Artificial Microvesicles: New Perspective on Healing Tendon Wounds

Tendons have a limited capacity to repair both naturally and following clinical interventions. Damaged tissue often presents with structural and functional differences, adversely affecting animal performance, mobility, health, and welfare. Advances in cell therapies have started to overcome some of these issues, however complications such as the formation of ectopic bone remain a complication of this technique. Regenerative medicine is therefore looking toward future therapies such as the introduction of microvesicles (MVs) derived from stem cells (SCs). The aim of the present study was to assess the characteristics of artificially derived MVs, from equine mesenchymal stem cells (MSCs), when delivered to rat tendon cells in vitro and damaged tendons in vivo. The initial stages of extracting MVs from equine MSCs and identifying and characterizing the cultured tendon stem/progenitor cells (TSCs) from rat Achilles tendons were undertaken successfully. The horse MSCs and the rat tendon cells were both capable of differentiating in 3 directions: adipogenic, osteogenic, and chondrogenic pathways. The artificially derived equine MVs successfully fused with the TSC membranes, and no cytotoxic or cytostimulating effects were observed. In addition, co-cultivation of TSCs with MVs led to stimulation of cell proliferation and migration, and cytokine VEGF and fractalkine expression levels were significantly increased. These experiments are the first to show that artificially derived MVs exhibited regeneration-stimulating effects in vitro, and that fusion of cytoplasmic membranes from diploid cell lines originating from different species was possible. The experiment in vivo demonstrated the influence of MVs on synthesis of collagen I and III types in damaged tendons of rats. Explorations in vivo showed accelerated regeneration of injured tendons after introduction of the MVs into damaged areas. The results from the studies performed indicated obvious positive modifying effects following the administration of MVs. This represents the initial successful step required prior to translating this regenerative medicine technique into clinical trials, such as for tendon repair in injured horses.

were both capable of differentiating in 3 directions: adipogenic, osteogenic, and chondrogenic pathways.The artificially derived equine MVs successfully fused with the TSC membranes, and no cytotoxic or cytostimulating effects were observed.In addition, co-cultivation of TSCs with MVs led to stimulation of cell proliferation and migration, and cytokine VEGF and fractalkine expression levels were significantly increased.These experiments are the first to show that artificially derived MVs exhibited regeneration-stimulating effects in vitro, and that fusion of cytoplasmic membranes from diploid cell lines originating from different species was possible.The experiment in vivo demonstrated the influence of MVs on synthesis of collagen I and III types in damaged tendons of rats.Explorations in vivo showed accelerated regeneration of injured tendons after introduction of the MVs into damaged areas.The results from the studies performed indicated obvious positive modifying effects following the administration of MVs.This represents the initial successful step required prior to translating this regenerative medicine technique into clinical trials, such as for tendon repair in injured horses.

Introduction
The restoration of normal tendon structure and function following damage is one of the biggest problems in both human and veterinary medicine, especially when assessing and treating injuries in sports horses [O'Brien et al., 2021].The economic cost of treating tendon injuries, alongside the losses in profits incurred by racehorse owners, has been estimated at 400 billion euros worldwide [Russo et al., 2020].Tendons have limited natural healing capacities, and as such are difficult to treat.Therefore, in most cases, long-term rehabilitation is required following treatment.Surgical treatment can restore tendon tissue integrity to a degree, however structural integrity and tissue properties following damage often differ to those observed prior to injury, frequently this is due to scarring and fibrosis.These outcomes lead to a higher risk of recurrent lacerations at the site of injury [Andarawis-Puri et al., 2015], the frequency of which can reach 67% [Bavin et al., 2015].Therefore, in order to treat these injuries, traditional treatment methods are used, but cell therapy can also be implemented to help restore damaged tissues [Smith, 2010].
However, cell therapy has some technical limitations.The main limitations are the possibility of ossification and calcification, as well as the risk of oncogenesis [Kusuma et al., 2015;Kovach et al., 2016;Gissi et al., 2020].To overcome these issues, one of the modern directions pertaining to regenerative medicine is the use of microvesicles (MVs) alongside stem cells (SCs; MV SCs).This technique exhibits the same regenerative potential as the original SCs but does not have the described undesirable side effects.In veterinary medicine, MV SCs are used for both regenerative and reproductive purposes [Zakirova et al., 2020b].However, they are mainly used alongside MVs from natural origins, as opposed to artificial MVs [Qamar et al., 2021].
This study was the first to examine the effects of artificial MVs, obtained from mesenchymal stem cells (MSCs), delivered in vitro to cells isolated from rat Achilles tendons (Tendo calcaneus) and into injured rat Achilles tendons in vivo.The results presented form part of the necessary preclinical studies within the field of regenerative veterinary medicine, which could ultimately be used to treat tendon injuries in companion and sports horses.With the long-term view of conducting clinical trials in horses, the MVs used in this study were obtained from horse MSCs, ensuring they would be compatible for these future interventions.
TSC differentiation analysis was performed following the previously described published method [Zakirova et al., 2021].Briefly, in order to analyze cell differentiation abilities within the cell cultures, cells in their third passage were seeded in 12-well plates, at a concentration of 30,000 cells per well, and then incubated in growth medium until a monolayer formed.To induce differentiation, the cells were cultured in special differentiation media (Gibco, USA).Differentiation was carried out in 3 different directions, namely osteogenic, adipogenic, and chondrogenic.The media were replaced every 3 days.After 21 days of incubation with differentiating media the cell cultures were fixed with 4% paraformaldehyde for 20 min at room temperature.To determine the mineralization (osteogenic differentiation), the von Kossa reaction was used [Naumenko et al., 2021].Silver nitrate solution (2%, w/v) was applied onto the cells with a subsequent incubation for 10 min under darkness.Then, the samples were washed with distilled water and put under bright illumination for 1 h.A mucopolysaccharidespecific stain was used to detect chondrogenic differentiation.Briefly, the fixed cells were washed 3 times with PBS for 5 min each, then stained for 1 h in Alcian blue (1 g Alcian blue/100 mL 0.1 M HCl), and then washed with PBS.To identify the oil droplets in the cells, they were incubated in adipogenic differentiation medium and a qualitative reaction for neutral fats was undertaken using the lipophilic fluorescent dye Nile red (Sigma, USA).The samples were incubated in Nile red solution for 30 min at a temperature of 37°C, then the nuclei were stained using DAPI fluorescent dye (0.1 μg DAPI/1 mL PBS).The differentiation results were recorded using an inverted AxioObserver Z1 microscope (Carl Zeiss, Germany).
To visualize the TSCs by scanning electron microscopy (SEM), the cells were seeded in a 6-well plate and cultured until the monolayer reached a density of 60%.Then the medium was removed, and the remnants of media were removed with a saline solution wash.The samples were then fixed in 2.5% phosphate buffered glutaraldehyde solution (Merck, Kenilworth, NJ) at +4°C for 24 h, dehydrated using an ethanol gradient increasing from 30% to 100%, vacuum dried in Quorum Q150T ES (Laughton, UK), then gold particles were sprayed onto the cells and the samples visualized using a Merlin SEM (Carl Zeiss, Oberkochen, Germany).

