Transcriptional regulation of the expression of ... - Semantic Scholar

Molecular and Cellular Endocrinology 160 (2000) 193 – 202 www.elsevier.com/locate/mce

Transcriptional regulation of the expression of macrophage colony stimulating factor J. Rubin a,*, D. Fan a, A. Wade a, T.C. Murphy a, H. Gewant a, M.S. Nanes a, X. Fan a, M. Moerenhout b, W. Hofstetter b a

Department of Medicine, Veterans Affairs Medical Center and Emory Uni6ersity School of Medicine, VAMC-151, 1670 Clairmont Road, Decatur, GA 30033, USA b Group for Bone Biology, Department Clinical Research, Uni6ersity of Bern, Bern, Switzerland Received 20 August 1999; accepted 8 November 1999

Abstract The regulatory regions for transcriptional control of the MCSF gene are unknown. We examined regulatory control in a 774-bp murine MCSF promoter transfected into MC3T3-E1 osteoblast-like and COS-7 cells. Deletion of upstream sequence from − 635 increased basal activity of the promoter by at least four-fold, an increase that was maintained when PU.1, NFkB and Egr1/Sp1 consensus sequences were subsequently removed. Mutagenesis identified a suppressor element between − 635 and − 642 from the transcriptional start site and an oligonucleotide representing this sequence was retarded by nuclear cell protein. TNFa (1 ng/ml), PTH (5 × 10 − 8 M), and IL-1a (100 pg/ml), which increased MCSF protein secretion, failed to enhance the transcriptional rate of the full-length promoter. TNFa was able to stimulate transcription of a heterologous reporter transfected into COS-7 containing multiple copies of the murine MCSF NFkB site inserted before a minimal promoter. In contrast, deletion of the same NFkB response element increased basal activity in the native promoter. Thus, the NFkB sequence may act as a negative regulator in the context of the endogenous promoter. Our results indicate that constitutive transcriptional activity conferred by the MCSF promoter may be damped by a suppressor protein. Transcriptional regulation, however, does not appear to be a major stimulatory mechanism for MCSF secretion. © 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Osteoclast; TNFa; Bone stromal cells; Macrophage colony-stimulating factor; Bone resorption

1. Introduction The growth factor macrophage colony-stimulating factor (MCSF) is an integral component in the recruitment and differentiation of osteoclast progenitors from the monocyte–macrophage lineage. A deficiency in MCSF leads to an osteopetrotic phenotype due to a virtual absence of osteoclasts (Felix et al., 1990a; Wiktor-Jedrzejczak et al., 1990). This phenotype can be reversed by reconstitution of the growth factor, demonstrating its necessity for the formation of functional osteoclasts (Felix et al., 1990b; Wiktor-Jedrzejczak et al., 1990; Kodama et al., 1993). The critical role of MCSF in the formation of osteoclasts in vivo is further * Corresponding author. Tel.: + 1-404-321-6111, ext. 2080; fax: +1-404-235-3011. E-mail address: [email protected] (J. Rubin)

supported by the close temporal and spatial association of the expression of the cytokine with the late step of osteoclastogenesis (Hofstetter et al., 1995), suggesting a tight regulation of the expression of this cytokine. Many skeletally active factors, such as 1,25-dihydroxyvitamin D3, parathyroid hormone and tumor necrosis factor-a (TNFa) cause increased MCSF mRNA and protein expression (Kaplan et al., 1996; Rubin et al., 1996; Weir et al., 1996; Yao et al., 1998). As little as 10 ng/ml MCSF can support survival of cells of macrophage lineage (Yasuda et al., 1998), while high amounts of MCSF may decrease the number of precursor cells that enter the osteoclast lineage (Perkins and Kling, 1995; Yasuda et al., 1998; Rubin et al., 1998). Thus, regulation of MCSF expression and secretion by stromal cells in bone should have implications for control both of bone resorption through recruitment of osteoclasts, and for recruitment of other cells in the macrophage lineage.

0303-7207/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 3 0 3 - 7 2 0 7 ( 9 9 ) 0 0 2 1 2 - 9

194

J. Rubin et al. / Molecular and Cellular Endocrinology 160 (2000) 193–202

The MCSF gene has been cloned and cis acting elements in the promoter have been identified by homology (Harrington et al., 1991). Induction of MCSF expression by TNFa has previously been assumed to be via activation of nuclear factor-kB (NFkB) and binding to a NFkB response element spanning nucleotides − 369– −378: (GGAAAGTCCC) (Harrington et al., 1991). Although we have shown that TNFa causes activation of this MCSF NFkB response element using EMSA in the ST2 murine bone stromal cell line, we have not found this element to be a functional enhancer in the context of the murine MCSF gene (Isaacs et al., 1999). This puzzling result led us to examine further the regulatory control over the murine MCSF promoter. The sequence examined in this work lies between − 774 and the transcription start site and includes recognition sites for NFkB, AP1, IL-6, Sp1/Egr1 and PU.1. We show in the work presented here that this span of the murine MCSF promoter transiently transfected into COS-7 cells behaves similarly as it does in MC3T3-E1 murine osteoblast like cells (Sudo et al., 1983). Surprisingly, hormones and cytokines, among them TNFa, known to increase MCSF mRNA, did not enhance the basal transcriptional rate of this promoter. In fact, mutagenesis showed that removal of the NFkB response element increased basal activity in COS-7 cells and MC3T3-E1 cells. Deletion analysis of the promoter region showed that promoters containing 200 bp from the transcriptional start site retained substantial basal activity in COS-7 cells, while these promoter fragments were not able to drive transcription in MC3T3-E1 cells. As well, in both cell lines a region between − 642 and − 635 from the transcriptional start site conferred a substantial suppressor function.

