 |
INTRODUCTION |
The growth hormone receptor
(GHR)1 is a 620-residue
transmembrane glycoprotein that binds GH in its extracellular domain
and transduces activating signals via its cytoplasmic domain (1-3). GH
signals a variety of metabolic and somatogenic effects in multiple target tissues (4). Because of conserved structural features in its
extracellular domain, GHR is considered a member of the cytokine
receptor superfamily, which includes receptors for prolactin, erythropoietin, and various interleukins and colony-stimulating factors, among others (5). Many of these receptors also share common
signaling features; most notably, they couple to members of the Janus
family of cytoplasmic tyrosine kinases (JAKs) to initiate their signal
cascades in response to ligand binding (6). GHR couples in particular
to JAK2 (7).
The interaction of GH with the cell surface GHR is complicated but is
now understood structurally with a high degree of resolution. Crystallographic and kinetic studies employing recombinant GH and GHR
extracellular domain have elegantly demonstrated that GH interacts with
the GHR to form a complex of 1:2 GH·GHR stoichiometry (8, 9).
Although GH is a molecule composed of four antiparallel helical bundles
without an axis of symmetry, two distinct sites within each molecule of
the hormone engage two GHR extracellular domains at nearly identical
contact points on each receptor (9). Binding of GH "site 1" to one
GHR facilitates interaction of "site 2" with the second receptor
(10). The GH·GHR2 complex is further stabilized by
interaction between receptor dimer partners in the region of the GHR
extracellular domain immediately proximal to the external face of the
plasma membrane (9). In addition, by using the human IM-9 lymphoblast
that homologously expresses GHRs as a model system, we have detected a
GH-dependent disulfide linkage of GHRs that is rapid and
quantitatively significant (11). Although the precise physiological
role(s) of the disulfide linkage is enigmatic, a large fraction of GHRs
activated by GH (as evidenced by tyrosine phosphorylation) undergo this
linkage, and receptors becoming detergent-insoluble in response to GH
are progressively accounted for by disulfide-linked GHRs (12).
In contrast to the degree to which the extracellular interaction of the
GHR with GH is known, the intracellular association of the receptor
with its important signal transducer, JAK2, is only partly understood.
Various lines of evidence indicate that physical interaction of the GHR
and JAK2 is necessary for GH-induced JAK2 tyrosine kinase activation,
tyrosine phosphorylation of JAK2, the GHR, and numerous signaling
molecules and biological effects (Refs. 2 and 3 and references
therein). Binding and functional assays demonstrate that within the
GHR, the proline-rich Box 1 region (residues 279-286) is necessary and
the amino-terminal one-third of the cytoplasmic domain is sufficient
for nearly full interaction with JAK2 (13-15), and reciprocally, that
regions of the amino terminus of JAK2 are required for association with
the GHR (16, 17). The GHR-JAK2 interaction can occur without GH occupancy of the GHR and can be detected in vitro without
tyrosine phosphorylation of JAK2 or the GHR (15); however, GH treatment of cells augments the degree and/or strength of association of JAK2
with the GHR (7, 15, 18).2
Whereas this enhanced GHR-JAK2 association likely relates to GH-induced
JAK2 signaling, the structural basis for this augmentation has yet to
be intensively investigated.
In this communication, we biochemically and immunologically examine in
several GHR-expressing cell lines the relationships between GH-induced
GHR dimerization, GHR disulfide linkage, and GHR-JAK2 interaction. In
so doing, we characterize several new reagents and refine the use of
our existing tools and techniques to develop approaches to further
study these proximal aspects of GH signaling. Our results indicate that
GH-induced GHR disulfide linkage is detectable for murine and rabbit
(as well as human) GHRs and can be detected in both non-adherent (IM-9)
and adherent (3T3-F442A, COS-7, and CHO fibroblast) cell lines. GHR
disulfide linkage as well as the loss of immunoprecipitability of GHRs
with a new anti-GHR extracellular domain monoclonal antibody are shown to be useful reflectors of GH-induced GHR dimerization. Furthermore, we
demonstrate that GH-induced enhancement of coimmunoprecipitation of
JAK2 with anti-GHR serum directed at the receptor cytoplasmic domain is
related to receptor dimerization, rather than receptor or JAK2
activation. Finally, by identifying an unpaired extracellular domain
cysteine (residue 241) as critical to GH-induced GHR disulfide linkage,
we differentiate noncovalent from disulfide-linked GHR dimerization,
allowing us to compare these processes as they apply to the GHR with
findings for other cytokine and growth factor receptors.
| |
EXPERIMENTAL PROCEDURES |
Materials--
Recombinant hGH was kindly provided by Lilly.
Recombinant hGH-G120K was kindly provided by Sensus Corp. (Austin, TX).
Routine reagents were purchased from Sigma unless otherwise noted.
Restriction endonucleases were obtained from New England Biolabs
(Beverly, MA). 125I-hGH (specific activity 85-130
µCi/µg) was purchased from NEN Life Science Products.
Cells, Cell Culture, Transfections, and 125I-hGH
Binding--
COS-7, IM-9, and 3T3-F442A (Ref. 19; a gift of Drs. H. Green (Harvard University) and C. Carter-Su (University of Michigan)) cells were maintained as described previously (12, 20). COS-7 cells
were transiently transfected at 75% confluency in 20 ml of DMEM in
150 × 25-mm dishes (Falcon) by the calcium phosphate precipitation method (21) as described previously (15, 17). Each dish
was transfected with either 50 µg of pSX rGHR, pSX rGHR C241S, pSX
rGHRd297-406, pSX rGHRd294-498, or pSX rGHR1-390, along with 25 µg of pSX JAK2 (in some
experiments), as indicated in the figure legends. For samples in which
no exogenous GHR or JAK2 was expressed, pSX alone was added to equalize
the amount of total DNA added in each transfection. Description of the
pSX vector and construction of the plasmids used is found below. Serum
starvation was begun at 24 h after transfection and continued for
16-20 h prior to stimulation, harvesting, and detergent extraction, as
below. 125I-hGH binding capacity of COS-7 cells transiently
expressing WT rGHR and rGHR C241S was assayed as described previously
(15).
CHO cells (a gift of Dr. J. Kudlow, University of Alabama, Birmingham)
were maintained in DMEM (1 g/liter glucose) (Cellgro, Inc.)
supplemented with 7% fetal bovine serum (Biofluids, Rockville, MD) and
50 mg/ml gentamicin sulfate, 100 units/ml penicillin, and 100 mg/ml
streptomycin (all Biofluids). Stable transfection of CHO cells was
achieved by introducing either pSX rGHR or pSX rGHR C241S (20 µg of
each in 3 ml of DMEM in 60 × 15-mm dishes), each along with 1 µg of pSP65-SR
.2-HAtag-HygroHA-Hygro (empty vector carrying the
hygromycin resistance marker, kindly provided by Dr. M. Streuli,
Dana-Farber Cancer Institute, Boston), using Lipofectin (Life
Technologies, Inc.) according to the manufacturer's protocol.
Transfected cells were grown in complete DMEM growth medium for 48 h. After dilution, clones were negatively selected in medium
supplemented with 500 µg/ml hygromycin B (Roche Molecular Biochemicals) and screened for GHR expression by
anti-GHRcyt-AL37 immunoblotting of detergent extracts, as
in text and figure legends (untransfected CHO cells express no
detectable GHR (13, 14).2
Antibodies--
The 4G10 monoclonal antiphosphotyrosine (APT)
antibody was purchased from Upstate Biotechnology, Inc. (Lake Placid,
NY). Anti-STAT5 monoclonal antibody was purchased from Transduction
Laboratories (Lexington, KY). Anti-phospho-STAT5 affinity-purified
rabbit polyclonal antibody (recognizing the tyrosine-phosphorylated
form (Y694) of STAT5a and STAT5b) was purchased from Zymed
Laboratories (San Francisco, CA).
The rabbit polyclonal sera, anti-GHRcyt (11), directed at
the residue 317-620 region of the human GHR cytoplasmic domain, and
anti-GHRcyt-AL37 (18), directed at residues 271-620 of the human GHR (the entire cytoplasmic domain), have been described. The
anti-GHRcyt-mAb mouse monoclonal antibody (IgG2b
) was
also raised against a GST fusion protein incorporating residues
271-620 of the human GHR (15) and screened by enzyme-linked
immunosorbent assay at the University of Alabama at Birmingham
Multipurpose Arthritis Center Hybridoma Facility (Dr. M. Accaviti).
