Directing the Differentiation of Parthenogenetic Stem Cells into Tenocytes for Tissue‐Engineered Tendon Regeneration

Abstract Uniparental parthenogenesis yields pluripotent stem cells without the political and ethical concerns surrounding the use of embryonic stem cells (ESCs) for biomedical applications. In the current study, we hypothesized that parthenogenetic stem cells (pSCs) could be directed to differentiate into tenocytes and applied for tissue‐engineered tendon. We showed that pSCs displayed fundamental properties similar to those of ESCs, including pluripotency, clonogenicity, and self‐renewal capacity. pSCs spontaneously differentiated into parthenogenetic mesenchymal stem cells (pMSCs), which were positive for mesenchymal stem cell surface markers and possessed osteogenic, chondrogenic, and adipogenic potential. Then, mechanical stretch was applied to improve the tenogenic differentiation of pMSCs, as indicated by the expression of tenogenic‐specific markers and an increasing COL1A1:3A1 ratio. The pSC‐derived tenocytes could proliferate and secrete extracellular matrix on the surface of poly(lactic‐co‐glycolic) acid scaffolds. Finally, engineered tendon‐like tissue was successfully generated after in vivo heterotopic implantation of a tenocyte‐scaffold composite. In conclusion, our experiment introduced an effective and practical strategy for applying pSCs for tendon regeneration. Stem Cells Translational Medicine 2017;6:196–208


INTRODUCTION
The critical function of the tendon is to transfer force from muscle to bone to support body movement. Tendon injury due to acute or chronic trauma is a common disorder and may cause pain and body morbidity in both professional and private life [1][2][3][4]. Injured tendon has limited selfhealing capacity owing to its avascularity and acellularity. Currently, clinical options to treat tendon injuries include autograft and allograft transplantation. However, harvesting of autografts will create a secondary injury site, and transplantation of allografts carries the risk of immune rejection and infectious pathogen transmission [5][6][7]. Furthermore, transplanted grafts rarely restore the structural and functional integrity of injured tendons owing to poor healing at the tendon-tendon or tendon-bone interface.
Cell therapy-based approaches (e.g., tissue engineering) show great potential for tendon regeneration and repair. Considerable efforts have been made to employ various types of cells for tendon regeneration and repair. However, to date, no clinically practical approaches have been developed. Tenocytes may be an adequate cell source for tendon regeneration. However, despite donor site morbidity, limited numbers of cells can be obtained from explanted tissue because tendons are relatively acellular [1], and increasing passage number may result in phenotypic drift [8]. Alternatively, dermal fibroblasts are more readily available through a simple dermal biopsy and are amenable to culture. Dermal fibroblasts could adhere to and proliferate on biomaterials and have been shown to synthesize many components of the extracellular matrix (ECM) [9,10]. The key concern with the use of fibroblast is scar formation, which may significantly influence the mechanical strength of the tendon [11].
Mesenchymal stem cells (MSCs) are characterized by their extensive proliferative ability and the potential to differentiate along various lineages, including bone, cartilage, adipose tissue, and tendon. Previous studies have shown that MSCs could form tendon-like tissue and improve the mechanical properties of injured tendon [12]. However, long-term in vivo observations have demonstrated that ossification frequently occurs in the repaired tissue; this may impair the structure and function of the tendon [13,14]. Ouyang's group established a strategy to induce differentiation of embryonic stem cells (ESCs) for tendon regeneration [15]. The results suggested that the progenitor cells derived from ESCs could promote tendon regeneration by secreting fetal tendon-specific matrix and growth factors. Furthermore, no teratoma or ossification was observed in the newly formed tissue [15]. However, harvesting of ESCs involves the destruction of viable embryos, limiting its use due to political and ethical concerns. Therefore, new cell-based model systems for tissue engineering and regeneration are needed.
Parthenogenesis refers to the embryonic development of eggs activated without fertilization. Mammalian oocytes can be artificially stimulated to develop into diploid nonembryonic blastocysts, and parthenogenetic stem cells (pSCs) can be obtained from the blastocoel inner cell mass. In primates, parthenotes are unable to grow into viable fetuses because genetic defects affect proper placenta formation [16][17][18][19][20][21]. pSCs are more histocompatible than other transplanted cells due to the presence of homozygous human leukocyte antigen (HLA) genotypes. These common HLA haplotype-matched pSCs may reduce the risk of immune rejection after transplantation of their differentiated derivatives, thus offering significant advantages for application to cell-based therapies compared with ESCs [22][23][24]. pSCs can develop into retinal pigment epithelium-like cells [25], muscle-like and bone-like cells [26], neuronal cells [27,28], and hepatocytes [29]. Moreover, cardiomyocytes derived from pSCs have been reported to facilitate engineering of myocardium and have been shown to enhance regional myocardial function after myocardial damage. MHC-haploidentical pSC allografts have been shown to be immunologically accepted in related and unrelated recipients [30]. However, no studies to date have identified the tenogenic differentiation capacity of pSCs.
Here, we examined whether pSCs exhibited properties similar to ESCs, including pluripotency, clonogenicity, self-renewal, and in vitro and in vivo differentiation capacity. By sequential differentiation of pSCs, parthenogenetic mesenchymal stem cells (pMSCs) had been established and further induced through cyclic mechanical stimulation. Cell proliferation, tendon-specific marker, and ECM-enhanced constructs were assessed in vitro and in vivo. Collectively, our data indicated that pSCs are an attractive cell source for tissue-engineered tendon.

