Conditional expression of Spry1 in neural crest causes craniofacial and cardiac defects
© Yang et al; licensee BioMed Central Ltd. 2010
Received: 20 November 2009
Accepted: 11 May 2010
Published: 11 May 2010
Skip to main content
© Yang et al; licensee BioMed Central Ltd. 2010
Received: 20 November 2009
Accepted: 11 May 2010
Published: 11 May 2010
Growth factors and their receptors are mediators of organogenesis and must be tightly regulated in a temporal and spatial manner for proper tissue morphogenesis. Intracellular regulators of growth factor signaling pathways provide an additional level of control. Members of the Sprouty family negatively regulate receptor tyrosine kinase pathways in several developmental contexts. To gain insight into the role of Spry1 in neural crest development, we analyzed the developmental effects of conditional expression of Spry1 in neural crest-derived tissues.
Here we report that conditional expression of Spry1 in neural crest cells causes defects in craniofacial and cardiac development in mice. Spry1;Wnt1-Cre embryos die perinatally and exhibit facial clefting, cleft palate, cardiac and cranial nerve defects. These defects appear to be the result of decreased proliferation and increased apoptosis of neural crest and neural crest-derived cell populations. In addition, the domains of expression of several key transcription factors important to normal craniofacial and cardiac development including AP2, Msx2, Dlx5, and Dlx6 were reduced in Spry1;Wnt1-Cre transgenic embryos.
Collectively, these data suggest that Spry1 is an important regulator of craniofacial and cardiac morphogenesis and perturbations in Spry1 levels may contribute to congenital disorders involving tissues of neural crest origin.
Neural crest cells (NCC) are pleuripotent cells that migrate out of the dorsal neural tube during early vertebrate embryogenesis to populate many anatomical structures along the dorsoventral axis [1, 2]. Cranial NCC migrate ventrolaterally from the forebrain and hindbrain region to populate craniofacial structures and branchial arches. The proliferation of cranial NCC results in a demarcation of each branchial arch. Once migration is complete, cranial NCC contribute to the maxilla, mandible, cranial ganglia, and other mesenchymally derived structures of the head and neck. Cardiac NCC emanating from rhombomeres 6-8 populate branchial arches 3, 4, and 6. Some cardiac NCC contributes to the development of the branchial arch arteries, cardiac outflow tract, and the spiral septum between the ascending aorta and the main pulmonary artery. Other cardiac NCC contribute to the formation of the outflow tract cushions/endocardial cushions and subsequently the semilunar valves and interventricular septum. Perturbations in normal neural crest development cause several congenital craniofacial and cardiac defects.
Cell-cell and tissue interactions are required for proper patterning of neural crest-derived structures. Several growth factors are important to NCC formation, migration, and differentiation, including members of the FGF family and their receptors [1, 2]. The identification of mutations in fibroblast growth receptors (FGFRs) that cause several craniosynostosis syndromes indicates a role for FGF signaling in the skeletogenic differentiation of NCC [3, 4]. Furthermore, NCC proliferate, migrate, and differentiate into cartilage and bone in vitro in response to FGF2 [5, 6]. In addition, tissue-specific deletion of FGF8 demonstrated a requirement for FGF8 in NCC cell survival and patterning of the first branchial arch . A hypomorphic allele of Fgfr1 has been used to demonstrate that FGFR1 is required for NCC migration into the second branchial arch . Mice carrying this allele showed severe abnormalities of the craniofacial bones and cartilage. These and other studies show that FGF signaling is important to craniofacial development and that gene dosage in components of the FGF pathway is important to normal craniofacial development.
Sprouty (Spry) was originally identified in Drosophila as a negative regulator of FGF signaling in tracheal development . Subsequently, Sprouty was demonstrated to inhibit EGF signaling in Drosophila eye development [10, 11]. In vertebrates, there are four Sprouty proteins that either inhibit or potentiate receptor tyrosine kinase (RTK) signaling in a context specific manner [12, 13]. For example, Spry2 can potentiate EGFR signaling by binding to c-Cbl and sequestering it away from the EGFR, thus preventing EGFR down regulation and degradation, consequently leading to sustained EGFR activation, and enhanced ERK signaling. Conversely, Spry2 inhibits ERK activation mediated by FGFR signaling. Thus, Spry proteins exhibit differential effects depending upon the cellular context.
