The active form of vitamin D, 1α,25-(OH)2D3, not only promotes intestinal calcium absorption, but also regulates the formation of osteoclasts (OCs) and their capacity for bone mineral dissolution. Gal-3 is a newly discovered bone metabolic regulator involved in the proliferation, differentiation, and apoptosis of various cells.
1. Introduction
OCs are derived from bone marrow mononuclear macrophages (BMMs) and are the only cells capable of bone resorption in the body. They not only degrade the bone organic and inorganic matrix but also cooperate with osteoblasts (OBs) to regulate bone formation and reconstruction
[1][2][3]. Overactive bone resorption by OCs in physiological processes (such as aging and menopause)
[4] and pathological processes (such as bone metastasis and rheumatoid arthritis)
[5][6] can lead to osteoporosis. OCPs, such as bone marrow cells, splenocytes, and RAW264.7 macrophages, co-cultured with stromal cells, OBs, or osteocytes in vitro can be induced into OCs by the parathyroid hormone (PTH), dexamethasone, tumor necrosis factor-alpha (TNF-α), interleukin-1 beta (IL-1β), and 1α,25-(OH)
2D
3, which regulate the expression of membrane-bound RANKL in OBs, stromal cells, and osteocytes
[1][7]. BMMs could also be induced into OCs directly by M-CSF and RANKL
[8]. M-CSF mainly promotes the proliferation and differentiation of OCPs, which further differentiate into OCs under the action of RANKL
[9][10].
The physiologically active form of vitamin D, 1α,25-(OH)
2D
3, regulates intestinal calcium absorption and acts on bone cells directly and a series of cytokines or signaling pathways in bone
[11][12][13]. OB-lineage cells express vitamin D receptor (VDR)
[14], and 1α,25-(OH)
2D
3 promotes OBs’ maturation and bone mineralization in vitro and in vivo via VDR and reduces the formation of unmineralized osteoids
[15]. However, bone tissue is in a state of dynamic equilibrium, and over-mineralization or absorption is not conducive to bone health. To prevent excessive bone mineralization, 1α,25-(OH
)2D
3 can enhance OC formation indirectly by promoting the expression of RANKL in a concentration-dependent manner
[15][16]. OC formation can also be directly regulated by 1α,25-(OH)
2D
3. Although mature OCs do not express VDR, OCPs do
[17][18][19]. The mechanism by which it regulates OC formation needs to be further clarified.
Gal-3 is a 29–35 kDa protein expressed in a variety of tissues and is a member of the β-galactosyl-binding protein family
[20]. It is a marker of chondrocyte and OB lineages in bone and is also present in OCs and BMMs
[21][22]. In proteomic studies, we found that 1α,25-(OH)
2D
3 activates gal-3 expression during OC formation in vitro
[23][24]. Simon et al. have also shown that gal-3 is a novel regulator of bone homeostasis directly or indirectly by regulating the association between OBs and OCs. Accordingly, gal-3 plays an important role in bone biology and is expected to be a potential target for the prevention of bone diseases. However, its role in the regulation of OC formation by 1α,25-(OH)
2D
3 needs to be further elucidated.
2. 1α,25-(OH)2D3 Had No Effect on Osteoclast Precursor Viability
We confirmed that 1α,25-(OH)
2D
3 upregulates VDR mRNA and protein expression in OCPs
[25]. In this study, we observed that adding 0.1, 1, and 10 nmol/L 1α,25-(OH)
2D
3 to the medium had no effect on OCPs’ viability (
Figure 1A). RANKL significantly inhibited cell proliferation during OCs formation (
p < 0.01). However, 1α,25-(OH)
2D
3 had no significant effect on OCPs’ viability in the absence or presence of RANKL (
Figure 1B).
