| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Molecular Genetics and the Comprehensive Cancer Center, Ohio State University, Columbus; and
Departments of Microbiology and Biochemistry and the Centre for Molecular and Cellular Biology, University of Queensland, Brisbane, Australia
Correspondence: Michael C. Ostrowski, Department of Molecular Genetics, Ohio State University, 484 W. 12th Ave., Columbus, OH 43210. E-mail: ostrowski.4{at}osu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: transcriptional regulation • myeloid differentiation • E-cadherin
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MITF regulates expression of genes in melanocytes through recognition of conserved E-box-related sequences found in the promoters and enhancers of target genes such as tyrosinase [9 ]. The core sequences TCATGTG or CATGTGA present in these E-boxes are crucial for high-affinity DNA recognition by MITF [10 ]. Recently, we have demonstrated that the proximal promoter region for the tartrate-resistant acid phosphatase (TRAP) gene, which serves as a marker for terminal osteoclast differentiation, contains a sequence closely related to the E-boxes defined in melanocyte promoters. This cis-element can mediate MITF regulation of TRAP during terminal differentiation of osteoclast-like cells (OCLs) [11 ].
The N-terminal region of the MITF protein contains an acidic domain, encoded by exon 4 of the MITF gene, necessary for transactivation of melanocyte target genes [6 , 12 ]. This region is conserved among MITF, TFEB, and TFE3 and is believed to be a binding site for the co-activator CBP/p300 [12 ]. Interaction of the acidic-activation domain with CBP/p300 is shown to be regulated by c-kit/mitogen-activated protein (MAP) kinase signaling through phosphorylation of serine residue 73 encoded by exon 2B [13 ].
Two previously described semi-dominant mutant alleles of MITF, white spot (ws) and x-38, have deletions of exons 2–4 and exon 4, respectively, and thus encode proteins lacking the acidic-activation domain [6 ]. Mice containing these alleles have defects in coat and eye pigmentation in heterozygous and homozygous combinations, and ws shows the more-severe phenotype, providing strong genetic evidence for the role of the N-terminal acidic-activation domain of MITF in melanocyte gene expression and differentiation [6 ]. However, despite the effects on melanocyte gene expression and differentiation, neither of these alleles leads to developmental defects in bone. One hypothesis to explain these results is that the acidic-activation domain is redundant for MITF action in osteoclasts.
To address this hypothesis, we set out to define transactivation domains in the MITF protein important for activation of osteoclast-expressed genes. We have defined a distinct region in MITF that behaves as an activation domain in transient transfection assays using two potential osteoclast targets, the E-cadherin promoter and the TRAP promoter. In addition, we show that this region of MITF is required for rescue of the defect in mi/mi osteoclasts by wild-type MITF.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B (NF-B) ligand
(RANKL; PeproTech, Rocky Hill, NJ), 20 ng/ml tumor necrosis factor-
(TNF-; R&D Systems, Minneapolis, MN), and 5 ng/ml colony stimulating
factor-1 (CSF-1). Cells were maintained in culture for 7 or 10 days.
Mice heterozygous for the mi mutation (B6C3Fe background)
were originally obtained from Jackson Labs (Bar Harbor, ME), and mice
of appropriate genotypes were used for the experiments. Procedures for
TRAP histochemical staining of cells and RNA extraction and analysis by
Northern blotting have been described previously [11
]. The procedure for preparation of thin sections from embryonic forelimbs and TRAP histochemical staining has been described previously [11 ].
Electrophoretic mobility gel shift assays (EMSAs)
Recombinant, 6X histidine (6Xhis) MITF (wild type) was produced
using the pET15b expression vector system and purified by
nickel-agarose chromatography as described previously
[11
]. Purified protein was used in EMSAs using standard
conditions as described previously [11
]. The sequences
of the E-cadherin oligonucleotides used were (sense strand in each
case): TTGGCTGCCACCTGCAGGTGCGTCCC (E-Pal),
TGCGGGCTCACCTGGCGGCCGC (E-Prox), and
TGCGGGCTCTCGAGGCGGCCGC (mutated E-Prox).
