Augmentation of BMP Signaling in Cranial Neural Crest Cells Leads to Premature Cranial Sutures Fusion through Endochondral Ossification in Mice

ABSTRACT Craniosynostosis is a congenital anomaly characterized by the premature fusion of cranial sutures. Sutures are a critical connective tissue that regulates bone growth; their aberrant fusion results in abnormal shapes of the head and face. The molecular and cellular mechanisms have been investigated for a long time, but knowledge gaps remain between genetic mutations and mechanisms of pathogenesis for craniosynostosis. We previously demonstrated that the augmentation of bone morphogenetic protein (BMP) signaling through constitutively active BMP type 1A receptor (caBmpr1a) in neural crest cells (NCCs) caused the development of premature fusion of the anterior frontal suture, leading to craniosynostosis in mice. In this study, we demonstrated that ectopic cartilage forms in sutures prior to premature fusion in caBmpr1a mice. The ectopic cartilage is subsequently replaced by bone nodules leading to premature fusion with similar but unique fusion patterns between two neural crest‐specific transgenic Cre mouse lines, P0‐Cre and Wnt1‐Cre mice, which coincides with patterns of premature fusion in each line. Histologic and molecular analyses suggest that endochondral ossification in the affected sutures. Both in vitro and in vivo observations suggest a greater chondrogenic capacity and reduced osteogenic capability of neural crest progenitor cells in mutant lines. These results suggest that the augmentation of BMP signaling alters the cell fate of cranial NCCs toward a chondrogenic lineage to prompt endochondral ossification to prematurely fuse cranial sutures. By comparing P0‐Cre;caBmpr1a and Wnt1‐Cre;caBmpr1a mice at the stage of neural crest formation, we found more cell death of cranial NCCs in P0‐Cre;caBmpr1a than Wnt1‐Cre;caBmpr1a mice at the developing facial primordia. These findings may provide a platform for understanding why mutations of broadly expressed genes result in the premature fusion of limited sutures. © 2022 The Authors. JBMR Plus published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research.


Introduction
C ranial sutures are fibrous tissues connecting bones of the head and face and function as signaling centers to support the continuous growth of facial and calvarial bones. (1)(2)(3) Many sutures eventually fuse during postnatal development in humans. When the sutures prematurely fuse as a result of genetic mutations or other biological reasons, especially in young ages, this causes a pathologic condition called craniosynostosis. (2)(3)(4) Craniosynostosis leads to abnormal skull shape and increased intracranial pressure, potentially resulting in blindness, deafness, and some cases mental retardation. The incidence is $1 in 2,500 births. (2) Current treatment is limited to repetitive surgeries to remove the prematurely fused sutures; these repeated surgeries also have negative impacts on patients' quality of life and impose a significant financial burden. (5)(6)(7) Thus, understanding the molecular and cellular pathogenesis of craniosynostosis may contribute to the development of strategies for early identification, prevention, and treatment of these conditions.
Cranial neural crest cells (NCCs) are the progenitor population giving rise to the major components of the face and the anterior part of the head. (8,9) Cranial NCCs are multipotent cells that differentiate into osteoblasts, chondrocytes, adipose, and other ectoderm-derived tissues. (10,11) Dysregulations of cranial NCCs, for example, aberrant migration, cell death, proliferation, differentiation, and cell fate specification, have been investigated as causative factors in craniofacial defects such as cleft palate and craniosynostosis. (12)(13)(14) Cranial NCCs and suture development in mice are similar to those in humans. Therefore, mouse models are valuable tools in determining the etiology and potential therapeutic strategies to treat craniosynostosis.
Several signaling pathways are reported to regulate the mechanisms of cranial NCCs. We and others have reported that transgenic mice with gain-of-function mutations or loss-of-function mutations for bone morphogenetic protein (BMP) signaling in NCCs develop craniofacial defects. (15)(16)(17) We generated transgenic mice expressing constitutively active BMP type 1A receptor (caBmpr1a) in a NCC-specific manner using a P0-Cre mouse line to demonstrate that P0-Cre;caBmpr1a mice prematurely fused the anterior frontal (AF) suture, leading to a craniosynostosis. (15,18,19) Interestingly, P0-Cre;caBmpr1a mice developed premature fusion in the AF suture but other sutures such as the coronal and sagittal sutures remained patent. (15) Despite broader expressions of the responsive genes in the craniofacial region, a fusion in limited sutures is frequently observed in patients with craniosynostosis and model mice for craniosynostosis. (20,21) As a result of the complexity of diseases, the underlying mechanisms remain poorly understood.
