Research ArticleTENDINOPATHY

Inflammation activation and resolution in human tendon disease

See allHide authors and affiliations

Science Translational Medicine  28 Oct 2015:
Vol. 7, Issue 311, pp. 311ra173
DOI: 10.1126/scitranslmed.aac4269

Tending toward resolution

Inflammation activation and resolution pathways remain poorly defined in tendon disease. In new work, Dakin et al. investigate shoulder tendons from patients before and after surgery. Diseased tendons showed different inflammation gene and protein signatures in early-stage disease compared to advanced-stage disease. The researchers identified pathways implicated in the resolution of tendon pain after surgical treatment. Investigation of inflammation activation pathways in cultured stromal cells derived from human diseased tendons revealed that the stromal cells may have been primed for inflammation. The authors also identified a stable metabolite of aspirin that may be therapeutically beneficial for the resolution of tendon inflammation.

Abstract

Improved understanding of the role of inflammation in tendon disease is required to facilitate therapeutic target discovery. We studied supraspinatus tendons from patients experiencing pain before and after surgical subacromial decompression treatment. Tendons were classified as having early, intermediate, or advanced disease, and inflammation was characterized through activation of pathways mediated by interferon (IFN), nuclear factor κB (NF-κB), glucocorticoid receptor, and signal transducer and activator of transcription 6 (STAT-6). Inflammation signatures revealed expression of genes and proteins induced by IFN and NF-κB in early-stage disease and genes and proteins induced by STAT-6 and glucocorticoid receptor activation in advanced-stage disease. The proresolving proteins FPR2/ALX and ChemR23 were increased in early-stage disease compared to intermediate- to advanced-stage disease. Patients who were pain-free after treatment had tendons with increased expression of CD206 and ALOX15 mRNA compared to tendons from patients who continued to experience pain after treatment, suggesting that these genes and their pathways may moderate tendon pain. Stromal cells from diseased tendons cultured in vitro showed increased expression of NF-κB and IFN target genes after treatment with lipopolysaccharide or IFNγ compared to stromal cells derived from healthy tendons. We identified 15-epi lipoxin A4, a stable lipoxin isoform derived from aspirin treatment, as potentially beneficial in the resolution of tendon inflammation.

INTRODUCTION

Pathology of musculoskeletal soft tissues is a common and significant health care problem both in sports injuries and in the ageing population. Shoulder pain is the third commonest orthopedic problem to present to clinicians (1, 2). Shoulder soft tissue pathology such as rotator cuff tears cause pain, loss of function, joint failure, and development of secondary osteoarthritis with a substantial social and economic burden. Rotator cuff pathology can be classified as early-stage disease in intact tendons or intermediate- to advanced-stage disease in torn tendons. Inflamed musculoskeletal soft tissues heal by the formation of a repair scar; however, the normal architecture, composition, and function of the tissues are not fully restored (3, 4), thus increasing susceptibility to reinjury. Multiple therapies have been advocated in patients with rotator cuff tendinopathy (most frequently affecting the supraspinatus tendon) with mixed results, including physiotherapy, nonsteroidal anti-inflammatory drugs (NSAIDs), and local injections of glucocorticoids and platelet-rich plasma. Surgical subacromial decompression and rotator cuff tear repair are performed in patients with persistent symptoms (about 25,000 subacromial decompression surgeries and 10,000 repairs in the UK, and 300,000 repairs in the United States annually) (5). Surgical repair of tendon tears is associated with high postoperative failure rates of about 40% (6, 7). Whereas the effects of repetitive wear and tear, daily exercise, ageing, and genetic factors are cited as important contributing factors (811), our incomplete understanding of the mechanisms underpinning tendon pathology hampers the development of efficacious new therapies.

The identification of key immune cell populations as master regulators of inflammation has advanced our understanding of pathogenic mechanisms in other rheumatic diseases such as rheumatoid arthritis and the spondyloarthropathies (1215). Growing evidence supports the contribution of inflammation to the development of tendinopathy (1621). Whereas animal models of induced tendinopathy have improved our understanding of acute inflammatory processes (22), the phenotypes of the key cells orchestrating inflammation and fibrosis in human soft tissue pathologies such as tendinopathy and enthesopathy have not been fully characterized. Macrophages play an essential role in homeostasis and pathology, orchestrating inflammation and repair in many tissues. The signaling pathways underpinning activation of macrophages to become M1 or M2 subtypes (2325) have been revised to highlight receptors and key signaling mediators in common and distinct pathways. These include proinflammatory pathways containing interferon (IFN) and nuclear factor κB (NF-κB), profibrotic pathways containing signal transducer and activator of transcription 6 (STAT-6), and inflammation resolving pathways involving glucocorticoid receptor activation (26).

Recent studies highlight the importance of tissue microenvironments and the innate immune response in perpetuating the inflammatory process. Nonmyeloid cell populations such as resident stromal cells also play a prominent role in the generation and maintenance of chronic inflammation (27, 28). Cancer-associated fibroblasts are known to become “polarized” within the tumor microenvironment. This “polarization” reflects the altered activation status and phenotype of these cells. The interactions between tumor-associated macrophages and stromal cells have been implicated in chronic inflammation (29, 30). However, inflammation activation pathways and the interplay between myeloid and stromal cells are understudied in diseased human musculoskeletal soft tissues from patients with tendinopathy or enthesopathy.

Inflammation resolution is a highly active and coordinated process (31). In health, a repertoire of proresolving lipids and proteins derived from myeloid and stromal cells promote the timely resolution of the inflammatory response and restoration of tissue homeostasis (3236). Inadequate or dysregulated resolution is thought to contribute to the development of many systemic chronic inflammatory diseases (37, 38). Resolution has been well studied in experimental mouse models of inflammation (3942) but not in diseased human tendons.

Here, we characterized inflammation activation pathways in diseased human supraspinatus tendons. We studied tendon tissue samples from a well-phenotyped longitudinal cohort of symptomatic patients with early to advanced disease before and after surgical subacromial decompression treatment. In some patients, the symptoms resolved after treatment, whereas other patients remained symptomatic; the posttreatment samples allowed us to compare the inflammatory signatures from these two groups. We also compared inflammation activation pathways in stromal cells derived from healthy asymptomatic human tendons and pathological painful human tendons. We hypothesized that inflammation signatures would differ throughout the spectrum of supraspinatus tendon pathology and after the resolution of clinical symptoms. We also tested whether stromal cells from diseased tendons exhibited a more proinflammatory phenotype after cytokine stimulation compared to cells derived from healthy tendons.

