Expression of chondro-osteogenic BMPs in ossified failed tendon healing model of tendinopathy

Authors

  • Pauline Po Yee Lui,

    Corresponding author
    1. Department of Orthopaedics and Traumatology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    2. The Hong Kong Jockey Club Sports Medicine and Health Sciences Centre, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    3. Program of Stem Cell and Regeneration, School of Biomedical Science, The Chinese University of Hong Kong, Hong Kong SAR, China
    • Department of Orthopaedics and Traumatology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China. T: (852)2632-3072; F: (852)2646-3020.
    Search for more papers by this author
  • Yin Mei Wong,

    1. Department of Orthopaedics and Traumatology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    2. The Hong Kong Jockey Club Sports Medicine and Health Sciences Centre, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Yun Feng Rui,

    1. Department of Orthopaedics and Traumatology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    2. The Hong Kong Jockey Club Sports Medicine and Health Sciences Centre, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Yuk Wa Lee,

    1. Department of Orthopaedics and Traumatology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    2. The Hong Kong Jockey Club Sports Medicine and Health Sciences Centre, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Lai Shan Chan,

    1. Department of Orthopaedics and Traumatology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    2. The Hong Kong Jockey Club Sports Medicine and Health Sciences Centre, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Kai Ming Chan

    1. Department of Orthopaedics and Traumatology, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    2. The Hong Kong Jockey Club Sports Medicine and Health Sciences Centre, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author

Abstract

Chondrocytes phenotype/markers were expressed in clinical samples of tendinopathy and calcifying tendinopathy. This study examined the spatial-temporal expression of chondro-osteogenic Bone Morphogenetic Proteins (BMPs), which might contribute to ectopic chondro-osteogenesis and failed healing process in tendinopathy. Collagenase was injected into patellar tendon of rats to induce ossified failed tendon healing. At week 2, 4, 8, 12, and 16, the patella tendon was harvested for immunohistochemical staining and analysis of BMP-2/4/7. BMP-4/7 showed similar expression patterns, which was different from BMP-2. The expression of BMP-2 in the tendon matrix increased at week 2 and was reduced to nearly undetectable level afterwards except at the chondro-ossification sites. However, the expression of BMP-4/7 in the healing tendon fibroblast-like cells and matrix increased at week 2, reduced at week 4 and 8 and increased again at week 12 and 16, consistent with transient healing at week 8 in this animal model. There was increasing strong expression of BMP-4/7 in the chondrocyte-like cells in the un-ossified and ossified areas from week 8–16. BMP-4/7, besides BMP-2, might also contribute to ectopic chondro-osteogenesis and failed healing in tendon injuries. BMP-4/7, but not BMP-2, might be involved in regulating late events in ossified failed tendon healing. © 2010 Orthopaedic Research Society Published by Wiley Periodicals, Inc. J Orthop Res 29:816–821

Chronic tendinopathy refers to a broad spectrum of pathological conditions in tendons and their insertion, with symptoms including activity-related chronic pain, which is prevalent among athletes and sedentary subjects. Despite its prevalence, its underlying pathogenesis is poorly understood and treatment is usually symptomatic. Failed healing was suggested as the pathological mechanism of tendinopathy.1 Chondrocyte phenotype/markers were expressed in clinical samples of tendinopathy and calcifying tendinopathy.2, 3 We reported the presence of chondrocyte phenotype and ectopic ossification in our collagenase-induced failed tendon healing model.4 We have shown this collagenase animal model to reproduce many key histopathological features of clinical tendinopathy and can be used to study the ossified failed healing process in tendinopathy.5 We further reported the expression of bone morphogenetic protein-2 (BMP-2) protein, an osteogenic growth factor, around the chondrogenic and ossifying sites in the same animal model,6 suggesting that BMP-2 might be involved in the pathogenesis. This was further supported by ectopic over-expression of BMPs in the subacromial bursa of patients with chronic degeneration of the rotator cuff.7 Our result also showed that repetitive cyclic loading could increase the expression of BMP-2 in tendon-derived stem cells (TDSCs) and that BMP-2 could induce the osteogenic differentiation of TDSCs in vitro.8 To date, more than 20 BMPs have been identified and many of them were chondro-osteogenic. We hypothesized that other chondro-osteogenic BMPs such as BMP-4 and BMP-7 might also express during the ossified failed healing process and there might be spatial and temporal difference in the expression patterns of these chondro-osteogenic BMPs which regulated the ossified failed healing process. This study therefore aimed to investigate the expressions of BMP-2/4/7 in this animal model by immunohistochemistry and image analysis.

