Stem cells from human exfoliated deciduous teeth attenuate mechanical allodynia in mice through distinct from the siglec-9/MCP-1-mediated tissue-repairing mechanism

The effects of stem cells from human exfoliated deciduous teeth (SHED) on mechanical allodynia were examined in mice. A single intravenous injection of SHED and conditioned medium from SHED (SHED-CM) through the left external jugular vein significantly reversed the established mechanical allodynia induced by spinal nerve transection at 6 days after injection. SHED or SHED-CM significantly decreased the mean numbers of activating transcription factor 3-positive neurons and macrophages in the ipsilateral side of the dorsal root ganglion (DRG) at 20 days after spinal nerve transection. SHED or SHED-CM also suppressed activation of microglia and astrocytes in the ipsilateral side of the dorsal spinal cord. A single intravenous injection of secreted ectodomain of sialic acid-binding Ig-like lectin-9 and monocyte chemoattractant protein-1 had no effect on the established mechanical allodynia, whereas a single intravenous injection of protein component(s) contained in SHED-CM with molecular weight of between 30 and 50 kDa reversed the pain. Therefore, it may be concluded that protein component(s) with molecular mass of 30–50 kDa secreted by SHED could protect and/or repair DRG neurons damaged by nerve transection, thereby ameliorating mechanical allodynia.

www.nature.com/scientificreports/ the peak of EAE through direct modulation of M2 macrophage polarization 15 . ED-Siglec-9 is a major component of SHED-CM, while it is barely detectable in bone marrow mesenchymal stem cell (BMMSC)-CM. Therefore, it is considered that mesenchymal stem cell-derived secreted factors directly convert the proinflammatory conditions prevalent in the damaged neurons to tissue-repairing ones by modulating the microglia/macrophage phenotype. Neuropathic pain is a debilitating symptom, which is caused by tissue or nerve damage. Neuropathic pain is characterized by severe pain produced by light touch. Accumulating evidence indicates that microglia in the spinal dorsal horn are highly activated following peripheral nerve injury 16,17 . Notably, daily administration of morphine, known as a powerful painkiller, leads to the activation of spinal microglia and causes hyperalgesia 18 . Currently available analgesics are not sufficiently treated neuropathic pain. A recent study revealed that macrophages in the dorsal root ganglion (DRG) are also required in the initiation and maintenance of neuropathic pain 19 . Thus, the importance of activated macrophages/microglia has been implicated in intractable neuropathic pain. Recent studies clarified the therapeutic significance of SHED in neuropathic pain caused by infraorbital nerve injury or diabetics in rats 12,20,21 . Analgesic mechanisms by SHED administration are revealed that SHED suppresses the upregulation of transient receptor potential vanilloid type 1 (TRPV1) in trigeminal ganglion neurons following infraorbital nerve injury 20 . However, SHED-mediated anti-allodynic effects are not fully understood. In the current study, we aimed to investigate the effects of SHED or SHED-CM on activated macrophage/microglia during neuropathic pain.

Results
Anti-allodynic effect of a single intravenous injection of SHED. A significant decrease in the paw withdrawal threshold (PWT) was observed from 1 day after peripheral nerve injury (PNI) in control mediumadministered group and the reduction of the PWT following PNI was continued throughout the experimental period (Fig. 1a, control medium, n = 5). No change was observed in the contralateral side of the hind paw (Fig. 1a). To evaluate the anti-allodynic effect of SHED, a single administration of SHED through the left external jugular vein was conducted 5 days after PNI. A significant PWT recovery was observed from day 8 after intrajuglar administration of SHED (Fig. 1a, SHED, n = 7).

