A clinical feasible stem cell encapsulation ensures an improved wound healing

Cell encapsulation has proven to be promising in stem cell therapy. However, there are issues needed to be addressed, including unsatisfied yield, unmet clinically friendly formulation, and unacceptable viability of stem cells after cryopreservation and thawing. We developed a novel biosynsphere technology to encapsulate stem cells in clinically-ready biomaterials with controlled microsphere size. We demonstrated that biosynspheres ensure the bioviability and functionality of adipose-derived stromal cells (ADSCs) encapsulated, as delineated by a series of testing procedures. We further demonstrated that biosynspheres protect ADSCs from the hardness of clinically handling such as cryopreservation, thawing, high-speed centrifugation and syringe/nozzle injection. In a swine full skin defect model, we showed that biosynspheres were integrated to the destined tissues and promoted the repair of injured tissues with an accelerating healing process, less scar tissue formation and normalized deposition of collagen type I and type III, the ratio similar to that found in normal skin. These findings underscore the potential of biosynsphere as an improved biofabrication technology for tissue regeneration in clinical setting.


Introduction
Stem cells, with the intelligence of differentiating to specialized cell types to compensate for somatic cell loss, are becoming a promising tool for the treatment of degenerative diseases [1][2][3]. Many efforts have been devoted to stem cell therapy, as evidenced by hundreds of clinical trials for the treatment of cardiovascular diseases, brain disorders, musculoskeletal defects, and osteoarthritis [4][5][6][7][8][9][10]. However, the therapeutic efficacy of systemic infusion of stem cells was unsatisfactory. A particular drawback for the systemic application of stem cells is the loss of cell viability post transplantation, caused by cell detachment from the extracellular matrix, termed 'anoikis' [11][12][13][14].
To solve this problem, cell encapsulation technology has been developed, encapsulating stem cells within biocompatible materials in a spherical geometry. This microscale encapsulation prevents cell loss as well as preserves stem cell function prior to their interaction with destined microenvironment [15,16]. The encapsulated stem cells were reported to survive longer and perform a long-term therapeutic function, mainly by paracrine actions [16][17][18] promoting surrounding cells proliferation and angiogenesis [19].
Stem cell encapsulation exhibits a considerable potential in experimental research setting. It, however, requires more comprehensive consideration of multiple features such as cell quality, production yield, and feasibility for clinical usage. Many studies reported some effective stem cell encapsulation with a good cell viability by taking the advantage of cell-friendly and highly controllable process of microfluidics [20][21][22]. However, microfluidic cell encapsulation is limited in yield (normally less than 1 ml) [23,24], which are not affordable for drug administration and quality control. To address the issue, many studies focus on chip parallelization [25,26], which is promising but has problems such as incompatibility, instability and complexity. There are a plenty of materials that can be used to encapsulate stem cells [27][28][29], but their clinical feasibility is limited. In addition, the cryopreservation and thawing of the stem cell encapsulates are essential for clinical setting, however, studies have shown that currently clinically-available encapsulates do not ensure a satisfactory cell viability (less than 70%) after thawing [30,31]. Therefore, a systemic and comprehensive biofabrication of clinically ready stem cell encapsulation is urgently needed.
To advance stem cell encapsulation for clinical use and address the above issues, we developed a novel stem cell encapsulation technology, biosynsphere. Stem cells are encapsulated in clinically-ready biomaterials via a mild microfluidic technology. This technology ensures the stability of stem cells post encapsulation. The yield and feasibility of the biosynspheres fabricated meet the requirement of drug administration. The bioviability and functionality of the encapsulated stem cells were well maintained as evidenced by their infusion with destined tissues and promotion of injured tissue repair after their transplantation.

Materials
Clinical grade (sterilized) bovine type I collagen solution was provided by Revotek Co., Ltd. Dulbecco's modification of Eagle's medium (DMEM), Ham's F12 nutrient medium, Dulbecco's phosphate buffered saline (DPBS), and fetal bovine serum were obtained from Gibco Life Technologies, USA. If not specified, all other reagents were purchased from Sigma-Aldrich, USA.

