The authors have no conflict of interest
Targeted Expression of SHH Affects Chondrocyte Differentiation, Growth Plate Organization, and Sox9 Expression†
Article first published online: 12 JUL 2004
Copyright © 2004 ASBMR
Journal of Bone and Mineral Research
Volume 19, Issue 10, pages 1678–1688, October 2004
How to Cite
Tavella, S., Biticchi, R., Schito, A., Minina, E., Di Martino, D., Pagano, A., Vortkamp, A., Horton, W. A., Cancedda, R. and Garofalo, S. (2004), Targeted Expression of SHH Affects Chondrocyte Differentiation, Growth Plate Organization, and Sox9 Expression. J Bone Miner Res, 19: 1678–1688. doi: 10.1359/JBMR.040706
- Issue published online: 2 DEC 2009
- Article first published online: 12 JUL 2004
- Manuscript Accepted: 11 MAY 2004
- Manuscript Revised: 30 APR 2004
- Manuscript Received: 6 AUG 2003
- rodent animal model;
- growth plate;
- parathyroid hormone/parathyroid hormone-related protein;
The role of Hedgehogs (Hh) in murine skeletal development was studied by overexpressing human Sonic Hedgehog (SHH) in chondrocytes of transgenic mice using the collagen II promoter/enhancer. Overexpression caused a lethal craniorachischisis with major alterations in long bones because of defects in chondrocyte differentiation.
Introduction: Hedgehogs (Hhs) are a family of secreted polypeptides that play important roles in vertebrate development, controlling many critical steps of cell differentiation and patterning. Skeletal development is affected in many different ways by Hhs. Genetic defects and anomalies of Hhs signaling pathways cause severe abnormalities in the appendicular, axial, and cranial skeleton in man and other vertebrates.
Materials and Methods: Genetic manipulation of mouse embryos was used to study in vivo the function of SHH in skeletal development. By DNA microinjection into pronuclei of fertilized oocytes, we have generated transgenic mice that express SHH specifically in chondrocytes using the cartilage-specific collagen II promoter/enhancer. Transgenic skeletal development was studied at different embryonic stages by histology. The expression pattern of specific chondrocyte molecules was studied by immunohistochemistry and in situ hybridization.
Results: Transgenic mice died at birth with severe craniorachischisis and other skeletal defects in ribs, sternum, and long bones. Detailed analysis of long bones showed that chondrocyte differentiation was blocked at prehypertrophic stages, hindering endochondral ossification and trabecular bone formation, with specific defects in different limb segments. The growth plate was highly disorganized in the tibia and was completely absent in the femur and humerus, leading to skeletal elements entirely made of cartilage surrounded by a thin layer of bone. In this cartilage, chondrocytes maintained a columnar organization that was perpendicular to the bone longitudinal axis and directed toward its outer surface. The expression of SHH receptor, Patched-1 (Ptc1), was greatly increased in all cartilage, as well as the expression of parathyroid hormone-related protein (PTHrP) at the articular surface; while the expression of Indian Hedgehog (Ihh), another member of Hh family that controls the rate of chondrocyte maturation, was greatly reduced and restricted to the displaced chondrocyte columns. Transgenic mice also revealed the ability of SHH to upregulate the expression of Sox9, a major transcription factor implicated in chondrocyte-specific gene expression, in vivo and in vitro, acting through the proximal 6.8-kb-long Sox9 promoter.
Conclusion: Transgenic mice show that continuous expression of SHH in chondrocytes interferes with cell differentiation and growth plate organization and induces high levels and diffuse expression of Sox9 in cartilaginous bones.
The family of vertebrate genes related to the Drosophila gene “Hedgehog” consists of three known members: Sonic (Shh), Indian (Ihh), and Desert (Dhh) hedgehog. Like its Drosophila homolog,(1)Shh encodes a signal that is instrumental in patterning the early embryo, and it is expressed in Hensen's node, the floorplate of the neural tube, the limb bud, and throughout the notochord. It has been implicated as the key inductive signal in patterning of the ventral neural tube, the anterior-posterior limb axis, and the ventral somites. Ihh, another member of this family, is mainly expressed in differentiating chondrocytes, where it plays a central role in controlling the rate of chondrocyte differentiation and longitudinal bone growth in a regulatory loop with PTHrP.(2)
Dhh is expressed in the male gonad, where it plays an important role in Leydig cell fate and testis organogenesis.(3)
All secreted Hedgehog proteins (Hhs) can bind to a 12-span transmembrane protein designated Patched-1 (Ptc1).(4) Ptc1 does not work as a functional receptor, but acts as an inhibitor of another transmembrane protein called Smoothened (Smo), which is believed to be the functional receptor, belonging to the superfamily of G-protein-coupled seven-pass receptors.(5) The binding of Hhs to Ptc1 can relieve Smo of the repression and activate signaling, suggesting that Hh proteins induce target gene transcription by removing Smo inhibition rather than by direct receptor activation.(6)Smo derepression culminates in a transcription factor acting as a transcriptional repressor. This factor is called cubitus interruptus (ci) in Drosophila(7) and Gli in vertebrates.(8) There are three Gli genes in vertebrates, each with distinct transcriptional functions.
