• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Reactivation of Hox Gene Expression During Bone Regeneration

Posted on: Sunday, 7 August 2005, 03:00 CDT

Abstract

Previous studies have explored the link between bone regeneration and skeletogenesis. Although a great deal is known regarding tissue and cell based events, especially those involving ossification and chondrogenesis, much remains unknown about the molecular similarity of repair and development. Since the functional significance of Homeobox (Hox) genes in embryonic skeletogenesis has been well documented through knockout and deficiency studies, we chose to investigate whether members of this family are reactivated during fracture repair. Specifically, we examined the temporal and spatial expression of Msx-1, Msx-2, rHox, Hoxa-2 and Hoxd-9, because of their involvement in limb development. Utilizing quantitative reverse transcriptase RT-PCR (qPCR), mRNA levels from all five genes were shown to be upregulated during fracture repair at all times tested (post-fracture day 3-21), as compared to intact bone. Further, using in situ hybridization and immunohistochemistry, spatial expression of these genes was localized to osteoblasts, chondrocytes and periosteal osteoprogenitor cells found within the fracture callus, the foremost cells responsible for the reparative phase of the healing process. Given the contribution of Hox genes in skeletal development, our results suggest that these genes are involved in either the patterning or formation of the fracture callus, further supporting the notion that bone regeneration recapitulates skeletal development.

2005 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved.

Keywords: Bone regeneration; Hox; Gene expression; Fracture repair

Introduction

It is well accepted that fracture repair is essentially a recapitulation of bone formation during development [7,10,13,23,33]. This connection is based on the notion that both processes involve similar molecular mechanisms that control cellular behavior and ultimately lead to the differentiation of mesenchymal cells into active osteoblasts and chondrocytes. Eventually, these cells direct the regeneration of skeletal tissues (bone and cartilage) that comprise the fracture callus and provide the ideal biomechanical environment for healing. Despite these marked similarities there are some clear differences, namely, the inflammatory response following a fracture, as well as remodeling, both processes which are absent during embryonic skeletal development.

In addition to these tissue and cell level morphological events, further similarities between embryonic bone development and regeneration can be found at the molecular level, especially in the activation patterns of specific genes. For example, the signaling genes Indian hedgehog (Ihh) [33] and Vascular Endothelial Growth Factor (VEGF), as well as matrix associated genes such as and Matrix metalloprotease (MMP) 13 and transcription factors such as Core binding factor alpha 1 (Cbfal/Runx2) and Glicentin 1 (Gli) all show related expression patterns throughout embryonic development and fracture repair [7,33]. The same is true for members of the Transforming Growth Factor beta (TGF-β) superfamily (reviewed in [10]). Further, through global transcriptional profiling, our laboratory has also shown the expression of a large number of other genes that are upregulated during both skeletal development and regeneration [13,21], including novel genes [23].

Based on these results, we hypothesized that other genes important during skeletal embryogenesis, such as Hox genes, are also reactivated during fracture repair. Hox genes represent transcription factors that were originally identified in Drosophila and, along with their mammalian homologs, have been implicated in a wide variety of processes during embryogenesis (reviewed in [17]). Specifically, we focused on Hox genes that are known to be involved in limb bud formation during embryogenesis [15], such as Msx-1, Msx- 2, Hoxa-2 and Hoxd-9. In addition, we analyzed the expression of another Hox candidate, rHox (also known as MHox, Prxl, Pmx and Prrx), which was initially identified in osteoblasts [18] and found to be upregulated during fracture repair [13]. Results from knockout studies suggest that some of these Hox genes play significant roles in bone development and as such, they may also have key roles in the fracture repair process. For example, Msx-2 mutant mice show abnormal cartilage and endochondral bone formation, reduced axial and appendicular skeletal length, as well as reduced numbers and altered morphology of osteoblasts and proliferating and hypertrophie chondrocytes [31]. Msx-1 and Msx-2 were also shown to play essential and possibly redundant, roles in craniofacial development including formation of sutures in the skull cap and tooth development [1]. Skeletal abnormalities were also observed for Hoxa-2 and Hoxd-9 deficiency. Specifically, absence of Hoxa-2 results in multiple cranial defects [9], while absence of Hoxd-9 creates malformations of the deltoid crest and reduction in humerus length. These observed effects suggest that Hoxd-9 and Hoxa-2 are involved in skeletal patterning [8,5]. Lastly, rHox is thought to direct formation of pre- skeletal condensations from undifferentiated mesenchyme because of the observation that rHox -/- mice die shortly after birth with loss or malformations of all skeletal components, including craniofacial, limb and vertebral structures [25].

