Variants in WASHC3, a component of the WASH complex, cause short stature, variable neurodevelopmental abnormalities, and distinctive facial dysmorphism

Abstract

Purpose: Genetic defects that impair growth plate chondrogenesis cause a phenotype that varies from skeletal dysplasia to mild short stature with or without other syndromic features. In many individuals with impaired skeletal growth, the genetic causes remain unknown.

Method: Exome sequence was performed in 3 unrelated families with short stature, distinctive facies, and neurodevelopmental abnormalities. The impact of identified variants was studied in vitro.

Results: Exome sequencing identified variants in WASHC3, a component of the WASH complex. In the first family, a de-novo-dominant missense variant (p.L69F) impaired WASHC3 participation in the WASH complex, altered PTH1R endosomal trafficking, diminished PTH1R signaling, and affected growth plate chondrocyte hypertrophic differentiation, providing a likely explanation for the short stature. Knockdown of other WASH complex components also diminished PTH1R signaling. In the second and third families, a homozygous variant in the start codon (p.M1?) markedly reduced WASHC3 protein expression.

Conclusion: In combination with prior studies of WASH complex proteins, our findings provide evidence that the WASH complex is required for normal skeletal growth and that, consequently, genetic abnormalities impairing the function of the WASH complex (WASHopathy) cause short stature, as well as distinctive facies and variable neurodevelopmental abnormalities.

Keywords: Growth; Neurodevelopmental abnormalities; Receptor trafficking; WASH complex; WASHC3.

Figure 1

Pedigrees of families with short stature and distinctive facies. A. Family 1. B. Family 2. Closed symbols, affected family members; open symbols, unaffected family members; black arrows, proband; double line indicates consanguinity. Height and height SDS are shown beneath each symbol. C. Family 3 pedigree; black arrows, proband. D. Distinctive facial dysmorphism. E. Lateral face to show low-set ear; F. Short stature and cubitus valgus. G. brachydactyly. H and I. Family 2 younger brother’s facial features with distinctive facial dysmorphism. I. lateral face to show low-set ear. J. Skeletal x-ray of the left hand in younger affected brother. A cone-shaped epiphysis is indicated with a yellow open arrow. K. Skeletal x-ray of right tibia in proband. An incidental cystic lesion is indicated with a yellow open arrow. L. Coronal T2 image showing left anterior temporal arachnoid cyst (white arrow). M. Midsagittal T1 image showing giant cisterna magna. The cortex, corpus callosum, and brainstem are normal. N and O. Family 3, distinctive facial dysmorphism. P. Brachydactyly. SDS, standard deviation score; WT, wild type.

Figure 1

Pedigrees of families with short stature and distinctive facies. A. Family 1. B. Family 2. Closed symbols, affected family members; open symbols, unaffected family members; black arrows, proband; double line indicates consanguinity. Height and height SDS are shown beneath each symbol. C. Family 3 pedigree; black arrows, proband. D. Distinctive facial dysmorphism. E. Lateral face to show low-set ear; F. Short stature and cubitus valgus. G. brachydactyly. H and I. Family 2 younger brother’s facial features with distinctive facial dysmorphism. I. lateral face to show low-set ear. J. Skeletal x-ray of the left hand in younger affected brother. A cone-shaped epiphysis is indicated with a yellow open arrow. K. Skeletal x-ray of right tibia in proband. An incidental cystic lesion is indicated with a yellow open arrow. L. Coronal T2 image showing left anterior temporal arachnoid cyst (white arrow). M. Midsagittal T1 image showing giant cisterna magna. The cortex, corpus callosum, and brainstem are normal. N and O. Family 3, distinctive facial dysmorphism. P. Brachydactyly. SDS, standard deviation score; WT, wild type.

Figure 2

The L69F variant in WASHC3 interferes with binding to WASH1. WASH1 and either empty vector (WASH1 only), wild-type-WASHC3-GFP (WASH1+WT), or L69F-WASHC3-GFP (WASH1+mut) were coexpressed in HEK293 cells, immunoprecipitated using anti-GFP, and analyzed by western blot using anti-WASH1 (to assess input) or anti-GFP (IP; to assess coprecipitation of WASHC3 and WASH1). GFP, green fluorescent protein.