Selection and Characterization of Horse MSCs
Horse MSCs with an adipogenic origin were obtained, cultured, and then characterized using the previously described method [Zakirova et al., 2021].Immunophenotyping of the equine MSCs was performed using antibodies (all from BioLegend) against Thy-1 (clone 5E10), CD44 (clone IM7), and CD73 (clone AD2) in accordance with the manufacturer's protocols.The immunophenotyping results were evaluated using a flow cytometer Guava EasyCyte 8HT (Millipore, USA).

Microvesicle Extraction
MVs from horse MSCs were obtained from a cell suspension.A total of 10 6 cells were removed from the substrate using a 0.25% trypsin-versene solution (PanEco, Russia).The cells were washed twice in DPBS (PanEco), then incubated in DMEM (PanEco) containing 10 μg/mL of cytochalasin B (Sigma-Aldrich, USA) for 30 min (37°C, 5% CO 2 ).At the end of the incubation period, the cell suspension was stirred vigorously for 30 s, then precipitated via centrifugation (100 g for 10 min).Thereafter the supernatant was subjected to 2 subsequent PBS centrifugation steps (100 g for 20 min and 2,000 g for 25 min).The precipitate contained the MVs.For the further experiments, the number of TSCs per well was calculated, then MVs obtained from the same number of equine MSCs were added into the experimental wells.

Impact of MVs from Equine MSCs on Rat TSC Viability in vitro, the MTS-Test
Intact rat TSCs were seeded in a 96-well plate at a concentration of 50,000 cells/mL, 200 μL per well.When the monolayer reached 80% confluence, MVs derived from equine MSCs were added into the test wells but not to the control wells.The cells were then incubated at 37°C, in 5% CO 2 .Viability was determined using a CellTiter 96 ® AQ ueous Non-Radioactive Cell Proliferation Assay (Promega, USA) for 24-72 h.A 2 mg/mL solution of MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2Htetrazolium] was prepared, as was phenazine methosulfate (PMS) at a concentration of 0.92 mg/mL in PBS.The resulting solutions were mixed at a ratio of 900 μL MTS:100 μL PMS:10 mL DMEM.This mixture was allowed to settle for 5 min at room temperature.The growth medium was then removed from the wells, and the plates washed with saline solution.A 100-µL volume of the MTS:PMS: DMEM solution was added to each well.The cell culture plate was then incubated for 2 h in a CO 2 incubator.Thereafter, optical density measurements were performed on a CERES 900HDi flatbed analyzer (Bio-TekInstruments Inc., USA), with the main filter set to 490 nm, and the reference filter set to 630 nm.