Several pieces of evidence from loss-of-function mice suggest an important role of Hhs molecules in skeletal development. Shh is essential for axis, limb, and digit formation as shown by the phenotype of knockout mice.(9)Ihh is critical for chondrocyte proliferation, differentiation, and for longitudinal growth of limbs, because homozygous gene inactivation in mice produces a lethal form of short-limbed dwarfism.(10) Furthermore, chondrocyte-specific removal of Smo activity by Cre-loxP recombination has generated mice with shorter limb bones and reduced chondrocyte proliferation rate.(11)
Gain-of-function experiments by retroviral infection of chick embryos have shown that Shh and Ihh act in a similar way during chick bone development and alter chondrocyte differentiation, leading to a faulty endochondral ossification.(2) Gain-of-function experiments in mice have not been very informative about the role of Hhs in skeletal development, because they have shown either embryonic lethality before skeletogenesis or minimal skeletal involvement because of restricted transgene expression.(12) Gain-of-function mice have been also generated by overexpression of either chick Ihh or a constitutively active Smo allele specifically in the cartilage using the bigenic UAS-Gal4 system.(11), (13) These studies showed that ectopic activation of Ihh signaling promotes chondrocyte proliferation differentially along the long axis of cartilaginous skeletal elements.(11)
To study the consequences of Hh gain-of-function in mouse skeletal development, we have overexpressed the SHH cDNA under the control of the tissue specific cis-element of the mouse α1(II) procollagen promoter/enhancer. The experiment was intended to clarify the mechanism of Hh control of skeletal development and the role of the Ihh-parathyroid hormone-related protein (PTHrP) regulatory loop in controlling the fate of chondrocytes. From these transgenic mice, we have discovered that Hh has an important role in controlling chondrocyte differentiation, growth plate organization, and Sox9 expression.
MATERIALS AND METHODS
Transgene construction and generation of transgenic mice
The full-length cDNA of the human SHH was cloned between the HindIII and XhoI sites of the mouse α1(II) procollagen promoter/enhancer plasmid as described in Figure 1A. The transgene was injected into the pronuclei of FVB fertilized eggs that were later implanted into CD1 pseudopregnant foster mothers.(14) Embryos were removed by cesarean section at 15.5 or 18.5 dpc. Transgenic embryos were identified by PCR, using the primers COL2XbaI (GenBank no. NT_039621): 5′-GGAGAGGGTCCAGCCCGAGCTAC-3′ and hSHH450 (GenBank no. NT_039300): 5′-TGAGTCATCAGCCTGT-CCGCTCCG-3′, and Southern blot of EcoRI-digested genomic placenta DNA hybridized with a transgene specific probe (data not shown). A Col2-LacZ reporter gene was co-injected in some experiments to facilitate the identification of transgenic embryos.
Transgene expression and quantitative PCR
Transgene expression was assessed by RT-PCR (Life Technologies) using total RNA from 15.5- and 18.5-dpc embryo limbs. The transgene-specific oligos were forward 5′-CTCGCAGCTCTACCAAATAGG-3′ and reverse 5′-TGGCACCTTCCAGGGTCAAGGAAG-3′. The level of Sox9 and Col2a1 mRNA was measured by real-time quantitative RT-PCR using the PE ABI PRISM 7700 Sequence Detection System (Applied Biosystems). Expression of the housekeeping gene GAPDH was measured in parallel as an endogenous control. The sequences of forward and reverse primers and of the TaqMan fluorogenic probes, as designed by the Primer Express 1.5 software, were Sox9 (GenBank no. NT_039521): forward 5′-CCGCATCTGCACAACGC-3′, reverse 5′-TCCTCCACGAAGGGTCTCTTC-3′, probe 5′-FAM-TCGTTCAGCAGCCTCCAGAGCTTGC-TAMRA-3′; GAPDH (GenBank no. NM_008084): forward 5′-TGTGTCCGTCGTGGATCTG-3′, reverse 5′-GATGCCTGCTTCACCACCTT-3′, probe 5′-TET-TGATGTCATCATACTTGGCAGGTTTCTCCA-TAMRA-3′; Col2a1 (isoform IIB; GenBank no. NT_039621): forward 5′-GCTGCTGACGCTGCTCATC-3′, reverse 5′-GTCCCTGAGGGCCAGGAG-3′, probe 5′-FAM-CTAATTTTCGGGCATCCTG-TAMRA-3′.
