Periprosthetic joint infections (PJI) are increasing in prevalence and are recognised as one of the most common modes of failure of joint replacements. Osteomyelitis arising from PJI is challenging to treat, difficult to cure and increases patient mortality 5-fold. PJI can have subtle symptoms and lie dormant or go undiagnosed for many years, suggesting persistent bacterial infection. Staphylococcus aureus is the most common pathogen causing PJI. Osteocytes are the most numerous and long-lived cell type in hard bone tissue. Our recent work has shown that S. aureus can infect and reside in human osteocytes without causing cell death, both experimentally and in bone samples from patients with PJI. Osteocytes respond to infection by the differential regulation of a large number of genes, suggesting previously unknown immune functions of this important cell type. S. aureus adapts during intracellular infection of osteocytes by adopting a quasi-dormant, small colony variant (SCV) phenotype, a property of several bacterial species known to cause PJI, which could contribute to persistent or silent infection. These findings shed new light on the aetiology of PJI and osteomyelitis in general. Further elucidation of the role of osteocytes in bone infection will hopefully lead to improved disease detection and management

Objectives. Osteoporosis and osteomalacia lead to increased fracture risk. Previous studies documented dysregulated osteoblast and osteoclast activity, leading to a high-turnover phenotype, reduced bone mass and low bone mineral content. Osteocytes, the most abundant bone cell type, are involved in bone metabolism by enabling cell to cell interaction. Osteocytes presence and viability are crucial for bone tissue homeostasis and mechanical integrity. Osseo-integration and implant degradation are the main problems in developing biomaterials for systemically diseased bone. This study examines osteocyte localisation, morphology and on the implant surface and at the implant bone interface. Furthermore, the study investigates ECM proteins regulation correlated to osteocytes and mechanical competence in an ovariectomised rat model with a critical size metaphyseal defect. Methodology. After induction of osteoporosis, 60 female Sprague-Dawley rats were randomised into five groups: SrCPC (n=15), CPC (n=15), ScB30 (n=15), ScB30Sr20 (n=15) and empty defect (n=15). The left femur of all animals underwent a 4mm wedge-shaped metaphyseal osteotomy that was internally fixed with a T-shaped plate. The defect was then either filled with the above mentioned implants or left empty. After six weeks, histomorphometric analysis showed a statistically significant increase in bone formation at the tissue-implant interface in the SrCPC group compared to the other groups (p<0.01). Osteocyte morphology and networks were detected using silver and staining. ECM proteins were investigated through immunohistochemistry. Cellular populations were tested using enzyme histochemistry. Mineralisation was assessed using time of flight secondary ion mass spectrometry (TOF-SIMS). Statistical analysis was performed using Mann Whitney U test with Bonferroni correction. Results. In the SrCPC and compared to other test groups, osteocytes presence and morphology was enhanced. An increased osteocytic activity was also seen in ScB30Sr20 when compared to SCB30 alone. Local osteomalatic lesions characterised by the presence of excessive unmineralised osteoid as revealed by the VKVG staining in the intact bone was also seen. A regular pattern of osteocytes distribution reflecting a better bone maturation was also seen in case of the Sr substituted cements. Whereas in case of the ScB30 degenerated osteocytes with a comparatively irregular arrangement were seen. Nonetheless, ECM proteins indicating discrepant bone turnover (RANKL, OPG, BMP2, OCN; ASMA) were noticed to increase within these regions and were accompanied by the presence of apoptotic osteocytes. Interestingly, osteocytes were also localised near the blood vessels within the newly formed woven bone. On the other hand, osteocytes allocation at implant bone interface and on the implant surface were qualitatively better in the Sr substituted groups when compared to the other test groups. Furthermore, this correlates with healing enhancement and implant retention results obtained from the histomorphometry (BV/TV and Osteoclasts count). The first qualitative results of the sclerostin visualisation showed a lower expression in the Sr supplemented biomaterials compared to the Sr free ones. Conclusion. Osteoblasts, osteoclast and osteocytes are the key players to bone metabolism through production and mineralisation of ECM or resorption. The current study indicates the importance in therapeutically targeting osteocytes to regulate bone metabolism in osteoporotic/osteomalatic bone. Sr inhibits osteoclast activity which is important for implant degradation. However, in osteoporotic bone osteoclasts inhibition is crucial to enhance the healing. Our data suggest that osteocytes allocation at the bone implant interface and on the implant surface is aiding in implant degradation through osteocytes dependent resorption. Currently, discrepancies in mechanosensitivity, proliferation and fibrotic tissue formation are being investigated together with several anchorage proteins to quench further effects of osteocyte presence at the implant bone interface

