 |
INTRODUCTION |
Myelopoiesis is characterized by differentiation stage-specific
transcription of defined sets of genes, resulting in acquisition of the
mature myeloid phenotype (1). For example, early myeloid differentiation is characterized by transcription of genes encoding hematopoietic cytokines, midmyeloid differentiation is characterized by
transcription of granule protein genes, and late myeloid
differentiation is characterized by transcription genes encoding
phagocyte respiratory burst oxidase proteins (1-3). Some
differentiation stage-specific myeloid gene transcription is regulated
by differentiation stage-restricted transcription factors, such as AML1
and CCAAT/enhancer-binding protein (1). However, other transcription
factors, such as PU.1, interferon consensus sequence-binding protein
(ICSBP),1 and HoxA10, are
involved in gene transcription throughout myelopoiesis but have
distinct differentiation stage-specific functions.
Recent investigations of myeloid gene transcription indicate that
post-translational modification alters function of some transcription
factors as myeloid differentiation proceeds. For example,
nonphosphorylated PU.1 interacts with AML1 and activates gene
transcription in early myeloid differentiation (4). In differentiated
myeloid cells, serine-phosphorylated PU.1 interacts with interferon
regulatory factors and activates transcription of genes characteristic
of the mature phenotype (5-7). The ICSBP is another factor with
differentiation stage-specific activity that depends on
post-translational modification. In undifferentiated myeloid cells,
nonphosphorylated ICSBP represses BCLX transcription (8). In
differentiated myeloid cells, tyrosine-phosphorylated ICSBP interacts
with PU.1, IRF1, and CBP to activate transcription of various
myeloid-specific genes (5-7).
The homeodomain protein HoxA10 is also functionally regulated by
post-translational modification during myeloid differentiation. Understanding the function of this transcription factor is important, because disregulation of HoxA10 is associated with leukemogenesis in
murine models and human disease (9, 10). In undifferentiated myeloid
cells, HoxA10 blocks differentiation by repressing transcription of
myeloid-specific genes, such as the genes encoding the respiratory burst oxidase proteins gp91PHOX and p67PHOX
(CYBB and NCF2 genes, respectively) (11). During
cytokine-induced differentiation of myeloid cell lines, HoxA10 is
tyrosine-phosphorylated, which decreases HoxA10 DNA binding affinity
(11). These results suggest that HoxA10 tyrosine phosphorylation
provides a mechanism for differentiation stage-specific HoxA10 function.
Previous investigations suggest that SHP1 protein-tyrosine phosphatase
(SHP1-PTP) regulates differentiation stage-specific transcription
factor phosphorylation. SHP1 is a "hematopoietic specific" PTP that
antagonizes myeloid differentiation (14-16). Consistent with this,
inhibition of SHP1-PTP activity, in undifferentiated myeloid cell
lines, increases transcription of the CYBB and
NCF2 genes (17). This suggests the hypothesis that SHP1-PTP
substrates include transcription factors involved in CYBB
and NCF2 transcription.
The current studies investigate regulation of HoxA10 tyrosine
phosphorylation and repression activity. We determine that interaction of phosphotyrosine residues in the HoxA10-homeodomain with an adjacent
domain in the HoxA10 protein decreases DNA binding affinity. Our
current studies indicate that SHP1-PTP activity impacts HoxA10-mediated transcriptional repression in myeloid cell lines. We also determine that SHP1-PTP activity influences the HoxA10 tyrosine phosphorylation state in myeloid cell lines and ex vivo cultured myeloid
cells from SHP1-deficient mice. Therefore, these investigations link post-translational modification that regulates HoxA10 function with a
phosphatase intermediate that effects this modification.
| |
MATERIALS AND METHODS |
Plasmids and PCR Mutagenesis: Reporter Constructs and
Plasmids for Protein Expression--
Artificial promoter/reporter
constructs were generated as previously described (18) in the minimal
promoter/reporter vector, p-TATACAT (19) (obtained from Dr. A. Kraft,
University of Colorado, Denver, CO). Constructs were generated with
three copies (in the forward orientation) of the consensus sequence for
HoxA10/Pbx binding (p-a10TATACAT) (2) or four copies (in the forward
direction) of the 
94 to 134 bp sequence from the CYBB
promoter (p-cybbTATACAT). This CYBB promoter sequence has
previously been demonstrated to function as a repressor element in
myeloid cell lines (11).
The cDNA for human HoxA10 was obtained from C. Largman (University
of California, San Francisco, CA) (20). This cDNA sequence represents the major transcript that is encoded in mammalian
hematopoietic cells, encoding a 393-amino acid, 55-kDa protein (21).
Wild type HoxA10 cDNA sequence was subcloned into the pcDNAamp
vector for in vitro translation and the pSR
and
pcDNA3.1his vectors for expression in mammalian cells (22).
HoxA10 5' truncation mutant cDNAs were generated by PCR, using
primers that incorporate the genuine HoxA10 Kozak consensus sequence
and an ATG. Truncation mutants were generated that included HoxA10
amino acids (aa) 140-393, 195-393, 267-393, 267-320, and 320-393.
These mutant cDNA sequences were subcloned into the pcDNAamp vector for in vitro translation. The 267-320 aa HoxA10
truncation mutant was subcloned into the pGEX3 vector for expression in
Escherichia coli as a GST fusion protein. A HoxA10 cDNA
sequence with point mutation of the two tyrosine residues in the
homeodomain (tyrosines 326 and 343 to phenylalanine) was generated by
site-directed mutagenesis, using the CLONTECH
QuikChange protocol (per the manufacturer's instructions
(Stratagene)). Y326F/Y343F HoxA10 cDNA was subcloned into
the pcDNAamp vector for in vitro translation and into
the pSR and pcDNA3.1his vectors for transfection experiments.
Mutant cDNAs were sequenced to verify that no unintended mutations
had been introduced.
The cDNAs for SHP1-PTP and CS453-SHP1-PTP (dominant negative),
subcloned into the pSX vector, were obtained from Dr. Stuart Frank
(Birmingham Veterans Affairs Hospital and the Department of Medicine,
University of Alabama, Birmingham) (24). These SHP1 cDNAs were also
subcloned into the pcDNAamp vector for in vitro translation.
Oligonucleotides--
Oligonucleotides were synthesized by the
Core Facility of the University of Alabama, Birmingham Comprehensive
Cancer Center, or the Core Facility of the Robert H. Lurie
Comprehensive Cancer Center at Northwestern University.
Oligonucleotides were as follows: derived consensus sequence for
Pbx/HoxA10 binding (dsA10),
5'-tgcgatgatttatgaccgc-3'; the similar sequence
from the CYBB promoter (94 to 134 bp) (dscybbA10) (25),
5'-ttcagttgaccaatgattattagccaattttctgataaaa-3'. In these oligonucleotides, the HoxA10 core is in boldface type, the Pbx
core is in italics, and CCAAT boxes are underlined.
Myeloid Cell Line Culture--
The human myelomonocytic cell
line U937 (25) was obtained from Andrew Kraft. Cells were maintained
and differentiated as described (5, 6, 11). U937 cells were treated
with 200 units/ml human recombinant IFN
(Roche Molecular Biochemicals).
Electrophoretic Mobility Shift Assays (EMSAs)--
Nuclear
extract proteins were prepared by the method of Dignam (26) with
protease inhibitors (as described) (17) and phosphatase inhibitors (as
indicated). Oligonucleotide probes were prepared, and EMSA and antibody
supershift assays were performed as described (11). Antiserum to HoxA10
(not cross-reactive with other Hox proteins) was obtained from Covance
Research Products (Richmond, CA) and from Santa Cruz Biotechnology,
Inc. (Santa Cruz, CA).
