Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 May 15.
Published in final edited form as: Dev Cell. 2012 May 15;22(5):927–939. doi: 10.1016/j.devcel.2012.03.011

FoxA family members are crucial regulators of the hypertrophic chondrocyte differentiation program

Andreia Ionescu 1, Elena Kozhemyakina 1, Claudia Nicolae 2, Klaus H Kaestner 3, Bjorn R Olsen 2, Andrew B Lassar 1,4
PMCID: PMC3356573  NIHMSID: NIHMS367822  PMID: 22595668

Abstract

During endochondral ossification small immature chondrocytes enlarge to form hypertrophic chondrocytes, which express collagen X. In this work, we demonstrate that FoxA factors are induced during chondrogenesis, bind to conserved binding sites in the collagen X enhancer, and can promote the expression of a collagen X-luciferase reporter in both chondrocytes and fibroblasts. In addition, we demonstrate by both gain and loss of function analyses that FoxA factors play a crucial role driving the expression of both endogenous collagen X and other hypertrophic chondrocyte-specific genes. Mice engineered to lack expression of both FoxA2 and FoxA3 in their chondrocytes display defects in chondrocyte hypertrophy, alkaline phosphatase expression, and mineralization in their sternebrae and in addition exhibit postnatal dwarfism that is coupled to significantly decreased expression of both collagen X and MMP13 in their growth plates. Together, our findings indicate that FoxA family members are crucial regulators of the hypertrophic chondrocyte differentiation program.

Introduction

The axial and appendicular skeleton is formed by a process termed endochondral ossification. During this process, mesenchymal cells aggregate to form cartilage condensations consisting of immature chondrocytes. Cells lying within the central regions of the cartilage undergo a further differentiation process (hypertrophy), withdrawing from the cell cycle, enlarging their size and initiating synthesis of a new extracellular matrix containing collagen X. Chondrocyte hypertrophy is largely responsible for about 60% of skeletal growth, while 30% is due to matrix deposition and about 10% due to cellular proliferation (Wilsman et al., 1996). The early steps of chondrogenesis, including mesenchymal condensation and expression of chondrocyte-specific extracellular matrix proteins are critically dependent upon Sox-family transcription factors, including Sox9, Sox5, and Sox6 (de Crombrugghe et al., 2001; Lefebvre, 2002). In contrast, the process of chondrocyte hypertrophy is regulated by the Runx family of transcription factors. Runx2 and Runx3 are expressed in chondrocytes as they initiate differentiation, and loss of these factors (in genetically engineered mice) severely delays or blocks chondrocyte hypertrophy in a number of developing bones (Inada et al., 1999; Kim et al., 1999; Yoshida et al., 2004). Similarly, conditional deletion of Runx1 in the developing skeleton results in delay of chondrocyte hypertrophy and ossification in the axial skeletal elements while the appendicular ones are largely unaffected (Kimura et al., 2010). Conversely, ectopic expression of Runx2 in immature chondrocytes drives premature cellular maturation and induces expression of collagen X and other hypertrophic markers, both in vivo (Stricker et al., 2002; Takeda et al., 2001; Ueta et al., 2001) and in vitro (Enomoto et al., 2000). In addition to Runx family members, MEF2C and MEF2D also play a critical role in modulating chondrocyte hypertrophy. They do so either directly, by controlling expression of various differentiation markers (i.e., Indian Hedgehog, PTHrP Receptor, collagen X) or indirectly, by promoting Runx2 expression (Arnold et al., 2007).

Besides their role in promoting cartilage hypertrophy, both Runx and MEF2 families play critical roles in regulating either osteocyte formation (Runx2) (Komori et al., 1997) or osteocyte homeostasis (MEF2) (Leupin et al., 2007). Because both hypertrophic cartilage and osteocytes share a common precursor cell (Akiyama et al., 2005) and a common set of transcriptional regulators (i.e., Runx2 and MEF2), it seems likely that these two cell types may differentially express other regulators to account for their phenotypic differences. To investigate whether this phenotypic divergence can be explained by the existence of lineage-specific co-factors, we forced expression of Runx2 in undifferentiated mesenchymal progenitor cells (i.e., somitic explants) and followed the induction of bone versus cartilage markers in these cells (Kempf et al., 2007). We found that exogenous Runx2, while capable of inducing the bone marker osteopontin in mesenchymal progenitor cells, was able to induce expression of the hypertrophic chondrocyte marker, collagen X, only in a chondrocytic milieu (Kempf et al., 2007). These findings suggested that chondrocytes may contain a co-factor that is necessary for Runx2 to activate collagen X gene expression. In the present work, we identify members of the FoxA transcription factor family as crucial chondrogenic transcription factors that work in collaboration with Runx and MEF2 factors to drive expression of collagen X and other hypertrophic chondrocyte markers.

Results

Both a chondrogenic environment and the presence of upstream enhancer sequences are necessary for induction of a collagen X reporter by Runx2, Smad1 and MEF2C

The regulatory regions that drive the expression of human, mouse and avian collagen X have been analyzed by a number of groups (reviewed in (Gebhard et al., 2004)). These studies have established that both positive and negative regulatory regions which lie upstream of the collagen X gene play an important role in the expression of a transgene in either transfected chondrocytes or in transgenic mice. The avian collagen X enhancer contains a 4.2 kb sequence (termed the ABC enhancer) located upstream of the coding region that can drive hypertrophic chondrocyte-specific gene expression both in vitro (Lu Valle et al., 1993) and in vivo (Campbell et al., 2004). Within the ABC enhancer, a 642 bp subfragment (termed the b2 enhancer) is capable of driving BMP-dependent activation of the collagen X proximal promoter in hypertrophic chondrocytes in vitro (Volk et al., 1998). Therefore, we have employed the b2 enhancer region appended to a minimal 230 bp sequence of the proximal collagen X promoter in a luciferase reporter (b2-230-Luc) to study collagen X regulation in different cellular environments. Co-transfection with Runx2, Smad1 and MEF2C expression vehicles robustly activates this reporter in chondrocytes but not in fibroblasts (Figure 1A). As Runx2, Smad1 and MEF2 are competent transcriptional activators in either cell type ((Kempf et al., 2007), data not shown), this result suggests either that chondrocytes contain an activity that is absent from fibroblasts that allows exogenous Runx2, Smad1 and MEF2C to induce expression of a collagen X-luciferase reporter only in the former cell type, or that fibroblasts contain an inhibitor of collagen X expression.

Figure 1. Characterization of chondrocyte-specific transcription factor binding sites within the chicken collagen X enhancer (b2) that are critical for promoter activity.

Figure 1

Open in a new tab

(A) Either upper sternal chondrocytes or fibroblasts isolated from chicken embryos were co-transfected with a collagen X luciferase reporter (b2-230-Luc) plus Smad1/Runx2/MEF2C expression vectors as indicated. The reporter is comprised of the b2 collagen X enhancer appended to a 230 bp proximal promoter region driving the expression of the Renilla Luciferase gene. In both this and subsequent experiments, Renilla Luciferase Units were normalized to the expression of a co-transfected SV40-Firefly Luciferase vehicle. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.01. Error bars indicate standard error of the mean (SEM) where n=3. (B) Electrophoretic mobility shift assay (EMSA) was performed using nuclear extracts derived from either chondrocytes [C] or fibroblasts [F]. The proteins were incubated with selected sequences (oligos 1–4) within the b2 enhancer in the presence of either dIdC or dGdC as non-specific competitors. Two chondrocyte-specific complexes: a slow mobility complex (SMC) and a fast mobility complex (FMC) are shown by arrowheads. (C) The nucleotides critical for the interaction of the chondrocyte-specific complexes (SMC/FMC) with the DNA are displayed in blue. The common core motif ACAAA, present in each oligo, is outlined in the red box. (D) EMSA was performed using chondrocyte nuclear extracts incubated with either wild-type (W) oligos 1–4 or those containing mutations (M) in the ACAAA sequence, in the presence of dGdC or dIdC as indicated. (E) Schematic of wild-type and mutant b2-230-Luc reporter constructs with mutations of the ACAAA sequence located in oligos 1 to 4 as indicated (“X” represents site of mutation). Activity of these constructs in transfected chondrocytes is shown in either the absence or presence of co-transfected Smad1 and Runx2 expression vehicles plus BMP2 (S/R/B). Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.02. Error bars indicate standard error of the mean (SEM) where n=3. See also Figures S1 and S2.

