Impairment of PTX3 expression in osteoblasts: a key element for osteoporosis

Pentraxin 3 (PTX3) is a multifunctional glycoprotein regulating inflammatory response, cell proliferation and migration and deposition and remodelling of the extracellular matrix by a variety of cells. In this study, we investigated the possible role of PTX3 in bone homeostasis. To this end, we compared the expression and function of PTX3 in human osteoblasts of osteoporotic, osteoarthritic patients and young subjects not affected by bone diseases. Immunohistochemical analysis performed on bone head biopsies showed a close association between bone health and the number of osteoblasts expressing PTX3. Noteworthy, the proportion of PTX3-positive osteoblasts resulted to be significantly lower in osteoporotic patients compared with both young patients and osteoarthritic patients of the same age. Ex vivo culture of osteoblasts isolated from the three groups of patients confirmed in vivo observation. Specifically, we observed rare runt-related transcription factor 2 (RUNX2) immunopositive osteoblasts expressing PTX3 in cell cultures derived from osteoporotic patients and western blotting analysis showed 80% reduction of PTX3 in the corresponding culture extracts compared with young and osteoarthritic patients. The treatment of human osteoblast primary cultures derived from young patients with anti-PTX3 antibody dramatically affected osteoblast behaviour. Indeed, they lost the morphological and molecular features typical of mature osteoblasts, acquiring fibroblast-like shape and highly decreasing nuclear factor kappa-B ligand (RANKL) and RUNX2 expression. Also, the inhibition of PTX3 negatively affected osteoblast proliferation and their ability to form cell clusters and microhydroxyapatite crystals. Altogether, these results suggest a central role of PTX3 in bone homeostasis showing its involvement in osteoblast proliferation, differentiation and function.

and pathological conditions. PTX3 has been found to bind and sequester fibroblast growth factor 2 (FGF2) via its N-terminal extension and to suppress in vitro FGF-dependent proliferation of endothelial and smooth muscle cells and in vivo tissue neovascularization. 14,15 In addition, several lines of evidence have also established a prominent role of PTX3 in extracellular matrix composition and organization. It was recently demonstrated that PTX3 regulates the injury-induced thrombotic response 16 and promotes wound healing by favouring timely fibrinolysis. 12 PTX3 expression is also induced by hormones and local factors in the ovary where it has an essential role for assembling hyaluronan 17 in a matrix suitable for oocyte fertilization. 18,19 Few and conflicting data are available to date concerning a possible role of PTX3 in bone metabolism. It has been reported that PTX3 induces the expression of receptor activator of nuclear factor kappa-B ligand (RANKL) by human osteoblasts thereby promoting osteoclastogenesis in an in vitro culture system. 20 On the other hand, preliminary data by Kelava et al. 21 investigated the relationship between PTX3 expression and bone histomorphometry parameters in mice deficient for the PTX3 gene demonstrating that PTX3 null mice had lower bone mass than their WT littermates, implying PTX3 involvement in bone formation.
Based on these evidence, we investigated the possible role of PTX3 in the alteration of bone remodelling that occurs in OP patients. To this end, the expression and function of PTX3 in human osteoblasts of OP patients were compared with those from patients affected by osteoarthritis (OA), that is, of comparable old age but experiencing bone alterations not necessarily linked to bone density loss, and with those from young subjects not affected by bone diseases (CTRL). Analyses were made in both in vivo and in vitro systems.

Results
Clinical evaluation. The OP group included 25 patients with fragility hip fracture, T-score ≤ − 2.5 S.D. and Kellgren − Lawrence (K-L) score from 0 to 1. The OA group included 25 patients with radiographic evidence of hip OA with a K-L score 3 or 4 and T-score ≥ − 2.5 S.D. (Table 1). CTRL patients were characterized by a T-score ≥ − 1.0 S.D. and K-L score from 0 to 1. In addition, OA and OP patients displayed homogenous values of hematochemical exams of both bone and kidney metabolism (Table 1).
Bone morphometric analysis allowed us to estimate trabecular, fat and bone marrow areas (Figures 1g-i). OP patients showed a lower percentage of trabecular tissue (29.41%) compared with both the OA and CTRL groups (OA 72.15%; CTRL 87.23%) (OP versus OA P = 0.0021; OP versus CTRL Po0.0001).
In particular, in OP patients the trabeculae were predominantly replaced by adipose tissue (OP 70.48% versus CTRL 2.62%), and the bone marrow tissue was significantly reduced (OP 0.11% versus 10.13 % CTRL; Figures 1g and i). Conversely, in OA patients, adipose tissue increased much less (23.40 %) and the residual bone marrow area was adequately preserved (5.45 %) (Figure 1h).

