Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
1
Quantification of gene expression patterns to reveal the origins of abnormal morphogenesis 1
2
N. Martínez‐Abadías1,2,3,*,†, R. Mateu3, J. Sastre4, Motch Perrine S5, Yoon M5, A. Robert‐Moreno1,2,†, J. 3
Swoger1,2,†, L. Russo1,2, Kawasaki K5, J. Richtsmeier5, J. Sharpe1,2,6,† 4
5
1 Centre for Genomic Regulation (CRG), The Barcelona Institute for Science and Technology, Dr. Aiguader 6
88, 08003 Barcelona, Spain 7
2 Universitat Pompeu Fabra (UPF), Barcelona, Spain 8
3 Universitat de Barcelona, Barcelona, Spain 9
4 Universitat de les Illes Balears (UIB), Palma de Mallorca, Spain 10
5 Pennsylvania State University, University Park (PA), USA 11
6 Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain 12
* Corresponding author: [email protected]. Dr. Aiguader, 88, ‐1, 08003, Barcelona (Spain). 13
†Current address: European Molecular Biology Lab (EMBL), Barcelona, Spain 14
15
Abstract 16
The earliest developmental origins of dysmorphologies are poorly understood in many congenital 17
diseases. They often remain elusive because the first signs of genetic misregulation may initiate as 18
subtle changes in gene expression, which can be obscured later in development due to secondary 19
phenotypic effects. We here develop a method to trace back the origins of phenotypic abnormalities by 20
accurately quantifying the 3D spatial distribution of gene expression domains in developing organs. By 21
applying geometric morphometrics to 3D gene expression data obtained by Optical Projection 22
Tomography, our approach is sensitive enough to find regulatory abnormalities never previously 23
detected. We identified subtle but significant differences in gene expression of a downstream target of 24
the Fgfr2 mutation associated with Apert syndrome. Challenging previous reports, we demonstrate that 25
Apert syndrome mouse models can further our understanding of limb defects in the human condition. 26
Our method can be applied to other organ systems and models to investigate the etiology of 27
malformations. 28
29
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
2
Keywords: Developmental defects, Optical Projection Tomography (OPT), Whole‐Mount‐in‐Situ 30
Hybridization (WISH), Geometric Morphometrics (GM), quantitative shape analysis, limb development, 31
Apert syndrome. 32
33
34
Morphogenesis is guided by dynamic spatio‐temporal regulation of gene expression patterns1,2, the 35
regions within tissues where genes are expressed at specific developmental times. Critical changes in 36
the space, time or intensity of gene expression patterns can result in organ malformation, with reduced 37
or even loss of function. These errors of morphogenesis, which occur in approximately 3% of live births3, 38
are induced by environmental and/or genetic insults that alter the normal process of development. The 39
common approach to reveal the origin of these alterations is to look for phenotypic abnormalities in 40
animal models and visually assess the overall patterns of gene expression. However, this qualitative 41
approach has not been able to identify the earliest signs of dysmorphogenesis in many diseases because 42
the first changes in the gene expression patterns may be subtle, limited in time, and later abnormalities 43
may obscure the original genetic cause. To reveal the primary etiology of congenital malformations, 44
more rigorous methods are needed to quantify the 3D phenotypes of gene expression patterns. A useful 45
tool to trace back development should be able to perform a quantitative statistical comparison of 46
normal and disease‐altered embryogenesis and detect the earliest signs of genetic misregulation leading 47
to organ malformation. 48
Developing organs are shaped by regulating gene activity in space and time1. Cascades of gene 49
regulatory networks provide the detailed instructions to organize cell behaviour and orchestrate tissue 50
growth and differentiation. Gene expression patterns can be readily mapped within tissues in a true 51
three‐dimensional framework combining Whole‐Mount‐in‐Situ Hybridization (WISH)4 with Optical 52
Projection Tomography (OPT)5. Whereas WISH is a standard molecular technique for detecting the 53
expression of a specific gene using a labelled complementary RNA probe6,7, OPT is a mesoscopic imaging 54
procedure that can produce high resolution 3D reconstructions of whole developing embryos processed 55
by WISH8,9. These technologies represented breakthroughs in developmental biology and have provided 56
invaluable qualitative insights into gene function and development8. However, methods for quantifying 57
the 3D gene expression distributions in a systematic, objective manner are still lacking. 58
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
3
Expanding the potential of OPT from qualitative to quantitative analysis of gene expression patterns is 59
challenging. The expression of genes is characterized by highly dynamic patterns, with fast rates of 60
change over time and fuzzy boundaries that usually do not correspond to well‐defined anatomical 61
structures but to tissue regions where cells are dynamically up‐ and down‐regulating genes. The 62
quantification of gene expression patterns has been rarely done10–16. We propose to quantify the shape 63
of developing organs in association with their underlying gene expression patterns applying Geometric 64
Morphometrics (GM), a set of statistical tools for measuring and comparing shapes with increased 65
precision and efficiency17–22. This approach provides the ability to replace qualitative observations with 66
quantification of subtle yet significant biological differences that underlie the processes of 67
morphogenesis altered by disease. 68
Here we illustrate how our method can reveal the genetic origin of developmental defects by 69
investigating limb malformations in Apert syndrome [OMIM 101200]. Apert syndrome is a rare 70
congenital disease, with disease prevalence of 15–16 per million live births, that is characterized by 71
cranial, neural, limb, and visceral malformations23. Over 99% of Apert cases are associated with one of 72
two missense mutations, S252W and P253R, on Fibroblast Growth Factor Receptor 2 (FGFR2)24,25. The 73
mutations occur on neighbouring amino acids on the linker region between the second and third 74
extracellular immunoglobulin domains of FGFR2, and alter the ligand‐binding specificity of the 75
receptors26,27. Thus, the FGF receptors are activated inappropriately, altering the entire FGF/FGFR 76
signalling pathway and causing dysmorphologies of different organs and systems28. Apert syndrome 77
shares craniofacial dysmorphologies with other craniosynostosis syndromes but is differentiated on the 78
basis of limb defects of fore‐ and hindlimb digits. The craniofacial dysmorphology of Apert syndrome 79
(i.e. premature closure of cranial sutures and patent anterior fontanelle associated with atypical head 80
shape, midfacial retrusion and palatal defects)23 has been intensively investigated, especially because 81
mouse models show cranial phenotypes that correspond with the human condition29–33. However, the 82
associated limb defects are less well studied, in part because mouse models for Apert syndrome present 83
only subtle limb anomalies34–36, even in those carrying the FGFR2 P253R mutation34 that is associated 84
with the more severe limb malformations in Apert syndrome37,38. 85
Here we present precise phenotyping of limbs of newborn and embryonic specimens of the Fgfr2+/P253R 86
Apert syndrome mouse model and reveal significant differences that can be traced to as early as one 87
day after the initiation of limb development. To explore the molecular basis of these initial signs of limb 88
dysmorphology we applied our method combining OPT and GM to assess the expression pattern of a 89
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
4
direct target of Fgf signalling, Dusp6, a relevant gene for limb morphogenesis39,40. Our quantitative 90
analyses demonstrate that the Apert syndrome Fgfr2 P253R mutation induces changes in the expression 91
pattern of Dusp6 and that these genetic changes are associated with significant phenotypic alterations. 92