2. Methods

2.1. Cell culture COS-7 cells were grown in minimum essential medium (MEM) containing 10% FBS in 75 cm2 flasks at 37°C. For transfection, medium was removed and cells were dislodged with trypsin/EDTA. Cellular pellets were resuspended in MEM/10% FBS and identical cell numbers for each experiment were plated in 60-mm dishes. MC3T3-E1 cells were cultured in aMEM, supplemented with 10% FBS and released with trypsin/ EDTA for the transfection experiments.

2.2. MCSF promoter constructs The MCSF promoter-CAT reporter construct (containing − 774– +134 of the murine MCSF gene) was generously provided by M. Harrington along with deletion mutants set upstream of a CAT reporter gene. The

fragment from − 774 to + 10 sequence was inserted into the pGL3-luciferase reporter multi-cloning site (Promega, Madison, WI), and sequence confirmed by sequencing. Serial deletion constructs were then made by cutting with Kpn I (at the − 774 site) and restriction endonucleases at − 635 (Asp I), − 534 (Nsi I), − 466 (Pst I), − 377 (A6r II), − 280 (Sac I) and − 200 (Nhe I). Deletion constructs were selected by size and inability to be cut with appropriate enzymes.

2.3. Electrophoretic mobility shift assay EMSA was performed as previously published (Isaacs et al., 1999) with nuclear extracts made from MC3T3-E1 and COS-7 cells. Double stranded oligonucleotides representing either 50 bp of the MCSF promoter sequence 774–724, 729–679, 684–634 or 20 bp of the promoter sequence 684–664, 669–649 and 654– 634 were labeled with 32P-ATP and used to identify proteins retarded on acrylamide gels.

2.4. Mutagenesis We have previously published the method for deleting 7 bp of the NFkB site by PCR mutagenesis (Isaacs et al., 1999) creating the construct ‘−774dkBMCSFluc’. Two methods were used to make deletions in the region containing the putative suppressor sequence from − 654 to − 634. To create a 7-bp deletion from − 640 to − 634 a similar technique was used as previously: briefly, PCR was performed on 50 ng of the − 774MCSF-luc template with 100 ng each of primers flanking the 7 bp to be deleted (forward primer 5%-TAG-TTG-TCA-CAG-GCC-TTC-CTC-TCC-TACCC-3% and reverse primer 5%-TAG-GAG-AGG-AAGGCC - TGT - GAC - AAC - TAG - CTG - 3%). Sequencing confirmed the identity of selected clones. To delete 20 base pairs from − 655 to − 635, PCR was performed with primers beginning at 5% − 635 sense and at 3% − 655 antisense to create a double stranded PCR product that would skip the 20 bp (sense primer 5%-CCT-TCC-TCTCCT-ACC-CAG-GT-3% and antisense primer 5%-GCTGGG-TGT-TTC-TGC-ACT-CT-3%). When the reaction was completed, 15 units of DpnI were added to digest template DNA and the PCR product was purified using a QIAGEN spin column (QIAGEN, Valencia CA). The PCR product was phosphorylated with T4 kinase in a total volume of 20 ml. This PCR product was run on 0.7% agarose gel, the desired linear band was cut out and purified by GENECLENE spin kit (Bio 101,Vista CA). Two ml of the final eluted PCR product was taken and self-ligation performed in 10 ml volume with 1 ml of T4 ligase. The reaction was left at 16°C overnight. Two ml of ligation mixture was used to transform competent cells and alkaline lysis DNA minipreps were performed


using QIAprep Spin kit. The appropriate sequence was confirmed by sequencing.

2.5. Construction of heterologous promoter constructs containing murine MCSF NFkB domain Double stranded oligonucleotides were designed that contained two copies of the sequence containing the murine NFkB RE (total of 40 nucleotides) with Bgl II-restriction endonuclease sites added as adapter ends. The HSV-TK renilla promoter construct (Promega, Madison WI) was linearized with Bgl II. Mixtures containing the 2X-NFkB RE were self-ligated for 5 min prior to ligation with linearized HSV-TK renilla vector. Competent JM109 bacteria were transformed with the new construct. Clones were selected and grown for diagnosis. PCR with primers spanning both plasmid sequence and the murine MCSF sequence were designed to allow for diagnosis of plasmids containing variable copy number of the 2X NFkB RE.