MATERIALS AND METHODS

Collagenase-Induced Injury

This study was approved by the Animal Research Ethics Committee of the authors' institution. Thirty male Sprague-Dawley rats, (8 weeks, weight 200–250 g) were used.4 After anaesthesia with 2.5% pentobarbital (4.5 mg/kg body weight), hairs over the lower limb were shaved. Patellar tendon was located by positioning the knee at 90°. Twenty microliters (0.015 mg/µl in 0.9% saline, i.e., 0.3 mg) of bacterial collagenase I (Sigma–Aldrich, St Louis, MO) were injected into the patellar tendon intratendinously with a 30 G needle in one limb while the contralateral limb was injected with saline. The rats were allowed free cage movement immediately after injection. At week 2, 4, 8, 12, and 16 after injury, the rats were sacrificed and the patellar tendons in both limbs were harvested (n = 6/group/time point) for immunohistochemical staining and semi-quantitative analysis of BMP-2, BMP-4, and BMP-7.

Immunohistochemistry

The patellar tendon was washed in PBS, fixed in buffered formalin and 100% ethanol, and embedded in paraffin. Consecutive 5-µm thick sections in the middle of the patellar tendon were cut longitudinally and mounted on coated slides for immunohistochemical staining of BMP-2, -4, -7. Immunohistochemistry was done as described previously.5 After deparaffination, the sections were rehydrated, decalcified, quenched of endogenous peroxidase activity, and subjected to antigen retrieval. After blocking with 5% normal donkey serum (for BMP-2)/5% normal goat serum (for BMP-4, -7), the sections were incubated with specific antibodies against BMP-2 (Santa Cruz Biotechnology, Santa Cruz, CA; 1:100) and BMP-4, -7 (Abcam, Cambridge, MA) at 4°C overnight. For BMP-2, donkey anti-goat horseradish peroxidase (HRP)-conjugated secondary antibody (Santa Cruz Biotechnology; 1:100) was then added for an hour, followed by 3,3′ diaminobenzidine tetrahydrochloride (DAKO, Glostrup, Denmark) in the presence of H2O2. For BMP-4, -7, immunohistochemical localization was visualized with anti-rabbit HRP/DAB UltraVision Detection System (Lab Vision, Fremont, CA). Afterwards, the sections were rinsed, counterstained in hematoxylin, dehydrated with graded ethanol and xylene, and mounted with p-xylene-bis-pyridinium bromide (DPX) permount (Sigma–Aldrich). Primary antibody was replaced with blocking solution in the controls. All incubation times and conditions were strictly controlled. Samples from control and injury groups were stained in the same batch. The sections were examined under light microscopy (DMRXA2, Leica Microsystems Wetzlar GmbH, Germany). Positive controls from rat fracture callus, brain, and liver were included to verify the specificities of the protein expressions of BMP-2, -4, -7, respectively.

Image Analysis

The method is well-established in our laboratory.9 Briefly, to analyze the immunopositive signals of BMPs after tendon injury, one photomicrograph from one section of each tendon sample was taken at 12.5× magnification for the whole tendon mid-substance excluding the paratendon under the same camera setting. We measured the immunpositive signals using the Image Pro Plus software (MediaCybernetics, Bethesda, MD). First, the region of interest (ROI), which was the whole tendon mid-substance excluding the ossified regions was selected manually. The ossified region in the patellar tendon was excluded for analysis because it stained heavily with hematoxylin without decalcification and might affect the result of analysis. Segmentation of the image was then performed with the “Select Colors” command of the software. Three histograms on hue, saturation, and intensity, respectively, would appear. The image was then segmented with hue range of 1–33, saturation range of 1–255 and intensity range of 1–215 to select the light to dark brown colour. This huge-saturation-intensity combination was determined based on the samples showing the highest and the lowest expression to ensure that we covered the whole range of light to dark brown colour that we would like to measure while minimizing the interference from the background colour of the image. To ensure reproducibility, the same huge-saturation-intensity combination was used for all samples. The Integrated Optical Density (IOD) (in arbitrary unit) of the immunopositive signal of BMPs was then measured using the “Count” command of the software. The area of ROI was also measured. The IOD/µm2 for each tendon sample was reported. The assessor was blinded to the time points during image analysis.

Data analysis

The mean IOD/µm2 of BMP-2, -4, and -7 were presented in boxplots. To compare the immuno-positive signal among different time points, Kruskal–Wallis test followed by post-hoc test using Mann–Whitney U-test was performed. All the data analysis was done using SPSS (SPSS Inc, Chicago, IL, version 16.0). p < 0.05 was regarded as statistically significant.