Recently, using suppressive subtractive hybridization and microarray studies, our laboratory identified some of these genes to be upregulated during bone regeneration [13]. In this study, we further explore the temporal and spatial expression patterns of five Hox genes, Msx-1, and Msx-2, rHox, Hoxa-2, and Hoxd-9, using qPCR, in situ hybridization and immunohistochemistry. Herein we reveal that all of these genes, known to participate in embryonic skeletal patterning, are reactivated during the development of a mammalian fracture callus.

Methods

Fracture model

All methods and animal procedures were reviewed and approved by the University's Institutional Animal Care and Use Committee and met or exceeded all federal guidelines for the humane use of animals in research. Controlled fixed fractures were generated in the right femurs of 28 six-month old male Sprague-Dawley rats by a procedure previously described [4] and routinely performed in our laboratory [11-14,21]. A set of five animals was euthanized by CO2 inhalation at post-fracture (PF) day 3, 5, 7 and 10, and a set of 4 animals at PF day 14 and 21. Both the fractured and intact contralateral (control) femurs were dissected free and processed for RNA, protein or immunohistochemistry.

RNA isolation

Total RNA was isolated on PF days 3, 5, 7, 10, 14 and 21 from calluses as well as the contralateral control femurs. Calluses were harvested from diaphyses (approximately 5 mm proximal and distal to the fracture midpoint). These samples were homogenized in TriZol reagent (Invitrogen) or ToTALLY RNA (Ambion) as described previously [11-14,21,23]. As control, the intact femurs (including articular cartilage) were also homogenized in TriZol reagent (Invitrogen) or ToTALLY RNA (Ambion) as described above.

Quantitative reverse transcriplase real time polvmerase chain reaction (qPCR)

RNA in equivalent amounts from four animals was pooled to make the intact and PF day 3, 5, 7, 10 and 14 stock samples. Only two animals were combined to create the PF day 21 pool. Also, a reference pool (used in the calibration curve) was created by combining equal amounts of RNA from all samples of intact and PF day pools. Primers were custom-designed (Primer 3, http:// frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) to amplify 206- 315 bp sequences within the coding sequences of the experimental genes Msx-1, Msx-2, rHox, Hoxa-2 and Hoxd-9 (Table 1). All gene expression patterns were normalized to the expression pattern of β-actin (Table 1). All reactions utilized the One-Step QuantiTect SYBR Green RT-PCR kit (Qiagen) and were run using a Light Cycler (Roche). Optimization of reaction conditions was performed for each gene by altering annealing temperature (56-60 C). Each run contained intact and PF day 3, 5, 7, 10, 14 and 21 RNA for both the experimental genes and β-actin in addition to a five point calibration curve. The experimental gene values for all time points were normalized to their respective β-actin levels. Each experimental gene was run three times and results are reported as average fold change in expression relative to intact levels with standard deviation (SD).

Histochemistry and immunohistochemistry

Intact contralateral control and fractured lemurs from each PF day were dissected free of soft tissue, fixed in 10% buffered formalin, decalcified in 5% formic acid, and embedded in paraffin (Polysciences). Serial longitudinal sections (10 m) were then cut from each sample. Immunohistochemical analysis was performed on specific sections using primary polyclonal antibodies for Msx-1, Msx- 2, Hoxa-2 and Hoxd-9 at a concentration of 1:50 (Santa Cruz Biotech) in conjunction with biotinylated anti-goat (Jackson Immunoresearch) or anti-rabbit (Zymed) secondary antibodies at a concentration of 1:200. Sections were incubated at 60 C for 30 min, then deparafinized in xylene, and hydrated using an ethanol gradient. The sections were next permea\bilized in PBS containing 0.2% Triton X- 100 for 10 min, washed in PBS, and blocked (PBS with 3% horse serum) for 30 min. Sections were then incubated overnight at 4 C with primary antibodies in blocking buffer. The following day, sections were washed in PBS and incubated with biotinylated secondary antibodies (Jackson Immunoresearch, Zymed) in blocking buffer for 30 min. After additional washing. ABC reagent (Vector) was added for 30 min and sections were washed again before detection with DAB reagent (Vector). Negative controls were generated on adjacent sections by omitting the primary antibodies application. In addition Safranin O/ fast green staining was performed on adjacent sections using standard staining procedures to indicate the areas of cartilage and bone. All sections were analyzed under bright field microscopy using a Zeiss microscope (Axiovert 200) and images were captured with a CCD camera (AxioCam MRc).

