|Home | About | Journals | Submit | Contact Us | Français|
CCAAT/Enhancer Binding Proteins (C/EBPs) are expressed by osteoblasts and adipocytes during differentiation. C/EBPβ is critical for adipogenesis; however its role in osteoblastogenesis is unclear, and its function in the postnatal skeleton is not known. To study C/EBPβ in osteoblasts in vivo, we created transgenic mice expressing full length C/EBPβ under the control of a 3.8 kilobase fragment of the human osteocalcin promoter. Two transgenic lines were established in an FVB genetic background, and compared to wild type littermate controls. Both C/EBPβ transgenic lines exhibited osteopenia, with a 30% decrease in bone volume, due to a decrease in trabecular number. The number of osteoblasts and osteoclasts per bone perimeter was not changed. Bone marrow stromal cells from C/EBPβ transgenics showed reduced mineralization, and reduced alkaline phosphatase mRNA levels. Calvarial osteoblasts from C/EBPβ transgenics displayed reduced alkaline phosphatase activity. To determine the consequences of the Cebpb deletion in vivo, the phenotype of Cebpb null mice was compared to that of wild type controls of identical genetic composition. Cebpb null mice exhibited reduced weight, body fat, and bone mineral density, and decreased bone volume, due to a decrease in trabecular number. The number of osteoblasts and osteoclasts per bone perimeter was not changed. C/EBPβ down regulation by RNA interference in calvarial osteoblasts had no effect on osteoblast differentiation/function. The phenotype of the Cebpb inactivation may be secondary to systemic indirect effects, and to direct effects of C/EBPβ in osteoblasts. In conclusion, C/EBPβ plays a role in mesenchymal cell differentiation and its misexpression in vivo causes osteopenia.
CCAAT/Enhancer binding proteins (C/EBPs) are a family of transcription factors that play a role in cell proliferation, differentiation and apoptosis (Nerlov 2007). Six members of the C/EBP family have been characterized and are termed α, β, δ, γ, ε, and ζ, which is also known as C/EBP homologous protein (CHOP) (Ramji & Foka 2002). C/EBP proteins contain a highly conserved leucine-zipper DNA-binding domain and a leucine dimerization domain that enables the formation of homo- and hetero-dimers, which bind to similar DNA sequence motifs. C/EBPs act as regulators of gene expression either by direct DNA binding or by interacting with other transcription factors, including peroxisome proliferator-activated receptor-γ2, retinoic acid receptor α, runt-related transcription factor-2 (Runx-2), and nuclear factor of activated T-cells (Clarke et al. 1997; Yang & Chow 2003; Gutierrez et al. 2002; Nerlov 2007). C/EBPs are expressed in multiple cell types, including osteoblasts and adipocytes, and C/EBPβ and δ are required for adipogenesis (Wiper-Bergeron et al. 2007; Rosen et al. 2000; Hata et al. 2005; Pereira et al. 2002).
C/EBPβ exists in three different isoforms, which are translated from different initiation sites within the same mRNA. All three isoforms share the same DNA-binding domain; however, differences in the length of the amino terminal end, determine different functions (Calkhoven et al. 2000). The full length isoform, termed liver-enriched activating protein* or LAP*, consists of 296 amino acids and acts as a transcriptional activator; it is commonly referred to as C/EBPβ. A second isoform, termed liver-enriched activating protein (LAP), lacks the first 21 amino acids present in the amino terminal end of full length LAP*, and it is generated from a different translation start site. LAP and LAP* are functionally identical. A third isoform, termed liver-enriched inhibitory protein (LIP), consists of 144 amino acids and acts as a dominant negative inhibitor of both LAP* and LAP, since it lacks transactivating activity but retains the ability to bind DNA and to dimerize with the other C/EBPβ isoforms (Fig. 1) (Nerlov 2007). The function of C/EBPβ is regulated by acetylation and phosphorylation at specific residues (Cesena et al. 2007; Tang et al. 2005; Li et al. 2007). Acetylation at Lys-39 of murine C/EBPβ is necessary to activate the transcription of C-fos and Cebpa, whereas sequential phosphorylation is required for the acquisition of C/EBPβ DNA binding activity and transactivation of target genes, such as Cebpa and activating protein 2 during adipogenesis in vitro (Cesena et al. 2007; Li et al. 2007).