Anti-GHRcyt-mAb reacted by enzyme-linked immunosorbent
assay specifically with GST/hGHR271-620 and not with GST.
Similarly, immunoblotting of thrombin-cleaved
GST/hGHR271-620 indicated that anti-GHRcyt-mAb reacted with hGHR271-62, but not with GST. The
anti-GHRext mouse monoclonal antibody (IgG1) (20, 22),
raised against a GST fusion protein incorporating residues 1-245 of
the rabbit (r) GHR, has been described, as has the rabbit polyclonal
serum, anti-JAK2AL33 (18), raised against residues
746-1129 of murine JAK2 (23). Anti-GHRcyt-mAb and
anti-GHRext were purified from ascites using the Affi-Gel
Protein A MAPS II Kit (Bio-Rad), according to the manufacturer's suggestions.
Plasmid Construction--
The pSX plasmid (a kind gift of Dr.
J. Bonifacino, National Institutes of Health, and Dr. K. Arai,
DNAX), which drives eukaryotic protein expression from the SR
promoter (composed of the SV40 early promoter and the R-U5 segment of
the HTLV-1 long terminal repeat), has been described (20, 24).
Preparation of the rGHR cDNA (a kind gift of W. Wood, Genentech,
Inc.) and the cDNAs encoding rGHRd297-406 (where
subscript d indicates deletion) and rGHR1-390 and their
ligation into pSX have been described (15, 17). The cDNA encoding
rGHRd294-498, with internal in-frame deletion of residues
294-498, was the result of BsaB1 digestion and religation
of the wild-type rGHR cDNA. The cDNA encoding rGHR C241S
(changing the codon for residue 241 from cysteine to serine) was
generated using the unique site elimination in vitro
mutagenesis strategy (25), unique site elimination mutagenesis kit by
Amersham Pharmacia Biotech, according to the manufacturer's suggestions. Primer sequences are available upon request. Confirmation of the desired mutations and the absence of unwanted mutations in the
cDNA constructs was accomplished by dideoxy DNA sequencing.
Cell Stimulation and Protein Extraction--
Serum starvation of
3T3-F442A, IM-9, and transfected COS-7 and CHO cells was accomplished
by substitution of 0.5% (w/v) bovine serum albumin (fraction V, Roche
Molecular Biochemicals) for serum in their respective culture media for
16-20 h prior to experiments. Unless otherwise noted, hGH was used at
a final concentration of 500 ng/ml, and stimulations were performed at
37 °C. Details of the hGH treatment protocol have been described
(12, 20). Briefly, for IM-9 cells, cells were stimulated in suspension
at 10 million cells/ml in binding buffer (BB, consisting of 25 mM Tris-HCl (pH 7.4), 120 mM NaCl, 5 mM KCl, 1.2 mM MgCl2, 0.1% (w/v) bovine serum albumin, and 1 mM dextrose). Stimulations were
terminated, and cells were collected by centrifugation (800 × g for 1 min at 4 °C) and aspiration of the BB. 3T3-F442A
cells and transfected COS-7 cells were stimulated in confluent 150 × 20-mm dishes (Falcon) in BB. Stimulations were terminated by washing
the cells once with PBS and then harvesting by scraping in ice-cold PBS
in the absence or presence (PBS/vanadate) of 0.4 mM sodium
orthovanadate, as indicated in figure legends. Pelleted cells were
collected by brief centrifugation. For each cell type, pelleted cells
were solubilized for 15 min at 4 °C in fusion lysis buffer (FLB)
(1% (v/v) Triton X-100, 150 mM NaCl, 10% (v/v) glycerol,
50 mM Tris-HCl (pH 8.0), 100 mM NaF, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 10 mM benzamidine, 10 µg/ml aprotinin), as indicated. After centrifugation at 15,000 × g for 15 min at 4 °C, the detergent extracts were
either subjected to immunoprecipitation or were directly
electrophoresed under nonreducing or reducing conditions, as indicated
below. In the case of CHO-rGHR and CHO-rGHR C241S, harvested cells were
in some experiments solubilized directly in reducing SDS-PAGE sample
buffer, as indicated, prior to electrophoresis.
Immunoprecipitation, Electrophoresis, and
Immunoblotting--
For immunoprecipitation, the rabbit antisera
described above were used at the following volumes per precipitation:
anti-JAK2AL33, 3 µl; anti-GHRcyt-AL37, 3 µl. For immunoprecipitation of the GHR with the monoclonal
anti-GHRext and anti-GHRcyt-mAb antibodies, 0.6 µg of purified antibody was used per precipitation. Protein A-Sepharose (Amersham Pharmacia Biotech) (or, for
anti-GHRext, protein G-Sepharose (Amersham Pharmacia
Biotech)) was used to adsorb immune complexes, and after extensive
washing with lysis buffer, SDS sample buffer eluates were resolved by
SDS-PAGE and immunoblotted as indicated. Resolution of proteins under
reducing or nonreducing conditions by SDS-PAGE, Western transfer of
proteins, and blocking of Hybond-ECL (Amersham Pharmacia Biotech) with
2% bovine serum albumin were performed as described previously (12, 26). Immunoblotting with antibodies 4G10 (1:2500), anti-phospho-STAT5 (1:1000), anti-GHRcyt-AL37 (1:2000),
anti-GHRcyt (1:2000), or anti-JAK2AL33
(1:2000), with horseradish peroxidase-conjugated anti-mouse or
anti-rabbit secondary antibodies (1:2000) and ECL detection reagents
(all from Amersham Pharmacia Biotech) and stripping and reprobing of
blots were accomplished according to manufacturer's suggestions.
Densitometric Analysis--
Densitometry of ECL immunoblots was
performed using a solid state video camera (Sony 
77, Sony Corp.) and
a 28-mm MicroNikkor lens over a light box of variable intensity
(Northern Light Precision 890, Imaging Research Inc., Toronto, Canada).
Quantification was performed using a Macintosh II-based image analysis
program (Image 1.49, developed by W. S. Rasband, Research Services
Branch, NIMH, Bethesda). In graphically displaying the fractional
GH-induced loss of anti-GHRext precipitation of GHRs (Fig.
7D, comparing WT rGHR and rGHR C241S), precipitability of
GHRs for each condition was estimated by normalizing the densitometric
signal for the GHR abundance in anti-GHRext precipitates
(detected by anti-GHRcyt blotting) by the densitometric
signal for GHR abundance in the unprecipitated detergent extract
resolved under reducing conditions (also detected by
anti-GHRcyt blotting). This ratio for unstimulated samples
was considered equal to 100% for each receptor (WT and CS mutant) in
each of two separate experiments, as indicated.
Specific JAK2 tyrosine phosphorylation in COS-7 cells transiently
expressing both GHR (WT rGHR or rGHR C241S) and JAK2 (Fig. 8B,
lower panel) was estimated for each condition by normalizing the
densitometric signal for tyrosine phosphorylation (APT immunoblot) of
immunoprecipitated JAK2 by the abundance of JAK2
(anti-JAK2AL33 immunoblot) in that precipitate. This ratio
for unstimulated samples was considered equal to one for each receptor
(WT and CS mutant). The fold increase in this ratio in response to GH
is graphically displayed. Specific GHR tyrosine phosphorylation in the
same COS-7 cell transfectants (Fig. 8C, upper panel) was
estimated in the same fashion, except that precipitations were
performed with anti-GHRcyt-AL37, and anti-GHR
immunoblotting of the precipitated GHRs was accomplished with
anti-GHRcyt. Graphical assessment of coimmunoprecipitation of JAK2 with anti-GHRcyt-AL37 for each condition in the
same COS-7 cell transfectants (Fig. 8C, lower panel) was
accomplished by normalizing the densitometrically determined abundance
of JAK2 (anti-JAK2AL33 immunoblot) in
anti-GHRcyt-AL37 precipitates by both the densitometrically
determined abundance of GHR (anti-GHRcyt immunoblot) in the
same precipitates and the total JAK2 abundance (anti-JAK2AL33 immunoblot of anti-JAK2AL33
precipitates). This ratio for unstimulated samples was considered equal
to one for each receptor (WT and CS mutant), and the fold increase in
this ratio in response to GH is graphically displayed.
| |
RESULTS |
GH-induced Loss of Anti-GHRext Immunoprecipitability
Correlates to GH-induced GHR Dimerization--
Initial experiments
focused on more detailed characterization of our recently described
(20, 22) mouse monoclonal antibody against the rabbit GHR extracellular
domain, anti-GHRext. Although it is only minimally capable
of immunoblotting (data not shown), anti-GHRext, as
previously reported (22), can specifically recognize the human (h) GHR
by immunoprecipitation. We noticed during this early characterization,
however, that GH treatment of cells rendered the GHR less
immunoprecipitable by anti-GHRext (Fig.