SHED accumulates in the injured DRG following intravenous injection.
In order to examine the localization of SHED after an intravenous administration, 6-carboxyfluorescein diacetate (CFDA)-labeled SHED was injected into the left external jugular vein 2 days after nerve injury. 6-CFDA-labeled SHED was detected in the ipsilateral side but not contralateral of L4 DRG at 3 days after injection, but not either the ipsilateral L3 or L5 DRG or L4 spinal dorsal horn (Fig. 1b,c).
Anti-allodynic effect of protein components secreted from SHED. Next, the effect of a single intravenous injection of SHED-CM was examined at 5 days after PNI. SHED-CM administration significantly reversed the established mechanical allodynia at 4 days after injection (Fig. 1d, SHED-CM). Five-fold concentrated SHED-CM (5× SHED-CM) had almost the same anti-allodynic effect as SHED-CM (Fig. 1d, 5× SHED-CM). In contrast, boiled SHED-CM had no significant effect on the mean PWT (Fig. 1d, b-SHED-CM), suggesting that SHED-CM includes protein components. No change was observed in the contralateral side of the hind paw (Fig. 1d). Inhibitory effects of SHED or SHED-CM on the induction of activating transcription factor 3 (ATF3) and the accumulation of macrophages in the DRG. Peripheral nerve injury induces the expression of ATF3, a marker of neuronal injury 22 , and accumulation of macrophages in the injured DRG 19 . Therefore, the effects of intravenous administration of SHED or SHED-CM on the expression of ATF3 and the accumulation of macrophages in the DRG were investigated.The mean number of ATF3-positive neurons was significantly increased in the ipsilateral side of the L4 DRG at 20 days after PNI (Fig. 2a,b). Intravenous administration of SHED or SHED-CM significantly reduced the mean number of ATF3-positive neurons (Fig. 2a,b). Marked accumulation of macrophages was observed in the ipsilateral side of the L4 DRG at 20 days after PNI compared to the contralateral side. The mean density of ionized calcium-binding adapter molecule 1 (IBA1)positive cells was significantly suppressed by intravenous administration of SHED or SHED-CM (Fig. 2c,d).
SHED attenuates the activation of microglia and astrocytes in the spinal dorsal horn following PNI. It is well known that microglia and astrocytes are activated following PNI and inhibition of these cells suppresses the development of mechanical allodynia 16,17,23,24 . To analyze the effects of SHED on the activation of spinal microglia and astrocytes, we stained spinal cord slices with anti-IBA1 or glial fibrillary acidic protein (GFAP) antibodies. Both IBA1 and GFAP immunofluorescence was increased in the ipsilateral side compared to the contralateral side of the spinal dorsal horn (SDH) after 20 days of PNI ( Fig. 3a-f). In contrast, intravenous administration of SHED markedly inhibited the increment of IBA1 and GFAP immunofluorescence in the ipsilateral side of the spinal dorsal horn (Fig. 3a-f).
Mechanical allodynia was unaffected by a single intravenous administration of ED-siglec-9 and MCP-1. Next, the effects of combined injection of ED-siglec-9 and MCP-1 were examined on the PNIinduced mechanical allodynia, because ED-siglec-9 and MCP-1 synergistically promoted recovery in a rat model of spinal cord injury by altering the macrophage polarity toward the anti-inflammatory M2 state 11 . However, a combined intravenous administration of human recombinant ED-siglec-9 and MCP-1 had no significant effect on the mean PWT (Fig. 4a, saline, n = 4; ED-Siglec-9 + MCP-1, n = 5). www.nature.com/scientificreports/ www.nature.com/scientificreports/

Effects of a single intravenous administration of filtrated SHED-CM on mechanical allodynia.
In order to narrow down the molecular weight range of active components in SHED-CM, SHED-CM was subjected to ultrafiltration using the membrane filters to cut off molecules with a molecular mass larger than 30, 50, and 100 kDa. A single intravenous administration injection of SHED-CM, which cut off molecules with a molecular mass larger than approximately 30 kDa, showed no significant effect on the mean PWT (Fig. 4b,  CM30). On the other hand, the administration of SHED-CM, which cut off molecules with a molecular mass larger than approximately 50 or 100 kDa, reversed the established mechanical allodynia at 6 days after injection ( Fig. 4b, CM50, CM100). Furthermore, the administration of SHED-CM with a molecular mass between 30 and www.nature.com/scientificreports/ 50 kDa also reversed the established mechanical allodynia at 6 days after injection (Fig. 4b, CM30-50). These observations suggest that active components present in SHED-CM that exhibit an ant-allodynic effect is heatstable molecule(s) with the molecular weight ranging from 30 to 50 kDa.

