Allogeneic transplantation of mobilized dental pulp stem cells with the mismatched dog leukocyte antigen type is safe and efficacious for total pulp regeneration

Background We recently demonstrated that autologous transplantation of mobilized dental pulp stem cells (MDPSCs) was a safe and efficacious potential therapy for total pulp regeneration in a clinical study. The autologous MDPSCs, however, have some limitations to overcome, such as limited availability of discarded teeth from older patients. In the present study, we investigated whether MDPSCs can be used for allogeneic applications to expand their therapeutic use. Methods Analysis of dog leukocyte antigen (DLA) was performed using polymerase chain reaction from blood. Canine allogeneic MDPSCs with the matched and mismatched DLA were transplanted with granulocyte-colony stimulating factor in collagen into pulpectomized teeth respectively (n = 7, each). Results were evaluated by hematoxylin and eosin staining, Masson trichrome staining, PGP9.5 immunostaining, and BS-1 lectin immunostaining performed 12 weeks after transplantation. The MDPSCs of the same DLA used in the first transplantation were further transplanted into another pulpectomized tooth and evaluated 12 weeks after transplantation. Results There was no evidence of toxicity or adverse events of the allogeneic transplantation of the MDPSCs with the mismatched DLA. No adverse event of dual transplantation of the MDPSCs with the matched and mismatched DLA was observed. Regenerated pulp tissues including neovascularization and neuronal extension were quantitatively and qualitatively similar at 12 weeks in both matched and mismatched DLA transplants. Regenerated pulp tissue was similarly observed in the dual transplantation as in the single transplantation of MDPSCs both with the matched and mismatched DLA. Conclusions Dual allogeneic transplantation of MDPSCs with the mismatched DLA is a safe and efficacious method for total pulp regeneration. Electronic supplementary material The online version of this article (10.1186/s13287-018-0855-8) contains supplementary material, which is available to authorized users.