Four distinct oligos within the b2 enhancer contain binding sites for DNA binding activities that are enriched in chondrocytes

To identify a putative “chondrocyte-specific” transcription factor(s) that is required for collagen X expression, we performed an electrophoretic mobility shift assay (EMSA) with 40 bp oligos that were tiled across either the entire b2 enhancer or the proximal promoter region in 20 bp overlaps. These oligos were incubated with nuclear extracts made from either chicken embryonic upper sternal chondrocytes (C) or chicken embryonic fibroblasts (F). By performing this analysis, we identified four oligos which interacted with DNA binding factors that were either enriched or present only in chondrocyte nuclear extracts. These 4 sequences (labeled oligos 1–4 in Figure 1B) were distributed within the b2 enhancer and flanked binding sites for both Runx2 and MEF2C (Supplemental Figure 1A). These sequences bound two classes of protein complexes with distinct electrophoretic mobilities: a Fast Mobility Complex (FMC) and a Slow Mobility complex (SMC). While the FMC bound equally well to all four oligos in the presence of competitor dIdC or dGdC, the SMC bound these same oligos only in the presence of dGdC and was competed off the oligos by dIdC (Figure 1B). Additionally, the SMC bound to multiple additional sites on the b2 enhancer that were not bound by the FMC (Supplemental Figure 1B).

Mutation of FMC/SMC binding sites cripples b2 enhancer activity in chondrocytes

In order to identify the binding sites for the FMC and the SMC we performed a 5bp scanning mutagenesis on oligos 1–4 which interconverted Gs with Ts, and As with Cs (shown for oligo 1 in Supplementary Figures 1C and 1D). This analysis mapped the sequences within each oligo that are required for interaction with the chondrocyte-specific complexes and revealed that both FMC and SMC bound to a common motif, ACAAANA (outlined in Figure 1C). Not surprisingly, mutation of the ACAAA core motif eliminated both the FMC and SMC chondrocyte-specific gel shifts (Figure 1D). To determine whether loss of FMC/SMC binding sites would affect expression of a collagen X transgene, we built these mutations into the luciferase reporter driven by the b2 enhancer sequences (b2-230-Luc). We mutated the ACAAA sequence in either both sites 1 and 2, both sites 3 and 4, or the combination of all four binding sites (Figure 1E). We observed an additive decrease in b2 enhancer activity of the collagen X reporter as we mutated more sites, with a maximal 95% decrease when we mutated all four FMC/SMC binding sites (Figure 1E).

The Slow Mobility Complex contains Sox5

Mutagenic analyses revealed that Runx2 and MEF2 binding sites in the b2 collagen X enhancer work in combination with binding sites for the FMC/SMC complexes to promote maximal activity of this enhancer in chondrocytes ((Drissi et al., 2003); Supplementary Figure 2). Thus, we focused our attention on discovering the identity of the SMC and FMC transcription factors. While the FMC bound only to oligos 1–4, the SMC bound to several other sites in the b2 enhancer (Supplementary Figure 1B). A common feature shared between these other SMC binding sites and the core motif of the FMC/SMC binding site (ACAAANA) is that they both contain A and T rich regions. Interestingly, the Sox family of transcription factors are similarly known to bind to runs of As and Ts (Wegner, 2005). Additionally, like the SMC factor, interaction of Sox transcription factors with their cognate binding site is inhibited by competitor dIdC but not by dGdC (Wegner, 2005). In light of these similarities between members of the Sox family and the SMC factor, we investigated whether in vitro translated Sox5 or Sox9 would bind to oligos 1–4, and whether antibodies that recognize different Sox family members would “double shift” the SMC. We found that both in vitro translated Sox5 and Sox9 could bind to these oligos as assayed by EMSA (shown for oligo 1 in Figure 2A). Of all the antibodies employed (i.e., against Sox5, Sox6, Sox9 and Sox13) we found that antibodies that recognized two distinct regions of Sox5 could specifically double shift the SMC gel shift on oligos 1 to 4 (shown for oligo 1 in Figure 2B). These findings indicate that Sox 5 is a component of the Slow Mobility Complex.

Figure 2. The Slow Mobility Complex (SMC) Contains Sox 5.

Figure 2

Open in a new tab

(A) EMSA using in vitro-translated Sox5 and Sox9 incubated with oligo 1. (B) EMSA using chondrocyte nuclear extracts incubated with oligo 1 and specific anti-Sox family antibodies as indicated. Two antibodies made against different domains of the Sox5 protein (anti-Sox5a and anti-Sox5b) specifically super-shift the SMC, but not the FMC gel shift. (C) Co-transfection of the b2-230-Collagen X Luciferase reporter together with Smad1, Runx2 and either Sox5, Sox6, or Sox9 expression vehicles in either chondrocytes or fibroblasts. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.05. Error bars indicate standard error of the mean (SEM) where n=3. (D) Immunostaining for Sox5 (green) or Collagen X (red) is displayed in either chondrocytes or fibroblasts. DAPI staining of DNA is shown in blue.

Transfection of Sox factors decreases the activity of the collagen X enhancer in chondrocytes and fails to activate this enhancer in fibroblasts

As SMC/FMC binding sites are required for maximal b2 enhancer activity, we explored whether ectopic expression of the SMC factor, Sox5, would augment the activity of the collagen X-reporter driven by this enhancer (b2-230-Luc). We found that co-transfection of Runx2 and Smad1 with any of the chondrogenic Sox family members (i.e., Sox5, Sox6, or Sox9) failed to activate this reporter in fibroblasts, and that these same Sox factors dampened the induction of the collagen X-reporter by Runx2 and Smad1 in chondrocytes (Figure 2C). Together these results suggest that the decrease of collagen X enhancer activity following mutation of the combined SMC/FMC binding sites (Figure 1E) are unlikely to be due to a loss of interaction with Sox5 (which is present in the SMC) but rather may reflect the loss of interaction of the FMC with the collagen X enhancer. As Sox5 is present within the nuclei of chondrocytes that express collagen X (Figure 2D), it is possible that Sox5 serves to dampen the expression of collagen X by competing with the FMC factor for a common binding site.

FoxA2 is present in the Fast Mobility Complex

Sequence comparison of the FMC binding sites revealed that the ACAAANA core motif of the FMC site includes a consensus Fox (Forkhead) binding site (see Figure 1C). Prior in situ hybridization and immunohistochemistry studies in mice embryos have indicated that FoxA2 and FoxA3 are expressed in developing cartilages (Besnard et al., 2004; Hiemisch et al., 1997; Monaghan et al., 1993). Thus, we investigated if FoxA family members could bind to the FMC binding sites in oligos 1–4 and whether antibodies against different Fox family members would “recognize” the FMC. While mice have three FoxA family members (A1, A2, and A3), browsing the chicken genome revealed that this genome encodes only FoxA1 and FoxA2. Either in vitro translated FoxA1 or FoxA2 could bind to oligos 1–4 by EMSA (Figure 3A), and mutation of the ACAAA sequence within each oligo abrogated the interaction with FoxA2 (Figure 3D). Most importantly, an antibody that recognizes in vitro translated chicken FoxA2 could also specifically block formation of the FMC gel shift in chondrocyte nuclear extracts incubated with oligos 1–4 (Figure 3B). In contrast, anti-FoxA1 failed to affect the FMC gel shift (Figure 3B).

Figure 3. The Fast Mobility Complex (FMC) Contains FoxA2.

Figure 3

Open in a new tab

(A) EMSA using in vitro-translated FoxA1 and FoxA2 incubated with oligos 1–4. (B) EMSA using chondrocyte (C) or fibroblast (F) nuclear extracts (NE) or in vitro- translated FoxA2 (IVT A2) and incubated with oligos 1–4 in the presence or absence of either anti-FoxA1 or anti-FoxA2 antibodies as indicated. (C) Co-transfection of the ABC-230-Collagen X Luciferase reporter together with Smad1, Runx2 and either FoxA1 or FoxA2 as indicated, in either chondrocytes or fibroblasts. Significance was calculated using Student's t-test, there is not a significant statistical difference (p ≥ 0.05) between the luciferase activities of the collagen X luciferase reporter co-transfected with either FoxA1 or FoxA2 in chondrocytes versus in fibroblasts. Error bars indicate standard error of the mean (SEM) where n=3. (D) EMSA of in vitro-translated FoxA2 with either wild-type (W) oligos 1–4 or those containing mutations (M) in the ACAAA sequence. (E) Schematic of wild-type and mutant b2-230-Luc reporter constructs with mutations of the ACAAA sequence located in oligos 1 to 4 as indicated (“X” represents the site of mutation). Activity of these constructs in transfected chondrocytes is shown in either the absence or presence of a co-transfected FoxA2 expression vehicle. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.01. Error bars indicate standard error of the mean (SEM) where n=3.

