Developmental regulation of osteocalcin expression in MC3T3-E1 osteoblasts: minimal role of the proximal E-box cis-acting promoter elements.

Osteoblasts undergo a temporal sequence of development characterized by transcriptional upregulation of osteoblast-specific genes. Basic helix-loop-helix (bHLH) transcription factors may control this developmental process through binding to E-box cis-acting elements in developmentally regulated genes. To investigate the role of bHLH proteins in MC3T3-E1 osteoblasts, which undergo a developmental sequence in vitro, we analyzed the transcriptional control of osteocalcin gene expression by stable transfection of an osteocalcin promoter-luciferase chimeric gene (p637OC-luc) and assessed the role of E-box cis-acting elements in osteocalcin promoter by DNA binding assays. We compared our findings in MC3T3-E1 osteoblasts with transient DNA transfections and DNA binding experiments in Ros 17/2.8 osteoblasts. We found that the activity of 637-OC luciferase promoter was low in undifferentiated 5-day-old cultures but increased in parallel with endogenous osteocalcin message expression in mature MC3T3-E1 osteoblasts, consistent with developmental stage-specific transcriptional upregulation of the osteocalcin gene. We identified two putative E-box elements in the proximal osteocalcin promoter, E-box 1 (CACATG) at -102 and E-box 2 (CAGCTG) at position -149. In gel mobility shift assays, factors present in nuclear extracts derived from differentiated osteoblast bound to oligonucleotide probes containing the E-box 1 and E-box 2 elements. Binding to the E-box 2 probe was not specific for the core CAGCTG element, whereas the CACATG site in E-box 1 oligonucleotide was required for specific binding of these nuclear factors. Stable transfection of p637OC-luc containing a mutant E1 site (p637OC-luc E1m), however, did not alter the developmental upregulation of osteocalcin promoter activity in MC3T3-E1 osteoblasts. Moreover, the E-box 1 mutation had no effect on either basal or vitamin D-stimulated activity of the osteocalcin promoter in Ros 17/2.8 osteoblasts in transient transfection experiments. These data suggest that osteoblasts contain underfined factors that bind to the E-box 1 CACATG site in the proximal osteocalcin promoter; however, this E-box element does not play a significant role in the developmental stage-specific regulation of the osteocalcin gene in MC3T3-E1 osteoblasts.

A distinct cation-sensing mechanism in MC3T3-E1 osteoblasts functionally related to the calcium receptor.

The presence of a cation-sensing mechanism in osteoblasts is suggested by the ability of specific cations to stimulate osteoblastic proliferation in culture and to induce de novo bone formation in some experimental models. Our study examines whether extracellular cations stimulate osteoblasts through the recently identified G protein-coupled calcium receptor (CaR). We found that CaR agonists, calcium (Ca2+), gadolinium (Gd3+), aluminum (Al3+), and neomycin, stimulated DNA synthesis in murine-derived MC3T3-E1 preosteoblasts, whereas magnesium (Mg2+), nickel (Ni2+), cadmium (Cd2+), and zinc (Zn2+) had no effect. With the exception of Mg2+, the cation specificities and apparent affinities were similar to that reported for CaR. CaR agonists also stimulated DNA synthesis in C3HT10(1/2) fibroblasts, but not in mesangial PVG, CHO, hepatic HTC, COS-7 cells, or malignant transformed ROS17/2.8 and UMR-106 osteoblasts. In addition, similar to other growth factors, CaR agonists activated transcription of a serum response element luciferase reporter construct (SRE-Luc) stably transfected into MC3T3-E1 osteoblasts, but had no effect on SRE-Luc transfected into CHO and COS-7 cells. We were unable to detect CaR expression by Northern analysis using a mouse CaR-specific probe or to amplify CaR mRNA by reverse transcribed polymerase chain reaction in MC3T3-E1 osteoblasts. These findings suggest that an extra-cellular cation-sensing mechanism is present in murine-derived osteoblasts that is functionally similar to but molecularly distinct from CaR.

