Author Correction: Umbilical cord extracts improve osteoporotic abnormalities of bone marrow-derived mesenchymal stem cells and promote their therapeutic effects on ovariectomised rats

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

MSCs are considered a highly useful cell source for regenerative medicine because of their multi-potentiality and safety profile. In particular, MSCs have been focused upon because of their strong capacity for self-renewal, pluripotency, reduced antigenicity, immunoregulatory functions, and ease of isolation and in vitro culture to obtain large numbers of cells for treatment 4 . Significant efforts have already been made for their clinical application as a result of their safety and efficacy in systemic administration 5 . Several clinical trials investigating BM-MSC cell therapies have been reported for autoimmune diseases 6,7 , chronic inflammatory disease 8,9 , myocardial infarction 10 , spinal cord injury 11 , and osteoporosis 12 .
Autologous transplantation of BM-MSCs has great benefits because of a low risk of rejection and exogenous infection, as well as the availability of a stable source of MSCs. However, several functional abnormalities of BM-MSCs have been reported in osteoporosis patients [13][14][15] , which suggested that BM-MSCs derived from patients are unsuitable for cell therapy. Zhao and others reported that oestrogen potentially regulates the osteoblastic differentiation of human BM-MSCs via PI3K signalling or upregulation of oestrogen receptor alpha, which results in the diminished production of osteoblasts and excessive differentiation of adipocytes from BM-MSCs in postmenopausal osteoporosis patients. Li and others reported that BM-MSCs derived from osteoporosis rats had decreased proliferation ability and pluripotency-related gene expression compared with BM-MSCs derived from normal rats 16,17 . However, the abnormalities of BM-MSCs with respect to regulation of osteoclast activity have rarely investigated. Furthermore, the therapeutic effect of abnormal BM-MSCs on osteoporosis in vivo has yet to be clarified. Therefore, we first aimed to investigate whether abnormal BM-MSCs derived from an oestrogen-deficient osteoporosis model exhibit sufficient therapeutic effects on osteoporosis in vivo, and then clarified the abnormalities of BM-MSCs by focusing on the regulation of osteoclast activity in vitro.
To carry out autologous transplantation using abnormal BM-MSCs, these cells need to be remade into functional cells capable of exerting therapeutic effects on abnormal bone metabolism. In this study, we focused on a novel activator, human UC extract, which we named "Wharton's jelly extract supernatant" (WJS). The UC is composed of embryonic tissues including umbilical vessels, Wharton's jelly (WJ), and amniotic membranes. We have elucidated the effect of WJS on the functional improvement of impaired BM-MSCs with diabetes 18 . WJS contained a variety of growth factors, cytokines, extracellular matrix (ECM) proteins and micro-vesicles. WJS improved not only morphological and functional abnormalities of BM-MSCs but also the therapeutic effect on diabetic nephropathy. Moreover, An et al. reported that human UC blood-derived MSCs ameliorated bone mineral density in ovariectomised (OVX) nude mice 19 . Therefore, we hypothesised that these biological components might also activate abnormal BM-MSCs derived from an osteoporosis. Thus, we first aimed to investigate the efficacy of WJS on abnormal BM-MSCs in vitro. Finally, we investigated the therapeutic effects of activated BM-MSCs on the OVX model in vivo.
Our method may allow for autologous cell transplantation using a patient's own BM-MSCs not only for treatment of osteoporosis, but also other diseases in which autologous BM-MSCs are abnormal.

BM-MSCs derived from sham-operated rats (Sham-MSCs) but not OVX rats (OVX-MSCs) ameliorated osteoporosis in OVX rats.
