Matrix-bound vesicles (MBVs) are embedded within osteoid and function as the site of initial mineral formation. However, they remain insufficiently characterised in terms of biogenesis, composition and function while their relationship with secreted culture medium EVs (sEVs) such as exosomes remains debated. We aimed to define the biogenesis and pro-mineralisation capacity of MBVs and sEVs to understand their potential in regenerative orthopaedics.

sEVs and MBVs isolated from conditioned medium (differential ultracentrifugation) and ECM (collagenase digestion and differential ultracentrifugation) of mineralising MC3T3 pre-osteoblast and human bone marrow MSC cultures were characterised by nanoparticle tracking analysis, western blotting, nano-flow cytometry, super resolution microscopy (ONI) and TEM. Immunoprecipitated populations positive for alkaline phosphatase (ALP), a putative marker of mineralisation capacity, were also characterised. Collagen binding efficiency was evaluated using MemGlow staining.

Results reported were comparative across both cell lines. Western blots indicated MBV fractions were positive for markers of endosomal biogenesis (CD9, CD81, ALIX, TSG101) and pro-mineralising proteins (ALP, Pit1, Annexin II, Annexin V), with Annexin V and CD9 present in immunoprecipitated ALP-positive fractions. MBVs were significantly larger than sEVs (p<0.05) and contained a higher amount of ALP (p<0.05) with a significant increase from day 7 to day 14 of cellular mineralisation (p<0.05). This mirrored the pattern of electron-dense vesicles seen via TEM. Super resolution single vesicle analysis revealed for the first-time co-expression of ALP with markers of endosomal biogenesis (CD9, CD63, CD81, ALIX) and Annexin II in both vesicle types, with higher co-expression percentage in MBVs than sEVs. MBVs also exhibited preferential collagen binding.

Advanced imaging methods demonstrated that contrary to opinions in the field, MBVs appear to possess exosomal markers and may arise via endosomal biogenesis. However, it was evident that a higher proportion of MBVs possessed machinery to induce mineralisation and were enriched in mineral-dense material.

The implantation of endoprosthesis is a routine procedure in orthopaedics. Endoprosthesis are mainly manufactured from ceramics, polymers, metals or metal alloys. To ensure longevity of the implants they should be as biocompatible as possible and ideally have antibacterial properties, to avoid periprosthetic joint infections (PJI). Various antibacterial implant materials have been proposed, but have so far only been used sporadically in patients. PJI is one of the main risk factors for revision surgeries. The aim of the study was to identify novel implant coatings that both exhibit antibacterial properties whilst having optimal biocompatibility.

Six different novel implant coatings and surface modifications (EBM TiAl6V4, strontium, TiCuN, TiNbN, gentamicin phosphate (GP), gentamicin phosphate+cationic polymer (GP+CP)) were compared to standard CoCrMo-alloy. The coatings were further characterized with regard to the surface roughness. E. coli and S. capitis were cultured on the modified surfaces to investigate the antibacterial properties. To quantify bacterial proliferation the optical density (OD) was measured and viability was determined using colony forming units (CFU). Murine bone marrow derived macrophages (BMMs) were cultured on the surfaces and differentiated into osteoblasts to quantify the mineralisation using the alizarin red assay.

All novel coatings showed reduced bacterial proliferation and viability compared to standard CoCrMo-alloy. A significant reduction was observed for GP and GP+CP coated samples compared to CoCrMo (OD_GP,E.coli = 0.18±0.4; OD_GP+CP,E.coli = 0.13±0.3; p≤0.0002; N≥7-8). An increase in osteoblast-mediated mineralisation was observed on all surfaces tested compared to CoCrMo. Furthermore, GP and GP+CP coated samples showed a statistically significant increase (M_GP = 0.21±0.1; M_GP+CP = 0.25±0.2; p<0.0001; N≥3-6).

The preliminary data indicates that the gentamicin containing surfaces have the most effective antibacterial property and the highest osseointegrative capacity. The use of antibiotic coatings on prostheses could reduce the risk of PJI while being applied on osseointegrative implant surfaces.