2.6. Lipofectamine transfection Cells at 50–80% confluence were washed once with 2 ml MEM. Solutions containing 100 ml serum free MEM and 25 ml lipofectamine per transfection dish were added and mixed with 100 ml of serum free MEM per dish containing DNA to be transfected. Unless otherwise stated, transfections included 5 mg/dish of either − 774MCSF-luc promoter constructs along with a TKrenilla-luciferase reporter (0.1 mg/dish) as a transfection control. In some experiments 0.1 – 0.5 mg of an expression vector encoding p65 was added (Schmitz and Baeuerle, 1991). Forty five min later, 0.8 ml serum free MEM per dish was added to the transfection mixture and 1 ml of this mixture was added to each cell culture dish. Cells with transfection mixture were then incubated at 37°C for 5 h at which time 1 ml 20% MEM was added to the cells. The transfection mixture was removed 24 h after the start of transfection and replaced with MEM/10% FBS. Cells were incubated an additional 24 h before assay (t = 48 h after the start of transfection).

2.7. Electroporation Murine osteoblastic MC3T3-E1 cells were transfected by electroporation. For this purpose, cells were resuspended at 2× 106 cells/ml in aMEM/10% FBS. A volume of 0.4 ml was electroporated at 220 V and a capacitance of 1050 mF in the presence of 10 mg circular plasmid reporter DNA. After electroporation, cells were transferred immediately into 60-mm culture dishes (Falcon) and grown for 48 or 72 h before determination of luciferase activity. To control transfection efficiency, cells were co-transfected with 5 mg of the plasmid

195

pSV-b-Gal (Promega). Expression of b-gal was quantitated spectrophotometrically using the b-galactosidase assay system from Promega.

2.8. Modulation of MCSF expression by osteotropic hormones To investigate whether osteotropic hormones exert their effects on the expression of MCSF at the transcriptional level, cells were transfected with reporter constructs and allowed to adhere for 16 h. Thereafter, they were treated with recombinant mouse IL-1a (Pharmingen, Hamburg, Germany) at 100 pg/ml, recombinant mouse TNFa (Pharmingen, Hamburg, Ger) at 1 ng/ml, PTH (Sigma, St. Louis, MO) at 5 × 10 − 8 M and 1,25(OH)2D3 (Roche, Switzerland) at 10 − 8 M for 24 h. Subsequently, CAT assay was performed as described below. To assess the effect of the above hormones or cytokines on MCSF secretion by MC3T3-E1 cells, the cells were treated for 3 days with the agents. Thereafter, cells were cultured for another 3 days without cytokines/hormones to condition the medium (the conditioned medium was a-MEM containing 1% heat inactivated FBS) (Felix et al., 1989). For the proliferation assay, 20 000 bone marrow derived macrophages obtained from teflon bag cultures with L929 supernatants (Cooper et al., 1984) were seeded into 96-well plates. The cells were cultured 48 h in 50% a-MEM plus 1% heat inactivated FBS/50% DMEM plus 15% heat inactivated horse serum. The amount of conditioned medium added was corrected for the cell number (no treatment of MC3T3-E1 cells=control= 100%), and supplemented with a-MEM/1% FBS incubated for 3 days without cells (Elford et al., 1987). Macrophage cell proliferation was determined by using an MTT assay as previously shown (Rubin et al., 1998).

2.9. Firefly and Renilla luciferase assays For the transcriptional assay, cells were washed twice with 4 ml PBS and 250 ml of Promega Reporter Lysis Buffer was added to each dish. Following a 15-min incubation at room temperature, the cell lysate was scraped and cell debris spun down to collect supernatant. Twenty microlitres of cell lysate was used for each assay, and measurements for each transfection dish were made in duplicate. Lysate luciferase levels were assessed with a Promega kit. Renilla luciferase was assessed with Packard RenLite (Packard BioScience, Meriden, CT). Luminescence was measured with a Dynatech Laboratories ML3000 Luminometer. MCSF reporter construct activity was expressed as the ratio of luciferase activity to renilla luciferase activity in arbitrary units.

196


3. Results

3.1. Cis elements participating in the basal transcriptional acti6ity of the MCSF promoter MCSF promoter deletion constructs were transfected into either murine osteoblastic MC3T3-E1, which represent cells committed to bone cell lineage or COS-7 cells, used as a comparison cell line. Data from both systems revealed that deletion of the upstream sequence

Fig. 1. MCSF promoter deletion constructs transfected into MC3T3E1 (A) and COS-7 cells (B). Five mg (A) MC3T3-E1 cells were electroporated with 10 mg MCSF promoter constructs and 1 mg pSV-bGal reporter constructs as described in Section 2. At 72 h after transfection, reporter activity was assayed for luciferase and b-galactosidase, with activity reported as a percentage of the ratio obtained for − 774MCSF-luc 9 S.E.M. (as 100%). The experiment was repeated with similar results. (B) Each construct was applied to three wells in lipofectamine. At 48 h after transfection, cells were assayed for both firefly and renilla luciferase activity. Activity is expressed as the ratio of luciferase/renilla activity, as a percentage of the ratio obtained for − 774MCSF-luc 9 S.E.M. (as 100%). The experiment was repeated more than four times with similar results.