RESULTS

Diseased tendons show increased numbers of CD14+ and CD68+ myeloid cells

Resident macrophage populations in diseased human tendons are not well studied, and little is known of their phenotype. The macrophage infiltrate in diseased tendons could derive from the adult axis of inflammatory CD14+ monocytes that extravasate and mature into CD68+ macrophages. Immunohistochemistry was performed to determine whether monocytes (CD14+ cells) or tissue-resident macrophages (CD68+ cells) were present in samples of early, intermediate, and advanced stages of disease in supraspinatus tendons and was compared to healthy supraspinatus tendons. Morphologically, diseased samples showed a marked increase in cellularity compared to healthy tendons, as shown by quantitative analyses (Fig. 1, A and B) of immunopositive staining (Fig. 1, C to H). There was an increase in CD14+ monocytes/macrophages in early-stage (P = 0.0004) and intermediate- to advanced-stage diseased tendons (P = 0.0026) compared to healthy tendon samples (Fig. 1A). There was also an increase in CD68+ macrophages in early-stage (P = 0.0012) and intermediate- to advanced-stage diseased tendons (P = 0.0002) compared to healthy samples; CD68+ macrophages were increased in early-stage compared to intermediate- to advanced-stage diseased tendons (P = 0.004) (Fig. 1B). The proportion of myeloid cells expressing both CD14 and CD68 was 13.4% (±5%), suggesting that most monocytes and macrophages in diseased tendons were distinct populations (Fig. 1, I and J).

Fig. 1. Immunohistochemistry showing CD14+ and CD68+ monocytes/macrophages in healthy and diseased human supraspinatus tendons.

(A and B) Graphs show quantitative analysis of CD14+ cells (monocytes) and CD68+ cells (macrophages) in healthy tendons and in early-stage and intermediate- to advanced-stage diseased tendons. Bars show median values. Statistically significant differences were calculated using Kruskal-Wallis tests with pairwise post hoc Mann-Whitney U tests. **P < 0.01, ***P < 0.001. (C to H) Panels show representative images of 3,3′-diaminobenzidine immunostaining (brown) for CD14 and CD68 in healthy (C and F) and early- (D and G) and intermediate- to advanced-stage diseased tendons (E and H). Nuclear counterstain is hematoxylin. Scale bar, 50 μm. (I and J) Representative confocal immunofluorescence images showing dual labeling for CD14 (purple) and CD68 (green) in sections of early- and advanced-stage diseased tendons. Cyan represents POPO-1 nuclear counterstain. Scale bar, 20 μm.

Inflammation activation signatures are altered in different stages of tendon disease

Having demonstrated the presence of CD14+ and CD68+ macrophages in diseased human tendons, we investigated whether products associated with macrophage activation could be identified in whole tendon tissue derived from patients with tendon pathology. To study gene expression, we used a panel of genes including those known to be involved in activation of macrophages, such as IFN, NF-κB, interleukin (IL)–13, IL-4, STAT-6, and the genes encoding the glucocorticoid receptor pathway (table S1). We investigated these inflammation activation pathways in samples of early-, intermediate-, and advanced-stage tendon disease compared to healthy tendons and also in samples collected from patients who showed resolution of their clinical symptoms after surgical subacromial decompression treatment. Changes in inflammation activation gene expression signatures in diseased tendons relative to healthy tendons are shown in table S2. Supraspinatus tendon samples from patients with early-stage disease before surgical subacromial decompression treatment (n = 6) showed a mixed inflammation gene signature, including activation of IFN, NF-κB, and STAT-6 pathways, compared to subscapularis tendons from healthy patients (Fig. 2A). There was increased expression of the IFN-induced genes CXCL11 and serine/arginine repetitive matrix 2 (SRRM2), the NF-κB–induced gene indoleamine 2,3-dioxygenase 1 (IDO1), and the STAT-6–induced genes CD206 and ALOX15 in early-stage diseased supraspinatus tendons compared to healthy subscapularis tendons (Fig. 2A). The IFN-induced genes vesicle-associated membrane protein 5 (VAMP5) and tryptophanyl–transfer RNA synthetase (WARS) were overrepresented in early-stage compared to advanced-stage diseased tendons (table S2). IL-10 mRNA expression was reduced in early-stage diseased tendons compared to healthy tendons (P = 0.047, 2.7-fold decrease) (Fig. 2A). A mixed protein signature was also observed in sections of tendons with early-stage disease, including IRF5 (IFN regulatory factor 5; a marker of IFN activation), IDO-1 (NF-κB activation), CD206 (STAT-6 activation), and CD163 (glucocorticoid receptor activation) (Fig. 2, B and C). Quantitative analysis of immunostaining showed an IFN and NF-κB activation signature, with 64 and 44% of immunopositive cells expressing IRF5 and IDO-1 proteins, respectively (Fig. 2D).

Fig. 2. Expression of inflammation activation pathway genes and proteins in early-stage diseased supraspinatus tendons compared to healthy subscapularis tendons.

(A) A mixed gene expression signature was present in pretreatment tendon samples with early disease (n = 6) compared to healthy subscapularis control tendons (n = 3). Shown is the expression of genes induced by IFN (CXCL11 and SRRM2), NF-κB (IDO1), and STAT-6 (ALOX15 and CD206). Fold changes in gene expression are shown in table S2. There was greater expression of IFN-induced genes including CXCL11 and SRRM2 and the NF-κB–induced gene IDO1 in early-stage diseased supraspinatus tendons compared to healthy subscapularis control tendons. Gene expression is normalized to β-actin; bars show median values. (B and C) Representative immunofluorescence images of sections of pretreatment early-stage diseased tendons stained for inflammation activation markers including those of the STAT-6 pathway (CD206, green), the glucocorticoid receptor pathway (CD163, red), the IFN pathway (IRF5, purple), and the NF-κB pathway (IDO-1, red). MerTK (purple) represents Mer tyrosine kinase, a tissue-resident macrophage marker. Cyan represents POPO-1 nuclear counterstain. Scale bar, 20 μm. (D) Quantitative analysis showing the percentage of immunopositive staining for inflammation activation markers. Data are shown as means and SEM.

A mixed inflammation activation gene signature was also seen in intermediate-stage samples of diseased supraspinatus tendons (Fig. 3). These samples showed an NF-κB activation gene signature with increased expression of NF-κB pathway genes including tumor necrosis factor (TNF) mRNA (P = 0.047, mean 7.5-fold increase) and CCL20 mRNA (P = 0.047, mean 6.5-fold increase) compared to healthy subscapularis tendons (Fig. 3A). Genes expressed in the STAT-6 pathway (CD206, P = 0.024, mean 53-fold increase) and the glucocorticoid receptor activation pathway (pentraxin 3, P = 0.024, mean 8.9-fold increase) were also increased in these samples compared to healthy tendons (Fig. 3A). In support of this, a mixed protein signature was observed in intermediate-stage tendon disease with increased expression of IRF5 (IFN activation), IDO-1 (NF-κB activation), CD206 (STAT-6 activation), and CD163 (glucocorticoid receptor activation) (Fig. 3, B and C). Quantitative analysis of staining showed an NF-κB and IFN signature, with 90 and 70% of immunopositive cells expressing IDO-1 and IRF5 proteins, respectively (Fig. 3F). In early and intermediate stages of disease, tendon stromal cells that expressed neither CD68 nor CD206 expressed IRF5 and IDO-1, suggesting that, in addition to macrophages, a small proportion of non-macrophage cells such as tendon stromal cells also expressed these proteins (Figs. 2B and 3B).