1). IM-9 lymphoblasts were treated with
GH or vehicle for 10 min at 37 °C prior to detergent lysis. Extracts
were subjected to immunoprecipitation with anti-GHRcyt-AL37
(a rabbit polyclonal serum directed at the GHR cytoplasmic domain
(18)), nonimmune rabbit serum (a negative control),
anti-GHRcyt-mAb (a mouse monoclonal antibody directed at
the GHR cytoplasmic domain), or anti-GHRext. Eluates of
these precipitates (upper panel) as well as nonprecipitated
detergent extracts (middle panel) were resolved under
reducing conditions by SDS-PAGE, and GHRs present in each sample were
immunoblotted with anti-GHRcyt (an independently derived
rabbit polyclonal serum directed at the receptor cytoplasmic domain
(11)). Whereas GH treatment did not affect the abundance of GHRs
detectable in either the extracts (middle panel, lanes 1-8)
or the specific anti-GHRcyt-AL37 or
anti-GHRcyt-mAb precipitates (upper panel, lanes
3-6), notably less GHR was detected in the
anti-GHRext precipitate from GH-treated cells than in the
precipitate from cells not exposed to GH (upper panel, lane
8 versus lane 7).
View larger version (56K):
[in this window]
[in a new window]
|
Fig. 1.
GH-induced loss of anti-GHRext
precipitability and disulfide linkage of GHRs in IM-9 cells. FLB
extracts from IM-9 cells treated with hGH (+) for 10 min or exposed to
vehicle only () were immunoprecipitated (IP) (20 million
cells per condition) with nonimmune (NI) (lanes 1 and 2), anti-GHRcyt-AL37 (lanes 3 and
4), anti-GHRcyt-mAb (lanes 5 and
6), or anti-GHRext (lanes 7 and
8) antibodies, and eluates were resolved by SDS-PAGE under
reducing conditions (upper panel). Aliquots (5 million cells
per condition) of the same extracts were resolved without
immunoprecipitation under reducing (middle panel) or
nonreducing (lower panel) conditions. Resolved proteins were
immunoblotted (WB) in each case with anti-GHRcyt
serum. The positions of migration of the GHR and the disulfide-linked
GHR (dsl GHR) are indicated by brackets.
Positions of prestained molecular mass markers, in kDa, are indicated
on the left. Lanes 1-6 and 7 and
8 were from two separate experiments. The experiments shown
are representative of five such experiments.
|
|
In principle, this GH-induced loss of anti-GHRext
precipitability could be accounted for by any of three potential
possibilities as follows: 1) GH induces a loss of GHRs from the
detergent-soluble extract such that less receptors are available for
immunoprecipitation; 2) GH binding directly interferes with the
epitope(s) in the extracellular domain of the GHR that are recognized
by anti-GHRext; 3) GH, by promoting dimerization of GHRs,
promotes interference of the interaction of anti-GHRext
with dimerization-sensitive epitope(s) in the GHR extracellular domain
and thus results in less immunoprecipitation of GHRs. In our previous
studies, we have observed GH-induced loss of detergent-soluble GHRs and
corresponding accumulation of detergent-insoluble GHRs in IM-9 cells
(12). However, the GH-induced loss of anti-GHRext
immunoprecipitability that we now observe is not likely to be explained
by such a change in detergent solubility since the abundance of GHRs
detectable in both unprecipitated detergent extracts and
anti-GHRcyt-AL37 precipitates does not vary with the 10-min
GH treatment at 37 °C. This is consistent with our previous finding
that significant loss of detergent-soluble GHRs was difficult to
observe when GH stimulation was for less than 30 min at 37 °C (12).
In addition, in other experiments (not shown), GH-induced loss of
anti-GHRext immunoprecipitability was observed even when
cells were treated with GH at 4 °C; no GH-induced loss of GHR
detergent solubility was detectable in our previous studies under these
conditions (12).
To address the other two possibilities for the GH-induced loss of
anti-GHRext precipitability of the GHR (GH blockade of the epitope(s) versus GHR dimerization-sensitive blockade of the
epitope(s)), we used an hGH analog, G120K, with a mutation at residue
120 (glycine to lysine); this mutation results in elimination of its
binding to the GHR at site 2 but preserves site 1 interaction. G120K is thus able to bind to the GHR but is unable to dimerize GHRs; similarly, this type of point mutated GH functions as a GH antagonist in that by
competition for site 1 binding it inhibits the ability of normal GH to
promote GHR dimerization (10, 27). As seen in Fig.
2A, treatment of IM-9 cells
with G120K alone, unlike treatment with GH, had no effect on the
ability of anti-GHRext to immunoprecipitate the GHR
(upper panel, lanes 1-3). Furthermore, the GH-induced loss
of anti-GHRext precipitability was reversed when G120K was added with GH at a ratio of G120K:GH equal to 2:1 (Fig. 2B, upper panel, lanes 1-4). The antagonism of G120K to the ability of GH to inhibit anti-GHRext precipitability indicates that the
G120K was, as expected, able to bind to the GHR in these cells. This G120K antagonism, coupled with the inability of G120K to alone inhibit
anti-GHRext precipitability, indicates that the loss of anti-GHRext precipitability in response to GH is correlated
with hormone-induced GHR dimerization, rather than with GH binding to
the GHR. Furthermore, these results suggest that GH-induced loss of
anti-GHRext immunoprecipitability can be used to monitor GH-induced GHR dimerization.
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 2.
The GH antagonist, G120K, causes neither loss
of anti-GHRext precipitability nor GHR disulfide linkage
and antagonizes GH with regard to both phenomena in IM-9 cells.
A, IM-9 cells were treated with vehicle (), GH or G120K
(500 ng/ml for each), as indicated for 10 min. FLB extracts were either
immunoprecipitated (IP) with anti-GHRext and
eluates were resolved by SDS-PAGE under reducing conditions
(upper panel) or extracts were resolved under reducing
(middle panel) or nonreducing (lower panel)
conditions, as in Fig. 1. GHRs were detected by anti-GHRcyt
immunoblotting (WB). Note that G120K causes neither loss of
anti-GHRext precipitability nor GHR disulfide linkage.
B, IM-9 cells were treated with vehicle (0 GH and 0 G120K),
GH, or GH and G120K, at the indicated concentrations for 10 min. FLB
extracts were either immunoprecipitated with anti-GHRext
and eluates were resolved by SDS-PAGE under reducing conditions
(upper panel) or extracts were resolved under reducing
(middle panel) or nonreducing (lower panel)
conditions, as in A. GHRs were detected by
anti-GHRcyt immunoblotting. Note that G120K antagonizes
both GH-induced loss of anti-GHRext precipitability and GHR
disulfide linkage. The experiments shown in both A and
B are representative of three such experiments.
|
|
GHR Dimerization Is Correlated with GHR Disulfide Linkage--
We
have previously described GH-induced disulfide linkage of GHRs that
occurs even at 4 °C and have noted that roughly one-half of GHRs
undergoing tyrosine phosphorylation rapidly in response to GH at
37 °C exist in this disulfide-linked form (11, 12). Furthermore,
GHRs that accumulate in the detergent-insoluble fraction of IM-9 cells
in response to GH are enriched in this disulfide-linked form (12). This
GH-induced formation of the disulfide-linked GHR was again demonstrated
in the present study by resolving detergent extracts of untreated and
GH-treated cells by SDS-PAGE under nonreducing conditions and detecting
monomeric and disulfide-linked GHRs by immunoblotting with
anti-GHRcyt (Fig. 1, lower panel, lanes 1-8). By using the information derived from the experiments described above
in the upper panels of Figs. 1 and 2, we tested whether GH-induced disulfide linkage such as that shown in the lower
panel of Fig. 1 might reflect GH-induced GHR dimerization.