In Vitro Translated Proteins and Tyrosine
Dephosphorylation--
In vitro transcribed HoxA10 and
Y326F/Y343F HoxA10 mRNA were generated from linearized template
DNA, using the Riboprobe system, according to the manufacturer's
instructions (Promega, Madison, WI). In vitro translated
proteins were generated in rabbit reticulocyte lysate, according to the
manufacturer's instructions (Promega). Control (unprogrammed) lysates
were generated in similar reactions in the absence of input RNA.
In vitro translated proteins and nuclear proteins were
tyrosine-dephosphorylated with Yop protein-tyrosine
phosphatase (New England Biolabs, Beverly, MA). Proteins (either 10 µl of in vitro translated protein or 2 µg of nuclear
proteins) were incubated for 30 min at 30 °C, in a 20-µl reaction
volume with 50 units of Yop and 1× reaction buffer,
according to the manufacturer's instructions. Control proteins were
incubated, similarly, in 1× reaction buffer, without Yop.
In other experiments, in vitro translated HoxA10 proteins
were incubated as above but in reactions in which 10 µl of in
vitro translated SHP1 or CS453-SHP1 was substituted for
Yop PTP. In vitro translated proteins were
assayed for tyrosine phosphorylation state by anti-phosphotyrosine
immunoprecipitation of [35S]methionine-labeled proteins
(see below).
EMSA with the in vitro translated proteins was performed as
described (11). The amount of mutant and wild type HoxA10 proteins in
DNA-binding reactions was equalized by SDS-PAGE of
[35S]methionine-labeled proteins.
Transfection and Reporter Gene Assays--
Cells were
transfected by electroporation as described (11). U937 cells (32 × 106/sample) were transfected with 70 µg of p-TATACAT
or p-cybbTATACAT; 30 µg of pSR, HoxA10/pSR, or Y326F/Y343F
HoxA10/pSR; and 15 µg of p-CMV
-gal (to normalize for
transfection efficiency). Transfectants were incubated for 24 h at
37 °C, 5% CO2, followed by 24 h with or without
IFN (200 units/ml). Preparation of cell extracts, -galactosidase,
and chloramphenicol acetyltransferase (CAT) assays was performed as
described (27, 28).
To generate stable transfectants, U937 cells (32 × 106) were transfected with 50 µg of plasmid: SHP1/pSX,
CS453-SHP1/pSX, or control pSX. Cells were incubated in Dulbecco's
modified Eagle's medium, 10% fetal calf serum, 1%
penicillin/streptomycin, 1 mg/ml Geneticin (G418) to select for stable
transfectants. Expression of the various proteins was verified by
Western blots of cell lysates, according to standard procedures, with
commercially available antibodies. Functional expression of the
proteins was verified by a protein-tyrosine phosphatase assay, using a
commercially available kit (Upstate Biotechnology, Inc., Lake Placid,
NY). These U937 stable transfectants have been previously described (17).
Ex Vivo Culture of Murine Bone Marrow-derived Myeloid
Cells--
Bone marrow cells were obtained from the femurs of wild
type C57BL/6J or C57BL/6J-Hcphme-v (viable
moth-eaten or SHP1
/) mice, obtained from Jackson
Laboratories (Bar Harbor, MA). Bone marrow myeloid progenitor cells
were recovered and ex vivo cultured in cytokines, as
described in Ref. 29. Bone marrow-derived cells were cultured for 48 or
72 h in Dulbecco's modified Eagle's medium with 10% fetal calf
serum and 10% 5× WEHI conditioned medium, supplemented with
recombinant murine GM-CSF (10 ng/ml; Upstate Biotechnology). These
conditions result in a population of predominantly myeloid progenitor
cells (29).
Immunoprecipitation and Western
Blotting--
Immunoprecipitations were performed with 100 µg of
nuclear proteins extracted from U937 stable transfectants with
SHP1-PTP, CS453-SHP1, or vector control. Proteins were
immunoprecipitated with HoxA10 antiserum (Covance Research Products) or
control rabbit preimmune serum (11). Proteins were immunoprecipitated
for 4 h at 4 °C, with 2 µl of HoxA10 antiserum or 2 µl of
control rabbit preimmune serum, followed with a 1-h incubation with 30 µl of 50% staphylococcus protein A-Sepharose bead slurry, as
described (17). Immunoprecipitated proteins were washed with RIPA
buffer, eluted in SDS sample buffer, and separated on 12% SDS-PAGE.
Immunoprecipitated proteins were transferred to nitrocellulose, and the
blots were sequentially probed with anti-HoxA10 (goat polyclonal; Santa
Cruz Biotechnology) or anti-phosphotyrosine antibodies (murine
monoclonal 4G10; Upstate Biotechnology). Immunoreactive proteins were
detected by chemiluminescence, according to the manufacturer's
instructions (Amersham Biosciences).
Immunoprecipitation experiments were also performed with in
vitro translated HoxA10 proteins, with and without preincubation with SHP1 or CS435-SHP1. In these experiments, 20 µl of in
vitro translated, [35S]methionine-labeled HoxA10
protein was preincubated with 10 µl of in vitro
translated, unlabeled SHP1 protein (in Yop buffer, as
described above). In vitro translated proteins were diluted into "denaturing lysis buffer" and subjected to boiling, followed by dilution with RIPA buffer with protease inhibitors, as described (11). Proteins were incubated with either 1 µl of
anti-phosphotyrosine antibody (4G10; Upstate Biotechnology) or
irrelevant antibody (mouse anti-rabbit IgG) for 4 h at 4 °C,
followed by a 1-h incubation with 15 µl of 50% staphylococcus
protein A-Sepharose bead slurry. Beads were washed with RIPA buffer,
and proteins were eluted in SDS sample buffer, separated on 12 or 15%
SDS-PAGE. The SDS gels were fixed and dried, and immunoprecipitating
HoxA10 proteins were identified by autoradiography.
Immunoprecipitation experiments to detect the tyrosine
phosphorylation state of HoxA10 were performed with U937 cells, U937 transfectants, or ex vivo cultured murine bone marrow cells.
In some experiments, cells were disrupted in "denaturing lysis
buffer" (U937 transfectants with HoxA10/pcDNA31.his or
Y326F/Y343F HoxA10/pcDNA3.1his or murine bone marrow cells),
as previously described (17). In other experiments, U937 nuclear
proteins were diluted with denaturing lysis buffer and subjected to
boiling (17). Cellular proteins (100-300 µg) were diluted with RIPA
buffer (with protease inhibitors), incubated with either HoxA10
antiserum (rabbit polyclonal from Covance) or anti-phosphotyrosine
antibody (murine monoclonal 4G10; Upstate Biotechnology) (or irrelevant
control antibody, as described above), and collected with
staphylococcus protein A-Sepharose, as described (17).
Immunoprecipitates were separated by 12% SDS-PAGE, and proteins were
detected by Western blotting with HoxA10 antibody (goat polyclonal;
Santa Cruz Biotechnology) or anti-phosphotyrosine antibody, as above.
GST Fusion Protein "Pull-down" Assays--
JM109 E. coli transformed with 267-320-amino acid HoxA10/pGEX3 or control
pGEX3 was grown to log phase, supplemented with 0.1 mM
isopropyl-1-thio--D-galactopyranoside, and incubated for 3 h at 37 °C with shaking. The cells were harvested and
resuspended in HN buffer (20 mM HEPES (pH 7.4), 0.1 M NaCl, 2 mM MgCl2, 0.1 mM EDTA, 0.5% Nonidet P-40, 0.1% Triton X-100, 2 mM phenylmethylsulfonyl fluoride, 5 mM NaF) and
sonicated on ice. Debris was removed by centrifugation, and the lysate
was incubated for 30 min at 4 °C with glutathione-agarose beads
(Sigma) and washed extensively with HN buffer. The beads were
preincubated for 30 min at 4 °C with 5 µl of control rabbit
reticulocyte lysate and then for 1 h with 20 µl of
[35S]methionine-labeled in vitro translated
protein and washed extensively in HN buffer. Proteins were eluted with
SDS-PAGE sample buffer and separated on 15% SDS-PAGE, and an
autoradiograph was performed.
| |
RESULTS |
Phosphotyrosine Residues in the HoxA10 Homeodomain Regulate
DNA-binding Affinity--
Previously, we found that HoxA10 tyrosine
phosphorylation decreases HoxA10 binding to the CYBB
repressor element, the derived HoxA10/Pbx1 DNA-binding consensus
sequence, and a homologous sequence in the NCF2 promoter
(11). We demonstrated that IFN-induced differentiation of U937
myeloid cells results in HoxA10 tyrosine phosphorylation (Fig.
