|Home | About | Journals | Submit | Contact Us | Français|
The mammalian skull vault consists mainly of 5 flat bones, the paired frontals and parietals, and the unpaired interparietal. All of these bones are formed by intramembranous ossification within a layer of mesenchyme, the skeletogenic membrane, located between the dermal mesenchyme and the meninges surrounding the brain. While the frontal bones are of neural crest in origin, the parietal bones arise from mesoderm. The present study is a characterization of frontal and parietal bones at their molecular level, aiming to highlight distinct differences between the neural crest-derived frontal and mesodermal-derived parietal bone. We performed a detailed comparative gene expression profile of FGF ligands and their receptors known to play crucial role in skeletogenesis. This analysis revealed that a differential expression pattern of the major FGF osteogenic molecules and their receptors exists between the neural crest-derived frontal bone and the paraxial mesoderm-derived parietal bone. Particularly, the expression of ligands such as Fgf-2, Fgf-9 and Fgf-18 was upregulated in frontal bone on embryonic day 17.5, postnatal day 1 and postnatal day 60 mice. Frontal bone also elaborated higher levels of Fgf receptor 1, 2 and 3 transcripts versus parietal bone. Taken together, these data suggest that the frontal bone is a domain with higher FGF-signaling competence than parietal bone.
The mammalian skull is the product of an evolutionary process during which 4 skeletal components of independent origin have progressively integrated into a structure of exquisite structural and functional complexity [Morriss-Kay, 2001]. The 4 components of the vertebrate skull are the cartilaginous neurocranium, cartilaginous viscerocranium, dermal skull roof and the sclerotomal occipital region.
The vertebrate dermal skull roof is an ancient structure formed from membrane bones that are evolutionarily derived from the protective dermal plates of early jawless fishes. The mammalian skull vault is constructed principally from 5 bones: the paired frontals and parietals, and the unpaired interparietal. These calvarial bones arise from 2 tissues, neural crest and mesoderm. The distinct contributions of each tissue to the skull have been well established by combining mice with a Wnt1-Cre construct and a conditional reporter gene, R26R [Chai et al., 2000; Jiang et al., 2002]. These studies have defined the pattern of cranial neural crest cell migration in mouse embryos and demonstrated that the frontal and squamosal bones are of neural crest origin, in contrast to the parietal and interparietal bone, which are of mesoderm origin; the unossified sutural mesenchyme between the parietal bones is also of neural crest origin.
Development of the normal skull vault requires mechanisms to ensure that both its morphology and its rate of growth are precisely matched to those of the developing brain. This precise relationship suggests that there are important tissue interactions between the brain and the skeletogenic membranes also involving the mesenchymal layers between them (the developing meninges).
A multitude of signaling molecules, as well as their respective receptors and downstream transcriptional factors, play in concert to regulate bone development [Karsenty, 2001; Karsenty and Wagner, 2002]. In particular, fibroblast growth factor (FGF) signaling has gained much attention for its major role in skeletogenesis, including calvarial osteogenesis [Muenke and Schell, 1995; Naski and Ornitz, 1998; Ornitz and Itoh, 2001; Marie, 2003]. FGF signaling is known to play a critical role in regulating proliferation and differentiation of osteoblasts and osteogenic precursors [Mansukhani et al., 2000; Kim et al., 2003; Marie, 2003; Fakhry et al., 2005; Quarto and Longaker, 2006]. Its role in bone biology has been highlighted by the skeletal phenotypes of the numerous FGF ligand and receptor transgenic mice [Coffin et al., 1995; Deng et al., 1996; Eswarakumar et al., 2002; Liu et al., 2002; Ohbayashi et al., 2002; Govindarajan and Overbeek, 2006; Hung et al., 2007]. Thus, delineating regional differences in osteogenic potential based on embryonic origin and determining the role of FGF signaling in imparting superior regenerative potential on calvarial tissue is of paramount importance. However, little has been done to investigate whether regional differences in the expression of signaling molecules occurs between frontal and parietal bones.