Cebpb null mutant mice are viable, possibly because of partial rescue from other members of the C/EBP family, or because C/EBPβ is dispensable during development. Cebpb null mice are affected by a lymphoproliferative disorder, imbalanced lymphocyte helper response, and impaired lipid metabolism in brown fat tissue leading to defective thermoregulation (Screpanti et al. 1995; Tanaka et al. 1997). Cebpb null mice exhibit impaired regenerative capacity of hepatocytes, impaired liver function (Greenbaum et al. 1998; Croniger et al. 2001), and decreased levels of systemic insulin-like growth factor (IGF) I (Staiger et al. 2008).
Mesenchymal cells are multipotent cells that can differentiate into cells of various lineages, including osteoblasts, adipocytes and chondrocytes (Bianco & Gehron 2000). The fate of mesenchymal cells and their differentiation toward cells of the osteoblastic lineage is controlled by a network of intracellular and extracellular signals, and C/EBPβ regulates the commitment of cells of the osteoblastic and adipocytic lineage (Canalis et al. 2005; Franceschi et al. 2007; Winslow et al. 2006; Hata et al. 2005). Although the role of C/EBPβ in adipocyte differentiation is well known, its effects on osteoblastogenesis are less clear (Rosen et al. 2000). Previous studies from our laboratory have shown that cortisol inhibits osteoblastogenesis and induces the expression of adipocytic markers, which are associated with an increase in C/EBPβ expression in ST-2 stromal cells (Pereira et al. 2002). C/EBPβ represses Runx-2 transcription and osteoblastic differentiation of MC3T3 cells; however, C/EBPβ interacts with Runx-2 to induce osteocalcin expression (Wiper-Bergeron et al. 2007; Iyer et al. 2004; Gutierrez et al. 2002; Hata et al. 2005).
The intent of this study was to define the function of C/EBPβ in skeletal tissue in vivo. For this purpose, we created transgenic mice overexpressing C/EBPβ under the control of a 3.8 kilobase (kb) fragment of the osteoblast specific human osteocalcin promoter, and determined their body composition and skeletal phenotype. Cebpb null mice were also studied, and their body composition and skeletal phenotype determined.
After introduction of an optimal Kozak consensus sequence upstream of the translation initiation codon, an 890 base pair (bp) fragment, coding for full length murine C/EBPβ or LAP* (S.L. McKnight, Department of Biochemistry, University of Texas Southwestern Medical Center, Dallas, Texas), was cloned downstream of a 182 bp artificial intron and a 3.8 kb fragment of the human osteocalcin promoter (E. Gardiner, University of Queensland, Brisbane, Australia), and upstream of polyadenylation sequences and a 3.5 kb fragment of the 3′ untranslated region and flanking DNA of the Osteocalcin gene (Fig. 1) (Cao et al. 1991; Sims et al. 1997). Nucleotide sequence analysis confirmed the absence of mutations and correct orientation of the construct. Microinjection of linearized DNA into pronuclei of fertilized oocytes from FVB (for tropism to Friend Leukemia Virus Strain B) inbred mice, and transfer of microinjected fertilized eggs into pseudopregnant FVB mice were carried out at the transgenic facility of the University of Connecticut Health Center (Farmington, CT). Positive founders were identified by Southern blot analysis of tail DNA (Irwin 1989). Founder mice were bred to wild type FVB mice to generate transgenic lines, and their phenotype was examined.
To study the effects of the Cebpb deletion in vivo, Cebpb null heterozygous mice (Cebpb+/−) (S. Akira, Osaka University, Osaka, Japan) were re-derived in a C57BL/6 background, and backcrossed 5 times to obtain an uniform genetic background (Tanaka et al. 1995). Cebpb null homozygous mice (Cebpb−/−) were obtained by intermating Cebpb+/− mice, and their phenotype was examined. All animal experiments were approved by the Animal Care and Use Committee of Saint Francis Hospital and Medical Center.
