 |
INTRODUCTION |
Mer, a then novel receptor tyrosine kinase, was
identified by bacterial phosphotyrosine (Tyr(P)) expression cloning
(1). Ling and Kung (2) independently cloned this receptor and referred to it as Nyk. By sequence, Mer is the closest mammalian gene
to the chicken proto-oncogene, c-EYK, identified by Jia
et al. (3, 4) as the cellular homologue of a chicken
retroviral oncogene v-EYK. The extracellular domain
structural motifs of Mer place it within the Axl/Ark/UFO family of
receptor tyrosine kinases (5-7). In addition to Axl and Mer, this
family contains, at least, one other member Tyro3 (8) (also
named SKY (9), RSE (10), BRT (11), DTK (12), and TIF (13)) and its
potential chicken homologue, REK (14). Mer mRNA
is most highly expressed in testis, ovary, prostate, lung, and kidney
and is detected in peripheral blood monocytes but not in granulocytes
(1). It is expressed early during mouse embryonic development, being
present in the blastocyst (15). Interestingly, despite the fact that
Mer mRNA is readily detected in neoplastic B- and T-cell
lines (1) and is present in childhood acute lymphoid leukemia
samples,1 it is not expressed
in normal B- or T-cells even when they are forced to proliferate (15).
To date, evidence gathered by multiple groups hints at physiologic
roles for the individual family members, but a coherent view of Mer (or
Axl or Tyro3) signaling as it relates to function has yet to emerge.
A large scale biochemical purification approach, using conditioned
media from cell lines, identified Gas6, a transcript whose expression
increases in senescent cells as a ligand for Axl (17). Gas6 also binds
to Tyro3 as does protein S (18). Gas6 binds to Axl and
Tyro3, with nanomolar affinity (0.4 and 2.9 nM,
respectively (19-21)), as well as to Mer (21) although the affinity is
low enough, 29 nM, to raise questions as to whether it is a
physiologic or sole ligand for Mer.
Axl/Mer/Tyro3 are ectopically or overexpressed in a variety of
human tumor cells (15, 22-24), and Axl and Tyro3 when
substantially overexpressed in fibroblasts can transform these cells in
the absence of ligand (6, 14, 24). Axl signaling was studied in 32D
cells, an
IL-32-dependent
murine leukemic cell line (25), by constructing an EGFR extracellular
and transmembrane domain-Axl cytoplasmic domain chimera
(26). Upon IL-3 withdrawal, ligand-dependent activation of
the Axl chimera prevented apoptosis and caused proliferation (26). In fibroblasts grown in serum, full-length Axl when activated by
Gas6 stimulated growth. Gas6 Axl signaling prevented apoptosis without stimulating growth when cells were challenged by serum starvation, Myc overexpression, or tumor necrosis factor-
(27, 28).
Axl, like platelet-derived growth factor, can increase smooth muscle
cell motility (29).
Turning to genetic approaches for clues to the function of this family,
our group and two others produced gene-targeted mice, deleting
Mer (30), Axl (31), and Tyro3
(31), respectively. The subtle nature of the individual gene
knockout phenotypes, monocyte hyper-reactivity to stimuli, and splenic
enlargement in part due to apoptotic debris (Mer (30)) and
being prone to seizures (Tyro3 (31)) led to a
collaboration in which animals lacking all three members of the family
were produced. These "triple knockout" animals were infertile, a
phenotype due to a Sertoli cell defect (30).
We have shown recently (32) that the apoptotic debris in
Mer
/ animals and immune dysregulation characterized by
high levels of autoantibodies may be the result of a demonstrable
defect in triggering the ingestion of apoptotic cellular material. This defect, particularly the immune dysregulation, is even greater in the
triple knockout animals (33). A mer-specific, selective phagocytic defect toward apoptotic material is also exhibited in the
pigmented retinal epithelial cells of the eye (34, 35), leading to
retinal degeneration. Investigation of a rat model of adult blindness,
in which pigmented retinal epithelial cells fail to ingest the
apoptotic tips of rods and cones on an ongoing basis, first identified
an inherited mutation in the rat Mer tyrosine kinase
gene that truncates the gene product (34). Not surprisingly, the
Mer/ mice show retinal degeneration as they
age.1 Finally, at least three human families with inherited
retinitis pigmentosa have different mutations in the Mer
TK gene, all of which abolish Mer tyrosine kinase
activity (35).
To study the signal emanating from the Mer tyrosine kinase, we
constructed a chimeric receptor molecule, hoping to achieve stable
expression of a ligand-dependent Mer tyrosine kinase. This was accomplished by replacing the extracellular domain of Mer with a
ligand binding and transmembrane domain of the rat EGF receptor (EMC).
Several groups (36-39) had previously transfected chimera receptor
tyrosine kinases into the IL-3-dependent murine leukemic
cell line 32D c13 (32D) and had shown that these receptors could
replace the ability of IL-3 to suppress apoptosis. All tyrosine kinase
transmembrane receptors, except one, resulted in both survival and
rapid proliferation of 32D cells even after IL-3 withdrawal. The
exception, the IGF1 receptor, resulted in apoptosis
suppression and slow growth in 32D cells (40). The IGF1
receptor signal resulted in a differentiated granulocyte-type phenotype
over 6-8 days, mimicking the effect of a cytokine, G-CSF (41, 42).