Abstract. Objectives. Osteocytes function as critical regulators of bone homeostasis by sensing mechanical signals. Stimulation of the mechanosensitive ion channel, Piezo1 promotes bone anabolism and deletion of Piezo1 in osteoblasts and osteocytes decreases bone mass and bone strength in mice. This study determined whether loading of osteocytes in vitro results in upregulation of the Piezo1 pathway. Methods. Human MSC cells (Y201), embedded in type I collagen gels and differentiated to osteocytes in osteogenic media for 7-days, were subjected to pathophysiological load (5000 µstrain, 10Hz, 5 mins; n=6) with unloaded cells as controls (n=4). RNA was extracted 1-hr post load and Piezo1 activation assessed by RNAseq analysis (NovaSeq S1 flow cell 2 × 100bp PE reads). To mimic mechanical load and activate Piezo1, Y201s were differentiated to osteocytes in 3D gels for 13 days and treated, with Yoda1 (5µM, 2 hours, n=4); vehicle treated cells served as controls (n=4). Extracted RNA was subjected to RT-qPCR and data analysed by Minitab. Results. Low mRNA expression of PIEZO1 in unloaded cells was upregulated 5-fold following 1-hr of mechanical load (p=0.003). In addition, the transcription factor NFATc1, a known regulator of Piezo1 mechanotransduction, was also upregulated by load (2.4-fold; p=0.03). Y201 cells differentiated in gels expressed the osteocyte marker, SOST. Yoda1 upregulated PIEZO1 (1.7-fold; p=0.057), the early mechanical response gene, cFOS (4-fold; p=0.006), COL1A1 (3.9-fold; p=0.052), and IL-6 expression (7.7-fold; p=0.001). Discussion. This study reveals PIEZO1 as an important mechanosenser in osteocytes. Piezo 1 mediated increases in the bone matrix protein, type I collagen, and IL-6, a cytokine that drives inflammation and bone resorption. This provides a direct link between mechanical activation of Piezo 1, bone remodelling and inflammation, which may contribute to mechanically-induced joint degeneration in osteoarthritis. Mechanistically, we hypothesise this may occur through promoting Ca2+ influx and activation of the NFAT1 signalling pathway

Stimulation of the mechanosensitive ion channel, Piezo1 promotes bone anabolism and SNPs in the Piezo1 locus are associated with changes in fracture risk. Osteocytes function as critical regulators of bone homeostasis by sensing mechanical signals. The current study used a human, cell-based physiological, 3D in vitro model of bone to determine whether loading of osteocytes in vitro results in upregulation of the Piezo1 pathway. Human Y201 MSCs, embedded in type I collagen gels and differentiated to osteocytes for 7-days, were subjected to pathophysiological load (5000 µstrain, 10Hz, 5 mins; n=6) with unloaded cells as controls (n=4). RNA was extracted 1-hr post load and assessed by RNAseq analysis. To mimic mechanical load and activate Piezo1, cells were differentiated to osteocytes for 13 days and treated ± Yoda1 (5µM, 2- and 24-hs, n=4); vehicle treated cells served as controls (n=4). RNA was subjected to RT-qPCR and data normalised to the housekeeping gene, YWHAZ. Media was analysed for IL6 release by ELISA. Mechanical load upregulated Piezo1 gene expression (16.5-fold, p<0.001) and expression of the transcription factor NFATc1, and matricellular protein CYR61, known regulators of Piezo1 mechanotransduction (3-fold; p= 5.0E-5 and 6.8-fold; p= 6.0E-5, respectively). After 2-hrs, Yoda1 increased the expression of the early mechanical response gene, cFOS (11-fold; p=0.021), mean Piezo1 expression (2.3-fold) and IL-6 expression (103-fold, p<0.001). Yoda1 increased the release of IL6 protein after 24 hours (7.5-fold, p=0.001). This study confirms Piezo1 as an important mechanosensor in osteocytes. Piezo1 activation mediated an increase in IL6, a cytokine that drives inflammation and bone resorption providing a direct link between mechanical activation of Piezo1, bone remodeling and inflammation, which may contribute to mechanically induced joint degeneration in diseases such as osteoarthritis. Mechanistically, we hypothesize this may occur through promoting Ca2+ influx and activation of the NFATc1 signaling pathway

