a

RESEARCH ARTICLE

Embryonic Craniofacial Bone Volume andBone Mineral Density in Fgfr21/P253R andNonmutant MiceChristopher J. Percival,1 Yuan Huang,2 Ethylin Wang Jabs,3 Runze Li,2 and Joan T. Richtsmeier1*

Background: Quantifying multiple phenotypic aspects of individual craniofacial bones across early osteo-genesis illustrates differences in typical bone growth and maturation and provides a basis for under-standing the localized and overall influence of mutations associated with disease. We quantify the typicalpattern of bone growth and maturation during early craniofacial osteogenesis and determine how thispattern is modified in Fgfr21/P253R Apert syndrome mice. Results: Early differences in typical relativebone density increase are noted between intramembranous and endochondral bones, with endochondralbones normally maturing more quickly during the prenatal period. Several craniofacial bones, includingthe facial bones of Fgfr21/P253R mice, display lower volumes during the earliest days of osteogenesis andlower relative densities until the perinatal period relative to unaffected littermates. Conclusions: Esti-mates of bone volume and linear measures describing morphology do not necessarily covary, highlightingthe value of quantifying multiple facets of gross osteological phenotypes when exploring the influence ofa disease causing mutation. Differences in mechanisms of osteogenesis likely underlie differences inintramembranous and endochondral relative density increase. The influence of the FGFR2 P253R muta-tion on bone volume changes across the prenatal period and again after birth, while its influence on rela-tive bone density is more stable. Developmental Dynamics 000:000–000, 2014. VC 2013 Wiley Periodicals, Inc.

Key words: micro-computed tomography; bone development; functional regression analysis; Apert syndrome

Key findings:� Quantifying multiple aspects of gross bone phenotype can provide a more complete understanding of typical

bone development and the influence of known mutations.� Within the mouse skull, endochondrally ossified bones increase in relative density quickly during the prenatal

period, while intramembranously ossified bones do not.� There appear to be shifts in the influence of the FGFR2 P253R mutation on craniofacial bone volume during

both the prenatal and postnatal periods.

Submitted 1 August 2013; First Decision 8 October 2013; Accepted 25 October 2013

INTRODUCTION

The skull is morphologically, evolutio-narily, and developmentally complex,making the study of normal craniofa-cial osteogenesis relevant for a largenumber of research questions. More-over, an understanding of skull devel-

opment is important for the design oftherapeutic agents for the large num-ber of medical syndromes that includecraniofacial dysplasias. While thegenetic pathways and developmentalmechanisms associated with pheno-types such as cleft palate (e.g., Gritli-

Linde, 2008; Dixon et al., 2011; Bushand Jiang, 2012) and craniosynostosis(e.g., Cohen and Maclean, 2000;Morriss-Kay and Wilkie, 2005; Passos-Bueno et al., 2008) have been a focus ofresearch, osteogenesis of whole cranio-facial bones is less well understood,

Dev

elop

men

tal D

ynam

ics

1Department of Anthropology, Penn State University, University Park, Pennsylvania2Department of Statistics, Penn State University, University Park, Pennsylvania3Department of Genetics and Genomic Sciences, Icahn School of Medicine at Mount Sinai, New York, New YorkGrant sponsor: NSF; Grant numbers: BCS-1061554; BCS-0725227; Grant sponsor: NIH/NIDCR; Grant numbers: R01DE018500;3R01DE018500-02S1; R01DE022988..*Correspondence to: Joan T. Richtsmeier, 409 Carpenter Building, University Park, PA, 16802. E-mail: jta10@psu.edu

DOI: 10.1002/dvdy.24095Published online 00 Month 2013 in Wiley Online Library (wileyonlinelibrary.com).

DEVELOPMENTAL DYNAMICS 00:000–000, 2014

VC 2013 Wiley Periodicals, Inc.

particularly in the case of intramem-branously ossified bones.

Quantification of normal patternsof craniofacial bone growth and devel-opment can provide insight about howdifferences in cellular origin, mode ofossification, and initial timing of ossi-fication may be associated with varia-tion of skull phenotypes. In addition,quantifying how mutations associatedwith craniofacial dysmorphologychange these baseline patterns ofgrowth and development can helpreveal how mutations lead to changesin skeletal phenotype through dysre-gulation of developmental systems,thereby contributing to the develop-ment of therapeutic treatment. Land-mark based morphometrics hasbecome a standard way to quantifyimportant changes in gross bone mor-phology across the skull that arebased in known mutations associatedwith human syndromes (e.g., Olsonet al., 2004; Perlyn et al., 2006; Hillet al., 2007; Boughner et al., 2008;Hallgr�ımsson et al., 2009; Mart�ınez-Abad�ıas et al., 2010; Schmidt et al.,2010; Percival et al., 2012) and toidentify locations for targeting specificsites for cellular analyses (Mart�ınez-Abad�ıas et al., 2013). Landmark-based measures estimate importantaspects of craniofacial bone pheno-type. However, the quantification ofadditional facets of skull phenotype,including bone volume and density,can provide additional complemen-tary data on the nature of skeletalvariation and can serve as the basisfor additional nuanced hypothesesabout the origins of normal and syn-dromic skeletal variation.

In this study, we quantify importantaspects of pre- and perinatal growthand maturation for individual craniofa-

cial bones, and then determine howthis pattern is modified by the Apertsyndrome associated Fgfr2 P253Rmutation. Here, growth refers tochanges in bone quantity while matu-ration refers to changes in bone qualityover development. Previous analysis ofFgfr2þ/P253R Apert syndrome mice indi-cated that mutants can be identified asearly as embryonic day (E) 16.5 by theirrelatively domed heads (Yin et al.,2008), that coronal fusion is evidentalong with changes in cell activity dur-ing the fetal period and that skull sizeis smaller at birth (Holmes et al., 2009;Wang et al., 2010; Mart�ınez-Abad�ıaset al., 2010). Both intramembranousand endochondral ossification processesare known to be influenced by theP253R mutation (Yin et al., 2008; Wanget al., 2010) and although dysmorphol-ogy may be more apparent in the facialskeleton, dysmorphology appears else-where in the skull of Fgfr2þ/P253R mice(Mart�ınez-Abad�ıas et al., 2010; Duet al., 2010). Therefore, we hypothesizethat all individual craniofacial bones ofFgfr2þ/P253R mice will display reducedbone volume and relative bone mineraldensity from the time of initial ossifica-tion through the postnatal period,although bones previously associatedwith early suture fusion (frontal andparietal bones) and those with knownshape changes (maxilla, premaxilla,palatine, nasals) are expected to displaymore severe reductions.