Separate populations of 32D cells, which express neither Mer nor EGF
receptor, were transfected with either EMC or full-length EGFR, and
stable cell populations were selected. Interestingly and in contrast to
multiple tyrosine kinases tested in the 32D cell including the
EGFR/Axl chimera, ligand-dependent EMC
activation prevented apoptosis upon IL-3 withdrawal without stimulating
proliferation and caused cell shape changes. More surprisingly, in some
32D cell clones the Mer signal produced a dramatic morphologic change that was coincident with a blockade of IL-3-dependent
growth. Thus Mer, at least as a chimeric receptor, is capable of
producing an unusual signal in comparison to other receptor tyrosine
kinases; blocking apoptosis in the absence of increased cell
proliferation although altering the cytoskeleton. This latter
capability may explain the failure of monocytes (32) and pigmented
retinal epithelial cells (34, 35) to trigger the local cytoskeletal changes necessary to ingest apoptotic material in the absence of Mer
tyrosine kinase activity.
| |
MATERIALS AND METHODS |
Reagents--
Murine EGF (receptor grade) was purchased from BD
PharMingen. Interleukin-3 (IL-3) was obtained by culturing WEHI cells
in Dulbecco's modified Eagle's medium-high glucose (Invitrogen)
supplemented with 10% fetal bovine serum (FBS). The monoclonal
antibodies specific for phosphotyrosine (PT66 and RC20) were obtained
from Sigma and Transduction Laboratories. The anti-rat EGFR antibody
used for immunoprecipitation (1382) was made as described using
purified rat liver EGFR as an immunogen (43). EGFR immunoblotting was performed using an antibody (number 22) raised against a fusion protein glutathione S-transferase/EGFR C terminus (44).
Anti-Mer antibodies were obtained by immunizing New Zealand
White rabbits with a bacterially expressed glutathione
S-transferase fusion protein containing the C-terminal-most
100 amino acids of human Mer. Anti-AKT,
anti-phospho-AKT, anti-p38, anti-phospho-p38, anti-p44/42, and anti-phospho-p44/42 were obtained from Cell Signaling. Anti-Rb antibody was obtained from BD PharMingen. PI 3-kinase inhibitor LY294
was purchased from Sigma, the MEK inhibitor, U0126, was purchased from Promega.
Engineering of the EGFR/Mer Chimera--
The
EGFR/Mer chimera (EMC) was constructed by combining
the extracellular and transmembrane regions of the EGFR with the
intracellular domain of Mer. This was accomplished by PCR-directed
insertion of a SalI restriction site into a basic stretch of
amino acids (RRR or RKR) that are conserved just inside of the
transmembrane domain in many tyrosine kinase receptors, including the
EGFR and Mer. The EGFR extracellular domain was amplified with
oligonucleotide EMC-2R (5'-gagagagagtcgacgcatg
aagaggccgagccc-3') in conjunction with an extended T7 vector
primer (attgtaatacgactcactata), generating the EGFR extracellular
domain with a silent mutation in the juxtamembrane Arg-Arg-Arg
sequence. The PCR product was then cut with EcoRI and
SalI and cloned into pBSII SK+ (Stratagene). The Mer
intracellular domain was amplified with the oligonucleotide EMC-1F
(5'-gagagagagagacgtcgacgagtccaggagacaaagtttggg-3') in conjunction with
the extended T7 primer, and the PCR product was TA-cloned (Invitrogen)
as per manufacturer's instructions. An NsiI to
HindIII fragment from the original full-length cDNA was
used to replace all of the PCR-generated sequence except for the 28 bp
immediately downstream of the SalI site. The mammalian expression plasmid was generated by introduction of the full-length chimera EcoRI fragment into the EcoRI site of the
expression vector pLXSN (45).
Construction of Stable 32D Cell EMC and Kinase-inactive EMC Cell
Lines--
32D parental cells (25) were grown in RPMI 1640 containing
15% heat-inactivated FBS, 5% WEHI 3B conditioned media (to provide IL-3), and 1× penicillin/streptomycin. Cells were washed twice in
serum-free RPMI 1640, and then DNA was added, either vector, full-length EGFR, or EMC. The cells were electroporated using the
Bio-Rad Gene Pulser Apparatus and immediately transferred to complete
media. After 48 h 650 ng/ml geneticin (G418, Invitrogen) was added
to the media. The growth and apoptosis analysis experiments shown in
this report were replicated using two separate transfection populations
and four distinct clones, two from each of the independent populations.
Kinase-inactive EMC was made by site-directed mutation of lysine 619 to
methionine. The mutated EMC was packaged in a murine retrovirus; 32D
cells were infected, and a stable population was selected in G418.
Reverse Transcriptase-PCR Analysis of EMC
Expression--
Approximately 5 × 106 cells from
each cell line were resuspended in 0.5 ml of lysis solution (4 M guanidinium isothiocyanate, 25 mM sodium
citrate, pH 7.0, 0.5% sarcosyl, and 0.1 M
2-mercaptoethanol), and total RNA was isolated by the acid phenol
procedure (46). The first strand cDNA was prepared using Moloney
murine leukemia virus reverse transcriptase. Five microliters of the
first strand cDNA was amplified in a 50-µl PCR using
Taq DNA polymerase (Invitrogen). To test for the
integrity of the RNA samples and for template standardization,
amplification by PCR using actin primers (HACA-1F and HACA-1R) was
performed. The expression of Mer, EGFR, and EMC in the
samples was analyzed using primers specific for each cDNA. Amplifications were performed for 35 cycles with an annealing temperature of 60 °C. The sequence of the primers is as follows: actin-HACA-1F, 5'-CCTTCCTGGGCATGGAGTCCT-3'; HACA-1R,
5'-GGAGCAATGATCTTGATCTTC-3'; Mer-3F,
5'-CACCTCTGCCTTACCACATCT-3'; and 2R, 5'-ATCCACAAAAGCAGCAAAGA-3' (specific to the mer extracellular region absent from the
chimeric cDNA); EGFR-1F, 5'-AAAGACTGCAAGGCCGTGAA-3'; 2R,
5'-GCCCAGATGGCCACACTTC-3'; EMC-1F, 5'-TAACGTGTGCCACCTCTGC-3'; and 2R,
5'-TCCTCACTGACTCCCAAGC-3'. The PCR products were analyzed by
electrophoresis of 10 µl of each reaction on a 1.0 or 1.5% agarose
gel containing ethidium bromide.
