Reagents and antibodies

ATP, quinacrine mustard dihydrochloride, 3,4-dihydroxy-9,10-dioxo-2- anthracenesulfonic acid sodium salt (Alizarin red S), 3-[4,5-dimethylthiazol-2-yl]-2,5- diphenyltetrasodium bromide (MTT), phosphate-buffered saline solution (PBS), p-nitrophenyl phosphate (PNP) and cell culture reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA; RRID: SCR_008988). 3-[[5-(2,3-dichlorophenyl)-1 H-tetrazol-1-yl]methyl]pyridine hydrochloride (A438079) and N, N’’-1,4-butanediylbis[N’-(3- isothiocyanatophenyl) thiourea (MRS 2578) were obtained from Tocris Bioscience (Bristol, UK; RRID: SCR_003689). Bovine Collagen I (A10644-01) was supplied by Thermo Fisher Scientific (Grand Island, NY, USA; RRID: SCR_008452). All primary anti-human and secondary conjugated antibodies used in this study have been previously validated [8, 9]. Anti-human P2X7 (anti-rabbit, Cat. No. APR-008, RRID: AB_2040065) and P2Y6 (anti-rabbit, Cat. No. APR-011, RRID: AB_2040082) were purchased from Alomone (Jerusalem, Israel); anti-human Osterix (anti-rabbit, Cat No. ab22552, RRID: AB_2194492), anti-human β-Tubulin (anti-rabbit, Cat. No. ab6046, RRID: AB_2210370) and horseradish-peroxidase-conjugated secondary antibodies (anti-rabbit, Cat. No. ab7083, RRID: AB_955416 and anti-mouse, Cat No. ab6820, RRID: AB_955438) were purchased from Abcam (Cambridge, UK; RRID: SCR_012931); anti-human Osterix (anti-mouse, Cat. No. sc-393,325, RRID: AB_2895257) and anti-human Osteopontin (anti-mouse, Cat. No. sc-21,742, RRID: AB_2194997) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA; RRID: SCR_008987); secondary antibody Alexa Fluor 488-labelled (anti-rabbit, Cat. No. A21206, RRID: AB_2535792) was supplied by Molecular Probes (Invitrogen, Carlsbad, CA, USA; RRID: SCR_013318).

Cell culture conditions and phenotypic characterization of Pm BM-MSCs

Bone marrow samples were obtained from the neck of the femur of thirteen Pm women (70 ± 3 years old) undergoing total hip arthroplasty to resolve non-inflammatory degenerative osteoarthrosis. Handling of bone marrow samples and culture of adherent cells was performed until near confluence for 10–15 days, as previously described [8,9,10, 18]. In brief, bone marrow samples were placed immediately in fresh-frozen α-minimal essential medium (α-MEM) supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, 100 µg/ml streptomycin, and 2.5 µg/ml amphotericin B (standard culture medium) and transported to the laboratory on the day or following day of surgery. Bone marrow cells were dispersed on plastic dishes by repeated gently pippeting, cultured in α-MEM-based standard culture medium and incubated at 37 °C in a humidified atmosphere of 95% air and 5% CO2. Non-adherent cells were removed after 5 days. From this time point onwards, the culture medium of adherent cells was changed twice a week. To avoid the influence of in vitro cells senescence and phenotypic modifications, we used only first subcultures. BM-MSCs were plated at 2.5 × 104 cells/mL density and were allowed to grow for 35 days in α-MEM-based standard culture medium (see the composition above) supplemented with 50 µg/mL ascorbic acid, 10 mM β-glycerophosphate and 10 nM dexamethasone to promote the osteogenic differentiation. The experimental protocols described herein are individualized, i.e. no pooled samples obtained from different individuals have been used in any circumstance.

The phenotypic characterization of the cells (first subculture) was performed previously by flow cytometry [8]. These cells exhibited positive immunoreactivity against CD105 (SH2), CD29 (integrin ß1) and CD117 (tyrosine-protein kinase Kit), which have been identified as surface markers of bone marrow-derived MSCs [19, 20]. Conversely, the cells were negative for haematopoietic surface markers, like CD14 and CD45, which have been extensively used as a good argument to distinguish bone marrow haematopoietic cells from MSCs [20, 21]. Thus, first passage plastic-adherent human bone marrow cells obtained under the present experimental conditions are highly enriched in multipotent MSCs [8].

