A Machine Learning-Based Image Segmentation Method to Quantify In Vitro Osteoclast Culture Endpoints.

Quantification of in vitro osteoclast cultures (e.g. cell number) often relies on manual counting methods. These approaches are labour intensive, time consuming and result in substantial inter- and intra-user variability. This study aimed to develop and validate an automated workflow to robustly quantify in vitro osteoclast cultures. Using ilastik, a machine learning-based image analysis software, images of tartrate resistant acid phosphatase-stained mouse osteoclasts cultured on dentine discs were used to train the ilastik-based algorithm. Assessment of algorithm training showed that osteoclast numbers strongly correlated between manual- and automatically quantified values (r = 0.87). Osteoclasts were consistently faithfully segmented by the model when visually compared to the original reflective light images. The ability of this method to detect changes in osteoclast number in response to different treatments was validated using zoledronate, ticagrelor, and co-culture with MCF7 breast cancer cells. Manual and automated counting methods detected a 70% reduction (p < 0.05) in osteoclast number, when cultured with 10 nM zoledronate and a dose-dependent decrease with 1-10 µM ticagrelor (p < 0.05). Co-culture with MCF7 cells increased osteoclast number by ≥ 50% irrespective of quantification method. Overall, an automated image segmentation and analysis workflow, which consistently and sensitively identified in vitro osteoclasts, was developed. Advantages of this workflow are (1) significantly reduction in user variability of endpoint measurements (93%) and analysis time (80%); (2) detection of osteoclasts cultured on different substrates from different species; and (3) easy to use and freely available to use along with tutorial resources.

Evidence that pyrophosphate acts as an extracellular signalling molecule to exert direct functional effects in primary cultures of osteoblasts and osteoclasts.

Extracellular pyrophosphate (PPi) is well known for its fundamental role as a physiochemical mineralisation inhibitor. However, information about its direct actions on bone cells remains limited. This study shows that PPi decreased osteoclast formation and resorptive activity by ≤50 %. These inhibitory actions were associated with reduced expression of genes involved in osteoclastogenesis (Tnfrsf11a, Dcstamp) and bone resorption (Ctsk, Car2, Acp5). In osteoblasts, PPi present for the entire (0-21 days) or latter stages of culture (7-21/14-21 days) decreased bone mineralisation by ≤95 %. However, PPi present for the differentiation phase only (0-7/0-14 days) increased bone formation (≤70 %). Prolonged treatment with PPi resulted in earlier matrix deposition and increased soluble collagen levels (≤2.3-fold). Expression of osteoblast (RUNX2, Bglap) and early osteocyte (E11, Dmp1) genes along with mineralisation inhibitors (Spp1, Mgp) was increased by PPi (≤3-fold). PPi levels are regulated by tissue non-specific alkaline phosphatase (TNAP) and ecto-nucleotide pyrophosphatase/phosphodiesterase 1 (NPP1). PPi reduced NPP1 expression in both cell types whereas TNAP expression (≤2.5-fold) and activity (≤35 %) were increased in osteoblasts. Breakdown of extracellular ATP by NPP1 represents a key source of PPi. ATP release from osteoclasts and osteoblasts was decreased ≤60 % by PPi and by a selective TNAP inhibitor (CAS496014-12-2). Pertussis toxin, which prevents Gαi subunit activation, was used to investigate whether G-protein coupled receptor (GPCR) signalling mediates the effects of PPi. The actions of PPi on bone mineralisation, collagen production, ATP release, gene/protein expression and osteoclast formation were abolished or attenuated by pertussis toxin. Together these findings show that PPi, modulates differentiation, function and gene expression in osteoblasts and osteoclasts. The ability of PPi to alter ATP release and NPP1/TNAP expression and activity indicates that cells can detect PPi levels and respond accordingly. Our data also raise the possibility that some actions of PPi on bone cells could be mediated by a Gαi-linked GPCR.

Modulation of osteoblast differentiation and function by the P2X4 receptor.