Cell Proliferation Assay--
32D vector alone (control), EGFR,
and EMC were seeded at 20 × 104 cells per ml in RPMI
1640 supplemented with 15% heat-inactivated FBS with or without 5%
WEHI conditioned medium (source of IL-3) or 100 ng/ml EGF. At the
indicated times flasks were scraped, and cells were counted using a hemocytometer.
Western Blot Analysis of Stable Cell Lines--
The
EMC-expressing cell populations were resuspended at 5 × 105 cells/ml in serum/WEHI-free media and incubated at
37 °C for 1 h prior to each experiment. The cells were then
treated with 100 ng/ml EGF, pelleted, and lysed with ice-cold new lysis
buffer (NLB, 20 mM HEPES, pH 7.3, 50 mM NaF,
10% glycerol, 1% Triton X-100, 1 mM
Na3VO4, 500 mM NaCl, 5 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride,
1 µl of leupeptin per ml, 4 KIU of aprotinin per ml). Lysates were
clarified, divided, and either boiled in SDS Laemmli sample for lysate
Westerns or used for immunoprecipitations. Polyacrylamide gels were
loaded and immunoblotted.
FACS Analysis of Receptor Expression in 32D Cells--
Cells
were washed twice with 1× PBS, resuspended in anti-EGFR 1382 polyclonal antibody, and incubated on ice for 30 min. Cells were then
washed three times in 1× PBS, resuspended in anti-rabbit IgG
conjugated with FITC antibody, and incubated 30 min on ice, followed by
3 washes in 1× PBS and then analyzed by flow cytometry with FACSTAR
plus (BD PharMingen).
FACS Analysis of Apoptosis and Cell Cycle--
An annexin
V-FITC-labeled apoptosis detection kit was used as instructed by the
manufacturer (Genzyme). The procedure evaluates an early event in
apoptosis, in which phosphatidylserine (PS) is translocated from the
inner to the outer leaflet of the plasma membrane, thereby exposing PS
to the external environment. Annexin V has a high affinity for PS and
therefore binds to cells with exposed PS. Briefly, 32D vector alone or
32D EMC cells were cultured in RPMI 1640 plus 15% heat-inactivated FBS
supplemented with no addition, 5% WEHI-conditioned medium (source for
IL-3), or 50 ng/ml murine natural EGF (receptor grade) for 24, 48, or
72 h. Both adherent and non-adherent cells were harvested and
washed twice in 1× phosphate-buffered saline (PBS). To analyze
apoptosis, cells were then resuspended in annexin binding buffer at a
concentration of 5 × 105 cells/ml and incubated in
the presence of optimized amounts of FITC-conjugated annexin V and
propidium iodine (PI) for 15 min. The percentage of apoptotic cells was
evaluated by flow cytometry with a FACStar plus (BD PharMingen). To
analyze the cell cycle, cells at the indicated times were fixed in
ice-cold 70% ethanol overnight. The next day the cells were washed
twice in calcium- and magnesium-free PBS, followed by a 1-h incubation
with 50 µg/ml PI and 50 µg of DNase-free RNase, and analyzed by
flow cytometry.
Cell Staining--
32D cells were grown on coverslips in regular
media with 100 ng/ml EGF, 10 µM LY294, or 10 µM U0126. At the appropriate time, cells were washed with
1× PBS and then fixed with methanol. Cells were then stained with
Wright Giemsa Stain (Sigma), washed with 1× PBS and then briefly with
water. Cells were viewed through a ×20 objective, and images were
captured through a CCD camera and analyzed by Scion/Photoshop imaging software.
| |
RESULTS |
Activity and Stable Expression of EMC--
To investigate the
consequences of ligand-dependent Mer activation, we
constructed a chimeric receptor placing the Mer tyrosine kinase domain
under the control of the EGF receptor ligand-binding domain (Fig.
1A). The full-length receptor
chimera is predicted to be a protein 1143 amino acids in length, with a
molecular mass of 127 kDa. However, since the extracellular
domain of the EGFR is heavily glycosylated and contributes ~110 kDa
to the mature chimera, the estimated molecular mass of the mature EMC
is ~170 kDa. Transient transfection of the full-length EGFR and EMC
in CHO-K1 cells revealed tyrosine-phosphorylated receptors of the appropriate size on phosphotyrosine immunoblots. In addition, immunoprecipitation with anti-extracellular domain EGFR antibody 1382 precipitated active tyrosine kinases from EGF receptor or EMC-transfected CHO-K1 cells as assessed with in vitro
autokinase activity assays with added [32P]ATP (data not
shown).
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 1.
Engineering and expression of the
EGFR/Mer chimera. A,
schematic diagrams of the Mer, EGFR, and the chimeric
receptor (EMC). The hatched rectangles represent the
immunoglobulin light chain-like domains (IgL), fibronectin
III-like domains (FnIII), or cysteine-rich domains
(black rectangles) characteristic of the extracellular
domains of Mer or EGFR. B, analysis of stable
expression of the EMC receptor or full-length rat EGFR in 32D cells.
