Anabolic and Antiresorptive Modulation of Bone Homeostasis by the Epigenetic Modulator Sulforaphane, a Naturally Occurring Isothiocyanate

Abstract

Bone degenerative pathologies like osteoporosis may be initiated by age-related shifts in anabolic and catabolic responses that control bone homeostasis. Here we show that sulforaphane (SFN), a naturally occurring isothiocyanate, promotes osteoblast differentiation by epigenetic mechanisms. SFN enhances active DNA demethylation viaTet1andTet2and promotes preosteoblast differentiation by enhancing extracellular matrix mineralization and the expression of osteoblastic markers (Runx2,Col1a1,Bglap2,Sp7,Atf4, andAlpl). SFN decreases the expression of the osteoclast activator receptor activator of nuclear factor-κB ligand (RANKL) in osteocytes and mouse calvarial explants and preferentially induces apoptosis in preosteoclastic cells via up-regulation of theTet1/Fas/Caspase 8 and Caspase 3/7 pathway. These mechanistic effects correlate with higher bone volume (∼20%) in both normal and ovariectomized mice treated with SFN for 5 weeks compared with untreated mice as determined by microcomputed tomography. This effect is due to a higher trabecular number in these mice. Importantly, no shifts in mineral density distribution are observed upon SFN treatment as measured by quantitative backscattered electron imaging. Our data indicate that the food-derived compound SFN epigenetically stimulates osteoblast activity and diminishes osteoclast bone resorption, shifting the balance of bone homeostasis and favoring bone acquisition and/or mitigation of bone resorptionin vivo Thus, SFN is a member of a new class of epigenetic compounds that could be considered for novel strategies to counteract osteoporosis.

Keywords: bone; epigenetics; osteoblast; osteoclast; osteoporosis.

FIGURE 1.

DMSO and SFN show structural similarities and analogous biological effects. DMSO and SFN contain a polar sulfoxide functional group. SFN carries an additional butane group with a terminal isothiocyanate group (A). DMSO and SFN increase matrix mineralization in MC3T3-E1 cells at 14 days of differentiation as revealed by Alizarin Red staining (B).

FIGURE 2.

SFN shows cell growth-suppressive effects in cells of the osteoblastic lineage. l-SFN and dl-SFN effects on proliferation/viability in MC3T3-E1 cells (A) and MLO-Y4 cells (B) after 24 and 48 h of SFN treatment (C and D) are shown. The EC₅₀ for l-SFN is ∼48 μm and for dl-SFN is ∼13 μm in MC3T3-E1 cells (E) and ∼11 μm for l-SFN and ∼6 μm for dl-SFN in MLO-Y4 cells (F). A comparison of l-, d-, and dl-SFN on proliferation/viability in MC3T3-E1 cells after 24 (G) and 48 h (H) is shown. Cell proliferation in A–D, G, and H is measured by cell count; EC₅₀ in E and F is measured by a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide-like assay; n = 4. Bars represent mean ± S.D.; error bars represent S.D.

FIGURE 3.

SFN induces extrinsic apoptosis in cells of the osteoblastic lineage. Fas mRNA expression after 3 μm l- or dl-SFN treatment in MC3T3-E1 and MLO-Y4 cells after 8 (A) and 16 h (B) is shown. Caspase 8 activity after 3 μm l-SFN and dl-SFN treatment at 24 h in both cell lines is shown in C. The effect of 3 μm dl-SFN treatment in MC3T3-E1 cells and of 3 μm l-SFN and dl-SFN in MLO-Y4 cells on the activities of Caspase 3/7 after 24 h is shown in D. For RT-qPCR analysis, Fas expression is referred to as 18S rRNA expression (A and B). In all graphs, Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. For A and B, n = 3, and for C and D, n = 5. *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001.

FIGURE 4.

SNF enhances osteoblast differentiation and activity. ECM mineralization of MC3T3-E1 cells (day 21), BMSCs (day 21), or neonatal calvarial explants (day 14) as measured by Alizarin Red staining is shown in A. Concentration-dependent effects on mRNA expression of ECM-related genes by dl-SFN in MC3T3-E1 cells after 14 days is shown in B. mRNA expression of Alpl (C), Bglap2 (D), Col1a1 (E), and Lox (F) after treatment with 3 μm d-, dl-, or l-SFN in MC3T3-E1 cells for 14 or 28 days is shown. In A, for MC3T3-E1 and BMSCs, n = 3, and for neonatal calvariae, n = 9. For RT-qPCR analysis, gene expression is referred to 18S RNA expression. In B–E, n = 3. In B, very low, non-measurable expression values are referred to as N/A. In all graphs, Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. *, p ≤ 0.05; **, p ≤ 0.01.

FIGURE 5.