Discussion
The present study has provided evidence that a single intravenous injection of either SHED or SHED-CM through the left external jugular vein reversed the established mechanical allodynia in mice. Rather surprisingly, the antiallodynic effects of SHED or SHED-CM persisted throughout the observation period of 20 days. Consistent with the present observations, a single intraspinal administration of SHED-CM loaded in collagen hydrogel showed anti-allodynic effects in the spinal cord injury model in rats 12 . In addition, the mean head withdrawal threshold in rats with infraorbital nerve ligation was significantly increased by a single systemic or local injection of SHED 20 . Accordingly, SHED has a potential therapeutic effect on severe neuropathic pain in animals. Our histological analyses here showed that a single intravenous injection of SHED or SHED-CM significantly reduced the mean number of AFT3-positive neurons and the accumulation of macrophages in the ipsilateral side of DRG. ATF3 is invariably induced in DRG neurons after PNI 17,22 . Guan et al. found that after PNI, ATF3expressing DRG neurons expressed colony-stimulating factor (CSF-1) which was transported to the spinal dorsal horn and activated microglia 17 . This evidence implies that ATF3 might regulate the expression of CSF-1 in DRG neurons. Spatiotemporal analysis revealed activation patterns of microglia in the SDH after PNI. Following L4 nerve injury, an increased expression of calcitonin gene-related peptide associated with microglial activation was spread from the L4 toward the L3 and S1 SDH 25 . Thereby, activated spinal microglia in the L3 or L5 SDH potentiate nociceptive information from the hind paw. Activated microglia secrete several molecules including interleukin-18 and complement C1q which activate astrocytes, resulting in the exacerbation of neuropathic pain 26,27 . Considering the above evidence and our results that SHED and SHED-CM suppressed ATF3-expression in the DRG and the activation of microglia and astrocytes, it is conceivable that SHED and SHED-CM alter activated type of microglia and astrocytes into a quiescent phenotype by interrupting the sequential signal relay initiated by ATF3 induction in DRG neurons. More recently, critical involvement of accumulated macrophages in the DRG on maintenance of neuropathic pain has been reported 19 . In fact, SHED and SHED-CM have suppressed the accumulation of macrophages in the DRG following PNI. Through the above-mentioned mechanisms, SHED or SHED-CM are thought to have a long-term anti-allodynic effect by a single intravenous administration. Unfortunately, an intravenous injection of MCP-1 and ED-Siglec-9, which improved spinal cord injury 11 , had no effect on the mechanical allodynia. Future studies will identify the target analgesic molecules. In the present study, there is a gap in the onset of anti-allodynic effects between SHED and SHED-CM treatment. Considering that SHED was selectively accumulated in the damaged DRG, it is conceivable that SHED was anchored in the DRG and released anti-allodynic molecule(s), resulting in the suppression of ATF3 expression in DRG neurons and macrophage accumulation in the DRG.
Despite the anti-allodynic effects of SHED, we did not address how many SHED is anchored in the DRG. We assessed the accumulation of 6-CFDA-labeled SHED in the DRG 3 days after its administration because the Co-administration of ED-siglec-9 + MCP-1 showed no effect of the mean PWT. n = 4 (Saline); n = 5 (ED-siglec-9 + MCP-1), two-way ANOVA repeated measures ANOVA post hoc Bonferroni's test, not significant, Saline vs. ED-siglec-9 + MCP-1. (b) The effects of SHED-CM on the development of mechanical allodynia. Data represents the mean ± SEM. n = 5 (CM); n = 5 (CM30); n = 6 (CM50); n = 5 (CM100); n = 6 (CM30-50), two-way ANOVA repeated measures ANOVA post hoc Dunnet's test (*p < 0.05, **p < 0.01, ***p < 0.001, CM vs. CM30). www.nature.com/scientificreports/ fluorescence intensity of 6-CFDA could not maintain for a long time. However, 6-CFDA-labeled SHED selectively migrated in the damaged site but not in the intact site, and anti-allodynic effects of SHED lasted for 20 days. After SHED administration, SHED migrated into the injured but not the intact side of the trigeminal ganglion after infraorbital nerve injury (a model of orofacial neuropathic pain) in rats 20 . Furthermore, the anti-allodynic effects of SHED lasted for 8 weeks 20 . Accordingly, long-lasting anti-allodynic effects might be observed in our case. Following the migration into the damaged tissue, SHED can differentiate into functional cells 2,28 . In addition to the factors secreted by SHED, it is also possible that the differentiation of SHED is responsible for the anti-allodynic effect. In the future, we need to analyze the profile of migrated SHED in the injured side of DRG. The close examination revealed that the molecular weight of proteins component(s) responsible for the anti-allodynic effect of SHED-CM ranged from 30 to 50 kDa. SHED-CM contains growth factors, proteases, and immune modulators with the molecular mass of 30-50 kDa, including vascular endothelial growth factor (VEGF), placental growth factor, matrix metalloproteinase-3, marapsin, follistatin, osteopontin, and CD40 11 . Among them, VEGF is the most likely mediator of the anti-allodynic effect exerted by SHED-CM, because VEGF has an important neuroprotective effect during neuropathic conditions 29 and anti-nociceptive effects in experimental painful conditions 29,30 . However, VEGF has been also shown to induce pronociceptive effects 31 . Therefore, it is likely to consider that some of these protein factors may synergistically cause protective and/or anti-inflammatory activity to induce an anti-allodynic effect. We found a strong anti-allodynic activity SHED-CM ranged from 30 to 50 kDa. However, the anti-allodynic effect of SHED-CM was also observed when the molecular weight was cut above 100 kDa. From this observation, the possibility of the existence of analgesic factors in the range of 50 to 100 kDa cannot be ruled out.
The proliferation rate of stem cells from SHED is significantly higher than that of BMMSCs and dental pulp stem cells 2,32,33 . SHED can be easily obtained, without the need for invasive procedures and thus represent a large source of stem cells for potential clinical application. Therefore, SHED may have advantages over BMMSCs in pain control and potentially become an alternate and efficacious treatment for neuropathic pain. Furthermore, considering the short-lasting anti-nociceptive effect of gabapentin, a first-line agent for the treatment of neuropathic pain, the long-lasting property of SHED-CM has greater clinical advantages. Further studies will be needed to identify and characterize protein factors responsible for the anti-allodynic effect of SHED-CM.

