearby residues (highlighted in light blue). (C) Schematic diagram indicating the domains of the GTPBP1 and GTPBP2 proteins, containing 669 and 602 amino acid residues, respectively. The orange shape represents the GTP-binding domain. The green and yellow boxes indicate the two beta-barrel domains. Variants reported in this study are represented in black, while previously reported variants in red.

Figure 2

Distinctive craniofacial features associated with GTPBP1/2-related disorders Facial pictures of I-1 (A–C), I-2 (D), I-3 (E), I-5 (F), I-7 (G and H), I-8 (I), I-9 (J), I-10 (K), I-11 (L), I-12 (M), I-14 (N), I-16 (O), I-17 (P), and I-18 (Q). Note the distinct phenotype shared by individuals with variants in GTPBP1 and GTPBP2, including microcephaly, high frontal hairline, sparse eyebrows and scalp hair, prominent nasal bridge, deep-set eyes, full lips, full cheeks, and abnormal dentition.

Figure 3

Brain imaging features associated with GTPBP1/2-related disorders Montage of cross-sectional MR images representing the spectrum of findings on neuroimaging. I-1 (A–D), I-4 (E and F), I-9 (G and H), I-12 (I and J), I-14 (K and L), I-16 (M and N), I-17 (O and P), I-18 (Q and R). The primary abnormalities included global cerebellar hypoplasia and atrophy (I-1, I-4, I-5, I-6, I-7, I-8, I-9), cerebellar hypoplasia (I-2, I-3), callosal hypoplasia (I-3, I-4, I-6), and underdevelopment and/or atrophy of the frontotemporal regions with involvement of the opercula (all individuals).

Figure 4

Suppressing expression of the Drosophila GTPBP1/2 homolog CG2017 reduces locomotor activity (A) Schematic of the CG2017 locus. Gray regions: 3′ (left) and 5′ (right) UTRs. White regions: coding exons. Positions of the two piggyBac elements studied in this work are noted. Both leave the 5′ UTRs of CG2017 isoforms A and B intact, while likely impacting transcription from upstream 5′ UTRs (isoforms C–E). (B) qPCR analysis of expression of isoform C of CG2017 in adult male iso31 controls and in CG2017^0758-G4 and CG2017^0269-G4 homozygotes. (C) Gal4 encoded within the Pbac [0758-G4] and Pbac [0269-G4] elements was used to report neural CG2017 expression by driving transcription of membrane-targeted GFP (CD8::GFP). Scale bar, 100 μm. (D–F) Activity levels across a 12 h light:dark period in iso31 controls (D) and in CG2017^0758-G4 (E) and CG2017^0269-G4 (F) homozygotes. ZT: zeitgeber time. Arrows in (D) point to periods of peak activity in iso31 flies following lights-on and lights-off transitions. Arrows at equivalent positions in (E) and (F) illustrate changes in peak activity in CG2017^0758-G4 and CG2017^0269-G4 homozygotes. (G and H) Total beam breaks in iso31 control males, and in CG2017^0758-G4 and CG2017^0269-G4 homozygotes, measured between ZT0–1 (G) or ZT12–13 (H). Population sizes for (D)–(H) are as follows: Iso31, N = 47; CG2017^0758-G4, N = 10; CG2017^0269-G4, N = 31. (I and J) Total beam breaks in CG2017^0758-G4/CG2017^0269-G4 trans-heterozygous flies (N = 53) and heterozygote controls for each insertion (CG2017^0758-G4/+, N = 33; CG2017^0269-G4/+, N = 31), measured between ZT0–1 (I) or ZT12–13 (J). (K and L) Total beam break in flies co-expressing CG2017 shRNA and Dicer-2 in neurons (elav > UAS-CG2017 shRNA, UAS-dcr-2; N = 41) and controls (elav > UAS-dcr-2; N = 44; + > UAS-CG2017 shRNA: N = 15) measured between ZT0–1 (K) or ZT12–13 (L). ^∗p < 0.05, ^∗∗∗p < 0.0005, Kruskal-Wallis test with Dunn’s post-hoc test (G, H, J, L) or one-way ANOVA with Tukey’s multiple comparisons test (I, K).