Background
Pulp/dentin complex in teeth has a critical function in the maintenance of tooth homeostasis, and viable pulp is essential for the longevity of the tooth [1]. The ultimate goal for regenerative endodontics is to replace or restore the impaired or damaged tissues with viable pulp tissue in the case of pulpitis or apical periodontitis, leading to the reestablishment of the physiologic, structural, and mechanical integrity of the native dentin/pulp complex, including function of pulp immunity, dentin formation, pulp innervation, and vascular perfusion [1][2][3][4]. Stem cell therapy has been suggested as an effective regenerative technique for pulpitis and apical periodontitis. Autologous transplantation of dental pulp stem cell (DPSC) subsets, dental pulp-derived CD31 − side population (SP) cells, or CD105 + cells with SDF1 in orthotopic sites demonstrated complete pulp regeneration [5,6]. Subsequently, colony-derived DPSCs with platelet-rich plasma/fibrin (PRP/PRF) showed similar successful results [7]. The safety and efficacy of autologous DPSC therapy were demonstrated in the preclinical study harnessing DPSCs mobilized (MDPSCs) with granulocytecolony stimulating factor (G-CSF) harvested in good manufacturing practice conditions [8]. Furthermore, a recent clinical study suggested autologous MDPSC transplantation may be safe and efficacious for pulp regeneration in humans [9]. The autologous DPSCs, however, have certain limitations to overcome, such as limited availability of human discarded teeth and the high cost of the safety and quality control tests of individual cell products before transplantation. The further potential disadvantages of the autologous mesenchymal stem cells (MSCs) are their decreased biological activity from older patients and altered intrinsic stem cell properties from patients with some systemic diseases including diabetes and rheumatoid arthritis [10]. Transplantation of autologous mobilized adiposederived or bone marrow-derived MSCs resulted in lower volume of regenerated pulp tissue, less angiogenesis, and reinnervation compared with MDPSCs [11]. Furthermore, regenerated pulp tissues in adipose and bone marrow MSC transplants were more mineralized compared with MDPSC transplant, suggesting pulp MSCs were an optimal cell source for pulp regeneration. Thus, banked allogeneic DPSCs would be highly advantageous to save time and costs and to confirm high quality [12].
The use of allogeneic MSCs permits low immunogenicity with immunomodulatory and immunosuppressive properties. It is well known that MSCs have low immunogenicity due to no expression of class II major histocompatibility complex (MHC-II) proteins, and low or modest expression of MHC-I proteins and costimulatory molecules such as CD40, CD80, and CD86 on their cell surface [13,14]. Therefore, MSCs are unable to provoke a cytotoxic effect by allogeneic immune cells [15], and MSCs from MHC-mismatched donors may also be used for cell therapy [16]. Many studies recently focused on mechanisms of immunomodulation and immunosuppression of MSCs, especially in reducing inflammation, escaping from immune cell response, and modulating Tcell proliferation. MSCs can interfere with different pathways of the immune responses by means of direct cell-to-cell interactions and secretion of soluble factors such as transforming growth factor (TGF)-β, hepatocyte growth factor (HGF), prostaglandin E2 (PGE2), nitric oxide (NO), indoleamine-2,3 dioxygenase (IDO), tumor necrosis factor (TNF)-α stimulated gene-protein 6, interleukin (IL)-6, IL-10, semaphorin-3A, galectin (Gal)-1, and Gal-9 [17][18][19][20]. MSCs also possess the ability to generate regulatory T cells (Tregs) which suppress other immune cells [21,22]. The whole range of mechanisms of immunomodulation and immunosuppression mediated by MSCs remains incompletely understood. The immunosuppressive and immunomodulatory responses are, however, properties shared by MSCs from a variety of adult and fetal tissues including dental pulp [18,23].
A number of animal experiments have demonstrated that allogeneic MSCs improve acute myocardial infarction, chronic spinal cord injury, ischemic stroke, fracture healing, and osteoarthritis by local injection or intravenous infusion [24][25][26][27][28]. No adverse effects have been noticed in 291 equine recipients over a period of up to 1 year after intravenous injection of allogeneic peripheral blood-derived MSCs [29]. On the other hand, MHC class I mismatched MSCs induced CD8 +VE , CD16 +VE , and CD8 +VE /CD16 +VE lymphocyte subpopulations, dependent on the dose of administered MSCs in intracranial injection and the degree of antigenic mismatch between donor and recipient [30]. Allogeneic bone marrow MSC transplantation into infarcted rat myocardium improved ventricular function for 3 months and a delayed immune rejection/response has been reported within 5 months due to the shift from a hypoimmunogenic to an immunogenic state of the transplanted MSCs upon differentiation [31]. Thus, consistent results have not yet been obtained on the therapeutic effects of allogeneic MSCs, depending on routes, timing duration, and dosage of MSC administration in vivo [10].
It is desirable for MSC transplantation into the root canal of the tooth to be repeated for pulp regeneration in patients with multiple caries. Repeated injection of allogeneic adipose tissue-derived MSCs (AT-MSCs) or bone marrow-derived MSCs (BM-MSCs) demonstrated a lack of adverse effects. However, repeated BM-MSC injections resulted in an increase in blood CD8 + T-cell numbers and splenic regulatory T-cell numbers compared with AT-MSCs in healthy horses, indicating a mild alloantigen-directed cytotoxic response [32]. Repeated intravenous injection of allogeneic porcine bone marrow MSCs or human umbilical cord blood-derived MSCs also induced no immunological alterations including T-cell proliferation, high levels of IFN-γ, TNF-α, and human IgG and no adverse events due to low immunogenicity [33,34]. Dual allogeneic MSC treatment by transepicardial injection in the acute and the subacute period after myocardial infarction improved ventricular function with increased myocardial mass and anteriolar density more than single MSC treatment in rats [35].
There are no reports on MHC-mismatched allogeneic transplantation and on dual allogeneic MSC transplantation in pulp regenerative therapy. Thus, there is a need to address challenges to allogeneic MDPSC cell therapy for total pulp regeneration. The aim of this investigation is to assess the safety and efficacy of allogeneic transplantation of canine MDPSCs into the pulpectomized tooth. The dual consecutive transplantation was further evaluated for safety and efficacy compared with the single transplantation.

