© 2019. 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/4.0), which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.

Mandibular dysmorphology due to abnormal embryonic osteogenesis in FGFR2-related

craniosynostosis mice

Susan M. Motch Perrine1,*, Meng Wu2,*,‡, Nicholas B. Stephens1, Divya Kriti2, Harm van Bakel2,

Ethylin Wang Jabs2, Joan T. Richtsmeier1

1Department of Anthropology, Pennsylvania State University, University Park, PA, USA

2Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai, New

York, NY, USA

*These authors contributed equally to the manuscript.

‡Author for correspondence (meng.wu@mssm.edu)

ORCID ID: Meng Wu, 0000-0001-5512-9651

KEY WORDS: Apert syndrome, Crouzon syndrome, Pfeiffer syndrome, Osteoclast, Cartilage,

Transcriptome

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

http://dmm.biologists.org/lookup/doi/10.1242/dmm.038513Access the most recent version at First posted online on 7 May 2019 as 10.1242/dmm.038513

SUMMARY STATEMENT

Mandibular dysmorphology was observed in FGFR2-related craniosynostosis mouse models.

FGFR2 gain-of-function mutations differentially affect cartilage formation and

intramembranous ossification of dermal bone, resulting in abnormal embryonic osteogenesis of

the mandible.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

ABSTRACT

One diagnostic feature of craniosynostosis syndromes is mandibular dysgenesis. Using three

mouse models of Apert, Crouzon, and Pfeiffer craniosynostosis syndromes, we investigated

how embryonic development of the mandible is affected by fibroblast growth factor receptor 2

(FGFR2) mutations. Quantitative analysis of skeletal form at birth revealed differences in

mandibular morphology between mice carrying Fgfr2 mutations and their littermates that do

not carry the mutations. Murine embryos with the mutations associated with Apert syndrome

in humans showed an increase in the size of the osteogenic anlagen and Meckel’s cartilage

(MC). Changes in the microarchitecture and mineralization of the developing mandible were

visualized using histological staining. The mechanism for mandibular dysgenesis in the Apert

Fgfr2+/S252W mouse resulting in the most severe phenotypic effects was further analyzed in

detail and found to occur to a lesser degree in the other craniosynostosis mouse models. Laser

capture microdissection and RNA-Seq analysis revealed transcriptome changes in mandibular

bone at E16.5, highlighting increased expression of genes related to osteoclast differentiation

and dysregulated genes active in bone mineralization. Increased osteoclastic activity was

corroborated by TRAP assay and in situ hybridization of Csf1r and Itgb3. Upregulated expression

of Enpp1 and Ank was validated in the mandible of Fgfr2+/S252W embryos resulting in elevated

inorganic pyrophosphate concentration. Increased proliferation of osteoblasts in the mandible

and chondrocytes forming MC was identified in Fgfr2+/S252W embryos at E12.5. These findings

provide evidence that FGFR2 gain-of-function mutations differentially affect cartilage formation

and intramembranous ossification of dermal bone contributing to mandibular

dysmorphogenesis in craniosynostosis syndromes.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

INTRODUCTION

The mandible has been used as a model for investigating how complex morphological

structures arise during development and how they are altered during evolution (Atchley and

Hall, 1991) providing insight into how the spatial and temporal organization underlying the

development of a separate morphological component assimilates into a functioning whole.

Each hemimandible (or dentary) is composed of two functional areas that are mineralized

proximate to Meckel’s cartilage (MC): the anterior body (tooth bearing portion of the

hemimandible) and ramus (containing three prominent processes: coronoid process on the

dorsal aspect, the condylar (or condyloid) process caudally, and the angular process

caudoventrally). Mandible and maxilla are dermal bones derived from neural crest cells that

migrate to the first pharyngeal arch and are the result of complex developmental patterning

(Couly et al., 1996; Depew et al., 2002; Frisdal and Trainor, 2014; Noden, 1983). Together they

form the lower and upper jaws of the facial skeleton whose synchronous development and

proper occlusion is necessary for feeding, respiration, and craniofacial morphogenesis.

Mutations within fibroblast growth factor receptor 2 (FGFR2) are responsible for aberrant

signaling within the FGF-signaling pathway resulting in midface developmental anomalies that

are features of Apert, Crouzon, Pfeiffer, Beare-Stevenson cutis gyrata, Jackson-Weiss, and Bent

Bone Dysplasia syndromes (Azoury et al., 2017; Cohen and MacLean, 2000; Cunningham et al.,

2007). These complex conditions involve the premature fusion of one or more cranial sutures

and midfacial dysgenesis and are often associated with other skeletal and soft tissue

abnormalities (Flaherty et al., 2016; Heuzé et al., 2014). Midfacial dysgenesis can be severe but

is variable within and across FGFR2-related craniosynostosis syndromes. Surgical correction and

reconstruction are adaptable, targeting the midfacial skeleton, dental arcade, choanae, and/or

airway, often requiring significant and multiple reconstructive procedures. The mandible, the

major skeletal element of the lower face, is an important consideration in surgical planning and

orthodontic management in craniosynostosis syndromes to address severe anomalies affecting

mastication and airway anomalies.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Apert, Crouzon, and Pfeiffer syndromes (MIM #101200, MIM #123500, MIM #101600,

respectively) are autosomal dominant conditions sharing many phenotypic similarities,

including premature suture closure, abnormal facies, exophthalmos, midfacial retrusion, dental

malocclusion of varying intensities, cranial base anomalies, and dysmorphic mandibles whose

configuration is discordant with the upper jaw (Cohen and MacLean, 2000). Although the

mandible is not well-studied in these syndromes, adult mandibular morphology in Apert

patients is usually reported as intrinsically normal, and detected differences in mandibular

length are thought to be secondary to midfacial dysgenesis (Lemire, 2000; Wink et al., 2013).

The apparent mandibular prognathism is thought to be relative, a condition resulting in relation

to anomalies of the cranial base and severe retrusion of the midface (Costaras-Volarich and

Pruzansky, 1984). Why the degree and nature of developmental anomalies of the lower face

would be different from the midface when mandible and maxillae are both dermal bones of

neural crest origin derived from the first pharyngeal arch is not clear.

We have previously reported statistical differences in craniofacial bone morphology, brain

morphology, soft tissue and negative space (nasopharynx, inner ear) volumes, and

morphological integration of brain and skull in mouse models for Apert and Crouzon/Pfeiffer

syndromes relative to their respective littermates that do not carry the mutation and show no

phenotypic effects (Aldridge et al., 2010; Holmes et al., 2018; Martínez-Abadías et al., 2013;

Motch Perrine et al., 2014; Motch Perrine et al., 2017). However, investigations of the mandible

have not been included in any of these studies. To test the hypothesis that FGFR2 mutations

causative for craniosynostosis syndromes target processes and mechanisms of mandibular

genesis, we present data on the developmental and morphological consequences of three

unique FGFR2 mutations associated with syndromic craniosynostosis in the mandible of the

mouse. We performed quantitative morphometric analysis of 3D µCT image data of three

mouse models with differing activating Fgfr2 mutations to determine the differential effects of

these mutations on the mandible: two Apert syndrome mouse models, Fgfr2+/S252W (Wang et al.,

2005) and Fgfr2+/P253R(Wang et al., 2010) on a C57BL6/J background, and a mouse model with a

mutation associated with Crouzon and Pfeiffer syndromes, Fgfr2cC342Y/+ (Eswarakumar et al.,

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

2004) on a CD1 background. These analyses revealed significant differences in mandibular

morphology at P0 in all three of the mouse models. Based on these findings, we performed

histological analysis on the mandible at embryonic stages. Of these three mouse models,

mandibles of Fgfr2+/S252W mice showed the greatest magnitude of morphological change at P0

and histologic differences at embryonic day (E) 16.5 and were further analyzed by

transcriptome analysis to reveal cellular and molecular dysregulation contributing to

mandibular dysgenesis. Increased osteoclastogenesis causes abnormal bone resorption and

overexpression of Enpp1 and Ank that are key regulators for inorganic pyrophosphate levels

inhibiting bone mineralization in Fgfr2+/S252W mandible. FGFR2 S252W mutation was associated

with increased proliferation of osteoblasts and chondrocytes in the mandible as early as E12.5.

We provide new information about the molecular processes affecting the mandible in FGFR2-

related craniosynostosis syndromes to improve our understanding of craniofacial dysgenesis

and move us closer to therapeutic approaches for patients.

RESULTS

Mandibular dysmorphology of FGFR2-related craniosynostosis mouse models

The left and right hemimandibles of 182 newborn (P0) mice of each of the three

craniosynostosis models of interest were analyzed morphometrically using the 3D coordinates

of 32 landmarks (lms) (Fig. S1 and Table S1). Landmark datasets characterizing whole mandibles

(consisting of right and left sides (32 lms)), left hemimandibles (16 lms), and right

hemimandibles (16 lms) were analyzed using the same morphometric methods. Within each

model, morphometric analyses compared mice carrying a specific mutation to littermates that

did not carry the mutation. Results revealed a lack of asymmetry in mandibular dysmorphology

in all models, such that the right and left hemimandibles were similarly affected (Table S2). For

clarity of presentation, analysis of left hemimandible is presented graphically (Fig 1A-C).

Two morphometric methods were used. Euclidean Distance Matrix Analysis (EDMA) (Lele

and Richtsmeier, 2001) revealed significant differences in hemimandible shape between all

mice carrying Fgfr2 mutations and their respective littermates that did not carry the mutation

(referred to as “unaffected littermates”) for the full 32 lms set (P≤0.001) and for all three

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

regional landmark subsets representing major functional regions of the left dentary: left

hemimandible (16 lms) (P<0.001); ramus (10 lms) (P<0.001); body (8 lms) (P<0.05).

Bootstrapped confidence intervals for differences in each linear distance obtained from EDMA

reveal statistically significant differences in the localized patterns of mutational effects on the

3D morphology of the hemimandible (Fig. 1D-F). Fgfr2cC342Y/+ mice show effects that are

generally of a lesser magnitude and of a different pattern when compared to the effects of the

other two mutations (Fig. 1F). There are obvious similarities in the way hemimandibles of

Fgfr2+/S252W and Fgfr2+/P253R mutant mice differ from their respective unaffected littermates.

However, the significant phenotypic effects of the FGFR2 S252W mutation on mandibular

morphology are more numerous, of greater magnitude, and located primarily in the posterior

components of the hemimandible (Fig. 1D, E).

Principal components analysis (PCA) of the scale-free shape data shows obvious separation

of hemimandibles of mice carrying an Fgfr2 mutation from their respective unaffected

littermates for each of the three mutation groups (Fig. 1A-C). PCAs of shape were also

conducted using linear distances estimated from the landmark coordinates that define the

anterior body of the hemimandible and the ramus portion of the hemimandible (Table S1, Fig.

S1). While the anterior body of the Fgfr2cC342Y/+ Crouzon/Pfeiffer syndrome mouse mandible

showed a distinct morphology relative to their unaffected littermates, there was less difference

in the anterior body of Fgfr2+/S252W and Fgfr2+/P253R Apert syndrome mice relative to their

respective unaffected littermates (Fig. S2). All three mutation groups showed differences

between mutant mice and their respective unaffected littermates in the ramus portion of the

hemimandible (Fig. S2).

In every analysis, all mice carrying mutations revealed a mandibular morphology that

differed from littermates that did not carry the mutation in unique ways. The finding that

different Fgfr2 gain-of-function mouse models exhibit different mandibular phenotypes is

consistent with our previous work that shows that the cranial phenotype (not including the

mandible) of each Fgfr2 model is different from their respective unaffected littermates that do

not carry the mutation, and that these changes vary across mouse models.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Quantitative characterization of hemimandible bone

Bone volume and bone surface area were determined using the Material Statistics Module of

Avizo 9.4 by first segmenting the left hemimandible as the region of interest from µCT scans.

Bone volume and bone surface area were compared using the Mann-Whitney U Test and did

not differ significantly between mutant and unaffected littermates in any of the three

craniosynostosis mouse models (Table 1), although the Fgfr2+/S252W mice had the least bone

volume of all genotypes. This is consistent with previous findings indicating no difference in

bone volume of the hemimandibles in Fgfr2+/P253R mice and their controls at E15.5, E16.5, E17.5,

P0 and P2 (Percival et al., 2014). Mean bone mineral density maps of the left hemimandibles

reveal little variation between mutant mice and unaffected littermates across all models (Fig.

