(Received for publication, September 6, 1996, and in revised form, April 23, 1997)
From the Department of Biochemistry and Molecular Biology, St. Louis University School of Medicine, St. Louis, Missouri 63104
The type V transforming growth factor (TGF-
) is a 400-kDa nonproteoglycan membrane protein that
co-expresses with the type I, type II, and type III TGF-
receptors
in most cell types. The type V TGF-
receptor exhibits a
Ser/Thr-specific protein kinase activity with distinct substrate
specificity (Liu, Q., Huang, S. S., and Huang, J. (1994) J. Biol. Chem. 269, 9221-9226). In mink lung epithelial cells, the
type V TGF-
receptor was found to form heterocomplexes with the type
I TGF-
receptor by immunoprecipitation with antiserum to the type V
TGF-
receptor after 125I-TGF-
affinity labeling or
Trans35S-label metabolic labeling of the cells. The kinase
activity of the type V TGF-
receptor was stimulated after treatment
of mink lung epithelial cells with TGF-
. TGF-
stimulation
resulted in the growth inhibition of wild-type mink lung epithelial
cells and to a lesser extent of the type I and type II TGF-
receptor-defective mutants, although higher concentrations of TGF-
were required for the growth inhibition of these mutants. TGF-
was
unable to induce growth inhibition in human colorectal carcinoma cells
lacking the type V TGF-
receptor but expressing the type I and type
II TGF-
receptors. These results suggest that the type V TGF-
receptor can mediate the TGF-
-induced growth inhibitory response in
the absence of the type I or type II TGF-
receptor. These results also support the hypothesis that loss of the type V TGF-
receptor may contribute to the malignancy of certain carcinoma cells.
Transforming growth factor (TGF-
)1 is the most potent polypeptide
growth inhibitor for epithelial cells and plays an important role in
the pathophysiology of epithelial cells in human and other species
(1-3). The TGF-
-induced growth inhibition of epithelial cells is
mediated by specific cell surface receptors (1-4). The type V TGF-
receptor is a 400-kDa nonproteoglycan membrane glycoprotein that
co-expresses with the type I, type II, and type III TGF-
receptors
in epithelial cells and other cell types but not in certain carcinoma
cells (4, 5). The type V TGF-
receptor as well as the type I and
type II TGF-
receptors are members of a new class of
Ser/Thr-specific receptor protein kinases with distinct substrate
specificities (6-11). The exact roles of these receptor kinases in the
growth inhibition of epithelial cells induced by TGF-
are unknown.
Recent studies have demonstrated that the heterocomplex formation of
the type I and type II TGF-
receptors and the transphosphorylation
of the type I TGF-
receptor by the type II TGF-
receptor are
important in the TGF-
-induced growth inhibition (9-12). The role of
the type V TGF-
receptor in the growth inhibition induced by TGF-
has not been defined. These studies reported here suggest that the type
V TGF-
receptor forms heterocomplexes with the type I TGF-
receptor in mink lung epithelial cells and can mediate the
TGF-
-induced growth inhibition in mink lung epithelial cells in the
absence of the type I or type II TGF-
receptor. These studies
reported here also support the hypothesis that loss of the type V
TGF-
receptor may contribute to the malignancy of certain carcinoma
cells.
MaterialsNa125I (17 Ci/mg),
[
-32P]ATP (4,500 Ci/mmol), and
Trans35S-label (1,000 Ci/mmol) were obtained from ICN
Biochemicals, Inc. (Irvine, CA). [
-32P]ATP was diluted
with unlabeled ATP to have a specific radioactivity of
~104 cpm/pmol. Molecular mass protein standards (myosin,
205 kDa;
-galactosidase, 116 kDa; phosphorylase, 97 kDa; bovine
serum albumin, 66 kDa; ovalbumin, 45 kDa; carbonic anhydrase, 29 kDa), poly-L-lysine HBr (Mr
15,000-30,000),
-mercaptoethanol, glycerol, Triton X-100, and other
chemical reagents were obtained from Sigma. TGF-
1 was
purchased from Austral Biologicals (San Ramon, CA). Disuccinimidyl
suberate (DSS) was obainted from Pierce. Peptide antigen, a
hexadecapeptide containing the ATP binding site amino acid sequence (6)
was synthesized using tert-butoxycarbonyl chemistry on an
Applied Biosystems model 431A peptide synthesizer. Recombinant human
nonglycosylated insulin-like growth factor binding protein 3 (IGFBP-3)
was provided by Celtrix Pharmaceuticals, Inc. (Santa Clara, CA).
