Gene-teratogen interactions influence the penetrance of birth defects by altering Hedgehog signaling strength

Abstract

Birth defects result from interactions between genetic and environmental factors, but the mechanisms remain poorly understood. We find that mutations and teratogens interact in predictable ways to cause birth defects by changing target cell sensitivity to Hedgehog (Hh) ligands. These interactions converge on a membrane protein complex, the MMM complex, that promotes degradation of the Hh transducer Smoothened (SMO). Deficiency of the MMM component MOSMO results in elevated SMO and increased Hh signaling, causing multiple birth defects. In utero exposure to a teratogen that directly inhibits SMO reduces the penetrance and expressivity of birth defects in Mosmo-/- embryos. Additionally, tissues that develop normally in Mosmo-/- embryos are refractory to the teratogen. Thus, changes in the abundance of the protein target of a teratogen can change birth defect outcomes by quantitative shifts in Hh signaling. Consequently, small molecules that re-calibrate signaling strength could be harnessed to rescue structural birth defects.

Keywords: Gene-environment interactions; Hedgehog signaling; Left-right patterning; Morphogen; Smoothened; Structural birth defects.

Fig. 1.

Loss of Mosmo results in embryonic lethality and developmental defects. (A) Viability of offspring derived from Mosmo^+/−×Mosmo^+/− crosses at the indicated developmental stages. Statistical significance was determined by the chi-squared test (ns>0.05, ***P<0.001) and n=number of live embryos collected. See Table S1 for full details. (B) Whole-mount images of E16.5 control (Mosmo^+/−) and Mosmo^−/− littermates show that the latter have edema and preaxial polydactyly. See Table S2 for a detailed list of phenotypes in each embryo analyzed. Scale bars: 1 mm. (C) Skeletons from E16.5 control (Mosmo^+/−) and Mosmo^−/− littermates stained with Alcian Blue and Alizarin Red S to visualize cartilage and calcified bone, respectively. Polydactyly (asterisks), sternal clefting (middle column, arrows) and tibial truncation (tibial hemimelia, right column, arrowheads) were observed in Mosmo^−/− embryos. Scale bars: 1 mm. (D) Whole-mount images of E11.5 control (Mosmo^+/+) and Mosmo^−/− littermates show that the latter suffer from exencephaly (arrow). Scale bar: 1 mm. (E) Whole-mount lungs (ventral view) of E12.5 control (Mosmo^+/+) and Mosmo^−/− embryos immunostained for E-cadherin to show the airway epithelium and allow for detailed branching analysis. Normal mouse lungs have one lobe on the left (L.L1) and four lobes on the right (R.Acc, right accessory; R.Cd, right caudal; R.Cr, right cranial; R.Md, right middle). Mosmo^−/− lungs exhibit right pulmonary isomerism (RPI), a duplication of the right lung morphology on the left side (L.Acc, left accessory; L.Cd, left caudal; L.Cr, left cranial; L.Md, left middle). Further details are provided in Table S3. (F) Mosmo^−/− embryos exhibit complex CHDs associated with abnormal left-right patterning of the heart. Necropsy and episcopic confocal fluorescence microscopy (ECM) images of representative embryonic hearts from E16.5 Mosmo^+/+ (control, top) and Mosmo^−/− embryos (bottom). The Mosmo^−/− heart has no apparent apex (black arrow), indicating mesocardia, with the aorta (Ao) abnormally positioned anteriorly. The Ao is situated anterior to the pulmonary artery (PA) and inserted into the morphological right ventricle (mRV) situated on the left of the body, while the pulmonary artery (PA) emerges from the morphological left ventricle (mLV) positioned on the right of the body, findings that are diagnostic of transposition of the great arteries (TGAs). Other findings include non-compaction of the ventricular myocardium and an unbalanced atrioventricular septal defect (AVSD) with symmetric insertions of both inferior and superior vena cava suggesting right atrial isomerism (mRA). Scale bars: 0.5 mm. A detailed list of cardiac phenotypes observed in each embryo can be found in Table S2.

Fig. 2.