DNA constructs and DNA transfections
For all experiments presented here, the so-called melanocyte
form of the MITF cDNA containing the additional six codons encoded by
alternate exon 6a was used [2
, 8
]. The
melanocyte-form mRNA is an MITF isoform expressed in cultured mouse
osteoclasts (unpublished results). MITF N-terminal deletions, termed
MITF 
109 or 140, were made by deleting nucleotides 455–546 or
546–683, respectively, in the MITF cDNA [2
]. These
deletions correspond to amino acid residues 109–140 or 140–185,
respectively. Full-length or N-terminal deletions were cloned into a
vector providing the influenza hemagglutinin (HA) tag or into a
retroviral vector containing a 6Xhis tag. N-terminal point mutations
within the region defined by MITF 140 were made by replacing the
conserved amino acid residues with alanine residues. Three mutations,
corresponding to amino acid residues 148–151 (mut1), 171–174 (mut2),
and 181–184 (mut3), were constructed. All deletion mutants and point
mutants were verified by sequencing.
A fragment of the E-cadherin promoter between nucleotides -178 and +92 relative to the transcriptional start site [14 ] was placed into pGL2-Basic vector (Promega Biotech, Madison, WI). The E-Pal (palindromic) or E-Prox (proximal) E-box sequence was mutated to the sequence of an XhoI site (CTCGAG) by site-directed mutagenesis, as above. The TRAP reporter construct was described previously [11 ].
DNA transfections of RAW 264.7 cells have been described previously [11 ]. Data for all transfections were subjected to analysis of variance (ANOVA) comparison using SAS software to determine statistically significant differences in the data.
Immunological reagents and analysis
For 35S-methionine-labeling experiments, cells were
grown in media lacking methionine overnight. 35S methionine
(90 µCi; ICN Pharmaceuticals, Costa Mesa, CA) was added to
the cells, and the cells were labeled for 3 h. Cells were
lysed in radioimmunoprecipitation assay [50 mM Tris-HCl,
pH 7.6, 125 mM NaCl, 1% Nonidet P-40 (NP-40), 0.5% deoxycholate, and
6 mM MgCl2], which included 10 µg/ml antipain, 10
µg/ml leupeptin, 10 µg/ml aprotinin, 100 µg/ml
phenylmethylsulfonyl fluoride (PMSF), and 1 mM sodium vanadate. Cells
were incubated in RIPA at 4°C for 20 min and then centrifuged at
25,000 rpm for 20 min. Protein G beads (Pharmacia, Upsala, Sweden) were
added to the lysate, and the appropriate antibody [anti-6Xhis from
Santa Cruz (Santa Cruz Biotechnology, CA); HA from BabCo,
Berkeley, CA] was added. The immunoprecipitate was incubated
at 4°C overnight. The immunoprecipitate was washed three times in
RIPA and run on 8% sodium dodecyl sulfate-polyacrylamide gel
electrophoresis.
Immunohistochemistry was performed with OCLs that had been cultured on plastic coverslips for 7 days. Cells were washed in Tris-buffered saline (TBS; 50 mM Tris, pH 7.5, 150 mM NaCl), fixed for 30 min at 4°C in 4% paraformaldehyde prepared in TBS, washed three times with TBS, and permeabilized for 5 min with TBS containing 1% Triton X-100. Endogenous peroxide activity was blocked by incubation with 0.1% hydrogen peroxide and 10% methanol in TBS. Cells were washed and treated with 2% normal goat serum (Sigma Chemical Co., St. Louis, MO) in TBS for 1 h at room temperature. The commercial His antibody (1:500) was added and incubated overnight at 4°C. In control experiments, cells were incubated without primary antibody. Following incubation with primary antibody, the cells were washed three times with TBS containing 0.2% NP-40. Anti-rabbit immunoglobulin (Ig)G-biotin (Boehringer Mannheim, Mannheim, Germany) was added for 1 h, followed by three washes with TBS. Antibody binding was detected with the metal-enhanced 3'-diaminobenzidine tetrahydrochloride (DAB) substrate kit (Pierce Biochemicals, Rockford, IL).
