SEARCH

SEARCH BY CITATION

Keywords:

  • tendon injury;
  • healing, chondrocyte phenotype;
  • ossification;
  • extracellular matrix

Abstract

  1. Top of page
  2. Abstract
  3. METHODOLOGY
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

The acquisition of chondro-osteogenic phenotypes and erroneous matrix deposition may account for poor tissue quality after acute tendon injury. We investigated the presence of chondrocyte phenotype, ossification, and the changes in the expression of major collagens and proteoglycans in the window wound in a rat patellar tendon window injury model using histology, von Kossa staining and immunohistochemistry of Sox 9, major collagens, and proteoglycans. Our results showed that the repair tissue did not restore to normal after acute injury. Ectopic chondrogenesis was observed in 33% of samples inside wound at week 4 while ectopic ossification surrounded by chondrocyte-like cells were observed in the window wound in 50% of samples at week 12. There was sustained expression of biglycan and reduced expression of aggrecan and decorin in the tendon matrix in the repair tissue. The erroneous deposition of extracellular matrix and ectopic chondro-ossification in the repair tissue, both might influence each other, might account for the poor tissue quality after acute injury. Higher expression of biglycan and aggrecan were observed in the ectopic chondro-ossification sites in the repair tissue, suggesting that they might have roles in ectopic chondro-osteogenesis. © 2011 Orthopaedic Research Society Published by Wiley Periodicals, Inc. J Orthop Res 30:37–46, 2012

Tendons regenerate and repair slowly and inefficiently after injury. The crimp pattern of collagen fibers and fibrils was smaller than that of the control1 and the regenerated fibrotic scar tissue could not return to its original mechanical strength for a long time after injury.1–3 It was known that collagens and proteoglycans in the extracellular matrix (ECM) have roles in modulating the activities of tenocytes in addition to their structural roles.4 Despite studies about the tendon healing process, there is still limited and inconsistent understanding about the change in biochemical composition of ECM after tendon injury5–7 and how does the ECM regulate cellular functions. On the other hand, ectopic ossification after midpoint tenotomy of rat or mouse Achilles tendon has been reported.8–12 Clinical studies have occasionally reported ectopic calcification after Achilles tendon rupture13, 14 as well as calcification and tendinopathic-like changes in patellar tendon donor site after anterior cruciate ligament (ACL) reconstruction.15–21 We hypothesized that chondro-osteogenesis and change in the ECM composition of the repair tissue after acute tendon injury might contribute to the poor tissue quality. This study therefore aimed to examine the presence of chondrocyte phenotype and ossification inside the window wound of the patellar tendon. The spatial-temporal changes of major collagens including collagen types I and III and major proteoglycans including decorin, biglycan, fibromodulin, and aggrecan in the window wound after tendon injury were also investigated.

METHODOLOGY

  1. Top of page
  2. Abstract
  3. METHODOLOGY
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Tendon Injury Model

This study was approved by the Animal Research Ethics Committee of the authors' institution. Eighteen Sprague–Dawley male adult rats (6–8 weeks, body weight of 260–280 g) were operated according to our well-established protocol.22 Under general anesthesia, an incision was made to expose the patellar tendon. The central one-third of the patellar tendon (1 mm × 4 mm) was then removed. Care was taken not to remove tissue from the tendon–bone junction. The wound was then closed in layers. The contralateral knee with skin injury only served as sham control. The rats were allowed free cage movement immediately after surgery. Buprenorphine hydrochloride (Temgesic®, Reckitt Benckiser Healthcare UK Ltd., Hull, East Yorkshire, UK) was injected intramuscularly at 0.03 mg/kg body weight for postoperative analgesia. At weeks 2, 4, and 12 after injury, rats were killed and both patellar tendons were harvested (n = 6 for each group) for histology as well as immunohistochemical staining and analysis of collagen type X, collagen type II, Sox 9, collagen type I, collagen type III, decorin, biglycan, fibromodulin, and aggrecan.