EXPERIMENTAL

PROCEDURES

Sample and Imaging

We use the previously described Apertsyndrome Fgfr1/P253R mouse modelin this study (Wang et al., 2010).Fgfr21/P253Rneo mice were crossed with

EIIA Cre homozygotes so that roughlyhalf of our litters heterozygouslyexpress the Fgfr2 P253R mutation inall tissues (Fgfr21/P253R) and half donot express this mutation (Fgfr2þ/þ).Based upon timed mating and evidenceof pregnancy, litters were sacrificed onembryonic days (E) 14.5, 15.5, 16.5,17.5 and postnatal days (P) 0 and 2.The E17.5, P0, and P2 mice in thisstudy were used in previous landmarkbased morphometric analyses (Wanget al., 2010; Mart�ınez-Abad�ıas et al.,2010; Motch et al., 2012; Hill et al.,2013). After sacrifice, specimens werefixed in 4% paraformaldehyde andstored in 0.01 M phosphate bufferedsaline with sodium azide as an antibac-terial agent. Care and use of mice forthis study were in compliance withrelevant animal welfare guidelinesapproved by Mount Sinai School ofMedicine and Pennsylvania StateUniversity Animal Care and UseCommittees.

High resolution micro-computedtomography (mCT) images of mouseheads (Table 1) were acquired in airat the Center for Quantitative X-RayImaging at Pennsylvania State Uni-versity (www.cqi.psu.edu) using anOMNI-X Universal HD600 industrialx-ray computed tomography system(Varian Medical Systems, Palo AltoCA). Solid hydroxyapatite phantoms(QRM GmbH, M€oehrendorf, Ger-many) scanned with each set of skullsallowed us to linearly associate rela-tive x-ray attenuation values withbone mineral density estimates.

Bone Volume and Density

Histograms

Using Avizo software (VisualizationSciences Group, Burlington, MA),

Dev

elop

men

tal D

ynam

ics

TABLE 1. Sample Sizes of Fgfr21/P253R and Fgfr21/1 Mice, the Associated Image Resolution, the Minimum, and

mCT Scan Settings by Age

Age

No. of

Fgfr2þ/þ

specimens

No. of

Fgfr2þ/P253R

specimens

Voxel

width

(mm)

Slice

thickness

(mm) kVp mA

Frame

averaging

No. of

projections

E15.5 7 7 0.0107 0.0125 100 0.19 5 2,400E16.5 16 9 0.0107 0.0125 100 0.19 5 2,400E17.5 7 9 0.0137 0.0154 110 0.18 3 2,400P0 14 11 0.0151 0.0159 130 0.15 4 1,440P2 8 5 0.0151 0.0159 130 0.15 4 1,440

2 PERCIVAL ET AL.

craniofacial bones with a minimum of74.0 mg/mm3 partial density ofhydroxyapatite were manually identi-fied (segmented) from mCT images ofE15.5, E16.5, and E17.5 skulls (nobone was detected in the mCT imagesof E14.5 specimens). Manual segmen-tation minimizes segmentation errorassociated with the relatively low lev-els of ossification within embryonicbones. Craniofacial bones of thelarger and more highly ossified P0and P2 specimens were segmentedusing a modified version of a previ-ously described semi-automatic seg-mentation method (Percival et al.,2012). After segmentation of speci-mens of all ages, a histogram of 126bone density values between 74.0 and372.2 mg/mm3 partial density ofhydroxyapatite (Fig. 1) was estimatedfor each identified bone. Because ofimage saturation for some bones ofthe older specimens, the highest den-sity values may include volumes ofbone above the maximum densityvalue.

Measurements from 16 bones serveas the basis of our analysis (Fig. 2).For each age-genotype associatedgroup of mice, either the right or leftversion of bilateral bones (shown pre-viously to display similar measureswithin individuals; Percival, 2013)was chosen for analysis based on seg-mentation accuracy. Teeth were notsegmented during manual segmenta-tion of embryonic specimens to avoidconfounding them with surroundingbone. For older specimens, teeth weremanually segmented independentlyand their density histograms weresubtracted from associated bone den-sity histograms before analysis.

Total individual bone volumes, thesum of histogram values for all den-sities multiplied by the mCT imageresolution, were compared betweenthe genotypes within age categorieswith two-sample Wilcoxon (Mann-Whitney) tests (a¼ 0.05), includingBonferroni correction for multipletesting, using R (R DevelopmentalCore Team, 2008). Mean total bonevolumes, calculated for each age-genotype combination, serve as aproxy for bone size, while differencesin volume between ages representbone growth. Mean individual bonevolumes standardized by the totalbone volume of the 16 bones under

study represent relative bone size ateach age-genotype combination.

Standardized bone density histo-grams were calculated for each speci-men by dividing each entry of thevoxel count filled bone density histo-grams by the total number of voxelsassociated with that bone. This stand-ardizes voxel counts by total bone vol-ume, removing variation associatedwith differences in scale. Mean stand-ardized bone density histograms werecalculated for the bones of each age-genotype combination. Spline curvesapproximating these standardizedhistograms, called relative densitycurves, serve as the basis for subse-quent density analysis (Fig. 1) andprovide the basis for the quantitativeevaluation of bone maturation.

Using the Fgfr2þ/þ relative bonedensity curves as a comparative, nor-mative bone maturation baseline,functional regression analysis wasperformed with the fda package in R(Ramsay et al., 2009), to determinethe influence of the P253R mutationon bone maturation across the ear-liest stages of craniofacial ossification.Because image saturation in somebones of the older specimens has astrong artificial influence on functionsestimated from histograms, the fivehighest bone density values were notincluded in functional analysis. Usingthe remaining 121 values from thestandardized histograms, cubic splinefunctions were estimated for eachbone using five knots. Functionalmultivariate regressions were com-puted across density values (d) foreach bone with Age as a numericalcovariate, Genotype as a binary cova-riate, and with an optional Age*Geno-type interaction term using thefollowing model:

EðyiðdÞÞ ¼ b0ðdÞ þ b1ðdÞGenotypeþ b2ðdÞAge

EðyiðdÞÞ ¼ b0ðdÞ þ b1ðdÞGenotypeþ b2ðdÞAgeþ b3ðdÞ IntGenotype;Age

Three separate multivariateregressions of relative density curves(using all ages, only prenatal ages[E15.5–P0], or only postnatal ages[P0, P2]) were computed for eachbone to identify differences in theeffects of the P253R mutation beforeand after birth. For each bone,

regressions were estimated only forthose age classes in which the bonecould be measured in all specimens.The 95% confidence intervals of theresulting coefficient curves werecomputed to determine the statisticalsignificance of the effect of the associ-ated covariate.

