Mechanical loading and TGF-β change the expression of multiple miRNAs in tendon fibroblasts

Abstract

Tendons link skeletal muscles to bones and are important components of the musculoskeletal system. There has been much interest in the role of microRNA (miRNA) in the regulation of musculoskeletal tissues to mechanical loading, inactivity, and disease, but it was unknown whether miRNA is involved in the adaptation of tendons to mechanical loading. We hypothesized that mechanical loading and transforming growth factor-β (TGF-β) treatment would regulate the expression of several miRNA molecules with known roles in cell proliferation and extracellular matrix synthesis. To test our hypothesis, we subjected untrained adult rats to a single session of mechanical loading and measured the expression of several miRNA transcripts in Achilles tendons. Additionally, as TGF-β is known to be an important regulator of tendon growth and adaptation, we treated primary tendon fibroblasts with TGF-β and measured miRNA expression. Both mechanical loading and TGF-β treatment modulated the expression of several miRNAs that regulate cell proliferation and extracellular matrix synthesis. We also identified mechanosensitive miRNAs that may bind to the 3′-untranslated region of the basic helix-loop-helix transcription factor scleraxis, which is a master regulator of limb tendon development. The results from this study provide novel insight into the mechanobiology of tendons and indicate that miRNA could play an important role in the adaptation of tendons to growth stimuli.

tendons are connective tissue structures that transmit force developed by muscles to bones and are composed of a dense extracellular matrix (ECM) consisting mainly of type I collagen, type III collagen, elastin, and various proteoglycans (18). Fibroblasts are the predominant cell type in tendons and are responsible for the maintenance, repair, and modification of tendon ECM (20). Physiological loading of tendons leads to increases in ECM content and fibroblast cell density (33, 34), but the failure of tendons to adapt to physiological loading may lead to the development of painful tendinopathies that can severely restrict activities of daily living (43). Tendinosis is among the most common forms of tendinopathy and is widely thought to occur due to a failure of tendon fibroblasts to regenerate from mechanical loading-induced ECM damage (19). The matrices of tendons from patients with tendinosis grossly appear a dull gray color and show disorganized collagen fibrils, an increase in the amount of ground substance, and an increase in markers of chondrogenesis and neovascularization (2, 3, 19, 41, 53). At the cellular level, the fibroblasts of patients with tendinosis have enlarged nuclei and altered morphology and there is an overall decrease in cell density (3, 19). As fibroblasts are the central regulators of ECM in tendon, gaining a greater understanding of the cellular and molecular mechanisms that regulate tendon fibroblast activity is likely to improve the treatment of tendinopathies.

Transforming growth factor-β (TGF-β) is a signaling molecule that appears to be central for tendon growth and adaptation by promoting both tendon fibroblast proliferation and type I collagen synthesis (20). During embryonic development, TGF-β signaling induces the expression of the basic helix-loop-helix (bHLH) transcription factor scleraxis, which is required for the proper development of limb tendons (25, 37, 40). TGF-β also promotes the expression of the type II transmembrane glycoprotein tenomodulin (25), which also plays an important role in promoting tendon fibroblast proliferation (12). Active TGF-β is released from tendons in response to an acute bout of mechanical loading and likely stimulates many of the anabolic pathways that control exercise-mediated tendon growth and adaptation (17, 20).

MicroRNA (miRNA) molecules are small noncoding RNAs that participate in the posttranscriptional regulation of many protein-coding mRNA molecules, often times by binding to the 3′-untranslated region (UTR) of a mRNA and targeting the mRNA for rapid degradation (7). An emerging body of evidence suggests that miRNA molecules play important roles in regulating the adaptation of skeletal muscle and bone to mechanical loading and disease (16, 30), but while tendons are anatomically situated between muscle and bone, the miRNA molecules that regulate tendon fibroblast activity are unknown. To gain a greater understanding of the role of miRNA in regulating tendon mechanobiology, we measured the expression of several miRNA transcripts after a single bout of mechanical loading in rats. Additionally, due to the link between TGF-β and tendon growth, we treated cultured tendon fibroblasts with TGF-β and measured miRNA expression. We hypothesized that mechanical loading and TGF-β treatment would regulate the expression of several miRNA molecules with known roles in cell proliferation and ECM synthesis.