General Histology and Immunohistochemistry

The procedure for general histology has been established.22, 23 The patellar tendon was washed in PBS, fixed in buffered formalin and 100% ethanol, and embedded in paraffin. The paraffin block was trimmed to exclude the peritendon and sections in coronal plane were then collected. The slides were labeled sequentially according to the sectioning order. Around 50 sections at thickness of 5 µm were collected from each sample block. We selected the sections in the mid-portion of the tissue with similar depth for staining. The sections were mounted on 3-aminopropyl-triethoxy-silane (Sigma–Aldrich, St. Louis, MO)-coated slides. After deparaffination, the sections were stained with hematoxylin and eosin or von Kossa stain. Immunohistochemistry was done as described previously.22, 23 Briefly, after removal of paraffin and rehydration, the sections were decalcified with 9% formic acid and quenched with 3% hydrogen peroxide in methanol. Antigen retrieval was performed (Supplementary S1). After blocking with 5% normal donkey/goat/rabbit/donkey serum in 1% BSA/PBS, the sections were stained with specific primary antibodies at 4°C overnight (Supplementary S1). The spatial and temporal localization of the protein was visualized by incubating with secondary antibodies for an hour (Supplementary S1). For the use of biotinylated secondary antibodies, streptavidin and biotinylated-horseradish peroxidase (1:100) were added. 3,3′-Diaminobenzidine tetrahydrochloride (DAKO, Glostrup, Denamark) was then added in the presence of H2O2. 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. For good reproducibility and comparability, samples from control and injury groups were stained in the same batch under the same condition. The sections were examined under light microscopy (Leica DMRXA2) and polarization microscopy (Leica DMRB, both from Leica Microsystems Wetzlar GmbH, Wetzlar, Germany).

von Kossa Staining

One section from the mid-portion of each sample was deparaffinized and rehydrated with xylene and graded ethanol. The section was then incubated with 1% silver nitrate solution under light bulb illumination, followed by incubation in 5% sodium thiosulfate solution to remove the unreacted silver as described previously.22 The slide was then washed, counterstained with 0.2% nuclear fast red solution, dehydrated and mounted with DPX.

Image Analysis

Image analysis was done by the Image Pro Plus software (MediaCybernetics, Bethesda, MD). The procedure has been well-established and detail was shown in Supplementary S2.23, 24 The percentage area with positive von Kossa stain in the whole tendon was calculated. Since the window wound could be identified under the microscope at week 12 post-operation, the wound was selected for the semi-quantitative image analysis of immunopositive signals of ECM proteins. Briefly, the integrated optical density (IOD; in arbitrary unit) of the immunopositive signal was measured from the sampled views inside wound of each slide and mean IOD per µm2 was calculated. In the week 12 ossified samples, mean IOD per µm2 was calculated from the ossified and un-ossified regions inside wound, respectively. The assessor was blinded to the time points during image analysis. We did not perform semi-quantitative image analysis for the immunopositive signals of collagen type II, Sox 9, and collagen type X inside wound as they were only used as markers of chondrocyte-like cells (collagen type II and Sox 9) and process of endochondral ossification (collagen type X). The results were hence described qualitatively.

Data Analysis

The comparison of immunopositive signal of ECM protein inside wound in the injured side with that in the contralateral control side at different time points after injury was done using Wilcoxon signed rank test. The difference in the immunopositivity of ECM protein in the ossified and un-ossified regions inside wound in the injury group at week 12 was compared by Wilcoxon signed rank test. The changes in the percentage area of positive von Kossa stain and immunopositive signal of ECM protein inside wound across different time points was compared using Kruskal–Wallis test followed by post hoc pairwise comparison of different time points using Mann–Whitney U-test with Bonferroni correction. To perform the Bonferroni correction, Mann–Whitney U-test was performed for comparison of different time points. The p-value obtained was then multiplied by 3 to get the final p-value as there were three comparisons among different time points (i.e., week 2 vs. week 4, week 2 vs. week 12, week 4 vs. week 12). p ≤ 0.05 was regarded as statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. METHODOLOGY
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Chondrocyte Phenotype and Ossification Inside Wound in the Tendon Injury Model