TGF-
receptor-defective mutants (DR26 and R1-B cells) and type II
TGF-
receptor cDNA and neo-vector stably transfected hereditary
human colorectal carcinoma cells (RII-37 and HCT 116 Neo cells) were
provided by Drs. Joan Massagué and Michael G. Brattain,
respectively. Mink lung epithelial cells have been routinely maintained
in the laboratory. All cultured cells were grown in 10% fetal calf
serum in Dulbecco's modified Eagle's medium.
The
antiserum to the type V TGF- receptor was raised in rabbits with the
conjugate of bovine thyroglobulin and peptide antigen (a
hexadecapeptide), whose amino acid sequence was derived from the ATP
binding site amino acid sequence of the type V TGF-
receptor (6).
The peptide antigen was conjugated to bovine thyroglobulin according to
the procedure of Huang and Huang (13). The antiserum to the type V
TGF-
receptor did not show reactivity to the type I and type II
TGF-
receptors based on Western blot analysis and immunoprecipitation of 125I-TGF-
affinity labeled or
Trans35S-label metabolically labeled type I and type II
TGF-
receptors from mink lung epithelial cells in the presence of
0.1% SDS.
The type V TGF- receptor was purified by
DEAE-cellulose column chromatography after Triton X-100 extraction of
bovine liver plasma membranes and wheat germ lectin-Sepharose 4B
affinity column chromatography as described previously (4).
About 0.1 µg of the type V TGF-
receptor from bovine liver plasma membranes was subjected to 6%
SDS-polyacrylamide gel electrophoresis under reducing conditions
followed by electrophoretic transblotting onto nitrocellulose membranes
(Protran). Western blot analysis of the type V TGF-
receptor was
performed as described previously (14), using antiserum to the type V
TGF-
receptor (1:100 dilution) with or without peptide antigen (0.5 mM).
Mink lung epithelial cells (Mv1Lu cells) were
grown to confluence on P-60 Petri dishes and metabolically labeled with
[32P]orthophosphate according to the procedure of Huang
and Huang (13). The monolayers were then treated with 0.1 nM TGF-1 in Dulbecco's modified Eagle's
medium, pH 7.4, at 0 °C for 30 min. The cells were then detached and
lysed in 100 µl of 1% Triton X-100 in 10 mM Tris-HCl, pH
7.0, 125 mM NaCl, and 1 mM EDTA. After centrifugation, the Triton X-100 extracts were then diluted 10-fold with Triton X-100-free buffer and incubated with antiserum or nonimmune
serum (1:100 dilution) at 0 °C overnight. The immunocomplexes were
precipitated with 20 µl of protein A-Sepharose (50%, v/v). After
washing with 20 mM Tris-HCl, pH 7.4, 0.2% Triton X-100, the immunoprecipitates were analyzed by 6% SDS-polyacrylamide gel
electrophoresis under reducing conditions and autoradiography. The
relative intensity of 32P-labeled type V TGF-
receptor
on the autoradiogram was quantitated by a PhosphorImager.
Mink lung
epithelial cells (wild-type, DR26, and R1B cells) and hereditary human
colorectal carcinoma cells (HCT 116 Neo and RII-37 cells) grown on P-60
Petri dishes were incubated with 125I-TGF-1
(0.5, 1, and 1.5 nM) in binding buffer (50 mM
HEPES, pH 7.4, 128 mM NaCl, 5 mM KCl, 5 mM MgSO4, and 1.2 mM
CaCl2) containing 0.2% bovine serum albumin at 0 °C for
2.5 h. The affinity labeling of cell surface TGF-
receptors was
carried out as described previously (4, 5). The
125I-TGF-
1 affinity labeled receptors were
then analyzed by 5.5% SDS-polyacrylamide gel electrophoresis under
reducing conditions and autoradiography.