These results provide insight into the origins of limb malformations in Apert syndrome. 93
94
Results 95
Apert syndrome mice present limb malformations at birth. Previous studies reported that Apert 96
syndrome mice do not show obvious abnormalities of the limb35, and thus focused their molecular 97
analyses on the skull34. Histopathological analyses in Apert syndrome mice only revealed overall limb 98
shortening due to abnormal osteogenic differentiation, but no signs of limb disproportion or 99
syndactyly34,36. As a further test of whether or not the Fgfr2 P253R mutation affects limb development in 100
mice, we first performed an extensive quantitative analysis of the size and shape of individual forelimb 101
bones using data from high resolution microCT images of newborn (P0) mutant and unaffected 102
littermates (Fig. 1a‐f). Our results revealed many more significant differences between P0 unaffected 103
and Fgfr2+/P253R mutant littermates than previously reported. We found that the humerus, radius and 104
ulna were statistically significantly shorter in length but had increased bone volumes in Fgfr2+/P253R 105
mutant mice in comparison to unaffected littermates (Fig. 1g and Table SI_1). More localized size 106
differences were detected in the bones derived from the autopod that give rise to the hands. The distal 107
phalanx of digit I, the proximal phalanx of digit V, and metacarpals II, III and IV were significantly longer 108
in Fgfr2+/P253R mutant mice (Fig. 1g and Table SI_1). In contrast, the proximal phalanx of digit III was 109
shorter and lower in bone volume in Fgfr2+/P253R Apert syndrome mice relative to unaffected littermates 110
(Fig. 1g and Table SI_1). The scapula and the clavicle, the bones that form the shoulder girdle, were also 111
significantly affected: the scapula was longer, the clavicle was shorter and both bones showed increased 112
bone volumes in Fgfr2+/P253R Apert syndrome mice (Fig. 1g and Table SI_1) compared to unaffected 113
littermates. 114
The PCA based on the shape of the humerus did not show marked shape differences between 115
unaffected and Fgfr2+/P253R Apert syndrome mice (Fig. 1h). However, the PCA of the scapula indicated a 116
clear morphological differentiation between groups (Fig. 1i). The scapula of Fgfr2+/P253R Apert syndrome 117
mice presented a more robust phenotype, with wider and longer scapulae in comparison to their 118
unaffected littermates. 119
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
5
Overall, these size and shape differences demonstrate that Fgfr2+/P253R Apert syndrome mice present 120
widespread and significant limb dysmorphologies at P0 that were not previously reported and would not 121
have been revealed without quantitative statistical testing. Some defects have a direct correspondence 122
with the human phenotype, such as shoulder anomalies and short humeri24. However, in newborn mice 123
we did not detect any clear sign of syndactyly, which is the most prominent limb defect in people with 124
Apert syndrome41,42. Since the forelimb of mice is not yet completely ossified at P0 and Fgfr2+/P253R 125
mutant littermates die shortly after birth, we could not assess whether other limb abnormalities appear 126
later in development. 127
128
A quantitative morphometric method to assess embryonic gene expression patterns. To determine the 129
developmental basis of the limb anomalies quantified in newborn mice, we developed a quantitative 130
method to explore early embryonic limb development. First, to visualize the expression pattern of a 131
downstream target of Fgfr2, we obtained OPT scans of Fgfr2+/P253R Apert syndrome mouse embryos 132
analysed with WISH to reveal Dusp6 expression (Fig. 2). Qualitative assessment of the 3D 133
reconstructions showed that Dusp6 was widely expressed throughout the embryo from embryonic day 134
(E) 10.5 to E11.5, with highest intensity in the limbs, the head and the spinal cord (Fig. 2). Dusp6 was 135
also expressed in the heart with moderate intensity. Visually comparing the distribution of the Dusp6 136
gene expression pattern it was possible to distinguish between embryos at E10.5 and E11.5 stages of 137
development. At E10.5, Dusp6 was prominently expressed in the facial prominences and along the 138
entire spinal cord, whereas at E11.5 the expression of Dusp6 was more widespread in the brain and 139
limited to the tail. Focusing on the limbs, the expression of Dusp6 at the two different stages was also 140
readily distinguishable, with Dusp6 expression domains thinning into a more extended domain along the 141
limb outline as the limb buds grow from E10.5 to E11.5 (Fig. 2). However, due to large amount of 142
developmental variation within litters, Fgfr2+/P253R mutant and unaffected littermates were not 143
distinguishable from each other (Fig. 2). 144
Quantitative testing was thus required to more accurately evaluate limb alterations potentially 145
associated with Apert syndrome but undetectable by eye. We developed a method for 3D shape analysis 146
of the limb and associated gene expression pattern of Dusp6 (Fig. 3 and Video SI_1). This protocol 147
enabled us to determine differences in limb size and shape between genotype groups and whether 148
these phenotypic differences are associated with altered gene expression patterns (Fig. 3). Our 149
approach uses GM methods to directly measure the limb anatomy and gene expression domains 150
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
6
segmented from the 3D reconstructions of the embryo OPT scans. As expression of Dusp6 showed a 151
fuzzy spatial gradient, multiple thresholding was used to consistently define a high gene expression 152
pattern (Fig. 3, steps from 1 to 5). After manual and semiautomatic recording of 3D coordinates of 153
landmarks on the surfaces of the limb and the gene expression domains blinded to group allocation (Fig. 154
3, step 6), multivariate statistical analyses were performed to explore shape and size variation and 155
covariation patterns between the limb morphology and the Dusp6 domain (Fig. 3, steps 7 and 8). 156
The first signs of limb dysmorphology in Apert syndrome. Since gene expression patterns are highly 157
dynamic and rapidly change in size, shape and position within a few hours of development15, individual 158
limb buds from Fgfr2+/P253R mutant embryos and their unaffected littermates aged between E10.5 and 159
E11.5 were staged using a fine‐resolution staging system (http://limbstaging.crg.es)43. The staging 160
results showed that the analysed limbs represent a temporal continuum over development, with no 161
significant differences between the staging of unaffected and mutant littermates of the same litter (Fig. 162
SI_1). We partitioned the time span from E10 to E11.5 into four periods, each one approximately 163
representing 12 hours of development (see Table SI_2 and Methods for further details on sample 164
composition). The analysis of the complete dataset that considers hindlimbs and forelimbs from each 165
litter separately is available as Supplementary Information. 166
We first focused on analysing limb dysmorphology, aiming to determine the youngest stage which 167
showed morphological differences between mutant and unaffected limbs. To trace limb development 168
back in time we first analysed embryos from the oldest period (as the differences would be easier to 169
find) and from there proceeded towards the earlier (younger) periods. In this way we should confidently 170
identify the initiation of limb dysmorphogenesis associated with the Fgfr2 P253R mutation determining 171
using geometric morphometric methods. 172
During the “Late” period, we detected that Fgfr2+/P253R mice were already clearly separated from their 173
unaffected littermates in the morphospace defined by the Principal Component Analysis (PCA) (Fig. 4a). 174
Relative to their unaffected littermates, limbs of Fgfr2+/P253R mice presented subtle phenotypic limb 175
differences: limbs were shorter, wider and more robust, with limited development of the wrist (Fig. 4a). 176
Quantitative comparison of limb size showed that the limbs of mutant mice were also significantly 177
