Endocrinology and Metabolism

P2X7-mediated calcium influx triggers a sustained, PI3K-dependent increase in metabolic acid production by osteoblast-like cells

Matthew W. Grol, Irene Zelner, S. Jeffrey Dixon

Abstract

The P2X7 receptor is an ATP-gated cation channel expressed by a number of cell types, including osteoblasts. Genetically modified mice with loss of P2X7 function exhibit altered bone formation. Moreover, activation of P2X7 in vitro stimulates osteoblast differentiation and matrix mineralization, although the underlying mechanisms remain unclear. Because osteogenesis is associated with enhanced cellular metabolism, our goal was to characterize the effects of nucleotides on metabolic acid production (proton efflux) by osteoblasts. The P2X7 agonist 2′,3′-O-(4-benzoylbenzoyl)ATP (BzATP; 300 μM) induced dynamic membrane blebbing in MC3T3-E1 osteoblast-like cells (consistent with activation of P2X7 receptors) but did not induce cell death. Using a Cytosensor microphysiometer, we found that 9-min exposure to BzATP (300 μM) caused a dramatic increase in proton efflux from MC3T3-E1 cells (∼2-fold), which was sustained for at least 1 h. In contrast, ATP or UTP (100 μM), which activate P2 receptors other than P2X7, failed to elicit a sustained increase in proton efflux. Specific P2X7 receptor antagonists A 438079 and A 740003 inhibited the sustained phase of the BzATP-induced response. Extracellular Ca2+ was required during P2X7 receptor stimulation for initiation of sustained proton efflux, and removal of extracellular glucose within the sustained phase abolished the elevation elicited by BzATP. In addition, inhibition of phosphatidylinositol 3-kinase blocked the maintenance but not initiation of the sustained phase. Taken together, we conclude that brief activation of P2X7 receptors on osteoblast-like cells triggers a dramatic, Ca2+-dependent stimulation of metabolic acid production. This increase in proton efflux is sustained and dependent on glucose and phosphatidylinositol 3-kinase activity.

  • adenosine 5′-triphosphate
  • microphysiometer
  • proton efflux
  • phosphatidylinositol 3-kinase
  • metabolism

bone remodeling is a dynamic process that relies on a delicate balance between formation and resorption by osteoblasts and osteoclasts, respectively. A combination of local and systemic factors, including parathyroid hormone (PTH), insulin-like growth factors (IGFs), estrogen, and bone morphogenetic proteins (BMPs), act in concert with mechanical stimuli to modulate skeletal homeostasis (27). Mechanical load enhances bone formation, resulting in improved skeletal strength (58). Osteoblasts release ATP and other nucleotides in response to mechanical stimuli (20, 36), and it has been suggested that these molecules mediate mechanotransduction in bone (12).

Extracellular nucleotides signal through P2 receptors expressed in a variety of cell types, including osteoblasts and osteoclasts. These receptors are divided into two families, the P2Y family of G protein-coupled receptors and the P2X family of ligand-gated cation channels (7, 30). Presently, eight P2Y subtypes (P2Y1, P2Y2, P2Y4, P2Y6, and P2Y11–14) and seven subtypes of P2X (P2X1–7) have been identified in mammals. The metabotropic P2Y receptors possess an extracellular NH2 terminus, an intracellular COOH terminus, and seven transmembrane domains and signal through heterotrimeric G proteins (7). In contrast, ionotropic P2X receptor subunits are composed of two transmembrane domains, a large N-glycosylated extracellular loop and intracellular NH2 and COOH termini. Functional channels, composed of homo- or heteromultimers of three P2X subunits (30), are permeable to cations, resulting in membrane depolarization and, in many cases, Ca2+ influx upon receptor activation. Unlike P2Y family members that are activated by adenine or uracil nucleotides, or both, P2X receptors are activated solely by adenine nucleotides (7).

Relative to other P2X family members, the P2X7 receptor is unique in that 1) it is activated only by relatively high concentrations of extracellular ATP (>100 μM), 2) it possesses an extended intracellular COOH terminus containing putative interaction domains for structural and signaling proteins (31), and 3) it causes dynamic blebbing of the plasma membrane in numerous cell types (56). The P2X7 receptor is expressed in cells of the osteoblast lineage (18, 24) together with other P2X and P2Y subtypes (12, 47). Ke et al. (29) demonstrated that genetically modified mice, in which P2X7 function is disrupted, exhibit diminished periosteal bone formation and excessive trabecular bone resorption. Moreover, Li et al. (36) found that these same mice also exhibit impaired skeletal responses to mechanical load, implicating P2X7 in mechanotransduction in vivo. Interestingly, other investigators have described a distinct skeletal phenotype in a second mouse model in which P2X7 expression is disrupted (17). However, recent studies have revealed this second model to be an inadvertent tissue-specific knockout (43, 63) in which the presence or absence of P2X7 in osteoblasts has yet to be determined. In keeping with the phenotype of the first mouse model (29), our laboratory has shown that activation of P2X7 receptors by exogenous nucleotides couples to production of the potent lipid mediator lysophosphatidic acid in cells of the osteoblast lineage, resulting in dynamic membrane blebbing and enhanced osteogenesis (50, 51). However, many of the signaling pathways and downstream effects of P2X7 activation in osteoblasts remain to be elucidated.

Numerous metabolic demands are placed on osteoblasts during differentiation and in the production and mineralization of osteoid. In this regard, osteoblast differentiation is associated with a striking increase in cellular metabolism (9, 34). Whereas osteoblast progenitors rely primarily on glycolytic metabolism, induction of osteoblast differentiation leads to increases in aerobic respiration and ATP production (34). In addition, differentiated osteoblasts possess an increased number of mitochondria with hyperpolarized transmembrane potentials, indicative of enhanced mitochondrial activity (9, 34). Heterologous expression of P2X7 in human embryonic kidney (HEK)-293 and HeLa cells hyperpolarizes the mitochondrial membrane and increases basal mitochondrial Ca2+ levels and intracellular ATP content (1). However, it is not known whether activation of P2X7 influences metabolism in cells such as osteoblasts, which express these receptors endogenously.

In the present study, we used a Cytosensor microphysiometer to monitor metabolic acid production (proton efflux) from osteoblasts in real time as a measure of cellular metabolism (38). We show that activation of P2X7 elicits a large and sustained increase in proton efflux. Initiation of the response is dependent on Ca2+ influx through activated P2X7 receptors, whereas its maintenance is dependent on phosphatidylinositol 3-kinase (PI3K) activity and the availability of glucose. These data show for the first time that stimulation of P2X7 elicits a dramatic increase in metabolic acid production and efflux, which is triggered by elevation of cytosolic free Ca2+ concentration ([Ca2+]i) and sustained in part by PI3K-activated glucose metabolism.

MATERIALS AND METHODS

Materials and solutions.

α-Minimum essential medium (α-MEM; catalog no. 12571) buffered with HCO3 (26 mM), heat-inactivated FBS (catalog no. 12483), antibiotic solution (10,000 U/ml penicillin, 10,000 μg/ml streptomycin, and 25 μg/ml amphotericin B, catalog no. 15240), trypsin solution (nominally Ca2+ and Mg2+ free, 0.05% trypsin, and 0.53 mM EDTA, catalog no. 25300), Dulbecco's phosphate-buffered saline (catalog no. 14040), HCO3-free MEM (catalog no. 41200 and 41500), MEM amino acid solution (×50, catalog no. 11130-051), l-glutamine (catalog no. 25030–081), HEPES (catalog no. 11344-033), HEPES buffer solution (1 M, catalog no. 15630), and UltraPure distilled water (DNase/RNase-free, catalog no. 10977-023) were obtained from GIBCO (Invitrogen Canada, Burlington, ON, Canada). BSA (fraction V, fatty acid free, catalog no. 10775835001) was from Roche Diagnostics (Indianapolis, IN). HCO3- and glucose-free DMEM (catalog no. D5030), ATP disodium salt (catalog no. A7699), UTP trisodium salt hydrate (UTP, catalog no. U6750), 2′-3′-O-(4-benzoylbenzoyl)ATP triethylammonium salt (BzATP; catalog no. B6396), lactate dehydrogenase (catalog no. L3916), β-NAD (catalog no. N8535), glycine buffer (catalog no. 8263), and EGTA (catalog no. E4378) were from Sigma (St. Louis, MO). FM4-64 (catalog no. T-13320) and indo-1 acetoxymethyl ester were from Molecular Probes (Eugene, OR). Click-iT terminal deoxynucleotidyl-mediated dUTP nick-end labeling (TUNEL) Alexa Fluor 488 imaging assay was from Invitrogen (catalog no. C10245). Vectashield mounting medium with 4,6-diamidino-2-phenylindole (DAPI) was from Vector Laboratories (Burlingame, CA). 3-[5-(2,3-Dichlorophenyl)-1H-tetrazol-1-yl]methyl pyridine hydrochloride (A 438079 HCl; catalog no. 2972) and N-{1-[(cyanoamino)(5-quinolinylamino)methylene amino]-2,2-dimethylpropyl}-3,4-dimethoxybenzene-acetamide (A 740003, catalog no. 3701) were from Tocris Bioscience (Ellisville, MO). Wortmannin (catalog no. 681675), 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY-294002; catalog no. 440202), and staurosporine (catalog no. 569397) were from Calbiochem (EMD Biosciences, San Diego, CA). Nucleotides were dissolved in the appropriate buffer, A 438079 HCl was dissolved in water and stored in aliquots at −20°C, and FM4-64, A 740003, wortmannin, LY-294002, and staurosporine were dissolved in dimethyl sulfoxide (DMSO) and stored in aliquots at −20°C.

