• osteoclast;
  • differentiation;
  • gene regulation;
  • microphthalmia;
  • TRAP


  1. Top of page
  2. Abstract
  7. Acknowledgements

The defective terminal differentiation of osteoclasts in mice homozygous for the mi allele of the microphthalmia transcription factor (MITF) gene implies that MITF plays a critical role in regulating gene expression during osteoclast ontogeny. To begin addressing the role of this transcription factor in the osteoclast, target genes need to be identified. In the present work, several lines of evidence show that the gene encoding the enzyme tartrate-resistant acid phosphatase (TRAP) is a target of MITF. Analysis of osteoclasts in vivo in the embryonic forelimb showed that MITF and TRAP RNA were coexpressed in a dynamic pattern during the process of endochondral ossification of long bone. Primary osteoclast-like cells (OCLs) produced from mi/mi mutant mice expressed TRAP messenger RNA (mRNA) at 8-fold lower levels than in OCLs derived from normal mice, indicating a direct link between MITF function and TRAP expression. The activity of mouse TRAP promoter-reporter genes was assayed in the primary OCLs by DNA-mediated transfection, and this activity was shown to depend on a conserved sequence (GGTCATGTGAG) located in the proximal promoter. Recombinant MITF protein recognized specifically this conserved sequence element Expression of a TRAP promoter–green fluorescent protein (GFP) transgene mimicked the expression of the endogenous TRAP gene during differentiation of osteoclast-like cells, and the expression of the transgene was decreased 8-fold when placed into the mutant mi/mi background. These results are consistent with a role for MITF in gene expression during terminal differentiation of the osteoclast and will allow osteoclast-specific mechanisms of gene regulation to be studied in greater detail.


  1. Top of page
  2. Abstract
  7. Acknowledgements

The osteoclast plays a central role in bone remodeling in vertebrates, including man, during development and throughout life.(1–3) The changes that occur during the growth phase of bone and during adaptive responses to external stimuli, including the effects of gravity on skeletal architecture, reveal the exquisite signaling mechanisms coupling the actions of the bone-producing cells, the osteoblasts, and the bone resorbing cells, the osteoclasts.(1–3) Ultimately, one consequence of osteoblast/osteoclast interactions must be the regulation of expression of specific genes within these cell types. However, little is known about how genes are regulated in the osteoclast.(4) This is in large part because permanent osteoclast cell lines do not exist and culture of primary osteoclasts in sufficient quantity for molecular genetic and biochemical analysis is difficult. Classically derived mouse mutations and more recently “knockout” mice derived by embryonic stem cell technology have provided genetic evidence for factors likely to be involved in osteoclast gene regulation.(3,4) For example, osteopetrosis results after the knockout of genes encoding the transcription factors c-fos, PU.1, or NF-κB, pointing to the likely role of these transcription factors in osteoclast differentiation and function.(5–7)

The microphthalmia transcription factor (MITF) locus in mouse is a well-characterized genetic locus with at least 16 different mutant alleles characterized.(8,9) Additionally, mutations in the human homologue of this gene can result in the genetic disease Waardenburg's syndrome type II.(10) The mouse MITF alleles can affect, to varying extent, coat color; development of the eye; ear, and colon; mast cell proliferation; osteoclast function; and bone resorption.(8,9) Only mice containing one of three dominant alleles (mi, or, and crc) or two recessive allele (di and vif) exhibit osteopetrosis.(9) The osteopetrosis seen in mutants arising from the recessive alleles is less severe than with the dominant alleles, with the mi allele exhibiting the most severe osteopetrotic phenotype.(8,9)

The gene-encoded MITF was cloned after fortuitous integration of transgenic constructs into the locus.(11,12) The MITF gene was found to encode a helix-loop-helix (HLH) zipper protein related to other known transcription factors, in particular the closely related HLH factors TFE3, TFEB, and TFEC.(11,12) The MITF protein can bind to an 11-base pair (bp) conserved sequence, termed the M-box, which is found in the promoters of melanocyte targets of MITF such as tyrosinase.(13,14) The molecular defect in several of the known mutant alleles has been defined.(8,9) For example, the mi allele has a three base deletion, which removes one of four arginine residues in the basic region of the encoded protein necessary for binding to DNA.(14)

The mi/mi mutant mouse model was of prime importance in showing the hematopoietic origin of the osteoclast, because bone marrow transplantation from normal donors could cure the osteopetrotic condition in mutant mice.(15–17) The defect induced by the mi mutation in the osteoclast appears to occur late in differentiation. Mononuclear osteoclasts can be detected in mi/mi mice but these cells are incapable of fusing to form multinucleate cells, lack a distinct ruffled border, and are defective in bone resorption.(18–22) This is in contrast to many other osteopetrotic mouse models, for example, the previously mentioned c-fos, PU.1, and NF-κB mouse knockout models, in which earlier steps of osteoclast differentiation are affected resulting in lower numbers of osteoclasts in vivo.(5–7) Thus, analysis of MITF may provide unique insights into the regulation of transcription in mature osteoclasts.

As a first step to understanding the role of MITF in the regulation of osteoclast-specific genes, we set out to determine targets for MITF regulation. One set of genes expressed in fully differentiated osteoclasts encode the enzymatic activities associated with osteoclast bone resorption, for example tartrate-resistant acid phosphatase (TRAP).(23–25) Although other cell types, in particular macrophages, also can express TRAP, expression of this enzyme is one of the earliest markers of osteoclast differentiation.(26–30) In the present report, we present evidence from several types of experiments, which together indicate that the TRAP gene is a target of MITF action. In particular, transfection studies with cultures of primary osteoclast-like cells (OCLs) and experiments with mice harboring TRAP-green fluorescent protein (GFP) transgenes provide strong evidence supporting MITF regulation of the TRAP gene. These studies afford a starting point for more detailed molecular analyses of gene transcription during terminal differentiation of osteoclasts.


