Examining the Neurodevelopmental Impact of Sonic Hedgehog Pathway Inhibition in Mice

Abstract

Background: Neurodevelopmental disorders (NDDs) are common, highly variable, and etiologically complex. Identifying environmental factors that adversely impact prenatal brain development is a direct path to NDD prevention. Small molecule disruption of the Sonic hedgehog (Shh) signaling pathway, a key regulator of craniofacial morphogenesis, can lead to overt face and forebrain malformations that produce profound neurological deficits. However, whether environmental disruption of Shh signaling can cause subtle neurodevelopmental outcomes in the absence of overt facial malformations was unknown.

Methods: We developed a dietary model of Shh signaling inhibition using the specific Shh pathway antagonist vismodegib. C57BL/6J mice were fed control chow or chow containing 25, 75, or 225 ppm vismodegib from gestational day (GD)4 through GD12 to target Shh signaling during craniofacial morphogenesis. Impacts of Shh pathway disruption on face and forebrain development were examined in exposed embryos and fetuses, and behavioral characteristics were assessed in adult mice.

Results: Exposure to chow containing 225 ppm vismodegib resulted in abnormal forebrain patterning at GD11, face and brain malformations at GD17, and early postnatal mortality, while lower treatment groups appeared phenotypically normal. Adult mice exposed to 25 and 75 ppm vismodegib outperformed control mice on repeated rotarod sessions, but treated mice did not significantly differ from control animals in open field exploration, marble burying, olfactory discrimination and detection, or fear conditioning assays.

Conclusions: Under the examined conditions, prenatal Shh disruption did not produce robust neurobehavioral differences in the absence of craniofacial malformations.

Keywords: behavioral assay; craniofacial morphogenesis; neurodevelopmental disorder; sonic hedgehog signaling.

FIGURE 1

Facial morphometric assessment in a dietary model of Shh pathway inhibition. (A–E) GD17 fetuses from dams fed vehicle‐ or vismodegib‐containing chow at the indicated concentrations. Interocular distance (IOD) and upper lip length (ULL) measurements are represented visually in A′ by dashed and dotted lines, respectively. (F, G) Plots of IOD and ULL measurements in fetuses of dams fed vehicle chow (n = 35) or chow containing 25 ppm (n = 28), 75 ppm (n = 36), or 225 ppm (n = 21) vismodegib. Individual data points are plotted along with bars representing mean ± SEM. IOD and ULL measurements for the fetuses shown in panels A–E are indicated by red borders. One‐way ANOVA with Tukey's post hoc test was used to compare the statistical significance of IOD and ULL measurements between treatment groups. *p < 0.05. Scale bar: 1 mm.

FIGURE 2

Evaluation of face and brain morphology in fetal mice exposed to vismodegib. (A–E) Facial images of GD17 fetuses from dams fed diets containing vehicle (n = 12) or 25 ppm (n = 20), 75 ppm (n = 19), or 225 ppm (n = 15) vismodegib. Midfacial hypoplasia was apparent in most individuals from the highest exposure group. (A′–E′) Dorsal view of brains from the same individuals shown in A–E. Olfactory bulbs (ofb) were hypoplastic and partially unseparated following high vismodegib exposure, and the cerebral cortices (cc) showed only shallow divisions in the most severely affected samples. (A″–E″) H&E staining of coronal sections through the forebrain of the individuals in A–E. The 225 ppm vismodegib diet altered the size and shape of the lateral ventricles (lv) and septal region (s) in moderately affected samples (D″) and induced true HPE (incomplete separation of the forebrain) in severely affected samples (E″). mb, midbrain. Scale bars: 1 mm.

FIGURE 3

Expression pattern of select genes within the forebrain of vismodegib‐exposed embryos. (A–E) Representative frontal images of GD11 embryos from dams provided chow containing vehicle (n = 24) or 25 ppm (n = 12), 75 ppm (n = 26), or 225 ppm (n = 19) vismodegib. The 225 ppm vismodegib group showed varying degrees of midfacial hypoplasia resulting in closely approximated facial processes, eyes (e), nostrils (n), and telencephalic vesicles (t). (F–T) Embryos were bisected sagittally and stained using in situ hybridization for the genes indicated. In control samples, Gli1 was most strongly expressed in the inferior aspect of the facial processes and dorsally into the diencephalon (d). Pax6 was expressed dorsally within the telencephalon and diencephalon. Nkx2‐1 was expressed in the medial ganglionic eminence (mge) and the ventral diencephalon. (U) An enlarged view of panel Q with the measured area of the MGE outlined with a dashed red line. (V) Plot of MGE area normalized to head area for a subset of embryos from each litter; vehicle (n = 9), 25 ppm vismodegib (n = 9), 75 ppm vismodegib (n = 10), and 225 ppm vismodegib (n = 12). Individual samples are plotted, and bars represent the mean ± SEM. One‐way ANOVA with Tukey's post hoc test was used to assess the statistical significance in MGE area. *p < 0.05. Scale bars: 500 μm.