As the first step toward understanding a mechanism leading to a suture-specific premature fusion by BMP signaling, we selected two neural crest-specific Cre transgenic mice lines, Wnt1-Cre and P0-Cre, of which Cre expression is driven by Wnt1 promoter and Protein zero (P0) promoter, respectively. (8,(22)(23)(24)(25) They show similar, but not identical, Cre expression patterns of cranial NCCs at early embryonic stages. (25)(26)(27) We previously showed that neural crest-specific disruption of Evc2, a causative gene for Ellis-van Creveld syndrome, by Wnt1-Cre mice or P0-Cre mice, develops a different severity of craniofacial deformity due to their different recombination efficiency in the skull base. (28) We thus expected that differential augmentation patterns of BMP signaling using two Cre transgenic lines might highlight a suture-specific behavior of cranial NCCs with BMP signaling.
Endochondral ossification is one of two mechanisms for bone formation, which develops cartilage tissues first. Mesenchymal cells at an early embryonic stage form high-density cell aggregations and start to express Sox9, a chondrogenic master gene, to initiate chondrogenesis. The resulting cartilage is gradually replaced by bone tissue. (29,30) Intramembranous ossification is another mechanism that plays a role for calvaria development. Mesenchymal cells in osteogenic fronts at cranial sutures differentiate into osteoblasts to form bones without the formation of cartilage. However, several reports demonstrated the potential involvement of endochondral ossification at the moment of suture fusion, both in physiologic and pathologic instances. Unlike humans, most sutures in mice stay patent throughout their life, except the posterior frontal (PF) suture, which normally fuses by postnatal day 15 (P15). (31) Prior to the fusion of the PF suture under normal physiological condition, the formation of cartilage tissue is demonstrated, and the subsequent disappearance of the cartilage coincides with suture fusion. (31) Similarly, the presence of cartilage in PF sutures in rats and metopic sutures in humans has been reported. (32,33) In pathological conditions, ectopic cartilage formation is found in the sagittal suture of Axin2 mutant mice before its premature fusion, likely due to increased Wnt signaling activity. (34) Interestingly, ectopic cartilage was also identified at the coronal suture in knock-in mouse carrying a gain-of-function mutation of fibroblast growth factor receptor 2 (FGFR2) before premature suture fusion. (35) These findings suggest that ectopic cartilage promotes endochondral ossification within the suture mesenchyme, resulting in premature fusion of cranial sutures in both rodents and humans.
In this study, we demonstrated that enhanced BMP signaling in NCCs leads to the development of ectopic cartilage in sutures that subsequently undergo premature fusion. We employed two NCC-specific Cre lines, P0-Cre and Wnt1-Cre, and the resulting mice show similar but unique patterns of ectopic cartilage formation that coincides with their respective premature fusion patterns. These results suggest that endochondral ossification through ectopic cartilage formation is a mechanism that may lead to premature suture fusion. Despite identical Cre recombination patterns between P0-Cre and Wnt1-Cre at the perinatal stages, we found different levels of cell death in the medial nasal process (MNP) at E10.5, which is a putative origin of suture mesenchyme. These results imply that cranial NCCs may be primed by the augmentation of BMP signaling soon after their birth, which subsequently influences their behavior at perinatal stages for craniofacial development.