Nonmyristoylated MARCKS complements some but not all of the developmental defects associated with MARCKS deficiency in mice.

The myristoylated alanine-rich C kinase substrate, or MARCKS protein, is a widely expressed, prominent substrate for protein kinase C. Although the exact function of MARCKS has not been elucidated, targeted disruption of the MARCKS gene (Macs) in mice has shown that MARCKS plays a crucial role in the development of the central nervous system. Mice deficient in MARCKS exhibited universal perinatal death with defects in neurulation, fusion of the cerebral hemispheres, formation of the great forebrain commissures, and retinal and cortical lamination (Stumpo et al., Proc. Natl. Acad. Sci. USA 92, 944-948, 1995). In the present studies, a transgene consisting of approximately 3.4 kb of promoter from the human MARCKS gene (MACS), with an epitope tag sequence inserted at the carboxyl terminus of the MARCKS coding region, was able to complement completely MARCKS deficiency in mice. Thus, the human transgene contained all of the elements necessary for normal developmental expression of MARCKS. To test the importance of MARCKS myristoylation to its developmental role, an otherwise identical transgene was constructed in which the glycine at the amino terminus of MARCKS was mutated to an alanine. This mutation, which resulted in the expression of nonmyristoylated MARCKS, was successful in partially rescuing the Macs null phenotype. Specifically, about 25% of these mice survived the perinatal period; these survivors appeared to develop normally except for slightly decreased body size. In both the survivors and the nonsurvivors, all of the known anatomical defects associated with MARCKS deficiency were corrected by expression of the nonmyristoylated human protein. These results indicate that myristoylation of MARCKS is not required for the protein to correct many of the developmental abnormalities characteristic of its deficiency.

Differential regulation of receptor-stimulated cyclic adenosine monophosphate production by polyvalent cations in MC3T3-E1 osteoblasts.

Extracellular cations have paradoxical trophic and toxic effects on osteoblast function. In an effort to explain these divergent actions, we investigated in MC3T3-E1 osteoblasts if polyvalent cations differentially modulate the agonist-stimulated cyclic adenosine monophosphate (cAMP) pathway, an important regulator of osteoblastic function. We found that a panel of cations, including gadolinium, aluminum, calcium, and neomycin, inhibited prostaglandin E1 (PGE)-stimulated cAMP accumulation but paradoxically potentiated parathyroid hormone (PTH)-stimulated cAMP production. In contrast, these cations had no effect on forskolin- or cholera toxin-induced increases in cAMP, suggesting actions proximal to adenylate cyclase and possible modulation of receptor interactions with G proteins. Phorbol 12-myristate 13-acetated (PMA) mimicked the effects of cations on PGE1- and PTH-stimulated cAMP accumulation in MC3T3-E1 cells, respectively, diminishing and augmenting the responses. Moreover, down-regulation of protein kinase C (PKC) by overnight treatment with PMA prevented gadolinium (Gd3+) from attenuating PGE1- and enhancing PTH-stimulated cAMP production, indicating involvement of PKC-dependent pathways. Cations, however, activated signal transduction pathways not coupled to phosphatidylinositol-specific phospholipase C (PI-PLC), since there was no corresponding increase in inositol phosphate formation or intracellular calcium concentrations. In addition, pertussis toxin treatment failed to prevent Gd(3+)-mediated suppression of PGE1-stimulated cAMP, suggesting actions independent of Gm. Thus, polyvalent cations may either stimulate or inhibit hormone-mediated cAMP accumulation in osteoblasts. These differential actions provide a potential explanation for the paradoxical trophic and toxic effects of cations on osteoblast function that occur in vivo under different hormonal conditions.

Forskolin inhibits protein kinase C-induced mitogen activated protein kinase activity in MC3T3-E1 osteoblasts.

