Mutational Landscapes and Phenotypic Spectrum of SWI/SNF-Related Intellectual Disability Disorders

Abstract

Mutations in genes that encode proteins of the SWI/SNF complex, called BAF complex in mammals, cause a spectrum of disorders that ranges from syndromic intellectual disability to Coffin-Siris syndrome (CSS) to Nicolaides-Baraitser syndrome (NCBRS). While NCBRS is known to be a recognizable and restricted phenotype, caused by missense mutations in SMARCA2, the term CSS has been used lately for a more heterogeneous group of phenotypes that are caused by mutations in either of the genes ARID1B, ARID1A, ARID2, SMARCA4, SMARCB1, SMARCE1, SOX11, or DPF2. In this review, we summarize the current knowledge on the phenotypic traits and molecular causes of the above named conditions, consider the question whether a clinical distinction of the phenotypes is still adequate, and suggest the term "SWI/SNF-related intellectual disability disorders" (SSRIDDs). We will also outline important features to identify the ARID1B-related phenotype in the absence of classic CSS features, and discuss distinctive and overlapping features of the SSRIDD subtypes. Moreover, we will briefly review the function of the SWI/SNF complex in development and describe the mutational landscapes of the genes involved in SSRIDD.

Keywords: ARID1A; ARID1B; BAF complex; Coffin-Siris syndrome; Nicolaides-Baraitser syndrome; SMARCA2; SMARCA4; SWI/SNF complex.

FIGURE 1

Diagram of SWI/SNF dependent chromatin remodeling. The SWI/SNF complex binds to DNA and histones and effectuates nucleosome displacement in order to enhance DNA accessibility and thereby initiation of the transcription machinery. SWI/SNF has a direct, SMARCA4-dependent PRC eviction capacity. PRC, polycomb repressive complex; TF, transcription factor.

FIGURE 2

Flow chart of SSRIDD phenotypic spectrum ranging from syndromic ID over classic and atypical/severe CSS to NCBRS. Clinical categories within the spectrum are associated with specific SWI/SNF-associated genes.

FIGURE 3

Concept chart of SSRIDD genes and their mutational characteristics and constraint metrics. LoF, loss of function; HI, haploinsufficiency; pLI, probability of a gene being LoF intolerant.

FIGURE 4

Proportions of genes mutated in SSRIDD. (A) ARID1B is the most commonly mutated gene in CSS. Pie-chart of percentages of genes mutated in patients with CSS from studies Tsurusaki et al. (2012), Santen et al. (2013), Wieczorek et al. (2013), and Tsurusaki et al. (2014b) (n = 103). (B) Pie-chart of all point mutations in the nine genes associated with SSRIDD described in the literature. Number of patients with a mutation in each gene. Numbers serve as an overview and do not reflect mutation detection rates, because the underlying studies used very differentially selected patient collectives (n = 190).

FIGURE 5

Overview of ARID family protein structure and distribution of mutations across the proteins. (A) Schematic representation of ARID1B. Mutations depicted from LOVD ARID1B database. Protein structure according to uniprot ID Q8NFD5. (B) Schematic representation of ARID1A. Mutations according to Tsurusaki et al. (2012); Santen et al. (2013), and Wieczorek et al. (2013). Protein structure according to uniprot ID O14497. (C) Schematic representation of ARID2. Mutations according to Shang et al. (2015) and Bramswig et al. (2017). The red line indicates the intragenic deletion reported by Van Paemel et al. (2017). Protein structure according to uniprot ID Q68CP9. Blue circle, frameshift mutation; red triangle, stop mutation; green circle, missense mutation; purple triangle, splice site mutation. LXXLL, nuclear receptor recognition motif; NLS, nuclear localization signal; RFX, RFX-type winged-helix; GLN, glycine rich region; ZNF, zinc finger.

FIGURE 6

Overview of SMARC family protein structure and distribution of mutations across the proteins. (A) Schematic representation of SMARCA2. Mutations according to Sousa et al. (2014), Sánchez and Rojas (2017). Protein structure according to uniprot ID P51531. (B) Schematic representation of SMARCA4. Mutations according to Tsurusaki et al. (2012), Santen et al. (2013), Tsurusaki et al. (2014b), and Tzeng et al. (2014). Protein structure according to uniprot ID P51532. (C) Schematic representation of SMARCB1. Mutations according to Tsurusaki et al. (2012), Santen et al. (2013), Tsurusaki et al. (2014b), Wieczorek et al. (2013), and Gossai et al. (2015). Protein structure according to uniprot ID Q12824. (D) Schematic representation of SMARCE1. Mutations according to Tsurusaki et al. (2012), Santen et al. (2013), Wieczorek et al. (2013), and Zarate et al. (2016). Protein structure according to uniprot ID Q969G3. Green circle, missense mutation; blue hexagon, in-frame deletion; red triangle, stop mutation; red bars, in-frame deletion. QLQ, Gln, Leu, Gln motif; has, RFX-type winged-helix; HELA, ATP-Helicase; HELC, Helicase C-terminal; Bromo, bromodomain; HMG, high mobility group box; CCD, coiled coil domain.

FIGURE 7

Overview of SOX11 and DPF2 protein structure and distribution of mutations across the proteins. (A) Schematic representation of SOX11. Point mutations according to Tsurusaki et al. (2014a), Hempel et al. (2016), Okamoto et al. (2017), and Khan et al. (2018). Protein structure according to uniprot ID P35716. Deletions not shown due to their size. (B) Schematic representation of DPF2. Mutations according to Vasileiou et al. (2018). Protein structure according to uniprot ID Q92785. Blue circle, frameshift mutation; red triangle, stop mutation; green circle, missense mutation; purple triangle, splice site mutation. HMG, high mobility group box; C2H2, zinc finger C2H2-type; PHD, plant homeo domain.