Mouse breeding
The mouse line carrying the Cre-inducible constitutively activated Bmpr1a (caBmpr1a) transgene was described previously. (15) caBmpr1a were bred with Wnt1-Cre mice (B6.Cg-Tg (Wnt1-cre)11Rth Tg(Wnt1-GAL4)11Rth/J, Jackson Lab, Stock No. 009107) (22) or P0-Cre mice (C57BL/6JTg(P0-Cre)94Imeg (ID 148) provided by CARD, Kumamoto University, Japan) (23) to obtain two distinct patterns of augmented BMP signaling in NCCs (henceforth Wnt1-Cre;caBmpr1a and P0-Cre;caBmpr1a). To visualize NCCs, Wnt1-Cre and P0-Cre mice were crossed with Cre reporter lines: Rosa26-LacZ (36) or tdTomato (B6.Cg-Gt (ROSA)26Sortm14(CAG-tdTomato)Hze/J, Jackson Lab, Stock No. 007914). (37) All mice used were group housed in specific pathogen-free conditions, fed a regular rodent diet, and kept in a healthy state. For each of the methods outlined, both male and female mice were utilized. All animals and embryos were allocated to experimental groups based on their genotypes, and there were no excluded animals and embryos. Investigators were blinded during allocation, animal handling, and endpoint measurements. All animal experiments were performed in accordance with the policies and federal laws associated with the judicious use of vertebrate animals as approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Michigan (PRO00009613) and were conducted in accordance with ARRIVE guidelines.

Micro-computed tomography
Mice were euthanized at postnatal day 17 (P17), and heads were harvested, debrided of extra tissues, and fixed with 10% formalin overnight. The heads were then embedded in 1% agarose, placed in a 19-mm-diameter tube, and scanned over their whole length using a micro-computed tomography (μCT) system (μCT40 Scanco Medical). Bone surface images were generated as described previously. (15,19) Image segmentation and surface models For quantification of the skull morphological differences, surface models were generated from CT scans through ITK-SNAP (www. itksnap.org). (38) Then anatomical landmarks were placed on the models according to Tables S1 and S2, which were also shown in Figure S1, (39,40) using 3D Slicer (www.slicer.org). The linear distances between landmarks were determined by the Q3DC module in 3D Slicer. There were no significant gender differences in either the control or the mutants, and the data from both genders are shown and compared between genotypes in Tables 1, S3, and S4.
Isolation of cranial neural crest stem cells and cells of nasal process at E11.5 We followed a previously published method. (11) In brief, we isolated the cranial region of P0-Cre;caBmpr1a;mTmG embryos at E8.5, and dissociated cells were cultured on a Matrigel-coated dish using media containing 10 3 units/mL of leukemia inhibitory factor and 25 ng/mL of FGF2. At passage 3, GFP-positive cells were isolated using fluorescence-activated cell sorting (FACS) and cultured further. Clones that were able to be cultured beyond passage 10 were assessed for expression of stem cell markers using quantitative RT-PCR primers listed in Table S5. Their multipotency was assessed using media to prompt osteogenesis and pellet culture for chondrogenesis. (11) The efficiency of establishing cranial neural crest stem cells (NCSCs) is summarized in Table S6. To isolate and differentiate cells of the nasal process (NP cells), we followed a previously published method. (11,41) Differentiation capacities were evaluated by alkaline phosphatase (ALP) staining, Alcian blue staining, and quantitative RT-PCR for differentiation marker genes. (11,42) Statistical analyses Statistical analyses were performed using GraphPad Prism version 9.0 software. For each dataset, Student's t test for the comparison of two groups of mice and one-way ANOVA with Tukey's test for multiple comparisons between groups of mice were performed. The p value is indicated by asterisks: * p < 0.05 and ** p < 0.01.
We then carried out a more detailed analysis of the skull shape. We placed landmarks on skull models according to Tables S1 and S2 to determine multiple linear distances. In the comparison of P0-Cre;caBmpr1a mice with littermate controls, we observed significant differences in 14 out of 21 measurements, including skull length, nasal bone length, frontal bone length, width of nasal bone at the intersection with premaxillae, viscerocranial length, width at zygomatic arch, width of temporal bone, cranial base length, presphenoid length, basisphenoid length, basioccipital length, viscerocranial height at intersphenoid synchondrosis (ISS), viscerocranial height at bason, and erupted lower incisor length (Tables 1 and S3). In contrast, only four measurements, skull length, nasal bone length, frontal bone length, and cranial base length, showed differences in Wnt1-Cre; caBmpr1a mice, and two other measurements, viscerocranial length and basisphenoid length, showed a tendency toward reduction (Table 1 and S2). These differences highlight similar but not identical skull defects in P0-Cre;caBmpr1a and Wnt1-Cre;caBmpr1a mice.