Radiography was either performed in C/EBPβ transgenic and wild type littermate controls or in Cebpb−/− mice and, age and sex, matched wild type controls of the same genetic composition. The mice were anesthetized with tribromoethanol (Sigma-Aldrich, St. Louis, MO), and the X-rays were performed at an intensity of 35 kW for 25 seconds on a Faxitron X-ray system (model MX 20, Faxitron X-Ray Corp., Wheeling, IL). Total BMD (g/cm2) and total body fat (g) were measured on anesthetized mice using the PIXImus small animal DEXA system (GE Medical System/LUNAR, Madison, WI) (Nagy et al. 2001). Calibrations were performed with a phantom of defined value, and quality assurance measurements were performed before each use. The coefficient of variation for total BMD is less than 1% (n = 9).
Static and dynamic histomorphometry was carried out on transgenic mice and wild type littermate controls and Cebpb−/− mice and wild type, sex and age, matched controls of the same genetic composition. Mice were injected with calcein, 20 mg/kg, and demeclocycline, 50 mg/kg, at an interval of 2 days for 1 and 1.5 month animals and 7 days for 3 month old animals. The mice were sacrificed by CO2 inhalation 2 days after the demeclocycline injection. Longitudinal sections, 5 μm thick, were cut on a microtome (Microm, Richards-Allan Scientific, Kalamazoo, MI) and stained with 0.1% toluidine blue or von Kossa. Static parameters of bone formation and resorption were measured in a defined area between 181 μm and 1080 μm from the growth plate, using an OsteoMeasure morphometry system (Osteometrics, Atlanta, GA) (Gazzerro et al. 2005). Osteoclasts were identified as multinucleated cells on the trabecular surface and their identity confirmed by the expression of tartrate resistant acid phosphatase determined by histochemical analysis of selected cells, using a commercial kit according to manufacturer’s instructions (Sigma-Aldrich). For dynamic histomorphometry, mineralizing surface per bone surface and mineral apposition rate were measured on unstained sections under ultraviolet light, using a B-2A set long pass filter, and bone formation rate was calculated. The terminology and units used are those recommended by the Histomorphometry Nomenclature Committee of the American Society for Bone and Mineral Research (Parfitt et al. 1987).
The bone remodeling marker CTX was measured in serum by ELISA using RatLaps ELISA kits (Nordic Bioscience Diagnostics, Herlev, Denmark), according to manufacturer’s instructions.
1 month old C/EBPβ heterozygous transgenic and wild type littermate controls from both sexes were sacrificed by CO2 asphyxiation, the femurs were aseptically removed, and used to recover bone marrow stromal cells by centrifugation, as described (Gazzerro et al. 2005). Cells were plated at a density of 5 × 105 cells/cm2 and cultured in minimum essential medium (α-MEM, Invitrogen, Carlsbad, CA) containing 15% fetal bovine serum (FBS, Atlanta Biologicals, Norcross, GA) at 37°C in a humidified 5% CO2 incubator. Confluent cells were switched to α-MEM containing 10% FBS in the presence of 100 μg/ml ascorbic acid and 5 mM β-glycerophosphate and cultured up to 2 weeks to assess changes in mineralized nodule formation and gene expression.
Osteoblasts were isolated from parietal bones of 3- to 5-day old C/EBPβ heterozygous transgenic and wild type control littermates from both sexes, by sequential collagenase digestions, as previously described (McCarthy et al. 1990). Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Invitrogen) supplemented with non-essential amino acids, 20 mM HEPES, 100 μg/ml ascorbic acid and 10% FBS. Osteoblasts were cultured to 70% confluence and trypsinized; these first passage cells were used for subsequent experiments. Cells were cultured for up to 10 days after confluence, in the presence of 100 μg/ml ascorbic acid and 5 mM β-glycerophosphate (Sigma-Aldrich) to assess changes in gene expression, and treated with bone morphogenetic protein-2 (BMP-2, Wyeth Research, Collegeville, PA) or control vehicle to determine alkaline phosphatase activity (APA).
