Bone anabolic modulation by sulforaphane

37 Pages·2016·2.02 MB·English

by  Roman Thaler

JBC Papers in Press. Published on January 12, 2016 as Manuscript M115.678235
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M115.678235
Bone anabolic modulation by sulforaphane

Anabolic and anti-resorptive modulation of bone homeostasis by the epigenetic modulator sulforaphane, a
naturally occurring isothiocyanate

Roman Thaler1,2, Antonio Maurizi3, Paul Roschger1, Ines Sturmlechner1, Farzaneh Khani2, Silvia Spitzer1, Monika
Rumpler1, Jochen Zwerina1, Heidrun Karlic1, Amel Dudakovic2, Klaus Klaushofer1, Anna Teti3, Nadia Rucci3*, Franz
Varga1*, Andre J. van Wijnen2*

1Ludwig Boltzmann Institute of Osteology at the Hanusch Hospital of WGKK and AUVA Trauma Center Meidling, 1st
Medical Department, Hanusch Hospital, Vienna, Austria.2Department of Orthopedic Surgery and Biochemistry and
Molecular Biology, Mayo Clinic, Rochester, MN, USA.
3Department of Biotechnological and Applied Clinical Sciences, University of L’Aquila, L’Aquila, Italy.

Running title: Bone anabolic modulation by sulforaphane

To whom correspondence should be addressed: Franz Varga, Ludwig Boltzmann Institute of Osteology at Hanusch
Hospital of WGKK and AUVA Trauma Center Meidling, 1st Medical Department, Hanusch Hospital, Vienna, Austria.
Email: [email protected]; Andre J. van Wijnen, Ph.D., Mayo Clinic, 200 First Street SW, Rochester, MN 55905,
Phone: [phone], Fax: [phone], Email: [email protected].
D
o
w
Keywords: Epigenetics, osteoblast, bone, osteogenesis, osteoporosis, ovariectomy n
lo
ad
e
d fro
balance of bone homeostasis and favor bone m
h
ABSTRACT acquisition and / or mitigation of bone resorption in ttp
vivo. Thus, SFN is a member of a new class of ://w
w
Bone degenerative pathologies like epigenetic compounds that could be considered for w
osteoporosis may be initiated by age-related shifts in novel strategies to counteract osteoporosis. .jbc
.o
anabolic and catabolic responses that control bone rg
homeostasis. Here we show that sulforaphane (SFN), by/
g
a naturally occurring isothiocyanate, promotes INTRODUCTION u
e
s
osteoblast differentiation by epigenetic mechanisms. t o
n
SFN enhances active DNA demethylation via Tet1 Bone loss disorders like osteoporosis are often A
p
and Tet2 and promotes pre-osteoblast differentiation multi-factorial diseases due to dysregulation of bone ril 1
2
by enhancing extracellular matrix mineralization and metabolism or homeostasis. To maintain bone density , 2
0
the expression of osteoblastic markers (Runx2, and integrity, complex networks and numerous 1
9
Col1a1, Bglap2, Sp7, Atf4 and Alpl). SFN decreases interactions between different bone cell types and their
the expression of the osteoclast activator RANKL in environment assure that bone remodeling is balanced
osteocytes and mouse calvarial explants and during growth and adulthood (1-3). These processes are
preferentially induces apoptosis in pre-osteoclastic orchestrated by many distinct regulators of gene
cells via up-regulation of the Tet1/Fas/Caspase 8 and expression like specific transcription factors, miRNAs
Caspase 3/7 pathway. These mechanistic effects and epigenetic chromatin modulators (4,5). These gene
correlate with higher bone volume (~20%) in both regulatory processes are controlled by paracrine events
normal and ovariectomized mice treated with SFN involving a variety of secreted ligands that affect both
for 5 weeks compared to untreated mice as bone-forming osteoblasts (e.g., parathormone (PTH),
determined by microCT. This effect is due to a higher bone morphogenetic proteins (BMPs), wingless-type
trabecular number in these mice. Importantly, no MMTV integration site family, members (WNTs)) or
shifts in mineral density distribution are observed bone resorbing osteoclasts (e.g., RANKL/TNSF11) or
upon SFN treatment as measured by quantitative both like members of the Serum amyloid A family (6,7).
back-scattered electron imaging (qBEI). Our data Both cell types are targeted by a number of different
indicate that the food-derived compound SFN therapeutic approaches that counteract bone
epigenetically stimulates osteoblast activity and degeneration in aging patients and other disorders
diminishes osteoclast bone resorption shifting the provoking bone loss.
1

Copyright 2016 by The American Society for Biochemistry and Molecular Biology, Inc.
Bone anabolic modulation by sulforaphane