RESULTS

Volume Analysis

Ossified portions of some of the 16bones identified from mCT images dis-play gross differences in extent andshape between Fgfr2þ/P253R Apertmodel mice and their unaffected lit-termates (Fgfr2þ/þ) (Fig. 2). The ossi-fied volume of the skull increasesevery day between E15.5 and post-natal day P2 (Table 2), with no signifi-cant difference in overall bone volumebetween the genotypes at any age.The proportional mean volume ofindividual elements varies betweenages (Table 2). Further illustrations ofbone extent and volume are availableat www.getahead.psu.edu/ApertDen-Vol for interested readers.

The volumes of specific elementsdiffer between genotypes at someages, suggesting strong trends (Figs.3, 4). At E15.5, Fgfr2þ/P253R mice dis-play (nonsignificant) lower mean vol-umes than Fgfr2þ/þ mice across allrelatively well-ossified bones (Fig.3A). Importantly, removing a singleFgfr2þ/P253R mouse of particularlyhigh ossification from the samplechanges to significant the differencesin volumes of SquT, Max, and PMaxbetween Fgfr2þ/þ and Fgfr2þ/P253R

mice, revealing the effects of strongintra-group variation on analyticalresults. The bones of E16.5 Fgfr2þ/P253R

mice have lower or similar mean vol-umes relative to Fgfr2þ/þ (Fig. 3B). AtE17.5, bone volumes are not signifi-cantly different between the geno-types (Fig. 3C). P0 Fgfr2þ/P253R micedisplay higher mean volumes thanFgfr2þ/þ for some bones and lower forothers (Fig. 4A). At P2, the volumes ofsix bones are significantly reduced inFgfr2þ/P253R mice (Fig. 4B). Becauseof relatively high levels of semi-automatic segmentation error for Pal,PSph, SphA, and SphB in P2mutants, the results for these bones

Dev

elop

men

tal D

ynam

ics

CRANIOFACIAL BONE VOLUME AND DENSITY 3

at P2 and in the postnatal regressionshould be accepted with caution.

Histogram Analysis

Bone density histograms summing alldensity values within identified boneswere standardized by total bone vol-ume to serve as a basis for comparingrelative bone densities across speci-mens. Functional representations ofthese standardized bone density histo-grams, called relative density curves,were used as the basis of statisticalcomparisons of relative bone density(Fig. 1). Relative density curves ofFgfr2þ/þ mice, representing each ageof a given bone, were plotted togetherto illustrate typical changes in relativebone mineral density. Most bones dis-play very low relative density atE15.5, but mature in one of threeways (Fig. 5; Table 3). Group 1 (contin-uous increase) bones display a consist-ent and continuous change from lowrelative bone density to higher density

across the age range (Fig. 5A,B).Group 2 bones (medium density) dis-play an intermediate relative bonedensity from E16.5 to P2 (Fig. 5C–E).Group 3 bones (low density) retain alow relative bone mineral density untilthe postnatal period, when increase inrelative density often occurs betweenP0 and P2 (Fig. 5F–H). A strongincrease between P0 and P2 is particu-

larly apparent for bones of group 3,but not limited to this group. Illustra-tions of relative density curves for allskull bones are available at www.geta-head.psu.edu/ApertDenVol for inter-ested readers.

Functional multivariate regressionsof relative density curves for eachspecimen indicated that Age andGenotype are significant factors in

Dev

elop

men

tal D

ynam

ics

Fig. 1. Simplified example illustrating relative density histogram and curve quantification; whereminimum bone density is 5 and the count of bone voxels is 24. Bone volume¼ 24 * mCT imageresolution. Standardized Histogram¼Histogram / 24. Relative density curves are estimated fromstandardized histograms using cubic spline functions.

TABLE 2. Comparison of Patterns of Bone Growth Between Genotypesa

Fgfr21/1 Fgfr21/P253R

Mean total bone volume (mm3) by age, with standard deviation

E15.5 E16.5 E17.5 P0 P2 E15.5 E16.5 E17.5 P0 P2

0.28(0.11)

1.23(0.29)

3.11(0.48)

5.44(1.05)

10.89(1.96)

0.20(0.17)

1.05(0.35)

3.62(0.45)

5.58(1.42)

9.99(2.15)

Mean % of total volume at each age (SD)

E15.5 E16.5 E17.5 P0 P2 E15.5 E16.5 E17.5 P0 P2

Man 53.3 (8.8) 35.7 (3.4) 31.2 (1.6) 26.8 (1.4) 22.2 (1.0) 54.1 (9.0) 31.7 (1.1) 28.6 (1.3) 24.8 (1.2) 23.7 (1.0)BasO 12.3 (4.4) 13.8 (1.1) 10.9 (0.9) 10.7 (0.9) 7.5 (0.4) 12.6 (3.1) 17.0 (0.8) 11.8 (0.7) 11.0 (1.4) 7.9 (0.8)Fro 10.3 (3.4) 12.6 (0.8) 12.9 (0.6) 10.3 (0.7) 10.7 (0.4) 14.1 (1.7) 13.1 (0.9) 12.9 (0.7) 10.1 (0.5) 10.6 (0.4)Max 11.5 (1.7) 11.1 (0.6) 11.6 (0.7) 10.0 (0.5) 9.6 (0.1) 8.3 (1.5) 9.7 (0.4) 12.7 (0.8) 12.2 (0.6) 10.8 (1.3)LatO 4.9 (2.4) 8.4 (1.0) 7.0 (0.8) 6.6 (0.6) 4.7 (0.2) 5.1 (3.5) 10.2 (0.6) 7.5 (0.6) 6.9 (0.9) 4.8 (0.2)PMax 2.4 (1.2) 3.5 (0.5) 4.3 (0.9) 5.5 (0.7) 5.9 (0.2) 1.0 (0.8) 2.1 (0.5) 4.1 (0.8) 6.0 (0.6) 5.8 (0.9)Pal 2.8 (0.6) 3.3 (0.2) 3.6 (0.2) 3.9 (0.3) 3.2 (0.2) 2.4 (0.5) 3.7 (0.4) 4.2 (0.3) 4.4 (0.3) 4.0 (0.2)Par 1.1 (0.9) 4.3 (0.9) 5.9 (1.2) 4.7 (1.0) 6.5 (0.4) 1.6 (1.7) 5.1 (0.8) 6.2 (0.7) 4.6 (1.0) 5.9 (0.4)SquT 0.8 (0.1) 1.5 (0.2) 2.2 (0.1) 2.8 (0.4) 3.0 (0.1) 0.5 (0.2) 1.1 (0.1) 2.0 (0.2) 2.6 (0.2) 2.8 (0.1)SphB 0.3 (0.3) 3.0 (0.8) 4.7 (0.4) 6.0 (0.6) 5.2 (0.3) 0 3.5 (0.5) 4.5 (0.3) 5.8 (0.7) 4.8 (0.3)SphA 0.3 (0.3) 1.7 (0.4) 2.9 (0.2) 3.6 (0.6) 3.3 (0.2) 0.4 (0.2) 1.9 (0.3) 2.8 (0.2) 3.7 (0.2) 3.2 (0.2)IPar 0 0.6 (0.3) 1.8 (0.5) 2.0 (0.7) 3.8 (0.2) 0 0.6 (0.2) 1.5 (0.2) 1.6 (0.5) 3.0 (0.3)PetT 0.1 (0.1) 0.5 (0.2) 0.5 (0.1) 1.1 (0.2) 5.3 (1.7) 0 0.3 (0.1) 0.4 (0.1) 0.9 (0.2) 3.7 (0.7)SquO 0 0 0 3.3 (0.5) 4.3 (0.3) 0 0 0 2.0 (1.0) 3.6 (0.5)PSph 0 0 0.2 (0.1) 2.4 (0.3) 3.1 (0.2) 0 0 0.4 (0.2) 2.6 (0.4) 3.7 (0.2)Nas 0 0 0.2 (0.1) 0.6 (0.3) 1.6 (0.1) 0 0 0.4 (0.2) 0.7 (0.6) 1.7 (0.4)