Four weeks after OVX, rats were used as a model of osteoporosis (OVX rats) to perform cell therapy using various types of BM-MSCs. At this time point, a significant decrease of bone volume ( Supplementary Fig. S1) indicated successful establishment of the OVX model. The experiment, carried out as shown in Fig. 1a, examined OVX rats treated with Vehicle (OVX-Vehicle), Sham-MSCs (OVX-Sham-MSCs), or OVX-MSCs (OVX-OVX-MSCs). Micro-computed tomography (micro-CT) showed that the number and volume of trabeculae were obviously reduced in the proximal tibia of OVX-Vehicle rats compared with Sham rats (Fig. 1b). In quantitative analysis of micro-CT images, bone volume fraction, trabecular thickness, and trabecular number were significantly decreased, and trabecular separation was significantly increased in OVX-Vehicle rats compared with Sham rats (P = 0.002, Fig. 1c; P = 0.02, Fig. 1d; P = 0.005, Fig. 1e; P = 0.005, Fig. 1f). As evaluated by these indicators, Sham-MSCs inhibited the progression of osteoporosis in the proximal tibia of OVX-Sham-MSCs rats compared with OVX-Vehicle rats (P = 0.008, Fig. 1c; P = 0.03, Fig. 1d; P = 0.014, Fig. 1e; P = 0.017, Fig. 1f). Conversely, OVX-MSCs did not exhibit adequate therapeutic effects in OVX rats similar to OVX-Vehicle rats, and in contrast to Sham-MSCs (P = 0.026, Fig. 1c; P = 0.019, Fig. 1d; P = 0.025, Fig. 1e; P = 0.044; Fig. 1f). Histological findings of haematoxylin and eosin (H&E) staining in the proximal tibia showed similar changes as observed in micro-CT. Specifically, narrowing of trabecular bone and numerous fat deposits in the bone marrow cavity were observed in OVX-Vehicle and OVX-OVX-MSC rats (Fig. 1g). Administration of Sham-MSCs improved these histological changes in OVX rats (Fig. 1g). Notably, the expression of receptor activator of nuclear factor κ-B (RANK) in osteoclasts was increased in OVX-Vehicle rats (Fig. 1h), as was the number of tartrate-resistant acid phosphatase (TRACP)-positive osteoclasts (Fig. 1i). In addition, both of them were decreased in OVX-Sham-MSCs rats compared with OVX-Vehicle rats, while these features were unchanged in OVX-OVX-MSCs rats (Fig. 1h,i). The average size of osteoclasts was larger in OVX-Vehicle rats compared with Sham rats, as determined by the intensity and area of TRACP expression, while it was smaller in OVX-Sham-MSC rats compared with OVX-Vehicle rats. Conversely, OVX-MSCs did not suppress the increased size of osteoclasts (Fig. 1i, right panels). Serum TRACP levels were also significantly higher in OVX-Vehicle rats and OVX-OVX-MSCs rats compared with Sham rats (P = 0.011, OVX-Vehicle vs. Sham; P = 0.047, OVX-OVX-MSCs vs. Sham; Fig. 1j), while they were significantly lower in OVX-Sham-MSCs rats compared with OVX-Vehicle and OVX-OVX-MSC rats 8 weeks after the administration of each type of BM-MSC (P = 0.004, OVX-Sham-MSC vs. OVX-Vehicle; P = 0.022, OVX-Sham-MSCs vs. OVX-OVX-MSCs; Fig. 1j).

Distribution of Sham-MSCs, OVX-MSCs-WJ(−) and OVX-MSCs-WJ(+) in OVX rats. PKH26-labelled
BM-MSCs were distributed in the bone marrow of the epiphysis in OVX rats ( Supplementary Fig. S4a). The number of cells distributed in bone was larger in OVX-Sham-MSCs rats and OVX-OVX-MSCs-WJ(+) rats compared with OVX-OVX-MSCs-WJ(−) rats on day 1, 3, and 7 after each cell administration ( Supplementary Fig. S4b). Numerous PKH26-labelled BM-MSCs were distributed in the lung on days 1, 3 and 7 after each cell administration, but there was no difference between BM-MSC types ( Supplementary Fig. S4c,d). Only a few cells were distributed in liver and spleen on day 1 and 3 after cell administration (data not shown).

Sham-MSCs and OVX-MSCs-WJ(+) ameliorated the maturation and excessive activation of macrophage-derived osteoclasts.