Indeed, obtaining a comprehensive understanding of unique cell biology and gene expression patterns of calvarial osteoblasts from dissimilar embryonic origins, and how those differences translate into variable potential for endogenous bone regeneration, will be critical to the optimization of cell-based skeletal tissue engineering strategies.
In this study, we have performed a detailed comparative gene expression profile of FGF ligands known to play crucial roles in skeletogenesis and their receptors. The analysis presented here reveals that a differential expression pattern of the major FGF osteogenic molecules and their receptors exists between the neural crest-derived frontal bone and the paraxial mesoderm-derived parietal bone. Particularly, the expression of ligands such Fgf-2, Fgf-9 and Fgf-18 is upregulated in frontal bone in embryonic day 17.5 (E17.5), postnatal day 1 (pN1) and postnatal day 60 (pN60) mice. Frontal bone also elaborated higher levels of Fgf receptors FgfR1,FgfR2 and FgfR3 transcripts versus parietal bone. Taken together, these results suggest that the neural crest-derived frontal bone is a highly activated FGF-signaling domain compared to parietal bone.
Thus, this gene profiling study begins to establish molecular differences between these 2 calvarial bones of different tissue origin. Moreover, the data indicate that differences in tissue origin may dictate a distinct FGF-signaling pattern between frontal and parietal bones.
All experiments using animals were performed in accordance with Stanford University Animal Care and Use Committee guidelines. CD-1 wild-type mice were purchased from Charles River Laboratories Inc. (Wilmington, Mass., USA). At birth, animals were considered 0.5 days old. Animals were housed in light- and temperature-controlled rooms and were given food and water ad libitum. After euthanizing the animals, skulls of E17.5, pN1 and pN60 mice were isolated using a stereo microscope.
Following sacrifice of animals as described above, the skulls were harvested and immediately fixed in fresh, chilled 4% phosphate-buffered paraformaldehyde overnight at 4°C followed by washing in phosphate-buffered saline-0.1% Tween-20 for 30 min.
For bony tissue detection, specimens were stained with 0.1% alizarin red and 3% potassium hydroxide overnight at room temperature. All specimens were cleared in 50% glycerol and 5% potassium hydroxide for 2–7 days. Specimens were analyzed under a dissecting microscope. Images were acquired using a Leica MZ16 equipped with a DC500 digital camera.
Fresh tissue was harvested from the frontal and parietal calvarial bones of E17.5, pN1 and pN60 mice (1 litter, on average 9 animals). The tissue was harvested from 3 independent littermates. After dissecting out the parietal and frontal bones meticulously to exclude dura mater, pericranium and suture-associated osteogenic fronts, the tissues were immediately snap frozen in liquid nitrogen. To extract RNA, the bone chips were mechanically homogenized in Trizol reagent using a pellet pestle motor (Kontes, Vineland, N.J., USA).
Total RNA was isolated by using Trizol reagent according to the manufacturer's instructions (Invitrogen, Carlsbad, Calif., USA). Purified and quantified RNA was treated with DNAse I (Ambion, Austin, Tex., USA) to clear genomic DNA. Two micrograms of total RNA was then reverse transcribed to cDNA using random hexamer primers (Invitrogen). Reverse transcription was performed for 1 h at 42°C, followed by incubation at 75°C for 5 min to inactivate reverse transcriptase. RNAse treatment was used to clear residual RNA. PCR were performed under the following conditions: 94°C for 5 min, 94°C for 30 s, annealing for 1 min (see table table11 for the annealing temperature), and 72°C for 1 min (25–30 cycles). Specific primers for the genes examined were designed based on their GenBank sequence. Primer sequence and PCR conditions for the Gapdh and Runx2 (Cbfa-1) genes have been previously described [Quarto and Longaker, 2006]. Primer sequences for the other genes are listed in table table1.1. For the densitometry analysis of RT-PCR, bands were scanned and quantified by using the ImageJ program 1.36b (National Institutes of Health, Bethesda, Md., USA). The densitometry results of each band were normalized to their respective loading control (Gapdh) and presented as percentage increase. The scanning was performed on 3 independent experiments (3 different littermates) and results are presented as means ± SE of the 3 independent experiments. Quantitative RT-PCR was performed as previously described [Quarto et al., 2005]. The results are presented as means ± SE of 2 independent experiments (online suppl. fig. 1, www.karger.com/doi/10.1159/000202789).