1A) (11), which is temporally
associated with decreased HoxA10 DNA binding and increased
CYBB and NCF2 transcription (11). We concluded
that post-translational modification modulates HoxA10 function
during myeloid differentiation. To understand regulation of HoxA10
transcriptional repression activity during myelopoiesis, it will be
important to determine the mechanism by which tyrosine phosphorylation
alters HoxA10 DNA binding affinity. In these investigations, we
initially approached this question by identifying phosphotyrosine
residues that modulate HoxA10 DNA binding affinity.
View larger version (36K):
[in this window]
[in a new window]
|
Fig. 1.
Multiple HoxA10 tyrosine residues are
phosphorylated during IFN-induced
differentiation of myeloid cells. A, HoxA10 is
tyrosine-phosphorylated during IFN differentiation of U937 myeloid
cells. U937 cells were cultured for 48 h, with and without IFN
(200 units/ml). Nuclear proteins were isolated and immunoprecipitated
with an antibody to HoxA10, under conditions that do not co-precipitate
other proteins. Proteins were analyzed by Western blot, as indicated.
IFN differentiation increases HoxA10 tyrosine phosphorylation state
but not total HoxA10 protein, consistent with our previous data (11).
B, distribution of tyrosine residues in the HoxA10 protein.
Graphic representation of the HoxA10 protein; tyrosine residues are
indicated with an asterisk, and the homeodomain is indicated
as HD. Location of the HoxA10 truncation mutants used in
these investigations relative to the tyrosine residues and the
DNA-binding homeodomain is also indicated.
|
|
To identify tyrosine residues involved in HoxA10 DNA binding, we
generated a series of HoxA10 cDNA truncation mutants. These mutant
cDNAs were designed to maintain the integrity of the HoxA10 DNA-binding homeodomain (amino acids 320-393) but sequentially delete
the 12 HoxA10 tyrosine residues. Truncated cDNAs were generated encoding HoxA10 amino acids 140-393, 195-393, and 267-393. The 140-393 truncation includes 7 of the 12 HoxA10 tyrosine residues, 195-393 includes 4 tyrosine residues, and the 267-393 truncation includes only the 2 tyrosine resides in the homeodomain (see Fig. 1B). These mutant HoxA10 cDNAs were used to generate
in vitro translated proteins in rabbit reticulocyte lysate.
These proteins were used in an EMSA with HoxA10/Pbx1-binding DNA
probes, as we have described (11).
Previously, we determined that in vitro translated HoxA10 is
tyrosine-phosphorylated and that Yop tyrosine phosphatase
treatment decreases HoxA10 tyrosine phosphorylation (11). In those
investigations, Yop tyrosine phosphatase treatment increased
in vitro binding of HoxA10 to DNA probes representing either
the CYBB repressor element or the derived HoxA10/Pbx1 DNA
binding consensus (11). To identify these regulatory phosphotyrosine
residues, we performed EMSA with HoxA10-DNA-binding probes and in
vitro translated HoxA10-truncation mutants, with and without
Yop tyrosine phosphatase treatment. In initial control
experiments, tyrosine phosphorylation of these HoxA10 truncation
mutants was demonstrated by anti-phosphotyrosine immunoprecipitation.
Additionally, Yop tyrosine phosphatase decreased tyrosine
phosphorylation of each of these truncation mutants (not shown).
As in our previous investigations, Yop treatment
increases binding of in vitro translated, wild type HoxA10
to DNA probes representing the derived HoxA10/Pbx1 consensus sequence
and the CYBB repressor element. Similarly, Yop
treatment increases DNA binding of all three HoxA10 truncation mutants
to these two HoxA10-binding DNA probes (shown for the 193-393 and
267-393 HoxA10 mutants (Fig. 2A)). Since only the
homeodomain tyrosine residues are common to all of these HoxA10
truncation mutants, we hypothesized that phosphorylation of tyrosine
residues in the HoxA10 homeodomain decreases DNA binding
affinity.
View larger version (40K):
[in this window]
[in a new window]
|
Fig. 2.
Phosphorylation of tyrosine residues in the
HoxA10 homeodomain decreases DNA binding affinity and transcriptional
repression. A, phosphorylation of the tyrosine residues in
the homeodomain decreases HoxA10 DNA binding affinity. EMSA was
performed with a DNA probe representing the derived HoxA10/Pbx1 binding
consensus sequence and HoxA10 truncation mutants, in vitro
translated in rabbit reticulocyte lysate. In vitro
translated proteins were incubated with or without Yop
tyrosine phosphatase as follows: 1) control HoxA10 aa 195-393; 2)
Yop-treated HoxA10 aa 195-393; 3) control HoxA10 aa
267-393; 4) Yop-treated HoxA10 aa 267-393. The
arrows indicate specific binding of in vitro
translated proteins. B, mutation of tyrosine residues 326 and 343 to phenylalanine abolishes Yop PTP-induced increase
in HoxA10 DNA binding affinity. EMSA was performed with a DNA probe
representing the derived HoxA10/Pbx1 binding consensus sequence and
HoxA10 or Y326/344F HoxA10, in vitro translated in rabbit
reticulocyte lysate. In vitro translated proteins were
incubated with or without Yop tyrosine phosphatase as
follows: 1) control HoxA10, 2) Yop-treated HoxA10; 3)
control Y326F/Y343F HoxA10; 4) Yop-treated Y326F/Y343F
HoxA10. The arrow indicates specific DNA binding of HoxA10
protein. C, HoxA10 with mutation of the homeodomain tyrosine
residues represses transcription in IFN-differentiated U937 cells.
U937 cells were co-transfected with an artificial promoter construct
with four copies of the repressor element from the CYBB gene
(p-cybbTATACAT) or empty vector control (p-TATACAT) and a vector to
overexpress HoxA10, Y326F/Y343F HoxA10, or empty vector control
(pSR). Reporter gene activity was assayed with and without
IFN-induced differentiation of the transfectants. Wild type
HoxA10 represses reporter expression from the
CYBB repressor element in undifferentiated but not
differentiated U937 cells, consistent with our previous results. In
contrast, Y326F/Y343F HoxA10 represses the CYBB repressor
containing an artificial promoter construct in U937 cells with and
without IFN differentiation. In undifferentiated U937 transfectants,
transcriptional repression by Y326F/Y343F HoxA10 was significantly more
efficient than wild type HoxA10. D, overexpressed HoxA10 and
Y326F/Y343F HoxA10 are tyrosine-phosphorylated during IFN
differentiation of U937 cells. U937 cells were transfected with a
vector to overexpress epitope-tagged HoxA10 or Y326F/Y343F HoxA10 (or
vector control) and incubated for 48 h, with or without IFN.
Cell lysates were immunoprecipitated with antibodies to the epitope tag
(anti-xpress tag and anti-His6 tags) under conditions that
do not co-precipitate other proteins. Western blots were analyzed for
tyrosine phosphorylation of the immunoprecipitated proteins, as
indicated. Both wild type and Y326F/Y343F HoxA10 were somewhat
tyrosine-phosphorylated in undifferentiated U937 cells, and tyrosine
phosphorylation (but not protein abundance) of both overexpressed
proteins was increased by IFN treatment of the transfectants.
|
|
To specifically investigate the role of the two HoxA10 homeodomain
tyrosine residues in DNA binding affinity, we mutated these residues
(Tyr326 and Tyr343) to phenylalanine
(Y326F/Y343F HoxA10). Y326F/Y343F HoxA10 was in vitro
translated in reticulocyte lysate and used in EMSA with HoxA10-binding
DNA probes, as above. Consistent with our hypothesis, Y326F/Y343F
HoxA10 demonstrated increased binding to these DNA probes, in
comparison with an equivalent amount of wild type HoxA10 (Fig.
