Stem cell application in rotator cuff repair: Interposition stem cell sheet versus overlaid stem cell sheet

Stem cells are an effective method of biologic healing and can be used to enhance the natural enthesis of the tendon-to-bone junction in rotator cuff repair. The purpose of this study was to investigate if the application of engineered stem cell sheets using adipose-derived cells (ADSCs) was effective in regeneration of natural enthesis and if there was a difference in the result of repair depending on the applied location


Introduction
The prevalence of the musculoskeletal disorders is increasing with the increase in the elderly population and participation in sports activities. These disorders not only cause physical pain but also have a considerable impact on social costs. In particular, rotator cuff tears, which can cause functional disorders in the shoulder, are common among tendon diseases. Despite the continued development of surgical treatment, there is still a high risk of re-rupture, which remains a clinically important issue (1,2). The tendon is an elastic tissue composed of collagen that adheres to the bone, maintains movement, and transmits power. As such, the success of suture after rotator cuff tear depends on the recovery of the normal connection between the tendon tissue and the bone. Following rupture, the lesions are limited in their self-regeneration ability due to the relatively low cell density, distribution of blood vessels, and microenvironment changes. For this reason, surgical treatment tends to be followed by slow regeneration and incomplete recovery of the biological properties and original functions (3)(4)(5). To prevent re-tear after rotator cuff repair, many studies have reported on surgical techniques, xation methods, and implant advancement to improve structural properties (6)(7)(8)(9). Nevertheless, high re-rupture rates are still reported in large-to-massive tears, and a number of biological studies have been conducted to reduce the re-tear rate (10,11).
Recently, as one of the methods for biologic healing enhancement, stem cell-based treatment has attracted attention as a means to improve the regeneration potential of tendon tissue (12)(13)(14)(15)(16). In particular, mesenchymal stem cells (MSCs) are widely used in tissue engineering because they are easy to separate and culture from various tissues. Indeed MSCs have been trialed variably as effective treatment methods for cell-based tendon regeneration research. However, a clear effect has not been demonstrated so far, and the success at the cellular level is limited for application to the preclinical or clinical stage. In terms of delivery, stem cells can be applied to lesions in various ways, such as injecting cells or injecting the culture medium of the cells to induce tissue regeneration (13,15,17). However, this method is di cult to apply in a speci c area and maintain for a long time. Moreover, it can also cause adhesion to surrounding tissues and the formation of scar tissue or ectopic bone, which eventually leads to poor quality of the damaged tendon tissue. However, by engineering stem cells in the form of a sheet, the stem cells can fairly easily adhere to the tissue and continuously function at the limited application site (18,19). The purpose of this study was to determine the most effective delivery system for applying stem cells for biologic augmentation in rotator cuff repair. Among the different systems, we attempted to determine whether overlaying the cell sheet on the repair site or interposition to the tendon-to-bone junction was more effective for tendon-to-bone regeneration.

Study design
For the recovery of the tendon-to-bone junction, this experiment was divided into three groups to compare healing augmentation using stem cell sheets (Fig. 1). Four Sprague-Dawley (SD) rats were used for each group, and 8 samples per group were processed with the same treatment on both shoulders per rat. The chronic rotator cuff tear models were made identical in all groups as a suitable animal model for this experiment. In order to exclude the possibility of spontaneous healing from the rotator cuff tear in the rat, an initial tear was made and maintained for 2 weeks to create a model of chronic cuff tear (20,21), before subsequently repairing. As a control, surgical repair was performed alone (Repair-only group), whereas the experimental group was divided into two groups according to the location of the stem cell sheet transplantation. In the rst experimental group, after surgical repair, the stem cell sheet was overlaid on the repair site so that it covered parts of both the humerus and the tendon (Overlay group, 8 shoulders). In the second group, the stem cell sheet was rst transplanted between the tendon and bone (greater tuberosity of humerus) in chronic tears, and then surgically repaired (Interposition group, 8 shoulders). After 2 weeks of repair, the repair quality was analyzed following sacri ce of all animals.
Isolation and culture of primary ADSCs in SD rat inguinal fats Rat ADSCs were isolated from inguinal fat tissue of SD rats aged 8 weeks (Orient Bio, South Korea). In order to track the cell location after transplantation, the GFP-expressing ADSCs were isolated from SD-GFP-transgenic rats (JapanSLC, Hamamatsu, JAPAN) and used to produce cell sheets. The ADSC isolation was performed using a previously published method (18). The inguinal fat tissue harvested from the rats was digested by 0.1% (w/v) collagenase type I (Worthington, USA) dissolved in warmphosphate buffered saline (PBS, Welgene) for 1 h at 37 °C. The isolated rat ADSCs were cultured in Dulbecco's Modi ed Eagle Medium (DMEM) -low glucose medium (GIBCO, ThermoFisher Scienti c, USA) containing 10% (v/v) fetal bovine serum (FBS) and 1% antibiotic-antimycotic (A-A) solution in an incubator (37 °C, 5% CO ). After incubation for 24 h, the medium was changed to fresh culture medium.
The cells were cultured up to passage 3.