All probes were located at the junction between two exons. Relative transcript levels were determined from the relative standard curve constructed from stock cDNA dilutions and divided by the target quantity of the calibrator according to the manufacturer's instructions. Col2a1 primers and probe were designed to amplify only the chondrocyte specific isoform IIB.
Histology and immunohistochemistry
The embryos were fixed in 4% formaldehyde overnight and stored in 70% ethanol before paraffin embedding. For immunohistochemistry, deparaffinized sections were treated with methanol and 3% H2O2 to inhibit endogenous peroxidase activity. For immunofluorescence, limbs were briefly fixed in 4% PAF/sucrose and embedded in OCT before cryosectioning. Primary antibodies were X53, a monoclonal antibody anti-collagen X(15); CIIE8, a monoclonal antibody against collagen II (gifts of K Von der Mark); and anti-Sox9 antibody, a rabbit antisera, a kind gift of B de Crombrugghe.(16)
In situ hybridization
Serial sections were processed for radioactive in situ hybridization using [33P]-UTP labeled antisense riboprobes. Hybridization was carried out at 70°C in 50% formamide as previously described.(2) Hybridization with digoxigenin (DIG-UTP)-labeled RNA probes (0.5 ng/μl) was carried out overnight in the same hybridization solution in a moist chamber. The riboprobes were subsequently visualized by immersion in the DIG-alkaline phosphatase for 2 h at 37°C and developed. Probes for in situ hybridization were as follows: rPTHrP,(17) a 440-bp SmaI fragment cloned into pBSK, rPTHR,(18)mIhh,(19)mPtc1,(19) and mCol10a1.(20)
Cell culture, transfection, and luciferase assay
Primary culture of chondrocytes were established as described.(21) To measure chondrocyte number in the whole limb, we established primary cultures from an entire limb with a modified protocol that preserves chondrocyte differentiation. Limb tissues were digested with Pronase (2 mg/ml in PBS) for 30 minutes, followed by 4-h digestion with collagenase D (3 mg/ml in DMEM with pyruvate and l-cysteine); then, the mixed cell population was grown in DMEM medium for chondrocytes as described.(21) To measure chondrocyte number, the cells were plated on chamber slides, grown 3 days in culture, and, after fixation, tested for procollagen II expression by immunochemistry with anti-collagen II antibody.
Transfection was performed in 6-well plates using Fugene 6 (Roche). The human SHH and the mouse Ihh cDNA were subcloned in pCDNA3 vector (Invitrogen). CMV-renilla was used as control of transfection efficiency. The Sox9-luciferase plasmid was obtained from P Koopman.(22) Forty-eight hours after transfection, cells were harvested, and firefly luciferase and Renilla luciferase activities were measured by the Dual-Luciferase reporter assay system (Promega).
Forced expression of SHH in embryonic cartilage results in severe maturation defects
We have generated mice expressing SHH in cartilage. The full-length cDNA of the human SHH gene was inserted in a plasmid containing the tissue-specific regulatory region of the mouse α1(II) procollagen gene (Col2a1; Fig. 1A), including the first intron where a strong enhancer is located.(23) This element targets expression of the LacZ reporter gene specifically to notochord, sclerotome, and chondrocytes of transgenic mice, and it has been frequently used to generate transgenic mice.(24) Transgenic mice were identified by PCR using transgene-specific primers and Southern blot of EcoRI-digested genomic DNA hybridized with a transgene specific probe (data not shown). Transgene expression was determined by RT-PCR of limb RNA using transgene specific primers. It showed that transgene expression persisted from 15.5 dpc until death at birth (Fig. 1B).
The phenotype of the Col2-SHH transgenic mice was neonatally lethal because of severe craniorachischisis, a condition in which the entire axis and the cartilaginous base of the skull denoted as the chondrocranium remained open (Figs. 1C and 1D). Twenty-six transgenic embryos were generated of which 22 (84%) showed severe craniorachischisis. No transgenic mouse survived, so it was impossible to establish a mouse line.
