3-D ultrastructure and collagen composition of healthy and overloaded human tendon: evidence of tenocyte and matrix buckling.

Participants

Six patients (age: 48 ± 6 years, mean and SE) suffering from chronic Achilles tendinopathy with focal mid-tendon pain, tendon thickening and ultrasonography-verified structural changes were recruited for this study. The recruitment of patients and isolation of biopsy material was approved by the regional ethical committee for The Capital Region of Copenhagen (H-1-2009-114). The subjects were either recreational athletes or manual workers with a history of Achilles tendon pain for more than 6 months (range 0.5–3 years), and they had all tried conventional treatments [i.e. eccentric strength training, oral anti-inflammatory drugs (non-steroidal anti-inflammatory drug, NSAID), peri-tendinous glucocorticoid injection] for tendinopathy without effect. Glucocorticoid injection or NSAID intake was not allowed for 6 months prior to the study. Subjects were recruited from the Department of Rheumatology, Silkeborg Hospital, Denmark.

Tissue collection

Biopsies of Achilles tendon were obtained as a standard routine diagnostic procedure in order to detect deposits of cholesterol, uric acid and amyloid; none was detected in any of the individuals examined. The subjects had two tendon biopsies taken in the diseased Achilles tendon, one where the focal tendinopathy changes and symptoms were present, and one biopsy in a presumably healthy area of the same tendon. Excess material from the tendon biopsies was used for this study, with patient consent. There was sufficient material from all subjects for protein analysis and from two subjects for electron microscopy investigation. In brief, 2 × 2 × 2 mm biopsies were obtained under local anesthesia applied peri-tendinously from both the medial and lateral side of the tendon with ultrasound-guided injection of 10 mL 1% lidocaine on both sides. The biopsies were obtained under ultrasonographic guidance using a semi-automatic biopsy needle (14 GA, 9 cm; Angiotech). The initial biopsy was obtained in the tendon area with the most increased tendon thickness, hypoeccogenicity and potential neovascularization located typically 3–5 cm above the calcaneic insertion of the tendon. The second biopsy was obtained 4 cm proximal to the first biopsy in an ultrasonographic normal region of the tendon.

Electron microscopy

Biopsy samples were prepared for electron microscopy as described previously (Starborg et al. 2013 ). In brief, 1 × 1 × 1 mm cubes of tendon were immersed in 2.5% glutaraldehyde prepared in 100 m m cacodylate buffer (pH 7.2), and processed using a double osmium protocol that is suitable for transmission electron microscopy and SBF-SEM (Starborg et al. 2013 ). Semi-thin (˜ 1 μm thick) sections were prepared, stained with toluidine blue and examined using a dissecting microscope to determine the orientation of the tendon within the biopsies. The resin blocks were trimmed for transverse sectioning (i.e. 90° to the tendon long axis). Ultrathin sections (70 nm thick) were prepared, and fibril diameter measurements were made using a FEI BioTwin transmission electron microscope. Resin blocks were trimmed and images were collected using an FEI Quanta 250 ESEM equipped with a Gatan 3View® for in-chamber ultramicrotome sectioning and image acquisition, as described previously (Starborg et al. 2013 ). Typically, 500–1000 × 100-nm-thick cuts were removed from the blocks during the imaging procedure, and image analysis and model reconstruction was performed using IMOD (Kremer et al. 1996 ).

Protein analysis

Tissue samples for protein analysis were snap-frozen in liquid nitrogen and stored at −80 °C prior to analysis. Collagen was extracted by pepsin, and run based on equal dry weight loads of starting tissue on 6% sodium dodecyl sulfate–polyacrylamide gel electrophoresis using an interrupted electrophoresis method that resolves type III collagen chains (Wu et al. 2010 , adapted to the method of Laemmli et al. 1970 ). Individual protein bands after Coomassie Blue staining were digested in-gel by trypsin (Kinter & Sherman, 2008 ; Eyre et al. 2008 ). The resulting peptides were subjected to microbore C8 column liquid chromatography (0.3 mm × 15 cm; Vydac) interfaced directly to a ThermoFinnigan LCQ Deca XP tandem mass spectrometer equipped with an electrospray ionization source. For protein identification, peptide fragments were compared with the NCBI non-redundant protein database using SEQUEST, an automated database search algorithm designed for use with tandem mass spectrometry (MS/MS) data.

Supporting Information Additional Supporting Information may be found in the online version of this article: Video S1 Step-through movie generated from 550 images of SBF-SEM analysis of healthy tendon. Video S2 Step-through movie generated from 550 images of SBF-SEM analysis of healthy tendon. Video S3 Three-dimensional reconstruction of healthy tendon showing the nuclei of adjacent cells in head-to-tail alignment. Video S4 Three-dimensional reconstruction of tendinopathic tendon showing the nuclei of adjacent cells. Disorganized cells are a feature of tendinopathic tendon.