Fig. 3. Inflammation activation pathway signatures in intermediate- and advanced-stage diseased tendons.

(A) A mixed inflammation activation gene signature in intermediate-stage diseased supraspinatus tendon samples (n = 6) compared to healthy subscapularis tendons (n = 3). (B and C) Representative immunofluorescence images of the inflammation activation protein signature in intermediate-stage diseased tendons. (B) Markers indicate activation of the following pathways: STAT-6 (CD206, green), NF-κB (IDO-1, red), and IFN (IRF5, purple). (C) Markers indicate activation of the glucocorticoid receptor pathway (CD163, red); also shown is the tissue-resident macrophage marker Mer tyrosine kinase (purple). (D) A STAT-6/glucocorticoid receptor gene signature predominates in advanced-stage diseased supraspinatus tendons (n = 5) compared to healthy subscapularis tendons (n = 3). Gene expression is normalized to β-actin. Statistically significant differences were calculated using pairwise Mann-Whitney U tests. Bars represent median values. *P < 0.05. (E) Representative immunofluorescence images of the inflammation activation protein signature in advanced-stage diseased supraspinatus tendons. Shown is the staining for the STAT-6 (CD206, green) and glucocorticoid receptor (CD163, red) activation pathways; tissue-resident macrophages are stained with Mer tyrosine kinase (purple). Cyan represents POPO-1 nuclear counterstain. Scale bar, 20 μm. (F) Quantitative analysis showing the percentage of cells staining immunopositive for inflammation pathway proteins. Data are shown as means and SEM.

A transition in inflammation activation signature was observed in advanced-stage diseased tendon samples (Fig. 3). A STAT-6/glucocorticoid receptor activation gene signature predominated in these samples compared to healthy subscapularis tendons, with increased expression of CD206 mRNA (P = 0.036, mean >500-fold increase), CD1D mRNA (mean 2.7-fold increase), complement component 1 mRNA (CIQA, mean 6.6-fold increase), and CD163 mRNA (mean 6.9-fold increase) (Fig. 3D). In support of this, proteins in the STAT-6 pathway (CD206) and the glucocorticoid receptor pathway (CD163) were highly expressed in advanced-stage diseased tendons (Fig. 3, E and F). The transition in inflammation signatures between intermediate- and advanced-stage diseased tendons was further supported by immunostaining of phosphorylated NF-κB–p65 and unphosphorylated NF-κB–p65. Tendon samples with intermediate-stage disease showed increased expression of both phosphorylated and unphosphorylated forms of NF-κB–p65 compared to advanced-stage diseased tendon samples (fig. S1A). Dual labeling demonstrated that NF-κB–p65 expression predominantly occurred on CD68 cells (fig. S1B). Low-level expression of phosphorylated STAT-6 (pSTAT-6) occurred in both intermediate- and advanced-stage diseased tendons with pSTAT-6 expressed by CD68+ macrophages (fig. S2).

Stromal cells from diseased tendons are primed for inflammation

Inflammation activation pathways identified in diseased human tendons were further studied in healthy and diseased stromal cells derived from tendons in vitro to investigate whether cytokines could induce a proinflammatory phenotype. Treatment of stromal cells derived from healthy human hamstring or torn supraspinatus tendons with IFNγ, IL-13 (each 20 ng ml−1), or lipopolysaccharide (LPS) (100 ng ml−1) for 96 hours increased the expression of inducible target genes compared to untreated controls. Both healthy and diseased tendon–derived stromal cells showed increased expression of IFN target genes including apolipoprotein L3 (APOL3), IFN regulatory factor 1 (IRF1), and VAMP5 in IFNγ-treated compared to untreated control tendon–derived stromal cells (Fig. 4A). IFNγ-treated diseased supraspinatus tendon cells showed increased expression of VAMP5 mRNA (P = 0.03) compared to IFNγ-treated healthy hamstring cells (Fig. 4A). Treatment of hamstring or supraspinatus tendon–derived stromal cells with LPS boosted the expression of NF-κB target genes including IDO1, IL6, and CCL2 compared to untreated control tendon cells (Fig. 4B). LPS-treated diseased supraspinatus tendon–derived stromal cells showed increased IDO1 (P = 0.04) and IL6 (P = 0.009) mRNA expression compared to LPS-treated healthy hamstring–derived stromal cells. IL-13 treatment induced the expression of STAT-6 pathway genes including ALOX15, transglutaminase-2 (TGM2), and fibrinogen-like protein 2 (FGL2) compared to untreated control tendon–derived stromal cells (Fig. 4C). Differences were observed between stromal cells derived from healthy hamstring tendons and stromal cells derived from diseased supraspinatus tendons. The expression of a potential marker of resolution, ALOX15, was increased in healthy hamstring tendon–derived stromal cells compared to those from diseased supraspinatus tendons (mean threefold increase). There was increased expression of the fibrogenic marker TGM2 (mean twofold increase) and reduced expression of FGL2 (P = 0.002, mean sevenfold decrease) in diseased supraspinatus tendon–derived stromal cells compared to healthy hamstring tendon–derived stromal cells (Fig. 4C).

Fig. 4. Gene expression after treatment of cultured tendon-derived stromal cells with IFNγ, LPS, or IL-13 for 96 hours.

Tendon stromal cells were derived from healthy hamstring or diseased supraspinatus tendons (n = 6 donors for LPS and IL-13 treatment; n = 5 donors for IFNγ treatment). (A) Treatment of cells from healthy hamstring (Ham) or diseased supraspinatus tendons (Supra) with IFNγ (20 ng ml−1). (B) Stimulation of cells from healthy hamstring or diseased supraspinatus tendons with LPS (100 ng ml−1). (C) Stimulation of cells from healthy hamstring or diseased supraspinatus tendons with IL-13 (20 ng ml−1). Statistically significant differences were calculated using pairwise Mann-Whitney U tests. Gene expression is normalized to β-actin; bars represent median values. *P < 0.05, **P < 0.01.