smaller (Fig. 5a and Table SI_3). Overall, these results confirmed that the Fgfr2 P253R Apert syndrome 178
mutation has an effect on limb development, altering both the size and shape of the limbs. Most likely, 179
these subtle but significant phenotypic differences would have remained undetected by a qualitative 180
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
7
approach. Our quantitative approach revealed their statistical significance and pointed to the origin of 181
the Apert syndrome limb malformation prior to E11.5, before the “Late” period. 182
At the “Mid late” period, the limbs of Fgfr2+/P253R mutant mice were still distinguishable from the limbs 183
of unaffected mice in the morphospace of the PCA (Fig. 4b). At this period, the limbs of Fgfr2+/P253R mice 184
lacked the antero‐posterior asymmetry and the narrowing of the wrist region more typical of unaffected 185
littermates. Instead, Fgfr2+/P253R mice showed a limb phenotype that was elongated in the proximo‐distal 186
axis and thickened in the dorso‐ventral axis (Fig. 4b), resembling the limb shape of younger unaffected 187
embryos. This shape difference coincided with reduced growth in Fgfr2+/P253R mice, as the limbs of 188
Fgfr2+/P253R mice tended to be smaller than unaffected limbs (Fig. 5a and Table SI_3). Therefore, 189
significant differences between unaffected and mutant limbs still could be detected at the “Mid late” 190
period of development and the origins of limb defects associated with Apert syndrome should be sought 191
earlier in development. 192
The first period where no significant differences could be detected between unaffected and Fgfr2+/P253R 193
Apert syndrome mice was at the “Mid early” period (Fig. 4c). Despite internal variation in limb shape, 194
with Fgfr2+/P253R mice spreading throughout the morphospace and unaffected littermates concentrated 195
on one region, mutant and unaffected littermates completely overlapped. Therefore, limb shape 196
differences could no longer be detected between groups. Limb size differences were not significant 197
either (Fig. 5a and Table SI_3). Therefore, our results suggest that the critical time point of limb 198
dysmorphogenesis associated with Apert syndrome occurred between the “Mid late” and “Mid early” 199
periods, corresponding to the transition period from E10.5 to E11 (Fig. 4b‐c). 200
Finally, no further sign of limb shape dysmorphology was detected at the “Early” period of development 201
(Fig. 4d). At this period there was a great range of developmental variation, with unaffected and 202
Fgfr2+/P253R mutant mice completely overlapping in the morphospace and all limbs displaying similar 203
incipient bud shapes (Fig. 4d). The limbs of Fgfr2+/P253R mice were significantly larger than the limbs of 204
their unaffected littermates (Fig. 5a and Table SI_3), suggesting that at this early time point there is a 205
significant effect of the Fgfr2 P253R mutation on limb size but not on limb shape (Fig. 5a). 206
Fgfr2 Apert syndrome mutation leads to aberrant overexpression of Dusp6 domains. We decided to 207
obtain direct evidence of altered genetic regulation that could explain the observed limb phenotype by 208
analyzing the shape dynamics of Dusp6 expression, a direct target gene of Fgf signaling. As with limb 209
shape, we first examined the gene expression of Dusp6 in the embryos from the latest period. We found 210
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
8
that at the “Late” period, Dusp6 expression was already different between Fgfr2+/P253R mutant mice and 211
unaffected littermates. The differences were significant both in shape (Fig. 4e) and size (Fig. 5b and 212
Table SI_3). In the limbs of unaffected mice, the Dusp6 expression domain appeared as a thin domain 213
underlying the apical ectodermal ridge, whereas in Fgfr2+/P253R mutant mice the shape of the Dusp6 214
domain was expanded in all directions (Fig. 4e). Accordingly, the volume of the Dusp6 expression 215
domain was significantly larger in Apert syndrome mice (Fig. 5b and Table SI_3), even when these mice 216
presented significantly smaller limbs (Fig. 5a). The Dusp6 expression domain thus grew 217
disproportionately in the limbs of Fgfr2+/P253R mutant mice in the latest period of development (Fig. 4e), 218
which is consistent with a reported whole‐body size reduction in Fgfr2+/P253R Apert mice and the over‐219
activation of Fgfr2 signaling by the Apert syndrome mutation27,44. 220
At the “Mid late” period, an expanded Dusp6 expression domain persisted on the dorsal and ventral 221
sides of mutant limbs, but was reduced on the anterior and posterior sides (Fig. 4f). The overall volume 222
of the Dusp6 expression domains remained larger in Fgfr2+/P253R mutant mice, but the difference was no 223
longer statistically significant (Fig. 5b and Table SI_3). 224
At the “Mid early” period, the separation between unaffected and Fgfr2+/P253R mutant mice was 225
maintained in the PCA analysis (Fig. 4g). Unaffected mice showed a Dusp6 expression domain expanded 226
towards the anterior and posterior edges of the expression domain (Fig. 4g). In contrast, Fgfr2+/P253R 227
mutant mice did not show the extension and the posterior asymmetry of the Dusp6 expression domain 228
typical of normal limb development, suggesting a lack of differentiation in the Dusp6 expression of 229
mutant limbs (Fig. 4g). 230
Finally, the “Early” period was the only time point where we did not detect a significant separation 231
between unaffected and Fgfr2+/P253R mice (Fig. 4h). The PCA showed variation in the expression domains 232
of Dusp6, with similar gene expression patterns in both shape (Fig. 4h) and size (Fig. 5b and Table SI_3) 233
across all mice. Therefore, the first observation of an alteration in the gene expression pattern (Fig. 4g) 234
occurred earlier than the alteration in the limb shape change (Fig. 4b). Our analyses provide evidence 235
that differences in the Dusp6 gene expression pattern occurred first, at the “Mid early period”, 236
preceding the phenotypic limb differentiation, which occurred a few hours later in development, during 237
the “Mid late period”. Overall, the time course showing the dynamics of limb and gene expression shape 238
changes over development (Figs. 4, 5 and SI_3, 4 and 5) confirmed that the Fgfr2 Apert syndrome 239
mutation causes an aberrant overexpression of Dusp6 early in development that could later lead to 240
significant limb malformations. 241
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
9
Altered Dusp6 expression and limb dysmorphology are highly associated. Finally, we explored the 242
correlation patterns between the limb phenotype and the gene expression pattern to further test 243
whether altered Fgf signaling underlies the limb malformations induced by Apert syndrome Fgfr2 P253R 244
mutation. First, we assessed the relationship between the size of the limbs and the volume of the Dusp6 245
expression domain pooling all the forelimbs and hindlimbs and assessing the correlation between these 246
two traits (Fig. 5c, d). The trend line showed that for the same limb size, Fgfr2+/P253R mutant mice showed 247
larger Dusp6 expression domains, both in forelimbs (R2=0.4) and hindlimbs (R2=0.6). If the extension of 248
the Dusp6 expression only depended on limb growth, a high correlation between the size of the limb 249
and the gene would be expected. However, the moderate correlation found here suggests that the size 250
of the Dusp6 gene expression is not dependent solely on limb size but is also influenced by other factors 251
and could be under further genetic regulatory control. 252
Second, we assessed the morphological integration between the shape of the limbs and the shape of the 253
Dusp6 expression domain. The statistical analysis of the covariance pattern between these shapes can 254
reflect the interaction of the phenotype and the gene expression pattern during limb development. As 255
shown by analysis of additional genes expressed during limb development15, even when a gene is 256
expressed within the limb, the shape of the limb and the shape of the gene expression domain are not 257
correlated by definition, and the integration pattern can change from a strong association to no 258