Production of retroviruses and infection of OCLs
The retroviral DNA constructs were transfected into the PA317
packaging cell line [15
]. The PA317-infected cells were
selected with puromycin, and individual clones were picked. The titer
of the virus was determined by infecting Rat-1 cells with various
dilutions of supernatant from PA317 cells. Because the titers for
viruses used were quite variable, a co-culture procedure was used to
insure efficient infection of osteoclast precursors. For this purpose,
PA137 cells producing virus were treated with 2 µg/ml mitomycin C for
16 h prior to use with OCLs [16
]. Producer cells
treated in this fashion can no longer replicate but are still competent
for virus production.
For retrovirus infection of OCLs, 1.8 x 107 cells
obtained from spleens of mi/mi mice were mixed with 1.5 x 105 mitomycin C-treated PA317 cells that produced the
various viruses and were tested in triplicate. The mixture of cells was
cultured in Dulbecco’s modified Eagles medium (DMEM) containing 10%
fetal bovine serum and 50 ng/ml CSF-1 on tissue culture-treated plastic
coverslips placed in 24-well tissue-culture dishes. After 3 days,
nonadherent spleen cells were removed by washing with TBS, and the
cultures were treated briefly with 0.1% trypsin. This treatment
removed the mitomycin C-treated fibroblasts, but the mononuclear
osteoclast precursors remained adherent. DMEM containing 10% fetal
bovine serum, 80 ng/ml RANKL, 20 ng/ml TNF-, and 5 ng/ml CSF-1 was
added, and cells were cultured for an additional 7 days.
After the 7-day incubation, cells were fixed, and expression of the 6Xhis-tagged MITF proteins was determined by immunohistochemistry as described above. The cell nuclei were stained with bis-benzamide (1 µg/ml), and cells positive for expression of exogenous 6Xhis-tagged MITF were counted and scored as mononuclear or multinuclear. The ratio of multinuclear, 6Xhis-tagged-positive cells over the total number of 6Xhis-tagged-positive cells (mononuclear and multinuclear) was used as a measure of the rescue of differentiation of mi/mi OCLs. Some cultures were subjected to TRAP histochemical staining instead of immunohistochemistry, revealing that multinuclear cells were strongly positive for TRAP expression as well (unpublished results).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
, and CSF-1
(see Materials and Methods) resulted in the formation of multinuclear
cells, sometimes containing several dozen nuclei. These OCLs exhibit
high levels of TRAP activity as determined by histochemical staining
(Fig. 1B , bottom panel). The cultures prepared from mi/mi
mice were composed nearly entirely of mononuclear cells that stain
weakly for TRAP activity (Fig. 1B
, top panel). The mi cells
fail to fuse to form multinuclear cells even after 2 weeks in culture
(unpublished results). Thus, the in vitro culture system accurately
reflects the differentiation of OCLs observed in vivo. The transmembrane cell-adhesion molecule E-cadherin is expressed in cultured OCLs as cell fusion occurs, and neutralizing antibodies to E-cadherin reduced the number of multinuclear, TRAP-positive OCLs obtained during in vitro differentiation [20 ]. Because of the obvious fusion defect in mi/mi OCLs (Fig. 1A and 1B) , experiments to test the hypothesis that MITF might regulate E-cadherin expression were undertaken. As a first step, the expression of E-cadherin mRNA in OCLs derived from mi/mi mice was determined using Northern analysis (Fig. 1C) . Expression of E-cadherin mRNA was detected in cells derived from wild-type mice but was expressed at an approximate 12-fold lower level in cells derived from mi/mi mutant mice (Fig. 1C) . As we demonstrated previously in an osteoclast/osteoblast co-culture in vitro system [11 ], two genes considered as definitive marker for osteoclasts, TRAP and calcitonin receptor (CTR), are expressed in the cultured OCLs. However, TRAP expression is six- to eightfold lower in the mi/mi cells versus wild-type cells, and CTR is expressed in both cell populations at equivalent levels and thus serves as a convenient RNA loading control (Fig. 1C) .