Reverse transcriptase-PCR was utilized to confirm expression of
transfected receptors and the absence of full-length Mer or
EGFR in 32D cells. RNA was isolated from parental 32D cells (1st
lane), and cells transfected with the EGFR (2nd lane),
EMC (3rd and 4th lanes, two separate populations
of transfected 32D cells). RNA from mouse spleen was also isolated and
subjected to PCR (5th lane). C, cycling 32D
parental, vector, EGFR, and EMC cells were incubated with rabbit
anti-rat extracellular EGFR antisera 1382, followed by incubation with
FITC-labeled and anti-rabbit IgG and subjected to analysis by flow
cytometry.
|
|
Following transfection of the vector alone (pLXSN), full-length EGFR,
or EMC and selection in G418, expression of EGFR and EMC
mRNA was assessed by reverse transcriptase-PCR. Mer
mRNA expression from murine spleen is shown for comparison. Murine
Mer was not expressed in any 32D cell populations, and the EGF receptor
was not detected except in the EGF receptor-transfected line. EMC was
expressed only in the lines transfected with EMC cDNA (Fig. 1B). Oligonucleotides utilized to detect EMC expression were
specifically designed to span the junction between chimeric partners.
Likewise, EGFR and Mer expression were verified by utilizing
oligonucleotides designed to regions of mRNA not expressed in the
chimera (see "Materials and Methods"). To detect surface receptor
protein expression, flow cytometry was performed using an antibody that
recognized the extracellular domain of the EGFR, the extracellular
portion of both EGFR and EMC. Control 32D cells did not express either receptor, whereas the amount of EGFR and EMC on the respective transfected clones was similar (Fig. 1C).
Ligand-dependent Signaling by the EGFR, Mer Chimeric
Receptor--
Neither EGFR nor EMC were expressed at levels that
produced constitutive, ligand-independent receptor tyrosine
phosphorylation. Fig. 2A shows
the absence of EMC tyrosine phosphorylation in untreated cells and the
rapid accumulation of tyrosine phosphate in EMC immunoprecipitated from
EGF-treated cells. Other experiments showed EMC tyrosine
phosphorylation within 60 s of EGF addition. The results indicate
that EMC-expressing 32D cells are an excellent model in which to assess
ligand-dependent Mer signaling. To study whether the EGF
receptor or the Mer chimera activated downstream events differentially, we examined phosphotyrosine (Tyr(P)) containing substrates after a 5-min EGF stimulation of full-length EGFR or EMC-expressing 32D cell lines. EGF stimulated tyrosine phosphorylation of multiple substrates in both cell lines; the pattern of Tyr(P) substrates clearly differed between the two receptors. As expected, EGF
did not stimulate tyrosine phosphorylation in parental or vector-transfected 32D cells. The similarity of
ligand-dependent receptor autophosphorylation in the EGF
and EMC-expressing 32D cell lines substantiates the comparable
expression of receptor as indicated by flow cytometry (Fig. 2). The
pattern of tyrosine-phosphorylated substrates in EMC-bearing cells
was both more substantial and complex than that of the EGFR-bearing
cells. Thus, despite similar receptor levels on the two cell lines, the
differences in tyrosine phosphorylation, in some way, engendered the
distinct biologic actions elicited by Mer and EGFR signaling (see
below).
View larger version (52K):
[in this window]
[in a new window]
|
Fig. 2.
Ligand-induced tyrosine phosphorylation in
32D cells. Comparison of phosphotyrosine substrates from EMC,
EGFR, or vector expression cells. A, 32D EMC cells were
stimulated with 100 ng/ml EGF and lysed 0, 5, 30, or 60 min later. EMC
was immunoprecipitated (IP) with anti-EGFR antibody 1382 followed by gel electrophoresis and immunoblotting (IB) with
anti-phosphotyrosine monoclonal antibody (RC20). B,
vector, EGFR, EMC 32D cells were left unstimulated (no addition,
NA) or stimulated with 100 ng/ml or IL-3 (5% WEHI
conditioned medium) for 10 min. Cell lysates were immunoprecipitated
with anti-phosphotyrosine monoclonal antibody (PT66), and precipitates
were subjected to gel electrophoresis and immunoblotting with an
anti-phosphotyrosine monoclonal antibody (RC20).
|
|
To determine the downstream signaling pathways stimulated by EMC, we
assessed the activation of members of the mitogen-activated protein
kinase family, ERKs 1 and 2 (p44/p42), JNK, and p38, as well as an
indication of PI 3-kinase activation, AKT. Both p44/42 gel shift (data
not shown) and phospho-ERK 1 and 2 immunoblotting showed that
ligand-dependent EMC activation rapidly stimulated ERKs 1 and 2 (Fig. 3A). To see this
activation by EMC alone, it was necessary to withdraw IL-3 for at least
1 h because IL-3 by itself stimulates ERKs 1 and 2. The
combination of EMC and IL-3 receptor signaling was, at best, minimally
more effective than either ligand alone, with respect to the ERK
activation. The EGF-dependent activation of ERKs 1 and 2 in
EGFR-expressing cells was similar to that of EMC (data not shown).