MATERIALS AND METHODS

Animals.

This study was approved by the University of Michigan Institutional Animal Care and Use Committee and is consistent with the American Physiological Society guidelines for the appropriate use and care of animals in research. Six-month-old male retired breeder Sprague-Dawley rats were used in this study. Rats were housed under specific-pathogen-free conditions and provided with food and water ad libidum. To mechanically load Achilles tendons, untrained rats (n = 5) underwent a 30-min uphill treadmill running session at a fixed velocity of 12 m/min, with the height of the treadmill progressively increasing in elevation from 0° for 10 min, 5° for 10 min, and then to 10° for the final 10 min. Age-matched sedentary male rats that did not undergo this single bout of training (n = 5) were used as controls. Eight hours after the training session, the rats were killed for RNA isolation.

Cell culture.

Tendon fibroblasts were isolated and cultured as described previously (32). Rats were anesthetized with sodium pentobarbital, and the Achilles tendons were carefully removed and trimmed of muscle and connective tissue, finely minced, and placed in DMEM with 0.2% type II collagenase (Invitrogen), in a shaking water bath for 2 h at 37°C. After dissociation, fibroblasts were pelleted by centrifugation, resuspended in DMEM containing 10% FBS and 1% antibiotic-antimycotic (Invitrogen) and plated in 100-mm dishes coated with type I collagen (BD Biosciences). Fibroblasts were passaged upon reaching 70% confluence onto 60-mm dishes coated with type I collagen (BD Biosciences), and the media were switched to DMEM containing 2% FBS and 1% antibiotic-antimycotic. Upon reaching 70% confluence, fibroblasts were treated with DMEM containing 2% FBS and 1% antibiotic-antimycotic supplemented with 2 ng/ml of recombinant human TGF-β1 (R&D Systems) for 4 h. For control cells, upon reaching 70% confluence the media consisted simply of DMEM containing 2% FBS and 1% antibiotic-antimycotic with no recombinant TGF-β1. Following treatment, cells were scraped from their dishes and prepared for total RNA isolation.

RNA isolation and PCR.

All RNA isolation, reverse transcription and gene expression reagents and primers were purchased from Qiagen. Total RNA was isolated from whole Achilles tendons and cultured tendon fibroblasts using an miRNeasy kit supplemented with DNase I treatment. RNA was also isolated from the gastrocnemius muscle of one rat to serve as a control. During the isolation of tendons, care was taken to avoid any contaminating skeletal muscle tissue, and the lack of contaminating muscle tissue was verified using end-point PCR for the skeletal muscle-specific gene MyoD (26) (Fig. 1). Total RNA concentration was determined using a NanoDrop (Thermo Scientific), and RNA integrity was verified using a Bioanalyzer (Agilent). For each reverse transcription reaction, 1 ng of total RNA was used. mRNA was reverse transcribed using a RT² kit and miRNA was reverse transcribed using a miScript II high spec system. Quantitative PCR was conducted using QuantiTect SYBR Green master mix and primers specific for target mRNA and miRNAs in a Bio-Rad CFX96 real time thermal cycler. The cycling program consisted of a denaturing cycle of 15 s at 94°C, an annealing cycle of 30 s at 55°C, and an extension cycle of 30 s at 70°C for 40 total cycles. Expression of mRNA transcripts was normalized to the stable housekeeping gene GAPDH, and miRNA transcripts were normalized to the stable housekeeping ncRNA Rnu6 using the methods of Schmittgen and Livak (47). The presence of single amplicons for mRNA reactions was verified by melt curve analysis. For endpoint PCR, cDNA was amplified using similar parameters, but products were subjected to electrophoresis in a 2% agarose gel stained with ethidium bromide. A list of transcripts and corresponding RefSeq and miRBase information is provided in Tables 1 and 2.