from − 634 to − 774 caused an increase in measured basal transcription (Fig. 1). In MC3T3-E1 cells, deletion of the upstream sequences induced a rise in expression of the reporter activity with maximal expression found with the construct containing the − 466 base pair promoter fragment, dropping thereafter (Fig. 1A). These results were not changed when the MC3T3-E1 cells were transfected using a calcium phosphate method, a secondary method used to assess whether the mode of transfection affected results. Deletion of the farthest upstream region increased basal transcription by about four-fold in the COS-7 cells, which did not vary once sequence from − 634 upstream was removed (Fig. 1B). To ascertain if our − 774-+10MCSF-luc promoter functioned differently than Harrington’s original constructs that contained Intron 1 sequence (to + 134), we also studied deletion constructs possessing base pairs from − 774-+134 linked to a CAT reporter. Highest CAT expression was again observed after removal of promoter fragment − 774 through − 509; if the promoter fragment was shortened further, CAT activity decreased further and was virtually absent in the − 43– + 134 base pair construct (data not shown). Since the removal of the far upstream sequence in both cell types enhanced basal transcription, we investigated whether any protein might be damping transcription by interacting with this area. Electrophoretic mobility shift assays using MC3T3-E1 cell nuclear extract mixed with double stranded oligonucleotides representing this area showed that the proximal third of the sequence from − 774 to − 634 was retarded by an associated protein (Fig. 2). As shown in the figure, further analysis of the ability to bind this protein allowed refinement of the involved sequence to the region containing nucleotides from 654 to 634. The remaining sequences (see Section 2) were not retarded. The same banding pattern was seen with nuclear extracts from COS-7 cells. (NB: ST2 cells showed transfection results indistinguishable from COS-7 cells with respect to basal transcription as well as retardation of a protein band by nucleotides 654–634, data not shown.) To evaluate whether this region did indeed inhibit basal transcription, two mutations were made in the − 774MCSF-luc promoter construct. As described in Section 2, either 20 base pairs representing the sequence from 634 to 655 (‘20 bp’), or seven base pairs representing − 635– −642 were deleted entirely (‘7 bp’). As shown in Fig. 3, these constructs had increased basal promoter activity when compared with the − 774MCSF-luc promoter construct, within the range of the other deletion constructs. Fig. 3A shows results in MC3T3-E1 cells, and Fig. 3B in COS-7 cells; the same relationship existed whether the constructs were added at 2.0 or 5 mg of DNA. The − 466MCSF-luc construct, which has the suppressor region deleted, was added as a secondary control.


197

Fig. 2. EMSA with MC3T3-E1 nuclear extracts show a protein is associated with a previously undescribed region of the MCSF-promoter. EMSA were run using 5 mg of MC3T3-E1 cell nuclear extract. Labeled double stranded-oligonucleotides representing designated sequences were used to assess protein association with that sequence in duplicate. Lane 1 contains probe while other lanes are labeled. A retarded band appeared when promoter sequence representing bases − 655– −635 was present (lanes 6, 7, 12, and 13).

3.2. Regulation of MCSF expression by cytokines and hormones in bone cells To identify cis-elements that might mediate regulatory effects on the expression of MCSF upon exposure of the cells to osteotropic factors, MC3T3-E1 cells were treated with TNFa at 1 ng/ml, PTH at 5 × 10 − 8 M, IL-1a at 100 pg/ml, or 1,25(OH)2D3 at 10 − 8 M for 24 h before determination of reporter activity. No effect on the expression of reporter activity was observed when the cells were treated with any of these osteotropic agents (Fig. 4). To assure that our MC3T3-E1 cells were functioning as predicted, we assessed their ability to respond to agents known to cause secretion of MCSF by bone cells (Felix et al., 1989; Kaplan et al., 1996; Rubin et al., 1996). As shown in Table 1, TNFa, PTH, and IL-1a enhanced macrophage proliferation. The relatively small difference between treated and non-treated cultures lies in the fact that MC3T3-E1 cells secrete MCSF constitutively, so the increases are above basal. The lack of effect of 1,25(OH)2D3, which has been shown to stimulate MCSF secretion previously (Rubin et al., 1996) may have resulted from a lack of effect of the hormone on this particular MC3T3-E1 clone, or a shorter effective stimulus, which became insignificant

during the 3-day collection after the hormone was removed.

3.3. The putati6e NFkB response element does not confer TNFa responsi6eness but can function as an enhancer in a heterologous construct Similar to our findings in ST2 bone stromal cells (Isaacs et al., 1999), a − 774MCSF-luc reporter lacking a NFkB RE ( − 774dkBMCSF-luc) had higher basal activity when transfected into COS-7 cells (Fig. 5). Furthermore, consistent with results shown above with MC3T3-E1 cells, TNFa from 25 to 100 ng/ml added for 24 h prior to measuring luciferase activity failed to stimulate MCSF promoter constructs. Because TNFa causes translocation of NFkB to the nucleus in COS-7 cells similarly to ST2 cells (Isaacs et al., 1999, data in COS-7 cells not shown), we asked whether the murine MCSF NFkB response element functioned differently than those previously shown to enhance transcription. To answer this question the murine MCSF NFkB response element was engineered into a weak HSVthymidine kinase promoter driving a renilla-luciferase reporter. As shown in Fig. 6, possession of four or six copies of the response element was required before a positive TNFa stimulatory response was measurable in

