Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

© 2016. Published by The Company of Biologists Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.

Gli2 gene dosage and gene-environment interaction illuminate

the etiological complexity of holoprosencephaly

Galen W. Heyne1,*, Joshua L. Everson1, 2,*, Lydia J. Ansen-Wilson1, Cal G. Melberg1, Dustin

M. Fink1, Kia F. Parins1, Padydeh Doroodchi1, Caden M. Ulschmid1, and Robert J. Lipinski1, 2,‡

1Department of Comparative Biosciences, School of Veterinary Medicine, University of

Wisconsin-Madison, Madison, WI, United States of America

2Molecular and Environmental Toxicology Center, School of Medicine and Public Health,

University of Wisconsin-Madison, Madison, WI, United States of America

* These authors contributed equally to this work

‡Corresponding author:

E-mail: rjlipinski@wisc.edu

Keywords: Holoprosencephaly, birth defects, gene-environment, Hedgehog signaling

Summary Statement: This work illustrates how a specific genetic predisposition in

combination with an environmental exposure can result in a severe birth defect, providing a

novel opportunity to develop prevention strategies.

http://dmm.biologists.org/lookup/doi/10.1242/dmm.026328Access the most recent version at DMM Advance Online Articles. Posted 1 September 2016 as doi: 10.1242/dmm.026328http://dmm.biologists.org/lookup/doi/10.1242/dmm.026328Access the most recent version at

First posted online on 1 September 2016 as 10.1242/dmm.026328

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Abstract

Holoprosencephaly (HPE) is a common and severe human developmental abnormality

marked by malformations of the forebrain and face. While several genetic mutations have

been linked to HPE, phenotypic outcomes range dramatically and most cases cannot be

attributed to a specific cause. Gene-environment interaction has been invoked as a premise

to explain the etiological complexity of HPE but identification of interacting factors has been

extremely limited. Here, we demonstrate that mutations in Gli2, which encodes a Hedgehog

pathway transcription factor, can cause or predispose to HPE depending upon gene dosage.

On the C57BL/6J background, homozygous GLI2 loss of function results in the characteristic

brain and facial features of severe human HPE, including midfacial hypoplasia,

hypotelorism, and medial forebrain deficiency with loss of ventral neurospecification. While

normally indistinguishable from wildtype littermates, we demonstrate that mice with single-

allele Gli2 mutations exhibit increased penetrance and severity of HPE in response to low-

dose teratogen exposure. This genetic predisposition is associated with a Gli2 dosage-

dependent attenuation of Hedgehog ligand responsiveness at the cellular level. In addition

to revealing a causative role for GLI2 in HPE genesis, these studies demonstrate a

mechanism by which normally silent genetic and environmental factors can interact to

produce severe outcomes. Together, these findings provide a framework for understanding

the extreme phenotypic variability observed in human mutation carriers, and a paradigm for

reducing the incidence of this morbid birth defect.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Introduction

Holoprosencephaly (HPE) is the most prevalent human forebrain malformation, and one of

the most common of all developmental abnormalities, with an estimated prevalence of 1 in

250 conceptuses (Leoncini et al., 2008, Orioli and Castilla, 2010, Matsunaga and Shiota,

1977). HPE is defined by incomplete midline division of the embryonic forebrain and

frequently co-occurs with facial abnormalities, including hypotelorism, ophthalmologic

anomalies, midfacial hypoplasia, and orofacial clefts (Richieri-Costa and Ribeiro, 2010,

Cohen, 2006, Pineda-Alvarez et al., 2011). In surviving patients, HPE causes severe

intellectual disability and learning, behavior, and motor impairment (Cohen, 2006). More

often, the severity of effects results in prenatal or perinatal mortality (Solomon et al., 2010).

Clinical presentation of HPE is extremely variable. True HPE ranges from alobar forms,

marked by a single ventricle and no separation of the cerebral hemispheres, to lobar forms

with minor midline deficiencies (Solomon et al., 2010). However, obligate carriers of HPE-

associated mutations frequently exhibit facial dysmorphology in the absence of detectable

brain abnormalities or no apparent phenotype (Solomon et al., 2012). The dramatic

variability in phenotypic expression is believed to stem from a complex, heterogeneous

etiology involving the interaction of genetic and environmental factors (Krauss and Hong,

2016). While this premise has become widely accepted, supportive experimental evidence

demonstrating specific interacting factors is limited. This knowledge gap hinders clinical

management of HPE by limiting the accuracy of genetic counseling and stymying

development of prevention strategies (Mercier et al., 2010).

In humans and animal models, HPE has been linked to chemical and genetic disruption of

the Hedgehog (Hh) signaling pathway (Roessler et al., 1996, Chiang et al., 1996, Heyne et al.,

2015a). Initiated by the Sonic Hedgehog (SHH) ligand, Hh signaling is required for ventral

patterning and expansion of the medial forebrain, as well as outgrowth of the processes that

form the midface. Hh signal transduction culminates in regulation of tissue-specific target

genes by the Gli family of zinc finger transcription factors, with GLI2 acting as the dominant

transcriptional activator (Lipinski et al., 2006). Serving as a poignant example of the

complexity that has frustrated basic and translational research efforts, the role of GLI2 in

HPE has remained unclear for two reasons. First, while single-allele GLI2 mutations have

been detected in individuals with HPE-like phenotypes, the majority of mutation carriers do

not exhibit the full manifestation of HPE or are clinically unaffected (Bear et al., 2014,

Franca et al., 2010, Roessler et al., 2003, Bertolacini et al., 2012, Rahimov et al., 2006).

Second, Gli2 knockout mice generated on an outbred CD-1 background do not recapitulate

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

the forebrain and facial abnormalities that characterize human HPE (Matise et al., 1998,

Park et al., 2000).

We examined the effect of GLI2 loss of function by backcrossing a null allele to the C57BL/6J

(B6) background, which has been shown to exacerbate craniofacial phenotypes, including

HPE. We found that a homozygous Gli2 mutation on the B6 background causes the salient

features of severe human HPE, including abnormalities of the face and forebrain. B6 Gli2+/-

mice and cells were then used to test whether normally silent single-allele mutations increase

sensitivity to a class of teratogens that includes environmental compounds. These functional

in vivo and mechanistic in vitro assays demonstrated a gene-environment interaction that

provides a basis to potentially reduce the incidence of this etiologically-complex disease in

susceptible human populations.

