SEARCH

SEARCH BY CITATION

Keywords:

  • DLK1;
  • FA1;
  • PREF-1;
  • BONE REMODELING;
  • OSTEOCLAST;
  • OVX;
  • T CELLS

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Delta-like 1/fetal antigen 1 (DLK1/FA-1) is a transmembrane protein belonging to the Notch/Delta family that acts as a membrane-associated or a soluble protein to regulate regeneration of a number of adult tissues. Here we examined the role of DLK1/FA-1 in bone biology using osteoblast-specific Dlk1-overexpressing mice (Col1-Dlk1). Col1-Dlk1 mice displayed growth retardation and significantly reduced total body weight and bone mineral density (BMD). Micro–computed tomographis (µCT) scanning revealed a reduced trabecular and cortical bone volume fraction. Tissue-level histomorphometric analysis demonstrated decreased bone-formation rate and enhanced bone resorption in Col1-Dlk1 mice compared with wild-type mice. At a cellular level, Dlk1 markedly reduced the total number of bone marrow (BM)–derived colony-forming units fibroblasts (CFU-Fs), as well as their osteogenic capacity. In a number of in vitro culture systems, Dlk1 stimulated osteoclastogenesis indirectly through osteoblast-dependent increased production of proinflammatory bone-resorbing cytokines (eg, Il7, Tnfa, and Ccl3). We found that ovariectomy (ovx)–induced bone loss was associated with increased production of Dlk1 in the bone marrow by activated T cells. Interestingly, Dlk1−/− mice were significantly protected from ovx-induced bone loss compared with wild-type mice. Thus we identified Dlk1 as a novel regulator of bone mass that functions to inhibit bone formation and to stimulate bone resorption. Increasing DLK1 production by T cells under estrogen deficiency suggests its possible use as a therapeutic target for preventing postmenopausal bone loss. © 2011 American Society for Bone and Mineral Research.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Bone remodeling is a regenerative process that maintains bone mass and mechanical competence of the postnatal skeleton. It involves bone resorption by osteoclasts to remove old bone followed by bone formation by osteoblasts of new bone in a balanced fashion, and both processes are coupled in time and space.1, 2 During the bone-formation phase, osteoblasts are recruited to bone-forming surfaces from stem and progenitor cells present in the bone marrow.3 In contrast, osteoclasts are recruited to bone-resorbing surfaces from hematopoietic cells that fuse to form mature multinucleated osteoclasts.4 The control of bone remodeling and coupling of bone resorption and bone formation are mediated by a number of hormones [eg, parathyroid hormone (PTH)], osteoblast-produced factors [eg, receptor activator of nuclear factor-κB ligand (RANKL)] or osteoclast-produced factors.5 Recently, several cytokines and immune regulatory molecules, produced by T- and B-lymphocytes in the bone marrow microenvironment, have been demonstrated to play a role in the control of bone remodeling under physiologic or pathologic conditions.6 For example, estrogen (E) deficiency–related bone loss has been suggested to be mediated by a number of cytokines produced by immune cells.7, 8

Dlk1 is an imprinted gene encoding for a transmembrane protein with six epidermal growth factor (EGF)–like repeats belonging to the Notch/Serrate/Delta family.9 DLK1 is cleaved at the juxtamembrane region to generate one of the biologically active soluble forms of this protein named fetal antigen 1 (FA-1), which is present in blood and body fluids.10 Therefore, DLK1 can function either as a membrane-bound (DLK1; paracrine or autocrine action) or soluble circulating protein (FA-1; endocrine action).11 While Dlk1 has been cloned as a factor regulating adipogenesis and named preadipocyte factor 1 (Pref-1),12 various studies have demonstrated that Dlk1 participates in several regenerative and differentiation processes (eg, hematopoiesis,13, 14 wound healing,15 and muscle16 and liver regeneration17).

Several lines of evidence suggest an important role of Dlk1 in skeletal development and function.9 Mice deficient in Dlk1 exhibit high prenatal mortality, growth retardation, and skeletal malformations.18 Additionally, in the human maternal uniparental disomy (UPD) 14 syndrome (where DLK1 is silent), patients exhibit skeletal abnormalities, including short stature and skeletal malformations.19 We also have demonstrated in ex vivo culture conditions that DLK1/FA-1 regulates the differentiation of human mesenchymal (skeletal) stem cells (MSCs) into osteoblastic or adipocytic cells20 by influencing their microenvironment composition.20, 21 Thus we examined the physiologicl role of DLK1/FA-1 in bone biology employing transgenic animal models under normal physiologic conditions and in conditions of bone loss resulting from E deficiency.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Animals

All experimental procedures were approved by the Danish Animal Ethical Committee. The mice, including Dlk1−/− mice,22 were bred and housed under standard conditions (21°C, 55% relative humidity) on a 12-hour light/12-hour dark cycle. Ad libitum food (Altromin Spezialfutter GmbH & Co. KG, Lage, Germany) and water were provided.

Generation of Col1-Dlk1 transgenic mice

For generating mice overexpressing Dlk1 in preosteoblasts (Col1-Dlk1), the full-length human DLK1/PREF-1 cDNA20 was subcloned into the pKBpA plasmid (kindly provided by Thomas L Clemens, University of Cincinnati, Concinnati, OH, USA), and the 3.6-kb fragment of the rat type 1 α1 collagen promoter (pOBCol3.6) was subcloned upstream of hDlk1 cDNA in the same vector. DNA was injected into C57BL/6J oocytes, and transgene-positive mice were mated with C57BL/6J mice to generate Col1-Dlk1 mice.

Ovariectomy

Ovariectomy (ovx) was performed on 4-month-old female mice anesthetized with ketamine (100 mg/kg of body weight) and xylazine (10 mg/kg of body weight). Ovaries were removed through the dorsal side, and sham-operated mice had both ovaries externalized briefly and then gently reinserted.

Skeletal analysis

Bone mass and body composition measurements

Fat mass (g), bone mineral content (BMC, g), bone mineral density (BMD, g/cm2), and femur length were measured using dual-energy X-ray absorptiometry (DXA) using a PIXImus2 (Version 1.44; Lunar Corporation, Madison, WI, USA). DXA scans were acquired every month after sedating mice with isoflurane starting at 2 months of age and ending at 14 months of age. Body weight was recorded at the same time using an XL-300 balance (Denver Instrument GmbH, Göttingen, Germany).