aAbove are mean total volumes of the 16 bones under study by age measured as mm3, with standard deviation in parentheses.Below is the mean relative volume of each bone by age and genotype, calculated as the percentage of mean total bone volume atthe associated age, with standard deviation in parentheses. Bone abbreviations are defined in Figure 2. A graphical illustrationof this table can be found at www.getahead.psu.edu/ApertDenVol by interested readers.

4 PERCIVAL ET AL.

influencing relative bone density. Theaddition of an interaction termbetween Age and Genotype in theregressions did not qualitativelychange the results, so we report theresults of the regressions without theinteraction term. Most bones show asignificant age effect for which theproportion of low density bonestrongly decreases with age and theproportion of higher density boneincreases with age more subtly(Figs. 6, 7). The bones that displayweak or nonsignificant age effects areIPar, Man (Fig. 6D), PetT, and Maxfor the prenatal regression (E15.5–P0), and LatO (Fig. 7C) for the post-natal regression (P0–P2). Bones thattypically display more obviouschanges in relative density curves

across the ages and proportionallymore high density bone (Fig. 5) alsotend to have age effect coefficientsthat remain significant across a widerrange of density values (Figs. 6, 7).

In all cases of a significant effect ofGenotype on relative bone density, theFGFR2 P253R mutation is associatedwith a higher proportion of low den-sity bone and a lower proportion ofhigher density bone, which is similarto a reduction in the value of the Agecovariate. For prenatal regressions,genotype has a significant effect forMan (Fig. 6D), Max, SquT (Fig. 6A),and PMax, while PetT and Pal displaya weakly significant genotype effect.For postnatal regressions, genotypehas a significant effect on Max, SquT,Pal, and a weak effect on Man (Fig.

7D), Fro (Fig. 7B), PetT, and SquO.For bones that display a significantgenotype effect, the relative densitycurves of Fgfr2þ/P253R mice tend tocluster separately from the Fgfr2þ/þ

mice for all ages. This contrasts withbone volumes that show inconsistentdifferences between Fgfr2þ/P253R andFgfr2þ/þ across the ages (Figs. 3, 4).

DISCUSSION

Bone Volume and Relative

Density Quantification

We have delineated a method thattakes advantage of the properties ofmCT to provide novel measurementsof bone growth and maturation acrossthe earliest periods of craniofacialossification, filling a void left by moretraditional approaches. mCT providesdirect 3D representations of ossifiedvolumes, allowing quantification oftheir shape, size, relative density, andspatial association. The variation inmeasures noted between specimens ofthe same genotype is typical ofquantitative analyses of unscaledphenotypes in mouse models (e.g.,Hill et al., 2007; Boughner et al.,2008).

Measurements of bone volume anddensity based on phantom-calibratedmCT images are likely to be lower thanvalues produced by other methods,including whole-mount staining andhistology. This is due to comparativelylower spatial resolution of mCT andbecause mineralized tissue is identifiedin mCT images rather than initial colla-gen or calcium deposits or expression ofother early bone markers visualizedwith some staining techniques.

Normal Craniofacial Bone

Maturation and Growth

Our estimates of “typical” bone den-sity, based on unaffected littermates,vary considerably within litters andmay be specific to the C57BL/6 back-ground (Beamer et al., 1996, 2001).Improvements in identifying the rela-tive developmental age of the embry-onic craniofacial complex, as done forlimb buds (Boehm et al., 2011), mayenable more precise age-phenotypeassociations.

Ontogenetic changes revealedin mean relative density curves ofFgfr2þ/þ mice serve as a proxy for

Dev

elop

men

tal D

ynam

ics

Fig. 2. Analysis of bones identified from the right lateral view of mCT images of representativespecimens at embryonic day (E) 15.5, E16.5, E17.5, and postnatal day (P) 0 for both genotypes.Bones being analyzed are brightly colored, while the rest of the skull is transparent gray. Thereare some noticeable differences in individual bone shape and extent between the genotypes ateach age. Bones are identified by name and abbreviation on the Fgfr2þ/þ P0 specimen. The ros-tral aspect of the skulls is to the right and the dorsal aspect is toward the top of the page. Imagesare not to scale. Higher resolution images displaying bone extent for all ages from right lateraland ventral views are available at www.getahead.psu.edu/ApertDenVol for interested readers.

CRANIOFACIAL BONE VOLUME AND DENSITY 5

normative bone maturation trajecto-ries. Qualitative comparisons of thesecurves revealed three patterns of mat-uration. All cranial bones in group 1(Fig. 5A,B) ossify endochondrally anddisplay an increase in relative bonedensity for each age interval from ini-tial ossification until P2, as well as apeak in their curves at later ages. Sig-nificant age coefficients from prenatal,but not necessarily postnatal, multi-variate regressions of early ossifyingmembers of group 1 match this inter-pretation. Ossification within a carti-lage model appears to promote a quickincrease in relative bone density afterinitial ossification, although the paceof density increase may slow afterbirth.