Standard superfusion medium, used in experiments monitoring proton efflux, was HCO3-free MEM supplemented with HEPES (1 mM) and BSA (1 mg/ml). Glucose dependence was investigated using DMEM supplemented with HEPES (1 mM) and BSA (1 mg/ml) with or without glucose (5.5 mM, glucose-containing and glucose-free medium, respectively). Ca2+-dependent experiments were performed using Ca2+-containing and Ca2+-free buffers. Ca2+-containing buffer consisted of (in mM) 5.4 KCl, 1.8 CaCl2, 0.8 MgSO4, 116.4 NaCl, 1 NaH2PO4·H2O, 2 l-glutamine, 5.5 glucose, and 1 HEPES supplemented with MEM amino acid solution (1×) and BSA (1 mg/ml). Ca2+-free buffer contained no added CaCl2 and was supplemented with EGTA (0.5 mM). For morphological assessments, culture medium was removed and replaced with nominally divalent cation-free buffer containing (in mM) 140 NaCl, 5 KCl, 20 HEPES, and 10 glucose. For measurements of [Ca2+]i, indo-1-loaded cells were suspended in HEPES buffer containing (in mM) 135 NaCl, 5 KCl, 1 MgCl2, 1 CaCl2, 10 glucose, and 20 HEPES. All media and buffers described above were adjusted to 290 ± 5 mOsm/l and pH 7.30 ± 0.02.

Cells and culture.

The MC3T3-E1 osteoblast-like cell line was obtained from the American Type Culture Collection (Rockville, MD). A clonal nontransformed cell line established from newborn mouse calvaria (62), variants of MC3T3-E1 cells that exhibit different phenotypic characteristics in vitro have since been isolated (64). For the present studies, the MC3T3-E1 subclone 4 line was selected because these cells exhibit properties of osteoblasts, including elevation of cAMP in response to PTH and expression of transcripts for Runx2, bone sialoprotein, and osteocalcin. Moreover, cultures form mineralized bone-like nodules upon supplementation with ascorbic acid and phosphate (64). Expression of P2X7 receptors in these cells has been demonstrated previously (36, 46, 55).

The UMR-106 cell line, isolated from a rat osteosarcoma, was also obtained from American Type Culture Collection. These cells exhibit several properties of osteoblasts, including type I collagen production, high alkaline phosphatase activity, responsiveness to PTH, and formation of mineralized tumors in vivo (52). However, unlike the MC3T3-E1 cells, UMR-106 cells do not express P2X7 (50). MC3T3-E1 and UMR-106 cells were subcultured twice weekly and maintained in α-MEM containing HCO3 (26 mM) supplemented with 10% FBS and 1% antibiotic solution in humidified 5% CO2 at 37°C.

Morphological assessments.

For time-lapse recordings, cells were trypsinized, plated at a density of 1.0–1.5 × 104 cells/cm2 on 35-mm culture dishes (Nunc, Thermo Fisher Scientific, Rochester, NY) in supplemented α-MEM, and cultured for 24–48 h. For recordings, culture medium was removed and replaced with nominally divalent cation-free buffer. Dishes were placed on a heater stage (∼35°C) mounted on an inverted-phase contrast microscope (Nikon Plan-Fluor ×20 objective, 0.45 numerical aperture). For quantitative analysis, the percentages of cells in each field exhibiting membrane blebbing before and after addition of test substances were determined. A single field was analyzed from each cell preparation. Experiments were performed on at least three independent preparations.

For live cell confocal microscopy, MC3T3-E1 cells were plated at a density of 1.5 × 104 cells/cm2 on 35-mm glass bottom culture dishes (MatTek, Ashland, MA) for 24 h in supplemented α-MEM. Membranes were labeled by incubation with lipophilic fluorescent probe FM4-64 (2 µg/ml) in medium for 15 min at 37°C. Medium was then replaced with nominally divalent cation-free buffer, and cells were observed by confocal microscopy at ∼28–30°C (model LSM 510; Zeiss, Jena, Germany). Images were acquired using Zeiss Plan-Apochromat ×63 objective (1.4 numerical aperture) and 488-nm Ar+ ion laser excitation. The emission was filtered at 560 nm long pass, and images were captured using time-lapse mode.

Measurement of proton efflux.

MC3T3-E1 and UMR-106 cells were seeded on porous polycarbonate membranes (Transwell, 12-mm diameter, 3-μm pore size; Corning Costar, Corning, NY) in supplemented α-MEM at a density of 8.8 × 104 and 3.5 × 104 cells/cm2, respectively. After 48 h, cells adhering to the polycarbonate membranes were placed in microflow chambers and positioned above silicon-based potentiometric sensors, which detect changes in extracellular pH (pHo) of as little as 10−3 units (Cytosensor microphysiometer; Molecular Devices, Sunnyvale, CA) (38). Cells were superfused continuously at a rate of 100 μl/min with medium at 37°C. Superfusion media with low buffering power were used to enhance the small alterations in pHo arising by efflux of protons from cells. Each chamber was supplied with medium from one of two reservoirs selected by a computer-controlled valve. Test substances were directly introduced in superfusion medium, and changes in proton efflux were monitored. The lag time between a valve switch and the arrival of test solutions at the microflow chambers was 4–5 s.

Surface potential of each silicon sensor, corresponding to the pHo, was plotted as a voltage-time trace. At 37°C, 61 mV corresponds to 1 pH unit. To measure the rate of acidification (net cellular efflux of proton equivalents), fluid flow to cells was stopped periodically for 30 s. During this time, acid accumulated in the microflow chamber (volume 2.8 μl), causing pHo to decrease. Measurement of acidification rate was obtained by linear least-squares fit to the slope of the pHo-time trace during the time when fluid flow to the cells was stopped. Because of an artefact arising from changing medium, the first data point after superfusion with agonist began was sometimes omitted from the trace.

Measurement of lactate efflux.

Extracellular lactate was measured using a spectrophotometric assay based on generation of NADH via the catalytic action of lactate dehydrogenase (modified from Ref. 59). Briefly, cells were seeded at a density of 1.5 × 104 cells/cm2 on 60-mm culture dishes (BD Biosciences, Franklin Lakes, NJ) in supplemented α-MEM for 48 h. On the day of the experiment, growth medium was replaced by α-MEM supplemented with BSA (1 mg/ml) and 1% antibiotic solution (serum free) in 5% CO2 at 37°C. After 3 h, cultures were incubated with BzATP or vehicle. Samples of medium (100 μl) were then collected for lactate determination. To measure lactate concentration, 6.9 μl of each sample was combined with 193.1 μl of lactate assay solution (50 mg of β-NAD dissolved in 10 ml of glycine buffer, 20 ml of deionized H2O, and 17 U/ml lactate dehydrogenase) in a 96-well UV plate (Corning Costar). The plate was incubated at 37°C for 15–30 min, and absorbance at 340 nm was quantified with a microplate reader (Tecan, Durham, NC). The concentration of lactate in each sample was determined from a standard curve of absorbance for known lactate concentrations.

Measurement of [Ca2+]i.