The immunoprecipitation of the type V TGF- receptor
in mink lung epithelial cells treated with and without 0.1 nM TGF-
1 at 0 °C for 30 min was carried
out in 0.1% Triton X-100 as described above, except the cells were not
metabolically labeled. The immunoprecipitates were incubated with 0.2 µM recombinant nonglycosylated human IGFBP-3 in 50 µl
of 20 mM HEPES, pH 7.4, containing 10% glycerol, 0.1% Triton X-100, 5 µM [
-32P]ATP, 0.1%
mercaptoethanol, and 2.5 mM MnCl2. An aliquot
of the reaction mixture was then analyzed by 7.5% SDS-polyacrylamide gel electrophoresis under reducing conditions and autoradiography. The
relative intensity of 32P-IGFBP-3 on the autoradiogram was
quantitated by a PhosphorImager.
Mink lung epithelial cells were grown to confluence on P-60
Petri dishes and metabolically labeled with Trans35S-label
([35S]methionine + [35S]cysteine) according
to the procedure of Huang and Huang (13). The cells were then detached
and lysed with Triton X-100 buffer (1% Triton X-100 in 100 mM Tris-HCl, pH 7.0, 125 mM NaCl2,
and 1 mM EDTA) or RIPA buffer (1% sodium droxycholate, 1%
Nonidet P-40, 0.1% SDS, 25 mM Tris-HCl, pH 7.4, and 0.15 M NaCl). After centrifugation, the Triton X-100 (final
concentration, 0.1%) or RIPA buffer extracts were immunoprecipitated
with antiserum to the type V TGF- receptor (1:100 dilution). The
immunoprecipitates were then analyzed by 5.5% SDS-polyacrylamide gel
electrophoresis and fluorography.
Mink lung epithelial cells were grown to confluence on P-60
Petri dishes and affinity labeled with 125I-TGF- in the
presence of DSS as described previously (4, 5). The cells were then
lysed with Triton X-100 buffer as described above. After
centrifugation, the Triton X-100 buffer extracts were
immunoprecipitated with antiserum to the type V TGF-
receptor (1:100
dilution). The immunoprecipitates were analyzed by 5.5% SDS-polyacrylamide gel electrophoresis and autoradiography.
Wild-type mink lung epithelial cells (Mv1Lu cells), mutants
(R-1B and DR26 cells), and human colorectal carcinoma cells (HCT 116 Neo and RII-37 cells) were plated on 24-well clustered dishes and
incubated with various concentrations of TGF-1 in
Dulbecco's modified Eagle's medium containing 0.1% fetal calf serum.
After incubation at 37 °C for 16 h, the cells were
pulse-labeled with 1 µCi of
[methyl-3H]thymidine at 37 °C for 4 h.
The cells were then washed twice with 1 ml of 10% trichloroacetic
acid, washed once with 0.5 ml of ethanol:ether (2:1, v/v), and
dissolved in 0.2 N NaOH. The [methyl-3H]thymidine incorporation was
determined by a scintillation counter.
The type I, type II, and type III TGF- receptors have been
shown to form heterocomplexes upon ligand binding (12, 15-17). The
formation of heterocomplexes appears to be important for initiation of
signaling (type I and type II TGF-
receptor complex) (9-12, 15) or
presentation of ligand (type III and type I/II TGF-
receptor
complexes) (16, 17). To see whether the type V TGF-
receptor formed
heterocomplexes with other TGF-
receptors, we performed
immunoprecipitations of the type V TGF-
receptor and other TGF-
receptors using specific antiserum to the type V TGF-
receptor. The
specificity of the antiserum to the type V TGF-
receptor had been
validated by two types of evidence. First, the antiserum reacted with
the type V TGF-
receptor in Western blot analysis. Fig.
1 shows that the antiserum specifically reacted with the
type V TGF-
receptor and its proteolytic product (~300 kDa) from
bovine liver plasma membranes on the Western blot analysis (Fig.
1A, lane 1). The reaction was blocked in the
presence of 0.5 mM peptide antigen (Fig. 1A,
lane 2). Second, the antiserum was also able to
immunoprecipitate the type V TGF-
receptor in normal mink lung
epithelial cells (Mv1Lu cells) that were metabolically labeled with
[32P]orthophosphate in response to TGF-
1
stimulation (Fig. 1B). The immunoprecipitated type V
receptor exhibited a kinase activity toward IGFBP-3 (Fig.