At week 2 post-injury, there was an increase in cellularity and vascularity inside wound (Fig. 1B). Collagen birefringence inside wound was weak under polarization microscopy at week 2 (Fig. 1F). At week 4 post-injury, both cellularity and vascularity inside wound decreased (Fig. 1C). Higher collagen birefringence in the wound at week 4 was seen compared to that at week 2 (Fig. 1G). Chondrocyte-like cells, as indicated by the presence of lacauna in the cells, were firstly observed inside wound at week 4 in two samples (i.e., 33%) (Figs. 1C and 3C, arrows). At week 12 after injury, the cellularity and vascularity inside wound decreased further (Fig. 1D) while the collagen birefringence inside wound was higher (Fig. 1H). Ossified deposits surrounded by chondrocyte-like cells were observed in the window wound in three samples (i.e., 50%; Fig. 1D, arrows, CR). Inside the ossified deposits and area with chondrocyte-like cells, the organization of collagen fiber was disrupted (Fig. 1H). The chondrocyte-like cells and ossified deposits were found in the tendon mid-substance inside wound and were not due to harvesting of the tendon–bone junction (Fig. 1I). The presence of ossification in the repair tissue was confirmed by von Kossa staining and immunohistochemical staining of collagen type X in the ossified deposits at week 12 post-injury in three out of six samples (Fig. 2D, CR). For the three samples with positive von Kossa signal, the mean percentage area was 0.50 ± 0.84%. The median percentage area with positive von Kossa signal for six samples was 0.001% (range: 0–1.46%). Because only three samples showed ossification at week 12, there was marginally insignificant difference in the percentage area with von Kossa stain in the week 12 injury group compared to that in the control group (overall: p = 0.020; post hoc comparison: p = 0.059). Immunopositive signal of collagen type X was observed in the matrix of ossified deposits (Fig. 2H, CR) and chondrocyte-like cells (Fig. 2H, arrows) at week 12 after injury. The presence of chondrocyte phenotype inside wound was confirmed by immunohistochemical staining of collagen type II and Sox 9. At week 4 after injury, immunopositive signal of collagen type II was observed in the matrix surrounding the chondrocyte-like cells (Fig. 3C, arrows) and healing tendon cells (Fig. 3C, arrowhead) in the wound in two samples (i.e., 33%). At week 12 after injury, immunopositive signal was observed in the ossified deposits (Fig. 3D, CR) and the chondrocyte-like cells surrounding the ossified deposits in three samples (i.e., 50%) (Fig. 3D, arrows). No immunopositive signal of collagen type II was observed in the week 12 control (Fig. 3A) and the tissue surrounding the wound (result not shown). For the expression of Sox 9, there was intense staining in the healing tendon cells and chondrocyte-like cells inside wound in four samples (i.e., 66%) with (n = 2) and without (n = 2) chondrocyte-like cells at week 4 after injury (Fig. 3G). At week 12 after injury, immunopositivity of Sox 9 was found in the ossified deposits (Fig. 3H, CR) and the surrounding chondrocyte-like cells inside wound (Fig. 4H, arrows), as well as in the healing tendon cells inside wound in three samples (i.e., 50%). A few healing tendon cells outside the window wound were also immunopositive for Sox 9 (result not shown). Immunopositive signal of Sox 9 was observed in the week 12 control along a few tendon cells (Fig. 3E).

thumbnail image

Figure 1. Photomicrographs showing H&E staining (A–D, I) and polarization (E–H) in the patellar tendon in the intact contralateral control at week 12 (A,E), and the window wound at week 2 (B,F), week 4 (C,G), and week 12 (D,H,I). Magnification: ×200 (A–D); ×1.25 (E–I); bar: 100 µm (A–D); 1,000 µm (E–I); arrows: chondrocyte-like cells; CR: ossified region; dashed lines: area with chondrocyte-like cells. [Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/jor]

Download figure to PowerPoint

thumbnail image

Figure 2. Photomicrographs showing von Kossa staining (A–D) and immunohistochemical staining of collagen type X (E–H) in the patellar tendon in the intact contralateral control at week 12 (A,E), and the window wound at week 2 (B,F), week 4 (C,G), and week 12 (D,H). Magnification: ×200; bar: 100 µm; arrows: chondrocyte-like cells; CR: ossified region. [Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/jor]

Download figure to PowerPoint

thumbnail image

Figure 3. Photomicrographs showing immunohistochemical staining of collagen type II (A–D) and Sox 9 (E–H) in the patellar tendon in the intact contralateral control at week 12 (A,E), and the window wound at week 2 (B,F), week 4 (C,G), and week 12 (D,H). Magnification: ×200; bar: 100 µm; arrows: chondrocyte-like cells; arrowheads: healing tendon cells; CR: ossified region. [Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/jor]

Download figure to PowerPoint

thumbnail image

Figure 4. Photomicrographs showing immunohistochemical staining of collagen type I (A–D) and collagen type III (E–H) in the patellar tendon in the intact contralateral control at week 12 (A,E), and the window wound at week 2 (B,F), week 4 (C,G), and week 12 (D,H). Magnification: ×200; bar: 100 µm; arrows: chondrocyte-like cells; CR: ossified region.[Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/jor]

