 |
INTRODUCTION |
Secretory and membrane glycoproteins of eukaryotic cells are
co-translationally translocated into the lumen of the endoplasmic reticulum (ER)1 from where
they travel to Golgi complex on the secretory pathway and to other
destinations. Recently, it became evident that the transport of
proteins out of the ER is limited by a unique "quality control"
system that involves recognition and retention of misfolded or
misassembled proteins. If further attempts of acquiring the correct
folding fail, these proteins may be directed into a degradation pathway
(1-4). In the case of oligomeric proteins, the correct formation of
disulfide bonds plays an important role in the assembly of secretory
and membrane proteins (5-7), which in turn determines stability,
intracellular transport, maturation, and function.
The disulfide bonds are generated through oxidation in the ER. The ER
lumen is unique among the various compartments in the eukaryotic cells
because it provides an oxidizing environment for the disulfide bond
formation with the help of the protein disulfide isomerase that
promotes the disulfide bond formation (8, 9). Recently, it was
demonstrated that the co-translational disulfide bond formation,
folding, and oligomerization of proteins within the ER can be
reversibly inhibited by the addition of the disulfide bridge disrupting
agent dithiothreitol (DTT) to living cells (10-12). Interestingly,
upon the removal of DTT, the disulfide bond formation, folding via
normal ER folding intermediates, and oligomerization seems to take
place (10-17). Moreover, DTT does not inhibit the transport within the
secretory pathway (13, 17).
Human chorionic gonadotropin (hCG), a glycoprotein hormone, is composed
of two noncovalently linked and glycosylated 
- and 
-subunits
(18). It is synthesized by the trophoblast cells of the placenta as
well as by malignant trophoblast cells and tumors of various origins
(19-29). Both subunits are transcribed from separate genes and
assembled post-translationally in the ER. JEG-3 choriocarcinoma cells
not only secrete hCG but also an excess of free - and a minor
quantity of -subunit (26). The - and -subunits are synthesized
via precursors. Five -subunit intermediates have been characterized
(30, 31). These intermediates represent discrete steps in the folding
process that are apparently coupled with the formation of individual
disulfide bonds (32). The disulfide bond formation seems to take place
post-translationally (30). No attempt has yet been made to understand
the effect of DTT on the biosynthesis of - and -hCG subunits
in vivo. Here, we communicate our investigations on the
effect of prevention of the disulfide bridge formation in
vivo by the use of DTT on the association of the - and
-subunits. Moreover, we have studied the effect of reduction and
reoxidation on the N-glycosylation of the -subunit as well as on the
maturation of the -subunit.
| |
EXPERIMENTAL PROCEDURES |
Cell Culture--
JEG-3 cells (American Type Culture Collection,
Manassas, VA) monolayer cultures were maintained in DMEM medium
(Sigma), containing 10% (v/v) fetal calf serum (Linaris Corp.,
Bettingen, Germany). The medium was supplemented with 3.7 g/liter
sodium bicarbonate, 100 IU/ml penicillin, and 100 µg/ml streptomycin
(Linaris Corp.). Confluent cell monolayers grown in 25-cm2
plastic flasks (Nunc GmbH, Wiesbaden-Biebrich, Germany) were used for
pulse labeling and chase experiments.
Radioactive Labeling and Cell Lysate Preparation--
Confluent
(<95%) JEG-3 cells grown at 37 °C in DMEM containing 10% (v/v)
fetal calf serum were used for the pulse labeling and chase
experiments. The cells, kept for 30 min in deficient medium (DMEM
lacking cysteine and methionine) were pulse labeled (for time, see
"Results") with 100 µCi/ml [35S]Met/Cys mixture
(Amersham Pharmacia Biotech) and chased (for time, see "Results")
in the DMEM containing 5 mM excess of Met and Cys in the
presence or absence of 5 mM DTT (as indicated below). After
incubation, the medium was removed, and the cells were chilled on ice
and incubated for 5-10 min with ice cold phosphate-buffered saline (10 mM sodium phosphate, pH 7.2, 150 mM NaCl)
containing 40 mM N-ethylmaleimide to prevent the
rearrangement of -S-S- bonds by a blockade of the free thiol groups.