Osteocytes direct bone adaptation to mechanical loading (e.g., exercise), but the ways in which osteocytes detect loading remain unclear. We recently showed that osteocytes develop repairable plasma membrane disruptions (PMD) in response to treadmill-running exercise, and that these PMD initiate mechanotransduction. As treadmill running is a non-voluntary activity for rodents, our current goal was to determine whether osteocytes develop PMD with voluntary wheel running as a better model of physiological exercise. Male and female Hsd:ICR mice from lines selectively bred (>75 generations) to demonstrate high voluntary wheel running (“High Runners”) or non-selected control lines (“Control”) were studied (n=9 to 12 mice per sex per line, 4 lines each). At 12 weeks of age, half of the animals within each group were provided access to running wheels for 6 days while remaining mice had no wheel access. Tibias were collected at sacrifice and bone mineral density was analyzed by DXA. Osteocyte PMD were quantified by immunochemistry for intracellular albumin. Groups were compared with 3-factor ANOVA. Voluntary exercise (wheel access) significantly increased osteocyte PMD (+16.4%, p=0.013). PMD-labelled osteocytes did not differ between sexes (p=0.415). Male mice had significantly greater BMD (p=0.0007) and BMC (<0.0001) than females. Interestingly, mice with wheel access had significantly lower BMD and BMC compared to mice without wheel access (p<0.004), and high runner lines had significantly lower BMD (p=0.001) and BMC (p<0.0001) than control lines. This may reflect new bone formation in the exercising mice, as newly formed bone is less mineralized than older bone. Data from this experiment support the idea that loading-induced disruptions develop in the osteocyte plasma membrane during both voluntary (wheel running) and forced (treadmill, shown previously) physical activity. These studies support the role of plasma membrane disruptions as a mechanosensation mechanism in osteocytes

Osteoarthritis (OA) is a common cause of chronic pain. Subchondral bone is highly innervated, and bone structural changes directly correlate with pain in OA. Mechanisms underlying skeletal–neural interactions are under-investigated. Bone derived axon guidance molecules are known to regulate bone remodelling. Such signals in the nervous system regulate neural plasticity, branching and neural inflammation. Perturbation of these signals during OA disease progression may disrupt sensory afferents activity, affecting tissue integrity, nociception, and proprioception.

Osteocyte mechanical loading and IL-6 stimulation alters axon guidance signalling influencing innervation, proprioception, and nociception.

Human Y201 MSC cells, embedded in 3D type I collagen gels (0.05 × 106 cell/gel) in 48 well plastic or silicone (load) plates, were differentiated to osteocytes for 7 days before stimulation with IL-6 (5ng/ml) with soluble IL-6 receptor (sIL-6r (40ng/ml) or unstimulated (n=5/group), or mechanically loaded (5000 μstrain, 10Hz, 3000 cycles) or not loaded (n=5/group). RNA extracted 1hr and 24hrs post load was quantified by RNAseq whole transcriptome analysis (NovaSeq S1 flow cell 2 × 100bp PE reads and differentially expressed neurotransmitters identified (>2-fold change in DEseq2 analysis on normalised count data with FDR p<0.05). After 24 hours, extracted IL-6 stimulated RNA was quantified by RT-qPCR for neurotrophic factors using 2–∆∆Ct method (efficiency=94-106%) normalised to reference gene GAPDH (stability = 1.12 REfinder). Normally distributed data with homogenous variances was analysed by two-tailed t test.

All detected axonal guidance genes were regulated by mechanical load. Axonal guidance genes were both down-regulated (Netrin1 0.16-fold, p=0.001; Sema3A 0.4-fold, p<0.001; SEMA3C (0.4-fold, p<0.001), and up-regulated (SLIT2 2.3-fold, p<0.001; CXCL12 5-fold, p<0.001; SEMA3B 13-fold, p<0.001; SEMA4F 2-fold, p<0.001) by mechanical load. IL6 and IL6sR stimulation upregulated SEMA3A (7-fold, p=0.01), its receptor Plexin1 (3-fold, p=0.03). Neutrophins analysed in IL6 stimulated RNA did not show regulation.

Here we show osteocytes regulate multiple factors which may influence innervation, nociception, and proprioception upon inflammatory or mechanical insult. Future studies will establish how these factors may combine and affect nerve activity during OA disease progression.

