SHED isolated from a 9-year-old girl’s mobile deciduous tooth were provided by Kidswell Bio Corporation (formerly Gene Techno Science Co., Ltd., Tokyo, Japan).

The specifications of SHED such as surface protein expression and multilineage potential are available at https://www.summitpharma.co.jp/japanese/service/s_ATCC_msc.html. Regarding surface protein expression, SHED were positive for CD73, CD90, and CD105 and negative for CD14, CD19, CD34, and HLA-DR. SHED were cultured in expansion medium consisting of Minimum Essential Medium-alpha (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 20% fetal bovine serum (FBS; Nichirei, Tokyo, Japan), 1% Penicillin–Streptomycin (10,000 U/mL; Thermo Fisher Scientific), and 100 $\mu$M ascorbic acid (Wako, Osaka, Japan) at 37 °C under 5% CO${}_2$. SHED were dissociated using TrypLE™ Select (Thermo Fisher Scientific). Culture medium was changed every 3 d, and the cells were passaged when they reached 70% to 80% confluence. SHED were used within 10 passages [25].

The human glioma cell lines U87 and U251 were purchased from the American Type Culture Collection (Manassas, VA, USA) and from the Japanese Collection of Research Bioresources (Ibaraki, Osaka, Japan), respectively. The human glioma cell line U87-luc (Bioware Ultra Cell Line U-87 MG-luciferase2) was purchased from Caliper Life Sciences (Waltham, MA, USA). Cells were cultured in Dulbecco’s Modified Eagle’s medium (DMEM) (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% FBS at 37 °C in a humidified atmosphere of 5% CO${}_2$. Glioma cell lines were dissociated using Accutase (Sigma-Aldrich). Half of the medium as changed every 3 or 4 d. When a confluent cell monolayer was observed, the cells were passaged at a ratio of 1:3. Glioma cell lines were used within 15 passages [26].

Human fibroblasts derived from fetal skin (human dermal fibroblast: hDF; RCB1139) were obtained from Riken BioResource Center (Tsukuba, Japan) and cultured in DMEM supplemented with 10% FBS at 37 °C in a humidified atmosphere of 5% CO${}_2$. Human MSC from bone marrow (hMSC-BM; RCB2163) were obtained from Riken BioResource Center (Tsukuba, Japan), and human MSC from adipose tissue (hMSC-AT; #C-12977, PromoCell, Heidelberg, Germany) were provided by Kidswell Bio Corporation (formerly Gene Techno Science Co., Ltd., Tokyo, Japan). Cells were cultured in Mesenchymal Stem Cell Growth Medium II (Ready-to-use) (Takara Bio, Shiga, Japan) at 37 °C in a humidified atmosphere of 5% CO${}_2$. All cells were used within 10 passages [11, 27].

Human astrocytes were obtained from Lonza (Basel, Switzerland). Cells were cultured in Astrocyte Growth Medium SingleQuots™ Supplements and Growth Factors (Ready-to-use) (Lonza, Basel, Switzerland) at 37 °C in a humidified atmosphere of 5% CO${}_2$. All cells were used within 10 passages.

To evaluate the migration, hMSC-AT, hMSC-BM, SHED, and hDF alone cells were incubated in 60-mm dishes at 1 $\times$ 10${}^5$ cells per dish, such that the cellular confluence reached approximately 80% after 24-h culture. Then a wound was created with a 1000-$\mu$L pipette, and the cellular debris was washed away with phosphate-buffered saline (PBS). Fresh serum-free DMEM was then added to each dish for subsequent culture. The wound healing progress was observed and photographed under an inverted microscope (IX83, Cell Sens, Olympus, Tokyo, Japan) at 12 h, and the rate of wound healing was calculated, which indicated migration ability. The cells were stained with the Diff-Quick kit (International Reagents, Hyogo, Japan). The number of cells that migrated towards the scratch were counted and evaluated as a percentage of the total number of cells. All experiments were repeated three times [28, 29].