In the skull, exoccipital and supraoccipital bones were missing (Fig. 1D, arrow), but other membranous bones seemed normal. The otic capsule and the ossicles were hyperplastic, whereas the mandible was hypoplastic.
The axial skeleton showed the most striking alteration. Besides the absence of the posterior arches and spinous processes, histology showed enlarged vertebral bodies with very large lateral processes fused to the body (Fig. 1E). Ganglia were displaced posteriorly, and nerves appeared compressed by cartilage overgrowth. The neural tube was completely closed, wrapped by meninges that were in direct contact with the thin skin layer (Figs. 1C and 1E). The ribs were almost entirely stained by Alizarin red (Figs. 1D and 1F). They seemed distorted and of increased and variable diameter, in some cases fused to each other and to the vertebrae (Fig. 1F). No costo-chondral junction was distinguishable. The sternum was composed entirely of cartilage and apparently split in half at the midline (Fig. 1F, arrow). All four ossification centers of the sternum were missing (Figs. 1D and 1F).
SHH inhibits formation of trabecular bone without altering periosteal and membranous bone formation
The appendicular skeleton also showed major defects. The length of long bones was near normal; however, their thickness was increased. The metaphyses and adjacent diaphyses of these transgenic tubular bones stained as bone with Alizarin red; however, they also stained as cartilage with Alcian blue (Fig. 2A). This double staining was most apparent in the humerus and femur, where it gave the impression of a layer of periosteal bone covering a core of cartilage. The elbow and knee (Figs. 2A and 2N) joints were fused, as were the carpal and tarsal bones, because of defects in joint cavitation (data not shown).
Previous experiments of Ihh overexpression by retroviral infection in developing tibia of chick embryos showed that bone formation is severely impaired.(2) Staining of the cleared Col2-SHH embryos suggested an abnormal cartilage-within-a-bone arrangement in the transgenic long bones. This arrangement was confirmed by histological sectioning, which revealed that the transgenic humerus (Fig. 2B, right) and femur (Fig. 2N) were comprised almost completely of cartilage surrounded by a layer of dense connective tissue that stained for Alizarin red, where it covered the would be diaphyses and metaphyses of the bones (Fig. 2C, right).
The cartilage occupying these areas was highly disorganized, without vertical columns of proliferating chondrocytes or a distinguishable hypertrophic zone (Fig. 2E). However, at the presumptive border of the epiphyses and would be diaphyses, there was an area of reduced staining that was depleted of cells and extracellular matrix that roughly corresponded to the location of the growth plate (Fig. 2E, arrow). Chondrocytes radiated from this location toward the bone surface to produce column-like clusters of chondrocytes perpendicular to the surface of the bone (Fig. 2F). In contrast, skeletal elements of the middle segment of the limbs (tibia, radius, and ulna) displayed a different organization with the presence of a disorganized growth plate containing some hypertrophic chondrocytes and formation of abnormal trabecular bone with hematopoietic marrow (Figs. 2B, 2C, and 2H).
The cartilage of transgenic femur and humerus was covered by an external layer of dense and broad tissue (Fig. 2N). Its increased thickness was caused by the formation of a periosteal bone in the tissue that directly abuts cartilage (Figs. 2O and 2P). Staining of sections of humerus and femur with Alizarin red and von Kossa and immunohistochemistry with anti-tenascin and osteocalcin-specific antibodies (data not shown) clearly showed that this collar was mineralized and closely resembled periosteal bone of the comparable bones from wildtype embryos (Figs. 2I, 2L, and 2M). This peculiar bone organization explains why long bones in cleared specimens were strongly stained by Alizarin red, although they were internally composed entirely of cartilage.
The discrepancy between the chondrocyte differentiation defect in elements of the upper segment of the limbs (stylopod) and those of the middle segment (zeugopod) was consistent and observed in both forelimbs and hindlimbs of all transgenic mice with craniorachischisis. To better define this discrepancy, we compared the morphology of the two segments during limb development. In Col2-SHH 15.5-dpc embryos, blood vessel invasion, cartilage matrix degradation, and hypertrophic cartilage replacement with bone were highly defective in the stylopod (femur and humerus) because bone was made of a homogeneous cartilage template (Figs. 2S and 2T). However, at the same developmental stages, the skeletal components of the zeugopod (tibia, radius, and ulna) showed signs of maturation, with a certain degree of hypertrophy and vascular invasion (Fig. 2T, arrow), leading to formation of irregular growth plates and mineralized, but defective, trabecular bone. Consequently, forced chondrocyte expression of SHH-affected limb segments differentially after 15 dpc, with a more severe delay of hypertrophy in the stylopod where trabecular bone and growth plate were missing and chondrocytes changed their organization, forming columns of cells rotated of 90°.