Changes in [Ca2+]i were measured using spectrofluorimetry, as described previously (59). Briefly, MC3T3-E1 cells were seeded at a density of 1.5 × 104 cells/cm2 on 60-mm culture dishes (BD Biosciences) in supplemented α-MEM for 48 h. On the day of the experiment, cells were loaded with indo-1 by incubation in serum-supplemented medium with indo-1 acetoxy-methyl ester (2 μg/ml) for 30 min at 37°C and 5% CO2. After loading, cells were washed with PBS and harvested by trypsinization. Supplemented medium was added to neutralize the trypsin, and cells were then sedimented and resuspended in HEPES-buffered MEM (catalog no. 41200). For measurement of [Ca2+]i, 1-ml aliquots of indo-1-loaded cell suspensions (∼1.0 × 106 cells/ml) maintained at room temperature in HEPES-buffered MEM were sedimented and resuspended in 2 ml of HEPES buffer in a fluorometric cuvette at room temperature. Test substances were added directly to the cuvette.

[Ca2+]i was monitored using a dual-wavelength spectrofluorimeter (Model RF-M2004; Photon Technology International, South Brunswick, NJ) at 355 nm excitation and emission wavelengths of 405 and 485 nm. The system software was used to calculate the ratio R, which is the fluorescence intensity at 405 nm divided by the intensity at 485 nm. [Ca2+]i was determined from the relationship [Ca2+] = Kd[(R − Rmin)/(Rmax − R)]β, where Kd (for the indo-1-Ca2+ complex) was taken as 250 nm, Rmin and Rmax were the values of R at low and saturating concentrations of Ca2+, respectively, and β was the ratio fluorescence at 485 nm measured at low and saturating Ca2+ concentrations.

Assessment of apoptosis.

MC3T3-E1 cells were seeded at a density of 1.5 × 104 cells/cm2 on 12-mm glass coverslips (Fisher Scientific, Ottawa, ON, Canada) in 24-well culture plates (BD Biosciences) in supplemented α-MEM for 48 h. On the day of the experiment, growth medium was replaced by α-MEM supplemented with BSA (1 mg/ml) and 1% antibiotic solution (serum free) in 5% CO2 at 37°C. Three hours later, cultures were incubated with test substances. After 24 h, cells were fixed with paraformaldehyde (4%) in sucrose solution (2%), permeabilized with 0.25% Triton X-100 in PBS, and assessed for apoptosis using the Click-iT TUNEL Alexa Fluor 488 imaging assay according to the manufacturer's instructions. Selected untreated samples from each experiment were incubated with DNase I before staining as a positive control for the TUNEL assay. After staining, coverslips were sealed using Vectashield mounting medium with DAPI and visualized by confocal microscopy (Zeiss model LSM 510). Images were acquired using a Zeiss Plan-Apochromat ×40 objective (1.2 numerical aperture) with a 730-nm Chameleon multiphoton or 488-nm Ar+ ion laser excitation, and the emission was filtered at 390–465 or 500–550 nm band pass, respectively. The total number of cells and the number of TUNEL-positive cells were determined from randomly selected fields.

Statistical analyses.

Proton efflux was normalized as a percentage of basal efflux in standard superfusion medium before the addition of a test substance or change in superfusion solution. This normalization compensated for differences in cell numbers among the chambers. Results are presented as means ± SE. Differences between two groups were evaluated by t-test. Differences among three or more groups were evaluated by one-way or two-way analysis of variance followed by Tukey or Bonferroni multiple comparison tests. Differences were accepted as statistically significant at P < 0.05.

RESULTS

The P2X7 agonist BzATP induces dynamic membrane blebbing in MC3T3-E1 osteoblast-like cells.

Among P2 nucleotide receptors, the ability to induce membrane blebbing is a unique property of P2X7 (44). Using primary cultures of rat and murine calvarial cells, we have shown previously that activation of P2X7 elicits dynamic and reversible blebbing of the plasma membrane in only a subpopulation of these cells (51). To investigate the proportion of MC3T3-E1 cells expressing functional P2X7 receptors, we monitored morphology using time-lapse phase contrast microscopy. Prior to addition of agonist, few cells exhibited membrane blebs [Fig. 1A (before)]. Cultures were then treated with the P2X7 agonist BzATP (300 μM) or vehicle for 20 min. BzATP caused cellular retraction followed by dynamic plasma membrane blebbing [Fig. 1A (BzATP) and Supplemental Video S1]. In contrast, UMR-106 cells exhibited cellular retraction but no membrane blebbing in response to BzATP [Fig. 1A (BzATP) and Supplemental Video S2], consistent with the absence of P2X7 in these cells.

Fig. 1.

The P2X7 agonist 2′,3′-O-(4-benzoylbenzoyl)ATP (BzATP) induces membrane blebbing in MC3T3-E1 osteoblast-like but not UMR-106 osteosarcoma cells. A: morphology of MC3T3-E1 and UMR-106 osteoblast-like cells was monitored by time-lapse phase contrast microscopy. Cultures were bathed in nominally divalent cation-free buffer at ∼35°C for 10 min (before). MC3T3-E1 cells, which endogenously express P2X7 receptors, exhibited dramatic retraction and dynamic blebbing of the plasma membrane following application of BzATP (300 μM); see also Supplemental Video S1. In contrast, BzATP (300 μM) caused retraction but failed to induce membrane blebbing in UMR-106 cells, which do not express P2X7 receptors; see also Supplemental Video S2. White arrows indicate blebbing cells. B: to examine bleb morphology, MC3T3-E1 cells were loaded with FM4-64 to stain membranes and observed using confocal microscopy. Cells were bathed in nominally divalent cation-free buffer. BzATP (300 μM) was added at time 0 min. Images show 2-μm-thick optical sections through a single cell at 5-min intervals. N, nucleus. Images are representative of cells demonstrating dynamic membrane blebbing from 4 independent preparations. C: percentage of MC3T3-E1 and UMR-106 cells exhibiting dynamic membrane blebbing determined from time-lapse movies obtained before and during 20 min of treatment with vehicle or BzATP (300 μM). *Significant difference compared with respective vehicle for each cell line (P < 0.05). Data are means ± SE (n = 8 samples from 6 independent preparations for MC3T3-E1 cells, and n = 6 samples from 4 independent preparations for UMR-106 cells). Scale bars = 20 μm in A and B.

To further characterize the dynamic changes in morphology of individual MC3T3-E1 cells exposed to BzATP, membranes were stained with the fluorescent lipophilic dye FM4-64, and cells were observed by live cell confocal microscopy. BzATP (300 μM) induced the formation of multiple membrane blebs that enlarged and shrank asynchronously (Fig. 1B). These blebs occurred over the entire cell surface and did not contain FM4-64-stained intracellular membranes.

We used the time-lapse recordings to quantify the percentage of cells exhibiting membrane blebbing. BzATP induced significant blebbing of the plasma membrane in MC3T3-E1 cells (85 ± 3%; Fig. 1C), thereby establishing the expression of functional P2X7 receptors in the majority of these cells. As expected, no significant increase in the percentage of blebbing cells was seen following treatment of UMR-106 cells with BzATP (Fig. 1C). Thus, in contrast to UMR-106 cells (which do not express P2X7) and primary bone cell systems (in which only a subpopulation of cells express functional P2X7 receptors), the vast majority MC3T3-E1 osteoblasts expressed functional P2X7 receptors. Consequently, we used the MC3T3-E1 cell line to characterize the effects of P2X7 activation on metabolic acid production by osteoblasts.

Effects of BzATP on proton efflux from MC3T3-E1 cells.

To assess changes in cellular metabolism, proton efflux from MC3T3-E1 cells was monitored by microphysiometry. Basal proton efflux in standard superfusion medium remained steady for periods of >60 min. Superfusion with BzATP (300 μM) for 9 min caused dramatic changes in proton efflux (Fig. 2). The first phase of the BzATP-induced response (initial phase) consisted of a variable, transient depression in proton efflux, which reached its maximum at 1.5–3 min after the addition of BzATP, followed by a slow increase to levels above basal by 9 min in the presence of agonist. The second phase (sustained phase) occurred following removal of BzATP and was characterized by a large sustained increase in proton efflux that lasted for >1 h following agonist removal (Fig. 2). A transient overshoot in proton efflux was also observed immediately upon washout of BzATP1. Treatment with vehicle did not cause any appreciable change in proton efflux. To exclude the possibility that responses to BzATP were due to direct interaction with the silicon sensor, cells were devitalized with 1% Triton X-100 and then superfused with agonist. The basal proton efflux from nonvital cells was zero, and no changes were observed upon superfusion with BzATP (300 μM), ruling out the possibility of interaction with the silicon sensor.

Fig. 2.

