Craniosynostosis, the premature fusion of one or more sutures between the calvarial bones, is a common birth defect. Mutations in genes encoding receptors for the transforming growth factor-beta (TGFβ) family of signaling molecules have been associated with craniosynostosis, but how TGFβ signaling is regulated during suture development is not known. In the present study, we found that expression of Smad2 and Smad3, intracellular mediators of canonical TGFβ signaling, gradually increases during early postnatal suture development in rat in both the coronal suture (CS), which remains patent throughout life, and the posterior frontal suture (PFS), which undergoes programmed closure by postnatal day 22. The amounts of phosphorylated Smad2 and Smad3 proteins showed a similar gradual increase in the PFS and CS, but in the CS, Smad2/3 activation was suppressed after neonatal day 10. The suppression of Smad2/3 activation in the CS correlated with upregulation of Smad7 expression. We demonstrate that siRNA-mediated knockdown of Smad7 caused increased phosphorylation of Smad2 and Smad3 and induced osseous obliteration of the CS from postnatal days 10 to 22. The Smad7 siRNA-induced CS closure was associated with significantly increased levels of Fgf10 and phosphorylated ERK1/2 in the suture mesenchyme. Moreover, addition of the Erk1/2 inhibitor U0126 partially blocked Smad7-siRNA–induced CS closure. These findings suggest that canonical TGFβ signaling induces suture closure at least in part through activation of FGF and ERK signaling and that Smad7 plays an important role in maintaining suture patency by suppressing canonical TGFβ signaling during suture development. © 2014 American Society for Bone and Mineral Research.
Cranial sutures are major sites of bone growth during postnatal craniofacial development. Abnormal suture development can result in premature fusion of one or more of these sutures, a common birth defect known as craniosynostosis, which is often associated with uncoordinated compensatory growth and deformity of the calvaria, orbits, face, and dental malocclusion.[2-4] Although a large number of genes have been linked to bone development and/or craniofacial growth, the molecular mechanisms regulating cranial suture development remain largely elusive.[5, 6]
Several laboratories have used rodent models to investigate the morphogenetic and molecular mechanisms regulating cranial suture development. In Sprague-Dawley rats, the posterior frontal suture (PFS) between the frontal bones undergoes physiological fusion by the weaning age, whereas all other cranial sutures remain patent throughout life. By surgically separating the bone and suture from the underlying dura mater in experimental Sprague-Dawley rats, Roth and colleagues demonstrated that PFS fusion depended on signaling interactions between the suture and dura mater. Levine and colleagues investigated the effect of surgically rotating the rat PFS and sagittal suture with respect to the underlying dura and showed that regional differences in the dura mater determined the whether the overlying suture underwent fusion or remained patent. Opperman and colleagues showed that rat coronal suture (CS) cultured in vitro requires soluble, heparin-binding factors present in the dura mater to resist osseous obliteration.[10-12] Subsequently, several authors detected increased immunoreactivity for members of the transforming growth factor-beta (TGFβ) family in the fusing rat PFS.[13, 14] Increased expression of TGFβ1 and TGFβ2 was also detected in prematurely fusing human CS in comparison with patent sutures. More recently, mutations in the TGFβ receptor genes TGFBR1 and TGFBR2 have been associated with the Loeys-Dietz syndrome, an autosomal dominant aortic aneurysm syndrome of which many patients were found to have craniosynostosis phenotypes.[16, 17] These mutations have been shown to increase cellular response to TGFβ, indicating that increase in TGFβ signaling is a shared mechanism for physiological PFS fusion in rodents and pathological cranial suture fusion in humans.[16, 17] The molecular mechanisms regulating TGFβ signaling during cranial suture development and fusion remain largely uncharacterized, however.