As the number and activity of osteoclasts in bone tissue were suppressed by Sham-MSCs and OVX-MSCs-WJ(+), we investigated whether regulation of osteoclast activity by BM-MSCs was the mechanism of MSC therapy. The macrophage cell line RAW264.7 was differentiated into macrophage-derived osteoclasts by adding receptor activator of nuclear factor κ-B ligand (RANKL) alone or in combination with the MEK inhibitor PD98059 in vitro ( Supplementary Fig. S5a). Morphologically, macrophages were fused into multinucleated osteoclasts, which became larger and more mature with the addition of RANKL and PD98059 in combination ( Supplementary Fig. S5b). TRACP activity in the culture supernatant of macrophage-derived osteoclasts was increased 72 h after induction with RANKL and PD98059 in combination than RANKL alone (P < 0.001, RANKL vs. None; P < 0.001, RANKL + PD98059 vs. None; P < 0.001, RANKL + PD98059 vs. RANKL; Supplementary Fig. S5c). Next, macrophage-derived osteoclasts and each type of BM-MSC were co-cultured indirectly for 24 h (Fig. 5a). The increased average size of osteoclasts was suppressed by co-culturing with Sham-MSCs or OVX-MSCs-WJ(+) compared with Vehicle and OVX-MSCs-WJ(−) (Fig. 5b). In addition, TRACP activity in the culture supernatant was significantly suppressed by co-culture with    These expression was not suppressed by OVX-MSCs-WJ(−). In addition, the expression of Clcn7 and Dc-stamp was significantly downregulated by OVX-MSCs-WJ(+) compared with OVX-MSCs-WJ(−) (P = 0.032, Clcn7; P = 0.047, Dc-stamp; Fig. 5d).

Sham-MSCs and OVX-MSCs-WJ(+) inhibited maturation and excessive activation of primary mouse bone marrow cell (BMC)-derived osteoclasts. Sham-MSCs and OVX-MSCs-WJ(+) inhib-
ited the maturation and excessive activation of primary osteoclasts. Co-culture experiments performed using primary osteoclasts derived from mouse BMCs cultured with macrophage colony-stimulating factor (M-CSF) showed differentiation into osteoclasts by the addition of RANKL alone or in combination with PD98059 in vitro ( Supplementary Fig. S6a). Morphologically, osteoclast precursor cells derived from BMCs fused to form multinucleated osteoclasts, which became larger and more mature upon the addition of RANKL and PD98059 in combination (Supplementary Fig. S6b). TRACP activity in the culture supernatant of mouse BMC-derived osteoclasts was increased 10 days after induction with RANKL and PD98059 in combination than RANKL alone (P < 0.001, RANKL vs. None; P < 0.001, RANKL + PD98059 vs. None; P = 0.03, RANKL + PD98059 vs. RANKL; Supplementary Fig. S6c). Next, primary mouse BMC-derived osteoclasts and each type of BM-MSC were co-cultured indirectly for 48 h (Fig. 6a). The increased size of osteoclasts was suppressed by co-culture with Sham-MSCs or OVX-MSCs-WJ(+) compared with Vehicle or OVX-MSCs-WJ(−) (Fig. 6b). TRACP activity in the culture supernatant was significantly suppressed by co-culture with Sham-MSCs, OVX-MSCs-WJ(−), or OVX-MSCs-WJ(+) compared with Vehicle [P < 0.001, Sham-MSCs; P < 0.001, OVX-MSCs-WJ(−); P < 0.001, OVX-MSCs-WJ(+), Fig. 6c]. There was no obvious difference between these BM-MSCs in their ability to inhibit TRACP activity. To clarify the reason why there was no difference in TRACP activity within supernatant between different BM-MSC types, TRACP expression in co-cultured primary mouse BMCs was investigated by immunostaining. TRACP was expressed not only in the margin of matured osteoclasts, but also in small BMCs surrounding the fused osteoclasts (Fig. 6d). That is, TRACP activity within supernatant may not reflect the activity of mature osteoclasts.  Atp6i, Fig. 6e). In addition, the expression of Nfatc1 and Atp6i was significantly downregulated by OVX-MSCs-WJ(+) compared with OVX-MSCs-WJ(−) (P = 0.029, Nfatc1; P < 0.001, Atp6i, Fig. 6e).