2B). Additionally, Yop treatment did not further
increase DNA binding of Y326F/Y343F HoxA10. Therefore, these
investigations suggested that we had correctly identified the
phosphotyrosine residues that decrease DNA binding of in
vitro translated HoxA10.
Phosphorylation of Homeodomain Tyrosine Residues Regulates HoxA10
Repression Activity during IFN-induced U937 Differentiation--
In
our previous studies, we determined that overexpression of HoxA10
represses expression of the endogenous CYBB and
NCF2 genes in undifferentiated U937 myeloid cells (11).
These genes encode the respiratory burst oxidase proteins
gp91PHOX and p67PHOX, which are expressed at
low levels in undifferentiated U937 cells. Consistent with this, we
previously found that HoxA10 overexpression represses reporter gene
expression from an artificial promoter/reporter construct containing a
minimal promoter linked to multiple copies of either the derived
HoxA10/Pbx1 binding consensus or the proximal repressor element from
the CYBB gene (referred to as p-a10TATACAT and
p-cybbTATACAT, respectively) (11).
IFN differentiation of U937 cells results in acquisition of
functional characteristics of mature myeloid cells, including respiratory burst activity and phagocytosis. Consistent with this, IFN differentiation of U937 cells increases transcription of the
CYBB and NCF2 genes (5). In our previous
investigations, IFN-differentiation of U937 cells abolished HoxA10
repression of reporter gene expression from HoxA10-binding artificial
promoter/reporter constructs (11). We hypothesized that the inability
of HoxA10 to repress transcription in IFN-treated U937 cells was due
to decreased DNA binding affinity of tyrosine-phosphorylated HoxA10 in
these cells.
Therefore, we were interested in determining whether phosphorylation of
tyrosine residues in the homeodomain regulates HoxA10 repression
activity in differentiating U937 cells. To test the functional
significance of HoxA10 Tyr326 and Tyr343, we
performed U937 transfection experiments, similar to those described
above for wild type HoxA10. If the phosphotyrosine residues that
influence HoxA10 DNA binding have been correctly identified, we
hypothesized that Y326F/Y343F HoxA10 would repress artificial promoter
constructs with HoxA10-DNA-binding sites in both undifferentiated and
IFN-differentiated U937 cells. We co-transfected U937 cells with an
artificial promoter/reporter construct with multiple copies of the
HoxA10 binding site from the CYBB gene (or control minimal promoter/reporter vector; p-TATACAT) and a plasmid to overexpress either wild type HoxA10 or Y326F/Y343F HoxA10.
Consistent with our previous results, reporter expression from
p-cybbTATACAT was significantly repressed by overexpression of wild
type HoxA10, in undifferentiated U937 cells (54.9 ± 18.8% decrease in CAT activity; p = 0.02, n = 3). Additionally, overexpressed Y326F/Y343F HoxA10 significantly
represses reporter expression from the HoxA10-binding artificial
promoter construct (p-cybbTATACAT) (90.9 ± 18.2% decrease in CAT
activity; p = 0.0086, n = 3) (Fig. 2C). Interestingly, overexpressed Y326F/Y343F HoxA10
represses p-cybbTATACAT significantly more efficiently than
overexpressed wild type HoxA10 (79.8 ± 8.2% decrease in CAT
activity with Y326F/Y343F Hoxa10 versus wild type HoxA10;
p = 0.003, n = 3).
We next investigated whether overexpressed Y326F/Y343F HoxA10 represses
transcription from the HoxA10-binding artificial promoter construct in
IFN-treated U937 cells. According to our previous investigations,
HoxA10 is tyrosine-phosphorylated under these conditions (11).
Consistent with our previous results, overexpressed wild type HoxA10
does not significantly repress reporter expression from p-cybbTATACAT
in IFN-differentiated U937 cells (p = 0.17, n = 3) (Fig. 2C). In contrast, overexpressed
Y326F/Y343F HoxA10 significantly represses p-cybbTATACAT reporter
expression in IFN-differentiated U937 transfectants (55.1 ± 6.6% decrease in CAT activity with Y326F/Y343F HoxA10
versus vector control; p = 0.002, n = 3). These results are consistent with our
hypothesis that phosphorylation of the two homeodomain tyrosine
residues decreases HoxA10 DNA binding affinity, and therefore
transcriptional repression activity, during myeloid differentiation.
However, an alternative explanation for these results is that mutation
of tyrosines 326 and 343 prevents phosphorylation of the remaining 10 HoxA10 tyrosine residues. To investigate this, we determined the
tyrosine phosphorylation state of overexpressed wild type and
Y326F/Y343F HoxA10 in U937 cells. U937 cells were transfected with
mammalian expression vectors to overexpress an epitope-tagged form of
wild type or Y326F/Y343F HoxA10 (or empty vector control). In these
experiments, the pcDNA3.1his vector was used to express HoxA10 with
six-histidine and "xpress" epitope tags. The transfectants were
incubated for 48 h, either with or without IFN treatment, as in
the experiments described above. Cells were harvested and lysed under
conditions that prevent co-precipitation of other proteins. Lysates
were subjected to anti-tag immunoprecipitation, and Western blots were
evaluated for tyrosine phosphorylation state of the overexpressed,
epitope-tagged protein (Fig. 2D).
We found that there was some tyrosine phosphorylation of overexpressed
wild type and Y326F/Y343F HoxA10 in U937 cells. However, this tyrosine
phosphorylation increased during IFN differentiation of the
transfectants (although the abundance of the overexpressed protein did
not). This result is consistent with mutation of tyrosine residues
326/343 resulting in a specific, but not general, decrease in HoxA10
tyrosine phosphorylation. These results also indicate that HoxA10
overexpression results in some tyrosine phosphorylation. This is
consistent with overexpression of other transcription factors involved
in myeloid gene transcription (17), as will be discussed below.
Phosphorylated Homeodomain Tyrosine Residues Interact with HoxA10
Amino Acids 267-320--
Therefore, our data suggest that
phosphorylation of the two homeodomain tyrosine residues decreases
HoxA10 DNA binding affinity during IFN differentiation of U937
cells. However, this new data do not immediately appear consistent with
data from our previous investigations. Previously, we investigated the
function of an alternatively spliced form of HoxA10, referred to as
short A10 (21). Short A10 transcripts are found in myeloid leukemia
cell lines but not in primary leukemia cells or in normal myeloid cells (21). In the short A10 transcript, a unique sequence encoding 4 amino
acids not found in "wild type" HoxA10 is spliced to sequence encoding HoxA10 amino acids 317-393 (21). Therefore, the 15-kDa short
A10 protein includes the DNA-binding homeodomain (including Tyr326 and Tyr343 of HoxA10). Our previous
investigations indicate that Yop tyrosine phosphatase
treatment did not increase DNA binding affinity of in vitro
translated short A10 (11). In transfection experiments, short A10
repressed reporter expression from artificial promoter constructs with
HoxA10 DNA-binding sites, although significantly less efficiently than
the 55-kDa form of HoxA10 (11). Consistent with these in
vitro DNA binding results, short A10 repression activity was not
abrogated by IFN-differentiation of U937 cells. These results
suggest that the phosphotyrosine residues that decrease HoxA10 DNA
binding affinity are not within the short A10 protein.