Loss of Mosmo results in elevated Hh signaling activity. Hh signaling activity in primary mouse embryonic fibroblasts (pMEFs) and embryonic tissues was assessed using expression of Gli1, a direct Hh target gene, or accumulation of SMO in primary cilia. (A) Gli1 mRNA abundance in wild-type (Mosmo^+/+) and Mosmo^−/− pMEFs was measured using qRT-PCR. Data are the median Gli1 mRNA values derived from the two or three individual measurements shown as circles. The statistical analysis between the two groups was determined using an unpaired two-tailed t-test (**P<0.01,***P<0.001 and ****P<0.0001). (B) X-gal staining was used to visualize Gli1-lacZ expression in whole-mount preparations of mouse embryos. Arrowheads indicate areas of elevated Hh signaling activity in the cardiac outflow tract and anterior hindlimb. Scale bars: 0.5 mm. (C) Immunoblot used to measure protein abundance of SMO, GLI3, PTCH1 and ɑTUB (a loading control) in whole-embryo lysates prepared from E12.5 wild-type (Mosmo^+/+) and Mosmo^−/− littermates. (D,E) Immunofluorescence (IF) was used to measure SMO abundance (red) in primary cilia (ARL13B, green) in pMEFs (D) and various embryonic tissues (E). Violin plots in D summarize ciliary SMO fluorescence data from wild-type (n=108) and Mosmo^−/− (n=156) pMEFs, with horizontal lines indicating the median and interquartile range. Statistical significance was determined using a two-tailed Mann–Whitney test (****P<0.0001). (E) DAPI (blue) marks nuclei. White arrowheads indicate the primary cilium enlarged in the inset. Scale bars: 10 µm in merged panels; 1 µm in insets.

Fig. 3.

MOSMO interacts with MEGF8 and regulates its abundance at the cell surface. (A) Abundance of SMO and MEGF8 protein in whole-cell lysates (top, 6.25% input) and on the cell surface [bottom, cell surface biotinylation and streptavidin immunoprecipitation (IP), 50% elution] in NIH/3T3 cells of the indicated genotypes. The cytoplasmic protein p38 serves as a loading control. (B,C) SMO ubiquitylation was assessed after transient co-expression of the indicated proteins in HEK293T cells, as described in our previous study (Kong et al., 2020). Cells were lysed under denaturing conditions, SMO was purified by immunoprecipitation, and the amount of HA-tagged ubiquitin (HA-UB) covalently conjugated to SMO was assessed using immunoblotting with an anti-HA antibody. In B, assays were carried out in the presence of endogenous MGRN1 and increasing amounts of transfected MOSMO. Assays in D co-transfected either wild-type MGRN1 or catalytically inactive MGRN1 (MGRN1^Mut) to show that SMO ubiquitylation was dependent on the function of MGRN1. (D) Endogenous MEGF8 and MGRN1 co-purified with 1D4-tagged MOSMO immunoprecipitated from Mosmo^−/− NIH/3T3 cells stably expressing MOSMO-1D4 (Fig. S4A,B). (E) Domain graphics depict the simplified modular architecture of wild-type MEGF8 (leftmost image) and its engineered variants (right three images). Gray circles denote the linearly connected EGFL and PSI repeats, with interspersed six-bladed β-propeller domains (hexagons) and CUB domains (diamonds), while the juxtamembrane M-Stem domain is a blue oval. The larger globular structures of the β-propeller, CUB and M-Stem folds all have closely spaced N- and C-termini, so they appear as pendant-like inserts into the long EGFL and PSI chain. A proposed mode of interaction between the M-Stem domain of MEGF8 and the extracellular domain of MOSMO is depicted (leftmost), based on modeling described in Fig. S5. (F) A series of truncation mutants of MEGF8 (shown in F) were used to identify the MEGF8 domain that binds to MOSMO. The interaction between HA-tagged MOSMO and these 1D4-tagged MEGF8 variants was assessed by transient expression in HEK293T cells followed by an anti-1D4 IP and immunoblotting to measure the amount of co-precipitated MOSMO-HA (right).

Fig. 4.

Reducing Hh signaling strength with the SMO rescues limb defects in Mosmo^−/− embryos. (A) Vismodegib administered every 12 h for ∼2 days from E9.75 to E11.5 (diagrammed on the left, with purple used to shade developmental windows for drug exposure) reduced Gli1-lacZ expression in wild-type embryos (right). Scale bar: 1 mm. (B) Numbers of digits per limb of E14.5 embryos from six individual litters exposed to increasing amounts of vismodegib (represented by purple shading in the developmental time courses depicted on the x-axis). The area of each bubble is proportional to the number of limbs included in the analysis. Bold horizontal lines represent the median number of digits per limb in control (Mosmo^+/+ and Mosmo^+/−, blue) and Mosmo^−/− (red) embryos. (C) Representative images (left) of E14.5 control and Mosmo^−/− embryos treated with or without vismodegib every 12 h for 3 (E8.25-E11.25) or 4 (E7.25-E11.25) days. Scale bars: 1 mm. Bubble plot (right) used to depict the numbers of digits per limb of control (Mosmo^+/+ and Mosmo^+/−, blue) and Mosmo^−/− (red) embryos treated with vismodegib. Statistical significance was determined using the Kruskal–Wallis test (****P<0.0001).