The results are presented as means ± SD of 2 or 3 independent experiments performed on different littermates. Statistical differences between themeans were examined by Student's t test. p < 0.05 was consideredstatistically significant.
Calvarial osteoblasts were harvested from the frontal and parietal bones of E17.5, pN1 and pN60 mice. Briefly, the frontal and parietal bones were removed under direct visualization using a dissecting microscope, taking great care to exclude suture-associated bone. The periosteum and dura mater were meticulously removed from the calvaria, and only nonsuture-associated frontal and parietal bone was dissected free from surrounding tissue. Calvarial bones were subsequently mechanically minced and subjected to enzymatic digestion in serum-free media. Frontal and parietal bones depletedof periosteum, suture mesenchyme and dura mater were digestedwith 0.2% dispase II and 0.1% collagenase A (Roche Diagnostics, Indianapolis, Ind., USA) in serum-free medium. The digestion was repeated 6 times for 10 min each for frontal and parietal bones from E17.5 and pN1 mice and 15 min each for pN60 mice, at 37°C in a waterbath shaker. The first 2 digestions were discarded. The latter4 digestions were pooled, pelleted and resuspended in α-minimal essential medium (α-MEM) supplemented with 10% FCS. First-passage osteoblasts were plated at equal density and cultured to subconfluence in growth media (α-MEM, 10% FCS, 100 IU/ml penicillin and streptomycin).
The growth rates of frontal and parietal calvarial osteoblasts were assessed by BrdU labeling assay. Frontal and parietal osteoblasts (at passage 1) were plated at 1,500 cells/well, in 96-multiwell cultureplates with flat bottoms (Corning, New York, N.Y., USA). Cells were washedtwice with sterile phosphate-buffered saline and starved in serum-free α-MEM overnight. BrdU incorporationwas carried out for 24 h (Roche Diagnostics) according to themanufacturer's instructions. Photometric detection was donewith an ELISA reader at 370 nm wavelength. The background wassubtracted when the resulting data were processed. Each time point was run in triplicate.
Mouse calvaria were harvested at pN7 and pN60. Samples were fixed in 10% buffered formalin overnight at 4°C, decalcified, processed for paraffin embedding and cut in 10-μm sections. Selected paraffin sections were chosen and heat antigen retrieval was performed. For proliferating cell nuclear antigen (PCNA) staining, the PCNA staining kit was used according to the manufacturer's instructions (Invitrogen). Biotinylated secondary anti-rabbit antibody was used at 1:200 dilutions. Immunostaining with normal (irrelevant) rabbit IgG (sc-2027;, Santa Cruz Biotechnology Inc., Santa Cruz, Calif., USA) did not detect any staining (data not shown). The slides were analyzed and photographed with a Zeiss Axioplan 2 microscope equipped with an Axiocam HRc digital camera.
The mammalian skull vault consists of 5 bones that mineralized directly in membrane: a pair of frontal bones, a pair of parietal bones and an interparietal bone; the interparietal bone later fuses with the endochondral component of the occipital bone to became the superior part of that bone. Figure Figure1a1a illustrates schematically the anatomy of skull vault and the neural and mesodermal origin of calvarial bones.