198


COS-7 cells. A similar positive response in a heterologous promoter with a strong CMV enhancer where six copies of the NFkB were added was also seen (data not shown). Interestingly, the heterologous promoter containing only two copies of the murine MCSF NFkB response element reliably responded with decreased luciferase activity in the presence of TNFa as shown in the figure. In an effort to further define a possibly negative role of the endogenous murine NFkB response element, the active partner of NFkB, p65 (Schmitz and Baeuerle, 1991), was concurrently transfected into COS-7 cells containing the − 774MCSF-luc or the − 774dkBMCSFluc constructs. Although 0.1 mg of this vector indeed

Fig. 3. Deletion of the putative suppressor region increases basal transcription of the MCSF promoter in MC3T3-E1 (A) and COS-7 cells (B). As described in Section 2, the putative suppressor region containing sequence − 655– −635 (20 bp) or − 642– −635 (7 bp) was deleted from the endogenous − 774MCSF-luc promoter construct. MC3T3-E1 (A) or COS-7 (B) cells were transfected as described in Section 2. Both constructs with the suppressor region mutated showed increased basal transcriptional activity compared with the − 774MCSF-luc construct. The − 466 construct was studied at the same time for comparison.

caused stimulation of a concurrently transfected heterologous promoter containing six copies of the NFkB RE (data not shown) it did not stimulate the 774MCSF-luc construct, and in fact tended to decrease slightly the basal transcription rate by about 10–25%. However, the same decrease in transcription was seen when the concurrently transfected construct was − 774dkBMCSF-luc, suggesting that the p65 protein had a general inhibitory effect on transcription unrelated to proximity of the NFkB response element.

4. Discussion The regulation of osteoclastogenesis and bone resorption is crucial to maintain bone homeostasis. Osteoblasts are central in this process, since it was shown that bone active hormones and cytokines such as PTH (McSheehy and Chambers, 1986) or TNFa (Thomson et al., 1987) bind to surface receptors on osteoblasts, which in turn elaborate soluble or membrane-associated molecules that directly modulate the formation of osteoclasts. Two factors have been identified that are an obligate requirement for osteoclast formation, activity and survival: colony-stimulating factor-1 (MCSF) (Felix et al., 1990a; Kodama et al., 1991) and osteoclast differentiation factor (ODF/TRANCE) (Fuller et al., 1998; Nakagawa et al., 1998; Kong et al., 1999). These factors are not only essential, but together they are also sufficient to induce the formation of osteoclasts from hemopoietic precursors (Quinn et al., 1998). Extensive data has been presented regarding pretranslational regulation of MCSF (Kaplan et al., 1996; Kimble et al., 1996; Rubin et al., 1996) and ODF/TRANCE (Horwood et al., 1998) expression by osteotropic hormones and cytokines in bone cells. Little is known, however, regarding the mechanisms of transcriptional control and on the critical cis-acting regulatory elements involved. The goal of the present study was to investigate the mechanisms of transcriptional regulation of the MCSF gene. The data presented here demonstrate that 774 base pairs of the MCSF promoter confer a basal transcription of the MCSF gene that is not modulated by PTH, TNFa, IL-1a and 1,25(OH)2D3 despite their ability to increase MCSF secretion. Furthermore, several key consensus sequences do not function as predicted, in particular, the NFkB site. Comparing the activity of the MCSF promoter in MC3T3-E1 and COS-7 cells reveals some striking differences. Although both cells showed increases in basal transcription in the absence of sequence upstream to − 635, in MC3T3-E1 bone cells exhibited decreased expression when further sequences downstream, including the NFkB site, were removed. Thus, in MC3T3-E1 cells the presence of the NFkB site or other sequence


199

Fig. 4. Osteotropic cytokines and hormones do not stimulate the MCSF promoter constructs in MC3T3-E1 cells. MC3T3-E1 cells were, after electroporation, treated with mouse IL-1a at 100 pg/ml (A), mouse TNFa at 1 ng/ml (B), PTH at 5× 10 − 8 M (C) and 1,25(OH)2D3 at 10 − 8 M (D) for 24 h before determination of reporter activity. No stimulation of the transcriptional activity of any MCSF promoter fragment was detected upon treatment with IL-1a, TNFa, PTH or 1,25(OH)2D3. A slight decrease in transcriptional activity, however, was found after treatment.

