Therapeutic interactions between mesenchymal stem cells for healing medication-related osteonecrosis of the jaw

Abstract

Background: Mesenchymal stem cells (MSCs) have been isolated from a variety of tissues, including bone marrow, adipose, and mucosa. MSCs have the capacity for self-renewal and differentiation. Reports have been published on the systemic administration of MSCs leading to functional improvements by engraftment and differentiation, thus providing a new strategy to regenerate damaged tissues. Recently, it has become clear that MSCs possess immunomodulatory properties and can therefore be used to treat diseases. However, the therapeutic effect mechanisms of MSCs are yet to be determined. Here, we investigated these mechanisms using a medication-related osteonecrosis of the jaw (MRONJ)-like mouse model.

Methods: To generate MRONJ-like characteristics, mice received intravenous zoledronate and dexamethasone two times a week. At 1 week after intravenous injection, maxillary first molars were extracted, and at 1 week after tooth extraction, MSCs were isolated from the bone marrow of the mice femurs and tibias. To compare "diseased MSCs" from MRONJ-like mice (d-MSCs) with "control MSCs" from untreated mice (c-MSCs), the isolated MSCs were analyzed by differentiation and colony-forming unit-fibroblast (CFU-F) assays and systemic transplantation of either d-MSCs or c-MSCs into MRONJ-like mice. Furthermore, we observed the exchange of cell contents among d-MSCs and c-MSCs during coculture with all combinations of each MSC type.

Results: d-MSCs were inferior to c-MSCs in differentiation and CFU-F assays. Moreover, the d-MSC-treated group did not show earlier healing in MRONJ-like mice. In cocultures with any combination, MSC pairs formed cell-cell contacts and exchanged cell contents. Interestingly, the exchange among c-MSCs and d-MSCs was more frequently observed than other pairs, and d-MSCs were distinguishable from c-MSCs.

Conclusions: The interaction of c-MSCs and d-MSCs, including exchange of cell contents, contributes to the treatment potential of d-MSCs. This cellular behavior might be one therapeutic mechanism used by MSCs for MRONJ.

Keywords: Medication-related osteonecrosis of the jaw; Mesenchymal stem cell; Therapeutic mechanism.

Fig. 1

Multipotential differentiation of MSCs. a Experimental protocol for the in-vivo study. To isolate diseased MSCs (d-MSCs), wild-type C57BL/6 N mice received both Dex and Zol administered intravenously for 2 weeks with tooth extraction. Control MSCs (c-MSCs) were isolated from normal mice, which received no treatment (n = 4 × 2 groups). b (Top left) Osteogenic differentiation of MSCs. After culture under osteogenic differentiation conditions for 4 weeks, osteogenic differentiation was determined by Alizarin Red S staining. Quantification of the Alizarin Red S dye content in differentiated osteoblasts from independent experiments is shown (mean ± SD). Scale bar, 50 μm. (Top middle) Adipogenic differentiation of MSCs. After culture under adipogenic differentiation conditions for 2 weeks, adipocyte differentiation was determined by Oil Red O staining. Zol treatment decreased adipogenic differentiation of MSCs, as indicated by the decreased number of Oil Red O-positive cells (mean ± SD). *P < 0.05. (Top right) Endothelial differentiation of MSCs. After culture under endothelial differentiation conditions for 2 weeks, differentiation was determined by the formation of typical vessel-like-structures. Scale bar, 50 μm. (Bottom left) MSC migration determined by fluorescence microscopy. Larger numbers of c-MSCs went through the transwell insert compared with d-MSCs. (Bottom middle) Rate of proliferation as determined by the EdU assay. The number of EdU-positive cells was significantly decreased in the d-MSCs from the Zol-treated group compared with c-MSCs from the normal group. (Bottom right) MSCs from normal mice generated fewer CFU-Fs compared with MSCs from MRONJ-like mice. CFU-F colony-forming unit-fibroblast, d days, EdU 5-ethynyl-2′-deoxyuridine, i.p. intraperitoneal, i.v. intravenous