Ectopic cartilage formed in suture where prematurely fused We recently reported that constitutively activated ACVR1, another BMP receptor type1, in NCCs (henceforth caAcvr1; P0-Cre mice) developed ectopic cartilage in the developing face. (41) It has been reported that the PF suture fuses via endochondral ossification under normal physiological conditions. (31) Based on these reports, we hypothesized that enhanced BMP signaling in NCCs altered cell fate to develop ectopic cartilages within suture mesenchyme, which prompts endochondral ossification, resulting in premature suture fusion.
We next identified cartilage tissues in the sutures after the birth of P0-Cre;caBmpr1a mice by whole mount cartilage staining with Alcian blue. Ectopic cartilages formed at positions of the AF suture and the naso-premaxillary suture of P0-Cre;caBmpr1a mice at NB and postnatal day 1 (P1), and the ectopic cartilage disappeared at P2 (Fig. 3A arrows). A small portion of bone nodule concomitantly formed at the AF suture at the NB stage, where the ectopic cartilage formed; subsequently, the size of bone nodules gradually increased in association with the disappearing ectopic cartilage (Fig. 3B, arrowheads). These results suggest that the ectopic cartilage was replaced by bone nodules. Histological observation confirmed Alcian blue-stained cartilage-like tissues were present in the AF suture of P0-Cre;caBmpr1a mice (Fig. 3C). Interestingly, in Wnt1-Cre;caBmpr1a mice, we identified ectopic cartilage only in the AF sutures, but not in the nasopremaxillary suture at the NB stage (Fig. 3D). We further histologically assessed fusion patterns of sutures in control, Wnt1-Cre;caBmpr1a, and P0-Cre;caBmpr1a mice. The naso-premaxillary suture, which prematurely fused only in P0-Cre;caBmpr1a mice, developed a small portion of ectopic cartilage tissue at the NB stage of P0-Cre;caBmpr1a mice, but not in control and Wnt1-Cre;caBmpr1a mice ( Fig. 3E, arrowheads). There were also ectopic SOX9-positive chondrogenic progenitor cells that are positive for phospho-SMAD1/5/9 in the nasopremaxillary suture (Fig. 3E, arrowheads). On the other hand, the AF suture, which prematurely fused in both Wnt1-Cre;caBm-pr1a and P0-Cre;caBmpr1a mice, formed ectopic cartilage with cells positive for SOX9 and phospho-SMAD1/5/9 (Fig. 3F, arrows). The AF suture in control mice did not develop ectopic cartilage.
Next, we examined whether the ectopic cartilage contributed to ectopic bone formation through endochondral ossification. We histologically assessed the ectopic cartilage in the AF suture in control and P0-Cre;caBmpr1a mice at NB, P1, P3, and P6 (Fig. 3G). The ectopic cartilage was enclosed in ectopic bone tissue in P0-Cre;caBmpr1a mice at P3; then the ectopic cartilage became smaller at P6 (Fig. 3G, black allow heads). Endochondral ossification progresses by chondrocyte differentiation into matured hypertrophic chondrocytes and invasion of osteoblasts and blood vessels into the cartilage template. At the NB stage, a few COL2-positive chondrocytes appeared at the ectopic cartilage. We also observed a small number of COL1-positive cells in the ectopic cartilage at NB (Fig. 3H, yellow arrowhead). The numbers of COL2-positive chondrocytes and COL1-positive cells were higher at P3 than the NB stage. The number of COL2-positive chondrocytes became lower at P6, while the number of COL1-positive osteoblasts became higher (Fig. 3H). We further detected COL10-positive cells, a marker for hypertrophic chondrocytes, and CD31-positive cells, a marker for vascular endothelial cells, at P6 (Fig. 3I). Taken together, these data strongly suggest that ectopic cartilage is formed only in sutures that cause premature suture fusion and contribute to premature suture fusion by prompting endochondral ossification.