Immunohistochemical analysis of osteoblast markers.
We analysed the osteoblast differentiation rate in OP, OA and CTRL patients by immunodetection of specific markers in cells lining the endostium. Immunohistochemical positivity was evaluated on digital images (Iscan Coreo, Ventana, Tucson, AZ, USA) by a semiquantitative approach. We assigned a score from 0 to 3 according to the number of positive osteoblasts on the total analysed for runt-related  Figure 2). Similar results were obtained regarding the expression of VDR (CTRL 1.48 ± 0.14; OP 0.88 ± 0.15; OA 1.00 ± 0.11; OP versus OA P = 0.4974; OP versus CTRL Po0.0001; OA versus CTRL Po0.0001; Figure 2). Immunohistochemistry for OPG showed a significant difference among the groups (Po0.0001). In particular, we observed a relative lower number of osteoblasts expressing OPG in OP patients compared with OA and CTRL (OP 0.87 ± 0.12; OA 1.32 ± 0.15; CTRL 2.28 ± 0.14; OP versus OA P = 0.0355; OP versus CTRL Po0.0001; OA versus CTRL Po0.0001; Figure 3). Also for osteocyte sclerostin expression, we found a significant group difference (Po0.0001). As expected, there was a significantly higher relative amount of sclerostin-positive cells in OP patients compared with OA and CTRL patients (OP 2.24 ± 0.14; OA 1.36 ± 0.13; CTRL 0.96 ± 0.09; OP versus OA Po0.0001; OP versus CTRL Po0.0001; OA versus CTRL P = 0.0242; Figure 3). These data are in agreement with a decreased osteoblast differentiation in OP patients.
In vivo PTX3 expression and bone metabolism. Immunohistochemical analysis of biopsies revealed that PTX3 was expressed in human osteoblasts of CTRL. Noteworthy, the relative number of PTX3-positive osteoblasts was significantly lower in OP compared with OA and CTRL patients   Figure 4). In addition, we plotted PTX3 score and BV/TV or Tb.Th values for each OP patient. We numbered OP patients in increasing order of BV/TV or Tb.Th values, respectively. Interestingly, we found that the few OP patients with relative elevated number of PTX3-expressing osteoblasts, that is, score of 2 or 3, showed higher values of BV/TV and Tb.Th (Figures 4b and c). Altogether, these results suggest a correlation between bone density and PTX3 expression.
In vitro study of PTX3 expression. We then isolated osteoblasts from the cancellous bone of OP, OA and CTRL patients and cultured them in vitro. In all cases, after 4 weeks of culture, cells became confluent and 495% were immunopositive for RUNX2 ( Figure 4j). Osteoblasts isolated from CTRL and OA patients were characterized by the presence of numerous micro-HA crystals, whereas primary osteoblast cultures derived from OP showed a poor propensity to form micro-HA crystals (Figure 4k). Dual-colour immunofluorescence reaction showed that PTX3 was expressed by 475% of RUNX2-positive osteoblasts in primary osteoblast cultures derived from both OA and CTRL patients (Figures 4j, l and m). Conversely, we observed rare RUNX2-positive osteoblasts expressing PTX3 in cell cultures derived from OP ( Figure 4j). This result was confirmed by western blotting analysis with PTX3 antibody. Protein extracts of OP, OA and CTRL osteoblast cultures showed a positive band at about 42 kDa corresponding to the molecular weight of monomeric PTX3. Noteworthy, the signal was remarkably less intense in OP than in OA and CTRL samples, corresponding to almost 80% reduction after normalization for housekeeping protein actin (Figure 4n). At the genomic level, we also demonstrated the decrease of PTX3 mRNA in osteoblast derived from OP patients with respect to both osteoblasts of OA and CTRL by real-time PCR (Figure 4o).
These in vitro results further support the evidence of impaired PTX3 expression in osteoblasts of OP patients.