Smad proteins are the principal intracellular mediators of TGFβ signaling.[18-20] Eight Smads have been identified in mammalian cells, with Smads 1, 5, and 8 mediating bone morphogenetic protein (BMP) signaling and with Smads 2 and 3 mediating TGFβ signaling.[18, 21] Typically, upon TGFβ binding to the type II receptor (TbRII), it recruits and phosphrylates the type I receptor (TbRI). The activated receptor complex then phosphorylates Smad2 and/or Smad3, causing a conformational change that facilitates their interaction with Smad4 and translocation into the nucleus where they activate or inhibit the transcription of TGFβ-responsive genes. Smads 6 and 7 are known as inhibitory Smads, with Smad6 specifically regulating BMP signaling while Smad7 functions as a general antagonist of TGFβ signaling. A recent study examined expression of TGFβ-responsive Smads in cranial suture development in rats and found that Smads 2, 3, and 4 are all present in the suture and that Smad2/4 activity decreased in areas where presumptive bone will form. Although various biochemical mechanisms involving Smad7 in the modulation of TGFβ signaling have been studied, whether Smad7 plays a role in regulating TGFβ signaling during cranial suture development is not known. To gain insight into the role and regulation of TGFβ signaling in development of cranial sutures and in the etiology of premature cranial suture fusion, we compared the expression patterns of Smad2, Smad3, Smad4 and Smad7 in the developing rat CS and PFS by using quantitative real-time RT-PCR and western blot assays. We found that physiological fusion of the PFS was associated with increased levels of phosphorylated Smad2/3 whereas persistent patency of CS was associated with suppression of Smad2/3 phosphorylation concomitant with increased Smad7 expression. Moreover, we found that siRNA knockdown of Smad7 expression in the CS induced suture closure, indicating a critical physiological role for Smad7 in maintaining suture patency.
Materials and Methods
Harvesting and preparation of tissues
Sprague-Dawley rats were obtained from Laboratory Animal Center of Sichuan University (Chengdu, China) and housed in a temperature- and moisture-controlled animal facility. All animals were handled in accordance with guidelines of the Chancellor's Animal Research Committee of the Office for Protection of Research Subjects at our institution. Pregnant females (day of mating plug = day 0) and neonates at days N1, N5, N10, N15, and N22 (day of birth = day N0) were euthanized by carbon dioxide asphyxiation overdose followed by cervical dislocation. Immediately after euthanization, the scalp was removed before dissection of the calvaria. Calvaria were removed and trimmed such that the frontal and parietal bones and incorporated frontonasal, sagittal, and coronal sutures remained intact (see Fig. 1A for a schematic diagram of the cranial sutures). Calvaria to be used for examination of PFS were dissected through the coronal sutures, separating the frontal and parietal bones. Calvaria to be used for examination of CS were dissected through the sagittal and frontonasal sutures, separating left and right frontal and parietal bones. Equivalent areas of the PFS-dura complex and the CS-dura complex were isolated, including osteogenic fronts, intervening mesenchyme, overlying periosteum, and the underlying dura mater (Fig. 1B). The tissue was handled only with RNase-free instruments, avoiding direct contact with sutures and underlying dura to minimize mRNA degradation. Half of harvested tissues were immediately snap-frozen in a sterile tube floating in liquid nitrogen and then frozen at –80°C, and the other half were fixed by 4% paraformaldehyde overnight and demineralized in 0.5 M ethylenediaminetetra-acetic acid (EDTA) for 15 days at 4°C before they were embedded in paraffin.
Construction of a recombinant lentivirus vector expressing siRNA for Smad7
The Lenti-X shRNA expression system (Clontech, Mountain View, CA, USA) was used for the construction of the lentiviral expression construct according to the manufacturer's instructions. The design of siRNA for Smad7 (GCAGCCTAACCAGACCTTT) strictly followed the previously described design rules. The cDNA oligos were annealed and ligated into the BamHI/EcoRI-digested pLVX-shRNA1 vector (Clontech). A control vector was constructed using a 19-nucleotide sequence (AACCTGCGGGAAGAAGTGG) with no significant homology to any mammalian or mouse gene by BLAST, serving as a negative control. The recombinant vectors were purified and cotransfected with Lenti-X HT Packaging Mix (Clontech) into HEK293T packaging cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). The virus-containing cell culture supernatants were collected 48 hours after transfection, passed through a 0.45-µm filter, and stored at –80°C. The virus titration was determined using puromycin selection following the manufacturer's protocol.