may be essential for promoter activity, while this is not the case in COS-7 cells. This finding is supported by data presented previously showing that expression of MCSF in different cell types is mediated by common and by cell type specific transcription factors (Konicek et al., 1998). Interestingly, deletion of PU.1, CK-1 and Sp1/Egr1 sites had no effect on basal transcription. Presence or absence of the Sp1/Egr1 site also had no effect on reporter expression in the presence of TNFa or IL-1a as can be seen by the inability to stimulate any deletion promoter construct (Fig. 4). This differs from the findings of Srivastava et al. (1998) who demonstrated that the Sp-1 site was critical for the transcriptional control of MCSF gene expression in the context of TNF/IL-1a stimulated cells. In studies performed using fibroblasts, Harrington, et al. (1993), also suggested that the product of the Wilms’ tumor locus, which recognizes the Egr-1 consensus sequence, repressed transcription of the MCSF promoter when co-expressed at very high levels, perhaps by competing with Sp1 or Sp3 nuclear proteins. It is possible that to uncover subtle effects of these cis acting sequences, high levels of transcriptional activators must be present to disrupt the usual equilibrium of binding factors present in the cell nucleus. A novel suppressor region appears to reside in and around the − 635– −642 sequence. Double stranded oligonucleotides ( − 635 – −685 and − 635 – −655) that contained this sequence were retarded on gel chromatography by an unknown protein. When seven or 20 base pairs containing this sequence were mutated in the

context of the − 774MCSF-luc promoter, basal transcription increased to a rate consistent with that seen in deletion mutants missing the entire upstream portion of the promoter. This suggests that this region interacts with a protein that negatively regulates MCSF gene expression. Although the identity of the associated protein is not known, increased or regulated expression may confer another level of control over MCSF expression, especially in bone cells. The role of the NFkB consensus sequence continues to be puzzling. We have previously shown that deletion of the putative NFkB sequence from the 774MCSF-luc promoter increases basal transcription in ST2 bone stromal cells (Isaacs et al., 1999) and we have now shown the same effect in both MC3T3-E1 and COS-7 Table 1 MC3T3-E1 secreted MCSF promotes proliferation of murine macrophagesa Treatment

Delta OD, 600 nm

Control IL-1a (100 pg/ml) TNFa (1 ng/ml) PTH (50 nM) 1,25(OH)2D3 (10−8 M)

0.696 9 0.011 0.951 9 0.013 0.926 9 0.017 0.861 9 0.013 0.702 9 0.015

a MC3T3-E1 cells were treated for 3 days with specified agents and cultured for another 3 days without cytokines/hormones for collection of conditioned medium (CM). CM was applied to bone marrow derived macrophages in 96-well plates. Macrophages were cultured for 48 h and cell proliferation determined using an MTT assay (see Section 2).

200


Fig. 5. TNFa does not stimulate MCSF promoter constructs. MCSF promoter constructs ( − 774MCSF-luc, dark bars or − 774dkBMCSFluc, light bars) were transfected into COS-7 cells. Cells were treated with TNFa (25, 50, and 100 ng/ml) after transfection. Luciferase/bgal values are expressed as a percentage of the luciferase/b-gal ratio of the basal (no TNFa) MCSF promoter constructs. In no case did TNFa cause an increase in transcription.

Fig. 6. Multiple copies of the murine MCSF NFkB RE can be stimulated by TNFa in a heterologous construct. COS-7 cells were transfected with 5.0 mg of either an empty HSV-TK Renilla-luc vector or one that contained 2, 4 or 6 X repeats of a 20 bp sequence containing the murine MCSF NFkB RE. Twenty four h after transfection, cells were treated 9 25 ng/ml TNFa, and renilla luciferase assayed 24 h later. Open bars represent control and gray bars cultures treated with TNFa. TNFa had no effect on the empty vector and increased the 6X NFkB four-fold. The heterologous promoter construct containing only two copies of the murine MCSF NFkB RE invariably showed decreased transcription. This experiment was repeated with similar results two more times.

cells. When multiple copies of this consensus sequence were expressed in a heterologous HSV-thymidine kinase promoter, they conferred TNFa responsiveness on the previously unresponsive reporter construct. Co-expression of the active partner of NFkB, p65, stimulated the heterologous construct containing six copies of the murine MCSF NFkB response element. This might suggest that multiple copies are necessary for NFkB to cause a positive effect, and that the putative NFkB sequence in the MCSF gene simply does not function as a classical response element. This agrees with our findings that show that despite the ability of TNFa to

increase MCSF expression in many cell types including MC3T3-E1 (Kaplan et al., 1996) and ST2 (Isaacs et al., 1999), it does not increase transcription of the endogenous promoter. Alternatively, the effects of TNFa on this gene may be post-transcriptional. Testing regulatory effects of TNFa on human HL-60 monocytic cells, Sherman et al. (1990) found that TNF increased MCSF transcription measured by run-on, but also that half-life of the protein was increased. Evidence for a labile (cycloheximide inhibited) protein affecting half-life of the MCSF message was also published by Chambers et al. (1993) to account for the increase in MCSF in phorbol treated HL-60 cells. In human blood monocytes, MCSF mRNA increases caused by IL-3 and GM-CSF also appear to be post-transcriptional, as nuclear run on failed to show increases in the presence of these growth factors (Ernst et al., 1989). Dexamethasone also increases the half-life of MCSF protein in bone cells (Rubin et al., 1998). Furthermore, regulation of MCSF expression may also occur at the level of alternative splicing (Stanley, 1994). It has been reported that ovariectomy in rats induces the synthesis of the membrane-bound cytokine (Lea et al., 1999). Although, as mentioned above, Egr-1/Sp1 was postulated to play a crucial role in MCSF production under conditions of hormone deficiency, a shift to the production of the membrane-bound form of the cytokine does not necessarily require regulation at the transcriptional but rather at the post-transcriptional level. Taken together, it is possible that transcriptional control of the MCSF promoter does not underlie the significant increases in MCSF protein seen after cells are stimulated with osteotropic factors. Thus, many questions remain regarding the regulation of MCSF gene expression in bone cells. MCSF is, indeed, a critical growth and differentiation factor for both macrophages and osteoclasts. Our results suggest, however, that MCSF will be expressed constitutively in the basal state by bone stromal cells, and transcriptional regulation is not the major mechanism by which the production of the cytokine is normally controlled.