Table 1

Oligonucleotide sequence of primers, amplicon size and melting temperature for each gene assayed by qPCR

In situ hybridization

In situ hybridization was carried out as previously described [23]. Briefly, prior to hybridization, all tissue sections were deparaffinized in xylene, washed, and hydrated using ethanol gradients. Protein digestion was accomplished by incubation in 1 N HCl, followed by incubation with varying concentrations of proteinase K (1-100 ng/mL, Roche). The sections were acetylated with 0.5% acetic anhydride in PBS (pH 8.0) for 10 min with continuous stirring. Riboprobes (antisense and sense) for rHox in hybridization buffer were heated at 80 C for 3 min, followed by quick cooling in ice water. The hybridization mixture contained each riboprobe (1.0 ng/L), 50% deionized formamide, 10% dextran sulfate, 2x SSC, 0.02% SDS, 0.01% salmon sperm DNA. The slides were incubated for 16 h at 60 C in a humidified chamber. After hybridization the sections were washed and processed using an anti-DIG detection assay (Roche). Finally, the sections were rinsed with tap water, mounted, viewed under bright field microscopy using a Zeiss microscope (Axiovert 200) and images were captured with a CCD camera (AxioCam MRc).

Results

Temporal mRNA expression analyses

Previous transcriptional profiling experiments from our laboratory identified rHox as differentially regulated during bone regeneration with mRNA levels in the fracture callus at 2.1-, 3.8-, 4-, 5.2-, 5.7- and 2.4-fold greater (relative to intact bone) for PF day 3, 5, 7, 10, 14 and 21, respectively [13]. To verify these results and to conclusively determine rHox expression pattern during callus progression, we carried out qPCR. This analysis revealed a similar trend in expression, that is, a steady increase in rHox expression peaking at 19.1-fold on PF day 10 followed by a decrease in expression, but still at much higher levels than intact bone, at PF day 14 (9.3-fold) and 21(11.2-fold) (Fig. 1A). Linear regression and correlation analysis of the two data sets yielded an R^sup 2^ value of 0.45, indicating that microarray data on its own is not very conclusive (although it did show a similar trend of upregulation) and should be confirmed with more quantitative assays.

Similarly, we sought to determine the exact expression levels of the remaining Hox candidate genes. For simplification of presentation and analysis, data from Hoxa-2 and Hoxd-9 analysis are grouped together in a single graph (Fig. 1B) while those for Msx-1 and Msx-2 is shown in Fig. 1C. Hoxa-2 shows a steady increase (2.7- 5.4-fold) in expression for the first 14 days and then a dramatic peak on PF day 21 with a 13.7-fold increase in expression relative to intact. In contrast, Hoxd-9 shows peaks in expression at PF days 3 and 10 (6.1-and 7.2-fold, respectively) (Fig. 1B). In the case of Msx-1 and Msx-2, both genes exhibited peaks of mRNA expression on PF day 7 with 6.1- and 12.6-fold increases, respectively, and again the level of upregulation was evident throughout all points tested (Fig. 1C). In summary, all of the genes monitored exhibited a clear increase in mRNA expression as compared to those detected in intact bone throughout the 21 post-fracture days monitored.

Spatial mRNA/protein expression analysis

Given that the RNA isolated from fracture calluses was derived from a heterogeneous population of cells, we sought to determine which particular cell type(s) in the callus is/are responsible for specifically expressing each of the Hox genes tested. We utilized both in situ hybridization (for rHox) and immunohistochemistry (for Msx-1, Msx-2, Hoxa-2 and Hoxd-9) to fulfill this objective.

Since no antibody was available for rHox, spatial localization analysis was performed via in situ hybridization on PF day 10 callus sections. We chose PF day 10 because it is the time point where the highest level of rHox expression was detected. In addition, on PF day 10 many of the biological processes (i.e., intramembranous ossification, chondrogenesis, endochondral ossification) are occurring simultaneously within the fracture callus. Strong hybridization was detected in areas of fibrocartilage, cartilage, periosteum, and woven bone as compared to sense control (Fig. 2). More specifically, in areas of cartilage, proliferating and pre- hypertrophic chondrocytes stained much more robustly than terminally differentiated hypertrophie chondrocytes (Fig. 2A and C). Osteoprogenitor cells lining the periosteum also demonstrated strong staining (Fig. 2D). Lastly, intense hybridization was detected in active osteoblasts found in areas of woven bone (Fig. 2D-F). In contrast, trapped osteoblasts that presumably differentiated into osteocytes, do not show robust reactivity to the rHox anti-sense riboprobe (Fig. 2D and F). Lastly, hybridization of an adjacent section (to that shown in Fig. 2A) with the control sense rHox riboprobe, showed no labeling (Fig. 2B).

Fig. 1. Transcriptional activation of Hox genes during fracture repair. All qPCR analyses were performed on pooled RNA samples (n = 4 animals/time point) from intact femurs and post-fracture days 3, 5, 7, 10 and 14. For PF day 21, n = 2. All values represent average fold change relative to intact bone with error bars indicating standard deviation among replicates (n = 3) in qPCR experiments. (A) Temporal mRNA analysis of rHox by qPCR (Black) and microarray (white). Linear regression and correlation analysis was used to derive the R^sup 2^ value. (B) Temporal mRNA analysis of Hoxa-2 (Black) and Hoxd-9 (White). (C) Temporal mRNA analysis of Msx-1 (White) and Msx-2 (Black).