BzATP causes a sustained increase in proton efflux from MC3T3-E1 cells. MC3T3-E1 cells were cultured on porous polycarbonate membranes, and proton efflux was monitored by microphysiometry. Cells were superfused with standard medium, and at 1.5-min intervals, superfusion was interrupted for 30 s to measure acidification rate. Net efflux of proton equivalents (proton efflux) was calculated from the acidification rate and expressed as a percentage of basal proton efflux. Where indicated by the shaded area, MC3T3-E1 cells were superfused with BzATP (300 μM; ●) or vehicle (○) in standard medium for 9 min. During superfusion, proton efflux first decreased and then slowly increased. After agonist removal, a sharp elevation was elicited followed by a large sustained increase in proton efflux that persisted for a period of >60 min. In contrast, superfusion with vehicle had no effect. A: data are representative traces. B: data are means ± SE of 8 samples from 4 independent preparations. Data points for BzATP from 22.5 to 60 min are significantly different compared with vehicle at the same time points (P < 0.05).

We next investigated the dependence of these changes in proton efflux on BzATP concentration (Fig. 3). The initial phase of the BzATP response was quantified as the average increase in proton efflux above basal within the 9 min during which BzATP was applied. The sustained phase was quantified as the average increase in proton efflux above basal 12–30 min after washout of BzATP. BzATP (100 μM, a concentration sufficient to fully activate several P2 receptors but only partially activate P2X7) elicited an initial transient increase in proton efflux (49 ± 19%, P < 0.05; Fig. 3B) but no significant sustained phase (Fig. 3C). In contrast, BzATP (300 μM, a concentration sufficient to fully activate P2X7) induced a large sustained elevation in proton efflux (86 ± 18%, P < 0.05; Fig. 3C). These data are consistent with a role for P2X7 in triggering the sustained increase in proton efflux, and subsequent studies were performed using a concentration of 300 μM BzATP.

Fig. 3.

Increases in proton efflux are dependent on BzATP concentration. MC3T3-E1 cells were superfused with standard medium, and proton efflux was monitored. A: where indicated by the shaded areas, MC3T3-E1 cells were challenged with indicated concentrations (μM) of BzATP or vehicle (not shown) in standard medium for 9 min. B: amplitude of the initial phase, quantified as the average increase in proton efflux above basal during application of BzATP or vehicle. C: amplitude of the sustained phase, quantified as the average increase in proton efflux above basal, 12–30 min following removal of BzATP or vehicle [indicated by gray-outlined boxes in A (iiii)]. *Significant difference compared with vehicle (P < 0.05). Data are means ± SE (n = 4–8 samples from 4 independent preparations).

Role of P2X7 in mediating nucleotide-induced increases in proton efflux.

In addition to P2X7, cells of the osteoblast lineage express a number of other P2X and P2Y nucleotide receptors. BzATP is not specific for P2X7 because it also activates a number of other P2 receptors (44), including P2Y11, P2X1, P2X3, and P2X4 (6, 10, 41), with similar potency. Thus, responses to BzATP alone cannot be used to establish involvement of the P2X7 receptor. To investigate the nucleotide receptors mediating the effects of BzATP on proton efflux, we first tested the effects of other P2 receptor agonists. Parallel samples of cells were superfused with low concentrations of ATP (100 μM, which activates P2X1–5 and many P2Y receptors) and UTP (100 μM, which activates P2Y2, P2Y4, and P2Y6 receptors) and high concentrations of ATP (1 mM) and BzATP (300 μM), both of which also activate P2X7. Responses to UTP and a low concentration of ATP (100 μM) consisted only of a transient elevation in proton efflux during agonist application [Fig. 4A (i and ii)]. On the other hand, the response to a higher concentration of ATP (1 mM) consisted of both initial transient and sustained elevations in proton efflux [Fig. 4A (iii)]2. Transient elevations in proton efflux caused by UTP (100 μM) and ATP (0.1–1 mM) during the initial phase were significantly different compared with vehicle-treated cultures [Fig. 4B (i)]. Consistent with the involvement of P2X7, only a high concentration of ATP (1 mM) elicited a significant and sustained increase in proton efflux following washout of agonist, which was similar to that seen with BzATP (300 μM) [Fig. 4B (ii)]. Since UTP (and UDP) can activate only certain members of the P2Y receptor family, the initial transient phase is most likely due to activation of P2Y receptors. On the other hand, the sustained phase is most likely triggered by P2X7.

Fig. 4.

Sustained increases in proton efflux are seen only with agonists that activate the P2X7 nucleotide receptor. MC3T3-E1 cells were superfused with standard medium, and proton efflux was monitored. A: where indicated by the shaded areas, MC3T3-E1 cells were challenged with BzATP (300 μM), ATP (100 μM or 1 mM), UTP (100 μM), or vehicle (Veh; not shown) in standard medium for 9 min. B (i): amplitude of the initial phase, quantified as the average increase in proton efflux above basal during application of agonist. B (ii): amplitude of the sustained phase, quantified as the average increase in proton efflux above basal, 12–30 min after application of agonist (indicated by gray-outlined boxes in A). ATP (1 mM, a concentration sufficient to activate the P2X7 receptor) mimicked the effects of BzATP, whereas agonists for other P2 receptors (UTP and a lower concentration of ATP) failed to induce a sustained increase in metabolic acid production. *Significant difference compared with Veh (P < 0.05). Data are means ± SE (n = 5–8 samples from 3 to 5 independent preparations).

We used two approaches to confirm the involvement of P2X7 receptors in triggering the sustained phase of BzATP-induced proton efflux. First, we compared responses of MC3T3-E1 cells (which express P2X7) and UMR-106 cells (which do not express P2X7). Parallel samples of MC3T3-E1 and UMR-106 cells were superfused with BzATP (300 μM) or vehicle (Fig. 5A). Proton efflux from UMR-106 cells was suppressed in the presence of BzATP (∼60% of basal levels). However, BzATP failed to elicit the sustained increase seen in MC3T3-E1 cells (Fig. 5B). Second, we examined the effects of A 438079 and A 740003, two specific P2X7 antagonists (14, 42). Although treatment with A 438079 or A 740003 (10 μM) had no appreciable effect on basal proton efflux (Fig. 6, A and C), both significantly blocked the sustained phase of the BzATP-induced response (Fig. 6, B and D). Taken together, we conclude that the sustained phase of nucleotide-induced increase in proton efflux from osteoblast-like cells is triggered by activation of P2X7.

Fig. 5.

BzATP (Bz) induces an increase in proton efflux from MC3T3-E1 but not UMR-106 cells. MC3T3-E1 and UMR-106 cells were superfused with standard medium, and proton efflux was monitored. A: parallel samples of MC3T3-E1 [A (i)] and UMR-106 cells [A (ii)] were challenged with BzATP (300 μM; ●) or Veh (○) in standard medium for 9 min where indicated by the shaded areas. During superfusion with BzATP, proton efflux from MC3T3-E1 cells first decreased and then slowly increased. In contrast, UMR-106 cells exhibited a sustained decrease in proton efflux in the presence of BzATP. Following removal of agonist, a sustained increase in metabolic acid production from MC3T3-E1 cells was observed, whereas proton efflux from UMR-106 cultures returned to basal levels. B: amplitude of the sustained phase, quantified as the average increase in proton efflux above basal, 12–30 min following application of Veh or Bz [indicated by gray-outlined boxes in A (iii)]. Superfusion with Bz caused a sustained increase in metabolic acid production from MC3T3-E1 cells but not UMR-106 cells. Thus, the metabolic response correlates with expression of the P2X7 nucleotide receptor. *Significant difference compared with respective Veh for each cell line (P < 0.05). Data are means ± SE (n = 3–9 samples from 3 independent preparations).

Fig. 6.

