Novel Application Method for Mesenchymal Stem Cell Therapy Utilizing Its Attractant-Responsive Accumulation Property

Featured Application: We developed a remote administration method of Mesenchymal Stem Cells with therapeutic characteristic that used collagen gel as a sca ﬀ old. Abstract: Stem cell therapy is an emerging treatment modality for various diseases. Because mesenchymal stem cells (MSCs) are known to accumulate at the site of damage, their possible clinical application has been investigated. MSCs are usually administered using intravenous injection, but this route carries a risk of pulmonary embolism. In contrast, topical injection of MSCs reportedly has an inferior therapeutic e ﬀ ect. We developed a remote administration method that uses collagen gel as a sca ﬀ old and investigated the e ﬀ ect of this sca ﬀ old on the retention of stemness, homing ability, and therapeutic e ﬀ ect using a mouse tooth extraction model. After verifying the retention of stemness of MSCs isolated from the bone marrow of donor mice in the sca ﬀ old, we administered MSCs subcutaneously into the back of the recipient mice with sca ﬀ old and observed the accumulation and the acceleration of healing of the extraction socket of the maxillary ﬁrst molar. The MSCs cultured with sca ﬀ old retained stemness, the MSCs injected into back skin with sca ﬀ old successfully accumulated around the extraction socket, and socket healing was signiﬁcantly enhanced. In conclusion, administration of MSCs with collagen sca ﬀ old at a remote site enhanced the lesion healing without the drawbacks of currently used administration methods. signiﬁcantly


Introduction
Mesenchymal stem cells (MSCs), which are known to contribute to tissue regeneration and repair [1] are normally present in a dormant state in almost all organs [2]. Upon the release of cytokines by a stimulus such as inflammation or tissue injury, MSCs begin to proliferate and migrate to where cytokines are being released [3]. However, proliferation and migration of MSCs take time, and a lack of MSCs at required sites may impede tissue healing. Although some methods utilize the regenerative properties of stem cells for therapy [4], in terms of cell delivery methods, a gold standard has not been established and further study is needed. For current cell therapies, MSC administration routes include intra-arterial/intravenous, intraperitoneal, and direct administration into tissues and organs. These methods can roughly be classified into two groups: systemic administration and local administration. Systemic administration is a well-documented method [5,6] whereby MSCs are applied in the cardiovascular system and reach the damaged area through circulating blood. However, fewer than 10% of administered MSCs accumulate at the site of damage, with many cells becoming captured in the lungs [7][8][9][10].
Local administration of MSCs to the injury site has advantages, such as a rapid and localized reaction [11]. However, it involves risks such as cells inducing apoptosis when administered at high density [12], or bleeding and secondary damage caused by the administration itself. For these reasons, treatment outcomes are not always stable and the ideal administration method has not yet been established for MSC-based therapies.
Therefore, we focused on the potential of MSCs to autonomously accumulate in damaged tissue [13][14][15]. Our hypothesis is that subcutaneous administration of a mixture of scaffold and MSCs into the body, such as under the easily accessible back skin, will result in MSCs autonomously migrating from the scaffold to the site where cytokines are produced and enhancing tissue repair [16]. Collagen gel, which has high biocompatibility and is easily applied in the clinic, was used as a scaffold to administer a sufficient number of stem cells in this study. We investigated the migration of delivered stem cells and the healing of a tooth extraction site, where both soft and hard tissue healing are involved.

Animals
Male C57BL/6N mice (6 weeks of age; Kyudo Laboratories, Tosu, Japan) and male green fluorescent protein (GFP)-transgenic C57BL/6N mice (CAG-EGFP; 6 weeks of age; Japan SLC, Shizuoka, Japan) were used in this study. All animal experiments were performed under an institutionally approved protocol for the use of animals in research at Kyushu University (approval number: A29-237-0), (approval day: 29 July 2016).