  1. Top of page
  2. Abstract
  7. Acknowledgements

In situ hybridization of developing limbs

Expression of RNA in developing mouse limbs was analyzed by in situ hybridization using single stranded deoxydigenin (DIG)-labeled RNA probes. Antisense (detection) and sense (control) complementary RNA probes from 400 to 900 ribonucleotides in length were synthesized using T3, T7, or SP6 RNA polymerases from linearized templates. The templates were prepared from plasmids containing suitable complementary DNA (cDNA) sequences for the MITF, TRAP, and calcitonin receptor (CTR) (provided by Debra Galson, Harvard University, Cambridge MA, U.S.A.).(11,31)

Embryos staged from 14.5 days to 18.5 days postcoitus were isolated and fixed in 4% paraformaldehyde at 4°C for 16 h. Serial sections (12 μm) of limbs were prepared and distributed sequentially among four silicon-treated slides and sections were dried for 30 minutes at 37°C. Slide surfaces were covered with 300 μl of hybridization solution (50% formamide, 5× SSC, 2% blocking powder [cat. no. 1096176, Boehringer-Mannheim, Indianapolis, IN, U.S.A.], 0.1% Triton X-100, 0.5% 3-[(3-chdlamidopropyl)dimethyl-ammonio]-1-propanesulfonate (CHAPS), 1 mg/ml yeast RNA, 5 mM ethylenediaminetetraacetic aciol (EDTA), and 50 μg/ml heparin) overnight at 42°C. Fresh hybridization solution (300 μl) containing 500 ng/ml DIG(see above)-labeled probe then replaced the prehybridization solution and hybridization was carried out overnight at 42°C. After hybridization sections were washed as follows: 1× 5 minutes in 2× SSC at 42°C, 3× 5 minutes in 60% formamide in 0.2× SSC at 42°C, 2× 5 minutes in 2× SSC at room temperature.

Sections were next washed 2× 5 minutes with 100 mM Tris-HCl (pH 7.5) and 150 mM NaCl and then incubated for 30 minutes with blocking buffer containing 100 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% Triton X-100, and 2% normal sheep serum. Sections were incubated for 2 h with 300 μl of a 1:200 dilution (1.25 U/ml) of alkaline phosphatase-conjugated anti-DIG antibody in the blocking buffer with 1% sheep serum. Sections were washed 2× 5 minutes at room temperature with 100 mM Tris-HCl (pH 7.5) and 150 mM NaCl and then l× 10 min with freshly prepared detection buffer (100 mM Tris-HCl [pH 9.5], 100 mM NaCl, and 150 mM MgC12). Sections on slides were immersed in detection buffer containing dyes (0.18 mg/ml x-phosphate/5-Bromo-4-chloro-3-indolyl-phosphate (BCIP) and 0.34 mg/ml 4-nitro blue tetrazolium [NBT]). The color reaction was stopped by washing with 10 mM Tris-HCl (pH 8.0) and 1 mM EDTA. Sections were counterstained with 0.02% fast green FCF (wt/vol) in distilled water and then mounted with an aqueous mounting solution when completely dry.

Coculture of OCLs and DNA transfection

Procedures for preparing cultures of OCLs have been described previously.(32–35) Mice heterozygous for the mi mutation (B6C3Fe background) originally were obtained from Jackson Labs (Bar Harbor, ME, U.S.A.), and mice of appropriate genotypes were used for the coculture experiments.

For transfection of primary OCLs, spleen/osteoblast cocultures were carried out in 30-mm dishes. After 6 days of coculture the cells were transfected using the Superfect reagent (Qiagen, Inc., Valencia, CA, U.S.A.). TRAP-luciferase reporter plasmids (5 μg) were mixed with 30 μl of the Superfect reagent and 150 μl of Dulbecco's modified Eagle medium (DMEM) (without serum). After incubation of the mixture for 10 minutes at room temperature, 1.0 ml of DMEM (+15% fetal calf serum [FCS]) was added and the entire cocktail was placed on the cells for 2.5 h at 37°C and 7% CO2. The DNA mixture was removed and replaced with normal coculture media.(35) Twenty-four hours after transfection osteoblasts were removed using 2 mM EDTA treatment.(35) Extracts were prepared from adherent OCLs using Cell Culture Lysis Reagent (Promega, Madison, WI, U.S.A.). In control experiments run in parallel, calvarial osteoblasts were grown in the absence of added spleen cells for 6 days and transfected as above. After 24 h, these cells were harvested and luciferase activity was determined. Luciferase assays were performed as previously described.(36) Luciferase activity is reported as relative light units per micrograms of protein extract assayed.

For transfection assays in RAW 264, cells were grown in 5% FCS (heat inactivated) RPMI and were plated at a density of 3 × 105 cells/ml 24 h before transfection. Cells were then harvested and resuspended at 25 × 106 cells/ml and 5 × 106 cells were used per transfection. RAW 264 were transfected via electroporation at 0.26 kV and 960 microfarad (μF) capacitance. In general, 5 μg of TRAP-luciferase reporter and 2 μg of MITF or mi expression vector were used per transfection.

The luciferase reporter plasmids contained murine TRAP genomic sequences located from 1700 bp upstream of noncoding exon 1 and included exon 1 and the first intron (sequences −2049 to + 1).(31) The conserved core of the M-box sequence (CACATG, sense strand, located at −559 to −554 relative to the ATG start codon) was altered to an Xho I site (CTCGAG) by site-directed mutagenesis.(31) MITF cDNAs have been previously described.(8,11) The short and long (which contains an additional six amino acids adjacent to the DNA-binding basic region) forms of MITF melanocyte cDNA were used for these experiments; however, only the results with the long form are reported.(8,11) DNA transfection data were analyzed by analysis of variance (ANOVA) with multiple comparisons using SAS software to determine statistically significant differences in the data.

Electrophoretic mobility gel shift assays

Recombinant, six his-tagged MITF (wild type and mi/mi mutated form) were produced using the pET15b expression vector system and purified by nickel-agarose chromatography (Novagen, Madison, WI, U.S.A.). Purified protein was used in electrophoretic mobility gel shift assays (EMSAs) using standard conditions as previously described.(14) Double stranded oligonucleotide probes were end-labeled with polynucleotide kinase. The sequences of oligonucleotides used were (sense strand) CAGTTCTGGGGAAGTCCAGTGCTCACATGACCCA (–582 to −549 relative to the ATG, sense strand of the M-box related sequence underlined) and CAGTTCTGGGGAAGTCCAGTGCTctcgagACCCA (mutated TRAP).