To down regulate C/EBPβ expression in vitro, a 19-mer double stranded small interfering (si) RNA targeted to the murine C/EBPβ mRNA sequence was obtained commercially, and a scrambled 19-mer siRNA with no homology to known mouse or rat sequences was used as a control (Applied Biosystems, Foster City, CA) (Sharp 2001; Elbashir et al. 2001). C/EBPβ or scrambled siRNA, both at 20 nM, were transfected into sub-confluent primary murine osteoblasts using siLentFect lipid reagent, in accordance with manufacturer’s instructions (BioRad, Hercules, CA). Cells were allowed to recover for 24 h prior to the determination of APA. To ensure adequate C/EBPβ down regulation, and to measure mRNA levels of osteoblast markers, total RNA was extracted in parallel cell cultures 96 h after the transfection of siRNAs, and mRNA levels determined by real time reverse transcription-polymerase chain reaction (RT-PCR).
To determine mineralized nodule formation, bone marrow stromal cells were fixed with 3.7% formaldehyde and stained with 2% Alizarin Red (Sigma-Aldrich) (DAHL 1952). APA was determined in 0.5% Triton X-100 cell extracts by the hydrolysis of p-nitrophenyl phosphate to p-nitrophenol measured by spectroscopy at 405 nm after 10 min of incubation at room temperature according to manufacturer’s instructions (Sigma-Aldrich). Data are expressed as picomoles of p-nitrophenol released per minute per icrogram of protein. The total protein content was determined in cell extracts by the DC protein assay, in accordance with manufacturer’s instructions (Bio-Rad).
Total RNA was extracted from cell cultures and mRNA levels determined by real time RT-PCR (Nazarenko et al. 2002b; Nazarenko et al. 2002a). For this purpose, 5 μg of RNA were reverse-transcribed using SuperScript III Platinum Two-Step qRT-PCR kit (Invitrogen), according to manufacturer’s instructions, and amplified in the presence of CGGTTAGGGCGTCTCCACAGTAAC[FAM]G-3′ and 5′-CTTGGAGAGGGCCACAAAGG-3′ primers for alkaline phosphatase; 5′-CGGATTGAACCGAGACAACTACATA[FAM]G-3′ and 5′-CGCCACCTACAATCAGGCTCT-3′ primers for C/EBPβ; 5′-CGAAGTTACATGACACTGGGCTT[FAM]G-3′ and 5′-CCCAGCACAACTCCTCCCTA-3′ primers for osteocalcin; 5′-CGAACCGGATAATGTGAAGTTCAAGGTT[FAM]G-3′ and 5′-CTGCTTCAGCTTCTCTGCCTTT-3′ primers for ribosomal protein L38 (RPL38), and Platinum Quantitative PCR SuperMix-UDG (Invitrogen) at 60°C for 45 cycles. Transcript copy number was estimated by comparison with a standard curve constructed using C/EBPβ (S.L. McKnight), osteocalcin (J.B. Lian, University of Massachusetts, Worcester, MA) (Lian et al. 1989), and alkaline phosphatase (American Type Culture Collection; ATCC, Manassas, VA) cDNAs, corrected for RPL38 copy number (ATCC) (Kouadjo et al. 2007). Reactions were conducted in a 96-well spectrofluorometric thermal iCycler (Bio-Rad), and fluorescence was monitored during every PCR cycle at the annealing step.
Data are expressed as means ± SEM. Statistical differences were determined by Student’s t test or ANOVA.
Two lines of FVB mice overexpressing C/EBPβ under the control of the 3.8 kb human osteocalcin promoter were established. The transgenic lines expressed 8 and 16 copies of the transgene. Both lines had a similar phenotype and the line expressing 16 copies of the transgene was characterized in detail. C/EBPβ transgenic mice were compared to wild type littermate controls of identical sex, at the age of 1 month, a time of high activity of the osteocalcin promoter, or at the age of 3 months (Frenkel et al. 1997). In the offspring of C/EBPβ heterozygous mice intermatings, the ratio of wild type, C/EBPβ heterozygous, and C/EBPβ homozygous was 29:60:11 for females and 32:63:5 for males. As a consequence, the number of C/EBPβ homozygous mice was limited, and this only allowed a detailed histomorphometric analysis of homozygous female mice. We suspect that the non Mendelian distribution of the transgenic allele is due to intrauterine lethality, which in turn is most likely because of transgene expression in the central nervous system, where the osteocalcin promoter is also active (Frenkel et al. 1997). There was no significant difference in weight and total body fat in either 1 or 3 month old C/EBPβ transgenic mice, when compared to wild type littermate controls (Table 1). Contact radiography did not reveal any obvious skeletal abnormality, and there was no difference in BMD between transgenics and controls (Table 1).