Abstract

Abstract

Introduction and Objective. The osteocyte, recognized as a major orchestrator of osteoblast and osteoclast activity, is the most important key player during bone remodeling processes. Imbalances that occur during bone remodeling, caused by hormone perturbations or alterations in mechanical loading, can induce bone disease as osteoporosis. Due to limited understanding of the underlying mechanisms, current therapies for osteoporosis cannot adequately address this imbalance because current studies of osteocytes rely on conventional cell culture that cannot recapitulate local in vivo microenvironments for the lack of control of the spatial/temporal distribution of cells and biomolecules. Microfluidics is the science and technology of microscale fluid manipulating and sensing and can help fill this gap. Materials and Methods. We used a microfluidic device to enable the culture of osteocyte-like cells (MLO-Y4 and MLO-A5) in a 3D fashion. Osteocytes were cultured in a perfused and 160 μm high channel and embedded in a bone-like extracellular matrix: osteocytes were embedded in a matrigel- and collagen-based hydrogel enriched with nanostructured hydroxypatite crystals (HA-NP) to mimic bone. To set up the best combination of matrigel enriched with Type I collagen we used fluorescent microspheres and confocal analysis. To evaluate the viability and the expression of osteocytic markers, we used live-dead assay amd immunofluorescent staining and confocal analysis combined with automated quantification. For mineralization, we performed alizarin red staining. Results. Osteocytes in the organ-on-a-chip model showed high viability and, in respect to 2D conventional cell cultures an increased differentiation, as assessed by a live-dead assay and the staining of the osteocytic markers connexin-43 and alkaline phosphatase and the increased mineralization activity. Furthermore, the addition of HA-NP significantly increased the formation of dendrite-like structures spreading through the xyz-axes, as assessed after G-actin immunofluorescence. Conclusions. Using a microfluidic device for MLO-Y4 and MLO-A5 cell cultures, compared to the 2D surfaces, we demonstrated a significant difference in cell differentiation and morphology. In particular, 3D cultures allowed the formation of 3D cell networks and the osteogenic phenotype. As a platform technology, this microfluidic device can function as a novel cell culture model that enables further studies of osteocytes and 3D co-culturing with other bone cells for the screening of anti-osteoporotic drugs

Osteoporosis is a progressive, chronic disease of bone metabolism, characterized by decreased bone mass and mineral density, predisposing individuals to an increased risk of fractures. The use of animal models, which is the gold standard for the screening of anti-osteoporosis drugs, raises numerous ethical concerns and is highly debated because the composition and structure of animal bones is very different from human bones. In addition, there is currently a poor translation of pre-clinical efficacy in animal models to human trials, meaning that there is a need for an alternative method of screening and evaluating new therapeutics for metabolic bone disorders, in vitro. The aim of this project is to develop a 3D Bone-On-A-Chip that summarizes the spatial orientation and mutual influences of the key cellular components of bone tissue, in a citrate and hydroxyapatite-enriched 3D matrix, acting as a 3D model of osteoporosis. To this purpose, a polydimethylsiloxane microfluidic device was developed by CAD modelling, stereolithography and replica molding. The device is composed by two layers: (i) a bottom layer for a 3D culture of osteocytes embedded in an osteomimetic collagen-enriched matrigel matrix with citrate-doped hydroxyapatite nanocrystals, and (ii) a upper layer for a 2D perfused co-culture of osteoblasts and osteoclasts seeded on a microporous PET membrane. Cell vitality was evaluated via live/dead assay. Bone deposition and bone resorption was analysed respectively with ALP, Alizarin RED and TRACP staining. Osteocytes dendrite expression was evaluated via immunofluorescence. Subsequently, the model was validated as drug screening platform inducing osteocytes apoptosis and administrating standard anti-osteoporotic drugs. This device has the potential to substitute or minimize animal models in pre-clinical studies of osteoporosis, contributing to pave the way for a more precise and punctual personalized treatment