IFNγ, IL-13, or LPS treatment of cocultures between tendon-derived stromal cells (from healthy hamstring or torn supraspinatus tendons) and macrophages for 96 hours also increased the expression of respective inducible genes compared to untreated control cocultures. IFNγ treatment boosted the expression of IFN target genes APOL3, IRF1, and VAMP5 compared to untreated controls (fig. S3A). There was a trend for increased expression of VAMP5 mRNA in supraspinatus tendon cell/macrophage cocultures. LPS treatment induced the expression of NF-κB target genes CCL2, IL6, and IL8 compared to untreated controls in both hamstring and supraspinatus tendon–derived stromal cell/macrophage cocultures (fig. S3B). IL-13 treatment induced expression of STAT-6 pathway target genes compared to untreated controls, with a trend toward increased expression of cytokine-inducible SH2-containing protein (CISH), IL17RB, and TGM2 genes in healthy hamstring tendon–derived stromal cell/macrophage cocultures compared to diseased supraspinatus tendon–derived stromal cells (fig. S3C).

Increased expression of proresolving proteins in early-stage diseased tendons

In addition to characterizing the plasticity of inflammation activation signatures in the different stages of tendon disease, we also investigated whether inflamed supraspinatus tendons were capable of mounting a resolution response. We investigated the expression of the proresolving proteins formyl peptide receptor 2 (FPR2/ALX) and chemerin receptor 23 (ChemR23) in sections of healthy tendons and tendons at the early and intermediate to advanced stages of disease. There was increased expression of FPR2/ALX in early-stage diseased tendons compared to healthy tendons (P = 0.01) and intermediate- to advanced-stage diseased tendons (P = 0.0002) (Fig. 5A). There was also increased expression of ChemR23 in early-stage diseased tendons compared to healthy tendons (P = 0.0025) and intermediate- to advanced-stage diseased tendons (P = 0.0006) (Fig. 5B). Immunofluorescence labeling showed low-level expression of FPR2/ALX and ChemR23 by CD68+ and CD206+ macrophages in early-stage diseased tendons (Fig. 5, C and D). Tendon-derived stromal cells that expressed neither CD68 nor CD206 also expressed the proresolving proteins FPR2/ALX and ChemR23.

Fig. 5. Expression of proresolving proteins in healthy and diseased supraspinatus tendons.

(A and B) Quantitative analyses show increased expression of FPR2/ALX (A) and ChemR23 (B) in early-stage diseased tendons compared to intermediate- to advanced-stage diseased tendons or healthy tendons. Bars represent median values. Statistically significant differences were calculated using Kruskal-Wallis tests with pairwise post hoc Mann-Whitney U tests. *P < 0.05, **P < 0.01, ***P < 0.001. (C and D) Representative confocal immunofluorescence images showing staining for macrophages (CD206+, CD68+, and MerTK+) and the proresolving proteins ChemR23 (red, C) and FPR2/ALX (purple, D) in sections of early-stage diseased tendons. Cyan represents nuclear counterstain. Scale bar, 20 μm.

ALOX15 and CD206 pathways are implicated in the resolution of pain after surgical treatment

To investigate whether inflammation activation pathways in diseased tendons are altered after surgical treatment, we compared biopsies from patients with supraspinatus tendinopathy who remained in pain after treatment (n = 5) with biopsies from patients who became pain-free after treatment (n = 6). The interval between surgical intervention and repeat biopsy was 2 to 4 years in both pain-free and painful posttreatment groups. The IFN activation gene signature seen in early-stage tendon disease before treatment persisted in samples from tendinopathy patients from both groups after treatment, with the expression of the IFN target genes APOL3, CXCL11, IRF1, VAMP5, WARS, and SRRM2 remaining unchanged (fig. S4). However, tendon samples from pain-free posttreatment patients showed increased expression of ALOX15 mRNA (P = 0.004, mean 4.3-fold increase) and CD206 mRNA (P = 0.017, mean 6.3-fold increase) compared to tendon samples from patients who continued to experience pain after treatment (Fig. 6). CD206 mRNA expression was also increased in the pain-free posttreatment group compared to patients before treatment (P = 0.026, mean 4.9-fold increase) (Fig. 6). CCL4 mRNA expression was reduced in samples from pain-free posttreatment patients compared to the painful pretreatment group (P = 0.009, mean 4.3-fold decrease) (Fig. 6).

Fig. 6. Expression of ALOX15, CD206, and CCL4 mRNA in biopsies from patients with early-stage tendon disease before and after surgical subacromial decompression treatment.

Pretreatment tendinopathy samples from patients experiencing pain (n = 6) were obtained after presentation to a referral clinic and before surgery. Biopsies were also collected from patients 2 to 4 years after treatment. Posttreatment patients either had persistent pain (n = 5) or were pain-free (n = 6). Statistically significant differences were calculated using pairwise Mann-Whitney U tests. Gene expression is normalized to β-actin; bars represent median values. *P < 0.05, **P < 0.01.

15-Epi lipoxin A4 promotes resolution of inflammation in LPS-treated stromal cells derived from diseased tendons

Given that we found increased CD206 and ALOX15 mRNA expression in tendon samples from patients whose clinical symptoms had resolved after treatment, we sought to therapeutically augment these pathways in stromal cells derived from diseased supraspinatus tendons in vitro. Low-dose aspirin is known to promote the release of stable aspirin-triggered isoforms of lipoxin A4 (LXA4), which potentiate resolution of inflammation (33, 43). In light of the importance of promoting timely resolution of tendon inflammation and the proinflammatory phenotype exhibited by cells from early- and intermediate-stage diseased tendons, we investigated whether a stable lipoxin isoform derived from aspirin treatment (15-epi LXA4) could potentiate CD206 and ALOX15 mRNA expression in an in vitro model of tendon inflammation. Treatment of stromal cells derived from intermediate-stage diseased tendons [n = 5 (minimum)] with LPS (100 ng ml−1) in the presence of 50 nM 15-epi LXA4 increased CD206 mRNA expression after 96 hours compared to tendon stromal cells treated with LPS alone (P = 0.002, 20.2-fold increase) (Fig. 7A). Treatment of diseased tendon cells with LPS in the presence of 100 nM 15-epi LXA4 increased the expression of both ALOX15 mRNA (P = 0.016, 4.8-fold increase) and CCL22 mRNA (P = 0.03, 3.1-fold increase) and reduced the expression of IL12B mRNA (P = 0.008, 9.6-fold decrease) after 96 hours, compared to tendon cells stimulated with LPS alone (Fig. 7, B to D).

Fig. 7. The aspirin metabolite 15-epi LXA4 modulates inflammation in LPS-treated stromal cells derived from diseased tendons.