Results

GLI2 loss of function causes HPE

Mating of B6 Gli2+/- mice generated Gli2-/- fetuses at the expected Mendelian ratio (n=13/49)

at gestational day (GD)15. Midfacial hypoplasia and hypotelorism were observed in all Gli2-/-

animals on the B6 background (Fig. 1). These abnormalities co-occurred with absence of the

upper lip notch and a single central or two closely opposed nostrils. The range of facial

phenotypes observed in Gli2-/- fetuses is shown in Figure S1. Significant reductions in both

snout width and interocular distance were identified by linear measurement. Gli2-/- animals

also displayed microphthalmia and decreased head width suggestive of microcephaly. Facial

morphology in B6 Gli2+/- fetuses was indistinguishable from that of wildtype littermates.

Crown-rump length and limb morphology were not different between B6 Gli2+/+, Gli2+/-, and

Gli2-/- fetuses (Fig. S2).

Gli2-/- animals were also compared to wildtype B6 fetuses exposed to a single 40mg kg-1 dose

of the potent Hh signaling pathway antagonist vismodegib at GD7.75. This teratogenic

exposure regimen was recently reported to result in severe HPE phenotypes (Heyne et al.,

2015a). A remarkable degree of overlap in facial dysmorphology was observed among Gli2-/-

fetuses and those acutely exposed to vismodegib (Fig. 1). Importantly, these facial

phenotypes in mice closely mimic those observed in humans with severe forms of true HPE.

This is illustrated in the neonate shown in Figure 1G, who exhibits severe midfacial

hypoplasia, hypotelorism, and a single central nostril. MRI conducted prior to birth showed

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

that these facial features co-occurred with HPE, illustrated by an undivided cerebral cortex

and single ventricle.

We next investigated whether facial dysmorphology resulting from GLI2 loss of function co-

occurs with forebrain abnormalities. Gli2-/- fetuses exhibited deficiency of the midbrain and

forebrain (Fig. 2). Narrowing of the anterior aspect of the cerebral cortices extended to the

olfactory bulbs, which were hypoplastic and abnormally closely spaced. A ventral view of the

forebrain further illustrates the medial deficiency observed in Gli2-/- fetuses (Fig. S3).

Vismodegib exposure in wildtype animals resulted in more pronounced deficiency of the

cerebral cortices and olfactory bulb aplasia. Gross brain morphology was indistinguishable

between Gli2+/- fetuses and their wildtype littermates.

The relationship between face and forebrain abnormalities resulting from GLI2 loss of

function was investigated in more detail by histologic examination. This revealed specific

features of midfacial hypoplasia including absence of the nasal septal cartilage, closely

spaced or fused nasal passages, and vomeronasal organ deficiency observed in each

examined Gli2-/- fetus (n=6/6) (Fig. 3C). One Gli2-/- fetus exhibited olfactory bulb agenesis,

while in each of the others (n=5/6), the olfactory bulbs were present but spaced abnormally

close (Fig. 3G). Medial forebrain deficiency, evidenced by absence of the septal region, was

observed in each B6 Gli2-/- fetus that was examined. The majority (n=4/6) exhibited

incomplete division of the cerebral cortices with a single communicating ventricle, the

defining feature of HPE (Fig. 3K). Severe abnormalities in the diencephalic region of the

forebrain were also observed. In all Gli2-/- fetuses, dysmorphology of the hypothalamus

included attenuation of the third ventricle, along with absence of both the anterior and

posterior lobes of the pituitary (Fig. 3O). The brains and faces of Gli2+/- fetuses were

histologically similar to those of their wildtype littermates.

The face and brain phenotypes observed in Gli2-/- fetuses were largely recapitulated by

vismodegib exposure at GD7.75. Teratogen-exposed B6 wildtype mice exhibited absence of

the nasal septal cartilage and forebrain septal region, and incomplete division of the cerebral

cortices (Fig. 3D,L). Notable differences detected in vismodegib-exposed fetuses included

higher penetrance of olfactory bulb agenesis, subtler diencephalic dysmorphology, and a

hypoplastic anterior pituitary (Fig. 3H,P).

Both GLI2 and the protein target of vismodegib, Smoothened, are expressed in SHH ligand-

responding cells, where they are required for transduction of the downstream signaling

cascade (Lipinski et al., 2006, Robarge et al., 2009). Given the overlap in face and brain

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

phenotypes between Gli2-/- and vismodegib-exposed wildtype embryos, we hypothesized that

tissue deficiency in these animals involves populations of cells that are responding to SHH

signaling at or shortly after GD7.75. To test this hypothesis, we used a Tamoxifen-inducible

lineage reporter model to trace temporally-specific Hh-responsive cells (Ahn and Joyner,

2005). Tamoxifen administration at GD7.75 revealed Hh-responsive cell lineages populating

tissues that are deficient in Gli2-/- mice (Fig. 3, Q-T). Regions with positive staining included

the vomeronasal organs, connective tissue between the olfactory bulbs, cerebral cortices,

hypothalamus, and anterior pituitary gland, indicating that these areas contain the progeny

of cells that are responsive to Hh signaling at or shortly after GD7.75. This premise is

further supported by the observation that expression of the conserved Hh pathway target

gene Gli1 is reduced in the anterior neural plate/folds as early as GD8.0 in both Gli2-/- and

vismodegib exposed embryos (Fig. S4).

Hh signaling plays two distinct roles in early forebrain development. It acts as a mitogen,

promoting expansion of medial forebrain tissue and separation of the initially singular eye

field. At the same time, SHH ligand secretion from the notochord and floor plate of the

neural plate/tube specifies ventral neuroprogenitor cells. We tested whether attenuated

forebrain expansion caused by GLI2 loss of function coincides with disruptions in early

dorsal-ventral forebrain specification. At GD11.0, the morphogenesis of HPE was clearly

evident in Gli2-/- embryos (Fig. 4C). This was highlighted by incomplete separation and

hypoplasia of the telencephalic vesicles, along with absence of the medial nasal processes

that form the median aspect of the upper lip and nose. As shown in a wildtype control that

was hemisected in the sagittal plane, Nkx2.1 is normally expressed in the medial ganglionic

eminences and the ventral aspect of the diencephalon, while Pax6 is present in dorsal

aspects of the telencephalon and diencephalon (Corbin et al., 2003). Gli2-/- embryos

exhibited agenesis of the medial ganglionic eminences and absence of Nkx2.1 expression in

the telencephalon (Fig. 4G). Concurrently, the expression domain of Pax6 was expanded

into the ventral telencephalon (Fig. 4K). The altered forebrain patterning observed in Gli2-/-

embryos was largely recapitulated by vismodegib exposure. However, the diencephalic

domain of Nkx2.1 expression was drastically reduced in Gli2-/- embryos compared to those

exposed to vismodegib. The expression domains of Nkx2.1 and Pax6 were indistinguishable

between Gli2+/- and Gli2+/+ embryos.