Bone cells are known to express multiple P2 receptor subtypes, and the functional effects of receptor activation have been described for many of these. One exception is the P2X4 receptor, which despite strong expression in osteoblasts and osteoclasts, has no defined functional activity. This study used the selective P2X4 receptor antagonists, 5-BDBD and PSB-12062, to investigate the role of this receptor in bone. Both antagonists (≥ 0.1 µM) dose-dependently decreased bone formation by 60-100%. This was accompanied by a ≤ 70% decrease in alkaline phosphatase activity, a ≤ 40% reduction in cell number, and a ≤ 80% increase in the number of adipocytes present in the culture. The analysis of gene expression showed that levels of osteoblast marker genes (e.g. Alpl, Bglap) were decreased in 5-BDBD treated cells. Conversely, expression of the adipogenic transcription factor PPARG was increased 10-fold. In osteoclasts, high doses of both antagonists were associated with a reduction in osteoclast formation and resorptive activity by ≤ 95% and ≤ 90%, respectively. Taken together, these data suggest that the P2X4 receptor plays a role in modulating bone cell function. In particular, it appears to influence osteoblast differentiation favouring the osteogenic lineage over the adipogenic lineage.

N-acetylcysteine (NAC) differentially affects arterial medial calcification and bone formation: The role of l-cysteine and hydrogen sulphide.

Arterial medial calcification (AMC) is the deposition of calcium phosphate in the arteries. AMC is widely thought to share similarities with physiological bone formation; however, emerging evidence suggests several key differences between these processes. N-acetylcysteine (NAC) displays antioxidant properties and can generate hydrogen sulphide (H2 S) and glutathione (GSH) from its deacetylation to l-cysteine. This study found that NAC exerts divergent effects in vitro, increasing osteoblast differentiation and bone formation by up to 5.5-fold but reducing vascular smooth muscle cell (VSMC) calcification and cell death by up to 80%. In vivo, NAC reduced AMC in a site-specific manner by 25% but had no effect on the bone. The actions of l-cysteine and H2 S mimicked those of NAC; however, the effects of H2 S were much less efficacious than NAC and l-cysteine. Pharmacological inhibition of H2 S-generating enzymes did not alter the actions of NAC or l-cysteine; endogenous production of H2 S was also unaffected. In contrast, NAC and l-cysteine increased GSH levels in calcifying VSMCs and osteoblasts by up to 3-fold. This suggests that the beneficial actions of NAC are likely to be mediated via the breakdown of l-cysteine and the subsequent GSH generation. Together, these data show that while the molecular mechanisms driving the actions of NAC appear similar, the downstream effects on cell function differ significantly between osteoblasts and calcifying VSMCs. The ability of NAC to exert these differential actions further supports the notion that there are differences between the development of pathological AMC and physiological bone formation. NAC could represent a therapeutic option for treating AMC without exerting negative effects on bone.

Differing calcification processes in cultured vascular smooth muscle cells and osteoblasts.

Arterial medial calcification (AMC) is the deposition of calcium phosphate mineral, often as hydroxyapatite, in the medial layer of the arteries. AMC shares some similarities to skeletal mineralisation and has been associated with the transdifferentiation of vascular smooth muscle cells (VSMCs) towards an osteoblast-like phenotype. This study used primary mouse VSMCs and calvarial osteoblasts to directly compare the established and widely used in vitro models of AMC and bone formation. Significant differences were identified between osteoblasts and calcifying VSMCs. First, osteoblasts formed large mineralised bone nodules that were associated with widespread deposition of an extracellular collagenous matrix. In contrast, VSMCs formed small discrete regions of calcification that were not associated with collagen deposition and did not resemble bone. Second, calcifying VSMCs displayed a progressive reduction in cell viability over time (≤7-fold), with a 50% increase in apoptosis, whereas osteoblast and control VSMCs viability remained unchanged. Third, osteoblasts expressed high levels of alkaline phosphatase (TNAP) activity and TNAP inhibition reduced bone formation by to 90%. TNAP activity in calcifying VSMCs was ∼100-fold lower than that of bone-forming osteoblasts and cultures treated with ß-glycerophosphate, a TNAP substrate, did not calcify. Furthermore, TNAP inhibition had no effect on VSMC calcification. Although, VSMC calcification was associated with increased mRNA expression of osteoblast-related genes (e.g. Runx2, osterix, osteocalcin, osteopontin), the relative expression of these genes was up to 40-fold lower in calcifying VSMCs versus bone-forming osteoblasts. In summary, calcifying VSMCs in vitro display some limited osteoblast-like characteristics but also differ in several key respects: 1) their inability to form collagen-containing bone; 2) their lack of reliance on TNAP to promote mineral deposition; and, 3) the deleterious effect of calcification on their viability.