Figure 1. OCP viability was not affected by 1α,25-(OH)2D3. (A) Cell viability detected by CCK-8 24 h after treatment with different concentrations of 1α,25-(OH)2D3. (B) Cell viability detected by CCK-8 24 h after treatment with 10 nmol/L 1α,25-(OH)2D3 in the absence or presence of 50 ng/mL RANKL. Data are shown as means ± SD. n = 6, ns, p > 0.05; ** p < 0.01.
3. 1α,25-(OH)2D3 Promoted Gal-3 Expression
To elucidate the effect of 1α,25-(OH)2D3 on gal-3 protein expression, 0.1, 1, and 10 nmol/L 1α,25-(OH)2D3 were added to the culture medium during OC formation induced by 25 ng/mL M-CSF and 50 ng/mL RANKL for 3 days. 1α,25-(OH)2D3 upregulated gal-3 protein expression in a dose-dependent manner (Figure 2). The 10 nmol/L 1α,25-(OH)2D3 group had a higher level of gal-3 protein expression than that in the other groups.
Figure 2. The expression of gal-3 protein was upregulated by 1α,25-(OH)2D3 dose-dependently on day 3 as determined by Western blotting. Histograms show gray values of gal-3 protein. Data are shown as means ± SD. n = 5, * p < 0.05, ** p < 0.01.
To confirm the effect of 1α,25-(OH)2D3 on gal-3 protein expression at different time points, OCPs induced by 25 ng/mL M-CSF and 50 ng/mL RANKL were treated with 10 nmol/L 1α,25-(OH)2D3 for 0, 1, 3, and 5 days. Anhydrous ethanol was used as a control. Compared with the level in the control group, 1α,25-(OH)2D3 significantly increased gal-3 protein expression on days 3 and 5 (p < 0.01) (Figure 3). No significant difference was observed between the control group and the 1α,25-(OH)2D3 group on day 1 (p > 0.05). Compared with day 1, 10 nmol/L 1α,25-(OH)2D3 significantly increased gal-3 protein expression on days 3 and 5 (p < 0.01). However, there was no significant change between days 3 and 5 1α,25-(OH)2D3 groups (p > 0.05). These data indicated that 1α,25-(OH)2D3 promoted the protein expression of gal-3 at the same cultivation time.
Figure 3. Gal-3 protein expression was upregulated by 1α,25-(OH)2D3 at the same cultivation time. Histograms show relative expression levels of gal-3 protein. Data are shown as means ± SD. n = 3, ** p < 0.01 vs. the 0 d group; $$ p < 0.01 vs. the 1α,25-(OH)2D3 treatment group on day 1; ## p < 0.01 vs. different groups on the same day.
To further confirm the effect of 1α,25-(OH)2D3 on gal-3 protein distribution, an immunofluorescence assay was performed. The gal-3 protein was visualized by green fluorescence, while F-actin was visualized in red on day 6 after treatment with 10 nmol/L 1α,25-(OH)2D3. Anhydrous ethanol was used as a control. Gal-3 mainly distributed in the nuclei (cyan) and cell membranes (yellow) of OCs (large cells with more than three nuclei marked by white triangles) and in the whole OCPs (small cells with one nucleus marked by white arrow) (Figure 4). Compared with OCs, OCPs had a wider green fluorescence distribution of gal-3. These data confirmed that 1α,25-(OH)2D3 changed the protein distribution of gal-3.
Figure 4. Immunofluorescence showed the distribution and expression of gal-3 after the treatment of 1α,25-(OH)2D3. Gal-3 mainly distributed in the nuclei (cyan) and cell membranes (yellow) of OCs (marked by white triangle) and in the whole cell of OCPs (white arrows). Gal-3 protein distribution was regulated by 1α,25-(OH)2D3. In non-merged images, green indicates gal-3, red indicates F-actin, and blue indicates nuclei. Bars = 25 μm.