Download figure to PowerPoint

Expression of Collagen Types I and III

There was weak expression of collagen type I in the tendon matrix and cells in the sham controls (Fig. 4A) and there was no significant change in the mean IOD across different time points (Fig. 5A; p = 0.076). At week 2 after injury, the expression of collagen type I increased in the tendon matrix and cells inside wound and was significantly higher than that at the contralateral controls (Figs. 4B and 5A; p = 0.028). The expression of collagen type I inside wound was significantly and insignificantly reduced at week 4 (Figs. 4C and 5A; p = 0.015) and week 12, respectively, compared to that at week 2 (Figs. 4D and 5A; p = 0.219) but remained higher than that in the contralateral controls (Fig. 5A; week 4: p = 0.028; week 12: p = 0.046). The expression of collagen type I was observed in chondrocyte-like cells and matrix around and inside the ossified region in addition to tendon cells and matrix in the repair tissue at week 12 (Fig. 4D, arrows and CR). There was no significant difference in mean IOD between the ossified and un-ossified regions inside wound at week 12 (p = 0.180). Similarly, there was weak expression of collagen type III in the tendon matrix and cells in the sham controls (Fig. 4E) and there was no significant change in the mean IOD across different time points (Fig. 5B, p = 0.402). There was increased and intense expression of collagen type III in tendon fibroblasts and matrix inside wound at week 2 after injury (Fig. 4F). The mean IOD inside wound at week 2 post-injury was significantly higher than that in the contralateral control (Fig. 5B; p = 0.018). The expression of collagen type III inside wound was significantly reduced at week 4 (Figs. 4G and 5B; p = 0.042) and week 12 (Figs. 4H and 5B; p = 0.015) compared to that at week 2 (Fig. 4F) but remained significantly higher than that in the contralateral controls (Fig. 5B; week 4: p = 0.028; week 12: p = 0.046). There was mild expression of collagen type III in chondrocyte-like cells and matrix around and inside the ossified region inside wound in addition to tendon cells and matrix inside wound at week 12 (Fig. 4H, arrows and CR). While the mean IOD at the ossified region was higher than that at the non-ossified region inside wound at week 12, it was not significantly different (p = 0.180). There was no significant change in the collagen type III/type I ratio across different time points in the control group (Fig. 5C; p = 0.067). The collagen type III/type I ratio in the window wound in the injury group peaked at week 4 and then significantly reduced at week 12 (week 4 vs. week 12 in the injury group: p = 0.045). The collagen type III/type I ratio inside wound in the injury group was insignificantly lower than that in the contralateral control at week 12 (Fig. 5C, p = 0.249). While the ratio at the ossified region was higher than that in the un-ossified region inside wound at week 12, it was not significantly different (p = 0.180).

thumbnail image

Figure 5. Boxplots showing the immunopositive signal of (A) collagen type I, (B) collagen type III, and (C) the collagen type III/collagen type I ratio at weeks 2, 4, and 12 in the control group and window wound in the injury group. *p ≤ 0.05.

Download figure to PowerPoint

Expression of Decorin

There was moderate and no significant change in the expression of decorin in the controls at all time points (Figs. 6A and 7A; p = 0.047). After injury, the expression of decorin in the tendon matrix inside wound reduced (Fig. 6B and C) and was significantly lower than that in the contralateral control at both weeks 2 and 4 (Fig. 7A; both p = 0.028). The expression of decorin inside wound significantly increased at week 12 after injury compared to that at week 2 (Figs. 6D and 7A; p = 0.006) but remained significantly lower than that at the contralateral control (Fig. 7A; p = 0.028). Absence to weak immunopositive signal of decorin was observed in the matrix surrounding the chondrocyte-like cells and ossified deposits inside wound at week 12 (Fig. 6D, arrows and CR) and there was no significant difference in mean IOD of decorin in the ossified and un-ossified regions inside the window wound (Fig. 7A; p = 0.655).

thumbnail image

Figure 6. Photomicrographs showing immunohistochemical staining of decorin (A–D), biglycan (E–H), fibromodulin (I–L), and aggrecan (M–P) in the intact contralateral control at week 12 (A,E,I,M), and the window wound at week 2 (B,F,J,N), week 4 (C,G,K,O), and week 12 (D,H,L,P). Magnification: ×200; bar: 100 µm; arrows: chondrocyte-like cells; CR: ossified region.

Download figure to PowerPoint

thumbnail image

Figure 7. Boxplots showing the immunopositive signal of (A) decorin, (B) biglycan, (C) fibromodulin, and (D) aggrecan at weeks 2, 4, and 12 in the control group and window wound in the injury group. *p ≤ 0.05.