As a first step in rigorously probing the relationship between
GH-induced GHR disulfide linkage and GHR dimerization, we tested the
effect of the non-dimerizing G120K antagonist on receptor disulfide
linkage. As seen in Fig. 2A (lower panel),
treatment of IM-9 cells with G120K did not promote appearance of the
disulfide-linked GHR (lane 3 versus lane
2). Furthermore, in a pattern that mirrored its ability to inhibit
GH-induced loss of anti-GHRext immunoprecipitability, G120K
significantly antagonized GH-induced GHR disulfide linkage with nearly
complete inhibition at a G120K:GH ratio of 2:1 (Fig. 2B, lower
panel, lanes 2-4). Thus, the capacity of GH to dimerize GHRs
appears to be required to cause GHR disulfide linkage.
This point was further verified by comparing the GH concentration
dependence of the two phenomena (loss of anti-GHRext
precipitability and GHR disulfide linkage). IM-9 cells were either left
unstimulated or treated with GH (1-500 ng/ml in Fig.
3A; 500-8000 ng/ml in Fig.
3B) for 10 min at 37 °C. Detergent extracts were either
immunoprecipitated with anti-GHRext and resolved under
reducing conditions (upper panels) or directly resolved
without immunoprecipitation under reducing conditions (middle
panels) or nonreducing conditions (lower panels) by
SDS-PAGE. GHRs in each instance were detected by immunoblotting with
anti-GHRcyt (Fig. 3A) or
anti-GHRcyt-AL37 (Fig. 3B). Notably, the GH
concentration dependence of the loss of anti-GHRext
immunoprecipitability (which reflects GHR dimerization, as shown above
in Figs. 1 and 2) and the appearance of the disulfide-linked GHR were
quite similar in both the 1-500 and 500-8000 ng/ml ranges. In
particular, both the GH-induced loss of anti-GHRext
precipitability and the GH-induced accumulation of disulfide-linked
GHRs were attenuated at GH concentrations of 2000-8000 ng/ml compared
with 500 ng/ml (Fig. 3B, lanes 3-5
versus lane 2). Such a suppression of these two
GH-induced changes at very high GH concentrations (which is concert
with our previous findings for the disulfide linkage (11)) is
consistent with dimerization dependence of both (8). Together, the
results of the experiments shown in Figs. 2 (lower panels)
and 3 clearly indicate the correlation between GH-induced GHR
dimerization and accumulation of the disulfide-linked form of the
receptor.
View larger version (38K):
[in this window]
[in a new window]
|
Fig. 3.
GH concentration dependence for loss of
anti-GHRext precipitability and GHR disulfide linkage in
IM-9 cells. A and B, IM-9 cells were treated
with vehicle (0) or GH at the indicated concentrations
(A, low dose; B, high dose) for 10 min. FLB
extracts were either immunoprecipitated (IP) with
anti-GHRext and eluates were resolved by SDS-PAGE under
reducing conditions (upper panel) or extracts were resolved
under reducing (middle panel) or nonreducing (lower
panel) conditions, as in Figs. 1 and 2. GHRs were detected by
anti-GHRcyt immunoblotting (WB). Note the
parallel concentration dependences for GH-induced loss of
anti-GHRext precipitability and GHR disulfide linkage
(A) and the reversal of both phenomena at GH concentrations
of 2000 ng/ml and greater (B). The experiments shown in
both A and B are representative of two such
experiments.
|
|
While GHR dimerization is believed to be a critical step in GH
signaling (10), detailed understanding of the relationship between the
dimerization state of the receptor and its capacity to interact with
key intracellular signaling molecules does not yet exist, nor do we yet
know the structural basis for the impact of GHR dimerization on GH
signaling. In part, the difficulty in fully understanding these issues
is in biochemically assessing GHR dimerization in cells under the usual
conditions employed in protein extraction, immunoprecipitation, and
SDS-PAGE. Our observations that both GHR disulfide linkage and loss of
anti-GHRext precipitability reflect the dimerization state
of the receptor are of potential utility in that assessments of each
process might serve as relatively simple biochemical proxies for GHR dimerization.
GH-induced Enhancement of Association of JAK2 with the GHR Is
Related to the Dimerization State of the Receptor--
One interaction
critical for effective GH signaling, but incompletely understood, is
between the GHR and the cytoplasmic tyrosine kinase, JAK2 (7, 13-15).
The bacterially expressed cytoplasmic domain of the GHR can interact
in vitro via its Box 1 region with JAK2 derived from cells
not treated with GH prior to extraction (15). Similarly, basal
(GH-independent) association of JAK2 with the GHR can be detected in
extracts of cells that express both molecules (15); however, our work
and that of others (7, 15, 20)2 indicate that treatment of
cells with GH enhances the association of the GHR and JAK2, as assessed
by coimmunoprecipitation of JAK2 with various anti-GHR antibodies. This
effect is pronounced in the highly GH-responsive murine 3T3-F442A
fibroblast cell line (7, 20). We explored the mechanism of this
GH-dependent enhancement of GHR-JAK2 association in this
cell type.
We first tested the effect of G120K versus GH on GHR
dimerization status and GHR-JAK2 association (Fig.
4). In preliminary experiments (not
shown), we determined that anti-GHRext did not detect the
murine GHR by either immunoprecipitation or immunoblotting; thus, we
could not use this antibody to probe the dimerization state of the
endogenous GHR in mouse cells. However, as with IM-9 cells, treatment
of 3T3-F442A cells with GH did promote appearance of the
disulfide-linked form of the GHR (Fig. 4A, lane 2 versus 1), confirming that this change in the GHR can be detected
for murine as well as human receptors (as shown below, GH-induced disulfide linkage occurs with the rabbit GHR as well). Also similar to
our findings in IM-9 cells, treatment with G120K alone could not
support GHR disulfide linkage; however, when added with GH at a
G120K:GH ratio of 5:1, G120K inhibited GH-induced GHR disulfide linkage
(Fig. 4A, lanes 3 and 4 versus
2). In other experiments, a G120K:GH ratio of 3:1 was
sufficient to inhibit GH-induced GHR disulfide linkage (not shown).
Thus, in 3T3-F442A cells, disulfide linkage (a structural correlate of
receptor dimerization) displays similar sensitivity to the GH
antagonist as it does in IM-9 cells.
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 4.
GH-induced disulfide linkage of GHRs and
GHR-JAK2 association are related to GHR dimerization in 3T3-F442A
cells. A, G120K does not cause GHR disulfide linkage
and antagonizes GH-induced GHR disulfide linkage in 3T3-F442A cells.
Cells were treated with vehicle (0 GH and 0 G120K), GH, or GH and
G120K, at the indicated concentrations for 15 min. FLB extracts
(one-fourth 150 × 25-mm dish equivalent per condition) were
resolved under nonreducing conditions. GHRs were detected by
anti-GHRcyt immunoblotting (WB). The positions
of the monomeric and disulfide-linked (dsl) murine GHR are
bracketed. B and C, GH-induced
coimmunoprecipitation of JAK2 and tyrosine phosphorylated JAK2 with GHR
is not caused by G120K but is antagonized by G120K. Cells were treated
with vehicle (0 GH and 0 G120K), GH, or GH and G120K, at the indicated
concentrations for 15 min. FLB extracts (one 150 × 25-mm dish
equivalent per condition) immunoprecipitated with
anti-GHRcyt-AL37 were resolved by SDS-PAGE under reducing
conditions, and sequentially immunoblotted with anti-GHRcyt
(B, upper panel), anti-JAK2AL33 (B, lower
panel), and antiphosphotyrosine (APT) antibody
(C, lower panel). Extracts from the same cells (one-half
150 × 25-mm dish equivalent per condition) were
immunoprecipitated with anti-JAK2AL33, resolved by SDS-PAGE
under reducing conditions, and sequentially immunoblotted with
anti-JAK2AL33 (C, upper panel) and
antiphosphotyrosine (C, middle panel). The positions of the
GHR, JAK2, tyrosine-phosphorylated JAK2 (P-JAK2), and a
nonspecifically detected protein (NS) are indicated. (As
described in the legend for Fig. 5, sodium orthovanadate was not
in- cluded in the cell wash and harvesting buffer in these
experiments.) The experiments shown in A C are
representative of four such experiments.
|
|
We next examined the coimmunoprecipitation of JAK2 with the GHR in
3T3-F442A cells. For this purpose, we employed
anti-GHRcyt-AL37 as the immunoprecipitating antibody. As
seen in Fig. 4B, when assessed by immunoblotting with
anti-GHRcyt, the 3T3-F442A GHR was specifically
precipitated by anti-GHRcyt-AL37 to a similar degree from
detergent extracts of unstimulated cells and cells treated with GH,
G120K, or the combination of GH and G120K (lanes 1-4, upper
panel). Despite this similarity in the abundance of directly
precipitated GHR, GH treatment, as anticipated, promoted a marked
enhancement in JAK2 coimmunoprecipitated with
anti-GHRcyt-AL37, as assessed by reprobing of the blot with
anti-JAK2AL33 (Fig. 4B, lower panel, lane 2 versus 1). Notably, G120K alone promoted no such
augmented association of JAK2 with the GHR (lane 3) but effectively competed away GH-induced coprecipitation of JAK2 with the
receptor (lane 4).
