Impact of retinoic acid exposure on midfacial shape variation and manifestation of holoprosencephaly in Twsg1 mutant mice

Abstract

Holoprosencephaly (HPE) is a developmental anomaly characterized by inadequate or absent midline division of the embryonic forebrain and midline facial defects. It is believed that interactions between genes and the environment play a role in the widely variable penetrance and expressivity of HPE, although direct investigation of such effects has been limited. The goal of this study was to examine whether mice carrying a mutation in a gene encoding the bone morphogenetic protein (BMP) antagonist twisted gastrulation (Twsg1), which is associated with a low penetrance of HPE, are sensitized to retinoic acid (RA) teratogenesis. Pregnant Twsg1(+/-) dams were treated by gavage with a low dose of all-trans RA (3.75 mg/kg of body weight). Embryos were analyzed between embryonic day (E)9.5 and E11.5 by microscopy and geometric morphometric analysis by micro-computed tomography. P19 embryonal carcinoma cells were used to examine potential mechanisms mediating the combined effects of increased BMP and retinoid signaling. Although only 7% of wild-type embryos exposed to RA showed overt HPE or neural tube defects (NTDs), 100% of Twsg1(-/-) mutants exposed to RA manifested severe HPE compared to 17% without RA. Remarkably, up to 30% of Twsg1(+/-) mutants also showed HPE (23%) or NTDs (7%). The majority of shape variation among Twsg1(+/-) mutants was associated with narrowing of the midface. In P19 cells, RA induced the expression of Bmp2, acted in concert with BMP2 to increase p53 expression, caspase activation and oxidative stress. This study provides direct evidence for modifying effects of the environment in a genetic mouse model carrying a predisposing mutation for HPE in the Twsg1 gene. Further study of the mechanisms underlying these gene-environment interactions in vivo will contribute to better understanding of the pathogenesis of birth defects and present an opportunity to explore potential preventive interventions.

Keywords: Apoptosis; Bone morphogenetic protein; Holoprosencephaly; Oxidative stress; Retinoic acid; Twisted gastrulation; Twsg1.

Fig. 1.

Phenotypic analysis of Twsg1 mutant embryos exposed in utero to a low dose (3.75 mg/kg of body weight) of ATRA. (A–D) Lateral views of E9.5 embryos. (A) WT untreated embryo; (B) normal appearance of ATRA treated WT embryo; (C,D) Twsg1 mutant embryos treated with ATRA showing proboscis (P) and absence of telencephalic vesicles. (E-H) Frontal views of E10.5 embryos. (E) WT untreated embryo; (F) normal telencephalic vesicles (T; outlined by a dotted line) of an ATRA-treated WT embryo; (G,H) HPE (marked by *) in Twsg1 embryos treated with ATRA.

Fig. 2.

GM analysis of facial shape. (A) Principal component analysis (PCA) based on a GM analysis of facial shape in WT and Twsg1^+/− mouse embryos at E11.5 with or without in utero exposure to a low dose of ATRA (RA) at E7.5. PC analysis of landmark data shows that the majority of shape variation is associated with narrowing of the midface and that PC1 discriminates between treatment groups. PC1 distinguishes between treated embryos and both WT and heterozygous Twsg1 mutant embryos. (B) 3D morphing showing variation along PC1. (C) Procrustes distances between groups. P-values were obtained by permutation of the Procrustes distance.

Fig. 3.

RA target gene expression in response to ATRA treatment in P19 cells as assessed by qPCR. The P19 cell cultures were treated with either DMSO vehicle (Ctrl) or 1 μM ATRA (RA) for 24 hours, with transcript levels quantified by qPCR. Gene expression was normalized to the expression of Gapdh, and is shown relative to the DMSO-vehicle-treated control expression (set at 1). The induction of RA targets in P19 cells indicates the suitability of the P19 cell model for investigating responses to RA signaling. Results are mean±s.e.m. (n=6). ***P<0.001 (Student’s t-test).

Fig. 4.

Gene expression levels in response to exogenous BMP2 and ATRA as assessed by qPCR. P19 cell cultures were treated for 24 hours with vehicle control (0.00001% BSA, 0.001% DMSO, Ctrl), 10 ng/ml recombinant human BMP2 (BMP2), 1 μM ATRA (RA), or 10 ng/ml rhBMP2 and 1 μM ATRA (RA+BMP). Gene expression was normalized to the expression of Gapdh, and is shown relative to the mean vehicle-treated control expression (set at 1), with transcript levels quantified by qPCR. ATRA induced both BMP2 and its downstream targets in P19 cells. Combined BMP and ATRA treatment resulted in a significant increase in Trp53inp1 expression. Results are mean±s.e.m. (n=6). ***P<0.001 compared with Ctrl; ^†††P<0.001 compared with BMP treatment; ^††P<0.01 compared with BMP treatment; ^‡‡‡P<0.001 compared with BMP+ATRA treatment (Tukey’s test).

Fig. 5.

Caspase 3 and 7 activity in P19 cells in response to exogenous BMP2 and ATRA. P19 cells were treated for 24 hours with vehicle control (0.00001% BSA, 0.001% DMSO, Ctrl), 10 ng/ml recombinant human BMP2 (BMP2), 1 μM ATRA (RA), or 10 ng/ml BMP and 1 μM ATRA (RA+BMP), no treatment (Ctrl) or 70 μM AMA as a positive control. RA acted in concert with BMP2 to increase caspase 3 and 7 activation in P19 cells. Results are mean±s.e.m. (n=3). *P<0.05, **P<0.01, ***P<0.001 (Student’s t-test).

Fig. 6.

Quantitation of oxidative stress in P19 cells treated with BMP2 and ATRA. The ratio of reduced glutathione (GSH) to oxidized glutathione (GSSG) was assayed in P19 cells treated for 16 hours with vehicle control (0.0001% BSA, 0.001% DMSO, Ctrl), 10 ng/ml recombinant human BMP2 (BMP2), 1 μM ATRA or 10 ng/ml BMP combined with 1 μM ATRA (BMP+RA). 70 μM AMA was tested as a positive control and compared to its vehicle, propylene glycol. As expected, treatment of P19 cells with the prooxidant AMA resulted in a significant decrease in GSH:GSSG. In addition, oxidative stress was increased in P19 cells treated with ATRA in combination with BMP2. Results are mean±s.e.m. (n=3). *P<0.05, **P<0.01 (Tukey’s test).