RESULTS

Immunohistochemistry of BMP-2

No immunopositivity of BMP-2 was observed in the tendon mid-substance in the saline controls (Figure 1A). At week 2 after collagenase-induced injury, moderate, patchy expression of BMP-2 was observed in the tendon matrix, particularly in areas with poor matrix organization and high cellularity (Figure 1B). At week 4, the immunopositive signal was reduced, and no to slight signal was observed in the tendon matrix (Figure 1C). The immunopositive signal at the tendon matrix remained low from week 8 to week 16 (Figure 1D–F). Weak to moderate expression of BMP-2 was observed in the chondrocyte-like cells in un-ossified area in one sample with such cells at week 4 and all samples from week 8 to week 16 (week 8: Figure 1D, arrowheads). Chondrocyte-like cells surrounding and inside the ossified deposits appeared in one sample at week 8 and all samples at week 12 and week 16. There was weak BMP-2 signal in these cells and their surrounding matrix at week 8 and week 12 (Figure 1E, CR) and the expression increased at week 16 (Figure 1F, CR). There was moderate expression of BMP-2 in the marrow-like cells inside the ossified deposits at week 16 (Figure 1F, insert, ▴). Weak to moderate expression of BMP-2 was observed in blood vessels in the tendon mid-substance at week 16 (Figure 1F, insert, △). Overall, the expression of BMP-2 increased significantly at week 2 after collagenase injection (post-hoc comparison with week 16 control: p = 0.004; overall p = 0.006) (Figure 2A). Its expression was reduced steadily afterwards except in areas with chondrocyte-like cells and ossified deposits, though it remained significantly higher than that in the saline control (post-hoc p-values compared to control ranged from 0.004 to 0.014) (Figure 2A).

Figure 1.

Photomicrographs showing the immunohistochemical staining of BMP-2 at week-16 saline control (A), week 2 (B), week 4 (C), week 8 (D), week 12 (E), and week 16 (F) after collagenase-induced tendon injury. Magnification: 400×, arrowhead: chondrocyte-like cells; CR: ossified region; ▴: marrow-like cells; △: blood vessels.

Figure 2.

Boxplot showing the changes of mean IOD/µm2 of BMP-2 (A), BMP-4 (B), and BMP-7 (C) at different times after collagenase-induced tendon injury. Expression of BMP at week 16 in the saline group was plotted for comparison. a: p < 0.050 compared with saline group; (b) p < 0.050 compared with week 2 after collagenase injection; (c) p < 0.050 compared with week 8 after collagenase injection in post-hoc comparisons.

Immunohistochemistry of BMP-4

There was no immunopositive signal of BMP-4 in the tendon mid-substance in the saline controls (Figure 3A). At week 2, after collagenase injection, there was intense expression of BMP-4 in the healing tendon fibroblast-like cells and matrix (Figure 3B). The expression in the tendon fibroblast-like cells and matrix was reduced at week 4 (Figure 3C) and week 8 (Figure 3D), consistent with the transient healing at week 8 in this animal model. However, the expression was elevated again at week 12 (Figure 3E) and week 16 (Figure 3F). There was weak immunopositive signal in the chondrocyte-like cells that appeared in one sample at week 4. The expression in the un-ossified chondrocyte-like cells increased afterwards, with strong expression from week 8 (Figure 3D, arrowheads) to week 16. There was strong expression in the chondrocyte-like cells and matrix surrounding and inside the ossified deposits from week 8 to week 16 (week 12 and week 16: Figure 3E,F, CR). Strong expression was also observed in the marrow-like cells inside the ossified deposits at week 12 and week 16 (Figure 3E,F, insert, ▴). There was also strong immunopositive signal in the blood vessels in the tendon mid-substance at week 12 and 16 (Figure 3E,F, insert, △). Overall, the expression of BMP-4 increased significantly at week 2 after collagenase injection, reached a minimum at week 8, consistent with transient healing in this animal model, and then increased again at week 12 and week 16 (overall p = 0.004) (Figure 2B). The expression of BMP-4 at all time points after injury was significantly higher than that in the saline control (post-hoc p-values compared to control ranged from 0.004 to 0.014) (Figure 2B). There were significantly lower expression of BMP-4 at week 8 (post-hoc p = 0.028) and week 16 (post-hoc p = 0.015) compared to that at week 2 (Figure 2B).

Figure 3.