Cranial NCSCs with enhanced BMP signaling showed higher capacity for chondrogenic differentiation Based on the ectopic cartilage formation in the sutures of mutant mice, we hypothesized that cranial NCCs with enhanced BMP signaling would show greater chondrogenic potential. To examine the chondrogenic differentiation potential of cranial NCCs, we isolated cranial NCCs from E8.5 mouse embryos to establish NCSC lines according to a previously published protocol. (11) The cells were isolated from either control or P0-Cre;caBmpr1a embryos labeled with an R26R mTmG reporter and GFP-positive cells were purified at passage 3 using FACS (Fig. 4A). We found that the cell lines established expressed neural crest markers, Snail1 and Twist1, and progenitor cell markers, Nestin, CD44, and Sca-1, at 18 passages (Fig. 4B) by RT-PCR analysis listed in Table S5. The cells were capable of differentiating into osteoblasts and chondrocytes, assessed by ALP staining and safranin O staining (Fig. 4C). We found that cranial NCCs from about 40% of embryos could grow over 10 passages regardless of their genotypes (Table S6). We further used 3D chondrogenic pellet cultures to examine the chondrogenic differentiation potential of NCSC lines from control and P0-Cre;caBmpr1a embryos. After 14 days of chondrogenic induction, NCSCs from P0-Cre;caBm-pr1a embryos demonstrated more cartilage matrix deposition and larger pellet size than controls by Alcian blue staining (Fig. 4D) and higher expression levels of chondrogenic markers Sox9, Col2a1, Acan, and Col10 by quantitative RT-PCR (Fig. 4E). These results from NCSCs strongly suggest that cranial NCCs in P0-Cre;caBmpr1a mice have a higher capacity for chondrogenic differentiation.
It is well known that BMP signaling promotes osteogenic differentiation. Thus, we further examined whether BMP signaling affected the differentiation capability of osteogenic differentiation. We isolated the cells of nasal process (NP) cells at E11.5 then successively induced osteogenic and chondrogenic differentiation (Fig. 5A). The differentiation was assessed by ALP staining, Alcian blue staining, and quantitative RT-PCR for osteogenic and chondrogenic marker genes. Chondrogenic differentiation was higher in P0-Cre;caBmpr1a mice, as expected, and there was a lower osteogenic differentiation tendency in P0-Cre;caBm-pr1a mice in vitro (Fig. 5B). Frontal bones were hypomorphic in P0-Cre;caBmpr1a mice at E15.5 (Fig. 5C, left). Taken together with a wider gap between the frontal bones at NB-P2 (Fig. 3B), these results suggest that cranial NCCs in P0-Cre;caBmpr1a mice have a lower capacity for osteogenic differentiation. Of note, cartilage formation in P0-Cre;caBmpr1a mice was comparable to controls at E15.5 (Fig. 5C, right), suggesting that the impact of augmented BMP signaling in cell fate specification in suture progenitor cells becomes visible in later stages.
Cranial NCCs marked by Wnt1-Cre and P0-Cre mice behave differently during early embryonic stage Augmentation of BMP signaling in NCCs at an early embryonic stage alters cell fate toward a chondrogenic fate that develops ectopic cartilage. However, there is still a knowledge gap as to why the naso-premaxillary suture prematurely fuses in P0-Cre; caBmpr1a mice but not in Wnt1-Cre;caBmpr1a mice, even though both Cre lines target NCCs. We then hypothesized that Cremediated recombination patterns at early embryonic stages between Wnt1-Cre and P0-Cre mice are similar but not identical, which results in region-specific premature suture fusions between two lines. To examine the differences in recombination patterns in neural crest (NC) derivatives between Wnt1-Cre and P0-Cre mice, we used R26-LacZ mice to observe Cre recombination patterns by LacZ staining (hereafter, Wnt1-Cre;LacZ, and P0-Cre;LacZ). At E16.5, there were no overt differences in LacZ activities within the suture mesenchyme at the nasal suture, the naso-premaxillary suture, and the AF suture (Fig. 6A,B). To detect small differences, we next employed tdTomato Crereporter mice (P0-Cre;tdTomato and Wnt1-Cre;tdTomato mice) to observe the spatial distribution of red fluorescence at early embryonic stages (Fig. 6C). At E12.5, tdTomato fluorescence localized in the face in Wnt1-Cre and P0-Cre mice. P0-Cre mice at E12.5 showed more intense tdTomato fluorescence than Wnt1-Cre mice in the area that includes the presumptive nasopremaxillary suture (Fig. 6C, arrows). Histological observation at E12.5 confirmed this notion (Fig. 6C, right). These small differences persisted at the three stages we observed (Fig. 6C). Wnt1-Cre mice at 12.5 had tdTomato fluorescence at the hindbrain, which is known to be non-NC-derived (Fig. 6C, arrowheads). (25,43,44) We and others have reported unique and overlapping patterns of P0-Cre expressing and Wnt1-Cre expressing cells in cranial NCCs at early embryonic stages when they start to emerge. (25) Our result may suggest that cranial NCCs labeled by P0-Cre are more involved in developing sutures in the anterior part of the face than those labeled by Wnt1-Cre in a temporal-specific manner before E16.5, which leads to the premature suture fusion of naso-premaxillary sutures only in P0-Cre mice.