Forced expression of SHH disrupts chondrocyte hypertrophy and cartilage growth plate polarity
Chondrocytes in the growth plate undergo a complex maturation process characterized by the expression of stage-specific molecules. Collagen II is an early marker of cartilage differentiation expressed in all cartilages of the developing skeleton starting at 10.5 dpc. Collagen X is expressed starting at 14.5 dpc in the restricted subpopulation of chondrocytes undergoing hypertrophic changes.
Immunohistochemistry revealed that the continuous expression of SHH in chondrocytes arrested bone formation, and the persisting cartilage was characterized by abundant collagen II expression (Fig. 3B) in the entire femur, even at 18.5 dpc, when collagen II expression is normally restricted to epiphyseal cartilage (Fig. 3A). Specific antibodies directed against collagen X, a marker of chondrocyte hypertrophy, showed that, in transgenic mice, this molecule was present only in a peculiar localization extending laterally from the would be growth plate (Figs. 3D and 3E). In situ hybridization confirmed this data, showing that, in contrast to its typical localization to cells in the hypertrophic zone (Fig. 3F), collagen X mRNA was identified in a subpopulation of chondrocytes radiating out from the would be growth plate toward the perichondrium (Figs. 3G and 3H). However, these cells did not resemble hypertrophic chondrocytes in size or shape and had the ability to form short columns of no more than 20 differentiating cells perpendicular to the longitudinal axis of the bone (Fig. 3E). The axis of these columns was thus rotated 90° from that typical of growth plate cells.
SHH expression alters the Ihh/PTHrP regulatory feedback loop
Ihh and PTHrP are thought to interact in a regulatory feedback loop that controls the rate of chondrocyte maturation and hypertrophic differentiation.(2) Ihh is normally expressed in the prehypertrophic region of growth plate cartilage (Fig. 4C); its receptor Ptc1 shows a wider distribution in the growth plate and in the perichondrium (Fig. 4A). Ihh induces the expression of PTHrP at the periarticular cartilage. From there, PTHrP acts, through its receptor throughout the growth plate, and in particular, in the prehypertrophic zone, to prevent hypertrophic differentiation and to negatively regulate Ihh expression.
In situ hybridization showed that SHH chondrocyte expression induces abundant expression of its receptor, Ptc1 (Fig. 4B), which largely overlaps transgenic SHH expression (data not shown) and extends over all cartilage. Forced activation of the SHH signal pathway through Ptc1 interferes with the Ihh-PTHrP regulatory loop, downregulating Ihh expression and changing the expression pattern to a restricted region that largely overlaps with the zone of columns of collagen X-expressing chondrocytes (Fig. 4D, arrows). PTHrP expression was greatly increased only at the articular surface (Fig. 4F, arrows). Expression of PTHrP receptor (PTHR) mRNA was very weak in transgenic cartilage, and it was evident in the perichondrium, bone collar, and collagen X- and Ihh-expressing chondrocytes (Fig. 4H, arrows).
Sox9 is upregulated in vitro and in vivo by SHH through cis-elements located in the 6.8-kb promoter
The phenotype of Col2-SHH long bones is characterized by the humerus and femur being composed entirely of cartilage extracellular matrix and chondrocytes. This observation led us to investigate how SHH could promote the formation of a larger cartilage template. Sox9 is believed to be an essential master gene for the proper formation of the cartilaginous skeleton, and its mutations in humans and mice significantly alter cartilage differentiation and skeletal development.(25) Therefore, we investigated whether SHH overexpression could directly affect Sox9 expression level in transgenic mouse chondrocytes.
The amount of Sox9 mRNA relative to GAPDH in whole limbs of 15.5-dpc embryos measured by real-time quantitative RT-PCR showed a 3-fold increase in transgenic compared with wildtype. This increase in Sox9 mRNA level in the whole limb of Col2-SHH mice could be only partially explained by the increase in chondrocyte number. Primary cultures of the contralateral transgenic limbs showed that the increase in the number of procollagen II-positive chondrocytes was only 1.6 ± 0.2-fold compared with control nontransgenic limbs. Therefore, we considered that the SHH expression in chondrocytes could promote higher Sox9 mRNA levels in the limbs, not only by an increase in chondrocyte number, but also by direct transcriptional activation. Accordingly, the amount of Sox9 mRNA in 15.5-dpc transgenic limbs, after normalization per chondrocyte number, was corrected to a statistically significant 1.8 ± 0.2-fold increase compared with wildtype controls, supporting the hypothesis that SHH has a direct role in increasing Sox9 expression per cell (Fig. 5A).