Dog leukocyte antigen genotyping and matching analysis
We used beagle dogs (Kitayama Labes, Iwakuni and Ina, Japan) owned by Shin Nippon Kagaku Biomedical Laboratories Ltd (n = 26). Total genomic DNA was extracted from whole blood of dogs by NuclesaseMag®96 Blood (Marcherey-Nagel, Düren, Germany) according to the protocol. All of the dogs in the investigation were not in a sibling relationship. Genotyping was performed by direct sequencing and sequencing of the polymerase chain reaction (PCR) product. PCR was performed using primers DLA-88 exon 1-3 (1100 bp), DLA-DQA exon 2 (300 bp), DLA-DQB exon 2 (350 bp), and DLA-DRB exon 2 (350 bp) [36,37] (Table 1) with KOD Fx (TOYOBO Co., Ltd, Osaka, Japan) in a GeneAmp PCR system 9700 (Thermo Fisher Scientific K.K., Yokohama, Japan). PCR products were subcloned into a ZeroBlunt®-TOPO PCR Cloning Kit (Thermo Fisher Scientific K.K.). Sequencing was carried out using a ABI PRISM BigDye Terminator v3.0 Ready Reaction Cycle Sequencing Kit (Thermo Fisher Scientific K.K.) with an ABI PRISM 3730 DNA Analyzer (Thermo Fisher Scientific K.K.), and the raw data were analyzed by Sequencer Ver 4.8 (Gene Codes Corp., Ann Arbor, MI, USA). The allele names were determined according to the universal nomenclature found in the Immuno Polymorphism Database (EMBL-EBI, Cambridge, UK).

Cell isolation and culture
Upper left lateral incisors were freshly extracted from each beagle dog at 8 months of age. After making a longitudinal cut, they were transported by air within 30 h to the National Center for Geriatrics and Gerontology (NCGG) from Shin Nippon Biomedical Laboratories Ltd, Drug Safety Research Laboratories (Kagoshima, Japan), and RaQualia Pharma Inc. (Rental Laboratories of NCGG) in Hank's balanced salt solution (Invitrogen, Carlsbad, CA, USA) with 2.5 mg/ml amphotericin B (Bristol-Myers Squibb, Tokyo, Japan) and 0.3% gentamicin (Nitten, Nagoya, Japan). Mobilized dental pulp stem cells (MDPSCs) were isolated using the similar procedure used for a previous autologous preclinical study [8]. In brief, dental pulp tissues isolated from the teeth were enzymatically digested in 0.04 mg/ml Liberase (Roche, Mannheim, Germany) for 30 min at 37°C. The isolated pulp cells were plated at 2 × 10 4 cells on T-25 (Asahi Technoglass, Funabashi, Japan) in Dulbecco's Modified Eagle's Medium (DMEM) (Sigma, St. Louis, MO, USA) supplemented with 10% autologous canine serum, 2.5 mg/ml amphotericin B, and 0.3% gentamicin. They were detached by incubation with TrypLE™ Select (Invitrogen) prior to 70% confluence. The colony-formed DPSCs were further isolated by G-CSF-induced stem cell mobilization method with G-CSF (Neutrogin; Chugai Pharmaceutical Co., Ltd, Tokyo, Japan) at 100 ng/ml, with 2 × 10 4 cells/100 μl on the Transwell (Corning, Lowell, MA, USA), and inserted into 24-well tissue culture plates for 48-h incubation [38]. The isolated MDPSCs were detached by incubation with Try-pLE™ Select at 60-70% confluence and subcultured at a 1: 3 dilution into cell culture flasks (25 cm 2 and further 75 cm 2 ) (Asahi Technoglass) in DMEM supplemented with 10% autologous canine serum without antibiotics. The cells were cryopreserved at a cell concentration of 1 × 10 6 cells/ml in an extracellular cryoprotectant (CP-1; Kyokuto Pharmaceutical Industrial Co., Ltd, Tokyo, Japan) at the 7th passage of culture.

Characterization of mobilized dental pulp stem cells
The quality of cryopreserved MDPSCs was confirmed by cell viability and proliferation abilities at the 7th passage To further characterize the immunomodulation ability of MDPSCs, mobilized adipose-derived stem cells (MADSCs) were isolated from the abdominal subcutaneous adipose tissue [11] from the same individual dog similarly to MDPSCs as already described for comparison with MDPSCs. Total RNA was extracted with TRIzol (Life Technologies). First-strand cDNA syntheses were performed on the total RNA of these cells by reverse transcription using ReverTra Ace-a (Toyobo, Tokyo, Japan) after DNase I treatment (Roche Diagnostics) at 37°C for 20 min. Real-time RT-PCR was performed using primers for the immunomodulatory factors (Table 2) prostaglandin E synthase (PTGES), cyclooxygenase-2 (COX-2), IL-6, TGF-β and indoleamine 2, 3-dioxygenase-1 (IDO-1) labeled with AmpliTaq Gold master mix (Thermo Fisher Scientific) in an Applied Biosystems® 7500 Real-Time PCR (Life Technologies). After normalizing with β-actin, the mRNA level of each immunomodulatory factor in MDPSCs was compared with that in MADSCs derived from the same individual dog (n = 3).