Western blot analysis
Suture tissue protein was extracted using the Protein Extraction Kit (KeyGEN, Nanjing, China). Protein concentration was determined using BCA Protein Assay Reagent Kit (Pierce, Thermo Scientific, Rockford, IL, USA). Equal amounts of protein (30 µg/lane) for each sample were analyzed by electrophoresis under reduced conditions on 10% acrylamide gels and transferred to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, CA, USA). To block nonspecific binding, the membrane was incubated in Tris-buffered saline (TBS) with 0.1% Tween-20 (TBST) containing 5% nonfat milk for 2 hours. Subsequently, the membrane was incubated for 2 hours with antibodies against Smad2, Smad3, Smad4, Smad7, pSmad2, pSmad3, FGF10, p-ERK1/2, osteocalcin, caspase3, and caspase7 (Santa Cruz Biotechnology, Santa Cruz, CA, USA, 1:200), respectively, in TBST containing 5% nonfat milk. After washing in TBST, the membrane was incubated for 1 hour with horseradish peroxidase-conjugated secondary antibody (1:10,000) (Santa Cruz Biotechnology) in TBST containing 2% bovine serum albumin (BSA). Each membrane was also probed with an anti-GAPDH antibody (Santa Cruz Biotechnology) as a loading control. Bands were scanned using a densitometer (GS-700, Bio-Rad), and quantification was performed using Quantity One 4.6.3 software (Bio-Rad).
Quantitative real-time RT-PCR
Total RNA was isolated from frozen tissues using the TRIzol reagent (Invitrogen) according to the manufacturer's instructions. To avoid DNA contamination, total RNA was treated with RNase-free DNase I (Takara, Kyoto, Japan) for 60 minutes at 37°C, and extracted with the TRIzol reagent again. The purity of total RNA was determined by the 260/280 absorbance ratio, and the RNA integrity by the intensity of the 28S and 18S rRNA bands after formaldehyde agarose gel electrophoresis. Two micrograms of total RNA was subjected to reverse transcription using RevertAid First-Strand cDNA Synthesis Kit (Thermo Scientific, Pittsburgh, PA, USA) with random hexamer primers. One microliter of each cDNA sample was used for real-time PCR in a 25-µL reaction using TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA) on ABI PRISM 7500 (Applied Biosystems). PCR reactions were run in triplicate, and results were averaged. Quantification of RT-PCR products was normalized against the level of glyceraldehyde 3-phosphatase dehydrogenase (GAPDH) expression. The sequences of PCR primers and Taqman probes used in this study were synthesized commercially (Invitrogen) and are shown in Table 1. The relative expression level of the genes was calculated using the 2-ΔΔCt (Livak) method.
|Primer (5′-3′)||Taqman Probe (5′-3′)|
Subcutaneous injection of siRNA
Twenty-four rats were divided into three groups. The same volume (0.1 mL) of saline solution (Mock) (n = 8), control vector (siRNA-neg) (n = 8), or Smad7-specific siRNA (siRNA-S7) (n = 8) were injected subcutaneously into the area near the CS at day N10 and repeated every 5 days for a total of three injections per animal. All injected rats were euthanized at N22 and processed for micro-computed tomography (CT) and histological analysis. After the findings of siRNA-S7–induced CS fusion, 16 additional rats were injected subcutaneously into the area near the CS with siRNA-S7 alone (n = 8) or siRNA-S7 plus the Erk pathway inhibitor U0126 (Calbiochem, San Diego, CA, USA) at 0.1 mg/kg body weight (n = 8) at day N10, N15, and N20. The injected rats were euthanized at N22 and processed for micro-CT and histological analysis.
