MSL2 variants lead to a neurodevelopmental syndrome with lack of coordination, epilepsy, specific dysmorphisms, and a distinct episignature

Abstract

Epigenetic dysregulation has emerged as an important etiological mechanism of neurodevelopmental disorders (NDDs). Pathogenic variation in epigenetic regulators can impair deposition of histone post-translational modifications leading to aberrant spatiotemporal gene expression during neurodevelopment. The male-specific lethal (MSL) complex is a prominent multi-subunit epigenetic regulator of gene expression and is responsible for histone 4 lysine 16 acetylation (H4K16ac). Using exome sequencing, here we identify a cohort of 25 individuals with heterozygous de novo variants in MSL complex member MSL2. MSL2 variants were associated with NDD phenotypes including global developmental delay, intellectual disability, hypotonia, and motor issues such as coordination problems, feeding difficulties, and gait disturbance. Dysmorphisms and behavioral and/or psychiatric conditions, including autism spectrum disorder, and to a lesser extent, seizures, connective tissue disease signs, sleep disturbance, vision problems, and other organ anomalies, were observed in affected individuals. As a molecular biomarker, a sensitive and specific DNA methylation episignature has been established. Induced pluripotent stem cells (iPSCs) derived from three members of our cohort exhibited reduced MSL2 levels. Remarkably, while NDD-associated variants in two other members of the MSL complex (MOF and MSL3) result in reduced H4K16ac, global H4K16ac levels are unchanged in iPSCs with MSL2 variants. Regardless, MSL2 variants altered the expression of MSL2 targets in iPSCs and upon their differentiation to early germ layers. Our study defines an MSL2-related disorder as an NDD with distinguishable clinical features, a specific blood DNA episignature, and a distinct, MSL2-specific molecular etiology compared to other MSL complex-related disorders.

Keywords: MSL2; autism; connective tissue; epigenetics; epilepsy; episignature; iPSC; male-specific lethal complex; neurodevelopmental syndrome.

Graphical abstract

Figure 1

Photos of individuals with the MSL2-related disorder (A) Individual 4 (p.Gln171^∗). Note prognathism as well as broad and high forehead. (B) Individual 7 (p.Gln353^∗). Note high forehead and retrognathia. (C) Individual 10 (p.Arg39Leufs^∗34). Note high forehead and prominent ridge of the nose. (D) Individual 21 (p.Leu266Valfs^∗5). Note mild telecanthus and retrognathia. (E) Individual 25 (complex variant). Note downslanting palpebral fissures. (F) Individual 1 (p.Met1?) fingers. Note thumb and 5th finger leading to a Beighton score of 4. Note mild malar hypoplasia in all facial photos.

Figure 2

Variant location along MSL2 (A) The variants studied in our cohort are represented on the MSL2 protein, excluding the complex variant of individual 25. The active regions are named and highlighted in different colors according to the legend. Variant types are also indicated, and if the same one was shared by multiple individuals, their total number is in a circle. Illustration adapted from ProteinPaint and domain sequences obtained from Pfam. (B) Amino acid conservation of the two missense variants (individuals 2 and 3). (C) Homology model of human MSL2 as experimentally crystalized (PDB: 4B86), represented by UCSF Chimera, from two angles. MSL2 is in yellow, the missense variants of our cohort in red, the zinc finger domain in blue, the two zinc atoms in fuchsia, and the MSL1 protein in green. MSL2 is illustrated from the 1^st to the 115^th residues, so Met317 is not represented.

Figure 3

HPO analysis of the MSL2-related disorder cohort (A) Hierarchical clustering with Ward method of the phenotypes of our 25 probands. (B) Gap statistic curve serving to determine the optimal number of clusters (vertical dashed line); error bars represent the uncertainty in the estimation of the Gap statistic parameter (k). (C) Quantitative phenotypic similarity among individuals based on HAC analysis.