We also examined expression of Shh and Gli1, with the latter serving as a reliable indicator of

Hh pathway activity. At GD11, Shh is normally expressed in the mantle region of the medial

ganglionic eminences, along the ventral aspect of the diencephalon, and in the zona limitans

intrathalamica (Fig. 4M). Gli1 expression reflects its paracrine responsiveness to stimulation

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

by secreted SHH. In Gli2-/- embryos, expression of Shh and Gli1 was nearly absent (Fig.

4O,S). In vismodegib-exposed wildtype embryos, Shh and Gli1 expression was diminished

but detectable in the ventral diencephalon and zona limitans intrathalamica (Fig. 4P, T).

Gli2 heterozygosity increases sensitivity to teratogen-induced HPE

In surviving human cohorts with HPE-associated phenotypes, variants in the GLI2 gene have

been identified as heterozygous loss of function mutations (Roessler et al., 2003). While

severe phenotypes have been observed in some GLI2 mutation carriers, most exhibit

incomplete HPE expressivity or are clinically normal. The data described above show that

mice with single-allele Gli2 mutations on the C57BL/6J background are phenotypically

indistinguishable from their wildtype littermates. These observations argue that GLI2 is

haplosufficient in the absence of additional genetic or environmental influences. To test

whether normally silent single-allele Gli2 mutations increase teratogenic sensitivity, direct

comparison was made between wildtype and heterozygous littermates by mating Gli2+/+

female and Gli2+/- male mice. As opposed to a 40mg/kg dose utilized in the experiments

described above, pregnant mice were exposed to a single dose of 2.5 or 10mg kg-1 vismodegib

at GD7.75. In the 2.5mg kg-1 vismodegib exposure group, HPE-associated facial

dysmorphology was detected in 15% (n=2/13) of Gli2+/- fetuses but in none (n=0/16) of the

wildtype littermates (Fig. 5). In the 10mg/kg vismodegib exposure group, dysmorphology

was detected in 67% (n=8/12) of Gli2+/- fetuses versus 17% (n=3/18) of wildtype fetuses. The

severity of vismodegib-induced facial dysmorphology was also dramatically increased in Gli2

heterozygous fetuses compared to their wildtype littermates. Histologic examination

confirmed that the degree of observed facial dysmorphology closely corresponds with the

severity of forebrain abnormalities.

We next investigated the effect of Gli2 gene dosage on cellular responsiveness to SHH ligand

using embryonic fibroblasts from B6 Gli2+/+, Gli2+/-, and Gli2-/- mice. As expected, Gli2

mRNA abundance was dependent upon the number of functional Gli2 alleles (Fig. 6A). In

Gli2+/+ cells, stimulation with SHH peptide resulted in significant upregulation of Hh target

genes Gli1 and Ptc1. Comparison of expression levels in Gli2+/+, Gli2+/- and Gli2-/- cells

revealed a significant gene dosage effect. Specifically, the SHH-induced expression of these

genes incrementally decreased with loss of each functional Gli2 allele (Fig. 6B-C). We then

examined the effect of Gli2 haploinsufficiency on pathway sensitivity to vismodegib. Gli2+/+

and Gli2+/- cells appeared equally responsive to pathway inhibition by vismodegib. However,

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

because of diminished capacity to respond to SHH stimulation, target gene expression was

lower in Gli2+/- cells at all vismodegib concentrations (Fig. 6D). Gli2+/- cells reached the level

of gene expression observed in Gli2-/- cells at a lower concentration of vismodegib than

Gli2+/+ cells. This in vitro result is consistent with the increased teratogenic sensitivity

observed in Gli2 heterozygous mice.

Discussion

Considerable effort has been made to elucidate the complex etiology of HPE, with recent

studies focusing on underlying genetic factors. However, most identified mutations are

heterozygous and demonstrate incomplete penetrance/expressivity. The tenuous connection

between GLI2 and HPE had typified this paradigm. While the defining medial forebrain

deficiency of true HPE has been identified in several individuals with single allele GLI2

mutations, most cases involve more subtle abnormalities, such as pituitary deficiency and/or

facial dysmorphology (Bear et al., 2014, Roessler et al., 2003, Bertolacini et al., 2012, Franca

et al., 2010, Rahimov et al., 2006). One explanation, proposed by Roessler et al., is that HPE

results from additional environmental or genetic influences superimposed on

the GLI2 haploinsufficent state (Roessler et al., 2003). This has become a widely accepted

overarching premise of HPE etiology, but while some interacting factors have been

identified, the majority of cases cannot be explained (Kietzman et al., 2014, Hong and

Krauss, 2012, Roessler and Muenke, 2010).

The studies described herein directly address the complex etiology of HPE, and provide a

framework to understand the extreme phenotypic variability observed in human GLI2

mutation carriers. We provide the first evidence definitively linking GLI2 to HPE, by

demonstrating that a homozygous mutation causes the defining brain and face

malformations. We also show that a single-allele mutation is normally silent but

dramatically increases sensitivity to HPE induced by low-dose teratogen exposure. These

findings establish a novel model to further elucidate the intricate gene-environment

interactions that have frustrated clinical management of this common human birth defect.

In addition to variable clinical findings, the uncertain role of GLI2 in HPE pathogenesis was

obfuscated by the initial characterization of Gli2 knockout mice. These mice were generated

and maintained on an outbred CD-1/129 background and described to have incompletely

penetrant abnormalities of the vertebrae and limbs, as well as secondary palate clefts,

pituitary hypoplasia, and midbrain deficiency (Mo et al., 1997, Matise et al., 1998, Park et al.,

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

2000). However, the forebrain and midfacial deficiency phenotypes characteristic of HPE

were not observed (Blaess et al., 2006). This discordance extended to cellular and molecular

findings. Gli2-/- fetuses on an outbred CD-1 background were found to have intact

expression of the ventral neurospecification marker Nkx2.1 and grossly normal medial

ganglionic eminences (Park et al., 2000). These transient developmental structures give rise

to inhibitory cortical interneurons, cell populations that are depleted in individuals with true

HPE (Fertuzinhos et al., 2009). Here, we demonstrate that when backcrossed to the

C57BL/6J background, homozygous Gli2 mutations cause the characteristic brain and face

abnormalities of severe HPE, including severe diminishment of the medial ganglionic

eminences, medial forebrain deficiency, hypotelorism, and midfacial hypoplasia. These

structural malformations followed diminished Hh signaling pathway activity and severely

disrupted forebrain patterning.

Backcrossing gene mutations to the C57BL/6 background has previously been shown to

reveal or exacerbate HPE penetrance and/or expressivity. Specific examples include

homozygous mutations in the Hh ligand co-receptor Cdo, a predicted truncating mutation in

Tgif, and single- allele mutations in Six3, a transcription factor upstream of Shh (Kuang et

al., 2006, Zhang et al., 2006, Geng et al., 2008). These reports, along with the findings

presented in this study, suggest that the C57BL/6 background includes one or more yet-to-

be-identified HPE modifier genes. Demonstration that GLI2 loss of function on the B6

inbred strain results in severe HPE phenotypes with complete phenotypic penetrance

provides a new opportunity to uncover background-specific interacting genetic variations

that may represent novel and clinically-relevant predisposing factors.