Materials and methods
Isolation and culture of SHED and preparation of conditioned medium from SHED. Experiments using human samples were reviewed and approved by the Kyushu University Institutional Review Board for Human Genome/Gene Research (Protocol Number: 677-00) and were conducted in accordance with the Declaration of Helsinki. Informed consent was obtained from the patients' guardians. Human exfoliated deciduous teeth were collected as discarded biological/clinical samples from children (5-7-year-old) at the Department of Pediatrics of Kyushu University Hospital. The SHED was isolated as previously described 34 and cultured in Alpha Modification of Eagle's Medium (Sigma-Aldrich, MO, USA) containing 15% fetal bovine serum (Sigma-Aldrich), 100 μM l-ascorbic acid 2-phosphate (Wako Pure Chemical Industries, Osaka, Japan), 2 mM l-glutamine (Thermo Fisher Scientific, MA, USA), 250 μg/ml fungizone (Thermo Fisher Scientific), 100 U/ ml penicillin, and 100 μg/ml streptomycin (Thermo Fisher Scientific), at 37 °C, in an atmosphere containing 5% CO 2 . The cells were used for further experiments as a heterogeneous cell population. In some experiments, SHED was incubated with 6-CFDA (100 nM, 15 min, Thermo Fisher Scientific), and then injected through the external jugular vein as described below.
To obtain SHED-CM, the cells were cultured to subconfluency and incubated in 10 ml of serum-free medium for 48 h. The supernatant was collected and centrifuged at 500 rpm for 5 min. The supernatant was recentrifuged at 3000 rpm for 3 min and the sample was collected. SHED-CM was subjected to a fivefold concentration (5× SHED-CM) using Amicon Ultra (Merck Millipore, MA, USA) according to the manufacture's protocol. To inactivate SHED-CM, SHED-CM was heated at 95 °C for 5 min. Fractionation of SHED-CM based on molecular size was conducted using Amicon Ultra 30 k, 50 k, or 100 k according to the manufactures' protocol. Serum-free alpha Modification of Eagle's Medium was used as the control medium.
Animals. Male C57BL/6J mice (8-10 weeks old) were purchased from CLEA Japan, Inc. The mice were maintained in a 12 h light/dark cycle (light beginning at 08:00) at 22 ± 1 °C ambient temperature with food and water provided ad libitum. All mice were handled daily for 5 days prior to the initiation of the experiment to minimize their stress reactions to manipulation. Animal protocols were approved by the Experimentation com- Surgical procedure. For the neuropathic pain model, PNI was made by the transection of the L4 spinal nerve without injury to the adjacent nerve under 2% isoflurane inhalation according to the method as described previously 16 .
For the intravenous administration, the left external jugular vein was exposed with blunt dissection. Intrajuglar injection was made using a 31-gauge needle after passing through the pectoral muscle. SHED (2 × 10 5 cells/100 μl serum-free medium), boiled SHED (100 μl), 5 × SHED-CM (100 μl), filtered SHED-CM (100 μl), or human recombinant ED-siglec-9 and MCP-1 (100 ng/ml in each) were intrajuglarly delivered 5 days after PNI. 6-CFDA-labeled SHED (2 × 10 5 cells/100 μl serum-free medium) was intravenously injected 2 days after PNI. ). The images were acquired on a Nikon C2 scanning confocal microscope using a 20× objective lens (NA 0.75; Nikon Corporation, Tokyo, Japan). 6-CFDAlabeled SHED and ATF3-positive and Nissl-positive DRG neurons and IBA1-positive cells in the images (image size: 460 × 460 μm 2 ) were counted by Image J software (version 1.53k) plugin Cell count (NIH; http:// rsbweb. nih. gov/ ij/). The signal was defined as the fluorescence intensity that was five times greater than the noise intensity. The ratio of ATF3-positive neurons in total DRG neurons was calculated. The area occupied by IBA1 fluorescence in the DRG or IBA1 or GFAP immunofluorescence in the SDH was measured by Image J following making binarized images from original images. Three sections from one mouse were stained, and the mean value was taken as the value for one mouse. www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.