Mechanical stimulation (MS) of Pm BM-MSCs

Pm BM-MSCs were cultured as described above for 35 days in an osteogenic inducing medium either in the absence or presence of test drugs, namely A438079 (3 µM, P2X7 receptor antagonist) and MRS 2578 (0.1 µM, P2Y6 receptor antagonist), which were added to the culture medium on day 1 (first subculture). MS of the cells consisted of a previously validated shear stress (SS) protocol (see e.g [22, 23] and used with minor modifications; the see-saw system used in previous works and in this study produces low-magnitude fluid-SS in standard culture dishes/plates. After allowing cells adhesion to the bottom of the culture dishes for 4 days, SS cells were mechanically-stimulated twice a week using a microplate shaker placed inside an incubator (980121EU-VWR, 90 r.p.m. for 30 min, at 37ºC). Culture media changes were made twice a week, always taking care to allow a 24 h-recovery time before submitting the cells to SS protocols. Cells proliferation and differentiation were assessed at culture days 7, 14 and 21; bone nodules formation were evaluated at culture day 35. Protein was collected from the cultures at day 21 for Western blot analysis of osterix (OSX) and osteopontin (OPN), as previously described [9, 10]. Intra- and extracellular ATP amounts were evaluated using the quinacrine staining assay and the luciferin-luciferase ATP bioluminescence assay, respectively, at culture days 7 and 21. The density of P2Y6 and P2X7 purinoceptors in the cells was assessed by immunofluorescence confocal microscopy at culture days 7 and 21. The kinetics of the extracellular breakdown of ATP/UDP and metabolites formation was measured by HPLC at days 7 and 21. All these assays were repeated in non-stimulated (CTR) cultures for adequate comparisons (see below).

Viability/proliferation and osteogenic differentiation of Pm BM-MSCs

Cell viability/proliferation was evaluated by the MTT assay [8,9,10, 18]. Data from this assay correlates positively with the results measuring cell proliferation from total DNA quantification per culture well [cf. in [8]].

The osteogenic differentiation of BM-MSCs was inferred as increases in the alkaline phosphatase (ALP) activity and in the expression of osteogenic transcription factors, osterix and osteopontin. The ALP activity was determined in cell lysates by colorimetric determination of p-nitrophenyl phosphate (PNP) hydrolysis, as previously described [8,9,10, 18]; obtained values were expressed in nanomole of PNP per min normalized by the MTT value (nmol min− 1 MTT− 1).

The total amounts of osterix (OSX) and osteopontin (OPN) proteins were determined by Western blot analysis at culture day 21, as previously described [9, 10]. Equal protein amounts (25 µg) were loaded into sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) (12%) gels and transferred onto a polyvinylidene fluoride membrane using a Mini-Protean Tetra Cell coupled to a Mini-Trans-Blot module (Bio-Rad, Hercules, CA, USA; RRID: SCR_008426). Blocked membranes were incubated with anti-human primary antibodies: anti-OSX (1:1000, anti-rabbit, Cat No. ab22552, RRID: AB_2194492 or 1:200, anti-mouse, Cat. No. sc-393,325, RRID: AB_2895257) and anti-OPN (1:400, anti-mouse, Cat. No. sc-21,742, RRID: AB_2194997). Anti-β-tubulin (anti-rabbit, Cat. No. ab6046, RRID: AB_2210370) was used for normalization purposes (i.e., OSX/β-Tubulin and OPN/β-Tubulin). The peroxidase detection system (1.25 mM luminol; 0.2 mM coumaric acid; 0.1 M Tris, pH 8.5; and 0.032% hydrogen peroxide) was used for visualization of the immunoreactivity using the horseradish-peroxidase-conjugated secondary antibodies (1:70000, anti-rabbit, Cat. No. ab7083, RRID: AB_955416 and anti-mouse, Cat No. ab6820, RRID: AB_955438). Gels were analysed using a gel blot imaging system (ChemiDoc MP, RRID: SCR_019037; Bio-Rad, Hercules, CA, USA; RRID: SCR_008426).