While this study focused on gene-environment interaction, gene-gene interactions have been

postulated to play a role in the complex etiology of HPE. As a basic test of this concept, we

generated mice heterozygous for Gli2 and Shh, the latter being the most commonly mutated

gene identified in in non-chromosomal HPE. When backcrossed to the C57BL/6J

background, Shh-/- embryos exhibited severe HPE phenotypes, including severe midfacial

hypoplasia and a proboscis (Fig. S5). However, Shh+/-Gli2+/- fetuses were apparently normal

and indistinguishable from single heterozygotes and wildtype littermates. This suggests that

Shh heterozygosity does not interact with Gli2 haploinsufficiency. A functional interaction

with a downstream gene is perhaps more likely given the role of GLI2 as a direct activator of

Hh target genes. However, the tissue specific Hh target genes that are regulated by GLI2 and

mediate the pathogenesis of HPE are not known. Moving forward, use of the

complementary genetic and teratogenic models of HPE characterized here provides an ideal

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

platform to identify Hh target genes involved in HPE pathogenesis and to test novel gene-

gene interactions.

While GLI2 loss of function causes chronic Hh pathway disruption, the effect of acute

pathway disruption can be examined by targeted exposure of the potent and specific

inhibitor vismodegib. We found that the anatomical and molecular phenotypes caused by

GLI2 loss of function were largely recapitulated by acute exposure to vismodegib at GD7.75.

This observation suggests that the abnormalities we observed in GD15 fetuses are caused in

part by changes to cell populations that respond to SHH signaling during the early

neurulation stage of embryogenesis. This period of development (GD8.0 and shortly after)

corresponds to the fourth week of human gestation and has been shown to be sensitive to

other chemicals that cause HPE (Lipinski et al., 2010). This premise was further supported

by genetic fate mapping using an inducible system that labels Hh-responsive cells and their

progeny at discrete periods of development. Tamoxifen exposure at GD7.75 demonstrated

Hh-responsive cell lineages in specific medial facial and forebrain compartments

corresponding with tissues that were deficient in Gli2-/- and vismodegib-exposed mice.

However, several differences were also observed between the genetic and teratogenic models

of HPE that we investigated. Gli2-/- animals exhibited more severe pituitary and

diencephalic abnormalities but subtler olfactory bulb deficiency than those exposed to

vismodegib. Loss of GLI2 function also resulted in a more extensive disruption of Hh

pathway activity and ventral specification in the diencephalic region of the forebrain. These

differences likely reflect, at least in part, the pathogenic effects of chronic versus acute

attenuation of the Hh pathway inherent in these models. Additional factors contributing to

these differences, like differential effects on Gli3 repression, are also possible and should be

considered in future investigation.

Directly addressing the premise of gene-environment interaction, our study demonstrates

that normally silent single-allele Gli2 mutations dramatically increase teratogenic sensitivity.

This result is congruous with our previous study examining the effect of prenatal alcohol

exposure in the context of Gli2 heterozygosity (Kietzman et al., 2014). These findings

illustrate that a functional predisposition can exacerbate the teratogenic effects of two

unique classes of environmental influences, resulting in severe birth defects with largely

overlapping phenotypes. Together these findings provide a construct for understanding the

extremely variable phenotypes exhibited in clinical populations, while highlighting the

emerging consensus that HPE and other birth defects likely result from complex interactions

of genetic and environmental factors.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

We utilized the potent Hh pathway antagonist vismodegib to test the concept of gene-

environment interaction in the context of Gli2 heterozygosity. Importantly, vismodegib is

part of a larger class of compounds that inhibit Hh signaling by binding the transmembrane

protein Smoothened (Robarge et al., 2009, Chen et al., 2002, Keeler, 1975, Chen, 2016).

With both synthetic and natural small molecules, this group includes more than 20

compounds and continues to grow with increasing attention focused on the biological effects

of Hh pathway inhibition. With respect to HPE etiology, perhaps the most intriguing Hh

pathway inhibitors are identified environmental small molecules, including dietary alkaloids,

an antifungal agent, a human dietary supplement, and a pesticide synergist present in

hundreds of insecticide formulations (Lipinski et al., 2007, Lipinski and Bushman, 2010,

Wang et al., 2012). These compounds are less potent than vismodegib and appear unlikely

to act independently to cause birth defects at typical environmental concentrations.

However, the findings presented here illustrate the “multiple hit” hypothesis of complex

disease and should prompt careful investigation of the etiological role of environmental Hh

pathway inhibitors in the context of predisposing mutations in genes like Gli2. Building

upon the study presented here, continued elucidation of the intricate gene-environment

interactions that cause HPE provides a direct path to improving clinical management and

developing effective prevention strategies.

Materials and Methods

Animals models. Mouse (Mus musculus) studies were carried out in strict accordance with

the recommendations in the Guide for the Care and Use of Laboratory Animals of the

National Institutes of Health. The protocol was approved by the University of Wisconsin

School of Veterinary Medicine Institutional Animal Care and Use Committee (protocol

number 13-081.0). Gli2+/- mice (Matise et al., 1998) were backcrossed to the C57BL/6J

background for more than 15 generations. C57BL/6J wildtype mice were purchased from

The Jackson Laboratory (Bar Harbor, ME, USA). Gli1CreERT2 (stock no: 007913) and

Rosa26lacZ (stock no: 003474) mice were purchased from The Jackson Laboratory. All mice

were housed under specific pathogen-free conditions in disposable, ventilated cages

(Innovive, San Diego, CA) in rooms maintained at 22 ±2 degrees Celsius and 30-70%

humidity on a 12 hr light, 12 hr dark cycle. Mice were fed 1919x Irradiated Harlan Teklad

Global Soy Protein-Free Extruded Rodent Diet. For timed matings, 1-3 female mice between

8 and 20 weeks of age were placed with a single male for 1-2 hrs and subsequently examined

for the presence of copulation plugs. The beginning of the mating period was designated as

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

GD0. True pregnancy was confirmed by assessing weight gain between GD7 and 10 before

dam treatment or embryo harvest as previously described (Heyne et al., 2015b).

Hedgehog signaling antagonist exposure. Vismodegib was purchased from LC Laboratories

(Woburn, MA, USA) was suspended as 3 mg ml-1 in 0.5% methyl cellulose with 0.2% tween

as previously described (Heyne et al., 2015a). Individual suspensions were prepared within

30 min of administration. Pregnant mice were administered 2.5, 10, or 40 mg kg-1

vismodegib (aka GDC-0449) by oral gavage at GD7.75.