Assessment of Cell Membrane Fusion between the MVs Obtained from Horse MSCs and the Rat TSCs in vitro
Prior to co-cultivation, 1 million rat TSCs were stained with a red DiD membrane dye (Life Technologies, USA) in accordance with the manufacturer's instructions.Then the cells were seeded onto culture plastic and cultured for 24 h, before the MVs were added, using the method described above.The MVs were stained with DiO (Life Technologies, USA) in order to identify them as green in color.The DiO (green)-stained MVs from the horse MSCs were then added to the wells containing the DiD (red)-stained TSCs, these were incubated under standard conditions for 24 h.Thereafter, the recipient cells were washed 3 times in FBS in order to remove any MVs which had not fused with TSCs.The fusion efficiency was then evaluated using an AxioObserver Z1 fluorescence microscope and flow cytometry (Guava EasyCyte 8HT, Millipore, USA).The cell nuclei within the monolayer had been prestained with DAPI (Sigma, USA).During cytometry, native, nonstained TSCs from rats were used as a negative control, and rat TSCs stained with only the cytoplasmic dye DiD were used as a positive control.

Determination of Cellular Mitotic Activity
To determine the mitotic activity effects of rat TSCs combined with MVs obtained from equine MSCs, E-Plates with 8 wells containing 5,000 cells/well were seeded with rat TSCs, which were then incubated for 24 h with the MVs.After that, cell proliferation activity was recorded for 72 h using xCELLigence RTCA DP (ACEA Biosciences, USA).

Determination of Cell Migration Capacity Using the Scratch Method
Six wells of a 24-well culture plate were seeded with 100,000 TSCs per well and cultured until the monolayer reached 90% confluence.The 3 experimental wells then received MVs extracted from equine MSCs, and the 3 control wells did not receive the MVs.After a 24-h incubation period, the wells were washed and a 1.0-1.3mm wide scratch created with a sterile pipette tip within all of the wells.The culture medium was then immediately removed (along with the detached cells), and fresh medium administered.Using an AxioVision 4.8 inverted microscope (Carl Zeiss), the scratch width was measured at 3 locations within each well (left edge, middle, right edge).These measurements were repeated after 6, 12, 24, 48, and 72 h, until the monolayer had completely covered the scratch zone.

Impact Assessment of MVs from Equine MSCs on Expression of Collagen Types I and III following Co-Cultivation with TSCs
TSCs were seeded on a 96-well plate at a concentration of 50,000 cells/well.After 24 h, the MVs from the equine MSCs were added, then incubated under standard conditions for 24 h.To enable quantitative determination of specific DNA sequences within the cell samples, a real-time polymerase chain reaction (RT-PCR) method was used alongside TaqMan technology, in accordance with the previously developed technique [Zakirova et al., 2019].Specific primers were used (Evrogen CJSC, Russia).Sequential dilutions of cDNA were synthesized from TSC mRNA, and from those treated with MVs from horse MSCs, and were used to construct a standard curve and determine the level of gene expression.The gene expression levels in native cells were assumed to be 100%.Three replicates were performed for all of the reactions.
Growth medium cytokine profiles were determined according to a method described previously [Zakirova et al., 2020a].In brief, the TSCs were seeded at 100,000/well in a 24-well plate.After a 90% monolayer had formed in the experimental group, the cells were co-incubated with MVs for 24 h.Then the medium was replaced, thus removing the remaining MVs, and the plates were incubated for a further 5 days.The cytokine content within the growth medium was then determined using xMapLuminex technology.Analysis was then performed using the MILLIPLEX MAP Rat Cytokine/Chemokine kit (Merck Millipore, Germany), in accordance with the manufacturer's protocol.The data collected were processed using a control software BioPlex Manager v.6.1, and analytical software MasterPlex QT (MiraiBio, USA).

Simulation of Tendon Injury
Currently, there are no animal models that would correspond to the injuries that occur in reality with horses or people.In racehorses, tendon rupture occurs because they were trained under physiological conditions, this does not happen in nature [Voleti et al., 2012].
Traumatic injury АТ was performed on adult rats of Wistar line weighing around 200-250 g.Animals were maintained in polycarbonate cages under a LD 14:10 h photoperiod, water and chow provided ad libitum.For this experiment rats were anesthetized by isoflurane (Wako, Osaka, Japan).The hair from the side surface of the hind paw was removed by depilation with depilatory cream.The skin was treated using antiseptic solution (ChloraPrep).After that, the longitudinal skin dissection was 0.5 cm long, thus providing medial access to the central area of the АТ (Fig. 1).A straight toothed clamp was superimposed onto the area of the lower third of the AT for 10 s, then it was taken off, the tendon was not cut [Yerofeyeva et al., 2020].The skin was returned using interrupted sutures using resorbable suture material Surgikrol (Footberg, Belarus).The animals were divided into groups: (1) group I without treatment, the damaged AT was injected with 0.9% saline solution at a volume of 100 μL (n = 42); (2) group II treated with MVs.They received a single transplantation of equine MVs from MSCs (1 × 10 6 MSC/animal) into the damaged area of the AT (n = 42).MVs were resuspended in 100 μL 0.9% saline solution.
Histological analyses were performed to study the effects of the MVs on the damaged tissue.At 7, 14, 21, 28, and 42 days after administration of the MVs or 0.9% saline solution, the animals were euthanized, and the tendons were extracted.The obtained samples were fixed in 10% formalin (Tathimreaktiv, Russia) and sections were stained with Mallory (Biovitrum, Russia) staining.Mallory staining was necessary to identify fibrous structures of tendon tissue.Following histological methods, light microscopy structure and arrangement of fibers, density and appearance, infiltration of inflammatory cells, neovascularization, and fat deposits were analyzed.The analysis of collagen fibers synthesized de novo in the tendons was performed with Picro-Sirius Red set according to the manufacturers' instructions (Polysciences, Inc., PA).Frozen slices of AT (7 μm thick) were fixed 15 min in 10% buffered formalin at room temperature.Light microscopy with polarized light was used for imaging.Picro-Sirius Red stained collagen I fibers red and collagen III fibers green.The calculation of the area of the colored areas (%) of the obtained samples was carried out with Image J program.The determination of the ratio of collagen types (CT) was defined at the presented samples, and it is based on differences in the color scheme characteristic of each type and transitive forms.