Exogenous FoxA2 can activate expression of the collagen X reporter in both chondrocytes and fibroblasts

As all four FMC binding sites are required for maximal b2 enhancer activity, we investigated whether ectopic expression of the FMC factor, FoxA2, would augment the activity of the full-length collagen X enhancer/promoter (i.e., ABC-230-Luc). In striking contrast to Runx2/Smad1 over-expression, which failed to stimulate expression of the collagen X reporter in fibroblasts, co-transfection of either FoxA1 or FoxA2 robustly activated expression of this reporter to approximately equivalent levels (in either the absence or presence of exogenous Runx2/Smad1) in both transfected upper sternal chondrocytes and in fibroblasts (Figure 3C). Mutation of the ACAAA sequence in the various FMC binding sites of the b2 enhancer both blocked interaction of the DNA with FoxA2 (Figure 3D) and led to loss of induction of b2 enhancer activity by co-transfected FoxA2 (Figure 3E). The combined loss of all 4 binding sites showed the most dramatic decrease of enhancer activity (Figure 3E). Together, these findings indicate that FoxA2 is a component of the chondrocyte-specific FMC, and that exogenous FoxA2 can activate expression of the collagen X reporter in both chondrocytes and fibroblasts.

Chondrogenic signals induce the expression of FoxA factors

To determine whether chondrogenic cues induce the expression of FoxA family members, we assayed their expression during chondrogenesis in high density micromass cultures of chicken limb bud mesenchymal cells. The expression of both FoxA1 and FoxA2 was robustly induced during chondrogenesis of these cells, slightly trailing that of Runx2 and roughly paralleling that of VEGF, collagen X and MMP13 (Figure 4A). As the FMC factor is present in nuclear extracts made from chicken sternal chondrocytes and absent from those made from chicken embryo fibroblasts, we also evaluated the relative expression of FoxA1 and FoxA2 in these cell types. We found that upper sternal chondrocytes, but not fibroblasts, express both FoxA1 and FoxA2 as assayed by RT-qPCR (Figure 4B). Interestingly, we noted that MEF2C is also induced during chondrogenic differentiation of limb bud cells and is specifically expressed in chondrocytes but not in fibroblasts (Figures 4A and B). However, while FoxA1/A2 could induce expression of the collagen X luciferase reporter in both chondrocytes and fibroblasts, MEF2C could only induce expression of this reporter in chondrocytes (Figure 4C), and in this cell type to a much lower level than did co-transfected FoxA factors. Together these findings suggest that FoxA factors, which are present in chondrocytes yet absent from fibroblasts, may confer competence for Runx2, Smad1, and MEF2C to activate expression of the collagen X luciferase reporter in the former cell type.

Figure 4. Over-expression or knock-down of FoxA2 increases or decreases expression, respectively, of several hypertrophic chondrocyte markers.

Figure 4

Open in a new tab

(A) Limb bud mesenchymal cells were cultured under high density micromass conditions for the indicated number of days. Gene expression was assayed by RT-qPCR. In both this and subsequent experiments, transcript levels were normalized to GAPDH levels. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.05 when gene expression in high density micromass cultures at various time points is compared with gene expression at the initial time of plating. Error bars indicat standard error of the mean (SEM) where n=3. (B) FoxA1,FoxA2, and MEF2C are expressed in chondrocytes (C) but not in fibroblasts (F) as assayed by RT-qPCR. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.01 when gene expression in chondrocyte cultures is compared with gene expression in fibroblast cultures. Error bars indicate standard error of the mean (SEM) where n=3. (C) Either chondrocytes or fibroblasts were co-transfected with the b2-230-Luc reporter in either the absence or presence of exogenous FoxA2 or MEF2C, as indicated. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.05 when luciferase activites from either FoxA2 or MEF2C co-transfected cells are compared to luciferase activities of control vehicle transfected cells. Error bars indicate standard error of the mean (SEM) where n=3. (D) Over-expression of FoxA2 increases expression of several hypertrophic chondrocyte markers. Chondrocytes were infected with an avian retrovirus encoding either GFP (RCAS-GFP; black bars) or FoxA2 (RCAS-mFoxA2; grey bars). Cells were cultured in BMP2 for the indicated number of days and subsequently gene expression was assayed by quantitative real time PCR. Transcript levels relative to GAPDH are displayed. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.05 when gene expression in chondrocyte cultures infected with RCAS-mFoxA2 is compared with gene expression in chondrocyte cultures infected with RCAS-GFP. Error bars indicate standard error of the mean (SEM) where n=3. (E) Knock-down of FoxA2 decreases expression of several hypertrophic chondrocyte markers. Chondrocytes were infected with an avian retrovirus encoding either a scrambled shRNA (RCAS-sh-nonGFP; black bars) or shRNA targeting FoxA2 (RCAS-sh-cFoxA2; grey bars).Cells were cultured in BMP2 for the indicated number of days and subsequently gene expression was assayed by quantitative real time PCR. Transcript levels relative to GAPDH are displayed. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.05 when gene expression in chondrocyte cultures infected with RCAS-sh-cFoxA2 is compared with gene expression in chondrocyte cultures infected with RCAS-sh-nonGFP. Error bars indicate standard error of the mean (SEM) where n=3.

Over-expression of FoxA2 increases the expression of endogenous markers of chondrocyte hypertrophy while knockdown of FoxA2 delays chondrocyte hypertrophy

Because forced expression of FoxA2 could boost the expression of the collagen X luciferase reporter in chondrocytes, we investigated whether exogenous FoxA2 would similarly increase the expression of endogenous markers of chondrocyte hypertrophy. We infected chicken sternal chondrocytes with an avian retrovirus encoding either GFP (RCAS-GFP) or FoxA2 (RCAS-FoxA2). Because BMP signals are known to induce chondrocyte hypertrophy (Volk et al., 2000; Volk et al., 1998), we cultured the cells in BMP2 for increasing periods of time. Expression of retroviral-encoded FoxA2 specifically increased the expression of endogenous collagen X up to 3-fold following 8 days culture in BMP2 (Figure 4D). Interestingly, RCAS-FoxA2 infection strikingly boosted the expression of the late chondrocyte hypertrophic marker MMP13 up to 60-fold, and in this case, robust induction was observed in either the absence or presence of exogenous BMP2 (Figure 4D). The fact that viral FoxA2-induced expression of collagen X was delayed relative to that of MMP13, suggests that induction of collagen X by FoxA2 may require other factors (i.e., Runx2 and MEF2C), whose expression/activity is rate-limiting in upper sternal chondrocytes until after 8 days culture in BMP2. Consistent with this notion, ectopic FoxA2 only boosted the expression of endogenous collagen X in chondrocytes and not in other cell types (data not shown). In contrast, FoxA2-mediated induction of MMP13 was observed in both chondrocytes and in the osteoblast line MC3T3 (data not shown). In addition, we noted that exogenous FoxA2 also induced the expression of VEGF up to 6.3-fold, alkaline phosphatase up to 14.7-fold (Figure 4D), and also boosted collagen II expression (2.5 fold) but only at early only at early times of culture (data not shown). Conversely, infection of chondrocytes with an avian retrovirus programmed to express shFoxA2 (RCAS-shFoxA2) significantly decreased BMP2-mediated induction of endogenous FoxA2, MMP13, collagen X, VEGF and alkaline phosphatase, but had little effect on collagen II expression (Figure 4E and data not shown). The loss of collagen X and other hypertrophic markers following the knockdown of FoxA2 in chondrocytes suggests that FoxA2 is necessary (directly or indirectly) to promote maximal expression of these genes in hypertrophic chondrocytes.