Characterization of rat ADSCs
Flow cytometry analysis using surface makers To analyze MSCs surface markers, the rat ADSCs were seeded at passage 3 in 6-well plates (Nunc, Denmark). After 24 h, the cells were incubated with antibodies for 1 h at 4 °C in darkness and washed twice in PBS. Flow cytometry (Cando, BD Bioscience, USA) analysis was performed, and the following antibodies were used: PE-isotype, PE-CD29, PE-CD31, PE-CD45, FITC-isotype, FITC-CD73 (BD, USA), FITC-CD90 and MHC class I (AbD serotec, BIO-RAD, CA, USA).

Con rmation of differentiation potential
To con rm the adipogenic and osteogenic differentiation ability of the rat ADSCs, the cells were seeded in 6-well plates and incubated until there were approximately 5 × 10 5 cells/well. For adipogenic differentiation, the cells were incubated in adipogenesis medium (Gibco, Life Technologies, USA) for 2 weeks. The cells were then stained with Oil red O solution (Sigma-Aldrich, USA) for 5 min. For osteogenic differentiation, the cells were incubated in DMEM-low glucose medium containing 1% A-A, 10 nM dexamethasone, 10 mM β-glycerophosphate, and 50 µM ascorbic acid. After 4 weeks, the cells were stained with 2% (w/v) Arizarin Red solution (Sigma-Aldrich, USA) for 30 min.

Quanti cation of genes by RT-qPCR
The rat ADSCs were seeded at 2 × 10 5 cells/well in 6-well plates for 24 h. The cells were incubated with GDF-7 in the same conditions as those used for immunocytochemistry. After 2 weeks, the treated cells were harvested and total RNA was extracted using TRIzol reagent (ThermoFisher Scienti c, USA). RT-qPCR was performed using SYBR™ Green PCR Master Mix (applied biosystems, UK). Primers were prepared using sequences veri ed in previously published papers (22,23); the sequences are summarized in Table 1. GAPDH was used for normalization. The gene expression levels were measured using an ABI prism 7900HT (applied biosystems) PCR machine, and the relative gene expression levels were quanti ed using 2 −ΔΔCt values. The above sequences were identified in the Primer-BLAST-NCBI.

Cell sheet fabrication
The rat ADSCs were seeded at passage 3 at 1.2 × 10 6 cells/dish in a temperature-responsive dish (35 mm, ThermoFisher Scienti c, USA), and incubated for 24 h at 37 °C. Then, the cells were removed from the dish by changing to a lower temperature of 20 °C (Fig. 2A). The removed cells shrunk to form a small round sheet (19).

Animal model
Chronic rotator cuff tear model Eight-week-old male SD rats with a weight of 250 g were used for this study. Prior to surgical treatment, 0.3 cc of a mixed solution with 50 mg/kg of Tiletamine/zolazepam (Virbac, Carros, France) and the supraspinatus tendon (SSP) attached to the humerus head was cut as close to the footprint as possible. A plastic drain was used to block the spontaneous healing in both the torn tendon stump and the humeral greater tuberosity. In particular, a stump on the torn tendon side was sealed with the plastic drain using 5 − 0 Ethibond (Somerville, USA) to block spontaneous healing (Fig. 2B). 5 − 0 Vicryl and 4 − 0 Prolene (Somerville, USA) were used for muscle and skin suture, respectively. After making a tear on the supraspinatus tendon, the rats were maintained with free cage activity for 2 weeks. Thus, the rat model of the chronic tear was completed (20,21). For the prevention of infection and pain relief, 50 mg/kg of ampicillin and 3 mg/kg of ketorolac were IM injected every day for 3 days after the tear model was created.
Rotator cuff repair and cell sheet application All plastic drains were removed 2 weeks after the tear was made. The scar tissues around the tear were removed as much as possible and repair was performed. In the control group, repair-only, the greater tuberosity of humerus head was drilled with a 0.5 mm drill, and the isolated SSP was transosseous suture repaired using 5 − 0 Ethibond (24,25). In the interposition group of the experimental group, a drill hole of greater tuberosity was made identically, 5 − 0 Ethibond was passed through to prepare a repair, and a stem cell sheet was applied. The prepared rat ADSC sheet was transferred on a tendon footprint of greater tuberosity using a thin membrane shifter (ThermoFisher, USA). In order to stably attach the sheet to the tissue, only the shifter was removed carefully after waiting for approximately 1 min (Fig. 3).