Fig. 2

Epithelial healing after tooth extraction in a histological study. a Experimental protocol for the in-vivo experiments. Wild-type C57BL/6 N mice received both dexamethasone (Dex) and zoledronate (Zol) administered intravenously for 2 weeks. c-MSCs or d-MSCs were injected into the mice a day after tooth extraction (c-MSC(+) and d-MSC(+)). Control mice received no treatment and MRONJ mice received both Dex and Zol but had no MSC treatment. b MRONJ model mice without treatment and with diseased MSCs had a lack of epithelial lining in the alveolar socket. In contrast, 100 % of the MSC-treated mice showed complete epithelial coverage, which was confirmed by the histological study. Moreover, MRONJ mice had higher IL-2 and IL-6 levels, and lower IL-10 levels than the controls. Inflammatory cytokine levels of the c-MSC(+) model were similar to those of the controls. The lung, kidney, liver, and spleen of the MRONJ model had obvious destruction of the typical structure, invasion of inflammatory cells, and necrosis of components. These inflammatory organ characteristics were recovered by c-MSCs but not d-MSCs. n = 5 × 4 groups. c-MSC control MSC, d days, d-MSC diseased MSC, i.p. intraperitoneal, i.v. intravenous, MRONJ medication-related osteonecrosis of the jaw, MSC mesenchymal stem cell

Fig. 3

Accumulation of GFP-transgenic injected MSCs into diseased mice. a Experimental protocol for the in-vivo experiments. Wild-type C57BL/6 N mice received both Dex and Zol administered intravenously. MSCs from GFP-transgenic mice were injected into the mice 1 day after tooth extraction. b In various organs and around the extraction site, injected MSCs (CD-90/GFP-FITC double-positive cells) selectively accumulated into the connective tissue close to the blood vessels. However, no double-positive MSCs were observed in the kidney and gingival mucosa. Scale bar, 100 μm. n = 5 × 2 groups. c-MSC control mesenchymal stem cell, d days, d-MSC diseased mesenchymal stem cell, GFP green fluorescent protein, i.p. intraperitoneal, i.v. intravenous. The right panel showed graphially the number of GFP and CD90 positive cells in each organs. Bars represent the mean ± SD of six independent experiments. *P<0.05

Fig. 4

Characteristics of MSCs from MRONJ mice treated with MSCs. a Experimental protocol. Wild-type C57BL/6 J mice received both Dex and Zol administered intravenously for 2 weeks (n = 5 × 2 groups). MSCs were isolated from MRONJ model mice treated with c-MSCs or d-MSCs a day after tooth extraction (M-cMSC and M-dMSC(+)). b MSC abnormalities in MRONJ were reversed by c-MSCs but not d-MSCs. CFU-F colony-forming unit-fibroblast, c-MSC control MSC, d days, d-MSC diseased MSC, EdU 5-ethynyl-2′-deoxyuridine, i.p. intraperitoneal, i.v. intravenous, MRONJ medication-related osteonecrosis of the jaw, MSC mesenchymal stem cell

Fig. 5

Intercellular exchange of MSCs. a Explanatory schema of the various MSCs in this experiment (n = 3 × 4 groups). b Experimental protocol for the in-vitro experiments. d-MSCs were cocultured with c-MSCs or d-MSCs for 24, 48, and 72 hours to analyze intercellular exchanges between MSCs. c Typical image of MSCs with mitochondria exchange. MSC-GFP(+), green; MSC-mito(+), red. d, e d-MSCs stained (red) for visualization of mitochondria were placed into direct coculture with c-MSCs stained with GFP (green). The mitochondria from the c-MSCs moved actively to the d-MSCs. Scale bar, 50 μm. White arrows show the intercellular exchanges in early stage. c-MSC control MSC, DAPI 4′,6-diamidino-2-phenylindole, d-MSC diseased MSC, GFP green fluorescent protein, i.p. intraperitoneal, i.v. intravenous, MRONJ medication-related osteonecrosis of the jaw, MSC mesenchymal stem cell

Fig. 6

Intercellular exchange between endogenous and exogenous MSCs. a Experimental protocol for the in-vivo experiments. Wild-type C57BL/6 N mice received both Dex and Zol administered intravenously. MSCs from GFP-transgenic mice were injected into the mice 1 day after tooth extraction. b A pair of continuous sections was considered the same section. A section was stained with an anti-GFP antibody, while the other was stained with an antibody against CD90 and subjected to mitochondrial staining. As a result, many CD90(+) (total MSCs) and GFP(−) cells (endogenous MSCs) with stained mitochondria were located at the upper portion of the socket (white arrows). Mitochondrial transfer from exogenous MSCs to endogenous MSC was observed in MRONJ model mice. n = 5 for sample, n = 3 for MSC injection. d days, DAPI 4′,6-diamidino-2-phenylindole, GFP green fluorescent protein, i.p. intraperitoneal, i.v. intravenous, MSC mesenchymal stem cell