Discussion
As BM-MSCs derived from an OVX osteoporosis rat model exhibited reduced therapeutic efficacy, we created a new activation method to improve the therapeutic effects of BM-MSCs for autologous transplantation. We focused on a UC extract, WJS, and demonstrated its ability to improve abnormalities in BM-MSCs arising from oestrogen deficiency. This is the first study to investigate the significant potential of WJS to ameliorate abnormalities in BM-MSCs caused by osteoporosis. BM-MSCs activated by WJS exhibited improved proliferative ability, cell mobilisation, and regulatory effects on the excessive osteolytic properties of osteoclasts.
OVX-MSCs had morphological abnormalities. Cell area was expanded in an irregular manner, with OVX-MSCs exhibiting a flattened appearance and increased cytosolic actin filaments compared with Sham-MSCs, which maintained their slim, spindle shape and uniform cell size. Furthermore, cell proliferation in OVX-MSCs was significantly decreased. The size of BM-MSCs reportedly correlates with their stemness, proliferation, differentiation, and tissue regenerative capacity 20 . Accordingly, abnormalities in the size and shape of OVX-MSCs predicted the following functional disorders of these cells.
Cell mobilisation was downregulated in OVX-MSCs. According to the results of scratch and chemotaxis assays, not only mobilisation itself, but also the response to chemokine stimulants, such as SDF-1, IL-1β and IL-6, were extremely downregulated in OVX-MSCs compared with Sham-MSCs. BM-MSCs are known to migrate in response to many chemotactic factors, including platelet-derived growth factor-AB (PDGF-AB), insulin-like growth factor 1 (IGF-1) and SDF-1 21 . The migratory capacity of BM-MSCs is under the control of receptor tyrosine kinases, growth factors, and CC and CXC chemokines. The mobilisation of MSCs, as well as their subsequent homing to damaged bone tissue, may depend on the systemic and local inflammatory state. Ovariectomy leads to a significant increase in IL-1β and IL-6 levels in tibia 22 and serum 23 . In our study, expression of IL-1β and IL-6 were increased in the tibias of OVX rats, which might induce enhanced mobilisation of BM-MSCs. However, due to the low reactivity of OVX-MSCs to these stimuli, an insufficient number of cells were able to migrate to local lesions. Reflecting these cell characteristics, the number of OVX-MSCs distributed in the bone epiphysis might contribute to the therapeutic effects of BM-MSCs.
OVX-MSCs could not regulate excessively activated osteoclasts with high osteolytic capacity in vivo, as indicated by reduced expression and secretion of OPG. By strongly binding to RANK to form a decoy receptor for RANKL, OPG triggers osteoclast differentiation and subsequently suppresses the activity of the essential transcription factor NFATc1 24 . Expression of TGF-β1, which was also downregulated in OVX-MSCs, is known to increase Opg transcription in osteoblasts 25 and promote osteoblastic cell proliferation, function, and survival 26 , which ultimately encourages bone formation and regulates osteoclast activity. As the expression of OPG and TGF-β1 were downregulated in OVX-MSCs, they could not sufficiently suppress osteoclast activity, which may cause the reduction of bone volume and loss of therapeutic effects observed in OVX rats. As expected, OVX-MSCs elicited reduced therapeutic effects in OVX rats. Bone strength is decreased in osteoporosis, as indicated by reduced bone volume fraction and bone quality. Bone strength and microstructure have been evaluated by micro-CT in the OVX model 27 , and bone quality can be determined by assessing trabecular thickness, number, and separation. In this study, OVX-MSCs did not improve any of these indicators, while Sham-MSCs improved them sufficiently. This result was consistent with histological findings showing the absence of an improvement of trabecular bone and the presence of fat deposits in the bone marrow cavity. Furthermore, serum TRACP levels and expression of RANK and TRACP in osteoclasts was not sufficiently decreased in OVX rats treated with OVX-MSCs compared with Sham-MSCs in vivo. Thus, the reduced therapeutic effects of OVX-MSCs was considered to arise from the functional abnormalities of these cells, as shown in vitro.