Calcium deposition in mineralized nodules was revealed by the Alizarin Red staining and photographed using an optic microscope (Olympus CKX41, RRID: SCR_023725; Tokyo, Japan; RRID: SCR_017564) equipped with a digital camera (Olympus SC30, Tokyo, Japan; RRID: SCR_017564), running an image acquisition software (Olympus Analysis GetIT 5.1, Tokyo, Japan; RRID: SCR_017564), at culture day 35. Calibrated images were exported to Image J 1.37c software (RRID: SCR_003070; NIH, Bethesda, MD, USA) for quantification of the total bone-nodule areas [9, 10].

Immunofluorescence staining and confocal microscopy observation of Pm BM-MSCs

Pm BM-MSCs were allowed to grow in glass chamber slides for 7 or 21 days. Paraformaldehyde fixed cells were incubated in the dark for 2 h with the following primary antibodies: rabbit anti-human P2X7 (1:75, Cat. No. APR-008, RRID: AB_2040065) and rabbit anti-human P2Y6 (1:75, Cat. No. APR-011, RRID: AB_2040082). Alexa Fluor 488 (1:1500, anti-rabbit, Cat. No. A21206, RRID: AB_2535792) was applied as secondary antibody for 1 h in the dark. The VectaShield mounting medium with DAPI was used to mount the glass slides, which were then stored at 4°C until visualization. Observation of the slides was made using a laser-scanning spectral confocal microscope (Olympus FV1000, RRID: SCR_020337; Tokyo, Japan; RRID: SCR_017564) built on an IX81SF-3 inverted motorized microscope with four laser lines controlled by an AOTF laser combiner. Both multi-argon laser and diode laser 405 lines, filtered by barrier filters Ion Coating for OLYMPUS UIS-2 optics, through a UPLSAPO40xOl / NA 0.5–1.0 WD 0.12 mm objective lens (Olympus, Tokyo, Japan; RRID: SCR_017564), were used to acquire images unless otherwise stated. The Fluoview FV1000 Advanced Software (4.0.3.4 version, RRID: SCR_014215; Olympus, Tokyo, Japan; RRID: SCR_017564) was used to analyse data and to control image acquisition parameters, which were set to one-way XY repeat scanning mode at 12.5 s/pixel speed with the pinhole set to 250 μm at an image resolution of 640 × 640 pixel (317.583 × 317.583 μm given that 1 pixel = 0.497 μm). Acquired micrographs were stored in the Olympus Multi TIFF format (Tokyo, Japan; RRID: SCR_017564) [8,9,10, 18]. For comparison purposes, confocal microscope settings and image acquisition parameters were kept unaltered throughout parallel documentation procedures. Negative controls were made in the absence of the primary antibodies or by replacing them by pre-immune sera. Unspecific fluorescence was not detected under these circumstances (data not shown). Five microscopic fields (area of approx. 93,000 µm2 each) were photographed per well. Standardization of XY image coordinates was as follows: the first image was taken at the centre of the well (X = 0; Y = 0) and the next four images were obtained sequentially from each corner of a hypothetical square enclosed in a circumference of 0.275 cm radius. The obtained five independent images were exported to Image J 1.37c software (RRID: SCR_003070; NIH, Bethesda, MD, USA) for quantification analysis. Regions of interest (ROIs) outlining complete individual cells were done manually and the average intensity of the pixels inside each cell was calculated per micrograph. The background fluorescence estimated from outlined ROIs drawn without transecting any cell was subtracted from all monitored ROIs. The computed analysis of the five individual images was expressed as the average fluorescence intensity (arbitrary units, a.u.) for each experimental condition. Shown in the figures are typical immunofluorescence images for each experimental condition, taken from a single representative microscopic field without juxtaposition. When necessary, software adjustments were applied to the entire image [10].