In vivo Imaging of MVs
The duration that the MV MSCs stayed at the injection site was determined as follows: the MV suspension was injected into the traumatized rat АТs, stained with the vital dye Dil, a day after the damage (the method is described above).A 0.9% saline solution was injected at the same volume as the group receiving treatment.Visualization was performed using an IVIS Spectrum in vivo imaging system (PerkinElmer Inc., CA) at days 0, 3, 7, 14, and 21 after administration.

Statistical Analysis
Statistical processing of the results was carried out using primary statistical analysis methods, the results of which are presented as mean ± standard deviation.Secondary statistical data processing was performed using the nonparametric Wilcoxon-Mann-Whitney test.Differences were considered significant at p < 0.05.

Identification and Differentiation of Cells
Once a dense monolayer had been achieved, the cells isolated from rat ATs had an epithelial-like shape, with relatively small dimensions and large nuclei (Fig. 2).The cells isolated from the horse adipose tissue expressed MSC markers: 99% Thy-1, 83% CD44, and 94% CD73.In addition, both the horse MSCs and the rat tendon cells were capable of differentiating in 3 directions: adipogenic, osteogenic, and chondrogenic (Fig. 3).

Cell Viability, Toxicity, and Membrane Fusion
Research was conducted to determine whether the MVs from the horse MSCs had any effect on the viability of rat TSCs in vitro.Effects were detected using the MTS test at 2 different time points; the rat TSCs were tested 24 and 72 h after the addition of the MVs extracted from equine MSCs.No statistically significant differences were observed for either cytotoxic or cytostimulating effects (Fig. 4).
Co-cultivation of TSCs and MVs extracted from the equine MSCs showed it was possible for their membranes to undergo fusion (Fig. 5).In addition, flow cytofluorometry showed that 88% of the TSCs in the positive control samples had been colored red by the DiD membrane dye compared to the native control (set at 100%).Of the 88% TSCs, 90% were recipient cells, identified as they contained green-colored areas (from the DiO membrane dye) on their cytoplasmic membrane.

Cell Proliferation, Migration, and Cytokine Expression
According to our data, incubation with MVs from horse MSCs significantly stimulated the proliferative activities of the rat TSCs (p ≤ 0.05; Fig. 6).
The scratch method results showed that the co-cultivation of TSCs with MVs from equine MSCs had a positive effect on the migration properties of the TSCs.Within 3 h of the scratch being created, the resulting gap had decreased by 19%, and after 7 h by 62% when MVs were added.In the group without treatment, a significant decrease in the gap was only noted after 7 h, and this amounted to a decrease of just 55%.After 48 h, and following co-cultivation with MVs, the TSC gap had completely overgrown, no clear space was present.In contrast, although the control group gap had decreased by 74%, it took until 72 h for the gap to become completely invaded by cells again.RT-PCR showed increased expression of collagen types I and III genes in cells isolated from the rat tendon, following co-cultivation with MVs, compared to the intact control.The increased expression of the collagen I gene was 2.36 relative to the control, whereas collagen III increased to 1.8 relative to the control.
Analysis of the cytokine content before and after cocultivation of the TSCs with the MVs showed that the concentration levels of cytokines G-CSF, Eotaxin, GM-CSF, IL-1a, Leptin, MIP-1a, IL-1b, IL-2, IL-6, EGF, IL-13, IL-10, IL-12p70, IFN-g, IL-5, IL-17a, IL-18, MCP-1, GRO, LIX, MIP-2, TNFa, and RANTES within the culture medium did not change and remained below the determined threshold values.Upon analysis of the medium from the intact TSCs, the concentration of the IL-4 cytokines was determined, and VEGF and FKN levels within the medium were below the defined set thresholds.After co-cultivation, these concentrations increased to 38 ± 6 pg/mL for IL-4, with VEGF at 3,901 ± 251 pg/mL and FKN expressed at 107 ± 4 pg/mL.