Bisphosphonates, which are potent inhibitors of osteoblasts by epigenetic mechanisms (30-36).
osteoclastic bone resorption, are widely prescribed and Epigenetic reprogramming of osteoblastic differentiation
very effective at limiting bone loss. Yet, there is or of other bone related cells may represent a novel
considerable debate about the efficacy versus the approach to effectively regulate bone remodeling and
potential adverse effects of these compounds. In mice, homeostasis in certain disorders. Indeed, an increasing
the bisphosphonate ibandronate was reported to strongly number of studies emphasize the central role of
reduce bone formation by decreasing osteoblast numbers epigenetic regulatory mechanisms in bone biology (4).
and to disturb the bone marrow plasma cell niche, thus Based on the molecular similarities of SFN and DMSO,
negatively affecting bone remodeling (8). Use of as well as their bone-related biological properties, we
bisphosphonates is also associated with a higher risk of hypothesized that SFN would exhibit bone anabolic
atypical femoral fractures in women (9,10), with partly effects. Our results show that SFN has beneficial effects
elevated bone mineralization patterns (11,12) and with on bone and may act through an epigenetic mechanism
significantly longer union times of distal radius fractures that promotes Tet1/Tet2 dependent hydroxymethylation
(13). Current therapeutic strategies to promote of DNA re-activating gene expression.
osteoblastogenesis in osteoporosis aim to increase
osteoblast number and/or increasing osteoblast activity
by the administration of bone anabolic molecules like EXPERIMENTAL PROCEDURES
intermittent parathyroid hormone treatment (iPTH) or
bone morphogenetic proteins (BMPs) (14). However, Cell culture D
o
w
there are only a limited number of bone anabolic The following murine cell lines were used: n
lo
strategies that have shown promise in clinical MC3T3-E1, a clonal pre-osteoblastic cell line derived ad
e
applications. from newborn mouse calvaria (kindly provided by Dr. d fro
Sulforaphane (SFN) is an organosulfur Kumegawa, Department of Oral Anatomy, Meikai m
h
compound belonging to the isothiocyanate group. Its University, Sakado, Japan), the osteocyte-like MLO-Y4 ttp
precursor molecule glucoraphanin naturally occurs at cell line (kindly provided by Lynda Bonewald, ://w
w
high concentrations in cruciferous vegetables like University of Missouri-Kansas City, USA) and the pre- w
broccoli or cabbages. Highest concentrations are found osteoclastic, macrophage-like RAW 264.7 cell line .jbc
.o
in sprouts of broccoli and cauliflower (15). SFN is (ATCC, Manassas, VA, USA). rg
generated from glucoraphanin by the enzyme All cell lines were cultured in a humidified by/
g
myrosinase upon damage to the plant (16) such as from atmosphere with 5% CO at 37°C and were sub-cultured u
2 es
chewing. Due to its anti-oxidative potential, to its ability twice per week using 0.001% Pronase E (Roche Applied t o
n
to selectively induce activating phase II enzymes (17), as Science, Penzberg, Germany) and 0.02% EDTA in Ca2+- A
p
well as to inhibit histone deacetylase (HDAC) activity and Mg2+-free PBS before achieving confluence. ril 1
2
(18-20), SFN is predominantly studied for its anti- MC3T3-E1 and MLO-Y4 cells were cultured in α- , 2
0
carcinogenic and anti-microbial properties. Furthermore, minimum essential medium (α-MEM; Biochrom, Berlin, 1
9
recent studies suggest that SFN has potential beneficial Germany) containing 10 μg/ml gentamicin (Sigma-
effects for the treatment of diabetes type 1 and type 2 Aldrich, St. Louis, MO, USA); for MC3T3-E1 culture
(21-23), as well as osteoarthritis (24,25), and rheumatoid media was supplemented with 10% heat inactivated fetal
arthritis. In the latter context, SFN was found to inhibit calf serum (FCS; Biochrom) and MLO-Y4 cells were
synovial hyperplasia and T-cell activation (26). In cultured on rat-tail derived collagen Type I (Roche)
addition, SFN represses matrix-degrading proteases and coated dishes (final concentration of 0.15 mg/ml) and
protects cartilage from destruction in vitro and in vivo culture media was supplemented with 2.5% FCS
(27). SFN also may have effects on bone resorption, (Hyclone, GE Healthcare Life Sciences, Logan, UT,
because it inhibits the RANKL-dependent differentiation USA) and 2.5% calf serum (CS, Hyclone).
of osteoclasts in vitro by suppressing nuclear factor- Differentiation of MC3T3-E1 cells was induced
kappaB (28) and activation of the transcription factor with 50 μg/ml ascorbic acid (Sigma-Aldrich) and 5 mM
NRF2 (Nfe2l2) (29). β-glycerophosphate (Sigma-Aldrich) in medium
SFN shares molecular properties with the containing 5% FBS. RAW 264.7 cells were cultured in
cryopreservant dimethyl sulfoxide (DMSO). As we and Dulbecco's modified Eagle's medium (DMEM,
others have previously shown, DMSO causes phenotypic Invitrogen, Carlsbad, CA, USA) supplemented with 10%
changes and supports differentiation of several cell types fetal bovine serum (Hyclone) and 2 mM glutamine
like erythroid or embryonic stem cells, as well as (Gibco, Carlsbad, CA, USA). Cells were seeded in
2
Bone anabolic modulation by sulforaphane

culture dishes at a density of 20,000 cells/cm2 and measured by Alizarin Red S stain (Sigma-Aldrich). For
cultured overnight unless stated otherwise. Before cells this purpose, Alizarin Red S dye was extracted using
were treated with compounds, the culture medium was 10% cetylpyridinium chloride (CPC) in 10 mM
changed. sodiumphosphate (pH= 7.0) for 45 min at room
temperature. Alizarin Red S absorbance was measured at
Primary cells and tissues 562 nm in a multi-plate reader (Tecan, Maennedorf,
Primary mouse tissues or cells were obtained Switzerland) and normalized to total protein amount
according to procedures that conform to the regulatory measured by bicinchoninic acid assay (Thermo Fisher
guidelines of the Institutional Animal Care and Use Scientific, Waltham, MA, USA). From the other part of
Committee of the Medical University of Vienna. the calvariae total RNA was extracted (see below).
Primary mouse osteoclasts were harvested from 8-week
old C57BL/6 mice which were sacrificed and bone Histomorphometric measurements/assessment of
marrow cells were isolated from tibias and femurs under osteoclastic resorption
aseptic conditions and cultured in α-MEM (Biochrom) After culturing the primary mouse osteoclasts
containing 10% FCS (Biochrom), 10 mg/ml gentamycin for 14 days on dentin slices in the presence or absence of
(Sigma-Aldrich) and macrophage colony stimulating 3 μM L- or DL- SFN, substrates were put into water,
factor (MCSF1, 30 ng/mL). After 24 hours, non- sonicated for 10 min to remove living cells and air-dried.
adherent cells were seeded onto sterile, 300 μm thick Photographs were obtained by reflected light microscopy
dentin slices (elephant ivory) at 700,000 cells/cm² in (objective 20×) of the entire substrate surface, and D
o
w
αMEM supplemented with 10% FBS, 2 mM l-glutamine, resorption trails and pits were analyzed and quantified n
lo
30 μg/mL gentamycin, 20 ng/mL M-CSF, and 2 ng/mL with standard image analysis software (ImageJ, ad
e
RANKL (R&D Systems, Minneapolis, MN, USA). rsbweb.nih.gov/ij/). Resorption is shown as % area d fro
Culture medium was changed twice per week, cells were resorbed to total area of the dentin slice. m
h
cultured for 14 days. ttp
Mouse bone marrow mesenchymal stem cells Cell metabolic activity ://w
w
were isolated from aseptically dissected long bones of 6- To assess cell metabolic activity, a commercially w
week old C57BL/6 mice (Himberg, Austria). The available, MTT similar assay (EZ4U; Biomedica, .jbc
.o
marrow cavities were flushed with sterile medium using Vienna, Austria) was used. For this purpose, the cell rg
a 25-gauge needle and the culture was established in α- lines were incubated with increasing concentrations of by/
g
MEM supplemented with 10% FCS and 10 mg/ml L- or DL-SFN. After a comparable doubling time for all u
e
s
gentamycin. After 48 hours of culture at 5% CO and three cell lines the assay was performed following the t o
2 n
37°C, the non-adherent cells were removed by gentle protocol of the supplier. A
p
rinsing with PBS. At confluence, cells were harvested ril 1
and seeded at a density of 3,000/cm2 for cell Cell count 2, 2
0
experiments. For osteogenic differentiation cells were Cell-lines were seeded in 24-well culture dishes 1
9
cultured in α-MEM supplemented with 10% FBS, 10 at a density of 20,000/cm2 and were either left untreated
mg/ml gentamycin, 100 nM dexamethasone, 10 mM β- (controls) or treated with L-, D-, or DL-SFN at the
glycerophosphate, and 50 µg/ml ascorbic acid with or indicated concentrations for up to 48 hours. Thereafter,
without L-, D-, or DL-SFN. Culture medium was cells were detached with 0.001% pronase E and the
renewed twice a week and cells were cultured for the number of viable cells was assessed with Casy cell
periods indicated in the result section. counter (Schaerfe Systems, Reutlingen, Germany). Each
Calvariae from 2 to 3-day-old and from 7-week experiment was performed in biological quadruplicates
old C57BL/6 mice were dissected aseptically. The and experiments were repeated twice. For long term
calvarial bone explants were cultured in 48-well plates in experiments, cell number was determined using DNA
α-MEM (Biochrom) containing 10% FCS (Biochrom), amount as surrogate. Cell layers were washed with PBS
50 μg/ml ascorbic acid (Sigma-Aldrich), 5 mM β- and fixed for 20 min with 4% PFA. Thereafter, Hoechst
glycerophosphate (Sigma-Aldrich) and 10 μg/ml 33258 dye (Polysciences, Warrington, PA, USA) was
gentamicin (Sigma-Aldrich). The day after dissection, added (1 μg/mL) and, after an incubation of 15 min at
medium was changed and a part of the explants was room temperature, the fluorescence was measured
treated with 3 µM L- or DL-SFN for 12 days. Thereafter, (excitation 360, emission 465 nm). The amount of DNA
one part of the calvariae were fixed for 1 hour in 4% was estimated using a standard curve prepared from calf
para-formaldehyde (PFA) and mineralization was thymus DNA (Roche).
3
Bone anabolic modulation by sulforaphane