Download figure to PowerPoint

Expression of Biglycan

There was weak expression of biglycan in the tendon matrix in the controls and the level remained stable at all time points (Figs. 6E and 7B; p = 0.249). At week 2 post-injury, there was strong and significant increase in the expression of biglycan in the tendon matrix inside wound compared to that in the contralateral control (Figs. 6F and 7B; p = 0.018). The expression of biglycan in the tendon matrix inside wound sustained (no significant change in the Kruskal–Wallis test across different time points; p = 0.276) and remained higher than that in the contralateral control at week 4 (Figs. 6G and 7B; p = 0.028) and week 12 (Figs. 6H and 7B; p = 0.028). At week 12, biglycan was expressed around the chondrocyte-like cells (Fig. 6H, arrows) and the ossified region inside wound in a patchy pattern in addition to its expression in the tendon matrix inside the window wound (Fig. 6H, CR). There was significantly higher mean IOD in the ossified compared to that in the un-ossified regions inside the window wound (p = 0.018; Fig. 7B).

Expression of Fibromodulin

There was moderate expression of fibromodulin in the controls and the level remained stable at all time points (Figs. 6I and 7C; p = 0.246). At week 2 post-injury, the expression of fibromodulin decreased and became undetectable in the tendon cells and matrix inside wound (Fig. 6J). The mean IOD inside wound at week 2 post-injury was significantly lower than that in the contralateral controls (Fig. 7C; p = 0.018). At week 4 post-injury, the expression of fibromodulin in the tendon cells and matrix inside wound increased slightly compared to that at week 2 (Figs. 6K and 7C; p = 0.006 compared to week 2). It remained lower than that in the contralateral controls though it was not significant (Fig. 7C; p = 0.249). The expression of fibromodulin inside wound increased (Fig. 7C; p = 0.003 compared to week 2) and became comparable to that in the contralateral controls at week 12 (Figs. 6L and 7C; p = 0.753). Fibromodulin was expressed in chondrocyte-like cells around and inside the ossified region in the window wound (Fig. 6L, arrows and CR) in addition to tendon cells and matrix inside wound at week 12. There was no significant difference in mean IOD in the ossified and un-ossified regions inside wound at week 12 (Fig. 7C; p = 0.180).

Expression of Aggrecan

There was moderate expression of aggrecan in the matrix in the controls and the level remained stable at all time points (Figs. 6M and 7D; p = 0.077). The expression of aggrecan inside wound decreased significantly from weeks 2 to 12 (Figs. 6N–P and 7D; week 2 vs. week 12: p = 0.003; week 4 vs. week 12: p = 0.006) and it was significantly lower than that in the contralateral control at week 12 (Fig. 7D; p = 0.028). Despite this, there was higher focal expression of aggrecan in chondrocyte-like cells around and inside the ossified region inside wound (Fig. 6P, arrows and CR) compared to that in the tendon cells and matrix inside wound at week 12 thought it was not statistically significant (Fig. 7D; p = 0.180).

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODOLOGY
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Despite the vast amount of research on tendon healing, much remained unknown about the process. Our result showed that chondrocyte-like cells were observed in 33% samples starting at week 4 after injury while ectopic ossification surrounded by chondrocyte-like cells was observed in 50% of samples at week 12. The ectopic chondrogenesis and ossification observed were not due to technical errors due to harvesting of the tendon–bone junction as the window wound was created without damaging the tendon–bone injunction. We tried to avoid the tendon–bone junction during sample collection. Although the normal tendon–bone junction could be observed partly in a few samples, they were not included in all analyses. The chondrocyte-like cells and the ossified deposits were observed only in the mid-substances of window wound of patellar tendon and there was no specific localization of the chondrocyte-like cells and ossified deposits in the tendon mid-substances, supporting that they were unlikely to be originated from the intact tendon–bone junction. The chondrocyte-like cells and ossified deposits inside wound were confirmed to be in the tendon mid-substance at low magnification of the histological images. There was no difference in the generation of defect in the animals that showed ectopic chondrogenesis and ossification. The ossified deposits inside wound were formed by endochondral ossification as shown by the expression of collagen type X. Ectopic ossification after midpoint tenotomy of rodent Achilles tendon has been reported in previous studies.8–12 We have not performed this surgery in other species of animal. Whether ectopic chondrogenesis and ossification occur only in rodent tendons after injury is not clear and this study should be repeated in other species of animal to confirm the present findings. However, calcification was also reported clinically in the follow-up of some, but not all, patients with Achilles tendon rupture13 and open Achilles tendon repair.14 The erroneous deposition of ECM as a result of the presence of chondrocyte-like cells and ossification might negatively impact the material property of patellar tendon after acute injury. The presence of ossified deposits in tendon could theoretically increase the stiffness. Both could decrease the tensile strength of the repair tissue. We observed more ossification after 6 months post-injury (Supplementary S3) though the sample size was small, suggesting that chondrocyte phenotype and ossification persisted for at least 6 months after injury.