S3). Cortical bone thickness was mapped in an identical manner and showed little variation (Fig.

S4).

Impaired microarchitecture in Fgfr2+/S252W mandible

The mandibles of Fgfr2+/S252W Apert syndrome mice showed the greatest difference relative to

their unaffected littermates in our analyses of 3D geometry and physical properties, and

therefore we initially focused on this mouse model for additional analyses of histological

properties of embryonic mandibular bone. We present histological analysis on the mandible at

E16.5 when differences were more obvious and consistent than at earlier stages. Osteogenic

tissue was defined operationally as cells that stain with alkaline phosphatase (ALP) that include

osteoprogenitors, preosteoblasts and differentiated osteoblasts (Huang et al., 2007). We

performed serial coronal sections (Fig. 2A) and stained them with ALP and alcian blue for

cartilage (Fig. 2B). Fgfr2+/S252W embryos showed significant morphological changes in the

developing mandible at E16.5. Both the osteogenic tissue and MC were enlarged in Fgfr2+/S252W

E16.5 embryos relative to Fgfr2+/+ littermates (Fig. 2B, C), prefiguring localization of the more

severe differences determined by morphometric analysis of µCT data of P0 mandibles.

Quantification of ALP-positive areas and cell numbers revealed the ALP-positive area was 38.5%

larger in Fgfr2+/S252W embryos relative to Fgfr2+/+ littermates (Fig. 2D) and cell number was

increased by 61.3% (Fig. 2E) at E16.5, while the cell density was not significantly changed (Fig.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

2F; P=0.135). The area of MC was 94.5% larger in Fgfr2+/S252W embryos relative to Fgfr2+/+

littermates (Fig. 2G) and cell number was increased by 56.6% (Fig. 2H), but there was no

significant change in cell density (Fig. 2I, P=0.354).

It has been reported that postnatal and adult Fgfr2S252W/+ mutant mice showed changes in

mineral apposition rate and microarchitecture of the mandible (Zhou et al., 2013). To analyze

the mineralized tissues of embryonic Fgfr2+/S252W mandibles, we performed alizarin red S and

von Kossa staining to detect calcium deposits and the presence of phosphate, respectively. In

the Fgfr2+/+ littermate, trabecular bone of the mandible was a well-organized network, while

the Fgfr2+/S252W trabecular bone had a disorganized and loose structure (Fig. 2J). Quantitative

analysis of alizarin red S staining showed that, although the total areas of mineralized tissues

were not significantly changed (Fig. 2K, P=0.414), the amount of staining versus osteogenic area

was significantly smaller in Fgfr2+/S252W embryos (Fig. 2L, P=0.043). Von Kossa staining revealed

a similar difference in mineralization of the mandible of Fgfr2+/+ and Fgfr2+/S252W embryos (Fig.

2M) and the quantification of the staining showed the similar result (Fig. 2N-O). Thus, the

FGFR2 S252W mutation is associated with impaired microarchitecture affecting mandibular

morphogenesis as early as E16.5.

To determine if these changes were found in the other FGFR2-related mouse models, we

performed histological analysis for mandibles of Fgfr2+/P253R and Fgfr2cC342Y/+ embryos at E16.5.

Fgfr2+/P253R embryos exhibited similar histological changes in the mandible of Fgfr2+/S252W

embryos, including increased osteogenic tissue, MC (Fig. S5B), and disorganized mineralization

pattern (Fig. S5F, J), consistent with the mandibular dysmorphology observed for these mutants

at P0. Fgfr2cC342Y/+ embryos showed more subtle, localized changes in mineralization in the

mandible (Fig. S5H, L).

Abnormal osteogenesis in Fgfr2+/S252W mandible

To understand the molecular mechanism for the changes of bone formation in the mandible by

the FGFR2 S252W mutation, we collected tissues from the mandibular bone and MC of

Fgfr2+/S252W embryos and their Fgfr2+/+ littermates at E16.5 by laser capture microdissection

(Fig. 3A). Total RNA from the two specific tissues was isolated and used for RNA-Seq analysis.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

We identified 122 genes that were significantly differentially expressed in the mandibular

bone of Fgfr2+/S252W embryos compared to their Fgfr2+/+ littermates (Table S3). Sixty-seven

differentially expressed genes (DEGs) were up-regulated and 55 DEGs were down-regulated in

the mutant (Fig. 3B, C). Gene ontology analysis of DEGs in biological process (Fig. S6A) and

cellular component (Fig. S6B) identified several GO terms including osteoclast differentiation

(GO:0030316), bone remodeling (GO:0046849), bone resorption (GO:0045453), ossification

(GO:0001503), and extracellular matrix (GO:0031012), all relevant to the histological phenotype

of Fgfr2+/S252W embryonic mandibles. No significant DEGs were found in MC of Fgfr2+/S252W

embryos compared to their Fgfr2+/+ littermates (data not shown).

Increased osteoclastogenesis in Fgfr2+/S252W hemimandibles

A group of genes active in osteoclastogenesis, including Acp5, Calcr, Csf1r, Ctsk, Il1r1, Itgb3,

Oscar and Tnfrsf11a (Table S3), were identified by transcriptome analysis as significantly

upregulated in Fgfr2+/S252W embryos relative to their Fgfr2+/+ littermates. To validate our

sequencing analyses, we tested the expression of Csfr1 and Itgb3 that are critical regulators of

the osteoclast lineage by section in situ hybridization (ISH) at E16.5. Csf1r encodes a tyrosine

kinase growth factor that is the receptor for the ligand colony stimulating factor-1 (CSF1).

CSF1R-mediated signaling plays an important role in osteoclastogenesis and Csf1r-/- mice exhibit

severe osteoclast deficiency (Dai et al., 2002). Integrin beta 3, encoded by Itgb3, forms a

complex with integrin alpha V, and integrin αvβ3 is essential for normal osteoclast function

(McHugh et al., 2000). ISH showed that Csf1r exhibited a scattered expression pattern in the

mandibular area of Fgfr2+/+ littermates (Fig. 4A, B), labeling preosteoclasts and osteoclasts in

the mandible. In the mandible of Fgfr2+/S252W embryos, increased Csf1r-positive cells were

detected (Fig. 4C, D), consistent with the results of the transcriptome analysis. Similarly, there

was increased expression of Itgb3 in the mandible of Fgfr2+/S252W embryos (Fig. 4G, H) relative

to their Fgfr2+/+ littermates (Fig. 4E, F), indicating increased expression of osteoclast genes in

mice carrying the FGFR2 S252W mutation. To confirm this result, we analyzed the tartrate-

resistant acid phosphatase (TRAP) activity as a functional osteoclastic marker. TRAP staining

revealed increased osteoclastic activity in the mandibular tissues of Fgfr2+/S252W embryos

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

relative to Fgfr2+/+ littermates at E16.5 (Fig. 4I, J). The numbers of osteoclasts per bone area in

mutant embryos were significantly increased compared to Fgfr2+/+ littermates (Fig. 4K,

38.5±8.7/mm2 for Fgfr2+/+, 117.9±12.3/mm2 for Fgfr2+/S252W, P=0.0158), and the percentage of

osteoclasts in the bone area was significantly increased as well (Fig. 4L, 2.61±0.23% for Fgfr2+/+,

7.19±0.89% for Fgfr2+/S252W, P=0.0288), suggesting higher bone resorption activity in mice

carrying the mutation relative to the controls.

To test if increased osteoclastic activity in the mandible observed in the Fgfr2+/S252W mice

was a common mechanism for mandibular dysmorphogenesis in the other FGFR2-related

mouse models, we performed the TRAP assay (Fig. S7). The Fgfr2+/P253R (Fig. S7C) like the

Fgfr2+/S252W (Fig. 4J and Fig. S7B) embryos showed increased osteoclastic activity compared with

unaffected littermates (Fig. S7A). The Fgfr2cC342Y/+ embryos showed increased staining in a

smaller area of the mandible (Fig. S7E) which also exhibited impaired mineralization (Fig. S5H,

L) compared to the unaffected littermates (Fig. S7D). The increase of osteoclastic activity was

consistent to the relative magnitude of changes in mandibular morphology and histology in the

three mouse models (Fgfr2+/S252W > Fgfr2+/P253R > Fgfr2cC342Y/+), indicating abnormal

osteoclastogenesis in the mandible is an important process affecting the relative

morphogenesis in FGFR2-related mouse models.

Increased expression of Enpp1 and Ank and elevated inorganic pyrophosphate (PPi)

concentration in Fgfr2+/S252W mandible

Gene ontology analysis (Fig. S6) showed that the expression of many genes that contribute to

ossification (GO:0001503) were down regulated in the mandible of Fgfr2+/S252W embryos,

including Aspn, Chrdl1, Igf1, Mgp, Ptn, Sfrp2, Thbs3 and Tnn, suggesting that osteogenesis was

inhibited in the Fgfr2+/S252W mandible. Mineralization plays a pivotal role in bone formation and

is initiated within matrix vesicles (MVs) where Ca2+ ions and inorganic phosphate (Pi) crystallize

to form hydroxyapatite (HA). The extracellular PPi (ePPi) adsorbs tightly to HA and potently

antagonizes the ability of Pi to crystallize with calcium to form HA, inhibiting HA crystal

propagation (Terkeltaub, 2001). Enpp1 and Ank are essential in regulating levels of PPi

(Mackenzie et al., 2012). Enpp1 is expressed in differentiated osteoblasts and encodes

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

ectonucleotide pyrophosphatase/phosphodiesterase 1 (ENPP1) protein that influences matrix

mineralization by increasing extracellular levels of PPi and regulates osteoblast differentiation

(Nam et al., 2011). Enpp1 is essential for normal bone development and control of physiological

bone mineralization and Enpp1−/− mice are characterized by severe disruption to the

architecture and mineralization of long bones, dysregulation of calcium/phosphate

homeostasis, and changes in Fgf23 expression (Mackenzie et al., 2012). Ank encodes the

progressive ankylosis protein, which is a highly conserved transmembrane pyrophosphate

transporter that channels PPi into extracellular matrix (Chen et al., 2011). Mutations located in

cytoplasmic domains close to the C-terminus of the human ANK gene (ANKH) were identified

for the autosomal dominant form of craniometaphyseal dysplasia (CMD) (Nürnberg et al., 2001;

Reichenberger et al., 2001). Overexpression of Ank in tissue culture cells leads to an increase in

the total level of ePPi (Ho et al., 2000).

Enpp1 and Ank were among the most significantly upregulated genes in the mandible of

Fgfr2+/S252W embryos at E16.5 (Fig. 3C), with log2 fold-change=1.83 and 1.33, respectively (Fig.

3C and Table S3). The differential expression was validated by ISH (Fig. 5A-H). To test if

upregulated expression of Enpp1 and Ank was associated with elevated PPi levels, the E16.5

mandibles were dissected and the weight and amount of PPi were quantified (Fig. 5I, J). The

weight of Fgfr2+/S252W mandibles (3.75±0.48 mg) were significantly lower than controls

(5.41±0.51 mg), while PPi concentrations in the mutant embryonic mandibles were higher

(4.22±0.93 nmol/mg) relative to those of the Fgfr2+/+ littermates (1.52±0.37 nmol/mg). These

findings suggest that genes that were transcriptionally dysregulated by the FGFR2 S252W

mutation changed the Pi/PPi balance toward reduced bone formation and mineralization.

Increased proliferation of osteoblasts and chondrocytes in the mandible of Fgfr2+/S252W

embryos

The osteogenic tissue and MC of Fgfr2+/S252W embryos at E16.5 were enlarged relative to their

littermates that did not carry the mutation (Fig. 2B). IHC for RUNX2 as an osteoblast marker

confirmed that osteoblasts are increased in the osteogenic region of Fgfr2+/S252W embryos (Fig.