Micro–computed tomographic (µCT) scanning

The tibias of 5-month-old mice were scanned with a high-resolution µCT system (vivaCT40, Scanco Medical, Bassersdorf, Switzerland), resulting in 3D reconstruction of cubic voxel sizes 12 × 12 × 12 µm3. A detailed description for quantification of the 3D microarchitecture of cortical bone has been presented previously.23

Static and dynamic histomorphometry

All histomorphometric analyses were performed according to the protocol described previously.24 Briefly, skeletons were fixed in PBS-buffered 3.7% formaldehyde, incubated in 70% ethanol, dehydrated, embedded in methyl methacrylate, sectioned at 5 µm, and stained as indicated using standard protocols. Parameters of static and dynamic histomorphometry were quantified on toluidine blue–stained undecalcified proximal tibia and lumbar vertebral sections using the Osteo-Measure Histomorphometry System (Osteometrics, Atlanta, GA, USA). Fluorochrome measurements for determination of the bone-formation rate were performed on two nonconsecutive 12-µm sections for each animal. Statistical differences between the groups (n = 5) were assessed by the Student t test.

Molecular biology techniques

Total RNA isolation and real-time PCR

RNA was isolated from organs or cultured cells using Trizol (Invitrogen A/S, Taastrup, Denmark) according to the manufacture's protocol. cDNA was synthesized using a commercial revertAid H minus first-strand cDNA synthesis kit (Fermentas, Copenhagen, Denmark) according to the instruction manual. Real-time polymerase chain reaction (PCR) was performed in the iCycler IQ detection system (Bio-Rad, Herlev, Denmark) by using SYBR Green I, as described previously.25 Gene expression level for each target gene was calculated using a comparative Ct method [(1/(2ΔCt) formula], and data were represented as relative expression to β-actin or as fold induction over control expression. Data were analyzed using Optical System Software Version 3.1 (Bio-Rad) and Microsoft Excel 2000 (Microsoft, Copenhagen, Denmark) to generate relative expression values.

Microarray analysis

Total RNA was isolated from cultured calvaria wild-type osteoblast (OB) and DLK1+ OB cells using an RNeasy Mini Kit (Qiagen, Copenhagen, Denmark). Gene expression profiles were determined using an Affymetrix MOE430 2.0 array GeneChip (Affymetrix, Voyager, Mercury Park, Wycombe Lane, UK) according to the manufacturer's recommendations, as described previously.26 Gene ontology classifications were performed according to their molecular function and relevant biologic process using DAVID 2.0 Software 1 (Affymetrix, Copenhagen, Denmark). Microarray data have been submitted to the Gene Expression Omnibus (GEO, NCBI, NIH) repository and can be found with accession number GSE22113 at the following link: www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc = GSE22113.

Transfection and NF-κB luciferase assay

Transfections and luciferase assay were done using Fugene 6 (Roche, Hvidovre, Denmark), a Cignal NF-κB reporter Assay Kit (Qiagen Nordic, Copenhagen, Denmark), and the Dual-Luciferase Assay (Promega Biotech AB, Nacka, Sweden), respectively, following the protocols recommended by the manufacturers. Briefly, neonatal calvarial wild-type OBs and DLK+ OB cells were transfected with the NF-κB reporter system and assayed for luciferase activity after 24 hours of transfection. To study the effect of FA-1, wild-type OB cells were transfected with reporter system, and 6 hours after transfection, the cells were treated with purified FA-1 for 24 hours and then assayed for luciferase activity.

In vitro cell culture

Bone marrow–derived mesenchymal stem cell (MSC) culture and CFU-F assay

MSCs were isolated from mouse bone marrow and cultured as described previously.27 For colony-forming unit fibroblast (CFU-F) assay, bone marrow cells (106) were cultured in chamber flasks as described earlier in complete isolation medium (CIM) for 5 days. Colonies with more than 50 cells were scored as CFU-Fs after staining with crystal violet. For osteogenic induction, bone marrow cells were cultured in CIM supplemented with 50 µg/mL of ascorbic acid and 5 mM β-glycerophosphate for 5 days, and the CFU-Fs were stained for alkaline phosphatase (ALP) activity to determine the number of ALP+CFU-Fs.

Osteoblastic cell culture

Primary osteoprogenitors (OBs) were isolated from the calvaria of neonatal (4- to 5-day-old) mice as described previously.28 ALP activity measurement and matrix mineralization staining by alizarin red (Sigma-Aldrich, St Louis, MO, USA) were performed as described previously.29

Heterotopic bone-formation assay in vivo

OB cells (5 × 105) were mixed with hydroxyapatite/tricalcium phosphate ceramic powder (HA/TCP, 40 mg; Zimmer Scandinavia, Albertslund, Denmark) and implanted subcutaneously in the dorsal surface of 2-month-old female in NOD/MrkBomTac-Prkdcscid mice (Taconic, Ry, Denmark). Implants were recovered 8 weeks after transplantation, paraffin embedded, sectioned, and stained by eosin/hematoxylin as described previously.30

Osteoclast culture

Bone marrow (BM) cells derived from 8-week-old mice were plated in 96-well plates in osteoclastic medium (OCM) containing α-MEM (Gibco BRL, Carlsbad, CA, USA) supplemented with 10% FBS (Gibco BRL), 100 U/mL of penicillin (Gibco BRL), 100 µg/mL of streptomycin (Gibco BRL), 25 ng/mL of recombinant human macrophage colony-stimulating factor (rhM-CSF; R&D Systems, Minneapolis, MN, USA), and 25 ng/mL of rhRANKL (Pepro-Tech, Rocky Hill, NJ, USA) to generate osteoclast-like cells.31 Tartrate-resistant acid phosphatase (TRACP) staining was performed 5 days after culture, and TRACP+ multinucleated cells (MNCs) with more than four nuclei were scored as osteoclasts (OCs). Osteoclast resorption pit assay was performed by culturing whole BM cells on bovine bone slices 6 mm in diameter and 0.65 mm thick (Nordic Bioscience, Copenhagen, Denmark) in OCM for 6 days. Resorption pit area was measured on nine randomly placed digital images per bone slice.