Tendon cells were derived from intermediate-stage diseased supraspinatus tendons from a minimum of five patients. (A) Tendon-derived cells were treated with LPS (100 ng ml−1) in medium containing 50 nM 15-epi LXA4 for 96 hours. There was increased expression of CD206 mRNA after treatment with LPS and 15-epi LXA4 compared to LPS alone (cells derived from tendons from six patients). (B to D) Tendon cells were treated with LPS (100 ng ml−1) in medium containing 100 nM 15-epi LXA4 for 96 hours. There was increased expression of ALOX15 mRNA (B) and CCL22 mRNA (C) and reduced expression of IL12B mRNA (D) compared to LPS-alone treatment (cells derived from tendons from five patients). Statistically significant differences were calculated using pairwise Mann-Whitney U tests. Gene expression is normalized to β-actin. Bars show median values. *P < 0.05, **P < 0.01.

DISCUSSION

Monocytes and macrophages exert dichotomous roles during the stages of inflammation, repair, and remodeling of tissues (4446). Diseased tendons from patients with tendinopathy showed an abundance of CD14+ and CD68+ myeloid cells, supporting the concept of an active axis of recruitment of CD14+ monocytes from blood and a process of maturation of these cells into CD68+ tissue-resident macrophages. Although it is increasingly accepted that these inflammatory cell populations contribute to the initiation of tendon pathology, it is not known how inflammation activation pathways in tendon tissue change during the progression of pathology or after clinical treatment. This study provides new insights into the pathogenesis of diseases affecting musculoskeletal soft tissues. This longitudinal cohort study of phenotyped human tendon tissues collected before and after treatment revealed that the gene and protein signatures of inflammation are complex and show a high degree of plasticity, with an activation pattern that parallels that of macrophages. Samples from patients with early-stage tendon disease showed a mixed inflammation signature with increased expression of genes and proteins regulated by IFN and NF-κB activation pathways. Intermediate-stage diseased tendons showed a mixed inflammation signature with increased expression of NF-κB target genes and proteins; advanced-stage diseased tendons had an inflammation signature where expression of STAT-6 and glucocorticoid receptor pathway components predominated. This transition in inflammation activation signatures suggests that different therapeutic interventions may be applicable at different disease stages. Furthermore, the plasticity of inflammation signatures in diseased tendons has important implications for the surgical repair of tendon tears with biomaterials, as immune cell populations have been shown to influence biomaterial compatibility (47, 48).

Further investigation into the mechanisms by which inflammation becomes persistent during tendinopathy will be essential to drive development of new therapies to target chronic inflammatory diseases of musculoskeletal soft tissues. In the current study, inflammation activation pathways identified in samples of diseased tendons were further studied in vitro to investigate whether activation of these inflammatory pathways induced tendon-derived stromal cells to become proinflammatory. Treatment of healthy and diseased tendon stromal cells with IFNγ, LPS, or IL-13 induced the expression of respective target genes compared to untreated controls. This effect was observed in experiments where tendon cells were stimulated in isolation and in direct cocultures with macrophages. Notably, LPS induced increases in IDO1 and IL6 mRNA expression and IFNγ induced increases in VAMP5 mRNA expression in diseased compared to healthy tendon–derived stromal cells. These findings suggest that inflammation alters the activation status of tendon-derived stromal cells. This concept is further supported by the expression of IFN and NF-κB target proteins such as IRF5, IDO1, and phosphorylated NF-κB–p65 by nonmyeloid cells in sections of diseased human tendons. We suggest that tendon-derived stromal cells become altered by the inflammatory milieu and adopt a proinflammatory phenotype. It is possible that diseased tendon cells that have been previously exposed to an inflammatory milieu will be primed and become hyperresponsive on subsequent exposure to proinflammatory mediators. In support of this, studies of cells in the tumor microenvironment (29, 30) and of fibroblasts derived from rheumatoid arthritis tissue (27, 28) indicate that resident stromal cell populations contribute to the development of chronic inflammation.

Studies of rheumatoid arthritis and the spondyloarthropathies have revealed interactions between myeloid and stromal cell populations during inflammation, identifying potential disease stage–specific therapeutic targets. To identify factors implicated in the resolution of tendon pain after surgical subacromial decompression treatment, we compared inflammation signatures in tendon samples collected from patients whose symptoms resolved after treatment and compared them with samples from patients who remained symptomatic. Key differences in the expression of NF-κB and STAT-6 activation pathways were identified in patients with tendon pathology before and after treatment. The expression of CCL4 mRNA was increased in tendons from pretreatment patients experiencing pain compared to posttreatment patients who were pain-free. In support of this, increased expression of CCL4 mRNA in injured nerves has been shown to correlate with neuropathic pain in a mouse model of sciatic nerve injury (49). We identified activation of inflammation-resolving pathways in tendon samples from patients 2 to 4 years after surgery who were pain-free. These tendon samples showed increased expression of CD206 and ALOX15 mRNA compared to tendons from patients who remained symptomatic after treatment. Macrophages expressing CD206 are frequently associated with the repair and remodeling of healing tissues (50) and have also been identified in healing equine digital flexor tendons (18). ALOX15 is a critical enzyme in the biosynthesis of lipid mediators such as the lipoxins (LXA4 and LXB4), promoting resolution and restoration of homeostasis in inflamed tissues (51, 52). These findings implicate the CD206 and ALOX15 pathways in the resolution of inflammatory-associated pain in tendon samples from patients whose clinical symptoms resolved after treatment.

In the current study, the expression of proresolving proteins FPR2/ALX and ChemR23 was increased in early-stage tendon pathology, suggesting that tendons mount a counterresolution response to inflammation. However, this activation of proresolving pathways was not sustained in intermediate- to advanced-stage diseased tendons. Samples from patients with advanced-stage tendinopathy showed increased numbers of CD14+ and CD68+ macrophages with diminished expression of the proresolving proteins FPR2/ALX and ChemR23 and marked disorganization of the extracellular matrix, implying failure of the tissue to return to homeostasis. In support of the findings from the current study, FPR2/ALX has also been shown to increase in the early stages of equine digital flexor tendon healing (18, 53). Experimental murine models of systemic inflammation have illustrated that treatment with cyclooxygenase-2 (COX-2) selective NSAIDs diminishes endogenous resolution responses (37, 54). COX-2 selective NSAIDs are frequently used in the management of musculoskeletal pain and, in light of this, may also impede the healing response of inflamed soft tissues such as tendons. Instead of total inflammatory blockade, an alternative approach to moderating inflammation while simultaneously potentiating resolution may have therapeutic benefit for patients with tendon pathology. Low-dose aspirin is known to promote the release of stable aspirin-triggered isoforms of LXA4, which potentiate the resolution of inflammation (33, 43). Stable lipid mediators of aspirin-induced eicosanoid metabolism such as 15-epi LXA4 have proven efficacious in other chronic inflammatory diseases including in murine models of pulmonary inflammation (55) and in the treatment of human infantile eczema (56). In the current study, incubation of LPS-treated diseased tendon–derived stromal cells in 50 or 100 nM 15-epi LXA4 increased the expression of CD206, ALOX15, and CCL22 mRNA and reduced the expression of IL12B mRNA compared to LPS-stimulated tendon cells not exposed to 15-epi LXA4. It is possible that low-dose aspirin and its metabolites may act by triggering the expression of anti-inflammatory and proresolving mediators (43, 57). These stable aspirin-induced lipid mediators may hold promise for treating patients with tendinopathy by potentiating resolution of tendon inflammation. Limitations of our study include the relatively small number of patients studied and the single posttreatment tendon tissue biopsy. It was not ethically possible to obtain multiple sequential biopsies. Future studies using animal models of tendinopathy and randomized controlled human clinical trials with a placebo arm will be necessary to further investigate the therapeutic potential of inflammation resolving mediators in vivo.

