Effects Of Bone Marrow Stromal Cells In The Periprosthetic Osteolysis Associated With Particulate Wear Debris

Periprosthetic osteolysis is the most common long-term complication of a total joint arthroplasty, often resulting in aseptic loosening of the implant, occurring in up to34%of younger implant recipients and usually requiring surgical revision. Particulate wear debris, continuously generated by articulating motions at the bearing surfaces, has been implicated as one of the primary causes of periprosthetic bone loss and implant loosening. As developing implants or bearing surfaces designs, variant types of wear particles with specific ingredient, dimension, and shape are formed, which may initiate different immune or inflammatory responses. Wear debris induces down-regulation or up-regulation of various pro-inflammatory cytokines and chemokines in a range of cell types at interface between implant component and the surrounding bone, such as macrophages, osteoclast precursor cells, osteoblasts, lymphocytes, fibroblasts etc. Concomitantly, these mediators further affect functions of cells, eventually leading to osteoclast activation and osteolysis. Bone marrow stromal cells (BMSCs), as origin of osteoblast, which locate in trabecular bone and closely contact with implant, are critical contributors to maintain homeostasis preserved by both formation and resorption of bone. Perturbation of MSC by implants and wear debris may thus affect bone ingrowth and interface stability, leading to increased risk of loosening.In this study, We first investigate variant types of orthopaedic biomaterial particles (Ti, PMMA, UHMWPE, Co-Cr) on differentiation and functions of bone marrow stromal cells (BMSCs), further we estimate the effect of Ti-particle challenged BMSCs/osteoblastic cells on the differentiation of osteoclast from peripheral blood mononuclear cells (PBMCs), a murine model of periprosthestic osteolysis is used to further investigate the regulatory effect of BMSCs onosteoclasts. Cells were isolated from femurs of BALB/c mice by density centrifugation over Histopaque(?)-1083and cultured in DMEM culture medium supplemented with10%fetal bovine serum (FBS) at37℃and5%CO2atmosphere, or cultured in complete-osteoblast-induction medium including lOmM β-glycerol phosphate,100nM L-ascorbic acid, and10nM dexamethasone after24h. Immunocytochemical stain of Osteocalcin, and Alkaline phosphatase (ALP) stain were performed to detect the successful induction of osteoblast phenotypes. The cells were respectively co-cultured with micron-sized Ti, PMMA, UHMWPE, and Co-Cr particles at the concentration of0.63,1.25,2.5or5 mg/ml. The supernatant was saved for induction of PBMCs. MTT assay was performed periodically (1,3,5,7day) for cell viability and proliferation, and cytotoxicity assay according to lactate dehydrogenase (LDH) activity was also performed; ALP assay in cell lysate and immunocytochemistry stain of osteocalcin, RANKL, and ALP were achieved to identify osteoblast differentiation. Levels of murine IL-1, IL-6and TNFa in culture media were determined by ELISA, while real-time PCR performed to examine gene expression of RANKL, OPG, LRP5, OSX, NFATcl and Runx2. Peripheral blood monocytes (PBMCs) were isolated from Balb/c mice and cultured in osteoblast-conditioned media. TRAP stain was performed to identify osteoclastogenesis. In vivo study, The mice with titanium-pin implants and Ti-particle challenges were divided into2groups post-op:(1) mice in the OB group were given an intra-articular injection of50ul medium containing5×105Ti particles-challenged OBs, and reinjection was performed at2weeks post-op (n=12);(2)12loosening control mice received an intra-articular injection of50ul cells-free medium;(3) The mice with neither Ti-particle challenge nor OBs injection were kept as stable control (n=12). MicroCT scans were used to evaluate bone mineral density (BMD). Mice were sacrificed at4weeks, and the implanted knee joint of each group was collected for pin-pullout testing, histological evaluation and RT-PCR analysis. The width of the soft-tissue membrane at the implant and surrounding bone interface was measured using Image-Pro+software. A commercial kit was used for histological TRAP staining on paraffin sectioned