Quinacrine-stained intracellular ATP deposits and extracellular of ATP bioluminescence

We used quinacrine fluorescence staining (ex: 476 nm / em: 500–540 nm) to visualize ATP intracellular stores. To this end, the cells were allowed to grow for 7 or 21 days in an osteogenic-inducible medium. After removing the incubation medium, the cells were washed three times with phosphate-buffered saline (PBS, 1x) and subsequently incubated for 1 h with quinacrine (30 µM), at 37ºC [24]. Images were acquired using an epifluorescence microscope equipped with a XBO 75 W Xenon arc lamp (Achroplan; Zeiss, Oberkochen, Germany; RRID: SCR_023607). The light path included ET460/30 x excitation / ET520/40 m emission filters (Chroma Technology Corp, Bellows Falls, VT, USA) and a LUMPLFLN40XW/0.80NA/3.3WD water dipping objective lens (Olympus, Tokyo, Japan; RRID: SCR_017564). A high-resolution cooled CCD camera (CoolSnap HQ, Roper Scientific Photometrics, Tucson, AZ, USA) connected to a computer running a digital image acquisition software (MetaFluor 6.3, RRID: SCR_014294; Molecular Devices Inc., Sunnyvale, CA, USA) was used to record images in the TIFF format. The CCD exposure time was set to 100 ms, binning was adjusted to 2 and gain to 1.

Extracellular ATP levels were quantified using the luciferin-luciferase ATP bioluminescence assay kit HS II (Roche Applied Science, Indianapolis, Indiana, USA) in a multi-detection microplate reader (Synergy HT, RRID: SCR_020536; BioTek Instruments, Vermont, USA), as described elsewhere [24]. Briefly, the cells were seeded onto 96-well microplates, at a density of 2.5 × 104 cells/mL, for 7 or 21 days (4–8 replicas were performed per individual experiment). At the beginning of the experiment, the cells were washed twice with a Tyrode’s solution. The cells where then incubated with a fresh Tyrode’s solution for 30 min, at 37ºC. Then, the incubation fluid was removed and aliquots were snap-frozen in liquid nitrogen. Before adding the luciferin-luciferase mixture, the collected samples were defrosted until 25ºC according to the manufacturer’s instructions. Sample bioluminescence was compared to external high-purity ATP standards; these were made daily within the same concentration range; all samples were analysed in duplicates. The remaining incubation medium was used to quantify the lactate dehydrogenase (LDH, EC 1.1.1.27) activity [25] to evaluate cell integrity during the experimental period (see [24]). The LDH activity was negligible (between 0.071 and 0.12 mU/mL) in all measured samples indicating the integrity of the cells during the experimental period.

Enzymatic kinetic experiments and HPLC analysis of extracellular nucleotides and its metabolites

The kinetics of the extracellular catabolism of ATP and UDP in Pm BM-MSCs cultures submitted or not to mechanical stimulation was evaluated on days 7 and 21 [8, 18]. After a 30-min equilibration period, the cells were incubated, at 37ºC, with gassed (95% O2 plus 5% CO2) Tyrode’s solution (137 mM NaCl, 2.7 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 0.4 mM NaH2PO4, 11.9 mM NaHCO3, and 11.2 mM glucose, pH 7.4) supplemented with 100 µM ATP or UDP (zero time). Samples (75 µl) were collected from each well at different times up to 30 min for HPLC analysis (LaChrom Elite, Merck, Frankfurt, Germany) of the variation of substrate disappearance and products formation; 20-µl injection volumes were used for the analysis. The concentrations of the substrates and their respective metabolites were plotted as a function of time (progress curves). The following parameters were analysed for each progress curve: half-life time (t1/2) of the initial substrate, time of appearance of the different concentrations of the products, the concentration of the substrate or any product remaining at the end of the experiment. The spontaneous degradation of ATP and UDP, at 37ºC, was negligible over a period of 30-min in the absence of the cells. At the end of the experiments, the remaining incubation medium was collected and used to measure the lactate dehydrogenase (LDH, EC 1.1.1.27) activity [25]. The negligible activity of LDH in the samples collected at the end of the experiments is an indication of the integrity of the cells during the experimental procedure.