Review of the literatures showed that calcification of patellar tendon, in fact, has been reported in some patients with ACL reconstruction.15, 16, 21 Other tendinopathic-like features such as mild degenerative changes,18 increase in cellularity and vascularity,17 hypoechogeneity, and hyperechogeneity,15, 20, 21 increase in tendon width15, 17–20, 21 as observed with MRI and ultrasonography have also been reported. Bayar et al.15 further described their patients having changes similar to tendinopathy in 6 out of 20 patients. This suggested that acute injury as created by the removal of the central one-third of patellar tendon could develop tendinopathic-like changes. Acquisition of chondrocyte phenotypes and ectopic ossification has been reported in clinical samples and animal models of tendinopathy.23, 25 Given unfavorable factors, acute injury might develop tendinopathic-like changes.

We observed transient increases in the expression of collagen types III and I inside wound after injury, consistent with repair. The increase in the expression of collagen types I and III after tendon injury has also been reported in previous studies.5–7 We observed sustained expression of biglycan and reduced expression of aggrecan and decorin in the tendon matrix inside wound in our injury model. The decrease in the expression of decorin after injury was consistent with previous work studying the mRNA expression of decorin after tendon injury5, 6 but inconsistent with the observation of others.7 Our result was also consistent with the results of Berglund et al.5 whom reported sustained mRNA expression of biglycan in tendon and tendon sheath after injury but was inconsistent with them on the expression of aggrecan. The discrepancy might be due to different injury models, length of follow-up and the assessment method. The turnover of mRNA after injury might not follow exactly the changes of protein expression. Despite the differences, the change in the composition of proteoglycans was confirmed and the alteration in composition of the ECM might contribute to reduced mechanical property of the repair tissue as reported in our previous study.2

Biglycan and decorin both belong to class I of small leucine-rich proteoglycan (SLRP) family. They are highly homologous and co-expressed in various tissues. Young et al.26 reported that there was higher expression of decorin in tissues from the biglycan knock-out mouse model compared to that in wild-type animals, suggesting that these two related class I SLRPs could share common functions and, possibly, compensate for each other's functions when the other one was absent. The sustained expression of biglycan in the healing tissue in our study might be to compensate for the reduced expression of decorin in the healing tissue after injury. The reciprocal changes in decorin and biglycan was also reported in a transected unrepaired rabbit medial collateral ligament model, which the expression of decorin was barely detectable while the expression of biglycan accumulated and the post-translational modification of biglycan was altered in the repairing ligament over 2 years.27

Young et al.26 reported that knock-down of biglycan in a mouse model resulted in low bone mass and biglycan was essential for bone formation while the knock-down of decorin in a mouse model resulted in normal bone mass. The increase in biglycan in the repair tissue therefore might contribute to the formation of ossified deposits in our animal model. Young et al.26 also reported that biglycan knock-out mice displayed larger irregular fibrils while decorin knock-out mice displayed smaller fibrils in bone. This supported our hypothesis that the sustained expression of biglycan and reduced expression of decorin in the healing tendon as observed in our animal model might be associated with smaller fibrils and hence poor quantity of healing after injury. The accumulation of biglycan might interfere with proper collagen network remodeling as decorin and biglycan could compete for binding on collagen type I.28 We observed earlier expression of Sox 9 and collagen type II in healing tendon fibroblasts and this preceded their expression in the chondrocyte-like cells and ossified area. The alteration of ECM composition and growth factors after injury might favor erroneous chondrogenic and osteogenic differentiation of tendon progenitor cells to chondrocytes and osteoblasts, respectively, and negatively affect the mechanical property of the regenerated tissue.23, 29 Aggrecan and biglycan, the key ECM proteins of cartilaginous tissue, were highly expressed in the chondrocyte-like cells and ossified deposits at week 12 in our study. Apart from their roles in ECM organization, proteoglycans could have roles in modulating the activity of tendon cells.4 There was rapid upregulation of expression of proteoglycans during chondrogenic differentiation of bone marrow mesenchymal stem cells (BMSCs).30 The expression of collagen types II and X occurred only at the late stage of the process, suggesting the importance of proteoglycans in modulating stem cell differentiation.30 The differentiation of marrow stromal cells was reported to depend on the expression of biglycan.26 We have reported increased expression of BMP-2 in the same patellar tendon window injury model, which might contribute to erroneous cell differentiaion.22 Regarding the possible source of tendon progenitor cells undergoing erroneous differentiation, it may be BMSCs which migrated into the wound or tendon-derived stem cells (TDSCs) identified recently.31, 32 BMP-2 has been reported to induce osteogenic32 and chondrogenic (unreported observation) differentiation of TDSCs in vitro. Tendon stem cells isolated from the biglycan and fibromodulin double knock-out mouse model were also reported to show increased sensitivity to BMP-2 signaling.31

Based on our observation, the expression of ECM in the intact region outside the window wound was similar to that in the intact contralateral control except that the expression of fibromodulin seemed to be slightly higher while the expression of aggrecan seemed to be slightly lower in the adjacent normal region compared to those in the intact contralateral control. However, we have not semi-quantified the expression levels in adjacent normal region and it would be interesting to look at it in future study.