To evaluate more precisely the capacity of SHH to induce Sox9 mRNA, we established primary cultures of wildtype and transgenic chondrocytes and measured Sox9 level by quantitative RT-PCR at day 3 in culture in a homogeneous population of chondrocytes. Transgenic chondrocytes in culture showed a 2.1 ± 0.1-fold increase of Sox9 mRNA over the wildtype control (Fig. 5B). In the same chondrocyte cultures, we also found that the mRNA of Col2a1, a target gene downstream of Sox9, was increased 2.3 ± 0.1-fold over the level of wildtype (Fig. 5C).
To understand whether the ability of SHH to control Sox9 occurred at the transcriptional level, we used the CH310T1/2 cell line that expresses a baseline level of Sox9 mRNA and is unique in the ability to respond to SHH through the specific signaling pathway.(26) We transfected the cells with increasing amounts of the SHH expression plasmid and measured endogenous Sox9 mRNA by real-time quantitative RT-PCR. Cells that were stably (data not shown) or transiently transfected with SHH (Fig. 5D) and Ihh (data not shown) expression plasmids showed a dose response with upregulation of endogenous Sox9 mRNA level from 3.2- to 10.5-fold above the baseline level (Fig. 5D).
To gain further insight into this mechanism, we co-transfected the SHH expression plasmid with a Sox9 reporter construct in which a 6.8-kb-long Sox9 promoter drives a firefly luciferase reporter gene.(22) The experiment showed that Sox9-luciferase is upregulated up to 8.1-fold by increasing the amount of SHH plasmid (Fig. 5E). Therefore, the 6.8-kb promoter behaves as the endogenous Sox9 promoter and contains the cis-elements needed for SHH-mediated Sox9 induction.
At the protein level, Sox9 can be detected in all epiphyseal chondrocytes, with the highest level in the prehypertrophic zone (Figs. 6A and 6C). In Col2-SHH mice, anti-Sox9 specific antibody showed an intense and diffuse fluorescence in the nuclei of the entire humerus of 15.5-dpc transgenic embryos compared with wildtype littermates, (Figs. 6B and 6D). Fluorescence intensity was equally distributed over the whole humerus without areas of reduced or increased expression, in contrast to control mice, which lacked expression in the hypertrophic zone or bony diaphysis but displayed more intense signal in prehypertrophic chondrocytes (Figs. 6A and 6C). An increase in Sox9 level is expected to upregulate downstream target genes and to increase the respective protein level per cell. When we counted the number of collagen II-expressing chondrocytes in whole limb culture, we observed that, in transgenic chondrocytes, there was a more intense and diffuse immunochemical staining, suggesting a relative increase in the amount of intracellular procollagen II (Figs. 6E and 6F). Our experiments suggest that SHH can induce Sox9 expression both in vivo in transgenic limbs and in vitro in a mesenchymal cell line and that this is achieved by upregulation of Sox9 at transcriptional level.
SHH gain-of-function in mouse skeletal development
Hh gain-of-function experiments during skeletal development were previously performed by retroviral infection of the chick limb bud.(2) These experiments allowed detailed study of long bone (tibia) defects, but were unable to give a general view of the developmental defects affecting the entire skeleton. We show here that, in the mouse, the same chondrocyte defects are found mainly in the humerus and femur (stylopod), while the skeletal elements of the tibia and radius/ulna (zeugopod) are less affected. In addition, the SHH gain-of-function in the mouse chondrocyte affects joint formation, interfering with the delimitation of joint cavity, and consequently leading to fusion of knee and elbow articular surfaces. Intermediate and distal limb joints were more severely affected than proximal joints, in apparent discordance with the severity of bone anomalies. These topographical differences were not identified by localized retroviral infection in chick limb bud, although we cannot exclude that they may be species-specific. Conversely, it was not reported that Ihh chondrocyte gain-of-function mice show segment specific defects of long bones ossification nor that Ihh or constitutively active Smo can modify the expression of the endogenous Ihh, ColX, and PTHR genes or change the growth plate polarity.