prosthetic joints alter pin removal. Gene expression profiles of RunX2, Osterix/Sp7, MMP-2,IL-1, TNF-α, RANKL and TRAP in in the peri-implant bone were evaluated by real-time PCR.As differentiation and proliferation proceeded, the cells gradually exhibited osteoblasts-like spindle or polygonal morphology in osteoblastogenic medium. Osteocalcin expression was visualized in majority of the conditioned cells (76.3±5.3%), and the positive staining for ALP was observed in85.3±3.9%cells. Challenge with low doses of Titanium, UHMWPE, or Co-Cr particles markedly promoted bone marrow cell proliferation while high dose of Co-Cr significantly inhibited cell growth (p<0.05). Interestingly, the MSCs co-cultured with PMMA particles appeared in slow growth pattern, even at low concentration (0.63mg/ml). However, co-culturing the cells with low dose of PMMA or UHMWPE particles (0.63mg/ml) in complete induction medium revealed relatively stronger ALP activity whereas Ti, and Co-Cr groups showed minimal ALP activity, in comparison with non-particle controls (p<0.05). Further, immunocytochemistry stain exhibited considerably more ALP/osteocalcin-positive cells in PMMA or UHMWPE challenged cultures than other groups. UHMWPE and Ti particles initiated higher expression of pro-inflammatory cytokines in challenged cells, in comparison with control group. Exposure to various particulate stimulus exhibited entirely disparate gene expression profiles, whereas cells treated by particles at low dose (0.5mg/ml) exhibited distinctly reduced RANKL generation than high concentration (3mg/ml), implicating that wear debris-interacted BMSCs may involve in osteoclast activation via RANKL. When the culture media collected from short-term challenged (48hours) osteoblasts was added to blood monocyte cultures, significantly more TRAP+cells were identified in comparison with cells receiving naive osteoblasts media (p<0.05). MicroCT scans indicated that the implants were well fixed with no obvious migration in the surgery cases without particle challenge, while titanium particle injections resulted in marked periprosthetic bone resorption, and the introduction of Ti-particles challenged OBs further reduced the BMD. Intra-articular introduction of the osteoblastic cells to the mouse pin-implant failure model resulted in reduced implant interfacial shear strength and thicker soft-tissue formation. Comparison of the gene expression profiles among the peri-implant bone following osteoblast injection did not find significant difference in RunX2or Osterix/Sp7between the groups. However, MMP-2, IL-1, TNF-α RANKL, and TRAP gene expressions were elevated in the challenged-osteoblast group (p<0.05), suggesting that titanium particles-challenged osteoblasts contributed to periprosthetic osteolysis.In conclusion, our data suggests that BMSCs co-cultured with various biomaterials that provide a particulate stimulus may result in completely different behaviours in cell proliferation, osteogenic effect, and pro-inflammatory response. Each type of particle exhibits varying degrees of suppressive effects on osteogenesis from BMSCs through the regulation of multiple gene expression associated with osteogenesis or osteoblast differentiation. Particle-stimulated BMSCs also respectively express a serial of pro-inflammatory genes, which could facilitate the peri-prosthesis inflammation and osteoclasts activation, and ultimately result in osteolysis and bone resorption. Significantly, osteoblastic cells stimulated by wear particles expressed higher level of RANKL, the key cytokine regulator of osteoclast differentiation and activation. Besides direct effect of Ti particles stimulating osteoblast to promote osteoclastogenesis through RANKL, several other possibilities may be considered in the peri-implant osteolysis problem, including TNF-a induce osteoclastogenesis by direct stimulation of macrophages and enzymatic bone resorption through MMPs. Indeed, it might be possible that these different scenarios (RANKL, MCP-1, TNF-α, MMPs) are simultaneously involved in the peri-implant osteolysis. Thus, the important result of the present study was to show that Ti particles might stimulate osteoblast to induce osteoclastogenesis and then be at the origin of the peri-implant osteolysis.