Colony-forming unit granulocyte/macrophage (CFU-GM) assay

BM cells collected from wild-type and Col1-Dlk1 mice were washed, diluted in 30% FBS– Iscove's modified Dulbecco's medium (IMDM) and then seeded in Mouse Methylcellulose Base Media (R&D Systems Europe Ltd., Abingdon, UK) supplemented with 10 ng/mL of CSF (R&D Systems). The clonogenic assay was performed as described previously32 by adding 300 µL of 3 × 105 cells directly into a 3-mL methylcellulose tube and seeded in triplicate in 35-mm dishes. The cultures were incubated at 37°C in 5% CO2 for 7 days, and the number of colonies (>50 cells) was scored at ×20 magnification.

Coculture experiments

Calvarial wild-type OBs or flow-activated cell sorting (FACS)–sorted DLK1+OBs (20,000 cells/well) were cocultured with wild-type-derived BM cells (2.5 × 105 cells/well) in 96-well plates in α-MEM containing 10% FBS, 25 ng/mL of rhM-CSF, and either 20 nM 1,25-dihydroxycholecalciferol (vitamin D3; kindly provided by Leo Pharma, Ballerup, Denmark) or rhRANKL (5, 10, or 15 ng/mL) for 5 days. OC differentiation was assessed by TRACP staining. For resorption activity, cocultures were performed on bovine bone slices and stained with 1% toluidine blue O (Sigma-Aldrich) as described earlier.

RNA interference

siRNAs were transfected into calvarial cell DLK1+OBs in 12-well plates according to the protocol provided by Applied Biosystems and Dharmacon/Thermo Scientific/Nordic Biolabs (Thermo Fisher Scientific, Slangerup, Denmark). Final concentration of each siRNA was 200 nM. The following siRNAs were used: mouse Ccl3 (s73439), mouse Tnfa (s75248), mouse Il7 (s68301), and control from Applied Biosystems.

Protein analysis

FA-1 enzyme-linked immunosorbent assay (ELISA)

Blood samples were collected from the retro-orbital plexus. The mouse and human FA-1 levels in serum were measured using a previously described sandwich ELISA.33, 34

FA-1 staining

To confirm the localization and expression of human DLK1 in the tissues and osteoblastic cells of the Col1-Dlk1 transgenic mice, paraffin-embedded sections from different tissues and cells cultured on chamber slides were stained with a rabbit anti-hFA-1 antibody as described previously.10

Flow-activated cell sorting (FACS)

Neonatal calvarial osteoprogenitors from Col1-Dlk1 mice were resuspended in blocking buffer, incubated with polyclonal rabbit anti-human DLK1,10 washed twice in PBS plus 0.5% BSA, incubated for 30 minutes on ice with FITC-conjugated polyclonal swine anti-rabbit immunoglobulin (1:10; Dako A/S, Glostrup, Denmark), and sorted on a FACSVantage with BD FACSDiVa option (Becton Dickinson Biosciences, Brøndby, Denmark). FACS data were analyzed using Cell-Quest 3.1 software (Becton Dickinson Biosciences). FACS analysis of the BM T-cell subpopulation was performed using four-color flow cytometric analysis, as described in Supplemental Materials and Methods.

TRACP staining of bone

Tibias harvested from 12-week-old mice were fixed in 4% paraformaldehyde (VWR - Bie & Berntsen A/S, Herlev, Denmark) for 24 hours, decalcified in 12% EDTA (pH 7.0), and embedded in paraffin. Sections 4 µm thick were rehydrated through descending concentrations of ethanol (99%, 93%, and 77%) before being stained for TRACP activity with an enzyme-histochemical method as described previously.35

Microscope imaging

Stained cells and paraffin-embedded sections were photographed using a DM4500 B microscope equipped with a Leica DFC300 FX Digital Color Camera (Leica Microsystems A/S, Herlev, Denmark). Turboscan mosaic images were created using Objective Imaging's Surveyor software (Version 5.04.01) with automated Turboscan using a ×20 objective lens in mosaic acquisition (Objective Imaging, Ltd., Cambridge, UK).

Statistical analysis

All values are expressed as mean ± SD or mean ± SEM. Comparison between groups was performed using an unpaired student's t test (two-tailed). p < .05 was considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Stage-dependent expression of Dlk1 during osteoblast differentiation

To determine the role of DLK1 in regulating postnatal bone mass, we studied the Dlk1 expression pattern during osteoblast (OB) differentiation of ex vivo cultured osteoblastic progenitors derived from neonatal calvaria (Calv-OB). Quantitative RT-PCR analysis demonstrated that Dlk1 expression was upregulated 30-fold during the early phases of OB differentiation and declined dramatically after 10 days in association with osteoblast maturation and expression of the late osteogenic markers osteocalcin and osteopontin (Fig. 1A).

thumbnail image

Figure 1. Osteoblast-specific expression of Dlk1 leads to growth retardation and skeletal abnormalities. (A) Real-time PCR analysis of mDlk1 expression in association with the expression of osteogenic markers Alp, Col1a1, and Oc during osteoblast differentiation of neonatal calvaria osteoprogenitor cells from wild-type mice. Expression of each target gene was normalized to β-actin and represented as fold induction over noninduced control cells. (B) Immunohistochemical staining analysis of the hDLK1 (transgene) expression in the metaphyseal regions of tibia and in cultured BM-derived MSCs and neonatal calvaria OBs from Col1-Dlk1 transgenic mice compared with wild-type mice. (C) ELISA measurements of DLK1 secreted form (hFA-1) in sera obtained from male and female 2-month-old transgenic mice. (D) X-ray radiographs showing the growth retardation in Col1-Dlk1 transgenic mice compared with the wild-type control (2 months old). (E) Whole-mount staining for bone and cartilage in E17.5 wild-type and Col1-Dlk1 embryos showing skeletal patterning defects in transgenic embryos. Delayed ossification and delayed closure of calvarial sutures are indicated by blue arrows and circle.

Download figure to PowerPoint

Col1-Dlk1 mice exhibit skeletal patterning defects and reduced body weight

We generated mice overexpressing Dlk1 in the whole osteoblast lineage (mesenchymal progenitors, osteoprogenitors, and osteoblasts) using the pOBCol3.6 promoter.36 We also employed human full-length Dlk1 cDNA as a transgene in order to distinguish its expression from endogenous Dlk1 using DLK1/FA-1 human-specific antibodies.37 Several transgenic lines (Col1-Dlk1 mice) were generated. The mice were healthy and transmitted the transgene at the expected Mendelian ratio. Immunohistochemical staining for hDLK1/FA-1 revealed expression of the transgene specifically in the osteoblastic cells lining bone-formation sites in trabecular bone (Fig. 1B). We did not detect any expression of the DLK1 transgene in osteocytes in bone sections from transgeneic mice. The transgene was found to be expressed in cultured BM-derived mMSCs (BM-MSCs) and Calv-OBs from Col1-Dlk1 mice (Fig. 1B). In addition, the full ectodomain of transgenic DLK1 was proteolytically cleaved and shed into the circulation, as indicated by high serum levels of hFA-1 (>100-fold) in Col1-Dlk1 mice compared with wild-type mice (Fig. 1C).