Isolation and Culture of MSCs
MSCs were isolated from the bone marrow of mice as described previously [17]. Briefly, bone marrow cells were flushed out from the femoral and tibial bone cavities of the mice. The cells were passed through a 40-µm cell strainer to obtain a single-cell suspension, which was seeded in 100-mm culture dishes at 1 × 10 6 cells/dish. One day after seeding, cells were washed with phosphate-buffered saline (PBS) and cultured in growth medium comprising alpha-minimum essential medium (α-MEM; Invitrogen, Grand Island, NY, USA) supplemented with 20% fetal bovine serum (FBS; Equitech-Bio, Kerrville, TX, USA), 2 mM L-glutamine (Invitrogen), 100 U/mL penicillin (Invitrogen), and 100 µg/mL streptomycin (Invitrogen).

Osteogenic Differentiation Assay
To confirm calcium deposition, MSCs were seeded on 35-mm dishes at 2.5 × 10 5 cells/dish, grown to confluence in growth medium, and then cultured in osteogenic culture medium (growth medium containing 1.8 mM KH 2 PO 4 and 10 nM dexamethasone; both from Sigma-Aldrich, St. Louis, MO, USA). After 28 days of osteogenic induction, cultures were stained with 1% alizarin red S solution (Sigma-Aldrich). Measurement of the positive area of alizarin red was performed using Image J (National Institutes of Health, Bethesda, MD, USA) [18].

Adipogenic Differentiation Assay
MSCs were seeded on 35-mm dishes at 2.5 × 10 5 cells/dish, grown to confluence in growth medium, and then cultured in adipogenic culture medium (growth medium containing 0.5 mM isobutylmethylxanthine, 60 µM indomethacin, 0.5 µM hydrocortisone, and 10 µg/mL insulin; all from Sigma-Aldrich). After 14 days of adipogenic induction, cultures were stained with oil red O. The oil red O-positive lipid droplets were observed using an inverted microscope (BZ-9000; Keyence, Osaka, Japan).

CFU-F Assay
The CFU-F assay was performed as described previously [17]. Passage 1 MSCs were seeded into culture dishes (Nalge Nunc, Rochester, NY, USA). After 16 days of culture, cells were stained with a solution of 0.1% toluidine blue and 2% paraformaldehyde. Total colony numbers were counted per dish. Three independent experiments were performed. Measurement of the positive area of toluidine blue was performed using Image J [18].

MSC Injection Via the Tail Vein
GFP-MSCs isolated from the bone marrow of transgenic mice were cultured and passaged three times before injection. Mice (n = 5 per group) were anesthetized by combined anesthesia (0.3 mg/kg medetomidine, 4.0 mg/kg midazolam, 5.0 mg/kg butorphanol) and ex vivo-expanded MSCs (1 × 10 6 cells in 2 mL of PBS) were administered via the tail vein at 48 h before tooth extraction. Control mice (n = 5) were injected with PBS via the tail vein.

MSC Injection into the Back of Mice
GFP-MSCs and mice were prepared as above. The ex vivo-expanded MSCs (1 × 10 6 cells in 2 mL of PBS) were then administered into the back of the mice subcutaneously at 48 h before tooth extraction. Control mice (n = 5) were injected with PBS via the same route.

Cell Cultures
For in vitro assessment, MSCs were seeded on 35-mm dishes at 2 × 10 4 cells/dish and incubated for 12 h at 37 • C under 5% CO 2 . A subset of MSCs was incubated in medium containing 100 nM chloromethyltetramethylrhodamine (MitoTracker Orange CMTMRos; Molecular Probes, Eugene, OR, USA) for 20 min at 37 • C in the dark. After three washes with medium, these cells were examined under an inverted fluorescence microscope (BZ-9000).

Collagen Gel Culture
MSCs were embedded three-dimensionally in the collagen gel. Neutralized collagen solution was prepared by mixing eight volumes of Cellmatrix type I-A (Nitta Gelatin, Osaka, Japan) with one volume of 10× α-MEM and reconstitution buffer (2.2 g NaHCO 3 in 100 mL of 0.05 N NaOH and 200 mM HEPES). The solution contained type I collagen at a final concentration of 2.4 mg/mL. To prevent cell attachment and growth directly on cell culture dishes, 1 mL of plain collagen solution was poured into each culture dish as a base layer. MSCs were suspended in neutralized collagen mixture at a density of 1.0 × 10 6 cells/mL of gel. After allowing the base layer to solidify with a short incubation at 37 • C, 1 mL of neutralized collagen mixture containing cells was added. After the top gel layer was formed by additional incubation at 37 • C for 15 min, 1.5 mL of culture medium (α-MEM with 2.5% FBS and antibiotics) was added into the culture dish. The medium was changed every 3 days [19,20].