The cells were again washed three times with the ice-cold
phosphate-buffered saline and lysed with 50 mM Tris-HCl
lysis buffer, pH 7.6, containing 200 mM NaCl, 0.1% (w/v)
SDS, 0.5% (w/v) sodium deoxycholate, 1.0% (w/v) Nonidet P-40, 20 mM N-ethylmaleimide, 20 mM EDTA, and
2 mM phenylmethylsulfonyl fluoride. A post-nuclear
supernatant (PNS) was prepared by centrifugation of the lysate at
19,900 × g for 5 min (Biofuge 15, Heraeus, Osterode, Germany). To reduce the nonspecific coprecipitation, the PNS was precleared by shaking with protein A Staphylococcus aureus
cells (Sigma; 100 µl cell suspension/1500 µl lysate) for 30 min at
4 °C. The S. aureus cells were pelleted for 5 min at
19,900 × g (Biofuge 15), and the supernatant was used
for sequential immunoprecipitation by using two different anti-hCG
antibodies as given below.
Immunoprecipitation and SDS-PAGE Analysis--
Two different
polyclonal antibodies against the hCG subunits were used in
immunoprecipitation of cell lysate. The various -subunit forms were
purified by using an antibody (G10, kindly provided by Dr. E. Bedows,
Omaha, NE) that recognizes all forms of free and -subunit
precursors but does not cross-react with the -subunit (33). The
fraction of the -subunit being associated with the -subunit in a
dimer, however, is coprecipitated with this antibody. After depleting
the PNS of the -subunit, their precursors, and the -dimer, the
supernatant of the same PNS was used for subsequent immunoprecipitation
with a goat anti--hCG antibody (34, 35). The immunoprecipitation
with each of the antibodies was carried out for 2 h at 4 °C.
The immune complexes were collected on the protein A-agarose beads
(Roche Molecular Biochemicals) and washed three times with the lysis
buffer and once with 20 mM Tris-HCl buffer, pH 6.8. The
immune complexes were eluted by the addition of elution buffer (20 mM Tris-HCl, pH 6.8, containing 1% SDS) and heating of the
samples for 1 min in a boiling water bath. Subsequently, the samples
were centrifuged at 19,900 × g for 15 min, and
aliquots of the supernatant were mixed with the equal volume of the
nonreducing sample buffer (100 mM Tris-HCl, pH 6.8, containing 4% (w/v) SDS, 0.2% (w/v) bromphenol blue, and 20% (v/v)
glycerol) and reducing sample buffer (containing 10% (v/v)
2-mercaptoethanol), respectively. The samples were separated on
SDS-PAGE (Mini-Protean II, Bio-Rad). In all gels, the
14C-labeled molecular weight marker (Rainbow, Amersham
Pharmacia Biotech) was run together with the samples. It contained
myosine (Mr = 220,000), phosphorylase b
(Mr = 97, 400), bovine serum albumin (Mr = 66,000), ovalbumin
(Mr = 46,000), carbonic anhydrase
(Mr = 30,000), trypsin inhibitor
Mr = 21,500), lysozyme
(Mr = 14,300), aprotinin
(Mr = 6,500), insulin chain B
(Mr = 3,400), and insulin chain A
(Mr = 2,350) as molecular weight markers.
Proteins were precipitated by incubation of the polyacrylamide gels in
20% (w/v) trichloroacetic acid. After two washings (20 min each) in
dimethyl sulfoxide, the gels were incubated in 22% (w/v) 2,5 diphenyloxazole (dissolved in dimethyl sulfoxide) for 90 min with
gentle shaking. After several washings with water, the gels were
transferred to Whatman 3MM paper and dried in a gel dryer (Bio-Rad).
The gels were exposed at 
80 °C to x-ray film (Fuji RX) in the
presence of an intensifying screen. The quantitative evaluation of the x-ray films was performed by laser densitometry (2202 Ultro scan, LKB)
and computer-supported calculation of the band intensities of the
fluorograms, or by the Kodak Digital Science one-dimensional system
(Amersham Pharmacia Biotech).