Evaluation
Sampling Two weeks after repair, all animals in this study were sacri ced using carbon dioxide gas. Based on the tendon-to-bone junction on both shoulders, tissues were completely removed, with the exception of the humerus head and tendon part.

Histological estimation
We obtained 4-µm thick tissue sections through the following series of processes: Formalin xation, para n embedding, depara nization, and dehydration with changed alcohol concentration. Immunohistochemistry staining was carried out using a GFP-antibody (Abcam, UK). The primary antibody was diluted at a 1:500 ratio. The procedures requiring DAB staining were performed by the pathology department. The sectioned tissues were stained with Hematoxylin and Eosin (H&E) solution (Sigma-Aldrich, USA). To con rm of the presence of collagen bers, the sectioned tissues were stained using Masson trichrome (MT) kit and Safranin O (SA) kit (Polysciences, Inc., USA). The images were taken using an EVOS microscope (ThermoFisher Scienti c, USA).
The brocartilage region was analyzed by H&E, Masson, and SA staining using four criteria: bone, tendon, mineralized brocartilage, and vascularity ( Table 2). The results were scored by two independent observers and nally evaluated as the average value (25,26). In particular, a sample stained with safranin O was observed under a microscope to evaluate the brocartilage formed by repair after tear was made. The brocartilage formation area was quantitatively analyzed using the ImageJ software program (National Institutes of Health, MD) based on the surgical site (27).

Biomechanical comparisons
A biomechanical test was performed using a machine capable of measuring the uniaxial tensile stress test (ST-1001; SALT, Korea). The proximal humerus side was xed to the distal part of the machine using sandpaper, and the tendon side was xed to the proximal part of the machine using a clamp. A gradient tension of 1 mm/min was applied to measure the load to failure and stiffness (28,29).

Statistical analysis
Statistical analysis was performed using GraphPad Prism 5 (GraphPad Software, CA, USA). For comparison between groups, one-way analysis of variance (ANOVA) and post-hoc analysis were performed, and errors were corrected using Tukey's method. The data were presented as mean and standard deviation (SD). P < 0.05 was considered as statistically signi cant, and P < 0.05 was marked as * and P < 0.001 as ** on the graph.

Properties of rat ADSCs
Cell morphology, MSC-speci c marker, and differentiation function were used to determine whether the stem cells isolated from 8-week-old SD rats can be de ned as MSCs. Firstly, as a result of the cell morphology of rat ADSC sheet made using rat ADSCs cultured at passage 3, it was con rmed through a microscope that the cells were attached and appeared in the broblast-like morphology (Fig. 4A, left). In addition, the rat ADSC-S were cultured to form a single layer in a round circle (Fig. 4A, right). Next, the rat ADSCs were found to be 99.9%, 100%, and 99.9% positive for the MSC-speci c surface markers CD90, CD29, and CD73, respectively. Moreover, ADSCs did not express CD45, CD31 (hematopoietic stem cell [HSC]), and MHC class I markers (Fig. 4B). To con rm osteogenic and adipogenic differentiation, ADSCs were cultured in differentiation medium and stained with oil red O and alizarin red. As a result, oil formation and calcium accumulation were con rmed (Fig. 4C). Thus, based on these results, it was found that rat ADSCs used in this study have potential activity as MSC.

Locations of transplanted rat ADSC sheet in vivo
To con rm that the transplanted rat ADSC sheet was located and remained at the applied site, GFPuorescence was detected in the cell sheet made from the rat ADSCs isolated from GFP-transgenic rats (Fig. 5A). When the cell sheet remained and GFP was overexpressed, it was stained with a brown dot and its location was con rmed. In the case of overlaid transplantation of cell sheets on repaired tendon-tobone junctions, it was con rmed that the brown spots were distributed outside the junction. On the other hand, when the sheet was transplanted in the tendon and bone interposition, uorescence could be con rmed inside the junction. It was found that rat ADSC sheet was still present at the applied site 2 weeks after transplantation and could affect healing.