To address the remaining question of why the region-specific premature suture fusion patterns developed even though all cranial NCCs at the craniofacial region were similarly recombined, we then hypothesized that cranial NCCs in the facial area at the early embryonic stage might behave differently by BMP signaling between Wnt1-Cre and P0-Cre mice. The origin of cranial NCCs in the cranial sutures remains unclear; however, some reports have shown that cranial NCCs at the first branchial arch and the MNP migrate above the developing eye, then contribute to suture development. (45,46) We thus focused on the MNP at an early embryonic stage. Most cranial NCCs in the MNP in Wnt1-Cre;LacZ and P0-Cre;LacZ mice were positive for LacZ (Fig. 7A). BMP-SMAD signaling was significantly increased in the MNP of Wnt1-cre;caBmpr1a and P0-Cre;caBmpr1a mice at E10.5 than the MNP of control mice (Fig. 7B,D). We have reported that the augmentation of BMP signaling in cranial NCCs caused enhanced cell death. (47) Cell death detected with TUNEL staining was significantly increased in Wnt1-Cre;caBmpr1a and P0-Cre; caBmpr1a mice at E10.5 compared with controls; furthermore, P0-Cre;caBmpr1a mice showed more cell death than Wnt1-Cre; caBmpr1a mice (Fig. 7C,E). It is interesting that BMP-SMAD signaling levels were similarly upregulated in both mutant mice, while impacts on cell death were different (Fig. 7D,E). These results suggest that the difference we reported in Cre expressions during the emergence of cranial NCCs (E8-E9) (25) may impact cell survivability at E10.5, which will subsequently affect the behavior of NCC-derived cells at perinatal stages to differentially contribute suture fusions.

Discussion
Excessive osteogenic differentiation of suture mesenchymal cells has long been considered the reason for premature suture fusion resulting in craniosynostosis. However, some reports have  demonstrated the presence of ectopic cartilage in sutures before their pathologic premature fusion. (35,48) Here we demonstrated that cranial NCCs with enhanced BMP signaling developed ectopic cartilage in sutures preceding premature fusion. Our two transgenic mice lines, Wnt1-Cre;caBmpr1a and P0-Cre;caBm-pr1a, showed common and specific premature suture fusion patterns, which coincide with the patterns of ectopic cartilage formation found in each mouse line. The ectopic cartilage is replaced by bone nodules, suggesting that the ectopic cartilage prompts endochondral ossification, which is one of the mechanisms of premature suture fusion. Our histological observations at early embryonic stages, together with the fact that NCSCs isolated from P0-Cre;caBmpr1a mice showed robust chondrogenic ability, imply that augmentation of BMP signaling at early embryonic stages alters the behaviors of cranial NCCs toward chondrogenic differentiation. BMPs are known as bone inducers (49,50) ; we found hypomorphic calvarial bone formation in both Wnt1-Cre;caBmpr1a and P0-Cre;caBmpr1a mice. Previously, we reported that augmented BMP signaling in cranial neural crest cells resulted in p53-dependent cell death (47) and that in osteoblasts it does not affect bone mass. (51) Together with new findings in this report, we speculate that BMP signaling in neural crest progenitors positively affects their commitment to chondrogenic differentiation and BMP signaling in cells committed to the osteogenic lineage does not affect their bone-forming activity but rather prompts p53-dependent cell death, leading to hypomorphic bone formation in the mutant mice we investigated in this report.