In vitro migration was examined using 24-well Matrigel Invasion Chambers (Corning Discovery Labware, Bedford, MA, USA), which contained 8-$\mu$m-pore-sized polyethylene terephthalate membranes. First, DMEM was added to rehydrate the membrane for 2 h at 37 °C in a 5% CO${}_2$ humidified atmosphere. The DMEM was then carefully removed without disturbing the Matrigel on the membrane. SHED, MSC and hDF (control) were washed twice in PBS, and then resuspended at 1 $\times$ 10${}^5$ cells/mL. Then 0.5 mL of the cell suspension (5 $\times$ 10${}^4$ cells) was added to the upper insert. To examine the ability of SHED or other MSC to migrate towards glioma-conditioned medium (CM), the lower chamber was filled with 0.75 mL of CM from the U87 and U251 glioma cell lines. CM were obtained by collecting, centrifuging, and filtering media from U87 or U251 cells (1 $\times$ 10${}^6$), which were cultured in 10 mL serum-free DMEM for 48 h [15].

To determine the ability of SHED to migrate towards specific growth factors, the lower chamber was filled with 0.75 mL of DMEM and specific growth factors, including stem cell factor (SCF), platelet-derived growth factor BB (PDGF-BB), C-X-C Motif Chemokine Ligand 12 (CXCL12), and vascular endothelial growth factor (VEGF), (Abcam PLC, Tokyo, Japan), which were added at concentrations ranging from 0.1 to 100 ng/mL.

For experiments that involved blocking specific growth factors, U87 and U251 CM were treated with anti-SCF, anti-PDGF-BB, anti-CXCL12, or anti-VEGF neutralizing antibodies (Abcam PLC) at concentrations of 1 to 10 $\mu$g/mL, which were added into the lower chamber. Following incubation of the Matrigel invasion chambers for 24 h at 37 °C in 5% CO${}_2$, the non-invading cells and Matrigel were removed from the upper surface of the insert membranes with a cotton swab. Meanwhile, cells that had migrated to the lower surface of the membrane were stained with the Diff-Quick kit (International Reagents) by sequentially transferring the inserts to air dry. The number of cells that had migrated was counted in four high-power fields per membrane (200$\times$ magnification) using an inverted microscope (IX83, Cell Sens, Olympus) [15].

The following experiments were performed and supervised as per guidelines approved by the Animal Care Committee of the Hamamatsu University School of Medicine Animal Care Facility (approval number 2020089).

The mouse model of glioma was based on protocols used in previous reports [8, 30]. Briefly, female BALB/c slc nu/nu mice (18–22 g, 8–9-weeks-old, Nippon SLC, Shizuoka, Japan) were subcutaneously injected with an anesthetic mixture comprising 0.75 mg/kg medetomidine (Nippon Zenyaku Kogyo, Fukushima, Japan), 4.0 mg/kg midazolam (SANDOZ, Tokyo, Japan), and 5.0 mg/kg butorphanol (Meiji Seika Pharma, Tokyo, Japan). The human glioma cells U87-luc (1 $\times$ 10${}^5$), which emit luminescence when D-luciferin is administered into the abdominal cavity, were transplanted in 5 $\mu$L of PBS. Similarly, human astrocytes (1 $\times$ 10${}^5$) in 5 $\mu$L of PBS were injected into the right hemisphere of nude mice on day 0 as a control [9]. A burr hole was placed at the point 2.2-mm lateral for anterior-posterior to bregma. Then, a 10-$\mu$L micro-syringe (Hamilton Company, Reno, NV, USA) was inserted at a depth of 3.0 mm from the brain surface and was left in position for 1 min, and then withdrawn to a depth of 2.5 mm, after which cells were infused at a rate of 2 $\times$ 10${}^4$ cells/min by a stereotaxic guidance system [9, 31].

To monitor bioluminescence signals of U87-luc cells, XenoLight Rediject D-luciferin (Summit Pharmaceuticals International, Tokyo, Japan) was injected intraperitoneally at 150 mg/kg body weight, followed by subcutaneous injection of the anesthetic mixture. Then the mice were placed on a stage inside the camera box of the IVIS200 imaging system coupled with a cooled CCD camera (Caliper Life Sciences, Waltham, MA, USA) 20 min after D-luciferin injection. The detected light emitted from U87-luc cells was digitized and electronically displayed as a pseudo color overlay onto a grayscale image of the animal. Images and measurements of luminescence signals were analyzed with Living Image software version 3.0 (Caliper Life Sciences, Waltham, MA, USA) and quantified as photons per second [9]. Images and measurements of luminescence signals were acquired on day 5.