Thus, EMC can signal to the ERK pathway, but this in no way
distinguishes it from EGFR or IL-3. With respect to JNK, IL-3 did not
produce detectable JNK activity when added to withdrawn cells. Both EMC
and EGFR signaling produced a small, very transient activation that was
observed at 10 min but was nearly gone at 30 min (data not shown). AKT
activation was also assessed by immunoblotting and showed that EMC
signaling clearly increased phospho-AKT. IL-3 also increased
phospho-AKT, but again the addition of both signals was less than
additive (Fig. 3B). Assessment of p38 using activated p38
antibodies for immunoblotting showed some persistent p38 activation in
IL-3 withdrawn cells (Fig. 3C). EMC activation led to a
modest increase in phospho-p38 that was no greater than that of
re-addition of IL-3. As with ERK and AKT, the combination of EMC and
IL-3 was not substantially increased over either alone. In summary, the
study of these four signaling pathways did not reveal substantial
differences among EMC, EGFR, and IL-3 signaling that would explain the
differences in biologic and functional output noted below.
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 3.
EMC activates ERK, AKT, and
p38 mitogen-activated protein kinase. IL-3 was
withdrawn from cells for 1 h followed by stimulation with 100 ng/ml EGF, IL-3 (5% WEHI conditioned medium), or both EGF and IL-3.
Cells were lysed at indicated times, and lysates were subjected to gel
electrophoresis and immunoblotted with anti-phospho ERK and anti-ERK
(A); anti-phospho-AKT and anti-AKT
(B); and anti-phospho-p38 and anti-p38 antibody
(C).
|
|
Ligand-dependent Activation of EMC Results in an
Anti-apoptotic Signal without Stimulating Proliferation--
IL-3
withdrawal results in rapid death (due to apoptosis) of parental 32D
cells as well as transfected cells expressing either vector
alone, full-length EGFR, or EMC (circles, Fig.
4) with virtually all cells dying within
24-48 h. Addition of EGF upon IL-3 withdrawal had no effect on
vector-transfected 32D cell lines. EGF not only prevented cell death in
EGFR-expressing cells but mimicked the proliferative effect of IL-3. In
EMC-expressing cells, EGF activation of the Mer signal resulted in a
nearly stable cell number over 96 h (Fig. 4 and Table
I). In contrast a 32D cell stably
expressing a kinase-inactive tyrosine kinase EMC (K619M) failed to
prevent apoptosis when treated with EGF after IL-3 withdrawal (Table
I).
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4.
EGFR signaling stimulates 32D cell
proliferation; EMC signaling does not. 32D vector, EGFR, or EMC
cells were cultured in the absence (NA) or presence of IL-3
(5% WEHI conditioned medium) or 100 ng/ml EGF for 4 days as described
under "Materials and Methods." Viable cells were counted every
12 h through day 4. Withdrawal of IL-3 without other additions
leads to apoptotic cell death. Stimulation of EGFR in EGFR-transfected
cells and IL-3 in all cell lines increased cell numbers substantially.
Activation of EMC resulted in a stable number of surviving cells. These
data are representative of six separate experiments.
|
|
View this table:
[in this window]
[in a new window]
|
Table I
Kinase-inactive Mer does not protect 32D cells from IL-3 withdrawal
32D cells stably transfected with pLXSN (vector) EMC or a
kinase-inactive EMC created by site-directed mutagenesis of lysine 619 to methionine 619 were treated with no additions (NA), EGF 100 ng/ml,
and IL-3 (5% WEHI conditioned media). IL-3 caused proliferation in all
three cell lines, but EGF only prevented cell death in the line
expressing kinase active EMC.
|
|
The stable cell number in the 32D EMC cell cultures treated with EGF
could result from, at least, two mechanisms as follows: Mer activation
could prevent apoptosis without stimulating proliferation or a large
percentage of the 32D EMC cells might undergo apoptosis, whereas the
surviving cells rapidly proliferate. To address this, vector control
and EMC 32D cell lines were cultured in medium devoid of any
supplements (no addition) or in medium containing IL-3 or EGF. At 24-h
intervals, cells were analyzed by flow cytometry for apoptosis and
necrosis after staining with annexin-V-FITC and propidium iodide (Fig.
5). By 72 h virtually all vector
alone or EMC-expressing cells were stained with both annexin V and
propidium iodide in the absence of any treatment (Fig. 5A).
Conversely, in the presence of IL-3 both cell lines were >90% viable.
EGF could not rescue vector alone cells (almost 100% apoptotic and necrotic) whereas >80% of the EGF-treated EMC cells were viable (Fig.
5A). Greater than 90% of the cells without IL-3 or EGF were dead by 48 and 72 h (Fig. 5B), and apoptosis was
apparent within 24 h. These data (Figs. 4 and 5) indicate
that the Mer signal partially replaces IL-3 action preventing apoptosis
but does so without stimulating proliferation.
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 5.
EMC signaling prevents apoptosis and flow
cytometric analysis. A, healthy cycling 32D vector
(top) and EMC (bottom)-expressing cells were
cultured in medium without additions (NA) or in the presence
of IL-3 (5% WEHI conditioned medium) or 100 ng/ml EGF for 72 h.
Cells were harvested and incubated with annexin V-FITC in a buffer
containing PI and analyzed by flow cytometry. Apoptotic cells were
found in quadrants labeled with both annexin V and propidium iodide.
B, calculated percent of apoptotic cells after 24, 48, and 72 h incubation with no additions, IL-3, or EGF. Withdrawal of
IL-3 led to apoptosis (annexin V-positive) and death (propidium
iodide-positive). When cultured in IL-3 both cell lines continued to
cycle and were largely annexin V-FITC- and propidium iodide-negative.