Acknowledgements Supported by VA Merit, NIH AR21910 and SNFgrant 31-49255-96. We appreciate the excellent technical help of J. Portenier.

References Chambers, S., Gilmore-Hebert, M., Wang, Y., Rodov, S., Benz, E., Kacinski, B., 1993. Posttranscriptional regulation of CSF-1 and CSF-1 receptor gene expression during inhibition of phorbol-es-

J. Rubin et al. / Molecular and Cellular Endocrinology 160 (2000) 193–202 ter-induced monocytic differentiation by dexamethasone and cyclosporin A; potential involvement of a destabilizing protein. Exp. Hematol. 21, 1328–1334. Cooper, P.H., Mayer, P., Baggiolini, M., 1984. Stimulation of phagocytosis in bone marrow-derived mouse macrophages by bacterial lipopolysaccharide: correlation with biochemical and functional parameters. J. Immunol. 133, 913–922. Elford, P.R.R.F., Cecchini, M., Trechsel, U., Fleisch, H., 1987. Murine osteoblast-like cells and the osteogenic cell MC3T3-E1 release a macrophage colony-stimulating activity in culture. Calcif. Tissue Int. 41, 151–156. Ernst, T., Ritchie, A., Demetri, G., Griffin, J., 1989. Regulation of granulocyte- and monocyte-CSF mRNA levels in human blood monocytes is mediated primarily at a post-transcriptional level. J. Biol. Chem. 264, 5700–5703. Felix, R., Fleisch, H., Elford, P.R., 1989. Bone-resorbing cytokines enhance release of macrophage colony-stimulating activity by the osteoblastic cell MC3T3-E1. Calcif. Tissue Int. 44, 356 – 360. Felix, R., Cecchini, M., Fleisch, H., 1990a. Macrophage colony stimulating factor restores in vivo bone resorption in the op/op osteopetrotic mouse. Endocrinology 127, 2592–2594. Felix, R., Cecchini, M.G., Hofstetter, W., Elford, P.R., Stutzer, A., Fleisch, H., 1990b. Impairment of macrophage colony-stimulating factor production and lack of resident bone marrow macrophages in the osteopetrotic op/op mouse. J. Bone. Miner. Res. 5, 781 – 789. Fuller, K., Wong, B., Fox, S., Choi, Y., Chambers, T.J., 1998. TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J. Exp. Med. 188, 997 – 1001. Harrington, M., Edeberg, H., Saxman, S., Pedigo, L., Daub, R., Broxmeyer, H., 1991. Coning and characterization of the murine promoter for the colony-stimulating factor-1-encoding gene. Gene 102, 165 –170. Harrington, M., Konicek, B., Song, A., Xia, I., Fredericks, W., Rauscher, F., 1993. Inhibition of CSF-1 promoter activity by the product of the Wilms’ tumor locus. J. Biol. Chem. 268, 21271 – 21275. Hofstetter, W., Wetterwald, A., Cecchini, M.G., Mueller, C., Felix, R., 1995. Detection of transcripts and binding sites for colonystimulating factor-1 during bone development. Bone 17, 145 – 151. Horwood, N., Elliott, J., Martin, T., Gillespie, M., 1998. Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells, Endocrinology 4743 –4746. Isaacs, S., Fan, X., Nanes, M., Rubin, J., 1999. The role of NFkB in the regulation of macrophage colony stimulating factor. J. Cell. Physiol. 179, 193–200. Kaplan, D., Eielson, C., Horowitz, M., Insogna, K., Weir, E., 1996. Tumor necrosis factor alpha induces transcription of the colony stimulating factor-1 gene in murine osteoblasts. J. Cell. Physiol. 165, 199 – 208. Kimble, R., Srivasta, S., Ross, F., Matayoshi, A., Pacifici, R., 1996. Estrogen deficiency increases the ability of stromal cells to support murine osteoclastogenesis via an interleukin-1 and tumor necrosis factor-mediated stimulation of macrophage colony stimulating factor production. J. Biol. Chem. 271, 28890– 28897. Kodama, H., Yamasaki, A., Nose, M., Niida, S., Ohgame, Y., Abe, M., Kumegawa, M., Suda T., 1991. Congenital osteoclast deficiency in osteopetrotic (op/op) mice is cured by injections of macrophage colony-stimulating factor, J. Exp. Med. 173. Kodama, H., Yamasaki, A., Abe, M., Nida, S., Hakeda, Y., Kawashima, H., 1993. Transient recruitment of osteoclasts and expression of their function in osteopetrotic mice by a single injection of MCSF. J. Bone. Miner. Res. 8, 45–50.