Bone histomorphometric analysis of distal femurs from 1 month old female C/EBPβ transgenics revealed a decrease in bone volume/tissue volume, secondary to a reduction in the number of trabeculae (Table 2, Fig. 2A). A similar decrease in bone volume was observed in male heterozygous transgenics (Table 3, Fig. 2B). The number of osteoblasts per tissue area was decreased in C/EBPβ transgenics, possibly due to the decreased bone volume, since the number of osteoblasts per bone perimeter, and osteoblast surface/bone surface were not different from controls (Tables 2 and and3).3). Changes in bone volume were not associated with an increase in bone resorption since the number of osteoclasts per bone perimeter and eroded surface were not increased in transgenics when compared to controls (Tables 2 and and3).3). A modest decrease in the number of osteoclasts per tissue area was detected in female transgenics (Table 2). Serum concentrations of CTX (means ± SEM; n = 3–6) were 26.5 ± 3.1 ng/ml in female heterozygous C/EBPβ transgenic mice and 33.3 ± 1.9 ng/ml in wild type littermate controls; 30.7 ± 5.6 ng/ml in female homozygous transgenics and 32.3 ± 4.0 ng/ml in wild type littermate controls; and 49.4 ± 15.5 ng/ml in male heterozygous transgenics and 50.4 ± 7.4 ng/ml in wild type littermate controls. The results indicate that C/EBPβ overexpression does not affect bone remodeling. There were no substantial differences between female and male phenotypes, and no change in mineral apposition rate (not shown) or bone formation rate (Fig. 2) was detected in C/EBPβ transgenics. No significant change in the number of adipocytes per tissue area was observed in either female or male transgenics when compared to wild type littermate controls (Tables 2 and and3).3). This could be expected since C/EBPβ transcription was directed by the osteocalcin promoter, which is not active in adipocytic cells.
Bone histomorphometric analysis of distal femurs from 3 month old female C/EBPβ heterozygous transgenics revealed a phenotype analogous to that observed at 1 month, and characterized by a decrease in bone volume/tissue volume, due to a reduction in the number of bone trabeculae (Table 4). Bone formation rate and mineral apposition rate were not affected (data not shown). Serum concentrations of CTX (means ± SEM; n = 4) were 17.8 ± 1.9 ng/ml in female heterozygous C/EBPβ transgenic mice and 16.3 ± 1.1 ng/ml in wild type littermate controls, confirming that C/EBPβ overexpression does not affect bone remodeling.
Both transgenic lines studied displayed analogous skeletal phenotypes. In accordance with the line described, 1 month old male C/EBPβ heterozygous transgenics, from the line expressing 8 copies of the transgene, exhibited a decrease in bone volume. Bone volume/tissue volume was (means ± SEM; n = 6 to 8) 8.0 ± 0.4% in controls and 5.3 ± 0.5% in transgenics, p < 0.05; trabecular number was 7.9 ± 0.3 mm−1 in controls and 5.9 ± 0.5 mm−1 in transgenics, p < 0.05. In summary, C/EBPβ transgenics exhibited osteopenia secondary to a decrease in trabecular number without apparent changes in bone formation or bone resorption.
To investigate the impact of C/EBPβ on osteoblastic cell function, bone marrow stromal cells from heterozygous transgenic mice and wild type littermate controls were cultured for up to two weeks after confluence, under conditions favoring osteoblastogenesis. C/EBPβ overexpression was confirmed by real time RT-PCR, and C/EBPβ transcripts were 2 fold greater in transgenic than in control cells (Fig. 3A). Alkaline phosphatase mRNA levels were suppressed by about 40% in cells from transgenics, when compared to control cells (Fig. 3A). Accordingly, mineralization of the culture, tested after 14 days of incubation by alizarin red staining, was reduced in C/EBPβ overexpressing cells, confirming an inhibitory effect of C/EBPβ on osteoblastic differentiation (Fig. 3B). Conversely, osteocalcin mRNA levels were increased in C/EBPβ overexpressing cells (Fig. 3A). This may be explained by observations demonstrating that interactions between Runx-2 and C/EBPβ can increase osteocalcin expression (Gutierrez et al. 2002; Hata et al. 2005).
