| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
INSERM U344: Endocrinologie Moléculaire (H.B., A.P., F.F.,
P.A.K., M.E.) Faculté de Médecine Necker 75730
Paris, France
INSERM U460: Remodelage Vasculaire (M.P-A.)
CHU Xavier Bichat 75870 Paris, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Previous studies have shown a differential contribution of the carbohydrate groups of several cytokine receptors on their cellular distribution, ligand binding, and functional activity. The extracellular domain of the PRLR contains three N-linked glycosylation sites (3, 12) defined by the presence of the consensus sequence Asn-X-Ser/Thr-X, where X is any amino acid except a proline (13). First reports demonstrated that deglycosylation by enzymatic treatment of the short form of PRLR does not impair the PRL-binding capacity (14); similarly, PRLRs presenting individual mutations of the three asparagine (Asn) residues are correctly expressed at the cell surface and exhibit a normal affinity (15). Using glycosylation blockers, it has been shown that the glycosidic core of the PRLR appears necessary to achieve an active conformation of the protein (14); however, interaction of these inhibitors with cellular metabolism at levels other than the glycosylation process is not excluded. Indeed, tunicamycin has been shown to activate an endoplasmic reticulum stress pathway, leading to induction of specific proteins, as CCAAT/enhancer binding protein homologous proteins (16), or chaperones that are known to influence the folding and/or trafficking of proteins. However, the functional activity of deglycosylated PRLR has not been investigated. Since specific signals emerging from the extracellular domain of cytokine receptors have been demonstrated to modulate signaling (17, 18), the influence of N-glycosylation of the PRLR on signal transduction is thus of potential interest.
To assess the functional importance of glycosylation of the PRLR, different combinations of mutations were performed affecting the three N-glycosylation sites of the long form of PRLR: Asn35, Asn80, and Asn108. These mutated PRLRs were expressed in several fibroblast cell lines (COS-7, 293, and CHO), and their trafficking, ligand-binding capacity, and functional activity were analyzed. Our results demonstrate that the carbohydrate groups are not essential per se for functional activity, since the aglycosylated PRLR, once partially addressed to the 293 cell membrane, activates transcription of the ß-casein gene. The cell surface targeting of the aglycosylated PRLR appears, however, to be cell-dependent and may affect its biological activity.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
) were incubated with iodinated human GH
(hGH) in the presence of disuccinyl suberate (DSS), and cross-linked
complexes were analyzed by SDS-PAGE. A specifically radiolabeled band
with a molecular mass of 117 kDa was apparent for the WT PRLR (Fig. 2); N35,80,108D and N80,108D receptors
migrated as smaller specific bands of apparent molecular mass of 109
and 113 kDa, respectively. Excess of unlabeled ovine PRL (oPRL) was
able to compete with the [125I]hGH binding of all
mutants. After the molecular mass of the complex was corrected by
subtracting the mass of the hGH (22 kDa), the molecular mass for WT
PRLR, N80,108D, and N35,80,108D receptors was determined to be,
respectively, 95, 91, and 87 kDa. These results suggest that each of
the three N-linked glycosylation sites of the PRLR undergoes
glycosylation in COS-7 cells and that 
8 kDa of N-glycosylation
contributes to the molecular mass of the PRLR, as previously described
(15). Four kilodaltons of glycosylation occur at Asn35 (difference
between the molecular mass of the WT PRLR and the N80,108D mutant), and
2 kDa for each of the two other sites as shown by individual mutations
(see below in Fig. 6A).
|
|
|
, mutants N35D, N80D, N108D, and
N80,108D showed a similar level of binding to that of the WT PRLR
(100% activity of the native receptor corresponds to 20% specific
binding of the ligand), which suggests that about the same number of
receptors are expressed at the cell surface. In contrast, a 90%
decrease in binding of the N35,80,108D mutant was observed at 4 C. To
determine whether this altered binding was due to a modification of
binding affinity, competition studies of oPRL binding were performed in
COS-7 cells expressing the mutated forms of the PRLR. Experiments were
performed at 20 C to access to the intracellular pool of receptors.