1C). IGFBP-3 is a nonphysiological substrate for the type V
TGF-
receptor but contains several SXE motifs that serve
as the phosphorylation sites for the type V TGF-
receptor kinase
activity (7). The radioactive band at the top of the separating gel
(Fig. 1C, lanes 1 and 2) was
identified as the TGF-
1-stimulated and -unstimulated
autophosphorylated type V TGF-
receptor that could be removed by
subsequent immunoprecipitation with antiserum to the type V receptor in
the presence of 0.1% SDS. The TGF-
1-stimulated
phosphorylation of the type V TGF-
receptor and the
TGF-
1-stimulated kinase activity of the type V TGF-
receptor on IGFBP-3 were both estimated to be ~1.5-fold greater than
those observed without TGF-
1 stimulation (Fig. 1, B and C, lanes 1 and 2).
This 1.5-fold stimulation was comparable with that observed for the
TGF-
1-stimulated kinase activity of the type V TGF-
receptor purified from bovine liver plasma membranes (7).
Normal mink lung epithelial cells (Mv1Lu cells) have been commonly used
to examine the cross-interaction of the TGF- receptors and
TGF-
-induced signal transduction. They express all the known TGF-
receptors (type I, type II, type III, and type V TGF-
receptors) and
exhibit both growth inhibition and transcriptional activation in
response to TGF-
stimulation (5, 18-20). In this study, the
125I-labeled TGF-
1 affinity labeled cell
surface receptors were immunoprecipitated with the antiserum to the
type V TGF-
receptor, after mink lung epithelial cells (Mv1Lu cells)
had been incubated with 0.1 nM
125I-TGF-
1 for 2 1/2 h at 0 °C,
exposed to the cross-linking reagent DSS, and extracted with Triton
X-100 buffer. The immunoprecipitates of the Triton X-100 buffer
extracts were then analyzed by 5.5% SDS-polyacrylamide gel
electrophoresis under reducing conditions and autoradiography. As shown
in Fig. 2A, the type I TGF-
receptor was
co-immunoprecipitated with the type V TGF-
receptor (Fig. 2A, lane 2). The immunoprecipitation efficiency
for the type V TGF-
receptor was estimated to be approximately 30%
of the total labeled receptor (Fig. 2A, lane 1).
Excess peptide antigen blocked the immunoprecipitation of the TGF-
receptors (Fig. 2A, lane 3). Nonimmune serum did
not immunoprecipitate the TGF-
receptors (Fig. 2A,
lane 4). These results suggest that on the cell surface of
mink lung epithelial cells, the type V TGF-
receptor forms heterocomplexes with the type I TGF-
receptor in the presence of the
ligand.
To see whether the ligand was required for the heterocomplex formation,
the TGF- receptors were immunoprecipitated with antiserum to the
type V TGF-
receptor after mink lung epithelial cells (Mv1Lu cells)
had been metabolically labeled with Trans35S-label in the
absence of exogenous ligand and extracted with Triton X-100 buffer or
RIPA buffer (13). The immunoprecipitates of the Triton X-100 or RIPA
buffer extracts were analyzed by 5.5% SDS-polyacrylamide gel
electrophoresis under reducing conditions and fluorography. As shown in
Fig. 2B, the type V and type I (53 kDa) TGF-
receptors
were found in the immunoprecipitates of the Triton X-100 buffer
extracts (Fig. 2B, lane 3), whereas only the type
V TGF-
receptor was detected in the immunoprecipitates of the RIPA
buffer extracts (Fig. 2B, lane 1). A 68-kDa
protein was also identified in the immunoprecipitates of the Triton
X-100 buffer extracts (Fig. 2B, lane 3). This
protein was possibly an isoform or differentially glycosylated form of
the type I TGF-
receptor because two different molecular mass
isoforms of the type I TGF-
receptor were also identified in the
125I-TGF-
1 affinity labeling experiment
(Fig. 2A, lanes 1 and 2). RIPA buffer
contained 0.1% SDS that destabilized the heterocomplexes of the type V
and type I TGF-
receptors. In the presence of RIPA buffer, only the
type V TGF-
receptor was immunoprecipitated (Fig. 2B,
lane 1). Because no endogenous TGF-
activity (growth inhibitory activity) was detected under the cultured conditions and
because exogenous TGF-
1 did not affect the complexation
of 35S-labeled type V and type I TGF-
receptors (data
not shown), these results suggest that the type V TGF-
receptor can
form heterocomplexes with the type I TGF-
receptor in the absence of
ligand.