To confirm that C/EBPβ inhibits osteoblast differentiation, calvarial osteoblasts were harvested from heterozygous transgenic mice and wild type littermate controls, and cultured for up to 10 days in differentiating conditions. C/EBPβ overexpression was confirmed by real time RT-PCR, and C/EBPβ transcripts at confluence were 1.5 fold greater in transgenic than in control cells (Fig. 3C). Alkaline phosphatase mRNA levels were suppressed by about 50% in cells from transgenics when compared to control cells (Fig. 3C). Accordingly, calvarial osteoblasts overexpressing C/EBPβ exhibited decreased levels of APA when compared to control cells cultured with vehicle or BMP-2 (Fig. 3D). Osteocalcin transcript levels did not change in osteoblasts from transgenics when compared to control cells (not shown), possibly because the inhibition of differentiated osteoblast function precluded the stimulatory effect of C/EBPβ and Runx-2 on osteocalcin transcription in mature osteoblasts.
Cebpb+/− mice were intermated to obtain Cebpb−/− mice, and their phenotype was examined. The homozygous Cebpb deletion causes female infertility, not allowing the establishment of a Cebpb−/− colony, and limiting the availability of mice for subsequent studies (Sterneck et al. 1997). Male Cebpb−/− mice showed a tendency toward a reduced body weight and total body fat as compared to wild type age matched controls of identical genetic composition at 1.5 and 3 months. The difference in weight and body fat was statistically significant at 6 months of age (Fig. 4A, B). Contact radiography of male Cebpb−/− mice did not reveal obvious skeletal abnormalities (data not shown). However, male Cebpb−/− mice at 1.5, 3 and 6 months displayed a tendency toward a decrease in BMD, which was statistically significant at 1.5 months of age (Fig. 4C). Male Cebpb−/− mice at 1.5, 3 and 6 months exhibited decreased trabecular bone volume that was statistically significant at 3 months of age (Fig. 4D).
A complete histomorphometric analysis, performed in 3 month old Cebpb−/− male mice, revealed a 30% decrease in bone volume/tissue volume, when compared to wild type age matched controls of the same genetic composition (Table 5, Fig. 5). The decrease was secondary to a reduced number of trabeculae. Cebpb−/− male mice did not exhibit differences in the number of osteoblasts per bone perimeter and osteoblast surface/bone surface. Similarly, there were no significant changes in the number of osteoclasts/perimeter or in eroded surface in Cebpb null mice, when compared to wild type controls. There was no change in the number of adipocytes per tissue area possibly because C/EBPδ alone is sufficient to support adipogenesis (Table 5) (Lane et al. 1999). No changes in mineral apposition rate (not shown) and bone formation rate (Fig. 5) were detected. In summary, Cebpb null mice are osteopenic secondary to a decrease in trabecular number, but do not exhibit changes in bone formation or bone resorption.
Skeletal cells from Cebpb−/− mice were not readily available because of the low fertility of Cebpb+/− mice and infertility of Cebpb−/− mice. To investigate the function of C/EBPβ, its mRNA levels were down regulated by RNAi in primary calvarial osteoblasts. C/EBPβ down regulation was confirmed by real time RT-PCR, and C/EBPβ transcripts were reduced by 70% in cells transfected with the C/EBPβ siRNA, when compared to control cells transfected with scrambled siRNA (Table 6). There were no changes in alkaline phosphatase or osteocalcin mRNA levels, measured by real time RT-PCR, in C/EBPβ down regulated cells (Table 6). Accordingly, there was no effect on APA in cells transfected with C/EBPβ siRNA when compared to control cells (Table 6). These results indicate that C/EBPβ may be dispensable for osteoblast differentiation in vitro, and suggest that the osteopenic phenotype of the null mutation could be secondary to indirect mechanisms.