Table 1 shows that the mutants exhibited
the same relative affinity for oPRL compared with the wild-type PRLR
[association constant (Ka) = 3.82
nM-1]; this constant was similar to the
affinity constant deduced from Scatchard analysis performed on
microsomal preparations from COS-7 or 293 cells (data not shown).
Similar levels of ligand-binding sites were detected in cells
expressing WT PRLR and single and double mutants (10,000–20,000 sites
per cell); the weaker (but not significantly different) number of
N35,80,108D receptors can be explained by variations of the level of
expression in transiently transfected cells. Taken together, these
results indicate that the aglycosylated PRLRs appear to be normally
expressed in the cell and exhibit a high-affinity binding site similar
to the WT receptor. The altered cell surface binding observed at 4 C
may thus reflect a defect in the transfer of the aglycosylated receptor
to the cell surface.
|
|
, a and d), confirming previous results
(19). The localization at the plasma membrane level was also observed
in nonpermeabilized cells with the N80,108D receptor mutant, whereas in
permeabilized cells, immunostaining was observed in the endoplasmic
reticulum, with a partial accumulation in the juxtanuclear area
depending on the cell and in vesicles (Fig. 4, b and e).
Immunofluorescence results for N35D, N80D, and N108D receptors showed
localization (data not shown) and cell surface binding similar to that
of the WT and N80,108D PRLRs. In contrast, in COS-7 cells expressing
the N35,80,108D receptor, we detected immunostaining only in the
juxtanuclear area (Fig. 4f) without any significant immunostaining at
the level of the plasma membrane (Fig. 4c). This staining was lower
than that observed for the WT PRLR or single mutant receptor,
suggesting a lower level of protein expressed. It also suggests that
this receptor is located in or close to the Golgi compartment. Similar
results were obtained in 293 cells transfected with this mutant,
especially the absence of immunofluorescence in the cell surface of
nonpermeabilized transfected cells (Fig. 4i). Immunostaining at the
cell surface observed in cells transfected with the WT receptor (Fig. 4g) or the N80,108D receptor (Fig. 4h) confirms results obtained with
the COS-7 cell line.
|
, b and c). In some cells, staining was
also visible in the rough endoplasmic reticulum (Fig. 5d).
Immunofluorescence experiment using antibody directed against the
protein disulfide isomerase [PDI, a luminal rough endoplasmic
reticulum foldase (21)] was performed in parallel on COS-7 cells (Fig. 5f): the pattern of endoplasmic immunostaining markedly differs from
the perinuclear staining pattern observed for Golgi protein (Fig. 5, c
and e) or the triple mutant (Fig. 5a).
|
.
|
and 7A). After stripping the membrane
and reprobing with anti-JAK2 antibody, JAK2 (130 kDa) was shown to
associate with all forms of PRLR (Figs. 6B and 7B). In response to PRL,
all receptors are phosphorylated and JAK2 kinase is activated. The
intensity of phosphorylation of both N80,108D receptor and associated
JAK2 was similar to the WT PRLR (Fig. 7C). However, we observed reduced amounts
of N35,80,108D receptor (Fig. 7A) and JAK2 protein (Fig. 7B). These
results may reflect a lower expression of the aglycosylated receptor
and, consequently, lower activation of the kinase. Thus, aglycosylated
PRLR retains active conformation for its phosphorylation and
association with JAK2, but its intracellular retention can indirectly
prevent coupling of the receptor with JAK2 kinase.
|
). Since this receptor was shown to be
phosphorylated and could associate with the JAK2 protein in 293
transfected cells, this suggested that the aglycosylated receptor could
be functional in this cell line. This mutant was thus expressed in 293
cells to test its functional activity; a significant activity (40% of
the WT activity) was observed when cells were transfected with 0.1 µg
of N35,80,108D mutant cDNA (Fig. 8B); the overexpression of this
receptor (0.5 µg) permitted an almost complete rescue of activity.