To evaluate the ability of SHED to target glioma cells in the brain and to understand the relationship between the number of SHED that accumulated around a glioma lesion and tumor size, we labeled SHED with ferucarbotran (Resovist® Inc., Fujifilm, Tokyo, Japan), in which SPIO was used to label the cell membrane with iron, and the number of SHED surrounding the lesion at the maximum cross-sectional area were counted. SHED were counted within the tumor and within 100 $\mu$m of the surrounding area. Counts were performed under a microscope (IX83, Cell Sens, Olympus). These data were then used to evaluate the relationship between tumor size and SHED migration. U87-luc cells (1 $\times$ 10${}^5$) in 5 $\mu$L of PBS or Human astrocytes (1 $\times$ 10${}^5$) in 5 $\mu$L of PBS (as control) were implanted in the right hemisphere on day 0 (n = 8). On the 5th day after transplantation, bioluminescence signal from the IVIS200 Imaging System confirmed tumor implantation (Supplementary Data 1). SPIO-labeled SHED (5 $\times$ 10${}^5$) in 5 $\mu$L of PBS or SPIO-labeled hDF (5 $\times$ 10${}^5$) in 5 $\mu$L of PBS (as control) were implanted into the left hemisphere of mice on day 7 (n = 4) or 14 (n = 4). On day 14 or 21 (after U87-luc implantation; 7 d after SHED implantation) the mice were sacrificed after the previously mentioned anesthesia was fully in effect by cardiac perfusion with PBS, and their brains were obtained, fixed in 4% paraformaldehyde in 0.1 M PBS, and stained for histologic examination with hematoxylin and eosin (H&E) or Prussian Blue [32].

For tumor sections from the SPIO experiment, while the Prussian Blue might stain macrophages, Qtracker probe (Thermo Fisher Scientific) was also used to evaluate the in vivo migratory ability of SHED. Malignant glioma cells (U87-luc) were transplanted into the right hemisphere of the mice (n = 3), and then SHED were transplanted into the contralateral hemisphere. U87-luc cells (1 $\times$ 10${}^5$) in 5 $\mu$L of PBS or PBS only (control) were implanted in the right hemisphere on day 0. On the 5th day after transplantation, bioluminescence signal obtained from the IVIS200 Imaging System confirmed tumor implantation. SHED (5 $\times$ 10${}^5$), labeled with the Qtracker 525 Cell Labeling Kit (Q25049, Thermo Fisher Scientific) following the manufacturer’s instructions in 5 $\mu$L of PBS, were implanted in the left hemisphere on day 7. On day 14 (after U87-luc implantation; 7 d after SHED implantation), the mice were sacrificed after the previously mentioned anesthesia was fully in effect by cardiac perfusion with PBS, and their brains were obtained for histological examination with H&E staining. The sections were then examined under a fluorescence microscope (IX83, Cell Sens, Olympus) [9].

To use SHED as a transporter for gene therapy in malignant glioma patients, we first assessed whether SHED had tumorigenic potential in vivo. SHED (5 $\times$ 10${}^5$ cells) in 5 $\mu$L of PBS or 5 $\mu$L of PBS alone (control) were implanted in the right hemisphere of female BALB/c slc nu/nu mice to observe alterations in weight and/or behavior. After 7 or 150 d, brain fragments were created, and immunostaining was performed with human anti-mitochondrial antibodies (rabbit monoclonal, Abcam plc, Tokyo, Japan) and secondary antibodies (goat anti-rabbit IgG H&L; Abcam plc, Tokyo, Japan). Tumor formation was evaluated by H&E staining [33].

The Excel Statistical Program File “ystat 2008.xls” (programmed by Shinya Yamazaki, DDS, Ph.D., Igaku-Tosho-Syuppan, Tokyo, Japan, ISBN4-87151-351-3) was employed for statistical analyses. All data are presented as mean $\pm$ standard deviation. p 0.05 was used to indicate significant differences. Enumeration data were analyzed using Friedman’s $\chi$${}^2$ test, and quantitative data were analyzed by repeated-measures analysis of variance (ANOVA), followed by Dunnett’s test or the Bonferroni method for post-hoc corrections.