TRAP-GFP transgenic mice

For transgenic experiments, coding exons 2–4 of the murine TRAP gene were replaced by T65S GFP cDNA.(31) This vector contains TRAP sequences 2100 bp 5′ to exon 1, noncoding exon 1, intron 1, the authentic TRAP ATG codon in exon 2 fused in frame to the GFP sequences (sequences from −2064 to +1), and 1260 bp of sequence 3′ to exon 5, including the TRAP polyadenylation sequence (+2750 to +4010).(31) Procedures for the creation and identification of transgenic mice has been previously described.(37) The F1 TRAP-GFP transgenic mice were interbred to obtain progeny homozygous for the transgene. These homozygous transgenic mice were bred subsequently with mi/+ heterozygous mice. Mice obtained from this cross were interbred to produce TRAP-GFP mice with the mi/mi genotype.

RNA was isolated from cells and analyzed by Northern blotting as previously described.(37) Microscopy on osteoclast cultures was performed using Eclipse E800 microscope (Nikon, Melville, NY, U.S.A.). Images were captured with MicroMax camera (Princeton Instruments, Inc., Trenton, NJ, U.S.A.) and analyzed with IP Lab Spectrum software (Signal Analytics, Vienna, VA, U.S.A.).


  1. Top of page
  2. Abstract
  7. Acknowledgements

MITF and TRAP messenger RNA expression patterns are similar in developing mouse forelimb

The expression of MITF messenger RNA (mRNA) was examined in developing mouse long bones by in situ hybridization. At 14.5 days of gestation MITF expression is limited to the periosteum in the developing radius (Fig. 1A, panel 1). MITF is expressed in a pattern identical to TRAP (Fig. 1A, panels 3 and 4, respectively). In contrast, the CTR is not expressed at this stage of radius development (Fig. 1A, panel 2).

One day later, the pattern of MITF mRNA expression has altered and cells that express MITF mRNA can be seen migrating into the developing bone (Fig. 1B, panel 1). The signal detected with the MITF probe is specific, because it is detected by the antisense RNA probe but not by the sense RNA probe (Fig. 1B, panel 1 vs. panel 2). The pattern of MITF-positive cells is again coincident with cells positive for TRAP mRNA expression (Fig. 1B, panels 3 and 4, respectively). At this stage of radius development, the pattern of CTR expression also is identical to the MITF and TRAP patterns (data not shown).

The dynamics of MITF expression were similar in other long bones examined (ulna, metacarpal, tibia, and femur), although the exact developmental timing of endochondral ossification varied for all of these long bone types (data not shown).(27) For example, endochondral ossification of metacarpal bone does not initiate until day 17 of embryonic development.(27)

thumbnail image

Figure FIG. 1.. MITF and TRAP expression during long bone development in mouse embryo limbs. MITF, TRAP, and CTR mRNA expression were analyzed in fixed serial sections of (A) 14.5 dpc or (B) 15.5 dpc mouse embryo forelimbs by in situ hybridization as described in Materials and Methods. (A) Riboprobes used were antisense MITF (panels 1 and 3), CTR antisense (panel 2), and TRAP antisense (panel 4). (B) Riboprobes used were MITF antisense (panels 1 and 3), MITF sense control (panel 2), and TRAP antisense (panel 4). Dark grey areas indicate hybridization with riboprobes. In both (A) and (B) the arrows in panel 1 indicate the region that is enlarged in panel 3 (again indicated by arrow) or panel 4. Magnification was ×10 for top panels and ×40 for the bottom.

Download figure to PowerPoint

Diminished expression of TRAP mRNA in OCLs derived from mutant mi/mi mice

To further examine whether MITF might regulate TRAP expression, we studied expression of TRAP mRNA in OCLs derived from spleen of normal and mi/mi mutant mice. These OCLs are produced after coculture of spleen with primary calvaria cultures containing immature osteoblasts in the presence of 1,25-dihydroxyvitamin D3.(32–35) During the course of coculture of cells obtained from wild-type spleen, immature osteoclast precursors differentiate into multi-nuclear OCLs capable of bone resorption.(32–35) After 7 days wild-type cultures yield large, multinuclear TRAP-positive OCLs (Fig. 2, panels 2 and 4). Treatment of the cocultures with EDTA differentially removes the calvaria-derived osteoblasts leaving relatively pure population of OCLs (Fig. 2 panel 2 vs. panel 4).(34,35) In contrast to OCLs derived from wild-type mice, OCLs cultured from mi/mi mutant mice do not fuse to form multinuclear cells (Fig. 2, panels 1 and 3). The mi/mi cells stain weakly for TRAP activity by histo-chemical analysis but can be separated from the calvarial cells by EDTA treatment (Fig. 2, panel 3). These immature cells derived from mi/mi mutant mice often adhere into groups of cells but fail to fuse and form multinuclear cells. The cells from mi/mi mice have been cocultured for up to 14 days without detecting further differentiation.

Northern analysis of total RNA isolated from wild-type and mi/mi OCLs that remain adherent after EDTA treatment of cocultures shows that mRNA for the osteoclast marker CTR is detected while expression of the osteoblast-specific mRNA encoding osteocalcin is not expressed (Fig. 3A, lanes 2 and 5 for wild-type, lanes 3 and 6 for mi/mi).(38–41) Analysis of RNA from the fraction of cells released by EDTA treatment yields the complementary pattern: CTR mRNA is not expressed but osteocalcin mRNA is expressed (Fig. 3A, lanes 1 and 4, respectively). Expression of the macrophage marker scavenger receptor mRNA is not detected in OCLs of either wild-type or mi/mi genotype (data not shown). Taken together, these data provide evidence showing that OCL cultures prepared by EDTA treatment are not significantly contaminated either by osteoblasts or macrophages.

The expression of TRAP mRNA was studied in cocultured OCLs derived from wild-type or mi/mi mutant mice by Northern blotting (Fig. 3B). After 7 days of coculture, TRAP mRNA is expressed abundantly in cells derived from wild-type mice but is expressed at approximately 8-fold lower levels in cells derived from the mi/mi mice. The same blots were rehybridized to a probe specific for the colony-stimulating growth factor-1 receptor, encoded by the c-fms gene. Unlike TRAP mRNA expression, the expression of c-fms is not affected by the mi mutation and serves as a convenient RNA loading control. The small increase seen in c-fms expression in mi/mi OCLs is not reproducible (e.g., see Fig. 6B) and is likely caused by RNA loading differences between samples used for the experiment depicted in Fig. 3B.