To elucidate the effect of 1α,25-(OH)2D3 on the expression of Lgals3, which encodes the gal-3 protein, 10 nmol/L 1α,25-(OH)2D3 was added to the culture medium during OC formation induced by 25 ng/mL M-CSF and 50 ng/mL RANKL for 0, 1, 3, and 5 days. Anhydrous ethanol was used as a control. The expression of Lgals3 first increased and then decreased over time. Compared with control groups (without 1α,25-(OH)2D3), 10 nmol/L 1α,25-(OH)2D3 significantly increased Lgals3 expression on days 3 and 5 (p < 0.01). Compared with day 1, 10 nmol/L 1α,25-(OH)2D3 significantly increased Lgals3 expression on days 3 and 5 (p < 0.01). However, Lgals3 expression on day 5 was lower than day 3 in the groups with or without 1α,25-(OH)2D3 (Figure 5).
Figure 5. Lgals3 expression is upregulated by 1α,25-(OH)2D3. Data are shown as means ± SD. n = 3, ** p < 0.01 vs. the 0 d group; $$ p < 0.01 vs. the 1α,25-(OH)2D3 treatment group on day 1; ## p < 0.01 vs. different groups on the same day; ¥¥ p < 0.01 vs. the 1α,25-(OH)2D3 treatment group on day 3.
4. Gal-3 Contributed to Osteoclasts Formation and Activation Regulated by 1α,25-(OH)2D3
We found that 1α,25-(OH)2D3 increased gal-3 expression at the mRNA and protein levels. To confirm the role of gal-3 in 1α,25-(OH)2D3-mediated OC formation and bone resorption, we constructed stable Lgals3 knockdown OCPs using gal-3 siRNA. Negative control (NC) siRNA was used as the control. These OCPs were treated with 10 nM 1α,25-(OH)2D3 in the presence of 25 ng/mL M-CSF and 50 ng/mL RANKL. Anhydrous ethanol was used as a control.
First, OC formation was detected by TRAP staining on day 6 after the treatment of 1α,25-(OH)2D3. In all groups, large cells with wine-red granules regarded as OCs were found. In the NC group, the volume of OCs treated with 1α,25-(OH)2D3 and the number and the size of OCs decreased significantly (p < 0.01). In the gal-3 knockdown group, 1α,25-(OH)2D3 had no significant effect on OC formation, but significantly decreased the size of OCs. These data confirmed that gal-3 contributes to the regulation of OC formation by 1α,25-(OH)2D3. Additionally, gal-3 knockdown significantly promoted OC formation and average size (p < 0.01) (Figure 6A–C). This suggests that gal-3 is a negative regulator of OC formation and average size.
Figure 6. Gal-3 knockdown attenuated the inhibitory effect of 1α,25-(OH)2D3 on OC formation. (A) TRAP staining. Large multinuclear cells (MNCs) with wine-red granules were regarded as OCs (yellow arrows). Bars = 400 μm. (B) Quantitative analysis of OC quantity. (C) Quantitative analysis of OC size. Data are shown as means ± SD. n = 3, ** p < 0.01, ns means p > 0.05.
Gal-3 and OC-related proteins (NFATc1 and MMP-9) were investigated by Western blotting on day 3 after treatment with 1α,25-(OH)2D3. Compared to the NC group, cells with Lgals3 knockdown exhibited a significant decrease in gal-3 protein level (p < 0.01) (Figure 7). In the NC group, the expression of NFATc1 and MMP-9 proteins were significantly inhibited by 1α,25-(OH)2D3 (p < 0.01). In gal-3 knockdown groups, 1α,25-(OH)2D3 had no significant effect on the expression of NFATc1 and MMP-9 proteins. These data confirmed that gal-3 contributed to OC-related protein expression regulated by 1α,25-(OH)2D3. Gal-3 knockdown significantly increased OC-related protein expression levels (p < 0.01) (Figure 7). These findings further suggest that gal-3 is a negative regulator of OC formation.