We developed a UC extract, namely WJS, as an activator of OVX-MSCs to improve their abnormalities. WJS contains various physiologically active substances including growth factors, ECM, amino acids, exosomes, and nucleic acids represented by microRNAs (miRNAs) originating in these tissue components 18,[28][29][30] . In addition, the cellular components of UC, represented by UC-MSCs, produce large amounts of autocrine/paracrine factors, such as IGF-1, basic fibroblast growth factor (b-FGF), TGF-β, PDGF, epidermal growth factor (EGF), and miRNAs 28,31  Morphological and functional abnormalities of OVX-MSCs were improved by WJS. Morphologically, expansion of cell area, flattening, and irregularities in the size and shape of cells were improved. The proliferation of OVX-MSCs was increased by the addition of WJS, as indicated by PDT and MTT assay. An improvement of proliferation is thought to be beneficial in cell therapy to secure an adequate number of cells for transplantation from bone marrow within a short period of time. Cell mobilisation was also improved by the addition of WJS, as indicated by scratch and chemotaxis assays. The improvement of cell mobilisation to damaged tissues would greatly benefit the efficiency of cell therapies by promoting cell distribution. In addition, WJS improved the expression of OPG and TGF-β1 in OVX-MSCs, suggesting OVX-MSCs-WJ(+) more effectively regulate osteoclast activity.
In vivo treatment with WJS-activated OVX-MSCs significantly improved trabecular bone volume and thickness as observed with micro-CT. These improvements were consistent with the recovery of histological findings for bone tissues and observed reductions of RANK, TRACP, IL-1β, and IL-6 expression. RANK is a receptor expressed on osteoclast precursor cells that transmits intracellular signals essential for the differentiation and activation of osteoclasts by binding with RANKL. In postmenopausal osteoporosis, bone resorption is also increased by the production of monocyte-related cytokines, such as IL-1, IL-6, and TNF-α, which induce RANKL expression in bone tissue and enhance RANKL-RANK-mediated osteoclastogenesis 39 . Conversely, UC-MSCs are known to exert immunosuppressive effects on monocytes by regulating cytokine production 40 . Considering the similar characteristics of OVX-MSCs-WJ(+) and UC-MSCs, WJS might change the function of OVX-MSCs to assist them in exerting immunoregulatory effects on osteoclasts by regulating RANK signalling with increasing OPG and cytokine production. These findings correlated with decreased TRACP-positive osteoclasts and expression of IL-1β and IL-6 in the OVX-MSCs-WJ(+) group compared with Vehicle and OVX-MSCs-WJ(−) groups in vivo.
The therapeutic effects continued for 8 weeks despite the disappearance of the administered cells in bone tissue within few days. Several studies have reported that systemic administration of MSCs effectively prevented bone loss despite poor BMC homing and short-term engraftment MSCs 41,42 . In addition, systemic injection of human umbilical cord blood derived-MSCs and their conditioned medium were effective on OVX-induced bone loss in mice 19 . Based on the previous reports, we speculated two reasons for these phenomena. First, transplanted BM-MSCs migrated to the epiphysis of bone tissues and triggered osteoclast regulation. Once the trigger for repair was turned on, the effect continued even after the cells disappeared. Second, administered BM-MSCs were also distributed to the lung and were still detectable 7 days after transplantation. As we showed that osteoclast activity was regulated by a paracrine effect of BM-MSCs, the therapeutic effect may have been sustained by paracrine factors derived from BM-MSCs remaining in the lung. Moreover, as the levels of OPG secretion and TGF-β expression were higher in Sham-MSCs and OVX-MSCs-WJ(+) compared to OVX-MSCs-WJ(−), the sufficient therapeutic effect was obtained in OVX rats which were treated with these functional BM-MSCs.