Blockade of P2X7 receptors suppresses the increase in proton efflux induced by BzATP. MC3T3-E1 cells were superfused with standard medium, and proton efflux was monitored. A: parallel samples of MC3T3-E1 cells were superfused initially with standard medium. Subsequently, samples were superfused with either control [ddH2O; A (i)] or the P2X7 antagonist 3-[5-(2,3-dichlorophenyl)-1H-tetrazol-1-yl]methyl pyridine hydrochloride [A 438079; 10 μM, A (ii)] for the period indicated by the horizontal bar beneath the graph. After 6 min, cultures in both A (i) and A (ii) were superfused with either vehicle (○) or BzATP (300 μM; ●) where indicated by the shaded area in the continued presence of the appropriate medium. B: amplitude of the sustained phase, quantified as the average increase in proton efflux above basal, 12–30 min after application of vehicle or BzATP [indicated by gray-outlined boxes in A (i) and (ii)]. C: parallel samples of MC3T3-E1 cells were superfused initially with standard medium. Subsequently, samples were superfused with either control [DMSO; C (i)] or the P2X7 antagonist N-{1-[(cyanoamino)(5-quinolinylamino)methylene amino]-2,2-dimethylpropyl}-3,4-dimethoxybenzene-acetamide [A 740003; 10 μM, C (ii)] for the period indicated by the horizontal bar beneath the graph. After 6 min, cultures in both C (i) and C (ii) were superfused with either vehicle (○) or BzATP (300 μM; ●) where indicated by the shaded area in the continued presence of the appropriate medium. D: amplitude of the sustained phase, quantified as the average increase in proton efflux above basal, 12–30 min after application of vehicle or BzATP [indicated by gray-outlined boxes in C (i and ii)]. *Significant difference compared with respective vehicle (P < 0.05). Data are means ± SE (n = 5–12 samples from 3 to 4 independent preparations).

Effects of P2X7 agonists on survival and apoptosis of MC3T3-E1 cells.

Activation of P2X7 promotes apoptosis in a number of cell systems (2), although its ability to promote cell death in osteoblasts is controversial. ATP and BzATP have been shown to induce delayed release of lactate dehydrogenase from osteoblast-like cells (18), indicating cell death. However, we have reported previously that BzATP-induced membrane blebbing of murine calvarial cells was reversible upon removal of agonist, suggesting that P2X7 receptors do not induce acute cell death (51). Moreover, stimulation of P2X7 receptors in MC3T3-E1 cells does not activate caspase 3 (36), a key mediator of apoptosis, arguing against a proapoptotic effect. To determine whether the increased proton efflux elicited by activation of P2X7 might be associated with induction of apoptosis, we investigated the effects of prolonged P2X7 stimulation on apoptosis and survival of MC3T3-E1 cells. Cultures were incubated with vehicle, BzATP (300 μM), ATP (1 mM), or staurosporine (1 μM, positive control) for 24 h in the absence of serum. Cultures were then fixed, permeabilized, subjected to TUNEL, and visualized by confocal microscopy (Fig. 7A). Staurosporine induced apoptosis in close to 99% of cells, whereas BzATP or ATP caused only a slight increase in the percentage of TUNEL-positive cells compared with vehicle (Fig. 7B). In this regard, staurosporine alone induced a significant decrease in total cell number, whereas treatment with BzATP had no significant effect (Fig. 7C). Interestingly, ATP slightly increased the total number of cells per coverslip compared with vehicle (Fig. 7C), consistent with an increase in proliferation. In summary, 1) the percentage of cells exhibiting TUNEL staining after treatment with BzATP or ATP (∼5%) was substantially less than the total number of cells expressing functional P2X7 receptors (∼85%; Fig. 1), and 2) neither BzATP nor a high concentration of ATP caused a decrease in the total number of cells. These findings rule out the possibility that the P2X7-induced increase in proton efflux is associated with induction of apoptosis or cell death.

Fig. 7.

P2X7 agonists do not induce death of MC3T3-E1 cells. Cultures of MC3T3-E1 cells were bathed in serum-free α-MEM supplemented with bovine albumin (1 mg/ml) and 1% antibiotic solution for 3 h and subsequently treated with Veh, Bz (300 μM), ATP (ATP high; 1 mM), or staurosporine (St; 1 μM). After 24 h, cells were fixed, permeabilized, and subjected to terminal deoxynucleotidyl-mediated dUTP nick-end labeling (TUNEL) to identify apoptotic cells (green). Nuclei were stained with 4,6-diamidino-2-phenylindole (DAPI; blue). A: samples were visualized by confocal microscopy. Images are representative of 3 independent preparations performed in duplicate. Scale bar = 20 μm. B: to quantify apoptosis, the number of TUNEL-positive cells was expressed as a percentage of the total number of cells in a field. St (positive control) induced apoptosis in nearly all cells, whereas Bz or ATP high elicited only a small increase in the percentage of TUNEL-positive cells. When treated with DNase I, 100% of cells were TUNEL positive, confirming validity of the assay. C: the total number of cells per coverslip for each treatment was determined. St decreased cell number significantly, whereas Bz had no significant effect. Interestingly, ATP high caused an increase in cell number compared with vehicle. *Significant difference compared with vehicle (P < 0.05); **significant difference compared with St. For both B and C, data are means ± SE (n = 3 independent preparations, each performed in duplicate).

Role of glucose metabolism in generating sustained nucleotide-induced proton efflux.

Transient changes in proton efflux can arise from the activation or inhibition of transporters mediating the movement of proton equivalents across the plasma membrane; however, sustained increases in proton efflux reflect increases in the production of proton equivalents and their subsequent extrusion. Therefore, we considered the possibility that the sustained phase was due to enhanced glucose metabolism. We first investigated dependence of the sustained phase on extracellular glucose. Parallel samples of MC3T3-E1 cells were treated with BzATP (300 μM) or vehicle in standard superfusion medium (containing 5.5 mM glucose). Subsequently, when BzATP-treated cells were within the sustained phase, samples were superfused with glucose-free medium, resulting in a large, rapid decrease in proton efflux in both vehicle- and BzATP-treated samples to comparable levels (Fig. 8). Thus, most if not all of the sustained phase of the nucleotide-induced response is dependent on the presence of extracellular glucose.

Fig. 8.

Sustained increase in proton efflux induced by Bz is dependent on extracellular glucose. MC3T3-E1 cells were superfused with standard medium, and proton efflux was monitored. A: where indicated by the shaded area, parallel samples of MC3T3-E1 cells were challenged with Bz (300 μM; ●) or Veh (○) in standard running medium for 9 min. Subsequently, both samples were superfused with glucose-free medium for 9 min where indicated by the horizontal bar beneath the graph. B: amplitude of the response, quantified as the change in proton efflux relative to basal, calculated at the times indicated by I and II in A. Proton efflux from both Bz- and Veh-treated cultures exhibited a large, rapid decrease to comparable levels when superfused with glucose-free medium, indicating that the sustained phase is dependent on the presence of glucose. *Significant difference compared with respective Veh (P < 0.05). Data are means ± SE (n = 6–8 samples from 3 independent preparations).

Next, we assessed the effects of BzATP on extracellular lactate levels. MC3T3-E1 cells were incubated with BzATP (300 μM) or vehicle, and medium samples were collected for analysis over a 3-h period. BzATP caused a significant increase in extracellular lactate concentration (indicative of lactate efflux) compared with vehicle-treated cultures (Fig. 9)3. In MC3T3-E1 cells, the observed ∼250% increase in lactate efflux (from 0.5 to 3 h) is comparable with the magnitude of the sustained increase in proton efflux induced by BzATP. Taken together, these data are consistent with the sustained phase arising from enhanced rates of glycolytic metabolism.

Fig. 9.

BzATP increases lactate production by MC3T3-E1 cells. Cultures of MC3T3-E1 cells were bathed in serum-free α-MEM supplemented with bovine albumin (1 mg/ml) and 1% antibiotic solution for 3 h and subsequently treated with BzATP (300 μM) or vehicle (beginning at time 0). Samples of extracellular medium were collected for lactate determination at the indicated times. To calculate lactate efflux (change in extracellular lactate concentration/h), data were fit by linear regression from 0.5 to 3 h. BzATP caused an ∼250% increase in lactate efflux compared with vehicle-treated cultures. *Significant difference compared with vehicle at the same time point (P < 0.05). Data are means ± SE (n = 3 samples from 3 independent preparations).

Dependence of nucleotide-induced proton efflux on Ca2+.

To characterize signaling mechanisms underlying the actions of nucleotides on proton efflux, we investigated the role of Ca2+ influx through P2X7 receptors. Since P2X7-induced Ca2+ signaling in cells of the osteoblast lineage had not been characterized previously, we first examined changes in [Ca2+]i of indo-1-loaded cells by fluorescence spectrophotometry. In MC3T3-E1 cells, BzATP (300 μM) induced both a transient and sustained increase in [Ca2+]i, whereas UTP or ATP (100 μM) elicited only a transient elevation (Fig. 10A). The initial transient increase in [Ca2+]i is likely due to activation of P2Y receptors, causing release of Ca2+ from intracellular stores. A 438079 (10 μM) blocked the sustained increase in [Ca2+]i elicited by BzATP in MC3T3-E1 cells (not shown). Moreover, both UTP and ATP (100 μM) elicited a transient elevation in [Ca2+]i in UMR-106 cells, whereas BzATP (300 μM) had no effect (Fig. 10B), consistent with a role for P2X7 in sustained Ca2+ signaling.

