The effects of dexamethasone on human patellar tendon stem cells: Implications for dexamethasone treatment of tendon injury

Authors

  • Jianying Zhang,

    1. MechanoBiology Laboratory, Departments of Orthopaedic Surgery, Bioengineering, and Mechanical Engineering and Materials Science, University of Pittsburgh School of Medicine, 210 Lothrop Street, BST, E1640, Pittsburgh, Pennsylvania 15213
    Search for more papers by this author
  • Camille Keenan,

    1. MechanoBiology Laboratory, Departments of Orthopaedic Surgery, Bioengineering, and Mechanical Engineering and Materials Science, University of Pittsburgh School of Medicine, 210 Lothrop Street, BST, E1640, Pittsburgh, Pennsylvania 15213
    Search for more papers by this author
  • James H-C. Wang

    Corresponding author
    1. MechanoBiology Laboratory, Departments of Orthopaedic Surgery, Bioengineering, and Mechanical Engineering and Materials Science, University of Pittsburgh School of Medicine, 210 Lothrop Street, BST, E1640, Pittsburgh, Pennsylvania 15213
    • MechanoBiology Laboratory, Departments of Orthopaedic Surgery, Bioengineering, and Mechanical Engineering and Materials Science, University of Pittsburgh School of Medicine, 210 Lothrop Street, BST, E1640, Pittsburgh, Pennsylvania 15213. T: 412-648-9102; F: 412-648-8548
    Search for more papers by this author

Abstract

Injection of Dexamethasone (Dex) is commonly used in clinics to treat tendon injury such as tendinopathy because of its anti-inflammatory capabilities. However, serious adverse effects have been reported as a result of Dex treatment, such as impaired tendon healing and tendon rupture. Using both in vitro and in vivo approaches, this study was to determine the effects of Dex treatment on the proliferation and differentiation of human tendon stem cells (hTSCs), which can directly impact tendon healing. We found that Dex treatment stimulated cell proliferation at lower concentrations (<1,000 nM), whereas a high concentration (1,000 nM) decreased cell proliferation. Moreover, at all concentrations used (5, 10, 100, and 1,000 nM), Dex treatment induced non-tenocyte differentiation of hTSCs, as evidenced by a change in cell shape, a nearly complete suppression of collagen type I expression, and an upregulation of non-tenocyte related genes (PPARγ and Sox-9), which was especially evident when higher concentrations (>10 nM) of Dex were used. Implantation of Dex-treated hTSCs for a short time (3 weeks) resulted in the extensive formation of fatty tissues, cartilage-like tissues, and bony tissues. These findings suggest that Dex treatment in clinics may cause a paradoxical effect on the injured tendons it is supposed to treat: by inducing non-tenocyte differentiation of hTSCs, Dex treatment depletes the stem cell pool and leads to the formation of non-tendinous tissues (e.g., fatty and cartilage-like tissues), which make tendon susceptible to rupture. © 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 31:105–110, 2012

Tendinopathy, a common tendon disorder, has drawn extensive attention due to its complexity in etiology and the current lack of effective treatments for tendinopathy. Tendinopathy may involve chronic inflammation and degeneration of tendons, and injections of potent synthetic glucocorticoids such as Dexamethasone (Dex) have been commonly used to alleviate such inflammation and pain.1 However, adverse effects of glucocorticoid treatments, such as impaired tendon healing and tendon rupture, have been reported.2–4

Over the years, extensive research has been conducted to investigate the effects of glucocorticoid treatments. Previous studies showed that the use of corticosteroids reduces tendon strength5 and suppresses proliferation and collagen synthesis of tenocytes derived from animal tendons.6 Moreover, glucocorticoids were also found to decrease viability, migration, proliferation, and collagen synthesis of both human and animal tenocytes.7–11

While tendons are considered to be mainly comprised of tenocytes, studies in recent years have shown that tendons from humans, mice, rats, and rabbits also contain tendon-specific stem cells called tendon stem/progenitor cells (TSCs).12–15 Like other adult stem cells, TSCs are essential in tendon homeostasis and the repair of injured tendons; TSCs have the ability to self-renew and differentiate into tenocytes to replenish tendon cells lost due to injury.14, 16, 17 On the other hand, because of their multi-differentiation potential, TSCs are considered to participate in the development of tendinopathy by undergoing aberrant differentiation into non-tenocytes (e.g., adipocytes, chondrocytes, and osteocytes).18