201

Kong, Y.Y., Yoshida, H., Sarosi, I., Tan, H.L., Timms, E., Capparelli, C., Morony, S., Oliveira dos Santos, A.J., Van, G., Itie, A., Khoo, W., Wakeham, A., Dunstan, C.R., Lacey, D.L., Mak, T.W., Boyle, W.J., Penninger, J.M., 1999. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315 – 323. Konicek, B.W., Xia, X., Rajavashisth, T., Harrington, M.A., 1998. Regulation of mouse colony-stimulating factor-1 gene promoter activity by AP1 and cellular nucleic acid-binding protein. DNA Cell Biol. 17, 799 – 809. Lea, C.K., Sarma, U., Flanagan, A.M., 1999. Macrophage colony stimulating-factor transcripts are differentially regulated in rat bone-marrow by gender hormones. Endocrinology 140, 273– 279. McSheehy, P.M., Chambers, T.J., 1986. Osteoblast-like cells in the presence of parathyroid hormone release soluble factor that stimulates osteoclastic bone resorption. Endocrinology 119, 1654 – 1659. Nakagawa, N., Kinosaki, M., Yamaguchi, K., Shima, N., Yasuda, H., Yano, K., Morinaga, T., Higashio, K., 1998. RANK is the essential signaling receptor for osteoclast differentiation factor in osteoclastogenesis. Biochem. Biophys. Res. Commun. 253, 395 – 400. Perkins, S., Kling, S., 1995. Local concentrations of MCSF mediate osteoclastic differentiation. Am. J. Physiol. 269, E1024 –1030. Quinn, J.M., Elliott, J., Gillespie, M.T., Martin, T.J., 1998. A combination of osteoclast differentiation factor and macrophage colony-stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424– 4427. Rubin, J., Fan, X., Thornton, D., Bryant, R., Biskobing, D., 1996. Regulation of murine osteoblast macrophage colony-stimulating factor production by 1,25(OH)2D3, 59, 291 – 296 Rubin, J., Biskobing, D.M., Jadhav, L., Fan, D., Nanes, M.S., Perkins, S., Fan, X., 1998. Dexamethasone promotes expression of membrane-bound macrophage colony- stimulating factor in murine osteoblast-like cells. Endocrinology 139, 1006 – 1012. Schmitz, M., Baeuerle, P., 1991. The p65 subunit is responsible for the strong transcription activating potential of NF-kappa B. EMBO J. 10, 3805 – 3817. Sherman, M., Weber, B., Datta, R., Kufe, D., 1990. Transcriptional and posttranscriptional regulation of MCSF gene expression by TNFa. J. Clin. Invest. 85, 442 – 447. Srivastava, S., Weitzmann, M.N., Kimble, R.B., Rizzo, M., Zahner, M., Milbrandt, J., Ross, F.P., Pacifici, R., 1998. Estrogen blocks M-CSF gene expression and osteoclast formation by regulating phosphorylation of Egr-1 and its interaction with Sp-1. J. Clin. Invest. 102, 1850 – 1859. Stanley, E.R., 1994. Colony stimulating factor-1 (Macrophage colony stimulating factor). In: Thomson, A. (Ed.), The Cytokine Handbook. Academic Press, London, pp. 387 – 418. Sudo, H., Kodama, H.A., Amagai, Y., Yamamoto, S., Kasai, S., 1983. In vitro differentiation and calcification in a new clonal osteogenic cell line derived from newborn mouse calvaria. J. Cell Biol. 96, 191 – 198. Thomson, B.M., Mundy, G.R., Chambers, T.J., 1987. Tumor necrosis factors alpha and beta induce osteoblastic cells to stimulate osteoclastic bone resorption. J. Immunol. 138, 775 –779. Weir, E., Lowik, C., Paliwal, I., Insogna, K., 1996. Colony stimulating factor-1 plays a role in osteoclast formation and function in bone resorption induced by parathyroid hormone and parathyroid hormone-related protein. J. Bone Miner. Res. 11, 1474 – 1481. Wiktor-Jedrzejczak, W., Bartocci, A., Ferrante, A.J., AhmedAnsari, A., Sell, K., Pollard, J., Stanley, E., 1990. Total absence of colony-stimulating factor 1 in the macrophage-deficient osteopetrotic (op/op) mouse. Proc. Natl. Acad. Sci. USA 87, 4828 – 4832.

202


Yao, G.Q., Sun, B., Hammond, E.E., Spencer, E.N., Horowitz, M.C., Insogna, K.L., Weir, E.C., 1998. The cell-surface form of colony-stimulating factor-1 is regulated by osteotropic agents and supports formation of multinucleated osteoclast-like cells. J. Biol. Chem. 273, 4119 – 4128. Yasuda, H., Shima, N., Nakagawa, N., Yamaguchi, K., Kinosaki,

.

M., Mochizuki, S., Tomoyasu, A., Yano, K., Goto, M., Murakami, A., Tsuda, E., Morinaga, T., Higashio, K., Udagawa, N., Takahashi, N., Suda, T., 1998. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597 – 3602.