In some experiments, cells treated with 5 mM DTT prior to
(5-40 min) as well as during the pulse (15 min) were lysed at the end
of the pulse. The lysate was purified as described above, however,
using three different monoclonal antibodies (INN-hCG-45, INN-hCG-53,
and INN-hCG-55, kindly provided by Dr. P. Berger, Innsbruck, Austria)
directed against epitopes that are exposed on hCG but not on the free
subunits to assess the immunologic properties of the -dimer
formed in the presence of DTT.
Separation of -Dimers and Free Subunits by Gel Filtration
Chromatography--
Three cultures of JEG-3 cells grown in
75-cm2 flasks were treated 45 min in Met/Cys-deficient DMEM
medium (Sigma) prior to labeling with 88.3 MBq
[35S]Met/Cys per flask for 45 min followed by 45 min of
chase. The cells were rinsed five times with ice-cold
phosphate-buffered saline containing each of 5 mM Met and
Cys. Cell lysis, preparation of the post-nuclear supernatant, and
preabsorption with the protein A S. aureus cells was
performed as described above. 2 ml of the lysate were applied to a
Sephadex G150 column (0.75 × 100 cm) equilibrated with 50 mM Tris-HCl buffer, pH 7.5, containing 0.05% (w/v) SDS, 20 mM EDTA, 10 mM N-ethylmaleimide,
0.02% (w/v) Nonidet-P40, 2 mM phenylmethylsulfonyl
fluoride, and 0.1% (w/v) bovine serum albumin. The column was run in a
fast protein liquid chromatography system (Amersham Pharmacia Biotech)
at a flow rate of 5 ml/h. Fractions of 750 µl were collected, and 20 µl of aliquot of each fraction was counted in a Tricarb 2450 scintillation counter (Canberra Packard, Dreieich, Germany). The pooled
fractions (see below) were purified by immunoprecipitation and analyzed
by SDS-PAGE as described above except that boiling of the unreduced
samples was omitted. The electrophoresis was performed in duplicate.
One gel for the preparation of a fluorogram. The other gel was exposed at 4 °C to a x-ray film (Fuji RX). The bands were excised, and the
radioactive material was eluted by breaking up the gels in to small
pieces with a glass rod and incubation overnight in reducing sample
buffer as well as a freezing-thawing cycle. The eluted proteins were
analyzed by SDS-PAGE.
| |
RESULTS |
Pulse-Chase Kinetics of the hCG Subunits in the Unperturbed
JEG-3 Cells--
The pulse-chase kinetics of the hCG--subunit,
their intermediates, and the -subunit contained in -dimers is
shown in Fig. 1. The band pattern is very
similar as observed by others in the case of JAR cells (see below). At
least four different precursor forms of the mature -subunit can be
discerned, designated as p0, p1,
p2, and *. The p0 form disappeared
completely within the first 15 min of chase (Fig. 1A). The
p1 and p2 forms show higher apparent
molecular weights (Mr = 30,300 and 32,600, respectively) than p0 (Mr = 25,200) in the SDS-PAGE under nonreducing conditions. Upon reduction of
the samples prior to electrophoresis, all the -subunit intermediates
collapse into one band (Fig. 1B) with the same apparent
molecular weight as the p0 band
(Mr = 25,200). This suggests that the different
electrophoretic mobilities of the precursors under unreduced conditions
display differences in the number of disulfide bridges formed. Small
amounts of a -subunit with an apparent molecular weight of 36,900 (*) emerged at a chase time of 30 min. The apparent molecular weight
of this precursor form in the SDS-PAGE is almost identical with that of the mature hCG--subunit; however, * showed an apparent molecular weight of 26,500 upon reduction.
View larger version (39K):
[in this window]
[in a new window]
|
Fig. 1.
Pulse-chase kinetics of hCG subunits in the
unperturbed JEG-3 cells. The cells were incubated in a
Met/Cys-free medium for 45 min prior to a 15-min pulse with
[35S]Met/Cys mixture (100 µCi/ml). Thereafter, the
pulse medium was replaced with the chase medium containing 5 mM Cys and Met for the indicated time. For a 0-min chase
(lane 2), the pulse labeled cells were immediately chilled
on ice and processed for further washing and lysis. The preabsorbed PNS
was immunoprecipitated with the anti- antibody (G10). The immune
complexes, collected on the protein A-agarose were washed, eluted, and
separated on SDS-PAGE under nonreducing (A) and reducing
(B) conditions. For further details see "Experimental
Procedures." In the right panel part of an experiment is
shown with a shorter pulse (30 min of Met/Cys-deficient medium, 2-min
pulse) and the first chase time point 2 min after the end of the pulse.