Fig. 2. In situ hybridization of rHox during fracture repair. (A) PF 10 callus section bordering the fracture site hybridized to rHox anti-sense riboprobe illustrating regions of muscle (Mu), woven bone (Wb) and cartilage (Ca). (B) Adjacent section within the fracture callus to (A) hybridized with the sense rHox riboprobe as a control. (C) Higher magnification view of the cartilaginous zone from (A) (white box) displaying fibrocartilage (FCa) with proliferating chondrocytes (arrows) seen staining strongly and hypertrophic chondrocytes (arrowheads) indicating little or no staining. (D) rHox expression in periosteum (P) containing strongly staining osteoprogenitor cells (black arrows) and osteoblasts (black arrowheads). Very little or no staining is detected in osteoblasts/ osteocytes within areas of woven bone (white arrowheads). (E) Low magnification of rHox expression in areas of woven bone (Wb) and fibrocartilage (FCa) adjacent to fracture site. (F) Higher magnification view of the region in (E) (white box) of woven bone signifying intensely stained osteoblasts (black arrowheads) and unstained osteocytes (white arrowheads). Scale bars represent 100 m in (A), (B) and (E), 50 m in (C) and (D) and 20 m in (F).

For the remaining genes, Msx-1, Msx-2, Hoxa-2 and Hoxd-9, immunohistochemistry was performed using polyclonal antibodies specific for each protein. Again, we chose callus sections from PF day 10 because they represent the multiple biological processes proceeding simultaneously during fracture repair. To help us histologically distinguish between these different tissues, Safranin O/fast green staining was used on adjacent callus sections which stains areas of bone as blue/green and cartilage as red (Fig. 3B, H and Fig. 4B).

Fig. 3. Spatial expression of Hoxa-2 and Hoxd-9. Immunohistochemical analysis of PF day 10 callus sections for Hoxa- 2 and Hoxd-9. (A) Low magnification view of Hoxa-2 staining adjacent to the fracture site with areas of woven bone (Wb), fibrocartilage (Fca), and cartilage (Ca) demonstrating intense staining. (B) Safranin O fast/green staining of adjacent section from the same fracture callus with same regions identified. (C) Low magnification view of adjacent section stained for Hoxd-9 with similar areas (fibrocartilage, cartilage, woven bone) stained. (D) Higher magnification view of cartilaginous zone in (A) (white box) showing robustly stained proliferating chondrocytes (black arrows) and osteoblasts in areas of woven bone (white arrows). (E) Adjacent section to (A) and (C) within the fracture callus stained with secondary antibody (negative control) section depicting no staining. (F) Higher magnification view of cartilaginous zone in (C) (white box) showing robustly stained proliferating chondrocytes (black arrows) and osteoblasts/osteocytes trapped in areas of woven bone (white arrows). (G) Low magnification view of the same section stained for Hoxa-2 and showing intense staining in the periosteum (P), as well as areas of woven bone (Wb), cartilage (Ca) and fibrocartilage (Fca). (H) Adjacent section within the fracture callus stained with Safranin O fast/green depicting the same regions. (I) Adjoining section to (G) and (H) stained for Hoxd-9 and showing intense staining in the periosteum (P), as well as areas of woven bone (Wb), cartilage (Ca) and fibrocartilage (Fca). Scale bars represent 100 min (A), (B), (C), (E), (G), (H) and (I) while scale bars in D and F represent 50 m.

Hoxa-2 and Hoxd-9 showed very similar spatial expression patterns, especially when viewed at low magnification (Fig. 3A and C) and as compared to the negative control (secondary antibody only) that is devoid of immunostaining (Fig. 3E). Both Hoxa-2 and Hoxd-9 displayed strong protein expression in proliferating chondrocytes/ pre-hypertrophic chondrocytes but strikingly diminished expression in hypertrophie chondrocytes typically found at the core of the soft callus (Fig. 3D and F). In contrast, intense staining is detected in areas of fibrocartilage (Fig. 3A and C). Similarly, strong immunoreactivity was also observed in osteoblasts found in the surrounding newly formed woven bone (Fig. 3D, F, G and I), as well as in osteoprogenitor cells of the periosteum (Fig. 3G and I). Lastly, Hoxa-2 and Hoxd-9 staining persists in osteoblasts/ osteocytes that have been trapped in the newly made woven bone (Fig. 3D, F, G and I).