Collagen-I encapsulation of cultured Pm BM-MSCs for xenotransplantation

For xenotransplantation, Pm BM-MSCs were cultured as described above and submitted (or not) to mechanical stimulation either in the absence or presence of selective P2X7 and P2Y6 receptor antagonists (A438079 3 µM and MRS 2578 0.1 µM, respectively). After allowing Pm BM-MSCs to grow in culture for day 21, they were detached and encapsulated into collagen I matrix (A10644-01, Thermo Fischer Scientific, NY, USA; RRID: SCR_008452) at a density of approximately 1 × 106 cells/mL. For encapsulation, the cells were pelleted and re-suspended into sterile 15-mL Falcon tubes containing 30-µL collagen I scaffolds, which were kept for 30–40 min at 37 °C in a humidified atmosphere under 95% air and 5% CO2. After adding 200-µL of the osteogenic culture medium to each tube, the encapsulated cells were maintained in culture for 24 h before xenotransplantation [26].

Critical bone defect repair using an “in vivo” animal model

Twenty female Wistar rats (Rattus norvegicus; Cat. No. 13,508,588, RRID: RGD_13508588; Charles River, Barcelona, Spain; RRID: SCR_003792) of about one-year old weighting 230–360 g were housed in groups of three to four animals inside ventilated Double Decker (38 cm high) cages with enriched environment and access to food and water ad libitum. The room temperature was kept constant (21 °C) and a regular light (07:30–19:30 h)–dark (19:30–07:30 h) cycle was imposed. The animals were acclimatized to these conditions for at least 10 days before their assignment to the experimental groups. Critical bone defects were made under general anaesthesia using 75 mg/kg ketamine (Imalgene 100 mg/mL, Boehringer Ingelheim, Germany; RRID: SCR_004791) plus 0.5 mg/kg medetomidine (Domtor 1 mg/mL, Ecuphar, Portugal) intraperitoneally. After achieving deep anaesthesia (absence of reflexes), the hind limbs were shaved and the skin disinfected. The rats were then positioned in lateral decubitus and a small incision was made in the skin below the greater trochanter of femora to expose the insertion of the rectus femoris and vastus lateralis muscles. The lateral surface of the femoral diaphysis was exposed by blunt debridement to create a critical bone defect (hole of 2.1 mm diameter) in the greater trochanter oriented towards the lesser trochanter using an electric drill (OmniDrill35, WPI, UK; RRID: SCR_008593) (see a schematic representation in Fig. 7). Drilling was made at slow speed with continuous irrigation using physiological saline to avoid heating and damage of the cells surrounding bone defects. After removing bone debris, the defect was loaded with encapsulated Pm BM-MSCs prepared as described above. All the cells used in the present study underwent osteogenic differentiation for 21 days, either in the absence (CTR group) or in presence of mechanical stimulation (SS group). The mechanically stimulated Pm BM-MSCs were subdivided into 3 groups comprising cells cultured in the absence (SS group) and in the presence of selective P2X7 (SS + A438079, 3 µM) and P2Y6 (SS + MRS 2578, 0.1 µM) receptor antagonists; at least 4 animals were used per experimental group and xenotransplantation of Pm BM-MSCs groups was made blindly by only one operator. Contralateral critical bone defects either empty or filled with collagen I (no added cells) scaffolds were used as controls. Bone defects were then covered with bone wax and the wound was closed by suturing soft tissue plans. After surgery, reversal of anaesthesia was achieved using 1 mg/kg IM atipamezole hydrochloride (Antisedan 5 mg/mL, Ecuphar, Portugal). Immediately after surgery and in the next following days analgesia was warranted with tramadol 10 mg/kg IM (tramadol 100 mg/2mL, Labesfal, Portugal). The follow-up recovery was monitored for 10 days after which the rats were sacrificed by decapitation followed by exsanguination and femora removed for histological analysis [26, 27].