32D vector cells cultured in the presence of EGF were positive for
annexin V and propidium iodide indicating cell death, and EMC cells
cultured in EGF were ~80% viable.
|
|
Mer Signaling Blocks IL-3-dependent Growth and Produces
Morphologic Changes in 32D Cells--
In the presence of IL-3, 32D
cells grow rapidly in suspension, and in fact, the medium must be
changed frequently to prevent apoptosis from overgrowth. Addition of
EGF to EMC-expressing cells not only keeps them alive without
proliferation but changes these suspension growing cells producing an
adherent, slightly flattened phenotype (not shown). This is clearly
distinguishable from the effects of persistent IL-3 or the addition of
EGF in EGFR-expressing cells, both of which promote continued growth of
round cell populations. A surprising result was obtained when EGF and
IL-3 were added together to EMC-expressing cells. The adherent
phenotype became more prominent, and IL-3-dependent growth
was substantially reduced so that by 24-48 h cell numbers no longer
increased. This was not seen when EGF and IL-3 were added to
EGFR-expressing cells. Fig. 6
demonstrates the lack of proliferation in EGF- and IL-3-treated EMC 32D
cells over a 7-day period. Vector-transfected cells, which do not
express EGFR or EMC, grow in response to IL-3 or IL-3 plus EGF and die
when IL-3 is withdrawn even if there is EGF in the media.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 6.
EMC stimulation prevented apoptosis and also
inhibited IL-3-dependent proliferation. 32D vector or
32D EMC-expressing cells were cultured with no additions
(NA), 100 ng/ml EGF, IL-3 (5% WEHI conditioned medium), or
both EGF and IL-3 for 7 days. Cells were counted every 24 h. IL-3
caused proliferation in both cell lines. EGF prevented apoptosis in EMC
cells without substantial proliferation. Surprisingly, EMC signaling
abrogated IL-3-dependent proliferation. These data are
representative of three separate experiments. In the vector alone
cells, all transfected cells with no additions or with EGF treatment
were dead by day 3. In vector alone cells IL-3 and EGF plus IL-3 gave
similar cell numbers at day 7. In the EMC-expressing cells, no addition
(withdrawal of IL-3) also led to 100% cell death by day 3. With EGF or
EGF plus IL-3 cell number increased from 2.0 × 105 to
2.3 × 105 and 2.2 × 105,
respectively, at day 7. Addition of IL-3 alone increased cell number
from 2.0 × 105 to 3.35 × 107 by day
7.
|
|
To study the state of the cell cycle, we investigated the
phosphorylation of the retinoblastoma protein (Rb) and performed cell
cycle analysis. Immunoblots of Rb show that addition of EGF and IL-3
led to Rb dephosphorylation as evidenced by a loss of slower mobility
forms of Rb progressively from 8 to 48 h (Fig. 7A). Flow cytometric analysis
showed that greater than 60% of 32D cells are in S phase during IL-3
treatment. Addition of EGF alone, which does not significantly
stimulate growth (Fig. 4 and Table I), decreased the percentage of
cells in S phase from 60 to 40% over time with the majority of cells
now in the G1 phase. The addition of EGF plus IL-3, which
more rapidly blocks proliferation and changes cell shape (see below),
produces a rapid decrease in the percentage of cells in S phase
(to 7%) and a striking increase in the percentage in G1
(Fig. 7B).
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 7.
The effect of EMC and
IL-3-dependent signaling on cell cycle progression.
A, IL-3 was with drawn from 32D EMC cells (clone A2)
for 1 h. Cells were then incubated with 100 ng/ml EGF and IL-3
(5% WEHI conditioned medium) cells for 0, 1, 8, 24, or 48 h. At
the indicated times cells were lysed; lysates were subjected to gel
electrophoresis, followed by Rb immunoblotting. By 8 h the near
uniform slow electrophoretic mobility of phosphorylated Rb protein
began to change with the appearance of the lower molecular weight
hypophosphorylated species. By 24 h both isoforms were detected,
and by 48 h most Rb was in the hypophosphorylated state.
B, cell cycle analysis by flow cytometry shows the
effect on cell cycle stage of additions (IL-3, EGF, or IL-3 plus EGF)
to 32D EMC cells at 24 and 48 h. Treatment with IL-3 maintained
>60% cells in S phase. EGF activation of EMC which did not sustain
proliferation over 5-7 days but which prevented apoptosis (see Fig. 4
and Table I) led to a slow decrease in S phase fraction. Combined
addition of IL-3 plus EGF led to a decrease in S phase fraction at
24 h and a dramatic reduction by 48 h.
|
|
Photomicrographs of stained EMC 32D cells over the course of a 24-h
treatment with IL-3 and EGF show the rapid morphologic change (Fig.
8). The morphology of typical IL-3 (or
EGF-treated EGF receptor-bearing 32D cells) is shown in the 0-h frame.
Within 8 h the initial flattening, caused by EGF and IL-3 addition
to the EMC-containing cells, was visible, and within 12 h the
dendritic-like processes that are characteristic of EMC-driven
morphologic change in this clone became apparent. By 24 h the
cells were adherent and dramatically different in shape. This
EMC-stimulated shape change was coincident with (or proceeded) the
suppression of IL-3-dependent growth, and to date we have
not determined whether they are separable events, i.e. does
the morphologic change represent differentiation that inhibits growth?
Unlike EMC-expressing cells, the addition of EGF and IL-3 to
EGFR-expressing cells does not alter cell morphology (not shown).
View larger version (80K):
[in this window]
[in a new window]
|
Fig. 8.