During vertebrate development, Spry proteins exhibit overlapping patterns of expression, particularly in craniofacial structures and limb buds . Gene targeting studies have revealed both distinct and redundant functions for Spry proteins during development. Targeted deletion of Spry2 results in defects of inner ear and in tooth development [15, 16]. Deletion of Spry1 results in defects in kidney development where supernumerary branching of the ureteric buds occurs resulting in multiple ureters . Spry4 null mice show defects in development of the mandible, polydactyly, and small size . Mice that are null for both Spry2 and Spry4 alleles exhibit very severe craniofacial defects and dwarfism . In addition, mice homozygous for a 1 MB deletion of chromosome 14, a region that encompasses the Spry2 gene, exhibited cleft palate and cleft lip of variable penetrance . Interestingly, a mouse carrying a Spry2-BAC transgene rescued the cleft palate defect. However, the Spry2-BAC transgenic line expressed Spry2 at reduced levels suggesting that palate development is Spry2 dosage sensitive .
Due to the complex nature of Spry function and the possible redundancies during development, we developed a conditional Spry1 transgenic mouse. To investigate the role of Spry1 in regulating NCC during development, we induced tissue-specific expression of Spry1 using Cre/loxP recombination in the neural crest lineage by using Wnt1-Cre transgenic mice . Our study shows that Spry1 expression in Wnt1-expressing neural crest cells in vivo results in facial clefting, cleft plate, failure of formation of the nasal and frontal bones as well as cardiovascular defects including ventricular septal defects, and outflow tract defects. Mutant embryos also exhibited hypoplastic thyroid, thymus, and cranial ganglia. Spry1 expression in NCC cells resulted in decreased proliferation and increased apoptosis. We conclude that Spry1 is a regulator of NCC cell proliferation and survival and that this occurs in both NCC cells and NCC-derived mesodermal cells that result in craniofacial and cardiac structures.
In situ hybridization and β-gal staining patterns of Spry1 suggests that it plays a role in the development neural crest derived sutures. To investigate this further we used transgenic mice with a floxed mSpry1 transgene, which efficiently undergoes Cre-mediated recombination as we have previously demonstrated . To enable tissue specific expression in neural crest, we crossed CAGGFP-Spry1 transgenic females with male Wnt1-Cre transgenic mice. Bitransgenic mouse embryos were designated Spry1;Wnt1-Cre and were confirmed by genotyping for GFP and Cre by PCR . Littermates carrying the Spry1 transgene, but lacking the Wnt1-Cre transgene served as controls in these studies. All Spry1;Wnt1-Cre mutant embryos died at birth and exhibited severe craniofacial defects (data not shown). Prenatal lethality was not observed. Attempts to show transgenic protein expression by immunohistochemistry proved difficult due to issues of sensitivity and background of the anti-myc antibody used to detect the epitope tagged transgenic protein. However, we have previously demonstrated Spry1 transgenic protein expression using the same transgenic mice when crossed with other transgenic Cre driver strains . qPCR analysis revealed expression levels of the transgene to be 2-8 fold above endogenous expression levels (Yang, data not shown).
The homeobox genes Dlx5 and Dlx6 play important roles in craniofacial and limb development . Mice that are null for both Dlx5 and Dlx6 exhibit severe craniofacial, axial, and appendicular skeletal abnormalities, resulting in perinatal lethality. Whole mount in situ hybridization of E10.5 Spry1;Wnt1-Cre embryos show that domains of expression of Dlx5 in the first and second branchial arches are greatly reduced. Similarly, Dlx6 expression domains were reduced in the first and second branchial arches. Dlx5 and Dlx6 expression in the limbs was variable but often reduced.