FIGURE 4

Evaluation of open field exploration in adult mice with prenatal vismodegib exposure. (A–D) Plots of total distance traveled, the percentage of time spent in the center of the arena versus the peripheries, and counts of stereotypic behavior and vertical activity in mice from dams provided chow containing vehicle (n = 23) or 25 ppm (n = 24) or 75 ppm (n = 24) vismodegib. The 225 ppm vismodegib group was excluded from behavioral testing due to low litter viability. Data are presented as mean ± SEM. For statistical testing, time in center data were log‐transformed to achieve a normal distribution. Stereotypic activity results failedShapiro‐Wilk normality testing, so a nonparametric aligned rank transform was performed prior to analysis. Two‐way ANOVA with Tukey's post hoc test was used to assess differences by treatment and sex.

FIGURE 5

Marble burying behavior in adult mice with prenatal vismodegib exposure. The number of marbles buried by mice prenatally exposed to diets containing vehicle (n = 23) or 25 ppm (n = 24) or 75 ppm (n = 24) vismodegib. Data are presented as mean ± SEM. The data failed Shapiro–Wilk normality testing, so a nonparametric aligned rank transform two‐way ANOVA with post hoc test was used to assess differences by treatment and sex.

FIGURE 6

Rotarod performance in adult mice with prenatal vismodegib exposure. Mice prenatally exposed to diets containing vehicle (n = 23) or 25 ppm (n = 24) or 75 ppm (n = 24) vismodegib were placed on a rotarod apparatus for four sessions, and the time it took for mice to fall from the apparatus was recorded. Many mice from each treatment group successfully remained on the rotarod for the full duration of each 300‐s session. Values represent the mean ± SEM for each treatment and session. The data failed Shapiro–Wilk normality testing, so a nonparametric aligned rank transform ANOVA with post hoc test was used to compare differences between treatment groups, sexes, and sessions. Treatment‐dependent changes are indicated within sessions, and a session‐dependent difference was observed between the first and fourth sessions. *p < 0.05.

FIGURE 7

Assessment of olfactory detection and discrimination in adult mice exposed prenatally to vismodegib. (A) Olfactory discrimination in mice prenatally exposed to diets containing vehicle (n = 23) or 25 ppm (n = 24) or 75 ppm (n = 24) vismodegib was evaluated in two phases. In the first phase, the amount of time mice spent investigating a swab scented with water (familiar) or a swab scented with either lemon or mint (novel) was recorded. During the second phase, mice were provided with two scented swabs, one with the now familiar scent from phase one and the other with a novel scent. The amount of time spent investigating each swab was again recorded. (B) The latency to find a scented object buried beneath bedding was used to evaluate the ability of treated mice to detect odors. Data are presented as mean ± SEM. For olfactory discrimination, the data failed Shapiro–Wilk normality testing, so a nonparametric aligned rank transform ANOVA with post hoc test was used to compare differences by treatment group, sexes, and scent within each session. Olfactory detection data were log‐normalized to achieve a normal distribution, and two‐way ANOVA with Tukey's post hoc test was used to assess differences by treatment and sex.

FIGURE 8

Fear conditioning behavior in adult mice exposed prenatally to vismodegib. (A) To test contextual fear recall, mice prenatally exposed to diets of vehicle (n = 23) or 25 ppm (n = 24) or 75 ppm (n = 24) vismodegib were placed in the same fear conditioning chamber in which they first experienced the CS‐US pairing 22 h prior, and time spent freezing was recorded over a 360‐s period. (B) Cued fear recall was assessed by administering the CS only to mice in a novel context 24 h after the CS‐US training. Time spent freezing during each 30 s CS event was recorded. Bars represent mean ± SEM. Two‐way ANOVA with Tukey's post hoc test was used to assess differences by treatment and sex. *p < 0.05.