Our hypothesis is that even though only a small fraction of
aglycosylated PRLRs are transported to the cell surface of 293 cells,
these receptors are able to mediate the milk protein gene
transcription, while the absence of detectable transcriptional activity
in CHO cells could be due to the inhibition of the transport of the
triple mutant, resulting in the almost complete absence of receptors at
the cell surface, as compared with 293 cells. The cell surface
targeting of the aglycosylated receptor is thus differently regulated
in these two cell lines and may influence its further signaling.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Hormone Binding of Glycosylation Site Mutated PRLRs
Our cross-linking studies suggest that the three Asn-linked
carbohydrate moieties contributed 8 kDa of glycosylation to the total
molecular mass of the long form of PRLR. If the theoretical molecular
mass of the nonglycosylated PRLR is considered to be 67 kDa, the
molecular mass of the N-glycosylated form should be 75 kDa. The
difference in the molecular mass between the predicted (75 kDa) and
apparent (95 kDa) glycosylated form observed on SDS gel may be due to
the presence of ubiquitin chains, as it was recently described for the
rabbit PRLR (22) or to other posttranscriptional modifications such as
phosphorylation. Our results suggest a contribution of 4 kDa of
the Asn35-linked glycosylation chain, and 2 kDa,
respectively, for Asn80 and Asn108-linked
glycosylation chains to the molecular mass of the long form of the
PRLR, in agreement with previous results obtained with the short form
of PRLR (15). The glycosidic core of the receptor is in fact composed
of similar tri- or tetra-antennary complex carbohydrates, with terminal
sialic acid residues (14): the higher molecular mass of glycosylated
chain linked to Asn35 compared with those linked to
Asn80 or Asn108 can be attributed to a
different number of antennae.
Our results showing that aglycosylated PRLR binds ligand with a high affinity, similar to that of the glycosylated form (Ka = 1.8 nM-1) indicate that the binding of the hormone to these receptors does not depend on the glycosylation. This is in agreement with Lascols (24), who showed that deglycosylation of mature mouse PRLR did not affect its ligand-binding properties. On the other hand, deglycosylated hepatic rat receptor resulting from tunicamycin treatment is not able to bind PRL (14). In fact, tunicamycin may act at other levels, more important for cellular processing than glycosylation, and thus may indirectly modify other processes, such as trafficking of the receptor. This result may also be explained by a specific process in liver cells. Nevertheless, the site-directed mutagenesis approach used in this study as well as previous results indicate clearly that Asn-linked glycosylation of PRLR is not implicated in the binding of the hormone to the PRLR.
Cellular Expression of N-Glycosylation Site Mutated PRLRs
Cell surface PRL binding experiments showed that removal of all
N-glycosylation sites from the PRLR resulted in alteration in its
distribution, since low levels of aglycosylated PRLR are present at the
cell surface; the majority of this receptor accumulates instead in the
perinuclear Golgi-like compartment, as shown by indirect
immunofluorescence studies. Presence of vesicles are never observed in
cells transfected with this mutant using immunofluorescence, but this
does not exclude that a small undetectable fraction is targeted to the
cell surface and reinternalized after which it accumulates in the Golgi
compartment.
Although the translocation of the PRLR from the endoplasmic reticulum to the Golgi apparatus occurs in the absence of N-glycosylation, further transport to the cell membrane seems to depend essentially on the presence of carbohydrate groups on the receptor. Several studies using N-glycosylation inhibitors, such as tunicamycin, have also demonstrated that deglycosylated receptors cannot be transported to the cell surface and accumulate in the endoplasmic reticulum and pre-Golgi compartment (14). This selective retention is accounted for by a quality control mechanism present in the endoplasmic reticulum (24), by which the transport of incompletely folded or misassembled proteins into the secretory pathway is inhibited; these defective proteins retained in the endoplasmic reticulum or in other early compartments of the secretory pathway accumulate and aggregate and are usually degraded (25, 26). Absence of glycosylation may thus result in protein alteration at the level of folding of the PRLR in the endoplasmic reticulum, which is probably further targeted to a degradation process, as suggested by the lower cytoplasmic immunostaining observed with the aglycosylated PRLR. The glycosylated chains of the receptor may interact with specific chaperone proteins, promoting correct folding of the receptor for its cell membrane targeting. Inhibition of such association may lead to an unstable protein that may be targeted to a degradation process. On the other hand, it cannot be excluded that the substitutions of Asn residues result in some modifications of the three- dimensional structure of the aglycosylated receptor that consequently prevent its surface expression. Thus, glycosylation appears crucial for the correct folding of the PRLR, allowing transport to the plasma membrane.