These differences in coimmunoprecipitation of JAK2 with the GHR were
not accounted for by differences in the abundance of JAK2 present in
each extract, as confirmed by immunoprecipitation of the same extracts
with anti-JAK2AL33 followed by immunoblotting of JAK2 (Fig.
4C, upper panel, lanes 1-4). Similar to these findings regarding GHR disulfide linkage and association of JAK2 with the GHR,
G120K also failed to promote tyrosine phosphorylation of JAK2 such as
that detected in response to GH in both anti-JAK2AL33 direct precipitates and anti-GHRcyt-AL37 coprecipitates.
However, G120K did substantially antagonize GH-induced JAK2 tyrosine
phosphorylation (Fig. 4C, middle and lower panels,
lanes 1-4).
The results of the experiments in Fig. 4 indicate that in 3T3-F442A
cells GH-induced GHR dimerization (conveniently monitored by tracking
GHR disulfide linkage) is required for both JAK2 tyrosine phosphorylation and JAK2's association with the receptor. These experiments did not, however, address the question of whether tyrosine
phosphorylation of the GHR and/or JAK2 is required to support the
GH-enhanced coimmunoprecipitation of JAK2 with the GHR. In other words,
is it the GHR dimerization that fosters enhanced GHR-JAK2 association
or is it the JAK2 activation and tyrosine phosphorylation of the GHR
and JAK2 that result from GHR dimerization that allow enhanced GHR-JAK2
association? We pursued this issue in two ways.
First, we tested whether the degree to which preservation of the
GH-induced tyrosine phosphorylation of JAK2 and the GHR during the
processing and extraction of the cells affected the abundance of JAK2
coimmunoprecipitated with the GHR. In preliminary experiments (not
shown), we observed that GH-induced tyrosine phosphorylation of the
GHR, JAK2, and other substrates was significantly more detectable after
detergent extraction, SDS-PAGE, and APT immunoblotting if the
GH-treated cells were washed prior to harvesting and lysis with
ice-cold PBS that contained sodium orthovanadate. Inclusion of
orthovanadate in the cell wash and harvesting buffer presumably better
"traps" proteins in their tyrosine-phosphorylated state by
inhibiting cellular protein tyrosine phosphatases that are either
always active or are adventitiously activated during harvesting and
collection of the cells.
In the experiment shown in Fig. 5,
3T3-F442A cells were treated in duplicate with (+) or without () GH
for 15 min and then harvested and collected in PBS that either
contained (+) or did not contain () sodium orthovanadate (0.4 mM). In all samples, cells were lysed in the standard lysis
buffer that contained sodium orthovanadate and under the usual
solubilization conditions (as under "Experimental Procedures").
Immunoprecipitation with anti-GHRcyt-AL37 followed by
SDS-PAGE and immunoblotting with anti-JAK2AL-33 revealed that inclusion of orthovanadate in the cell wash and harvesting buffer
had no effect on the degree to which GH treatment promoted coimmunoprecipitation of JAK2 with the GHR (upper panel, lane 4 versus lane 2). However, as anticipated,
stripping and reprobing of this blot with APT antibody revealed that
the degree to which the directly precipitated GHR and coprecipitated
JAK2 were detectably tyrosine-phosphorylated was markedly diminished
when orthovanadate was omitted from the buffer (lower panel, lane
4 versus lane 2). (We verified in other
experiments (not shown) that inclusion of orthovanadate in the wash and
harvesting buffer had no effect on the abundance of JAK2 or GHR present
in the cell extracts or in direct immunoprecipitates of each.) This
result indicates that differential degrees of preservation of
GH-induced tyrosine phosphorylation of GHR and JAK2 do not affect the
stability of the GH-induced enhanced interaction of the GHR and JAK2,
as assessed by coimmunoprecipitation.
View larger version (42K):
[in this window]
[in a new window]
|
Fig. 5.
GH-induced coimmunoprecipitation of JAK2 with
GHR is not affected by tyrosine dephosphorylation during cell washing
and harvesting after stimulation. 3T3-F442A cells were treated
with vehicle () or GH (+), as indicated, for 15 min. After
stimulation, cells were washed with and harvested in ice-cold PBS
containing (+ Na3VO4) or not
containing ( Na3VO4)
0.4 mM sodium orthovanadate, as under "Experimental
Procedures." FLB extracts (one 150 × 25-mm dish equivalent per
condition) (all containing 1 mM sodium orthovanadate) were
immunoprecipitated with anti-GHRcyt-AL37, were resolved by
SDS-PAGE under reducing conditions, and were sequentially immunoblotted
with anti-JAK2AL33 (upper panel) and
antiphosphotyrosine (APT) antibody (lower panel).
The positions of the JAK2 and tyrosine-phosphorylated JAK2
(P-JAK2) and GHR (P-GHR) are indicated. Note that
despite markedly reduced tyrosine phosphorylation level of both
directly precipitated and coimmunoprecipitated JAK2 in the samples in
which orthovanadate was omitted during washing and harvesting, the
abundance of coimmunoprecipitated JAK2 was unchanged (as was the
abundance of directly precipitated GHR (data not shown)).
|
|
We further addressed the issue of whether GHR dimerization
versus the consequent tyrosine phosphorylations were
responsible for enhanced GHR-JAK2 association by attempting to uncouple
the two phenomena. We have previously observed in IM-9 cells that GH-induced GHR disulfide linkage was not affected by pretreatment of
the cells with staurosporine but that GH-induced tyrosine
phosphorylation of cellular proteins was markedly blunted under the
same conditions (11). In the experiment shown in Fig.
6, 3T3-F442A cells were treated with GH
either in the presence or absence of staurosporine (1.25 µM). Anti-GHRcyt immunoblotting of detergent
extracts subjected to SDS-PAGE under nonreducing (Fig. 6A, upper
panel) or reducing (Fig. 6A, lower panel) conditions
confirmed that, as in IM-9 cells, staurosporine pretreatment did not
affect either the degree of GH-induced GHR disulfide linkage or the GHR
abundance in these cells (lane 3 versus
lane 2 in each panel). Staurosporine pretreatment, however,
significantly blunted the degree of GH-induced tyrosine phosphorylation
of JAK2, as assessed by anti-JAK2AL33 immunoprecipitation and APT immunoblotting (Fig. 6B, upper panel, lane
3 versus lane 2) without affecting the
abundance of the JAK2 that was immunoprecipitated (Fig. 6B, lower
panel, lane 3 versus lane 2). Similarly,
when the same extracts were immunoprecipitated with
anti-GHRcyt-AL37, the level of GH-induced tyrosine
phosphorylation of both directly precipitated GHR and
coimmunoprecipitated JAK2 was also markedly diminished by staurosporine
pretreatment (Fig. 6C, upper panel, lane 3 versus lane 2). Notably, despite the diminished
GH-induced JAK2 tyrosine phosphorylation, staurosporine pretreatment
did not affect the abundance of JAK2 coimmunoprecipitated with the GHR
in response to GH (Fig. 6C, lower panel, lane 3 versus lane 2). This result complements the
findings displayed in Fig. 5; collectively, they indicate that the
GH-enhanced association of JAK2 with the GHR is related more to GHR
dimerization, per se, than to consequent JAK2 and/or GHR
tyrosine phosphorylation.
View larger version (32K):
[in this window]
[in a new window]
|
Fig. 6.