Col1-Dlk1 mice were smaller in size than their wild-type littermates (Fig. 1D). Examining the whole-mount skeleton of a fetal mouse at E17.5 revealed mild skeletal patterning abnormalities in Col1-Dlk1 embryos (Fig. 1E) with delayed closure of cranial sutures, flattening of the skull, hypomineralization of the parietal bones, and the formation of a “kink” at the border between the proximal and distal parts of the ribs. Compared with wild-type mice, Col1-Dlk1 mice exhibited decreased body weight (p < .005 for all time points; Fig. 2A) and reduced fat mass (Fig. 2B) throughout the life span of both males and females (p < .05 for each individual time point). At 6 months of age, the total body fat of wild-type mice was 20.22% ± 3.83% compared with 16.31% ± 3.29% in Col1-Dlk1 mice (p < .05). In addition, the adipose tissue depots at different locations (eg, renal fat pads; Fig. 2B) seemed to be reduced in size in Col1-Dlk1 mice compared with their wild-type littermates. Histomorphometric analysis of BM fat revealed significant reduction in marrow fat of adult Col1-Dlk1 mice compared with age-matched wild-type controls (Fig. 2C). Furthermore, the expression levels of adipogenic markers, including adiponectin (Apm1), fatty acid synthase (Fas), and adipsin, were markedly reduced in the white adipose tissue obtained from Col1-Dlk1 mice compared with wild-type mice (Supplemental Fig. S1A).

thumbnail image

Figure 2. Reduced fat mass and bone mass in Col1-Dlk1 mice. (A) Growth curves of both genders of Col1-Dlk1 mice and their wild-type littermates over a period of 14 months. Mice were weighted gravimetrically at different time points on an XL-300 balance (Denver Instrument GmbH). (B) Total fat mass during growth of mice was determined by PIXImus2 imager (Lunar). Histologic analysis of white fat obtained from renal pad fat of 3-month-old mouse stained with hematoxylin and eosin (H&E). (C, a) Histologic analysis of BM fat in tibia sections from 3-month-old transgenic mouse. (b) Histomorphometric analysis of marrow fat. (D) Total BMD of both genders of Col1-Dlk1 mice and their wild-type littermates during their growth up to 1 year of age. Total BMD (g/cm2) was measured using a PIXImus2 (Lunar). (E) Histomorphometric analysis of 2-month-old wild-type and Col1-Dlk1 vertebrae shows decreased bone volume/total volume (BV/TV), trabecular thickness (Tb.Th), and osteoblast surface/bone surface (ObS/BS). (F) Dynamic histomorphometric analysis showing reduced bone-formation rate (BFR) in Col1-Dlk1 mice. Values are represented as mean ± SD; n = 10 to 12 mice. *p < .05; **p < .005 versus wild-type mice.

Download figure to PowerPoint

Col1-Dlk1 mice exhibit a low-bone-mass phenotype

Both male and female Col1-Dlk1 mice displayed a marked reduction of whole-body bone mineral density (BMD; 19%) and bone mineral content (BMC; 33%) during the first 3 months of age (Fig. 2D and Supplemental Fig. S1B). At 1 year of age, the differences were less pronounced in male mice [total BMD and BMC were reduced by 3.5% (p < .05) and 3.9% (NS), respectively; Supplemental Table S1], but they were still highly statistically significant for female Col1-Dlk1 mice: Whole-body BMD and BMC were reduced by 7.3% (p < .005) and 11.6% (p < .005), respectively, compared with wild-type mice. In addition, micro–computed tomographic (µCT) analysis of cortical bone in the proximal tibias of 5-month-old mice showed significantly decreased bone volume fraction, decreased cortical thickness, and significantly decreased material density in Col1-Dlk1 mice compared with wild-type mice (Supplemental Fig. S2).

Histomorphometric analysis performed on the lumber vertebra of 10-week-old mice revealed a marked reduction (47.4%, p < .01) of the trabecular bone volume/total volume (BV/TV) ratio, a reduction of trabecular thickness by 13.2% (p < .01), and a decrease of the osteoblast surface/bone surface ratio (Ob.S/BS) of 28% (p < .05) in Col1-Dlk1 mice compared with wild-type controls (Fig. 2E). Dynamic histomorphometry revealed decreased bone-formation rate (BFR) by 41% (p < .05) in Col1-Dlk1 mice compared with wild-type mice (Fig. 2F). These data demonstrate that high levels of DLK1/FA-1 are associated with impaired osteoblastic bone formation.

Impaired osteoblast differentiation in Col1-Dlk1 mice

To study the cellular mechanisms of reduced osteoblastic bone formation in Col1-Dlk1 mice, osteoblast proliferation and differentiation were evaluated in a number of in vitro osteoprogenitor culture systems. BM-derived stromal (mesenchymal) stem cells (MSCs) obtained from 8-week-old Col1-Dlk1 mice exhibited reduced total number of CFU-Fs by 71% (p < .01) and decreased alkaline phosphatase–positive CFUs (ALP+-CFU-Fs) by 47% (p < .01) compared with wild-type mice (Fig. 3A, B), indicating the inhibitory effect of DLK1 on both the number of bone progenitors and the osteogenic commitment of BM-MSCs.

thumbnail image

Figure 3. DLK1 inhibits osteoblast differentiation of MSC and calvarial OB precursor cells. (A) Analysis of the colony-forming unit fibroblast (CFU-F) number derived from 2-month old Col1-Dlk1 and wild-type BM. (B) Analysis of osteogenic CFU-F number (ALP+CFU-F). Alkaline phosphatase (ALP) staining of CFU-Fs was performed after culturing CFU-Fs for 6 days in osteogenic medium. (C) ALP and alizarin red staining of calvarial osteoprogenitor cultures (calvarial-OBs) generated from 4-day-old Col1-Dlk1 and wild-type mice on day 14 of induction in osteogenic medium. (D) ALP activity measurements during the osteogenic differentiation of calvarial OBs. (E) Real-time PCR analysis of osteogenic markers (Alp, Oc, Cbfa1, Osx, and Bsp) in calvarial OBs on day 14 of induction in osteogenic medium. (F) FACS of DLK1+OBs from freshly cultured calvarial OBs obtained from Col1-Dlk1 mice. Positive cells are gated by region P4 (purple), whereas negative sorted cells are gated by region P5 (yellow). (G) Analysis of in vivo ectopic bone-formation capacity of DLK1+OBs versus wild-type OBs. Histologic analysis of paraffin-embedded implants stained with H&E demonstrating the impaired capacity of transplanted DLK1+OBs to generate bone (B) compared with wild-type OBs (magnification ×100).