significant correlation within few hours of development15. The dynamics of the integration pattern can 259
identify the key periods during which the expression of a gene is relevant for determining the shape of 260
the limb. If the morphological integration is low, the expression of the gene will not be as relevant for 261
the shape of the limb as if the integration is high. If the integration is low, the impact of the altered gene 262
expression on the phenotype will be minimal. If the morphological integration is high, the impact of the 263
genetic mutation will be maximized (i.e., changes of the gene expression pattern will produce changes in 264
the limb shape). Our results showed that the shape of the limb and the shape of the Dusp6 domain were 265
indeed highly correlated (RV=0.88 in forelimbs; RV=0.91 in hindlimbs). This is evidence that altered Fgf 266
signaling induced by the Fgfr2 P253R Apert syndrome mutation will have a direct effect on the limb 267
phenotype. 268
By comparing the morphological integration pattern in mutant and unaffected littermates at different 269
periods, we can test whether this interaction is maintained or disrupted by the disease during 270
development. If the pattern or magnitude of morphological integration was different in mutant mice, it 271
would reveal further mechanisms underlying the etiology of the disease. Our analyses showed that the 272
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
10
pattern of morphological integration of the shape of the limbs and the shape of the Dusp6 expression 273
domains was similar in unaffected and Fgfr2+/P253R mutant mice (Fig. 5e and Fig SI_6). Our results 274
confirmed that the Fgfr2 P253R mutation does not disrupt the strong association between limb shape 275
and the Dusp6 expression domain. Therefore, the alteration of the Dusp6 expression pattern caused by 276
the Fgfr2 mutation between E10 and E11.5 will produce the limb dysmorphologies associated with Apert 277
syndrome. Overall, the high correlation between the shapes of the limb and the Dusp6 expression 278
domain provides further evidence that altered Fgf expression due to the Fgfr2 mutation is strongly 279
associated with limb defects in Apert syndrome. 280
281
Discussion 282
By definition, to reveal the primary etiology of an abnormality requires going back in time to the earliest 283
moment when abnormal development can be found. Typically, the earliest changes will be the most 284
subtle, and so the most statistically sensitive techniques are necessary. The methods currently used to 285
assess gene expression patterns are mainly qualitative and only focus on the shape and size differences 286
that can be simply detected by eye. Therefore, slight changes in gene expression domains, even if they 287
may have large effects on the phenotype45, can remain undetected. To reveal these subtle changes we 288
have developed a precise method combining OPT and GM for quantifying embryo morphology and the 289
underlying 3D gene expression patterns in a systematic, objective manner. This enables visualization and 290
quantification of how the genotype translates into the phenotype during embryonic development, to 291
compare normal and disease‐altered patterns of genetic and phenotypic variation and, eventually, 292
identify the origins of abnormal morphogenesis. This approach can further our understanding of the 293
etiology of genetic diseases in research using animal models46,47, even in those that do not seem to 294
recapitulate the human disease faithfully48. 295
Our study of the Fgfr2 P253R mouse model for Apert syndrome is an exemplary case of how 296
quantitative assessment can overcome the shortcomings of traditional qualitative morphological 297
assessment and lead to new discoveries. So far, the molecular and developmental mechanisms 298
underlying the limb defects associated with Apert syndrome remained obscure, even when these limb 299
abnormalities clinically differentiate Apert syndrome from other craniosynostosis syndromes (e.g. 300
Pfeiffer, Crouzon and Saethre‐Chotzen syndromes)23,24,41. Most Apert syndrome research focused on 301
premature fusion of cranial sutures and craniofacial malformations23,29–33, clinical traits that are 302
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
11
consistently phenocopied in mouse models. However, little research about the limb defects in Apert 303
syndrome has been done using the same animal models, mainly because previous research reported the 304
absence or subtle malformation of the limbs in the different mouse models for Apert syndrome34–36. 305
Contrary to these previous results, our quantitative morphometric analyses demonstrate that the limbs 306
of Fgfr2+/P253R Apert syndrome mice present with significant defects that are detectable in newborn mice 307
and can be traced back to early embryogenesis (Figs. 1, 4 and 5). 308
Our analyses provide insight into the genetic origins of these limb defects, showing that altered gene 309
expression patterns in the Fgf signaling pathway precede and contribute to limb dysmorphogenesis in 310
Fgfr2+/P253R Apert syndrome mice. In fact, we detected that Dusp6 expression patterns were different in 311
unaffected and mutant littermates a few hours before the first limb dysmorphologies appeared (Fig. 4), 312
and confirmed that limb shape and Dusp6 expression patterns were highly correlated (Fig. 5e). The 313
altered Fgf signaling observed was due to the Fgfr2 P253R Apert syndrome mutation, which causes loss 314
of ligand specificity of the Fgfr2 isoform and increased affinity of various Fgfs to Fgfr2. Our analyses 315
showed that over‐activation of the Fgf signaling pathway results in more expanded (Fig. 4e‐g) and larger 316
(Fig. 5b) expression domains of Dusp6, a gene that acts as a negative‐feedback control of Fgf signaling49. 317
A delay in misregulation of the expression of Dusp6 may explain the lack of antero‐posterior asymmetry 318
and the shape deficiencies observed in Fgfr2+/P253R mutant mice (Fig. 4). We found evidence that limb 319
shape and Dusp6 expression were highly associated (Fig. 5c‐e), but it is likely that other downstream 320
genes of the Fgf signaling pathway also contribute to the limb shape malformations associated with 321
Apert syndrome. 322
Our analyses also demonstrated that the embryonic limb defects persisted until birth (Fig. 1) and thus 323
did not disappear over development. For instance, we found that Fgfr2+/P253R Apert syndrome mice 324
presented postnatal limb malformations involving the shape, length and volume of many bones of the 325
forelimb, including the scapula, humerus, ulna, radius, metacarpals and phalanges (Fig. 1 and Table 326
SI_1). Though subtle, these malformations suggest that further research into the origins and causes of 327
limb dysmorphologies in Apert syndrome using these and other mouse models is warranted. 328
Our quantitative approach could be similarly applied to investigate other developmental defects and 329
dysmorphologies 50. OPT and GM can be potentially used to analyze any organ and animal model that 330
can be visualized during development using light microscopy, and any gene whose gene expression 331
pattern can be detected by WISH and show a continuous expression domain over development15. We 332
exemplified the method analyzing limb defects in mouse models, but it could be applied to 333
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
12
malformations affecting other organs such as the face, the brain, the heart, etc, in mouse models as well 334
as in any other vertebrate animal model, such as zebrafish, chicken and Xenopus. This is relevant as 335
major developmental defects represent a leading cause of infant mortality and affect a small but 336
relevant percentage of the population, severely compromising their quality of life3. 337
Through its increased quantitative sensitivity, our method has allowed us to reveal that the mouse 338
model for Apert syndrome does indeed show early defects in limb development. We detected that 339
dysregulation of an Fgf target gene precedes measurable changes in limb bud morphology, thus 340
identifying a relevant component of its genetic etiology. Quantitative evaluation of size and shape 341
