De Novo Variants in CDK19 Are Associated with a Syndrome Involving Intellectual Disability and Epileptic Encephalopathy

Abstract

We identified three unrelated individuals with de novo missense variants in CDK19, encoding a cyclin-dependent kinase protein family member that predominantly regulates gene transcription. These individuals presented with hypotonia, global developmental delay, epileptic encephalopathy, and dysmorphic features. CDK19 is conserved between vertebrate and invertebrate model organisms, but currently abnormalities in CDK19 are not known to be associated with a human disorder. Loss of Cdk8, the fly homolog of CDK19, causes larval lethality, which is suppressed by expression of human CDK19 reference cDNA. In contrast, the CDK19 p.Tyr32His and p.Thr196Ala variants identified in the affected individuals fail to rescue the loss of Cdk8 and behave as null alleles. Additionally, neuronal RNAi-mediated knockdown of Cdk8 in flies results in semi-lethality. The few eclosing flies exhibit severe seizures and a reduced lifespan. Both phenotypes are fully suppressed by moderate expression of the CDK19 reference cDNA but not by expression of the two variants. Finally, loss of Cdk8 causes an obvious loss of boutons and synapses at larval neuromuscular junctions (NMJs). Together, our findings demonstrate that human CDK19 fully replaces the function of Cdk8 in the fly, the human disease-associated CDK19 variants behave as strong loss-of-function variants, and deleterious CDK19 variants underlie a syndromic neurodevelopmental disorder.

Keywords: Cdk8; Drosophila; West syndrome; bang sensitivity; de novo; dominant variants; genetic disease; infantile spasms; rare disease; seizure.

Figure 1

Clinical Features of Probands (A) Front view of proband 1 at age 21 years. Distinctive features include thin, sparse, arched eyebrows; curly hair; mild hypotelorism; a prominent bulbous nose; a wide mouth with wide spaced teeth; a thin upper lip; and a long smooth philtrum. (B) Facial profile of proband 1 at age 23 years shows midface retrusion. (C) Photograph of proband 1 from the newborn period shows mild hypotelorism and fleshy nose. (D) Dysmorphic facial features, including hypotelorism and a prominent nose with a bulbous tip, were observed in proband 2. (E and F) T1 (E) and T2 (F) brain MRI images for proband 2 show mild atrophy. (G) Dysmorphic features, including ocular hypertelorism; a prominent nose with a bulbous tip,;U-shaped vermillion and arched upper lip; and a large mouth were observed in proband 3. (H and I) T1 (H) and T2 (I) brain MRI images for proband 3 show delayed myelination.

Figure 2

Cdk8 Is Functional Fly Homolog of Human CDK19 (A) The domain that includes the variants is fully conserved from fly to human. (B) Strong ubiquitous expression of CDK19 reference is not toxic, but expression of the variants is toxic to flies. Numbers indicate the ratio of observed/expected flies. (C) Ubiquitous expression of reference CDK19 rescued the larval lethality observed in Cdk8^K185/Cdk8 Df flies, whereas the expression of variants CDK19 failed to rescue the larval lethality. Numbers indicate the ratio of observed/expected flies.

Figure 3

Loss of Cdk8 Resulted in Lethality, Seizure, and Lifespan Decrease in Flies (A) Ubiquitous expression of Cdk8 RNAi caused lethality, and it can be rescued by co-expression of human reference CDK19, whereas it cannot be rescued by co-expression of CDK19 variants (n = 8 crosses per each genotype). Statistical analyses were performed via one-way ANOVA followed by a Tukey post-hoc test. Results are means ± SEM, ^∗∗∗p < 0.001. (B) Neuronal expression of Cdk8 RNAi caused severe lethality, but glial expression of Cdk8 RNAi did not cause any lethality. Co-expression of human reference CDK19 rescued the lethality, whereas co-expression of CDK19 variants failed to rescue the lethality (n = 9 crosses per each genotype). Statistical analyses were performed via one-way ANOVA followed by a Tukey post-hoc test. Results are means ± SEM, ^∗∗∗p < 0.001; n.s., not significant. (C) Flies that lost Cdk8 in neurons exhibit strong bang sensitivity, which can be fully rescued by expression of human reference CDK19, whereas they fail to be rescued by expression of the variants. Statistical analyses were performed via one-way ANOVA followed by a Tukey post-hoc test. Results are means ± SEM, ^∗∗∗p < 0.001; n.s., not significant. (D) Lifespan of flies that co-express human reference CDK19 or the variants with Cdk8 RNAi (n > 50 per each genotype).

Figure 4

CDK19 Is Localized to the Perinuclear Space and Cytoplasm of Most Neurons but Expressed in the Nucleus in a Few Neurons (A) Localization of CDK19 in humanized flies (elav > Cdk8 RNAi; UAS-CDK19 reference). Dashed boxes indicate the region that are magnified in images below. White arrow indicates the neuron that expresses CDK19 in nucleus. Scale bar: 10 μm. (B) Localization of variant CDK19 in the flies (elav > Cdk8 RNAi; UAS-CDK19 p.Thr196Ala). (C) Localization of variant CDK19 in the flies (elav > Cdk8 RNAi; UAS-CDK19 p.Tyr32His). Scale bar: 10 μm.

Figure 5

Loss of Cdk8 Caused a Severe Synapse Loss in Larval NMJs (A) Representative images of larval NMJs of each genotype (Control, elav > luciferase RNAi [top]; Cdk8 loss, elav > Cdk8 RNAi;UAS-LacZ [middle]; and rescued flies, elav > Cdk8 RNAi;UAS-CDK19 reference [bottom]). (B) Quantification of the number of total boutons in NMJs (n = 14 [control], n = 23 [Cdk8 loss], n = 25 [rescued flies]). Statistical analyses were performed via one-way ANOVA followed by a Tukey post-hoc test. Results are means ± SEM, ^∗∗∗p < 0.001. (C) Quantification of the length of boutons of NMJs (n = 14 [control], n = 23 [Cdk8 loss], n = 25 [rescued flies]). Statistical analyses were performed via one-way ANOVA followed by a Tukey post-hoc test. Results are means ± SEM, n.s., not significant.