In summary, we found that inflammation activation signatures displayed a high degree of plasticity throughout the stages of tendon disease. We have implicated the ALOX15 and CD206 pathways in the resolution of tendon pain in patients whose symptoms resolved after treatment. We suggest that the aspirin metabolite 15-epi LXA4 may contribute to resolving tendon inflammation.

MATERIALS AND METHODS

Study design

The objective of this study was to investigate inflammation activation pathways in samples of early-, intermediate-, and advanced-stage diseased tendons from patients with tendinopathy. Tissue samples were obtained using high-definition ultrasound guidance from well-phenotyped patient cohorts before and after treatment. We compared inflammation signatures in patients whose clinical symptoms resolved after surgical subacromial decompression treatment with those that remained persistently painful. Inflammation activation pathways were further studied in stromal cells derived from healthy and diseased human tendons in vitro. Statistical analysis and sample size justification were derived from previous studies (18, 53) that were sufficiently powered to gain insights from studying inflammation in samples of native tendon disease and using an in vitro model of inflammation. For histological studies, a single blinded investigator undertook image capturing. Further details of sample size/replicates are given in the figure legends.

Study approval

Ethical approval for this study was granted by the local research ethics committee, Oxfordshire REC B references 10/H0402/24, 09/H0605/111, and 10/H0606/60 and South Central Oxford B reference 14/SC/0222. Full informed consent according to the Declaration of Helsinki was obtained from all patients.

Tissue collection from patients with shoulder tendinopathy

All patients were recruited from orthopedic referral clinics where the structural integrity of the rotator cuff was determined ultrasonographically. Patients completed the Oxford Shoulder Score (OSS), a validated and widely used clinical outcome measure scoring from 0 (severe pathology) to 48 (normal function) (58). Samples of healthy and early-stage tendinopathic supraspinatus were collected via percutaneous ultrasound-guided biopsy under local anesthesia by a senior consultant shoulder surgeon. The biopsy specimen was taken using a trucut needle about 5 to 10 mm posterior to the anterior edge of the supraspinatus tendon. This validated biopsy technique is described in detail elsewhere (59). Biopsies of healthy supraspinatus tendons (n = 9) were collected from male or female patients (ages 20 to 30 years; mean, 23 ± 3.8 years) who underwent shoulder surgery for posttraumatic instability. These patients had intact nondegenerative supraspinatus tendons on ultrasound, which was confirmed at surgery. Healthy subscapularis tendons were also collected from three male or female patients undergoing shoulder surgery for posttraumatic instability (ages 61 to 77 years; mean, 66 ± 8 years).

Biopsies of early-stage pretreatment tendinopathic supraspinatus were taken from 21 male and female patients undergoing subacromial decompression surgery (painful pretreatment group). Biopsies were also taken from patients between 2 and 4 years after undergoing subacromial decompression surgery, in whom pain had resolved completely (pain-free after treatment, n = 6) or pain persisted (painful after treatment, n = 5). Patients in the pain-free group had significant pain before their surgery was performed as evidenced by a median Oxford Shoulder Score of 24 (range, 20 to 40) before surgery, and the biopsy was taken away from the surgical margins. All early-stage tendinopathic patients were of ages between 38 and 65 years (mean, 52 ± 8 years).

Intermediate- to advanced-stage diseased tendons (supraspinatus tears) were collected at the time of surgical debridement of the edges of the torn tendons from 11 male and female patients of ages between 50 and 78 years (mean, 61 ± 6.5 years). All patients were symptomatic and had full-thickness tendon tears. Tear sizes were classified as follows: small (≤1 cm), medium (>1 and ≤3 cm), large (>3 and ≤5 cm), and massive (>5 cm in anterior-posterior length) (60). Increasing tear size has been shown to correlate with the progression of pathology with reduced likelihood of repair with chronic disease (61). Torn tendons were collected under research ethics from the Oxford Musculoskeletal Biobank (09/H0606/11) or from patients participating in the United Kingdom Rotator Cuff Trial (UKUFF), a multicenter randomized controlled trial investigating the efficacy of open versus arthroscopic surgical repair for supraspinatus tendon tears (REC reference 10/H0402/24). Exclusion criteria for all patients in this study included previous shoulder surgery, other shoulder pathology, rheumatoid arthritis, and systemic inflammatory disease (7).

Tissue collection from patients with healthy hamstring tendons

Healthy hamstring tendons were collected from 10 male and female patients undergoing surgical reconstruction of their anterior cruciate ligament. All patients were of ages between 18 and 48 years (mean, 25.5 ± 11 years). Tissues were collected under research ethics from the Oxford Musculoskeletal Biobank (09/H0606/11). Hamstring tendons were immediately placed in Dulbecco’s minimum essential medium (DMEM)/F12 (Lonza) and processed in tissue culture to isolate tendon-derived stromal cells.

Processing of human shoulder tendon samples

Immunohistochemistry and immunofluorescence. Samples of healthy and diseased supraspinatus tendons were immersed in 10% buffered formalin and left for about 0.5 mm/hour. After fixation, tendons were processed using a Leica ASP300S tissue processor and embedded in paraffin wax. Tissues were sectioned at 4 μm using a rotary RM2135 microtome (Leica Microsystems Ltd.) onto adhesive glass slides and baked at 60°C for 30 min and 37°C for 60 min.

Gene expression. Samples of healthy subscapularis and diseased supraspinatus tendons were immediately snap-frozen in liquid nitrogen and stored at −80°C until RNA extraction.