6A). However, there is no indication from RNA-Seq analysis in mandibular bone or MC at E16.5,

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

suggesting that the molecular changes resulting from the FGFR2 S252W mutation might have

occurred at an earlier stage to cause the enlargement. As osteoblasts with FGFR2 S252W

mutation have an increased capacity for proliferation and differentiation in vitro (Holmes et al.,

2009; Yang et al., 2008) and FGF signaling plays an important role in chondrocyte proliferation

(Brewer et al., 2016), we hypothesized that the FGFR2 S252W mutation would increase cell

proliferation during the earlier stages of mandible development, resulting in increased cell

numbers and enlargement of the osteogenic tissue and MC. To test this, the EdU assay was

performed at E12.5 when ALP, an early osteoblast marker, can be detected in the jaw (Funato

et al., 2016) and adjacent mesenchymal cells are condensing to form MC (Parada and Chai,

2015). RUNX2 was used as a marker for osteoblasts (Fig. 6B), and chondrocytes in the MC were

visualized by staining SOX9 (Fig. 6D), which was strongly expressed in immature/proliferating

chondrocytes (Leung et al., 2011). The percentage of proliferating cells (EdU-positive) in

osteoblasts (SOX9-positive) of Fgfr2+/S252W embryos was 59.7±2.1%, significantly increased

compared with 45.7±2.0% in Fgfr2+/+ littermates (Fig. 6C). The percentage of proliferating cells

(EdU-positive) in MC (SOX9-positive) of Fgfr2+/S252W embryos was 57.4±6.1%, significantly

increased compared with 32.6±3.7% in Fgfr2+/+ littermates (Fig. 6E).

As apoptosis is another possible mechanism affecting cell numbers, the TUNEL assay was

performed on the mandible of Fgfr2+/S252W and Fgfr2+/+ embryos at E16.5. TUNEL signal was

detected in the mandibular bone of both groups (Fig. S8A). No apoptotic chondrocytes in MC

were observed but signals were detected in perichondrium (Fig. S8A), similar to a previous

study on normal MC development (Amano et al., 2010). Quantitative analysis shows

insignificant increase of TUNEL-positive cells in mandibular bone (Fig. S8B, 33.1±5.7 for Fgfr2+/+

and 60.3±7.8 for Fgfr2+/S252W, P=0.089) and perichondrium (Fig. S8C, 6.8±0.6 for Fgfr2+/+ and

9.0±0.5 for Fgfr2+/S252W, P=0.083) of Fgfr2+/S252W embryos compared with Fgfr2+/+ littermates,

indicating apoptosis does not contribute to the enlargement of the osteogenic tissue or MC. D

isea

se M

ode

ls &

Mec

hani

sms

• D

MM

• A

ccep

ted

man

uscr

ipt

DISCUSSION

The findings of this study demonstrate that the 3D dysmorphology of hemimandibles of three

FGFR2-related mouse models for craniosynostosis syndromes are easily distinguished from the

mandibles of their respective unaffected littermates at birth. Focusing on the hemimandibles of

Fgfr2+/S252W mice, we demonstrated significant dysmorphology, increase in MC size and

dysmorphic bony microarchitecture in embryos at E16.5. Our analyses showed that these

changes are due in part to inhibited osteogenic activity with increased osteoclastogenesis of the

mandible and earlier increased proliferation of osteoblasts and chondrocytes.

There are comparative data relevant to this study from previous analyses of humans and

adult mice. Most studies of the mandible in humans with craniosynostosis syndromes have

concluded that the mandible is intrinsically normal and that the morphological differences that

are noted likely represent a developmental response to the composite of structural variations

in the basicranium and midface in these conditions (Enlow, 2000; Wink et al., 2013). In humans,

the data are necessarily postnatal and analyses of mandibular shape are often conducted after

individuals have undergone reconstructive surgery of the midface. Because the maxilla and

mandible function as a unit in phonation and mastication, surgically-induced changes of

midfacial morphology could affect mandibular growth and morphology through changes in

functional relationships.

Previous mandibular analyses in mouse models for craniosynostosis include little or no

quantitative or gene expression information for embryonic and newborn mice, although it has

been reported that Fgfr2 is expressed in mandibular osteoblasts (Rice et al., 2003). A recent

study using only mandibular morphometry of adult mice carrying the FGFR2c C342Y mutation

on a mixed genetic background (72% C57BL6/J and 28% Swiss) reported significantly reduced

ascending (measured from the apex of coronoid process to gonion) and descending (measured

from the apex of coronoid process to menton) mandibular heights, mandibular lengths

(measured from condylion to pognion and from condylion to pogonion), and intercoronoid and

intercondylar widths, but increased intergonial widths (Khominsky et al., 2018). Though the

method of measurement differs from ours, these observations generally agree with our data

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

for newborn mice carrying the same mutation on a CD1 background. Mice aged 4 and 8 weeks

carrying the FGFR2 P253R mutation causative for Apert syndrome display globally reduced

mandibular dimensions (Du et al., 2010). Results presented here show that the mandibles of

newborn Fgfr2+/P253R mice are generally reduced in size relative to their normal littermates at

birth, and that the reduction is of greater magnitude in the posterior portion (including the

posterior mandibular body, and coronoid, condylar and angular processes), contributing to a

complex change in shape. We also reported that adult Fgfr2+/S252W Apert syndrome mice have a

very small mandible with a dysmorphic angular process (Wang et al., 2005; Wang et al., 2010).

The molecular changes that result from FGFR mutations are complex and include

constitutive (ligand-independent) or ligand-dependent FGFR activation, loss of function, and

altered cellular trafficking of receptors (Ornitz and Itoh, 2015). The FGFR2c C342Y mutation

associated with Crouzon/Pfeiffer syndrome lies within the Ig‐III domain of FGFR2c, and results

in constitutive activation of the receptor (Mangasarian et al., 1997). FGFR2 S252W and P253R

mutations are in the linker region, resulting in increased ligand affinity and altered specificity

(Andersen et al., 1998; Ibrahimi et al., 2001; Yu et al., 2000). Crystal structures of the FGFR2

S252W and P253R mutations indicate that P253R indiscriminately increases the affinity of

FGFR2 toward any FGF, while the S252W mutation selectively enhances the affinity of FGFR2

toward a limited subset of FGFs (Ibrahimi et al., 2001). These mutations in FGFR2 then

differentially affect FGFR2 intracellular signaling pathways (e.g. ERK1/2, PLCγ/PKCα, and

PI3K/Akt), resulting in alterations in cell proliferation, differentiation, and apoptosis, depending

on the stage of cell differentiation (Ornitz and Marie, 2015), forming the basis for different

mandibular phenotypes.

Zhou et al. demonstrated a significant decrease in mandibular cortical bone, decreased

bone mass, a significant decrease in calcein labeling of mineralizing surfaces, and a reduced

mineral apposition rate in the postnatal mandibles of Fgfr2S252W/+ mice at P28 and P56 (Zhou et

al., 2013). An observed increase in the number of osteoclasts, and a decreased number of

osteoblasts per bone surface area suggested lower bone formation capacities in Fgfr2S252W/+

adult mandibles relative to those of Fgfr2+/+ littermates. Bone modeling increases bone mass

and changes the shape of bones and occurs throughout life while bone remodeling functions to

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

renew bone (Allen and Burr, 2015). If the cellular activities reported in the mandibles of adult

Fgfr2S252W/+ mice (Zhou et al., 2013) are functioning primarily to renew bone, while our results

reflect changes in cellular activities that increase bone mass and change its shape, then the

balance of the amount of tissue resorbed and formed at any particular site may be disrupted by

the effect of FGFR2 S252W in modeling and remodeling.

Our results demonstrate an intrinsic difference in mandibular morphology of newborn mice

carrying FGFR2-related craniosynostosis mutations. We used ALP staining to detect mature

osteoprogenitor cells, preosteoblasts and differentiated osteoblasts (Huang et al., 2007), and

found that cell numbers in ALP-positive areas were increased in Fgfr2+/S252W embryos, indicating

that the FGFR2 S252W mutation promotes osteoblastic proliferation and differentiation,

consistent with results of previous analyses of adult cranial bone response to FGFR2 mutations

(Holmes et al., 2009; Yang et al., 2008). However, the overall mandible is reduced in size in

mutant mice both pre- and post-natally. Our finding of increased osteoclastogenesis is a

mechanism that can account for the overall reduction in size of mandibles of mice carrying

FGFR2 mutations.

Investigation of the transcriptome of the mandible in embryonic mice carrying the FGFR2

S252W mutation revealed dysregulation of genes involved in bone formation, bone

mineralization, and osteoclastogenesis, highlighting increased expression of genes undergoing

osteoclast differentiation and dysregulated genes active in bone mineralization. Bone formation

and bone resorption are important determinants of bone size and shape, whether osteoblast

and osteoclastic activity are coupled (as in remodeling) or uncoupled (as in modeling) (Allen and

Burr, 2015). Altered signaling pathways result in the dysregulation of genes that are involved in

osteoclastogenesis, bone formation, and bone mineralization, contributing to impaired

mandibular morphogenesis and microarchitecture. During normal growth, as bone mass

increases, resorption is required to alter bone shape and maintain a functioning skeletal

element. Our data show that improper regulation of osteoblastogenesis and osteoclastogenesis

can offset the balance required for bone modeling contributing to changes in the

developmental trajectory of individual embryonic bones and resulting in altered bone

phenotypes. The FGFR2 S252W mutation may retard mandibular bone formation, contribute to

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

decreased bone volume and compromised skeletal architecture by regulating both

osteoblastogenesis and osteoclastogenesis.

MC plays a crucial role as a supportive tissue for mandible formation and early growth.

Chondrocytes that form endochondral bone are differentiated from mesodermal cells in

general, whereas cells forming MC are differentiated from cells of neural crest origin (Amano et

al., 2010). In addition, the boundary between neural crest and mesoderm cells of the

chondrocranium lies between the hypophyseal and parachordal cartilages (McBratney-Owen et

al., 2008), indicating that the cartilages caudal to the hypophyseal cartilage are of mesodermal

origin. Abnormal MC development is associated with dysmorphogenesis of the mandible. For

example, activating FGFR3 mutations associated with achondroplasia lead to structural

anomalies of MC and condylar cartilages of the mandible, resulting in mandibular hypoplasia

and dysmorphogenesis (Duplan et al., 2016). We observed dramatically increased size and cell

number in MC of Fgfr2+/S252W embryos, with increased proliferation of chondrocytes detected as

early as E12.5. Previous studies have shown that Fgfr2+/S252W mutant mice have increased

cartilage of the basicranium (posterior to the hypophyseal cartilage) and thickened nasal

cartilage due to increased chondrocyte proliferation (Wang et al., 2005; Holmes et al., 2018).

These results suggest a common mechanism of increased proliferation by the FGFR2 S252W

mutation in these cranial cartilages whether derived from cranial neural crest cells or

mesoderm.

In summary, we quantitatively analyzed prenatal mandibular morphology in mouse models

carrying mutations by variants of Fgfr2 that are associated with craniosynostosis syndromes

when present in humans. Finding that the mandibles of Fgfr2+/S252W mice were quantitatively

the most different from their unaffected littermates we further studied the mandibles of these

mice using histology, immunochemistry and transcriptome analyses to understand the source

of altered morphology and abnormal microarchitecture of these mice with altered ligand

affinity and specificity of FGFR2 (Cunningham et al., 2007). We have previously shown that the

craniofacial phenotype (not including the mandible) of mice carrying the mutation in each Fgfr2

model is different compared to that of littermates not carrying the mutation (Martínez-Abadías

et al., 2013; Motch Perrine et al., 2014; Wang et al., 2005; Wang et al., 2010). We suggest a

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

model that mutation-induced changes in activated FGF signaling and downstream pathways are

associated with dysregulation of osteoblastogenesis, osteoclastogenesis, resorption,

mineralization, and the formation of MC, resulting in dysmorphogenesis of the mandible (Fig.

7). Mandibular development is directly affected by the FGFR2 mutations in these mouse

mutants, as was first suggested morphologically and later shown histologically by our results.

Mandibular dysmorphogenesis in these mouse models for craniosynostosis results, at least in

part, from the intrinsic effects of the mutation, and are not solely related to the functional

relationship of the mandible with the midface and cranial base as previously deduced from

human data.