Fig. 10.

Patterns of cytosolic free Ca2+ concentration ([Ca2+]i) elevation elicited by P2 receptor agonists. MC3T3-E1 (A) or UMR-106 cells (B) were loaded with the Ca2+-sensitive dye indo-1 and suspended in Ca2+-containing HEPES buffer in a fluorometric cuvette with continuous stirring. [Ca2+]i of parallel samples was monitored by fluorescence spectrophotometry. Where indicated by the arrows, vehicle, UTP (100 μM), ATP (ATP low; 100 μM), or BzATP (300 μM) was added directly to the cuvette. UTP and ATP low induced a transient increase in [Ca2+]i in both cell types. In contrast, BzATP induced a sustained elevation of [Ca2+]i in MC3T3-E1 cells and no response in UMR-106 cells. Treatment with vehicle had no effect. Traces are representative of responses from 3 to 4 independent preparations, each assayed in duplicate.

We next assessed the role of Ca2+ influx through P2X7 in initiating nucleotide-induced proton efflux. Parallel samples of MC3T3-E1 cells were superfused with Ca2+-containing or Ca2+-free medium. Superfusion with Ca2+-free medium had no appreciable effect on basal proton efflux (Fig. 11A). Subsequently, cells were treated with BzATP (300 μM) or vehicle in the continued presence or absence of extracellular Ca2+. Removal of Ca2+ from the superfusion medium abolished the BzATP-induced response (Fig. 11A). Upon reintroduction of Ca2+ to the superfusion medium, BzATP-induced proton efflux remained completely inhibited (10 ± 9%) when compared with the sustained phase in control cells exposed to BzATP in the presence of extracellular Ca2+ (149 ± 22%, P < 0.05; Fig. 11B).

Fig. 11.

Initiation of the BzATP-induced increase in proton efflux is dependent on extracellular calcium. MC3T3-E1 cells were superfused with standard medium, and proton efflux was monitored. A: parallel samples of MC3T3-E1 cells were superfused initially with standard medium. Subsequently, samples were superfused with either Ca2+-containing medium [1.8 mM Ca2+ (+Ca2+); A (i)] or calcium-free medium with 0.5 mM EGTA [−Ca2+; A (ii)] for the period indicated by the horizontal bar beneath the graph. After 6 min, cultures in both A (i) and A (ii) were superfused with vehicle (○) or BzATP (300 μM; ●) where indicated by the shaded areas in the continued presence of the appropriate medium. B: amplitude of the sustained phase, quantified as the average increase in proton efflux above basal, 12–30 min after application of vehicle or BzATP [indicated by gray-outlined boxes in A (i) and A (ii)]. BzATP did not induce an increase in proton efflux in the absence of extracellular Ca2+, indicating that initiation of the metabolic response is dependent on Ca2+ influx through the P2X7 receptor. *Significant difference compared with respective vehicle (P < 0.05). Data are means ± SE (n = 9–10 samples from 5 independent preparations). C: where indicated by the shaded area, parallel samples of MC3T3-E1 cells were challenged with BzATP (300 μM) in standard running medium for 9 min. Subsequently, samples were superfused with either Ca2+-containing medium [1.8 mM Ca2+ (+Ca2+); ●] or calcium-free medium with 0.5 mM EGTA (−Ca2+; ○) for 9 min where indicated by the horizontal bar beneath the graph. D: amplitude of the response, quantified as the change in proton efflux relative to basal, calculated at the times indicated by I and II in C. Filled bars represent responses of cells in the continuous presence of 1.8 mM Ca2+. Open bars represent responses of cells exposed to Ca2+-free medium. No difference in proton efflux was observed in cultures superfused with medium with or without Ca2+, indicating that maintenance of the sustained phase is not dependent on extracellular Ca2+. Data are means ± SE (n = 6–13 samples from 3 to 5 independent preparations).

We then investigated the role of extracellular Ca2+ in maintaining the sustained phase of the BzATP-induced response (subsequent to its initiation). Parallel samples of MC3T3-E1 cells were treated with BzATP (300 μM) in standard medium, and during the sustained phase BzATP-treated cells were superfused with either standard or Ca2+-free medium. Removal of extracellular Ca2+ had no appreciable effect on proton efflux (Fig. 11, C and D). Thus, initiation (but not maintenance) of the sustained phase requires the presence of extracellular Ca2+, indicating that influx of Ca2+ through activated P2X7 receptors is necessary to trigger sustained proton efflux.

Dependence of nucleotide-induced proton efflux on PI3K signaling.

Many growth factors and hormones stimulate cellular metabolism through activation of the PI3K/Akt pathway (15, 26). PI3K signaling also plays an important role in osteoblast differentiation (16, 21, 53). Moreover, the PI3K pathway in osteoblasts mediates the increase in metabolic acid production induced by IGF-I (60). Consequently, it was of interest to explore whether P2X7 activates PI3K signaling in osteoblasts. To characterize the role of PI3K signaling in mediating the effects of nucleotides on proton efflux, parallel samples of MC3T3-E1 cells were first treated with the irreversible PI3K inhibitor wortmannin (100 nM) or its vehicle (DMSO) in standard superfusion medium for 9 min. Wortmannin slightly decreased basal proton efflux (Fig. 12A), suggesting that unstimulated metabolism in osteoblast-like cells is partially dependent on PI3K activity. Subsequently, cells were treated with BzATP (300 μM) or vehicle in the continued presence of wortmannin or DMSO (Fig. 12A). The PI3K inhibitor partially blocked the sustained phase of the BzATP response (41 ± 8%) compared with parallel cultures treated with BzATP alone (104 ± 13%, P < 0.05; Fig. 12B)4.

Fig. 12.

Wortmannin (Wort) inhibits the sustained increase in proton efflux induced by Bz. MC3T3-E1 cells were superfused with standard medium, and proton efflux was monitored. A: parallel samples of MC3T3-E1 cells were superfused initially with standard medium. Subsequently, samples were superfused with either control (−Wort; ●) or the irreversible phosphatidylinositol 3-kinase (PI3K) inhibitor Wort (100 nM, +Wort; ○) for the period indicated by the horizontal bars beneath the graphs. After 9 min, cultures were superfused with either Bz [300 μM; A (i)] or vehicle [A (ii)] where indicated by the shaded areas in the continued presence of the appropriate medium. B: amplitude of the sustained phase, quantified as the average increase in proton efflux above basal, 12–30 min after application of vehicle or Bz (indicated by gray-outlined boxes in A). *Significant difference compared with respective vehicle (P < 0.05); **significant effect of Wort (P < 0.05). Data are means ± SE (n = 6–8 samples from 4 independent preparations).

To examine the role of PI3K signaling in initiation and maintenance of the sustained phase of the BzATP-induced response, we used LY-294002, a specific and reversible PI3K inhibitor structurally unrelated to wortmannin. To assess initiation, parallel samples of MC3T3-E1 cells were treated with LY-294002 (30 μM) or its vehicle (DMSO) in standard superfusion medium for 15 min. Like wortmannin, LY-294002 decreased basal proton efflux (Fig. 13A), confirming that baseline metabolism in osteoblast-like cells is partially dependent on the PI3K activity. Subsequently, cells were treated with BzATP (300 μM) or vehicle in the continued presence of LY-294002 or DMSO (Fig. 13A). LY-294002 significantly inhibited the sustained phase of the BzATP response (34 ± 6%; Fig. 13B, point I) compared with parallel cultures treated with BzATP in the absence of the inhibitor (82 ± 10%; Fig. 13B, point I)5. Interestingly, upon washout of LY-294002, proton efflux recovered to levels comparable with those observed in control cells exposed to BzATP in the presence of DMSO (Fig. 13B, point II). Thus, PI3K activity is not essential to trigger the sustained phase of nucleotide-induced proton efflux.

Fig. 13.