In conclusion, ectopic chondrogenesis and ossification were observed inside wound in 50% samples at week 12. There was sustained expression of biglycan and reduced expression of aggrecan and decorin in the tendon matrix in the window wound. The erroneous deposition of ECM and ectopic chondro-ossification, both might influence each other, might account for the poor tissue quality after acute injury. Higher expression of biglycan and aggrecan at the ectopic chondro-ossification sites suggested that they might have roles in ectopic chondro-osteogenesis in the repair tissue.

Acknowledgements

  1. Top of page
  2. Abstract
  3. METHODOLOGY
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

This research project was supported by the CUHK Direct Grants (2008.1.062 and 2009.1.049).

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODOLOGY
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information
  • 1
    Miyashita H, Ochi M, Ikuta Y, 1997. Histological and biomechanical observations of the rabbit patellar tendon after removal of its central one-third. Arch Orthop Trauma Surg 116: 454462.
  • 2
    Chan BP, Fu SC, Qin L, et al. 1998. Pyridinoline in relation to ultimate stress of the patellar tendon during healing: an animal study. J Orthop Res 16: 597603.
  • 3
    Tohyama H, Yasuda K, Kitamura Y, et al. 2003. The changes in mechanical properties of regenerated and residual tissues in the patellar tendon after removal of its central portion. Clin Biomech (Bristol, Avon) 18: 765772.
  • 4
    Riley GP, 2005. Gene expression and matrix turnover in overused and damaged tendons. Scand J Med Sci Sports 15: 241251.
  • 5
    Berglund M, Reno C, Hart DA, et al. 2006. Patterns of mRNA expression for matrix molecules and growth factors in flexor tendon injury: differences in the regulation between tendon and tendon sheath. J Hand Surg Am 13: 12791287.
  • 6
    Lo IK, Boorman R, Marchuk L, et al. 2005. Matrix molecule mRNA levels in the bursa and rotator cuff of patients with full-thickness rotator cuff tears. Arthroscopy 21: 645651.
  • 7
    Yokota A, Gimbel JA, Williams GR, et al. 2005. Supraspinatus tendon composition remains altered long after tendon detachment. J Shoulder Elbow Surg 14 (1 Suppl): 72S78S.
  • 8
    Brown H, Ehrlich HP, Newberne PM, et al. 1986. Para osteo arthropathy-ectopic ossification of healing tendon about the rodent ankle joint: histologic and type V collagen changes. Proc Soc Exp Biol Med 183: 214220.
  • 9
    Rooney P, Walker D, Grant ME, et al. 1993. Cartilage and bone formation in repairing Achilles tendons within diffusion chambers: evidence for tendon–cartilage and cartilage–bone conversion in vivo. J Pathol 169: 375381.
  • 10
    Lin L, Shen Q, Xue T, et al. 2010. Heterotopic ossification induced by Achilles tenotomy via endochondral bone formation: expression of bone and cartilage related genes. Bone 46: 425431.
  • 11
    Xu JC, Wu T, Wu GH, et al. 2009. Leptin expression by heterotopic ossification-isolated tissue in rats with Achilles' tenotomy. Saudi Med J 30: 605610.
  • 12
    Hannallah D, Peng H, Young B, et al. 2004. Retroviral delivery of noggin inhibits the formation of heterotopic ossification by BMP-4, demineralized bone matrix, and trauma in an animal model. J Bone Joint Surg Am 86-A: 8091.
  • 13
    Bleakney RR, Tallon C, Wong JK, et al. 2002. Long-term ultrasonographic features of the Achilles tendon after rupture. Clin J Sport Med 12: 273278.
  • 14
    Kraus R, Stahl JP, Meyer C, et al. 2004. Frequency and effects of intratendinous and peritendinous calcifications after open Achilles tendon repair. Foot Ankle Int 25: 827832.
  • 15
    Bayar A, Turhan E, Ozer T, et al. 2008. The fate of patellar tendon and infrapatellar fat pad after arthroscopy via central portal. Knee Surg Sports Traumatol Arthrosc 16: 11141120.
  • 16
    Erdogan F, Aydingoz O, Kesmezacar H, et al. 2004. Calcification of the patellar tendon after ACL reconstruction. A case report with long-term follow-up. Knee Surg Sports Traumatol Arthrosc 12: 277279.