Significant inhibition of GH-induced tyrosine
phosphorylation of GHR and JAK2 prevents neither GH-induced GHR
disulfide linkage nor coimmunoprecipitation of JAK2 with GHR in
3T3-F442A cells. A, staurosporine pretreatment does not
prevent GH-induced GHR disulfide linkage. Cells were pretreated
Me2SO vehicle (0 staurosporine) or staurosporine (1.25 µM) for 15 min prior to treatment with vehicle (0 GH) or
GH (500 ng/ml) for 15 min, as indicated. FLB extracts (one-fourth
150 × 25-mm dish equivalent per condition) were resolved by
SDS-PAGE under nonreducing (upper panel) or reducing
(lower panel) conditions. GHRs were detected by
anti-GHRcyt immunoblotting (WB). The positions
of the monomeric and disulfide-linked (dsl) murine GHR are
bracketed. B and C, GH-induced
coimmunoprecipitation of JAK2 and tyrosine-phosphorylated JAK2 with GHR
is not prevented by pretreatment with staurosporine, despite marked
inhibition of GH-induced tyrosine phosphorylation of both GHR and JAK2.
Cells were treated as in A. FLB extracts immunoprecipitated
(IP) with anti-JAK2AL33 (B, one-half
150 × 25-mm dish equivalent per condition) or anti-GHRcyt-AL37 (C, one
150 × 25-mm dish equivalent per condition) were resolved by
SDS-PAGE under reducing conditions, and sequentially immunoblotted
(WB) with antiphosphotyrosine (APT) antibody
(B and C, upper panels) and
anti-JAK2AL33 (B and C, lower
panels). The positions of JAK2, tyrosine-phosphorylated JAK2
(P-JAK2), tyrosine-phosphorylated GHR (P-GHR),
and a nonspecifically detected protein (NS) are indicated.
(Sodium orthovanadate (0.4 mM) was included in the cell
wash and harvesting buffer in these experiments.) The experiments shown
in AC are representative of four such experiments.
|
|
GHR Disulfide Linkage Involves Cysteine 241 and Is Not Required for
GHR Dimerization--
The above results indicated that GH-induced GHR
disulfide linkage correlates with GHR dimerization and that receptor
dimerization allows enhanced association of the GHR and JAK2 and is
required for effective signaling. It therefore became relevant to ask
whether disulfide linkage is required for GHR dimerization and
consequent GHR-JAK2 association and signaling. To address this issue,
we sought to abrogate GH-induced disulfide linkage by mutation of relevant receptor cysteine residue(s) that might be involved in such a
linkage. It is known from both mutagenic and crystallographic studies
that there are six cysteine residues conserved in the GHR extracellular
domains of various species that are engaged in three pairs of
intrachain disulfide linkages (9); these linkages are critical to the
structural integrity of receptor independent of its occupancy by GH and
would thus not be predicted to be candidates for participating in
GH-induced GHR disulfide linkage. The remaining cysteine residues in
the rGHR are indicated in the diagram of the receptor in Fig.
7A. Without any prior
information, we presumed each of these remaining cysteines (residue 241 in the extracellular domain and residues 371, 387, 407, 496, 506, 518, 525, and 607 in the cytoplasmic domain) could be unpaired and thus a
candidate for being involved in the GH-induced disulfide linkage (all
but cysteine 387 are conserved in the human GHR).
View larger version (51K):
[in this window]
[in a new window]
|
Fig. 7.
Cysteine 241 is critical for GH-induced
disulfide linkage, but not GH-induced dimerization, of the rabbit
GHR. A, diagram of the structure of WT rGHR and rGHR
mutants used in experiments in Figs. 7 and 8 and in data not shown.
rGHR mutants (rGHR C241S, rGHRd297-406,
rGHRd294-498, and rGHR1-390) are described
under "Experimental Procedures" and "Results." The positions of
the extracellular, transmembrane (white space), and
intracellular domains are indicated, as are the unpaired cysteine
residues referred to in the text. Results of experiments in
B and in data not shown are summarized with + or to
indicate whether GH can or cannot, respectively, promote disulfide
linkage of the rGHR form shown. B, WT rGHR, but not rGHR
C241S, transiently expressed in COS-7 cells, undergoes disulfide
linkage in response to GH. Transient transfections of COS-7 cells with
expression vectors encoding WT rGHR or rGHR C241S (or empty vector as a
negative control) were accomplished as under "Experimental
Procedures." Cells were treated without () or with (+) GH (500 ng/ml) for 10 min. FLB extracts (one-fourth 150 × 25-mm dish
equivalent per condition) were resolved by SDS-PAGE under reducing
(upper panel) or nonreducing (lower panel)
conditions. GHRs were detected by anti-GHRcyt or
anti-GHRcyt-AL37 immunoblotting (WB), as
indicated. The positions of the monomeric and disulfide-linked
(dsl) rabbit GHR are bracketed. The large band
(bracketed and marked NS) migrating just below
the monomeric GHR in the lower panel is a mixture of the
underglycosylated GHR and a comigrating nonspecific band uniquely
recognized by anti-GHRcyt-AL37 immunoblotting (Y. Zhang, J. Jiang, and S. J. Frank, unpublished observations). C,
rGHR C241S, although incapable of undergoing GH-induced disulfide
linkage, is, like WT rGHR, rendered markedly less precipitable by
anti-GHRext in response to GH. COS-7 cells were transfected
and treated as in B. FLB extracts (one 150 × 25-mm
dish equivalent per condition) were immunoprecipitated (IP)
with anti-GHRext, and eluates were resolved by SDS-PAGE
under reducing conditions. GHRs were detected by
anti-GHRcyt immunoblotting. D,
anti-GHRext precipitability of WT rGHR and rGHR C241S was
assessed densitometrically, as detailed under "Experimental
Procedures," from two separate experiments such as that shown in
C. Average GH-induced loss of anti-GHRext
precipitation for each receptor form is graphically displayed
(error bars are range about the mean in two independent
experiments).
|
|
Because of the relatively reducing intracellular environment, we
thought it unlikely that such a linkage would be favored to occur via
the cytoplasmic domain. Nevertheless, in preliminary experiments, we
evaluated the ability of several of our previously characterized rGHR
mutants (see Fig. 7A) to undergo GH-induced GHR disulfide
linkage when transiently expressed in COS-7 cells. These experiments
indicated that rGHR mutants harboring either internal deletions of
residues 297-406 (eliminating cysteines 371 and 387) or 294-498
(eliminating cysteines 371, 387, 407, and 496) or carboxyl-terminal
truncation of residues 391-620 (eliminating cysteines 407, 496, 506, 518, 525, and 607) were each capable of undergoing GH-induced GHR
disulfide linkage (data not shown). Thus, none of the cysteine residues
present in the rGHR cytoplasmic domain could be experimentally
implicated as involved in the disulfide linkage. We therefore mutated
the only remaining unpaired extracellular cysteine residue (cysteine
241) to serine and compared the capacity of the resulting mutant,
designated rGHR C241S (diagrammed in Fig. 7A), and the
wild-type (WT) rGHR to undergo GH-induced disulfide linkage.
Expression vectors with the cDNAs encoding WT rGHR and rGHR C241S
were transiently expressed in COS-7 cells. As assessed by anti-GHRcyt immunoblotting of reduced cell extracts (Fig.
7B, upper panel) and by 125I-hGH binding studies
(not shown), both WT and C241S receptors were similarly detectable and
capable of binding to GH. To test for disulfide linkage, transfected
cells were treated without () or with (+) GH for 10 min, and
detergent extracts of the cells were subjected to SDS-PAGE under
nonreducing conditions and immunoblotted with
anti-GHRcyt-AL37 (Fig. 7B, lower panel).
GH-induced appearance of the disulfide-linked receptor form was
detected for the WT rGHR (lane 3 versus
lane 2) but not for rGHR C241S (lane 5 versus lane 4).
We also tested whether the C241S mutant, despite its inability to
undergo GH-induced disulfide linkage, exhibited evidence of GH-induced
dimerization. For this, we employed anti-GHRext for
immunoprecipitation of aliquots of the same extracts analyzed in Fig.
7B. For both WT rGHR and rGHR C241S, there was a substantial loss of receptors precipitated by anti-GHRext after
treatment of the cells with GH (Fig. 7C, upper panel, lanes
2-5). When the results of two such experiments were subjected to
densitometric analysis and normalized for the GHR expression level in
each extract, GHRs precipitated by anti-GHRext fell in
abundance on average by 86% for WT rGHR and 48% for rGHR C241S in
response to GH (Fig. 7D). This substantial GH-induced loss
of anti-GHRext precipitability of both WT rGHR and rGHR
C241S strongly suggests that each underwent GH-induced receptor
dimerization, although dimer formation and/or stability were likely
more significant for the WT rGHR. Mutation of cysteine 241, therefore,
abrogates GH-induced GHR disulfide linkage but not GH-induced receptor dimerization.