Photomicrographs showing the immunohistochemical staining of BMP-4 at week-16 saline control (A), week 2 (B), week 4 (C), week 8 (D), week 12 (E), and week 16 (F) after collagenase-induced tendon injury. Magnification: 400×, arrowhead: chondrocyte-like cells; CR: ossified region; ▴: marrow-like cells; △: blood vessels.

Immunohistochemistry of BMP-7

There was no immunopositive signal of BMP-7 in the tendon mid-substance in the saline controls (Figure 4A). The spatial and temporal expression profile of BMP-7 was similar to that of BMP-4 after collagenase injection. At week 2, there was weak (3/6) to strong (3/6) expression of BMP-7 in the healing tendon fibroblast-like cells and matrix (Figure 4B). At week 4 (Figure 4C) and week 8 (Figure 4D), the expression in the healing tendon fibroblast-like cells and matrix was reduced and only weak expression was observed, consistent with transient repair. However, the expression in the healing tendon fibroblast-like cells and matrix increased again to moderate level at week 12 (Figure 4E) and week 16 (Figure 4F). There was weak expression of BMP-7 in un-ossified chondrocyte-like cells at week 4 and week 8 (week 8: Figure 4D, arrowheads) and the expression in these cells increased at week 12 and week 16. There was strong expression of BMP-7 in the chondrocyte-like cells and matrix surrounding and inside the ossified deposits from week 8 to week 16 (week 12 and week 16: Figure 4E,F, CR). Strong immunopositive signal was also observed in marrow-like cells inside the ossified deposits at week 12 and week 16 (Figure 4E,F, insert, ▴). There was strong expression of BMP-7 in the blood vessels in the tendon mid-substance at week 12 (Figure 4E, insert, △) and 16 (Figure 4F, △). Overall, the expression of BMP-7 increased significantly at week 2 after collagenase injection, reduced at week 4 and week 8 and then increased again at week 12 and week 16 (overall p = 0.001) (Figure 2C). The expression of BMP-7 at all time points after collagenase injection was significantly higher than that in the saline control in the post-hoc comparison (post-hoc p-values compared with control ranged from 0.004 to 0.028) (Figure 2C). The expression of BMP-7 was significantly reduced at week 8 compared to that at week 2 (post-hoc p = 0.028) and then significantly increased at week 12 (post-hoc p = 0.009) and week 16 (post-hoc p = 0.006) compared to that at week 8 (Figure 2C).

Figure 4.

Photomicrographs showing the immunohistochemical staining of BMP-7 at week-16 saline control (A), week 2 (B), week 4 (C), week 8 (D), week 12 (E), and week 16 (F) after collagenase-induced tendon injury. Magnification: 400×, arrowhead: chondrocyte-like cells; CR: ossified region; ▴: marrow-like cells; △: blood vessels.

DISCUSSION

The pathogenesis of tendinopathy, including the cause of tendon pain, tendon degeneration, and ossification, remained largely unknown and hence current treatments are usually symptomatic. Failed healing was suggested as the pathological mechanism of tendinopathy. Change of tendon loading due to mechanical overload, compression or disuse have been implicated as the possible etiologies,10 but they do not completely explain the cellular and molecular alternations seen in the diseased tendon such as chondrometaplasia and ectopic ossification, hypercellularity, vascularity, and extracellular matrix degeneration. We speculated that the ectopic expression of chondro-osteogenic BMPs might contribute to ectopic chondro-ossification and failed healing in tendinopathy. This study therefore aimed to investigate the spatial and temporal expression of BMP-2, -4, and -7 in the failed tendon healing animal model.

Our results showed that there was increased expression of BMP-4 and -7, besides BMP-2, in the chondrocyte-like cells and ossified deposits in this animal model, suggesting that they might also contribute to ectopic chondro-ossification and failed healing in tendon injuries. Both the spatial and temporal expression profiles of BMP-4 and BMP-7 were very similar, with a transient decrease at week 4 and week 8. This was consistent with the transient healing at week 8, followed by the development of failed healing in this animal model.4 On the other hand, the expression of BMP-2 increased at week 2 after collagenase injection and was reduced to nearly undetectable level afterwards except in areas with chondrocyte-like cells and ossified deposits. It indicated that BMP-4 and -7, but not BMP-2, might be involved in regulating late events in ossified failed tendon healing. The expression of BMP-2,-4, and -7 has also been reported in the subacromial bursa tissue of patients with chronic tears of the rotator cuff,7 supporting the observation in our animal model about the role of these chondro-osteogenic BMPs in tendinopathy. BMP-2, -4, -7 were also reported to express in ossified matrix, chondrocytes, and fibroblasts near the ossified areas and have also been suggested to play roles in the ectopic ossification of spinal ligaments.11 Recently, we have also observed the expression of BMP-2, -4, and -7 in clinical samples of un-calcifying and calcifying tendinopathy, further supporting the findings in this animal model and the potential roles of chondro-osteogenic BMPs in the pathogenesis of tendinopathy (unpublished results). Rui YF, Lui PPY, Rdf CG, Wong YM, Lee YW, Chan KM, Expression of chondro osteogenic BMPs in clinical samples of Patellar Tendinopathy – A Histopatho – logical study. (Unpublished)