Interestingly, Msx-1 and Msx-2 show similar pattern of protein expression to Hoxa-2 and Hoxd-9. Specifically, intense immunostaining in proliferating/prehypertrophic chondrocytes can be observed (Fig. 4A, C, D and F) as compared to the negative control that shows no immunostaining (Fig. 4E). In contrast to Hoxa-2 and Hoxd-9, the expression pattern for both Msx-1 and Msx-2 persists even into the core cartilaginous regions where hypertrophie chondrocytes are predominantly found (Fig. 4D and F). Although not all hypertrophie chondrocytes stained positively, a large number of them are immunoreactive. It is worthwhile to mention that the staining of both of these transcription factors is concentrated within the nuclei of the stained cells (Fig. 4D and F). Lastly, osteoprogenitor cells of the periosteum, as well as osteoblasts and osteocytes in areas of newly made woven bone were highly immunoreactive (Fig. 4G and I). In contrast, no staining was detected in these cells on a control adjacent section (secondary antibody alone), as expected (Fig. 4H).

Fig. 4. Spatial localization of Max-1 and Msx-2. Immunohistochemical analysis of PF day 10 callus sections for Msx-1 and Msx-2. (A) Section near fracture site stained for Msx-1 showing positive labeling in cartilage (Ca), and fibrocartilage (Fca). (B) Safranin O fast/green staining of adjacent section within the fracture callus with same regions identified. (C) Adjacent section to (A) within the fracture callus illustrating the same regions (cartilage, fibrocartilage) stained for Msx-2. (D) Higher magnification view of cartilaginous zone in (A) (white box) demonstrating intense labeling in proliferating chondrocytes (black arrows) and pre-hypertrophic chondrocytes (white arrows). (E) Adjacent section to (A) and (7C) within the fracture callus stained with secondary antibody (negative control) section depicting no staining. (F) Higher magnification view of cartilaginous area in (C) (white box) showing robust staining in proliferating/pre- hypertrophic chondrocytes (black arrows) but no/weak staining in hypertrophic chondrocytes (white arrows). (G) Low magnification view of the callus (proximal to the fracture site) showing areas of muscle (Mu) and woven bone (Wb) with intense Msx-1 staining in periosteum (P) and osteoblasts, especially those that are still active below the periosteum (black arrows). Osteocytes that are found further away from the periosteum (bottom) do not express Msx- 1 as strongly. (H) Adjacent section to (G) and (I) stained with secondary antibody (negative control) and devoid of staining. (I) Adjacent sections to (G) illustrating positive Msx-2 immunoreactivity in periosteum (P) and woven bone (Wb), particularly in osteoblasts (black arrows) and even osteocytes throughout the entire Wb region. Scale bars represent 50 m in (A), (B), (C), (E), (G, (H), (I) and 20 M in (D) and (F).

Discussion

In this study, we chose to examine the temporal and spatial expression of five Hox genes, Msx-1, Msx-2, rHox, Hoxa-2 and Hoxd- 9. We specifically focused on these genes since they have been implicated in limb bud formation during embryogenesis [15] and because other genes that are expressed during limb formation have also been found to be reactivated during fracture repair [7,33]. Thus, we hypothesized that Hox genes, critical regulators of the embryonic skeleton, would also be reactivated during bone regeneration.

Hox gene expression was demonstrated as crucial to body segmentation and patterning during embryogenesis with specific Hox members involved in morphogenesis by activating other transcription factors and/or signaling molecules [17]. Mutations in these genes result in severe malformations of specific organs/tissues and even death [35]. More relevant to this study, Hox genes were also implicated in the process of embryonic skeletogenesis. For example, Pbx-1 deficient mice showed defects in their axial and appendicular skeleton, while disruption of rHox resulted in deformity in, or lack of, craniofacial and limb development as well as vertebral skeletal structures [25,32].

rHox deficient mice also revealed that this gene is active in regulating skeletal development and osteoblastic differentiation (binds to both collagen type I and osteocalcin promoters) [18,19]. Upon further examination of the mutant phenotype, a defect in the formation and growth of chondrogenic and osteogenic precursors was identified and was suggested that rHox regulates the formation of pre-skeletal condensations from undifferentiated mesenchyme [25]. These conclusions are consistent with findings presented here, especially our in situ hybridization data that reveal increased rHox expression in periosteal osteoprogenitor cells and young active osteoblasts found in areas of woven bone. In addition, we also observed strong expression of rHox in proliferating/pre- hypertrophic chondrocytes, consistent with previous data that localized rHox in chondrocytes of developing long bones at embryonic day 15, as well as newborn mice [25]. Further, the steady increase of rHox expression at PF day 10, a time where chondrogenesis and intramembranous ossification are occurring simultaneously, also supports the notion that rHox serves as a regulator for both osteoblast and chondrocyte differentiation.