The double mutant, glycosylated only on Asn35 residue, showed no significant difference in its ligand binding or cellular localization compared with the native PRLR and is still targeted to the cell surface along with partial accumulation in a perinuclear compartment, depending on the cell (and possibly on the level of expression). Individual mutations confirmed that no specific glycosylated chain is involved in the cellular localization of the long form of the PRLR; however, comparison of the expression of the N35D and the N80,108D mutants demonstrates that correct expression of the protein can be achieved with partial glycosylation equivalent to 4 kDa either on Asn35 or on both Asn80 and Asn108.
Of interest is the residual cell surface expression of aglycosylated PRLR (only detectable in COS-7 and 293 cells using binding experiments), which is sufficient to allow the functional activity of this mutant in 293 cells. This activity cannot be explained by existence of other glycosylation chains associated with the PRLR, such as O-glycosylation, since it has been shown that this receptor is only N-glycosylated (12, 14). Overexpression of receptor, which leads to the saturation of the endoplasmic reticulum and Golgi compartments, could modify the environment conducive to conformational rescue of the misfolded protein necessary for its cell surface expression. Other motifs in the primary sequence of the PRLR could also be implicated in this export process, e.g. the conserved WSXWS motif of the extracellular domain (15, 27) or potential cytoplasmic sequences, as has been described for other receptors (28). Interaction of the aglycosylated receptor with other chaperones that might assist the protein to its cell membrane localization cannot be excluded. Lastly, the cell surface expression of an aglycosylated protein may reflect a more specific cell line-dependent mechanism.
The influence of N-glycosylation on the transport of receptors has been
described for other members of the cytokine receptor superfamily; for
example, the erythropoietin receptor is more abundantly expressed than
a glycosylation-defective erythropoietin receptor mutant in a lymphoid
BA/F3 cell line (29); interferon-
receptor and GH receptor transport
also appears to be glycosylation-dependent (30, 31). On the other hand,
N-glycosylation of the human granulocyte macrophage-colony
stimulating factor-
(GM-CSF) receptor (32) or the interleukin-2
receptor (33) does not participate in cell surface receptor expression.
These latter studies have not focused, however, either on the
importance of other glycosylated receptor-associated chains in the
trafficking of the receptor or in differences in the cell lines
used.
Previous data have suggested a role of N-glycosylation in the internalization process; for example, all N-linked glycosylation chains of the ectodomain of the ß-subunit of insulin receptor apparently participate in the control of receptor internalization (34); total inhibition of N-glycosylation of the GH receptor also results in a more efficient internalization of the protein in stably transfected mouse T cells (31). In the present report, the low level of aglycosylated receptors present at the cell surface at 4 C make the internalization study difficult to analyze (data not shown); nevertheless, a modification in recycling of the receptor cannot be excluded.
JAK2 Activation and Transcriptional Activity of Glycosylation Site
Mutants
Recently, it has been demonstrated that the extracellular domain
of cytokine receptors can modulate the activity of the cytoplasmic
domain (17, 18). Several studies have analyzed the implication of
glycosylation of the extracellular domain on receptor function.