OVX-MSCs-WJ(+) suppressed the fusion and enlargement of osteoclasts derived from both RAW cells and primary mouse BMCs. Expression of osteoclast-activating factors (i.e. Nfatc1, cathepsin-k, Clcn7, and Atp6i), a fusion-promoting factor (i.e. Dc-stamp) and TRACP levels in the culture supernatants of osteoclasts were also suppressed by culture with OVX-MSCs-WJ(+). Significant decreases in Nfatc1, a master transcription factor of osteoclasts located downstream of the RANK pathway, cathepsin-k, a degradation factor of type 1 collagen, Clcn7, a chlorine transporter that promotes bone resorption, Atp6i, an acid transporter that promotes bone resorption, and Dc-stamp, which specifically promotes the fusion of macrophages and multinucleation of osteoclasts, were observed in osteoclasts cultured with OVX-MSCs-WJ(+) compared with Vehicle. In addition, Nfatc1 and Atp6i were significantly downregulated in mouse BMC-derived osteoclasts, while Clcn7 and Dc-stamp were significantly downregulated in macrophage-derived osteoclasts in response to administration of OVX-MSCs-WJ(+) compared with OVX-MSCs-WJ(−). These differences were thought to arise from the presence of BMCs other than osteoclasts producing physiologically active substances such as cytokines. Considering these findings,

Methods
Animal model of osteoporosis. Eight-week-old female Wistar rats weighing 135-145 g were purchased from Japan SLC, Inc. (Shizuoka, Japan). Rats were housed in a temperature-controlled room (24 ± 1 °C) with a 12-h light/dark cycle and given free access to food and water. Rats received either a sham operation (Sham) or OVX under general anaesthesia. The sham operation was performed using the same surgical procedure as for OVX, but without removing the ovaries. Both Sham and OVX rats underwent minimal surgery through a dorsal approach 43   Evaluation of bone mass and microarchitecture by micro-CT. The right tibia of each rat was isolated and fixed in 4% ethanol for micro-CT with a ScanXmate-L090 (Comscantecno, Yokohama, Japan) operated at a lamp voltage of 75 kV and current of 100 μA, using the software X sys FP Version 1.7 and coneCTexpressIV 1.54 (Comscantecno). Samples were scanned at a magnification factor of 5.263 and spatial resolution of 19.001 μm/ pixel. Captured images were rendered using the machine software TRI/3D BON (Ratoc System Engineering Co., Ltd., Tokyo, Japan).
Histological findings of bone tissues. The left tibia from each rat was fixed with 4% paraformaldehyde in phosphate-buffered saline and decalcified with 10% ethylenediaminetetraacetic acid. Bone tissues were cut into thin sections (7 µm) and stained with H&E (Wako, Osaka, Japan). Stained sections were observed with a light microscope (NIS element BR 3.0; Nikon, Tokyo, Japan).
Immunofluorescence staining of bone tissues. Bone tissues were cut into thin sections (7 µm).

Measurement of TRACP levels in serum and culture supernatant of osteoclasts. Blood samples
were obtained through cardiac puncture at sacrifice, and serum was separated and stored at −80 °C until use. Culture supernatants of osteoclasts were collected and stored at −80 °C until use. TRACP levels in serum and culture supernatants were measured with a TRACP & ALP Assay Kit (Takara Bio, Inc., Shiga, Japan) according to the manufacturer's instructions.
Preparation of WJS from UC tissues. WJS was prepared as previously described 18 . Briefly, whole UC tissues were cut into short pieces, and all of the sectioned sheathes of amnion, UC vessels, and WJ were collected. Tissues were suspended in serum-free medium, and the supernatants of tissue suspensions were obtained as WJS. Proliferation assays of BM-MSCs. PDT was measured at P2 and calculated using the formula: PDT = tplg2/(lgNH − lgNI), where NI is the inoculum cell number, NH is the cell harvest number, and t is the time of the culture (in hours). The mean and standard deviation were calculated for three independent experiments. The PDT assay was performed using P2 because cells other than MSCs, such as macrophages, fibrocytes and haematopoietic stem cells, contaminated passages 0 to 1 in rat BMC cultures.