Introduction. Bone and joint infection (BJI) is often characterized by severe inflammation and progressive bone destruction. Osteocytes are the most numerous and long-lived bone cell type, and therefore represent a potentially important long-term reservoir of bacterial infection. Staphylococcus aureus is known to establish stable intracellular osteocytic infections, however, little is known about the less virulent yet second most prevalent BJI pathogen, S. epidermidis, associated with late-diagnosed, chronic BJI. Thus, this study sought to establish an in vitro model to study the infection characteristics of S. epidermidis in human osteocyte-like cells. Methods. SaOS2 cells (1 ×10. 4. cells/cm. 2. ) were grown to confluence either without differentiation, representing an osteoblast-like (OB) state (SaOS2-OB) or differentiated to an osteocyte-like stage (SaOS2-OY), using established methods. Four S. epidermidis strains used (ATCC-12228, ATCC-14990, ATCC-35984 and a clinical osteomyelitis strain RAH-SE1) were tested to be Lysostaphin-resistant, necessitating antibiotic (Levofloxacin) control of extracellular bacteria. Infection of host cells (OB or OY) was tested at three multiplicities of infection (MOI: 10, 100 and 1000). Extracellular bacteria were controlled by overnight incubation at a 10X minimum inhibitory concentration (MIC) of Levofloxacin and thereafter at 1XMIC. At each time point (days 1, 3, 5) viable intra- and extracellular bacteria were quantified. Result. All strains displayed similar intracellular infection and persistence capabilities in SaOS2-OB and SaOS2-OY. Independent of MOI, intracellular bacteria in SaOS2-OB decreased over time, becoming non-culturable by day 5. In contrast, SaOs2-OY displayed enhanced intracellular bacterial persistence at each time point. In the presence of increased Levofloxacin concentration (10XMIC), S. epidermidis could persist intracellularly for at least 14 days. Conclusion. This study showed for the first time that S. epidermidis can infect human osteocytes and persist intracellularly. Additionally, even a 10xMIC antibiotic concentration failed to eradicate intracellular bacteria, suggesting that persistence within osteocytes could contribute to treatment failure and establishment of chronic BJI

Introduction and Objective. Klinefelter Syndrome (KS, karyotype 47,XXY) is the most frequent chromosomal aneuploidy in males, as well as the most common cause of infertility in men. Patients suffer from a lack of testosterone, i.e. hypergonadotropic hypogonadism provoking infertility, but KS men also show an increased predisposition to osteoporosis and a higher risk of bone fracture. In a mouse model for human KS, bone analysis of adult mice revealed a decrease in bone mass that could not be rescued by testosterone replacement, suggesting a gene dosage effect originating from the supernumerary X-chromosome on bone metabolism. Usually, X chromosome inactivation (XCI) compensates for the dosage imbalance of X-chromosomal genes between sexes. Some studies suggested that expression of genes that escape silencing of the supernumerary X-chromosome (e.g. androgen receptor) has an impact on sex differences, but may also cause pathological changes in males. As a promising new such candidate for a musculoskeletal escape gene, we identified the integral membrane protein (ITM) 2a, which is encoded on the X-chromosome and related to enchondral ossification. The aim of the project was to characterize systemic bone loss in the course of aging in our KS mouse model, and whether the supernumerary X-chromosome causes differences in expression of genes related to bone development. Materials and Methods. Bone structure of 24 month (=aged) old male wild type (WT) and 41, XXY mice (B6Ei.Lt-Y) were analysed by μCT. Afterwards bones were paraffin embedded and cut. In addition, tissue of brain, liver, kidney, lung and heart were also isolated and embedded for IHC staining. Using an anti-ITM2a antibody, expression and cellular localization of ITM2a was evaluated. IHC was also performed on musculoskeletal tissue of WT embryos (E18.5) and neonatal mice to determine possible age-related differences. Results. In 24 month old mice, the analysis of the lumbar vertebrae revealed a significantly lower BV/TV, trabecular bone volume and trabecular number in the XXY- group compared to WT. Trabecular thickness appeared lower but did not reach significance, with the cortical thickness being significantly higher in the XXY- group. High expression of ITM2a was detected in bone slices of both karyotypes in the chondrocytes inside the growth plate, as well as in megakaryocytes and leucocytes as well as endothelial cells of blood vessels inside the bone marrow. Osteocytes, along with erythrocytes and erythropoetic stem cells were negative for ITM2a. Other organs that showed ITM2a positive staining were kidney (blood vessels), heart (muscle) and brain (different structures). Liver and lung tissue were negative for ITM2a. No obvious difference in the intensity of the ITM2a-expression was observed between the WT and the XXY-karyotype. Analyses of embryotic bone tissue (WT) showed high expression of ITM2a in proliferating, hypertrophic and resting chondrocytes in the growth plates of tibia and femur. In comparison, the neonatal animals (WT) did not show any protein-expression in chondrocytes. Furthermore, within the metaphysis of both, embryotic and neonatal bones, endothelial cells and osteoblasts were ITM2a-positive. Further analyses of bones and tissues from young mice (4–6 month) are ongoing. Conclusions. Bone analyses revealed a significant reduction in trabecular bone mass along with fewer and thinner trabeculae in XXY mice compared to the WT, especially in the spine. ITM2a expression was visible in different cell types inside the bone, and in addition, different expression patterns at different stages of development (embryonic/neonatal) were observed. However, we have not found a significant difference in the quantity of ITM2a between tissues of XXY-karyotypes and WT. Further analyses of X-chromosomal encoded and therefore dysregulated modulators in XXY-karyotype mice and patients may reveal new sex chromosomal effector proteins in bone metabolism