MATERIALS AND METHODS

Mouse models

The generation of Fgfr2+/S252W, Fgfr2+/P253R and Fgfr2cC342Y/+ models were previously described

(Eswarakumar et al., 2004; Wang et al., 2005; Wang et al., 2010). Fgfr2+/S252W and Fgfr2+/P253R

models were maintained on a C57BL6/J background. The Fgfr2cC342Y/+ model was maintained on

a CD1 background for viability and breeding. Our samples for µCT analyses consisted of 182

newborn (P0) (Fgfr2+/S252W model: 22 mutants (9 females (F):13 males (M)); 24 unaffected

littermates (20 F:4 M); Fgfr2+/P253R model: 35 mutants (20 F:15 M); 28 unaffected littermates

(16 F:12 M); Fgfr2cC342Y/+ model: 34 mutants (18 F:16 M); 39 unaffected littermates (24 F:15

M)). Newborn mice (P0) were euthanized by inhalation anesthetics and fixed in 4%

paraformaldehyde. Gestation time was 19.0±0.5 days. Samples for histological and

transcriptome analysis consisted of 91 embryos (Fgfr2+/S252W model: 29 mutants (17 F:12 M); 37

unaffected littermates (21 F:16 M); Fgfr2+/P253R model: 8 mutants (2 F:6 M); 5 unaffected

littermates (3 F:2 M); Fgfr2cC342Y/+ model: 8 mutants (4 F:4 M); 4 unaffected littermates (3 F:1

M)). Genotyping of tail DNA by PCR was performed to distinguish mutants and unaffected

littermates. Mouse litters were produced in compliance with animal welfare guidelines

approved by Icahn School of Medicine at Mount Sinai and Pennsylvania State University Animal

Care and Use Committees.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Imaging protocols

High resolution micro computed tomography (µCT) images with pixel size and slice thickness

ranging from 0.014 to 0.025 mm were acquired by the Center for Quantitative X-Ray Imaging at

the Pennsylvania State University (www.cqi.psu.edu) using the HD-600 OMNI-X high-resolution

X-ray computed tomography system (Bio-Imaging Research Inc, Lincolnshire, IL, USA). A

minimum threshold of 70-100 mg/cm3 partial density hydroxyapatite (HA) (based on HA

phantoms imaged with specimens) was used to reconstruct isosurfaces in Avizo 6.3

(Visualization Sciences Group, VSG, Burlington, MA, USA). 3D coordinates of 32 biologically

relevant landmarks (Fig. S1 and Table S1) were collected from the isosurfaces. Specimens were

digitized twice, and measurement error was minimized by averaging coordinates of the two

trials (maximum accepted error in landmark placement=0.05 mm).

Statistical evaluation of shape differences

Variation in mandible shape was assessed by principal component analysis (PCA) using SAS 9.4

(SAS Institute, Cary, NC, USA). PCA summarizes the variation of large numbers of variables into

a lower-dimensional space defined by principal component (PC) axes that are mutually-

orthogonal, linear combinations of the linear distance data. The scores of an observation

(mandible or mandibular region) on the PC axes map that observation into the space. We

performed PCA (Darroch and Mosimann, 1985; Falsetti et al., 1993) of form (size and shape

together) using inter-landmark linear distances estimated using the full mandible landmark set

and subsets defining the left hemimandible (results of PCA analyses of left and right

hemimandibles were similar), anterior body, and ramus. Inter-landmark distances were ln-

transformed, and their variance-covariance matrix was used as the basis for the PCA.

Euclidean Distance Matrix Analysis (EDMA) was used to statistically evaluate mandibular

shape differences by hypothesis test and confidence interval estimation (Lele and Richtsmeier,

2001). EDMA is a 3D morphometric technique that is invariant to the group of transformations

including translation, rotation, and reflection (Lele and Richtsmeier, 1995; Richtsmeier and Lele,

1993). Briefly, the original 3D coordinates of landmark locations are re-written and analyzed as

a matrix of all unique linear distances among landmarks called the form matrix (FM). An

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

average FM is estimated for each sample (Lele and Richtsmeier, 1995). The difference between

samples is evaluated by estimating ratios of like-linear distances using sample-specific average

FMs. The resulting matrix of ratios, the form difference matrix (FDM), is a collection of relative

differences among landmarks used to define the forms. A non-parametric bootstrap procedure

(100,000 resamples) is used to obtain confidence intervals for elements (each corresponding to

a linear distance) of the FDM (Lele and Richtsmeier, 2001) that reveals the localized effects of

the mutations on the mandible. We also include a non-parametric bootstrap assessment of the

null hypothesis that the mean forms of two samples are the same (Lele and Richtsmeier, 2001).

We tested for form difference of the entire left hemimandible, the anterior body, and the

ramus portion using WinEDMA (Cole III, 2002).

Bone volume, surface area, and bone mineral density analyses

Bone volumes and surface areas were determined using the high resolution µCT scans

described above using Avizo 9.4 (ThermoFischer Scientific, Materials & Structural Analysis

Division, Hillsboro, OR, USA). The minimum thresholds used to create isosurfaces ranged from

70-100 mg/cm3 partial density hydroxyapatite. The isosurfaces were then analyzed using the

Material Statistics module of Avizo 9.4 software to determine bone volumes and bone surface

area. Stradwin v5.3 (http://mi.eng.cam.ac.uk/~rwp/stradwin) was used to create isosurfaces

from 30 of the hemimandibles (5 of each group) (Treece et al., 2010; Treece et al., 2012). The

tooth was excluded manually by placing guiding contours every five tomographic slices along

the hemimandibles. Density values were determined from the partial density hydroxyapatite

phantom normalized grey values at each isosurface vertex (~89,000-112,000 measurements).

Isosurfaces and their associated density values were registered using wxRegSurf v17

(http://mi.eng.cam.ac.uk/~ahg/wxRegSurf) (Gee et al., 2015; Stephens et al., 2018). A statistical

shape model (SSM) was generated for the full dataset and for each mouse model. Mean density

was calculated for each corresponding vertex and mapped onto the pertinent SSM. Pairwise

statistical differences between littermates were determined by performing a linear model

comparison at each vertex of the full SSM in R, with the resulting P values being mapped for

visual comparison.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Histological analysis

Mouse embryos at E16.5 were dissected and the heads were fixed in 4% paraformaldehyde

overnight at 4°C and then washed for 3 times with PBS. Samples were infused in 0.5 M sucrose

in PBS until tissue sank, and then quick-frozen in OCT. Samples were sectioned at a thickness of

10 μm. For ALP and alcian blue staining, cryosection samples were incubated for 5 min at room

temperature (RT) in 100 mM Tris-maleate buffer, pH 9.2, and then incubated for 5 min at RT in

freshly prepared ALP substrate solution (100 mM Tris-maleate buffer, pH 9.2, 0.2 mg/mL

naphthol AS-MX phosphate and 0.4 mg/mL Fast Red TR). The slides were washed briefly and

then stained with alcian blue solution (1% alcian blue, 3% acetic acid, pH 2.5) for 5 min at RT.

The slides were washed with water and then stained with 1 µg/mL Hoechst 33258 (Invitrogen

Life Technologies, Carlsbad, CA, USA) in PBS for 5 minutes. Images were collected in bright field

for ALP and alcian blue staining and then in UV for Hoechst 33258-stained nuclei. Calcium

deposits were detected with alizarin red S staining solution (MilliporeSigma, Burlington, MA,

USA). Presence of phosphate was detected with von Kossa staining kit (American MasterTech

Scientific, Lodi, CA, USA). TRAP staining was performed using cryosections with Acid

Phosphatase, Leukocyte (TRAP) Kit (MilliporeSigma, Burlington, MA, USA) following the

manufacturer’s instructions. Images were analyzed with ImageJ for stained particles and areas.

Laser capture microdissection (LCM) and RNA sequencing (RNA-Seq)

LCM was performed as previously described in detail (Holmes et al., 2018). The heads of female

Fgfr2+/+ littermates (n=3) and female Apert Fgfr2+/S252W (n=3) embryos at E16.5 were embedded

in OCT without fixation and rapidly frozen. Coronal cryosection was performed for the mandible

with 12 µm thickness. The mandibular tissue and MC were captured and collected respectively.

RNA was isolated with Arcturus Picopure RNA Isolation Kit (Thermo Fisher Scientific, Waltham,

MA, USA). Library preparation with NuGEN Ovation RNA-Seq System v2 (NuGEN Technologies,

San Carlos, CA, USA) and Nextera XT Library Prep kit (Illumina, San Diego, CA, USA) was

performed by the Gene Expression Core Facility at the Cincinnati Children's Hospital Medical

Center as described (Holmes et al., 2018). Library sequencing was performed on an Illumina

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

HiSeq 2500 instrument using standard protocols for paired-end 100 bp sequencing by the

Genetic Resources Core Facility at the Johns Hopkins School of Medicine.

Differential gene expression analysis and gene ontology (GO) enrichment analyses

RNA-Seq data processing, differential gene expression analysis and GO enrichment analyses

were done as described previously (Holmes et al., 2018). Briefly, paired-end reads were

mapped to the mouse (mm10) reference genome using STAR (Dobin et al., 2013) and gene

count summaries were generated using featureCounts (Liao et al., 2014). Only genes with

expression levels above 1 FPKM in at least 50% of samples were retained for further analysis.

Normalization factors were computed on the filtered data matrix using the weighted trimmed

mean of M-values (TMM) method (Robinson and Oshlack, 2010), followed by voom mean-

variance transformation in preparation for Limma linear modeling (Law et al., 2014). Data was

fitted to a design matrix containing all sample groups, and pairwise comparisons were

performed between sample groups. Finally, eBayes adjusted P values were corrected for

multiple testing using the Benjamin-Hochberg (BH) method and used to select genes with

significant expression differences (q<0.05). For GO enrichment analyses, the ‘elim’ algorithm

and ‘Fisher exact’ test were used to identify statistically over-represented GO categories at an

FDR corrected P value threshold of 0.05.

RNA in situ hybridization

Differential gene expression identified by RNA-Seq was validated by RNA in situ hybridization.

Riboprobe templates were generated by PCR using primers from published literature or

designed by Primer3 (http://primer3.ut.ee and Table S4), using cDNA derived from mouse

embryonic total RNA at E11.5. Riboprobes were prepared with DIG RNA Labeling Mix (Roche

Applied Science, Mannheim, Germany) as described by the manufacturer. RNA in situ

hybridization (ISH) was performed as described (Xu and Wilkinson, 1999).

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Quantification of PPi levels

Mandibles were isolated from embryos at E16.5 in cold PBS and dried briefly on delicate task

wipers before the weight was measured. Quantification of PPi in the mandible was performed

as described (Murali et al., 2016) with modification. Briefly, each mandible was incubated in

100 µl of 1.2M HCl at 4°C overnight and neutralized with NaOH and diluted with water.

Extracted PPi was quantified using the PPiLight Inorganic Pyrophosphate Assay (Lonza,

Walkersville, MD, USA) according to the manufacturer's protocol.

EdU assay and immunohistochemistry

EdU in vivo labelling was performed by single intraperitoneal injections of EdU to pregnant mice

at E12.5 at a dose of 50 mg/kg body weight in a solution of 10 mg/mL PBS (pH 7.35)

(Chehrehasa et al., 2009). The dams were sacrificed 30 min after the injection and embryos

were dissected for cryosection. The heads were sectioned at a thickness of 10 μm and EdU-

labeled cells were detected with Click-iT™ EdU Alexa Fluor™ 488 Imaging Kit (Thermo Fisher

Scientific, Waltham, MA, USA) followed by immunostaining for SOX9 (1:500, AB5535,

MilliporeSigma, Burlington, MA, USA) or RUNX2 (1: 200, HPA022040, MilliporeSigma,

Burlington, MA, USA) and Hoechst 33258 staining.

TUNEL assay

TdT-mediated dUTP nick end labeling (TUNEL) staining was performed using the In Situ Cell

Death Detection Kit, Fluorescein (MilliporeSigma, Burlington, MA, USA) according to the

manufacturer's protocol.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Acknowledgements

The authors would like to thank Dr. Tim Ryan and Mr. Tim Stecko at Penn State’s Center for

Quantitative Imaging for obtaining high quality µCT images and assisting in protocol review for

bone mineral density analysis. This work was supported in part through the computational

resources and staff expertise provided by Scientific Computing and special thanks to the

Pathology Core for access to the Leica laser capture microdissection platform at the Icahn

School of Medicine at Mount Sinai.

Competing interests

The authors declare no competing or financial interests.