Fig. 1.

RNA isolated from cultured tendon fibroblasts and whole achilles tendons did not contain contaminating RNA from skeletal muscle tissue, as indicated by the absence of expression of the myogenic basic helix-loop-helix transcription factor MyoD.

Table 1.

mRNA transcripts evaluated by PCR or quantitative PCR

mRNA	RefSeq No.	Qiagen Catalog No.
Col1a1	NM_053304	PPR42922A
GAPDH	NM_017008	PPR06557A
MyoD	NM_176079	PPR44362
PCNA	NM_022381	PPR06518A
Scx	NM_001130508	PPR68624A
Tnmd	NM_022290	PPR48089A

Col1a1, type I collagen; PCNA, proliferating cell nuclear antigen; Scx, scleraxis; Tnmd, tenomodulin.

Table 2.

miRNA transcripts evaluated by quantitative PCR

miRNA	miRBase Accession No.	Qiagen Catalog No.
rno-let-7a	MIMAT0000774	MS00033131
rno-let-7b	MIMAT0000775	MS00000007
rno-let-7c	MIMAT0000776	MS00005467
rno-let-7d	MIMAT0000562	MS00012915
rno-let-7e	MIMAT0000777	MS00033145
rno-let-7f	MIMAT0000778	MS00005481
rno-miR-1	MIMAT0003125	MS00012943
rno-miR-100	MIMAT0000822	MS00033152
rno-miR-126	MIMAT0000832	MS00000329
rno-miR-126*	MIMAT0000831	MS00005607
rno-miR-133a	MIMAT0000839	MS00033208
rno-miR-133a*	MIMAT0017124	MS00026712
rno-miR-133b	MIMAT0003126	MS00033215
rno-miR-140	MIMAT0000573	MS00000406
rno-miR-140*	MIMAT0000574	MS00005635
rno-miR-143	MIMAT0000849	MS00000420
rno-miR-16	MIMAT0000785	MS00033229
rno-miR-205	MIMAT0000878	MS00028833
rno-miR-206	MIMAT0000879	MS00000623
rno-miR-21	MIMAT0000790	MS00013216
rno-miR-221	MIMAT0000890	MS00033313
rno-miR-222	MIMAT0000891	MS00008169
rno-miR-296	MIMAT0004742	MS00033369
rno-miR-29a	MIMAT0000802	MS00033397
rno-miR-29b	MIMAT0000801	MS00005544
rno-miR-29c	MIMAT0000803	MS00000175
rno-miR-30a	MIMAT0000808	MS00013363
rno-miR-335	MIMAT0000575	MS00000847
rno-miR-338	MIMAT0000581	MS00000868
rno-miR-378	MIMAT0003379	MS00005810
rno-miR-381	MIMAT0003199	MS00028903
rno-miR-450a	MIMAT0001547	MS00028931
rno-miR-743a	MIMAT0005334	MS00013832
rno-miR-743b	MIMAT0005280	MS00013839
rno-miR-872	MIMAT0005282	MS00013881
RNU6		MS00033740

miRNA, micro-RNA.

Statistical analyses.

Data are presented as means ± SD. Differences between sedentary and mechanically loaded tendons, and control and TGF-β-treated tendon fibroblasts were tested using Student's t-tests (α =0.05) in GraphPad Prism 5.0.