Osteocytes are terminally differentiated long-lived cells and account for greater than 95% of the bone cell population. It has been established that osteocytes are connected through their highly developed dendritic network, which is necessary for the maintenance of optimal bone homeostasis. However, little is known on how osteocytes use the network to coordinate their cellular function and communication that requires energy and protein turnover. Here using super-resolution confocal imaging on both live and fixed osteocytes, we demonstrated conclusively that mitochondria are widely distributed and dynamically shared between osteocytes. Using confocal live cell imaging analysis we showed that inhibiting the contact between mitochondria and endoplasmic reticulum (ER) by the knockdown of MFN2 in osteocytes impedes the transfer of mitochondria suggesting the involvement of ER contact with mitochondria in the transfer process. Moreover, we showed that transport of mitochondria between osteocytes within the network enables rescue of osteocytes with dysfunction of mitochondria. Using the 3D tetraculture system with confocal imaging, we identify the transfer of mitochondria from healthy osteocytes enables recovery of mitochondria activities in osteocytes that devoid of mitochondrial DNA by ethidium bromide. The results indicated that when osteocytes are depleted of functional mitochondria, normal parental osteocytes can transfer mitochondria to these stressed osteocytes to provide them with energy. Collectively we show for the first time that the utilisation of mitochondrial transfer enables osteocytes to function with a network and coordinate their cellular activities in response to different energy demands

Introduction. Elevated remodelling of subchondral bone and marrow tissues has been firmly established as diagnostic and prognostic radiological imaging marker for human osteoarthritis. While these tissues are considered as promising targets for disease-modifying OA drugs, the development of novel treatment approaches is complicated by the lack of knowledge whether similar tissue changes occur in rodent OA models and poor understanding of joint-specific molecular and cellular pathomechanisms in human OA. Here, we describe the establishment of a human OA explant model to address this crucial niche in translational preclinical OA research. Methods. Osteochondral (knee, spine) and bone (iliac crest) clinical specimens were acquired from patients undergoing total knee arthroplasty (n=4) or lumbar spine fusion using bone autografts (n=6). Fresh specimens were immediately cut in equal-sized samples (50–500 mg wet weight) and cultured in 8 mL osteogenic medium for one week. Samples were either left untreated (control) or stimulated with lipopolysaccharide (LPS, 100 ng/mL) in the absence and presence of transforming growth factor-beta inhibitor (SB-505124, 10 μm). Pro-collagen-I (Col-I), interleukin-6 (IL-6) and monocyte chemoattractant protein 1 (MCP-1) secretion was determined in conditioned medium by ELISA. Tissue viability was assessed using MTT and alkaline phosphatase (ALP) activity staining. Results. Explanted tissues remained viable after one week culture in control and treatment conditions. Osteocytes, subchondral marrow spaces and calcified cartilage stained positive for ALP activity without gross morphological differences between groups. Median basal secretion levels were Col-I (2.3 ng/mg), IL-6 (90 pg/mg) and MCP-1 (25 pg/mg). LPS treatment led to a significant increase of IL-6 (330 pg/mg) and MCP-1 (70 pg/mg), but not Col-I secretion. Interestingly, inhibition of TGF-beta signalling in osteochondral tissues specifically reduced Col-I levels (0.4 ng/mg) compared to controls and LPS-treated samples. LPS-induced IL-6 and MCP-1 levels were slightly reduced (−120 pg/mg, p=0.03) and increased (+50 pg/mg) by SB-505124 treatment, respectively. IL-6 and MCP-1 levels were strongly correlated under basal (r=0.80) and treatment conditions (r=0.62). Conclusion. In this study, we provided proof of concept for the first ex vivo explant model of human osteoarthritis. Osteochondral tissue specimens can readily be cultured without loss of tissue viability and mount a robust inflammatory response upon LPS challenge. Treatment with a potential disease-modifying agent (TGF-beta signalling inhibitor) reduced collagen metabolism in bone and marrow and modified cytokine and chemokine expression. The osteochondral explant model might be highly valuable to evaluate disease-modifying OA drugs