Fig. 1

Transplanted BM-MSCs submitted to mechanical stimulation (SS) accelerate closure of critical bone defects in an “in vivo” animal model compared to non-stimulated cells and this effect depends on tonic P2X7 and P2Y6 receptors activation. Panel A shows a schematic representation of a rat femur with a critical bone defect drilled in the greater trochanter prepared to receive the encapsulated Pm BM-MSCs (for details see Materials and Methods). Panel B shows representative micrograph sections of critical bone defects in rat femora filled with collagen I-encapsulated BM-MSCs of a 60-years old woman. Before xenotransplantation, BM-MSCs were differentiated using an osteoblast-inducing medium for 21 days in the absence (CTR) or presence of shear stress (SS) mechanical stimulation only or together with of selective P2X7 (SS + A438079, 3 µM) and P2Y6 (SS + MRS 2578, 0.1 µM) receptor antagonists. Shown are images stained with hematoxylin and eosin (HE) and Masson’s trichrome highlight the bone defect area (dashed yellow line). The mesenchymal tissue (MT) consists of BM-MSCs transplanted and/or recruited to the injury site; endochondral ossification (EO) is evidenced as the result of secondary bone repair. Panel C shows three graphs computed from histological digital images, representing: (i) the percentage of MSCs area, endochondral ossification area and woven bone area as a function of the bone defect filled area (100%); (ii) the ratio between the area occupied by BM-MSCs and the total bone defect area; and (iii) the ratio between the endochondral ossification area and the total bone defect area, obtained for each experimental condition. Boxes and whiskers represent pooled data from 4 individual experiments; the cells used were obtained from 4 Pm women (60, 68, 76 and 79 years old); *P < 0.05 (ordinary one-way ANOVA with Fisher’s LSD test) represents significant differences. The insert part in panel Ci shows a higher magnification (410x) of a micrograph section stained with Masson’s trichrome in which the bone defect area shows spicules of newly formed woven bone (WB) through endochondral ossification (EO)

Histological staining and analysis

After sacrifice of the animals, both femora were removed and fixed in 10% buffered formalin for at least 72 h for histological analysis, as described previously [26, 27]. The bones were then decalcified in fresh Shandon TBD-1™ Decalcifier (Thermo Fischer Scientific, NY, USA; RRID: SCR_008452) for 48 h and, then, sectioned through the middle line after gross examination. The tissue samples were routinely processed for histological analysis. Paraffin-embedded serial sections of 4-µm thick were stained with hematoxylin and eosin (HE; general overview) and Masson’s trichrome. The latter staining was used to highlight collagen fibres allowing the distinction of the four overlapping stages of secondary bone repair [28], namely the initial inflammatory response, the soft callus formation, the hard callus formation, and the bone-remodelling phase. The NanoZoomer 2.0HT (Hamamatsu Photonics K.K., Japan; RRID: SCR_017105) was used to visualize and scan the histological glass slides, which were then converted into high-quality/high-resolution digital images using the SlideViewer 2.7 software (RRID: SCR_017654; 3DHISTECH, Budapest, Hungary). This software was also used to manually outline and quantify the following histological parameters: the total area of the bone defect, the area filled by Pm BM-MSCs (initial stage of bone regeneration characterized by MSCs recruitment), the endochondral ossification area (second stage of bone regeneration through soft callus formation), and the woven bone area (third stage of bone regeneration by hard callus formation). Two independent observers made histological analysis blindly.

Presentation of data and statistical analysis

Data are expressed as scatter dot plot (with mean ± S.E.M.) or as Box and Whiskers (Min to Max) from an n number of individuals. No predetermined sample size calculation was performed. Due to restricted access to similar human samples and a limited pool of initial cell density, we were unable to perform all the indicated assays in all collected human samples. Normality tests included D’Agostino & Pearson and Shapiro-Wilk, depending on sample size. According to normality test results, statistical analyses included parametric (two-way ANOVA with Tukey’s test for multiple comparisons and ordinary one-way ANOVA with Fisher’s LSD test) or non-parametric tests (two-tailed Mann Whitney, Kruskal-Wallis with Dunn’s multiple comparison or uncorrected Dunn’s test and Wilcoxon matched-pairs signed rank test), with a confidence level of 0.05 (95% confidence interval). Values of P < 0.05 were considered to represent significant differences. Data analysis was performed using the Prism 10.0.2 TM software (RRID: SCR_002798; GraphPad Software, CA, United States).