Adipose-derived mesenchymal stem cells (ADSCs)
ADSCs were isolated following a procedure described previously [32,33]. Briefly, ADSCs were isolated from 20 ml abdominal adipose tissue of Bama miniature pigs, treated with collagenase type I, and then cultured in a customized complete medium (Gibco Life Technologies, USA) without animal-derived components for 1 d in a T-75 flask (Thermo Fisher, Carlsbad, CA, USA). Floating cells were removed the next day by replacing the medium. The isolated ADSCs were propagated for three generations within 14 d, and the propagated ADSCs were used to make biosynspheres.

Biosynsphere preparation
Sterilized acidic collagen solution (6 mg ml −1 , kept in 4 • C) was firstly neutralized to pH 6.7-7.2 by adding 1 M NaOH in 10x DPBS with thorough shaking. Neutralized collagen solution was centrifuged at 2000 g for 1 min to remove air bubbles and kept at 4 • C until further use. ADSCs solution was centrifuged at 500 g for 5 min to acquire the cell pellet. After removing the supernatant, a certain amount of the prepared collagen solution was added on top of the pellet and gently mixed using a pipette to prepare the cell-laden collagen solution. The cellladen collagen solution was transferred into a syringe covered by ice bags and pumped into the customized multi-channel microfluidic chip (1.0 mm in width and 0.3 mm in depth) as the dispersed phase, while room temperature mineral oil with 1.5% (v v −1 ) Span 80 was pumped as the continuous phase, using syringe pumps (LSP01-1A, Longer Precision Pump, China). For cell viability preservation, up to 10 ml cell-laden collagen solution can be propelled at the maximum flow rate of 0.2 ml h −1 per channel for each fabrication batch. The cell-laden collagen solution was dispersed by mineral oil to form droplets and flow out of the microfluidic chip to be collected by a customized oil bath. The oil bath was set to 37 • C and filled with mineral oil for 30 min collagen sol-gel transition. After gelation, biosynsphere cores were immediately washed using DPBS to remove the unreacted mixtures and mineral oil to avoid cell damage caused by hypoxia. After that, shells were formed by mild electrostatic deposition using poly-L-lysine and sodium alginate solution, accordingly ( figure 1(b)). To avoid contamination, all the preparation steps were performed in biosafety cabinets.

Characterizations of biosynspheres
Morphology. Biosynspheres were characterized for morphological information by an inverted fluorescence microscope (IX 83, Olympus). To quantify the diameters of the biosynspheres, three batches of biosynspheres were fabricated. For each batch, a dilution of 100 µl of biosynspheres in 2 ml DPBS was prepared. For every dilution, at least 100 biosynspheres were captured and measured for appearance and diameter.
Cell viability assay. Biosynsphere pellets of 100 µl were suspended in 1.0 ml Low glucose-DMEM (L-DMEM). LIVE/DEAD TM cell imaging kit (Cat No. R37601, Thermo Fisher Scientific, USA) and NucBlue TM live cell stain (Cat No. R37605, Thermo Fisher Scientific, USA) were prepared and added into the biosynsphere solution as per manufacturer reference. The solution was then incubated at room temperature for 20 min in darkness. After incubation, the solution was transferred to a 24-well plate and imaged by Nikon A1 laser confocal microscope (Nikon, Japan).
Determination of the stemness of ADSCs. ADSCs in biosynspheres at different stages were collected to analyze the trilineage differentiation as the measure  Cell differentiation. Biosynspheres of 200 µl were suspended in a 15 ml centrifuge tube filled with 2 ml complete media. The tube was centrifuged at 500 g for 5 min and placed in the incubator with cap slightly twist-off. After 3 d, the solution in the tube was replaced by either of the trilineage induction medium (Cyagen, USA) and kept culturing for another 21 d with solution replacement every 3 d (for adipogenic differentiation every 3 d culturing using solution A was followed by 1 d of solution B) before staining. The osteogenic differentiation of ADSCs was observed by alizarin red staining. Adipogenic differentiation of ADSCs was observed by oil Red O staining. The chondrogenic differentiation of ADSCs was observed by Alcian blue staining.
Cell adhesion, stretching, proliferation, and connection. Every 200 µl of biosynspheres were suspended in a 15 ml centrifuge tube filled with 2 ml complete media and placed in the incubator with cap slightly twist-off culturing for 3 d for further use. Cells in the biosynspheres were labeled with cell tracker Green CMFDA and expressed green fluorescence for cell adhesion and stretching evaluation. Or the cells were stained using EdU (red channel), and cell nuclei were stained by 4 ′ ,6-diamidino-2phenylindole (DAPI) (blue channel) for proliferation evaluation. Or cells were labeled by Connexin 43 antibody (green) and DAPI for cell connection evaluation. Labeled cells were imaged by laser confocal microscopy.
Cell migration. Cells in the biosynsphere were labelled by PKH26 fluorescent dye. About 50 biosynspheres with 500 µl complete medium were transferred into one well of a 24-well plate and imaged by laser scanning confocal microscopy every 10 min for 4 h.
Biosynsphere fusion. Every 50 µl of biosynsphere pellet was transferred to a 15 ml centrifuge tube and diluted by 10 ml of complete medium. The tube was then centrifuged for biosynsphere precipitation. After that, the tube with its cap slightly twist-off was placed for incubation at 37 • C and 5% CO 2 without changing the medium. After culturing for 7 d, the fused biosynsphere pellet was stained by calcium-AM and observed under laser confocal microscope.