Engineered cartilage is poorly organized and fails to recapitulate physiologic organization in a hyaline upper and a mineralizing bottom zone deemed important for proper function. Objective was to grow bizonal human cartilage constructs in which in vivo mineralization is self-restricted to the bottom zone. Self-assembling biomaterial-free cell discs were generated from mesenchymal stroma cells and allowed to accumulate proteoglycans and collagen-type II over 3 weeks. In vitro mineralization of the cell discs with four mineralization media for up to 8 weeks showed that calcification was supported in all media containing ß-glycerophosphate. However, proteoglycans were retained only in media containing insulin. Bizonal cartilage constructs were made from 3-week non-mineralized cell discs overlaid with chondrocyte-seeded starPEG-heparin hydrogel or with a fibrin-gel layer to select the best design for upper zone development. Freshly prepared zonal constructs were implanted into subcutaneous pouches of immuno-deficient mice to compare in vivo development. After 6 weeks in vivo, both construct types were rich in collagen-type II in the upper zone and contained a mineralized bottom zone. However, solely for starPEG constructs, tissue volume of the upper zone remained high and alkaline phosphatase, alizarin red, and collagen-type X staining were restricted to the bottom zone. StarPEG zonal constructs were superior to fibrin constructs due to self-restriction of mineralization and hypertrophic markers to the bottom zone. This innovative design of bizonal constructs offers the successful generation of an organized cartilage resembling the native cartilage with the chance for immediate use of autogenous chondrocytes in a one-step surgical joint intervention.

Osteophyte deposition and subchondral bone damage are notable features of osteoarthritis (OA). Deregulated mineralization contributes to osteophyte and subchondral irregularity. The microRNA-29 (miR-29) family is associated with arthritic disorders. This study is aimed to investigate miR-29a function to OA osteophyte formation and subchondral integrity. Intact and damaged articular cartilage in patients with end-stage knee OA who required total knee arthroplasty were harvested to probe miR-29a, cartilage, and mineralized matrix expression using RT-PCR and in situ hybridization. Osteophyte volume and subchondral morphometry of collagenase-induced OA knees in mice were quantified using μCT and histomorphometry. Increased bone matrix expression (collagen I and bone alkaline phosphatase) and reduced cartilage matrix (collagen II and aggrecan) along with low miR-29a expression existed in human OA specimens. Aged miR-29a knockout mice showed spontaneous osteophyte formation and articular cartilage erosion. In primary articular chondrocytes, miR-29a deficiency significantly reduced cartilage matrix synthesis, whereas von Kossa staining-positive mineralized matrix production was increased. Of interest, the severity of collagenase-induced osteophyte accumulation and subchondral damage along with serum cartilage breakdown products CTX-II and COMP levels were significantly compromised in mice overexpressing miR-29a. Intra-articularly injecting miR-29a significantly reduced osteophyte volume and subchondral integrity and retained cartilage morphology in collagenase-injured knees. Reduced miR-29a signalling worsens osteophyte and subchondral destruction in OA through increasing mineralized matrix formation of chondrocytes. Restoring miR-29a shields joints from cartilage degradation, osteophyte and subchondral destruction. This study conveys new mechanistic underlying OA osteophyte pathogenesis and shines light on the remedial potential of miR-29a to OA.

During remodelling, osteoclasts produce discrete bone cavities filled with bone and this is associated with the dimensions of the cavity. The aim of this study is to investigate the effect of pores of similar size to those produced by osteoclasts on the morphology, proliferation and osteogenic differentiation of bone marrow mesenchymal stem cells (BMSCs) in vitro. The hypothesis is that a porous surface similar in morphology to a bone surface prepared by osteoclasts will increase cell proliferation and osteogenic differentiation of MSCs.

Sheep BMSCs were seeded onto plain titanium surfaces and 100µm, 250µm and 500µm discrete pores surfaces. Cell metabolic activity was investigated using Presto Blue on days 3, 7 and 10. Bone mineralisation was quantified by Alizarin red staining at days 3, 7 and 14. Cell morphology was observed by scanning electron microscopy (SEM). Data was statistically analysed using one-way analysis of variance and a Bonferroni correction method.

Cells on porous discs had a three dimensional phenotype and aligned on the circumference of each pore. Metabolic activity was significantly higher by day 10 on plain discs compared to all porous discs. Bone mineralization was significantly higher on 100µm pores by day 3 (0.545mM±0.66; p=0.047) than plain discs and significantly higher on both 100µm and 250µm pores by day 7(p=0.000 and p=0.005) than plain discs. Substantial mineralised bone matrix was found on 100µm discs without being treated with osteogenic supplements, compared to other control disc types (p=0.043, p=0.003, p=0.000).

The different topographies altered cell behaviour and migration.100µm pores demonstrated earlier and enhanced bone mineralisation even in the absence of osteogenic supplements. This pore size is aligned to the size of individual resorption bays that osteoclasts produce on bone surfaces and is considerably lower than the pore sizes used to enhance osteo-integration of implant surfaces.