Measurement of caspase activity Whole cell protein extracts were prepared using
Caspase 3/7 and caspase 8 activities were SDS sample buffer (2% SDS, 100 mM β-
measured by using the Caspase-Glo 3/7 and Caspase- mercaptoethanol, and 125 mM Tris-HCl, pH= 6.8) and
Glo 8 assay Kit (Promega, Madison, WI, USA) heated at 95°C for 5 min.
following manufactures instructions. Briefly, after For immunoblotting analysis, 15 μg of protein
treatments, cells were lysed and substrate cleavage by extracts were separated on 10% SDS poly-acryl amide
caspases was measured by the generated luminescent gels, transferred to nitrocellulose membranes (Millipore)
signal with a 96 multi-well luminometer (Glomax, and equal protein loading was confirmed by Ponceau
Promega). Each experiment was performed in Red staining. Subsequently, membranes were blocked
quintuplicate and experiments were carried out twice. overnight with 10% blocking reagent (Roche) in TN
buffer (50 mM Tris and 125 mM NaCl, pH= 8.0). The
Tet1 and Tet2 siRNA transfections following primary antibodies were used: rabbit, anti
For Tet1 and Tet2 depletion by siRNA, cells Runx2 (sc-10758, Santa Cruz Biotechnologies, Santa
were seeded at 20,000 cells/cm2 in six-well culture Cruz, CA) rabbit, anti Tet1 (Active Motif, Carlsbad, CA,
plates. Six hours after seeding, cells were transfected USA, #61741) and rabbit, anti Tet2 (Santa Cruz, sc-
with 40 pmol of Tet1 or Tet2 siRNA (Sigma-Aldrich) 136926). Washing was performed with TN buffer
using X-tremeGENE siRNA Transfection Reagent containing 0.01% Tween. Binding of the secondary
(Roche) as described by the supplier. One day after antibody (anti-rabbit IgG/anti-mouse IgG horseradish
transfection, medium change was performed and one peroxidase-coupled) (Roche) diluted 1:10,000 in 10% D
o
w
part of the cells was treated with medium containing 3 blocking solution followed by detection with the BM n
lo
µM DL-SFN while the other part was left untreated. chemo-luminescence immunoblotting kit (Roche) was ad
e
After incubation time of 24 hours proteins or nucleic carried out as described by the supplier. Chemo- d fro
acids were isolated as described below and subjected to luminescence was measured with an image acquisition m
h
qRT-PCR, western blot or cell count was performed. system (Vilber Lourmat, Marne-la-Vallée, France). ttp
Measurements are given as means of 3 immuno-blots ://w
w
Isolation of RNA and reverse-transcriptase normalized to total protein loading and representative w
quantitative polymerase chain reaction (RT-qPCR) blots are shown. .jbc
.o
Total RNA was extracted using the SV Total rg
RNA Isolation kit (Promega) following the supplier's Immunostaining and quantifications of global by/
g
instructions. cDNA was synthesized from 0.5 μg RNA cytosine 5-hydroxymethylation u
e
s
using the First Strand cDNA Synthesis kit (Roche) as Cells were fixed with 4% paraformaldehyde in t o
n
described by the supplier. The resulting cDNAs were PBS for 20 min at room temperature and permeabilized A
p
subjected to quantitative PCR amplification with a real- with 0.2% triton in 4% paraformaldehyde in PBS. ril 1
2
time cycler using the QuantiTect SYBR-Green PCR Kit Thereafter cell layers were blocked for 20 min with 10% , 2
0
(Qiagen, Hilden, Germany) for the genes Alpl, Fas, Lox, blocking reagent (Roche) and incubated for one further 1
9
Tet1, Tet2, and TaqMan Gene expression Master Mix hour with 0.5 µg/ml anti-5hmC antibody (Abgent).
(Applied Biosystems, Foster City, CA, USA) for Afterwards, the cells were washed thrice with PBS and
measuring Runx2, Bglap2, Col1a1, Tnfsf11, and 18S incubated for one further hour with an Alexa-488 labeled
rRNA expression (for primers, see Table 1). SYBR- secondary antibody (Invitrogen) diluted 1:300 in
Green RT-qPCR was started with an initial denaturation blocking buffer. Finally nuclei were stained with
step at 95°C for 10 min and then continued with 45 Hoechst 33258 dye (Polysciences).
cycles consisting of 30 sec denaturation at 95°C, 30 sec For immunostaining, cell slides were mounted
annealing at primer-specific temperatures, and extension with Vectashield mounting media and
at 72°C. For measurement of the TaqMan assays, we immunofluorescence was visualized on a Leica laser-
applied an initial denaturation at 95°C for 10 min, scanning microscope. For better visualization of the
followed by 45 cycles alternating 60 sec at 60°C and 15 small and faint 5hmC spots in the nuclei, gain for
sec at 95°C (primers or TaqMan probe references are nuclear visualization (Hoechst, blue) was strongly
listed in Table 2). All RT-qPCR assays were performed reduced to facilitate 5hmC visualization in euchromatin
in triplicate and expression was evaluated using the (black spots in Low gain representation) and gain for
comparative quantification method (37). 5hmC signals (Alexa-488, green) was strongly
enhanced. Still, under this conditions, no background
Protein isolation and immunoblotting signal and no signal was found by using only the
4
Bone anabolic modulation by sulforaphane