We recently demonstrated that stimulation of DNA synthesis in MC3T3-E1 osteoblasts involves cross-talk between protein kinase C (PKC)-dependent pathways and activation of possible nonreceptor tyrosine kinases. In the current investigation we examined whether the Raf-1/MAP kinase kinase (MKK)/mitogen-activated protein kinase (MAPK) cascade integrates cross-talk between G protein-coupled second messengers and protein tyrosine phosphorylation in osteoblasts. We investigated the effects on DNA synthesis, protein tyrosine phosphorylation, and Raf-1, MKK, and MAPK activities of PKC activation by phorbol 12-myristate 13-acetate (PMA) and of cAMP elevation by forskolin (FSK) in MC3T3-E1 osteoblasts. We found that PMA-stimulated DNA synthesis was associated with increments in tyrosine phosphorylation of p44mapk (ERK1) and p42mapk (ERK2) and activation of Raf-1, MKK, and MAPK in these cells. FSK treatment of osteoblasts, which raised intracellular cAMP levels and inhibited DNA synthesis, blocked PKC-stimulated tyrosine phosphorylation of p44mapk (ERK1) and p42mapk (ERK2) as well as inhibited PKC-stimulated MAPK and Raf-1 activities. Despite this, PMA activated the intermediate MKK step of the Raf-1/MKK/MAPK cascade in the presence of FSK. The differential inhibition of PMA-stimulated Raf-1 and MKK activities by FSK suggests that PKC activates both Raf-1-dependent and -independent pathways in MC3T3-E1 osteoblasts. Moreover, the noncoordinate effects of FSK on PMA-stimulated MKK and MAPK activities indicates the presence of a additional distal cAMP-dependent inhibitory mechanisms.

Differentiation of MC3T3-E1 osteoblasts is associated with temporal changes in the expression of IGF-I and IGFBPs.

We examined the relationship between osteoblast maturation and temporal changes in the secretion of IGF-I and the IGF-binding proteins (IGFBPs) in the MC3T3-E1 model of osteoblast development. IGF-I was present at low levels in conditioned media in proliferating preosteoblasts (3.7 +/- 1.7 ng/micrograms DNA and 3.9 +/- 0.6 at culture (days 3 and 9) and increased progressively in postmitotic differentiating osteoblasts, reaching a maximal concentration of 13.1 +/- 1.5 ng/micrograms DNA by day 25 of culture. We also observed an increase in IGF-I mRNA expression. Using Western ligand blot and immunoblot techniques, we found that IGFBP-2, -4, and -5 also displayed temporal differences in expression during MC3T3-E1 development. We observed a sustained increase in IGFBP-2, -4, and -5 mRNA expression between days 10-14, coincident with the onset of differentiation. IGFBP-2 and IGFBP-4 protein concentrations increased in parallel with IGFBP mRNA expression, but IGFBP-5 levels peaked between days 8-14 of culture, and declined thereafter in spite of persistent IGFBP-5 mRNA levels. These findings suggest complex transcriptional and post-transcriptional regulation of IGFBP metabolism during osteoblast development. Thus, IGF-I and IGFBP production are regulated during osteoblast development. In turn, time-dependent changes in IGF-I and modulation of IGF-I bioavailability by IGFBPs may regulate the osteoblastic developmental sequence.

Molecular to pharmacologic control of osteoblast proliferation and differentiation.

Control of osteoblast growth and development can be characterized from receptor mediated events to nuclear messengers controlling gene transcription. From this analysis it is possible to formulate a model to explain the reciprocal relationship between growth and differentiation as well as differential cytokine modulation of osteoblast function. Central to this model are putative tissue specific transcriptional switches (possibly of the bHLH gene superfamily) that may repress proliferation and permit the regulation of mature osteoblast phenotypic characteristics. This model proposes that in post-mitotic differentiated osteoblasts, tissue specific transcription factors determine the capacity to express osteoblastic characteristic, whereas receptor activated signalling cascades, namely, cAMP/protein kinase A, receptor serine/threonine kinase, and vitamin D receptor-dependent pathways, regulate mature osteoblast-specific gene expression. Activated differentiation switches also may feedback to transcriptionally repress proliferation. Conversely, in preosteoblasts, in which differentiation switches are turned off, distinct signalling cascades involving tyrosine kinases, PKC, and calcium/calmodulin regulate proliferation. Proliferating preosteoblasts also exhibit negative modulation of maturation either through inactivation of putative tissue-specific transcription factors and/or through AP-1 dependent phenotype suppression of genes expressed in mature osteoblast. Thus, the final outcome of transcriptional regulation of osteoblast function results from complex interactions between signalling pathways and permissive differentiating transcription factors. Though many aspects of this model remain speculative and require confirmation, it serves as a useful conceptual framework to further investigate the differential control of osteoblast proliferation and differentiation that may lead to improved pharmacologic ways to manipulate bone formation in vivo.

