Combined ascorbic acid and T₃ produce better healing compared to bone marrow mesenchymal stem cells in an Achilles tendon injury rat model: a proof of concept study

rBMSCs were isolated from the bone marrow of the tibia and femur of three 6–8-week-old male inbred Lewis rats (ENVIGO, Wyton, UK), as previously described [26]. The bone marrow cells were seeded into a tissue culture flask in MesenCult Basal Medium (STEMCELL Technologies, Vancouver, Canada), supplemented with penicillin/streptomycin (100 U/mL–100 μg/mL) and 10% FBS, and incubated at 37 °C, in a 5% CO₂ atmosphere with 95% relative humidity. Three days after seeding, floating cells were removed, and the medium was replaced with a fresh medium. Adherent cells were allowed to reach approximately 80% confluence (12–17 days for the first passage). Cells were detached following treatment with StemPro accutase (ThermoFischer Scientific, USA) and replaced every 6–8 days at approximately 80% confluence. Cells at second passage (P2) were used unless otherwise stated. When the cells in the culture flasks were nearly confluent, the cells in the primary culture were sub-cultured and collected for the study. In addition, mesenchymal stem cells (MSCs) were tested by flow cytometry (Becton Dickinson, Franklin Lakes, New Jersey, USA) using specific surface markers, being negative for CD45 and positive for CD90 [27]. Debris particles were excluded from the analysis by gating on forward and side scatter (FSC and SSC) morphological parameters. Dead cells were gated out by propidium iodide (0.07 μg/mL, Sigma-Aldrich, Saint Louis, USA) staining; 10.000 viable and non-debris events in the morphological gate were recorded for each sample. Cells were acquired on a CytoFLEX flow Cytometer (Beckman Dickinson, CA, USA). The analysis was performed using CytExpert Software (Beckman Dickinson). The stem cell characteristics of bone MSCs were routinely confirmed by their trilineage differentiation abilities (adipocytes, osteoblasts, and chondrocytes) when cultured under appropriate conditions in vitro before being used for experimental procedures as previously described [27, 28]. Adipocytes were visualized by Oil Red-O staining, osteoblasts evidenced by alizarin red, and chondrocytes were visualized with Alcian Blue. Isolated rBMSCs (P0) were cryopreserved according to standard procedures in liquid nitrogen in 90% fetal bovine serum + 10% DMSO and stored in liquid nitrogen in vials until transplantation.

Twenty-four 6–8-week-old male inbred Lewis rats (ENVIGO) were housed under controlled conditions in the Interdepartmental Service Centre - Station for Animal Technology, University of Rome “Tor Vergata” (Italy), supplied with standard diets (4RF18; Mucedola srl, Italy) and water ad libitum. Under general anesthesia induced using an intramuscular injection of tiletamine/zolazepam (50 mg/kg) (Zoletil 100, Virbac Italia SRL, Italy) and xylazine (10 mg/kg) (Rompun Bayer AG, Germany), the right Achilles tendon was laid open by a skin incision approximately 10-mm long on the medial aspect of the right hind limb of each rat. In each animal, an Achilles tendon defect 2 mm in diameter was produced as previously described [29] with some modifications. Briefly, the Achilles tendon was separated from the surrounding tissue, and the medial, more prominent component of the Achilles tendon, the m. flexor digitorum superficialis, was isolated. Using a sterile punch (diameter 2 mm), which was placed around the Achilles tendon at a distance of approximately 2 mm from the calcaneus, a full-thickness hole, 2 mm in diameter, was made. The skin was closed by a continuous suture using absorbable 4.0 absorbable monofilament sutures. The animals were divided into eight groups (with three rats per group). The injured tendon was filled with 50 μL of phosphate-buffered saline (PBS), a physiological buffer solution which contained (1) 50 μg/mL AA (AA group), (2) 10⁻⁷ M T₃ (T₃ group), (3) 4 × 10⁶ rBMSCs (rBMSC group), (4) 50 μg/mL AA + 10⁻⁷ M T₃ (AA + T₃ group), (5) 4 × 10⁶ rBMSCs + 50 μg/mL AA (rBMSC + AA group), (6) 4 × 10⁶ rBMSCs + 10⁻⁷ M T₃ (rBMSC + T₃ group), (7) 4 × 10⁶ rBMSCS + 50 μg/mL AA + 10⁻⁷ M T₃ (rBMSC + AA + T₃ group), and (8) PBS only (control group: CTRL). Additionally, AA was injected again on the second and fourth day following the initial injection (for groups 1, 4, 5, and 7), while T₃ was injected again every day for 4 days (for groups 2, 4, 6, and 7). Post-operatively, antibiotics and analgesics were administered: 0.5 mL/kg Baytril (Bayer AG, Germany) and 0.1 mL/kg/day Rimadyl (Veterfarma SpA, Italy). After 30 days, the animals were anesthetized and then euthanized using CO₂ in specially designed chambers. Tendons were explanted, embedded in optimal cutting temperature (O.C.T.) compound medium (Sakura Finetek USA, Inc., Torrance, CA, USA), and quickly frozen in liquid nitrogen-cooled isopentane for sectioning at a thickness of 12 μm using a Leica (Leica Biosystems, Wetzlar, Germany) cryostat, and then stained with hematoxylin and eosin or Picro-Sirius red solution for histological and histomorphometric analyses and examined under white light and polarized light microscopy, respectively. Sections were examined independently by two pathologists in a blinded, random order.