BM-MSCs at P2 were plated in 96-well cell culture plates (Corning Costar; Sigma-Aldrich, St. Louis, MO, USA) at a density of 2.5 × 10 3 cells/well. Triplicate wells were used for each sample. Proliferation of BM-MSCs was analysed using a Cell Counting Kit-8 (Dojindo Laboratories). The absorbance at 450 nm (OD450) was measured with a microplate reader (Infinite M1000 Pro; TECAN, Männedorf, Switzerland).

Quantitative real-time reverse transcription polymerase chain reaction (RT-PCR) of BM-MSCs.
Quantitative RT-PCR analysis was performed as described in Supplementary Methods in detail. Specific primers for rat Opg was used, which are described in Supplementary Table S5. Rat Gapdh primers acted as an internal standard for RNA integrity and quantity. Scratch assay and Chemotaxis assay. Scratch assay and Chemotaxis assay were carried out as described in Supplementary Methods in detail. The mobilisation ability of BM-MSCs was evaluated by measuring the area of an open wound after scratching the cell monolayer. The chemotaxis assay was carried out using culture membrane inserts using 100 ng/ml of SDF-1 (Peprotech, Inc., Rocky Hill, NJ, USA), 100 ng/ml of IL-1β (BioLegend, Inc. San Diego, USA), or 100 ng/ml of IL-6 (BioLegend). Recombinant chemokines used for chemotaxis assay were listed in Supplementary Table S6.

Measurement of OPG levels in
Regulation of RAW264.7 cell-derived osteoclast activities by BM-MSCs in vitro. Osteoclastogenesis of monocytes/macrophages was conducted by modifying the method of Yonezawa et al. 45 . The murine RAW264.7 monocyte/macrophage cell line (ATCC, Manassas, VA, USA) was used as osteoclast precursors by employing culture methods described in the Supplementary Methods in detail. Subsequently, RAW264.7 cells were cultured in the presence of RANKL and PD98059 to investigate the functional effect of BM-MSCs on osteoclasts. After 48 h, osteoclastic cells differentiated from RAW264.7 cells were co-cultured indirectly with BM-MSCs. After 24 h of co-culture, cell morphology, supernatant TRACP levels, and osteoclast-related gene expression were evaluated. mRNA expression of osteoclast-related genes was determined by quantitative real-time RT-PCR using primers listed in Supplementary Table S7.

Regulation of mouse BMC-derived osteoclast activity by BM-MSCs in vitro.
Osteoclastogenesis of mouse BMCs was conducted by modifying the method of Yonezawa et al. 45 . Briefly, mouse BMCs from 7-week-old C57BL/6 mice were harvested as described in the Supplementary Methods for use as osteoclast precursors. To investigate the functional effect of BM-MSCs on osteoclasts, mouse BMCs were cultured in the presence of M-CSF (100 ng/mL), RANKL (100 ng/mL), and PD98059 (20 mM). After 10 days, osteoclastic cells differentiated from mouse BMCs were co-cultured indirectly with BM-MSCs. After 48 h of co-culture, cell morphology, TRACP staining, supernatant TRACP levels, and osteoclast-related genes were evaluated. mRNA expression of osteoclast-related genes was determined by quantitative real-time RT-PCR using primers listed in Supplementary Table S7.
Ethics committee approval. The human study was conducted in accordance with the ethical principles of the Declaration of Helsinki and approved by the Ethics Committee of Sapporo Medical University (Registration numbers: 24-142, 25-1227, 262-1031, 262-1046, 262-110). Written informed consent was received from participants prior to their inclusion in the study.
Statistical analysis. Data from quantitative experiments are expressed as mean ± standard error (SE) values. One-way analysis of variance was employed for multiple comparisons. Two-way repeated measures (mixed between-within subjects) analysis of variance followed by Bonferroni's test was used for serial assessment. Differences were considered significant at P < 0.05 in all two-tailed tests. Statistical analysis was performed using SPSS software (version 16.0; SPSS, Inc., Chicago, IL, USA).