FoxA2 and FoxA3 exhibit distinct chondrocyte-specific expression patterns

Prior in situ hybridization and immunohistochemistry studies in mice embryos have indicated that FoxA2 and FoxA3 are expressed in the developing skeleton in vivo (Besnard et al., 2004; Hiemisch et al., 1997; Monaghan et al., 1993). To further characterize FoxA3 expression in the developing skeleton, we genetically labeled all descendants of cells that have expressed FoxA3. We used a transgenic mouse line (FoxA3-Cre) containing a 170 kb yeast artificial chromosome that encompasses the entire FoxA3 locus, which faithfully reproduces the expression of the endogenous gene, and drives the expression of Cre DNA recombinase in vivo (Lee et al., 2005). Mating of FoxA3-Cre males with ROSA26 reporter females, which contain a Cre-inducible LacZ locus (Soriano, 1999), results in beta-galactosidase expression in all cells (and their descendents) that express FoxA3. In FoxA3-Cre; ROSA26 newborn mice, X-gal staining, which detects beta-galactosidase activity, was apparent in all cartilaginous elements of the skeleton but was absent from bone tissue (Figure 5A, panels a, b, c ). Histological analysis of newborn tibias revealed the presence of X-gal positive cells in all three layers of the growth plate: immature round cells, columnar flat chondrocytes, and hypertrophic chondrocytes (Figure 5B, panels 5–7). In contrast, no labeled cells were detected in trabecular bone (Figure 5B, panel 8), nor was staining observed in control ROSA26 reporter mice alone (Figure 5B, panels 1–4). These results indicate that FoxA3 is expressed in immature chondrocytes and possibly in their descendents (discussed below).

Figure 5. Chondrocyte-specific knockout of FoxA2 and deletion of FoxA3 in mice results in delayed hypertrophy and postnatal dwarfism.

Figure 5

Open in a new tab

(A) Whole mount X-gal staining of the skeleton of newborn FoxA3-Cre; ROSA26 mice as compared with control ROSA26 littermates (panels a, b, c). (B) X-gal/ Fast Red staining of tibial sections of newborn FoxA3-Cre; ROSA26 mice (panels 5–8) compared with control littermates (panels 1–4). (C) Hematoxylin/Eosin (HE) and X-gal staining on sections of either vertebrae, second sternebra, or forelimb digit from FoxA2-CreERT2; ROSA26 mice that were administered tamoxifen by intraperitoneal (IP) injection at either 7 or 14 dpc and harvested at the various times as indicated. (D) Alcian Blue/Alizarin Red skeletal preparations of either sternum, vertebral column, tail, or forelimb from E19.5 Col2-Cre; FoxA2flox/flox (2 null alleles; 2NA) mice embryos, Col2-Cre; FoxA2flox/flox; FoxA3−/− (4 null alleles; 4NA) mice embryos or their control littermates.

For a detailed analysis of FoxA2 expression in the developing skeleton, we employed a tamoxifen-inducible Cre driver (CreERT2) knocked-into the 3' UTR of the FoxA2 gene behind an IRES sequence (Park et al., 2008). FoxA2-CreERT2 males were crossed with ROSA26 females and a single IP injection was administered to the pregnant females at either 7 dpc or 14 dpc; the embryos were harvested after either 2, 4, or 11 days following tamoxifen administration. In the axial skeleton (i.e. vertebrae) of embryos administered tamoxifen at 7 dpc and harvested 11 days later, X-gal positive cells were present in both the intervertebral discs as well as throughout the vertebral bodies (Figure 5C, top panels 4–6). In contrast, in the axial skeleton of embryos that were administered tamoxifen at 14 dpc and harvested 2 or 4 days later, only the hypertrophic chondrocytes located in the center of the vertebral bodies were X-gal positive (Figure 5C, top panel 3), while cells in either the nucleus pulposus or less mature chondrocytes exhibited no staining (Figure 5C, top panels 1–2). Together, these findings suggest that FoxA2 is initially expressed in a population of progenitor cells that gives rise to various structures of the vertebral column, while at later times expression of this gene is restricted to hypertrophic chondrocytes.

Sternum development is delayed in comparison to vertebral column development. In the sternum, X-gal staining was only detectable in the hypertrophic chondrocytes of mice embryos injected with tamoxifen at 14 dpc and harvested 4 days later (Figure 5C, middle panels 1–4). In the appendicular skeleton of mice embryos injected with tamoxifen at 14 dpc, X-gal staining was also present in hypertrophic chondrocytes of the growth plate and in a few sparse cells in the prehypertrophic region, but was absent from either perichondrial cells or immature chondrocytes (Figure 5C, bottom panels 1–3). Interestingly, in embryos injected with tamoxifen at 7 dpc and harvested 11 days later, no cells were labeled in either the sternum (Figure 5C, large middle right panel) or the appendicular skeleton (Figure 5C, large bottom right panel). Thus, in contrast to the axial skeleton, FoxA2 expression in both the sternum and appendicular skeleton is apparently restricted to hypertrophic chondrocytes.

Chondrocyte-specific knockout of FoxA2 and FoxA3 in mice results in postnatal dwarfism, loss of collagen X expression and decreased expression of other chondrocyte hypertrophy markers

Mice homozygous for a FoxA2 null allele die at E10–11 due to severe defects in the node, notochord, neural tube and gut tube (Ang and Rossant, 1994; Weinstein et al., 1994), precluding an analysis of the role of FoxA2 at later stages of development. Thus, to evaluate the role of FoxA2 in cartilage development, we conditionally deleted FoxA2 in developing chondrocytes using a mouse line that expresses Cre recombinase controlled by the collagen2a1 promoter (Col2-Cre; (Long et al., 2001)) plus engineered alleles of FoxA2 that contain loxP sites (Sund et al., 2000). In order to test for potential redundancy between different FoxA family members, we conditionally deleted FoxA2 alone or in a FoxA3−/− background, since FoxA3 null mice are viable and display no skeletal abnormalities (Kaestner et al., 1998; Shen et al., 2001). Both two-allele deletion mutant mice (Col2-Cre; FoxA2flox/flox) (a.k.a., 2 null alleles; 2NA) and four-allele deletion mutant mice (Col2-Cre; FoxA2flox/flox; FoxA3−/−) (a.k.a., 4 null alleles; 4NA) were recovered with the expected Mendelian ratio, but they died at birth, or in the immediate postnatal period, and very few mice survived post-weaning time. Conditional FoxA2 null mice were readily recognizable by their misshaped “curly” tails (Figures 5D and 7B). Staining of cartilage and mineralized bone using Alcian blue and Alizarin red respectively, revealed multiple skeletal defects in both axial and appendicular skeletons of E19.5 embryos lacking either FoxA2 or both FoxA2 and FoxA3 in their developing cartilages (Figure 5D). While loss of FoxA2 alone led to a slight delay in cartilage calcification, loss of both FoxA2 and FoxA3 led to a profound loss of mineralization in both sternebrae and vertebrae (Figure 5D). A small subset of mutant mice exhibited kyphosis due to misshaped or fused thoracic vertebrae (Figure 5D). Moreover, in all mutants, loss of FoxA2/3 led to an extreme fusion of caudal vertebrae, mostly towards the distal end of the tail (Figure 5D). This was also apparent, but less extreme, in mice lacking only FoxA2 in their chondrocytes (Figure 5D). Interestingly, cartilage calcification in the appendicular skeleton was also decreased but less affected by loss of FoxA2/3 than the axial cartilages (Figure 5D).

Figure 7. Chondrocyte-specific knockout of FoxA2 and deletion of FoxA3 in mice results in loss of hypertrophic chondrocyte gene expression in the growth plate.

Figure 7

Open in a new tab

(A) In situ hybridization analysis of tibial growth plates for the indicated genes is displayed for either E19.5 Col2-Cre; FoxA2flox/flox; FoxA3−/− (4NA) embryos or their control littermates. (B) Alcian Blue/Alizarin Red skeletal preparations of 2 week old Col2-Cre; FoxA2flox/flox; FoxA3−/− (4NA) mice or their control littermates. (C) Hematoxylin and Eosin (HE) staining of tibial epiphysis of either 2 week old Col2-Cre; FoxA2flox/flox; FoxA3−/− (4NA) mice or control littermates. Quantification of the thickness of either the proliferative zone (PZ) or the hypertrophic zone (HZ) in the tibial growth plates is displayed. Significance was calculated using Student's t-test, * denotes statistical significance at p ≤ 0.001 when tibial growth plate zones from 2 week old 4NA mutant mice are compared to the tibial growth plate zones from control animals. Error bars indicate standard error of the mean (SEM) where n=6. (D) In situ hybridization analysis of tibial growth plates for the indicated genes is displayed for either 2 week old Col2-Cre; FoxA2flox/flox; FoxA3−/− (4NA) mice or their control littermates.