Micro-CT analysis of the skull samples
Skulls were scanned using a high-resolution micro-CT scanner (µCT80, Scanco Medical, Bassersdorf, Switzerland) and reconstructed with an isotropic voxel size of 10 µm. Micro-CT data were collected at 50 kV and 145 µA and reconstructed using the cone-beam algorithm supplied with the micro-CT scanner by Scanco (thresholds = 180).
For histological examination, paraffin-embedded suture samples were sectioned at 6-µm thickness, stained with hematoxylin and eosin (H&E), and observed under a bright-field microscope. Histomorphometric measurements were conducted as described previously. The width of the CS were measured in 10 serial sections from each sample.
Apoptotic cells in suture tissues were detected by using TUNEL assay using the In Situ Cell Death Detection Kit (Roche Applied Science, Basel, Switzerland) according to the manufacturer's protocol. The apoptotic cells display the dark brown nuclei, karyopyknosis, and conglomeration. TUNEL-positive cells were quantified in three sections per animal and five different fields (not including ossification area) per section by three pathologists under a bright-field microscope.
Data are expressed as mean ± SD when normally distributed. The statistical significance of differences was determined by Student's two-tailed t test in two groups and one-way ANOVA in multiple groups. One-way ANOVA was conducted to assess differences among different groups. Student-Newman-Keuls (SNK) was applied for multiple comparisons. A p value <0.05 was considered statistically significant. All data were analyzed with SPSS 13.0 software (SPSS, Inc., Chicago, IL, USA).
Expression of TGFβ pathway Smads in normal suture development
During normal growth of the rat cranial vault bones, both the CS and PFS are initiated at fetal day 19 (F19). In the first 2 weeks after birth, both CS and PFS remain unossified to allow continued expansion of the cranial vault during early postnatal development. By postnatal day 22 (N22), the PFS becomes obliterated by bone, whereas the CS remains patent throughout the life of the animal.[7, 14] To investigate the role of Smad-mediated TGFβ signaling in suture development, we first compared the expression patterns of Smad2, Smad3, p-Smad2, p-Smad3, Smad4, and Smad7 proteins in the developing CS and PFS from F19 to N22 by western blot analyses. In CS, the expression of Smad2 and Smad3 increased gradually from F19 to N22, whereas expression of Smad4 remained unchanged. However, the levels of p-Smad2 and p-Smad3 declined from N5 to N22 (Fig. 2A). Concomitantly, the levels of Smad7 increased dramatically from N5 to N22 in the CS (Fig. 2A). In PFS, the expression patterns of Smad2, Smad3, and Smad4 proteins were similar to those in CS from F19 to N22, but the expression levels of Smad7 did not change significantly (Fig. 2B). The levels of both p-Smad2 and p-Smad3 increased significantly from F19 to N22 in the developing PFS (Fig. 2B).
The mRNA levels of Smad2, Smad3, Smad4, and Smad7 in the developing CS and PFS were analyzed by using quantitative real-time RT-PCR. The mRNA expression levels of Smad2, Smad3, Smad4, and Smad7 were consistent with the protein expression levels (Fig. 2 C, D).
SiRNA knockdown of Smad7 in the CS
Because the recombinant lentivirus vectors express green fluorescent protein (GFP) independently of the siRNA constructs, we first monitored GFP expression to examine the delivery efficiency of the gene-silencing construct. The results demonstrated that the lentivirus vectors could efficiently infect the cells in the suture and surrounding tissues in siRNA-neg and siRNA-S7 groups, whereas no GFP-positive cells were observed in the Mock control group (Fig. 3A–C).