Alizarin red staining (fig. (fig.1b)1b) shows the mineralizing pattern of skull vault at E17.5. By E17.5, the frontal and parietal bones are mineralized in the lateral part of their future domains; interparietal mineralization can be observed as a narrow transverse line at some distance caudal to the parietal bones. Mineralization starts from the lateral areas. By pN1, mineralization of frontal and parietal bone is further progressed, both pairs of bone are almost completely stained with alizarin red (fig. (fig.1c).1c). Intense uniform staining is observed in pN60 skull vault, as a result of a completed mineralization (fig. (fig.1d1d).
As an initial approach to define potential molecular and biological differences between the 2 calvarial bones of different tissue origin, we analyzed the gene expression profile of FGF ligands known to play a major role in osteogenesis. RT-PCR was performed on fresh tissues harvested from embryos at E17.5. In order to analyze the endogenous expression of osteoblasts, tissues were depleted of dura mater and pericranium. We found substantially elevated levels of Fgf-2, Fgf-9 and Fgf-18 gene expression in the frontal bone relative to the parietal bone (fig. (fig.2a–c).2a–c). Under these PCR experimental conditions we did not detect any expression of Fgf-4 and Fgf-8 either in frontal or parietal bone (data not shown). Because FGF ligands such as Fgf-2 and Fgf-9 are known to trigger proliferation of osteoprogenitor cells [Fakhry et al., 2005; Quarto and Longaker, 2006], we investigated the expression level of Runx2, a marker of osteoprogenitor cells in frontal and parietal bones. Interestingly, embryonic frontal bone also elaborated higher levels of Runx2 (fig. (fig.2d).2d). The above results were further confirmed by quantitative real- time PCR analysis (online supplementary fig. 1, www.karger.com/doi/10.1159/000202789).
Next, we analyzed the Fgf-2, Fgf-9 and Fgf-18 gene profiles in newborn mice. Fresh tissue was harvested from pN1 mice and RT-PCR was performed as above. We found that also in pN1 frontal bone the expression of all 3 Fgf genes was higher compared to the parietal bone (fig. (fig.3a–c).3a–c). However, overall, the levels of Fgf-2 and Fgf-9 transcripts in pN1 frontal bonewere lower than that observed in embryonic frontal bone. Similarly to E17.5, Runx2 transcript was elevated in the frontal bone (fig. (fig.3d).3d). The above results were further confirmed by quantitative real-time PCR analysis (online supplementary fig. 1, www.karger.com/doi/10.1159/000202789).
The trend of Fgf-2 and Fgf-18 genes profile was maintained throughout adult organisms. We observed a quantitative differential expression pattern of Fgf-2 and Fgf-18, between adult frontal and parietal bones, with frontal bone having higher expression of Fgf-2 and Fgf-18 (fig. (fig.4a,4a, c). However, the threshold level of these genes was much lower than that observed either in embryonic or newborn mice. No Fgf-9 transcript was detected in pN60 frontal and parietal bones at least under these PCR experimental conditions (fig. (fig.4b).4b). Runx2 gene expression was also upregulated in the frontal bone of adult mice (fig. (fig.4d).4d). The above results were further confirmed by quantitative real-time PCR analysis (online supplementary fig. 1, www.karger.com/doi/10.1159/000202789).
Because the initial study revealed that the neural crest-derived frontal bone expressed elevated levels of FGF ligands compared to the mesoderm-derived parietal bone, we decided to investigate whether there were also regional differences in the expression of FgfRs. Again, fresh tissue was harvested from the frontal and parietal bones of E17.5 mice. RT-PCR analysis demonstrated elevated levels of FgfR1, FgfR2 and FgfR3 in frontal bone relative to parietal bone, paralleling observed differences in Fgf-2, Fgf-9 and Fgf-18 expression in frontal and parietal bone (fig. (fig.5a–c).5a–c). The expression of FgfR2c isoform was much more abundant than that of FgfR2b. In addition, absence of FgfR4 transcript was observed in both frontal and parietal bones at all time points analyzed (data not shown). The above results were further confirmed by quantitative real-time PCR analysis (online supplementary fig. 1, www.karger.com/doi/10.1159/000202789).