TRAP mRNA expression also can be detected in thioglycolate elicited peritoneal macrophages.(37) TRAP mRNA expression in peritoneal macrophages is approximately 5-fold lower than observed in differentiated OCLs when compared with levels of c-fms mRNA, which again serves as an RNA loading control (Fig. 3B). In contrast to OCLs, the levels of TRAP mRNA in peritoneal macrophages isolated from wild-type or mi/mi mice are identical (Fig. 3B).

thumbnail image

Figure FIG. 2.. Differentiation of OCLs after in vitro coculture of primary calvarial and spleen cells. Comparison of OCLs derived from mi/mi mutant mice (panels 1 and 3) or wild-type mice (panels 2 and 4). After day 7 of coculture, the entire culture was fixed with 3.7% paraformaldehyde and TRAP activity detected by histochemical staining (panel 1 vs. panel 2), or osteoblasts were first removed by treatment with 2 mM EDTA,(34) the adherent OCLs fixed and TRAP activity determined by histochemistry (panel 3 vs. panel 4). Grey shading of cells indicates histochemical detection of TRAP activity. Magnification was ×5 for panels 1, 2, and 4 and ×20 for panel 3.

Download figure to PowerPoint

thumbnail image

Figure FIG. 3.. Analysis of mRNA expression in OCLs. (A) Cocultures of osteoblasts and osteoclasts were treated with 2 mM EDTA after 7 days of culture. RNA was prepared from the adherent OCLs derived from wild-type mice (lanes 2 and 5) and from mi/mi mice (lanes 3 and 6). RNA also was extracted from the osteoblast cell population removed by 2 mM EDTA treatment of wild-type cocultures (lanes 1 and lane 4). Expression of CTR (lanes 1–3) and osteocalcin (OSC; lanes 4–6) were compared in the RNA preparations. (B) After 7 days of coculture, RNA was prepared from adherent OCLs after 2 mM EDTA treatment of the cultures. OCLs were derived from mi/mi mutant mice (mi/mi) or from wild-type (wt) litter mates. After Northern transfer, the blot was hybridized first to a probe specific for c-fms and then stripped and probed with a probe specific for TRAP, as indicated in the figure (left two panels, OCL). RNA also was isolated from peritoneal macrophages prepared from mi/mi and wild-type mice, and c-fms and TRAP RNA expression were analyzed after Northern analysis of the two samples (right two panels, MAC).

Download figure to PowerPoint

A conserved sequence in the TRAP promoter is a target for MITF

Targets for MITF action in melanocytes such as the tyrosinase gene contain a conserved sequence motif in their promoters that has been termed a M-box.(13) Inspection of the published sequences for the TRAP proximal promoter region for three species (human, mouse, and pig) revealed a conserved sequence that resembles this M-box (Fig. 4A, antisense DNA strand is shown for the TRAP sequences). This core of the conserved sequence CATGTG is located at −559 to −564 relative to the TRAP ATG codon and is located on the antisense (bottom) DNA strand.(31) To test the functional significance of this conserved sequence in TRAP regulation, reporter genes composed of the mouse TRAP promoter fused to the firefly luciferase gene were constructed. Two such reporters were constructed, one that represented wild-type mouse TRAP promoter sequences from −2000 bp upstream of the TRAP ATG codon, and a second, termed TRAP-Mmut, in which the M-box sequences were mutated to the sequence CTCGAG, which is not recognized by HLH factors.(31)

In the first set of experiments, the activity of the TRAP wild-type sequence was compared with the TRAP-Mmut sequence after transfection into primary OCLs (Fig. 4B). These experiments revealed that the wild-type TRAP promoter exhibited 5- to 6-fold more activity than the M-box-mutated version when transfections were performed in OCLs derived from normal mice (p < 0.0001 by ANOVA multiple comparison tests). However, in OCLs derived from mi/mi mice, both promoter constructs had the same activity, and this activity was identical to the TRAP-mMut activity in normal cells. Additionally, the activity of the promoters was identical when transfections were performed using primary calvaria cells to which spleen cells were not added. As with the mi/mi OCLs, the TRAP promoter activity in the osteoblast cultures is 5- to 6-fold lower than in normal OCLs (p < 0.0001; Fig. 4B). Thus, efficient TRAP promoter activity in OCLs depends on a functional M-box sequence and also on expression of functional MITF protein.

The activity of the TRAP promoter constructs also was assayed in heterologous cell types, including in the macrophage cell line RAW 264 (Fig. 4C). These cells express MITF protein, but at levels 5-fold lower than osteoclasts (Michael C. Ostrowski, unpublished observation, 1999). The activity of the TRAP-Mmut promoter is about 2-fold lower than the wild-type promoter in RAW 264 cells, but this difference is not statistically significant (p = 0.6688 by ANOVA multiple comparison analysis; Fig. 4C, white bars). Additionally, cotransfection of an MITF expression vector along with the wild-type TRAP promoter results in a 6- to 8-fold increase in promoter activity (p < 0.0001; Fig. 4C, grey bars), whereas cotransfection of a vector that contains the mi/mi mutant protein increased TRAP promoter activity about 2-fold, a difference that is not statistically significant (p = 0.2419; Fig. AC, black bars). Cotransfection of either form of MITF expression vector was unable to increase significantly the activity of the TRAP-Mmut reporter (Fig. 4C).