As described previously, mink lung epithelial cells (Mv1Lu cells) have
been a useful system for studying the roles of the type I and type II
TGF- receptors in TGF-
-induced cellular responses. Mutant cells
defective for the type I or type II TGF-
receptor have been reported
to be unable to exhibit growth inhibition and transcriptional
activation following TGF-
stimulation (18-21). However, the growth
inhibition and transcriptional activation could be restored by genetic
complementation between cells defective in the type I and type II
TGF-
receptors (21). No study on the expression of the type V
TGF-
receptor in these mutants has been reported. To test its
presence, we investigated the expression of the type V TGF-
receptor
in these mutants.
As shown in Fig. 3 (A and B), both
the type II and type I TGF- receptor-defective mutants (DR26 and
R-1B cells, respectively) did express the type V TGF-
receptor,
which was identified as the 400-kDa
125I-TGF-
1 affinity labeled protein. It is
of importance to note that the type I TGF-
receptor-defective mutant
(R-1B cells) expressed less of the type V TGF-
receptor when
compared with the type II TGF-
receptor-defective mutant (DR26
cells).
Because both mutants (DR26 and R-1B cells) express the type V TGF-
receptor, an important role of the type V TGF-
receptor in
TGF-
-induced cellular responses cannot be excluded in the genetic
complementation experiments of the mutants (21). To determine whether
the type V TGF-
receptor mediated a TGF-
-induced growth
inhibition, we investigated the effect of higher concentrations of
TGF-
1 on [methyl-3H]thymidine
incorporation into DNA of mink lung epithelial cells (Fig.
3C). The Kd values for
TGF-
1 or TGF-
2 binding to the type V
receptor are approximately 40-fold higher than those for the type I and
type II TGF-
receptors (6). If the type V TGF-
receptor can
mediate a TGF-
-induced growth inhibition, TGF-
at higher
concentrations should inhibit the DNA synthesis in the type I and type
II TGF-
receptor-defective mutants. At ~50-100 pM,
TGF-
1 induced a small but significant inhibition of DNA
synthesis (Fig. 3C) in both mutants (~20-40%). In type II TGF-
receptor-defective mutant cells (DR26 cells), the inhibition of DNA synthesis showed a small downward trend at
125 pM
(Fig. 3C), but higher concentrations of TGF-
1
(~500 pM) did not cause a further decrease (30 ± 4% inhibition). A smaller degree of inhibition of DNA synthesis by
TGF-
1 in type II TGF-
receptor-defective mutants was
previously reported but was not discussed (20). These results suggest
that the type V TGF-
receptor may mediate the growth inhibition in
the absence of the type I or type II TGF-
receptor, although a
higher level of ligand is required for the type V TGF-
receptor-mediated growth inhibition. In the wild-type cells (Mv1Lu
cells) containing all TGF-
receptors, the maximal inhibition was
observed at ~1 pM of TGF-
1.
In a previous study, we reported that several carcinoma cells lacked
the type V TGF- receptor and other TGF-
receptors (5). These
cells lacking the type V TGF-
receptor (MCF-7 and PC-12 cells) have
been found not to respond to TGF-
stimulation with respect to growth
inhibitory response (5).2 Recently,
heriditary human colorectal carcinoma cells (HCT 116 cells) were shown
to be deficient in the type II TGF-
receptor (22). Stable
transfection of these colorectal carcinoma cells with the type II
TGF-
receptor cDNA was found to rescue the transcriptional response but failed to restore the growth inhibitory response to
exogenous TGF-
(22). The inability of the transfected colorectal carcinoma cells to exert growth inhibitory response to TGF-
stimulation was also confirmed in our laboratory. One of the
possibilities for the failure to restore the growth inhibitory response
could be the lack of the type V TGF-
receptor expression in cells
stably transfected with the neo-vector only (HCT 116 Neo cells) or with vector expressing the type II TGF-
receptor cDNA (RII-37 cells). To test this possibility, we performed the
125I-TGF-
1 affinity labeling of the TGF-
receptors in these cells. No detectable type V TGF-
receptor was
found in these cells (Fig. 4, lanes 1-4).
