Megakaryocytes promote osteoclastogenesis in aging

Abstract

Megakaryocytes (MKs) support bone formation by stimulating osteoblasts (OBs) and inhibiting osteoclasts (OCs). Aging results in higher bone resorption, leading to bone loss. Whereas previous studies showed the effects of aging on MK-mediated bone formation, the effects of aging on MK-mediated OC formation is poorly understood. Here we examined the effect of thrombopoietin (TPO) and MK-derived conditioned media (CM) from young (3-4 months) and aged (22-25 months) mice on OC precursors. Our findings showed that aging significantly increased OC formation in vitro. Moreover, the expression of the TPO receptor, Mpl, and circulating TPO levels were elevated in the bone marrow cavity. We previously showed that MKs from young mice secrete factors that inhibit OC differentiation. However, rather than inhibiting OC development, we found that MKs from aged mice promote OC formation. Interestingly, these age-related changes in MK functionality were only observed using female MKs, potentially implicating the sex steroid, estrogen, in signaling. Further, RANKL expression was highly elevated in aged MKs suggesting MK-derived RANKL signaling may promote osteoclastogenesis in aging. Taken together, these data suggest that modulation in TPO-Mpl expression in bone marrow and age-related changes in the MK secretome promote osteoclastogenesis to impact skeletal aging.

Keywords: aging; bone marrow macrophage; megakaryocyte; osteoclast; thrombopoietin.

Figure 1

Aging increases OC formation from sorted CD45+F4/80+ bone marrow macrophages. Similar numbers of sorted CD45+F4/80+ macrophages from young and aged C57BL/6J mice were cultured in the presence of RANKL (80 ng/ml) and M-CSF (20 ng/ml). Growth media was changed every second day for 5-7 days until mature, multinucleated OCs (>3 nuclei) were formed. (A) Images of TRAP+ OCs generated. Scale bar is 100 μm. (B) The number of TRAP+ OCs formed in each well was quantified, showing that more OCs were generated from aged CD45+F4/80+ macrophages compared to young macrophages. Four independent experiments gave similar results, and a representative experiment is shown. Graphs are mean ± SD with ***p<0.001 (N=4/group). (C) Frequency of CD45+F4/80+ macrophage (%) in bone marrow cells isolated from young and aged C57BL/6J mice (males and females combined). Data are mean ± SD of four independent experiments (p=0.08, ns indicates non-significant).

Figure 2

Aging increases bone marrow MK number. (A) Flow cytometry was used to determine the total number of CD45+CD41+ MKs in the bone marrow of young and aged C57BL/6J female and male mice. The data are presented as mean ± SD of four independent experiments with *p<0.05 (N=4/group). (B) Bone marrow CD45+CD41+ MKs were isolated by FACS from young and aged mice (males and females were combined) and used for real-time PCR analysis of relative Mpl mRNA expression between groups. The Ct cycle range was 28.99–33.83. The data are presented as mean ± SD of three independent experiment (****p<0.0001; N=3-6/group).

Figure 3

Changes in TPO levels in young and aged mice. (A) TPO concentration was measured by ELISA in serum from young and aged, male and female C57BL/6J mice. Serum TPO was significantly reduced in aged female and male mice compared to sex-matched young mice. The data are presented as mean ± SD (*p<0.05, **p<0.01, N=9-10/group). (B) Bone marrow supernatant was collected from young and aged femur and tibia (male and female mice were combined) and used to determine TPO concentrations by ELISA. The data are presented as mean ± SD (*p<0.05; N=5/group).

Figure 4

TPO stimulates OC formation. (A) FACS sorted CD45+F4/80+ macrophages from young and aged mice were subject to real-time PCR to assess the relative mRNA expression of Mpl. The Ct cycle range was 34.06–35.15. The data shown are mean ± SD (*p<0.05). (B, C) FACS sorted CD45+F4/80+ macrophages were cultured with 20 ng/ml M-CSF and 80 ng/ml RANKL in the presence or absence of recombinant human TPO (100 ng/ml). Growth media was changed every second day for 5–7 days until mature, multinucleated OCs (>3 nuclei) were formed. Mature OCs were fixed and TRAP+ OCs were counted. The data are presented as mean ± SD of four (female) and two (male) independent experiments (*p<0.05, ***p<0.001; N=3/group).

Figure 5

Effect of MK conditioned media on OC formation. (A–E) Conditioned medium (CM) was prepared from MKs isolated from young and aged, female and male mice. (A–C) Unsorted BMMs from mice (6-10 weeks) were cultured in media containing 25% (v/v) MK CM plus RANKL (80 ng/ml) and M-CSF (20 ng/ml). The control media was prepared without MKs. (A) Micrographs showing mature OCs. Scale bar is 50 μm. (B, C) The number of TRAP+ multinucleated OC (>3 nuclei) was quantified. The data show that MK CM from young female mice inhibits OC formation compared to control and aged female MK CM. The data are presented as mean ± SD of three experiments (****p<0.0001 compared to the control group). No significant difference in OC formation was observed between male groups. (D, E) Unsorted BMMs from mice (6-10 weeks) were differentiated into mature OCs in media containing 25% MK CM plus normal or reduced RANKL (80 or 26.6 ng/ml, respectively) and M-CSF (6.6 or 20 ng/ml, respectively). Mature OCs were fixed and TRAP+ OCs were counted. (F) Comparison of OCs formed in control media (no MK CM) with normal or reduced RANKL and M-CSF. Three independent experiments showed similar results and a representative figure is shown. The data are presented as mean ± SD (**p<0.01, ***p<0.001, ****p<0.0001; N= 3/group). P-values were calculated by one-way ANOVA followed by Tukey’s post-hoc test.

Figure 6

Expression of key OC factors in MKs from young and aged mice. (A–D) MKs were prepared from the bone marrow of young and aged, male and female mice using a BSA gradient and used for mRNA expression analyses by real-time PCR. (A) M-CSF mRNA (Ct cycle range was 25.2–30.12). (B) OPG mRNA (Ct range 35.09–39.53). (C) RANKL mRNA (Ct range 28.6–35.5). (D) The RANKL/OPG ratio was calculated for young and aged, male and female MKs (N=4-5/group). (E) RANKL concentrations in MK CM was quantified by ELISA. MK CM from aged females was significantly higher than MK CM from young females. P-values were calculated by one-way ANOVA followed by Tukey’s post-hoc test and the data are presented as mean ± SD. (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001).

Figure 7

Model showing OC stimulation by TPO and MKs in aging. The schema illustrates that aging increases MK number and alters MK-secreted factors, including RANKL, which together promote osteoclastogenesis. In addition, elevated TPO in the bone marrow cavity promotes direct and MK-mediated effects on OC formation in aged mice.