Author contributions

Imaging studies were performed at Pennsylvania State University and mouse and wet lab

studies were performed at Icahn School of Medicine at Mount Sinai. Study design: S.M.M.P.,

M.W., E.W.J., J.T.R.; Methodology: S.M.M.P., M.W., E.W.J., J.T.R.; Software: S.M.M.P., N.S., D.K.,

H.V.B.; Validation: S.M.M.P., M.W. and N.S.; Formal analysis: S.M.M.P., M.W., N.S., D.K., H.V.B.;

Investigation: S.M.M.P., M.W., N.S.; Resources: J.T.R., E.W.J.; Data curation: S.M.M.P., M.W.,

N.S., D.K., H.V.B.; Writing-original draft preparation: S.M.M.P., M.W., and J.T.R.; Visualization:

S.M.M.P, M.W., N.S., D.K., H.V.B.; Supervision: E.W.J., J.T.R., H.V.B.; Project administration:

S.M.M.P., M.W., E.W.J., J.T.R.; Funding acquisition: E.W.J., J.T.R.

Funding

This work was supported by National Institute of Dental and Craniofacial Research

[R01DE022988 to J.T.R. and E.W.J.] and Eunice Kennedy Shriver National Institute of Child

Health and Human Development [P01HD078233 to J.T.R. and E.W.J.].

Data availability

RNA-Seq data has been deposited in Gene Expression Omnibus (GEO) with accession number

GSE121780. Reviewers can access the private GEO records with the token: shoveicupbkdbsv.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

References

Aldridge, K., Hill, C. A., Austin, J. R., Percival, C., Martinez-Abadias, N., Neuberger, T., Wang,

Y., Jabs, E. W. and Richtsmeier, J. T. (2010). Brain phenotypes in two FGFR2 mouse models

for Apert syndrome. Dev. Dyn. 239, 987–997.

Allen, M. R. and Burr, D. B. (2015). Bone modeling and remodeling. In Basic and applied bone

biology, pp. 75–92. Elsevier Science.

Amano, O., Doi, T., Yamada, T., Sasaki, A., Sakiyama, K., Kanegae, H. and Kindaichi, K. (2010).

Meckel’s cartilage: Discovery, embryology and evolution. J. Oral Biosci. 52, 125–135.

Andersen, J., Burns, H. D., Enriquez-Harris, P., Wilkie, A. O. M. and Heath, J. K. (1998). Apert

syndrome mutations in fibroblast growth factor receptor 2 exhibit increased affinity for

FGF ligand. Hum. Mol. Genet. 7, 1475–1483.

Atchley, W. R. and Hall, B. K. (1991). A model for development and evolution of complex

morphological structures. Biol. Rev. Camb. Philos. Soc. 66, 101–57.

Azoury, S. C., Reddy, S., Shukla, V. and Deng, C.-X. (2017). Fibroblast growth factor receptor 2

(FGFR2) mutation related syndromic craniosynostosis. Int. J. Biol. Sci. 13, 1479–1488.

Brewer, J. R., Mazot, P. and Soriano, P. (2016). Genetic insights into the mechanisms of Fgf

signaling. Genes Dev. 30, 751–771.

Chehrehasa, F., Meedeniya, A. C. B., Dwyer, P., Abrahamsen, G. and Mackay-Sim, A. (2009).

EdU, a new thymidine analogue for labelling proliferating cells in the nervous system. J.

Neurosci. Methods 177, 122–130.

Chen, I. P., Wang, L., Jiang, X., Aguila, H. L. and Reichenberger, E. J. (2011). A Phe377del

mutation in ANK leads to impaired osteoblastogenesis and osteoclastogenesis in a mouse

model for craniometaphyseal dysplasia (CMD). Hum. Mol. Genet. 20, 948–961.

Cohen, M. M. and MacLean, R. E. (2000). Craniosynostosis: diagnosis, evaluation, and

management. Oxford University Press.

Cole III, T. M. (2002). WinEDMA: Softw. euclidean distance matrix Anal. Version 1.0.1 beta.

Kansas City Univ. Missouri – Kansas City Sch. Med.

Costaras-Volarich, M. and Pruzansky, S. (1984). Is the mandible intrinsically different in Apert

and Crouzon syndromes? Am. J. Orthod. 85, 475–487.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Couly, G., Grapin-Botton, A., Coltey, P. and Le Douarin, N. M. (1996). The regeneration of the

cephalic neural crest, a problem revisited: the regenerating cells originate from the

contralateral or from the anterior and posterior neural fold. Development 122, 3393–1407.

Cunningham, M. L., Seto, M. L., Ratisoontorn, C., Heike, C. L. and Hing, A. V. (2007). Syndromic

craniosynostosis: From history to hydrogen bonds. Orthod. Craniofacial Res. 10, 67–81.

Dai, X.-M., Ryan, G. R., Hapel, A. J., Dominguez, M. G., Russell, R. G., Kapp, S., Sylvestre, V.

and Stanley, E. R. (2002). Targeted disruption of the mouse colony-stimulating factor 1

receptor gene results in osteopetrosis, mononuclear phagocyte deficiency, increased

primitive progenitor cell frequencies, and reproductive defects. Blood 99, 111–120.

Darroch, J. N. and Mosimann, J. E. (1985). Canonical and principal components of shape.

Biometrika 72, 241–252.

Depew, M. J., Lufkin, T. and Rubenstein, J. L. R. (2002). Specification of jaw subdivisions by Dlx

genes. Science 298, 381–5.

Dobin, A., Davis, C. A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson, M.

and Gingeras, T. R. (2013). STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29,

15–21.

Du, X., Weng, T., Sun, Q., Su, N., Chen, Z., Qi, H., Jin, M., Yin, L., He, Q. and Chen, L. (2010).

Dynamic morphological changes in the skulls of mice mimicking human Apert syndrome

resulting from gain-of-function mutation of FGFR2 (P253R). J. Anat. 217, 97–105.

Duplan, M. B., Komla-Ebri, D., Heuzé, Y., Estibals, V., Gaudas, E., Kaci, N., Benoist-Lasselin, C.,

Zerah, M., Kramer, I., Kneissel, M., et al. (2016). Meckel’s and condylar cartilages

anomalies in achondroplasia result in defective development and growth of the mandible.

Hum. Mol. Genet. 25, 2997–3010.

Enlow, D. H. (2000). Normal craniofacial growth. In Craniosynostosis: diagnosis, evaluation, and

management, 2nd ed. New York: Oxford University Press. p (ed. Cohen, M. M. J.) and

MacLean, R. E.), pp. 35–50. New York: Oxford University Press.

Eswarakumar, V. P., Horowitz, M. C., Locklin, R., Morriss-Kay, G. M. and Lonai, P. (2004). A

gain-of-function mutation of Fgfr2c demonstrates the roles of this receptor variant in

osteogenesis. Proc Natl Acad Sci U S A. 101, 12555–12560.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Falsetti, A. B., Jungers, W. L. and Colle III, T. M. (1993). Morphometrics of the callitrichid

forelimb: a case study in size and shape. Int J Primatol 14, 551–72.

Flaherty, K., Singh, N. and Richtsmeier, J. T. (2016). Understanding craniosynostosis as a

growth disorder. Wiley Interdiscip. Rev. Dev. Biol. 5, 429–459.

Frisdal, A. and Trainor, P. A. (2014). Development and evolution of the pharyngeal apparatus.

Wiley Interdiscip. Rev. Dev. Biol. 3, 403–418.

Funato, N., Kokubo, H., Nakamura, M., Yanagisawa, H. and Saga, Y. (2016). Specification of

jaw identity by the Hand2 transcription factor. Sci. Rep. 6, 28405.

Gee, A. H., Treece, G. M., Tonkin, C. J., Black, D. M. and Poole, K. E. S. (2015). Association

between femur size and a focal defect of the superior femoral neck. Bone 81, 60–66.

Heuzé, Y., Holmes, G., Peter, I., Richtsmeier, J. T. and Jabs, E. W. (2014). Closing the gap:

Genetic and genomic continuum from syndromic to nonsyndromic craniosynostoses. Curr.

Genet. Med. Rep. 2, 135–145.

Ho, A. M., Johnson, M. D. and Kingsley, D. M. (2000). Role of the mouse ank gene in control of

tissue calcification and arthritis. 289, 265–271.

Holmes, G., Rothschild, G., Basu, U., Deng, C., Mansukhani, A. and Basilico, C. (2009). Early

onset of craniosynostosis in an Apert mouse model reveals critical features of this

pathology. Dev. Biol. 328, 273–284.

Holmes, G., O’Rourke, C., Perrine, S. M. M., Lu, N., van Bakel, H., Richtsmeier, J. T. and Jabs, E.

W. (2018). Midface and upper airway dysgenesis in FGFR2-craniosynostosis involves

multiple tissue-specific and cell cycle effects. Development dev.166488.

Huang, W., Yang, S., Shao, J. and Li, Y. P. (2007). Signaling and transcriptional regulation in

osteoblast commitment and differentiation. Front. Biosci. 12, 3068–3092.

Ibrahimi, O. A., Eliseenkova, A. V., Plotnikov, A. N., Yu, K., Ornitz, D. M. and Mohammadi, M.

(2001). Structural basis for fibroblast growth factor receptor 2 activation in Apert

syndrome. Proc. Natl. Acad. Sci. U. S. A. 98, 7182–7187.

Khominsky, A., Yong, R., Ranjitkar, S., Townsend, G. and Anderson, P. J. (2018). Extensive

phenotyping of the orofacial and dental complex in Crouzon syndrome. Arch. Oral Biol. 86,

123–130.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Law, C. W., Chen, Y., Shi, W. and Smyth, G. K. (2014). voom: Precision weights unlock linear

model analysis tools for RNA-seq read counts. Genome Biol. 15, R29.

Lele, S. and Richtsmeier, J. T. (1995). Euclidean distance matrix analysis: Confidence intervals

for form and growth differences. Am. J. Phys. Anthropol. 98, 73–86.

Lele, S. and Richtsmeier, J. T. (2001). An invariant approach to statistical analysis of shapes.

Chapman & Hall/CRC.

Lemire, R. J. (2000). Embryology of the skull. In Craniosynostosis: diagnosis, evaluation, and

management (ed. Cohen, M. M. (Meyer M.) and MacLean, R. E.), pp. 24–32. New York:

Oxford University Press.

Leung, V. Y. L., Gao, B., Leung, K. K. H., Melhado, I. G., Wynn, S. L., Au, T. Y. K., Dung, N. W. F.,

Lau, J. Y. B., Mak, A. C. Y., Chan, D., et al. (2011). SOX9 governs differentiation stage-

specific gene expression in growth plate chondrocytes via direct concomitant

transactivation and repression. PLoS Genet. 7, e1002356.

Liao, Y., Smyth, G. K. and Shi, W. (2014). featureCounts: an efficient general purpose program

for assigning sequence reads to genomic features. Bioinformatics 30, 923–930.

Mackenzie, N. C. W., Zhu, D., Milne, E. M., van ’t Hof, R., Martin, A., Quarles, D. L., Millán, J.

L., Farquharson, C. and MacRae, V. E. (2012). Altered bone development and an increase

in FGF-23 Expression in Enpp1−/− Mice. PLoS One 7, e32177.

Mangasarian, K., Li, Y., Mansukhani, A. and Basilico, C. (1997). Mutation associated with

crouzon syndrome causes ligand-independent dimerization and activation of FGF receptor-

2. J. Cell. Physiol. 172, 117–125.

Martínez-Abadías, N., Motch, S. M., Pankratz, T. L., Wang, Y., Aldridge, K., Jabs, E. W. and

Richtsmeier, J. T. (2013). Tissue-specific responses to aberrant FGF signaling in complex

head phenotypes. Dev. Dyn. 242, 80–94.

McBratney-Owen, B., Iseki, S., Bamforth, S. D., Olsen, B. R. and Morriss-Kay, G. M. (2008).

Development and tissue origins of the mammalian cranial base. Dev. Biol. 322, 121–132.

McHugh, K. P., Hodivala-Dilke, K., Zheng, M. H., Namba, N., Lam, J., Novack, D., Feng, X., Ross,

F. P., Hynes, R. O. and Teitelbaum, S. L. (2000). Mice lacking beta3 integrins are

osteosclerotic because of dysfunctional osteoclasts. J. Clin. Invest. 105, 433–440.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Motch Perrine, S. M., Cole, T. M., Martínez-Abadías, N., Aldridge, K., Jabs, E. W. and

Richtsmeier, J. T. (2014). Craniofacial divergence by distinct prenatal growth patterns in

Fgfr2 mutant mice. BMC Dev. Biol. 14, 8.