Alexa488 labeled secondary antibody as negative control injected with 7.5 mM DL-SFN (200 µl of SFN at a
(Invitrogen). For quantification of global cytosine 5hmC, concentration of 63.8 mg/ml/kg dissolved in ethanol and
the fluorescence in the plates was measured with a further diluted in PBS) intraperitoneally every other day
multiwell plate reader (Tecan). The amount of DNA for five weeks. Control mice received vehicle alone. As
(nuclei signal) was estimated using a standard curve described by Kong JS et al., the injection of this dose of
prepared from calf thymus DNA (Roche). The signal of DL-SFN does not show apparent adverse effects,
the fluorescence of the 5hmC staining was normalized to including weight loss, alterations in physical appearance,
the amount of DNA. Also in this case, no signals were or changes in behavior in the treated mice (26).
found when only the Alexa-488 labeled secondary
antibody (Invitrogen) was used. Micro computed tomography (μCT) analysis
Images from proximal tibias fixed in 4%
Specific promoter hydroxymethylation analysis formaldehyde were acquired in a SkyScan 1174 with a
To analyze the 5hmC content in the resolution of 6 m (X-ray voltage 50 kV). Image
Atf4 proximal promoter/first untranslated exon, reconstruction was performed by applying a modified
appropriate fragments of the targeted promoter regions Feldkamp algorithm using the Skyscan Nrecon software.
were generated by digestion of 3 µg of genomic DNA 3D and 2D morphometric parameters were calculated for
from cells cultured for 16 hours or for 1 day with 3 μM the trabecular bone (350 consecutive slides, 6 m thick)
of DL-SFN with 30 U of BstBI and XmaI for 1 hours at starting from 300 m from the growth plate. Threshold
37°C. Subsequently, the enzyme was heat inactivated at values were applied for segmenting trabecular bone Do
w
65°C for 20 min and DNA was cleaned up with the corresponding to bone mineral density values of 0.6/cm3 n
lo
qiaquick DNA purification kit (Qiagen). Digestion of calcium hydroxyapatite. 3D parameters were based on ad
e
non-5mC DNA with the mentioned enzymes results in analysis of a Marching Cubes type model with a d fro
two fragments which over-spann a CpG rich region in rendered surface. Calculation all of 2D areas and m
h
the proximal promoter/first untranslated exon of the Atf4 perimeters was based on the Pratt algorithm (38). Bone ttp
gene (see Fig. 9D, Seq ref. CH466550). After DNA structural variables and nomenclature were those ://w
w
purification, hydroxymethylated DNA fragments were suggested in Bouxsein et al. 2010 (39). w
extracted by immuno-precipitation as described .jbc
.o
previously (30). In brief, hydroxymethylated DNA of Quantitative backscattered electron imaging (qBEI) rg
three independent biological experiments was captured Un-decalcified distal femora from 13 weeks old by/
g
by a specific hydroxymethyl-cytosine antibody (Active mice, sham/Crt (n=5), sham/DL-SFN (n=5), OVX/ctrl ue
s
Motiv #39769) coupled to magnetic beads to separate (n=4) and OVX/DL-SFN (n=5), were fixed and t o
n
the 5hmC containing DNA fragments. DNA was eluted dehydrated in alcohol, embedded in Ap
from the magnetic beads with 75 µl of 2 M NaCl polymethylmethacrylate (PMMA) and prepared for ril 1
2
solution. Finally, the mean methylation or backscattered electron imaging as previously described. , 2
0
hydroxymethylation status of the fragments was qBEI is based on backscattering of electrons from the 19
determined by amplifying the fragments by quantitative surface-layer (i.e., the initial ~1.5 micron) of a bone
real-time PCR. Amplification ratios of the bound section. The rate of these backscattered electrons is
(hydroxymethylated) DNA fraction to unbound (non proportional to the weight concentration of mineral
hydroxymethylated) DNA fraction were calculated (for (hydroxyapatite) and thus of that of calcium in bone.
primers see Table 2). For assessment of the specificity of Details of the method have been published elsewhere
the 5-hmC antibody, either 40 ng of non-methylated, (11,40,41). A scanning electron microscope (DSM 962,
fully methylated or fully hydroxymethylated DNA Zeiss, Oberkochen, Germany) equipped with a four
standards (Zymo Research) were run in parallel and quadrant semiconductor backscattered electron detector
amplified by real-time PCR using the supplied PCR was used. The accelerating voltage of the electron beam
Primers. was adjusted to 20 kV, the probe current to 110 pA, and
the working distance to 15 mm. The cancellous and
In vivo analysis of SFN effects on bone cortical bone areas were imaged at 200x nominal
Analysis of SFN on mice was based on a magnification (corresponding to a pixel resolution of
protocol published before by Kong JS et al. (26). In our 1µm/pixel). The BE-signal (gray scale) was calibrated
experimental setting, eight week old sham operated or using the “atomic number contrast” between carbon (C,
ovariectomized (OVX) C57BL6/J female mice were Z=6) and aluminum (Al, Z=13) as reference materials.
purchased from Charles River laboratories. Mice were From the calibrated digital images frequency distribution
5
Bone anabolic modulation by sulforaphane