should thus be performed before discarding any animal models as useful for investigating human 342
congenital malformations51. Our method has the potential to become a high‐throughput tool for 343
biomedical research, providing insight into the processes that cause malformations and lead to 344
malfunction, which is essential for understanding diseases and discovering potential therapies. 345
346
Methods 347
Mouse model. We analysed the Fgfr2+/P253R Apert syndrome mouse model, an inbred model backcrossed 348
on C57BL/6J genetic background for more than 10 generations carrying a mutation that in humans with 349
Apert syndrome is associated with more severe syndactyly. This gain of function mutation, which is 350
embryonically lethal in homozygosis, involves a proline to arginine amino acid change at position 253 of 351
the Fgfr2 protein that alters the ligand binding specificity of the receptor and causes a stronger receptor 352
signaling. Further details of the mouse model on generation of targeting construct can be found 353
elsewhere34. All the experiments were performed in compliance with the animal welfare guidelines 354
approved by the Pennsylvania State University Animal Care and Use Committees (IACUC46558, 355
IBC46590). 356
Micro‐CT imaging. High resolution micro‐computed tomography (μCT) scans were acquired by the 357
Center for Quantitative Imaging at the Pennsylvania State University (www.cqi.psu.edu) using the HD‐358
600 OMNI‐X high‐resolution X‐ray computed tomography system (Varian Medical Systems, Palo Alto, 359
CA). Pixel sizes ranged from 0.01487 to 0.01503 mm, and all slice thicknesses were 0.016 mm. Image 360
data were reconstructed on a 1024×1024 pixel grid as a 16‐bit TIFF. To reconstruct forelimb morphology 361
from the μCT images, isosurfaces were produced with median image filter using the software package 362
Avizo 6.3 (Visualization Sciences Group, VSG) (Fig. 1a‐f). 363
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
13
Morphometrics in P0 mice. We assessed forelimb morphology at P0 using unaffected (n=10) and 364
mutant (n=12) littermates of Apert syndrome mouse models. A set of 16 landmarks were collected on 365
each left forelimb, including points on the main bones of the forelimb (phalanges, metacarpals, radius, 366
ulna, humerus, scapula and clavicle), as shown in Fig. 1b‐f. To minimize measurement error, each 367
landmark was collected twice by the same observer restricting the deviations between the two trials to 368
0.05 mm. 369
At P0, we estimated the dimensions of the long bones of the forelimbs using the 3D coordinates of the 370
landmarks located at the proximal and distal ends of the bones (Fig. 1b‐f). We also estimated the bone 371
volumes from volume data collected from the μCT scans. To assess size differences in bone length and 372
bone volume between mutant and unaffected P0 mice of the Fgfr2+/P253R Apert syndrome mouse model, 373
we performed a two‐tailed one‐way ANOVA on those variables showing a normal distribution, and the 374
non‐parametric Mann‐Whitney U‐Test on those variables that deviated from a normal distribution. The 375
shape of the humerus and the scapula was comparatively assessed in unaffected and Fgfr2+/P253R Apert 376
syndrome littermates using Geometric Morphometrics. Shape information was extracted using a 377
General Procrustes Analysis (GPA)52, in which configurations of landmarks are superimposed by shifting 378
them to a common position, rotating and scaling them to a standard size until a best fit of corresponding 379
landmarks is achieved53. The resulting Procrustes coordinates from the GPA were the input for further 380
statistical analysis to compare the shape of the bones in unaffected and Fgfr2+/P253R mice. A Principal 381
Component Analysis (PCA) was used to explore the morphological variation within each bone. PCA 382
performs an orthogonal decomposition of the data and transforms variance covariance matrices into a 383
smaller number of uncorrelated variables called Principal Components (PCs), which successively account 384
for the largest amount of variation in the data20. Each specimen is scored for every principal component 385
and the specimens can be plotted using these scores along the morphospace defined by the principal 386
axes. 387
WISH and OPT scanning. To examine early embryonic mouse limb development in Apert syndrome we 388
bred 4 litters of the Fgfr2+/P253R Apert syndrome mouse model and collected them between E10.5 and 389
E11.5. In total, 32 mouse embryos were harvested and classified by PCR genotyping into unaffected 390
(n=16) and mutant (n=16) littermates (see Fig. SI_1 and Table SI_2 for further details on sample size and 391
composition). Dusp6 gene expression was assessed by Whole‐Mount‐In‐Situ Hybridization (WISH). 392
Mouse embryos were dissected in cold phosphate‐buffered saline, 0.1% tween 20 (PBT), fixed overnight 393
in 4% paraformaldehyde (Sigma), dehydrated in a graded PBT/methanol series and stored at ‐20ºC in 394
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
14
methanol. The mouse embryos recovered their original size after rehydration in decreasing series of 395
methanol/PBT. WISH was carried out using Dusp6 antisense RNA probes labelled with digoxigenin‐UTP 396
(Roche), following standard protocols7. Alkaline phosphatase coupled anti‐digoxigenin (anti‐DIG‐AP, 397
Roche) and NBT/BCIP staining (Roche) were used to reveal the expression pattern for Dusp6. To 398
minimize variations during experimental procedures, all embryos were processed systematically within 399
the same batch, processing unaffected and mutant littermates from different litters in separate tubes, 400
but simultaneously using the same probe, timings and concentration reagents. 401
After embedding in agarose, dehydrating in methanol and chemically clearing the samples with benzyl 402
alcohol and benzyl benzoate (BABB), the stained whole embryos were scanned with both fluorescence 403
and transmission light with a CFP (Cyan Fluorescent Protein) filter using our home‐build OPT imaging 404
system mounted on a Leica MZ 16 FA microscope5. The embryos were 3D‐reconstructed from the 405
resulting 2D images using Matlab and visualized using Amira 6.3 (Visualization Sciences Group, FEI). 406
From the OPT fluorescence scans we produced 3D reconstructions of the embryo surface and we 407
dissected the available right and left fore‐ and hind limbs of each specimen, resulting in a sample of 101 408
embryonic limbs (Table SI_2). From the OPT transmission scans, we recovered the Dusp6 expression 409
domain. As Dusp6 is expressed in a fuzzy spatial gradient, as already shown by other genes15, we used 410
3D multiple thresholding to visualize the gene expression domain at different intensities (Fig. 3 and 411
Video SI_1). To comparatively analyse the gene expression domains across the sample, we inspected for 412
each limb the whole range of threshold values under which the gene expression could be visualized, 413
from the threshold showing its first appearance to the threshold under which it disappeared and was no 414
longer detectable. We analysed the 3D reconstruction based on a threshold computed as 1/3 of the last 415
grey value showing the Dusp6 expression domain, which displayed a Dusp6 domain at high gene 416
expression (Fig. 3, step 3). Finally, we obtained 80 limbs (46 forelimbs and 34 hindlimbs) with associated 417
gene expression patterns for Dusp6. 418
Embryo staging. To account for breeding and developmental variation, individual limb buds were staged 419
using our publicly available web‐based staging system (http://limbstaging.crg.es)43. Considering the 420
spline curve along the outline of the limb, this tool provides a stage estimate with a reproducibility of ± 2 421
hours. According to the staging results, the different mouse litters were ordered following a continuous 422
temporal sequence from E10 to E11.5 (Fig. SI_1). To minimize high developmental variation within and 423
among litters of mice, hindlimbs and forelimbs from each litter were analysed separately, except for 424
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
15
those from two litters from the earliest stage that were pooled into the same group because their 425