Biosynsphere dissociation
Every 100 µl of biosynsphere pellet was suspended using 2 ml complete medium and preheated at 37 • C for 30 min. The solution was centrifuged to remove the supernatant. 1 ml of collagenase solution (0.1% w v −1 in complete medium) was then added to suspend the pellet, and the mixture was incubated at 37 • C for 10 min. After that, 10 ml complete medium was added to the mixture, and the whole solution was filtered using a 40 µm cell mesh to harvest the encapsulated cells.

Biosynsphere cryopreservation, thawing and extrusion
Freshly fabricated biosynspheres were suspended in complete medium and cryopreservation reagent (CD-50, Origen Biomedical, USA) in a volume ratio of 1: 4: 1. Each 1 ml of the mixture was transferred to cryogenic vials (Thermo Fisher Scientific, USA), which were pre-cooled at −80 • C for 4 h. The vials were then moved into liquid nitrogen for long-term storage.
For thawing, a vial of biosynspheres was taken from liquid nitrogen and placed in a water bath (Fisher Scientific, US) with temperature of 37 • C. With shaking, the biosynsphere solution thawed and was transferred to a centrifuge tube. The solution was centrifuged to remove cryopreservation reagent, which was replaced by complete medium, for further experiments.
After thawing, the solution was transferred to a 5 ml syringe (Cat No. 302135, BD, USA) and centrifuged to remove the supernatant. The remaining biosynsphere precipitation was then extruded using a syringe pump (LSP01-1A, Longer Precision Pump, China) through either a pipette tip (inner diameter 0.41 mm equivalent to 22G needle) or a 22G syringe needle at the speed of 12 ml min −1 to evaluate the characteristics of biosynspheres after 3D bioprinter extrusion or clinical injection. After wound creation, for analgesia and infection prevention, each pig was administered 20 mg of nefopam hydrochloride daily and 1 g of cefazolin sodium daily via intramuscular injection for five consecutive days.

Animal experiment
For the treatment, 1 ml of biosynspheres and cellladen alginate microspheres (both allogenic ADSCs, and cell density 3 × 10 7 ml −1 ) were injected on top of the wound to completely cover it, correspondingly. One wound was left untreated as control. The treatments were repeated at day 3 and day 5. Images to record the recovery of the wounds were taken at day 1, day 3, day 5, and day 7, followed by once every week until week 6. At day 42, animals were sacrificed, and the recovered wounds were dissected for histology analysis.

Histological and histochemical staining
Swine skin tissue was collected on day 42, fixed with 4% paraformaldehyde, embedded in paraffin by the usual method, sliced to a thickness of 4 µm, stained with HE and Masson, Sirius red according to the manufacturer's instructions. All sections were histologically analyzed and photographed with an optical microscope.

