Prenatal ethanol exposure in mice phenocopies Cdon mutation by impeding Shh function in the etiology of optic nerve hypoplasia

Abstract

Septo-optic dysplasia (SOD) is a congenital disorder characterized by optic nerve, pituitary and midline brain malformations. The clinical presentation of SOD is highly variable with a poorly understood etiology. The majority of SOD cases are sporadic, but in rare instances inherited mutations have been identified in a small number of transcription factors, some of which regulate the expression of Sonic hedgehog (Shh) during mouse forebrain development. SOD is also associated with young maternal age, suggesting that environmental factors, including alcohol consumption at early stages of pregnancy, might increase the risk of developing this condition. Here, we address the hypothesis that SOD is a multifactorial disorder stemming from interactions between mutations in Shh pathway genes and prenatal ethanol exposure. Mouse embryos with mutations in the Shh co-receptor, Cdon, were treated in utero with ethanol or saline at embryonic day 8 (E8.0) and evaluated for optic nerve hypoplasia (ONH), a prominent feature of SOD. We show that both Cdon^-/- mutation and prenatal ethanol exposure independently cause ONH through a similar pathogenic mechanism that involves selective inhibition of Shh signaling in retinal progenitor cells, resulting in their premature cell-cycle arrest, precocious differentiation and failure to properly extend axons to the optic nerve. The ONH phenotype was not exacerbated in Cdon^-/- embryos treated with ethanol, suggesting that an intact Shh signaling pathway is required for ethanol to exert its teratogenic effects. These results support a model whereby mutations in Cdon and prenatal ethanol exposure increase SOD risk through spatiotemporal perturbations in Shh signaling activity.

Keywords: Cdon; Ethanol; Optic nerve hypoplasia; Septo-optic dysplasia; Shh.

Fig. 1.

Cdon mutation and ethanol exposure independently cause optic nerve hypoplasia. (A-D) Immunostaining for neurofilament (green) on transverse sections through the eye at E14.5 labels the optic nerve (arrow). Compared with (A) saline­-treated wild-type (Cdon^+/+) embryos (n=9), the diameter of the optic nerve (white line) is significantly reduced in (B) saline­-treated Cdon^−/− mutants (n=8), (C) ethanol-treated wild-type embryos (n=9), and (D) ethanol-treated Cdon^−/− mutants (n=9). Scale bar: 200 µm. (E-H) Quantification of (E) optic nerve diameter, (F) axial length of eye, (G) axial width of eye, and (H) optic nerve diameter (OND) normalized to axial width (AW) of the eye. Error bars represent s.d. **P<0.01, ***P<0.001 by Student's t-test.

Fig. 2.

The optic disc is not compromised in Cdon^−/− or ethanol-treated embryos. (A-D) Immunostaining for Pax2 on transverse sections through the eye at E14.5 marks the optic disc (arrows). No significant differences were observed in the average number of Pax2⁺ optic disc cells per section from (A) saline­-treated wild-type (Cdon^+/+) embryos (n=6), (B) saline­-treated Cdon^−/− mutants (n=8), (C) ethanol-treated wild-type embryos (n=8), and (D) ethanol-treated Cdon^−/− mutants (n=8). Scale bar: 200 µm. (E) Quantification of Pax2⁺ optic disc cells. Error bars represent s.d. Student's t-test.

Fig. 3.

Selective reduction of Gli1 expression in the eyes of Cdon^−/− and ethanol-treated embryos. In situ hybridization for Gli1 (A-H), and Shh (I-P) on transverse sections through the eye (A-D,I-L) and hypothalamus (E-H,M-P) of E14.5 wild-type (Cdon^+/+) and Cdon^−/− embryos treated with saline or ethanol at E8.0. Gli1 expression is detected in retinal progenitor cells (RPCs, area marked by dotted white line) of (A) saline­-treated wild-type embryos (n=5). Gli1 expression is markedly reduced in RPCs of (B) saline­-treated Cdon^−/− mutants (n=5), (C) ethanol-treated wild-type embryos (n=9), and (D) ethanol-treated Cdon^−/− mutants (n=5). No differences were observed in the expression of Gli1 in the hypothalamus between genotypes or treatment groups (E-H). No differences were observed in the level of Shh expression in retinal ganglion cells (RGCs, area marked by dotted black line, I-L) or the hypothalamus (M-P) between genotypes or treatment groups. Scale bar: 200 µm.