Morphologic changes in 32D EMC following
stimulation with EGF IL-3. Cells (clone A2) were treated with EGF
100 ng/ml plus IL-3 (5% WEHI conditioned medium) for the indicated
times and were then fixed and stained with Wright Giemsa stain. The
round cell morphology characteristic of proliferating 32D cells
(0 h) was altered with adherence to the tissue culture dish
beginning at ~8 h and by spreading and elongation of processes
(12-16 h). By 24 h the majority of cells had undergone flattening
and process elongation.
|
|
To attempt to relate the differentiated morphology, growth suppression,
or anti-apoptotic signals to known signaling actions of EMC (and IL-3),
we incubated EMC-expressing 32D cells in separate experiments with the
PI 3-kinase inhibitor, LY294, and the MEK inhibitor, U0126. LY294
completely abolished the generation of active phospho-AKT for up to
24 h (Fig. 9A). Fig.
9B shows that U0126 completely blocked the stimulation of
ERKs 1 and 2 caused by the addition of EGF and IL-3 in cells that had
been withdrawn from IL-3 1 h previously. The effect of U0126
inhibition lasts at least 24 h (data not shown). Interestingly,
the loss of either the ERK or the AKT pathway does not
result in apoptosis in IL-3 or EGF-treated cells (data not shown).
Thus, neither the ERK nor the PI 3-kinase pathways are necessary to
prevent apoptosis under these conditions. Moreover, neither the ERK nor
PI 3-kinase pathways are necessary for the dramatic morphologic change
seen in EGF- and IL-3 treated cells (Fig. 9C). However, if
both the MEK and PI 3-kinase inhibitors are added together, even IL-3-
and EGF-treated cells begin to die within 24 h.
View larger version (61K):
[in this window]
[in a new window]
|
Fig. 9.
The effects of inhibiting AKT
and ERK p44/42 on 32D cell morphology. IL-3 was withdrawn
from 32D EMC cells. They were subsequently incubated with either 10 µM LY294 (A and C) or 10 µM U0126 (B and C) for 1 h.
100 ng/ml EGF and IL-3 (5% WEHI conditioned medium) were added back.
Cells were lysed at 0, 1, 6, or 24 h (A) or 24 h
(B). Lysates were subjected to get electrophoresis and
immunoblotted with anti-phospho-AKT (A) or
anti-phospho-p44/42 or anti-p44/42 (B). C,
after 24 h cells were fixed, stained with Wright Giemsa stain, and
visualized. LY294 and UO126 blocked AKT and ERK activation,
respectively, and did not change the flattened elongated morphology
produced by activating EMC and the IL-3 receptor for 24 h.
|
|
| |
DISCUSSION |
The Axl/Ufo receptor tyrosine kinase family consists of
at least three members, each of which has several different names as
follows: (i) Axl/Ufo/Ark (5-7); (ii)
Tyro3/SKY/RSE/BRT/TIF/DTK/REK (8-13); and (iii) MER/NYK/EYK
(1-4, 15). Gas6, the growth arrest-specific gene 6, binds to each
mammalian member of this receptor family (16, 19-21), but the affinity
for Mer is, at least, 1 log lower (10
9 versus
108 M) than it is for Axl and Tyro3
(21). The physiological consequences of one ligand binding to divergent
members of a receptor family are not fully established although it is
not without precedent (e.g. EGF receptor family ligands)
(47).
The Mer chimera, EMC, was generated by using the
extracellular and transmembrane domains of the EGF receptor, and the
construction and transfection produced a tightly
ligand-dependent Mer signal. This construction allowed us
comparison of the acute Mer signal to the EGFR/Axl chimera
made by the same method (26). Multiple receptors including HER2, EGFR,
Ret, and the EGFR/Axl chimera all stimulate 32D cell growth
and prevent apoptosis (26, 36-39), whereas EMC prevents apoptosis
without stimulating proliferation. In conjunction with IL-3, the Mer
signal alters morphology, dramatically, and in 32D cell clones prevents
further cell growth. Clearly, the signaling output from Mer is unusual
for a receptor tyrosine kinase although certain aspects of this signal
may be due to its emanation from a chimera rather than wild-type
Mer. Further studies defining differences between full-length and
chimeric Mer signaling are needed.
Liu and co-workers (48) have shown EGF/Axl chimera
stimulates 32D cell proliferation but that Gas6 activation of
full-length Axl does not. They also showed that full-length Axl is
cleaved at the surface upon the addition of Gas6 ligand to 32D cells. This change may alter signaling duration or the range of the substrates as an alternative mechanism of down-regulation is used in the Axl-bearing 32D cells as compared with those expressing the
EGFR/Axl chimera. Whether full-length Mer and EMC have
different effects remains to be determined, but it is obvious that the
Mer chimera signal differs substantially from the
many receptor tyrosine kinases that stimulate proliferation including
the EGFR/Axl chimera.
The dramatic adhesive and morphologic changes that EMC induces in the
32D leukemic cell line derived from the mono-myelocytic lineage
resembles a form of differentiation. In turn, this
"differentiation" could explain the cessation of
IL-3-dependent growth. Prior studies of 32D cells have
shown that G-CSF and IGF1 can sustain 32D cells in the
absence of IL-3 and that G-CSF and IGF-treated cells grow slowly and
differentiate toward a granulocyte phenotype over 4-8 days (40-42).
Thus there is similarity between the G-CSF and IGF1 receptor differentiation pathways, both of which are distinct from that
of Mer. In our 32D cell clones, G-CSF was added, and after a period of
weaning from IL-3, G-CSF prevented apoptosis over 4-8 days (data not
shown). Under these conditions, G-CSF-treated cells were
morphologically distinct from those receiving a Mer signal. These
results point to a different end point of Mer signaling compared with
that of the G-CSF and the IGF1 receptors.