In frontal and parietal bones derived from pN1 mice the expression of FgfR1 and FgfR3 genes was similar to that observed in embryonic bones, both transcripts were upregulated in the frontal bone relative to parietal bone (fig. (fig.6a,6a, c). Elevated levels of FgfR2b isoform were detected in the frontal bone, while similar levels of FgfR2c isoform were observed in both frontal and parietal bones (fig. (fig.6b).6b). The above results were further confirmed by quantitative real-time PCR analysis (online supplementary fig. 1, www.karger.com/doi/10.1159/000202789).
A unique expression pattern was observed for FgfR1 gene in pN60 mice. In contrast with what was observed in E17.5 and pN1 mice, pN60 mice expressed higher levels of Fgfr1 transcript in the parietal rather than in the frontal bone (fig. (fig.7a).7a). In contrast, both FgfR2 isoforms were upregulated in the frontal bone (fig. (fig.7b).7b). Moreover, in parietal bone the expression level of FgfR2c isoform was markedly lower compared to that of pN1 parietal bone. Frontal bone also elaborated elevated levels of FgfR3 transcripts versus parietal bone (fig. (fig.7c).7c). The above results were further confirmed by quantitative real-time PCR analysis (online supplementary fig. 1, www.karger.com/doi/10.1159/000202789).
The higher expression of Runx2 observed in frontal bone compared to parietal bone would suggest the presence of a larger pool of osteoprogenitor cells in frontal bone. Because osteoprogenitor cells are less differentiated and therefore more proliferative, we reasoned to analyze the proliferation activity of frontal and parietal primary osteoblasts harvested from E17.5, pN1 and pN60 mice. Proliferation was evaluated by BrdU incorporation. After 24 h, osteoblasts of frontal bone origin demonstrated significantly greater incorporation of BrdU than osteoblasts of parietal origin (fig. (fig.8a).8a). Thus, frontal bone-derived osteoblasts from E17.5, pN1 and pN60 mice have superior proliferative activity than those from parietal bone (p < 0.01). In addition, immunohistochemistry analysis performed on frontal and parietal bones using anti-PCNA antibody also detected more proliferative osteoblasts in frontal than parietal bone (fig. (fig.8b).8b). Furthermore, PCNA staining was also observed in the periostium and dura mater of both, frontal and parietal bones of pN1 mice, as well as in the periostium of frontal bone of pN60 mice.
The mammalian skull vault has a mixed developmental tissue origin that reflects its heterogenous evolutionary origin. Indeed, this complex history has functional consequences for the pattern of skull vault growth, and for the genetic control of the tissue interactions that are responsible for maintaining a balance between osteogenic cell proliferation and differentiation. Studies using a transgenic mouse with a permanent neural crest cell lineage marker, Wnt1-Cre/R26R, have revealed that the frontal bones are neural crest derived and the parietal bones mesodermal.
The present study is a characterization of frontal and parietal bones at their molecular level. In order to define potential differences between these 2 calvarial bones, we have analyzed their gene expression pattern for the main Fgf ligands and receptors known to be important for the osteogenic differentiation program. The importance of FGF/FGFR signaling in human skull development has been revealed by the identification of mutations in FGFR genes in several craniosynostoses [Wilkie et al., 1995a, b] as well as by transgenic mouse models [Coffin et al., 1995; Deng et al., 1996; Eswarakumar et al., 2002; Liu et al., 2002; Ohbayashi et al., 2002; Govindarajan and Overbeek, 2006; Hung et al., 2007].