We have previously reported the earlier expression of chondrogenic markers, sox9, and collagen type II, in the healing tendon fibroblast-like cells at week 2 followed by their expression in the chondrocyte-like cells and ossified deposits, which appeared at week 4 and week 12, respectively, in the failed tendon healing animal model.4 These observations led us to speculate that erroneous differentiation of tendon progenitor cells to chondrocytes or osteoblasts might lead to ectopic chondro-ossification and failed healing in this animal model.12 This speculation was further supported by the results in this study as we also observed earlier expression of BMP-2, -4, and -7 in the healing tendon fibroblast-like cells prior to their expression in the chondrocyte-like cells and ossified deposits. The present data further suggested that ectopic expression of chondro-osteogenic BMPs might be one group of factors leading to erroneous differentiation of tendon progenitors during tendon injury.

Both mechanical and biological factors might contribute to the ectopic expression of BMPs. It was known that the expression of BMP-2, -4, and -7 were sensitive to mechanical load.13, 14 A share stress-responsive element was found in the promoter region of mouse BMP-215 and mouse BMP-4.16 Repetitive cyclic loading increased the expression of BMP-2 in TDSCs and BMP-2 could induce the osteogenic differentiation of TDSCs with increase alkaline phosphatase (ALP) activity and matrix mineralization in vitro.8 The role of mechanical loading and/or BMP signaling in erroneous stem cell differentiation and heterotropic calcification was also reported in other diseases.17, 18

It was reported that tendon progenitor cells isolated from biglycan- and fibromodulin-knockout mice were more sensitive to BMP-2 stimulation with increased phosphorylation of Smad1, Smad5, and Smad8 as well as with more abundant nuclear localization of phosphorylated Smad1 than those of wild type cells.19 Changes in the composition of the extracellular matrix therefore might also affect the cellular response of tendon stem/progenitor cells and promote their aberrant differentiation to osteoblasts and chondroblasts. We and others have reported the changes in the extraceullar matrix (ECM) composition with increase in proteogylcans (PG) in collagenase-induced tendon injury animal model and increase in sulphated glycosaminoglycans (GAGs) in clinical samples of tendinopathy, respectively.20, 21 Reduction of sulphated GAGs in PG on Mesenchymal Stem Cells (MSCs) surface was reported to reduce the effect of endogenous BMPs and exogenous BMP-2 on osteogenic gene expression and ALP activity, suggesting that changes in ECM composition might modulate the effects of BMPs.22 The damage to the ECM as a result of intratendinous collagenase injection in this animal model and the increased expression of collagenase in clinical samples of tendinopathy23 might increase the sensitivity of tendon progenitor cells to mechanical loading, resulting in increased production of chondro-osteogenic BMPs which promoted ectopic chondro-ossification and failed healing. We need further experiments to prove this hypothesis.

Activities of BMPs are inhibited extracellularly by BMP-binding proteins such as noggin and chordin as well as intracellularly by Smad6, tob, and Smurf1.24 Implantation of noggine gene or muscle-derived stem cells over-expressing noggin were reported to inhibit heterotopic ossification induced by BMP-4, demineralized bone matrix, and trauma in animal models.25, 26 The use of these inhibitors might inhibit ectopic chondro-osteogenesis and promote tendon healing in our collagenase-induced injury model and tendinopathy. Further studies are required to understand the effectiveness of these therapies for the management of tendinopathy.

CONCLUSION

In conclusion, we reported ectopic expression of BMP-4 and BMP-7, in addition to BMP-2, in the chondro-ossification sites in the failed tendon healing animal model. The ectopic expression of these chondro-osteogenic BMPs might induce tissue transformation into ectopic bone/cartilage and promoted structural degeneration during failed tendon healing. BMP-4/7 showed similar expression patterns. BMP4/7, but not BMP2, might be involved in regulating late events in ossified failed tendon healing.

Acknowledgements

This work was supported by equipment/resources donated by the Hong Kong Jockey Club Charities Trust.

Ancillary