Motch Perrine, S. M., Stecko, T., Neuberger, T., Jabs, E. W., Ryan, T. M. and Richtsmeier, J. T.

(2017). Integration of brain and skull in prenatal mouse models of Apert and Crouzon

syndromes. Front. Hum. Neurosci. 11, 369.

Murali, S. K., Andrukhova, O., Clinkenbeard, E. L., White, K. E. and Erben, R. G. (2016).

Excessive osteocytic Fgf23 secretion contributes to pyrophosphate accumulation and

mineralization defect in Hyp Mice. PLOS Biol. 14, e1002427.

Nam, H. K., Liu, J., Li, Y., Kragor, A. and Hatch, N. E. (2011). Ectonucleotide

pyrophosphatase/phosphodiesterase-1 (ENPP1) protein regulates osteoblast

differentiation. J. Biol. Chem. 286, 39059–39071.

Noden, D. M. (1983). The role of the neural crest in patterning of avian cranial skeletal,

connective, and muscle tissues. Dev. Biol. 96, 144–165.

Nürnberg, P., Thiele, H., Chandler, D., Höhne, W., Cunningham, M. L., Ritter, H., Leschik, G.,

Uhlmann, K., Mischung, C., Harrop, K., et al. (2001). Heterozygous mutations in ANKH, the

human ortholog of the mouse progressive ankylosis gene, result in craniometaphyseal

dysplasia. Nat. Genet. 28, 37–41.

Ornitz, D. M. and Itoh, N. (2015). The Fibroblast Growth Factor signaling pathway. Wiley

Interdiscip. Rev. Dev. Biol. 4, 215–266.

Ornitz, D. M. and Marie, P. J. (2015). Fibroblast growth factor signaling in skeletal development

and disease. 29, 1463–1486.

Parada, C. and Chai, Y. (2015). Mandible and tongue development. Craniofacial Dev. 115, 31–

58.

Percival, C. J., Huang, Y., Jabs, E. W., Li, R. and Richtsmeier, J. T. (2014). Embryonic craniofacial

bone volume and bone mineral density in Fgfr2 +/P253R and nonmutant Mice. Dev Dyn.

243, 541–551.

Reichenberger, E., Tiziani, V., Watanabe, S., Park, L., Ueki, Y., Santanna, C., Baur, S. T., Shiang,

R., Grange, D. K., Beighton, P., et al. (2001). Autosomal dominant craniometaphyseal

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

dysplasia is caused by mutations in the transmembrane protein ANK. Am. J. Hum. Genet.

68, 1321–1326.

Rice, D. P. C., Rice, R. and Thesleff, I. (2003). Fgfr mRNA isoforms in craniofacial bone

development. Bone 33, 14–27.

Richtsmeier, J. T. and Lele, S. (1993). A coordinate-free approach to the analysis of growth

patterns: Models and theoretical considerations. Biol. Rev. Camb. Philos. Soc. 68, 381–411.

Robinson, M. D. and Oshlack, A. (2010). A scaling normalization method for differential

expression analysis of RNA-seq data. Genome Biol. 11, R25.

Stephens, N. B., Kivell, T. L., Pahr, D. H., Hublin, J. J. and Skinner, M. M. (2018). Trabecular

bone patterning across the human hand. J. Hum. Evol. 123, 1–23.

Terkeltaub, R. A. (2001). Inorganic pyrophosphate generation and disposition in

pathophysiology. Am. J. Physiol. - Cell Physiol. 281, C1–C11.

Treece, G. M., Gee, A. H., Mayhew, P. M. and Poole, K. E. S. (2010). High resolution cortical

bone thickness measurement from clinical CT data. Med. Image Anal. 14, 276–290.

Treece, G. M., Poole, K. E. S. and Gee, A. H. (2012). Imaging the femoral cortex: Thickness,

density and mass from clinical CT. Med. Image Anal. 16, 952–965.

Wang, Y., Xiao, R., Yang, F., Karim, B. O., Iacovelli, A. J., Cai, J., Lerner, C. P., Richtsmeier, J. T.,

Leszl, J. M., Hill, C. A., et al. (2005). Abnormalities in cartilage and bone development in

the Apert syndrome FGFR2(+/S252W) mouse. Development 132, 3537–3548.

Wang, Y., Sun, M., Uhlhorn, V. L., Zhou, X., Peter, I., Martinez-Abadias, N., Hill, C. A., Percival,

C. J., Richtsmeier, J. T., Huso, D. L., et al. (2010). Activation of p38 MAPK pathway in the

skull abnormalities of Apert syndrome Fgfr2+P253R mice. BMC Dev. Biol. 10, 22.

Wink, J. D., Bastidas, N. and Bartlett, S. P. (2013). Analysis of the long-term growth of the

mandible in apert syndrome. J. Craniofac. Surg. 24, 1408–1410.

Xu, Q. and Wilkinson, D. G. (1999). In situ hybridization of mRNA with hapten labelled probes.

In In Situ Hybridization: a Practical Approach. (ed. Wilkinson, D. G.), pp. 87–106. Oxford:

Oxford University Press.

Yang, F., Wang, Y., Zhang, Z., Hsu, B., Jabs, E. W. and Elisseeff, J. H. (2008). The study of

abnormal bone development in the Apert syndrome Fgfr2+/S252W mouse using a 3D

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

hydrogel culture model. Bone 43, 55–63.

Yu, K., Herr, A. B., Waksman, G. and Ornitz, D. M. (2000). Loss of fibroblast growth factor

receptor 2 ligand-binding specificity in Apert syndrome. Proc. Natl. Acad. Sci. U. S. A. 97,

14536–14541.

Zhou, X., Pu, D., Liu, R., Li, X., Wen, X., Zhang, L., Chen, L., Deng, M. and Liu, L. (2013). The

Fgfr2(S252W/+) mutation in mice retards mandible formation and reduces bone mass as in

human Apert syndrome. Am. J. Med. Genet. Part A 161, 983–992.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Figures

Fig. 1. Morphological differences in newborn (P0) mice carrying mutations associated with

three FGFR2-related craniosynostosis syndromes and their unaffected littermates. Results of

PCA of mandibles based on unique linear distances among 3D landmarks (A-C) and Euclidean

Distance Matrix Analysis (EDMA) of landmark coordinates (D-F). Scatter plots of individual

scores on first and second Principal Components axes (PC1 and PC2) of linear distance based

PCAs of the hemimandibles of mutant and unaffected littermates of Fgfr2+/S252W and Fgfr2+/P253R

Apert syndrome mouse models (A, B, respectively) and Fgfr2cC342Y/+ Crouzon/Pfeiffer syndrome

mouse model (C). Results of EDMA of each craniosynostosis mouse model and unaffected

littermates showing linear distances within each model that are significantly different by at

least 5% between mutant and unaffected littermates (D, E, F). Blue lines are significantly larger

in mutant mice relative to unaffected littermates; fuchsia lines are significantly smaller in

mutant mice. Thin lines indicate linear distances that are increased/decreased by 5-10% in mice

carrying one of the Fgfr2 mutations while thick lines indicate linear distances that differ by

>10% between mutant and unaffected mice. The buccal aspects of the left hemimandibles of

the models were used for illustration. Hemimandibles were segmented into an anterior portion

(anterior body, shown in blue) and posterior portion (ramus, shown in red) to indicate

functional areas. Scale bar=1 mm.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Fig. 2. Histological analysis of mandible of Fgfr2+/S252W embryos at E16.5. (A) Schematic

embryonic mouse head at E16.5 modified from the e-Mouse Atlas Project

(http://www.emouseatlas.org/emap/eHistology). The red line indicates the location of sections

used for B-O. (B) Cryosections of Fgfr2+/+ and Fgfr2+/S252W embryos were stained with the ALP

assay (red) and alcian blue. MB, mandibular bone. MC, Meckel’s cartilage. T, tongue. (C) The

ALP-positive regions (red) were selected to quantify the areas and numbers of nuclei stained

with Hoechst 33258 (blue). (D-I) The areas (D, G), cell numbers (E, H) and cell density (F, I) in the

ALP-positive regions for Fgfr2+/+ (n=6) and Fgfr2+/S252W (n=6) embryos and MC of Fgfr2+/+ (n=6)

and Fgfr2+/S252W (n=6) embryos. (J, M) Alizarin red S staining (J) and von Kossa staining (M)

showed ossification in the mandible of Fgfr2+/+ and Fgfr2+/S252W embryos. The areas and the

percentages of the stained area in osteogenic tissue were measured for (K, L) alizarin red S

staining and (N, O) von Kossa staining. Scale bar=100 µm. The experimental data were analyzed

by two-tailed Welch’s t-test and expressed as the mean±standard error of the mean (s.e.m.).

*P<0.05.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Fig. 3. Laser capture microdissection (LCM) and RNA-Seq analysis of mandibular bone of

Fgfr2+/S252W embryos at E16.5. (A) A representative mandibular region in cryosection was

dissected by laser and collected for RNA-Seq (left, before LCM; right, after LCM). Scale bar=400

µm. (B) Hierarchical clustering of 122 genes significantly differentially expressed in the

mandibular bone between Fgfr2+/S252W and Fgfr2+/+ littermate embryos. Three biological

replicates were used for each genotype. (C) Volcano plot shows P values and fold changes of

DEGs in the mandibular bone between Fgfr2+/S252W and Fgfr2+/+ littermate embryos. Some of

the most significantly differentially expressed genes (-log10(P-value)>4.5) implicated in

mandibular dysmorphology are shown in blue.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Fig. 4. Increased osteoclastogenesis in the mandibular bone of Fgfr2+/S252W embryos at E16.5.

(A-H) The differential expression of Csf1r and Itgb3 in the mandible of Fgfr2+/+ (A, B, E, F) and

Fgfr2+/S252W embryos (C, D, G, H) were validated by in situ hybridization (ISH). The areas in the

boxes in A, C, E and G are shown in B, D, F and H, respectively, with higher magnification. (I-J)

TRAP assay stained osteoclasts (purple) in the mandible of Fgfr2+/+ (I) and Fgfr2+/S252W (J)

embryos. Scale bar=100 µm. (K-L) Quantitative measurements of the density (K) and

percentage (L) of osteoclasts in the bone area of Fgfr2+/+ (n=3) and Fgfr2+/S252W (n=3) embryos.

The experimental data were analyzed by two-tailed Welch’s t-test and expressed as the mean±

s.e.m. *P<0.05.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Fig. 5. Increased expression of Enpp1 and Ank and elevated PPi concentration in the mandible

of Fgfr2+/S252W embryos at E16.5. RNA expression of Enpp1 and Ank in the mandible of Fgfr2+/+

littermate (A, B, E, F) and Fgfr2+/S252W embryos (C, D, G, H) was validated by in situ hybridization

(ISH). The areas in the boxes in A, C, E and G are shown in B, D, F and H, respectively, with

higher magnification. Scale bar=100 µm. The weight of the mandible (I) and PPi concentration

in the mandible (J) were measured for Fgfr2+/+ littermates (n=7) and Fgfr2+/S252W (n=11)

embryos at E16.5. The experimental data were analyzed by two-tailed Welch’s t-test and

expressed as the mean±s.e.m. *P<0.05.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Fig. 6. Increased cell proliferation of osteoblasts and chondrocytes in the mandible of

Fgfr2+/S252W embryos. (A) The osteoblasts in the mandibular bone at E16.5 were visualized by

IHC for RUNX2. The areas in the boxes are shown on the right in higher magnification,

respectively. (B) Double staining with EdU assay (green) and IHC for RUNX2 (red) at E12.5. (C)

The percentage of proliferating osteoblasts (EdU-positive) in the total osteoblasts (RUNX2-

positive) is shown for Fgfr2+/+ (n=4) and Fgfr2+/S252W (n=4) embryos. (D) EdU assay (green) with

IHC for SOX9 (red) in MC at E12.5. (E) The percentage of proliferating cells (EdU-positive) in MC

(SOX9-positive) is shown for Fgfr2+/+ (n=4) and Fgfr2+/S252W (n=4) embryos. Scale bar=100 µm.