Because tendons contain stem cells, it seems certain that injections of Dex into tendons (e.g., intratendinous injections) will expose TSCs to the drug. Nevertheless, the effects of such Dex treatment on TSCs have not been defined. Therefore, this study aims to determine the proliferation and differentiation of human TSCs (hTSCs) subjected to Dex treatments with clinically relevant dosages. After performing hTSC culture and in vivo implantation experiments, we found that Dex treatment altered cell proliferation and induced cell differentiation in a dose-dependent manner; especially at high doses, Dex treatment decreased hTSC proliferation and caused hTSCs to differentiate into non-tenocytes. Moreover, implanted hTSCs that had been treated with Dex resulted in the formation of fatty, cartilage, and bone-like tissues in vivo. These findings indicate that Dex treatment causes detrimental side effects on tendons, which explains why clinical Dex treatments in clinics are associated with impaired tendon healing, reduced tendon strength, and tendon rupture.

METHODS

Preparation of hTSC Culture

hTSCs were isolated from the patellar tendon tissues of seven human donors (age 28 ± 6.7 years), five males and two females, using a previously published method.13 The cells were cultured in growth medium consisting of Dulbecco's modified Eagle's medium (DMEM; Lonza, Walkersville, MD) supplemented with 20% fetal bovine serum (FBS; Atlanta Biologicals, Lawrenceville, GA), 100 U/ml penicillin, and 100 µg/ml streptomycin (Atlanta Biologicals, Lawrenceville, GA). The stemness of the hTSCs was confirmed because these cells formed colonies at low cell densities and exhibited a cobblestone morphology in confluent conditions.15 In addition, the expression of stem cell markers Oct-4, SSEA-4, and nucleostemin was detected on hTSCs using immunocytochemistry.19

hTSC Culture Experiments with Dex Treatments

hTSCs at the first passage were grown in 6-well culture plates and divided into five groups (one control and four treatments), and each group was cultured in triplicate. After 24 h of culture, attachment of the hTSCs was confirmed by microscopic observation. hTSCs were then treated with four concentrations (5, 10, 100, and 1,000 nM) of Dex solutions in culture media. Control cells were those grown in culture media with the addition of appropriate amounts of DMSO, which was used for preparation of Dex stock solution.

To assess proliferation effects after Dex treatment of hTSCs, we calculated the population doubling time (PDT) of each group using a previous method.13 A digital cellometer (Nexcelcom Bioscience, Lawrence, MA) was used to count cell numbers before culture and after Dex treatment.

To determine differentiation effects after Dex treatment of hTSCs, we performed quantitative real time RT-PCR (qRT-PCR) to quantify gene expression using a QIAGEN QuantiTect SYBR Green PCR Kit (QIAGEN). Briefly, in a 25 µl PCR reaction mixture, 2 µl cDNA (total 100 ng RNA) was amplified in a Chromo 4 Detector (MJ Research, Maltham, MA) with incubation at 94°C for 5 min, followed by 30–60 cycles of a three temperature program consisting of 1 min at 94°C, 40 s at 57°C, and 40 s at 72°C. The PCR reaction was terminated after a 10-min extension at 70°C and stored at 4°C until analysis. Using the qRT-PCR, expression of the following genes was measured: collagen type I (marker for tenocytes), PPARγ (marker for adipocytes), Sox-9 (marker for chondrocytes), and Runx-2 (marker for osteocytes). The expression of GAPDH, which served as an internal control, was also determined. All primers for RT-PCR analysis were adopted from previous studies20 and synthesized by Invitrogen (Grand Island, NY).

hTSC Implantation Experiment

Eight female nude rats (10 weeks old; 200–250 g) were used to test the effects of Dex on hTSC differentiation in vivo. Rats were housed individually on a 12 h:12 h light–dark cycle and were cared for in accordance with the Guide to the Care and Use of Experimental Animals. hTSCs at passage 2 in a 24-well plate (6 × 104/well) were treated with three concentrations of Dex (10, 100, and 1,000 nM) for 1 week. Cells without Dex treatment were used for control. After detachment with trypsin, all cells in both treatment and control groups were collected and mixed with 0.5 ml of Matrigel (Cat. # 354234, BD Biosciences, Bedford, MA) in a 24-well plate.