In the case of Msx-1 and Msx-2, homologous genes to the Drosophila muscle segment homeobox gene (Msh) [16,29], prior findings report that they may have redundant functions as transcriptional repressors [1]. This similarity is linked to their identical DNA coding sequence, except for a discrepancy found in the N-terminus of each gene which bestows Msx-2 a greater affinity for DNA binding but makes Msx-1 a more effective represser [1]. Both genes were found strongly expressed in developing craniofacial regions in an overlapping manner [24], with the exception that Msx- 1 is expressed even during post-natal stages, whereas Msx-2 expression declines after birth. Consistent with these data, our results also indicate that both Msx-1 and Msx-2 are expressed by the identical cell types within the fracture callus. The expression of both genes was found to be very high (~3- to 13-fold) above mRNA levels seen in intact bone, and localized in periosteal osteoprogenitor cells, young active osteoblasts and proliferating/ prehypertrophic chondrocytes. Furthermore, Msx-1 was showed to be downregulated in periosteal osteoblasts at birth but expressed postnatally in alveolar bone processes [27]. Although initially this might be contradictory to our results that show both Msx-1 and Msx- 2 abundantly expressed by periosteum osteoblasts, when one considers the notion that fracture repair recapitulates embryonic development, reactivation of Msx-1 and Msx-2 in periosteum-derived osteoblasts is then consistent and even expected since it is known that these cells originate from mesenchymal progenitors.

Further, Msx genes promote proliferation while blocking terminal differentiation, as was shown by the downregulation of terminal differentiation markers osteocalcin and collagen type I through Msx- 2 [6,26] and cbfa/Runx2 by Msx-1 [3]. This is also supported by our data, especially by the reduced expression of both Msx-1 and Msx-2 detected in terminally differentiated osteocytes and hypertrophie chondrocytes. In contrast, robust expression of both genes was detected in active osteoblasts and proliferating chondrocytes. Further, suppression of terminal differentiation by Msx-1 and Msx-2 would explain the incomplete or delayed ossification of calvarial bones in Msx-2 deficient mice [31], as well as the complete deficiency of calvarial ossification in the Msx-1 and Msx-2 double mutant mice [1]. This is probably due to the overall decrease in differentiated osteoblasts required to seal the cranial sutures. Lastly, abnormal cartilage and endochondral ossification were also observed in Msx-2 knockout mice leading to a reduction in both axial and appendicular skeletal lengths [31]. Clearly, additional experiments are required to conclusively determine the precise role(s) of Msx-1 and Msx-2, especially in periosteal osteoprogenitor differentiation during the progression of the fracture callus.

While Hoxa-2 knockouts and deletions were generated, studies on these animals primarily focused on the patterning of the branchial arches and the mammalian hindbrain [2,9,28]. However, even in these studies a link can be found between Hoxa-2 and skeletal development. The Hoxa-2 mutant mice exhibited multiple cranial defects, including duplication of ossification centers of middle ear bones as well as replacement of the second branchial arch elements with the first set [9]. In subsequent experiments it was demonstrated that Hoxa-2 directs proper formation of the second arch by preventing chondrogenesis and intramembranous ossification [20]. Our data is in agreement with the fact that Hoxa-2 is involved in chondrogenesis andosteogenesis (ossification) during facture repair, since we mapped the upregulated expression of this gene in both active osteoblasts and chondrocytes. In addition, osteoprogenitor cells in the periosteum also express high levels of Hoxa-2. However, an interesting inference may be made from the temporal peak in expression on PF day 21. Given the critical role of Hoxa-2 in skeletal patterning, this late peak in expression during fracture repair may indicate a yet undefined role of Hoxa-2 in endochondral ossification or bone remodeling, known to occur at this time in the callus.

While Hoxd-9 has been examined in systems ranging from locomotor behavior to cervical cancer [5,22], its role has not been very well described during skeletogenesis. The single Hoxd-9 deficiency study conducted showed malformations of deltoid crest and reduction of the humerus in Hoxd-9 -/- mice [8]. In addition, a targeted disruption study of Hoxd-9 and Hoxd-10 found alterations in axial and appendicular skeletal structures, specifically the presence of an additional S1 vertebrae, as well as fusion of transverse processes with a high rate of four fused sacral vertebrae [5]. Collectively, these results support our findings that Hoxd-9 is involved in fracture callus development, especially the early stages (PF day 3- 10), consistent with the spatial analysis where we mapped the expression of Hoxd-9 in proliferating chondrocytes and osteoblasts, as well as in the periosteum. These results suggest that Hoxd-9 plays a role in intramembranous ossification and chondrogenesis rather than endochondral ossification which occurs at later times (PF 14 and 21).