Bone homeostasis is a highly regulated process involving pathways in bone as WNT, FGF or BMP, but also requiring support from surrounding tissues as vessels and nerves. In bone diseases, the bone-vessel-nerve triad is impacted. Recently, new players appeared as regulators of bone homeostasis: microRNAs (miRNA). Five miRNAs associated with osteoporotic fractures are already known, among which miR-125b is decreasing bone formation by downregulating human mesenchymal stem cells (hMSCs) differentiation. Other miRNAs, as miR-214 (in cluster with miR-199a), are secreted by osteoclasts to regulate osteoblasts and inhibit bone formation. This forms a very complex regulatory network. hMSCs and osteoblasts (n=3) were transfected with mimic/antagomiR of miR-125b, miR-199a-5p or miR-214, or with a scrambled miRNA (negative control) in osteogenic differentiation calcium-enriched medium (Ca++). Mineralization was assessed by Alizarin Red/CPC staining, miRNA expression by qPCR and protein by western blotting. Exposure of hMSCs or osteoblasts to Ca++ increased mineralization compared to basal medium. hMSCs transfected with miR-125b mimic in Ca++ presented less mineralization compared to scramble. This correlated with decreased levels of BMPR2 and RUNX2. hMSCs transfected with miR-125b inhibitor presented higher mineralization. Interestingly, hMSCs transfected with miR-214 mimic in Ca++ presented no mineralization while miR-214 inhibitor increased mineralization. No differences were observed in hMSCs transfected with miR-199a-5p modulators. On the contrary, osteoblasts transfected with miR-199a-5p mimic present less mineralization than scrambled-transfected and same was observed for miR-214 and miR-125b mimics. We highlight that miR-125b and miR-214 decrease mineralization of hMSCs in calcium-enriched medium. We noticed that miR-199a-5p is able to regulate mineralization in osteoblasts but not in hMSCs suggesting that this effect is cell-specific. Interestingly, the cluster miR-199a/214 is known as modulator of vascular function and could thus contribute to bone remodeling via different ways. With this work we slightly open the door to possible therapeutic approaches for bone diseases

Chondrogenic differentiation and cartilage homeostasis requires a high cellular translational capacity to meet the demands for cartilaginous extracellular matrix production. Box C/D and H/ACA snoRNAs guide post-transcriptional 2′-O ribose methylation and pseudouridylation of specific ribosomal RNA (rRNA) nucleotides, respectively. How specific rRNA modifications influence rRNA function is poorly documented, but modifications are thought to tune rRNA folding and interaction with ribosomal proteins, which is critical for ribosome function. We hypothesise that chondrocyte translational capacity is supported by snoRNA-mediated post-transcriptional fine-tuning of rRNAs. ATDC5 progenitor cells were differentiated into the chondrogenic lineage, resembling mature and mineralising chondrocytes after 7 or 14 days, respectively. UBF-1 (rRNA transcription factor), fibrillarin (box C/D methyltransferase) and dyskerin (box H/ACA pseudouridylase) expression displayed highest fold induction at day 5/6 in differentiation. Ribosomal RNA content per cell was increased at day 7, but not at day 14 in differentiation. These data suggest that ribosome biogenesis adapts to the chondrocyte's differentiation status. RNA-Seq of RNA species <200 nt revealed expression of at least 224 individual snoRNAs. Due to initiation of chondrogenic differentiation (Δt0-t7), 21 snoRNAs were differentially expressed (DE; FDRadj-p<0.05, logFC>1or<−1). Mineralization (Δt7-t14) induced DE of 23 snoRNAs. Comparing t0 with t14 resulted in DE of 43 snoRNAs. To anticipate on the biological relevance of DE snoRNAs, their rRNA target nucleotides were plotted in 18S, 5.8S and 28S rRNA secondary structures. This revealed that DE snoRNAs, amongst others, target nucleotide modifications in the 28S peptidyl transferase center and the 18S decoding center (DC). Snora40 was DE, targeting helix 27/18S rRNA. Helix 27 controls DC function. Helix 68 of 28S rRNA is part of the ribosome's E-site, therefore, DE snord36c and snora31 (targeting helix 68) could potentially fine-tune the translation mechanism. As a final example we found snord46 to be DE (target: helix 69/28S rRNA). Mutations in helix 69 have been shown to severely affect cell viability. Our data show that increased demand for translational capacity during chondrogenic differentiation is associated with differential expression of snoRNAs, potentially controlling ribosome fidelity via site-specific rRNA-modifications. These data enable us to determine the role of individual snoRNAs in tuning the chondrocyte's translational properties and current efforts focus on confirming site-specific rRNA-modifications and determine their biological relevance