Group 2 and 3 bones maintain a sta-ble level of relative bone density evenas their volumes increase during theearliest ages of ossification (Figs. 3, 4).The intramembranous bones of group2 (Fig. 5C–E) sustain weakly convexcurves between E16.5 and P0 or P2,which represent moderate relativebone density without much increase.Group 3 bones (Fig. 5F–H), whichtend to ossify in the more peripheralregions of the skull, retain a steep con-vex curve of relatively low densityuntil an increase between P0 and P2.PetT is the exception, having a morecentral location in the cranial base butretaining low density prenatally.

Fundamental differences betweenintramembranous and endochondral

osteogenesis may account for increasedrelative density of most endochondralbones during the embryonic period andmoderate or low relative density ofintramembranous bones before post-natal density increases. Cartilage mod-els of endochondral bones roughlyreflect the shapes of associated adultbones (Eames and Schneider, 2008).Intramembranous ossification requiressustained proliferation of precursor cellswithin mesenchymal condensations thatdo not resemble their ossified versionsuntil after ossification has alreadybegun within the condensations (Lana-Elola et al., 2007; Yoshida et al., 2008).Perhaps this early proliferation ofmesenchyme precursors occurs at theexpense of higher densities of differen-tiating osteoblasts within intramem-branous bones during the earlieststages of ossification.

The PetT may represent an impor-tant exception to the idea that endo-chondral bones mature more quicklyduring initial ossification. The PetTexists primarily as a relatively smallossified annulus for much of theembryonic period until a swift andsubstantial increase in PetT volumeoccurs between P0 and P2 (Table 2).In this way, it may be similar to SquOand PSph, which do not ossify untillater in development, but mature veryquickly once initiated. If the annulusis disregarded from our measurementof the PetT, because its developmentalorigin is distinct from the rest of thebone (Kardong, 2012), it is possiblethat the curves of PetT would fall ingroup 1, particularly if measurementsafter P2 were available.

Effect of the P253R Mutation

on Bone Growth

We reject our hypothesis that allindividual craniofacial bones ofFgfr2þ/P253R mice display reducedbone volume and relative bone min-eral density from initial ossificationinto the postnatal period. Instead,certain bones are significantlyreduced while others are not. In addi-tion, the influence of the P253R muta-tion on bone volume changes acrossthe prenatal period and then againpostnatally, while the influence ofgenotype on relative density curves ofeach bone is stable across ages. Thesedifferences in volume and relative

Dev

elop

men

tal D

ynam

ics

Fig. 3. Boxplots of bone volumes by age, comparing the bone volume distributions for eachbone between both genotypes at (A) embryonic day (E) 15.5, (B) E16.5, and (C) E17.5. Whiteboxes represent Fgfr2þ/þ and grey boxes represent Fgfr2þ/P253R. Dots represent outlier valuesthat are more than 1.5 times the interquartile range from the box. Bone abbreviations aredefined in Fig 2.

6 PERCIVAL ET AL.

density suggest more nuanced hypoth-eses about the bases of craniofacialdysmorphology in these mice and inhumans with Apert syndrome.

Below, we discuss the implicationsof our results and compare them withthe results of other studies. Differen-ces in background strain may accountfor some of the differences betweenthe results of our study and that ofothers. Direct comparisons betweenour whole bone measures and thosestemming from landmark based mor-phometrics or cranial suture basedhistology may not always be appropri-ate. We attempt to be explicit aboutdifferences in the data on which thecited results are based throughout thediscussion.

Prenatal Effect

Most cranial bones of Fgfr2þ/P253R micedisplay lower mean volumes at E15.5relative to unaffected littermates (notstatistically significant), while severaldisplay significantly lower volumes atE16.5. These same bones appearreduced in ossified extent within mCTimages at these ages (Fig. 2). These

reductions in whole bone volumeappear to disagree with observationsmade at the coronal suture of differentApert syndrome model mice, whereincreased osteoid deposition at approx-imating bone fronts are observed byE15.5 or E16.5 (Holmes et al., 2009).The potential reasons for these conflict-ing results include: differencesbetween the two Apert mouse models;acute and specific dysregulation at thecoronal suture; and/or the fact thatmCT images display relatively wellossified bone as opposed to the osteoidand collagen deposition typicallyvisualized by cell biology techniques.

Similar volumes across genotypesfor most bones at E17.5 suggesta period of volume catch-up forFgfr2þ/P253R mice before E17.5.Continued increase leads to largermean volumes for some bones in P0Fgfr2þ/P253R mice, including Max.Although the effect of the P253R muta-tion on individual bone volume variesacross prenatal ages, facial bones andPetT of Fgfr2þ/P253R mice that displaysignificantly lower volume at E16.5also display lower relative densityacross the prenatal period. This sug-

gests that the P253R mutation retardsbone maturation starting at initialossification and that relative densityremains reduced even as bone volumesapproach or exceed normal levels.

Previous histological observations ofthe influence of the P253R mutationon local cell activity near cranialsutures may provide a partial expla-nation for the catch-up in bone volumeobserved for some bones by E17.5 andP0. Proliferation at the osteogenicfronts of the coronal sutures ofFgfr21/S252W mice was reduced signif-icantly at E16.5, although it did notdiffer at E13.5 or E15.5 (Holmes et al.,2009). Proliferation was also reducedat E17.5 at the coronal sutures ofFgfr2þ/P253R mice, but not at E19 orP5 (Wang et al., 2010). Concurrentwith reduced proliferation, increasedosteoblast differentiation was noted atthe coronal sutures in Fgfr2þ/P253R

mice between E16.5 and E17.5 (Yinet al., 2008) and between E17.5 andP5 (Wang et al., 2010). Decreased pro-liferation of preosteoblastic cells andincreased differentiation associatedwith the P253R mutation may coin-cide with known shifts in FGFRexpression during osteoblast matura-tion (Morriss-Kay and Wilkie, 2005)and/or the switch from an expandingpopulation of mesenchymal progeni-tors to the more typically describedossification front moving outwardfrom intramembranous bone ossifica-tion centers (Lana-Elola et al., 2007;Yoshida et al., 2008).

Dev

elop

men

tal D

ynam

ics

Fig. 4. Boxplots of bone volumes by age, comparing the bone volume distributions for eachbone between both genotypes at (A) postnatal day (P) 0 and (B) P2. White boxes representFgfr2þ/þ and gray boxes represent Fgfr2þ/P253R. Dots represent outlier values that are more than1.5 times the interquartile range from the box. Bone abbreviations are defined in Fig 2.

