Adenosine A1 receptors (A1Rs) play a critical role in osteoclast formation and function

Abstract

Adenosine regulates a wide variety of physiological processes via interaction with one or more G-protein-coupled receptors (A(1)R, A(2A)R, A(2B)R, and A(3)R). Because A(1)R occupancy promotes fusion of human monocytes to form giant cells in vitro, we determined whether A(1)R occupancy similarly promotes osteoclast function and formation. Bone marrow cells (BMCs) were harvested from C57Bl/6 female mice or A(1)R-knockout mice and their wild-type (WT) littermates and differentiated into osteoclasts in the presence of colony stimulating factor-1 and receptor activator of NF-kappaB ligand in the presence or absence of the A(1)R antagonist 1,3-dipropyl-8-cyclopentyl xanthine (DPCPX). Osteoclast morphology was analyzed in tartrate-resistant acid phosphatase or F-actin-stained samples, and bone resorption was evaluated by toluidine blue staining of dentin. BMCs from A(1)R-knockout mice form fewer osteoclasts than BMCs from WT mice, and the A(1)R antagonist DPCPX inhibits osteoclast formation (IC(50)=1 nM), with altered morphology and reduced ability to resorb bone. A(1)R blockade increased ubiquitination and degradation of TRAF6 in RAW264.7 cells induced to differentiate into osteoclasts. These studies suggest a critical role for adenosine in bone homeostasis via interaction with adenosine A(1)R and further suggest that A(1)R may be a novel pharmacologic target to prevent the bone loss associated with inflammatory diseases and menopause.

Figure 1.

mRNA for all 4 adenosine receptor subtypes is expressed in osteoclasts and their precursors. A) Total RNA was isolated from murine BMCs, splenocytes, and RAW264.7 cells, reversed transcribed, and subjected to RT-PCR, as described in Materials and Methods. Message for all 4 adenosine receptors was detected in all cell types studied. B) Adenosine A₁R expression in WT bone marrow-derived macrophages induced to differentiate into osteoclasts was examined by RT-PCR at indicated time points. GAPDH is shown as a loading control.

Figure 2.

A₁KO BMCs generate fewer, morphologically altered osteoclasts in vitro. A) BMCs were induced to differentiate into osteoclasts in the presence of CSF-1 and RANKL for 6 d, fixed, and stained for TRAP. Osteoclastic cells from A₁KO mice are less intensely TRAP⁺ and much smaller and contain fewer nuclei than control osteoclasts. B) Quantitative evaluation of osteoclast differentiation in A₁KO and WT BMC cultures.

Figure 3.

A₁R antagonist DPCPX inhibits osteoclast development. BMCs from WT mice were induced to differentiate to osteoclasts by culture in CSF-1 and RANKL in the presence or absence of various concentrations of DPCPX. After 7 d, cells were fixed and stained for TRAP, and the number of MNCs per well was quantitated. A) DPCPX (1 mM) inhibits CSF-1 and RANKL-induced formation of TRAP-positive MNCs; representative photomicrographs (×200) of cells cultured for 7 d in the presence of RANKL and DPCPX. B) DPCPX inhibits osteoclast formation in a dose-dependent fashion. Average ± sem value of TRAP + MNCs/control well was 248 ± 24. Each point represents results of ≥3 separate determinations carried out in triplicate with BMCs from different mice. *P < 0.05; ANOVA. C) Osteoclast precursors were induced to differentiate, and DPCPX was added to the culture on d 0, 3, and 6 after addition of RANKL to cultures. All cultures were fixed and stained after 7 d of incubation in the presence of RANKL. Each point represents 3 determinations carried out with BMCs from different mice in triplicate.

Figure 4.

Abnormal phenotype of osteoclasts formed from A₁KO mouse bone marrow precursors. BMCs were cultured on glass coverslips and treated with CSF-1 and RANKL with or without DPCPX (1 mM), as described in Materials and Methods. F-actin was detected by Alexa 555-Phalloidin staining, and nuclei were labeled with DAPI. A) Morphology of the least mature (arrow indicates osteoclast), maturing (arrow indicates cytoplasmic bridge), and mature osteoclasts cultured on glass. B) Quantitative evaluation of number of least mature, maturing, and mature osteoclasts in WT and A₁KO cultures. C) Quantitative evaluation of the number of least mature, maturing, and mature osteoclasts in BMC cultures treated or not with DPCPX (1 mM). Data represent mean ± sd proportion of different types of osteoclasts. *P < 0.05 vs. WT, **P < 0.01 vs. DPCPX; ANOVA.

Figure 5.

Osteoclasts from A₁KO mice cultured on dentin form abnormal actin rings and resorb less bone. A) Osteoclast formation on dentin coverslips (left panels) and bone resorption (center panels) by WT and A₁KO-deficient cells were visualized by TRAP and toluidine blue staining, respectively. Morphology of the actin cytoskeleton was evaluated in Alexa 555-Phalloidin-stained samples (right panels). B) Cross-sections through 3-dimensional projections of WT and A₁KO osteoclasts grown on dentin coverslips and labeled with Alexa 555-Phalloidin were visualized by confocal microscopy. Arrows indicate borders of the sealing zone. C) Osteoclast height was determined by confocal microscopy as mean ± sd distance from dentin surface to apex of dorsal membrane of osteoclasts (n=12 cells/genotype, samples from 3 mice). *P < 0.05; Student’s t test.

Figure 6.

Effect of A₁R blockade on intracellular signaling in osteoclasts. RAW 264.7 cells were incubated for 3 d, washed, and stimulated with RANKL for 30 min in the presence or absence of DPCPX, as described. Cells were lysed, and proteins were separated by SDS-PAGE and transferred to Immobilon membranes for blotting with anti-TRAF6 antibody. Blots were stripped and blotted with β-actin antibody (loading control). Representative experiment from 6 discrete experiments is shown.