We further examined the effect of mutation of cysteine 241 on
GH-induced signaling and enhancement of GHR-JAK2 association in both
transiently and stably transfected cells. Our previous studies
indicated that transient cotransfection of COS-7 cells with JAK2 and
the GHR allowed detection of GH-dependent JAK2 tyrosine phosphorylation (15, 17). In the experiment shown in Fig. 8, cells were transiently transfected
with expression vectors encoding murine JAK2 and either WT rGHR
(lanes 2, 3, and 6) or rGHR C241S (lanes
4 and 5). As a negative control, empty vector alone was
transfected (lane 1). Transfected cells were treated without
() or with (+) GH for 10 min. Anti-GHRcyt immunoblotting of detergent extracts of each sample resolved under reducing (Fig. 8A, upper panel) and nonreducing (Fig. 8A, lower
panel) conditions again verified similarity of expression of each
receptor and the inability of rGHR C241S to undergo GH-induced receptor
disulfide linkage. Notably, GH induced a subtle, but discernible,
retardation of migration of the monomeric WT and C241S rGHRs, both
under reducing and nonreducing conditions (compare lanes 3 and 6 versus lane 2 and lane
5 versus lane 4 in both upper and
lower panels). This shift is consistent with both WT rGHR
and rGHR C241S undergoing GH-induced receptor tyrosine phosphorylation
(more below).
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 8.
Tyrosine phosphorylation of GHR and JAK2 and
enhanced coimmunoprecipitation of JAK2 with GHR occur in response to GH
in cells transiently expressing either WT rGHR or rGHR C241S.
A, lack of GH-induced disulfide linkage of rGHR C241S when
cotransfected with JAK2. Transient cotransfections of COS-7 cells with
expression vectors encoding WT rGHR or rGHR C241S, in each case along
with an expression vector encoding murine JAK2 (or empty vector as a
negative control), as indicated, were accomplished as in Fig. 7 and
"Experimental Procedures." Cells were treated without () or with
(+) GH (500 ng/ml) for 10 min, as indicated. (Note that the
transfection and stimulation of the sample in lane 6 constitutes a duplicate of that in lane 3, so that it could
be used for nonspecific immunoprecipitation in B.) FLB
extracts (one-fourth 150 × 25-mm dish equivalent per condition)
were resolved by SDS-PAGE under reducing (upper panel) or
nonreducing (lower panel) conditions. GHRs were detected by
anti-GHRcyt immunoblotting (WB). (Sodium
orthovanadate (0.4 mM) was included in the cell wash and
harvesting buffer in these experiments.) The positions of the monomeric
and disulfide-linked (dsl) rabbit GHR are bracketed. Note the slight retardation
of migration of monomeric GHRs under both reducing and nonreducing
conditions in response to GH (lanes 3, 5, and 6),
consistent with GH-induced tyrosine phosphorylation of the GHR (see
B and C). B, GH-induced tyrosine
phosphorylation of JAK2. As described in A, the same pools
of transfected cells were treated without () or with (+) GH (500 ng/ml) for 10 min, as indicated. FLB extracts (one-half 150 × 25-mm dish equivalent per condition) were immunoprecipitated
(IP) with anti-JAK2AL33 (lanes 1-5)
or nonimmune (NI) serum (lane 6), resolved by
SDS-PAGE under reducing conditions, and sequentially immunoblotted with
anti-JAK2AL33 (upper panel) and
antiphosphotyrosine (APT) antibody (middle
panel). The positions of JAK2 and tyrosine-phosphorylated JAK2
(P-JAK2) are indicated. Specific JAK2 tyrosine
phosphorylation induced by GH was densitometrically determined from the
data in the upper and middle panels of this
experiment, as detailed under "Experimental Procedures." The
GH-induced fold increase in specific JAK2 tyrosine phosphorylation for
both WT rGHR and rGHR C241S is graphically displayed in the lower
panel. Note that GH-induced tyrosine phosphorylation of JAK2 is
not prevented by expression of rGHR C241S, although rGHR C241S does not
undergo disulfide linkage. C, GH-induced tyrosine
phosphorylation of WT rGHR and rGHR C241S and the GH-induced
enhancement of coimmunoprecipitation of JAK2 with each receptor. The
same extracts as those used in A and B were
subjected to anti-GHRcyt-AL37 immunoprecipitation and
eluates were resolved under reducing conditions and sequentially
immunoblotted with anti-GHRcyt, APT antibody, and
anti-JAK2AL33. Graphically displayed are the relative
specific GHR tyrosine phosphorylations induced by GH for WT rGHR and
rGHR C241S (upper panel) and the relative GH-induced
enhancements in the levels of JAK2 coimmunoprecipitated with GHR form
(lower panel) in this experiment, in each case estimated as
detailed under "Experimental Procedures." Note that neither
GH-induced tyrosine phosphorylation of GHR nor GH-induced enhancement
of GHR-JAK2 association is prevented by expression of rGHR C241S. The
experiments shown in AC are representative of two such
experiments.
|
|
These same extracts were immunoprecipitated with
anti-JAK2AL33, and as a negative control, the extract from
a duplicate pool of cells transfected with JAK2 and WT rGHR and
stimulated with GH was precipitated with nonimmune serum (lane
6). Anti-JAK2AL33 immunoblotting of these precipitates
(Fig. 8B, upper panel) showed that the transfected JAK2 was
similarly abundant in the transfected cells (lanes 2-5),
was specifically precipitated by anti-JAK2AL33 (lane
6 versus lanes 2-5), and was much more abundant than
the endogenous COS-7 JAK2 (lane 1 versus
lanes 2-5). Stripping and reprobing of this blot with APT
antibody revealed that both WT rGHR and rGHR C241S mediated GH-induced
tyrosine phosphorylation of JAK2 (Fig. 8B, middle panel, lane
3 versus lane 2 and lane 5 versus lane 4 and densitometrically analyzed and
graphically displayed in Fig. 8B, lower panel).
Anti-GHRcyt-AL37 immunoprecipitation was also performed on
extracts of these transiently transfected cells. Densitometric analysis
of APT and anti-GHRcyt immunoblots of the precipitated receptors verified the conclusions suggested by the retarded migration of GHRs described above in that similar degrees of relative GH-induced GHR tyrosine phosphorylation were seen for specifically precipitated WT
rGHR and rGHR C241S (Fig. 8C, upper panel). When
anti-JAK2AL33 immunoblotting of JAK2 coprecipitated with
anti-GHRcyt-AL37 was examined, GH-induced association of
JAK2 with the GHR was also detected for both receptors (Fig. 8C,
lower panel). Collectively, the results of the transient
transfection experiments shown in Figs. 7 and 8 lead us to conclude
that GH-induced GHR dimerization, rather than receptor disulfide
linkage, is required for GHR-JAK2 association and JAK2 and GHR tyrosine phosphorylation.
To verify and extend these findings in a system in which GHRs are
stably reconstituted, we used Chinese hamster ovary (CHO) cells, which
others have shown to express no endogenous GHRs and to be a useful
vehicle for GHR expression and studies of GH signaling (13, 14). CHO
cells were stably transfected with WT rGHR or rGHR C241S, as under
"Experimental Procedures"; two clones, designated CHO-rGHR
(expressing WT rGHR) and CHO-rGHR C241S (expressing rGHR C241S), were
selected for further evaluation based on their similar levels of
receptors detected by immunoblotting (as shown in Fig. 9A and in
anti-GHRcyt-AL37 immunoprecipitates in Fig. 9B,
upper panel). As seen in Fig. 9A, GH-induced GHR
disulfide linkage was easily detected in CHO-rGHR, but not in CHO-rGHR
C241S, by anti-GHRcyt-AL37 immunoblotting of detergent cell
extracts resolved by nonreducing SDS-PAGE (compare lane 2 versus lane 1 and lane 4 versus lane 3), in concert with the findings in
COS-7 cells. (This cysteine 241-dependence for GH-induced GHR disulfide
linkage was also seen in multiple other CHO-rGHR versus
CHO-rGHR C241S clones.)
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 9.
Acute GH-induced tyrosine phosphorylation of
STAT5 is similar in CHO-rGHR and CHO-rGHR C241S cells.
A, lack of GH-induced disulfide linkage of rGHR C241S.