Effects of rat ADSC sheet on attachment locations
Gross morphology and histology After 2 weeks of repair, it was con rmed that the group transplanted with the rat ADSC sheet overlay had a higher amount of tissue in the region above the junction compared to the repair-only group. Furthermore, the group transplanted with interposition demonstrated a hardened junction region and a relatively thickened SSP.
On histologic evaluation, the total score was signi cantly higher in the interposition group (5.17 ± 1.14) than in the repair group (3.21 ± 1.42) and overlay group (3.29 ± 1.87). No bone formation or structural differences were observed in any of the three groups, and there was no statistically signi cant difference in tendon and blood vessel formation. However, the tendon-to-bone junction showed a signi cant difference (P < 0.05) in the interposition group compared to the other two groups. A considerable amount of brocartilage was formed, and the boundary at the mineralized brocartilage became clear in the two groups transplanted with the stem cell sheet. In particular, with Masson staining, it was found that the interposition transplanted group had a broader and more clearly formed collagen formation range and direction than the overlaid group (Fig. 5B, C). This was con rmed by quantifying the area of the redstained glycosaminoglycan region in safranin O staining and showing statistically signi cant differences between the groups (Fig. 5D).

Biomechanical test
The maximum load to failures were signi cantly different between the three groups, especially the repair group and the overlay group (12.51 ± 5.19 N and 17.02 ± 5.32 N), and the repair group and the interposition group (12.51 ± 5.19 N and 19.20 ± 5.63 N). In terms of stiffness, the repair group (5.00 ± 1.07 N/mm) was signi cantly different compared to the interposition group (9.89 ± 0.83 N/mm). However, in the overlay group (7.06 ± 1.80 N/mm), the numerical value was increased, but there was no statistically signi cant difference (Fig. 6).

Discussion
The results of this study demonstrated that engineered stem cell sheets are an effective way to allow MSCs to act in su cient quantities for a relatively long period of time. In addition, in the application method, it was shown that the interposition of the tendon-bone interface was more effective in the regeneration of the tendon-to-bone junction than the overlay.
It is known that scar tissue healing occurs mostly in the tendon-to-bone junctions, even in situations when rotator cuff tearing is performed. This is considered to explain the reason for early stage re-ruption of suture surgery. Therefore, many studies have attempted to overcome the limitation of scar tissue healing and recovery native enthesis. Among these, stem cells are getting attention on account of their paragenic effects, as well as their tenogenesis potential. However, it is di cult to fully determine whether these ADSCs have the capacity to augment healing (34, 35), and how to deliver these stem cells to the lesion remains under debate. Whether su cient stem cells can be administered, whether the cells have high activity and viability, the extent to which the lesion can be reached, and how long the cells remain in the target region have not yet been studied adequately. The stem cell sheets applied in this study were able to supply a su cient quantity of stem cells in a controlled manner while also maintaining cell activity. In addition, as a characteristic of engineering itself, it was easy to move to the application site using a cell shifter without a special adhesion method, and the cells appeared to stay at the application site for a relatively long time. Histological examination, con rmed by GFP expression, also showed that the stem cells remained at the applied site after 2 weeks. This compared to the stem cell injection method, stem cell delivery without worrying such as wash out. It was found that it is an effective method that can be applied (Fig. 5A).
One of the characteristic histological ndings of native enthesis at the interface between the tendon and bone that we attempted to recover is the formation of an unmineralized and mineralized brocartilage layer. We demonstrated that stem cell sheets enhance regeneration with this native enthesis (Fig. 5B). Particularly, when the stem cells were interposition between the tendon and bone, an enthesis site, an increase in the brocartilage layer was apparent on the histological exam that showed statistically and substantially superior biomechanical properties. Even when the stem cell was overlaid on the adjacent part of the repair, brocartilage formation was abundant compared to the repair only; this result may be interpreted as a paracrine effect. However, with regards to interposition, it could be observed to a more robust brocartilage layer and collagen alignment, and this case can be said that not only the direct differentiation potential of the stem cells but also the possibility that the paracrine effect acts simultaneously.
This study had a limitation in that it is di cult to evaluate the mechanism of the effect of stem cells because the tendon-to-bone junction tissue could not directly track the tenogenic and adipo-or osteogenic differentiation of stem cells. In addition, although 8 shoulders were compared per group, the sample size was relatively small for both the histology and biomechanical tests. However, despite the sample size limitations, we were able to con rm the obvious histology and biomechanical statistical differences.

Conclusion
In this study, engineered stem cell sheets were shown to be effective for the recovery process of enthesis (tendon-to-bone junction) in a chronic rotator cuff repair rat model. In particular, we were able to demonstrate that the recovery effect was further improved when the stem cell sheet was interposition between the tendon and bone.  Figure 1 Animal experiment design. The experiment was divided into three groups, with 8 rats in each group  The formation of repair and cell sheet transplantation. The comparison of cell sheet transplantation among the three groups. Control group: Repair only, Experiment group: Cell sheet transplantation on top of the repair site or between tendon and bone. Figure 4