of mineral concentration, the so called bone suggesting that the methanesulfonyl group of DMSO is
mineralization density distribution (BMDD), was important but not sufficient for full biological activity.
derived. The BMDD was characterized by 5 parameters: Because SFN has cell growth suppressive effects
CaMean (weighted mean Calcium content), CaPeak (43-45), we evaluated the impact of SFN on cell
(mode Ca content – peak position), CaWidth (full width proliferation and viability of bone related immortalized
at half maximum of the BMDD peak - reflecting the cell-lines. MC3T3-E1 osteoblasts and the MLO-Y4
heterogeneity in matrix mineralization), CaLow (the osteocytes were treated with increasing concentrations of
percentage of lowly mineralized bone area - below 17.68 SFN for up to 48 hours (Fig. 2). Because SFN is a chiral
weight% Ca), CaHigh (the percentage of highly molecule, we tested three different preparations: L-, D-,
mineralized bone area - above the 95th percentile value and DL-SFN, which is a mixture of both D and L
of the corresponding control animals BMDD. enantiomers. Both L-SFN and DL-SFN have modest
effects (<50%) on cell number in MC3T3-E1 and in
Statistical analysis MLO-Y4 after 24 hours of treatment at a concentration
Statistical analyses were performed using of 3 μM (Fig. 2A-B). At higher concentrations, both the
ANOVA or Student's t test in Prism 4.03 (GraphPad L-enantiomer and the DL-mixture compromise cell
Software, La Jolla, CA, USA). Values of P ≤ 0.05 were viability with the greatest decrease in cell number
considered significant. Each experiment consisted of at observed for MLO-Y4 cells (Fig. 2B). Furthermore, DL-
least three biological replicates. For RT-qPCR data, SFN has a somewhat stronger effect on cell survival
results from technical triplicates were averaged and the compared to L-SFN in MLO-Y4 cells at most D
o
w
mean value was treated as a single, biological statistical concentrations. Effects of L-SFN or DL-SFN on cell n
lo
unit. Results are presented as means ± SD. viability are slightly more pronounced after 48 hour ad
e
treatment in both MC3T3-E1 and MLO-Y4 cell lines d fro
(Fig. 2C-D). m
h
RESULTS To assess whether SFN affects cellular ttp
metabolic activity after 24 hours of treatment, we ://w
w
SFN affects cell viability of bone related cells performed EZ4U assays, an MTT-like assay that w
Our group and others have shown previously measures the capability of living cells to reduce .jbc
.o
that the widely used cryopreservant dimethylsulfoxide tetrazolium salts in the mitochondria into formazan rg
(DMSO) triggers differentiation of MC3T3-E1 derivatives that absorb at 450 nm. Titration curves for by/
g
osteoblasts. Although use of DMSO is restricted in both L-SFN and DL-SFN using either cell line revealed u
e
s
medical applications, there are intriguing analogies in that the half-maximal effective concentration (EC50) of t o
n
the chemical structures of DMSO and plant secondary L-SFN is ~48 μM and DL-SFN ~13 μM in MC3T3-E1 A
p
metabolites. As deduced from the IUPAC nomenclature, cells (Fig 2E), while lower doses are needed in MLO-Y4 ril 1
2
DMSO (methanesulfinyl methane) and the natural food cells (i.e., ~11 μM for L-SFN and ~6 μM for DL-SFN) , 2
0
compound SFN (1‐isothiocyanato‐5‐ (Fig. 2F). Hence, SFN has less metabolic effects on 1
9
methanesulfinylpentane) share striking molecular MC3T3-E1 osteoblasts than in MLO-Y4 osteocytes.
similarities. SFN carries an additional pentane group Remarkably, the preparation with both
with terminal Isothiocyanate (Fig. 1A). These structural enantiomers (L- and DL-SFN) shows stronger biological
similarities predict similarities in the biological effects effects compared to the sole L- enantiomer, suggesting
of DMSO and SFN. Therefore, we tested the effects of that the biological potency of SFN depends on steric
DMSO and SFN on MC3T3-E1 differentiation by considerations. Indeed, direct comparison of the effects
assessing the extent of Extra Cellular Matrix (ECM) of L- DL- and D-SFN enantiomers on cell proliferation
mineralization. Under our conditions, staining for the in MC3T3-E1 cells shows that the D-isoform strongest
late osteoblast biomarker Alizarin Red is not observed in affects cell proliferation after 24 hours (Fig. 2G) or 48
our cultures for the first two weeks of culture, but hours (Fig. 2H) at the most concentrations tested. Taken
typically is evident after three to four weeks (42). Our together, the results show that 3 μM SFN is a relatively
results show that both DMSO and SFN have similar non-toxic dose (Fig. 2) that is biologically effective (see
biological effects as revealed by increased matrix Fig. 1).
mineralization as early as 14 days of treatment (Fig. 1B).
Strikingly, SFN is at least four orders of magnitude more SFN induces extrinsic apoptosis in cells of the
potent than DMSO in stimulating matrix mineralization, osteoblastic lineage
6
Bone anabolic modulation by sulforaphane