For implantation experiment, the nude rats were placed under general anesthesia using ketamine hydrochloride (75 mg/kg body weight) and xylazine hydrochloride (5 mg/kg body weight), administrated by intramuscular injection. Wounds (1 cm diameter/each wound) were created on the back of each rat, and three pieces of cell-Matrigel were placed in three distinct wounds on each side of the rat's back. The approximate distance between two wound sites was 2 cm, with a total of six cell-Matrigel composites implanted into each rat. Each group had two rats and a total of eight rats were used for four groups (Dex 0, 10, 100, and 1,000 nM). At 3 weeks after implantation, tissue samples were harvested and placed in pre-labeled base molds filled with frozen section medium (Neg 50; Richard-Allan Scientific; Kalamazoo, MI). The base mold with tissue samples was quickly immersed in liquid nitrogen cold 2-methylbutane and allowed to solidify completely. The tissue blocks were then placed on dry ice and subsequently stored in a deep freezer (−80°C) until used for histological analysis.

Histochemical Analysis

The tissue block was cut into 10 µm thick sections, which were then placed on glass slides and allowed to dry overnight at room temperature. Oil Red O staining, Safranin O staining, and Alizarin Red S staining were used to detect fatty, cartilage-like, and bone-like tissues, respectively, using previously described procedures.15

Data Analysis

All data are presented in mean ± standard deviation (SD), unless otherwise indicated. For statistical analysis of the data, one-way ANOVA was used, followed by Fisher's PLSD test for multiple comparisons. A t-test was also used for statistical analysis wherever appropriate. A p-value <0.05 was considered to indicate a significant difference between two comparison groups (e.g., a control group vs. a treatment group).

RESULTS

Cell Proliferation after Dex Treatment

We investigated cell proliferation after Dex treatment of hTSCs by determining their PDTs. Compared to control cells without Dex treatment, PDT of the Dex-treated cells was altered in a Dex concentration-dependent manner (Fig. 1), indicating that cell proliferation changed in response to Dex treatments. Specifically, Dex treatment at 5 nM increased cell proliferation, and Dex treatment at higher concentrations (10 and 100 nM) induced a smaller, but similar, dose-dependent pro-proliferative effect. However, Dex treatment at the highest concentration (1,000 nM) used in the culture experiment decreased cell proliferation, as evidenced by a higher PDT value than that of control cells.

Figure 1.

The effects of Dex treatment on the PDT of hTSCs. Note that compared to non-treated controls cells, lower concentrations (<1,000 nM) of Dex treatment led to lower PDT values, or increased cell proliferation, whereas a high concentration (1,000 nM) resulted in a higher PDT, or decreased cell proliferation. Thus, Dex affected cell proliferation in a dose-dependent manner (*p < 0.05, compared to control cells).

Cell Differentiation after Dex Treatment

We first assessed hTSC differentiation in response to Dex treatment by examining changes in cell morphology. We noticed that after Dex treatments of hTSCs for 1 week, cell morphology changed from a generally cobblestone shape to an elongated shape (Fig. 2A and B). This change in cell morphology suggested concentration-dependent hTSC differentiation as a result of Dex treatments.

Figure 2.

The effects of Dex treatment on cell morphology and collagen type I gene expression. hTSCs were cultured with various concentrations of Dex (5, 10, 100, and 1,000 nM) for 1 week. Without Dex treatment, hTSCs were cobblestone-like (A). But with 5 nM Dex treatment, the cell shape already became elongated (B). Higher concentrations (10, 100, and 1,000 nM) of Dex treatment resulted in a similar change in cell shape (data not shown). Moreover, at all concentrations (5–1,000 nM), Dex treatment of hTSCs for 1 week almost completely suppressed collagen type I expression (C) (*p < 0.01 compared to control cells). Scale bars: 100 µm. [Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/jor]

We next examined TSC differentiation by determining tenocyte and non-tenocyte related gene expressions. It was found that the expression of collagen type I was almost completely suppressed in all four Dex treatment groups after 1 week of culture (Fig. 2C). After Dex treatment, the gene expression of PPARγ also changed: higher concentrations of Dex treatment led to higher gene expression of PPARγ (Fig. 3). Moreover, Dex treatment of hTSCs led to the gene expression of Sox-9 in a concentration-dependent manner (Fig. 4). However, Dex treatment did not induce much change in the gene expression of Runx-2, an osteogenesis marker (data not shown).