Download figure to PowerPoint

In addition, ALP activity, in vitro formation of mineralized bone nodules, and the expression levels of various osteoblastic marker genes were dramatically reduced in ex vivo cultures of neonatal calvarial osteoprogenitors from Col1-Dlk1 mice compared with wild-type mice (Fig. 3CE). To study the effect of DLK1 overexpression in a homogeneous culture during OB differentiation, we isolated DLK1+ oseoproginators from neonatal calvaria of Col1-Dlk1 mice (DLK1+OBs) by FACS using human-specific antibody for hDLK1 protein (Fig. 3F). DLK1+OBs did not form bone in an ectopic bone-formation assay compared with wild-type cells (Fig. 3G). These data suggest the presence of a cell-autonomous osteoblast differentiation and bone-formation defect in DLK1+OBs.

Enhanced osteoclastic bone resorption in Col1-Dlk1 mice

Histomorphometric analysis of proximal tibias from Col1-Dlk1 mice revealed an increase in the osteoclast surface/bone surface (Oc.S/BS) ratio by 62% (p < .01) compared with wild-type mice (Fig. 4A). Interestingly, mature osteoclasts present in Col1-Dlk1 bone were bigger and contained more than 10 nuclei, whereas we did not observe osteoclasts with more than 5 nuclei in wild-type mice. Cultured BM cells derived from Col1-Dlk1 mice in the presence of M-CSF and RANKL formed 40% more TRACP+ osteoclasts than BM cells from wild-type animals (p < .01; Fig. 4B, a). Furthermore, in a pit-formation assay, Col1-Dlk1-generated osteoclasts showed higher bone-resorption activity than those from wild-type mice (p < .01; Fig. 4B, b). In addition, BM mononuclear cells harvested from Col1-Dlk1 mice showed significantly increased numbers of CFU-GM colonies in vitro compared with wild-type BM cells, indicating increased numbers of osteoclast precursors in Col1-Dlk1 mice BM (Fig. 4C). To study whether DLK1 exerts a direct effect on osteoclastogenesis, we added a purified soluble form of DLK1 (FA-1) at different concentrations (5 to 20 µg/mL) to cultured wild-type BM cells in the presence of M-CSF. We did not observe an increased number of TRACP+ osteoclasts compared with nontreated controls (Fig. 4D). Taken together, these data suggest that DLK1 enhances osteoclast precursor cell proliferation and differentiation through an indirect mechanism.

thumbnail image

Figure 4. Increased osteoclastic bone resorption in Col1-Dlk1 mice. (A) Marked stimulation of TRACP-stained osteoclastic bone-resorption area in the proximal tibia of 2-month-old Col1-Dlk1 mice compared with their wild-type littermattes. Histomorphometric analysis of OcS/BS in tibia sections obtained from 2-month-old mice. (B, a) Enhanced in vitro osteoclast differentiation of BM cells obtained from 2-month-old Col1-Dlk1 mice compared with wild-type cultures. The TRACP+ cells with more than five nuclei per cell were counted as multinucleated cells (MNCs). (b) Increased resorption pit forming ability of cultured transgenic BM cells on dentin disks stained with 1% toluidine blue O. Scale bar = 200 µm. (C) Increased number of CFU-GM colonies of cultured mononuclear cells from Col1-Dlk1 bone marrow compared with wild-type BM. Cells were cultured in semisolid methylcellulose medium for 7 days, as described in “Materials and Methods.” (D) Wild-type BM cells were cultured in the presence of rhM-CSF and rhRANKL without (non) or with different concentration of purified FA-1 protein for 5 days and stained for TRACP. (E) Increased number of TRACP+ MNCs in coculture system of wild-type BM with DLK1+OBs. Whole BM cells from wild-type mice were cocultured either with wild-type OBs or transgenic OBs for 5 days in the presence of vitamin D3 and prostaglandin E2. (F) FA-1 stimulates the number of TRACP+ MNCs in the coculture of wild-type BM with wild-type OBs. (G) Real-time PCR analysis of Rankl, Opg, and Nfat-c1 expression in bone samples obtained from Col1-Dlk1 and wild-type mice. Values are represented as mean ± SD; n = 4 to 5 mice. *p < .05; **p < .005 versus wild-type mice.

Download figure to PowerPoint

To investigate whether the stimulatory effect of DLK1 on osteoclastogenesis is mediated via an osteoblast-dependent mechanism, we performed coculture experiments using mononuclear cells derived from wild-type BM and OBs derived from neonatal calvaria of either wild-type or Col1-Dlk1 mice in the presence of vitamin D3 and prostaglandin E2 (inducers of RANKL expression in osteoblasts). Interestingly, DLK1+OB cells significantly stimulated osteoclast differentiation of wild-type BM cells by more than 52% (p < .01) compared with wild-type OBs (Fig. 4E). Moreover, purified FA-1 protein markedly increased the TRACP+ osteoclastic cells when added to a coculture of wild-type BM cells and wild-type OBs (Fig. 4F).

To test whether the observed increase in the number of osteoclasts in Col1-Dlk1 mice was due to changes in the components of the RANKL-RANK pathway,38 we examined the expression of Rankl, Opg, and Nfat-c1 in tibias from Col1-Dlk1 and wild-type mice. As shown in Fig. 4G, no significant changes in the expression of Rankl, Opg, and Nfat-c1 were observed in wild-type and transgenic mice, as assessed by real-time polymerase chain reaction (PCR). In addition, the expression of Rankl, Opg, and Nfat-c1 from cultured OBs or whole BM cells was comparable between transgenic and wild-type mice (data not shown).