In vitro study of PTX3 function. To establish whether PTX3 affects osteoblast behaviour, osteoblasts derived from CTRL patients cultured for 4 weeks were seeded at a density of 30 × 10 3 cells/well. At the beginning of culture (T0), 0.1 or 1 μg/ml PTX3 antibody or rabbit IgG at the same concentrations were added to the medium, and cells were cultured for additional 72 h. Cell proliferation, morphology, micro-HA crystals and osteoblast characteristics were then evaluated. The results showed that treatment with anti-PTX3 antibody significantly inhibited cellular proliferation in a dosedependent manner (Figure 5a), without affecting cell viability, as assessed by nuclear feature (Figure 5g) and trypan blue Conversely, osteoblasts treated with anti-PTX3 antibody lost the ability to aggregate in clusters (Figures 5f and g) and showed rare or absent micro-HA crystal deposition ( Figure 5h). In addition, to further elucidate the effects of PTX3 on osteoblasts behaviour, we treated osteoblast primary cultures derived from OP patients with recombinant human PTX3 (20 ng/ml) for 72 h. As shown in Figure 5, the effect of exogenous PTX3 induce a significant increase of both cell proliferation (18.37%) ( Figure 5i) and formation of HA microcrystals (Figure 5j) with respect to cultures treated with vehicle. Noteworthy, already after 72 h it is possible to observe the formation of calcified nodules in osteoblast primary cultures derived from OP patients treated with recombinant human PTX3.

Discussion
PTX3 is induced by a variety of cytokines in immune cells and has a non-redundant role in the regulation of inflammation. 22 It acts as an extrinsic onco-suppressor gene in mouse and human by regulating complement-dependent, macrophagesustained, tumor-promoting inflammation. 23 Moreover, PTX3 is also produced by several cell types under appropriate stimuli and, likely owing to its complex quaternary structure, it is able to interact with several molecules, including growth factors, extracellular matrix components and fibrinolytic enzymes, thereby regulating cell proliferation and matrix remodelling in a variety of tissues. 22 It has been recently shown that human mesenchymal cells derived by bone marrow express PTX3 when induced to differentiate in vitro into osteogenic lineage. 20 It has been proposed that PTX3 elevation during bone inflammatory conditions promotes RANKL production and favour osteoclastogenic potential by osteoblasts, implying its involvement in bone resorption. 20 We now explored the role of PTX3 in bone homeostasis by studying PTX3 expression in osteoblasts from healthy and OP patients who experience bone mass loss without apparent The first aim of this study was to determine whether the expression of PTX3 in human osteoblasts from OP patients differed from those from patients affected by OA, that is, of comparable age (74-76 years) but experiencing bone alterations not linked to bone density loss (surrogate control), and with those from young (from 18 to 46 years) subjects not affected by bone diseases (CTRL).
As expected, the mean value of bone morphometric parameters showed that the bone from OP patients had a significant reduction of bone mass (BV/TV and Tb.Th). Clinical analysis excluded the occurrence of osteoporosis in OA and CTRL patients enrolled. Then immunohistochemical analysis performed on bone head biopsies of young healthy patients displayed that, besides hematopoietic bone marrow cells, 22 a high percentage of osteoblasts in the endostium expresses PTX3. This in vivo observation confirms and extends previous studies showing that PTX3 expressed by osteoblasts differentiated in vitro from bone marrow mesenchymal cells. 20 Interestingly, the comparison of immunohistochemical results obtained by OP, OA and CTRL showed a close association between bone health and PTX3. The number of PTX3expressing osteoblasts was positively correlated with BV/TV and Tb.Th values reaching the maximum in the CTRL group. Noteworthy, the proportion of PTX3-positive osteoblasts resulted to be significantly lower in OP patients compared with both young patients and old OA patients. In addition, the few OP patients with relative elevated number of PTX3expressing osteoblasts also showed higher values of BV/TV and Tb.Th. Also, preliminary data indicate a positive putative association between PTX3 serum levels and bone quality in CD1 mice (data not shown); a significant reduction of BV/TV and Tb.Th were observed in concomitance with a 15% decrease of PTX3 serum levels. These results suggested an active involvement of PTX3 in bone formation.