Dissection, imaging, and fetal phenotyping. Pregnant dams were euthanized at indicated

stages of development by CO2 asphyxiation and subsequent cervical dislocation. Fetal

specimens were fixed in 10% formalin or Bouin’s solution for at least 1 week prior to imaging.

Images were captured using a micropublisher 5.0 camera connected to a Nikon SZX-10

stereomicroscope. Linear facial measurements of formalin-fixed GD15 fetuses, were

produced in Photoshop v14.1.2 as previously described (Lipinski et al., 2014). For

comparison of face and brain morphology, images of Bouin’s-fixed tissue were captured and

converted to grey-scale. The severity of HPE phenotypes in GD17 fetuses was assessed by a

single rater blinded to treatment based upon a semi quantitative scale as previously

described (Heyne et al., 2015a, Kietzman et al., 2014).

Histology. GD15 fetuses were fixed in Bouin’s solution for at least one week, and then

transferred to 70% ethanol. GD17 fetuses were fixed in 10% formalin. Following paraffin

embedding, 10μm sections were produced and stained with H&E by standard protocols.

Fate mapping. Gli1CreERT2+/- male mice were mated with Rosa26lacZflox/flox females. 50mg

ml-1 Tamoxifen dissolved in corn oil was administered to pregnant dams by IP injection at

GD7.75. GD15 fetuses were fixed overnight in 2% paraformaldehyde with 0.2%

glutaraldehyde. Embryonic tissues were then embedded in 4% agarose and 150μm coronal

sections were produced by vibrating microtome and x-gal stained as described previously

(Mehta et al., 2011).

In situ hybridization. ISH was performed using an established high-throughput technique

that allows multiple treatment groups to be processed identically and as a single unit (Abler

et al., 2011). For GD11 embryos, mid-sagittal hemisection using a scalpel was performed

before staining. Hemisected embryos were incubated in proteinase K (2.5µg ml-

1)/collagenase (250µg ml-1) for 2 minutes prior to initial washes. Younger embryos were not

subjected to proteinase K/collagenase treatment. ISH probe primers were resuspended as

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

stock solutions of 100mM in TE buffer pH7.0 (Ambion). Working stocks were made as

10mM solutions containing both forward and reverse primers. Primer sequence for probe

generation are listed in supplemental table S1.

Mouse embryonic fibroblast isolation and treatment. Mouse embryonic fibroblasts were

harvested from GD15 embryos as previously described (Lipinski et al., 2006). Cells were

grown to confluence in DMEM [with l-glutamine, 4.5g L-1 glucose, without sodium pyruvate],

with 10% FCS and 1% Pen/Strep. They were then trypsinized and plated in 24 or 48 well

tissue culture plates (Falcon, Franklin Lakes, NJ, USA) at 4.0 × 105 cells ml-1 media. Cells

were allowed to attach for 24 hr and media were replaced with DMEM containing 1% FBS ±

SHH-N peptide (R&D Systems, Minneapolis, MN, USA) at 0.4 μg ml-1, ± vismodegib (LC

Laboratories) at indicated concentrations. For cell culture experiments, vismodegib was

dissolved in DMSO.

RNA isolation and Real Time RT-PCR. RNA was isolated using GE Illustra RNAspin

kits. 100 ng of RNA was reverse transcribed using the Promega GoScript Reverse

Transcription System. Both kits were used according to manufacturer's instructions. Real-

time PCR was performed using a BioRad CFX96 Touch real-time PCR detection system.

Reaction mixtures contained 6 µl SSoFast EvaGreen Supermix (BioRad Laboratories,

Hercules, CA, USA), 4.75 µl ddH2O, 0.75 µl cDNA, and 0.5 µl 10 mM combined forward and

reverse gene-specific primers. Primers were resuspended as stock solutions of 100mM in TE

buffer pH7.0. Working stocks were made as 10mM solutions containing both forward and

reverse primers. Primer sequences are listed in Table S2. Reaction conditions were as

follows: 1 cycle at 95°C for 3 min, then 40 cycles of 95°C for 10 s followed by 30 s at 60°C

(annealing temperature). To confirm specificity, primer sequences were analyzed with

BLAST and melt curves were examined for a single peak in the expected temperature range.

Gapdh was used as the housekeeping gene and analysis was conducted with the 2^-ddCt

method.

Statistics. Analysis of linear measurements was made using one-way ANOVA followed by

Tukey's HSD test using Graphpad Prism software (v6.04). Differences in the frequency of

facial dysmorphology between experimental groups were assessed by one-tailed Fisher’s

exact test using Graphpad Prism software. T-tests were used to determine whether gene

expression was changed by SHH stimulation in mouse embryonic fibroblasts. To test

whether basal and Shh-induced target gene expression preserved the ordering implicitly

suggested by the genotypes, a Jonckheere–Terpstra test was performed using Mstat version

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

6.1.4 (http:/mcardle.oncology.wisc.edu/mstat/download/index.html). An alpha value of

0.05 was maintained for all analyses.

Human images. Parental written informed consent was received for inclusion and

publication of patient images in Figure 1.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Acknowledgements

The authors thank Dr. Ruth Sullivan for consultation of study design, discussion of results,

and review of the manuscript. We also thank Dr. Paul Kruszka and the patient’s family for

use of images depicting the clinical manifestation of HPE.

Competing Interests

No competing interests declared.

Author Contributions

RJL, GWH, LAW designed studies; JLE, CGM, LAW, GWH, DMF, KFP, PD, CMU, and RJL

conducted experiments and acquired data; JLE, KFP, DMF, and RJL analyzed data; RJL

wrote the manuscript.

Funding

This work was supported by grants R00DE022101 to RJL from the National Institute of

Dental and Craniofacial Research/National Institutes of Health, T32ES007015-37 to JLE

from the National Institute of Environmental Health Sciences/National Institutes of Health,

T35OD011078 to LAW from the National Institutes of Health/Office of the Director, and

pilot funding from the University of Wisconsin Molecular and Environmental Toxicology

Center.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

References ABLER, L. L., MEHTA, V., KEIL, K. P., JOSHI, P. S., FLUCUS, C. L., HARDIN, H. A., SCHMITZ, C. T. &

VEZINA, C. M. 2011. A high throughput in situ hybridization method to characterize mRNA expression patterns in the fetal mouse lower urogenital tract. J Vis Exp.

AHN, S. & JOYNER, A. L. 2005. In vivo analysis of quiescent adult neural stem cells responding to Sonic hedgehog. Nature, 437, 894-7.