To verify whether Smad7 mRNA expression in the CS was downregulated by siRNA, we analyzed the mRNA levels in CS tissues in Mock, siRNA-neg, and siRNA-S7 groups by quantitative real-time RT-PCR, with GAPDH as internal control. The results showed that suture tissues transfected with siRNA-S7 had significantly reduced Smad7 mRNA levels (knockdown by 74%) compared with those in Mock group (Fig. 3D). Western blot analyses showed that the Smad7 protein levels in the siRNA-S7–infected CS tissues were reduced by 71% compared with the Mock group (Fig. 3E).
Smad7 knockdown induced CS obliteration accompanied by significantly increased p-Smad2 and p-Smad3 levels
Micro-CT analysis showed that the CS underwent osseous fusion in all siRNA-S7–treated rats (n = 7), whereas none of the Mock-treated (n = 8) or siRNA-neg–treated (n = 8) rats showed CS fusion (Fig. 4A–C). To quantify the extent of CS fusion in the siRNA-S7–treated rats, we examined serial histological sections through the CS of the rat samples from all three groups and measured the width of CS. A significant decrease in CS width was observed in the siRNA-S7 group (11.2 ± 2.3 µm) compared with the Mock (72.4 ± 13.4 µm) and siRNA-neg (69.3 ± 11.7 µm) groups (Fig. 4D–G). Consistent with the osseous fusion phenotype, western blot analysis showed that the level of osteocalcin, an osteoblast differentiation marker, is significantly increased in the siRNA-S7–treated CS compared with the control samples (Fig. 5).
To investigate the effect of Smad7 knockdown on TGFβ signaling, we compared the levels of p-Smad2 and p-Smad3 in the CS of siRNA-S7 and the control groups. Although the siRNA-neg vector had no effect on p-Smad2 and p-Smad3 levels, siRNA-S7 caused a significant increase in the amounts of p-Smad2 and p-Smad3 in the CS (Fig. 5). These data, together with our finding that the levels of p-Smad2 and p-Smad3 in the developing CS is suppressed in association with increased endogenous Smad7 expression, indicate that Smad7 is required for maintenance of CS patency and functions at least in part through inhibition of TGFβ signaling in the developing suture.
Smad7 knockdown caused increase in expression of Fgf10 and p-ERK1/2 and decrease in apoptosis in the CS mesenchyme
It was shown previously that TGFβ2-induced suture closure was nearly completely inhibited by an ERK kinase inhibitor. The levels of p-ERK1/2 were indeed significantly increased in the siRNA-S7–treated CS compared with the Mock and siRNA-neg–treated samples (Fig. 5). Because ERK1/2 phosphorylation is also directly regulated by FGF signaling and gain-of-function of FGF receptors have been associated with craniosynostosis in humans and mice, we investigated whether Smad7 knockdown also affected the FGF signaling pathway. As shown in Fig. 5, we found that FGF10 expression was significantly increased in the siRNA-S7–treated CS mesenchyme compared with the control samples (Fig. 5).
To investigate whether siRNA-S7 induction of CS was mediated by activation of Erk signaling, we compared the effects of siRNA-S7 alone with that of siRNA-S7 plus a specific inhibitor of Erk kinase, U0126, on CS patency. As shown in Fig. 6, co-injection of U0126 partially inhibited siRNA-S7–induced CS fusion. Examination of serial histological sections showed that the width of CS was significantly increased in the U0126 group (42.3 ± 8.3 µm) compared with the siRNA-S7 (11.7 ± 2.5 µm) group (Fig. 6E). These data suggest that maintenance of suture patency by Smad7 involves preventing activation of Erk signaling during suture development.
We further analyzed the effect of Smad7 knockdown on suture cell survival. The number of TUNEL-positive apoptotic cells was significantly decreased in the siRNA-S7–treated CS compared with that in the Mock and siRNA-neg groups (Fig. 7A–D). Western blot analysis showed that the levels of Caspase3 and Caspase7 were significantly reduced in the CS suture in siRNA-S7–treated rats compared with those in Mock and siRNA-neg–treated animals (Fig. 7E). These data suggest that Smad7 normally maintains suture patency in part by stimulating suture mesenchyme apoptosis.