Animals
Six-week-old nude mice and C57BL/6 mice were purchased from the Shanghai Experimental Animal Centre, Chinese Academy of Sciences (Shanghai, China, http://english.cas.cn). All animal studies and protocols were carried out following the guidelines of the Animal Holding Unit of Northwest University.

Derivation and Expansion of pMSCs
The 5-day embryoid bodies (EBs) were plated onto dishes coated with 0.1% gelatin (Sigma-Aldrich, St. Louis, MO, https://www. sigmaaldrich.com) and cultured with complete growth medium for 7 days with medium changes every 3 days. Spindle-shaped cells were observed in the outgrowths. The cells were then selectively separated by cell scrapers and collected the scrapes to sediment at the bottom of the tube, subcultured in MesenCult MSC Basal Medium (Stemcell, Cambridge, U.K., https://www.stemcell.com) for 21 days. The culture medium was changed every other day. Cells were subcultured for additional two to three passages before use.

Gene Expression Analysis by QuantitativePolymerase Chain Reaction
Total RNA was extracted from the cells using an RNA Isolation Reagent (Takara Bio Inc., Kusatsu, Japan, http://www.takara-bio.com) according to the manufacturer's protocol. The extracted RNA was quantified using a GeneQuant pro (GE Healthcare Life Sciences, Chicago, IL, http://www.gehealthcare.com). A RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, https://www.thermofisher.com) was used to convert the RNA template into cDNA. Quantitative polymerase chain reaction (q-PCR) was performed using a Bio-Rad Q-PCR system (Bio-Rad, Hercules, CA, http://www.bio-rad.com). The relative levels of gene expression were conducted, using the comparative DDCT method, with b-actin as an internal control and normalized to the control group. The primers used are listed in the supplemental online data.

Mechanical Stretch Stimulation
Mechanical stretch was applied to induce cellular tenogenic lineage differentiation with the Flexcell FX-5000 Tension System (Flexcell International Corporation, Burlington, NC, http://www. flexcellint.com). pMSCs, dermal fibroblasts, and bone marrow stem cells (BMSCs) were seeded onto collagen type 1-coated BioFlex plates at a density of 1 3 10 5 cells per well. When the cultures reached approximately 70%-80% confluence, the cells were subjected to cyclic mechanical stretch with 10% elongation for 24 hours or 10 days (16 hours/day). Each cycle consisted of a 10-second stretch and 10-second relaxation. Control cultures were grown under the same conditions but without stretch.