Figure 3.

The effects of Dex treatment on adipogenesis of hTSCs. hTSCs were treated with Dex (5–1,000 nM) for 1 week. It was found that Dex treatment increased PPARγ gene expression with increasing Dex concentrations (*p < 0.05 with respect to control cells). PPARγ is a specific marker for adipogenesis of adult stem cells; hence, Dex treatment induced concentration-dependent adipogenesis of hTSCs.

Figure 4.

The effects of Dex treatment on chondrogenesis of hTSCs. After treating hTSCs with low and high concentrations of Dex for 1 week, Sox-9 gene expression increased with increasing Dex concentrations (*p < 0.05 with respect to control cells). Sox-9 expression is required for chondrogenesis of stem cells; hence, Dex treatment induced concentration-dependent chondrogenesis of hTSCs.

Non-Tendinous Tissue Formation after Implantation of Dex-Treated hTSCs

In normal conditions, hTSCs differentiate into tenocytes. To determine whether Dex-treated hTSCs underwent abnormal differentiation, namely, non-tenogenic differentiation, we implanted Dex-treated hTSCs into nude rats subcutaneously. We found that 3 weeks after implantation, fatty, cartilage-like, and bone-like tissues were extensively formed, with the extent of such tissue formation apparently depending on the Dex concentration; in contrast, control cells without Dex treatment formed little such tissues (Fig. 5).

Figure 5.

The formation of non-tendinous tissues after implantation of Dex-treated hTSCs. The cells had been treated with Dex at three concentrations (10, 100, and 1,000 nM) for a week before they were mixed with Matrigel and then implanted subcutaneously into nude rats for 3 weeks. The same cells without Dex treatment, but with Matrigel, were implanted for controls. It was found that Dex-treated hTSCs resulted in the formation of extensive non-tendinous tissues, with the degree being apparently dependent on Dex dosage. Implantation of control hTSCs led to minimal tissue formation. A, E, and I: controls; B–D: fatty tissues (Oil Red O staining); F–H: cartilage-like tissues (Safranin O staining); and J–L: bone-like tissues (Alizarin Red S staining). Scale bars: 100 µm.

DISCUSSION

Dex treatment is frequently used in clinics to treat tendon injury (e.g., tendinopathy) because of its potent ability to suppress inflammation, and hence alleviate associated pain. However, severe side effects, including impaired tendon healing and tendon rupture have been reported. This study was designed to gain a better understanding of the mechanisms of Dex treatment on tendons by examining the Dex effects on hTSCs, a new type of tendon cells that has been identified in recent years. Our findings showed that Dex treatment of hTSCs caused dose-dependent changes in cell proliferation and differentiation. Specifically, at low concentrations (5–100 nM), Dex treatment accelerated cell proliferation; however, at a high concentration (1,000 nM), Dex treatment decreased cell proliferation. Also, at all concentrations, Dex treatment of hTSCs blocked cellular gene expression of collagen type I, a major component of tendons. Moreover, higher concentrations of Dex treatment (>10 nM) induced hTSC differentiation into non-tenocytes (adipocytes and chondrocytes). Finally, implantation of Dex-treated hTSCs resulted in the formation of non-tendinous tissues (fatty, cartilage-like, and bone-like tissues). We also confirmed that these tissues resulted from implanted hTSCs using human specific antibodies (data not shown). Taken together, these findings suggest that Dex treatment of injured tendons, especially at high doses, may do more harm than good. Dex causes nearly exclusive differentiation of TSCs into non-tenocytes instead of tenocytes, the tendon cell responsible for repair of injured tendons. Based on these findings, we speculate that a large number of TSCs in tendon tissues could be depleted by Dex treatment that induces differentiation of TSCs into non-tenocytes, and non-tendinous tissues could be formed within the tendons treated with Dex. These events may lead to impaired tendon healing, reduced tendon strength, and tendon rupture.1, 5

Previous studies have examined the effects of Dex treatment on cell proliferation. For example, Dex treatment at a low concentration (10 nM) and above decreases proliferation of human bone marrow stromal cells (BMSCs).21 In addition, Dex treatment at the same dose (10 nM) inhibits EGF-stimulated epithelial cell proliferation.22 In tenocytes, Dex treatment at a range of doses (0.1–1,000 nM) decreases cell proliferation in a dose-dependent manner.10 This study, however, shows that Dex treatment of hTSCs is biphasic in terms of its effect on cell proliferation: at low concentrations, Dex stimulated cell growth, whereas a high concentration of Dex treatment suppressed cell proliferation. The contrasting results between these previous studies and this study indicate that cellular responses to Dex treatment likely depend on the type of cells (BMSCs vs. TSCs and TSCs vs. epithelial cells) and probably on the species where tendon cells are derived.