The Injection of Collagen Gel with MSC
To prepare a homogeneous mixture, MSCs (1.0 × 10 6 cells) were stirred on ice with a mixed solution of 1.6 mL of collagen gel (Cellmatrix type IA) and 0.4 mL of reconstitution buffer. Next, mice (n = 5 per group) were anesthetized using a combination anesthetic (described above), and 2 mL of collagen gel containing MSCs was injected into the back of each mouse subcutaneously using a 5-mL syringe (Terumo, Tokyo, Japan) and 25-G injection needle (Terumo). MSCs were administered at the intersection between the straight line connecting the median line and the shoulder blades on both sides of the mice.

Counting of Injected MSCs in Collagen Gel
Mice were deeply anesthetized and the scaffold was removed from the back of the mice. The collected scaffold was fixed in 4% paraformaldehyde (pH 7.4). Frozen samples were cut into 10 µm-thick sections with a cryostat at −20 • C. After choosing 10 sections from every 10 slices, they were stained and examined under an inverted fluorescence microscope (BZ-9000) by counting cells in a visual field region (500 × 500 µm) in random areas. Then, the average number of cells was determined.

Counting of Injected MSCs at Tooth Extraction Site and Organs
Mouse maxillary first molars were extracted under a combination anesthetic. Three days after tooth extraction, mice were deeply anesthetized and perfused intracardially with heparinized PBS. As controls, mice were subjected to tooth extraction but not administered MSCs. Intact maxilla, kidney, liver, lung, spleen, femur, and tibia were harvested en bloc. After preparing frozen sections, 10 sections were selected from every 10 slices to observe GFP-MSCs that accumulated in each organ. The sections were stained, and the number of cells were counted.

Tissue Preparation
For in vivo assessment, tissues were prepared according to our previously described methods [21,22]. At the end of each experimental period, mice were euthanized under combination anesthesia and perfused intracardially with heparinized PBS, followed by 4% paraformaldehyde (pH 7.4). Maxillae were demineralized in 5% EDTA for 4 days at 4 • C. Prepared sites were cut into 10-µm buccopalatal sections with a cryostat at −20 • C.

Immunofluorescence Microscopy
Immunofluorescence staining of in vitro and in vivo samples was carried out by blocking samples with normal serum (matched to the secondary antibody) for 1 h, followed by incubation with mouse anti-rat GFP, CD29, CD45, CD90, and CD105 antibodies (1:100; Sigma-Aldrich) overnight at 4 • C. Samples were then treated with a fluorescein isothiocyanate-conjugated secondary antibody (1:200; Jackson ImmunoResearch, West Grove, PA, USA) for 1 h at room temperature (RT) and mounted with 4 ,6-diamidino-2-phenylindole (DAPI). A subset of sections was stained with Azan stain [23].

Expression of Adhesion Protein
To observe the primary epithelial closure of the tooth extraction cavity, laminin (Ln)-332 was immunostained to confirm the continuity of the epithelium (the distance between edges of basement membranes). Tissue sections were blocked with 10% normal goat serum (Vector Laboratories, Burlingame, CA, USA) for 30 min at RT. Sections were stained using rabbit anti-mouse Ln-332 (1:100, Santa Cruz Biotechnology, Santa Cruz, CA, USA) antibodies (1:100, Sigma-Aldrich) diluted in antibody dilution solution for 18 h at 4 • C. Finally, samples were treated with fluorescein isothiocyanate-conjugated secondary antibody (1:200; Jackson ImmunoResearch) for 1 h at RT [21].