Figure 7. Gal-3 knockdown attenuated the inhibitory effect of 1α,25-(OH)2D3 on OC-related protein expression. The histograms show the relative expression level of proteins. Data are shown as means ± SD. n =3, ** p < 0.01, ns means p > 0.05.
mRNA expression levels of OC-related genes, Ctsk, and Mmp-9 were evaluated by qPCR on day 3 after treatment with 1α,25-(OH)2D3. In the NC group, Ctsk and MMP-9 levels were significantly inhibited by 1α,25-(OH)2D3 (p < 0.01). In the gal-3 knockdown groups, 1α,25-(OH)2D3 also inhibited Ctsk and MMP-9 expression. However, compared to levels in the NC groups, the inhibitory effects of 1α,25-(OH)2D3 on Ctsk and MMP-9 were significantly attenuated by gal-3 knockdown (p < 0.01). Additionally, gal-3 knockdown significantly increased OC-related gene expression levels (p< 0.01) (Figure 8). These findings were consistent with TRAP-positive OC formation and OC-related protein expression results.
Figure 8. Gal-3 knockdown attenuated the inhibitory effect of 1α,25-(OH)2D3 on OC-related gene expression. Histograms show relative expression levels of genes. Data are shown as means ± SD. n = 3, ** p < 0.01.
To evaluate the effects of gal-3 on bone resorption regulated by 1α,25-(OH)2D3, equal number of BMMs were cultured on an osteoassay surface multiple-well plate for each group. Bone resorption lacunae were observed using an inverted microscope on day 6 after the treatment with 1α,25-(OH)2D3. We observed bone resorption lacunae in each group (Figure 9A, black arrow). Based on the area of bone resorption lacunae, in the NC group, bone resorption was significantly inhibited by 1α,25-(OH)2D3 (p < 0.01). In the gal-3 knockdown groups, 1α,25-(OH)2D3 had no effect on bone resorption activity. Gal-3 knockdown significantly attenuated the inhibitory effect of 1α,25-(OH)2D3 on bone resorption (p < 0.01). Additionally, gal-3 knockdown significantly increased OC bone resorption (p < 0.01) (Figure 9B). These data confirmed that gal-3 is a negative regulator of OC bone resorption and contributes to the inhibitory effect of 1α,25-(OH)2D3 on OC bone resorption.
Figure 9. Gal-3 knockdown attenuated the inhibitory effect on bone resorption by 1α,25-(OH)2D3. (A) Bone resorption lacunae (marked by black arrows) observed by inverted microscopy. Bars = 400 μm. (B) Statistical analysis of the area of bone resorption lacunae. Data are shown as means ± SD. n = 3, ** p < 0.01, ns means p > 0.05.
5. Interaction between Gal-3 and VDR
To verify the relationship between gal-3 and VDR proteins, they were evaluated by co-immunoprecipitation and immunofluorescence double staining. The expression of VDR and gal-3 could be detected in the input group (Figure 10A). The expression of VDR and gal-3 was also detected in the protein samples precipitated by the anti-VDR antibody (Figure 10A). These results suggest that there is an interaction between gal-3 and the VDR proteins.
Figure 10. Images showing co-localization and possible interactions between gal-3 and VDR proteins. (A) Interaction between gal-3 and VDR proteins confirmed by co-immunoprecipitation. (B,C) Co-localization of gal-3 and VDR proteins detected by immunofluorescence. In non-merged images, gal-3 is green, VDR is red, and nuclei are blue. Bars = 10 μm.
The expression of gal-3 (green) and VDR (red) protein was also observed by confocal fluorescence microscopy. OCPs (small cells marked by white arrows in Figure 10B) showed high expression levels of gal-3 and VDR, while gal-3 and VDR expression levels were low in OCs (large cells with multiple nuclei, marked by white triangles). Gal-3 and VDR proteins were mainly co-localized (yellow) in the cell membrane (Figure 10B). The red and green curves change in the same way, which suggests that gal-3 and VDR are co-located. (Figure 10C). These results further supported the co-localization and possible interaction between gal-3 and VDR.