The reversible PI3K inhibitor 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY-294002) inhibits maintenance of the sustained increase in proton efflux induced by Bz. MC3T3-E1 cells were superfused with standard medium, and proton efflux was monitored. A: parallel samples of MC3T3-E1 cells were superfused initially with standard medium. Subsequently, samples were superfused with either control (−LY; ●) or the reversible PI3K inhibitor LY-294002 (30 μM, +LY; ○) for the period indicated by the horizontal bar beneath the graph. After 15 min, cultures were superfused with Bz (300 μM) or vehicle (not shown) where indicated by the shaded areas in the continued presence of the appropriate medium. B: amplitude of the response, quantified as the change in proton efflux relative to basal, calculated at the times indicated by I and II in A. Data are means ± SE (n = 7–10 samples from 5 independent preparations). C: where indicated by the shaded area, parallel samples of MC3T3-E1 cells were challenged with Bz (300 μM) in standard running medium for 9 min. Subsequently, samples were superfused with either control (−LY; ●) or LY-294002 (+LY; ○) for 9 min where indicated by the horizontal bar beneath the graph. D: amplitude of the response, quantified as the change in proton efflux relative to basal, calculated at the times indicated by III and IV in C. Data are means ± SE (n = 7 samples from 3 independent preparations). *Significant effect of LY-294002 (P < 0.05).

To assess the role of PI3K signaling in maintenance of the sustained phase of the BzATP-induced proton efflux, parallel samples were treated with BzATP (300 μM) and subsequently superfused with either LY-294002 (30 μM) or DMSO for 9 min before being returned to standard superfusion medium (Fig. 13C). LY-294002 caused a significant decrease in proton efflux (31 ± 4%; Fig. 13D, point IV) compared with cells treated with BzATP in the absence of an inhibitor (89 ± 14%; Fig. 13D, point IV)6. Consistent with its reversibility, removal of LY-294002 from the superfusion medium resulted in complete recovery of proton efflux to levels comparable with the sustained phase observed in cells exposed to BzATP in the absence of an inhibitor (Fig. 13C). Thus, maintenance of the sustained phase, but not its initiation, is dependent on PI3K activity.

DISCUSSION

The P2X7 receptor plays a role in the regulation of osteogenesis (29, 50) and is required for the full response of the skeleton to mechanical loading (36), but the mechanisms underlying its effects in osteoblasts are poorly understood. In this study, we examined changes in cellular metabolism triggered by activation of endogenous P2 nucleotide receptors in osteoblast-like cells. We show for the first time in any system that brief activation of P2X7 elicits a large and sustained increase in metabolic acid production that requires Ca2+ for initiation and is maintained by PI3K signaling, resulting in enhanced glucose metabolism. As well, this is the first report that, in osteoblast-like cells, P2X7 couples to sustained Ca2+ and PI3K signaling, pathways known to enhance osteoblast differentiation, leading to increased bone formation.

BzATP elicits membrane blebbing in osteoblast-like cells expressing the P2X7 receptor.

In this study, we investigated the effects of P2X7 receptor activation on osteoblast metabolism. Whereas UMR-106 cells do not express P2X7 (50), studies by others have confirmed expression of this receptor in MC3T3-E1 cells by RT-PCR (46, 55), Western blot, and pore formation assays (36). However, because pore formation has been observed for other P2X receptor subtypes, including P2X2, P2X2/3, and P2X4 (44), it cannot be used to demonstrate specifically the expression of functional P2X7 receptors. Membrane blebbing is a characteristic unique to activation of P2X7 (44). In the present study, we demonstrated that MC3T3-E1 osteoblast-like cells, but not UMR-106 osteosarcoma cells, display blebbing in response to BzATP. Previous work from our laboratory has shown that BzATP or high concentrations of ATP induce formation of plasma membrane blebs in ∼40% of cultured rat and murine calvarial cells (51), providing evidence for heterogeneous expression of functional P2X7 receptors in calvarial cell cultures. In contrast, more than 85% of MC3T3-E1 cells exhibited membrane blebbing following stimulation of P2X7, indicating that these cells represent a more homogenous population with respect to the expression of functional P2X7. Thus, the MC3T3-E1 cell line was an ideal model with which to assess signaling and metabolic pathways activated by stimulation of endogenous P2X7 receptors.

P2X7 receptors stimulate sustained proton efflux.

In the present study, we demonstrated that the P2X7 agonist BzATP (300 μM) elicits a sustained increase in metabolic acid production by osteoblast-like cells. However, because osteoblasts express multiple P2 receptor subtypes, it was possible that the sustained increase in proton efflux elicited by BzATP was not activated specifically by P2X7 receptors. In this regard, low concentrations of ATP or UTP (100 μM), which activate other P2 receptors but not P2X7, elicited only a transient increase in proton efflux that upon removal of either agonist returned to basal levels. These responses are in keeping with changes in proton efflux reported previously for ATP or UTP (10 μM) in primary astrocyte cultures (13). In contrast, a high concentration of ATP (1 mM), which is known to activate the P2X7 receptor, recapitulated the sustained increase in proton efflux elicited by BzATP. Providing further evidence for involvement of P2X7, BzATP did not elicit a sustained increase in proton efflux from UMR-106 cells, which do not express the P2X7 receptor. Moreover, we showed that the BzATP-induced response in MC3T3-E1 cells was blocked by A 438079 and A 740003, recently developed antagonists that specifically block P2X7 (14, 42).

The sustained phase of P2X7-induced proton efflux observed in this study is consistent with a net increase in metabolic acid production rather than simply activation of proton efflux pathways. Had only proton efflux mechanisms been activated, the response would have been transient, lasting only until a more alkaline steady-state cytosolic pH was established. Such transient responses are often mediated by activation of Na+/H+ exchangers, which are associated with phospholipase C-mediated mobilization of Ca2+ from intracellular stores. In contrast, sustained proton efflux with delayed onset that persists even after removal of the agonist (as observed in this study) generally involves activation of protein kinases (see below) (38).

The mechanisms underlying P2X7-induced acid production are unclear but may involve increased ATP hydrolysis or efflux (8), lipolysis, or glycogenolysis, followed by glycolysis (producing lactic acid) or oxidative phosphorylation (producing carbonic acid). In the present study, the sustained increase in proton efflux was dependent on the presence of extracellular glucose. Moreover, activation of P2X7 was associated with an increase in lactate efflux that was sustained for at least 3 h, establishing that the response was due, at least in part, to stimulation of glycolytic metabolism. These findings are in contrast to those reported previously for growth factor effects in osteoblasts, in which IGF-I-induced proton efflux was independent of extracellular glucose (60).

Dependence of P2X7-induced proton efflux on Ca2+.

Stimulation of P2X7 triggers rapid and sustained elevations of [Ca2+]i in HEK-293 cells heterologously expressing P2X7 receptors (25). Such elevations in [Ca2+]i mediate a number of important physiological functions in many cell types, including inhibition of neuritogenesis in the Neuro-2a neuroblastoma cell line (22), neuroprotection in primary cerebellar granular neurons (48), and regulation of IL-1β secretion in monocytes and macrophages (25). In the present study, we demonstrated that stimulation of osteoblasts with BzATP elicits a biphasic elevation in [Ca2+]i consisting of an initial transient and sustained phase. The initial transient increase in [Ca2+]i induced by nucleotides has been shown previously to be due to activation of P2Y receptors, leading to release of Ca2+ from intracellular stores (25, 57). Consistent with these reports, we showed that activation of P2Y receptors by UTP elicited only an initial transient increase in proton efflux in osteoblasts. That BzATP also elicited a rapid Ca2+ transient in osteoblasts is consistent with BzATP activating P2 receptors other than P2X7 in MC3T3-E1 cells. In contrast, the sustained Ca2+ elevation elicited by BzATP is mediated specifically by the P2X7 receptor and arises by influx of Ca2+ from the extracellular milieu.

The similarity in patterns of Ca2+ signaling and proton efflux induced by exogenous nucleotides suggests a role for Ca2+ in the regulation of proton production downstream of P2 receptors in osteoblasts. In this regard, we observed that extracellular Ca2+ is required to initiate but not maintain BzATP-induced proton efflux. Ca2+ influx through activated P2X7 receptors may directly stimulate proton efflux independent of cellular metabolism. For instance, Ca2+ can acidify the cytosol by displacing protons from common binding sites (4). Ca2+ also increases the rate at which Na+/H+ exchanger 1 transports protons from the cytosol to the extracellular fluid (45). However, these two mechanisms would only give rise to transient increases in proton efflux. Thus, influx of Ca2+ through the P2X7 receptor appears to modulate cellular metabolism.

