TBC1D24 genotype-phenotype correlation: Epilepsies and other neurologic features

Abstract

Objective: To evaluate the phenotypic spectrum associated with mutations in TBC1D24.

Methods: We acquired new clinical, EEG, and neuroimaging data of 11 previously unreported and 37 published patients. TBC1D24 mutations, identified through various sequencing methods, can be found online (http://lovd.nl/TBC1D24).

Results: Forty-eight patients were included (28 men, 20 women, average age 21 years) from 30 independent families. Eighteen patients (38%) had myoclonic epilepsies. The other patients carried diagnoses of focal (25%), multifocal (2%), generalized (4%), and unclassified epilepsy (6%), and early-onset epileptic encephalopathy (25%). Most patients had drug-resistant epilepsy. We detail EEG, neuroimaging, developmental, and cognitive features, treatment responsiveness, and physical examination. In silico evaluation revealed 7 different highly conserved motifs, with the most common pathogenic mutation located in the first. Neuronal outgrowth assays showed that some TBC1D24 mutations, associated with the most severe TBC1D24-associated disorders, are not necessarily the most disruptive to this gene function.

Conclusions: TBC1D24-related epilepsy syndromes show marked phenotypic pleiotropy, with multisystem involvement and severity spectrum ranging from isolated deafness (not studied here), benign myoclonic epilepsy restricted to childhood with complete seizure control and normal intellect, to early-onset epileptic encephalopathy with severe developmental delay and early death. There is no distinct correlation with mutation type or location yet, but patterns are emerging. Given the phenotypic breadth observed, TBC1D24 mutation screening is indicated in a wide variety of epilepsies. A TBC1D24 consortium was formed to develop further research on this gene and its associated phenotypes.

Figure 1. Genetic and phenotypic heterogeneity

The diagram illustrates the exonic structure of TBC1D24 isoform 1 (NM_001199107.1), with the introns as thin lines not drawn to scale. The noncoding exonic regions are drawn in gray, and the coding regions thicker and in color (orange for Rab GTPase-activating protein [Rab-GAP] domain, blue for TLDc domain, and yellow for the rest). The location of the mutations identified in various epilepsy syndromes is shown, according to the severity of the epilepsy phenotype. No clear pattern of genotype–phenotype correlation emerges. Circle = myoclonic epilepsy; square = generalized epilepsy; triangle = focal epilepsy; diamond = early-onset epileptic encephalopathy; hexagon = unclassified epilepsy. Black = death; red = drug-resistant epilepsy; blue = drug-responsive epilepsy or seizure-free. D = DOORS syndrome. Black arrows = missense mutations; red arrows = frameshift mutations; gray arrows = nonsense mutations; blue arrows = splice-site mutations; * = recurrent mutations.

Figure 2. Disease-causing mutations in evolutionarily conserved motifs

(A) Seven highly conserved motifs are identified in TBC1D24/Sky in an interspecies comparison against related Tre2/Bub2/Cdc16 (TBC) domain proteins. The Rab GTPase-activating protein (Rab-GAP) TBC domain is drawn in orange, the TLDc domain is drawn in blue, and the rest is in yellow, as in figure 1. (B) The consensus sequence of the TBC domain motif 2 contains an arginine residue (*) that is substituted in the most frequent pathogenic mutation seen in patients with deafness, onychodystrophy, osteodystrophy, mental retardation, and seizures. Further residues affected by pathogenic mutations are indicated by an asterisk above the residue. (C) Motif 1 in the TBC domain, motif 3 in the region between the TBC and TLDc domains, as well as motifs 4 and 7 in the TLDc domain contain further residues affected by pathogenic mutations, as indicated by an asterisk.

Figure 3. Neurite outgrowth associated with TBC1D24 mutants

Primary mouse cortical cells were transfected with wild-type (WT) or mutant TBC1D24 constructs as indicated and representative images of transfected neurons are shown. Quantification of neurite length shows that expression of Arg242Cys (R242C) and Arg360Leu (R360L), but not Arg40Leu (R40L) or Arg270Cys (R270C), results in a significantly reduced induction of outgrowth compared to WT. ***p < 0.001, 1-way analysis of variance; scale bar = 10 μM.

Figure 4. Phenotypic spectrum of patients with TBC1D24 mutation

The graph displays the age at epilepsy onset along the horizontal axis, the categorical outcome varying from seizure control to death along the vertical axis. The degree of intellectual disability is represented by the different fill of the circles, varying from white to black.