This report demonstrates that transgenic mice overexpressing C/EBPβ under the control of the 3.8 kb human osteocalcin promoter exhibit osteopenia. The osteopenia was secondary to a reduced number of trabeculae, but C/EBPβ transgenics exhibited no change in osteoblast number or bone formation rate. The osteopenic phenotype could not be attributed to changes in bone resorption, since osteoclast number, eroded surface and biochemical markers of bone remodeling were not different between transgenics and controls. The reduced number of trabeculae could be the result of impaired trabecular formation during skeletal development. This is possible since the osteocalcin promoter is expressed by mature osteoblasts as early as day 18 of embryonic life (Kalajzic et al. 2002; Frendo et al. 1998), and C/EBPβ overexpression can suppress the terminal differentiation of osteoblasts (Iyer et al. 2004). A possible explanation for a developmental impact of C/EBPβ could be a suppression of Runx-2 transcription, and, as a consequence, impaired trabecular bone formation (Otto et al. 1997; Wiper-Bergeron et al. 2007). It is possible that osteoblastic differentiation and function are more sensitive to transcriptional changes during development, or that compensatory mechanisms are established postnatally, since osteoblast number and bone formation were not affected in 1 month old C/EBPβ transgenics.
Cebpb null mice showed reduced weight, total body fat and total BMD. Although they were osteopenic, no changes in osteoblast or osteoclast number or bone formation rate were observed. The osteopenic phenotype of Cebpb null mice is in accordance with recent reports from other investigators (Staiger et al. 2008; Tominaga et al. 2008), and with a function of C/EBPβ in the early differentiation stages of mesenchymal cells, resulting in impaired chondrocyte and osteoblast maturation. This leads to impaired cartilage and bone development, and would explain the reduced number of trabeculae and osteopenic phenotype (Tominaga et al. 2008). However, other factors probably contribute to the bone loss caused by the Cebpb deletion, since C/EBPβ regulates gene expression in a variety of organs (Ramji & Foka 2002). Cebpb null mice have decreased serum levels of IGF-I, secondary to decreased IGF-I expression by the liver (Staiger et al. 2008), and systemic IGF-I plays an important role in skeletal homeostasis (Gazzerro & Canalis 2006; Giustina et al. 2008; Yakar et al. 2006). Other contributing factors to the osteopenic phenotype could be the consequences of lymphoproliferative and myeloproliferative alterations, and increased Interleukin-6 serum levels exhibited in the Cebpb null state (Screpanti et al. 1995). Uncontrolled proliferation of lymphocytes or myeloid precursors can lead to osteopenia (Roodman 1997), and overexpression of Interleukin-6 in vivo causes impaired growth plate development, increased bone resorption, and decreased osteoblastic differentiation (De et al. 2006). A systemic nature of the osteopenic phenotype, as opposed to an osteoblast specific effect, is supported by the absence of a cellular phenotype following the down regulation of C/EBPβ by RNAi in osteoblasts. However, the absence of a cellular phenotype is in contrast with recently reported findings in osteoblasts from Cebpb null mutants, where inactivation of Cebpb resulted in impaired osteoblast maturation (Tominaga et al. 2008). The discrepancy between our results and those of Tominaga et al., could be attributed to the transient and possibly incomplete down regulation of C/EBPβ by RNAi, causing sufficient C/EBPβ expression to support osteoblast differentiation/function for the period of study. It is also possible that under the experimental conditions used in the present work, the levels of C/EBPβ were sufficient to maintain osteocalcin expression (Gutierrez et al. 2002). C/EBPβ can favor the early stages of osteoblast differentiation through its hetero-dimerization with activating transcription factor (ATF) 4, and by enhancing the association of ATF 4 with Runx-2 to regulate osteocalcin transcription (Hata et al. 2005; Tominaga et al. 2008). Interactions of C/EBPβ with ATF 4 may affect other steps of osteoblastic differentiation, since ATF 4 plays an essential role in this process (Yang et al. 2004), and could also explain the findings reported in Cebpb null osteoblasts and in the Cebpb null mice.