N-Glycosylation does not appear to be critical to elicit either
tyrosine phosphorylation of proteins after growth hormone/receptor
interaction (31) or the erythropoietin-induced cell proliferation
signal (29). However, the absence of glycosylation of the GM-CSF
receptor markedly decreased GM-CSF-induced deoxyglucose uptake and
protein tyrosine phosphorylation in HL-60 (eos) cells (32). Our
experiments on the activation and association of the tyrosine kinase
JAK2, the initial step in PRLR signaling, indicate that all mutant
receptors were able to be phosphorylated and to activate the kinase in
a PRL-dependent manner. This indicates that the absence of
N-glycosylation does not interfere with the first key steps in PRLR
signaling. These results imply that homodimerization of the
aglycosylated PRLR occurs. In fact, previous studies have demonstrated
the formation of a complex consisting of one molecule of hormone and
two molecules of recombinant nonglycosylated extracellular domain of
the rat PRLR, expressed in a prokaryotic cell system (35). Functional
activity of the mutant receptors was further analyzed by testing their
ability to activate a PRL-dependent ß-casein gene promoter. Partially
deglycosylated receptors were all functional when tested in CHO cells
and also in 293 cells (data not shown); on the other hand, the
aglycosylated mutant, inactive in CHO cells, was fully active when
overexpressed in 293 cells. Trafficking and subsequent cell surface
expression of this mutated receptor appear to be different between
these two cell lines. The biological response may be elicited by a
limited number of receptors present at the cell surface as it has been
shown for insulin receptor (36). 293 cells do express limited but
detectable levels of aglycosylated receptors at the cell surface; it is
thus conceivable that this limited number of receptors is sufficient to
transduce the PRL signal into the cell. The situation is quite
different in CHO cells where much lower levels of receptors are
expressed (37, 38), and thus the level of aglycosylated receptors
present at the cell membrane may be below the threshold level required
for a biological response. Moreover, in 293 or COS cells,
overamplification of multiple cDNAs can create perturbations,
especially in the endoplasmic reticulum, as was recently described for
the expression of the human calcitonin receptor in COS cells (39). The
cell surface expression of the aglycosylated receptor can thus be
explained by an escape process linked to this overexpression, which
appears specific to the cell line.
Taken together, these results demonstrate that the N-glycosylation of the extracellular domain of the PRLR is not implicated in the acquisition of a correct conformation of the receptor essential for ligand binding, but its absence leads to different alterations in cell surface translocation of the protein, depending on the cell line, which can affect further signal transduction.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-JAK2, Upstate Biotechnology, dilution
1:4000). Mouse monoclonal anti-PRLR U5 has been described previously
and used at a concentration of 0.5 µg/ml in Western blot (12).
Monoclonal mouse antibodies specific for the 58K protein (Sigma, St.
Louis, MO; dilution 1:40) and the protein disulfide isomerase (PDI,
Stressgen Biotech Corp., Victoria, British Columbia, Canada; dilution
1:100) were used to localize the Golgi apparatus and the rough
endoplasmic reticulum, respectively. The rabbit polyclonal anti-rab6
antibody was generously provided by Dr. B. Goud (UMR 144, CNRS,
Institut Curie, Paris, France) (40). The secondary antibodies used in
these studies were the Texas-Red-conjugated and
fluorescein-isothiocyanate-conjugated (FITC) goat anti-mouse IgG
(dilution 1:40) (Vector Laboratories, Burlingame, CA; and Biosys,
Compiègne, France), the biotinylated donkey anti-rabbit IgG and
the Texas-Red-conjugated streptavidin (1:100; Amersham International,
Aylesbury, UK).
Construction of the Mutants
Short-form rat PRLRs presenting individual substitutions of the
asparagyl residues (Asn35, Asn80, and
Asn108) by aspartic residues were previously characterized
(15); the negative charge present on terminal sialic acid residues of
glycosylation chains of the native PRLR was then conserved.