Western Blotting
pMSCs were subjected to cyclic mechanical stretch with 10% elongation for 24 hours or 10 days (16 hours/day). Total protein was quantified using a BCA Assay Kit (Thermo Fisher). Protein blotting was performed in a Mini-PROTEAN Tetra (Bio-Rad) at a constant current of 110 V for 60-120 minutes. Membranes were blocked with 5% (w/v) BSA (Merck, Darmstadt, Germany, http://www.merckgroup.com) for 2 hours. Membranes were then incubated overnight with primary antibodies, followed by application of appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies (Santa Cruz Biotechnology) for 1 hour. Proteins were detected using the ECL Advance chemiluminescent substrate (GE Healthcare Life Sciences). The primary antibodies used were as follows: anti-COL1A1, anti-COL3A1, anti-TNMD, anti-SCX, antiglyceraldehyde 3-phosphate dehydrogenase (Santa Cruz Biotechnology), and anti-EYA2.

CM-Dil Cell Tracking and Detection
Cells were labeled with CM-Dil (Thermo Fisher) according to the manufacturer's instructions before seeding into poly (lacticcoglycolic) acid (PLGA) scaffolds to track cells in vitro and in vivo.

In Vitro Evaluation of Cell Growth on Scaffolds
For in vitro evaluation of cell growth, 1-cm crimp-like PLGA (LA/ GA 10:90; Ethicon, Somerville, NJ, http://www.ethicon.com) scaffolds were sterilized with 75% ethanol and air dried before use. Next, 1 3 10 5 PDTs (pMSCs after 10 days of stretch) in 10 ml medium (DMEM supplemented with 10% FBS) were then seeded onto scaffolds in 6-well plates. Two milliliters of medium was added 4 hours later to facilitate cell attachment. Eighteen cellscaffold constructs were prepared. Culture media were changed every 2-3 days during the 15-day culture period. Three specimens were harvested at predesignated time intervals of 5, 10, and 15 days. DNA was isolated from each construct with DNAiso reagents (Takara Bio) according to the manufacturer's protocol. The extracted DNA was quantified using a GeneQuant pro (GE Healthcare Life Sciences).

Scanning Electron Microscopy Examination
After observation by laser confocal microscope, three specimens from each scaffold group at different time intervals were prepared for scanning electron microscopy (SEM) examination. The specimens were prefixed with 2% glutaraldehyde, postfixed with 1% osmic acid, sputter-coated with gold (Bal-Tec, Philips, Eindhoven, The Netherlands, http://www.philips.com), and examined with a scanning electron microscope (Philips-XL-30; Philips) to observe cell attachment.
used under the same conditions as negative controls. Then the slides were incubated with suitable Alexa Fluor 488-or Alexa Fluor 594-labeled secondary antibodies (Thermo Fisher). Nuclei were counterstained with DAPI (Thermo Fisher). Images were taken with a laser confocal microscope.

Transmission Electron Microscopy Examination
The newly formed tendon tissues obtained from the in vivo experiments were fixed in fixative solution (2.5% glutaraldehyde and 2% formaldehyde) embedded in epoxy resin. The prepared samples were sectioned to 70-90-nm thicknesses. Ultrathin sections were stained in 3% aqueous uranyl acetate and then in Sato triple lead stain before examination using an FEI CM12 Electron Microscope (FEI Company, Hillsboro, OR, https://www.fei.com). Image-Pro Plus Analysis Software (version 6.0; Media Cybernetics, Rockville, MD, http://www.mediacy.com) was applied to quantify fibril diameter. Briefly, for each sample, three images and approximately 90-100 collagen fibrils from each image were randomly selected for measuring and obtaining the average fibril diameter of each examined sample.

Statistical Analysis
The data are expressed as means 6 SDs from three replicate experiments. Data sets that involved more than two groups were assessed by one-way analysis of variance (ANOVA) followed by Newman-Keuls post hoc tests ( Fig. 2A; supplemental online Figs. 19,20). In the figures, the data with different superscript letters are significantly different based on post hoc ANOVA statistical analysis. Differences with p , .05 were considered significant.