Evidence for direct activity of C/EBPβ in osteoblasts also can be derived from observations in vivo. Transgenics overexpressing either LIP, a physiological dominant negative form of C/EBPβ (Harrison et al. 2005), or CHOP, a member of the C/EBP family, are osteopenic (Pereira et al. 2007). The primary partner of CHOP is C/EBPβ and the osteopenia observed in CHOP transgenics is likely secondary to its dimerization with C/EBPβ and a decrease in C/EBPβ transactivating activity in the skeletal environment (Ron & Habener 1992; Vinson et al. 2002). It is of interest that inactivation of Chop causes decreased bone formation, but not an obvious osteopenic phenotype. The function of CHOP, like that of C/EBPβ, appears to be dependent on cell type, stage of mesenchymal/osteoblastic cell differentiation, and experimental conditions (Pereira et al. 2004; Pereira et al. 2006; Shirakawa et al. 2006)
To understand the mechanism leading to the osteopenic phenotype resulting from C/EBPβ overexpression, we examined primary osteoblasts and bone marrow stromal cells harvested from C/EBPβ transgenic mice. The cellular phenotype is in accordance with the osteopenic phenotype observed, and with previous work demonstrating that the stable transduction of C/EBPβ in MC3T3 and C3H10T1/2 cells, by repressing Runx-2 transcription, abrogates osteoblast differentiation induced by retinoic acid (Wiper-Bergeron et al. 2007). Accordingly, osteoblastic cells from C/EBPβ transgenics exhibited reduced mineralization, APA, and alkaline phosphatase mRNA expression. In contrast, transduction of cells of the osteoblastic lineage with adenoviral vectors expressing C/EBPβ resulted in enhanced osteoblastic differentiation (Hata et al. 2005), also suggesting that experimental conditions can have a significant impact on the effect of C/EBPs in osteoblasts. C/EBPβ/ATF 4/Runx-2 interactions enhance osteocalcin transcription, and osteocalcin transcripts are increased in conditions of C/EBPβ overexpression (Gutierrez et al. 2002; Tominaga et al. 2008). Accordingly, osteocalcin mRNA levels were increased in bone marrow stromal cell cultures from C/EBPβ transgenics. However, they were not changed in calvarial osteoblasts, possibly due to the inhibitory effect of C/EBPβ on Runx-2 expression and on terminal osteoblast differentiation (Iyer et al. 2004; Wiper-Bergeron et al. 2007).
C/EBPβ is involved in cell fate decisions and regulates the balance between osteoblasts and adipocytes in vitro, and the Cebpb deletion in vivo impairs the formation of white and brown adipose tissue (Hata et al. 2005; Tanaka et al. 1997). Accordingly, male Cebpb null mice exhibited reduced total body fat. The Cebpb deletion did not affect the number of bone marrow adipocytes, possibly because C/EBPδ is sufficient to sustain adipogenesis in the absence of C/EBPβ (Lane et al. 1999). However, the effect of the Cebpb deletion on the differentiation of bone marrow adipocytes in vivo is not known, and it is conceivable that genes regulating the development of brown and white adipose tissue could be different from genes regulating bone marrow adipogenesis. We did not detect an effect on total body fat or in the number of adipocytes in C/EBPβ transgenics. This should be expected, since C/EBPβ was expressed under the control of the osteocalcin promoter, which is expressed by mature osteoblasts and not by undifferentiated mesenchymal cells capable of adipocytic differentiation.
In conclusion, C/EBPβ plays an important regulatory role in osteoblast differentiation and skeletal function in vivo.
This work was supported by Grant DK042424 from the National Institute of Diabetes and Digestive and Kidney Diseases.
The authors thank S. Akira for Cebpb heterozygous null mice, S.L. McKnight for C/EBPβ cDNA, J.B. Lian for osteocalcin cDNA, E. Gardiner for human osteocalcin promoter construct, Melissa Burton, Trung X. Le and Kristen McBreairty for technical assistance, and Mary Yurczak for secretarial assistance.
Publisher's Disclaimer: Disclaimer. This is not the definitive version of record of this article. This manuscript has been accepted for publication in the Journal of Endocrinology, but the version presented here has not yet been copy edited, formatted or proofed. Consequently, the Society of Endocrinology accepts no responsibility for any errors or omissions it may contain. The definitive version is now freely available at http://dx.doi.org/10.1677/JOE-08-0514
DECLARATION OF INTEREST
The authors declare that there is no conflict of interest that could be perceived as prejudicing theimpartiality of the research reported.