  • 17
    Kartus J, Movin T, Papadogiannakis N, et al. 2000. A radiographic and histologic evaluation of the patellar tendon after harvesting its central third. Am J Sports Med 28: 218226.
  • 18
    Lidén M, Ejerhed L, Sernert N, et al. 2006. The course of the patellar tendon after reharvesting its central third for ACL revision surgery: a long-term clinical and radiographic study. Knee Surg Sports Traumatol Arthrosc 14: 11301138.
  • 19
    Svensson M, Kartus J, Ejerhed L, et al. 2004. Does the patellar tendon normalize after harvesting its central third? A prospective long-term MRI study. Am J Sports Med 32: 3438.
  • 20
    Wiley JP, Bray RC, Wiseman DA, et al. 1997. Serial ultrasonographic imaging evaluation of the patellar tendon after harvesting its central one third for anterior cruciate ligament reconstruction. J Ultrasound Med 16: 251255.
  • 21
    Järvelä T, Paakkala T, Kannus P, et al. 2004. Ultrasonographic and power Doppler evaluation of the patellar tendon ten years after harvesting its central third for reconstruction of the anterior cruciate ligament: comparison of patients without or with anterior knee pain. Am J Sports Med 32: 3946.
  • 22
    Lui PPY, Chan LS, Cheuk YC, et al. 2009. Expression of bone morphogenetic protein-2 in the chondrogenic and ossifying sites of calcific tendinopathy and traumatic tendon injury rat models. J Orthop Surg Res 21: 27.
  • 23
    Lui PPY, Fu SC, Chan LS, et al. 2009. Chondrocyte phenotype and ectopic ossification in collagenase-induced tendon degeneration. J Histochem Cytochem 57: 91100.
  • 24
    Yee Lui PP, Wong YM, Rui YF, et al. 2011. Expression of chondro-osteogenic BMPs in ossified failed tendon healing model of tendinopathy. J Orthop Res 29: 816821.
  • 25
    Maffulli N, Reaper J, Ewen SW, et al. 2006. Chondral metaplasia in calcified insertional tendinopathy of the Achilles tendon. Clin J Sports Med 16: 329334.
  • 26
    Young MF, Bi Y, Ameye L, et al. 2003. Biglycan knockout mice: new models for musculoskeletal diseases. Glycoconj J 19: 257262.
  • 27
    Plaas AH, Wong-Palms S, Koob T, et al. 2000. Proteoglycan metabolism during repair of the ruptured medial collateral ligament in skeletally mature rabbits. Arch Biochem Biophys 374: 3541.
  • 28
    Schonherr E, Witsch-Prehm P, Harrach B, et al. 1995. Interaction of biglycan with type I collagen. J Biol Chem 270: 27762783.
  • 29
    Rui YF, Lui PPY, Chan LS, et al. 2011. Does erroneous differentiation of tendon-derived stem cells (TDSCs) contribute to the pathogenesis of calcifying tendinopathy? Chin Med J (Engl) 124: 606610.
  • 30
    Barry F, Boynton RF, Liu B, et al. 2001. Chondrogenic differentiation of mesenchymal stem cells from bone marrow: differentiation-dependent gene expression of matrix components. Exp Cell Res 268: 189200.
  • 31
    Bi Y, Ehirchiou D, Kilts TM, et al. 2007. Identification of tendon stem/progenitor cells and the role of the extracellular matrix in their niche. Nat Med 13: 12191227.
  • 32
    Rui YF, Lui PPY, Ni M, et al. 2011. Mechanical loading increased BMP-2 expression which promoted osteo-chondrogenic and inhibited tenogenic differentiation of tendon-derived stem cells. J Orthop Res 29: 390396.

Supporting Information

  1. Top of page
  2. Abstract
  3. METHODOLOGY
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  8. Supporting Information

Additional Supporting information may be found in the online version of this article.

FilenameFormatSizeDescription
jor_21495_sm_SuppDataS1.doc38KSupplementary Data S1
jor_21495_sm_SuppDataS2.doc30KSupplementary Data S2
jor_21495_sm_SuppDataS3.tif13403KSupplementary S3. Photomicrographs showing the presence of chondrocyte-like cells and ossified deposits inside the patellar tendon window wound at 6 months after injury. Magnification: 200x; scale bar: 100μm; Arrows: chondrocyte-like cells; CR: ossified region

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.