Statistical Analysis
Data are expressed as mean ± standard deviation (SD). Student one-way analysis of variance with Tukey's range test or t-test was performed. Values of p < 0.05 were considered significant. Experiments were performed using triplicate samples and repeated three or more times to verify their reproducibility. Figure 1A shows the experimental schedule and method used in the study. The number of MSCs remaining in the scaffold at 21 days was significantly decreased compared with the number on other days ( Figure 1B). During the first 14 days, MSCs retained their original characteristics, including abilities for self-renewal and multi-directional differentiation. As shown in Figure 1C, MSCs maintained expression of MSC positive markers (CD29, CD90, and CD105) without expressing negative marker (CD 45) could differentiate into adipocytes or osteoblasts based on oil red O or alizarin red S staining, respectively, and formed colonies from single cells.

Statistical Analysis
Data are expressed as mean ± standard deviation (SD). Student one-way analysis of variance with Tukey's range test or t-test was performed. Values of p < 0.05 were considered significant. Experiments were performed using triplicate samples and repeated three or more times to verify their reproducibility. Figure 1A shows the experimental schedule and method used in the study. The number of MSCs remaining in the scaffold at 21 days was significantly decreased compared with the number on other days ( Figure 1B). During the first 14 days, MSCs retained their original characteristics, including abilities for self-renewal and multi-directional differentiation. As shown in Figure 1C, MSCs maintained expression of MSC positive markers (CD29, CD90, and CD105) without expressing negative marker (CD 45) could differentiate into adipocytes or osteoblasts based on oil red O or alizarin red S staining, respectively, and formed colonies from single cells.

MSC Behavior in Collagen Gel Scaffold (In Vitro Study)
Collagen gel  Figure 2A shows the experimental schedule and method used in the study. MSCs were injected with collagen gel into the subcutaneous tissue of mice. The appearances of mice backs immediately, 7 days, and 14 days after collagen injection are shown in Figure 2B. Until 7 days after collagen injection, the collagen retained its shape without significant changes. Additionally, as shown in Figure 2C, the gel showed some changes in the number of GFP/CD90-positive cells over 2 weeks. Thus, MSCs decreased gradually from the scaffold after 2 weeks in vivo.   Figure 3A shows the experimental schedule and method used in the study. After tooth extraction, the number of MSCs in the scaffold decreased significantly compared with the number before extraction ( Figure 3B). In particular, at 5 and 7 days after collagen injection with MSCs, there was a large difference between extraction and non-extraction groups.

Changes in Cell Number in Scaffolds before and after Tooth Extraction
for in vivo experiments. (B) (Upper) Appearances of mice backs after collagen injection (immediately, 7 days, and 14 days after injection). (Lower) Shapes of collagen gels in the backs of mice (immediately, 7 days, and 14 days after injection). The size of the collagen gel decreased over time. (C) Numbers of green fluorescent protein (GFP) positive cells in collagen gels at time of injection and upon removal from mice backs at 2 weeks after injection (Number of cells per 500 × 500 μm field in section). More than half the cells remained alive for 2 weeks. * p < 0.05. Figure 3A shows the experimental schedule and method used in the study. After tooth extraction, the number of MSCs in the scaffold decreased significantly compared with the number before extraction ( Figure 3B). In particular, at 5 and 7 days after collagen injection with MSCs, there was a large difference between extraction and non-extraction groups.  Figure 4A shows the experimental schedule and method used in the study. Numbers of accumulated MSCs in lung, kidney, liver, spleen, and bone marrow after MSC injection via the tail vein and injection of collagen gel with MSC are shown in Figure 4B. In particular, the lung accumulated a large number of MSCs in MSC injection via the tail vein. MSCs were not observed except in the lungs of the collagen group. As shown in Figure 4C, mice injected with MSCs in the scaffold or via the back accumulated much lower numbers of MSCs in the lung than those injected with MSCs via the tail vein. Figure 5A shows the experimental schedule and method used in the study. In mice injected with MSC via the tail vein and in the scaffold MSC groups, many GFP/CD90 double-positive cells were observed around the socket after tooth extraction ( Figure 5B). The numbers of MSCs were similar in the scaffold MSC and tail vein-injected MSC groups. Although both control groups exhibited poor healing, the scaffold MSC group showed good mucosal healing, similar to the injected MSC group ( Figure 5C). Immunohistochemical staining with Ln-332 revealed the degree of epithelial healing ( Figure 5D). The data indicated that scaffold MSC had epithelial continuity similar to that of tail vein-injected MSC.