RESULTS

To determine if a single session of mechanical loading induced the expression of mRNAs with well-known roles in tendon growth and adaptation, we subjected rats to a 30-min treadmill run and measured gene expression 8 h later. Compared with age-matched rats that did not undergo mechanical loading, there was an increase in the expression of the ECM gene type I collagen, the general cell proliferation gene proliferating cell nuclear antigen (49), and the tendon fibroblast proliferation and matrix synthesis genes scleraxis and tenomodulin (Fig. 2). We then sought to determine the expression of different miRNA molecules with known roles in the regulation of cell proliferation and ECM synthesis (Fig. 3A) (4, 11, 15, 21, 23, 24, 35, 45, 50), skeletal muscle growth and adaptation (Fig. 3B) (9, 30, 31, 38, 51, 52), chondrogenesis and neovascularization (Fig. 3C) (1, 36, 42), the let-7 cluster (Fig. 3D) (8), and miRNAs (Fig. 3E) predicted to bind to the 3′-UTR of scleraxis or tenomodulin by mirSVR algorithm analysis (5, 14). As shown in Fig. 2, 12 out of 35 miRNAs evaluated showed a change in expression following an acute bout of mechanical loading.

Fig. 2.

Mechanical loading induces the expression of type I collagen (Col1a1), proliferating cell nuclear antigen (PCNA), scleraxis (Scx), and tenomodulin (Tnmd) in Achilles tendons. Target gene expression is normalized to the stable housekeeping gene GAPDH. Values are means ± SD; n = 5 rats from each group. *P < 0.05, significantly different from sedentary rat Achilles tendons.

Fig. 3.

Mechanical loading changes the expression of several microRNA (miRNA) transcripts associated with cell proliferation and extracellular matrix (ECM) synthesis (A), skeletal muscle adaptation (B), chondrogenesis and neovascularization (C), the let-7 cluster (D), and miRNAs (E) predicted to bind to the 3′-untranslated region (UTR) of scleraxis or tenomodulin. Target miRNA expression is normalized to the stable housekeeping noncoding RNA Rnu6. Values are means ± SD; n = 5 rats from each group. *P < 0.05, significantly different from sedentary rat Achilles tendons.

Since TGF-β plays an important role in tendon growth and adaptation, we next determined whether treating cultured primary tendon fibroblasts with TGF-β would result in similar changes in mRNA and miRNA expression. Compared with control tendon fibroblasts, treatment of tendon fibroblasts with TGF-β resulted in an increase in the expression of type I collagen, proliferating cell nuclear antigen, scleraxis, and tenomodulin (Fig. 4). We then measured the expression of the same miRNA molecules that were evaluated in whole Achilles tendons. In response to TGF-β treatment, there was a change in the expression of 16 out of 35 miRNAs (Fig. 5).

Fig. 4.

Transforming growth factor-β (TGF-β) treatment induces the expression of type I collagen, PCNA, scleraxis, and tenomodulin in cultured tendon fibroblast cells. Target gene expression is normalized to the stable housekeeping gene GAPDH. Values are means ± SD; n = 5 replicates from each group. *P < 0.05, significantly different from fibroblasts that were not treated with TGF-β.

Fig. 5.

TGF-β treatment changes the expression of several miRNA transcripts associated with cell proliferation and ECM synthesis (A), skeletal muscle adaptation (B), chondrogenesis and neovascularization (C), the let-7 cluster (D), and miRNAs (E) predicted to bind to the 3′-UTR of scleraxis or tenomodulin. Target miRNA expression is normalized to the stable housekeeping noncoding RNA Rnu6. Values are means ± SD; n = 5 replicates from each group. *P < 0.05, significantly different from fibroblasts that were not treated with TGF-β.