Characterization of pSCs and Generation of MSCs From pSCs
The commonly used criteria to define stem cells are pluripotency, clonogenicity, and self-renewal. We first tested morphology and alkaline phosphatase (ALP) activity (supplemental online Fig. 1) and presence of stemness markers (supplemental online Fig. 2), and we found no differences in PSCs with J1 cells. A small population (approximately 14%-20%) of pSCs formed adherent cell colony (supplemental online Figs. 3, 4). WST-1 assays showed that J1 cells proliferated more rapidly than pSCs at all densities, as assessed at six time points. The molecular basis of these patterns of proliferation demonstrated that pSCs maintained low expression of the IGF2 gene We compared pSCs, J1 lines for their EB formation and multilineage differentiation capability in EB cultures using (a) limiting dilution for EB formation efficiencies (supplemental online Table 1); (b) TdT-mediated dUTP nick-end labeling (TUNEL) for EB apoptotic nuclei (supplemental online Fig. 6); (c) Q-PCR for ectodermal, mesodermal, endodermal, and epithelial-mesenchymal transition (EMT) markers (supplemental online Figs. 7,9); (d) immunofluorescence staining for early mesodermal (stem cells antigen-1 [SCA-1] and CD34), mesodermal (NT5E, MYOG, a-actin) and EMT (N-cadherin, vimentin, and E-cadherin markers [supplemental online Fig. 8]); (e) the process coincided with a multiplex migration of spindleshaped, brick-shaped cells and formed beating cardiomyocytes (supplemental online Fig. 10; supplemental online Movie 1); and (f) teratoma formation after subcutaneous injection in nude mice (supplemental online Fig. 11) .
According to gene expression levels for mesodermal commitment, the outgrowth of spindle-shaped cells was selectively isolated with a cell scraper 7 days after EB plating, and the scraped cells were subsequently expanded in MesenCult MSC Basal Medium to obtain MSCs ( Fig. 1A; supplemental online Fig. 12). The population of cells obtained was passaged at 14-21 days after plating. After two to three additional passages, the cells exhibited a uniform spindle-shaped pattern.
The multidifferentiation potential of pMSCs toward osteogenic, chondrogenic, and adipogenic lineages was analyzed. Q-PCR analysis showed that the expression of osteogenic markers, including ALP, bone sialoprotein (IBSP), osteocalcin (BGLAP1), secreted phosphoprotein 1 (OPN), and runt-related transcription factor 2 (RUNX2), was increased. Strong staining for Alizarin Red S and Von Kossa 21 days after induction demonstrated accumulated mineralization, indicating that these cells had osteogenic potential after induction ( Fig. 1D; supplemental online Fig. 15). Safranin O, immunocytochemistry staining, and Q-PCR verified the chondrogenic lineage differentiation ( Fig. 1E; supplemental online Fig. 16). The cells also had the capacity to undergo adipogenic differentiation. This was evident through the accumulation of lipid vacuoles and upregulated expression of the adipogenic markers aP2 and C/EBP after induction ( Fig. 1F; supplemental online Fig. 17).

Mechanical Stimulation Enhanced the Tenogenic Phenotype of pMSCs
To determine whether pMSCs were tenogenic, we used a mechanical stimulation assay. Dermal fibroblasts, bone marrow stem cells, and patellar tenocytes (PTCs) acted as controls (supplemental online Fig. 18). After exposure to 10% mechanical stretching for 8 hours or 10 days, the expression levels of the genes COL1A1, CO-L3A1, TNMD, and EYA2 exhibited an approximate 18-fold upregulation. No substantial upregulation of SCX was observed ( Fig. 2A; supplemental online Figs. 19,20). In addition, the apex ratios of COL1A1:3A1 for both groups were 9.81 6 0.59 (24 hours) versus 19.64  Figure 2A), suggesting that relatively longer mechanical stimulation will assist cell in differentiation and maturation. The increased ratio of COL1A1: 3A1 was achieved early on for fibroblasts and BMSCs, whereas pMSCs reached their apex at 10 days. Western blot analysis also showed that expression of the tenogenic markers COL1A1, COL3A1, TNMD, and EYA2 was increased after mechanical stimulation for 24 hours (Fig. 2B). Immunocytochemistry results confirmed that mechanical stretching significantly enhanced collagen biosynthesis (as shown by increased expression of COL1A1, COL3A1, COL5A1, and HSP47), induced tenocyte formation, and promoted tenocyte phenotype (as shown by increased expression of BMP2, BMP13, TNMD, EYA2, and TNC; Figure 2C). In addition, mechanical stretch had no significant influence on cell proliferation at the end-point of loading (Fig. 2D).
Taken together, these results provided evidence that mechanical stimulation could enhance the tenogenic phenotype of cells.