SFN enhances expression of the osteoblastic master transcription factors. Treatment with 3 μm dl-SFN stimulates Runx2 mRNA expression in MC3T3-E1 cells and in BMSCs (A). In MC3T3-E1 cells, 3 μm dl-SFN significantly increases Runx2 protein expression after both 24 h (B) and 14 days (C) of treatment. Representative immunoblots of Runx2 protein expression are shown in D. M, protein marker. Concentration-dependent mRNA regulation of osteoblastic transcription factors by dl-SFN in MC3T3-E1 cells after 14 days is shown in E. mRNA expression of Runx2 (F), Atf4 (G), Sp7 (H), and Dlx5 (I) after treatment with 3 μm d-, dl-, or l-SFN in MC3T3-E1 cells for 14 (d14) or 28 days (d28) is shown. In A and E–I, for RT-qPCR analysis gene expression is referred to 18S rRNA expression. In B, Runx2 protein expression is referred to total protein expression. In E, very low, non-measurable expressions values are referred to as N/A. In A–C and E–I, n = 3. In all graphs, Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. *, p ≤ 0.05; **, p ≤ 0.01.

FIGURE 6.

SFN attenuates RANKL/Tnfsf11 expression in MLO-Y4 cells and mouse calvarial explants. Tnfsf11 expression in osteocyte-like MLO-Y4 cells upon 3 μm dl-SFN treatment at 3 and 8 days of treatment is shown in A. In mouse calvarial explants from neonatal (5 days (d)) and from adult mice (7 weeks (we)), 3 μm dl-SFN decreases Tnfsf11 expression after 12 days of treatment (B). Tnfsf11 gene expression is referred to 18S rRNA expression; n = 3. In all graphs Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. *, p ≤ 0.05.

FIGURE 7.

SFN affects global DNA hydroxymethylation status. MC3T3-E1 nuclei were stained by Hoechst dye and are represented at high and low gain by confocal microscopy. Global nuclear 5hmC levels (green dots) in MC3T3-E1 cells after 16 h of treatment with 3 μm dl-SFN are shown. The gain for nuclear visualization is strongly reduced to facilitate 5hmC visualization in euchromatin (black spots in low gain representation) (A). Global DNA hydroxymethylation at 16 h after treatment with 3 μm l- and dl-SFN in MC3T3-E1 cells and in MLO-Y4 cells was analyzed spectrophotometrically (B). In A, representative images are shown. In B, n = 4. Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. *, p ≤ 0.05.

FIGURE 8.

SFN transiently enhances expression of Tet1 in osteoblastic cells. Time-dependent Tet1, Tet2, and Tet3 expression patterns in MC3T3-E1 cells (A) and MLO-Y4 cells (B) are shown. d, days. Expression of Tet1, Tet2, and Tet3 by 3 μm dl-SFN at 8 and 16 h after treatment in MC3T3-E1 cells is shown in C. 3 μm dl-SFN does not significantly increase mRNA expression of Tet1, Tet2, or Tet3 in MLO-Y4 cells (D). For RT-qPCR analysis, Tet1, Tet2, and Tet3 gene expressions are referred to 18S rRNA expression; n = 3. In C and D, Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. *, p ≤ 0.05.

FIGURE 9.

Effects of SFN-induced active DNA demethylation in osteoblasts. Validation of mRNA (A) and protein (B) knockdown by specific siRNAs against Tet1 and Tet2 in MC3T3-E1 cells is shown. The effect of Tet1 and Tet2 knockdown on the cytostatic effect induced by dl-SFN in MC3T3-E1 cells is shown in C. In the schematic representation of the Atf4 proximal promoter/gene region, untranslated exons, translated exons, and introns are represented by gray, black, and white boxes, respectively. Analyzed fragments that overspan a CpG-rich region (bold black line) are shown. Small black arrows represent primer binding sites to quantify the selected fragments (D). Relative 5hmC amounts on Fragment 1 (E) and Fragment 2 (F) of the Atf4 proximal promoter/first exon in MC3T3-E1 cells after 16 h and 1 day of 3 μm dl-SFN treatment are shown. In A, E, and F, n = 3. In C, n = 4. A representative image is shown in B. M, protein marker. For RT-qPCR analysis, Tet1 and Tet2 gene expressions are referred to 18S rRNA expression. Cell proliferation in C is measured by cell count. In all bar charts, Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. *, p ≤ 0.05; **, p ≤ 0.01. Neg, negative; fw, forward; rv, reverse; TSS, transcription start site.

FIGURE 10.