As we have previously shown, DMSO induces differentiation related genes that support mineralization
the extrinsic pathway of apoptosis (30), suggesting that in MC3T3-E1 cells. Treatment with increasing DL-SFN
the growth suppressive action of SFN (as reflected by a concentrations (1μM to 30μM) shows that the strongest
decrease in cell proliferation and metabolic activity; see up-regulating effects are induced at a concentration of 3
Fig. 1) may be caused by programmed cell death in both, μM DL-SFN for the ECM related genes Alpl, Bglap2,
MC3T3-E1 osteoblasts and MLO-Y4 osteocytes. To test Col1a1 and Lox after 14 days of treatment. DL-SFN
this possibility, we monitored expression levels and concentrations higher than 10-15 μM result in inhibition
activity of apoptotic biomarkers. We find that Fas of cell proliferation and thus in poor RNA yields for RT-
mRNA expression is not affected by either L- or DL- qPCR analysis (Fig 4B). Furthermore, we were
SFN in MC3T3-E1 and MLO-Y4 cells after 8 hours of interested in the effects of the different SFN enantiomers
treatment (Fig. 3A). However, after 16 hours of on the expression of these genes at day 14 and at day 28
treatment, Fas mRNA expression is significantly in MC3T3-E1 cells. Therefore, we directly compared the
increased in both cell types by DL-SFN treatment (Fig. three SFN-enantiomer preparations D-, DL, and L-SFN.
3B). Corroborating these results, the activity of caspase At day 14, Alpl expression is significantly up-regulated
8 is significantly increased by L-SFN and DL-SFN after by DL- and L-SFN preparations (Fig. 4C) and the
24 hours (Fig. 3C). Furthermore, we observe a racemic DL-SFN mixture significantly increases Lox
significant increase in the activities of the caspases 3/7 mRNA expression at this time point (Fig. 4F). Bglap2
upon DL-SFN treatment of MC3T3-E1 cells and MLO- (Fig. 4D) and Col1a1 (Fig. 4E) are significantly up-
Y4 cells for 24 hours (Fig. 3D). These data indicate that regulated by D- as well as by DL-SFN at day 14. D
o
w
SFN induces the extrinsic apoptotic pathway in Furthermore, at day 28 Bglap2 is the only gene which is n
lo
osteogenic cells. still significantly up-regulated by D-SFN (Fig. 4D). ad
e
Taken together, our results indicate that SFN promotes d fro
Osteogenic long-term effects of SFN the expression of osteoblastic genes and mineralization. m
h
Beyond effects on cell growth and survival, SNF ttp
may affect osteoblast differentiation and activity. The bone-anabolic effects of SFN are reflected by ://w
w
Therefore, we investigated the long-term effects (>20 enhanced expression of primary osteoblastic w
days) of L-SFN and DL-SFN on mineralization of transcription factors .jbc
.o
osteoblast, bone marrow stromal/stem cell (BMSC) SFN may play a role in cellular differentiation rg
cultures and neonatal calvarial explants maintained in by selectively modulating mRNA expression of by/
g
osteogenic media. Treatment of MC3T3-E1 cells or osteogenic transcription factors. Therefore, we examined u
e
s
BMSCs (Fig. 4A) significantly increases ECM whether SFN affects the mRNA and protein expression t o
n
mineralization, the principal marker for mature of the bone-related master regulator Runx2 in mouse A
p
osteoblast function. Because L-SFN and DL-SFN have MC3T3-E1 cells or BMSCs. Treatment of cells with ril 1
2
mild effects on cell survival (at 3 μM) in non-confluent DL-SFN stimulates Runx2 mRNA expression in both , 2
0
MC3T3-E1 osteoblasts (see Fig. 2A and 2C), we cell types (Fig. 5A). Immune blot analysis of DL-SFN- 1
9
assessed cell number of both MC3T3-E1 cells and treated MC3T3-E1 cells shows a significant increase in
BMSCs in response to both enantiomer preparations. As Runx2 protein expression after both 24 hours (Fig. 5B)
anticipated, MC3T3-E1 and BMSCs cultures treated and 14 days (Fig. 5C) of treatment thus validating the
with either SFN type exhibit modest lower cell numbers, pro-osteogenic effects of SFN (see Fig. 4).
but higher mineral content per DNA unit after long-term Representative immune blots of Runx2 protein
treatment, indicating that SFN treated cells are more expression are shown in figure 5D.
active in mineralization. To evaluate if this bone After 14 days of treatment, similar to the ECM
anabolic effect occurs in explants which are rich in post- related genes (see Fig. 4), also for the osteoblastic
proliferative mature osteoblasts, we measured ECM transcription factors Runx2, Atf4 and Sp7/Osx, 3 μM DL-
mineralization of newborn calvarial explants treated with SFN shows the strongest positive effect in MC3T3-E1
3 μM L- or DL-SFN for 14 days. As in the 2D cell cells (Fig. 5E). In regard to the potency of the different
culture models, L- and DL-SFN each significantly SFN- isomers, comparison of the three SFN-preparations
increase ECM mineralization in calvarial explants (Fig. shows that for Runx2 (Fig. 5F), Atf4 (Fig. 5G) and
4A). These data indicate that SFN is capable of Sp7/Osx (Fig. 5H) the strongest effects are induced by
biologically stimulating mineralization of neonatal bone the D- or by the DL-SFN mixture at day 14. At day 28
tissue in culture. Next, we assessed whether SFN significantly up-regulations are still observed for Atf4
selectively modulates mRNA expression of and for Sp7/Osx genes by the D- or by the L-SFN
7
Bone anabolic modulation by sulforaphane

preparations. In contrast to DMSO (30), SFN does not that are at rather early than at late stages of osteogenic
increase the expression of the osteoblastic transcription differentiation.
factor Dlx5 at the times measured (Fig. 5I).
SFN transiently enhances expression of Tet1 in
SFN attenuates RANKL/Tnfsf11 expression in mouse osteoblastic cells
calvarial explants Global changes in 5hmC levels are mediated by
Runx2 expressed in osteoblastic cells mediates enzymes that control DNA- hydroxymethylation and are
biological feed-back by promoting osteoclast encoded by Ten-eleven translocation (Tet) genes. All
differentiation via up-regulation of Tnfsf11 (RANKL) three Tet genes (Tet1-3) are expressed in MC3T3-E1
(46). The latter promotes osteoclast formation and cells and expression for Tet1 and Tet2 decreases whereas
survival by binding to the osteoclastic receptor expression of Tet3 increases within the first 8 days of
Tnfrsf11a (alias RANK) which in turn activates NFkB MC3T3-E1 osteoblast differentiation (Fig. 8A). Yet, the
(47). Furthermore, osteocyte-derived Tnfsf11 controls levels of the Tet genes are distinctly regulated during
bone remodeling during postnatal development and in early stages in MLO-Y4 cells. Tet1 is much more rapidly
adult mammals. In osteocyte-like MLO-Y4 cells, DL- down-regulated in MLO-Y4 cells compared to MC3T3-
SFN reduces Tnfsf11 mRNA expression at both 3 and 8 E1, while Tet2 and Tet3 levels are steadily up-regulated
days after treatment, although changes in expression are (Fig. 8B). We next examined mRNA expression of Tet
only significant at the latter time-point (Fig. 6A). More genes in the absence or presence of DL-SFN in both
importantly, SFN decreases Tnfsf11 expression in mouse MC3T3-E1 and MLO-Y4 cells. Beyond the modulations D
o
w
calvarial explants from neo-natal mice and to a greater in Tet1 and Tet2 levels in response to biological n
lo
extent in explants from adult mice after 12 days of induction of differentiation, we find that expression of ad
e
treatment with DL-SFN ex vivo (Fig 6B). These results Tet1 and Tet2 increases to a limited degree (between 20 d fro
suggest that SFN may potentially promote net bone and 40%) by DL-SFN at 8 hours and 16 hours after m
h
accumulation by suppressing RANKL/Tnfsf11 induced treatment in MC3T3-E1 cells, albeit that these values ttp
osteoclast formation. reach statistical significance only for Tet1 at the 16 hour ://w
w
time point (Fig. 8C). DL-SFN does not positively change w
SFN induces global DNA hydroxymethylation mRNA expression of Tet1 or Tet2 in MLO-Y4 cells .jbc
.o
changes (Fig. 8D) and neither of Tet3 in both cell-lines (Fig. 8C, rg
As we have shown previously, DMSO D). We conclude that mRNA levels of the hydroxylation by/
g
dramatically induces active DNA demethylation in pre- enzymes Tet1 and Tet2 are modestly regulated by SFN u
e
s
osteoblastic MC3T3-E1 cells within less than one day of during bone cell differentiation, and may perhaps t o
n
treatment (<16 hours) through formation of 5-hydroxy- contribute to DL-SFN induced changes in global A
p
methylcytosine (5hmC) (30). Due to the structural and hydroxymethylation. ril 1
2
biological similarities between DMSO and SFN, we , 2
0
analyzed whether SFN alters the level of active DNA Effects of SFN-induced active DNA demethylation in 1
9
demethylation in cells of the osteogenic linage. Indeed, MC3T3-E1 osteoblasts
we measured a significant global increase in 5hmC as The here presented results are consistent with
marker for ongoing active DNA demethylation upon the idea that Tet-dependent hydroxymethylation may
treatment with either L- or DL-SFN in MC3T3-E1 cells compromise cell survival as previously shown by our
at 16 hours after treatment based on group (30). To test this hypothesis, we treated MC3T3-
immunofluorescence laser-scanning microscopy (Fig. E1 cells with non-silencing RNA (negative control) or
7A) and spectrophotometry (Fig. 7B). As expected, siRNAs against either Tet1 or Tet2 (Fig. 9A, B) prior
5hmC spots are observed in the looser euchromatic treatment with 3 μM of DL-SFN. Depletion of Tet1 but
regions (dark nuclear parts in the low gain picture not Tet2 significantly reduces the cytostatic effect of
version in Fig. 7A) demonstrating opening of chromatin DL-SFN (Fig. 9C) in MC3T3-E1 cells concurring with
by active DNA demethylation after SFN treatment. our previous results with DMSO (30).
Interestingly, in MLO-Y4 osteocytes that are The early effect of SFN on osteoblastic gene expression
differentiated beyond the osteoblastic stage, both L- and (e.g. Runx2, Fig. 5B) suggests that SFN, similarly to
DL-SFN do not elevate 5hmC levels (Fig. 7B). These DMSO (30), promotes osteoblastic differentiation by
findings collectively suggest that SFN selectively active DNA demethylation. The Runx2 P1 promoter
induces active DNA demethylation in osteoblastic cells which is active in osteoblasts (48,49) is not CpG
methylation sensitive (not shown) which may indicate an
8
Bone anabolic modulation by sulforaphane