Figure 4

Verification of the identified episignature and MVP scores generated by the SVM classifier (A) The heatmap depicts hierarchical clustering, where the rows represent the 239 selected probes and the columns represent the affected and control individuals. The case and control samples are indicated by the colors red and blue, respectively. The color gradient ranges from blue (representing no methylation or 0) to red (representing full methylation or 1). (B) MDS plot, where the MSL2 case samples are represented by red circles and the control individuals by blue circles. Both plots (A and B) clearly demonstrate a distinct separation between the case and control groups. (C) The SVM-generated MVP scores ranging from 0 to 1, indicating similarity to the MSL2 episignature, with high scores indicating greater similarity. Blue circles represent training samples, while gray circles represent testing samples. The low MVP score of testing control samples and testing samples from other rare genetic disorders, except for two individuals, suggests high specificity of the classifier. ADCADN, cerebellar ataxia, deafness, and narcolepsy, autosomal dominant; ARTHS, Arboleda-Tham syndrome; ATRX, X-linked alpha-thalassemia/impaired intellectual development syndrome; AUTS18, autism, susceptibility to, 18; BAFopathy, Coffin-Siris syndrome-1,2,3,4 (CSS1,2,3,4) and Nicolaides-Baraitser syndrome (NCBRS); BEFAHRS, Beck-Fahrner syndrome; BFLS, Borjeson-Forssman-Lehmann syndrome; BIS, blepharophimosis intellectual disability SMARCA2 syndrome; CdLS, Cornelia de Lange syndromes 1,2,3,4 (CdLS1,2,3,4); CHARGE, CHARGE syndrome; Chr16p11.2del, chr16p11.2 deletion syndrome, 593-KB; CSS_c.6200, Coffin-Siris syndrome-1,2 (CSS1,2); CSS4_c.2656, Coffin-Siris syndrome-4 (CSS4); Down, Down syndrome; Dup7, Williams-Beuren duplication syndrome (chr7q11.23 duplication syndrome); DYT28, dystonia 28, childhood-onset; EEOC, epileptic encephalopathy, childhood-onset; FLHS, Floating Harbor syndrome; GADEVS, Gabriele-de Vries syndrome; GTPTS, genitopatellar syndrome; HMA, Hunter McAlpine craniosynostosis syndrome; HVDAS_C and HVDAS_T, Helsmoortel-van der Aa syndrome; ICF_1, immunodeficiency-centromeric instability-facial anomalies syndrome 1 (ICF1); ICF_2_3_4, immunodeficiency-centromeric instability-facial anomalies syndromes 2,3,4 (ICF2,3,4); IDDSELD, intellectual developmental disorder with seizures and language delay; Kabuki, Kabuki syndrome 1,2 (KABUK1,2); KDVS, Koolen de Vreis syndrome; Kleefstra, Kleefstra syndrome 1 (KLEFS1); LLS, Luscan-Lumish syndrome; MKHK_ID4, Menke-Hennekam syndromes 1,2 (MKHK1,2); MLASA2, myopathy, lactic acidosis, and sideroblastic anemia 2; MRD23, intellectual developmental disorder, autosomal dominant 23; MRD51, intellectual developmental disorder, autosomal dominant 51; MRX93, intellectual developmental disorder, X-linked 93; MRX97, intellectual developmental disorder, X-linked 97; MRXSA, intellectual developmental disorder, X-linked, syndromic, Armfield type; MRXSCJ, intellectual developmental disorder, X-linked, syndromic, Claes-Jensen type; MRXSN, intellectual developmental disorder, X-linked syndromic, Nascimento-type; MRXSSR, intellectual developmental disorder, X-linked, Snyder-Robinson type; PHMDS, Phelan-McDermid syndrome; PRC2, Cohen-Gibson syndrome (COGIS) and Weaver syndrome (WVS); RENS1, Renpenning syndrome; RMNS, Rahman syndrome; RSTS, Rubinstein-Taybi syndrome; SBBYSS, Ohdo syndrome, SBBYSS variant; Sotos, Sotos syndrome 1; TBRS, Tatton-Brown-Rahman syndrome; VCFS, velocardiofacial syndrome; WDSTS, Wiedemann-Steiner syndrome; WHS, Wolf-Hirschhorn syndrome; Williams, Williams-Beuren deletion syndrome.

Figure 5

Generation and molecular characterization of induced pluripotent stem cell (iPSC) lines from individuals with the MSL2-related disorder (A) Schematic highlighting the MSL2 variants of the three probands whose dermal fibroblasts have been reprogrammed into iPSCs. (B) Immunofluorescence images showing the expression of pluripotency markers (TRA-1060, SOX2, OCT4, and SSEA4) in patient-derived iPSC clones, two clones per individual. Scale bars, 100 μm. (C) Schematic depicting the predicted outcome of the variants on MSL2. Heterozygous de novo frameshift (ST) or nonsense (IX) variants lead to the generation of early stop codons leading to truncation of MSL2 whereas the intact allele would produce the full-length MSL2. The missense variant (MV) would not alter the size of the translated MSL2 with the variant. (D) Immunoblotting on nuclear soluble extracts from patient-derived iPSC clones showed lower levels of full-length MSL2 in ST and IX compared to a control (C) iPSC line. Note the comparable levels of detection of MSL2 in MV to the C replicates. The asterisk shows unspecific bands on the MSL2 blot. MOF levels are similar across patient-derived clones compared to Cs. DHX9 is used as the loading control. Immunoblotting on nuclear insoluble fractions shows similar levels of H4K16ac in all affected-individual-derived clones compared to the control. H4 is used as the loading control for the insoluble fractions.

Figure 6

Analysis of the expression of MSL2 target genes in affected-individual-derived iPSCs and early germ layers (A) Top, schematic of the differentiation timeline of iPSCs into all three germ layers. Bottom, RT-qPCR results showing the relative RNA expression (normalized to RPLP0, arbitrary units [a.u.]) of lineage-specific marker genes for ectoderm (PAX6, NES), mesoderm (TBXT, CXCR4), and endoderm (SOX17, FOXA2). Note that the expression of each marker gene is enriched for its respective lineage across all affected-individual-derived clones compared to the control indicating successful differentiation onto all germ layers. n = 3 per clone per differentiation. (B) RT-qPCR results showing the log2(fold change) of relative RNA expression of putative mammalian targets of MSL2 (ZNF185, BEX2, TSIX, BSCL2, GNG3) in patient-derived iPSCs (bar graphs in respective colors assigned to individuals) compared to control (gray). Expression levels are compared in iPSCs (top, n = 4 per clone) as well as upon trilineage differentiation to ectoderm, mesoderm, and endoderm (n = 3 per clone per lineage). Two-way ANOVA with post-hoc Fisher’s LSD test, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001. Data are represented as the mean ± SEM.