DISCUSSION

Previous studies have indicated that miRNA plays an important role in regulating the structure and function of many different musculoskeletal tissues, but to our knowledge this study is the first to examine changes in miRNA expression in tendons. As the positive adaptation of tendons to mechanical loading occurs due to an increase the amount of ECM and an increase in fibroblast density (33, 34), we evaluated the expression of mRNA and miRNAs that previous studies have indicated are involved with the regulation of ECM synthesis and cell proliferation. While we anticipated some overlap in differentially regulated miRNAs between mechanically loaded tendons and TGF-β-treated tendons, we found distinct changes in miRNA expression for whole tendons and cultured tendon fibroblasts. Inhibition of miR-100 in carcinoma cell lines is associated with increased cell proliferation (24), and overexpression of miR-378 enhances the proliferation of U87 and MT-1 cell lines (23). Consistent with these findings, we observed a decrease in miR-100 and an increase in miR-378 in mechanically loaded tendons. In Madin Darby canine kidney epithelial cells, TGF-β reduces the expression of miR-205, which increases the epithelial to mesenchymal transition and cell migration rates in these cells (15). miR-21 is expressed in cardiac fibroblasts, and increases in miR-21 are associated with increased collagen I synthesis in these cells (4). miR-221 and miR-222 promote cell proliferation by targeting the cell cycle inhibitor p27^Kip1 (35). In agreement with these findings, we observed a decrease in miR-205 and an increase in miR-21, miR-221, and miR-222 expression in TGF-β-treated fibroblasts. Due to their role in regulating cell proliferation, we also expected changes in the expression of miR-16 (45), miR-30a (21), miR-335 (50), and miR-450a (11) in whole tendons and treated fibroblasts but found no differences between control and treatment groups for these miRNAs. Combined, these results indicate that mechanical loading and TGF-β treatment regulate the expression of several miRNAs that regulate cell cycle progression and ECM synthesis.

The let-7 family were among the first miRNAs to be discovered and play important roles in the embryonic development of many tissues across various eukaryotic species (22, 39, 44). The let-7 family also appears to have important roles in regulating postnatal cell proliferation, and often there is an inverse relation between let-7 levels and cell proliferation rates (8). In skeletal muscle tissue, there is an aging-related increase in let-7b and let-7e that were postulated to play a role in the reduction in the proliferative capacity of cells from elderly subjects (13). In the current study, for both mechanically loaded tendons and TGF-β-treated fibroblasts, we observed a decrease in let-7a and let-7e expression. TGF-β-treated fibroblasts also had a reduction in let-7b. These results support the notion that the let-7 family is also likely important in controlling the postnatal adaptation of tendons to growth stimuli.

Skeletal muscle has an intimate anatomical association with tendon, and functional adaptations in muscle are often linked to adaptations in tendon (20). In biomechanical studies, the relationship between muscles and tendons is so linked that they are referred to as a combined “muscle-tendon unit” (28). Several studies have identified a group of miRNAs, miR-1, miR-133a, miR-133b, and miR-206 that are highly expressed in skeletal and cardiac muscle and play a role in the regulation of striated muscle to mechanical loading, unloading and various disease processes (30, 31). The high level of expression in striated muscle and the absence of their expression in liver, spleen, brain, and kidney led to the designation of miR-1, miR-133a, miR-133b, and miR-206 as “myomiRs” (29, 30, 48). In addition to the myomiRs, the miR-29 family has been implicated in the TGF-β-mediated differentiation of myoblast cells (52). For both mechanically loaded tendons and TGF-β-treated fibroblasts, we observed an increase in miR-1, miR-133a, miR-133a*, and miR-133b expression. Loaded Achilles tendons also had an increase in miR-206 expression. The current results from tendons are generally consistent with changes in myomiR expression observed in skeletal muscle following an acute bout of exercise (38, 46). In cultured tendon fibroblasts, TGF-β treatment induced the expression of miR-29, similar to what has been previously reported for myoblasts (52). Since TGF-β plays an important role in regulating the adaptation of both muscle and tendon to mechanical loading (6, 10, 17, 20), it is likely that there are overlapping molecular pathways in the regulation of myoblasts and tendon fibroblasts to mechanical loading. The results from the current study provide the first evidence that myomiRs are expressed in tendon cells, have similar changes in gene expression in tendons response to growth stimuli, and may therefore be important in the adaptation of the ECM of muscle-tendon units to mechanical loading.

