Molecular analysis, pathogenic mechanisms, and readthrough therapy on a large cohort of Kabuki syndrome patients

Abstract

Kabuki syndrome (KS) is a multiple congenital anomalies syndrome characterized by characteristic facial features and varying degrees of mental retardation, caused by mutations in KMT2D/MLL2 and KDM6A/UTX genes. In this study, we performed a mutational screening on 303 Kabuki patients by direct sequencing, MLPA, and quantitative PCR identifying 133 KMT2D, 62 never described before, and four KDM6A mutations, three of them are novel. We found that a number of KMT2D truncating mutations result in mRNA degradation through the nonsense-mediated mRNA decay, contributing to protein haploinsufficiency. Furthermore, we demonstrated that the reduction of KMT2D protein level in patients' lymphoblastoid and skin fibroblast cell lines carrying KMT2D-truncating mutations affects the expression levels of known KMT2D target genes. Finally, we hypothesized that the KS patients may benefit from a readthrough therapy to restore physiological levels of KMT2D and KDM6A proteins. To assess this, we performed a proof-of-principle study on 14 KMT2D and two KDM6A nonsense mutations using specific compounds that mediate translational readthrough and thereby stimulate the re-expression of full-length functional proteins. Our experimental data showed that both KMT2D and KDM6A nonsense mutations displayed high levels of readthrough in response to gentamicin treatment, paving the way to further studies aimed at eventually treating some Kabuki patients with readthrough inducers.

Keywords: KDM6A; KMT2D; Kabuki syndrome; haploinsufficiency; readthrough.

Figure 1

KMT2D methyltransferase activity is impaired in KS lymphoblastoid and fibroblast cell lines with KMT2D-truncating mutations. (A–F): Immunoblotting analysis by using anti-KMT2D antibody on whole protein lysate from six (A and B) KS lymphoblastoid (KB41, KB44, KB45, and KB82, KB83, KB153, and KB186) and three (C) fibroblast cell lines (KB186, KB153, and KB3) with nonsense mutations and frameshift mutations, compared with control cell lines. Pool: pooled protein lysates from two control cell lines. D–F: The density of each band was determined by densitometer. The expression level of KMT2D was determined by calculating the protein level for each sample, normalized to the corresponding GAPDH level. G: qPCR was performed to measure the expression of known target genes of KMT2D (HOXC6, S100A2, S100A4, S100A5, S100A6) in six KS–LCLs, compared with control cell lines. Scale bars represent standard errors. *P < 0.01.

Figure 2

Effect of KMT2D haploinsufficiency on E2-induced expression of HOXC6. A and B: HOXC6 transcriptional level was determined in three KS patient fibroblast cell lines (KB153, KB186, and KB3) exposed to 100 nM E2 for 8 hr or not by using qPCR analysis, in comparison to a pool of two normal fibroblast cell lines. C–E: HOXC6 promoter fragment spanning the ERE1–ERE2 regions was cloned into a luciferase-based reporter construct (pGL3) and transfected into three KS patient fibroblast cell lines and into two normal fibroblast cell lines along with a Renilla luciferase construct used as an internal transfection control. Cells were then treated with 100 nM E2 (E) for 8 hr and subjected to luciferase assay. The luciferase activities (normalized to Renilla activity) in the presence or not of E2 over normal fibroblast cell lines were plotted. The experiment with three replicates was repeated at least thrice. Scale bars represent standard errors. *P < 0.01.

Figure 3

Identification of KMT2D and KDM6A nonsense mutations responsive to gentamicin treatment. A and B: Comparison between basal and gentamicin-induced readthrough level. Readthrough efficiency for 14 nonsense mutations in the KMT2D gene and for two elsewhere published [Miyake et al., 2013b] nonsense mutations in the KDM6A gene was assessed in HEK293 cells exposed or not to 600–800–1,200 μg/ml of gentamicin for 48 hr. A pCRFL reporter vector harboring the 319d Duchenne muscular dystrophy mutation was used as positive control (CNT). Each value corresponds to the mean of four to six independent experiments. Scale bars represent standard errors. C and D: Gentamicin restored the expression of HOXC6 and S100A4 in patient cultured cells with different efficiency. qPCR analysis was performed to evaluated HOXC6 and S100A4 expression level in two patients LCLs, KB41 (C) and KB45 (D), and two control cell lines treated or not with 800 μg/ml of gentamicin for 48 hr. The results are the mean from at least two experiments performed in duplicate. *P < 0.01.