Fig. 5.

Reducing Hh signaling strength partially rescues heart defects in Mosmo^−/− embryos. (A) Representative necropsy (top row) and ECM (bottom row) images of embryonic hearts from E14.5 control (Mosmo^+/+ and Mosmo^+/−) and Mosmo^−/− embryos treated with vismodegib or a vehicle control. Vismodegib was administered every 12 h for 3 (E8.25-E11.25) or 4 (E7.25-E11.25) days. Scale bars: 200 µm. (B) Summary of heart malformations and left-right patterning defects in E14.5 control (Mosmo^+/+ and Mosmo^+/−) and Mosmo^−/− mouse embryos treated with either vehicle or vismodegib every 12 h for 3 (E8.25-E11.25) or 4 (E7.25-E11.25) days. Lung situs could not be determined (N/A) in vismodegib-treated control embryos due to cystic and hypoplastic lungs. Ao, aorta; DORV, double outlet right ventricle; LV, left ventricle; mLV, morphological left ventricle; mRV, morphological right ventricle; PA, pulmonary artery; PTA, persistent truncus arteriosus; RV, right ventricle; TGA, transposition of the great arteries. A detailed list of phenotypes observed in each embryo can be found in Table S4. (C) A proposed model for how a combination of genotype and SMO inhibition by vismodegib influences Hh signaling strength and consequently ventriculoarterial alignment in developing embryos. (D) Viability of E14.5 embryos from Mosmo^+/−×Mosmo^+/− crosses treated with vehicle or vismodegib every 12 h for 3 (E8.25-E11.25) or 4 (E7.25-E11.25) days. Statistical significance was determined using the chi-squared test (ns>0.05, **P<0.01) and n=number of live embryos collected. See Table S5 for full details.

Fig. 6.

The developing spinal cords of Mosmo^−/− embryos are resistant to SMO inhibition. Neural tube patterning was assessed using confocal fluorescence microscopy to image markers that define neural progenitor populations in sections of the ventral spinal cord from E11.5 control (Mosmo^+/+ and Mosmo^+/−) and Mosmo^−/− embryos treated with or without vismodegib. (A) Immunofluorescence (IF) was used to evaluate SMO abundance (red) in primary cilia (ARL13B, green) within neural progenitors (SOX2, blue) of untreated E11.5 embryos (left) and vismodegib-treated embryos (right). White arrows indicate regions where SMO is seen in cilia; red lines indicate the regions of the developing neural tube where ciliary SMO is observed. Scale bars: 100 µm in merged panels; 50 µm in zoomed displays. (B) IF was used to assess the abundance and distribution of NKX6.1⁺, OLIG2⁺ and NKX2.2⁺ (red) neural progenitor cells (SOX2, blue) of untreated E11.5 embryos (left) and vismodegib-treated embryos (right). Images represent three serial sections taken from a single representative embryo. Scale bars: 50 µm. (C) Summary of the dorsal-ventral distribution of ciliary SMO, PAX6, NKX6.1, OLIG2, NKX2.2 and HNF3β from untreated E10.5 control (Mosmo^+/+ and Mosmo^+/−) (n=3) and Mosmo^−/− (n=3) embryos. See Fig. S6B,C for representative images. (D) Quantification of NKX6.1⁺, OLIG2⁺ and NKX2.2⁺ spinal neural progenitors from vismodegib-treated E11.5 control (n=7) and Mosmo^−/− (n=4) embryos. (C,D) Each point represents one embryo. Data are the mean and the statistical analysis between the two groups was determined using an unpaired two-tailed t-test (ns>0.05, *P<0.05, **P<0.01 and ****P<0.0001).

Fig. 7.

The regulation of cell surface Smoothened (SMO) by the MMM complex. MOSMO binds to the M-Stem domain of MEGF8. Together, MOSMO, MEGF8 and MGRN1 form a membrane-tethered E3 ligase complex (the MMM complex) that regulates the sensitivity of a target cell to Hh ligands by regulating levels of SMO at the cell surface and primary cilium.