Tendinosis is a painful and often times debilitating disease of tendons with a poorly understood etiology but likely occurs due to a failure of the molecular pathways that regulate the proper adaptation of tendons to mechanical loading (19, 53). The tendons of patients with tendinosis are characterized by increases in cartilage-like ECM and dramatic increases in vascularity (2, 3, 19, 41). To determine if an acute bout of exercise or TGF-β treatment changed the expression of miRNAs with known roles in chondrogenesis and angiogenesis, we evaluated the expression of miR-140, miR-126, miR-143, and miR-296. miR-140 is required for the proper embryonic development of cartilage and also plays an important role in postnatal cartilage homeostasis (36). For whole tendons, there was a decrease in miR-140 expression in mechanically loaded tendons, but interestingly we observed an increase in miR-140 in tendon fibroblasts treated with TGF-β. While the results from whole tendons would be anticipated, the reason for an upregulation of miR-140 in cultured tendon fibroblasts is not clear. miR-126, miR-143 and miR-296 are associated with angiogenesis and neovascularization in other tissues (1, 42), but no difference in these miRNAs was observed in any groups in the current study. Combined, for whole tendons, the results suggest that a single bout of mechanical loading does not produce changes in the expression of miRNAs in a way that would promote chondrogenesis or neovascularization.

Scleraxis is a bHLH transcription factor that is required for the proper embryonic development of limb tendons (37). Scleraxis expression is induced in tendon fibroblast cells in response to treatment with both TGF-β and myostatin, which is a member of the TGF-β-superfamily (25, 27, 32, 40). Tendomodulin is a type II transmembrane glycoprotein that promotes tendon fibroblast proliferation (12) and is also downstream of TGF-β and myostatin signaling (25, 32). Scleraxis and tenomodulin are also expressed in adult tendons and are upregulated following a 6-wk mechanical loading regime (33), and the current results indicate they are also upregulated following an acute bout of mechanical loading. While scleraxis and tenomodulin both play important roles in regulating tendon fibroblast biology, it is not known whether miRNA molecules regulate the expression of these transcripts. We used the mirSVR algorithm and the Ensembl genome browser to identify potential miRNAs that bind to the 3′-UTR of scleraxis or tenomodulin (5, 14). miRNAs that were predicted to bind to the 3′-UTR of scleraxis include miR-338, miR-381, miR-743a, and miR-743b, while miR-872 was predicted to bind to the 3′-UTR of tenomodulin. For both mechanically loaded tendons and TGF-β-treated tendon fibroblasts, consistent with an increase in scleraxis expression, there was a decrease in miR-338 and miR-381. For whole tendons, there was also a decrease in miR-743a. No difference in miR-872 was observed for either loaded tendons or TGF-β-treated fibroblasts. While further studies are necessary, these results suggest that scleraxis may be regulated by miR-338 and miR-381.

There are several limitations to the current study. We only evaluated changes in miRNA expression after a single session of mechanical loading and a single dose of TGF-β. With the exception of miRNAs that are predicted to interact with scleraxis and tenomodulin, we chose to evaluate miRNAs that already have known roles in regulating cell proliferation and ECM production in other cell types. Further studies using a combination of microarrays, deep sequencing, bioinformatics pathway analyses and quantitative PCR, along with different time points, mechanical loading conditions, and TGF-β doses, would provide further insight into the role of miRNA in tendon adaptation. Additionally, studies looking at differences in miRNA expression patterns between healthy and diseased tendons are likely to provide important insight into the etiology of tendinopathies. Despite these limitations, the current results provide the first insight into miRNA and tendons, and we hope will be the first of many studies exploring the role of miRNA in tendon mechanobiology.

GRANTS

This work was supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant AR-058920.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: C.L.M., J.P.G., and E.B.L. conception and design of research; C.L.M., J.P.G., and E.B.L. performed experiments; C.L.M., J.P.G., and E.B.L. analyzed data; C.L.M., J.P.G., and E.B.L. interpreted results of experiments; C.L.M. prepared figures; C.L.M. drafted manuscript; C.L.M., J.P.G., and E.B.L. edited and revised manuscript; C.L.M., J.P.G., and E.B.L. approved final version of manuscript.