indirect or active DNA demethylation independent effect up-regulates expression of the pro–apoptotic gene Fas at
of SFN on Runx2 expression. However, a CpG rich 16 hours but not at 8 hours (Fig. 11A). In addition, either
region is present in the Atf4 proximal promoter/first exon SFN preparation stimulates Caspase 8 and Caspase 3/7
region of the gene (Fig. 9D). As shown in Fig. 9E and F, activity (Fig. 11B) indicating that SFN induces the Fas-
already after 16 hours and partially after 1 day of Caspase8-Caspase3/7 pathway that defines extrinsic
treatment, 3 μM of DL-SFN treatment significantly apoptosis. Similar to findings with MC3T3-E1
increase the 5hmC levels in the selected fragments of the osteoblasts and MLO-Y4 osteocytes (see Fig. 3), the
proximal promoter region/first-second exon of the Atf4 DL-SFN mixture is more potent than the pure L-SFN
gene by 3 to 4 fold. In summary, these results indicate enantiomer in inducing apoptosis (Figs. 11A, B). More
that by DL-SFN induced active DNA demethylation importantly from a biological perspective, we observe
affects primary cellular functions in MC3T3-E1 cells. that bone-resorbing RAW 264.7 osteoclasts are twice as
sensitive to apoptotic induction as the bone-anabolic
Effects of SFN on osteoclast resorption and apoptosis MC3T3-E1 osteoblasts and MLO-Y4 osteocytes (Fig.
Because SFN has been shown to inhibit 11C-E). These findings suggest that SFN may positively
osteoclastogenesis by suppressing nuclear factor-kappaB affect bone homeostasis in vivo by preferentially
(NFkB) (28) and it also has an anti-proliferative effect in inducing apoptosis in osteoclasts.
bone-anabolic MC3T3-E1 and MLO-Y4 cells, we
examined whether SFN affects proliferation and activity SFN induces global DNA hydroxymethylation and
of RAW 264.7 cells. Treatment of these cells with controls Tet1 dependent cell death in osteoclasts D
o
w
increasing concentrations of SFN for 24 hours with the Similar to MC3T3-E1 pre-osteoblasts, we tested n
lo
two different preparations of SFN (i.e., L-SFN and DL- if SFN alters active DNA demethylation through ad
e
SFN) reduces the number of viable cells at doses of 3 formation of 5hmC in RAW 264.7 pre-osteoclasts. d fro
μM and higher (Fig. 10A). This inhibitory effect appears Treatment of these cells with 3 μM DL-SFN for 16 m
h
more pronounced compared to observations made with hours strongly increases the global levels of 5hmC (~3 ttp
osteogenic cell lines (see Fig. 2). Furthermore, SFN fold) as measured by immunofluorescence microscopy ://w
w
strongly reduces the cellular metabolic activity of RAW (Fig. 12A) and spectrophotometry (Fig. 12B). The L- w
264.7 cells in EZ4U assays (Fig. 10B). Also for this SFN enantiomer that has lower biological activity .jbc
.o
parameter a stronger effect in the RAW 264.7 cells is exhibits a less pronounced effect (<2 fold) (Fig. 12B). rg
observed when compared with the osteogenic cell-lines The observed changes in 5hmC levels are most by/
g
(Fig. 10C). We next assessed whether these negative prominent in RAW 264.7 cells compared to MC3T3-E1 u
e
s
effects alter the functional capacity of osteoclasts in and MLO-Y4 cells (Fig. 12C). The mRNA levels of the t o
n
bone resorption. Treatment with SFN significantly hydroxymethylation-related gene Tet1 but not Tet2 or A
p
reduces the area resorbed by primary mouse osteoclast Tet3 are upregulated after DL-SFN treatment in RAW ril 1
2
cultures on ivory discs (Fig. 10D, E). We also 264.7 cells which becomes evident at 8 hours but is , 2
0
investigated whether SFN alters progression of more pronounced at 16 hours (Fig. 12D). Therefore, to 1
9
osteoclastic differentiation in RAW 264.7 cells upon test if the strong cytostatic effects induced by SFN are
induction with RANKL and MCSF by monitoring the dependent on the up-regulation of Tet1, similar to
temporal expression of classical osteoclastic markers MC3T3-E1 pre-osteoblasts we pre-treated RAW 264.7
(e.g., Acp5, Clcr, Ctsk). However, we did not observe cells with non-silencing RNA (negative control) or
appreciable changes in the expression of these genes siRNAs for either Tet1 or Tet2 (Fig. 12E). Again
upon co-treatment with SFN (data not shown). Hence, depletion of Tet1 but not Tet2 reduced the cytostatic
SFN primarily affects activity of osteoclastic cells via its effect of DL-SFN (Fig. 12F), concurring with the data
cytostatic effects. from the MC3T3-E1 cells and with our previous results
(30).
SFN induces FAS-dependent extrinsic apoptosis in Taken together, our results show that the two
pre-osteoclastic cells organosulfur compounds DMSO (30) and SFN induce
Because SFN induces apoptotic markers in active DNA demethylation via up-regulation of the Tet
MC3T3-E1 osteoblasts and MLO-Y4 osteocytes (see genes in vitro. This epigenetic reprogramming of the
Fig. 3), we examined whether RAW 264.7 are sensitive chromatin leads to apoptosis of pre-osteoclasts, but only
to programmed cell death. Similar to MC3T3-E1 and to a lower extent of pre-osteoblasts. More importantly,
MLO-Y4 cells, the results show that treatment of RAW SFN leads to enhanced osteoblast differentiation without
264.7 pre-osteoclastic cells with 3 μM of L- or DL-SFN promoting osteoclast differentiation. Furthermore, in
9
Bone anabolic modulation by sulforaphane