thumbnail image

Figure FIG. 4.. TRAP promoter activity depends on a conserved M-box sequence in OCLs. (A) A sequence conserved in TRAP promoters from mouse, pig, and human is similar to the tyrosinase promoter M-box. Sequences are displayed from 5′ to 3′ end and were obtained from Genbank (accession numbers U03039 for human tyrosinase promoter and M99054, M30283, and X67123, for TRAP mouse, pig, and human, respectively). The sense (top) strand is shown for the tyrosinase sequence, which is conserved in mouse and human tyrosinase, trp-1 and trp-2 genes. The antisense (bottom) strand is shown for the TRAP promoters (sense strand sequence is CACATG). (B) Transient transfections of TRAP-luciferase promoter-reporter constructs into osteoclast primary cultures. Wild-type TRAP promoter (TRAP) or TRAP promoter with mutated M-box (TRAP-Mmut) were transfected into primary OCLs, obtained from the wild-type mice (wt OCL) or mi/mi mice (mi/mi OCL). These reporters also were transfected into primary calvarial osteoblasts that were not cocultured with osteoblast-like cells (OBL). The results of the three independent experiments are shown. (C) Transient transfections of TRAP-luciferase promoter-reporter constructs into the macrophage cell line RAW 264. (Left) Activity of the TRAP luciferase reporter in RAW 264 in the absence or presence of MITF and mi expression vectors. The long form of MITF was used for these experiments. (Right) Activity of the mutated TRAP-Mmut luciferase reporter in RAW 264 in the absence and presence of MITF and mi expression vectors. Results from three independent transfections performed in duplicate are shown. Activity is expressed as relative luciferase activity. Error bars indicate standard deviation.

Download figure to PowerPoint

The transfection results indicated that MITF directly binds to the conserved M-box sequence located in the proximal mouse TRAP promoter. To directly test whether MITF recognized this sequence, recombinant his-tagged MITF protein was overexpressed in Escherichia coli and used in EMSAs with the TRAP-M-box sequence (Fig. 5). These experiments revealed that MITF could form a complex with the TRAP-M-box sequence. The formation of this complex was specifically competed by an excess of unlabeled TRAP-M-box sequence (Fig. 5A, lanes 2–6), but not by a competitor with the TRAP-Mmut sequence (Fig. 5A, lanes 7–10). The formation of this complex also was competed by the tyrosinase-box sequence (data not shown).

thumbnail image

Figure FIG. 5.. EMSA of the TRAP–M-box with recombinant MITF and mi proteins. (A) The TRAP–M-box oligonucleotide (see Materials and Methods) was end-labeled with P-32 and incubated with 0.1 μg of the six his-tagged MITF protein (lane 1). The formation of this complex could be specifically competed by addition of unlabeled TRAP–M-box oligonucleotide to the reaction (lanes 2–6, 2-, 4-, 8-, 16-, and 32-fold molar excess of cold competitor, respectively). An oligonucleotide with the M-box mutated could not compete for complex formation at similar molar ratios (lanes 7–10-, 4-, 8-, 16-, and 32-fold molar excess of competitor, respectively). (B) The recombinant mi protein (0.1 μg) does not form a complex with the TRAP–M-box oligonucleotide (lane 1) and can block the ability of wild-type MITF to form a complex with this probe (lanes 2–5). Lanes 2–5 contained 0.1 μg of wild-type MITF protein and 0 μg, 1.0, 0.4, and 0.2 μg of mutated mi protein, respectively.

Download figure to PowerPoint

The protein containing the mi mutation also was overexpressed and studied by EMSA (Fig. 5B). As was reported for melanocyte target sequences, this mutated version of MITF could not bind to the TRAP-M-box (Fig. 5B, lane 1).(14) When mixed with wild type MITF protein, a 10-fold excess of the mutated mi protein could prevent formation of the MITF-DNA complex (Fig. 5B, lanes 3–5), probably because inactive MITF/mi heterodimers incapable of binding to the target sequence are predominantly formed.(14)

Expression of a TRAP-GFP transgene in wild-type but not mi/mi, mice

A TRAP-GFP transgene was constructed and placed into the germ line of mice by microinjection of fertilized eggs. Of five transgenic lines derived, three expressed the GFP transgene in macrophages and osteoclasts (data not shown). Expression of GFP in OCLs derived from spleen of one of these transgenic lines was studied as a function of time in coculture (Fig. 6A). This analysis showed that expression of the transgene was low after 4 days in coculture, at a point when the majority of osteoclasts are immature, mononuclear cells. As cells begin to differentiate after 5 days of coculture, GFP expression begins to increase, whereas after 6 days in coculture, when multinucleate cells begin to appear, GFP expression is readily detectable. The expression of GFP was highest after 7 days in coculture, when significant numbers of multinuclear, large OCLs are detected, which parallels expression of the endogenous TRAP gene in these cocultures.

The TRAP-GFP transgenic mice were bred with heterozygous mi/+ mice, and F1 offspring from this cross-interbred in order to obtain transgenic TRAP-GFP mice in the mi/mi homozygous mutant background. The expression of transgenic RNA in cocultured osteoclasts was compared with expression of the endogenous TRAP gene (Fig. 6B). This analysis showed that TRAP-GFP expression was detected in cells derived as wild-type or mi/+ heterozygous mice, but that transgene expression was 8-fold lower in mi/mi homozygous mutant mice. TRAP mRNA expression and TRAP-GFP transgene expression in the three genotypes was identical, with lowest expression observed in mi/mi homozygous mutant mice. Expression of c-fms was not dependent on the MITF genotype and served as a control for RNA loading (Fig. 6B).


  1. Top of page
  2. Abstract
  7. Acknowledgements

The phenotype of osteoclasts observed in mi/mi mice suggest that this HLH transcription factor is involved in the regulation of gene expression during the terminal stages of osteoclast differentiation.(18–22) Identification of the TRAP gene as a target of MITF regulation is consistent with this idea. TRAP expression has long been associated with terminal differentiation of the osteoclast, and understanding how this gene is regulated provides a model for the process of osteoclast-specific gene expression.(27–30)

Several lines of evidence are presented indicating that TRAP is an authentic target for MITF action. MITF and TRAP RNA were observed to be expressed in a coincident and dynamic pattern in developing long bone. At 14.5 days of embryonic development expression of both genes is detected in long bone rudiments in the region of the presumptive bone collar adjacent to the area where endochondral ossification initiates.(27–29) Expression of MITF and TRAP at this stage of development precedes expression of the CTR, a definitive marker of the fully functional osteoclast, and MITF and TRAP probably are expressed in immature osteoclasts found in the periosteum at this stage of long bone development.(27–30,38,39) After the initiation of endochondral bone synthesis mature osteoclasts expressing MITF, TRAP, and CTR rapidly migrate into the interior of the developing bone.(27,28) The data indicate that MITF and TRAP mRNA expression patterns are similar in cells present in the developing long bones in mouse limbs.