Evaluation of safety of first and second allogeneic transplantation for pulp regeneration
The cryopreserved MDPSCs were transported by air to the operating room of the animal facility in Shin Nippon Biomedical Laboratories Ltd under strict temperature control. One day before transplantation, the root canal was open to the apex with a #25 K file after pulpectomy and shaped to 0.55 mm in width, 0.5 mm from the apex in the upper right lateral incisors in 14 dogs. For the first allogeneic transplantation, the dog leukocyte antigen (DLA) matched and mismatched MDPSCs (n = 5, respectively) were transplanted into the root canal, 5 × 10 5 cells together with 20 μl of collagen scaffold (atelocollagen implant; Koken, Tokyo, Japan) and 150 ng of G-CSF (Neutrogin), respectively (Fig. 1). For toxicology assessment, clinical signs of dogs were observed and their food consumption was measured daily, and their weights were recorded weekly. Urine chemistry examinations by Clinitek AtlasXL (Sparton Medical Systems, Strongsville, OH, USA) were performed at 4 and 12 weeks, and blood tests by ADIVIA 120 (Siemens Healthcare Diagnostics Manufacturing Ltd, Erlangen, Germany) and blood chemistry examinations by JCA-BM6070 (Japan Electron Optical Laboratory, Tokyo, Japan) were performed at 4 and 12 weeks after transplantation. Blood tests demonstrate the red blood cell count (RBC) and hematocrit (Ht) for homeostasis of blood cells, the platelet count (Plt) for inflammation, and the white blood cell count (WBC) for inflammation and infection. Blood chemistry examinations demonstrate aspartate transaminase (AST) and alanine transaminase (ALT) for abnormality of the liver, albumin and globulin for protein metabolism, total cholesterol (Tcho) for the lipid profile, and glucose for abnormality of hormone. After extraction of the upper right lateral incisors at 12 weeks, the second allogeneic transplantation of the same matched and mismatched MDPSCs as the first transplantation was performed in pulpectomized lower right third incisors in the same dogs, respectively (n = 5). The safety tests were further performed at 4 and 12 weeks. The transplanted teeth were extracted at 24 weeks followed by euthanization. All organs were weighed and macroscopically examined. Furthermore, histopathological examination of all organs and tissues including the transplanted teeth with surrounding periodontal tissue were also   For neovascularization and innervations analyses, 5μm-thick paraffin sections were deparaffinized and stained with Fluorescein Griffonia (Bandeiraea) Simplicifolia Lectin 1/fluorescein-galanthus nivalis (snowdrop) lectin (BS-1 lectin) (20 μg/ml; Vector laboratories, Inc., Youngstown, OH, USA) and anti-PGP9.5 (Ultra Clone) (1:10,000), respectively, as described previously [6]. The ratios of the BS-1 lectin-positive newly formed capillaries were calculated respectively using a Dynamic cell

Statistical analyses
Data are reported as mean ± SD. P values were calculated using Tukey's multiple comparison test method in SPSS 21.0 (IBM, Armonk, NY, USA).

DLA analysis
DLA genotyping and matching analyses in 26 dogs demonstrated a four homozygous allele profile (nine dogs), a three homozygous and one heterozygous allele profile (three dogs), a two homozygous and two heterozygous allele profile (four dogs), a one homozygous and three heterozygous allele profile (one dog), and a four heterozygous allele profile (nine dogs). In the four homozygous allele profile group, eight dogs had eight completely matched alleles (Group A) out of nine dogs. In the two homozygous and two heterozygous allele profile group, four dogs had seven matched alleles. In the four heterozygous haplotype group, four dogs had seven matched alleles (Group B) out of nine dogs (Table 3). We selected five identical and almost identical donors of the allele profiles (four dogs from Group A, one dog from Group B) and five nonidentical donors with at least four mismatched alleles for allogeneic transplantation (Table 4).