TABLE 3. Groups of Bones

Defined by the Pattern of Change

of Their Relative Density Curves

Between E15.5 and P2 for

Fgfr21/1 Micea

Group 1 – Group 2 – Group 3 –

Continuous Moderate Low

Increase Density Density

BasO Fro IParLatO Pal ParSphB Max PetTSquO Man NasPSph SquT PMax

SphA

aBone abbreviations are defined inFigure 2.

CRANIOFACIAL BONE VOLUME AND DENSITY 7

Landmark-based morphometricanalysis indicates that Fgfr2þ/P253R

mice can be differentiated fromFgfr2þ/þ at P0 by brachycephaly, mid-facial hypoplasia, an anteriorly dis-placed cranial base, as well assynostosis of the coronal, maxillary-zygomatic, and premax-maxillarysutures (Wang et al., 2010; Mart�ınez-Abad�ıas et al., 2010; Hill et al., 2013).The bones for which volume is moststrongly influenced at E15.5–E16.5and for which relative bone density isinfluenced across the prenatal periodinclude bones of the face. Landmark

based measures suggest shorter lin-ear scale of facial bones in Fgfr2þ/P253R

mice at P0 (Wang et al., 2010;Mart�ınez-Abad�ıas et al., 2010; Hillet al., 2013) as well as reduced growthof the palate and face between E17.5and P0 (Motch et al., 2012). However,we find that facial bones display simi-lar (or higher) volumes at E17.5 andP0, indicating a lack of correlationbetween landmark based and volumebased measures of bone scale. Overall,Fgfr2þ/P253R mice display thicker orsquatter facial bones with reduceddensity at E17.5 and P0.

If a shift in the balance betweenosteoblast lineage proliferation anddifferentiation occurs as describedabove across all facial bones, this mayexplain the bone volume catch-upnoted in our prenatal results. Anincrease in osteoblast differentiationat the expense of proliferation (andperhaps migration) may be directlyresponsible for the increased rate ofossification and the subsequentthicker morphology of many bones ofthe Fgfr2þ/P253R mice by E17.5 andP0. It is plausible that the quickincrease in volume and thickening ofmaxillae and premaxillae, which startout smaller than normal at E16.5,may contribute to the overall mor-phology of midfacial hypoplasia. Wealso suggest that the thickening ofthese bones without adequate prolif-eration and subsequent migration ofosteoblast precursors outward maycause the edges of the premaxilla andthe maxilla to approach and touch,contributing to the premature fusionof the premax-maxillary suture.

Our results of significant early dif-ferences in total bone volume and rel-ative density, followed by evidence ofcatch-up ossification before birth,highlight the importance of looking atthe effect of the P253R mutation onosteoblast lineage activity acrosswhole bones, rather than only focus-ing on the suture. In combinationwith landmark based linear measure-ments, bone volume and relative den-sity provide a more complete pictureof the effect of a mutation on grossskeletal phenotype and serve as abasis for hypotheses on changes incellular mechanisms that underliepathology.

Postnatal Effect

A second shift in the influence of theP253R mutation on bone volumegrowth occurs postnatally. BetweenP0 and P2, Fgfr2þ/P253R bone vol-umes increase slower than Fgfr2þ/þ

volumes, leaving several bones thatwere equal in size at P0, signifi-cantly smaller at P2. The postnatalreduction in bone volume growth inthis sample is similar to the reduc-tion noted in the Fgfr21/Y394C mousemodel of Beare-Stevenson craniosyn-ostosis syndrome between P0 and P8(Percival et al., 2012). Many bones

Dev

elop

men

tal D

ynam

ics

Fig. 5. A–H: Mean relative bone density curves for a subset of bones of Fgfr2þ/þ specimens ateach age, representing three patterns of bone maturation. The axis scales are standardizedacross plots to allow for easy comparison, but forcing high standardized volume values off theplots. Bone density is measured as mg/mm3 partial density of hydroxyapatite. Histogram valuesare standardized by total bone volume. Images illustrating relative density curves for all bonesare available at www.getahead.psu.edu/ApertDenVol for interested readers.

8 PERCIVAL ET AL.

Dev

elop

men

tal D

ynam

ics

Fig. 6. The functional intercept, age, and genotype coefficients and 95% confidence intervals from prenatal multivariate regressions of relativebone density curves for four bones. The genotype coefficient illustrates how the Fgfr2þ/P253R mice differ from unaffected littermates, independentof age. The y-axis differs for all plots. A: SquT displays a significant genotype effect between E15.5 and P0, which is opposite from the effect ofincreased age. The (B) Fro and (C) LatO do not display a significant genotype effect prenatally, but they do display a significant age effect over awide range of density values. D: Man is among a few bones that displays a weakly significant age effect during the prenatal period. However, itdisplays a strong genotype effect. Bone abbreviations defined in Figure 2.

Fig. 7. The functional intercept, age, and genotype coefficients and 95% confidence intervals from postnatal multivariate regressions of relative bonedensity curves for four bones. The genotype coefficient illustrates how the Fgfr2þ/P253R mice differ from unaffected littermates, independent of age.The y-axis differs for all plots. The (A) SquT and (B) Fro display significant genotype and age effects between P0 and P2, although the genotype coeffi-cient is not significant over a wide range of bone density values. C: LatO does not display a significant genotype or age effect postnatally. D: Man dis-plays a weakly significant effect of genotype and a strong effect of age during the postnatal period. Bone abbreviations defined in Figure 2.

CRANIOFACIAL BONE VOLUME AND DENSITY 9

that displayed lower relative bonedensity prenatally retain lower rela-tive density postnatally, although itis unclear whether this is based oncontinued differences in bone matu-ration between the genotypes or is acarryover from the prenatal period.

The smaller volumes of manyFgfr2þ/P253R bones at P2 appear tomatch the great reduction in skullsize noted between P0 and P2 in thismodel by previous morphometricanalyses (Hill et al., 2013), includingreduced size of the palate and rostralcranial base. In addition, the volumesof several neurocranial bones arereduced in P2 Fgfr2þ/P253R mice, sug-gesting that the previously notedincrease in relative neurocranialheight and width (Hill et al., 2013) isbased on relatively wider fontanellesand sutures rather than increasedossification of the neurocranial ele-ments. Such increased width is likelyassociated with wide midline calvarialdefects, such as that noted in humaninfants with Apert syndrome (Cohenand Kreiborg, 1998).