BEAR, K. A., SOLOMON, B. D., ANTONINI, S., ARNHOLD, I. J., FRANÇA, M. M., GERKES, E. H., GRANGE, D. K., HADLEY, D. W., JÄÄSKELÄINEN, J., PAULO, S. S., RUMP, P., STRATAKIS, C. A., THOMPSON, E. M., WILLIS, M., WINDER, T. L., JORGE, A. A., ROESSLER, E. & MUENKE, M. 2014. Pathogenic mutations in GLI2 cause a specific phenotype that is distinct from holoprosencephaly. J Med Genet, 51, 413-8.

BERTOLACINI, C. D., RIBEIRO-BICUDO, L. A., PETRIN, A., RICHIERI-COSTA, A. & MURRAY, J. C. 2012. Clinical findings in patients with GLI2 mutations--phenotypic variability. Clin Genet, 81, 70-5.

BLAESS, S., CORRALES, J. D. & JOYNER, A. L. 2006. Sonic hedgehog regulates Gli activator and repressor functions with spatial and temporal precision in the mid/hindbrain region. Development, 133, 1799-809.

CHEN, J. K. 2016. I only have eye for ewe: the discovery of cyclopamine and development of Hedgehog pathway-targeting drugs. Nat Prod Rep.

CHEN, J. K., TAIPALE, J., YOUNG, K. E., MAITI, T. & BEACHY, P. A. 2002. Small molecule modulation of Smoothened activity. Proc Natl Acad Sci U S A, 99, 14071-6.

CHIANG, C., LITINGTUNG, Y., LEE, E., YOUNG, K. E., CORDEN, J. L., WESTPHAL, H. & BEACHY, P. A. 1996. Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature, 383, 407-13.

COHEN, M. M. 2006. Holoprosencephaly: Clinical, anatomic, and molecular dimensions. Birth Defects Research Part a-Clinical and Molecular Teratology, 76, 658-673.

CORBIN, J. G., RUTLIN, M., GAIANO, N. & FISHELL, G. 2003. Combinatorial function of the homeodomain proteins Nkx2.1 and Gsh2 in ventral telencephalic patterning. Development, 130, 4895-906.

FERTUZINHOS, S., KRSNIK, Z., KAWASAWA, Y. I., RASIN, M. R., KWAN, K. Y., CHEN, J. G., JUDAS, M., HAYASHI, M. & SESTAN, N. 2009. Selective depletion of molecularly defined cortical interneurons in human holoprosencephaly with severe striatal hypoplasia. Cereb Cortex, 19, 2196-207.

FRANCA, M. M., JORGE, A. A., CARVALHO, L. R., COSTALONGA, E. F., VASQUES, G. A., LEITE, C. C., MENDONCA, B. B. & ARNHOLD, I. J. 2010. Novel Heterozygous Nonsense GLI2 Mutations in Patients with Hypopituitarism and Ectopic Posterior Pituitary Lobe without Holoprosencephaly. J Clin Endocrinol Metab, 95, E384-91.

GENG, X., SPEIRS, C., LAGUTIN, O., INBAL, A., LIU, W., SOLNICA-KREZEL, L., JEONG, Y., EPSTEIN, D. J. & OLIVER, G. 2008. Haploinsufficiency of Six3 fails to activate Sonic hedgehog expression in the ventral forebrain and causes holoprosencephaly. Dev Cell, 15, 236-47.

HEYNE, G. W., MELBERG, C. G., DOROODCHI, P., PARINS, K. F., KIETZMAN, H. W., EVERSON, J. L., ANSEN-WILSON, L. J. & LIPINSKI, R. J. 2015a. Definition of critical periods for Hedgehog pathway antagonist-induced holoprosencephaly, cleft lip, and cleft palate. PLoS One, 10, e0120517.

HEYNE, G. W., PLISCH, E. H., MELBERG, C. G., SANDGREN, E. P., PETER, J. A. & LIPINSKI, R. J. 2015b. A Simple and Reliable Method for Early Pregnancy Detection in Inbred Mice. J Am Assoc Lab Anim Sci, 54, 368-71.

HONG, M. & KRAUSS, R. S. 2012. Cdon mutation and fetal ethanol exposure synergize to produce midline signaling defects and holoprosencephaly spectrum disorders in mice. PLoS Genet, 8, e1002999.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

KEELER, R. F. 1975. Teratogenic effects of cyclopamine and jervine in rats, mice and hamsters. Proc Soc Exp Biol Med, 149, 302-6.

KIETZMAN, H. W., EVERSON, J. L., SULIK, K. K. & LIPINSKI, R. J. 2014. The teratogenic effects of prenatal ethanol exposure are exacerbated by Sonic Hedgehog or GLI2 haploinsufficiency in the mouse. PLoS One, 9, e89448.

KRAUSS, R. S. & HONG, M. 2016. Gene-Environment Interactions and the Etiology of Birth Defects. Curr Top Dev Biol, 116, 569-80.

KUANG, C., XIAO, Y., YANG, L., CHEN, Q., WANG, Z., CONWAY, S. J. & CHEN, Y. 2006. Intragenic deletion of Tgif causes defectsin brain development. Hum Mol Genet, 15, 3508-19.

LEONCINI, E., BARANELLO, G., ORIOLI, I. M., ANNERÉN, G., BAKKER, M., BIANCHI, F., BOWER, C., CANFIELD, M. A., CASTILLA, E. E., COCCHI, G., CORREA, A., DE VIGAN, C., DORAY, B., FELDKAMP, M. L., GATT, M., IRGENS, L. M., LOWRY, R. B., MARASCHINI, A., MC DONNELL, R., MORGAN, M., MUTCHINICK, O., POETZSCH, S., RILEY, M., RITVANEN, A., GNANSIA, E. R., SCARANO, G., SIPEK, A., TENCONI, R. & MASTROIACOVO, P. 2008. Frequency of holoprosencephaly in the International Clearinghouse Birth Defects Surveillance Systems: searching for population variations. Birth Defects Res A Clin Mol Teratol, 82, 585-91.

LIPINSKI, R. J. & BUSHMAN, W. 2010. Identification of Hedgehog signaling inhibitors with relevant human exposure by small molecule screening. Toxicol In Vitro, 24, 1404-9.

LIPINSKI, R. J., DENGLER, E., KIEHN, M., PETERSON, R. E. & BUSHMAN, W. 2007. Identification and characterization of several dietary alkaloids as weak inhibitors of hedgehog signaling. Toxicol Sci, 100, 456-63.

LIPINSKI, R. J., GIPP, J. J., ZHANG, J., DOLES, J. D. & BUSHMAN, W. 2006. Unique and complimentary activities of the Gli transcription factors in Hedgehog signaling. Exp Cell Res, 312, 1925-38.