In light of the skeletal abnormalities observed in the Col2-Cre; FoxA2flox/flox; FoxA3−/− mice, we further characterized cartilage hypertrophy in these animals. Histological analysis showed only residual hypertrophic chondrocytes in the sterna of E19.5 mutant mice. Moreover, these small islands of cells misaligned, indicating that chondrocyte hypertrophy was not properly coordinated in the progeny of the two fused sternal bars (Figure 6A panels a, b). To further characterize this population of hypertrophic cells, we assayed both alkaline phosphatase activity and matrix mineralization by von Kossa staining. Interestingly, while the residual hypertrophic cells in the sternebrae of E19.5 Col2-Cre; FoxA2flox/flox; FoxA3−/− embryos display an enlargement in cell volume (Figure 6A panels a, b), they lack significant alkaline phosphatase activity (Figures 6B and 6D panels a, b) and little if any matrix mineralization (Figure 6C panels a, b). Similarly, the cephalic and caudal vertebrae were severely misshapen, displayed a deficit of hypertrophic cells and the nucleus pulposus was frequently absent between vertebrae primordia in the tails of these mutant embryos (Figure 6A panels c, d, e, f). Alkaline phosphatase activity was present in the periosteal cells lining the vertebrae of both mutant and control mice, but was significantly diminished in the vertebral bodies of mutant mice (Figures 6B and 6D panels c, d, e, f) as was matrix mineralization, as assayed by von Kossa staining (Figures 6C panels c, d, e, f). In contrast to the axial cartilage which displayed gross maturation defects, the appendicular cartilage of E19.5 mutant mice was relatively normal in embryos lacking FoxA2/3 in their chondrocytes, but displayed less trabeculae, less mineralization and attenuated expression of alkaline phosphatase relative to their control litter mates (Figures 6A, 6B, 6C, 6D, panels g, h). The delay in cartilage degradation and bone invasion in the appendicular cartilage of embryos lacking FoxA2/3 was even more evident in E16 mutant embryos (Figures 6A, 6B, 6C, 6D panels i, j).

Figure 6. Chondrocyte-specific knockout of FoxA2 and deletion of FoxA3 in mice results in delayed hypertrophy, loss of alkaline phosphatase activity and delayed mineralization.

Figure 6

Open in a new tab

(A, B, and C) Sections of either E19.5 sternebrae (a,b), thoracic vertebrae (c,d), caudal vertebrae (e,f), femurs (g,h), or E16.5 femurs (i, j) taken from either control or Col2-Cre; FoxA2flox/flox; FoxA3−/− (4NA) embryos were stained with either Hematoxylin and Eosin (HE) (A), for alkaline phosphatase activity (B) or with von Kossa stain (C). See also Figure S3.

To further characterize cartilage maturation in embryos lacking FoxA2 and FoxA3 in their chondrocytes, we assayed the expression of collagen II, collagen X, and MMP13, which are expressed in either immature, early hypertrophic, or late hypertrophic chondrocytes, respectively. While expression of collagen II was not significantly affected in the tibial growth plates of E19.5 Col2-Cre; FoxA2flox/flox; FoxA3−/− embryos, expression of collagen X was severely decreased and that of MMP13 was markedly attenuated in these embryos (Figure 7A). A small fraction of Col2-Cre; FoxA2flox/flox; FoxA3−/− mice (4NA) survived postnatally. Skeletal staining of these two-week old 4NA mice indicated that they displayed dwarfism, had smaller rib cages, frequently lacked a xiphoid process, had fused vertebrae in their tail, and their long bones were relatively short (Figure 7B). Consistent with the smaller size of these mutant animals relative to their normal littermates at 2 weeks of age, the tibial growth plates in the Col2-Cre; FoxA2flox/flox; FoxA3−/− mice were smaller, with a specific deficiency of cells in the hypertrophic zone but not in the proliferating zone (Figure 7C). These results are supported by in situ hybridization analysis which revealed that expression of both collagen X and MMP13 were significantly diminished in the tibial growth plates of 2 week old Col2-Cre; FoxA2flox/flox; FoxA3−/− mice relative to their control littermates (Figure 7D). Together, these findings indicate that both FoxA2 and FoxA3 are necessary to promote high level expression of several hypertrophic chondrocyte markers in the growth plate, including collagen X, alkaline phosphatase, and MMP13.

Discussion

FoxA factors are required for expression of both collagen X and other hypertrophic chondrocyte markers

In prior work, we found that exogenous Runx2 can induce the expression of chondrocyte hypertrophy markers such as collagen X and Ihh only in chondrogenic cells (Kempf et al., 2007), suggesting that this cell type may uniquely express a competence factor that allows Runx2 to activate expression of chondrocyte hypertrophy markers. To identify this factor, we performed EMSA with oligos tiled across the minimal chicken collagen X enhancer, and thereby identified two chondrocyte-specific transcription factors, Sox5 and FoxA2, that bind to essential sequences that are reiterated within this enhancer. In this work, we demonstrate that FoxA factors are induced during chondrogenesis, that ectopic expression of FoxA factors can boost the expression of a collagen X-reporter in both chondrocytes and fibroblasts, and that expression of collagen X, MMP13, and alkaline phosphatase are markedly attenuated in mice lacking FoxA2 and FoxA3 in their cartilage.

Interestingly, we found that the axial and appendicular cartilage elements displayed differential sensitivities to the loss of FoxA2 and FoxA3. While chondrocyte hypertrophy, mineralization, and alkaline phosphatase activity were nearly extinguished in sternebrae and tail vertebrae of E19.5 Col2-Cre; FoxA2flox/flox; FoxA3−/− embryos, these markers of chondrocyte maturation were only attenuated or delayed in expression in appendicular skeletal elements. It is not clear why axial and appendicular cartilage elements display this differential sensitivity to the loss of FoxA2/3, however this finding suggests that there is differential expression of a factor(s) that compensates for the loss of FoxA2/3 in axial versus appendicular cartilage. It will be interesting to determine whether FoxA1, which is expressed along with FoxA2/3 in metatarsal explants (Supplementary 3), is differentially expressed in axial versus appendicular cartilage to allow hypertrophic chondrocyte gene expression in the absence of FoxA2/3 in these latter skeletal elements.

In addition, we also noted that not all markers of hypertrophic chondrocyte differentiation are equally affected by loss of FoxA2 and FoxA3. While the number of hypertrophic cells in the growth plates of Col2-Cre; FoxA2flox/flox; FoxA3−/− mice were diminished relative to control littermates, the hypertrophic chondrocytes within these growth plates clearly underwent an increase in cell size. In contrast, collagen X expression in the hypertrophic chondrocytes was nearly extinguished in Col2-Cre; FoxA2flox/flox; FoxA3−/− mice. In addition, while chondrocyte hypertrophy was markedly blunted in the sternebrae of these animals, residual pockets of large hypertrophic chondrocytes could be observed that lacked significant alkaline phosphatase activity and failed to mineralize. Thus the hypertrophic growth of chondrocytes can apparently be uncoupled from the expression of hypertrophic chondrocyte markers such as collagen X and alkaline phosphatase.

Some of the features of the Col2-Cre; FoxA2flox/flox; FoxA3−/− mice are reproduced in two mouse models of hypophosphatasia characterized by knock-out of the tissue-nonspecific isoenzyme of alkaline phosphatase (TNSALP) (Fedde et al., 1999). Mice deficient for this enzyme are normal during the first week of age, but develop osteopenia and multiple fractures shortly thereafter, when they exhibit severe growth retardation characterized by shortened limbs with delayed ossification centers and a markedly reduced hypertrophic zone. These characteristics are similar to the phenotype of the two-week old Col2-Cre; FoxA2flox/flox; FoxA3−/− mice, consistent with the depressed alkaline phosphatase activity in the growth plate of these animals.

Chondrogenic cues induce the expression of FoxA factors in mesenchymal cells

FoxA1/2 and FoxA3 expression are induced in primary cultures of either chicken or murine limb bud mesenchymal cells, respectively, when such cells are induced to differentiate into chondrocytes (this study and (Hoffman et al., 2006)). The induction of FoxA family members during in vitro chondrogenesis of limb bud mesenchymal cells is consistent with in vivo findings that FoxA family members are specifically expressed in the cartilage anlagen. FoxA3-Cre; ROSA26 mice exhibit X-gal staining in all cartilaginous elements of the skeleton but not in their bone tissue. In addition, FoxA3-Cre; ROSA26 mice display X-gal staining in all chondrocyte populations; including immature, prehypertrophic, and hypertrophic chondrocytes. However, since this Cre line is non-inducible, we cannot distinguish whether FoxA3 is uniquely expressed in cartilage progenitor cells (which will yield descendants that stably express beta-galactosidase after FoxA3 expression is extinguished) or if FoxA3 continues to be expressed in immature, prehypertrophic and hypertrophic chondrocytes. A pulse of tamoxifen to E7 FoxA2-Cre-ERT2; ROSA26 embryos gave rise to X-gal positive cells in both the intervertebral discs as well as throughout the vertebral bodies, but in neither the sternum nor limbs, suggesting that FoxA2 is initially expressed in a progenitor population that gives rise to different structures in the vertebral column. In contrast, if the window of tamoxifen exposure began at E14, only hypertrophic chondrocytes in both the axial and appendicular skeleton expressed beta-galactosidase, suggesting that eventually FoxA2 expression is restricted to hypertrophic chondrocytes.