Ex vivo culture of osteoblasts isolated from the three groups of patients confirmed in vivo observation. We found that 95% of the cells were immunopositive for the osteoblast lineage master gene RUNX2 regardless of their patient origin, but OP primary osteoblast cultures showed deep difference with the other two groups in PTX3 co-expression. Of all the RUNX2positive cells, 475% was also positive for PTX3 in CTRL and OA primary osteoblast cultures while the number dropped to 5-10% in OP osteoblast cultures. Accordingly, western blotting analysis showed 80% reduction of PTX3 in the cell culture extracts of OP compared with CTRL and OA samples. The evidence that the deficiency of PTX3 production is maintained by OP-derived osteoblasts when cultured in vitro suggests that PTX3 gene is not negatively affected by environmental factors but rather that an appropriate positive signal is missing or that the PTX3 promoter is stably silenced by epigenetic mechanisms, as recently demonstrated in human cancers. 24 That PTX3 is positively involved in bone metabolism is further supported by the evidence that cell cultures derived from OP patients displayed a poor propensity to produce mineralized matrix as demonstrated by rare presence of micro-HA crystals. Moreover, the treatment of human osteoblast primary cultures derived from CTRL patients with anti-PTX3 antibody induced considerable changes in osteoblast behaviour. Specifically, they lost the morphological and molecular features typical of mature osteoblasts, acquiring fibroblast-like shape and drastically decreasing RANKL and RUNX2 expression. Notably, the inhibition of PTX3 negatively affected osteoblast's proliferation and their ability to form both cellular clusters and micro-HA crystals. In addition, we reported the effects of exogenous PTX3 on osteoblast primary cultures derived from OP patients. Our data clearly demonstrated the ability to PTX3 to induce an increase of both cell proliferation and HA In this context, FGF2 could be a key mediator of the relationship between bone metabolism and PTX3. It is known that PTX3 contains two FGF2-binding sites and sequesters this growth factor thereby inhibiting its action on target cells. 25 This interaction can modulate the capability of FGF2 to interfere with osteoblast activity. Indeed, although FGF2 is generally considered to favour bone deposition, it exerts differentiationstage-specific effects on osteoblasts. 26 Thus, the ability of PTX3 to sequester FGF2 via its N-terminal extension can influence the osteoblastogenesis by regulating the activity of FGF2.
The correlation between osteoblast PTX3 deficiency and low bone density reported here in humans well matches with bone structure deficiency and impaired fracture healing observed in the animal model of PTX3 null mice. 21 The requirement of fine control of PTX3 expression for maintaining the bone in good health is also strengthened by the studies performed in conditions mimicking in vivo and in vitro acute inflammation. 20 In such inflamed situations, cytokines induced PTX3 overexpression by osteoblasts, which in turn increased their osteoclastogenic potential by elevating RANKL production. Altogether, these results allow to hypothesize an opposite effect of PTX3 in the delicate balance of bone remodelling depending on health conditions. In physiological conditions, the expression of PTX3 would stimulate osteoblast differentiation by bone marrow mesenchymal stem cells. On the other hand, a decreased PTX3 production, as that observed in OP patients, could result in inadequate bone formation and an excessive PTX3 elevation in inflammatory conditions, such as in rheumatoid arthritis, could promote bone resorption, in both cases leading to bone mass loss.
Limits of the study. The evaluation of PTX3 was performed using a semiquantitative approach (immunohistochemistry), as protein or mRNA extraction is difficult to perform on bone biopsies. In addition, it is not always possible to obtain suitable material from clinical sources. To corroborate our results, western blotting analysis was performed on primary osteoblast cultures derived from OP, OA and CTRL patients. In our laboratory, blood serum concentration of creatinine, nitrogen (BUN), phosphorus, calcium, Vit D (25OHD3) and intact parathyroid hormone (PTH) were not assessed in patients undergoing hip arthroplasty for high-energy hip fractures. Nevertheless, the bone quality of all these patients was evaluated by histomorphometric analysis (Figure 2). Unfortunately, it is very difficult to collect bone head biopsies of patients aged 470 years without OP or OA who underwent hip arthroplasty for high-energy hip fractures. However, the results demonstrated that OA patients were good surrogate controls in this study, showing several similarities with younger patients.

Conclusions
The identification of new determinants of bone loss in osteoporosis is a field in constant development. This study suggests an important role of PTX3 in normal bone homeostasis showing its involvement in osteoblast proliferation, differentiation and function. Its impaired expression by osteoblast cells in OP patients strongly support the hypothesis, that PTX3 is a novel regulator of bone metabolism with prominent effects on cellular processes that are essential for normal bone physiology. Further studies are needed to elucidate the molecular mechanisms through which PTX3 regulates the activities of osteoblasts.

Materials and Methods
All experiments described in the present study were approved by the ethics committee of 'Policlinico Tor Vergata' (approval reference number 85/12). All experimental procedures were carried out according to The Code of Ethics of the World Medical Association (Declaration of Helsinki). Informed consent was obtained from all patients prior to surgery. Specimens were handled and carried out in accordance with the approved guidelines.  Table 1).
Exclusion criteria were history of cancer, myopathies or other neuromuscular diseases or chronic administration of corticosteroid for autoimmune diseases (41 month), diabetes, alcohol abuse and HBV, HCV or HIV infections.