LIPINSKI, R. J., GODIN, E. A., O'LEARY-MOORE, S. K., PARNELL, S. E. & SULIK, K. K. 2010. Genesis of teratogen-induced holoprosencephaly in mice. Am J Med Genet C Semin Med Genet, 154C, 29-42.

LIPINSKI, R. J., HOLLOWAY, H. T., O'LEARY-MOORE, S. K., AMENT, J. J., PECEVICH, S. J., COFER, G. P., BUDIN, F., EVERSON, J. L., JOHNSON, G. A. & SULIK, K. K. 2014. Characterization of subtle brain abnormalities in a mouse model of hedgehog pathway antagonist-induced cleft lip and palate. PLoS One, 9, e102603.

MATISE, M. P., EPSTEIN, D. J., PARK, H. L., PLATT, K. A. & JOYNER, A. L. 1998. Gli2 is required for induction of floor plate and adjacent cells, but not most ventral neurons in the mouse central nervous system. Development, 125, 2759-70.

MATSUNAGA, E. & SHIOTA, K. 1977. Holoprosencephaly in human embryos: epidemiologic studies of 150 cases. Teratology, 16, 261-72.

MEHTA, V., ABLER, L. L., KEIL, K. P., SCHMITZ, C. T., JOSHI, P. S. & VEZINA, C. M. 2011. Atlas of Wnt and R-spondin gene expression in the developing male mouse lower urogenital tract. Dev Dyn, 240, 2548-60.

MERCIER, S., DUBOURG, C., BELLEGUIC, M., PASQUIER, L., LOGET, P., LUCAS, J., BENDAVID, C. & ODENT, S. 2010. Genetic counseling and "molecular" prenatal diagnosis of holoprosencephaly (HPE). Am J Med Genet C Semin Med Genet, 154C, 191-6.

MO, R., FREER, A. M., ZINYK, D. L., CRACKOWER, M. A., MICHAUD, J., HENG, H. H., CHIK, K. W., SHI, X. M., TSUI, L. C., CHENG, S. H., JOYNER, A. L. & HUI, C. 1997. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development, 124, 113-23.

ORIOLI, I. M. & CASTILLA, E. E. 2010. Epidemiology of holoprosencephaly: Prevalence and risk factors. Am J Med Genet C Semin Med Genet, 154C, 13-21.

PARK, H. L., BAI, C., PLATT, K. A., MATISE, M. P., BEEGHLY, A., HUI, C. C., NAKASHIMA, M. & JOYNER, A. L. 2000. Mouse Gli1 mutants are viable but have defects in SHH signaling in combination with a Gli2 mutation. Development, 127, 1593-605.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

PINEDA-ALVAREZ, D. E., SOLOMON, B. D., ROESSLER, E., BALOG, J. Z., HADLEY, D. W., ZEIN, W. M., HADSALL, C. K., BROOKS, B. P. & MUENKE, M. 2011. A broad range of ophthalmologic anomalies is part of the holoprosencephaly spectrum. Am J Med Genet A, 155A, 2713-20.

RAHIMOV, F., RIBEIRO, L. A., DE MIRANDA, E., RICHIERI-COSTA, A. & MURRAY, J. C. 2006. GLI2 mutations in four Brazilian patients: how wide is the phenotypic spectrum? Am J Med Genet A, 140, 2571-6.

RICHIERI-COSTA, A. & RIBEIRO, L. A. 2010. Holoprosencephaly and holoprosencephaly-like phenotypes: Review of facial and molecular findings in patients from a craniofacial hospital in Brazil. Am J Med Genet C Semin Med Genet, 154C, 149-57.

ROBARGE, K. D., BRUNTON, S. A., CASTANEDO, G. M., CUI, Y., DINA, M. S., GOLDSMITH, R., GOULD, S. E., GUICHERT, O., GUNZNER, J. L., HALLADAY, J., JIA, W., KHOJASTEH, C., KOEHLER, M. F., KOTKOW, K., LA, H., LALONDE, R. L., LAU, K., LEE, L., MARSHALL, D., MARSTERS, J. C., MURRAY, L. J., QIAN, C., RUBIN, L. L., SALPHATI, L., STANLEY, M. S., STIBBARD, J. H., SUTHERLIN, D. P., UBHAYAKER, S., WANG, S., WONG, S. & XIE, M. 2009. GDC-0449-a potent inhibitor of the hedgehog pathway. Bioorg Med Chem Lett, 19, 5576-81.

ROESSLER, E., BELLONI, E., GAUDENZ, K., JAY, P., BERTA, P., SCHERER, S. W., TSUI, L. C. & MUENKE, M. 1996. Mutations in the human Sonic Hedgehog gene cause holoprosencephaly. Nat Genet, 14, 357-60.

ROESSLER, E., DU, Y. Z., MULLOR, J. L., CASAS, E., ALLEN, W. P., GILLESSEN-KAESBACH, G., ROEDER, E. R., MING, J. E., RUIZ I ALTABA, A. & MUENKE, M. 2003. Loss-of-function mutations in the human GLI2 gene are associated with pituitary anomalies and holoprosencephaly-like features. Proc Natl Acad Sci U S A, 100, 13424-9.

ROESSLER, E. & MUENKE, M. 2010. The molecular genetics of holoprosencephaly. Am J Med Genet C Semin Med Genet, 154C, 52-61.

SOLOMON, B. D., BEAR, K. A., WYLLIE, A., KEATON, A. A., DUBOURG, C., DAVID, V., MERCIER, S., ODENT, S., HEHR, U., PAULUSSEN, A., CLEGG, N. J., DELGADO, M. R., BALE, S. J., LACBAWAN, F., ARDINGER, H. H., AYLSWORTH, A. S., BHENGU, N. L., BRADDOCK, S., BROOKHYSER, K., BURTON, B., GASPAR, H., GRIX, A., HOROVITZ, D., KANETZKE, E., KAYSERILI, H., LEV, D., NIKKEL, S. M., NORTON, M., ROBERTS, R., SAAL, H., SCHAEFER, G. B., SCHNEIDER, A., SMITH, E. K., SOWRY, E., SPENCE, M. A., SHALEV, S. A., STEINER, C. E., THOMPSON, E. M., WINDER, T. L., BALOG, J. Z., HADLEY, D. W., ZHOU, N., PINEDA-ALVAREZ, D. E., ROESSLER, E. & MUENKE, M. 2012. Genotypic and phenotypic analysis of 396 individuals with mutations in Sonic Hedgehog. J Med Genet, 49, 473-9.

SOLOMON, B. D., PINEDA-ALVAREZ, D. E., MERCIER, S., RAAM, M. S., ODENT, S. & MUENKE, M. 2010. Holoprosencephaly flashcards: A summary for the clinician. Am J Med Genet C Semin Med Genet, 154C, 3-7.