To further examine the function of DLK1 as a costimulatory factor for osteoclastogenesis, we studied the stimulatory effect of the addition of RANKL (ranging from 5 to 20 ng/mL) on transgenic OB-induced osteoclastogenesis in the coculture system with wild-type BM cells in the presence of M-CSF. Interestingly, in the presence of DLK1, RANKL exerted a dose-dependent synergistic effect on the formation of TRACP+ MNCs in cocultures of wild-type BM cells and DLK1+OBs (Supplemental Fig. S3) compared with the coculture of wild-type BM cells with wild-type OBs in the presence of RANKL. In conclusion, these data suggest that DLK1 stimulates osteoclastogenesis via an osteoblast-dependent mechanism.

DLK1+OB cells exhibit enriched gene expression of a large number of immune-modulatory factors and activation of NF-κB signaling pathway

We have shown recently that Dlk1 inhibited the differentiation of hMSCs into osteoblast and adipocyte lineages and increased the production of a large number of inflammatory and immune response–related factors.21 We therefore hypothesized that the observed effects of DLK1 on osteoclastogenesis and osteoblastogenesis could be mediated by increasing the expression of OB-derived bone-resorbing cytokines other than RANKL. To test this possibility, we performed microarray analysis to compare the basal gene expression pattern of DLK1+OBs versus wild-type OBs using an Affymetrix MOE430 2.0 Array (Affymetrix). A total of 153 genes were found to be differentially upregulated by more than twofold in DLK1+OBs compared with wild-type OBs. Gene annotation analysis revealed that the immune regulatory factor genes accounted for 30% of the total upregulated genes in DLK1+OBs (Fig. 5A). As shown in Supplemental Table S2, the upregulated immune response–related genes include Tnfa (5.7-fold) and Il7 (2.2-fold), both known to promote osteoclastogenesis and repress osteoblast differentiation and activity.39, 40 In addition, Dlk1 upregulated several proinflammatory mediators such as chemokines (Cxcl2 and Ccl3), cathepsins (Ctss and Ctsb), and complement components (C1qa). To examine whether the stimulatory effect of Dlk1 on cytokines expression is mediated through the induction of the transcription factor NF-κB, we transfected DLK1+OBs and wild-type OBs with an NF-κB-responsive luciferase construct. Interestingly, reporter luciferase activity was increased significantly in DLK1+OBs compared with wild-type OBs (Fig. 5B, a), suggesting an activation of the NF-κB signaling pathway (Fig. 5B, a). Furthermore, addition of purified FA-1 (1 µg/mL) to wild-type OBs increased the NF-κB reporter luciferase activity (Fig. 5B, b).

thumbnail image

Figure 5. DLK1 promotes osteoclastogenesis and inhibits osteoblast differentiation by increasing the production of inflammatory cytokines. (A) Microarray analysis of differentially expressed genes in DLK1+OBs versus wild-type OBs. Dlk1 upregulated genes were categorized based on their biologic function. (B) DLK1 induces activation of the NF-κB signaling pathway. (a) Wild-type OBs and DLK1+OBs were tansfected with a luciferase NF-κB reporter construct and assayed for luciferase activity. (b) Wild-type OBs transfected with reporter system were treated or not (non) with 1 µg/mL of purified FA-1 protein for 24 hours and assayed for luciferse activity. Measurements are represented as mean of ± SD of three independent experiments. (C) Knocking down of Ccl3, Tnfa, or Il7 expression in DLK1+OBs using siRNA. Real-time PCR analysis of Ccl3, Tnfa, or Il7 mRNA expression in wild-type OBs (open bar), DLK1+OBs (gray bar), and siRNA-transfected DLK1+OBs (black bar). (D) ALP activity measurement in siRNA-transfected DLK1+OBs after 7 days of culture in osteogenic medium. (E) Reduced osteoclast supportive ability of siRNA-transfected DLK1+OBs in coculture system with wild-type BM. Values are represented as mean ± SD of three independent experiments; n = 5 to 6 mice. *p < .05; **p < .005 versus wild-type mice.

Download figure to PowerPoint

To further study whether the inhibitory effect of DLK1 on OB differentiation is mediated by the increased production of inflammatory cytokines, we inhibited the expression of three significantly upregulated cytokines (Il7, Tnfa, and Ccl3) known to inhibit OB differentiation in DLK1+OBs using RNA interference technology. Individual silencing of Il7, Tnfa, or Ccl3 in DLKL1+OBs (Fig. 5C) rescued their ability to undergo osteoblast differentiation, as assessed by an increase in ALP activity by 2.8-, 4.0-, and 3.8-fold, respectively, over control siRNA-transfected cells (Fig. 5D). Furthermore, as shown in Fig. 5D, knocking down Il7, Tnfa, and Ccl3 individually in DLK1+OBs markedly reduced their capacity to enhance osteoclast formation to different degrees in coculture assays with BM MSCs from wild-type mice (Fig. 5E). These data support the dual role of DLK1 in inhibiting OB differentiation and stimulating osteoclastogenesis via upregulation of proinflammatory cytokines.

Ovariectomy increases the serum level of DLK1/FA-1 and upregulates the production of DLK1 by T cells in the bone marrow

To further study the role of DLK1 in normal bone physiology, we measured serum DLK1/FA-1 levels in mice during estrogen deficiency (E deficiency), which is a physiologic condition known to affect bone formation and bone resorption.8 We found that ovx-induced bone loss in mice (Fig. 6A) was associated with enhanced production of FA-1 up to 45% compared with sham-operated or control nonoperated groups (Fig. 6B). In addition, quantitative real-time PCR revealed a marked increase of Dlk1 expression in whole BM isolated from ovx mice compared with sham-operated mice (Fig. 6C).

thumbnail image

Figure 6. Ovariectomy increases DLK1-producing T cells in BM. (A) Total BMD (g/cm2) was measured by PIXImus2 (Lunar) at 10 weeks of age in intact mice and 4 weeks after ovariectomy or sham operation in the same group of mice. Data (n = 6/group) are represented as percent difference from intact mice. (B) ELISA measurements of serum mFA-1 (soluble form of DLK1) in control, sham-operated, and ovx mice. (C) Real-time PCR analysis of Dlk1 mRNA expression in whole BM samples. (D) FACS analysis to study the effect of ovariectomy on the expression of DLK1 by different subpopulations of T cells in BM. (a) Dot plot shows the expression of DLK1 by CD3+ T cells in whole BM mononuclear cells. (b) Ovariectomy increases the number of DLK1-producing CD3+ T cells. (c) Ovariectomy increases the percentage of both CD8+DLK1+ and CD4+DLK1+ cells in CD3+ BM T cells. (d) Ovariectomy increases the expression of DLK1 by activated CD69+ cells in both CD4+ and CD8+ BM T cells that are gated from whole CD3+ cells. (E) Ovariectomy increases the amount of DLK1 expression by CD4+ and CD8+ cells in BM CD3+ T cells. Mean equivalent fluorochrome (MEF). FACS data are mean ± SD of three independent experiments; n = 6 mice per group. *p < .05; **p < .005 compared with sham-operated mice.