Although increased TGFβ activity has been associated with PFS fusion in rodents[10-12] and gain-of-function mutations in TGFBR1 and TGFBR2 have been associated with craniosynostosis in humans,[16, 17] the molecular mechanisms involving TGFβ signaling in cranial suture development have not been well characterized. In this report, we demonstrate that both Smad2 and Smad3 are increased in expression as well as in activation by phosphorylation during PFS fusion in rats. Remarkably, we found that whereas the expression levels of both Smad2 and Smad3 were also increased in the developing CS from N1 to N22, the levels of p-Smad2 and p-Smad3 in the CS were decreased from N5 to N22. These data suggest that PFS fusion is induced by canonical TGFβ signaling and that maintenance of CS patency involves active suppression of this pathway.
In the canonical Smad pathways, Smad7 acts as a competitive inhibitor of receptor-activated Smad2/3 phosphorylation by forming stable complexes with activated type I receptors. Smad7 also recruits the HECT type of E3 ubiquitin ligases, Smurf1 and Smurf2. When Smurfs bind to Smad7, they induce the nuclear export of Smad7, which translocates into the cytoplasm in response to TGFβ and recruits the ubiquitin ligases to the activated type I receptor, leading to the degradation of the active receptors through the proteasomal pathway and inhibition of TGFβ signaling. In present study, we found that, although Smad2, 3, and 4 had similar expression patterns in the developing CS and PFS, Smad7 expression was increased dramatically in the CS but not in the PFS at the key period of PFS fusion (N10–N22). We found that the levels of p-Smad2 and p-Smad3 were significantly increased in the CS after Smad7 knockdown, indicating that Smad7 maintains suture patency through inhibition of Smad2/3 activation.
In addition to TGFβ signaling, the FGF signaling pathway plays critical roles in skull vault development. Gain-of-function mutations in FGF receptors are the most common cause of syndromic craniosynostosis in humans. Apert syndrome, characterized by a severe form of craniosynostosis, is caused by the Ser252Trp(S252W) or Pro253Arg(P253R) in FGFR2. FGF10 is an important mesenchymal Fgf ligand that can activate FGFR2. Genetic knockdown of FGF10 expression rescued the craniofacial defects in an Apert syndrome mouse model. Recently, Hosokawa and colleagues demonstrated that FGF10 acts downstream of TGFβ signaling in regulating tongue muscle development. We found that FGF10 expression was upregulated in the siRNA-S7–treated rat CS, suggesting that FGF10 may also act downstream of TGFβ signaling in induction of CS fusion in the siRNA-S7–treated rats.
Whereas FGF signaling is known to regulate cell fate specification during mammalian development through the Erk-MAPK pathway, a previous study suggested that Erk1/2 also acts downstream of TGFβ signaling during suture development. TGFβ2-induced suture closure was nearly completely inhibited by an Erk kinase inhibitor. We found that the levels of p-Erk1/2 were elevated after Smad7 knockdown in the CS. Co-injection of the Erk kinase inhibitor U0126 partially inhibited siRNA-S7–induced CS fusion. These results indicate that CS fusion induced by Smad7 knockout is mediated at least in part through activation of Erk signaling. Further studies are necessary to elucidate the mechanisms regulating Smad7 expression and function during suture development.
All authors state that they have no conflicts of interest.
This study was supported by a grant from National Nature Science Foundation of China (No. 81070859).
Authors' roles: Study design: HZ, RJ, and JH. Study conduct: HZ, SZ, and WF. Data collection: HZ, SZ, and WF. Data analyses: HZ, SZ, YL, and WF. Drafting manuscript: HZ, RJ, and JH. Revising manuscript content: SZ, WF, YL, RJ, and JH. Approving final version of manuscript: HZ, SZ, YL, RJ, WF, and JH.