osteocytes and bone tissue explants, SFN decreases the DL-SFN. The parameters CaPeak and CaWidth correlate
expression of RANKL/Tnfsf11, a major activator of significantly with the corresponding µCT parameters
osteoclastogenesis. Although the mechanism of this BV/TV, Tb.N and Tb.Sp (Table 1 and Fig. 15 A-F).
modulation remains to be investigated, it could be Independently of the treatment, it appears that mice with
associated with reduced reactive oxygen species required a higher trabecular number have higher matrix
for osteoclast differentiation (50) (Fig. 13). mineralization (CaPeak, Fig. 15C) which is less
heterogeneous (CaWidth, Fig. 15D), while mice with a
SFN has anabolic and anti-resorptive effect on bone low trabecular number have a less mineralized matrix
homeostasis in vivo which tends to be more heterogenous. Notably, the
Our in vitro results show that SFN stimulates correlations between structural indices and BMDD
osteogenic differentiation, as well as act as an anti- parameters show a clear separation between the sham
resorptive by blocking osteoclastogenesis and increasing and OVX group. Consistent with the expected lower
osteoclast apoptosis. Therefore, the attractive possibility bone turnover rates in sham versus OVX treated mice,
arises that SFN may affect bone homeostasis in vivo. As sham has higher CaPeak and lower CaWidth values than
shown by Kong JS et al, SFN treatment of young mice the OVX group. Taken together, our data suggest that
for five weeks with 7.5 mM DL-SFN (which results in DL-SFN has a beneficial effect on bone volume by
serum concentration of 24 µM and lower after 4 hours mitigating loss in trabecular bone number due to both
post injection) each other day, has no side effects on the anabolic and anti-resorptive cellular effects without
animals (26). We investigated the biological effects of altering bone matrix mineralization content and D
o
w
SFN in ovariectomized (OVX) mice that exhibit bone distribution. n
lo
loss due to estrogen deficiency (Fig. 14). Treatment of ad
e
sham-operated young adult mice (8 week old) with DL- d fro
SFN for five weeks shows significantly higher DISCUSSION m
h
metaphyseal cancellous bone volume per tissue volume ttp
(BV/TV) (Fig. 14A) and trabecular number (Tb.N) (Fig. In this study, we show that the broccoli-derived ://w
w
14B) in proximal tibias, while decreasing trabecular natural compound SFN has very similar osteoblast w
separation (Tb.Sp.) (Fig. 14C) but not affecting stimulatory effects as the structurally related organic .jbc
.o
trabecular thickness (Tb.Th.) (Fig. 14D) in these same solvent DMSO, which is known to promote osteoblast rg
samples. Furthermore, DL-SFN treatment for five weeks differentiation (30,36). Compared to DMSO, SFN by/
g
does not affect cortical thickness in any of the treated contains an aliphatic isothiocyanate moiety instead of a u
e
s
groups (data not shown). Thus, the relative higher methyl group and this modification apparently increases t o
n
BV/TV values found in the DL-SFN treated animal its biological potency in stimulating osteoblast A
p
groups are reflected by changes in trabecular number differentiation by >10,000 fold. We tested whether SFN ril 1
2
and not due to a thickening of the trabeculae by has anabolic effects on bone in vivo and find that SFN , 2
0
enhanced bone apposition. Importantly, SFN treatment increases bone volume by increasing trabecular number 1
9
of OVX mice reduces bone loss due to estrogen- albeit not cortical thickness. In addition, SFN mitigates
deficiency as reflected by mitigation of the observed bone-loss induced by estrogen depletion. Thus, SFN
OVX-dependent decrease of trabecular number (Fig. exhibits promising efficacy as a possible treatment
14A-B). modality for bone loss related pathologies like
To complement the µCT results on bone micro- osteoporosis.
structural indices, we measured the mineralization status Alteration of the epigenome by nutritional habits
of the bone matrix by assessing local mineral content or specific food compounds is associated with promoting
and distribution using quantitative back-scattered or repressing specific gene expression and pathologic
electron imaging (qBEI). Mice were examined for conditions (52-55). SFN has previously been shown to
changes in bone mineral density distribution (BMDD) inhibit histone deacetylase (HDAC) activity (18,45,56).
parameters CaMean, CaPeak, CaWidth, CaLow and In human volunteers, a single ingestion of a small
CaHigh reflecting bone turnover, mineralization kinetics amount of broccoli sprouts (~70g), which contain
and average bone matrix age (40,51) (Table 1 and Fig. particularly high amounts of SFN, inhibits HDAC
15). Two-way ANOVA analysis of CaPeak and activity in circulating peripheral blood mononuclear
CaWidth of trabecular bone revealed significant cells and consequently induces histone H3 and H4
differences between control (sham) and estrogen- acetylation few hours after intake (56). The epigenetic
depleted (OVX) mice, but no effects for treatment with landscape of cells can also be modified by other food
10