Our previous studies have revealed an important role for Spry1 in endochondral bone formation and chondrogenesis . We undertook the present study to determine the role of Spry1 in craniofacial development. Previous studies using gene-targeting strategies revealed that targeted deletion of Spry1  did not produce a craniofacial phenotype, and deletion of Spry2 produced defects in the inner ear  and dentition , however deletion of Spry2 and Spry4 results in several abnormalities including facial clefting and limb defects . In furtherance of these studies, we used Spry1 transgenic mice to gain additional insight into the role of Spry1 in craniofacial development. To gain insight into the role of Spry1 in development of NCC-derived structures, we used a floxed transgenic allele of Spry1 , and induced its expression in NCC by Cre-mediated recombination driven by the Wnt1 promoter . Mutant embryos died perinatally from multiple defects including severe facial clefting and cardiovascular defects including persistent truncus arteriosus and ventricular septation defects. We also observed hypoplasia of the thymus and thyroid glands (Kilgallen and Friesel, data not shown). Our results with Spry1;Wnt1-Cre embryos are consistent with insufficient neural crest-derived cells populations for normal craniofacial and cardiac morphogenesis. Our data indicate that increased apoptosis and decreased cell proliferation likely cause the NCC insufficiency. Although our analysis was performed at an embryonic stage where most cranial neural crest have emigrated from the neural tube, Wnt1-Cre mediated β-gal staining was still evident in Spry1;R26R;Wnt1-Cre embryos in the dorsal neural tube at this stage. This suggests that some residual neural crest cells or neural crest-derived cells remained in this region suggesting that decreased proliferation and increased apoptosis in the dorsal neural tube may be attributable to this population of β-gal positive cells.
In Spry1;R26R;Wnt1-Cre transgenic embryos β-gal positive cells were present in the branchial arches, and the cardiac region, however the number of β-gal positive cells, and the overall size of the branchial arches and neural crest-derived cardiac structures was greatly reduced. Immunostaining for phospho-histone H3, a marker of proliferation, and TUNEL assays revealed decreased proliferation and increased apoptosis respectively in the branchial arches of Spy1;Wnt1-Cre embryos but not their Cre-negative littermates suggesting that the increase in NCC apoptosis is a contributor to the observed craniofacial defects. Because fgf8 expression was essentially normal in Spry1;Wnt1-Cre embryos, fgf8 availability is likely not a factor in decreased NCC survival. Furthermore, we cannot rule out the possibility that the β-gal positive cells in the branchial arches and outflow tract did not undergo recombination of the Spry1 transgene, which may account for the β-gal positive cells that are present in this region. However, the increase in apoptosis and decrease in proliferation in this region suggest that Spry1 expression affected development of these structures possibly by affecting the reciprocal signaling between the NCC-derived mesoderm and the overlying ectoderm.
Palate development is a multistep process that involves the growth, elevation and midline fusion of the palatal shelves. The palatal shelves are comprised of NCC-derived ectomesenchyme and pharyngeal ectoderm [1, 20, 29]. The growth and development of the palate is controlled by several growth factors including members of the TGF-β family and members of the FGF family. Conditional loss-of-function of Tgfbr2 in NCC of Tgfbr2 fl/fl ; Wnt1-Cre mutant mice results in cleft palate . Mice carrying a large deletion of chromosome 14 (Pub36-/-), a region that contains the Spry2 gene exhibit cleft palate, excessive cell proliferation and up regulation of FGF target genes including Msx1, Etv5 and Barx1 . Interestingly, targeted disruption of Spry2 did not phenocopy the megabase deletion in chromosome 14; however a BAC Spry2 transgene expressing reduced levels of Spry2 completely rescued the facial clefting and cleft palate phenotype in Pub36-/- mice . These data suggest that palate development is sensitive to Spry2 gene dosage. Our data are consistent with this notion. Spry1 and Spry2 have overlapping domains of expression during development and current data suggest that they may be functionally redundant in regulating FGF signaling . Here we show that Spry1 over expression in neural crest derivatives partially phenocopies the palate defect of the Pub36-/- mutation. Together, these data suggest that normal palate development is in part dependent upon proper growth factor signaling thresholds, and that Spry1 and Spry2 play a key role in regulating these thresholds. Whether the roles of Spry1 and Spry2 are functionally redundant in palate development remains to be determined using tissue-specific loss-of-function approaches targeting two or more Spry family members in neural crest in vivo.