NcoI inserts encoding modified sequences were subcloned into
the pECE or pcDNA3 (Invitrogen, San Diego, CA) plasmids containing the
long form of the rat PRLR. The double mutant (N80,108D) was obtained
after BamHI digestion of pcDNA3 mutants. From this mutant
plasmid, a single-stranded DNA was generated by using the origin of
replication of the M13 phage present in the pcDNA3 vector, in the
Escherichia coli CJ236 strain in the presence of M13K07
helper phage. This single-stranded DNA was then used as a template for
mutagenesis with the primer 5'-CAGTGAATAATCGGTAAGGAAG-3' (Genosys
Biotechnologies, Cambridge, UK) where the Asn35 residue was
replaced by an aspartate residue, to obtain the triple mutant. The
modified region was verified by sequencing.
Cell Culture
Kidney monkey cells (COS-7) were grown as monolayers in DMEM nut
F12 medium containing 10% FCS. Cells were transiently transfected at
60% confluence using the diethylaminoethyl-dextran-chloroquine
procedure (38). A 10% solution of Me2SO4 in
HBSS was applied 4 h after transfection, after which cells were
incubated with complete medium. This cell line, known to generate a
high level of expression of recombinant receptors, was used for binding
and immunofluorescence studies. Chinese hamster ovary (CHO K1) cells,
routinely used for transcriptional assays (37, 38), were maintained in
Ham’s F-12 medium containing 10% FCS. Before transfection, cells were
starved in GC3 serum-free medium (1:1 mixture of MEM and Ham’s F-12
supplemented with 10 µg/ml transferrin, 3 µg/ml insulin, 2.5
mM glutamine, and 1x nonessential amino acids). 293
fibroblast cells were maintained in DMEM nut F12 containing 10% FCS.
Cells were transfected by the calcium phosphate technique in a rich
medium (2/3 DMEM nut F12, 1/3 DMEM containing 4.5 g/liter glucose with
10% FCS) and used for Western blot studies and transcriptional
assays.
Determination of Cell Surface Ligand Binding
[125I]hGH was prepared using chloramine T to a
specific activity of 80–140 µCi/µg (41). COS-7 (106)
cells were transfected with 1 µg of cDNAs encoding WT or mutant PRLR
in six-well plates. After 48 h, cells were starved in GC3 medium
for several hours and then incubated with 50,000 cpm of
[125I]hGH in the presence or absence of excess of
unlabeled oPRL (2.5 µg), in 1 ml PBS, 0.5% BSA overnight at 4 C.
After being washed twice with chilled PBS, cells were solubilized with
1 ml of 1 M NaOH, and cell-associated radioactivity was
counted using a -counter. Specific binding was determined by the
difference between total and nonspecific binding.
Competition Curves
Transfected COS-7 cells were incubated with 100,000 cpm of
[125I]hGH and increasing concentrations of unlabeled oPRL
(0–1 mg) in 1 ml of PBS/BSA 0.5% overnight at 20 C. Cells were washed
twice with chilled PBS and solubilized with 1 ml of 1 M
NaOH, after which radioactivity was counted. Binding data were analyzed
using the LIGAND program (Elsevier Biosoft, Cambridge, UK).