FoxA factors work in competition with Sox factors and in synergy with Runx2/3, MEF2C/D, and Smad1/4 to drive expression of collagen X

In contrast to the immature chondrocyte differentiation program which is driven by Sox transcription factors (de Crombrugghe et al., 2001; Lefebvre, 2002), the hypertrophic chondrocyte differentiation program is induced by the combined activities of Runx2/3 (Yoshida et al., 2004), MEF2C/D (Arnold et al., 2007), and FoxA2/3 (this work) (summarized in Supplemental Figure 4). While each of these transcription factor families is expressed in other cells types, the requirement for the combination of these transcription factors to induce collagen X expression renders expression of this gene unique to hypertrophic chondrocytes. In addition, we noted that Sox5 (and Sox 6 and 9) are able to bind to the same sequences in the chicken collagen X enhancer as FoxA family members. While ectopic expression of FoxA factors markedly boosted the expression of a collagen X-luciferase reporter in both chondrocytes and fibroblasts, ectopic expression of either Sox 5, 6, or 9 repressed induction of this reporter by Runx2/Smad1 in chondrocytes. These findings suggest that Sox transcription factors, which are expressed in immature chondrocytes and are down-regulated in hypertrophic chondrocytes, may act to compete for FoxA factor binding to the collagen X enhancer and thereby inhibit precocious expression of collagen X and/or other hypertrophic markers in immature chondrocytes. Indeed others have noted that mutation of a Sox9 binding site within the murine Col10a1 enhancer is Sox9 required to repress expression of a transgene driven by this enhancer in non-hypertrophic chondrocytes (Leung et al., 2011).

FoxA factors have been termed pioneer DNA binding factors that are able to bind nucleosomal covered DNA and to disrupt local chromatin structure (Cirillo et al., 2002). Indeed FoxA factor-chromatin interaction has been demonstrated to facilitate binding of other transcription factors including the androgen, estrogen, and glucocorticoid receptors to adjacent binding sites (reviewed in (Kaestner, 2010)). Thus it is possible that FoxA factors may provide competence of for expression of collagen X by promoting the interaction of other transcription factors such as Runx2/3 and MEF2C/D with adjacent binding sites in the regulatory region of this gene. Consistent with this notion, we have found that binding sites for Runx2/3 and MEF2C/D are flanked by FoxA binding sites in the chicken collagen X enhancer (summarized in Supplementary Figure 1A). It will be interesting to determine whether FoxA factors allow accessibility of the collagen X enhancer to these other families of transcription factors and/or promote the assembly of an enhancesome like structure containing these many classes of transcription factors on this regulatory region.

A 535 bp enhancer located upstream of the human collagen X gene was demonstrated to drive expression of a reporter gene specifically in hypertrophic chondrocytes (Chambers et al., 2002; Riemer et al., 2002). This enhancer sequence is conserved upstream of both human, mouse and bovine collagen X genes (Gebhard et al., 2004), and a 4.6 kb fragment upstream of the mouse collagen X gene containing this enhancer can drive hypertrophic chondrocyte-specific expression of beta-galactosidase in transgenic mice (Gebhard et al., 2004). While regions of extensive sequence similarity to the human collagen X enhancer are also present upstream of the mouse and bovine collagen X genes, the chicken collagen X ABC enhancer does not share continuous blocks of sequence similarity with this sequence (Gebhard et al., 2004). Interestingly however, 6 conserved FoxA binding sites are present within the conserved region of the human collagen X enhancer that are also present upstream of the murine and bovine collagen X genes (Supplemental Figure 5A; putative FoxA binding sites outlined in light blue) and flank Runx2 binding sites that are also conserved in the human, murine and bovine collagen X enhancers (Supplemental Figure 5A; outlined in yellow). Co-transfection of a luciferase reporter (4.6kb mcol10-Luc) driven by both the mouse collagen X enhancer and proximal promoter together with either murine FoxA1, FoxA2, or FoxA3 revealed that the enhancer activity of this regulatory region is significantly boosted by exogenous FoxA factors (Supplemental Figures 5B and 5C). In contrast, expression of a mouse collagen X reporter lacking the upstream conserved enhancer sequences (1kb mcol10-Luc) was only slightly augmented by co-transfected FoxA factors (Supplemental Figures 5B and 5C). Thus, the presence of conserved FoxA and Runx2 binding sites in both mammalian and avian collagen X enhancers suggests that these transcription factors may directly activate the expression of collagen X in hypertrophic chondrocytes of both mammals and birds.

Experimental Procedures

Cell culture and luciferase reporter assay

Chondrocytes were isolated from the cephalic portion of day 15 chicken embryo sterna and cultured as previously described (Leboy et al., 1989). Cells were subsequently plated at 150,000 cells/well into six-well plates and transfected with the indicated expression plasmids using Superfect transfection reagent (Qiagen) according to the manufacturer's protocol. Cells were lysed 48 h after transfection, and luciferase reporter activity was determined using the dual-luciferase reporter assay system (Promega). Limb bud mesenchymal cells were isolated from stage 22–24 chicken embryos and cultured under micromass conditions of extremely high cell density (Ahrens et al., 1977; Osdoby and Caplan, 1979) Metatarsals were isolated from 15.5-day postcoitum (dpc) pregnant female mice and cultured as described (Guo et al., 2006).

EMSA – electrophoretic mobility shift assay

Nuclear protein extracts were prepared as previously described (Lee et al., 2004) and in vitro translated FoxA2 protein was obtained using a TNT Quick Coupled Transcription/Translation System (Promega) according to manufacturer's protocol. The mobility shift assay was performed as previously described (Ionescu et al., 2001). Supershifts were performed with antibodies specific for Sox and Fox family members from Santa Cruz Biotechnologies: anti-Sox5a (sc-20091x), anti-Sox5b(sc-17329x), anti-Sox6a(sc-17332x), anti-Sox13a(sc-20009x), anti-Sox13b(sc-17349x), anti-Sox13c(sc-17350x), anti-Sox9a(sc-20095x), anti-Sox9b(sc-17340x), anti-Sox9c(sc-17341x), anti-FoxA1(sc-22841x), anti-FoxA2(sc-6554x).

Immunofluorescence

Chondrocytes were fixed in 2% paraformaldehyde for 15 minutes at room temperature, permeabilized with 0.25% Triton X-100, blocked in donkey serum and then incubated with primary antibodies: 1:500 goat anti-Sox5 (Santa Cruz, sc-20091x) and 1:50 mouse anti-collagen x (Developmental Studies Hybridoma Bank, X-AC9). The interaction was then visualized using 1:250 anti-goat-FITC and anti-mouse-TRITC from Jackson Labs.

Analysis of mutant mice

For FoxA2-CreERT2; ROSA26 and FoxA3-Cre; ROSA26 mice, whole mount beta-galactosidase staining was performed using a beta-galactosidase detection kit from Millipore. Alcian Blue-Alizarin Red staining of the Col2-Cre; FoxA2fl/fl; FoxA3−/− skeletons was performed as previously described (Depew, 2008). For immunohistochemistry, E19.5 embryos were fixed in 4% buffered paraformaldehyde overnight at 4°C, paraffin-embedded and sectioned at 7 μm, and stained with Hematoxylin /Eosin from Richard Allen Scientific. Von Kossa staining was performed using a “Von Kossa method for Calcium” kit from Polysciences according to manufacturer's protocol. Alkaline phosphatase staining was performed using NBT/BCIP (Roche). Non-radioactive in situ hybridization was performed as previously described (Kolpakova-Hart et al., 2008).

On line supplemental experimental procedures

Details of the construction of plasmids and viruses, RNA isolation and RT-PCR, and generation of mutant mice are provided in the on line supplement for this work.