CHO-rGHR and CHO-rGHR C241S cells were generated as described under
"Experimental Procedures." Cells were treated without () or with
(+) GH (500 ng/ml) for 1 min, as indicated. FLB extracts (one-half
100 × 20-mm dish equivalent per condition) were resolved by
SDS-PAGE under nonreducing conditions. GHRs were detected by
anti-GHRcyt-AL37 immunoblotting (WB). The
positions of the monomeric and disulfide-linked (dsl) rabbit
GHR are bracketed. The experiment shown is representative of
five such experiments. B, GH-induced enhancement of
coimmunoprecipitation of JAK2 with the GHR in CHO-rGHR and CHO-rGHR
C241S cells. The same extracts as those used in A were
subjected to anti-GHRcyt-AL37 immunoprecipitation
(IP). Eluates were resolved under reducing conditions and
sequentially immunoblotted with anti-GHRcyt-AL37
(upper panel) and anti-JAK2AL33 (lower
panel). The positions of the immunoprecipitated rGHR and JAK2 are
indicated. Note the slight retardation of migration of GHRs in response
to GH (lanes 2 and 4), indicative of the
GH-induced tyrosine phosphorylation of the GHR. C, GH time
and concentration dependence of tyrosine phosphorylation of STAT5 in
CHO-rGHR and CHO-rGHR C241S cells. Cells were treated with vehicle (0)
or GH (500 ng/ml) for the indicated duration (upper panel)
or for 10 min at the indicated GH concentration (lower
panel). Total cell extracts were resolved by SDS-PAGE under
reducing conditions and immunoblotted with anti-phospho-STAT5. The
position of the tyrosine-phosphorylated STAT5 (P-STAT5) is
indicated.
|
|
Also as observed in transient reconstitution experiments, GH promoted
enhanced coimmunoprecipitation of JAK2 with
anti-GHRcyt-AL37 in both CHO-rGHR and CHO-rGHR C241S cells
(Fig. 9B, lower panel), and GHR and JAK2 both became
tyrosine-phosphorylated acutely in response to GH in each cell line
(data not shown). To monitor further the ability of GH to acutely
activate signaling via the WT and C241S mutant receptors, we examined
the tyrosine phosphorylation of STAT5 in response to GH in CHO-rGHR and
CHO-rGHR C241S (Fig. 9C). Since levels of STAT5 did not vary
between the two cell lines (data not shown), we used a state-specific
antibody, anti-phospho-STAT5, which detects only the
tyrosine-phosphorylated form of STAT5, to compare by immunoblotting
of total cell extracts the time course (Fig. 9C, upper
panel) and GH concentration dependence (Fig. 9C, lower
panel) of STAT5 activation in each cell. The time course of
GH-induced STAT5 tyrosine phosphorylation was indistinguishable between
the two cells with detection as early as 1 min into GH stimulation in
both CHO-rGHR and CHO-rGHR C241S. Likewise, the concentration
dependence for STAT5 tyrosine phosphorylation after 10 min exposure to
GH was the same for both cells with detectable signal evident in
response to as little as 10 ng/ml GH and maximal with 100-250 ng/ml
GH. Thus, using both transient and stable transfection, no substantial
difference was observed in the ability of GH to dimerize acutely the
GHRs or to initiate proximal aspects of GHR signaling in cells
expressing the wild-type versus C241S mutant GHR.
| |
DISCUSSION |
It is generally appreciated that GH-induced dimerization of the
GHR is important for GH signaling. Many molecular details of the GH-GHR
interaction are understood structurally, but less is known about the
dimerization process as it occurs in cells and the effects of receptor
dimerization on the association of GHR with intracellular signaling
molecules. These issues are intrinsically difficult to study
biochemically in part because the standard immunological and
electrophoretic techniques useful in studies of signaling typically
promote denaturation and disassembly of noncovalently dimerized
receptors and proteins with which they interact. In this study, we
develop two tools for biochemically tracking GH-induced receptor
dimerization, GH-induced loss of anti-GHRext precipitation
and assessment of GH-induced disulfide linkage of the GHR, and we make
use of these and other reagents we have developed to examine further
both structural aspects of receptor dimerization and determinants for
GHR-JAK2 interaction.
Anti-GHRext monoclonal antibody was raised against the
rabbit GHR extracellular domain and can effectively immunoprecipitate rabbit, human, and bovine (data not shown) GHRs; it does not
significantly recognize murine or rat receptors (data not shown). We
found that GH treatment of cells that express the rabbit or human GHRs
led to a significant loss of immunoprecipitability of the receptors with anti-GHRext, a characteristic not shared by three
separate antibodies (data for one of them, anti-GHRcyt, not
shown in this study) directed at the cytoplasmic domain of the
receptor. The GH antagonist, G120K, while it could bind to the GHR (as
evidenced by its ability to compete with GH for
GH-dependent effects), did not render the GHR less
precipitable by anti-GHRext. Thus, the ability of
anti-GHRext to recognize the GHR depends on the
dimerization state of the receptor, rather than its state of ligand
occupancy. Although we do not yet know whether this loss of recognition
by anti-GHRext is related to a conformational change
occurring within receptor monomers upon their dimerization or whether
it is because of dimerization-induced blockade of an epitope located in
what becomes the GHR-GHR interface, testing of the ability of this antibody to immunoprecipitate the receptor appears a convenient means
of immunologically and electrophoretically assessing the degree of GHR dimerization.
We previously detected and partially characterized GH-induced GHR
disulfide linkage in human IM-9 lymphoblasts (11, 12), cells that
homologously express roughly 5000 GH-binding sites per cell (28). In
those cells, GH treatment induces rapid formation of a species
migrating in nonreducing SDS-PAGE at roughly 215-230 kDa that reacts
with several independently derived antisera directed at the GHR
cytoplasmic domain. In our earlier studies, two-dimensional nonreducing/reducing "diagonal" SDS-PAGE revealed that this
215-230-kDa species, when reduced, contained the monomeric GHR and was
therefore suspected of representing a dimerized GHR that underwent
disulfide linkage (11). Furthermore, a substantial fraction of the GHRs undergoing acute GH-induced tyrosine phosphorylation were found in the
disulfide-linked form, although GHR disulfide linkage occurred even if
GH-induced tyrosine phosphorylation was substantially attenuated (11).
Finally, although significantly slowed, GH-induced GHR disulfide
linkage still proceeded at 4 °C (12).
Our current studies substantiated further the conclusion that GHR
disulfide linkage is a reflection of receptor dimerization. The GH
concentration dependences for GHR disulfide linkage and loss of
anti-GHRext precipitability are quite similar, indicating that these two phenomena may share underlying determinants.
Furthermore, G120K could not support GHR disulfide linkage and
antagonized the ability of GH to promote this linkage at G120K:GH
ratios that also reverse the loss of anti-GHRext
precipitability and have previously been shown to antagonize GH-induced
biological activities (10, 27, 29). Notably, GH induced receptor
disulfide linkage of both mouse (3T3-F442A cells) and rabbit
(transiently and stably transfected into COS-7 and CHO cells,
respectively) GHRs, indicating that this phenomenon is generally found
among various species; G120K also antagonized GH-induced GHR disulfide
linkage in 3T3-F442A cells.
Although GH-induced GHR disulfide linkage reflects GHR dimerization, we
were able to uncouple the two phenomena. Our mutagenesis results
strongly point to cysteine 241 as critical in formation of the
disulfide-linked receptor; mutation to serine rendered the rabbit GHR
incapable of efficiently undergoing this linkage. Yet, consistent with
studies showing a lack of impairment of receptor extracellular domain
folding when residue 241 was mutated (30), rGHR C241S displayed a
capacity for 125I-hGH binding similar to that seen for the
wild-type rGHR when each was expressed in COS-7 cells. Our use of
anti-GHRext immunoprecipitation as a monitor of GH-induced
dimerization indicated that, despite its inability to become
disulfide-linked in response to GH, rGHR C241S still underwent
GH-induced dimerization (anti-GHRext precipitation of rGHR
C241S was substantially decreased by GH treatment). These findings
favor a model in which GH promotes rapid noncovalent receptor
dimerization which then quickly progresses to a disulfide-linked assembly of receptor dimers if cysteine 241 is intact. Our results indicate that receptor dimerization is not driven by disulfide linkage,
but they do not rule out the possibility that cov
§,
**, and
TOP
ABSTRACT
INTRODUCTION