Immunofluorescent staining
Frozen sections were fixed by 4% paraformaldehyde in 4 • C for 10 min, and then washed three times with PBS for every 3 min. Sections were incubated with 0.1% Triton X-100 for 10 min to punch at room temperature before blocking. After being incubated with 1% albumin from bovine serum at 37 • C for 30 min, sections were incubated with the first antibody (mouse anti-CD31, AF3628, RD; mouse anti-αactinin, ab9465, Abcam, and rabbit anti-Ki67, PA5-19442, Thermo Fisher) in 4 • C overnight. After a brief wash with PBS, sections were incubated with a second antibody anti-mouse Alexa Fluor 488, 594 (Invitrogen) for 1 h in 37 • C. After being washed with PBS, the sections were incubated with DAPI for 5 min at room temperature. After a brief wash, sections were mounted and examined under a confocal microscope (DXM1200/NIS-Elements, Nikon). Pictures were taken by an observer unaware of grouping. Five scattered images were taken in each area (scar area, and border zone) for every section. Images were analyzed by Image J. Both single-channel and mosaic images were used.

Statistical analysis
Data were obtained from three separate experiments and expressed as means ± standard deviation. A single factor design was applied to this study. After a significant interaction was detected by the analysis of variance Statistical Product and Service Solutions, (SPSS), the significance of the main effects was further determined by T-test. The level of significance was considered when P < 0.05.

The composition and biofabrication of biosynspheres
As illustrated in figure 1(a), a biosynsphere was composed of the core of the engulfed cells, a plasma-like nutrient milieus to support the viability and stability of the cells, and the envelop shell structure composed of poly-L-lysine and alginate. In this assembling process, we used swine ADSCs (pADSCs) as the engulfed cells. The ADSCs after propagation for three passages from freshly isolated fat tissues were analysed for their viability and surface markers before encapsulation (supplementary figure S1(a)). A two-step process including the core formation and the shell assembling was developed. As shown in figure 1(b), a customized microfluidic chip which can have multiple cross-junctions was employed to generate cell-laden collagen droplets aided by mineral oil as the continuous phase. The droplet formation was achieved by widening the fluidic channel to 1 mm to minimize fluidic shear to protect cells. The cell-laden droplets were collected and solidified in a customized oil bath triggering the sol-gel transition. With repeated washings to remove the oil residues, the acquired biosynsphere core, as illustrated in figure 1(b), showed a round shape with poly-dispersed diameters. Shells were formed by mild electrostatic deposition using poly-L-lysine and sodium alginate solution.
Three batches of biosynspheres were fabricated with averaged sizes of 306 ± 69, 304 ± 64, and 296 ± 73 µm, respectively (figure 1(c)). By adjusting the initial parameters such as inflow rates as well as cell density of the cell-collagen mixture, we achieved the desirably-sized biosynspheres with designated number of cells in the core (figure 1(d) and supplementary figure S1(b)). In general, for a set of one microfluidic chip with eight replicated cross junctions, an input of 10 ml cell-laden collagen solution generates 5 ml biosynspheres with high cell viability (containing up to 6 × 10 7 cells ml −1 ) (supplementary figure S1(b)). The prepared biosynspheres can then be cryopreserved and thawing on-demand ( figure 1(e)).

Biosynsphere preserves the viability and stemness of stem cells
As shown in figure 2, the ADSCs remained alive while they were engulfed in the biosynsphere ( figure 2(a)). Cells were released from the biosynsphere by collagenase, and were easily attached to tissue culture dishes as the same as the cells that did not undergo the encapsulation (supplementary figure S2(a)). Cell viability assessment showed that there were more than 95% of the cells remained alive after they were released from the biosynsphere, comparable to that of normal control cells (supplementary figure S2(b)).
The stemness of ADSCs was maintained after the encapsulation, as defined by the immunofluorescence  figure S2(c)), further evidencing the maintenance of stemness of the ADSCs after the encapsulation. The ADSCs within the biosynsphere were also able to differentiate into osteoblasts, chondroblasts, or adipoblasts ( figure 2(d)). Together, these results confirmed that the biosynsphere and its generation process did not change the stemness of the ADSCs.
The evaluation of cell attachment and stretching, proliferation, migration, and cell-cell interaction of the ADSCs in the biosynsphere indicated that all of these biological functions remained normal (figures 2(e)-(h)). They proliferated normally after they were released from the biosynsphere and cultured in the culture dishes as the same as normal control cells. (supplementary figure S2(d)). These cells also showed a strong cell migration to the plastic surface (supplementary figure S2(e)). The secretion capacity of the ADSCs, as judged by the production of vascular endothelial growth factor (VEGF) in suspension cultures, remained normal (supplement figure S2(f)).
The surface shell of the biosynsphere allowed infiltration of fluorescein isothiocyanate (FITC)labeled dextran (mW > 70 kDa) (supplement figure  S2(g)). The cells in biosynspheres were found to be released and integrated as a whole population after the biosynspheres were cultured under biological conditions for 3 d (figure 2(i)).