Histological Studies
Hypercellularity and hypervascularization were observed in rat tendon samples on the 7th day of the experiment in both treatment and control animals.The tendon fibers were disorganized at the damage sites, but preserved sites with a longitudinal orientation were marked.Concentrations of adipocytes and focal lymphocytic infiltration were present within the samples.
On the 14th day of the experiment, hypercellularity and hypervascularization continued in both groups.The fibers of the tendons were still disorganized at the damage sites, but there was a tendency toward the organization and longitudinal orientation of fibers in the form of soft unformed tendon tissue and tendon restoration.The area of disorganized sites was significantly lower than on the 7th day.In both the treatment and control groups on the 14th day fat infiltration and focal inflammatory infiltration continued, but in comparison to the 7th day there were fewer present and there was a tendency toward the growth of immature collagen fibers at the site of infiltration.
On the 21st day of the experiment, the control samples' (saline only treatment) hypercellularity had decreased.There was a tendency toward organization and maturation within the tendons.The treatment samples con- tained large numbers of fibers with densely formed tendon tissue which had a longitudinal structure, but these sites also contained disorganized fibers.In the group which had received MV treatment, there were more mature tendon fibers than observed in the control group.On the 28th day, the control group had hypercellularity and the fatty infiltration continued; there were still sites containing disorganized fibers.In comparison to the MV-treated samples, the control samples contained a significant number of immature fibers.The samples taken from the MV-treated group contained a significant number of longitudinal collagen fibers.Insignificant numbers of disorganized areas of tendon tissue and immature fibers were present in the MV-treated group.There was a continued tendency toward an earlier maturation of tendon fibers following the injection of MVs in comparison to the controls.
On the 42nd day, the samples receiving MV treatment showed almost complete recovery, a large area of samples was mature tendon.There were still come isolated areas showing fiber disorganization, which was especially evident via Mallory staining.In the samples not treated, similar observations were made, however, more immature fibers were observed in the controls compared to the MVtreated samples (Fig. 7).
After staining the samples with Picro-Sirius on the 7th day, it was observed that the tissue around the tendon in the control group consisted of mainly immature collagen, while in the MV-injection group, a combination of mature and immature collagen was observed.By the 14th day, there was more immature tendon tissue in the control group compared to the MV-treated group; in the treated group there was a tendency for earlier maturation of the collagen fibers.By the 21st day, histology was similar to that observed on the 14th day.In the control group there were more non-mature fibers than in the group treated with MVs.On the 28th day following the injection of MVs, insignificant tendon areas were disorganized and had immature fibers, however, as observed in the previous time periods, there was a tendency toward earlier maturation in the samples which had MVs injected.On the 35th day, there were more immature collagen fibers in the control samples compared to those injected with MVs.On the 42nd day, a similar situation was observed to that on the 35th day, however in the MV-treated group only single areas containing fiber disorganization remained, and the recovery process was almost complete.
Nearly the whole area (>90%) was mature tendon.In the group without MV treatment, the view was similar but more non-mature fibers were present in comparison to the MV-treated group.
The determination of the ratio of collagen types (the value of the CT ratio) was based on differences in the color scheme characteristic of each type and transitive form: type I collagen -red, type III collagen -green.On the 7th day, the ratio of collagen I to III types (CT) was 1.11 ± 0.80 in the group injected with MVs, and in the control group it was 0.67 ± 0.11.On the 14th day, the ratios of CTs I and III were 0.70 ± 0.20 and 1.04 ± 0.09; on the 21st day, the ratios were 1.03 ± 0.17 and 0.74 ± 0.14; on the 28th day the ratios were 0.87 ± 0.18 and 0.60 ± 0.23; and by the 42nd day, the ratios were 3.36 ± 0.70 and 1.47 ± 0.74 (Fig. 8).
Thus, a change in the ratio of CT in the forming tendon tissue during the healing of damaged tissue after the injection of MVs indicated an increase of type I collagen by day 7 of the experiment, this tendency persisted throughout the experiment.On day 42, there was a high content of type I collagen, compared with the control group.

In vivo Imaging of MVs
The analysis of migration potential of the MVs from the injection site at administration, at 1 × 10 6 MSC, in the damaged rat tendon showed no fluorescence for 21 days when using the IVIS Spectrum Perkin Elmer.This was probably caused by the low concentration of MVs.At injection of the MVs, at 3 × 10 6 MSC, there was no fluorescence during the first 3 days, but it was present on the 7th day and continued through until day 21 (Fig. 9).The absence of fluorescence at the injection site up until the 7th day was probably due to tissue edema caused by the surgical intervention.After the edema had subsided, the fluorescence emitted by the MVs was observed.