The type III and type II TGF-
receptors were detected in RII-37
cells (Fig. 4, lane 3). The type I TGF-
receptor migrated at the dye front and could not be identified in the 5% polyacrylamide system (Fig. 4, lane 3). Absence of the type V TGF-
receptor was confirmed by the observation that the type V TGF-
receptor antigen was not detected by Western blot analysis in HCT 116 Neo and RII-37 cells (data not shown). These results and the results from several experiments as shown in Fig. 3 can be summarized in Table
I. The type V TGF-
appears to be critical for
TGF-
-induced growth inhibition. The type I and type II TGF-
receptors are required for the maximal growth inhibition induced by
TGF-
. Together with the observation that only transformed epithelial
cells (carcinoma cells) have been found to lack the expression of
detectable type V TGF-
receptor (5), these results also support the
hypothesis that loss of the type V TGF-
receptor may contribute to
the transformed state of certain carcinoma cells (e.g.
hereditary human colorectal carcinoma cells) (5).
|
Together with the previous observations of distinct substrate
specificities of the type V, type I, and type II TGF- receptors (7,
23), the finding of heterocomplex formation of the type V TGF-
receptor and type I TGF-
receptor reported here raise an important
question: What role does the type V TGF-
receptor play in various
cellular responses induced by three different TGF-
isoforms
(TGF-
1, TGF-
2, and TGF-
3)?
The cellular responses induced by the TGF-
isoforms can vary
substantially (1-3, 24, 25). The molecular mechanisms of the opposite
effects of TGF-
3 versus TGF-
1
or TGF-
2 (25) are not easy to interpret within a model in which only the type I and type II TGF-
receptor heterocomplex mediates the signaling. The type V TGF-
receptor could be involved in these diverse cellular responses due to its distinct kinase substrate specificity (acidotrophic kinase activity versus
the basic-trophic kinase activities of the type I and type II TGF-
receptors) (7, 23) and its different binding affinities to TGF-
isoforms. The distinct substrate specificity of the type V TGF-
receptor implies a different signaling pathway from those of the type I
and type II TGF-
receptors. Segregation of growth inhibitory and
transcriptional responses induced by TGF-
has been reported (22, 26,
27). The signaling pathway mediated by the type V TGF-
receptor
could be important for growth inhibitory response but not obligatory
for transcriptional response (the activation of transcription of
fibronectin, collagen and plasminogen activator inhibitor-1 genes)
(22). The cross-modulation of the two pathways mediated by the type V/I
and type II/I TGF-
receptors should be determined by the affinities
of TGF-
isoforms to the TGF-
receptors and extracellular
concentrations of TGF-
isoforms. The binding affinities of
TGF-
1, -
2, and -
3 for the
type I and type II TGF-
receptors are very similar with
Kd of ~0.01 nM (24). The
Kd of the type V TGF-
receptor for
TGF-
1, -
2, and -
3 have
been estimated to be ~0.4 nM, ~0.4 nM, and
~5 nM, respectively (7).2 Low concentrations
of TGF-
3 might favor the formation of the type I and
type II TGF-
receptor heterocomplexes that do not include the type V
TGF-
receptor. Under the condition in which the amount of the type I
receptor is limiting compared with those of other TGF-
receptors,
TGF-
3 could alter the availability of the type I TGF-
receptor when the formation of type I and type V TGF-
receptor
heterocomplex is required for certain cellular responses induced by
TGF-
1 or TGF-
2.
We thank Dr. William S. Sly for critical
review of the manuscript, Drs. Joan Massagué and Michael G. Brattain for kindly providing TGF- receptor-defective mutants (DR26
and R-1B cells) and type II TGF-
receptor cDNA and neo-vector
transfected hereditary human colorectal carcinoma cells (HCT 116 cells), and Celtrix Pharmaceuticals, Inc., for providing recombinant
human nonglycosylated IGFBP-3. We also thank Tao Zhao for performing
the kinase assay of the immunoprecipitated type V receptor and Maggie
Klevorn for preparing the manuscript.