The intracellular - and -subunit forms are indicated. For further
details see the text. In lane 1 the bands of carbonic
anhydrase (Mr = 30,000) and trypsin inhibitor
(Mr = 21,500) are depicted as molecular weight
markers. Similar results were obtained in a total of five
experiments.
|
|
Moreover, the results also indicate that the observed -subunit
intermediates represent ER forms of the -subunit that have not
entered into a Golgi compartment where the O-glycan residues are attached. After this has taken place an apparent molecular weight
of 37,500 is reached and maintained also in the presence of reducing
agents (mature). At the end of the pulse period (Fig. 1A, lane 2), a significant fraction of the
-subunit was coprecipitated by the anti- antibody (G10),
indicating that an -dimer has already formed. The free
-subunit is not precipitated with G10.
Pulse-Chase Kinetics in the Presence of DTT--
Experiments to
find out the efficient DTT concentration needed to reduce the disulfide
bridges of the hCG subunits in JEG-3 cells were performed in a range of
0.1- 20 mM DTT. A 5 mM concentration of DTT
turned out to be completely sufficient to obtain the same picture of
the subunit bands in the SDS-PAGE as after complete reduction of the
sample with 1.3 M -mercaptoethanol prior to electrophoresis (data not shown). We, therefore used a concentration of
5 mM DTT in the subsequent experiments. The cells were
pulse labeled in the absence of DTT and subsequently chased in the
presence of 5 mM DTT up to 240 min. The pulse labeled cells
showed the presence of -subunit (contained in -dimers) and
-subunit intermediates (Fig. 2,
lane 2). The p0 form was the prominent
intermediate in the presence of DTT. Within a period of 5 min in the
presence of DTT in the chase medium, a distinctly higher intensity of
the p0 form (36.6 ± 13.0% (n = 4)
versus 12.8 ± 1.2% (n = 5) in the absence of DTT; intensity of all forms = 100%) was observed. Moreover, the p1 intermediate was missing in the
presence of DTT.
View larger version (54K):
[in this window]
[in a new window]
|
Fig. 2.
Pulse-chase kinetics of hCG subunits in JEG-3
cells in the presence of DTT. The samples were separated with
SDS-PAGE under nonreducing (A) and reducing (B)
conditions. The cells were pulse labeled under the same conditions as
used in Fig. 1. The chase medium contained 5 mM cysteine
and methionine and 5 mM DTT. Lane 1, molecular
weight markers (the two strong bands represent the molecular weight
markers carbonic anhydrase (Mr = 30,000) and
trypsin inhibitor (Mr = 21, 500)); lane
2, pulse labeled cells without chase (no DTT). Lanes
3-9 show the chase of the labeled subunits in the presence of 5 mM DTT for the indicated time. In three other experiments
similar results were obtained.
|
|
Reoxidation of Reduced - and -Subunits--
In a further
series of experiments, the DTT treatment was performed during the pulse
(15 min) and the first 15 min of chase. Fig.
3 shows the results of a representative
experiment. The reduced -subunit form (p0) was the
only precursor observed as long as DTT was present. Remarkably, an
-subunit band was also visible in the anti- precipitated samples
(lanes 1 and 2), indicating the presence of
-dimer even under the reducing conditions in vivo.
After 5 min of chase in a DTT-free medium, the -subunit precursors
seemed to be recovered and reoxidized (Fig. 3A, lane 3). The band pattern seems to be the same as observed in the
unperturbed cells.
View larger version (79K):
[in this window]
[in a new window]
|
Fig. 3.