However, the short A10 protein includes the two HoxA10 homeodomain
tyrosine residues found to be involved in DNA binding affinity in the
experiments above. Specifically, the short A10 EMSA results directly
contradict our results with the Yop-treated amino acid 267-393 HoxA10 truncation mutant. Comparison of short A10 and 267-393
aa HoxA10 cDNA sequences indicates that short A10 includes 4 amino
acids that are not found in 267-393 aa HoxA10 and that 267-393 aa
HoxA10 encodes 46 amino acids not present in short A10. We hypothesized
that these amino acids might influence DNA binding activity, resulting
in differences between the two truncated HoxA10 proteins. To
investigate this possibility, we generated an additional mutant HoxA10
cDNA sequence, which encodes amino acids 320-393. In
vitro translated 320-393 aa HoxA10 was used in EMSA with
HoxA10/Pbx binding DNA probes, as above.
We were somewhat surprised to discover that Yop tyrosine
phosphatase treatment did not increase in vitro binding of
320-393 aa HoxA10 to HoxA10-binding DNA probes (Fig.
3A). Consistent with this,
320-393 aa HoxA10 had increased DNA binding affinity, in comparison
with 55-kDa HoxA10. These results suggest that HoxA10 DNA binding
affinity is determined both by tyrosine residues in the homeodomain and
by amino acids 267-320. A possible mechanism for this is physical
interaction between phosphotyrosine residues 326/343 and amino acids
267-320. Such an interaction could alter accessibility of the HoxA10
homeodomain for potential DNA-binding sites. The corollary of this
hypothesis would be that nonphosphorylated Tyr326/Tyr343 does not interact with amino
acids 267-393.
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 3.
Amino acids 267-320 are necessary for
phosphotyrosine residues in the homeodomain to decrease HoxA10 DNA
binding affinity. A, HoxA10 sequences outside of the
homeodomain contribute to HoxA10 DNA binding affinity. EMSAs were
performed with a DNA probe representing the HoxA10/Pbx DNA-binding
consensus sequence and HoxA10 truncation mutants, in vitro
translated in rabbit reticulocyte lysate. In vitro
translated proteins were incubated with or without Yop
tyrosine phosphatase, as follows: 1) control HoxA10 aa 320-393
(homeodomain only); 2) Yop-treated HoxA10 aa 320-393; 3)
control HoxA10 aa 267-393; 4) Yop-treated HoxA10 aa
267-393. The arrows indicate specific shifted proteins.
B, interaction of HoxA10 amino acids 267-320 with amino
acids 320-393 requires phosphorylation of tyrosines 326 and 343. HoxA10 aa 276-320 were expressed in E. coli as a fusion
protein with GST. HoxA10 aa 320-393 and Y326F/Y343F 320-393 were
in vitro translated in rabbit reticulocyte lysate.
Interaction of HoxA10 267-320/GST or GST with HoxA10 320-393 or
Y326F/Y343F HoxA10 320-393 was identified by co-precipitation. HoxA10
267-320/GST interacts with HoxA10 320-393 but not Y326F/Y343F HoxA10
320-393, as indicated. Neither in vitro translated protein
interacts with control GST.
|
|
To investigate this hypothesis, we determined the ability of the HoxA10
homeodomain to directly interact with HoxA10 amino acids 267-320
in vitro. HoxA10 amino acids 267-320 were expressed in
E. coli as a fusion protein with glutathione
S-transferase (267-320 aa Hoxa10/GST) and purified by
affinity to glutathione-agarose. Interaction of HoxA10 amino acids
267-320 with 320-393 was demonstrated by co-purification of in
vitro translated, [35S]methionine-labeled 320-393
aa HoxA10 by 267-320 aa HoxA10/GST, but not control GST (Fig.
3B). This result is consistent with the first part of our
hypothesis, that the HoxA10 amino acids 267-320 interact with the
HoxA10 homeodomain. Therefore, we determined whether this interaction
was dependent on phosphorylation of the tyrosine residues in the homeodomain.
To do this, we investigated the ability of 267-320 aa HoxA10/GST to
affinity co-purify in vitro translated,
35S-labeled 320-393 aa Y326F/Y343F HoxA10 (a truncation
mutant with mutation of tyrosine residues 326 and 343 to
phenylalanine). In contrast to "wild type" 320-393 aa HoxA10,
Y326F/Y343F 320-393 aa HoxA10 did not affinity-co-purify with 267-320
aa HoxA10/GST (Fig. 3B). In initial control experiments, we
demonstrated that in vitro translated 320-393 aa HoxA10
(but not Y326F/Y343F 320-393 aa HoxA10) is tyrosine-phosphorylated (as
demonstrated below). This result suggests that the interaction of these
two HoxA10 domains is dependent on phosphorylation of the homeodomain
tyrosine residues. Therefore, this result is consistent with the second part of our hypothesis, that differentiation induced phosphorylation of
tyrosine residues in the HoxA10 homeodomain induces interaction with
another domain in the HoxA10 protein.
SHP1-PTP Activity Increases HoxA10 Transcriptional Repression in
Undifferentiated U937 Cells--
In previous investigations, we found
that activity of SHP1 protein-tyrosine phosphatase decreases
transcription of the CYBB and NCF2 genes in
undifferentiated myeloid cells (17). These results are consistent with
the phenotype of SHP1-PTP-deficient (viable moth-eaten, or me
v/) mice, which is characterized by an increase in
differentiating myeloid cells with activated neutrophils and monocytes
in the peripheral blood. Additionally, ex vivo cultured
myeloid cells from SHP1/ mice (me v/
mice) undergo accelerated differentiation to mature phagocytes in the
presence of cytokines, in comparison with normal murine myeloid cells
(12-14).
Consistent with this murine model, we previously found increased
CYBB and NCF2 transcription in undifferentiated
U937 cells, stably overexpressing dominant negative SHP1 (CS453-SHP1)
(17). Additionally, we found decreased CYBB and
NCF2 transcription in IFN-differentiated U937 cells,
stably overexpressing SHP1 (see Ref. 17 for characterization of these
lines). We found that overexpression of CS453-SHP1 increases, and
overexpression of SHP1 decreases, reporter gene expression from
constructs with either the CYBB or NCF2 promoter
(17). In these previous studies, we found abundant endogenous and
overexpressed SHP1 protein in the nucleus, suggesting that
transcription factors would be reasonable SHP1-PTP substrates
(17).
Based on these previous investigations, we hypothesized that SHP1-PTP
might decrease HoxA10 tyrosine phosphorylation. If this hypothesis is
correct, SHP1-PTP would increase HoxA10 transcriptional repression,
decreasing CYBB and NCF2 transcription in
undifferentiated myeloid cells. Such a result would link SHP1-PTP
activity to transcriptional repression of myeloid-specific genes,
identifying a mechanism regulating HoxA10 function. To investigate
this, we used U937 stable transfectants overexpressing SHP1,
CS453-SHP1, or vector control (pSX) (17). These U937 stable
transfectants were co-transfected with a vector to overexpress HoxA10
and an artificial promoter construct with multiple copies of a
HoxA10/Pbx binding site linked to a minimal promoter and reporter
(p-cybbTATACAT, described above).
We were intrigued to find that overexpression of SHP1 significantly
increases the ability of overexpressed HoxA10 to repress reporter gene
activity via the repressor element from the CYBB gene
(33.2 ± 6.2% decrease in CAT activity in the presence of SHP1
overexpression; p = 0.027, n = 3) (Fig.
4A). Consistent with this,
overexpression of CS453-SHP1 significantly decreases the efficiency of
HoxA10 repression of the p-cybbTATACAT construct (73.3 ± 10.2%
increase in CAT activity in the presence of dominant negative SHP1
overexpression; p = 0.031, n = 3).
Indeed, there was no significant difference in p-cybbTATACAT reporter
gene activity with and without overexpression of HoxA10, in U937 cells
overexpressing dominant negative SHP1 (HoxA10 versus control
vector; p = 0.34, n = 4). These results
suggest that SHP1-PTP activity influences HoxA10 repression
activity.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 4.