The experimental data were analyzed by two-tailed Welch’s t-test and expressed as the

mean±s.e.m. *P<0.05.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Fig. 7. Proposed model of mandibular dysmorphogenesis in prenatal development of

Fgfr2+/S252W mice.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Table

Table 1. Bone volume (mm3) and bone surface area (mm2) of newborn craniosynostosis

mouse models and their unaffected littermates.

Craniosynostosis Mouse Model

Mean Standard

Deviation

P value*

Statistic Std. Error

Fgfr2+/S252W

Fgfr2+/S252W Bone Volume 2.029372 0.1870266 0.5610799 0.606

Bone Area 51.1714 1.47767 4.43300 1.000

Fgfr2+/+ Bone Volume 2.174143 0.1742227 0.4609499

Bone Area 51.7267 2.22106 5.87638

Fgfr2+/P253R

Fgfr2+/P253R Bone Volume 2.2272 0.13911 0.31105 0.310

Bone Area 45.3520 2.12721 4.75659 1.000

Fgfr2+/+ Bone Volume 2.1741 0.17422 0.46095

Bone Area 51.7267 2.22106 5.87638

Fgfr2cC342Y/+

Fgfr2cC342Y/+ Bone Volume 2.378689 0.1105776 0.3317327 0.258

Bone Area 39.5811 1.77999 5.33997 0.258

Fgfr2c+/+ Bone Volume 2.203444 0.1123060 0.3369180

Bone Area 36.4152 1.04272 3.12815

*Exact significance is displayed for the independent samples Mann-Whitney U Test. The chosen

significance level is 0.05.

Dis

ease

Mo

dels

& M

echa

nism

s •

DM

M •

Acc

epte

d m

anus

crip

t

Fig. S1. Thirty-two anatomical landmarks located on left and right hemimandibles of newborn

(P0) mice. Bilateral landmarks are annotated only on the left hemimandible. Landmark

definitions are listed in Table S1. 3D coordinates of landmarks were collected from the buccal

view (A) and lingual (B) view of an isosurface of the left hemimandible. Landmarks shown in the

rostral (C) and caudal (D) views of the left and right hemimandibles are noted. Scale bar=1 mm.

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Fig. S2. Principal component analysis (PCA) of the anterior body and ramus portion of

hemimandibles of three mouse models of craniosynostosis. Scatter plots of individual scores

based on PCA from left hemimandibles of mutant and unaffected littermates of Fgfr2+/S252W (A)

and Fgfr2+/P253R (B) Apert syndrome mouse models; and Fgfr2cC342Y/+ (C) Crouzon/Pfeiffer

syndrome mouse model along first and second Principal Components axes (PC1 and PC2).

Separate PCAs were estimated for the anterior portion (anterior body, shown in blue in inset

image) and posterior portion (ramus, shown in red in inset image) of the hemimandibles. PCA

of the anterior body of the left hemimandible is shown in the top panel of each column while

PCA of the left ramus portion is shown in the bottom panel.

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Fig. S3. Bone mineral density mapping. Bone mineral density (BMD) was mapped for each of

the genotypes of interest: (A, B) buccal view of left hemimandibles of mice carrying Fgfr2

mutations (B) and their respective unaffected littermates (A); (C, D) lingual view of left

hemimandibles of mice carrying Fgfr2 mutations (D) and their respective unaffected littermates

(C). Though there are differences across the models, the contrast between mutant and

respective unaffected littermates is greatest in the Fgfr2+/S252W mice.

Fig. S4. Cortical bone thickness mapping. Cortical bone thickness was mapped for each of the

genotypes of interest: (A, B) buccal view of left hemimandibles of mice carrying Fgfr2 mutations

(B) and their respective unaffected littermates (A); (C, D) lingual view of left hemimandibles of

mice carrying Fgfr2 mutations (D) and their respective unaffected littermates (C).

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Fig. S5. Histological analysis of the mandibles of Fgfr2+/P253R and Fgfr2cC342Y/+ embryos at

E16.5. (A-D) ALP and alcian blue staining showed osteogenic tissues (red) and Meckel’s cartilage

(blue) in the mandible of Fgfr2+/P253R (B), Fgfr2cC342Y/+ (D) and their Fgfr2+/+ littermate embryos

(A, C, respectively). MB, mandibular bone. MC, Meckel’s cartilage. (E-H) Alizarin red S staining

and (I-L) von Kossa staining showed ossification in the mandible of Fgfr2+/P253R (F, J) and Fgfr2+/+

littermates (E, I), and Fgfr2cC342Y/+ (H, L) and Fgfr2+/+ littermates (G, K). Scale bar=100 µm. Boxes

in G, H, K and L indicate the small area that exhibits impaired microarchitecture at a higher

magnification in Fgfr2cC342Y/+ embryos compared with Fgfr2+/+ littermates.

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Fig. S6. Gene ontology analysis of DEGs in mandibular bone of Fgfr2+/S252W embryos at E16.5.

(A) Biological process. (B) Cellular component. (C) Molecular function.

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Fig. S7. Osteoclast activity in the mandibles of three FGFR2-related craniosynostosis mouse

models at E16.5. (A-E) TRAP staining (purple) in mandibular bone of Fgfr2+/+ (A), Fgfr2+/S252W (B)

and Fgfr2+/P253R (C) embryos on a C57BL6/J (B6) background; and Fgfr2c+/+ (D) and Fgfr2cC342Y/+

(E) on a CD1 background. Scale bar=100 µm. The boxes in D and E indicate the area that shows

increased signal in Fgfr2cC342Y/+ embryos compared with unaffected littermates.

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Fig. S8. Apoptosis in the mandible of Fgfr2+/S252W embryos at E16.5. (A) TUNEL assay in the

mandibular area. TUNEL signal was detected in the mandibular bone (MB) and the

perichondrium of Meckel’s cartilage (MC) of both Fgfr2+/+ and Fgfr2+/S252W embryos. Scale

bar=100 µm. (B-C) Quantification shows the number of TUNEL-positive cells in the mandibular

bone (B) and the perichondrium (C) of Fgfr2+/+ (n=3) and Fgfr2+/S252W (n=3) embryos. The

experimental data were analyzed by two-tailed Welch’s t-test and expressed as the

mean±s.e.m.

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Table S1. Anatomical definitions of landmarks displayed in Fig. 1.

Landmark (Left, Right)

Anatomical Definition Region

1, 2 Inferior-most point on incisor alveolar rim at midline of the incisor at bone-tooth junction

AB

3, 4 Junction of the rim of the alveolar process with incisor, most lateral and centered along the cranial-caudal axis (taken on bone, not tooth)

AB

5, 6 Superior-most point on incisor alveolar rim at midline of bone-tooth junction

AB

7, 8 Anterior lip of the mental foramen AB

9, 10 Antero-lateral edge of the molar alveolar process AB

11, 12 Intersection of molar alveolar rim and base of coronoid process (posterior molar alveolus)

R, AB

13, 14 Most dorsal point of the coronoid process R

15, 16 Most cranio-ventral point between the coronoid and condyloid processes

R

17, 18 Most caudal point on the cranial angle of the condyloid process

R

19, 20 Most caudal point on the ventral angel of the condyloid process

R

21, 22 Most anterior point between the angle that separates the condyloid and angular processes

R

23, 24 Midpoint on the cranial caudal axis of the most posterior aspect of the angular process

R

25, 26 Cranial-most point of the angular process along the ventral surface of the mandible

R, AB

27, 28 Dorsal mandibular foramen R

29, 30 Most ventral edge of the molar alveolar process, toward the midline

R

31, 32 Ventral mandibular foramen AB

Landmarks are classified as located on Anterior Body (AB; n=8) or Ramus (R; n=10) to create

subsets for analysis. Additional landmarks defined on the embryonic mouse skull can be found

at: http://www.getahead.la.psu.edu/landmarks.

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

Table S2. Results (P values) of nonparametric null hypothesis tests for form differences

(EDMA) between mice (P0) carrying a mutation for a specific craniosynostosis syndrome and

their littermates without the mutation.

Craniosynostosis

Model

Left and Right

Hemimandibles

(32 landmarks)

Left Hemimandible

(16 landmarks)

Right Hemimandible

(16 landmarks)

Fgfr2+/S252W 0.001 0.001 0.001

Fgfr2+/P253R 0.001 0.001 0.001

Fgfr2cC342Y/+ 0.001 0.001 0.001

Table S3. Differentially expressed genes in the mandibular bone of Fgfr2+/S252W embryos at

E16.5 compared to their Fgfr2+/+ littermates.

Ensembl ID Gene

Symbol

log2FC Average

Expression

t P Value Adjusted P

Value

ENSMUSG00000030257 Srgap3 1.27673267 6.610930486 9.640369417 5.20142E-07 0.003057916

ENSMUSG00000037370 Enpp1 1.831773279 8.59074871 10.20061207 2.82417E-07 0.003057916

ENSMUSG00000045573 Penk -1.928167528 4.158086959 -9.22055957 8.37469E-07 0.003188606

ENSMUSG00000034573 Ptpn13 1.133258416 6.830424831 8.998531675 1.08474E-06 0.003188606

ENSMUSG00000069917 Hba-a2 -1.622438006 4.106158794 -8.558032456 1.83915E-06 0.003902154

ENSMUSG00000022265 Ank 1.326616334 7.721340364 8.493163202 1.99123E-06 0.003902154

ENSMUSG00000073940 Hbb-bt -1.628190337 8.121203499 -8.263608921 2.64729E-06 0.004446698

ENSMUSG00000042436 Mfap4 -1.15168023 6.836719917 -8.135654118 3.11049E-06 0.004571637

ENSMUSG00000024593 Megf10 0.927309625 6.425005115 7.528605255 6.85637E-06 0.008957467

ENSMUSG00000036905 C1qb -2.263114164 2.013981763 -6.999425722 1.41583E-05 0.011193739

ENSMUSG00000052305 Hbb-bs -1.57470114 9.839937637 -7.157779 1.13549E-05 0.011193739

ENSMUSG00000035783 Acta2 -1.163348891 5.560032653 -6.993321688 1.42802E-05 0.011193739

ENSMUSG00000006369 Fbln1 0.877748065 5.903300431 7.109264313 1.21449E-05 0.011193739

ENSMUSG00000054594 Oscar 1.217936999 5.141501095 7.270388572 9.72461E-06 0.011193739

ENSMUSG00000040703 Cyp2s1 1.331780479 3.815520063 6.998767749 1.41714E-05 0.011193739

ENSMUSG00000031351 Zfp185 -2.478906784 3.159340528 -6.742466291 2.03933E-05 0.011887187

ENSMUSG00000033491 Prss35 -1.351275459 5.889560643 -6.721057207 2.1031E-05 0.011887187

ENSMUSG00000040569 Slc26a7 -1.106708548 6.18519432 -6.763093111 1.97983E-05 0.011887187

ENSMUSG00000037664 Cdkn1c -0.880989857 8.705576379 -6.714493501 2.12307E-05 0.011887187

ENSMUSG00000024621 Csf1r 0.932343075 7.271710056 6.790660552 1.90318E-05 0.011887187

ENSMUSG00000020340 Cyfip2 1.222456601 6.830403391 6.838067414 1.77866E-05 0.011887187

ENSMUSG00000028047 Thbs3 -1.146915292 4.350320931 -6.66151792 2.29185E-05 0.012248919

ENSMUSG00000052512 Nav2 0.954836312 6.560031841 6.618351607 2.43993E-05 0.012473328

ENSMUSG00000014846 Tppp3 -2.657022312 2.498697296 -6.399689044 3.36333E-05 0.015732053

ENSMUSG00000055013 Agap1 0.824170057 6.949305293 6.376950077 3.47877E-05 0.015732053

ENSMUSG00000023964 Calcr 1.39086817 4.385746777 6.386112042 3.43176E-05 0.015732053

ENSMUSG00000044667 Plppr4 -1.64880105 3.281452634 -6.337299974 3.69024E-05 0.016070294

ENSMUSG00000021388 Aspn -1.565468808 7.851433763 -6.235000826 4.30138E-05 0.017532802

ENSMUSG00000026321 Tnfrsf11a 0.878575936 6.328271017 6.231469435 4.3243E-05 0.017532802

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

ENSMUSG00000069516 Lyz2 -1.48397449 5.657315666 -5.984140315 6.30167E-05 0.021283798

ENSMUSG00000030218 Mgp -1.221549215 4.443285577 -6.029622637 5.87632E-05 0.021283798

ENSMUSG00000041559 Fmod -0.935657006 6.937668196 -6.047249988 5.71975E-05 0.021283798