temporal distribution completely overlapped, as shown in Fig. SI_1. 426
Morphometrics from E10 to E11.5. To capture the size and shape of the limbs and the expression 427
domains of Dusp6, we collected a set of anatomical landmarks as well as curve and surface 428
semilandmarks (Fig. SI_2), as recommended in structures devoid of homologous landmarks. 429
Semilandmarks are mathematical points located along a curve54 or a surface55 within the same object 430
that can be slid to corresponding equally spaced locations across the sample. We used Amira 6.3 431
(Visualization Sciences Group, FEI) to record the anatomical landmarks and Viewbox 4 (dHAL software, 432
Kifissia, Greece) to construct a limb template of surface and curve semilandmarks and to interpolate 433
them onto each target shape (Fig. SI_2). 434
The 3D landmark coordinates defining the shape of the limb and the Dusp6 expression domain were 435
analysed using Procrustes‐based landmark analysis54. Semilandmarks were allowed to slide in the GPA 436
by minimizing the bending energy54–56. Quantitative shape analyses based on PCA were performed as 437
explained above. 438
We estimated the size of the limb as the centroid size, calculated as the square root of the summed 439
squared distances between each landmark coordinate and the centroid of the limb configuration of 440
landmarks53. The volumes of the Dusp6 domains were estimated from the 3D reconstructions of the OPT 441
scans. Size differences in limb size and gene volume between mutant and unaffected embryonic mice 442
were tested for statistical significance using a two‐tailed Welch Two Sample t‐test. 443
We quantified the integration between the limb and the Dusp6 expression pattern and produced 444
visualizations of the patterns of associated shape changes between them using Two‐Block Partial Least 445
Squares analysis (PLS)57. This method performs a singular value decomposition of the covariance matrix 446
between the two blocks of shape data (i.e., the limb and the Dusp6 expression domain). Uncorrelated 447
pairs of new axes are derived as linear combinations of the original variables, with the first pair 448
accounting for the largest amount of inter‐block covariation, the second pair for the next largest amount 449
and so on. The amount of covariation is measured by the RV coefficient, which is a multivariate 450
analogue of the squared correlation58. Statistical significance was tested using permutation tests under 451
the null hypothesis of complete independence between the two blocks of variables. Separate analyses 452
for each developmental period, as well as for forelimbs and hindlimbs of all stages were computed. 453
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
16
All the analyses were performed using R (R Development Core Team, 2013; http://www.R‐project.org); 454
the R package geomorph59 ( http://cran.r‐project.org/web/packages/geomorph), SPSS Statistics 22 (IMB, 455
2013) and MorphoJ60. 456
457
458
References 459
1. Andrey, G. & Mundlos, S. The three‐dimensional genome: regulating gene expression during 460
pluripotency and development. Dev. Camb. Engl. 144, 3646–3658 (2017). 461
2. Chan, C. J., Heisenberg, C.‐P. & Hiiragi, T. Coordination of Morphogenesis and Cell‐Fate 462
Specification in Development. Curr. Biol. CB 27, R1024–R1035 (2017). 463
3. Toxicology, N. R. C. (US) C. on D. Developmental Defects and Their Causes. (National Academies 464
Press (US), 2000). 465
4. Rosen, B. & Beddington, R. S. Whole‐mount in situ hybridization in the mouse embryo: gene 466
expression in three dimensions. Trends Genet. TIG 9, 162–167 (1993). 467
5. Sharpe, J. et al. Optical projection tomography as a tool for 3D microscopy and gene expression 468
studies. Science 296, 541–545 (2002). 469
6. Correia, K. M. & Conlon, R. A. Whole‐mount in situ hybridization to mouse embryos. Methods 470
San Diego Calif 23, 335–338 (2001). 471
7. de la Pompa, J. L. et al. Conservation of the Notch signalling pathway in mammalian 472
neurogenesis. Dev. Camb. Engl. 124, 1139–1148 (1997). 473
8. Sharpe, J. Optical projection tomography as a new tool for studying embryo anatomy. J. Anat. 474
202, 175–181 (2003). 475
9. Boot, M. J. et al. In vitro whole‐organ imaging: 4D quantification of growing mouse limb buds. 476
Nat. Methods 5, 609–612 (2008). 477
10. Xu, Q. et al. Correlations Between the Morphology of Sonic Hedgehog Expression Domains and 478
Embryonic Craniofacial Shape. Evol. Biol. 1–8 (2015). doi:10.1007/s11692‐015‐9321‐z 479
11. Hu, D. et al. Signals from the brain induce variation in avian facial shape. Dev. Dyn. Off. Publ. Am. 480
Assoc. Anat. (2015). doi:10.1002/dvdy.24284 481
12. Jernvall, J., Keränen, S. V. & Thesleff, I. Evolutionary modification of development in mammalian 482
teeth: quantifying gene expression patterns and topography. Proc. Natl. Acad. Sci. U. S. A. 97, 14444–483
14448 (2000). 484
13. Salazar‐Ciudad, I. & Jernvall, J. A computational model of teeth and the developmental origins of 485
morphological variation. Nature 464, 583–586 (2010). 486
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
17
14. Mayer, C., Metscher, B. D., Müller, G. B. & Mitteroecker, P. Studying developmental variation 487
with Geometric Morphometric Image Analysis (GMIA). PloS One 9, e115076 (2014). 488
15. Martínez‐Abadías, N., Mateu, R., Niksic, M., Russo, L. & Sharpe, J. Geometric Morphometrics on 489
Gene Expression Patterns Within Phenotypes: A Case Example on Limb Development. Syst. Biol. 65, 194–490
211 (2016). 491
16. Airey, D. C., Wu, F., Guan, M. & Collins, C. E. Geometric morphometrics defines shape 492
differences in the cortical area map of C57BL/6J and DBA/2J inbred mice. BMC Neurosci. 7, 63 (2006). 493
17. James Rohlf, F. & Marcus, L. F. A revolution morphometrics. Trends Ecol. Evol. 8, 129–132 494
(1993). 495
18. Wiley: Statistical Shape Analysis ‐ Ian L. Dryden, Kanti V. Mardia. Available at: 496
http://eu.wiley.com/WileyCDA/WileyTitle/productCd‐0471958166.html. (Accessed: 14th September 497
2015) 498
19. Klingenberg, C. P. Evolution and development of shape: integrating quantitative approaches. 499
Nat. Rev. Genet. 11, 623–635 (2010). 500
20. Hallgrimsson, B. et al. Morphometrics, 3D Imaging, and Craniofacial Development. Curr. Top. 501
Dev. Biol. 115, 561–597 (2015). 502
21. Adams, D. C., Rohlf, F. J. & Slice, D. E. A field comes of age: geometric morphometrics in the 21st 503
century. Hystrix Ital. J. Mammal. 24, 7–14 (2013). 504
22. Klingenberg, C. P. Morphometrics and the role of the phenotype in studies of the evolution of 505
developmental mechanisms. Gene 287, 3–10 (2002). 506
23. Craniosynostosis: Diagnosis, Evaluation, and Management. (Oxford University Press, 2000). 507
24. Park, W. J. et al. Analysis of phenotypic features and FGFR2 mutations in Apert syndrome. Am. J. 508
Hum. Genet. 57, 321–328 (1995). 509
25. Wilkie, A. O. M. et al. Apert syndrome results from localized mutations of FGFR2 and is allelic 510
with Crouzon syndrome. Nat. Genet. 9, ng0295‐165–165 (1995). 511
26. Yu, K., Herr, A. B., Waksman, G. & Ornitz, D. M. Loss of fibroblast growth factor receptor 2 512
ligand‐binding specificity in Apert syndrome. Proc. Natl. Acad. Sci. U. S. A. 97, 14536–14541 (2000). 513
27. Yu, K. & Ornitz, D. M. Uncoupling fibroblast growth factor receptor 2 ligand binding specificity 514
leads to Apert syndrome‐like phenotypes. Proc. Natl. Acad. Sci. U. S. A. 98, 3641–3643 (2001). 515
28. McIntosh, I., Bellus, G. A. & Jab, E. W. The pleiotropic effects of fibroblast growth factor 516
receptors in mammalian development. Cell Struct. Funct. 25, 85–96 (2000). 517
29. Martínez‐Abadías, N. et al. Beyond the closed suture in apert syndrome mouse models: 518
evidence of primary effects of FGFR2 signaling on facial shape at birth. Dev. Dyn. Off. Publ. Am. Assoc. 519
Anat. 239, 3058–3071 (2010). 520
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
18
30. Hill, C. A. et al. Postnatal brain and skull growth in an Apert syndrome mouse model. Am. J. Med. 521
Genet. A. 161A, 745–757 (2013). 522
31. Holmes, G. et al. Early onset of craniosynostosis in an Apert mouse model reveals critical 523
features of this pathology. Dev. Biol. 328, 273–284 (2009). 524