Download figure to PowerPoint

Since T-lymphocytes and their secretory products have been shown to play a major role in mediating E-deficiency-induced bone loss,41 we examined whether ovariectomy upregulated the production of FA-1 by T-lymphocytes in BM. Thus whole BM was harvested 4 weeks after ovariectomy or sham operation and analyzed by FACS. Ovariectomy increased the number of BM T cells (CD3+) producing DLK1 by fivefold over sham-operated mice (Fig. 6D, a, b).

Analysis of fractionated CD3+ T cells by triple staining of BM cells with CD3, DLK1, and either CD4 or CD8 antibodies revealed a significant increase in the number of DLK1-producing CD8 and CD4 T-cell subsets by 7.5- and 4.4-fold, respectively, in ovx mice compared with sham-operated mice (Fig. 6D, c). Analysis of Dlk1 expression by activated CD69+ T cells revealed that ovariectomy increased the percentage of CD8+CD69+DLK1+ and CD4+CD69+DLK1+ T cells by 4.6- and 5.3-fold, respectively, over sham-operated mice (Fig. 6D, d). Furthermore, FACS analysis performed to determine the amount of DLK1 expression on CD4+ and CD8+ T cells (mean equivalent fluorochrome) revealed a marked upregulation of DLK1 expression per CD4+ and CD3+ T cell. Taken together, these data demonstrate that ovx-induced E deficiency increases the level of DLK1 in the BM by increasing the number of DLK1-producing T-cell fractions, as well as the level of DLK1 expression from each T cell (Fig. 6E). FACS analysis of T cells from the spleen revealed no difference in the total number of CD3+DLK1+ cells between ovx and sham-operated mice. However, the analysis of fractionated T cells showed a significant increase in the number of DLK1-producing CD8+ T cells but not in CD4+ cells in ovx mice compared with sham-operated mice (Supplemental Fig. S4). In support of the immune cells (T and B cells) as a crucial source for DLK1 production in vivo, we measured the serum level of DLK1/FA-1 and the expression of Dlk1 mRNA in the bone marrow of NOD/SCID (NOD/MrkBomTac-Prkdcscid) mice known to be deficient in B and T cells.42 Interestingly, serum DLK1/FA-1 levels and Dlk1 expression in bone marrow were markedly reduced by 37% and 75%, respectively, in SCID mice compared with wild-type mice (Supplemental Fig. S5).

Dlk1 knockout mice are significantly protected from ovx-induced bone loss

To investigate whether DLK1 functions to mediate the effect of E deficiency on bone loss, Dlk1−/− mice and their wild-type littermates underwent either ovariectomy or sham operation. Interestingly, 1 month after ovariectomy, the wild-type ovx mice displayed significantly bone loss (p < .005) compared with bone loss observed in Dlk1−/− ovx mice (p < .05; Fig. 7A). Dlk1−/− mice were protected from ovx-induced bone loss by 60% (p < .005) compared with wild-type ovx mice. These data indicating that E-deficiency-induced bone loss is mediated at least in part by DLK1.

thumbnail image

Figure 7. Dlk1−/− mice are protected against ovariectomy-induced bone loss. Four-month-old female Dlk1−/− mice and their wild-type littermates underwent either ovariectomy or sham operation. (A) Total-body BMD was measured at baseline 1 month after ovariectomy. All data are represented as percent change from baseline. Data are mean ± SEM of three independent experiments of n = 8 mice per experiment. *p < .05; **p < .005, compared with wild-type ovx mice. (B) Schematic model for the mode of action of DLK1 in promoting bone resorption and inhibiting bone formation. Increasing the expression of DLK1 (ie, under estrogen deficiency) stimulates activation of the NF-κB signaling pathway and the production of proinflammatory cytokines by stromal/osteoprogenitor cells in the BM microenvironment that show mediation of the effect of DLK1 on blocking osteoblastogenesis and stimulating osteoclastogenesis with a net effect of increasing bone loss.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

In this study we identified DLK1 as a regulator of bone mass that inhibits bone formation and stimulates bone resorption through a mechanism involving the increased production of inflammatory cytokines by stromal/osteoprogenitor cells in the BM microenvironment. Additionally, our data demonstrate that DLK1 mediates part of the E-deficiency-induced bone loss, which confirms its role in normal physiology.

Dlk1 is an imprinted gene broadly expressed by most fetal tissues and known for its role during embryonic development. In support of the role of Dlk1 during embryogenesis, our Col1-Dlk1 mouse embryos (E17.5) displayed growth retardation and mild skeletal abnormalities, including delayed ossification in the sternum and in closure of the sagittal suture of the skull. Similar to our findings, mice expressing double and triple doses of Dlk1, driven by its own endogenous gene regulators, revealed a dose-dependent effect of Dlk1 on the severity of skeletal defects and growth development.43 Moreover, recently, Dlk1 has been demonstrated to play a role in chondrogenesis44 and in embryonic chondrogenic development.45 Thus the level of Dlk1 expression during embryonic development appears crucial for skeletal maturation and needs a more extensive analysis.