Delayed chondrocyte differentiation in SHH gain-of-function mice
During endochondral ossification, Ihh and PTHrP have been shown to form a negative feedback loop regulating the onset of hypertrophic differentiation of chondrocytes. The bone morphogenetic protein (BMP) family of secreted factors regulating bone formation has been implicated as potential interactors of the Ihh/PTHrP feedback loop. However, BMPs do not act as a secondary signal of Ihh to induce PTHrP expression or to delay the onset of hypertrophic differentiation, being BMP signaling-independent of the Ihh/PTHrP pathway in long bones.(13)
In Col2-SHH transgenic mice, the Ihh-PTHrP loop is altered by continuous expression of SHH. Expression of SHH in chondrocytes can dissociate the regulation of these molecules, downregulating Ihh and upregulating PTHrP. Activation of Ptc1 in chondrocytes has two consequences. The expression of endogenous Ihh is inhibited by the negative feedback loop that attenuates hedgehog signaling in several known hedgehog systems.(27) PTHrP expression is increased, particularly at the articular surface, presumably using the Ihh signaling pathway, because SHH has an identical or even greater ability to bind Ptc1 and to activate Smo signaling. Therefore, extremely high levels of SHH and PTHrP are present in the developing transgenic cartilage, showing the delayed hypertrophic differentiation.
As a consequence of PTHrP overexpression, bone defects in Col2-SHH mice partially resemble the anomalies described in Col2-PTHrP transgenic mice that show a milder phenotype with temporary delay in chondrocyte differentiation and bone collar formation.(28) In Col2-PTHrP mice, hypertrophy occurred at the periphery of the developing long bones rather than in the middle, leading to a seeming reversal in the pattern of chondrocyte differentiation and ossification that was completely corrected in seven weeks, because the phenotype was not lethal. Besides, defects were found in all long bones,(28) without segmental differences as in Col2-SHH mice, and did not lead to the change of growth plate polarity that we have described.
Hedgehogs control Sox9 expression in chondrocytes
Much evidence suggests that SHH acts on paraxial mesoderm-inducing genes, such as Pax1, Pax9, and MFH1,(29) which control the number of sclerotome cells and genes that induce chondrocyte gene expression and maintain a committed state of chondrocyte precursors, such as Nkx3.2/Bapx1(30) and Sox9.(29) The concerted action of these two classes of transcription factors is needed to initiate and sustain the chondrogenic differentiation program. We show that Sox9 expression can be upregulated by SHH also in long bone chondrocytes and that cis-elements in the 6.8-kb Sox9 promoter are sufficient to elicit the SHH effect. However, it was not possible to establish whether Nkx3.2/Bapx1 is the mediator of this effect.
It is reported that PTHrP can increase Sox9 phosphorylation at a consensus protein kinase A phosphorylation site.(16) PTHrP can also increase the Sox9-dependent activity of chondrocyte specific Col2a1 enhancer in transient transfection experiments.(16) Sox9 phosphorylated at serine 181 (P-Sox9) was detected almost exclusively in chondrocytes of the prehypertrophic zone, coincident with the highest level of Sox9 protein. In preliminary experiments, we did not detect evidence of Sox9 phosphorylation in Col2-SHH transgenic mice using P-Sox9 specific antibody, notwithstanding elevated Sox9 levels in cartilage and PTHrP expression at articular surface. Sox9 protein distribution was also uniform in transgenic humerus, without any area of more intense staining. This is not surprising, because the expression of PTHR, which is normally detectable only in prehypertrophic chondrocytes, was greatly reduced and displaced in Col2-SHH mice. Therefore, we believe that increased amounts of unphosphorylated Sox9 can participate in the changes of chondrocyte differentiation, and it is directly dependent on SHH overexpression.
Chondrocyte-specific inactivation of Sox9 showed that when both Sox9 alleles were deleted, cells rapidly became hypertrophic, leading to the hypothesis that Sox9 prevents conversion of proliferating chondrocytes to hypertrophic chondrocytes.(25)Furthermore, an article by Akiyama et al.(25) mentions that unpublished experiments of Sox9 overexpression in transgenic mouse chondrocytes suggest that hypertrophic chondrocyte differentiation is delayed by Sox9. Consistent with this hypothesis, in Col2-SHH transgenic mice with high level of Sox9 expression, we have observed delayed chondrocyte differentiation. These higher Sox9 levels were present in Col2-SHH transgenic mice, and a 6.8-kb-long Sox9 promoter was sufficient for increasing Sox9 transcriptional levels. The identification of the promoter cis-elements that are implicated in the modulation of Sox9 expression will open the way to elucidate the molecules and the mechanism used by Hedgehogs to control cartilage formation and skeletal development.