Post-translational reduction and reoxidation
of hCG subunits. JEG-3 cells were incubated in the
Met/Cys-deficient medium in the absence of DTT for 30 min, pulse
labeled for 15 min (lane 1), and further incubated in the
chase medium for 15 min (lane 2), each in the presence of 5 mM DTT. Thereafter, the DTT containing chase medium was
removed and the cells were chased in the absence of DTT for the
indicated time (lanes 3-9). The cell lysates were first
immunoprecipitated with the anti- G10 antibody. The band pattern
SDS-PAGE under nonreducing is given in A. After the immune
complexes of the -subunit forms with the G10 antibody were removed,
the supernatants were treated with the polyclonal anti- antibodies
to isolate the free -subunit in the samples (B,
nonreducing conditions). The -subunit, immunopurified from the
culture medium is depicted in section C (SDS-PAGE under reducing
conditions). Similar results were obtained in three other experiments
of the same experimental design.
|
|
The supernatants of the samples preabsorbed with anti- antibody
(which removes -dimers) were immunoprecipitated with an antibody
that recognizes free -subunit. No free -subunit could be detected
in the presence of DTT (Fig. 3B, lanes 1 and
2). This is due to the fact that the anti- antibody used
does not react with the completely reduced
-subunit.2 Within 5 min of
DTT-free chase, the free -subunit ( not contained in the
-dimer) had regained a conformation that was recognized by the
anti- antibody (Fig. 3B, lane 3). The decrease
of the intracellular -subunit concentration in the later chase time (
60 min) was due to the export of the free -subunit via the secretory pathway into the culture medium (Fig. 3C). Besides
the mature free -subunit (m), a small amount of a
-subunit form with a lower apparent molecular weight
(f) was also secreted from DTT-treated cells (Fig.
3C). It might represent an immature - or a degraded
-subunit.
Association of - and -Subunits--
Gel filtration
chromatography of the cell lysates of unperturbed JEG-3 cells was used
to separate -dimers from free subunits (Fig.
4A). The fractions were pooled
as indicated and purified by immunoprecipitation as described above.
The immune complexes were eluted from the protein A-Sepharose and
applied to the SDS-PAGE with and without reduction of the samples (Fig.
4B). In the unreduced samples of the pooled fractions 1 and
2, the immune complexes were visible as a high molecular weight band
that did not enter into the separating gel. Moreover two other bands
(Mr = 52,200 and Mr = 37,700) were detected in the gel (Fig. 4B, pool
2). After reduction, the - and -subunit bands were visible.
Whereas the -subunit present in the pool fractions 1 and 2 was
coprecipitated with the -subunit (as -dimer), the bulk of the
free -subunit was eluted from the column in the fractions of pool 3 (Fig. 4B, right panel). The inability of the
anti- antiserum to precipitate the free -subunits is obvious from
the Fig. 4B (left panel). This shows clearly that
the -subunit eluted in the higher molecular weight fractions of pool
1 and 2 was indeed part of an -dimer complex, whereas the free
-subunit was eluted later. We cut the individual bands of the
unreduced samples from the gel and separated the eluted material again
in SDS-PAGE under reducing conditions to in identify the individual
components of each band (Fig. 4C). The immune complexes at
the top of the gel dissociated into the same subunit band pattern as
already seen in Fig. 4B. The band with
Mr = 52,000 turned out to represent an
*-dimer, whereas the Mr = 37,700 band
consisted of -p-dimers. In the pool fraction 3, in addition to
the free -subunit, a pre--band (see also below) was also clearly
visible (Fig. 4B). In the case of the -dimers the
pre--band was missing (Fig. 4B, pool 2).
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 4.
Separation of contained in -dimer and
free -subunit by gel filtration.
Unperturbed JEG-3 cells were pulse-chase labeled as described under
"Experimental Procedures" in detail. 2 ml of cell lysate were
applied to a Sephadex G150 column (0.75 × 100 cm). Further
details are described above. Fractions of 750 µl were collected,
pooled as indicated (A), purified by immunoprecipitation
with anti- and anti- antibodies, and analyzed in SDS-PAGE under
nonreducing and reducing conditions (B). Under less
stringent elution conditions, (more rapid elution at low temperature)
in the case of the unreduced samples, only the immune complexes at the
top of the gels are visible (bands with Mr = 53,000 and 37,000 were missing; data not shown). The bands of the
unreduced samples were cut from the gel, eluted by diffusion, reduced,
and separated in SDS-PAGE to identify their individual constituents
(C). Similar results were obtained in a total of six
independent experiments.