In addition to controlling palate development, TGFβ receptors and tyrosine kinase receptors (RTK) regulate the development of the calvarial bones of the skull that are derived from NCC [1, 6, 30–32]. Spry1;Wnt1-Cre mice show craniofacial and cardiac phenotypes that are very similar to Pdgfrα fl/fl ; Wnt1-Cre embryos including facial clefting, and aortic arch defects . The similarity in phenotypes between a loss-of-function Pdgfrα mutant and a gain-of-function Spry1 mutant are consistent with the notion that Spry 1 inhibits signaling downstream of RTKs. While the phenotypes of Spry1;Wnt1-Cre and Pdgfrα fl/fl ; Wnt1-Cre embryos are similar, Pdgfrα fl/fl ; Wnt1-Cre embryos did not show any changes in proliferation or apoptosis and the authors speculated that the defects were due to defects in NCC differentiation . Conversely, Spry1;Wnt1-Cre embryos showed decreased proliferation and increased apoptosis in NNC derived structures. While it is likely that increased proliferation and decreased apoptosis in NCC of Spry1;Wnt1-Cre embryos contributes to the phenotype, it is also possible that similar to the Pdgfrα fl/fl ; Wnt1-Cre, Spry1;Wnt1-Cre have defects in differentiation. Spry1;Wnt1-Cre mice also share phenotypic similarities to Alk5 fl/fl ; Wnt1-Cre mice including cardiac defects  and craniofacial defects including cleft palate . The Alk5 fl/fl ; Wnt1-Cre craniofacial defects are more severe in that they lack nasal and frontal bones and, parietal bones, whereas Spry1;Wnt1-Cre embryos lacked frontal and nasal bones but had nearly normal parietal bones. Whether Spry1 directly influences PDGFRα and Alk5 signaling in NCC directly will require further study.
Forced expression of Spry1 in Wnt1-expressing cells was also associated with defects in the development of cranial nerves including the glossopharyngeal nerve (IX) and the vagus nerve (X). Hypoplastic and patterning abnormalities of cranial nerves was revealed by immunostaining with neurofilament antibodies. Migrating Sox-10-expressing NCC contribute to cranial nerves IX and X, and these cells are reduced and their migration and guidance are defective in Hox3a-/-, Fbln1-/-, and Msx1-/-;Msx2-/- mice. It is likely the defects in cranial nerves in Spry1;Wnt1-Cre embryos are due to a combination of reduced NCC proliferation or survival or altered responses to local guidance cues due to forced expression of Spry1.
Spry1;Wnt1-Cre embryos die perinatally due to craniofacial and cardiac defects including persistent truncus arteriosus and aortic pulmonary trunk abnormalities. Fate mapping studies using Spry1;R26R;Wnt1-Cre mice show that NCC correctly migrate into the branchial arches. It is likely that NCC insufficiency due to decreased proliferation and increased apoptosis in this region is the cause for the failure of formation of the aorticopulmonary septum, resulting in an overriding truncus arteriosus and DORV.
Our results show that Spry1 is expressed in neural crest and neural crest derived craniofacial structures. Forced expression of Spry1 in Wnt1-Cre expressing cells resulted craniofacial and cardiac defects. Our data and that of others suggest that appropriate Spry1 levels are important to correct patterning of neural crest derived structures including bones of the face and the cardiac outflow tract. The similarity of the Spry1;Wnt1-Cre embryonic phenotype to the phenotypes of Pdgfrα fl/fl ; Wnt1-Cre are consistent with Spry1 inhibiting signaling downstream of RTKs. The similarity of the Spry1;Wnt1-Cre embryonic phenotype to that of Alk5 fl/fl ; Wnt1-Cre embryos suggests a possible interaction of Spry1 with the Alk5 pathway.