Indirect Immunofluorescence
COS-7 (2 x 105) or 293 (2 x 105) cells
transfected with 0.5 µg of WT or mutant PRLR cDNAs were grown as
subconfluent monolayer cultures in Labtek chambers for 2 days. Cells
were fixed with 4% paraformaldehyde in 0.1 M phosphate
buffer, pH 7.4, for 20 min at room temperature and permeabilized or not
with methanol (-20 C) for 5 min. Antibody dilutions were made in PBS
containing 0.1% of BSA. The nonspecific sites were saturated with
blocking solution (goat serum diluted 1:40). Slides were incubated with
the U5 antibody at 160 µg IgG/ml overnight at 4 C and then
extensively washed in PBS and further incubated with a 1:40 dilution of
FITC goat anti-mouse IgG, as previously described (19), for 1 h at
room temperature. Cells were incubated overnight at 4 C with 1:100
dilution of mouse monoclonal anti-PDI antibody to localize the rough
endoplasmic reticulum, and then extensively washed and further
incubated with a 1:40 dilution of Texas-Red-conjugated goat anti-mouse
IgG, for 1 h at room temperature. For the double labeling
experiment, transfected cells fixed with 4% paraformaldehyde were
permeabilized with Triton 0.1% for 3 min at room temperature. After
washing, donkey serum diluted in PBS-BSA 2.5% was applied during 15
min. Cells were then successively exposed to the anti-rab6 antibody
(1:40) overnight at 4 C followed by 1 h exposure at room
temperature to biotinylated donkey anti-rabbit IgG, and then washed and
incubated 1 h at room temperature with Texas-Red-conjugated
streptavidin. The same cells were then successively incubated with goat
serum for 15 min at room temperature, with anti-PRLR U5 antibody for
3 h at room temperature, and with the FITC goat anti-mouse IgG for
1 h at room temperature. Once washed in PBS, slides were mounted
in 50% glycerol in PBS and observed on a Zeiss microscope (Carl Zeiss,
Thornwood, NY). No immunofluorescence was detected in any of the
specificity control experiments (including incubation of cells in the
absence of primary and/or secondary antibodies) or with IgG control
monoclonal antibodies.
Cross-Linking Experiments
COS-7 cells (106 cells per plate) were transfected
with WT or mutant PRLR cDNAs and then incubated with 1 ml of DMEM/0.1%
BSA containing 500,000 cpm of [125I]hGH in the absence or
presence of excess of unlabeled oPRL (2.5 or 5 µg) at 37 C for 30
min. Cells were then incubated with 0.5 mM of the
cross-linking agent DSS at room temperature for 20 min. The
cross-linking reaction was quenched with 50 mM Tris-HCl, pH
7.4, 150 mM NaCl. Cells were lysed in 50 µl 2x Laemmli
buffer (42) for 15 min at 4 C. Protein complexes were electrophoresed
on a 7.5% SDS-PAGE. Gels were dried and submitted to autoradiography
for 3–5 days.
Purification of PRLR Complexes by Streptavidin-Agarose
293 cells (107/100 mm dish) were transfected with 4
µg cDNA encoding wild type or mutated forms of PRLR and 2 µg cDNA
encoding the human tyrosine JAK2, using the calcium phosphate
procedure. Twenty four hours after transfection, cells were deprived of
serum for 16 h, and then stimulated with 20 nM
biotinylated oPRL or bGH for 10 min at 37 C (Biotinylation kit,
Boehringer Mannheim, Mannheim, Germany). Cellular proteins were
extracted in 1 ml lysis buffer (150 mM NaCl, 50
mM Tris-HCl, pH 7.5, 5 mM EDTA, 1
mM Na3VO4, 30 mM Na
pyrophosphate, 50 mM NaF, 1 mM
phenylmethylsulfonylfluoride, 5 µg/ml aprotinin, 2 µg/ml leupeptin,
1 µg/ml pepstatin, 10% glycerol, and 0.2% Triton X-100) for 10 min
at 4 C. Equivalent quantities of proteins were then mixed with 20 µl
of streptavidin-agarose beads (50% vol/vol) 2 h at 4 C. The
agarose beads were washed three times with lysis buffer, eluted in SDS
sample buffer, and subjected to a 7.5% SDS-PAGE.
Western Blot
Proteins were transferred to a nitrocellulose membrane at room
temperature for 1 h, incubated in TBS (50 mM Tris HCl,
pH 7.5, 200 mM NaCl, 1% BSA) for 1 h, and then
incubated with the first antibody in TBS complemented with 0.05%
Tween-20 for 90 min. After washing, membranes were incubated for 1
h at room temperature with anti-rabbit or anti-mouse IgG conjugated to
horseradish peroxidase (Amersham Corp., Arlington Heights, IL) diluted
at 1:4000. Proteins were visualized using the enhanced
chemiluminescence detection system (Amersham International, Aylesbury,
UK). Membranes were stripped overnight at 4 C in an acid solution (0.1
M glycine, pH 3, 0.1 M NaCl) and reprobed.