SHP1-PTP activity in
U937 cells increases HoxA10 transcriptional repression and DNA binding
but decreases HoxA10 tyrosine phosphorylation. A, inhibition
of SHP1-PTP in undifferentiated U937 cells increases HoxA10
transcriptional repression activity. U937 cells stably overexpressing
SHP1, CS453-SHP1, or vector control were co-transfected with an
artificial promoter construct with four copies of the repressor element
from the CYBB gene (p-cybbTATACAT) or empty vector control
(p-TATACAT) and a vector to overexpress HoxA10, Y326F/Y343F HoxA10, or
empty vector control (pSR). Reporter gene assays demonstrate that
overexpression of SHP1 increases and CS453-SHP1 decreases repression of
p-cybbTATACAT by overexpressed HoxA10 (but not control p-TATACAT). In
contrast, neither SHP1 nor CS435-SHP1 overexpression influences
Y326F/Y343F HoxA10-mediated repression of p-cybbTATACAT. B,
inhibition of SHP1-PTP activity in undifferentiated U937 cells
increases HoxA10 tyrosine phosphorylation. Nuclear proteins were
isolated from U937 cells stably expressing CS453-SHP1 or control vector
pSX. HoxA10 was immunoprecipitated from cell lysate proteins under
conditions that do not co-precipitate other proteins and separated by
SDS-PAGE. Western blots were probed with antibodies to phosphotyrosine
and to HoxA10, as indicated. HoxA10 tyrosine phosphorylation was
significantly increased in U937 cells overexpressing dominant negative
SHP1-PTP in comparison with control vector transfectants. C,
inhibition of SHP1-PTP activity in undifferentiated U937 cells
decreases HoxA10-Pbx1 complex DNA binding. EMSAs were performed with a
DNA probe representing the HoxA10/Pbx1 binding CYBB
repressor element and nuclear proteins isolated from U937 cells stably
expressing the following: 1) control pSX vector; 2) CS453-SHP1; 3)
SHP1-PTP. Inhibition of SHP1-PTP activity decreases and overexpression
of SHP1-PTP increases HoxA10/Pbx1 binding to the CYBB
repressor element in vitro.
|
|
If we have correctly identified HoxA10 tyrosine residues that mediate
DNA binding affinity, our data suggest that these residues would be
substrates for SHP1-PTP. Therefore, we investigated whether the
influence of SHP1-PTP activity on HoxA10 repression involves the HoxA10
homeodomain tyrosine residues (326 and 343). To do this, we
co-transfected U937 cells, stably overexpressing SHP1, CS452-SHP1, or
control vector, with a vector to overexpress Y326F/Y343F HoxA10 and the
p-cybbTATACAT construct, as above. Although Y326F/Y344F HoxA10
represses the HoxA10-binding artificial promoter construct, neither
overexpressed SHP1-PTP nor CS453-SHP1-PTP alters Y326F/Y343F HoxA10
repression of p-cybbTATACAT reporter expression (Fig. 4A). Therefore, these results are consistent with a role for SHP1-PTP in
regulation of HoxA10 repression activity via the homeodomain tyrosine residues.
SHP1-PTP Influences HoxA10 Tyrosine Phosphorylation in U937 Myeloid
Cells--
The above investigations imply that SHP1-PTP activity
influences HoxA10 tyrosine phosphorylation in undifferentiated U937 cells. Therefore, we investigated HoxA10 tyrosine phosphorylation state
in U937 cells overexpressing CS453-SHP1. Nuclear protein fractions were
isolated from these cells and immunoprecipitated with an antibody to
HoxA10 (rabbit polyclonal) under conditions that do not co-precipitate
other proteins. Immunoprecipitates were analyzed for HoxA10 tyrosine
phosphorylation by Western blot (Fig. 4B). Tyrosine
phosphorylation of HoxA10 immunoprecipitated from dominant negative
SHP1-expressing U937 cells is increased in comparison with HoxA10
immunoprecipitated from control (empty) vector transfectants. However,
there is no difference in abundance of total HoxA10 protein in these
two cell lines (empty vector versus dominant negative SHP1).
In our previous investigations, we used DNA probes representing the
CYBB proximal repressor element, a homologous
NCF2 sequence, or the derived HoxA10/Pbx1 consensus
DNA-binding site in EMSA with nuclear proteins from undifferentiated
myeloid cell lines. We found that a low mobility complex that interacts
with each of these DNA probes is immunoreactive with antibodies to
HoxA10 and Pbx1 (11). We also found that either IFN differentiation or sodium orthovanadate treatment of U937 cells decreases
HoxA10-Pbx1 complex binding (11). Consistent with these
observations, Yop tyrosine phosphatase treatment of U937
nuclear proteins decreases binding of this complex to HoxA10/Pbx
binding probes (11).
To determine the role of SHP1-PTP activity in HoxA10 DNA binding, we
performed EMSA with HoxA10/Pbx binding probes and nuclear proteins from
U937 cells overexpressing SHP1, CS453-SHP1, or vector control. In EMSA
with nuclear proteins from CS453-SHP1-overexpressing cells, in
vitro binding of the HoxA10-Pbx protein complex to HoxA10/Pbx binding probes is decreased (Fig. 4C), consistent with
increased HoxA10 tyrosine phosphorylation in these cells (Fig.
4B). In contrast, overexpression of SHP1 slightly increases
interaction of the HoxA10-Pbx complex to these probes.
SHP1-PTP Influences HoxA10 Tyrosine Phosphorylation in Cultured
Murine Myeloid Cells--
In our previous experiments, the influence
of SHP1-PTP activity on HoxA10 tyrosine phosphorylation could be a
manifestation of the U937 transformed phenotype. To address this, we
investigated HoxA10 tyrosine phosphorylation in ex vivo
cultured myeloid cells from viable moth-eaten mice (me
v/ mice), comparing the results with cells cultured
from control wild type mice (C57BL/CJ). Viable moth-eaten
mice lack SHP1 and manifest myeloid abnormalities. In these
investigations, we took advantage of previous work indicating that a
population of predominantly committed myeloid progenitors can be
isolated from normal murine bone marrow cells and cultured, short term,
in IL-3 and GM-CSF (29).
Bone marrow cells from SHP1/ mice and control mice were
cultured in IL-3 with GM-CSF, and cells were harvested after 48 and 72 h. Lysate proteins were analyzed for SHP1-PTP activity (as described (17)). SHP1-PTP activity was present in control cultured myeloid cells but not in SHP1/ cells, consistent with
previous investigations (23, 30). Cultured cell lysates were
immunoprecipitated with an anti-phosphotyrosine antibody, under
conditions that do not co-precipitate other proteins. At both time
points, abundance of tyrosine-phosphorylated HoxA10 was greater in
SHP1/ cells cultured than in cells cultured from wild
type mice (Fig. 5A). However,
it is possible that the absence of SHP1 increases HoxA10 protein in
these cells. To investigate that possibility, we performed another
series of immunoprecipitation experiments.
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 5.
Absence of SHP1-PTP increases HoxA10 tyrosine
phosphorylation in ex vivo cultured murine myeloid
cells. A, tyrosine-phosphorylated HoxA10 is increased in
ex vivo cultured bone marrow cells from viable moth-eaten
mice in comparison with control mice. Murine bone marrow myeloid cells
were cultured in IL-3 and GM-CSF for 48 and 72 h. Cell lysates
were immunoprecipitated with antibody to phosphotyrosine, under
conditions that do not co-precipitate other proteins. Western blots
were probed with antibody to HoxA10, as indicated. Increased
tyrosine-phosphorylated HoxA10 is present in ex vivo
cultured bone marrow cells from viable moth-eaten mice in comparison
with control C57BL mice. B, absence of SHP1-PTP increases
HoxA10 tyrosine phosphorylation, but not HoxA10 protein abundance, in
cultured murine bone marrow myeloid cells. Murine bone marrow myeloid
cells from viable moth-eaten and control C57BL mice was cultured in
IL-3 plus GM-CSF for 48 or 72 h. Cell lysates were
immunoprecipitated with antibody to HoxA10 (goat polyclonal) under
conditions that do not co-precipitate other proteins. Western blots
were performed and probed with anti-phosphotyrosine antibody or HoxA10
antibody (rabbit polyclonal), as indicated. Abundance of
tyrosine-phosphorylated HoxA10, but not total HoxA10 protein, is
increased in ex vivo cultured myeloid cells from viable
moth-eaten mice in comparison with control mice.
|
|
Therefore, ex vivo murine bone marrow cells, cultured in
IL-3 and GM-CSF, were also analyzed for HoxA10 protein abundance. Cell
lysates were immunoprecipitated with an anti-HoxA10 antibody, under
conditions that do not co-precipitate other proteins.