Reduced postnatal Fgfr2þ/P253R bonevolume would not be expected ifincreased osteoblast differentiationand bone expansion continued until P5across all bones of the skull, asobserved local to the leading ossifica-tion edges of the frontal and parietalbones making up the coronal suture(Wang et al., 2010). In fact, Par, whichmakes up one half of the coronalsuture border, displays significantlylower volume at P2. These contrastingresults illustrate differences in locallyfocused histological measures and esti-mates derived from whole bones.Assuming that a decrease in preosteo-blastic cell proliferation and increasein differentiation occurs betweenE16.5 and E17.5, it is possible thatpostnatal populations of presumptiveosteoblasts will be smaller withinFgfr2þ/P253R bones. A small progenitorcell population may limit the speed ofbone growth and density increase viareduced potential for osteoblast differ-entiation and function. This hypothet-ical explanation for lower postnatalbone volumes for Fgfr2þ/P253R micehighlights the need for direct studiesof bone cell activity in regions otherthan the coronal sutures of Fgfr2mutant mice if we are to understandthe global effects of these mutations

on craniofacial development from theearliest embryonic periods onward.Given the increasing evidence thatfacial dysmorphology (Mart�ınez-Abad�ıas et al., 2010, 2013) and braindysmorphology (Aldridge et al., 2010,Hill et al., 2013) occur as primaryeffects of known FGFR mutations,rather than secondary to calvarialsuture fusion; studying the effect ofApert and other FGFR mutations ondevelopment across the head is a nec-essary step toward developing pre-ventative treatments.

Summary

Using a refined version of our methodto quantify bone volume and relativebone density (Percival et al., 2012),we quantify the “normal” pattern ofearly bone growth and maturation ofindividual mouse craniofacial bonesand the influence of the P253R muta-tion on this pattern. Our results sug-gest that endochondral (group 1) andintramembranous cranial bones(groups 2 and 3) have fundamentallydifferent patterns of bone maturationduring initial osteogenesis. Endo-chondral bone density increasesalmost immediately following firstossification, while intramembranousbones retain a moderate or low levelof bone density until the postnatalperiod, even as they significantlyincrease in volume. Although testingit is beyond the scope of this research,we hypothesize that the lower ‘normal’levels of relative density increase inintramembranous bones stems fromincreased osteoblast lineage cell prolif-eration at the expense of differentiationduring the earliest period of osteogene-sis. In addition, further work on bonevolume and relative density during thepostnatal period will be required todetermine whether the relative densityof bones from all three groups becomessimilar later in development.

Relative to our findings in unaf-fected littermates, mice carrying theFgfr2 P253R mutation show initiallyreduced volumes and relative densityof many craniofacial bones during theearliest stages of ossification, includ-ing bones of the face that displaysome of the most serious dysmorphol-ogy associated with Apert syndrome.A change in the influence of thismutation on bone volume occurs

between E16.5 and E17.5, whichcould be associated with a previouslynoted decrease in osteoblast prolifera-tion and increased differentiation atthe coronal suture (Yin et al., 2008;Holmes et al., 2009; Wang et al.,2010). After bone volumes similar tothose measured for unaffected litter-mates are noted across the skull atE17.5 and P0 in Fgfr2þ/P253R mice,the volumes and relative density ofvault bones decrease between P0 andP2, as the overall relative linear scaleof the Fgfr2þ/P253R skull is reducedand cranial dysmorphology increases.Early differences in the volume andrelative density of facial bones associ-ated with later suture fusion and mid-facial hypoplasia suggest that thedevelopmental changes underlying atleast some Apert syndrome associatedcraniofacial dysmorphology occur dur-ing the earliest period of osteogenesis.

We found that facial bones previ-ously shown by landmark basedmeasures to be smaller in perinatalFgfr2þ/P253R mice, had bone volumessimilar to unaffected littermates atbirth. This demonstrates that vol-umes of ossifying bones do not neces-sarily covary with landmark basedmeasures of bone extent. Instead ofrelying on one type of phenotypicmeasure, the combination of comple-mentary measures from landmark-based, volume, and relative densityanalyses provides a more completepicture of gross skeletal phenotype ofindividual craniofacial bones acrossthe developing skull and illustratesthe broader changes in individualbone development and maturationthat are associated with craniofacialdysmorphology in Apert syndrome.We have suggested several testablehypotheses throughout our discussionabout the cellular bases of noted dif-ferences in volume and density. Fur-ther work on the cellular activity inskull bones located distant to cranialvault and facial sutures are requiredto definitively connect the effects ofthis and other Fgfr mutations at thecellular level with noted differencesin gross craniofacial phenotype.

ACKNOWLEDGEMENTSMany thanks for the technical exper-tise and care by Tim Ryan and TimStecko at PSU CQI in producing the

Dev

elop

men

tal D

ynam

ics

10 PERCIVAL ET AL.

mCT images used in this study. TaliaPankratz completed valuable data col-lection on the P0 and P2 mice for thisstudy. Thanks to Xueyan Zhou andJunghye Kim who organized and car-ried out a large part of the mousebreeding at Icahn School of Medicineat Mount Sinai and PSU, respectively.This work was supported in part bygrants from the National ScienceFoundation to C.J.P. as well as toJ.T.R. and collaborators and from theNational Institute of Dental and Cra-niofacial Research and the AmericanRecovery and Reinvestment act toJ.T.R.

REFERENCES

Aldridge K, Hill CA, Austin JR, PercivalC, Mart�ınez-Abad�ıas N, Neuberger T,Wang Y, Jabs EW, Richtsmeier JT.2010. Brain phenotypes in two FGFR2mouse models for Apert syndrome. DevDyn 239:987–997.

Beamer WG, Donahue LR, Rosen CJ,Baylink DJ. 1996. Genetic variability inadult bone density among inbredstrains of mice. Bone 18:397–403.

Beamer WG, Shultz KL, Donahue LR,Churchill GA, Sen S, Wergedal JR,Baylink DJ, Rosen CJ. 2001. Quantita-tive trait loci for femoral and lumbarvertebral bone mineral density inC57BL/6J and C3H/HeJ inbred strainsof mice. J Bone Miner Res 16:1195–1206.

Boehm B, Rautschka M, Quintana L,Raspopovic J, Jan Z, Sharpe J. 2011. Alandmark-free morphometric stagingsystem for the mouse limb bud. Devel-opment 138:1227–1234.

Boughner JC, Wat S, Diewert VM, YoungNM, Browder LW, Hallgr�ımsson B.2008. Short-faced mice and develop-mental interactions between the brainand the face. J Anat 213:646–662.

Bush JO, Jiang R. 2012. Palatogenesis:morphogenetic and molecular mecha-nisms of secondary palate development.Development 139:231–243.