Discussion
Traditionally, tendons were thought to contain only tenocytes, the resident cells within the tendons.However, a recent study has shown that human and mouse tendons also contain SCs called TSCs, which play key roles in tendon development, homeostasis, and damaged tissue regeneration [Andarawis-Puriet al., 2015].Unlike tenocytes, the predominant resident cells within tendons, TSCs have the ability to self-renew and multidifferentiate.After staining, nuclei of cells appear red, collagen appears blue, and muscle cells become orange-red.There was a lower level of degenerative changes in tendons in the experimental animals (group II) than in the control (group I).Group II exhibited decreased levels of disrupted collagen fibers and cellularity.The thicker, mature collagen fibers appear red-orange (type I collagen); the thinner collagen fibers appear pale green (type III collagen).The quality of tendon repair was evaluated based on properties displayed.
There was a lower level of degenerative changes in tendons in the experimental animals (group II) than in the control (group I).Because of these distinctive properties, TSCs can play a crucial role in tendon physiology, as well as in pathologies such as tendinopathy, which is a common chronic tendon injury [Wang and Komatsu, 2016].It is also known that TSC exosomes are able to regulate matrix metabolism and stimulate TSC tenogenesis, which therefore stimulates tendon healing [Wang et al., 2019].Despite these biological properties, TSCs remain largely unexplored [Zhang and Wang, 2010].In addition, the lack of specific cellular markers for TSCs makes identification difficult in vitro [Li Y et al., 2021].The cells isolated from rat AT in this study represent a heterogeneous population, potentially reflecting a tendency to multilinear differentiation (Fig. 2, 3), expressing specific tendon tissue markers (Scleraxis, tenomodulin), and adult MSC markers (CD90, CD29), as well as markers for embryonic stem cells (Nanog, SOX2).The Scleraxis marker is known to be expressed throughout tendon development and plays an important role in both embryonic tendon development and adult tendon healing [Liu et al., 2021].However, the presence of these markers in the cells does not allow distinguishing TSCs from tendon progenitor cells or tenocytes.A wide panel of several SC markers, that meet the criteria for MSC markers, is now available for use in identifying TSCs [Yin et al., 2016].The isolated cells expressed markers of adult MSCs, embryonic SCs, and specific tendon markers.They, as well as rabbit TSCs, expressed SC markers including Oct-4 and SSEA-4 in vitro.In contrast rabbit tenocytes, which are considered fully differentiated fibroblast-like cells, did not express any of these markers.Moreover, rabbit TSCs have been shown to differentiate not only into tenocytes, but also into adipocytes, chondrocytes, and osteocytes both in vitro and in vivo [Zhang and Wang, 2010;Wang and Thampatty, 2018].The cells isolated in the present experiments also differentiated in 3 main directions (Fig. 3).According to the data obtained, a small number of the cells isolated expressed collagen I, a marker of TSCs and tenocytes.At the same time, the fact that they identify with TSCs was confirmed by the presence, in a predominant number of cells, of markers that are absent in tenocytes, such as collagen III [Yang et al., 2016], CD90, Nanog [Schneider et al., 2018], α-SMA [Citeroni et al., 2020], and SOX2 [Tan et al., 2013].According to the literature data, there are no TSCs expressing CD45 and CD34 on the cell membrane [Durgam et al., 2016], as observed in the isolated cells.Cultured rat cells expressed collagen II in the 4th passage.This does not contradict our conclusions regarding the cells belonging to the TSCs.In the literature, there are data relating to the appearance of collagen II expression in TSCs from rats after 3 passages [Rui et al., 2010].We have also shown the presence of vimentin in the rat TSCs.Vimentin is one of the cytoskeletal proteins characteristic of mesenchymal cells and is commonly used as a tenocyte marker [Fong et al., 2013].There are no data on its expression in TSCs in the literature, however, its presence in the isolated cells does not negate their belonging to the TSCs, because vimentin is a characteristic protein for cells of mesenchymal origin, and there are no data in the literature on when expression begins in tendon cells during ontogenesis.
Upon analysis of the present results, in addition to the available literature, we came to the conclusion that the cells isolated from the rat AT were TSCs.According to the results obtained, co-cultivation of the rat TSCs with MVs from equine MSCs did not affect the viability of the TSCs.Interestingly, it was shown that they are capable of merging their membranes.In this work, the fusion of cytoplasmic membranes from diploid cell lines originating from different animals was demonstrated for the first time.Previously, the possibility of merging in a single species was demonstrated, using MVs derived from transplantable human cell lines with transplantable human cells derived from tumors [Gomzikova et al., 2017].
As a result of co-cultivation with MVs, the expression of both the collagen I and III genes increased in comparison to the intact control.It should be noted that the expression of the collagen I gene was greater than that of collagen III.This is most likely a good prognostic sign for the treatment of tendon injuries in vivo, given the aid of MV treatment.Type I collagen is the predominant form in the extracellular matrix of an intact tendon, accounting for almost 60% of dry tissue mass and approximately 95% of the total collagen present.It provides the tendon with high tensile strength and elasticity.However, the damaged tendon heals using predominantly collagen III, a type that lacks the properties of collagen I [Docheva et al., 2015].
After co-cultivation of the TSCs with MVs from horse MSCs, we showed an increase in the content of the FKN cytokine in the culture medium, compared to the control.Inhibition of the FKN receptor is known to block the migration of tendon progenitor cells in vitro.FKN has also been described as able to induce the release of epiregulin (EREG), a 46-amino acid protein belonging to the epidermal growth factor (EGF) peptide hormone family [Lehner et al., 2019].According to our data, EGF did not increase in the medium following co-cultivation, but after co-cultivation with MVs the migratory activity of the TSCs increased in comparison to the control group.