PDTs Could Proliferate on PLGA Scaffolds
At 10 days after mechanical loading, PDTs were seeded on PLGA fibers and cultured for 15 days in vitro. CM-Dil staining and SEM observations indicated that PDTs adhered to and proliferated well on PLGA fibers, as shown in Figures 3B and 3C. PDTs bridged the fibers, covered the surface of the scaffolds and began to secrete fine collagen fibrils 15 days after seeding (Fig. 3C). DNA content increased steadily during in vitro incubation, further demonstrating cell proliferation on the scaffold. These results indicated that the PLGA scaffold was suitable for PDTs cultivation (Fig. 3D).

PDTs Regenerated Tendon-Like Tissue In Vivo
Next, we examined the ability of PDTs to regenerate tendon-like tissue in vivo. All animals survived throughout the experiment. The animals were sacrificed at weeks 6 and 12 after implantation to harvest specimens. Gross inspection showed that the newly formed tendon had a smooth and glistening white surface (Fig.  4A). Histology results showed that there was no teratoma formation after implantation. At 6 weeks, H&E staining revealed the tissue containing a large number of spindle-shaped cells, which were situated along the axis of PLGA fibers. Masson's trichrome staining revealed some positively stained area, indicating collagen deposition in the construct. PLGA scaffolds were partially degraded (Fig. 4B). Transmission electron microscopy (TEM) observations showed that some fine collagen fibers, having an average diameter of 43.34 6 9.77 nm, formed in the construct (Fig. 4A, 4G).
In contrast, the number of cells in newly formed tendon tissues was significantly lower at 12 weeks than at 6 weeks postoperation (Fig. 4C). H&E and Masson's trichrome staining revealed a structure of longitudinally aligned fibers and cells with increased matrix deposition. Typical crimp-patterned collagen fibers with proper cell density could be observed, and most of the PLGA scaffolds were degraded. TEM observations further showed the formation of wave-like collagen fibers, having an average diameter of 53.28 6 9.29 nm, in the specimen (Fig. 4B, 4G). The red fluorescent signal of CM-Dil-labeled cells was distributed in the newly formed tissue (Fig. 4B, 4D). Immunohistochemistry staining confirmed that the engineered tendon sustained expression of tendon-specific markers, including COL1A1, COL3A1, COL5A1, TNC, TNMD, SCX, HSP47, EYA2, BMP2, and BMP13 (Fig. 4C, 4E).