We thank R Arbicò, R Gulli, E Noviello, and G Campanile for expert technical assistance. The financial support of Telethon-Italy (Grant D.112) is gratefully acknowledged. This work was partially supported by COFIN 2001 (RC), COFIN 2003 (SG), and European (ESA-ERISTO) and Italian (ASI) Space Agencies (RC).
- 11980 Mutations affecting segment number and polarity in Drosophila. Nature 287: 795–801.,
- 21996 Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 273: 613–622., , , , ,
- 32002 Desert Hedgehog/Patched 1 signaling specifies fetal Leydig cell fate in testis organogenesis. Genes Dev 16: 1433–1440., ,
- 41996 Biochemical evidence that patched is the Hedgehog receptor. Nature 384: 176–179., , , ,
- 51996 The Drosophila smoothened gene encodes a seven-pass membrane protein, a putative receptor for the hedgehog signal. Cell 86: 221–232., , , ,
- 61997 The smoothened gene and hedgehog signal transduction in Drosophila and vertebrate development. Cold Spring Harb Symp Quant Biol 62: 217–226., , , , , ,
- 71993 Genetic analysis of hedgehog signalling in the Drosophila embryo. Development (Suppl): 115–124., , ,
- 81996 Sonic hedgehog differentially regulates expression of GLI and GLI3 during limb development. Dev Biol 180: 273–283., , ,
- 91996 Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383: 407–413., , , , , ,
- 101999 Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev 13: 2072–2086., ,
- 112001 Genetic manipulation of hedgehog signaling in the endochondral skeleton reveals a direct role in the regulation of chondrocyte proliferation. Development 128: 5099–5108., , , ,
- 121993 Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell 75: 1417–1430., , , , , ,
- 132001 BMP and Ihh/PTHrP signaling interact to coordinate chondrocyte proliferation and differentiation. Development 128: 4523–4534., , , , , ,
- 141994 Manipulating the Mouse Embryo: A Laboratory Manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, USA., , ,
- 151996 Immunolocalization of type X collagen in normal fetal and adult osteoarthritic cartilage with monoclonal antibodies. Matrix Biol 15: 231–238., , , , , ,
- 162001 The chondrogenic transcription factor Sox9 is a target of signaling by the parathyroid hormone-related peptide in the growth plate of endochondral bones. Proc Natl Acad Sci USA 98: 160–165., , ,
- 171990 Gene-encoding parathyroid hormone-like peptide: Nucleotide sequence of the rat gene and comparison with the human homologue. Mol Endocrinol 4: 441–446., , , ,
- 181996 PTH/PTHrP receptor in early development and Indian hedgehog-regulated bone growth. Science 273: 663–666., , , , , , , , , , , , ,
- 191995 Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev Biol 172: 126–138.,
- 201993 The mouse collagen X gene: Complete nucleotide sequence, exon structure and expression pattern. Biochem J 289: 247–253., , , , , , ,
- 211994 Characterization of primary cultures of chondrocytes from type II collagen/beta-galactosidase transgenic mice. Matrix Biol 14: 329–335., , , , ,
- 221999 Structural and functional characterization of the mouse Sox9 promoter: Implications for campomelic dysplasia. Hum Mol Genet 8: 691–696.,
- 231997 SOX9 is a potent activator of the chondrocyte-specific enhancer of the pro alpha1(II) collagen gene. Mol Cell Biol 17: 2336–2346., , , ,
- 241995 A 182 bp fragment of the mouse pro alpha 1(II) collagen gene is sufficient to direct chondrocyte expression in transgenic mice. J Cell Sci 108: 3677–3684., , , , , ,
- 252002 The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes Dev 16: 2813–2828., , , ,
- 261997 Induction of osteogenic differentiation by hedgehog proteins. Biochem Biophys Res Commun 237: 465–469., , , , , , , , , , ,
- 271996 Regulation of patched by sonic hedgehog in the developing neural tube. Proc Natl Acad Sci USA 93: 9346–9351.,
- 281996 Targeted overexpression of parathyroid hormone-related peptide in chondrocytes causes chondrodysplasia and delayed endochondral bone formation. Proc Natl Acad Sci USA 93: 10240–10245., , , , ,
- 292002 Shh establishes an Nkx3.2/Sox9 autoregulatory loop that is maintained by BMP signals to induce somitic chondrogenesis. Genes Dev 16: 1990–2005., , , ,
- 302001 The chick transcriptional repressor Nkx3.2 acts downstream of Shh to promote BMP-dependent axial chondrogenesis. Dev Cell 1: 411–422., , ,