Immunoprecipitates were analyzed for HoxA10 protein abundance and
tyrosine phosphorylation by Western blot (Fig. 5B). We
observed that HoxA10 protein abundance was approximately equivalent,
with or without endogenous SHP1-PTP. Consistent with the results above,
phosphorylated HoxA10 was more abundant in ex vivo cultured
myeloid cells from mice lacking SHP1 than in cells cultured from wild
type mice.
SHP1-PTP Activity Influences HoxA10 Tyrosine Phosphorylation State
in Vitro--
The preceding experiments employed cell lysates, which
are a complex mix of proteins. Therefore, we investigated the ability of recombinant SHP1 to dephosphorylate in vitro translated
HoxA10. SHP1 and CS453-SHP1 were in vitro translated in
rabbit reticulocyte lysate (17). Protein-tyrosine phosphatase activity
assays of the recombinant proteins were performed using a colorimetric
assay (as described (17)). As expected, SHP1, but not CS453-SHP1, demonstrates PTP activity. In vitro translated HoxA10
([35S]methionine-labeled) was incubated with either SHP1
or CS453-SHP1 (unlabeled), under the same conditions used for
Yop PTP dephosphorylation (11). Tyrosine phosphorylation of
in vitro translated HoxA10 was determined by
anti-phosphotyrosine immunoprecipitation under conditions that do not
co-immunoprecipitate SHP1 or CS453-SHP1. We found that incubation with
SHP1, but not catalytically inactive SHP1, decreases in
vitro translated HoxA10 tyrosine phosphorylation (Fig.
6A).
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 6.
SHP1-PTP activity
decreases HoxA10 tyrosine phosphorylation in vitro.
A, SHP1-PTP dephosphorylates HoxA10 in
vitro. SHP1 and CS453-SHP1 were in vitro translated in
rabbit reticulocyte lysate (unlabeled), and HoxA10 was in
vitro translated in rabbit reticulocyte lysate
([35S]methionine-labeled). HoxA10 was incubated with
either SHP1 or CS453-SHP1, followed by immunoprecipitation with
anti-phosphotyrosine antibody under conditions that do not
co-precipitate other proteins. Proteins were separated by SDS-PAGE, and
HoxA10 was detected by autoradiography. HoxA10 incubated with
catalytically inactive SHP1, but not wild type SHP1, was
immunoprecipitated by antibody to phosphotyrosine. B,
SHP1-PTP dephosphorylates amino acid 320-393 HoxA10 in
vitro. SHP1 and CS453-SHP1 were in vitro translated in
rabbit reticulocyte lysate (unlabeled) and amino acid 320-393 HoxA10
was in vitro translated in rabbit reticulocyte lysate
([35S]methionine-labeled). Amino acid 320-393 HoxA10 was
incubated with either SHP1 or CS453-SHP1, followed by
immunoprecipitation with anti-phosphotyrosine antibody under conditions
that do not co-precipitate other proteins. Proteins were separated by
SDS-PAGE, and 320-393 HoxA10 was detected by autoradiography. 320-393
HoxA10 incubated with catalytically inactive SHP1, but not wild type
SHP1, was immunoprecipitated by antibody to phosphotyrosine.
C, SHP1-PTP activity increases DNA binding of in
vitro translated HoxA10. EMSA was performed with a DNA probe
representing the derived HoxA10/Pbx binding sequence and in
vitro translated HoxA10 incubated with the following: 1)
in vitro translated CS453-SHP1; 2) in
vitro translated SHP1. The arrow represents
the specific shifted HoxA10 protein complex. SHP1-PTP activity is
associated with increased HoxA10 DNA binding affinity.
|
|
To determine whether the tyrosine residues in the HoxA10 homeodomain
are substrates for SHP1, similar experiments were performed with the
in vitro translated, 320-393 amino acid HoxA10 truncation mutant. In vitro translated 320-393 amino acid HoxA10
([35S]methionine-labeled) was incubated with either
in vitro translated SHP1 or CS453-SHP1 (unlabeled), and
tyrosine phosphorylation of 320-393 amino acid HoxA10 was determined
by anti-phosphotyrosine immunoprecipitation, as above. Similar to the
results above, incubation with in vitro translated SHP1, but
not catalytically inactive SHP1, decreases tyrosine phosphorylation of
in vitro translated 320-393 amino acid HoxA10 (Fig.
6B). This experiment is possible, because SHP1 interacts
with its substrates via the domain that includes the target tyrosine residues.
Based on these results, we investigated whether SHP1-PTP treatment
increases the DNA binding affinity of in vitro translated HoxA10. These experiments were designed to link dephosphorylation by
SHP1-PTP with change in HoxA10 DNA binding activity. EMSA were performed with a probe representing the derived HoxA10/Pbx consensus sequence and in vitro translated HoxA10, preincubated with
either in vitro translated SHP1 or CS453-SHP1 (Fig.
6C). Similar experiments were performed with a DNA probe
representing the HoxA10-binding CYBB repressor element (not
shown). Consistent with our hypothesis, SHP1-PTP treatment, but not
control CS453-SHP1 treatment, increases HoxA10 binding affinity to the
DNA probes.
SHP1-PTP Interacts Directly with HoxA10--
The preceding
experiments suggest that SHP1-PTP activity influences the tyrosine
phosphorylation state of HoxA10. The ability of HoxA10 to serve as a
substrate for SHP1-PTP would be further substantiated by demonstration
of direct interaction between HoxA10 and SHP1-PTP proteins. Therefore,
we expressed SHP1 in E. coli as a fusion protein with
glutathione S-transferase (SHP1/GST) and purified the fusion
protein by affinity to glutathione-agarose. We found that in
vitro translated HoxA10 co-precipitates with SHP1/GST fusion
protein but not with control GST protein (Fig. 7A).
View larger version (55K):
[in this window]
[in a new window]
|
Fig. 7.
HoxA10 interacts directly with SHP1-PTP.
A, in vitro translated HoxA10 interacts directly
with recombinant SHP1. HoxA10 was in vitro translated in
rabbit reticulocyte lysate (labeled with
[35S]methionine), and SHP1 was expressed in E. coli as a fusion protein with glutathione S-transferase
(SHP1/GST). Interaction of HoxA10 with SHP1/GST or GST control was
determined by co-precipitation. SHP1/GST, but not control GST,
co-precipitated HoxA10. B, HoxA10 co-immunoprecipitates with
SHP1 from undifferentiated U937 cell lysates. Nuclear proteins were
isolated from undifferentiated U937 cells, and protein was
immunoprecipitated with an antibody to SHP1 under nondenaturing
conditions. Proteins were separated by SDS-PAGE and analyzed by Western
blot, probed for SHP1 or HoxA10, as indicated. HoxA10 co-precipitated
with SHP1 from U937 nuclear proteins.
|
|
The results of this assay suggest that SHP1-PTP interacts directly with
HoxA10, in vitro. To determine whether SHP1 and HoxA10 interact in myeloid cells, we investigated co-precipitation of SHP1-PTP
and HoxA10 in U937 cell lysate proteins. Undifferentiated U937 lysates
were precipitated with an anti-SHP1 antibody and the immunoprecipitates
analyzed by Western blot. We found that HoxA10 co-immunoprecipitates
with antibody to SHP1 but not with irrelevant control antibody (Fig.