Wnt1-Cre transgenic mice and R26R reporter mice have been described previously . The generation of CAGGFP-Spry1 transgenic mice has been described elsewhere . Briefly, the mouse Spry1 open reading frame was tagged with a myc/his epitope and cloned into the CAG-loxP-GFP-loxP vector (gift of J. Yoon). Transgenic mice were generated by pronuclear injection of the linearized plasmid, and transgenic mice screened by PCR of genomic DNA with GFP specific primers. The resulting transgenic mice were designated CAGGFP-Spry1. This transgenic line was maintained on a FVB genetic background. For lineage tracer analysis, CAGGFP-Spry1 mice were crossed with R26R mice. Mice that were positive for both GFP and β-galactosidase were then crossed with Wnt1-Cre mice. These mice carry CNC cells labeled with β-galactosidase before CNC cells begin to migrate out of the neural tube . Additional Spry1 expression studies were performed on Spry1lacZ/+ mice were obtained from the Mutant Mouse Regional Resource, University of California, Davis, and recently described . Detection of β-galactosidase activity (β-gal) activity on whole embryos and tissue sections was carried by using standard procedures . To over express Spry1 in neural crest cells CAGGFP-Spry1 mice were crossed with Wnt1-Cre mice, and double transgenic mice were identified by PCR of genomic DNA from either tails or placenta using specific primer for GFP and Cre.
All mice were housed in a pathogen-free environment, under light, temperature, and humidity controlled conditions. The Maine Medical Center Research Institute Institutional Animal Care and Use Committee approved all procedures involving animals.
Skeletal preparations were performed as described . Briefly, timed pregnant females or newborn mice were euthanized by asphyxiation in CO2. Embryos and neonates were skinned, eviscerated, and fixed in 95% ethanol. The skeletons were stained with alcian blue, cleared in 1% KOH, and counterstained with alizarin red.
Magnetic resonance (MR) images were obtained with a BRUKER PharmaScan 7 T, 300 MHz scanner using a RARE 8 pulse sequence with the following parameters: TE 39.8 ms, TR 2571 ms, FOV 35 × 35 mm, Matrix 256 × 256, Slice 1 mm (total of 7 slices), 3 averages, total scan time 4 min 6 sec. Pregnant female mice were maintained under anesthesia using 2% isoflurane, a slightly higher percentage than used in other scans to minimize embryonic movement and to allow non-breathing gated image acquisition. The total anesthesia time was less than 30 min and the pregnant females recovered normally from the procedure. The orientation of the image slices was chosen such that two embryos could be imaged in the sagittal view.
For histological analysis, embryos were fixed in 4% paraformaldehyde, and were either embedded in OCT, and serial 7 μm-frozen sections were prepared, or embryos were embedded in paraffin and sectioned using standard procedures. For general morphology, deparaffinized sections were stained with hematoxylin and eosin using standard procedures.
For whole-mount in situ hybridization, plasmids were linearized with appropriate restriction enzyme; digoxygenin-labeled riboprobes were generated using a kit (Roche) according to manufacturer's protocol. Fgf8 probe was from P.H. Crossley, Msx1, Msx2, Dlx5, and Dlx6 were from Yang Chai, AP-2 was from Trevor Williams. In situ hybridization was processed according to established protocols. Briefly, embryos were washed with PTW (PBS + 0.1% Tween-20), treated with 10 μg/ml proteinase K briefly, prehybridized at 65°C, hybridized with indicated antisense probe at 65°C for overnight, then detected with alkaline phosphatase-conjugated anti-digoxygenin antibody (Roche), and developed with BM purple (Roche).
Mouse embryos at E9.5 or E10.5 were fixed in 4% paraformaldehyde for 10 mins, washed in PBS, and stained in X-gal solution at 37°C overnight. After staining, embryos were refixed and embedded in either OCT or paraffin; 7 μm cryostat sections were taken for microscopic analysis.
Embryos were collected at E9.5 or E10.5, fixed in 4% paraformaldehyde and embedded in OCT, and serial 7 μm sections were prepared for proliferation or apoptosis analysis. Immunofluorescent staining using anti-phosphor-Histone3 (Ser10) antibody (Upstate) was performed, phosphor-H3 positive cell was quantified and the proliferation rates were expressed as a percentage of total cells. For TUNEL labeling, the fluorescent in situ Cell Death Detection kit (Roche) was used according to the manufacturer's instructions, and the number of apoptotic cells per section was quantified.
The authors wish to thank Jeong Yoon, Leif Oxburgh, Yang Chai, and members of the Friesel lab for insightful comments and support during the course of this work. This work was supported by NIH grants R01DK73871, R01 HL65301, and COBRE grant P20RR15555 from the NCRR (to R.F). We also wish to thank the Kathleen Carrier and the histopathology laboratory supported by NIH grant P20RR018789 (D. Wojchowski, PI).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.