The mobility of SHED was compared with those of hMSC-AT and hMSC-BM in scratch assays, with hDF used as a control. The percentages of cells that moved into the scratch range were calculated as follows: SHED, 31.2% $\pm$ 2.45%; hMSC-AT, 25.5% $\pm$ 2.35%; and hMSC-BM, 24.2% $\pm$ 1.22%. All cell lines exhibited significantly more movement than hDF (17.7% $\pm$ 1.59%), and SHED were significantly more mobilized than hMSC-BM (p 0.05) (Fig. 1). SHED showed greater mobility than hMSC-AT and hMSC-BM.

Fig. 1.

Verifying the mobility of stem cells from different sources. (A) Quantitative data of cell motility. All stem cells showed some evidence of increase motility, including SHED, hMSC-AT, and hMSC-BM, which were all significantly more mobile than hDF. Data are presented as mean $\pm$ SD; *p 0.05. (B) Representative images of scratch assays. The white lines indicate the range of each scratch wound; scale bar: 200 $\mu$m.

The migration ability of SHED towards the CM of malignant glioma cells was compared with those of hMSC-AT and hMSC-BM in transwell assays. hDF were used as a control. The numbers of migrating cells towards the CM of malignant glioma were calculated as follows: SHED, U87: 221.7 $\pm$ 13.1 and U251: 194.7 $\pm$ 14.8; hMSC-AT, U87: 149.3 $\pm$ 10.47 and U251: 153.7 $\pm$ 5.55; hMSC-BM, U87: 177.3$\pm$ 3.18 and U251: 171.3 $\pm$ 4.41; hDF, U87: 55.3 $\pm$ 2.19 and U251: 51.7 $\pm$ 0.67. All SHED, hMSC-AT, hMSC-BM cells migrated significantly more towards the CM of malignant glioma cell lines compared with hDF. Additionally, SHED migrated more significantly towards glioma CM compared with hMSC-AT (U87 and U251) and hMSC-BM (U87) (p 0.05) (Fig. 2).

Fig. 2.

In vitro migration to glioma conditioned medium. Induced migration of mesenchymal stem cells (SHED, hMSC-AT, and hMSC-BM) and hDF (control) in 24-well Matrigel chambers by conditioned medium (CM) from U87 and U251 human glioma cells. After 24 h, the number of cells that had migrated was observed under a microscope. (A) The total number of migrated cells in four high power fields. The number of migrated mesenchymal stem cells was significantly increased compared with that in hDF. Data represent the mean number of migrated cells $\pm$ SD from experiments run in triplicate; *p 0.05. (B) Representative photomicrographs of the micropore membranes of U87CM and U251CM; scale bar: 100 $\mu$m.

We also evaluated the migration of SHED to four specific growth factors with transwell assays. For all four growth factors, the number of migrating SHED cells increased in a concentration-dependent manner (Fig. 3).

Fig. 3.

In vitro migration of SHED towards growth factors secreted by gliomas. (A) Migration of SHED towards four specific growth factors (SCF, PDGF-BB, CXCL12, and VEGF) in a 24-well Matrigel chambers. The migration of SHED was significantly increased following stimulation with each growth factor in a concentration-dependent manner (concentrations of 0.1, 1, 10, and 100 ng/mL compared with that with Dulbecco’s modified Eagle’s medium [DMEM]). Data represent the mean number of migrated cells $\pm$ SD from experiments run in triplicate; *p 0.05. (B) Representative photomicrographs of the micropore membranes of SCF at concentrations from 0.1 to 100 ng/mL. The number of SHED on the membrane was increased depending on the concentration of SCF; scale bar: 100 $\mu$m.

The number of migrating SHED towards CM of both U87 and U251 cells (malignant glioma cell lines) was decreased in a concentration-dependent manner by adding growth factor-neutralizing antibodies. This suggested that SHED might be attracted to these growth factors secreted by malignant gliomas (Fig. 4).

Fig. 4.