Endochondral ossification in cranial suture mesenchyme is a potential reason for premature suture fusion Cranial NCCs are multipotent cells differentiating into osteoblasts, chondrocytes, and others (52) ; however, cranial NCCs in calvaria do not differentiate into chondrocytes under physiological condition. The PF suture fuses in normal development, followed by cartilage formation in the area of the suture. (31) We recently reported that augmentation of BMP signaling in cranial NCCs alters cell fate toward the chondrogenic lineage at early embryonic stages. (41) In the present report, we showed a significant increase in the cell death of cranial NCCs from P0-Cre;caBmpr1a mice compared with Wnt1-Cre;caBmpr1a mice in the facial primordia (Fig. 7C,E) at much earlier stages than perinatal stages when ectopic cartilage formed (Fig. 3). This indicates that the changes in the behavior of cranial NCCs at early embryonic stages may influence chondrogenic differentiation at later stages. Such large temporal differences support the idea that a mechanism of ectopic cartilage formation may be due to cell fate switching toward chondrogenic lineage rather than prompting chondrogenic differentiation of committed progenitors.
Wnt1-Cre and P0-Cre label different subpopulations of cranial NCCs Cranial NCCs develop the face and anterior part of the head. (4) We have shown that Wnt1-Cre and P0-Cre markers overlap but with unique domains of NCCs at the stage of NCC formation (E8-E9); more Wnt1-Cre expressing cells are found in the midbrain region while more P0-Cre expressing cells are found in the hindbrain region. (25) Moreover, our study also demonstrated that Wnt1-Cre expressing cells appear at five somite stages and P0-Cre expressing cells appear at 11 somite stages, (25) suggesting that Wnt1-Cre expressing cells and P0-Cre expressing cells appear at different places and with different timings. It is known that NCCs have subpopulations, mainly cranial NCCs, cardiac NCCs, and trunk NCCs, and they are assigned by where they emerge. (52) Also, their differentiation potentials are different. It has been reported that cranial NCCs at the forebrain mainly give rise to frontal and nasal bone, including AF suture, while those at the midbrain mainly give rise to maxillary and dentary bones, including nasal-maxillary sutures. (52) Therefore, it is possible that cranial NCCs from different positions, depicted by Wnt1-Cre and P0-Cre recombination patterns, have a different character over time, resulting in premature suture fusion in different places through a mosaic activation pattern of BMP signaling. Fig. 7. Effect of augmented BMP signaling in NCCs at early embryonic stage. (A) Whole LacZ staining and histological LacZ staining for medial nasal process (MNP) at E10.5 (n = 6). Counterstaining for nuclei was performed with nuclear fast red staining. (B) Activation of BMP-SMAD signaling in MNP and lateral nasal process (LNP) was analyzed by phospho-SMAD 1/5/9 staining (n = 6). (C) Cell death in MNP was revealed by TUNEL staining (n = 6). (D, E) Number of positive cells for p-SMAD1/5/9 or TUNEL in each transgenic mouse was quantified. One-way ANOVA with Tukey's test was used for statistical analysis. NS, no significance; * p < 0.05, ** p < 0.01. Scale bars: 100 μm.
In summary, we propose that the augmentation of BMP signaling develops ectopic cartilage prompting ectopic endochondral ossification in the calvaria, ultimately resulting in premature suture fusion and a craniosynostosis phenotype. It is still unclear whether inhibition of ectopic endochondral ossification in the sutures can rescue premature suture fusion. It is an important future effort to identify the critical time window that enhanced BMP signaling in cranial NCCs influences cell fate specification. Although there are some gaps between the translation from findings in mouse models into clinical research, we hope these findings will contribute to defining the molecular cellular mechanisms that might contribute to the etiology of craniosynostosis, eventually leading to better treatment options for pathological conditions.