Effect of Delivering MSCs with a Scaffold on Epithelial Healing after Tooth Extraction
Appl. Sci. 2019, 9,4908 8 of 13 accumulated a large number of MSCs in MSC injection via the tail vein. MSCs were not observed except in the lungs of the collagen group. As shown in Figure 4C, mice injected with MSCs in the scaffold or via the back accumulated much lower numbers of MSCs in the lung than those injected with MSCs via the tail vein.   the scaffold MSC and tail vein-injected MSC groups. Although both control groups exhibited poor healing, the scaffold MSC group showed good mucosal healing, similar to the injected MSC group ( Figure 5C). Immunohistochemical staining with Ln-332 revealed the degree of epithelial healing ( Figure 5D). The data indicated that scaffold MSC had epithelial continuity similar to that of tail veininjected MSC.

Discussion
Recently, clinical research of stem cell therapies has been increasing. Although the mechanisms underlying the therapeutic benefit of stem cells remain unclear, many reports suggest the efficacy of related cell therapies [24,25]. In this study, two mechanisms of enhanced healing of the tooth extraction site can be assumed. First, administered MSCs differentiate into gingiva cells and facilitate wound closure. Second, MSCs secrete cytokines and growth factors that activate epithelial cells to migrate and proliferate to achieve primary closure of the damaged tissue [26].
We previously reported that intravenous administration of MSCs in rats contributed to the healing of damaged tissue [27]. However, delivering cells subcutaneously without a scaffold did not result in a positive effect on tissue healing [28]. As such, we focused on the scaffold as a foundation for cells.
First, we searched for a suitable scaffold material in which MSCs could retain their characteristics for long periods of time. Although we verified the competence of other materials (data