Taken together, the data gathered from this study strongly suggest that tissue origin translates in marked molecular differences between the neural crest-derived frontal and the mesodermal-derived parietal bone in terms of the expression levels of Fgfs/FgfRs genes. From embryonic stages (E17.5) through newborn and adult mice, frontal bone elaborates higher levels of Fgfs/FgfRs transcripts versus parietal bone. This observation is interesting and would suggest that the frontal bone is a domain with higher FGF-signaling competence than parietal bone. Our analysis also revealed that Runx-2 expression is higher in frontal bone compared to parietal bone at all time points investigated. The higher expression of Runx2 may reflect the presence of a larger pool of osteoprogenitor cells [Ducy et al., 1997] in the frontal bone due to the elevated levels of Fgfligands. Indeed, the higher proliferation activity observed in frontal bone-derived osteoblasts further supports this hypothesis. Several studies have indicated that Fgf ligands, such as Fgf-2 and Fgf-9, act as positive regulators of osteoprogenitor cells by triggering their proliferation, and therefore enriching a pool of osteoprogenitor cells [Kim et al., 2003; Fakhry et al., 2005; Quarto and Longaker, 2006]. The above observation leads us to speculate that the frontal bone may be endowed with higher potential for tissue repair due to the presence of a larger number of putative osteoprogenitor cells. This hypothesis is currently under investigation [Quarto et al., in preparation].
Mirroring the expression pattern of Fgf ligands, frontal bone also elaborates higher levels of FgfR transcripts than parietal bone. However, among the 3 stages analyzed, differences in the expression of FgfR1 gene profile were observed. Adult, pN60 mice elaborated higher levels of transcript in the parietal bone, while E17.5 and pN1 mice expressed more FgfR1 transcripts in the frontal bone.
Because our data indicated a consistent upregulation of Fgf-2, Fgf-9 and Fgf-18 as well as FgfR1, FgfR2 and FgfR3 genes expression in the neural crest-derived frontal bone, we reasoned to investigate whether the expression of other important osteogenic genes such as Bmp ligands and their receptors mirror that of Fgf/FgfRs in frontal and parietal bones. A preliminary analysis of Bmp-2 and Bmp-4, as well as receptors BmpRIA and BmpRIB indicated that the expression pattern of these osteogenic molecules in the frontal and parietal bones does not mirror that of Fgf/FgfRs. Bmp-2 was the only gene upregulated exclusively in frontal bone derived from pN1 mice, while elevated levels of BmpRIA transcript were detected in the parietal bone of pN60 mice. Overall no significant quantitative differences in the expression pattern of Bmp-4 and BmpRIB were observed between the frontal and parietal bones in E17.5, pN1 and pN60 mice (data not shown). Thus, the quantitative differential gene expression observed between the frontal and parietal bones appears unique to Fgf/FgfRs.
Previous studies performed by in situ hybridization have mainly focused on the expression pattern of FgfR genes during craniofacial bone development and cranial suture patterning in embryonic animals [Iseki et al., 1999; Rice et al., 2003]. To our knowledge, the study reported here represents the first comparative and quantitative analysis of the expression pattern of osteogenic Fgfs/ FgfRs genes, performed focusing on frontal and parietal bones in embryonic, newborn and adult mice. This study highlights distinct and sustained differences in the expression pattern of Fgfs/FgfRs genes between the 2 bones of different tissue origin. Indeed, the data suggest that embryonic tissue origin differences determine a distinct FGF-signaling pattern in frontal and parietal bones, and that the neural crest-derived frontal bone represents a more competent domain for the FGF signaling.
Quantitative analysis of Fgf ligands and Fgf receptors performed by real-time PCR. A standard curve method of quantitation was used to calculate expression of target genes relative to the house keeping geneGapdh. The data correlate with the above results obtained by semiquantitative RT-PCR showing the highest expression level of both Fgf ligands and Fgf receptors in frontal bone of E17.5, pN1 and pN60. The results are presented as the mean ±SE of two independent experiments. Statistical differences between the means are examined by Student's test. A. P value <0.05 was considered statistically significant.
This work was supported by NIH grants R01DE13194 and R01DE14526 to M.T.L. and by the Oak Foundation.