For histological and histomorphometric analyses, three sections per animal and three animals per group were evaluated for each experimental group. Three sections of each specimen were placed onto one slide at an approximate distance of 250 μm. Six areas were selected from each section to cover the entire surface of the tendon and were collected using an optical microscope (BX51, Olympus Italia Srl, Milano, Italy) at × 10 magnification, coupled to Aperio ScanScope (Aperio ScanScope CS, Aperio Technologies, Leica Biosystems, USA) digital image analysis system. A modified, semi-quantitative score obtained by the three different scores of Soslowsky, Svensson, and Cook’s scoring systems was used (Table 1) [30‐32]. These scores take into account four different parameters: fiber structure, cellularity, vascularity, and cartilage formation. Each parameter was scored on a 4-point scale of 0–3, as follows: 0, normal; 1, slightly abnormal; 2, abnormal; and 3, markedly abnormal. The samples were scored according to whether there was a significant abnormal appearance. The total score is the sum of the values of each parameter varied between 0 (normal tendon) and 12 (severest abnormality) [33]. For the validation study, pathologists were additionally asked to make overall examinations for each specimen. The specimens were anonymized and randomized by an IRCCS Rizzoli Orthopedic Institute employee not involved in the study. For polarization microscopy, each section (three sections per animal, three animals per group) were imaged using × 8 magnification. To prevent variations in illumination and magnification, all pictures were recorded by the same person. The sections underwent further evaluation using computerized image analysis software (Aperio Area Quantification FL Algorithm). Color intensity and color amount (i.e., positively stained area) were measured by Aperio ScanScope CS (Aperio Area Quantification FL Algorithm, Aperio ScanScope CS Leica Biosystems). Collagen type I and type III percentages were calculated by these automated measurement of the red-orange fiber area (collagen type I) and pale green fiber area (collagen type III) from Picro-Sirius red staining [34].

Once they had been isolated and cultured, the rBMSCs were analyzed for immunophenotype by flow cytometry, which confirmed the expression of CD90 and the lack of CD45 (Fig. 1a). In addition, rBMSCs were functionally characterized by differentiation into adipocytes, osteoblast, and chondrocytes (Fig. 1b–d), displaying the typical characteristics of MSCs [35].

Surgical procedures and treatments (Fig. 2) did not determine any additional accidental damage and no clinical signs of necrosis or tissue infection. At gross examination post mortem, scar tissue was visible at the site of the core lesion in the center of each tendon. Macroscopically, no differences were visible between the groups. No macroscopic signs of necrosis or tendon degeneration were observed. The paratenon covering the tendon was of normal thickness, as were the collagen fibers of the epitenon, and no inflammatory infiltrate or other morphological changes were seen. Intra-class correlation for histology scores was excellent for inter-observer reliability (0.79–0.97 depending on score/sub-score). The lesions could be identified clearly on H&E staining. Histological analysis indicated incomplete restoration of structural integrity (Fig. 3).

Semi-quantitative histomorphometric results are reported in Fig. 4. The total histological score was lower in the AA + T₃-treated group (score 1) than in all the other group treatments (AA score 7; T₃ score 5.3; rBMSC score 4.16; rBMSC + AA score 7.16; rBMSC + T₃ score 7.16; rBMSC + T₃ + AA score 9.66) including the CTRL group (score 2.3) (Fig. 4a). Specifically, in the group treated with AA + T₃, score values were lower for fiber structure (close to regular orientation, score 0.5) than those in the CTRL group (0.7); vasculature (vessels run inconspicuous coursing parallel to the collagen fiber bundles in the septa, like in a normal tendon, score 0) than those in the CTRL group (score 0.8); and cartilage formation (slight cartilage formation was only observed in one section, score 0.1) than those in the CTRL group (slight formation of cartilage in three sections for one of the three animals, and in two sections for the second of the three animals, score 0.5). The rBMSC group had the same histology score as the CTRL group for cartilage formation (slight cartilage formation was observed in two of the three sections for the first animal and in one section only for the other two animals in the group) (Fig. 4b). The AA + T₃ group total scores were significantly lower than those of the rBMSC + T₃ group (p < 0.005) and the rBMSC + AA + T₃ group (p < 0.0001). The rBMSC + AA + T₃ group scored significantly higher than the CTRL group. Specifically, the AA + T₃ group sub-scores were lower for fiber structure in comparison with the AA group, the rBMSC + AA group, the rBMSC + T₃ group, and the rBMSC + AA + T₃ group (p < 0.005); for cellularity scores, the AA + T₃ group sub-scores were lower than for all other groups and, most significantly, than for the rBMSC + AA group, the rBMSC + T₃ group, and the rBMSC + AA + T₃ group (p < 0.005); for vascularity, the AA + T₃ group sub-scores were significantly lower than for the AA group and the rBMSC + AA + T₃ group (p < 0.005), and the AA + T₃ group cartilage formation sub-scores were lower than all the other group treatments (Fig. 4b). In the AA + T₃ group, the vascularity sub-score value was similar to that of a healthy tendon (Fig. 4b), and the value was significantly lower than that for the AA + T₃ group and for the rBMSC + AA + T₃ group (p < 0.005) (Fig. 4b).

Analysis of the collagen ratio revealed that the association AA + T₃ exerted greater effect on tendon healing than all the other treatments. Specifically, the collagen type I value was significantly higher than in the rBMSC + AA group and the rBMSC + AA + T₃ group (p < 0.05; p < 0.01); the AA + T₃ group collagen type III value was significantly lower than in the rBMSC + AA group and the rBMSC + AA + T₃ group (p < 0.05; p < 0.01) (Fig. 5). The collagen type I value of the rBMSC + AA + T₃ group was significantly lower than the CTRL (p < 0.05), and the collagen type III value was significantly higher in the rBMSC+AA+T₃ group than the CTRL group (p < 0.05) (Fig. 5).