Supplementary Material

01

Highlights

  • Ectopic Runx2/Smad1/MEF2C only induce collagen X-luciferase in chondrocytes

  • FoxA2 and Sox5 bind to critical binding sites in the chicken collagen X enhancer

  • Ectopic FoxA2 boosts expression of coll X, MMP13, and alk. phos. in chondrocytes

  • Deletion of FoxA2 and FoxA3 in chondrocytes disrupts chondrocyte hypertrophy

Acknowledgements

This work was supported by grants to A.B.L. from NIAMS/NIH (AR048524, AR055552), grants to B. R. O. from the NIH (AR36819 and AR36820), and grants to K.H.K. from NIDDK/NIH (DK049210, DK054342). A. I. was supported by a fellowship from the Arthritis Foundation. We thank Drs. Leboy, Lefevbre, Olson, Sanders, Whitman, von der Mark, and Zaret for generously sharing plasmids with us, the Nikon Imaging Facility at Harvard Medical School for the use of their microscopes, and Jennifer Waters for her assistance with photomicroscopy.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Ahrens PB, Solursh M, Reiter RS. Stage-related capacity for limb chondrogenesis in cell culture. Dev Biol. 1977;60:69–82. doi: 10.1016/0012-1606(77)90110-5. [DOI] [PubMed] [Google Scholar]
  2. Akiyama H, Kim JE, Nakashima K, Balmes G, Iwai N, Deng JM, Zhang Z, Martin JF, Behringer RR, Nakamura T, et al. Osteo-chondroprogenitor cells are derived from Sox9 expressing precursors. Proc Natl Acad Sci U S A. 2005;102:14665–14670. doi: 10.1073/pnas.0504750102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Ang SL, Rossant J. HNF-3 beta is essential for node and notochord formation in mouse development. Cell. 1994;78:561–574. doi: 10.1016/0092-8674(94)90522-3. [DOI] [PubMed] [Google Scholar]
  4. Arnold MA, Kim Y, Czubryt MP, Phan D, McAnally J, Qi X, Shelton JM, Richardson JA, Bassel-Duby R, Olson EN. MEF2C Transcription Factor Controls Chondrocyte Hypertrophy and Bone Development. Dev Cell. 2007;12:377–389. doi: 10.1016/j.devcel.2007.02.004. [DOI] [PubMed] [Google Scholar]
  5. Besnard V, Wert SE, Hull WM, Whitsett JA. Immunohistochemical localization of Foxa1 and Foxa2 in mouse embryos and adult tissues. Gene Expr Patterns. 2004;5:193–208. doi: 10.1016/j.modgep.2004.08.006. [DOI] [PubMed] [Google Scholar]
  6. Campbell MR, Gress CJ, Appleman EH, Jacenko O. Chicken collagen X regulatory sequences restrict transgene expression to hypertrophic cartilage in mice. Am J Pathol. 2004;164:487–499. doi: 10.1016/S0002-9440(10)63139-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chambers D, Young DA, Howard C, Thomas JT, Boam DS, Grant ME, Wallis GA, Boot-Handford RP. An enhancer complex confers both high-level and cell-specific expression of the human type X collagen gene. FEBS Lett. 2002;531:505–508. doi: 10.1016/s0014-5793(02)03606-2. [DOI] [PubMed] [Google Scholar]
  8. Cirillo LA, Lin FR, Cuesta I, Friedman D, Jarnik M, Zaret KS. Opening of compacted chromatin by early developmental transcription factors HNF3 (FoxA) and GATA-4. Mol Cell. 2002;9:279–289. doi: 10.1016/s1097-2765(02)00459-8. [DOI] [PubMed] [Google Scholar]
  9. de Crombrugghe B, Lefebvre V, Nakashima K. Regulatory mechanisms in the pathways of cartilage and bone formation. Curr Opin Cell Biol. 2001;13:721–727. doi: 10.1016/s0955-0674(00)00276-3. [DOI] [PubMed] [Google Scholar]
  10. Depew MJ. Analysis of skeletal ontogenesis through differential staining of bone and cartilage. Methods Mol Biol. 2008;461:37–45. doi: 10.1007/978-1-60327-483-8_5. [DOI] [PubMed] [Google Scholar]
  11. Drissi MH, Li X, Sheu TJ, Zuscik MJ, Schwarz EM, Puzas JE, Rosier RN, O'Keefe RJ. Runx2/Cbfa1 stimulation by retinoic acid is potentiated by BMP2 signaling through interaction with Smad1 on the collagen X promoter in chondrocytes. J Cell Biochem. 2003;90:1287–1298. doi: 10.1002/jcb.10677. [DOI] [PubMed] [Google Scholar]
  12. Enomoto H, Enomoto-Iwamoto M, Iwamoto M, Nomura S, Himeno M, Kitamura Y, Kishimoto T, Komori T. Cbfa1 is a positive regulatory factor in chondrocyte maturation. J Biol Chem. 2000;275:8695–8702. doi: 10.1074/jbc.275.12.8695. [DOI] [PubMed] [Google Scholar]
  13. Fedde KN, Blair L, Silverstein J, Coburn SP, Ryan LM, Weinstein RS, Waymire K, Narisawa S, Millan JL, MacGregor GR, et al. Alkaline phosphatase knock-out mice recapitulate the metabolic and skeletal defects of infantile hypophosphatasia. J Bone Miner Res. 1999;14:2015–2026. doi: 10.1359/jbmr.1999.14.12.2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gebhard S, Poschl E, Riemer S, Bauer E, Hattori T, Eberspaecher H, Zhang Z, Lefebvre V, de Crombrugghe B, von der Mark K. A highly conserved enhancer in mammalian type X collagen genes drives high levels of tissue-specific expression in hypertrophic cartilage in vitro and in vivo. Matrix Biol. 2004;23:309–322. doi: 10.1016/j.matbio.2004.05.010. [DOI] [PubMed] [Google Scholar]
  15. Guo J, Chung UI, Yang D, Karsenty G, Bringhurst FR, Kronenberg HM. PTH/PTHrP receptor delays chondrocyte hypertrophy via both Runx2-dependent and -independent pathways. Dev Biol. 2006;292:116–128. doi: 10.1016/j.ydbio.2005.12.044. [DOI] [PubMed] [Google Scholar]
  16. Hiemisch H, Schutz G, Kaestner KH. Transcriptional regulation in endoderm development: characterization of an enhancer controlling Hnf3g expression by transgenesis and targeted mutagenesis. Embo J. 1997;16:3995–4006. doi: 10.1093/emboj/16.13.3995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hoffman LM, Garcha K, Karamboulas K, Cowan MF, Drysdale LM, Horton WA, Underhill TM. BMP action in skeletogenesis involves attenuation of retinoid signaling. J Cell Biol. 2006;174:101–113. doi: 10.1083/jcb.200604150. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Inada M, Yasui T, Nomura S, Miyake S, Deguchi K, Himeno M, Sato M, Yamagiwa H, Kimura T, Yasui N, et al. Maturational disturbance of chondrocytes in Cbfa1-deficient mice. Dev Dyn. 1999;214:279–290. doi: 10.1002/(SICI)1097-0177(199904)214:4<279::AID-AJA1>3.0.CO;2-W. [DOI] [PubMed] [Google Scholar]
  19. Ionescu AM, Schwarz EM, Vinson C, Puzas JE, Rosier R, Reynolds PR, O'Keefe RJ. PTHrP modulates chondrocyte differentiation through AP-1 and CREB signaling. J Biol Chem. 2001;276:11639–11647. doi: 10.1074/jbc.M006564200. [DOI] [PubMed] [Google Scholar]
  20. Kaestner KH. The FoxA factors in organogenesis and differentiation. Curr Opin Genet Dev. 2010;20:527–532. doi: 10.1016/j.gde.2010.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kaestner KH, Hiemisch H, Schutz G. Targeted disruption of the gene encoding hepatocyte nuclear factor 3gamma results in reduced transcription of hepatocyte-specific genes. Mol Cell Biol. 1998;18:4245–4251. doi: 10.1128/mcb.18.7.4245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kempf H, Ionescu A, Udager A, Lassar AB. Prochondrogenic signals induce a competence for Runx2 to activate hypertrophic chondrocyte gene expression. Developmental Dynamics. 2007;236:1954–1962. doi: 10.1002/dvdy.21205. [DOI] [PubMed] [Google Scholar]
  23. Kim IS, Otto F, Zabel B, Mundlos S. Regulation of chondrocyte differentiation by Cbfa1. Mech Dev. 1999;80:159–170. doi: 10.1016/s0925-4773(98)00210-x. [DOI] [PubMed] [Google Scholar]