thumbnail image

Figure FIG. 6.. GFP expression in primary OCLs derived from TRAP-GFP transgenic mice. (A) Cocultures were grown on coverslips and OCLs were prepared by treatment with 2 mM EDTA followed by fixation with 3.7% paraformaldehyde. OCLs were prepared after 4,5, 6, and 7 days of coculture. Cells were stained with bis-benzamide and observed on the fluorescent microscope using fluorescein isothiocyanate (FITC) filter for GFP (left panel) or 4.6-diamidino-2-phenylindole (DAPI) filter for bis-benzamide staining (right panel). Magnification was ×20. (B) Northern analysis of total RNA from OCLs obtained from TRAP-GFP transgenic mice with either +/+, +/mi, or mi/mi genotype. RNA was prepared from OCLs cells after 7 days of coculture. The same blot was hybridized sequentially with GFP, c-fms, and TRAP probes, as indicated. The genotypes of the mice are indicated at the top of the gel lanes.

Download figure to PowerPoint

More direct evidence indicating a role for MITF in TRAP gene expression was provided by experiments with OCLs. OCLs prepared by the coculture system accurately recapitulated the phenotype of osteoclasts present in vivo in mi/mi mutant mice, as coculture cells derived from these mice fail to fuse and form multinuclear cells. Analysis of TRAP mRNA expression in these cells reveals that TRAP expression is 8-fold lower in OCLs derived from mi/mi mice versus normal mice. TRAP also can be expressed in other cell types, in particular in certain classes of macrophages.(26,37) However, TRAP mRNA expression in peritoneal macrophages is not affected by the mi mutation, indicating that MITF regulation of TRAP is osteoclast specific.

Transient transfection of TRAP reporter genes into OCLs provides additional evidence for MITF regulation of TRAP. The wild-type TRAP reporter construct has 6-fold more activity in wild-type cells than in mi/mi cells or in primary osteoblasts. Mutation of a conserved M-box–like sequence in the TRAP promoter abrogates this increased activity in the wild-type cells. Gene transfer into the primary OCLs is relatively efficient, with approximately 10% of the cells successfully transfected, as determined by histochemical staining of cells that were cotransfected with a lacZ reporter (Michael C. Ostrowski, unpublished observations, 1999). This type of gene transfer technique will be of general utility in the molecular dissection of osteoclast-specific processes such as gene regulation, protein trafficking, and signal transduction.

The most compelling evidence indicating MITF regulation of TRAP gene expression was derived from experiments with mice that contained a TRAP-GFP transgene. The TRAP promoter has been used to drive osteoclast-specific expression of transgenes previously.(42,43) The present results confirm these observations and also show that in wild-type cells TRAP-GFP expression is similar to that of the endogenous TRAP gene with maximal expression observed after 6–7 days of coculture. When the TRAP-GFP transgene is placed into the mi/mi genetic background, expression of the transgene is decreased, again paralleling expression of the endogenous TRAP gene.

Deletion of both copies of the TRAP gene in mice results in mild osteopetrosis but multinuclear osteoclasts with ruffled borders are still observed in these mutant mice.(44) Thus, the lower expression of TRAP in mi/mi mice likely contributes to the overall osteopetrotic phenotype but is unlikely to account entirely for the disease. This data argues that additional targets for MITF must exist. The identification of the sequences in the TRAP promoter that mediate MITF activation should aid in identifying additional potential targets for MITF. Prime candidates for such targets include genes involved in cell fusion and in formation of the ruffled border.(18–20)

Mice heterozygous for the mi allele (mi/+ genotype) have no discernable differences in osteoclast ultrastructure or bone phenotype.(9,20,21) At the molecular level no affects were observed either on endogenous TRAP gene expression or TRAP-GFP transgene expression in mi/+ mice. In contrast, in melanocytes pigmentation and target gene expression are affected in mi/+ mice compared with wild-type mice.(8,9) These observations suggest that the mi mutation may not be behaving as a dominant negative allele in osteoclasts. A nonphysiological excess of mi protein can dominantly suppress specific DNA binding by wild-type MITF, including MITF binding to the TRAP gene target sequences identified here.(14) However, it is difficult to relate these in vitro observations to the in vivo situation in mi/+ heterozygotes in which mutated and wild-type protein concentrations are expected to be equal. The protein encoded by the mi allele also may be defective in nuclear localization, at least in heterologous cell types, and perhaps cell-type differences in nuclear localization could account for differences in dominant activity of the mi mutation.(45) Further work is required to determine the mechanism by which the mi allele causes osteopetrosis.

MITF regulates a different set of genes in osteoclasts and melanocytes, for example TRAP and tyrosinase, respectively. One hypothesis to account for the action of this transcription factor in distinct cell-specific programs of gene expression is that MITF interacts with different partners in the different cell types. Two classes of interacting proteins might be considered in this regard. First, there are other members of the MITF HLH subfamily, the proteins TFE-3, TFE-B, and TFE-C, which can form heterodimeric complexes with MITF that recognize M-box–related sequences.(14) Differential expression of these potential heterodimeric partners might alter the pattern of genes regulated by MITF. Consistent with this idea, MITF and TFE-B appear to be the predominant family members expressed in melanocyte cell lines, whereas expression of all four family members can be detected in OCLs.(36) In addition, MITF and TFE-3 can be detected in a complex in extracts prepared from rat OCLs.(46)

A second class of interacting partner could be transcription factors expressed in a cell-type–specific manner that together with MITF compose cell-specific enhancer elements. For example, the Ets family member PU.1 is expressed in OCLs but not in melanocytes, and it is possible that the combination of an osteoclast-specific factor such as PU.1 and MITF might be required for osteoclast-specific regulation of some genes.(36) Experiments with the TRAP promoter now can be designed to test the role of these types of protein interactions in osteoclast-specific gene expression.