One notable finding of this study is that Dex treatments at low and high concentrations all suppressed the gene expression of collagen type I (Fig. 3). The decreased expression of this tenocyte-related gene indicates that the normal differentiation of TSCs into tenocytes was blocked by Dex treatment. Previous studies also showed that Dex treatments reduced collagen type I expression.6, 10 Collagen type I is a major component of tendons, and the synthesis of collagen type I is critical to maintain tendon structure and repair the tendon in case of injury. Because of this, the suppression of collagen type I expression in Dex treated hTSCs suggests that fewer tenocytes will be available after Dex treatment to repair injured tendons, thus leading to impaired tendon healing and making the tendon susceptible to rupture.

It should be noted that even at low doses (<10 nM), Dex treatment was able to initiate non-tenocyte differentiation of hTSCs, as evidenced by changes in cell shape, suppression of collagen type I gene expression (Fig. 2B and C), and induction of non-tenocyte related gene expression (Figs. 3 and 4). Moreover, low doses of Dex treatment stimulated proliferation of these non-tenocyte-differentiation cells. Thus, similar to high doses, low doses of Dex treatment would do more harm than good: it could lead to more non-tenocytes in treated tendons, thus risking the formation of non-tendinous tissues within the tendons.

Dex has been reported to induce adipogenic differentiation,23–25 chondrogenic differentiation,26 or both14, 15 in stem cells. In fact, specific concentrations of Dex in culture media are used to induce a desired differentiation path (500 nM for adipogenic and 100 nM for chondrogenic). However, we did not find that Dex induced osteogenesis of hTSCs in vitro, while others did in previous studies.27, 28 It could be that in order for Dex to induce osteogenesis of stem cells, other components (e.g., ascorbic acid) are required in culture. In the in vivo experiment, bone-like tissues were found by implantation of Dex treated hTSCs. One possible reason is that cartilage-like tissue may have become calcified, as indicated by the formation of bone-like structures (Fig. 5F, arrows), which were nevertheless stained by Safranin O staining indicating that proteoglycans were present in these structures. Overall, our finding in this study that Dex induced non-tenocyte differentiation is consistent with the established role of Dex in previous studies. Note that Dex is a known component in adipogenic and chondrogenic differentiation media used to induce adipogenesis and chondrogenesis of rat TSCs14 and rabbit TSCs15. Dex-induced adipogenesis and chondrogenesis of TSCs also suggest that there were two sub-populations of TSCs (or their progenitor cells) that were inclined to differentiate into adipocytes and chondrocytes, respectively, in response to Dex treatment.

In addition to altering cell proliferation and inducing differentiation of stem cells, Dex treatment was shown to decrease tenocyte migration,7 induce apoptosis of tendon cells,11 and reduce cell viability in human tendon explants.8 Collectively, the findings of these previous studies and this study indicate that Dex treatment can be harmful for tendon healing and hence the repair of injured tendons.

For future work, the molecular mechanisms responsible for the effects of Dex on tendon stem cells should be investigated. One possible mechanism may involve PGE2, which is known to induce non-tenocyte differentiation of TSCs.17 In addition, because tendons contain both TSCs and tenocytes, the effects of Dex on co-cultures of TSCs and tenocytes should be investigated. Moreover, the effects of Dex treatments on tendinous tissues should be investigated in animal models. Specifically, the pathological differentiation, that is, the non-tenocyte differentiation due to high concentrations of Dex treatment, should be verified by examining the formation of non-tendinous tissues in Dex-treated tendons. Finally, efforts should also be made to prevent the deleterious effects of Dex treatment by adding simultaneous application of other treatments. Previous studies showed that the side effects of glucocorticoids were suppressed in vitro during co-treatment therapies where vitamin C,29 platelet-derived growth factors,9 or platelet-rich plasma30 was used.

Acknowledgements

The funding support from NIH (AR049921 and AR061395) for this work is gratefully acknowledged (J.H.W.).

Ancillary