The isolated canine MDPSCs
The isolated and cryopreserved MDPSCs at the 7th passage of culture were stellate with short processes or spindle-shaped. The cell viability was more than 90% following thawing of the frozen cells. The doubling time was approximately 30 h as previously isolated from canine teeth transported by land within 1 h [9], suggesting that the transportation of the extracted teeth by air within 30 h did not affect the cell proliferation ability. The mRNA expression levels of PTGES, COX-2, IL-6, TGF-β, and IDO-1 were similar in MDPSCs and MADSCs derived from the same individual dog (Table 5), suggesting similar immunomodulatory/immunosuppressive function of MDPSCs to MADSCs.

Safety of allogeneic transplantation
Toxicology assessment showed no adverse effects on appearance, clinical signs, food consumption, and body weight for 12 weeks after allogeneic first transplantation of the MDPSCs from four DLA-nonidentical donors as well as those from three DLA-identical and one almost DLA-identical donors. The blood test demonstrated no increase of white blood cell and platelet numbers (Table 6), demonstrating no alloreaction toward the transplanted cells. Serum and urine chemistry parameters showed values within normal ranges at 4 and 12 weeks after both first and second allogeneic transplantation (Table 6). Furthermore, there was also no evidence of toxicity or adverse events at 4 and 12 weeks after second DLA-nonidentical and DLA-identical transplantation of the same type of MDPSCs as in the first transplantation. No abnormalities were caused by the allogeneic transplantation in any organ or tissues assessed by histopathological examinations at 12 weeks after the second transplantation. These results demonstrate that DLA mismatched transplantation might be safe for pulp regeneration for 12 weeks in dogs not only the first time but also the second time.

Pulp regeneration after allogeneic transplantation
We next compared the regenerated tissue after DLA mismatched MDPSC transplantation with matched MDPSC transplantation into the pulpectomized root canal (Fig. 2). Pulp-like tissues with well-developed vasculature were observed at 12 weeks in both the allogeneic first transplants (Fig. 2a, b, d, e and Additional file 1: Figure S1). Similar pulp-like tissues in cell morphology, cell density, and architecture of the extracellular matrix with a few inflammatory cells were also regenerated at 12 weeks in both the DLA mismatched and matched second transplants (Fig. 2g, h, j, k). Odontoblast-like cells with a long process and osteodentinoblast-like cells were attached to the wall of newly formed osteo/tubular dentin (Fig. 2c, f, i, m) and neither inflammation nor internal/external resorption was detected (Fig. 2j, n). The regenerated tissue was filled in the root canal more than 80% to the dentin-enamel junction in all four transplants. The statistical analysis demonstrated no difference among the four transplants (Fig. 2o). Vascularization demonstrated by BS-1 lectin staining was also similar in density and orientation in the four groups of transplants (Fig. 3a-d), and there was no significant differences in the capillary density among them (Fig. 3e).
Nerve fibers stained by PGP9.5 antibody were similarly observed, indicating similar reinnervation (Fig. 3f-i). Dentin-like mineralized tissue formation was similarly observed along the dentinal wall (Fig. 3j-m). There was no significant difference in the matrix formation among the four transplants by morphometric analysis of the Masson trichrome positively stained area (Fig. 3n). These results suggest no qualitative and quantitative differences in the regenerated tissue between the DLA mismatched and matched transplants and between first and second transplants.