Discussion
Recently, clinical research of stem cell therapies has been increasing. Although the mechanisms underlying the therapeutic benefit of stem cells remain unclear, many reports suggest the efficacy of related cell therapies [24,25]. In this study, two mechanisms of enhanced healing of the tooth extraction site can be assumed. First, administered MSCs differentiate into gingiva cells and facilitate wound closure. Second, MSCs secrete cytokines and growth factors that activate epithelial cells to migrate and proliferate to achieve primary closure of the damaged tissue [26].
We previously reported that intravenous administration of MSCs in rats contributed to the healing of damaged tissue [27]. However, delivering cells subcutaneously without a scaffold did not result in a positive effect on tissue healing [28]. As such, we focused on the scaffold as a foundation for cells.
First, we searched for a suitable scaffold material in which MSCs could retain their characteristics for long periods of time. Although we verified the competence of other materials (data not shown), we selected collagen as a scaffold because it is fluid at low temperatures, which makes it easy to deliver by injection and less damaging to the cells. In addition, collagen forms a gel at body temperature, which facilitates retention at the application site [29]. Generally, if artificial materials exist in the body for a long time, problems such as infection or degradation may arise; thus, the characteristics of type I collagen, which can be absorbed by the body, might be suitable. In addition, collagen seems to provide an effective scaffold with little toxicity, as suggested by other studies [30,31].
Consistent with previous reports [32,33], MSCs could remain alive in the scaffold without major changes in cell number for 2 weeks. Meanwhile, as shown in Figure 1C, MSCs maintained expression of MSC markers (CD29, CD90, and CD105), and MSCs extracted at each stage exhibited differentiation and self-renewal potential. Therefore, MSCs could maintain their "stemness", which is an important characteristic for at least 14 days.
Second, we checked the competence of collagen as a scaffold. As shown in Figure 2A, the injected 2 mL of collagen retained its form at 2 weeks after insertion into back subcutaneous tissue. Our results were consistent with previous reports suggesting that a 2-week period was required to absorb approximately 5-mm diameter (corresponding to 2 mL) of collagen containing MSCs after insertion into the body as a scaffold for exogenous cells [34]. Safety of the collagen scaffold was determined by the survival rate of MSCs in the scaffold in vivo experiments. As a result, MSCs in collagen gels implanted in back subcutaneous tissue remained alive for at least 2 weeks, similar to the in vitro data shown. Thus, the injected collagen scaffold had the potential to act as a reservoir for exogenous MSCs in the body.
Third, we investigated whether MSCs in the collagen gel could maintain their chemotactic properties. A previous report indicated that systemically delivered MSCs accumulated at sites of inflammation [27,35]. As shown in Figure 3, MSCs could be guided from the collagen scaffold to the wound area after tooth extraction. However, most MSCs remained in the scaffold when tooth extraction was not performed. MSCs in the collagen gel seemed to have the ability to accumulate at a wound site via the systemic circulation. Therefore, we examined the localization of the MSCs after they migrated from the collagen scaffold, as shown in Figure 4.
Regarding the effect of administration method, we previously conducted a study comparing systemic and topical administration, which suggested that subcutaneous administration of MSCs was ineffective when not applied with a scaffold [28]. Similar to previous findings [36,37], the lung and liver accumulated many cells from the systemic circulation. In this experiment, many MSCs were observed in the lung following systemic administration without using scaffolds, but there was only a slight accumulation in the lung when the MSCs were injected within a collagen scaffold. After tooth extraction, fewer MSCs from the collagen scaffold were observed in the lung but could be seen in the wound area. Therefore, delivering MSCs within a scaffold may decrease the risk of lung or cardiac infarction. As MSCs are adherent cells, they readily form cell aggregates immediately after isolation from culture dishes [38]. Thus, when injected via the tail vein, most of the cells formed a mass. This might explain why intravascularly applied cells have higher risks for capture in capillary-like lung tissue. In contrast, cells migrating from the scaffold should move individually, making them less likely get clogged in the lung, which is supported by our results.
Finally, we investigated the effect of MSCs with a collagen scaffold. As mentioned before, many reports suggest that MSCs could accumulate in the wound area after systemic administration [39]. As shown in Figure 5B, many cells were observed around the extraction site following conventional systemic injection, but a similar number of cells were observed around the site in the scaffold group. Both groups showed similar promotion of wound healing after tooth extraction, compared with both control groups ( Figure 5C). On the contrary, we recently reported that the high density of MSCs at the injected site following local administration meant that it was hard for them to move from the administration site to the site of inflammation [28].
In the scaffold group, the cells present in a suitable space in the collagen apparently maintained their MSC properties. This indicates that side effects such as pulmonary embolism may be avoided by using collagen scaffolds, and that this process may also contribute to tissue healing.
The amount of cytokines released from the wound site is reported to closely correlate with the degree of inflammation [40]. In addition, MSC migration in response to cytokines has been shown to occur in a dose-dependent manner [41]. This suggests that when severe damage occurs in a tissue, increased inflammation and cytokine release elicit migration of MSCs to that area. Based on this concept, our delivery method should control the number of MSCs migrating to the damaged site depending on the degree of damage, which may overcome some of the disadvantages of existing cell delivery methods.

Conclusions
In summary, administering stem cells remotely via a collagen scaffold and allowing cells to migrate autonomously to the damaged area is advantageous for delivery of cell therapies. This novel method has great potential for both the treatment and prevention of tissue damage, given the ability of MSCs to detect abnormal and damaged cells. MSC-based therapy using a collagen scaffold may thus offer a safer and more effective therapeutic modality for various systemic diseases.
Author Contributions: N.U., I.A., Y.A., I.N., and Y.M. were involved in the practical aspects of experiments. N.U., I.A., Y.A., T.Y., R.K., Y.W., and X.Z. collected, analyzed, and interpreted the data. I.A. designed the study and provided financial and administrative support. I.A., Y.A., T.Y., A.F., and K.K. wrote the manuscript. I.A., T.Y., A.F., and K.K. revised the manuscript for publication. Each author participated sufficiently in the work to take public responsibility for appropriate portions of the content.