MITF transactivates the E-cadherin promoter through a conserved
E-box-like sequence
Targets for MITF in melanocytes and osteoclasts contain a
conserved E-box sequence motif in their promoters [9
,
11
]. Inspection of the published sequence for the
E-cadherin promoter for two species (human and mouse) revealed two
conserved E-box-related sequences (Fig. 2 A
). One of these E-box sequences, located at approximately position
-78 relative to the mRNA initiation site, has been described
previously to act as a negative element in mesenchymal cells
[14
]. This site consists of palindromic E-boxes in the
mouse sequence (termed E-Pal), but only one of these E-boxes is
conserved in the human promoter (Fig. 2A)
. The second conserved site is
located at position -18 in a GC-rich region of the promoter (termed
E-Prox in Fig. 2
). Two types of assays were used to determine the
functional significance of these conserved E-boxes.
|
To further test the functional significance of the conserved E-box sequences in the E-cadherin promoter, reporter genes composed of the mouse E-cadherin promoter fused to the firefly luciferase gene were constructed. Reporters representing the wild-type E-cadherin promoter sequences from -178 to +72 bp relative to the E-cadherin transcription initiation site or versions in which the distal or proximal E-box sequence was mutated (E-Pal and E-Prox in Fig. 2C , respectively) were constructed. The activity of the E-cadherin constructs was assayed in RAW 264.7 cells, a macrophage/osteoclast cell line [21 ] (Fig. 2C) .
There was no significant difference in the basal level activity of the wild type and mutant (E-Pal and E-Prox, Fig. 2C ). Co-transfection of the MITF expression vector along with wild-type E-cadherin promoter or the promoter mutated at the upstream E-Pal sequence resulted in a 10- to 12-fold increase in promoter activity. In contrast, co-transfection of the MITF expression vector with the promoter containing the mutated E-Prox sequence did not stimulate the promoter above the basal level (Fig. 2C) . The proximal conserved E-box is defined as a target for MITF action by EMSA and transient transfection assays.
Structure/function analysis of MITF transactivation using TRAP
and E-cadherin reporters
To determine the regions of MITF that were necessary for
transactivation of the osteoclast target promoters, a series of
deletion mutations in the N-terminal and C-terminal regions of MITF
were constructed. Three deletion mutations that represent the three
types of transactivation results observed in our studies are
represented in Figure 3 A
. The mutation 109 has an in-frame deletion of the region of
MITF between amino acids 109 and 140 that represents a previously
defined acidic-activation domain and potential CBP/p300 binding site
[12
]. Mutation 140 contains a deletion of the region
adjacent to the acidic-activation domain that has previously been shown
to be dispensable for MITF activation in a melanocyte cell line
[12
]. Lastly, 367 truncates a C-terminal region of
MITF that contains a proline-rich region. A similar proline-rich region
in the C-terminal of the related factor TFE3 has previously been shown
to function as a transactivation domain [22
].
|
109 MITF transactivated the E-cadherin promoter just as
well as wild-type MITF but exhibited five- to sixfold lower
transactivation of the TRAP promoter than wild-type (a significant
decrease; P<0.0001 by ANOVA multiple comparison test). The
140 MITF transactivated E-cadherin and TRAP reporters eight- to
tenfold less efficiently than wild-type MITF (Fig. 3B ;
P<0.0001). The 367 MITF protein activated both reporters
to the same extent as wild-type MITF, indicating that the C-terminal
proline-rich sequence was not necessary for transactivation of the OCL
reporter genes (Fig. 3B)
.