Transcriptional Assays
CHO cells (8 x 105) were transiently transfected in
six-well plates by the calcium phosphate procedure with 1 µg of cDNAs
encoding the WT or mutant PRLR, 0.5 µg of the ßCas-luc reporter
construct [containing the luciferase-coding sequence linked to the
sequence -2300/+490 of rat ß-casein gene promoter (21)], and 1 µg
of pCH110 (ß-galactosidase expression vector, Pharmacia, Bromma,
Sweden). Cells were submitted to a glycerol shock after 4 h of
incubation, and then incubated for 48 h in GC3 medium containing
20 nM oPRL and 250 nM dexamethasone (Sigma
Chemical Co.). 293 cells (106) were also transiently
transfected with 0.1 µg of receptor cDNA, 0.1 µg of reporter
construct, and 0.5 µg of pCH110 using the same precipitation
procedure. After an overnight 3% CO2 shock, cells were
starved in serum-free medium for several hours, and then stimulated
with 20 nM oPRL and 250 nM dexamethasone for
24 h. Cells were then lysed in lysis buffer (1% Triton X-100, 25
mM Tris-phosphate, pH 7.8, 2 mM dithiothreitol,
2 mM
1.2-diaminocyclohexane-N,N,N',N'-tetraacetic
acid, 10% glycerol). After centrifugation, 20 µl of supernatant were
used for determination of luciferase and ß-galactosidase activities.
A LUMAT LB9501 (Berthold, Wildbad, Germany) was used to quantify
luciferase activity into relative light units. Arbitrary luciferase
units were normalized for ß-galactosidase activity in each
experiment. Fold induction is calculated as the ratio between the
normalized light units of stimulated vs. unstimulated cells.
Activities of the mutants were calculated from the maximal activity
obtained with WT PRLR. Values are mean ± SEM of five
or three independent experiments performed on CHO cells or 293 cells,
respectively.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication July 28, 1997. Revision received December 3, 1997. Accepted for publication December 24, 1997.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
receptor. J Biol Chem 265:1710–1717 subunit is essential for ligand binding and signal transduction.
J Biol Chem 270:24580–24584 receptor. J Biol Chem 263:4900–4906This article has been cited by other articles:
 |
|
 |
|
  D. F. Fiol, E. Sanmarti, R. Sacchi, and D. Kultz A novel tilapia prolactin receptor is functionally distinct from its paralog J. Exp. Biol., July 1, 2009; 212(13): 2007 - 2015. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Freeman, B. Kanyicska, A. Lerant, and G. Nagy Prolactin: Structure, Function, and Regulation of Secretion Physiol Rev, October 1, 2000; 80(4): 1523 - 1631. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. B. Kline, H. Roehrs, and C. V. Clevenger Functional Characterization of the Intermediate Isoform of the Human Prolactin Receptor J. Biol. Chem., December 10, 1999; 274(50): 35461 - 35468. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Vleurick, A. Pezet, E. R. Kühn, E. Decuypere, and M. Edery A {beta}-Turn Endocytic Code Is Required for Optimal Internalization of the Growth Hormone Receptor but Not for {alpha}-Adaptin Association Mol. Endocrinol., November 1, 1999; 13(11): 1823 - 1831. [Abstract] [Full Text]
|
|
|
|
|
|
P. Touraine, I. Beau, A. Gougeon, G. Meduri, A. Desroches, C. Pichard, M. Detoeuf, B. Paniel, M. Prieur, J-R Zorn, et al. New Natural Inactivating Mutations of the Follicle-Stimulating Hormone Receptor: Correlations between Receptor Function and Phenotype Mol. Endocrinol., November 1, 1999; 13(11): 1844 - 1854. [Abstract] [Full Text]
|
|
|
|
|
|
J. B. Kline and C. V. Clevenger Identification and Characterization of the Prolactin-binding Protein in Human Serum and Milk J. Biol. Chem., June 29, 2001; 276(27): 24760 - 24766. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