Assessment of diseased supraspinatus tendons. Histological assessment of tendons collected from the study cohort was performed on hematoxylin and eosin–stained sections using the Bonar scoring system (0 to 12) that evaluates tissue structure (62). Healthy supraspinatus tendons exhibited a more normal tissue architecture (median, 2; interquartile range, 1 to 2) compared to early-stage (median, 7; interquartile range, 6 to 8) and intermediate- to advanced-stage diseased tendons (median, 10; interquartile range, 8.25 to 10).

Immunohistochemistry

For antigen retrieval, slides were baked at 60°C for 60 min, and tissue sections were obtained through deparaffinization and target retrieval steps (heat-mediated antigen retrieval at high pH) using an automated PT Link (Dako). Antibody staining was performed using the EnVision FLEX visualization system with an Autostainer Link 48 (Dako). Antibody binding was visualized using FLEX 3,3′-diaminobenzidine (DAB) substrate working solution and hematoxylin counterstain (Dako) as per protocols provided by the manufacturer. Details of antibodies and their working dilutions are shown in table S3. Positive controls consisted of sections of normal human spleen. For negative controls, the primary antibody was substituted for universal isotype control antibodies: cocktail of mouse immunoglobulin G (IgG1), IgG2a, IgG2b, IgG3, and IgM (Dako) and rabbit immunoglobulin fraction of serum from nonimmunized rabbits, solid-phase absorbed (Dako). Isotype control images are shown in fig. S5 (A and B). After staining, slides were taken through graded industrial methylated spirit and xylene and mounted in Pertex mounting medium (Histolab).

Image acquisition and quantitative analyses for immunohistochemistry

Images were acquired on an inverted microscope using Axiovision software (Zeiss). Twenty images (or until the tissue section was exhausted) were acquired in a systematic manner at ×100 magnification with oil immersion by a single blinded investigator. Image analysis was conducted using ImageJ (National Institutes of Health) as previously described (63). For each sample, immunopositive staining was normalized to the number of hematoxylin-counterstained nuclei within the field of view.

Immunofluorescence for colocalization of macrophage markers

After antigen retrieval steps, tissues were blocked in 5% normal goat serum (Sigma) in phosphate-buffered saline (PBS) for 60 min in a humid chamber at room temperature. After removal of blocking solution, sections were incubated with the primary antibody cocktail diluted in 5% normal goat serum in PBS for 2 hours at room temperature. Details of primary antibodies used for immunofluorescence are listed in table S3. After incubation with the primary antibody, sections were washed three times with PBS–Tween 20 (PBST) for 5 min. Slides were incubated in the secondary antibody cocktail, each diluted 1:200 in 5% normal equine serum (Sigma) in PBS for 2 hours, and shielded from light. The secondary antibodies were Alexa Fluor goat anti-mouse IgG2a or IgG2b or goat anti-rabbit IgG (Life Technologies) and goat anti-mouse IgG1 (Southern Biotech). After washing with PBST, sections were incubated in 2 μM POPO-1 nuclear counterstain (Life Technologies) diluted in PBS containing 0.05% saponin (Sigma) for 20 min. After washing with PBST, tissue autofluorescence was quenched with a solution of 0.1% Sudan Black B (Applichem) in 70% ethanol for 10 min (18). After washing, slides were mounted using fluorescent mounting medium (VectaShield), sealed, and stored at 4°C until image acquisition. For negative controls, the primary antibody was substituted for universal isotype control antibodies: cocktail of mouse IgG1, IgG2a, IgG2b, IgG3, and IgM (Dako) and rabbit immunoglobulin fraction of serum from nonimmunized rabbits, solid-phase absorbed (Dako). Isotype control images are shown in fig. S5 (C and D).

Immunofluorescence image acquisition

Immunofluorescence images were acquired on a Zeiss LSM 710 confocal microscope using a 40× oil immersion objective (numerical aperture, 0.95). The fluorophores POPO-1, Alexa Fluor 488, Alexa Fluor 568, and Alexa Fluor 633 were excited using the 405-, 488-, 561-, and 633-nm laser lines, respectively. To minimize bleed-through, all channels were acquired sequentially. Averaging was set to 2 and the pinhole was set to about 1 airy unit. Two-dimensional image reconstructions were created using ZEN 2009 (Zeiss).

Image analysis

Grayscale tiff files from confocal images were imported and analyzed in a custom MATLAB (MathWorks) script. Single cells were identified on the basis of the POPO-1 nuclear staining and converted to regions of interests. These were then dilated to encompass the cell to give new regions of interests. From the enlarged region, the average intensity of any surface marker staining in the green, red, and purple channels was measured. The average intensity of the antibody stain was then measured for each cell, with cells expressing low CD68 and/or CD206 staining being gated. The intensities of proteins of interest for these gated cells were then plotted. Isotype controls were used to set a threshold to distinguish between the negatively and positively stained cells. Colocalized cells were identified as having positive staining for two or more markers. To calculate the percentage of immunopositive cells for each surface marker, the number of cells exhibiting average intensities above the threshold was divided by the total number of POPO-1 nuclei.

Isolation of tendon-derived stromal cells from healthy hamstring and diseased supraspinatus tendons

Fresh samples of healthy hamstring or torn (intermediate-advanced stage) supraspinatus tendons were cut into 2-mm3 explants and incubated in DMEM/F12 (Lonza) containing 50% fetal calf serum (FCS; Labtech) and 1% penicillin-streptomycin (Lonza). Fresh media were replaced every 4 days, and cells were allowed to grow out from explants over time in a tissue culture incubator at 37°C and 5% CO2. Once cells were 90% confluent, explants were removed and media replaced with DMEM/F12 containing 10% FCS and 1% penicillin-streptomycin. Cells between passages 1 and 3 were used for all experiments.

Treatment of tendon-derived stromal cell–macrophage cocultures with cytokines

Tendon-derived stromal cells from healthy hamstring and diseased supraspinatus were seeded at a density of 15,000 cells per well in a 24-well plate. Tendon cells were allowed to reach 70% confluence before stimulation with cytokines or addition of monocytes. Human monocytes (98% CD14+, 13% CD16+) were obtained from healthy donor buffy coats by two-step gradient centrifugation as described in detail elsewhere (64). For coculture experiments, 150,000 monocytes per well were added and allowed to differentiate into macrophages for 2 days before cytokine stimulation. Tendon-derived stromal cells in isolation and tendon-derived stromal cell–macrophage cocultures were incubated in X-Vivo10 medium (Lonza) containing 1% heat-inactivated human serum (Sigma). Cells were treated as previously described (23) with either LPS (100 ng ml−1; Escherichia coli 055:B5, L2880, Sigma), IFNγ (20 ng ml−1; R&D Systems), or IL-13 (20 ng ml−1; BioLegend); nontreated (vehicle only) cells served as controls for each experiment. After treatment, cells were then incubated at 37°C and 5% CO2 until harvest of the cell lysate for mRNA isolation after 96 hours.