To gain some insight into the normal bone phenotype in ws
and x-38 MITF mutant mice, the 109 and 140 expression
vectors were co-transfected along with wild-type MITF to determine if
they could act in a dominant negative fashion (Fig. 3C)
. For these
experiments, equal amounts (1 µg each) of expression vectors encoding
mutated or wild-type proteins were co-transfected along with the two
reporter genes. Consistent with the results of the transactivation
assay, 109 and 140 mutations were able to decrease
transactivation by MITF approximately threefold when the TRAP reporter
was tested (Fig. 3C
, left panel). However, only the 140 mutation
inhibited MITF activation of the E-cadherin promoter (by fivefold, Fig. 3C , right panel). This result is consistent with the results of the
transactivation assays presented above.
Fine mapping of sequences contained within the 140 mutation
necessary for activation of target promoters
We reasoned that comparison of the sequences defined by the 140
mutation to the other members of the MITF bHLHzip family might reveal
conserved sequences necessary for activation of target genes. This
comparison revealed blocks of homology between MITF and TFE3 and to a
lesser extent, TFEB (Fig. 4A
). TFEC lacks the acidic-activation domain present in the other
MITF family members [23
] and had limited homology in the
residues most conserved in the 140–185 region of MITF (Fig. 4A)
. Based
on these comparisons, point mutations were constructed by substituting
stretches of four conserved amino acids with alanine in three different
regions of homology (highlighted in Fig. 4A
).
|
The 109 and 140 mutations are stably expressed and localized
to the nucleus of OCLs
The expression of proteins containing the 109 and 140
mutations was compared with wild type to insure that the alteration of
transactivation potential was not a result of lower protein stability.
In one set of experiments, expression of influenza HA-tagged versions
of wild-type MITF and the 109 and 140 MITF mutations was studied
in COS cells following transient transfection (Fig. 5A
, left panel). In another set of experiments, retrovirus vectors
were used to express 6Xhis-tagged versions of these three proteins as
well as a protein encoded by the mi allele in Rat1 cells
(Fig. 5A
, right panel). In both types of experiments, no major
differences in the level of 35S-labeled MITF proteins were
observed (Fig. 5A ; see Materials and Methods), indicating that protein
stabilities are not grossly altered by the deletion mutations. In both
cell types, the MITF wild type, 109, and 140 proteins appeared as
a doublet (Fig. 5A)
. The band with lower electrophoretic mobility has
been shown previously to be a version of MITF phosphorylated at
position ser73 [24
]. For unknown reasons, the
MITF-mi protein appears as only a single band, presumably
the unphosphorylated version of the protein.
|
Expression of the 140 mutation in mi/mi OCLs fails to
rescue differentiation
The results presented above suggested that the 140 mutation
defines a region of MITF necessary for transactivation of OCL target
genes and by extension, is important for OCL differentiation. To test
directly the biological role of this putative activation domain in OCL
differentiation, the ability of the 140 mutation to rescue
differentiation of mi/mi OCLs was compared with wild-type
MITF and the 109 mutation. The MITF genes were introduced into
spleen cells obtained from mi/mi mice using retrovirus
vectors as above. Following treatment of transduced OCLs with RANKL and
CSF-1 for 7 days, expression of 6Xhis-tagged proteins was determined by
immunohistochemistry. Multinuclearity of cells positive for expression
of tagged proteins was used to score differentiation. The average of
three experiments (each performed in triplicate) is represented in
graphical form in Figure 6.