In the medium surrounding the native TSCs, VEGF concentrations were lower than the standard determined values.Following co-cultivation of the rat TSCs with the MVs derived from equine MSCs, concentrations increased further.It is likely that this is a good prognostic sign regarding the treatment of tendon injuries in vivo.VEGF is known to stimulate both angiogenesis in tissues in vivo [Kovac et al., 2017] and the formation of capillarylike structures in vitro [Zakirova et al., 2020a].In addition, during ontogenesis, VEGF is intensively produced in the fetal tendon tissue, whereas late fetal and adult tendons show decreased expression during tendon development [Russo et al., 2015].Despite this, levels of VEGF are high in the early stages of tendon healing in these older animals [Li ZJ et al., 2021].
In the post-co-cultivation environment, the MVs from equine MSCs, cultured with rat TSCs, experienced an increase in IL-4 concentration.It is known that in vitro IL-4 increases collagen I production similar to human fibroblasts [Bou-Gharios and de Crombrugghe, 2008].In our case, following co-cultivation with MVs, there was increased expression of the type I collagen gene in the rat TSCs.IL-4 inhibits the production of proinflammatory molecules and is also a mitogen of microvascular and endothelial walls in large vessels [Shamriz et al., 2017].All this activity may therefore also help stimulate regeneration of the damaged tendons in vivo.
The tendons are important structures of the musculoskeletal system, consisting of an extracellular matrix and a cellular component.Type I collagen (70-80%) and elastic fibers (up to 2%) predominate in the extracellular matrix.The cellular component consists of tenoblasts and tenocytes.The extracellular matrix surrounding collagen and elastic fibers contains glycosaminoglycans, proteoglycans, and also other small molecules [Chisari et al., 2021].
The tendon healing consists of 3 main phases: the inflammatory phase, the proliferative phase, and the remodeling phase [Abate et al., 2009].
The inflammatory phase of tendon healing starts immediately after injury and is characterized by the formation of a thrombus in the damaged tissue.In this phase, a thrombus forms in the damaged vessels, inflammatory cells are activated, and fibroblasts are recruited.This thrombus is the initial carcass for attracted external inflammatory cells.Platelets and cells inside the thrombus release TGFβ, IGF-I, and PDGF, causing local inflammation.Growth factors involve neutrophils, which activate macrophages for the phagocytosis of necrotic debris.Furthermore, after damage, these cytokines are released from macrophages and internal cells of the endotenon and epitenon and initiate the proliferation stage by recruiting fibroblasts.TGFβ and IGF-I are responsible for the regulation of proteinase activity, stimulation of collagen production, recruitment of fibroblasts.PDGF enhances DNA and protein synthesis and also the expression of other growth factors [Chang et al., 1997].
The proliferative phase is characterized by an increase in the extracellular matrix, an increase of cellularity, and the deposition of a fibrovascular scar by fibroblasts.At the site of injury, fibroblasts migrate and proliferate.Endotenon and epitenon cells also proliferate.At this stage, expression of IGF-I, TGFβ, bFGF, and VEGF factors by cells increases.IGF-I and TGFβ factors are necessary to attract fibroblasts and increase extracellular matrix production.Factors bFGF and VEGF promote angiogenesis and cell proliferation [Hope and Saxby, 2007].
Remodeling of the damaged area starts with the reorganization of the newly deposited collagen.This process overlaps with the proliferative phase and leads to a gradual decrease in cellularity and an increase in the fibrous matrix.Tenocytes and collagen fibers align in the direction of stress.At the same time, there is a decrease in type III collagen, vascularization, cellularity, and also the water content in the forming scar decreases.There is also an activation of the action of collagen, which contributes to the resorption of type III collagen and its replacement with type I collagen, which has more cross-links and tensile strength.This process can continue for months and years after the injury [Oshiro et al., 2003;Chisari et al., 2021].
However, tendon healing is a complex process.Often, recovery occurs to an incomplete extent, in particular, scar formation occurs.Also, after tendon damage, there is an increased synthesis of type III collagen and a decrease in type I collagen synthesis.This leads to the formation of less stable fibers of tendon, which are more prone to rupture and repeated injury [Satomi et al., 2008].
Our results in vivo showed that the introduction of MVs into the site of injury led to stimulation and synthesis of type I collagen, remodeling of the tendon matrix, and regeneration of the entire damaged tissue.The potential regenerative effect of artificial MVs in experiments in vivo has been confirmed by our in vitro studies.MVs from horse MSCs accelerated cell migration and stimulated mitotic activity.In addition, MVs stimulated the expression of the type I collagen gene and cytokines VEGF, IL-4, and FKN in vitro.The obtained results indicate that artificial MVs have the same characteristics as those observed in natural MVs.According to literature data, natural exo-somes released from stem cells facilitate various types of wound healing, promoting collagen synthesis and angiogenesis in vivo [Zhang et al., 2015;Yu et al., 2020].
Analysis of migration and biodistribution of MVs in vivo is an important and necessary condition for development of MVs as therapeutic medicines.There are data available indicating that natural MSC exosomes showed accumulation in damaged tissues after being injected intravenously into mice [Grange et al., 2014].Several authors have shown that vesicles persisted in the damaged tissue for 15 days following local administration [Ren et al., 2019].In our study, a fluorescent signal was detected throughout the first 21 days after injection of artificial MVs.