SHED migration was inhibited by neutralizing antibodies against four growth factors. (A) The migration of SHED towards conditioned media (CM) of U87 and U251 cells was evaluated with inhibitory neutralizing monoclonal antibodies against the indicated growth factors at concentrations of 1 and 10 $\mu$g/mL. SHED migration was significantly attenuated by treatment with anti-SCF, anti-PDGF-BB, anti-CXCL12, and anti-VEGF antibodies. Data represent the mean number of migrated cells $\pm$ SD from experiments run in triplicate; *p 0.05. (B) Representative photomicrographs of the micropore membranes of U87CM and U251CM treated with anti-SCF antibody (0, 1, and 10 $\mu$g/mL); scale bar: 100 $\mu$m.

We evaluated the migration of SHED from the contralateral hemisphere to malignant glioma in the brains of nude mice. Brain sections harvested 7 d after SHED implantation showed that SHED accumulated around the tumor in the contralateral hemisphere of the brain (Fig. 5A). In contrast, hDF implanted as a control did not migrate to U87 tumors. Additionally, SHED did not migrate around human astrocytes transplanted as a control for U87 cells (Supplementary Data 2).

Fig. 5.

Tropism of SHED to glioma cells in vivo. (A) SPIO-labeled SHED were implanted into the left hemisphere of BALB/c nude mice (n = 8), and the number of SHED that accumulated around along the longest diameter of a section of U87-luc tumors previously implanted in the right hemisphere were counted. Representative Prussian Blue-stained images, in which SPIO-labeled SHED are blue. Hematoxylin and eosin (H&E) staining; scale bar: 1 mm. Prussian Blue-staining; scale bar: 100 $\mu$m. (B) The maximum tumor diameter and the number of SHED associated with the tumor were measured. Two slices with the largest tumor diameter were selected from each mouse. Tumor diameter was significantly correlated with the number of SHED (r = 0.74, p 0.05). (C) U87-luc were implanted in the right hemisphere of BALB/c nude mice on day 0, and Qtracker-labeled SHED were implanted in the left hemisphere on day 7; scale bar: 1 mm. (D) SHED were confirmed to have accumulated around the tumor after 7 d; scale bar: 200 $\mu$m.

By labeling SHED with SPIO, we could count the number of cells that had accumulated around the tumor. The number of SHED that accumulated was positively correlated with tumor size (r = 0.74, p 0.05) (Fig. 5B). Moreover, SHED labeled with Qtracker also migrated to the tumor area in the contralateral hemisphere (Fig. 5C,D).

SHED were implanted into the brains of nude mice, which were evaluated for tumor formation. The mice implanted with SHED showed the same trend of weight gain as the control group, and there were no obvious changes in their health conditions. SHED implanted in the mouse brain were confirmed to be present in the mouse brain after 7 d (Fig. 6A) and 150 d (Fig. 6B) with human mitochondrial antibodies. After 150 d of SHED transplantation, there was no evidence of SHED forming tumors in brain sections from the mice (Fig. 6B).

Fig. 6.

The situation of tumor formation after SHED implantation. Brain sections of mice at 7 d (A) and 150 d (B) after SHED implantation are presented. Upper column is hematoxylin and eosin (H&E) staining, and lower is immunostaining for anti-human mitochondrial antibodies. As shown in the H&E staining, SHED showed no malignant transformation or engraftment in the mouse brain. As shown by the immunostaining, SHED were still present in the mouse brain 150 d after transplantation. H&E staining; scale bar: 1 mm. Immunostaining for anti-human mitochondrial antibodies; scale bars: 100 $\mu$m.

MSC are self-renewing, pluripotent and have been identified in many human tissues, including bone marrow, fat, cord blood, amniotic fluid, and dental pulp. In recent studies, MSC have received particular attention in regenerative medicine and stem cell therapy [27]. Among the MSC, DPSC have recently been used for treatment as multipotent stem cells because they have neuroprotective and anti-inflammatory effects. DPSC are also being applied to treat trauma and stroke [34]. DPSC are also known to have high migratory capacity, and there have been reports on factors affecting DPSC migration [35]. SHED are a type of DPSC that were recently identified as an ideal cell type for regenerative medicine [36]. SHED are known to have no risk of tumorigenesis, high proliferative potential, multipotency, and a low risk of immunorejection [37, 38, 39]. SHED have also been investigated for treating neonatal hypoxic brain injury [40], ulcerative colitis, spinal cord injury, and Parkinson’s disease [41]. SHED can be managed by cryopreservation [36], and SHED banks have been established for autologous and allogeneic cell therapies [38, 39, 42].