Table 1
contains a more complete summary of the data. These experiments
demonstrated that expression of the wild-type MITF protein could rescue
differentiation in approximately 60% of the infected mi/mi
mononuclear OCLs. In contrast, <5% of cells infected with virus
encoding the mi allele or with viral vector alone was
multinuclear. Expression of the 109 mutation resulted in rescue of
35% of mi/mi OCLs, about 60% the level of rescue seen with
wild type—a difference that is significant (ANOVA with pair-wise
comparison; P<0.0001). However, the 140 mutation is more
seriously impaired, with only 7% of cells rescued. This is a
statistically significant difference from wild type or 109 mutation
(P<0.0001 for either) but is not different from rescue
observed in the mi/mi or viral vector controls
(P=0.012). The majority of cells rescued by 109 or 140
mutations was predominantly bi- or tri-nuclear, in contrast to
wild-type MITF, in which cells with four to six nuclei were encountered
frequently (see Fig. 5B
).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B, PU.1, or
c-fos, are mutated by targeted deletion because such
mutations affect earlier stages of myeloid differentiation, and they
result in an absence or defects in macrophages and osteoclasts in
mutant mice [26
27
28
]. In this study, we have confirmed
that osteoclast maturation appears to halt at the mononuclear cell
stage in vivo during early bone development in mi/mi
embryos, as well as in vitro using a cell-culture system that depends
on the addition of defined factors for OCL differentiation. Therefore,
studying MITF action in osteoclasts may provide a unique window into
understanding how cell-cycle exit, cell fusion, and expression of genes
necessary for bone resorption are coordinated during the terminal
differentiation of this cell type. Two semi-dominant MITF alleles, ws and x-38, encoding proteins that lack an acidic-activation domain, show no obvious osteoclast phenotype [6 ]. These data led us to consider that MITF might contain an additional activation domain that could compensate for the loss of the acidic-activation domain in osteoclasts. This hypothesis was confirmed by experiments demonstrating that a second N-terminal activation domain, located between amino acid residues 140 and 185, was required for transactivation of potential OCL target genes. Analysis of point mutations within this region highlighted the role of specific amino acid residues conserved with other closely related bHLH-zip family members, in particular with TFE3. A biological role for this N-terminal region of MITF was established in rescue experiments performed using the differentiation-defective mi/mi OCLs. Wild-type MITF or a version lacking the acidic-activation domain could effectively rescue differentiation of mi/mi OCLs, but a protein lacking the 140–185 activation domain could not.
Because of technical difficulties with the experiments, we could not obtain sufficient numbers of rescued cells to perform western blots to insure that the mutated forms of MITF were expressed at equivalent levels in retrovirus-transduced cells. Experiments performed in two heterologous cell types showed no gross differences in protein stability between wild-type and mutated MITF proteins. In addition, immunohistochemical analysis of infected OCLs demonstrated that all MITF protein versions were expressed and located in the nucleus. Although immunohistochemistry is not strictly quantitative, the levels of staining were equivalent for all MITF genes transduced into the OCLs, suggesting that protein levels did not vary widely. Although we believe the sum of our data makes the trivial explanation of our results unlikely, we cannot rule out the possibility that variable levels of protein expression between wild-type and mutated versions of MITF account for our results.
In the course of analysis of MITF function, the E-cadherin promoter was identified as a potential target of MITF. The gene was chosen as a candidate because, like TRAP, its expression greatly increases during OCL maturation, and more importantly, it has been causally linked to the formation of multinucleated giant cells [20 ], a process that does not occur in OCL-derived mi/mi mice. As in the TRAP gene [11 ], we have identified a conserved E-box that could be bound specifically by MITF and was necessary for promoter transactivation by this factor. Recent work has demonstrated that another potential target of MITF in osteoclasts, the human cathepsin K gene, contains similar cis-acting elements [29 ]. Although the data presented here are consistent with the proposal that E-cadherin is a target for MITF, additional types of experiments will have to be performed to confirm that this is the case. For example, experiments in transgenic mice, such as were performed with the TRAP promoter [11 ], would provide in vivo evidence supporting the hypothesis that E-cadherin is a target of TRAP.