The MITF system provides an attractive model to study gene regulation during the terminal differentiation of osteoclasts. Understanding how genes are regulated in mature osteoclasts has the potential to help define the molecular basis of human disorders such as osteoporosis in postmenopausal women or the osteolytic bone destruction and hypercalcemia that occurs in patients with multiple myeloma.(1–3)


  1. Top of page
  2. Abstract
  7. Acknowledgements

We would like to acknowledge the Ohio State University Keck Genetic Facility and Jan Parker Thornburg for producing transgenic mice, the James Comprehensive Cancer Center Biostatistics Unit for aid in performing statistical analyses of DNA transfection data, Carl Insogna for providing advice on coculture of OCLs, and Lori Nelsen for expert technical assistance. This work is supported by NIAMS grant AR-44719 (M.C.O.) and by the National Health and Medical Research Council of Australia (A.I.C. and D.A.H.).


  1. Top of page
  2. Abstract
  7. Acknowledgements
  • 1
    Raisz LG 1992 Mechanisms and regulation of bone resorption by osteoclastic cells. In: CoeFL, FavusMJ (eds.) Disorders of Bone; Mineral Metabolism. Raven Press, New York, NY, U.S.A., pp. 287311
  • 2
    Hall TJ, Chambers TJ 1996 Molecular aspects of osteoclast function. Inflamm Res 45: 19.
  • 3
    Reddy SV, Roodman GD 1998 Control of osteoclast differentiation. Crit Rev Eukaryot Gene Expr 8: 117.
  • 4
    Odgren PR, Hermey DC, Popoff SN, Marks SC Jr 1997 Cellular and molecular strategies for studying the regulation of bone resorption using the toothless (osteopetrotic) mutation in the rat. Histol Histopathol 12: 11511157.
  • 5
    Wang ZQ, Ovitt C, Grigoriadis AE, Mohle-Steinlein U, Ruther U, Wagner EF 1992 Bone and haematopoietic defects in mice lacking c-fos. Nature 360: 741745.
  • 6
    Tondravi MM, McKercher SR, Anderson K, Erdmann JM, Quiroz M, Maki R, Teitelbaum SL 1997 Osteopetrosis in mice lacking haematopoietic transcription factor PU.1. Nature 385: 8184.
  • 7
    Franzoso G, Carlson L, Xing L, Poljak L, Shores EW, Brown KD, Leonardi A, Tran T, Boyce BF, Siebenlist U 1997 Requirement for NF-kappaB in osteoclast and B-cell development. Genes Dev 11: 34823496.
  • 8
    Steingrimsson E, Moore KJ, Lamoreux ML, Ferre-D'Amare AR, Burley SK, Zimring DC, Skow LC, Hodgkinson CA, Arnheiter H, Copeland NG, Jenkins NA 1994 Molecular basis of mouse microphthalmia mi mutations helps explain their developmental and phenotypic consequences. Nat Genet 8: 256263.
  • 9
    Moore KJ 1995 Insight into the microphthalmia gene. Trends Genet 11: 442448.
  • 10
    Tassabehji M, Newton VE, Read AP 1994 Waardenburg syndrome type 2 caused by mutations in the human microphthalmia MITF gene. Nat Genet 8: 251255.
  • 11
    Hodgkinson CA, Moore KJ, Nakayama A, Steingrimsson E, Copeland NG, Jenkins NA, Arnheiter H 1993 Mutations at the mouse microphthalmia locus are associated with defects in a gene encoding a novel basic-helix-loop-helix-zipper protein. Cell 74: 395404.
  • 12
    Hughes MJ, Lingrel JB, Krakowsky JM, Anderson KP 1993 A helix-loop-helix transcription factor-like gene is located at the mi locus. J Biol Chem 268: 2068720690.
  • 13
    Bentley NJ, Eisen T, Goding CR 1994 Melanocyte-specific expression of the human tyrosinase promoter: activation by the microphthalmia gene product and role of the initiator. Mol Cell Biol 14: 79968006.
  • 14
    Hemesath TJ, Steingrimsson E, McGill G, Hansen MJ, Vaught J, Hodgkinson CA, Arnheiter H, Copeland NG, Jenkins NA, Fisher DE 1994 microphthalmia, a critical factor in melanocyte development, defines a discrete transcription factor family. Genes Dev 8: 27702780.
  • 15
    Walke, DG 1972 Congenital osteopetrosis in mice cured by parabiotic union with normal siblings. Endocrinology 91: 916920.
  • 16
    Walker DG 1975 Bone resorption restored in osteopetrotic mice by transplants of normal bone marrow and spleen cells. Science 190: 784785.
  • 17
    Marks SC Jr, Walker DG 1981 The hematogenous origin of osteoclasts: Experimental evidence from osteopetrotic (microphthalmic) mice treated with spleen cells from beige mouse donors. Am J Anat 161: 110.
  • 18
    Thesingh CW, Scherft JP 1985 Fusion disability of embryonic osteoclast precursor cells and macrophages in the microphthalmic osteopetrotic mouse. Bone 6: 4352.
  • 19
    Green PM, Marshall MJ, Nisbet NW 1986 A study of osteoclasts on calvaria of normal and osteopetrotic (mi/mi) mice by vital staining with acridine orange. Br J Exp Pathol 67: 8593.
  • 20
    Holtrop ME, Cox KA, Eilon G, Simmons HA, Raisz LG 1981 The ultrastructure of osteoclasts in microphthalmic mice. Metab Bone Dis Relat Res 3: 123129.
  • 21
    Glowacki J, Cox KA, Wilcon S 1989 Impaired osteoclast differentiation in subcutaneous implants of bone particles in osteopetrotic mutants. Bone Miner 5: 271278.
  • 22
    Graves L, Jilka RL 1990 Comparison of bone and parathyroid hormone as stimulators of osteoclast development and activity in calvarial cell cultures from normal and osteopetrotic (mi/mi) mice. J Cell Physiol 145: 102109.
  • 23
    Minkin C 1982 Bone acid phosphatase: Tartrate-resistant acid phosphatase as a marker of osteoclast function. Calcif Tissue Int 34: 285290.
  • 24
    Chappard D, Alexandre C, Riffat G 1983 Histochemical identification of osteoclasts. Review of current methods and reappraisal of a simple procedure for routine diagnosis on undecalcified human iliac bone biopsies. Basic Appl Histochem 27: 7585.
