Mode of action of interleukin-6 on mature osteoclasts. Novel interactions with extracellular Ca2+ sensing in the regulation of osteoclastic bone resorption

Abstract

We describe a physiologically significant mechanism through which interleukin-6 (IL-6) and a rising ambient Ca2+ interact to regulate osteoclastic bone resorption. VOXEL-based confocal microscopy of nonpermeabilized osteoclasts incubated with anti- IL-6 receptor antibodies revealed intense, strictly peripheral plasma membrane fluorescence. IL-6 receptor expression in single osteoclasts was confirmed by in situ reverse transcriptase PCR histochemistry. IL-6 (5 ng/l to 10 microg/l), but not IL-11 (10 and 100 microg/l), reversed the inhibition of osteoclastic bone resorption induced by high extracellular Ca2+ (15 mM). The IL-6 effect was abrogated by excess soluble IL-6 receptor (500 microg/l). Additionally, IL-6 (5 pg/l to 10 microg/l) inhibited cytosolic Ca2+ signals triggered by high Ca2+ or Ni2+. In separate experiments, osteoclasts incubated in 10 mM Ca2+ or on bone released more IL-6 than those in 1.25 mM Ca2+. Furthermore, IL-6 mRNA histostaining was more intense in osteoclasts in 10 or 20 mM Ca2+ than cells in 1.25 mM Ca2+. Similarly, IL-6 receptor mRNA histostaining was increased in osteoclasts incubated in 5 or 10 mM Ca2+. Thus, while high Ca2+ enhances IL-6 secretion, the released IL-6 attenuates Ca2+ sensing and reverses inhibition of resorption by Ca2+. Such an autocrine-paracrine loop may sustain osteoclastic activity in the face of an inhibitory Ca2+ level generated locally during resorption.

Figure 1

Confocal microscopic localization of the IL-6R on rat osteoclast plasma membrane. (a and c) Intense peripheral immunofluorescent staining of osteoclasts with two highly specific anti– IL-6R monoclonal antibodies, MT-18 (a; serial sections in coronal plane, 1 μm thick) and 15A7 (c; single coronal section). Note that osteoclasts incubated with no antiserum, nonimmune mouse serum, or an irrelevant antibody (Ab³⁴) did not stain (not shown). (b) Very faint staining of an osteoclast with MT-18 coincubated with soluble rhIL-6R (500 μg/l); this confirms IL-6R specificity of the antibody. (d) Lack of appreciable staining of IL-6R– negative MBA13.2 cells incubated with antibody, MT-18. For details on confocal microscopy, see Materials and Methods. (e and f) VOXEL (volume element) imaging of osteoclasts stained with the anti–IL-6R antibody, MT-18. This method allows three-dimensional optical sectioning in both the coronal (e) and sagittal (f) planes of the cell. Staining is typically localized to the cell's periphery with a sparing of the central cytoplasmic and nuclear areas. (g) A three-dimensional reconstruction of membrane staining from the coronal and sagittal sections.

Figure 2

(a) Bone resorption by isolated rat osteoclasts, expressed as the mean pit number per bone slice (±SEM), in the presence of different concentrations of IL-6 and CaCl₂ (Ca^{2 +}), as shown. Statistics by ANOVA with Bonferroni's correction for inequality. Compared with control: ^a P = 0.205, ^b P = 0.126, ^c P = 0.418, and ^d P = 0.007; compared with 15 mM Ca²⁺: ^e P = 0.002, ^f P = 0.026, and ^g P = 0.05. (b) Bone resorption by isolated rat osteoclasts, expressed as the mean pit number per bone slice (±SEM), in the presence of different concentrations of IL-11 and CaCl₂ (Ca^{2 +}), as shown. Statistics by ANOVA with Bonferroni's correction for inequality. Compared with control: ^a P = 0.621, ^b P = 0.317, and ^c–e P < 0.001. (c) Bone resorption by isolated rat osteoclasts, expressed as the mean pit number per bone slice (±SEM) in the presence of different concentrations of CaCl₂ (Ca^{2 +}), IL-6, and/or the soluble rhIL-6R, as shown. Statistics by ANOVA with Bonferroni's correction for inequality. Compared with control: ^a P = 0.847, ^b P = 0.536, ^c,e,f P < 0.001, and ^d P = 0.675.

Figure 3

Representative traces showing the effect of IL-6 on cytosolic Ca²⁺ levels elicited in fura-2–loaded rat osteoclasts in response to 10 mM Ca²⁺ applied extracellularly with either vehicle or 0.5, 5, or 10 μg/l IL-6 (n = 4–6 cells for each variable). For statistical analysis, see Table I.

Figure 4

Representative traces showing the effect of preincubation with either IL-6 (5 pg/l to 10 μg/l) or IL-11 (100 μg/l) on cytosolic Ca²⁺ levels in fura-2–loaded rat osteoclasts elicited in response to the application of 5 mM Ni²⁺ extracellularly (n = 4–6 cells for each variable). For statistical analysis, see Table II.

Figure 5

(a) Cumulative supernatant levels of IL-6 (μg/l) after incubation of disaggregated osteoclast cultures with either 1.25 mM Ca²⁺ (squares) or 10 mM Ca²⁺ (circles) (on glass coverslips) or on bone slices (Ca²⁺ = 1.25 mM) (triangles). Statistics by ANOVA and Bonferroni's correction for inequality: ^a,b,c P = 0.001, ^d P = 0.05, ^e P = 0.0021, ^f P = 0.47, ^g P = 0.05, ^h P = 0.217, ⁱ P = 0.13, ^j P = 0.064, ^k P = 0.005, and ^l P = 0.683, all compared with the corresponding time points for the 1.25 mM Ca²⁺ (on glass) variable. (b–d) IL-6 release rates (ng l⁻¹ min⁻¹) after incubation of disaggregated osteoclast cultures with either 1.25 mM Ca²⁺ (b, squares) or 10 mM Ca²⁺ (c, circles) (on glass coverslips) or on bone slices (Ca²⁺ = 1.25 mM) (d, triangles). Statistics by ANOVA and Bonferroni's correction for inequality: ^a P = 0.05, ^b P = 0.028, ^c P = 0.006, ^d P = 0.740, ^e P = 0.001, ^f P = 0.849, ^g P = 0.087, and ^h P = 0.092. Note that other contaminating cells in our cultures may contribute to the measured IL-6.

Figure 6

Histostained osteoclasts after in situ RT-PCR for detection of either IL-6 mRNA, IL-6R receptor mRNA, or GA3PD mRNA. The left panels represent controls, i.e., without added primer. For details and primer sequences, refer to Materials and Methods.

Figure 7

Semiquantitative representation in frequency histograms of corrected staining intensity after in situ RT-PCR for IL-6 mRNA (a) or IL-6R mRNA (b) for osteoclasts incubated with 1.25 mM Ca²⁺ (solid black bars), 5 mM Ca²⁺ (dotted bars), or 10 mM Ca²⁺ (diagonal bars).