Bone mineral density evaluation (DXA). DXA was performed with a Lunar DXA apparatus (GE Healthcare, Madison, WI, USA). Lumbar spine (L1-L4) and femoral (neck and total) scans were performed, and bone mineral density (BMD) was measured according to the manufacturer's recommendations. 27 Dualenergy X-ray absorptiometry measures BMD (in grams per square centimetre), with a coefficient of variation of 0.7%. For patients with fragility fractures, BMD was measured on the uninjured limb. For all the other patients, measurements were performed on the non-dominant side, with the participants supine on an examination table with their limbs slightly abducted. 28 DXA exam was performed 1 day before surgery for OA patients and 1 month after surgery for OP and control patients (CTRL). The results were expressed as T-scores.
Radiographic analysis. Anteroposterior radiographs of the pelvis of all groups were obtained using a standard, validated protocol. 29 Two orthopaedists independently assessed all radiographs using the K-L radiographic atlas. 30 Patients with a grade of K-L ≥ 2 were considered osteoarthritic.
Sampling. At surgery, the femoral head was removed to implant prosthesis. Bone samples were taken for histological analysis, excluding areas with macroscopic alterations of trabecular bone such as necrotic areas.
Histology. Bone biopsies of the femoral head were fixed in 4% paraformaldehyde for 24 h and paraffin embedded without decalcification. 31 From each patient, we obtained two paraffin blocks. Undecalcified tissues were cut by a tungsten carbide knife; 3 μm-thick sections were stained using H&E.
Histomorphometric analysis. Ten microscopic images, randomly selected, were evaluated for each biopsy sample. Images were acquired at × 40 magnification using a Nikon Eclipse E600 light microscope connected to a Nikon digital camera (Nikon Corp, Japan) and saved at a resolution of 1280 × 1024 pixels. Image analysis was performed using a BioQuant Osteo software (version7.20.10; BIOQUANT Image Analysis Corporation, Nashville, TN, USA) according to the manufacturer's instructions. 32 The following parameters: BV/TV, Tb.Th, and Tb.S, were evaluated according to Dempster et al. 33 Moreover, H&E slides were analysed using the Viewing software (Ventana, Tucson, AZ, USA) to evaluate bone tissue composition (trabecular, bone marrow and fat).
Human osteoblast primary cell cultures: Primary cultures of osteoblasts were obtained from the cancellous bone of: patients with high-energy femoral fracture (CTRL, Caucasian, 18 years), patients affected by osteoporosis (OP, Caucasian, 71 years), and patients affected by osteoarthrosis (AO, Caucasian, 72 years). The samples were dissected and treated to obtain a homogeneous population of osteoblasts. Briefly, after dissection, trabecular bone fragments were repeatedly washed in PBS. Then bone fragments were briefly incubated at 37°C with 1 mg/ml Trypsin from porcine pancreas ≥ 60 /mg (SERVA Electrophoresis GmbH, Heidelberg, Germany) diluted in DPBS. After washing, bone fragments were subjected to repeated digestions with 2.5 mg/ml Collagenase NB 4G Proved grade ≥ 0.18 U/mg (SERVA Electrophoresis GmbH) diluted in DPBS with calcium and magnesium. Supernatant were collected and centrifuged at 310 RCF for 5 min. Cell pellets were resuspended in DMEM with 15% FBS, seeded into a 24-well plate and incubated at 37°C, 5% CO 2 until reaching confluence (about 4 weeks). Medium was changed twice a week. Osteoblasts were characterized by alkaline phosphatase test and immunostained for RUNX2 and RANKL.
Immunostaining of primary cell cultures: Expression of RUNX2 and PTX3 was simultaneously evaluated by dual-colour immunofluorescence in confluent CTRL, OP and OA primary osteoblast cultures. Briefly, after fixation in PFA 4% for 30 min, cell cultures were pretreated with EDTA citrate pH 7.8 for 5 min at 95°C and incubated with mouse monoclonal anti-RUNX2 antibody for 30 min (1 μg/ml, clone EPR14334, AbCam). Reaction with anti-Runx2 was revealed by using FITCconjugated anti-mouse antibody. Afterwards, cell cultures were incubated with rat monoclonal anti-PTX3 (2 μg/ml, clone MNB1, AbCam) for 30 min. Reaction with anti-PTX3 was revealed by using Texas Red-conjugated anti-rat antibody. Washing was performed with PBS/Tween 20 pH 7.6 (UCS Diagnostic).