Glioma microenvironment plays an essential role in controlling tumor growth and spreading. This microenvironment is made up of different cell types: tissue-resident cells, myeloid cells, bone marrow-derived cells, immune cells, endothelial cells, pericytes, and fibroblasts. In addition, glioma cells produce cytokines, chemokines, and growth factors to form tumor microenvironment [43, 44, 45].

There have been many reports on the migration of MSC to gliomas, and it is likely that various factors secreted by gliomas affect the migration of MSC to this disease [17, 46]. In this study, SHED showed significant migratory ability to the CM of two human glioma cell lines and to four specific growth factors secreted by glioma. In blocking experiments performed on CM of human glioma cells with inhibitory monoclonal antibodies against the four growth factors, SHED migration to the CM of U87 and U251 cells was significantly inhibited by all four neutralizing antibodies, indicating that these four specific growth factors significantly affect the migration of SHED.

Previously, in vitro migration assays have reported that NSC and MSC migrate to specific growth factors, particularly SCF, PDGF-BB [16], CXCL12, and VEGF [47, 48]. We also found that iPS cells recognized a variety of factors secreted by malignant glioma cells as triggers for migration [15]. Hence, SHED could likely serve as carriers for gene therapy treatments for malignant gliomas if they are genetically modified to express therapeutic transgenes.

Regarding future studies, we are in the process of verifying the receptors for the four specific growth factors (SCF, PDGF-BB, CXCL12, and VEGF), and comparing the levels of their expression between SHED and other MSCs. Additionally, we also need to confirm other factors for migration into glioma.

In vivo, SHED accumulated around tumors implanted in the contralateral hemisphere. Moreover, the number of SHED that migrated to the tumor was increased with tumor size.

NSC and MSC are excellent carriers for gene transfer into gliomas, and thus could be used to transport localized gene therapy to residual tumor cells after surgical removal, and several studies have shown that this is true for NSC, MSC, iPS cells, and Muse cells [7, 8, 9, 15]. In rodent experiments, intratumoral injection of NSC or Muse cells expressing the TK gene followed by systemic GCV administration has shown therapeutic potential. However, ethical and practical issues with transporter collection, as well as the safety and stability of the vectors remain problems that need to be resolved. TK-transduced bone marrow cells or MSC-BM also demonstrate the bystander effect [7]; however, they also contain miscellaneous factors and lack stability [49].

Murakami et al. [50] reported the self-migrating ability of MSC harvested from the same individual dog and found that DPSC showed higher migratory ability than MSC-AT and MSC-BM. Both SHED and DPSC are derived from neural crest cells that migrate throughout the body during development to form tissues [51]. Neural crest cells have been found to have excellent migratory ability [52]. Therefore, we consider that SHED and DPSC derived from neural crest cells show higher migratory ability than MSC-AT and MSC-BM derived from mesoderm. In a study comparing SHED and DPSC, Wang et al. [53] found no significant difference in the migratory capacity of SHED and DPSC. Similarly, the age of the cell donor may affect the migration of the harvested stem cells.

SHED can be collected from deciduous teeth; hence, they have no invasiveness to the human body and no ethical issues are involved in their collection. Therefore, SHED are likely to be more ideal in terms of collection. Furthermore, dental pulp containing SHED are protected by hard teeth, and their DNA is not easily damaged; thus it remains stable [54]. SHED are therefore a potential practical candidate for a stem cell-based suicide gene therapy for glioma. With respect to the effects of SHED on the normal brain, histological sections of mouse brains 150 d after SHED implantation showed no tumorigenic tissue or remnant SHED, indicating that the safety of SHED transplantation.

In this study, we confirmed that SHED is a potential candidate carrier to deliver genes to glioma cells for gene therapy. However, there are several limitations to our study. We evaluated only four specific growth factors. The study used a nude mice model, which lacks the effect on the immune system. Furthermore, we did not investigate the migratory behavior of SHED in detail. In the future, the following issues should be further investigated: the relationship between the migratory ability of SHED and other triggering factors and receptors; the immune system and inflammation; migration pathways; and the time required for migration.