Discussion
The aim of the present investigation was to evaluate the safety and efficacy of allogeneic transplantation of DLA matched and mismatched MDPSCs in pulpectomized teeth with complete apical closure for pulp/dentin regeneration in dogs. A crucial challenge, however, is the limitation of genotyping major histocompatibility with relevance for humans using animal models in a preclinical study. The dog is a suitable animal model for pulp regenerative therapy, where the incisor tooth and its dental pulp tissue are similar in anatomy and developmental biology to humans [39]. Tissue regeneration in the dog may also be influenced by similar factors as in humans, including the immune system [40] and genetics [41]. Thus, dogs have served as an effective, directly translatable model for MSC transplantation [42,43]. Major histocompatibility complex (MHC) genes in mammals include class I and class II genes that are highly polymorphic and their donor-recipient matching is important for cell transplantation [43]. The genes for the dog MHC or DLA have been defined as a sequencebased nomenclature [44,45]. There are four complete class I genes: DLA-88, DLA-12, DLA-64, and DLA-79, in which DLA-88 is highly polymorphic (more than 72 alleles) and the others are less polymorphic. In the class II region, there are DLA-DQA1 (nine alleles), DLA-DQB1 (20 alleles), DLA-DRB1 (at least 24 alleles), and DLA-DRA (monomorphic) [43,46]. In the present work, DLA genotyping and matching analysis were performed in 26 dogs by PCR for the four genes orthologous to the human genes including DLA-88, DLA-DQA, DLA-DQB, and DLA-DRB [47]. The results demonstrated eight dogs with four completely matched alleles (four homozygous allele profile), four dogs with three completely matched alleles (two homozygous and two heterozygous allele profile), and four dogs with three completely matched alleles (four heterozygous allele profile). The similarity of the allele profiles was not due to related siblings, and all of the MDPSCs were transplanted into unrelated recipients. Based on an analysis of canine DLA diversity, the three-locus DLA haplotype, DQA1 00101;DQB1 00201; DRB1 00102 represented 40.3% and DQA1 00101;DQB1 00201;DRB1 00101 represented 11.9% in the beagle [37], which is higher than the present rate: eight dogs out of 26 (30.7%) and only one dog out of 26 (3.8%), respectively, in the present study, suggesting the possibility of a distinct breed. The true extent of diversity of DLA genes in canines, especially of the class I gene DLA-88, however, is still unknown [43]. A variety of animal and human studies have demonstrated that stem cell-based therapy with allogeneic MSCs is a potential therapeutic option to regenerate damaged tissue/organ [10]. The low immunogenicity and immunomodulatory/immunosuppressive properties of allogeneic MSCs may contribute to a reduced immune response [48]. We have previously demonstrated lack of expression of MHC class II and costimulatory molecules, such as CD40, CD80 (B7-1), and CD86 (B7-2), although MHC-I is expressed in human MDPSCs [38]. We have also demonstrated that human and canine MDPSC conditioned medium inhibits allogeneic peripheral blood mononuclear cell (PBMC) proliferation and  MPDSC mobilized dental pulp stem cell, MADSC mobilized adipose derived stem cell, PTGES prostaglandin E synthase, COX-2 cyclooxygenase-2, IL interleukin, TGF transforming growth factor, IDO-1 indoleamine 2,3-dioxygenase 1  demonstrates a dose-dependent inhibition of PBMC immune responses in mixed lymphocyte reaction (MLR) assays [8,38], confirming the work in DPSCs [49]. IFN-γ secreted by activated PBMCs induce the expression of soluble factors by DPSCs, which may play an important role in the immunosuppressive process [50]. Expression of PGE2, TGF-β, indolamine-2, 3-dioxygenase-1 (IDO1), IL-6, IL-10, and COX2 triggers the immunosuppressive activity of DPSCs [51][52][53][54][55]. The present study demonstrated that genes related to immunomodulation   including prostaglandin E synthase (PTGES), COX-2, IL-6, TGF-β, and IDO1 were similarly expressed in canine MDPSCs compared to canine mobilized adipose-derived MSCs. Fas ligand (FasL) associated with the immunoregulatory properties of DPSCs in the context of inducing T-cell apoptosis [56] was also expressed in MDPSCs (data not shown). The present results are consistent with the previous studies in human MSCs [57][58][59], demonstrating low immunogenicity of MDPSCs and potential to induce immune tolerance in the hosts. Furthermore, MSCs can inhibit the immune response not only by suppressing T cells and by inducing Tregs but also by converting macrophages into a regulatory phenotype [60]. Polarization of macrophages toward the anti-inflammatory M2 phenotype has been reported after DPSC transplantation in diabetic peripheral nerves [61]. In addition, similar findings were seen by injection of conditioned medium from stem cells of human exfoliated deciduous teeth (SHED) into acute lung injury [62], autoimmune encephalomyelitis [63], and rheumatoid arthritis [64]. Further investigation is necessary to examine the shift in the macrophage phenotype from M1 to M2 in the regenerated pulp and periapical tissue after allogeneic DPSC transplantation.