Fig. 4.

Reduced proliferation of RPCs in Cdon^−/− and ethanol-treated embryos. (A-H) Immunostaining for Ki67 on transverse sections through the eye (A-D) and hypothalamus (E-H) of E14.5 embryos labels proliferating progenitors. (A) The majority of retinal progenitor cells (RPCs, area marked by dotted white line) in saline­-treated wild-type (Cdon^+/+) embryos (n=3), are marked by Ki67. The number of Ki67⁺ RPCs is significantly reduced in (B) saline­-treated Cdon^−/− embryos (n=3), (C) ethanol-treated wild-type embryos (n=3), and (D) ethanol-treated Cdon^−/− embryos (n=3). No differences in the number of Ki67⁺ cells in the ventricular layer of the ventral hypothalamus (boxed area) were observed between genotypes or treatment groups (E-H). Scale bar: 200 µm. (I-K) Quantification of Ki67⁺ cells in (I) retina, (J) lens and (K) hypothalamus sections across all experimental groups. Error bars represent s.d. *P<0.05, **P<0.01, ***P<0.001 by Student's t-test.

Fig. 5.

Precocious differentiation of RGCs in Cdon^−/− and ethanol-treated embryos. (A-D) In situ hybridization for Math5 on transverse sections through the eye at E14.5 labels postmitotic progenitors in the ventricular zone (vz) of control embryos (n=3) (A). Math5 expression expands into the ganglion cell layer (gcl) of Cdon^−/− and ethanol-treated embryos (n=3 for each experimental group) (B-D). Dashed line marks the vz–gcl boundary. (E-H) Immunostaining for Isl1/2 on transverse sections through the eye of E14.5 embryos primarily labels differentiating retinal ganglion cells (RGCs, arrows). Compared with (E) saline­-treated wild-type (Cdon^+/+) embryos (n=3), the number of Isl1/2⁺ RGCs is significantly increased in (F) saline­-treated Cdon^−/− mutants (n=3), (G) ethanol-treated wild-type embryos (n=3), and (H) ethanol-treated Cdon^−/− mutants (n=3). (I-L) Immunostaining for AP-2α in amacrine cells. No significant differences were observed between the average number of AP-2α⁺ amacrine cells per section from (I) saline­-treated wild-type (Cdon^+/+) embryos (n=4), (J) saline­-treated Cdon^−/− mutants (n=3), (K) ethanol-treated wild-type embryos (n=4), and (L) ethanol-treated Cdon^−/− mutants (n=3). Scale bar: 50 µm. (M,N) Quantification of cells expressing Isl1/2 (M) and AP-2α (N). Error bars represent s.d. *P<0.05, **P<0.01 by Student's t-test.

Fig. 6.

Model depicting the influence of Cdon mutation and ethanol exposure on Shh signaling activity in the developing eye. (A) In wild-type embryos (E14.5), the binding of Shh to Cdon and Ptch1 releases the inhibition on smoothened (Smo), facilitating the transcription by Gli activator (GliA) of target genes involved in RPC proliferation. (B) In the eyes of 129S6.Cdon^−/− embryos, there is persistent inhibition of Smo by Ptch1, even in the presence of Shh, causing Gli repressor (GliR) to block transcription of genes involved in RPC proliferation, resulting in precocious RPC differentiation. (C) Ethanol exposure at E8.0 interferes with Shh signaling in the eye through a variety of proposed mechanisms. The lengthy delay between ethanol exposure (E8.0) and its negative effects on Shh signaling activity (E14.5), suggests that the epigenetic landscape of Shh target genes might be modified to suppress RPC proliferation. (D) Cdon mutation or ethanol exposure at E8.0 impedes Shh signaling activity in RPCs resulting in optic nerve hypoplasia.