Biosynsphere protects stem cells under harsh conditions
Biosynspheres were cryopreserved up to 6 months to further validate its physical and biological properties. Before and after cryopreservation and thawing, the biosynspheres maintained their shape without any sign of defection ( figure 3(a)). The average diameter of the biosynspheres were slightly reduced after thawing ( figure 3(b)). The viability of the stem cells in the biosynsphere subjected to cryopreservation and thawing remained satisfactory, as defined by few dead cells detected by BOBO-3 iodide staining ( figure 3(c)) and normal cell surface markers assessed by flow cytometry ( figure 3(d)). Cell proliferation test indicated that cells behaved as the same for the encapsulated stem cells before and after biosynsphere cryopreservation and thawing ( figure 3(e)). Successful trilineage differentiation were also performed for the ADSCs in the form of biosynsphere (figures 3(f)-(h)).
We extruded the biosynspheres from a 22G syringe needle or a pipette tip to mimic the clinical application conditions. As shown in supplementary figure S3(a), there were neither deformation nor damage found in the tested biosynspheres. The ADSCs in the tested biosynspheres maintained the same biological functions as described above (supplementary figures S3(b)-(f)).

Biosynsphere ensures the function of stem cells in tissue regeneration
A total of three Bama minipigs were used in the study of wound healing of full-thickness skin lesions. Each pig was surgically incised to create three 3 × 3 cm full-thickness defects on its back. One lesion was treated with sterile gauze as control (untreated), the other was treated with ADSCs-laden alginate microspheres filling and sterile gauze (alginic acid), and the third was filled by cryopreserved biosynspheres and sterile gauze (Biosynsphere). The healing process of the wounds on the back of all pigs was consistent during the observation period. Scabs formation starting on day 3 and wound healing on day 35 were observed on the untreated lesion, so did the lesion treated with the alginic acid. The lesion treated with biosynsphere displayed a much accelerated healing process, although a complete wound healing was achieved at the same time point (figure 4(a)).
Further analyses of the wound healing revealed that there was massive fibrous deposition across the healed region in both the untreated and alginic acid treated lesions ( figure 4(b) and supplementary figure  4(a)). Thus, the healed lesions were composed of scar tissue with a faint amount of normalized skin found near the boundary of uninjured tissue. However, in the biosynsphere treated lesions, dispersed net-like deposition of fibrous that are similar to normal skin were more easily found in the regenerated area (supplementary figure 4(b)). Moreover, the fibrotic area of the untreated and alginic acid groups were more than 30 mm 2 while it was less than 20 mm 2 in the Biosynsphere treated lesions. The epithelization rate was also much more predominant in the lesions treated with biosynspheres, which was more than 60% of the healed lesions covered by epithelium.
Particularly, Sirius Red staining indicated that at the junction zone between the regenerative tissue and the scar tissue the deposition of collagen type I and type III were different among the groups ( figure 4(c)). The deposition of collagen type I and type III in the Biosynsphere group was closely similar to that of the normal skin which was as a dispersed net with wider fibers. The area of the two collagens (in pixels) were measured using Image J software (figure 4(c) bottom left). The result showed that the three experimental groups had similar content of collagen type I, while the content of collagen type III was obviously higher in the biosynsphere group. The ratio of collagen type I to type III in normal skin was 3.3, and was significantly increased to 15 in the untreated and alginic acid treated lesions, indicating a scar tissue predomination, as previously demonstrated that collagen type I is one of the major extracellularmatrix (ECM) proteins in scar tissue [34,35]. In contrast, this ratio in the healed lesions treated with biosynsphere was 5.4, also more closing to that of normal skin.