SFN strongly inhibits viability and resorption of osteoclasts. l-SFN and dl-SFN effects on RAW 264.7 cells on proliferation/viability after 24 h of treatment are shown in A. For RAW 264.7 cells, the EC₅₀ for l-SFN is ∼10.10 μm and for dl-SFN is ∼6.01 μm (B). SFN affects cell metabolic activity most strongly in RAW 264.7 cells when compared with cells of the osteoblastic lineage (C). The effect of 3 μm l- or dl-SFN on dentin resorption by primary mouse osteoclast cultures is shown in D. A representative image of osteoclast resorption trails and pits on dentin is shown in E. Cell proliferation in A is measured by cell count. EC₅₀ in B is measured by a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide-like assay. For A and B, n = 4, and for C, n = 9. Bars represent mean ± S.D.; error bars represent S.D. **, p ≤ 0.01; ***, p ≤ 0.001.

FIGURE 11.

SFN induces FAS-dependent extrinsic apoptosis pathway in preosteoclastic cells. mRNA expression of Fas in RAW 264.7 preosteoclastic cells after treatment with 3 μm l- or dl-SFN for 16 h is shown in A. An increase of Caspase 8 and Caspase 3/7 activity in RAW 264.7 cells is seen after treatment with a 3 μm concentration of either SFN preparation after 24 h of treatment (B). A comparison for Fas induction (C) activation of Caspase (Casp) 8 (D) and Caspase 7 (E) activity by SFN between RAW 264.7 osteoclasts, MC3T3-E1 osteoblasts, and MLO-Y4 osteocytes is shown. For RT-qPCR analysis, Fas expression is referred to 18S rRNA expression. In all graphs, Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. For A, n = 3, and for B, n = 5. *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001.

FIGURE 12.

SFN induces global DNA hydroxymethylation and induces Tet1-dependent cell death in RAW 264.7 preosteoclasts. Global nuclear (blue) 5hmC levels (green dots) in RAW 264.7 cells after 16 h of treatment with 3 μm dl-SFN are shown. Nuclear gain settings are strongly reduced to facilitate 5hmC visualization in euchromatin (black spots in low gain representation). The original nuclear frame, as seen at high gain in confocal microscopy, is delineated by a dotted line for the enlarged nucleus (A). Quantitative analysis of global 5hmC by treatment with 3 μm l-SFN and dl-SFN in RAW 264.7 cells after 16 h is shown in B. A comparison of changes in global 5hmC levels among RAW 264.7, MC3T3-E1, and MLO-Y4 cells upon 3 μm dl-SFN treatment after 16 h is shown in C. The effect of 3 μm l- or dl-SFN on Tet1, Tet2, and Tet3 mRNA expression after 8 and 16 h of treatment in RAW 264.7 cells is shown in D. Validation of mRNA expression knockdown by specific siRNAs against Tet1 and Tet2 in RAW 264.7 cells is shown in E. The effect of Tet1 and Tet2 knockdown on the cytostatic effect induced by dl-SFN in RAW 264.7 cells is shown in F. In A, representative images are shown. In B, n = 4. For RT-qPCR analysis, Tet1, Tet2, and Tet3 gene expressions are referred to 18S rRNA expression; n = 3. Cell proliferation in F is measured by cell count; n = 4. In all graphs. Ctrl is set to 1, and treatments are referred to as -fold change to Ctrl. Values are represented as the mean ± S.D.; error bars represent S.D. *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001. Neg, negative.

FIGURE 13.

Schematic overview of the effects of SFN/DMSO on bone cells. DMSO and SFN induce active DNA demethylation via up-regulation of the Tet genes in vitro. This leads to apoptosis of preosteoclasts and to a lesser extent of preosteoblasts. Active DNA demethylation also enhances osteoblast differentiation. SFN further decreases the expression of RANKL/Tnfsf11 by a yet undefined mechanism. mCpG, methylated cytosines; hmCpG, hydroxymethylated cytosines; CpG, unmethylated CpGs; TSS, transcriptional start site.

FIGURE 14.

SFN has an anabolic effect on bone homeostasis in C57BL/6 mice. Effects of treatment of sham-operated and ovx young adult mice (8 weeks old) with 7.5 mm dl-SFN for 5 weeks on BVTV (A), Tb.N in proximal tibial bone (B), Tb.Sp (C), and Tb.Th (D) are shown. n = 9 for the sham group, n = 7 for the Ctrl OVX group, and n = 8 for the dl-SFN group. Values are represented as the mean ± S.D.; error bars represent S.D. *, p ≤ 0.05; ***, p ≤ 0.001.

FIGURE 15.

Evidence of correlations of matrix mineralization parameters (qBEI) with structural parameters (μCT). The parameters CaPeak (the most frequent calcium content) and CaWidth (the width of the calcium content distribution) correlate significantly with the corresponding μCT parameters BV/TV (A and B), Tb.N (C and D), and Tb.Sp (E and F). Trab, trabecular.