Okadaic acid-sensitive protein phosphatases dephosphorylate MARCKS, a major protein kinase C substrate.

The myristoylated alanine-rich C kinase substrate (MARCKS) undergoes a rapid and, in certain circumstances, transient increase in phosphorylation in response to stimuli that activate protein kinase C. We have investigated the protein-serine/threonine phosphatase activity responsible for reversing the phosphorylation of MARCKS. In cell-free assays, protein phosphatases 1, 2A and 2C (PP1, PP2A and PP2C) all dephosphorylate recombinant MARCKS or a synthetic peptide based on its phosphorylation site domain. In intact Swiss 3T3 cells, okadaic acid, a specific inhibitor of PP1 and PP2A, had little effect on MARCKS phosphorylation on its own, but largely prevented the dephosphorylation of MARCKS that occurred following activation of protein kinase C by bombesin with subsequent receptor blockade. These results indicate that although the dephosphorylation of MARCKS can be mediated by PP2C in vitro, this protein is dephosphorylated by okadaic acid-sensitive phosphatases in the intact cell.

Influence of dietary fat on factors in serum that regulate thyroid cell metabolism.

The type of dietary fat affected the production of cAMP by cultured thyroid cells incubated with mouse and rat sera. Greater amounts of cAMP were produced with serum from mice fed 30% rac-1(3)-palmitoyl glycerol and 4% safflower oil (PG + SO) than with serum from mice fed 30% rac-1(3)-palmitoyl glycerol (PG). The serum from mice fed PG + SO gave a response similar to that with calf serum. Sera were separated into lipoprotein and aqueous fractions by centrifugation. A combination of both the lipoprotein and aqueous fractions of serum from mice fed PG + SO was required for the increased response. Proteolysis of the aqueous fraction of the serum from either mice fed PG or rats fed 30% hydrogenated coconut oil (HCO) reduced the amount of cAMP produced by the thyroid cells. However, the same effect was not seen with aqueous fractions of calf serum or serum from mice fed PG + SO or rats fed HCO plus 5% corn oil. These findings suggest that there are at least three factors in serum capable of regulating thyroid cell metabolism that are controlled by the type of fat fed the animal.

Toxicity of palmitoyl glycerol to mice: hypothermia and reversal of the toxicity.

Adult and weanling mice kept at low ambient temperatures show an increased sensitivity to the toxicity of dietary rac-1(3)-palmitoyl glycerol. When fed the palmitoyl glycerol, mice less than 6 wk old show a pronounced hypothermia that is prevented by adding safflower oil to the diet. A more moderate degree of hypothermia is seen with older animals. Once body temperature fell below 28 degrees C, replacing the toxic monoacylglycerol with safflower oil and/or raising the environmental temperature to 34 degrees C did not reverse the ultimate fatality caused by palmitoyl glycerol ingestion. If hypothermia was between 28 and 32 degrees C, high mortality was not reversed by feeding the unsaturated fat or raising the environmental temperature to 34 degrees C. However, a combination of both treatments reduced the mortality. Irrespective of body temperature, the hypothermia was eliminated by the warm ambient temperature, but mortality was high. Thus, although hypothermia is a sign of the toxicity of rac-1(3)-palmitoyl glycerol, it is not the immediate cause of death.