Discussion
In experimental studies, cell encapsulation technology has been adopted for cell therapy to solve the problem of anoikis as observed in clinical trials of stem cell therapy [11][12][13][14]. An ideal stem cell encapsulation should include but not limit the features of cell viability, cell functionality, fabrication yield, mechanical stress protection, cryopreservation, etc to fulfill the clinical requirement. However, there has been no report systematically considered these features. In the present study, ADSCs were encapsulated in clinically ready biomaterials with the reservation of the viability and functionality of the cells. The cell-laden encapsulation, biosynsphere, was thoroughly studied to ensure the encapsulated stem cells, the product yield, and the accessibility are all meet clinical requirement. In vivo study indicated that the biosynspheres were convenient to use and easily integrated to the destined tissues and promoted the functional repair of injured tissues.
Microfluidic based cell encapsulation has been extensively studied because of its high encapsulation efficiency and maintenance of cell viability. For stem cell encapsulation, at least tens of milliliters of a product is required for drug quality control and clinical administration. However, the yield of microfluidic based cell encapsulation are normally reported at the sub-milliliter level. Chip parallelization is one of the most common method to scale up the yield, but it was not suitable for the use of high viscous cell-laden biomaterials due to the viscosity incompatibility and droplet coalescences [25,26,36]. Therefore, the favorable solution should be to increase the efficiency of the microfluidic chip itself. One of the common solutions is to increase the inflow rate but to reduce the shear stress that damage the cells, we designed the fluidic channels of the chip to be 1000 µm in width and 200 µm in depth, which was larger than currently available microfluidic chips involving cell-laden microgels preparation.
Mesenchymal stem cells (MSCs) have been extensively studied because of their differentiation potential and trophic properties for clinical use [37]. The behavior of encapsulated MSC has been studied for the cell viability and paracrine ability. A study showed that encapsulating bone marrowderived MSCs in Methacrylate Gelatin (GelMA) microspheres retained an acceptable cell viability, but needed several weeks for cells to migrate out of the microsphere [22]. Another study has shown that encapsulating MSCs in a hyaluronic acid and mussel adhesive protein liquid system helped maintaining a high survival rate, and the resulting encapsulation promoted myocardial angiogenesis as well as would healing [38]. There is a study that made an attempt to develop an encapsulation formula promoting the secretion of paracrine factors by MSCs [39]. Therefore, it is important to consider the overall integration of stem cell behavior including not only cell viability, proliferation, stemness, differentiation, attachment, stretching, and migration, but also cell-cell communications between the implanted cells themselves, and their interactions with host tissues.
Mechanical manipulation such as syringe injection, cryopreservation, and thawing are some of the fundamental features in stem cell therapy, which allows on-demand and off-the-shelf clinical application [40]. We cryopreserved the biosynspheres for 6 months, thawed, and then extrude the biosynspheres as a simulation of clinical manipulation. In the performed studies ADSCs remained alive, and the biosynsphere remained providing sufficient mechanical protection to the cells, and no abnormality was found in terms of the cell functionality. A slight shrinkage was found on the recovered biosynsphere, which could be attribute to water loss but did not cause any adverse effect.
Alginic acid is one of the most common materials for stem cell encapsulation, and ADSCs encapsulated in alginate hydrogels are found to promote tissue repair [41]. However, as observed in the present study, the wound healing associated with ADSCs encapsulated in the alginate displayed a predominant scar formation with a significant increase in the ratio of collagen type I to type III in the healed tissue. In contrast, the wound healing completed by biosynspheres showed a much smaller scar formation, along with only a slight increase in the ratio of collagen type I to type III. Alginate has been known for its lacking of cell adhesion sites [19]. Therefore, ADSCs encapsulated in alginate might not allow cell attachment so that diminishing the function of ADSCs. Therefore, an encapsulation method with a consideration of the biological complexity of stem cells would demonstrate the effectiveness in promoting regenerative process, as revealed in the present study of biosynspheres. Although one should note that the present study focuses on the encapsulation technology, the animal experiment presented here was preliminary. More animal experiments should be performed and assessed with extra immunohistochemical staining such as cytokeratin, vimentin, or alpha-smooth muscle actin.

Conclusions
In summary, we developed a biosynsphere technology here to improve stem cell encapsulation for clinical application. This procedure fulfills the requirements of cell viability, yield, material feasibility, and drug administration. This novel biosynsphere technology was further proved by an in vivo study of swine full skin defect to be more effective in wound healing. This study provides a promising procedure for clinical application of stem cells in regenerative medicine.

Data availability statements
The data underlying this article will be shared on reasonable request to the corresponding author.
The data that support the findings of this study are available upon reasonable request from the authors.