32. Heuzé, Y. et al. Morphological comparison of the craniofacial phenotypes of mouse models 525
expressing the Apert FGFR2 S252W mutation in neural crest‐ or mesoderm‐derived tissues. Bone 63, 526
101–109 (2014). 527
33. Holmes, G. & Basilico, C. Mesodermal expression of Fgfr2S252W is necessary and sufficient to 528
induce craniosynostosis in a mouse model of Apert syndrome. Dev. Biol. 368, 283–293 (2012). 529
34. Wang, Y. et al. Activation of p38 MAPK pathway in the skull abnormalities of Apert syndrome 530
Fgfr2(+P253R) mice. BMC Dev. Biol. 10, 22 (2010). 531
35. Chen, L., Li, D., Li, C., Engel, A. & Deng, C.‐X. A Ser252Trp [corrected] substitution in mouse 532
fibroblast growth factor receptor 2 (Fgfr2) results in craniosynostosis. Bone 33, 169–178 (2003). 533
36. Wang, Y. et al. Abnormalities in cartilage and bone development in the Apert syndrome 534
FGFR2(+/S252W) mouse. Dev. Camb. Engl. 132, 3537–3548 (2005). 535
37. Slaney, S. F. et al. Differential effects of FGFR2 mutations on syndactyly and cleft palate in Apert 536
syndrome. Am. J. Hum. Genet. 58, 923–932 (1996). 537
38. von Gernet, S., Golla, A., Ehrenfels, Y., Schuffenhauer, S. & Fairley, J. D. Genotype‐phenotype 538
analysis in Apert syndrome suggests opposite effects of the two recurrent mutations on syndactyly and 539
outcome of craniofacial surgery. Clin. Genet. 57, 137–139 (2000). 540
39. Kawakami, Y. et al. MKP3 mediates the cellular response to FGF8 signalling in the vertebrate 541
limb. Nat. Cell Biol. 5, 513–519 (2003). 542
40. Li, C., Scott, D. A., Hatch, E., Tian, X. & Mansour, S. L. Dusp6(Mkp3) is a negative feedback 543
regulator of FGF stimulated ERK signaling during mouse development. Dev. Camb. Engl. 134, 167–176 544
(2007). 545
41. Holten, I. W., Smith, A. W., Bourne, A. J. & David, D. J. The Apert syndrome hand: pathologic 546
anatomy and clinical manifestations. Plast. Reconstr. Surg. 99, 1681–1687 (1997). 547
42. Cohen, M. M. & Kreiborg, S. Hands and feet in the Apert syndrome. Am. J. Med. Genet. 57, 82–548
96 (1995). 549
43. Boehm, B. et al. A landmark‐free morphometric staging system for the mouse limb bud. Dev. 550
Camb. Engl. 138, 1227–1234 (2011). 551
44. Ibrahimi, O. A. et al. Structural basis for fibroblast growth factor receptor 2 activation in Apert 552
syndrome. Proc. Natl. Acad. Sci. U. S. A. 98, 7182–7187 (2001). 553
45. Honeycutt, R. L. Small changes, big results: evolution of morphological discontinuity in 554
mammals. J. Biol. 7, 9 (2008). 555
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
19
46. Rosenthal, N. & Brown, S. The mouse ascending: perspectives for human‐disease models. Nat. 556
Cell Biol. 9, 993–999 (2007). 557
47. Spradling, A. et al. New Roles for Model Genetic Organisms in Understanding and Treating 558
Human Disease: Report From The 2006 Genetics Society of America Meeting. Genetics 172, 2025–2032 559
(2006). 560
48. Guénet, J.‐L. Animal models of human genetic diseases: do they need to be faithful to be useful? 561
Mol. Genet. Genomics 286, 1–20 (2011). 562
49. Ekerot, M. et al. Negative‐feedback regulation of FGF signalling by DUSP6/MKP‐3 is driven by 563
ERK1/2 and mediated by Ets factor binding to a conserved site within the DUSP6/MKP‐3 gene promoter. 564
Biochem. J. 412, 287–298 (2008). 565
50. Winter, R. M. & Baraitser, M. The London Dysmorphology Database. J. Med. Genet. 24, 509–510 566
(1987). 567
51. Zuniga, A., Zeller, R. & Probst, S. The molecular basis of human congenital limb malformations. 568
Wiley Interdiscip. Rev. Dev. Biol. 1, 803–822 (2012). 569
52. Rohlf, F. J. & Slice, D. Extensions of the Procrustes Method for the Optimal Superimposition of 570
Landmarks. Syst. Zool. 39, 40–59 (1990). 571
53. Dryden, I. L. & Mardia, K. V. Statistical Shape Analysis. (Wiley, 1998). 572
54. Bookstein, F. L. Morphometric Tools for Landmark Data: Geometry and Biology. (Cambridge 573
University Press, 1997). 574
55. Gunz, P., Mitteroecker, P. & Bookstein, F. L. Semilandmarks in Three Dimensions. in Modern 575
Morphometrics in Physical Anthropology 73–98 (Springer, Boston, MA, 2005). doi:10.1007/0‐387‐27614‐576
9_3 577
56. Mitteroecker, P. & Gunz, P. Advances in Geometric Morphometrics. Evol. Biol. 36, 235–247 578
(2009). 579
57. Rohlf, F. J. & Corti, M. Use of Two‐Block Partial Least‐Squares to Study Covariation in Shape. 580
Syst. Biol. 49, 740–753 (2000). 581
58. Klingenberg, C. P. Morphometric integration and modularity in configurations of landmarks: 582
tools for evaluating a priori hypotheses. Evol. Dev. 11, 405–421 (2009). 583
59. Adams, D. C. & Otárola‐Castillo, E. geomorph: an r package for the collection and analysis of 584
geometric morphometric shape data. Methods Ecol. Evol. 4, 393–399 (2013). 585
60. Klingenberg, C. P. MorphoJ: an integrated software package for geometric morphometrics. Mol. 586
Ecol. Resour. 11, 353–357 (2011). 587
588
589
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
20
Acknowledgements 590
We acknowledge support of the Spanish Ministry of Economy and Competitiveness to the EMBL 591
partnership* and ‘Centro de Excelencia Severo Ochoa, and the CERCA Programme / Generalitat de 592
Catalunya. We acknowledge the support of the CERCA Programme / Generalitat de Catalunya. We 593
acknowledge Ethylin Wang Jabs for access to the Fgfr2+/P253R Apert syndrome mouse model. The 594
research leading to these results received funding from the following grants: European Union Seventh 595
Framework Program (FP7/2007‐2013) under grant agreement Marie Curie Fellowship FP7‐PEOPLE‐2012‐596
IIF 327382, National Institutes of Health NICHD P01HD078233 and NIDCR R01DE02298, and a Burroughs‐597
Welcome Fund 2013 Collaborative Research Travel Grant. 598
599
Author contributions 600
NM‐A, JS and JR conceived the project. JR contributed with experimental mouse models; AR‐M, JS, LS 601
and KK conducted experiments and scanning; and NM‐A, RM, JS, SMP and MY collected and analysed 602
the data. NM‐A and JS wrote the paper and all authors critically reviewed the manuscript. 603
604
Figure legends 605
606
Figure 1. Quantitative size and shape comparison of forelimb bones in Fgfr2+/P253R newborn mice (P0) 607
and unaffected littermates (P0). a) Mouse skeleton at P0. 3D isosurface reconstruction of the skeleton 608
of an unaffected littermate obtained from a high resolution µCT scan. b‐f) Anatomical landmarks 609
recorded on µCT scans of Apert syndrome mice at P0. b) Autopod (hand). Landmarks were recorded at 610
the midpoint of the proximal and distal tips of the distal, mid and proximal phalanges (P1‐P28) and the 611
metacarpals (P29‐P38). Proximal phalanx I, middle phalanx V, and metacarpal I are not displayed 612
because these bones have not yet mineralized at P0 and could not be visualized in many specimens. c) 613
Zeugopod. Landmarks were recorded at the midpoint of the proximal and distal tips of the radius (R1‐614
R2) and ulna (U1‐U2). d) Stylopod. Landmarks were recorded at the midpoint of the proximal and distal 615
tips of the humerus (H1‐H2), as well as the tip of the deltoid process (H3). e) Scapula. Landmarks were 616
recorded at the most superior and inferior lateral points of the scapula (S1‐S2), the most posterior point 617
of the spine (S3), most antero‐medial point of the acromion process (S4) and the medial, superior and 618
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
21
inferior points of the glenoid cavity (S5‐S7). f) Clavicle. Landmarks were recorded at the medial point of 619
the sternal and the acromial ends (C1‐C2). g) Length and volume differences in the forelimbs of Apert 620
syndrome mouse models. Schematic representation of the forelimb of a P0 mouse showing in different 621
colours statistically significant differences in bone length and volume as measured by two‐tailed one‐622
way ANOVA or Mann‐Whitney U‐test in Fgfr2+/P253R mice relative to unaffected littermates as specified in 623
Table SI_1. Longer/shorter refer to length, whereas larger/smaller refer to volume. Blue: Longer and 624
larger, green: longer, purple: shorter and larger, pink: shorter and smaller, orange: no difference, yellow: 625
not assessed because bones were not visible/ossified. h, i) Shape differences in the forelimbs of Apert 626