Modulating inflammation with 15-epi LXA4 in LPS-stimulated tendon-derived stromal cells

Tendon stromal cells were derived from intermediate-stage diseased supraspinatus tendons from a minimum of five patients. Cells were seeded at a density of 15,000 cells per well. Once cells were 70% confluent, they were preincubated in 50 or 100 nM 15-epi LXA4 (Cayman Chemical) for 24 hours in X-Vivo10 medium (Lonza) containing 1% heat-inactivated human serum (Sigma). Cells were subsequently treated with LPS (100 ng ml−1), as described earlier, in medium containing either 50 or 100 nM 15-epi LXA4 or vehicle-only control medium. After LPS treatment, cells were shielded from light and incubated at 37°C and 5% CO2 until harvest of the cell lysate for mRNA after 96 hours.

Extraction of RNA from human tendons

Tissues. Tendon samples (30 to 100 mg) were homogenized in 1 ml of RNA Bee (Ams Biotechnology) using an IKA Ultra Turrax T8 homogenizer (Fischer Scientific). Samples included healthy subscapularis (n = 3), early-stage tendinopathic supraspinatus (painful before treatment, n = 6; painful after treatment, n = 5; pain-free after treatment, n = 6), and intermediate- to advanced-stage diseased (torn) supraspinatus tendons (small to medium tears, n = 6; large to massive tears, n = 5). RNA extraction was performed according to the manufacturer’s protocol. RNA cleanup was subsequently performed using an RNeasy Mini Kit (74106, Qiagen) with an on-column DNA treatment using DNase I (EN0521, Thermo Scientific).

Cells. Cells were lysed in RLT lysis buffer, and RNA was isolated using an RNeasy Mini Kit (74106, Qiagen) with an on-column DNA treatment using DNase I (EN0521, Thermo Scientific). The quality of extracted and cleaned RNA from tissues and cells was determined using a NanoDrop 1000 spectrophotometer (Thermo Scientific). The ratio of absorbance of 260/280 nm was used to determine RNA quality, with all samples achieving a minimum of 1.80.

Complementary DNA synthesis and quantitative polymerase chain reaction. RNA (250 ng) was reverse-transcribed into complementary DNA (cDNA) using a High Capacity Reverse Transcription Kit (4368813, Applied Biosystems). cDNA was diluted to 2.5 ng/μl with ribonuclease-free water, and 1 μl was used in a 10-μl quantitative polymerase chain reaction (qPCR) volume with Fast SYBR Green Master Mix (4385612, Applied Biosystems) and primers diluted to 200 nM (Eurofins Genomics) in 384-well plates (4309849, Life Technologies). Gene signatures consisted of a panel of genes for IFN, NF-κB, STAT-6, and glucocorticoid receptor activation pathways. The primers for each gene are shown in table S1. The reaction efficiency was calculated by measuring the Ct values for both sets of genes in a cDNA mix dilution series and applying the following formula: Efficiency = 10(−1/slope) − 1 as previously described (64). Duplicate reactions for each gene were run on a ViiA7 qPCR machine (Applied Biosystems), and results were calculated using the ΔΔCt method using reference genes for human β-actin, GAPDH (glyceraldehyde-3-phosphate dehydrogenase), and β2-macroglobulin. Results were consistent using these three reference genes, and data are shown normalized to β-actin.

Statistical analyses

Statistical analyses were performed using GraphPad Prism 6 (GraphPad Software). Normality was tested using the Shapiro-Wilk normality test. Kruskal-Wallis tests followed by pairwise post hoc Mann-Whitney U tests were used to compare the expression of markers of monocytes, macrophages, and proresolving proteins in sections of healthy and diseased human supraspinatus tendons. Pairwise Mann-Whitney U tests were used to test for differences in gene expression in tendon samples from pain-free patients and patients experiencing pain before and after treatment. Pairwise Mann-Whitney U tests were used to test for differences between the expression of NF-κB, IFN, and STAT-6 target genes in cytokine-treated stromal cells derived from healthy and diseased tendons. Pairwise Mann-Whitney U tests were used to test for differences in gene expression between LPS-stimulated tendon-derived stromal cells treated with 15-epi LXA4 compared to LPS-stimulated cells not exposed to 15-epi LXA4. P < 0.05 was considered statistically significant.

SUPPLEMENTARY MATERIALS

www.sciencetranslationalmedicine.org/cgi/content/full/7/311/311ra173/DC1

Fig. S1. NF-κB–p65 staining in diseased human supraspinatus tendons.

Fig. S2. Phosphorylated STAT-6 staining in diseased human supraspinatus tendons.

Fig. S3. Gene expression signatures after treating tendon-derived stromal cell–macrophage cocultures with IFNγ, LPS, or IL-13 for 96 hours.

Fig. S4. Expression of IFN activation pathway genes in biopsies from patients with early-stage tendon disease before and after surgical subacromial decompression treatment.

Fig. S5. Isotype control staining of diseased human supraspinatus tendons.

Table S1. Primer sequences and validation for qPCR.

Table S2. Summary of inflammation activation pathway gene signatures in samples of early-, intermediate-, and advanced-stage diseased human supraspinatus tendons.

Table S3. Primary antibodies used for immunohistochemistry and immunofluorescence.

REFERENCES AND NOTES

  1. Acknowledgments: We thank S. Gordon and J. Sherlock for their review of the manuscript. Funding: S.G.D. is funded by Arthritis Research UK grant 20506. Research in our laboratories is supported through funding from Arthritis Research UK (program grant number 20522), the National Institute for Health Research (NIHR) Oxford Biomedical Research Unit, and the Rosetrees Trust. The Structural Genomics Consortium is a registered charity (number 1097737) that receives funds from Abbvie, Bayer HealthCare, Boehringer Ingelheim, the Canadian Institutes for Health Research, the Canadian Foundation for Innovation, Eli Lilly and Company, Genome Canada, GlaxoSmithKline, the Ontario Ministry of Economic Development and Innovation, Janssen, the Novartis Research Foundation, Pfizer, Takeda, and the Wellcome Trust. Author contributions: S.G.D. performed all experiments and wrote the manuscript with input from all coauthors. S.G.D., A.J.C., and F.O.M. designed the study. F.O.M., U.O., and G.W. provided qPCR reagents, and F.O.M. and U.O. performed array analysis. F.O.M. and G.W. provided human macrophages for coculture experiments. C.Y. facilitated confocal image acquisition. B.J.F.D., K.W., B.W., L.R., R.D.J.S., and A.J.C. facilitated procurement and collection of healthy and diseased shoulder tendons from patients. Competing interests: The authors declare that they have no competing interests. Data and materials availability: Primer sequences and validation data are shown in table S1.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article