  • 25
    Lindunger A, MacKay CA, Ek-Rylander, B, Andersson G, Marks S. C. Jr 1990 Histochemistry and biochemistry of tartrate-resistant acid phosphatase (TRAP) and tartrate-resistant acid adenosine triphosphatase (TrATPase) in bone, bone marrow and spleen: Implications for osteoclast ontogeny. Bone Miner 10: 109119.
  • 26
    Razdun HJ, Kreipe H, Parwaresch MR 1983 Tartrate-resistant acid phosphatase as a differentiation marker for the human mononuclear phagocyte system. Hematol Oncol 1: 321327.
  • 27
    Scheven BA, Kawilarang-De Haas EW, Wassenaar AM, Nijweide PJ 1986 Differentiation kinetics of osteoclasts in the periosteum of embryonic bones in vivo and in vitro. Anat Rec 214: 418423.
  • 28
    van de Wijngaert FP, Burger EH 1986 Demonstration of tartrate-resistant acid phosphatase in un-decalcified, glycolmethacrylate-embedded mouse bone: A possible marker for (pre)osteoclast identification. J Histochem Cytochem 34: 13171323.
  • 29
    Baron R, Neff L, Tran Van P, Nefussi JR, Vignery A 1986 Kinetic and cytochemical identification of osteoclast precursors and their differentiation into multinucleate osteoclasts. Am J Pathol 122: 363378.
  • 30
    Minkin C, Yu XH 1991 Calcitonin receptor expression and its regulation by 1 alpha-25-dihydroxyvitamin D3 during de novo osteoclast formation in organ cultures of fetal mouse metatarsals. Bone Miner 13: 191200.
  • 31
    Cassady AI, King AG, Cross NC, Hume DA 1993 Isolation and characterization of the genes encoding mouse and human type-5 acid phosphatase. Gene 130: 201207.
  • 32
    Akatsu T, Tamura T, Takahashi N, Udagawa N, Tanaka S, Sasaki T, Yamaguchi A, Nagata N, Suda T 1992 Preparation and characterization of a mouse osteoclast-like multinucleate cell population. J Bone Miner Res 7: 12971306.
  • 33
    Ikegame M, Rakopoulos M, Martin TJ, Moseley JM, Findlay DM 1996 Effects of continuous calcitonin treatment on osteoclast-like cell development and calcitonin receptor expression in mouse bone marrow cultures. J Bone Miner Res 11: 456465.
  • 34
    Tanaka S, Amling M, Neff L, Peyman A, Uhlmann E, Levy JB, Baron R 1996 c-Cbl is downstream of c-Src in a signaling pathway necessary for bone resorption. Nature 383: 528531.
  • 35
    Insogna KL, Sahni M, Grey AB, Tanaka S, Horne WC, Neff L, Mitnick M, Levy JB, Baron R 1997 Colony-stimulating factor-1 induces cytoskeletal reorganization and c-src-dependent tyrosine phosphorylation of selected cellular proteins in rodent osteoclasts. J Clin Invest 100: 24762485.
  • 36
    Rehli M, Lichanska A, Cassady AI, Ostrowski MC, Hume DA 1999 TFEC is a macrophage-restricted member of the microphthalmia-TFE subfamily of basic helix-loop-helix leucine zipper transcription factors. J Immun 162: 15591565.
  • 37
    Jin DI, Jameson SB, Reddy MA, Schenkman D, Ostrowski MC 1995 Alterations in differentiation and behavior of monocytic phagocytes in transgenic mice that express dominant suppressors of ras signaling. Mol Cell Biol 15: 693703.
  • 38
    Nicholson GC, Moseley JM, Sexton PM, Mendelsohn FA, Martin TJ 1986 Abundant calcitonin receptors in isolated rat osteoclasts. Biochemical and autoradiographic characterization. J Clin Invest 78: 355360.
  • 39
    Hattersley G, Chambers TJ 1989 Calcitonin receptors as markers for osteoclastic differentiation: correlation between generation of bone-resorptive cells and cells that express calcitonin receptors in mouse bone marrow cultures. Endocrinology 125: 16061612.
  • 40
    Kerner SA, Scott RA, Pike JW 1989 Sequence elements in the human osteocalcin gene confer basal activation and inducible response to hormonal vitamin D3. Proc Natl Acad Sci U S A 86: 44554459.
  • 41
    Lian J, Stewart C, Puchacz E, Mackowiak S, Shalhoub V, Collart D, Zambetti G, Stein G 1989 Structure of the rat osteocalcin gene and regulation of vitamin D-dependent expression. Proc Natl Acad Sci U S A 86: 11431147.
  • 42
    Boyce BF, Wright K, Reddy SV, Koop BA, Story B, Devlin R, Leach RJ, Roodman GD, Windle JJ 1995 Targeting simian virus 40 T antigen to the osteoclast in transgenic mice causes osteoclast tumors and transformation and apoptosis of osteoclasts. Endocrinology 136: 57515759.
  • 43
    Hentunen TA, Reddy SV, Boyce BF, Devlin R, Park HR, Chung H, Selander KS, Dallas M, Kurihara N, Galson DL, Goldring, SR, Koop BA., Windle JJ, Roodman GD 1998 Immortalization of osteoclast precursors by targeting Bcl -xL and Simian virus 40 large T antigen to the osteoclast lineage in transgenic mice. J Clin Invest 102: 8897.
  • 44
    Hayman AR, Jones SJ, Boyde A, Foster D, Colledge WH, Carlton MB, Evans MJ, Cox TM 1996 Mice lacking tartrate-resistant acid phosphatase (Acp 5) have disrupted endochondral ossification and mild osteopetrosis. Development 122: 31513162.
  • 45
    Takebayashi K, Chida K, Tsukamoto, I, Morii E, Munakata H, Arnheiter H, Kuroki T, Kitamura Y, Nomura S 1996 The recessive phenotype displayed by a dominant negative microphthalmia-associated transcription factor mutant is a result of impaired nucleation potential. Mol Cell Biol 16: 12031211.
  • 46
    Weilbaecher KN, Hershey CL, Takemoto CM, Horstmann MA, Hemesath TJ, Tashjian AH, Fisher DE 1998 Age-resolving osteopetrosis: A rat model implicating microphthalmia and the related transcription factor TFE3. J Exp Med 187: 775785.