The MITF-target E-box in E-cadherin is located close to the site where
transcription initiates, at position -18, in contrast to other
osteoclast target genes TRAP and cathepsin K, in which the MITF binding
site is more distal, at a position 
100 nucleotides 5' to the major
transcriptional start sites [11
, 29
]. This
difference correlates with a differential requirement for MITF
activation domains: The E-cadherin promoter required only the 140–185
activation domain defined by our experiments, and the acidic-activation
domain between 109 and 140 and the 140–185 element were required for
maximal TRAP activation. One explanation for this result is that
distinct types of activation domains are known to function differently
depending on their distance from core promoter elements such as TATA
boxes and initiator (Inr) elements [30
,
31
]. In this regard, MITF transactivation may be very
similar to the example provided by upstream regulatory factor-2
(USF-2), a distantly related bHLHzip factor. USF-2 has two N-terminal
activation domains that function differently depending on whether the
USF binding site is promoter-proximal or -distal [32
].
It is interesting that the USF-2 activation domain, termed USR, which
functions in a promoter-proximal, Inr-dependent manner, may be
important for cell type-specific activation of target genes
[33
].
In conclusion, results presented here likely identify an additional activation domain in the transcription factor MITF that may partly account for the ability of this factor to regulate different sets of genes in osteoclasts and melanocytes. The results suggest the hypothesis that MITF might recruit different co-activator molecules in melanocytes and osteoclasts. Identification and characterization of such putative co-activators will be required to determine their role in the molecular mechanisms underlying osteoclast-specific gene regulation.
|
ACKNOWLEDGEMENTS |
|---|
|
FOOTNOTES |
|---|
Current address of Georgia H. Purdom: Department of Biology, Mt. Vernon Nazarene College, Mt. Vernon, OH 43050.
Received April 30, 2001; revised August 18, 2001; accepted August 20, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B in osteoclast and B-cell development Genes Dev 11,3482-3496This article has been cited by other articles:
|
|
 |
|
  R. Hu, S. M. Sharma, A. Bronisz, R. Srinivasan, U. Sankar, and M. C. Ostrowski Eos, MITF, and PU.1 Recruit Corepressors to Osteoclast-Specific Genes in Committed Myeloid Progenitors Mol. Cell. Biol., June 1, 2007; 27(11): 4018 - 4027. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Sharma, A. Bronisz, R. Hu, K. Patel, K. C. Mansky, S. Sif, and M. C. Ostrowski MITF and PU.1 Recruit p38 MAPK and NFATc1 to Target Genes during Osteoclast Differentiation J. Biol. Chem., May 25, 2007; 282(21): 15921 - 15929. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. A. Meadows, S. M. Sharma, G. J. Faulkner, M. C. Ostrowski, D. A. Hume, and A. I. Cassady The Expression of Clcn7 and Ostm1 in Osteoclasts Is Coregulated by Microphthalmia Transcription Factor J. Biol. Chem., January 19, 2007; 282(3): 1891 - 1904. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Mbalaviele, G. Anderson, A. Jones, P. De Ciechi, S. Settle, S. Mnich, M. Thiede, Y. Abu-Amer, J. Portanova, and J. Monahan Inhibition of p38 Mitogen-Activated Protein Kinase Prevents Inflammatory Bone Destruction J. Pharmacol. Exp. Ther., June 1, 2006; 317(3): 1044 - 1053. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. M. Roundy, G. Spangrude, J. J. Weis, and J. H. Weis Partial rescue of B cells in microphthalmic osteopetrotic marrow by loss of response to type I IFNs Int. Immunol., November 1, 2005; 17(11): 1495 - 1503. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Huan, D. Sashital, T. Hailemariam, M. L. Kelly, and C. A. J. Roman Renal Carcinoma-associated Transcription Factors TFE3 and TFEB Are Leukemia Inhibitory Factor-responsive Transcription Activators of E-cadherin J. Biol. Chem., August 26, 2005; 280(34): 30225 - 30235. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. C. Mansky, U. Sankar, J. Han, and M. C. Ostrowski Microphthalmia Transcription Factor Is a Target of the p38 MAPK Pathway in Response to Receptor Activator of NF-kappa B Ligand Signaling J. Biol. Chem., March 22, 2002; 277(13): 11077 - 11083. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