The in vivo analyses of Msx-1, Msx-2, rHox, Hoxa-2 and Hoxd-9 described above, indicate that each gene is involved in skeletal development in a very specific way. As we expected, our hypothesis that Hox genes, critical regulators of the embryonic skeleton, will also be reactivated during bone regeneration was verified by the data presented here. To our knowledge, this study is one of the first to examine the reactivation of Hox genes during bone regeneration. Previously, another study showed that the rat homolog of the homeoprotein distal-less (Dlx) was found to be upregulated during fracture at PF day 2-30 [34], This was not a surprising finding since Dlx has been shown to regulate the expression of osteocalcin which promotes osteoblast differentiation [30]. Together with our results, these data indicate that all five Hox genes studied are reactivated during fracture repair and display differential upregulated patterns of expression by specific callus cell types. It is clear from the expression studies that all of the Hox genes analyzed, as well as other members of their respective families (including Dlx), may collectively function together to control the regeneration of the various anatomical tissues of the callus as they develop, mature and eventually remodel into lamellar cortical bone. While the precise molecular mechanisms of action and functional contributions for these genes during fracture repair remains unknown, this study provides supplementary evidence strongly supporting the notion that bone regeneration recapitulates development. Obviously, additional studies are needed to fully elucidate the exact roles of these genes in the fracture repair process.

Acknowledgments

We thank Alan Ka and Fayez Safadi for help in developing the in situ hybridization protocol, David Komatsu for insightful discussions, Jonathan Chiu for critically reading the manuscript and Rosemary Gaynor for secretarial support. This work was generously supported by a grant from NASA (NAG2-1517) and the Aircast Foundation (F600R) to MH.

References

[1] Alappat S, Zhang ZY, Chen YP. Msx homeobox gene family and craniofacial development. Cell Res 2003;13(6):429-42.

[2] Barrow JR, Stadler HS, Capecchi MR. Roles of Hoxal and Hoxa2 in patterning the early hindbrain of the mouse. Development 2000;127(5):933-44.

[3] Blin-Wakkach C, Lezot F, Ghoul-Mazgar S, et al. Endogenous Msx1 antisense transcript: in vivo and in vitro evidences, structure, and potential involvement in skeleton development in mammals. Proc Natl Acad Sci USA 2001;98(13):7336-41. E-pub: 2001 June 05.

[4] Bonnarens F, Einhorn TA. Production of a standard closed fracture in laboratory animal bone. J Orthop Res 1984;2:97-101.

[5] de la Cruz CC, Der-Avakian A, Spyropoulos DD, et al. Targeted disruption of Hoxd9 and Hoxd10 alters locomotor behavior, vertebral identity, and peripheral nervous system development. Dev Biol 1999;216(2):595-610.

[6] Dodig M, Kronenberg MS, Bedalov A. et al. Identification of a TAAT-containing motif required for high level expression of the COLlAl promoter in differentiated osteoblasts of transgenic mice. J Biol Chem 1996:271(27): 16422-9.

[7] Ferguson C, Alpern E. Miclau T, Helms JA. Docs adult fracture repair recapitulate embryonic skeletal formation? Mech Dev 1999;87(1- 2):57-66.

[8] Fromental-Ramain C, Warot X, Lakkaraju S, et al. Specific and redundant functions of the paralogous Hoxa-9 and Hoxd-9 genes in forelimb and axial skeleton patterning. Development 1996:122(2):461 72.

[9] Gendron-Maguire M. Mallo M, Zhang M, Gridley T. Hoxa-2 mutant mice exhibit homeotic transformation of skeletal elements derived from cranial neural crest. Cell 1993;75(7):1317-31.

[10] Gerstenfeld LC, Cullinane DM, Barnes GL, Graves DT, Einhorn TA. Fracture healing as a post-natal developmental process: molecular, spatial, and temporal aspects of its regulation. J Cell Biochem 2003;88(5):873-84.

[11] Hadjiargyrou M, Ahrens W, Rubin CT. Temporal expression of the chondrogenic and angiogenic growth factor CYR61 during fracture repair. J Bone Miner Res 2000:15:1014-23.

[12] Hadjiargyrou M, Halsey MF, Ahrens W, et al. Cloning of a novel cDNA expressed during the early stages of fracture healing. Biochem Biophys Res Commun 1998:249:879-84.

[13] Hadjiargyrou M. Lombarde F, Zhao S, et al. Transcriptional profiling of bone regeneration. Insight into the molecular complexity of wound repair. J Biol Chem 2002;277(33):30177-82.

[14] Hadjiargyrou M, Rightmire EP, Ando T, Lombarde FT. The El 1 osleoblastic lineage marker is differentially expressed during fracture healing. Bone 2001;29:149-54.

[15] Hall BK, Miyake T. All for one and one for all: condensations and the initiation of skeletal development. Bioessays 2000:22(2): 138-47.

[16] Hill RE, Jones PF, Rees AR, et al. A new family of mouse homeo box-containing genes: molecular structure, chromosomal location, and developmental expression of Hox-7.1. Genes Dev 1989:3(1): 26-37.