Figure 3

The L69F variant in WASHC3 alters trafficking of PTH1R. A. Endosomal localization. HEK293 cells that lack WASHC3 were transfected with vectors expressing PTH1R plus either with empty vector (EV), wild-type-WASHC3 (WT), or L69F-WASHC3 (Mut) and then treated with vehicle or PTH1-34 (a PTH1R agonist). Endosomes were isolated and analyzed by western blot for EEA1 (early endosome marker), FRS2 (plasma membrane marker), and PTH1R. PTH1R was identified in endosomes primarily in cells expressing wild-type WASHC3 and treated with PTH1-34. B and C. Golgi localization. HEK293 cells that lack WASHC3 were transfected with PTH1R-tGFP expression vector and either with EV, WT, or L69F-WASHC3 expression vector (Mut). Then, cells were treated with PTH1-34 for 2.5 h or without PTH (0 h), fixed, and stained with anti-TGN46 (trans-Golgi marker) (B) or anti-GM130 (cis-Golgi marker) (C). Representative images are shown. PTH1R-tGFP and TGN46 (trans-Golgi marker) showed similar colocalization with EV, WT, and Mut at 0 and 2.5 h. PTH1R-tGFP and GM130 (cis-Golgi marker) did not colocalize with WT but colocalize with EV and Mut at 2.5 h PTH treatment. The graphs next to confocal images show the fluorescence intensity profiles along the cross sections indicated by yellow line segments. Scale bars, 10 μm. EV, empty vector; FRS2, fibroblast growth factor receptor substrate 2; Mut, mutant; WT, wild type.

Figure 4

The L69F variant in WASHC3 decreases PTH1R-cAMP signal transduction and alters mRNA expression of hypertrophy-associated genes. A. HEK293 cells that lack WASHC3 were transfected with a PTH1R expression vector and either with empty vector (EV), wild-type-WASHC3 (WT), or L69F-WASHC3 (Mut) expression vectors and then treated with vehicle or PTH1-34. Intracellular cAMP levels were measured by ELISA and normalized to total protein concentration (n = 6). In EV-transfected cells, PTH1-34 increased cAMP levels. WT WASHC3 augmented PTH levels further, whereas L69F-WASHC3 did not. B. HEK293 cells were transfected with a PTH1R expression vector and with siRNA directed against various components of the WASH complex or with scrambled siRNA and then treated with PTH1-34 (+PTH) or vehicle (−PTH). Knockdown of each member of the WASH complex decreased cAMP levels (normalized to protein concentration, n = 9). WASH1 was analyzed separately but with an identical experimental design. C. Primary chondrocytes from 4-day-old rat epiphyses were transfected with siRNA directed against WASHC3 and either with empty vector (EV), a wild-type-WASHC3 (WT) or an L69F-WASHC3 (Mut) expression vector and then treated with PTH1-34 (a PTH1R agonist). Col10a1, Ihh, and Mmp13 relative expression were measured by real-time PCR, normalized to 18S rRNA (n = 7). Mut, mutant.

Figure 5

Start codon variant (c.1A>T) reduces WASHC3 protein expression. HEK293 cells that lack WASHC3 were transfected with vectors expressing WASHC3 with or without the start codon variant. Cells were lysed, and protein expression was assessed by western blot (A), quantitation of western blot was assessed using relative density by ImageJ (B), and mRNA levels were assessed by quantitative real-time PCR (C). The variant reduced the protein expression but not the mRNA expression. EV, empty vector; Mut, mutant; WT, wild type.

Figure 6

Diagram of the WASH complex and associated conditions. WASH complex components are depicted in the center of the diagram.^,, Pathogenic variants in SWIP, strumpellin, and WASHC3 cause phenotypes (text boxes), which all include short stature. GWA studies have found an association between height and single-nucleotide polymorphisms near genes participating in WASH. This conceptual diagram incorporates findings from multiple previous studies.^,,