ENSMUSG00000026469 Xpr1 0.756418685 8.588488909 5.965430507 6.48599E-05 0.021283798

ENSMUSG00000035311 Gnptab 0.877553017 7.078081271 6.048247317 5.71102E-05 0.021283798

ENSMUSG00000022303 Dcstamp 1.019517844 5.581854423 5.945315992 6.69054E-05 0.021283798

ENSMUSG00000040061 Plcb2 1.240168118 3.840512082 5.944636192 6.69757E-05 0.021283798

ENSMUSG00000021214 Akr1c18 1.299377329 4.04982874 5.9744297 6.39663E-05 0.021283798

ENSMUSG00000020053 Igf1 -1.273626277 6.98869413 -5.913249668 7.03087E-05 0.021754992

ENSMUSG00000028238 Atp6v0d2 0.857306076 8.426331011 5.891337145 7.27394E-05 0.021929993

ENSMUSG00000031283 Chrdl1 -0.834377044 5.949540909 -5.813408207 8.21296E-05 0.021947271

ENSMUSG00000001270 Ckb 0.899333325 8.11410426 5.841953781 7.85492E-05 0.021947271

ENSMUSG00000050390 C77080 1.100606755 4.008047025 5.824856679 8.06733E-05 0.021947271

ENSMUSG00000058897 Col25a1 1.145484143 6.457814805 5.817728948 8.15768E-05 0.021947271

ENSMUSG00000025089 Gfra1 1.467929152 4.267875815 5.853457969 7.71533E-05 0.021947271

ENSMUSG00000012405 Rpl15 -0.906856775 6.677200075 -5.773069762 8.74862E-05 0.022362233

ENSMUSG00000021728 Emb 0.914199758 5.859278388 5.781838677 8.62911E-05 0.022362233

ENSMUSG00000032948 Lipi -0.936729854 5.444723773 -5.663403914 0.000104002 0.025438301

ENSMUSG00000022324 Matn2 -0.694566276 5.740143219 -5.640097671 0.000107918 0.025438301

ENSMUSG00000052459 Atp6v1a 0.718305056 8.851933228 5.638604413 0.000108174 0.025438301

ENSMUSG00000044447 Dock5 1.048477353 6.571337781 5.639942804 0.000107945 0.025438301

ENSMUSG00000017737 Mmp9 1.19764316 9.755281083 5.569229207 0.000120814 0.027853587

ENSMUSG00000035296 Sgcg -4.411710792 -0.922779598 -5.517657881 0.000131215 0.028010594

ENSMUSG00000032291 Crabp1 -2.155264427 3.115239671 -5.497116825 0.000135618 0.028010594

ENSMUSG00000052854 Nrk -1.487662892 6.623819756 -5.547251564 0.000125136 0.028010594

ENSMUSG00000001348 Acp5 0.793398575 8.627342467 5.516153703 0.000131533 0.028010594

ENSMUSG00000021306 Gpr137b 0.861604003 6.046324047 5.519626396 0.000130802 0.028010594

ENSMUSG00000032122 Slc37a2 1.191073369 7.077239901 5.496333193 0.000135789 0.028010594

ENSMUSG00000069919 Hba-a1 -1.337990689 7.815219545 -5.422264917 0.000153024 0.030909916

ENSMUSG00000020689 Itgb3 1.110567359 7.880369008 5.41393388 0.000155102 0.030909916

ENSMUSG00000027562 Car2 0.745533042 6.189698692 5.394819185 0.000159983 0.031351269

ENSMUSG00000022519 Srl -0.874711972 4.426953711 -5.369870117 0.000166598 0.03211249

ENSMUSG00000040037 Negr1 -2.06759258 3.207329165 -5.347098779 0.000172888 0.032266944

ENSMUSG00000041362 Shtn1 0.856606818 5.623009126 5.348913307 0.000172378 0.032266944

ENSMUSG00000029765 Plxna4 1.049654389 5.846000814 5.319457787 0.000180861 0.033227474

ENSMUSG00000035258 Abi3bp -1.564181951 4.695670152 -5.29502116 0.000188231 0.033533688

ENSMUSG00000027962 Vcam1 -1.033135854 4.531254283 -5.295700595 0.000188022 0.033533688

ENSMUSG00000021390 Ogn -1.35118064 7.552490868 -5.285796718 0.000191095 0.033535785

ENSMUSG00000047139 Cd24a -0.988039433 5.207433327 -5.274863718 0.000194549 0.033639831

ENSMUSG00000026072 Il1r1 0.782735636 4.957639421 5.248885822 0.000203022 0.034101898

ENSMUSG00000028581 Laptm5 0.879244011 6.642567459 5.25154349 0.000202138 0.034101898

ENSMUSG00000032060 Cryab -1.110150001 3.610742127 -5.220757626 0.000212635 0.035213634

ENSMUSG00000024236 Svil 0.624813843 7.133248611 5.188180982 0.000224371 0.036640971

ENSMUSG00000026576 Atp1b1 1.17396213 4.596809676 5.172977602 0.000230079 0.037058409

ENSMUSG00000042082 Arsb 0.97021764 7.506223253 5.160310942 0.000234951 0.037331753

ENSMUSG00000004151 Etv1 0.730282583 4.988104973 5.132028881 0.000246224 0.038601426

ENSMUSG00000002014 Ssr4 -0.810564086 5.903376453 -5.097609777 0.000260716 0.039301265

ENSMUSG00000034707 Gns 0.727994379 7.459110088 5.104733751 0.000257645 0.039301265

ENSMUSG00000027670 Ocstamp 1.019536701 5.872442547 5.107174041 0.000256601 0.039301265

ENSMUSG00000061731 Ext1 0.663550112 7.436421895 5.069385813 0.000273267 0.04067181

ENSMUSG00000035357 Pdzrn3 -0.775776383 5.557112496 -5.047366705 0.000283499 0.041667193

ENSMUSG00000000290 Itgb2 0.939153536 5.118505724 5.024342179 0.000294628 0.042768391

ENSMUSG00000033220 Rac2 0.7747666 6.466527167 5.001754487 0.000305993 0.043876437

ENSMUSG00000023886 Smoc2 -1.076440655 5.017467359 -4.98683623 0.000313751 0.043917621

ENSMUSG00000039116 Gpr126 -0.981742058 4.952859094 -4.988178407 0.000313044 0.043917621

ENSMUSG00000015134 Aldh1a3 -1.172030326 5.661165159 -4.96080253 0.000327786 0.044910263

ENSMUSG00000026921 Egfl7 -0.844921515 4.284607323 -4.957154752 0.000329805 0.044910263

ENSMUSG00000018567 Gabarap -0.694331619 7.756610472 -4.952676418 0.000332301 0.044910263

ENSMUSG00000037533 Rapgef6 0.614649699 7.303891696 4.92217191 0.000349838 0.046743173

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n

ENSMUSG00000026596 Abl2 0.608326336 6.50724323 4.910707644 0.000356678 0.047121542

ENSMUSG00000044468 Fam46c -0.684580972 6.669624759 -4.897416701 0.000364783 0.047656834

ENSMUSG00000004730 Emr1 -2.239807676 1.348603199 -4.832587645 0.000407175 0.049783334

ENSMUSG00000027313 Chac1 -2.213689517 4.887600415 -4.764393172 0.000457375 0.049783334

ENSMUSG00000027996 Sfrp2 -1.171324394 4.710267004 -4.788885324 0.000438639 0.049783334

ENSMUSG00000062515 Fabp4 -1.05622216 3.509023446 -4.704234021 0.000507036 0.049783334

ENSMUSG00000028583 Pdpn -1.045235352 3.565172539 -4.802701535 0.000428426 0.049783334

ENSMUSG00000044337 Ackr3 -1.019213136 4.030412869 -4.708330855 0.000503482 0.049783334

ENSMUSG00000029838 Ptn -0.936047162 8.138301746 -4.752584264 0.000466706 0.049783334

ENSMUSG00000027204 Fbn1 -0.801254068 6.408244189 -4.808538301 0.000424186 0.049783334

ENSMUSG00000026421 Csrp1 -0.783880408 6.759861112 -4.787747037 0.000439492 0.049783334

ENSMUSG00000026725 Tnn -0.777064263 7.437115497 -4.847170068 0.000397209 0.049783334

ENSMUSG00000071856 Mcc -0.73828315 5.345518229 -4.714396914 0.000498266 0.049783334

ENSMUSG00000024534 Sncaip -0.723674183 4.868083805 -4.694867838 0.000515261 0.049783334

ENSMUSG00000038871 Bpgm -0.682666203 5.249711873 -4.7094474 0.000502517 0.049783334

ENSMUSG00000022912 Pros1 0.663131966 6.459136494 4.749861565 0.000468885 0.049783334

ENSMUSG00000033705 Stard9 0.706268577 6.374308981 4.698304549 0.000512227 0.049783334

ENSMUSG00000061175 Fnip2 0.710174275 5.074804605 4.84910423 0.000395907 0.049783334

ENSMUSG00000028111 Ctsk 0.713395398 9.9623244 4.816083029 0.000418771 0.049783334

ENSMUSG00000075324 Fign 0.737656804 5.609855492 4.780463665 0.000444989 0.049783334

ENSMUSG00000028962 Slc4a2 0.773555819 6.05593102 4.795937482 0.000433394 0.049783334

ENSMUSG00000041235 Chd7 0.793445297 5.723571848 4.736175534 0.000480003 0.049783334

ENSMUSG00000039062 Anpep 0.813927796 8.056912398 4.76404628 0.000457646 0.049783334

ENSMUSG00000068854 Hist2h2be 0.830691553 4.786180283 4.735427416 0.000480619 0.049783334

ENSMUSG00000038665 Dgki 0.831832322 4.196277449 4.707501683 0.000504199 0.049783334

ENSMUSG00000079057 Cyp4v3 0.83836864 4.764351948 4.838702637 0.000402965 0.049783334

ENSMUSG00000022231 Sema5a 0.839094343 9.365098271 4.837629137 0.0004037 0.049783334

ENSMUSG00000026437 Cdk18 0.88555389 3.867251827 4.693416906 0.000516548 0.049783334

ENSMUSG00000022012 Enox1 0.892698352 3.686678832 4.709026013 0.000502881 0.049783334

ENSMUSG00000018774 Cd68 0.959283668 7.245963357 4.703581561 0.000507605 0.049783334

ENSMUSG00000079625 Tm4sf19 1.030522587 3.924772294 4.762574467 0.000458799 0.049783334

ENSMUSG00000038418 Egr1 1.120357726 6.447009919 4.772930692 0.000450751 0.049783334

ENSMUSG00000049493 Pls1 1.373575186 2.360582503 4.733077184 0.000482558 0.049783334

ENSMUSG00000062151 Unc13c 1.540649612 3.574785275 4.723859106 0.000490244 0.049783334

Table S4. Primers used to generate riboprobe templates for RNA in situ hybridization.

Gene Primers References

Ank 5’-GAGTAATACGACTCACTATAGGGTGGGATGTGCCTCAATCTCA-3’

5’-GAGATTAACCCTCACTAAAGGGACACAGAGTTCTGCAAAGGCAA-3’

Uzuki et al., 2014

Csf1r 5’-GAGTAATACGACTCACTATAGGGAGGAGGTGTCTGTGGGTGAC-3’

5’-GAGATTAACCCTCACTAAAGGGATGGTACTTCGGCTTCTGCTT-3’

Designed using

Primer3

Enpp1 5’-GAGTAATACGACTCACTATAGGGGCTGTCTGAGACTCCCTTGG-3’

5’-GAGATTAACCCTCACTAAAGGGAGTCCCCAGACCACGTACACT-3’

Designed using

Primer3

Itgb3 5’-GAGTAATACGACTCACTATAGGGGAAAATGTCGTCAGCCTTTACC-3’

5’-GAGATTAACCCTCACTAAAGGGAGCAGGAGAAGTCATCGCACTC-3’

Diez-Roux et al.,

2011

Disease Models & Mechanisms: doi:10.1242/dmm.038513: Supplementary information

Dis

ease

Mo

dels

& M

echa

nism

s •

Sup

plem

enta

ry in

form

atio

n