syndrome mouse models. Scatterplots of PC1 and PC2 scores based on Procrustes analysis of anatomical 627
landmark locations representing the shape of the left humerus (h) and the left scapula (i) of unaffected 628
(n=10) and mutant (n=12) littermates of Apert syndrome mouse models. Convex hulls represent the 629
ranges of variation within each group of mice. 630
631
Figure 2. Qualitative visualisation of gene expression of Dusp6 in unaffected and Fgfr2+/P253R mouse 632
embryos at E10.5 and E11.5. OPT scans of embryos WISH stained for Dusp6 revealed the anatomical 633
location of gene expression (shown in yellow). For each stage, the main expression domains are 634
highlighted on the left for anatomical reference on a lateral view of a 3D reconstruction of a Fgfr2+/+ 635
unaffected embryo (fb: forebrain, is: isthmus, hb: hindbrain, mx: maxillary prominence, md: mandibular 636
prominence, lnp: lateral nasal process, mnp: medial nasal process, ht: heart, sc: spinal cord, fl: forelimb, 637
hl: hindlimb). On the right, 3D reconstructions of three unaffected and three Fgfr2+/P253R mutant 638
embryos from the same litter are displayed to represent the high degree of variation in developmental 639
age within litters. Embryos are not to scale. 640
641
Figure 3. New quantitative analysis method for 3D gene expression data, based on geometric 642
morphometrics. Mouse embryos between E10.5 and E11.5 were analysed with WISH to reveal the 643
expression of Dusp6 (1), and then cleared with BABB and OPT scanned using both fluorescence and 644
transmission light (2). The external surface of the embryo was obtained from the 3D reconstruction of 645
the fluorescence scan (2). Multiple thresholding of the transmission scan by choosing different levels of 646
grey values as shown by the histogram allowed to visualize gene expression patterns at different 647
intensities (3). Moderate gene expression (shown in green) was displayed as the isosurface obtained 648
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
22
using as a threshold the grey value computed as 2/3 of the last grey value showing the Dusp6 expression 649
domain (3). High gene expression (shown in yellow) was displayed as the isosurface obtained using as a 650
threshold the grey value computed as 1/3 of the last grey value showing the Dusp6 expression domain 651
(3). From the whole mouse embryo isosurfaces, all four limb buds were segmented (5a). From the high 652
gene expression isosurface, the Dusp6 domains from all the available limbs were segmented (5b). 653
Maximum curvature patterns were displayed to optimize landmark recording (5). For each limb we 654
captured the shape and size of the limb bud (6a) and the underlying high Dusp6 gene expression pattern 655
(6b) recording the 3D coordinates of anatomical landmarks (yellow dots), curve semilandmarks (red 656
dots) and surface semilandmarks (pink dots). Anatomical and curve landmarks were recorded manually 657
on each limb. Surface landmarks were recorded on one template limb and interpolated onto target 658
limbs. Landmark coordinates were the input for Geometric Morphometric (GM) quantitative shape 659
analysis (7, 8) to superimpose the landmark data (GPA, General Procrustes Analysis), compute limb size 660
(Centroid size), and explore shape variation within limbs and gene expression domains by litters (PCA, 661
Principal Component Analysis). Finally, the covariation patterns between the shape of the limb and the 662
shape of the gene expression domain were also explored (PLS, Partial Least squares). 663
664
Figure 4. Tracing of limb phenotypes (anatomical and molecular) back through developmental time to 665
the earliest moment of appearance. Principal Component Analyses based on the Procrustes‐based 666
semilandmark was used to analyse the shape of the limbs and the corresponding Dusp6 expression 667
domains at each developmental period. Each period was analysed separately for the shape of the limb 668
(a‐d) and the Dusp6 expression domain (e‐h), as specified in Methods and Table SI_2. Scatterplots of PC1 669
and PC2 axes with the corresponding percentage of total morphological variation explained are 670
displayed for each analysis, along with morphings associated with the negative, mid and positive values 671
of the PC axis that separates mutant and unaffected littermates (PC1 or PC2, as highlighted in bold black 672
letters in the corresponding axis). Morphings are displayed in grey tones when the analysis showed no 673
differentiation between mutant and unaffected littermates. Morphs are displayed in colour when the 674
analysis revealed differentiation between mice (blue: unaffected littermates; pink: mutant littermates). 675
Limb buds and Dusp6 domains are not to scale and oriented with the distal aspect to the left, the 676
proximal aspect to the right, the anterior aspect at the top and the posterior aspect at the bottom of all 677
images. Convex hulls represent the ranges of variation within each group of mice. 678
679
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
23
Figure 5. Quantitative correspondence between the size and shape of the limbs and the Dusp6 680
expression pattern. a‐d) Comparison of limb bud size and Dusp6 volume in unaffected and Fgfr2+/P253R 681
mutant littermates across development. Limb size was measured as limb centroid size (a), whereas the 682
size of the Dusp6 expression was measured as the volume of the gene domain (b), as specified in 683
Methods and Tables SI_2 and 3. The association between the size of the limbs and the volume of the 684
Dusp6 domain was assessed separately for forelimbs (c) and hindlimbs (d). Statistical significant 685
differences as revealed by two‐tailed t‐tests are marked with asterisks, representing the degree of 686
significance: *P‐value=0.03, ** P‐value=0.01. e) Time course assessing the morphological integration 687
pattern between the limb phenotype and the shape of the gene expression pattern using Partial Least 688
Squares analyses. Associated shape changes from late to early limb development are shown from 689
morphings associated with the negative, mid and positive values of PLS1, which accounted for almost all 690
the covariation (97.6% in forelimbs and 99.5% in hindlimbs) between the limb buds and the Dusp6 gene 691
expression domains. Limb buds and Dusp6 domains are not to scale and oriented distally to the left, 692
proximally to the right, anteriorly to the top and posteriorly to the bottom. On the right, representing 693
the positive extreme of PLS1 axis, typical early limb buds showed a protruding shape (i.e. short in the 694
proximo‐distal axis and symmetrical in the antero‐posterior axis) associated with a flat‐bean shaped 695
Dusp6 expression pattern localized in the distal limb region, underlying the apical ectodermal ridge, and 696
spreading proximally towards the dorsal and ventral sides of the limb. On the left, representing the 697
negative extreme of the PLS1 axis, limb buds were elongated in the proximal axis, asymmetric on the 698
antero‐posterior axis, with an expansion of the distal limb region and a contraction of the proximal 699
region, at the wrist level. This typical limb shape of more developed limbs was associated with a Dusp6 700
expression that was extended underneath the apical ectodermal ridge towards the anterior and the 701
posterior ends of the gene expression, but reduced on the dorsal and ventral sides of the limb. 702
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
24
FIGURE 1 703
704
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
25
FIGURE 2 705
706
707
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
26
FIGURE 3 708
709
710
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint
27
FIGURE 4 711
712
.C
C-B
Y-N
C 4.0 International license
was not certified by peer review
) is the author/funder. It is made available under a
The copyright holder for this preprint (w
hichthis version posted January 11, 2018.
. https://doi.org/10.1101/246256
doi: bioR
xiv preprint
28
FIGURE 5 713
714
.CC-BY-NC 4.0 International licensewas not certified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (whichthis version posted January 11, 2018. . https://doi.org/10.1101/246256doi: bioRxiv preprint