WANG, J., LU, J., MOOK, R. A., ZHANG, M., ZHAO, S., BARAK, L. S., FREEDMAN, J. H., LYERLY, H. K. & CHEN, W. 2012. The insecticide synergist piperonyl butoxide inhibits hedgehog signaling: assessing chemical risks. Toxicol Sci, 128, 517-23.

ZHANG, W., KANG, J. S., COLE, F., YI, M. J. & KRAUSS, R. S. 2006. Cdo functions at multiple points in the Sonic Hedgehog pathway, and Cdo-deficient mice accurately model human holoprosencephaly. Dev Cell, 10, 657-65.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Figures

Fig. 1. GLI2 loss of function causes HPE-associated facial dysmorphology. (A-D)

GD15 B6 Gli2+/+, Gli2+/-, Gli2-/- fetuses are shown along with a Gli2+/+

fetus exposed to 40mg

kg-1 vismodegib (Vis) at GD7.75. Snout width (SW), interocular distance (IOD), and head

width (HW) was measured as illustrated by dashed lines shown in E. (F) Linear

measurements were normalized to Gli2+/+ control group values and shown on a semi-log

plot. Values represent the mean ± s.e.m. (Gli2+/+ n=6; Gli2+/- n=9; Gli2-/-

n=6; Gli2+/++ Vis,

n=12). * p≤0.05 by one-way ANOVA followed by Tukey's HSD. Scale bar = 1 mm.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Fig. 2. Facial dysmorphology in Gli2-/- fetuses co-occurs with brain

abnormalities. (A-D) Shown are animals from the same experimental groups described in

Fig. 1. To visualize correlative phenotypes, images of the dorsal surface of the brain along

with facial images of the same animal. Images were captured following Bouin’s fixation and

converted to grayscale. In the Gli2-/- brain, hypoplasia of the cerebral cortices (cc) and

midbrain (mb) is apparent. The olfactory bulbs (ofb) are hypoplastic and abnormally

approximated. In the Gli2+/+ fetus exposed to vismodegib, more severe deficiency of the

cerebral cortices and absence of the olfactory bulbs is observed. Scale bar = 1 mm.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Fig. 3. GLI2 loss of function results in deficiency of medial forebrain and facial

tissue. (A-P) Serial histological images are shown from the same experimental groups

described in Figs. 1 and 2. Dashed boxes in low magnification images on the left provide

context for the coronal sections shown in each row. Gli2-/- and Gli2+/+

vismodegib-exposed

fetuses exhibit absence of nasal septal (ns) cartilage and diminished vomeronasal organs

(vo). The midline connective tissue separating the olfactory bulbs (ofb) marked by the (*) is

deficient in the Gi2-/- fetus, while the olfactory bulbs are absent in the Gli2+/+ vismodegib-

exposed fetus. These

fetuses also have severe medial forebrain deficiencies, including

absence of the septal region (s) and communicating lateral ventricles (lv). In the Gli2-/- fetus,

attenuation of the third ventricle (tv) and absence of the anterior pituitary (ap) lobe are also

apparent. The anterior pituitary is present but hypoplastic in the Gli2+/+ vismodegib-

exposed fetus, which also exhibits subtle attenuation of the third ventricle. (Q-T) A genetic

fate mapping system (23) was used to identify Hh-responsive cell lineages. Embryos

exposed to Tamoxifen at GD7.75 were sectioned and stained with X-gal to visualize Hh-

responsive cells and their progeny. 100 uM sections were produced at planes comparable to

those stained by H&E. Tongue (t). Scale bar = 0.25 mm.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Fig. 4. GLI2 loss of function results in altered dorsal-ventral forebrain

patterning. Shown are GD11 embryos of the same experimental groups described in Figs 1-

3. (A-D) Frontal images show forebrain and facial morphology. Gli2-/- and Gli2+/+ fetuses

exposed to vismodegib exhibit hypoplasia and abnormal approximation of the telencephalic

vesicles (t), absence of the medial nasal processes (mnp), and a single central nostril (n).

Embryos were hemisected near the midline and subjected to ISH staining for the indicated

genes. In the Gli2+/+ control, Nkx2.1 is expressed in the medial ganglionic eminences (mge)

and the ventral aspect of the diencephalon (d), while Pax6 is present in the dorsal aspect of

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

the telencephalon and diencephalon. Shh is expressed in the mantle region of the medial

ganglionic eminence, along the ventral aspect of the diencephalon, and in the zona limitans

intrathalamica (zli). Gli1 expression reflects its paracrine responsiveness to stimulation by

secreted SHH. Eye (e). Scale bar = 1 mm.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Fig. 5. Gli2 heterozygosity increases teratogenic sensitivity. Wildtype dams

carrying Gli2+/+ and Gli2+/- embryos were exposed to 2.5 or 10mg kg-1 vismodegib at GD7.75.

The graph at the left shows percentage of GD17 fetuses classified as having mild, moderate,

or severe HPE-associated facial dysmorphology. * p<0.05 by Fisher’s exact test. n=5

litters/treatment group. (A-D) Categories of facial morphology are illustrated by GD17

fetuses of indicated genotype and dose group. Animals that were indistinguishable from

controls were assigned “normal” (A). Those with a diminished area of pigmentation between

the nostrils were assigned as “mild” (B). Animals with a gross deficiency of the median lip

notch but some remaining pigment remaining at the tip of the nose were assigned as

“moderate” (C). Those with an absent lip notch and single central nostril were assigned

“severe” (D). (A’-D’) Histologic sections through the cerebral cortices illustrate that the

degree of facial dysmorphology corresponds to forebrain abnormalities. Note deficiency of

the septal region (s) in (B’) and absent septal region and single communicating lateral

ventricle in (C’) and (D’). Scale bar = 1 mm.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Adv

ance

art

icle

Fig. 6. Gli2 heterozygosity diminishes SHH-ligand responsiveness. (A-C) Gli2+/+,

Gli2+/-, Gli2-/- mouse embryonic fibroblasts (MEFs) were treated ± SHH ligand for 48 hrs.

Expression of Gli2 and the conserved Hh target genes Gli1 and Ptc1 were measured by Real

Time RT-PCR and normalized to Gapdh. *p<0.05 as determined by t-test for genotype-

specific gene expression. Arrows indicate a significant gene dosage effect determined by

Jonckheere-Terpstra test. (D) Gli2+/+ and Gli2+/- fibroblasts were treated ± SHH ligand and

graded concentrations of vismodegib for 48 hrs. Induction of Ptc1 (SHH/Vehicle) relative to

Gli2+/+ cells is shown in the semi-log plot. The dashed line shows relative Ptc1 induction in

Gli2-/- fibroblasts to illustrate a theoretical threshold of teratogenicity. For each graph,

values represent the mean ± s.e.m. of 6 replicate experiments.