Cohen MM Jr, Kreiborg S. 1998. Sutureformation, premature sutural fusion,and suture default zones in Apert syn-drome. Am J Med Genet Part A 62:339–344.

Cohen MM Jr, Maclean RE. 2000. Cranio-synostosis: diagnosis, evaluation andmanagement. 2nd ed. New York: OxfordUniversity Press.

Dixon MJ, Marazita ML, Beaty TH,Murray JC. 2011. Cleft lip and palate:understanding genetic and environmen-tal influences. Nat Rev Genet 12:167–178.

Du X, Weng T, Sun Q, Su N, Chen Z, QiH, Jin M, Yin L, He Q, Chen L. 2010.Dynamic morphological changes in theskulls of mice mimicking human Apertsyndrome resulting from gain-of-function mutation of FGFR2 (P253R). JAnat 217:97–105.

Eames BF, Schneider RA. 2008. The gene-sis of cartilage size and shape duringdevelopment and evolution. Develop-ment 135:3947–3958.

Gritli-Linde A. 2008. The etiopathogene-sis of cleft lip and cleft palate: useful-ness and caveats of mouse models. CurrTop Dev Biol 84:37–138.

Hallgr�ımsson B, Boughner JC, TurinskyA, Parsons TE, Logan C, Sensen CW.2009. Geometric morphometrics and thestudy of development. In: Sensen CW,Hallgr�ımsson B, editors. Advancedimaging in biology and medicine. NewYork: Springer. p 319–336.

Hill CA, Mart�ınez-Abad�ıas N, Motch SM,Austin JR, Wang Y, Jabs EW,Richtsmeier JT, Aldridge K. 2013. Post-natal brain and skull growth in anApert syndrome mouse model. Am JMed Genet Part A 161:745–757.

Hill CA, Reeves RH, Richtsmeier JT.2007. Effects of aneuploidy on skullgrowth in a mouse model of Down syn-drome. J Anat 210:394–405.

Holmes G, Rothschild G, Roy UB, DengCX, Mansukhani A, Basilico C. 2009.Early onset of craniosynostosis in anApert mouse model reveals critical fea-tures of this pathology. Dev Biol 328:273–284.

Kardong KV. 2012. Vertebrates: compara-tive anatomy, function, evolution. 6thed. New York, NY: McGraw-Hill.

Lana-Elola E, Rice R, Grigoriadis AE,Rice DPC. 2007. Cell fate specificationduring calvarial bone and suture devel-opment. Dev Biol 311:335–346.

Mart�ınez-Abad�ıas N, Holmes G, PankratzT, Wang Y, Zhou X, Jabs EW, RichtsmeierJT. 2013. From shape to cells: mouse mod-els reveal mechanisms altering palatedevelopment in Apert syndrome. DisModel Mech 6:768–779.

Mart�ınez-Abad�ıas N, Percival C, AldridgeK, Hill C, Ryan T, Sirivunnabood S,Wang Y, Jabs E, Richtsmeier J. 2010.Beyond the closed suture in Apertmouse models: evidence of primaryeffects of FGFR2 signaling on facialshape at P0. Dev Dyn 239:3058–3071.

Morriss-Kay GM, Wilkie AOM. 2005.Growth of the normal skull vault and itsalteration in craniosynostosis: insightsfrom human genetics and experimentalstudies. J Anat 207:637–653.

Motch SM, Pankratz T, Percival CJ, JabsEW, Richtsmeier JT. 2012. Developmentof the skull in Fgfr2þ/P253R andFgfr2þ/S252W mouse models for Apert

syndrome in late embryogenesis. Pre-sentation at the 2012 Meeting of theSociety for Craniofacial Genetics andDevelopmental Biology, San Francisco,California.

Olson LE, Richtsmeier JT, Leszl J, ReevesRH. 2004. A chromosome 21 criticalregion does not cause specific Down syn-drome phenotypes. Science 306:687–690.

Passos-Bueno MR, Serti�e AL, Jehee FS,Fanganiello R, Yeh E. 2008. Genetics ofcraniosynostosis: genes, syndromes,mutations and genotype-phenotype cor-relations. In: Rice DP, editor. Craniofa-cial sutures: development, disease andtreatment. Vol. 12 Basel: Karger. p107–143.

Percival CJ, Wang Y, Zhou X, Jabs EW,Richtsmeier JT. 2012. The effect of aBeare-Stevenson syndrome Fgfr2Y394C mutation on early craniofacialbone volume and relative bone mineraldensity in mice. J Anat 221:434–442.

Percival CJ. 2013. The influence of angio-genesis on craniofacial developmentand evolution. Available from: https://etda.libraries.psu.edu/paper/16838/.

Perlyn CA, DeLeon VB, Babbs C, GovierD, Burell L, Darvann T, Kreiborg S,Morriss-Kay G. 2006. The craniofacialphenotype of the Crouzon mouse:analysis of a model for syndromic cra-niosynostosis using three-dimensionalMicroCT. Cleft Palate Craniofac J 43:740–748.

Ramsay JO, Hooker G, Graves S. 2009.Functional data analysis with R andMATLAB. New York, NY: Springer.

Schmidt EJ, Parsons TE, Jamniczky HA,Gitelman J, Trpkov C, Boughner JC,Logan CC, Sensen CW, Hallgr�ımsson B.2010. Micro-computed tomography-basedphenotypic approaches in embryology:procedural artifacts on assessments ofembryonic craniofacial growth and devel-opment. BMC Dev Biol 10:1–14.

Wang Y, Sun M, Uhlhorn VL, Zhou X,Peter I, Martinez-Abadias N, HillCA, Percival C, Richtsmeier JT,Huso DL, Jabs EW. 2010. Activationof p38 MAPK pathway in the skullabnormalities of Apert syndromeFgfr2þ/P253R mice. BMC Dev Biol10:22.

Yin L, Du X, Li C, Xu X, Chen Z, Su N,Zhao L, Qi H, Li F, Xue J. 2008. APro253Arg mutation in fibroblast growthfactor receptor 2 (Fgfr2) causes skeletonmalformation mimicking human Apertsyndrome by affecting both chondrogen-esis and osteogenesis. Bone 42:631–643.

Yoshida T, Vivatbutsiri P, Morriss-Kay G,Saga Y, Iseki S. 2008. Cell lineage inmammalian craniofacial mesenchyme.Mech Dev 125:797–808.

Dev

elop

men

tal D

ynam

ics

CRANIOFACIAL BONE VOLUME AND DENSITY 11