Conclusion
This study highlights a variety of prospects in relation to the potential application of artificial MVs in novel therapies for tendon injury.The results obtained showed there were no negative impacts following treatment with the MVs in relation to the viability of the TSCs in vitro.Concomitantly, the co-cultivation of TSCs with MVs stimulated the proliferative and migratory activities of the TSCs.In addition, the MVs also stimulated gene expression of type I collagen to a greater extent, and type III collagen to a lesser extent, and increased levels of cytokines with proangiogenic, mitogenic, and anti-inflammatory properties.
Histological studies have shown that the injection of MVs into the damaged tendon significantly improved the structure of the tissue.Overall, we demonstrated the beneficial role of artificial MVs in tendon regeneration of rats.This work indicates that MVs from equine MSCs would potentiate regeneration-stimulating effects following injection into damaged horse tendons.

Statement of Ethics
The protocol for this study was approved by the Biomedicine Ethic Expert Committee of Kazan Federal University (Protocol 3; date May 5, 2015) under both institutional and international ethical ARRIVE guidelines.Injections and care were given in accordance with standard veterinary practice recommendations, and undertaken by qualified clinicians, with additional health and welfare checks and clinical observations conducted.

Fig. 1 .
Fig. 1.Model of tendon injury and introduction of MVs. a Performing surgical access to the Achilles tendon.b Injury by a surgical clamp.c Suture of surgical wound.d Introduction of MVs.

Fig. 2 .
Fig. 2. Cellular morphology of stem/progenitor cells (TSCs) and tenocytes cultured from rat Achilles tendons.a Light micrograph of the monolayer.b Scanning electron microscopy.The arrows 1 and 2 indicate the rounded shape of the TSCs, with a large number of cytoplasmic outgrowths, and differentiated tenocytes, with a fusiform shape, elongated nuclei, and thin cytoplasmic protrusions, respectively.Scale bars, 100 µm (a) and 10 µm (b).

Fig. 3 .
Fig.3.Differentiation results for horse MSCs from an adipogenic origin (ADSCs) and of rat tendon stem/progenitor cells (TSCs).Adipogenic differentiation was noted using a lipophilic fluorescent dye Nile red stain (orange-red); the nuclei were pre-stained with the fluorescent dye DAPI (blue).Osteogenic differentiation was highlighted using von Kossa staining, which revealed calcium phosphate deposits in the extracellular matrix, that turned black upon staining.Chondrogenic differentiation was assessed using Alcian blue staining; the acidic mucopolysaccharides in the extracellular matrix turned blue.

Fig. 4 .
Fig. 4. Determination of rat TSC mitochondrial dehydrogenase activity in both the control group (intact TSCs only) and in the experimental group containing TSCs after co-cultivation with MVs from horse mesenchymal stem cells.No significant differences were observed.p > 0.05.Values represent mean ± standard deviation.TSCs, tendon stem/progenitor cells; MVs, microvesicles.

Fig. 5 .
Fig. 5. Membrane fusion occurred between rat TSCs co-cultivated with MVs. a The recipients were rat tendon cells (red from DiD).b Photomicrographs demonstrated merged cytoplasmic membrane of MVs (green) with cytoplasmic membrane of rat tendon cells.c TSC nuclei were stained blue with the nuclear dye DAPI.d The merged image shows green inclusions (DiO dye), representing the cytoplasmic membrane of MVs from equine MSCs following fusion with TSC cytoplasmic membranes (red DiD dye).Scale bars, 50 µm.

Fig. 6 .Fig. 7 .
Fig. 6.Graph depicting the changes in the proliferation index of both intact rat tendon stem/progenitor cells (TSCs, red) and TSCs following incubation with microvesicles from equine mesenchymal stem cells (blue).* p ≤ 0.05 according to the Wilcoxon-Mann-Whitney test.

Fig. 8 .
Fig.8.Histological analysis of tendon healing through 42 days from start of treatment.The injured Achilles tendons of rats were stained by Picro-Sirius Red staining.The thicker, mature collagen fibers appear red-orange (type I collagen); the thinner collagen fibers appear pale green (type III collagen).The quality of tendon repair was evaluated based on properties displayed.There was a lower level of degenerative changes in tendons in the experimental animals (group II) than in the control (group I).

Fig. 9 .
Fig. 9.In vivo artificial-origin MV biodistribution in rats by optical imaging.Representative images, acquired in the posterior position following the acute damage of tendons and injection of MVs (experimental rats) or 0.9% saline solution at the same volume (control rats) stained by Dil.