Overexpression of Dlk1 in the postnatal period led to decreased bone formation. This inhibitory effect of Dlk1 was associated with decreased numbers of osteogenic stem cells, with an impaired ex vivo differentiation into mature osteoblasts or with their inability to form ectopic bone in vivo compared with wild-type cells. In addition, during the osteoblast differentiation program, Dlk1 gene expression is downregulated in parallel with upregulation of mature osteoblastic markers. This pattern also has been demonstrated in other cell models, including murine preadipocytic cell lines 3T3-L1 and Balb/c 3T3, where downregulation of Dlk1 expression was necessary to allow adipocyte differentiation in response to proadipogenic signals.46, 47 Our findings suggest a similar regulatory function for DLK1 in the osteoblast differentiation process and are consistent with our previous data demonstrating a role of DLK1 in maintaining the BM pool of undifferentiated hMSCs by blocking their differentiation into mature osteoblasts and adipocytes.20

Our data showed that reduced bone mass in Dlk1-overexpressing mice was caused by reduced bone formation and increased bone resorption. Using different in vitro cell culture models, we demonstrated that DLK1 functions as a costimulatory factor for osteoclastogenesis via an osteoblast-dependent mechanism. This effect was not due to increased production of Rankl/Rank because we did not detect any changes in expression of Rankl (or Rankl/Opg ratio) in bone samples or in cultured OBs or BM MSCs from transgenic mice. Conversely, the effect of DLK1 on osteoclast formation was enhanced in the presence of RANKL. One explanation is that the main effect of DLK1 is to enhance the production of committed preosteoclastic precursor cells (Fig. 4C) that fuse to form mature osteoclasts in the presence of RANKL. Our data also demonstrated that DLK1 can participate in the later differentiation stage by increasing the production of a number of proinflammatory cytokines (eg, TNF-α) that are known to control osteoclast generation.

To identify targets mediating the biologic effects of DLK1, we studied the expression profile of Dlk1-induced genes in DLK1+OBs. Interestingly, Dlk1 expression in osteoprogenitors increases the expression of several proinflammatory cytokines known to act as “coupling factors” that inhibit osteoblast formation and stimulate osteoclast formation, such as Il7, Tnfa, and Ccl3 (MIP-1α).8, 48–50 Indeed, knocking down the expression of the Dlk1 upregulated Il7, Tnfa, and Ccl3 individually in DLK1+OBs, rescued their osteoblastic differentiation capacity, and reduced their ability to stimulate osteoclastogenesis. However, we were not able to detect significant differences among the effects of Il7, Tnfa, and Ccl3 siRNA on ALP activity or TRACP+ MNC number, probably owing to the general nature of the experimental setup, which has limited responsive range.

Dlk1 increased the production of other inflammatory-related factors that are known to attract, differentiate, and activate osteoclastic cells, including Cxcl4 [platelet factor 4 (PF4)]51; Ccl9 (MIP-1γ)52; complement components C1qα, C1qc, and C3αR153, 54; cathepsins such as Ctss, Ctsc, and Ctsb55, 56; integrin β2 (ITGB2, CD18)57; and Tlr2 (Toll-like receptor 2), an important receptor required for eliciting inflammatory bone loss in response to bacterial infections.58 The association between Dlk1 expression and inflammation also has been demonstrated in a recent report showing the stimulatory effect of DLK1 protein on proinflammatory cytokines by immune and adipose cells.59 Furthermore, several studies support the involvement of DLK1 in general biologic processes associated with stimulating inflammatory responses. These studies demonstrated the increased expression of Dlk1 in association with inflammatory responses during liver17 and muscle regeneration16 following tissue injury and during the progression of liver fibrosis in biliary atresia disease.60 It is possible that DLK1, with its receptor, induces the expression of proinflammatory cytokines. One of the candidate signaling pathways that can mediate these effects is NF-κB signaling, known for its important role in inflammation.61 Indeed, we have demonstrated previously that the inhibitory effect of DLK1 on hMSC differentiation into osteoblasts and adipocytes is mediated through activation of the NF-κB pathway and increasing the production of inflammatory cytokines.11 Consistent with this, using NF-κB-responsive luciferase construct, we demonstrate that Dlk1 functions as an NF-κB activator in osteoprogenitor cells. Thus this finding, together with our previous data, strongly suggests that the stimulatory effect of Dlk1 on increasing the production of inflammatory cytokines in BM is mediated by the NF-κB signaling pathway.

In this report we have demonstrated a physiologic role for DLK1 in mediating bone loss owing to E deficiency based on the following: (1) the increased expression levels of Dlk1 under E deficiency, (2) the increased production of DLK1 in ovx mice by both activated CD4 and CD8 T cells, known to be involved in enhanced E-deficiency-based osteoclastogensis,62, 63 (3) the partial protection of Dlk1-deficient mice from ovx-induced bone loss, and (4) the ability of Dlk1 to induce bone loss by stimulating osteoclastogenesis and inhibiting osteoblastogenesis, which is similar to the known effects of E deficiency on bone biology. What are the sources of DLK1 in the BM microenvironment? There are several possible cell types that produce DLK1 in BM, including osteoprogenitors/stromal cells, preadipocytes, and B cells.9, 20 Our study demonstrates for the first time that T cells are an important source for DLK1 production, especially in the context of E deficiency. This evidence is supported by significant reduction of DLK1 serum levels and reduced Dlk1 expression in the BM of mice deficient in T cells (SCID mice). Moreover, T cells have been demonstrated to be an important target for estrogen action in the BM and to play a crucial role in the mechanism by which E deficiency induces bone loss by increasing the production of several osteoclastogenic cytokines (eg, RANKL, TNF, and IFN-γ).48, 63 It is thus possible that under E deficiency, production of DLK1 by T-lymphocytes is upregulated. DLK1 participates in the creation of the proinflammatory microenvironment through its autocrine or paracrine effects (see model in Fig. 7B). The mechanism by which ovariectomy increases DLK1 production by T cells remains to be determined.

Our data demonstrate that the upregulation of DLK1 in BM creates an inflammatory microenvironment that leads to bone loss owing to enhanced bone resorption and inhibition of bone formation and demonstrate a physiologic role for DLK1 in mediating E-deficiency-related bone loss. Our data may suggest a link between inflammation, DLK1, and bone loss and suggest a role for DLK1 in other conditions where inflammation is associated with bone loss (eg, during infectious diseases and inflammatory diseases such as rheumatoid arthritis).63 Further studies are needed to explore the possibility that DLK1 may be a good therapeutic target for suppressing inflammation-related bone loss.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

We are grateful to Tina Nielsen, Lone Christiansen, and Bianca Jørgensen for excellent technical assistance and Linda Harkness for critical reading of the manuscript. This study was supported by grants from the Lundbeck Foundation, the Danish Medical Research Council, the Novo Nordisk Foundation, and the Region of Southern Denmark.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Additional Supporting information can be found in the online version of this article.

FilenameFormatSizeDescription
JBMR_346_sm_SuppData.pdf357KSupplemental Data

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.