A description of novel variants and review of phenotypic spectrum in UBA5-related early epileptic encephalopathy

Abstract

Early infantile epileptic encephalopathy-44 (EIEE44, MIM: 617132) is a previously described condition resulting from biallelic variants in UBA5, a gene involved in a ubiquitin-like post-translational modification system called UFMylation. Here we report five children from four families with biallelic pathogenic variants in UBA5 All five children presented with global developmental delay, epilepsy, axial hypotonia, appendicular hypertonia, and a movement disorder, including dystonia in four. Affected individuals in all four families have compound heterozygous pathogenic variants in UBA5 All have the recurrent mild c.1111G > A (p.Ala371Thr) variant in trans with a second UBA5 variant. One patient has the previously described c.562C > T (p. Arg188*) variant, two other unrelated patients have a novel missense variant, c.907T > C (p.Cys303Arg), and the two siblings have a novel missense variant, c.761T > C (p.Leu254Pro). Functional analyses demonstrate that both the p.Cys303Arg variant and the p.Leu254Pro variants result in a significant decrease in protein function. We also review the phenotypes and genotypes of all 15 previously reported families with biallelic UBA5 variants, of which two families have presented with distinct phenotypes, and we describe evidence for some limited genotype-phenotype correlation. The overlap of motor and developmental phenotypes noted in our cohort and literature review adds to the increasing understanding of genetic syndromes with movement disorders-epilepsy.

Keywords: central hypotonia; epileptic encephalopathy; generalized dystonia; limb hypertonia; profound global developmental delay; severe global developmental delay.

Figure 1.

In the first step of UFMylation, UFM1 is activated by UBA5 (E1-like enzyme) and forms a thioester bond with the Cys250 residue of UBA5. In the second step, UFC1 (E2-like enzyme) binds to UBA5, and UFM1 is transferred to the Cys116 residue of UFC1 by a trans-esterification reaction. In the third step, the UFL1 complex (E3-like enzyme) brings a substrate to the UFC1–UFM1 complex, and UFM1 is covalently conjugated to the substrate. The functional studies performed in this study measure the presence of the UFM1–UBA5 and UFM1–UFC1 conjugates.

Figure 2.

Brain MRI images: Patient 1 and control. Vertically oriented right hippocampus and blurring of the stratum radiatum in Patient 1 (top left, arrow) and age-matched control with normal hippocampi (top right). Attenuated subcortical U-fibers compared to white matter in Patient 1 (middle left, two white arrows) and age-matched normal control (middle right). Diminutive thalami in Patient 1 (bottom left, two white arrows) and normal thalami in age-matched control (bottom right).

Figure 3.

Immunoblot and in vitro thioester formation assays. (A,B) Immunoblot assay of UBA5 mutants in UBA5^−/− HEK293T cells. Indicated constructs (0.1 µg for UBA5 or UBA5 mutant, 0.5 µg for UFC1 or UFC mutant, and 2 µg for UFM1ΔC2) were expressed in UBA5-deficient HEK293T cells. Twenty-four hours after transfection, the cell lysates were subjected to immunoblot analysis with indicated antibodies. Bar graphs indicate the quantitative densitometric analyses of FLAG-UBA5-MYC-UFM1ΔC2 and FLAG-UFC1-MYC-UFM1ΔC2 intermediates relative to free FLAG-UBA5 and FLAG-UFC1, respectively. Statistical analyses were performed using the unpaired t-test (Welch test). The data represents the means ± SE of three separate experiments. (**) P < 0.01. (C,D) In vitro thioester formation assay of UFM1 by UBA5 (C) and of UFM1 by UFC1 (D). The assay was conducted as described in Methods. Data shown are representative of three separate experiments.