Cytosolic Ca2+ is a known regulator of ATP synthesis and utilization, thereby affecting the formation of acid metabolites (lactic and carbonic acids) (5, 61). Several dehydrogenases of the citric acid cycle, including pyruvate dehydrogenase, isocitrate dehydrogenase, α-ketoglutarate dehydrogenase, succinate dehydrogenase, and malate dehydrogenase, are positively regulated by elevations in mitochondrial [Ca2+], resulting in enhanced rates of oxidative phosphorylation (11, 23). Mitochondrial [Ca2+] may also directly regulate the ATP synthase complex to increase ATP production (11). In this regard, HEK-293 and HeLa cells heterologously expressing P2X7 exhibit more mitochondria with hyperpolarized transmembrane potentials and greater basal mitochondrial Ca2+ and intracellular ATP content than nontransfected cells (1). These effects are presumably dependent on tonic stimulation of P2X7 by secreted ATP and implicate P2X7 in the positive regulation of cellular metabolism. On the other hand, Adinolfi et al. (1) provide evidence that stimulation of these heterologous P2X7 receptors by exogenous ATP causes depolarization of the mitochondrial transmembrane potential that is accompanied by a large increase in mitochondrial [Ca2+], mitochondrial fragmentation, and cell death. Interestingly, we confirmed a previous report (36) that stimulation of endogenous P2X7 receptors in MC3T3-E1 cells does not elicit cell death. This discrepancy highlights potential differences in signaling by endogenously vs. heterologously expressed P2X7 receptors.

Role of PI3K in mediating P2X7-induced proton efflux.

There are few reports concerning regulation of the PI3K/Akt pathway by P2X7 in any cell type, and opposing effects have been described. For instance, stimulation of P2X7 in non-small-cell lung cancer, HepG2 liver hepatocyte, and Panc1 human pancreatic cell lines decreases constitutive and insulin-induced total and nuclear phosphorylated Akt (39, 40). Moreover, in Neuro-2a neuroblastoma cells, inhibition or knockdown of P2X7 leads to increased Akt phosphorylation and activity (22). In contrast, BzATP acts through P2X7 receptors to stimulate PI3K-dependent Akt phosphorylation in cultures of rat astrocytes and cerebellar granular neurons (28, 37, 49). Consistent with the latter, we observed that PI3K activity is required to maintain proton efflux downstream of the P2X7 receptor, demonstrating that PI3K signaling is sustained for an extended period of time following activation of P2X7 in osteoblast-like cells.

Several mechanisms may underlie the coupling of P2X7 to activation of the PI3K/Akt pathway. In cultures of rat cortical astrocytes, activation of P2X7, as well as other P2 receptors, stimulated Akt phosphorylation in a manner dependent upon extracellular and cytosolic Ca2+, PI3K, and a Src family kinase (28, 37). The lipid kinase PI4K is one of many signaling proteins that form the P2X7 receptor complex, a complex that is in part responsible for signaling downstream of P2X7 (31). It has been suggested that phosphatidylinositol 4-phosphate (generated by PI4K following stimulation of the P2X7 receptor) may serve as a substrate for PI3K, leading to activation of Akt (28). Elevations in [Ca2+]i also appear to be critical in coupling P2X7 to activation of the PI3K/Akt pathway. In this regard, heterologously expressed P2X7 activates the proline-rich/Ca2+-activated tyrosine kinase Pyk2 (19), a kinase that associates with Src tyrosine kinases to stimulate the Ras/PI3K pathway (32). Ca2+ can also trigger Akt phosphorylation through CaMKK, independent of Pyk2, Src tyrosine kinases, or PI3K (65). Ultimately, additional studies will be needed to determine how P2X7 couples to PI3K signaling in cells of the osteoblast lineage.

The PI3K/Akt pathway regulates several cellular processes that may contribute to proton efflux; these include uptake of nutrients such as amino acids and glucose, the activity of several glycolytic enzymes, and ATP production (15, 26). For example, Akt directly increases activities of the glycolytic enzymes hexokinase and phosphofructokinase-2 in Rat1a fibroblasts and cardiac muscle, respectively. The importance of the PI3K/Akt pathway in regulation of glycolytic metabolism is in keeping with our results demonstrating that BzATP-induced increase in proton efflux is dependent on extracellular glucose and PI3K activity and arises at least in part from increased rates of lactic acid production and efflux.

Potential physiological roles of P2X7-induced Ca2+ elevation, PI3K signaling, and proton efflux in osteoblast regulation and function.

To the best of our knowledge, the present study is the first to report that P2X7 couples to sustained Ca2+ and PI3K signaling in osteoblasts. Interestingly, both the Ca2+/NFATc1 (33) and PI3K/Akt (16) pathways play important roles in osteoblast differentiation. Activation of the PI3K/Akt pathway also enhances survival in many cell types (15, 26), a phenomenon that may explain the resistance of osteoblasts to P2X7-induced apoptosis seen in several other cell types (1).

It is possible that metabolic acid production induced by P2X7 plays a role in osteoblast function. During the synthesis and secretion of osteoid by active osteoblasts in vivo, the pHo in the region between the osteoblast layer and the mineralizing front affects both the rate of mineral formation and its phase transformation. An acidic zone beneath the active osteoblast layer may prevent premature mineralization of the osteoid seam during bone formation. It is also possible that acid production by cells of the osteoblast lineage activates osteoclastic bone resorption (3, 35, 54). In vitro studies of pit formation by rat osteoclasts have shown that there is little, if any, resorptive activity at values of pHo >7.3. Slight decreases in pHo markedly stimulate osteoclastic resorption, which is maximally active at values <7.0 (3). Taken together, the effects of P2X7 on Ca2+ and PI3K signaling, as well as metabolic acid production, described in the present study may help to explain the mechanism by which P2X7 promotes osteogenesis in vivo.

GRANTS

This study was funded by the Canadian Institutes of Health Research (CIHR). M. W. Grol is supported by a CIHR Frederick Banting and Charles Best Canada Graduate Scholarship Doctoral Award.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

M.W.G. and S.J.D. did the conception and design of the research; M.W.G. and I.Z. performed the experiments; M.W.G. and I.Z. analyzed the data; M.W.G. and S.J.D. interpreted the results of the experiments; M.W.G. prepared the figures; M.W.G. drafted the manuscript; M.W.G., I.Z., and S.J.D. approved the final version of the manuscript; S.J.D. edited and revised the manuscript.

ACKNOWLEDGMENTS

We thank Elizabeth Pruski for technical assistance with Cytosensor microphysiometry and spectrofluorimetry and Kim Beaucage and Dr. Stephen M. Sims (University of Western Ontario) for constructive comments on the manuscript.

Present address of I. Zelner: Department of Pharmacology & Toxicology, University of Toronto, Toronto, ON, Canada, M5S 1A8.

Footnotes

  • 1 The transient depression in proton efflux upon exposure to BzATP and subsequent overshoot in proton efflux upon washout of BzATP are due to the presence of triethylammonium in the BzATP preparation. Upon superfusion with BzATP solution, influx of unprotonated triethylammonium transiently increases pHi, leading to brief suppression of proton efflux. On the other hand, upon washout, efflux of unprotonated triethylammonium transiently decreases pHi, leading to a transitory increase in proton efflux (Reyes JP, Grol MW, Sims SM, and Dixon SJ, unpublished data).

  • 2 The transient depression and overshoot in proton efflux observed with the triethylammonium+ salt of BzATP were not observed with ATP, which was applied as a Na+ salt.

  • 3 When tested in UMR-106 cells, neither BzATP (300 μM) nor ATP (1 mM) induced a significant increase in extracellular lactate concentration relative to vehicle-treated cultures (data not shown), in keeping with a role for the P2X7 receptor in mediating this response.

  • 4 Since wortmannin suppressed basal proton efflux compared with DMSO alone, data were also corrected for the effects of wortmannin on baseline at the sustained phase. Even with values corrected, the PI3K inhibitor still significantly blocked the sustained phase of the BzATP response (69 ± 16%) compared with parallel cultures treated with BzATP in the absence of an inhibitor (104 ± 20%).

  • 5 Since LY-294002 suppressed basal proton efflux compared with DMSO alone, data were also corrected for the effects of LY-294002 on baseline. Even with values corrected, the PI3K inhibitor still significantly blocked the sustained phase of the BzATP response (50 ± 12%) compared with parallel cultures treated with BzATP in absence of an inhibitor (90 ± 17%).

  • 6 Since LY-294002 suppressed basal proton efflux compared with DMSO alone (not shown), data were corrected for the effects of LY-294002 on baseline. With values corrected, LY-294002 still significantly blocked the sustained phase of the BzATP response (37 ± 6%) compared with parallel cultures treated with BzATP alone (105 ± 21%).

REFERENCES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
View Abstract