DISCUSSION
An optimal cell source for tendon regeneration should possess high proliferative and biosynthetic activity. In this study, we examined the capacity of pSCs to maintain basic biological characteristics of biparental stem cells (J1 cells). We found that pSCs could be differentiated into pMSCs, which had the potential to differentiate into three mesenchymal lineages. With in vitro cyclical mechanical stretch, we then directed pMSCs to differentiate into PDTs, which expressed tenocyte-specific markers. Finally, we successfully engineered tendon tissue in vivo. These findings, together with the fundamental advantages (technical, ethical, and immunological) of pSC application, demonstrated that pSCs may represent an efficient and practical strategy for tendon healing and regeneration. It would be ideal to repair injured tendon with autologous cells. However, treatment generally occurs under conditions of acute injury. Therefore, because application of autologous cells requires a time-consuming expansion process (within weeks or months) to obtain sufficient numbers of tenogenic cells, autologous transplantation may not be practical [31]. Furthermore, the fundamental limitations of the available autologous cells, including tenocytes, BMSCs, and fibroblasts, will significantly   Both pSCs and J1 cells expressed pluripotent markers, formed colonies, proliferated actively, and generated EBs effectively (supplemental online Figs. 1-7). pSCs had relatively lower saturation densities and proliferation rates than J1 cells, consistent with previous studies [32]. The expression of maternally imprinted genes (i.e., IGF2) is unequal expression in pSCs from various species and may influence cell proliferation and growth patterns [33][34][35]. Whereas pSCs only exhibited a dot-patterned colony shape, J1 cells exhibited dot-and ring-patterned colony shapes. We hypothesize that this may be due to the higher expression levels of vitronectin (which is involved in the spreading of cells and stabilization of cell adhesion), or decreased E-cadherin rapidly in J1, suggesting an EMT that exhibited the differentiation of early commitment to mesenchymal fates and may increase cell migration [36].
The development of mesodermal lineages is compromised in parthenotes [35,37]. Our results showed that there were no obvious differences in differentiation potential between pSCs and J1 cells (supplemental online Figs. 7-11). Moreover, pSCs could spontaneously differentiate into three germ lines in EBs and their outgrowths, which may be particularly attractive for cellbased therapy. As demonstrated by Q-PCR, pSC derivatives expressed various mesodermal lineages markers, such as lateral Even the alteration trends of most detective gene were the same. This was somewhat unexpected but attractive. In addition, the formation mesodermal lineages components were observed (Figs. 1, 2), and pSCs were shown to form mature cartilage, muscle, and osseous tissue (supplemental online Fig. 11). These results were consistent with the report by Didié et al. and Liu, who reported that murine PSCs and ESCs had similar fundamental properties, despite notable differences in genetic (allelic variability) and epigenetic (differential imprinting) characteristics [30,38], provided a solid basis for subsequent pSC differentiation into mesodermal lineages, including tenocytes. A number of factors have been identified as important in maintaining the tenogenic phenotype, including ECM, growth factors, mechanical stimulation, and oxygen tension. The mechanical niches have been identified with in a variety of cell [39][40][41]. We found that mechanical stretch was efficient in improving the tenogenic phenotype of pMSCs, BMSCs, and fibroblast. A combination of Q-PCR, immunocytochemistry, and Western blot analyses showed that mechanical stimulation was sufficient in upregulating tenocyte-specific markers at both the mRNA and protein level; these effects were considered important for the maturation of tenocytes (Fig. 2, supplemental online Figs. 19,20). Cell responses to mechanical stimulation varied among the different types of cells. Future studies should examine whether the differences in mechanical stretch patterns are associated with changes in the efficiency of differentiation into tenocytes.
Finally, we tested the applicability of PDTs in tendon regeneration. PLGA-scaffold used in this study suited cellular attachment, growth, and ECM formation (Fig. 3). In vivo tests confirmed that the newly formed tissue possessed the similar structure and ECM components of primary tendons. Histological observation showed that wave-like collagen fibers formed in the specimen, and the implanted cells remained viable in vivo for up to 12 weeks. No teratoma formation could be observed in any specimen (Fig. 4). Compared with the native patellar tendon, the newly formed tendons tissue in the current experiment exhibited looser structures (Fig.  4D) and smaller collagen fiber diameters (Fig. 4G), as evidenced by Masson's trichrome staining and TEM analysis. The less ideal structure of the newly formed tendons may be partially due to the lack of physiological mechanical stimulation in subcutaneous sites.
In the current experiment, we show that PDTs could form tendon-like tissue in ectopic site in vivo. Human PDTs may have implications for the treatment of tendon injury by local delivery, or for the treatment of tendon defect with the combination with scaffold. Therefore, the future goal of our study should be to further investigate the tendon regeneration potential of PDTs in situ in immunocompetent animal models.

CONCLUSION
Our study successfully demonstrated that pSCs function as a unique cell type capable of differentiation into tenocytes. pSCs showed almost the same characteristics as biparental stem cells. Under in vitro cyclical mechanical stretch, pMSCs differentiated and formed tenocytes. Moreover, PDTs seeded on PLGA scaffolds formed neo-tendon tissues in mice. Our studies of pSC properties showed that these cells are attractive alternatives and that may represent a novel tool for tissue engineering.

ACKNOWLEDGMENTS
This study was supported by National Natural Science Foundation Grants 31271026 and 31300797 and by Shaanxi Province