[17] Hombria JC, Lovegrove B. Beyond homeosis HOX function in morphogenesis and organogenesis. Differentiation 2003:71(8): 461- 76.

[18] Hu Y, Flanagan J, Brennan DP, et al. rHox: a homeobox gene expressed in osteoblastic cells. J Cell Biochem 1995;59(4):486-97.

[19] Hu YS. Zhou H, Kartsogiannis V, et al. Expression of rat homeobox gene, rHOX, in developing and adult tissues in mice and regulation of its mRNA expression in osteoblasts by bone morphogenelic protein 2 and parathyroid hormone-related protein. Mol Endocrinol 1998:12(11): 1721-32.

[20] Kanzler B, Kuschert SJ, Liu YH, Mallo M. Hoxa-2 restricts the chondrogenic domain and inhibits bone formation during development of the branchial area. Development 1998:125(14): 2587- 97.

[21] Komatsu DE, Hadjiargyrou M. Activation of the transcription factor HIF-I and its target genes, VEGF. HO-1. iNOS, during fracture repair. Bone 2004;34(4):680-8.

[22] Li H, Huang CJ, Choo KB. Expression of homeobox genes in cervical cancer. Gynecol Oncol 2002;84(2):216-21.

[23] Lombardo F, Komatsu D, Hadjiargyrou M. Molecular cloning and characterization of Mustang, a novel nuclear protein expressed during skeletal development and regeneration. FASEB J 2004;18(1):52- 61.

[24] MacKenzie A, Ferguson MW, Sharpe PT. Expression patterns of the homeobox gene, Hox-8, in the mouse embryo suggest a role in specifying tooth initiation and shape. Development 1992:115(2): 403- 20.

[25] Martin JF, Bradley A, Olson EN. The paired-like homeo box gene MHox is required for early events of skeletogenesis in multiple lineages. Genes Dev 1995;9(10): 1237-49.

[26] Newberry EP, Boudreaux JM, Towler DA. Stimulus-selective inhibition of rat osteocalcin promoter induction and protein-DNA interactions by the homeodomain represser Msx2. J Biol Chem 1997:272(47):29607-13.

[27] Orestes-Cardoso S, Nefussi JR, Eezot F, et al. Msx1 is a regulator of bone formation during development and postnatal growth: in vivo investigations in a transgenic mouse model. Connect Tissue Res 2002:43(2-3): 153-60.

[28] Rijli FM, Mark M, Lakkaraju S, et al. A homeotic transformation is generated in the rostral branchial region of the head by disruption of Hoxa-2, which acts as a selector gene. Cell 1993; 75(7): 1333-49.

[29] Robert B, Sassoon D, Jacq B, et al. Hox-7, a mouse homeobox gene with a novel pattern of expression during embryogenesis. EMBOJ 1989;8(1):91-100.

[30] Ryoo HM, Hoffmann HM, Beumer T, et al. Stage-specific expression of Dlx-5 during osleoblast differentiation: involvement in regulation of osteocalcin gene expression. Mol Endocrinol 1997;11(11):1681-94.

[31] Satokata I, Ma L, Ohshima H, et al. Msx2 deficiency in mice causes pleiotropic defects in bone growth and ectodermal organ formation. Nat Genet 2000;24(4):391-5.

[32] Selleri L, Depew MJ, Jacobs Y, et al. Requirement for Pbx1 in skeletal patterning and programming chondrocyte proliferation and differentiation. Development 2001;128(18):3543-57.

[33] Vortkamp A, Pathi S, Peretti GM, et al. Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev 1998;71(1-2):65-76.

[34] Yaoita H, Orimo H, Shirai Y, Shimada T. Expression of bone morphogenetic proteins and rat distal-less homolog genes following rat femoral fracture. J Bone Miner Metab 2000;18(2):63-70.

[35] Zakany J, Kmita M, Duboule D. A du\al role for Hox genes in limb anterior-posterior asymmetry. Science 2004;304(5677): 1669-72.

Robert P. Gersch, Frank Lombardo1, Scott C. McGovern 2, Michael Hadjiargyrou *

Department of Biomedical Engineering, Stony Brook University, Psychology A Building, Stony Brook, NY 11794-2580, USA

* Corresponding author. Tel.: +1 631 632 1480; fax: +1 631 6328577.

E-mail address: michael.hadjiargyrou@sunysb.edu (M. Hadjiargyrou).

1 Present address: Department of Orthopaedic Surgery, Long Island Jewish Medical Center, 270-05, 76th Avenue, New Hyde Park, New York 11040, USA.

2 Present address: Department of Orthopaedic Surgery, Mayo Clinic, Rochester, MN 55905, USA.

Copyright Journal of Bone and Joint Surgery, Inc. Jul 2005

Source: Journal of Orthopaedic Research

More News in this Category