In another 32D cell model system, activation of an intracellular
tyrosine kinase cascade by an expressed, mutated erythropoietin receptor did result in more rapid (24-48 h) growth cessation in G1, an effect similar to that of Mer (49). This
erythropoietin receptor was constructed to activate JAK2 and leads to
the tyrosine phosphorylation of STAT 3. The biologic result, which
occurred in 24-48 h, was induction of the surface proteins, ICAM and
CD18, and a homotypic cell adhesion phenotype resulting in clumping of
cells in suspension. This clumped suspension cell phenotype is clearly
different from the Mer-induced adhesive and extended process phenotype
(Fig. 8). The Mer signal is thus distinct from that of G-CSF, IGF, and
a mutant erythropoietin receptor.
The cellular response to Axl/Mer/Tyro3 with regard to
anti-apoptotic and proliferative signaling may be cell type-specific. Substantial overexpression of Axl and Tyro3/REK can transform fibroblasts (6, 14, 24). We have not been able to transform NIH 3T3
cells with Mer even in side-by-side transfection experiments in which
Axl, using the same expression vector, was fully capable of
transforming NIH 3T3 cells. Thus, the Mer and Axl signaling output upon
overexpression appears to be different. Mer signaling as shown by EMC
(Fig. 3) can stimulate ERK activation, and Kung and co-workers (2)
demonstrated ligand-dependent (CSF1) ERK activation and
proliferation in NIH 3T3 cells when they stably transfected a
c-Fms extracellular domain and Mer transmembrane and
cytoplasmic domain chimera into these mouse fibroblasts. Thus, a Mer
signal under some circumstances can make fibroblasts grow. Hanafusa and
colleagues (50) produced and stably transfected a constitutively active
CD8 extracellular domain Mer chimera into the
BaF3 mouse lymphoid cell line. In these cells a
constitutively active, ligand-independent Mer signal not only
maintained survival but stimulated growth. It is interesting to note
that Mer is ectopically expressed in most lymphoid leukemic cells lines
(1) and in over two-thirds of childhood Acute Lymphatic
Leukemia samples.1 Perhaps Mer is mitogenic in
lymphoid cells in which it is never expressed physiologically,
whereas its function in monocytic type cells (32D cell is a
myelo-monocytic cell line) is more restrained or stimulates a
cytoskeletal rearrangement that leads to growth cessation. Cell type
specificity is also seen in the fact that Axl in some cells can prevent
apoptosis without stimulating growth (27, 28).
Recent data (30) from our own and other laboratories indicate that Mer
in addition to its role in repressing immune activation may have a
function in some cellular contexts in the regulation of specialized
cytoskeletal rearrangement. First our group has shown that Mer plays an
obligatory role in the monocyte ingestion of apoptotic material
including apoptotic thymocytes induced into programmed cell death by
dexamethasone (32). Particle phagocytosis proceeds normally in
monocytes from Mer/ mice, but although apoptotic cells
bind to the Mer-deficient monocytes, they are not internalized (32).
Additional genetic evidence demonstrating the role of Mer in the
ingestion of apoptotic rod and cone tips by pigmented retinal
epithelial cells also suggests that Mer signaling can provide a
stimulus for specialized cytoskeletal control. Gene-targeted mice (30,
31), a naturally occurring rat model (34), and a human genetic disease
(retinitis pigmentosa) (35) all have mutations abrogating the tyrosine
kinase activity of Mer resulting in retinal degeneration. The dramatic
shape changes and alteration in adhesion brought about by Mer in the
32D cell clones suggest a direct or indirect receptor-mediated signal
to the cytoskeleton. The anti-apoptotic signaling may well be shared by
all members of this family, but specific morphologic changes induced in
at least some monocytic and epithelial cells may be more characteristic of Mer.
The intracellular signaling pathways responsible for these EMC-specific
effects in 32D cells remain to be determined and must by definition
differ to some extent from the EGFR Axl chimera and the EGF
receptor. The Tyr(P) substrates for Mer differ from those of the EGF
receptor (Fig. 2), and their identification may provide clues as to the
mechanism by which Mer changes cell shape and stops cell cycle
progression. The "usual suspects," ERKs 1 and 2, JNK, p38, and PI
3-kinase may well contribute to some of the effects of EMC but are not
central to the shape changes nor do they discriminate between the EGF
receptor and Mer. All of these pathways are activated by
EMC, EGFR, and IL-3, and neither the ERKs nor PI 3-kinase by themselves
appear necessary for the morphologic changes engendered by EMC (Fig.
9). One cannot even invoke the duration of activation of ERKs as a key
component. Whereas the length of ERK activation appears to help
distinguish the growth stimulatory effect of the EGF receptor and the
differentiation effect of the nerve growth factor receptor in PC12
cells (51), IL-3 sustains ERK signaling in 32D cells without causing
the morphologic changes induced by Mer. Thus, the duration of ERK
activation alone cannot regulate this cytoskeletal-signaling pathway.
To date we cannot separate temporally or by inhibitor experiments the
"differentiation" and morphologic changes from the abrogation of
IL-3-dependent growth. Until we can do so, it is cautious
to suggest that the growth cessation is due to the cytoskeletal changes
rather than a product of a distinct cell cycle regulatory mechanism.
,
TOP
ABSTRACT
INTRODUCTION
These authors contributed equally to this work.
To whom correspondence should be addressed. Fax: 919-966-3015;
E-mail: hse@med.unc.edu.
CiteULike 
Complore 