MSCs express detectable levels of MHC class I and low levels of MHC class II to avoid recognition by a host immune system in allogeneic therapies [65,66]. Several animal and clinical studies, however, have demonstrated that MSCs are weakly immunogenic in vivo in the case of transplantation across MHC class I barriers and that MSCs are rejected to induce chronic immune responses [31,[67][68][69]. Such responses could restrict the effectiveness of repeated transplantation of allogeneic MSCs [69]. Syngeneic and minor mismatched transplantation of synovial MSCs demonstrated more optimal meniscus regeneration compared with major mismatched transplantation in a meniscectomized model [70]. On the other hand, the correlation between the number of donorhost MHC mismatches and the efficacy of treatment was not detected in local injection in osteoarthritis and degenerative disc disease [71]. Furthermore, osteogenically differentiated MSCs are immunomodulatory and lack immunogenicity, demonstrating potential use in bone repair. However, maintenance of these properties in vivo is still open to question since immunogenic markers are upregulated after transplantation of the differentiated MSCs [72]. In the present study there was no correlation between the number of donor-host mismatches and efficacy, demonstrating the lack of immune response. This may be due to reduced host immune responses by the transplanted MDPSCs and effective confinement of these cells into the root canal of the tooth. Another possibility is that the transplanted MDPSCs are not differentiated into any host cells in the regenerated pulp tissue. Our previous study demonstrated that injected DiI-labeled autologous MDPSCs did not differentiate into host cells and induced pulp regeneration by secreting trophic factors to elicit migration and proliferation and inhibit apoptosis of endogenous MSCs [8]. We further demonstrated that transplanted porcine pulp MSCs were not directly incorporated into endothelial cells, neuronal cells, or host pulp cells in mouse ectopic tooth transplantation models [6].
Although MSCs are known to be immunoprivileged, repeated transplantation of mismatched MSCs has been reported to lead to alloimmunization and subsequent refractoriness in mice [67,73]. Multiple intravenous injections of allogeneic MSCs are well tolerated in healthy horses, indicating no clinical signs of organ toxicity and systemic inflammatory response. A mild alloantigeninduced cytotoxic response, however, is suggested by an increased numbers of circulating CD8 + T cells [32]. Dual injection of allogeneic MSCs into joint and articular cartilage induces an adverse clinical response, suggesting immune recognition of allogeneic MSCs after the second infection [74,75]. Allogeneic MSCs are weakly immunogenic when transplanted across MHC boundaries in rhesus monkeys, indicating negative impact by dual transplantation [76], while repeated intravenous injection of human umbilical cord bloodderived MSCs has low immunogenicity and no adverse events detected in mice [34] and humans [77]. Furthermore, there are no toxicological abnormalities and no obvious pathological changes although CD3 + and IL-6 levels are significantly increased after repeated intravenous injection of monkey umbilical cord MSCs [78]. In the present study, there were no toxicological abnormalities and no significant difference in tissue volume of regenerated dental pulp and inflammatory cell infiltration between the first and second transplants for both (See figure on previous page.) Fig. 3 Histochemical analyses of regenerated pulp tissue. Immunostaining with (a-d) BS-1 lectin, (f-i) PGP 9.5 (arrows indicate neurite extension), and (j-m) Masson trichrome staining (hatched lines indicate newly formed osteo/tubular dentin (d)). a, b, f, g, j, k First transplantation. c, d, h, i, l, m Second transplantation. a, c, f, h, j, l DLA matched transplant. b, d, g, i, k, m DLA mismatched transplant. e Ratio of positively stained area by BS-1 lectin in a frame 310 μm × 240 μm in size. Data are mean ± standard deviation (n = 4). Tukey's multiple comparison test method. n Ratio of Masson trichrome positively stained area to pulp regenerated area. Data are mean ± standard deviation (n = 3). Tukey's multiple comparison test method DLA matched and mismatched MDPSCs. This result demonstrates that MDPSCs are immunologically safe for use in allogeneic applications.

Conclusion
In this preclinical study, the safety of allogeneic mismatched MDPSC transplantation in pulpectomized teeth was demonstrated. Regenerated pulp tissues including neovascularization and neuronal extension were similar in the DLA mismatched transplants compared to the DLA matched transplants even after dual transplantation of MDPSCs, suggesting efficacy for total pulp regeneration.