7B). Therefore, these results are consistent with the
hypothesis that HoxA10 interacts with SHP1-PTP in undifferentiated
myeloid cells.
| |
DISCUSSION |
Previously, we found that tyrosine phosphorylation decreases
HoxA10 DNA binding, thereby decreasing HoxA10-mediated transcriptional repression of myeloid-specific genes (11). Since HoxA10 is
tyrosine-phosphorylated during myeloid differentiation (11), this
provides a mechanism for regulating HoxA10 activity in differentiating
cells. In the current investigations, we identify the phosphotyrosine
residues that decrease HoxA10 DNA-binding affinity and a mechanism by
which phosphorylation of these residues impacts DNA binding. In the current studies, we also determine that these regulatory tyrosine residues are a substrate for SHP1-PTP. Therefore, these investigations indicate that SHP1-PTP activity, in undifferentiated myeloid cells, contributes to HoxA10-mediated transcriptional repression.
We found that DNA binding affinity is decreased by phosphorylation of
two tyrosine residues in the HoxA10 homeodomain. Consistent with this,
a homeodomain tyrosine mutant form of HoxA10 represses transcription,
via the CYBB HoxA10-binding repressor element, in both
undifferentiated and IFN-differentiated U937 cells. In contrast,
wild type HoxA10 represses, via this repressor element, only in
undifferentiated U937 transfectants. These results suggest that the
HoxA10 homeodomain tyrosine residues are phosphorylated during
IFN-induced U937 differentiation, decreasing DNA binding and
transcriptional repression. This is of interest because homeodomain sequences are more than 80% identical in HoxA9, -10, and -11 (three "Abd" Hox proteins), including conservation of the two tyrosine residues. Therefore, our results suggest that phosphorylation of
homeodomain tyrosine residues the may be a general mechanism for
regulation of Abd HoxA protein activity.
In undifferentiated U937 cells, overexpressed, homeodomain tyrosine
mutant HoxA10 represses transcription, via the CYBB
repressor element, more efficiently than overexpressed, wild type
HoxA10. Consistent with this, overexpressed HoxA10 is somewhat
tyrosine-phosphorylated in undifferentiated U937 cells. This suggests
that basal phosphorylation of tyrosines 326 and 344 in
undifferentiated U937 cells decreases transcriptional repression
by overexpressed, wild type HoxA10. Alternatively, substitution of
phenylalanine for tyrosine may increases HoxA10 DNA binding affinity by
causing phosphorylation-independent alteration of the homeodomain.
Although homeodomain tyrosine mutant HoxA10 represses in both
IFN-treated and untreated U937 cells, repression is less efficient
in differentiated transfectants. One possibility is that additional
HoxA10 tyrosine residues influence transcriptional repression,
independent of DNA binding (e.g. residues involved in
protein-protein interactions). Alternatively, IFN may activate
positive acting transcription factors, which interact with the
CYBB sequence in the artificial promoter construct. If such
hypothetical factors do not compete with HoxA10 for DNA binding, activation could occur in IFN-treated cells, despite HoxA10 binding to the repressor element.
Since HoxA10 binds to DNA via the homeodomain, we initially
hypothesized that tyrosine phosphorylation directly interferes with
DNA-protein interaction. However, tyrosine phosphorylation did not
influence in vitro DNA binding affinity of a HoxA10
homeodomain-only truncation mutant. Further investigations indicated
that both the homeodomain and amino acids 267-320 (with no tyrosines)
are necessary for tyrosine phosphorylation to decrease HoxA10 DNA binding affinity. Consistent with this, phosphotyrosine residues in the
homeodomain interact with HoxA10 amino acids 267-320. These results
suggest two possible mechanisms by which HoxA10 homeodomain tyrosine
phosphorylation might decrease DNA-binding affinity. The first is that
homeodomain tyrosine phosphorylation induces HoxA10 folding, rendering
the homeodomain unavailable for DNA binding. The second possibility is
that tyrosine phosphorylation leads to formation of HoxA10 homodimers,
unable to participate in DNA binding. Such homodimers have not been
identified but would represent an interesting mechanism of regulation.
Investigations are planned to identify specific HoxA10 residues that
interact with homeodomain phosphotyrosines. Such results might clarify functional differences among Abd HoxA proteins, which are not well
conserved outside of the homeodomain.
To understand the events regulating HoxA10-mediated repression, we
investigated regulation of HoxA10 tyrosine phosphorylation. Previously,
we found that SHP1-PTP activity decreases transcription of the
CYBB and NCF2 genes in undifferentiated myeloid
cell lines (17). Additionally, we found that SHP1-PTP activity
decreases tyrosine phosphorylation of IRF1 and ICSBP, providing one
mechanism for SHP1 inhibition of CYBB and NCF2
transcription (17). In these studies, we determined that SHP1-PTP also
regulates the HoxA10 tyrosine phosphorylation state. We found that
inhibition of SHP1-PTP activity decreases wild type HoxA10-mediated
repression, via the repressor element from the CYBB gene. In
contrast, inhibition of SHP1-PTP activity did not influence the
transcriptional repression by homeodomain tyrosine mutant HoxA10 in
this assay, suggesting that SHP1-PTP influences tyrosine
phosphorylation of the HoxA10 homeodomain. Consistent with these
results, tyrosine phosphorylation of HoxA10 is increased in dominant
negative SHP1-expressing U937 cells.
However, these experiments were performed in a myeloid leukemia cell
line cell line. It is possible that the results represent a
manifestation of the transformed phenotype. To investigate the influence of SHP1 on HoxA10 tyrosine phosphorylation in a
nontransformed model of myelopoiesis, we employed cultured myeloid
cells from viable moth-eaten (SHP1/) mice. We found
that HoxA10 tyrosine phosphorylation (but not abundance) is increased
in cultured myeloid cells from SHP1/ mice, in
comparison with control mice. In these experiments, we used unsorted
populations of cultured myeloid cells, derived from total murine bone
marrow by Ficoll-Hypaque (after red cell lysis), followed by culture in
IL-3 and GM-CSF. Based on experience with normal murine bone marrow
cells (29), these cultures would be anticipated to contain committed
myeloid progenitors. However, based on experience with ex
vivo cultured murine spleen cells (23, 30), it is possible that
cultured SHP/ bone marrow myeloid cells are more
differentiated than control cells under the same cytokine conditions.
This possibility does not adversely affect interpretation of our
results, since we hypothesize that HoxA10-induced repression of
myeloid-specific genes antagonizes differentiation. This hypothesis
implies that events that reverse HoxA10 transcriptional repression also
induce differentiation. Therefore, accelerated ex vivo
differentiation of SHP1/ monocytes (as in Refs. 29 and
30) may be related to the absence of SHP1 dephosphorylation of
transcription factors that regulate progression of myeloid
differentiation, such as HoxA10, IRF1, and ICSBP.
Therefore, these investigations indicate that HoxA10 DNA binding is
regulated by interaction of phosphotyrosine residues in the homeodomain
with an adjacent domain in HoxA10, that HoxA10 is a substrate for SHP1,
and that SHP1-PTP activity regulates HoxA10 repression activity in
differentiating myeloid cells. Since homeodomain tyrosine residues are
conserved over the Abd group of HoxA proteins (HoxA9 to -13), our
investigations suggest a mechanism for regulating the functional
activities of this important group of transcription factors.
§¶
,
TOP
ABSTRACT
INTRODUCTION
To whom correspondence should be addressed: Dept. of Medicine,
Northwestern University Medical School, Olson Pavilion, Rm. 8527, 710 N. Fairbanks Ct., Chicago, IL. E-mail:
e-eklund@northwestern.edu.