  24. Kimura A, Inose H, Yano F, Fujita K, Ikeda T, Sato S, Iwasaki M, Jinno T, Ae K, Fukumoto S, et al. Runx1 and Runx2 cooperate during sternal morphogenesis. Development. 2010;137:1159–1167. doi: 10.1242/dev.045005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kolpakova-Hart E, Nicolae C, Zhou J, Olsen BR. Col2-Cre recombinase is co-expressed with endogenous type II collagen in embryonic renal epithelium and drives development of polycystic kidney disease following inactivation of ciliary genes. Matrix Biol. 2008;27:505–512. doi: 10.1016/j.matbio.2008.05.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Komori T, Yagi H, Nomura S, Yamaguchi A, Sasaki K, Deguchi K, Shimizu Y, Bronson RT, Gao YH, Inada M, et al. Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell. 1997;89:755–764. doi: 10.1016/s0092-8674(00)80258-5. [DOI] [PubMed] [Google Scholar]
  27. Leboy PS, Vaias L, Uschmann B, Golub E, Adams SL, Pacifici M. Ascorbic acid induces alkaline phosphatase, type X collagen, and calcium deposition in cultured chick chondrocytes. J Biol Chem. 1989;264:17281–17286. [PubMed] [Google Scholar]
  28. Lee CS, Sund NJ, Behr R, Herrera PL, Kaestner KH. Foxa2 is required for the differentiation of pancreatic alpha-cells. Dev Biol. 2005;278:484–495. doi: 10.1016/j.ydbio.2004.10.012. [DOI] [PubMed] [Google Scholar]
  29. Lee KH, Evans S, Ruan TY, Lassar AB. SMAD-mediated modulation of YY1 activity regulates the BMP response and cardiac-specific expression of a GATA4/5/6-dependent chick Nkx2.5 enhancer. Development. 2004;131:4709–4723. doi: 10.1242/dev.01344. [DOI] [PubMed] [Google Scholar]
  30. Lefebvre V. Toward understanding the functions of the two highly related Sox5 and Sox6 genes. J Bone Miner Metab. 2002;20:121–130. doi: 10.1007/s007740200017. [DOI] [PubMed] [Google Scholar]
  31. Leung VY, Gao B, Leung KK, Melhado IG, Wynn SL, Au TY, Dung NW, Lau JY, Mak AC, Chan D, et al. SOX9 governs differentiation stagespecific gene expression in growth plate chondrocytes via direct concomitant transactivation and repression. PLoS Genet. 2011;7:e1002356. doi: 10.1371/journal.pgen.1002356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Leupin O, Kramer I, Collette NM, Loots GG, Natt F, Kneissel M, Keller H. Control of the SOST bone enhancer by PTH using MEF2 transcription factors. J Bone Miner Res. 2007;22:1957–1967. doi: 10.1359/jbmr.070804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Long F, Zhang XM, Karp S, Yang Y, McMahon AP. Genetic manipulation of hedgehog signaling in the endochondral skeleton reveals a direct role in the regulation of chondrocyte proliferation. Development. 2001;128:5099–5108. doi: 10.1242/dev.128.24.5099. [DOI] [PubMed] [Google Scholar]
  34. Lu Valle P, Iwamoto M, Fanning P, Pacifici M, Olsen BR. Multiple negative elements in a gene that codes for an extracellular matrix protein, collagen X, restrict expression to hypertrophic chondrocytes. J Cell Biol. 1993;121:1173–1179. doi: 10.1083/jcb.121.5.1173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Monaghan AP, Kaestner KH, Grau E, Schutz G. Postimplantation expression patterns indicate a role for the mouse forkhead/HNF-3 alpha, beta and gamma genes in determination of the definitive endoderm, chordamesoderm and neuroectoderm. Development. 1993;119:567–578. doi: 10.1242/dev.119.3.567. [DOI] [PubMed] [Google Scholar]
  36. Osdoby P, Caplan AI. Osteogenesis in cultures of limb mesenchymal cells. Dev Biol. 1979;73:84–102. doi: 10.1016/0012-1606(79)90140-4. [DOI] [PubMed] [Google Scholar]
  37. Park EJ, Sun X, Nichol P, Saijoh Y, Martin JF, Moon AM. System for tamoxifen-inducible expression of cre-recombinase from the Foxa2 locus in mice. Dev Dyn. 2008;237:447–453. doi: 10.1002/dvdy.21415. [DOI] [PubMed] [Google Scholar]
  38. Riemer S, Gebhard S, Beier F, Poschl E, von der Mark K. Role of c-fos in the regulation of type X collagen gene expression by PTH and PTHrP: localization of a PTH/PTHrP-responsive region in the human COL10A1 enhancer. J Cell Biochem. 2002;86:688–699. doi: 10.1002/jcb.10260. [DOI] [PubMed] [Google Scholar]
  39. Shen W, Scearce LM, Brestelli JE, Sund NJ, Kaestner KH. Foxa3 (hepatocyte nuclear factor 3gamma ) is required for the regulation of hepatic GLUT2 expression and the maintenance of glucose homeostasis during a prolonged fast. J Biol Chem. 2001;276:42812–42817. doi: 10.1074/jbc.M106344200. [DOI] [PubMed] [Google Scholar]
  40. Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet. 1999;21:70–71. doi: 10.1038/5007. [DOI] [PubMed] [Google Scholar]
  41. Stricker S, Fundele R, Vortkamp A, Mundlos S. Role of Runx genes in chondrocyte differentiation. Dev Biol. 2002;245:95–108. doi: 10.1006/dbio.2002.0640. [DOI] [PubMed] [Google Scholar]
  42. Sund NJ, Ang SL, Sackett SD, Shen W, Daigle N, Magnuson MA, Kaestner KH. Hepatocyte nuclear factor 3beta (Foxa2) is dispensable for maintaining the differentiated state of the adult hepatocyte. Mol Cell Biol. 2000;20:5175–5183. doi: 10.1128/mcb.20.14.5175-5183.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Takeda S, Bonnamy JP, Owen MJ, Ducy P, Karsenty G. Continuous expression of Cbfa1 in nonhypertrophic chondrocytes uncovers its ability to induce hypertrophic chondrocyte differentiation and partially rescues Cbfa1-deficient mice. Genes Dev. 2001;15:467–481. doi: 10.1101/gad.845101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Ueta C, Iwamoto M, Kanatani N, Yoshida C, Liu Y, Enomoto-Iwamoto M, Ohmori T, Enomoto H, Nakata K, Takada K, et al. Skeletal malformations caused by overexpression of Cbfa1 or its dominant negative form in chondrocytes. J Cell Biol. 2001;153:87–100. doi: 10.1083/jcb.153.1.87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Volk SW, D'Angelo M, Diefenderfer D, Leboy PS. Utilization of bone morphogenetic protein receptors during chondrocyte maturation. J Bone Miner Res. 2000;15:1630–1639. doi: 10.1359/jbmr.2000.15.8.1630. [DOI] [PubMed] [Google Scholar]
  46. Volk SW, Luvalle P, Leask T, Leboy PS. A BMP responsive transcriptional region in the chicken type X collagen gene. J Bone Miner Res. 1998;13:1521–1529. doi: 10.1359/jbmr.1998.13.10.1521. [DOI] [PubMed] [Google Scholar]
  47. Wegner M. Secrets to a healthy Sox life: lessons for melanocytes. Pigment Cell Res. 2005;18:74–85. doi: 10.1111/j.1600-0749.2005.00218.x. [DOI] [PubMed] [Google Scholar]
  48. Weinstein DC, Ruiz i Altaba A, Chen WS, Hoodless P, Prezioso VR, Jessell TM, Darnell JE., Jr. The winged-helix transcription factor HNF-3 beta is required for notochord development in the mouse embryo. Cell. 1994;78:575–588. doi: 10.1016/0092-8674(94)90523-1. [DOI] [PubMed] [Google Scholar]
  49. Wilsman NJ, Farnum CE, Leiferman EM, Fry M, Barreto C. Differential growth by growth plates as a function of multiple parameters of chondrocytic kinetics. J Orthop Res. 1996;14:927–936. doi: 10.1002/jor.1100140613. [DOI] [PubMed] [Google Scholar]
  50. Yoshida CA, Yamamoto H, Fujita T, Furuichi T, Ito K, Inoue K, Yamana K, Zanma A, Takada K, Ito Y, et al. Runx2 and Runx3 are essential for chondrocyte maturation, and Runx2 regulates limb growth through induction of Indian hedgehog. Genes Dev. 2004;18:952–963. doi: 10.1101/gad.1174704. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS367822-supplement-01.doc (1.8MB, doc)

RESOURCES