|
|
Effect of DTT on the N-Glycosylation of the Free
-Subunit--
In the free -subunit fraction (after having
removed -dimers by immunoprecipitation), besides the regular
-subunit (Mr = 20, 500), a molecular variant
designated as pre- was observed (apparent Mr = 18,300; Fig. 5). It represents an
-subunit with only one of the two carbohydrate residues attached to
the protein. Both -subunit forms collapsed into one band with a
molecular weight of 10,000 after digestion with peptide
N-glycanase F (data not shown). Interestingly, the
pre--subunit was missing in the DTT-treated JEG-3 cells (Fig.
5B). This seems to indicate the accelerated linkage of the
second N-linked carbohydrate residue of the -subunit when
the disulfide bridges are not formed. In the case of
N-glycosylation of the -subunit a similar process was not
observed.
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 5.
Facilitated N-glycosylation
in the free -subunit fraction in the presence
of DTT. JEG-3 cells were pulse labeled for 2 min and chased as
indicated in the absence (A) or in the presence of 5 mM DTT during the first 15 min of chase (B).
Cell lysate was immunoprecipitated with anti-. Similar results were
observed in two experiments.
|
|
| |
DISCUSSION |
We have studied the effect of a prevention of the disulfide bridge
formation on the subunit association in hCG biosynthesis. This was
achieved by shifting the oxidizing into a reducing milieu in the ER by
means of DTT. The DTT concentrations used in the present investigation
do not influence protein synthesis and translocation along the
secretory pathway (36). In the presence of DTT disulfide bridge
formation is delayed and reinitiated post-translationally without a
loss in efficiency after removal of DTT (11). The hCG contains as much
as 11 disulfide bridges (five in the -subunit and six in the
-subunit). The significance of the sequence of disulfide bridge
formation for the folding and the association of the subunits is not
yet fully elucidated. In this context the effect of a replacement in
the -subunit of individual pairs of cysteine by alanine residues was
thoroughly studied (32, 41). This methodology is of great value.
However, at least theoretically the possibility cannot be excluded that
the elimination of single or even more disulfide bridges might have an
impact on the conformation of the subunit as well as on the folding and
association with the -subunit. Reduction and reoxidation studies of
the wild-type subunits in vivo by the use of DTT provides an
independent way to study the interdependences between disulfide bridge
formation, folding, subunit association, and
N-glycosylation. In this publication, we have addressed the
question whether disulfide bridge formation is an essential requirement
for subunit association. To our knowledge, no experiments have been
published carried out to study the in vivo effects of DTT on
the hCG subunit folding pattern and subunit association.
Precursors of the -Subunit--
In unperturbed JEG-3 cells, the
mature hCG--subunit is formed through well defined intermediates
that seem to acquire distinct conformations that allow separation in
the SDS-PAGE. Obviously these intermediates are defined by the
formation of disulfide bridges (Fig. 1). These -subunit precursors
seem to resemble very closely or are even identical to the pattern
observed and extensively studied in JAR cells (31, 37-39) and Chinese
hamster ovary cells transfected with hCG subunits (32, 40, 41). We have
purified the -subunit intermediates with the same antibodies (G10)
as used in the literature (30-33, 39). In JAR cells, the following
sequence of -subunit intermediates, leading to a form that combines
with the -subunit, was described: p0 
p1early p1late p1early p2free p2combined-early p2combined-late (30-32).
This model implies that the subunit association is achieved when the
-subunit has attained an association competent form as a result of
extensive conformational changes as well as disulfide bridge formation
as requirements. Here, we did not intend to study the maturation of the
-subunit in JEG-3 cells but to follow the DTT-induced changes during
reduction and reoxidation. DTT has a marked influence on the half-lives
of the subunit precursors. The -subunit precursors showed very
different stability in the presence of DTT. The p1 was
most sensitive to a reducing environment because it disappeared
completely within 2 min of in vivo treatment with DTT but
again reappeared within 5 min after a change to DTT-free culture medium
(Fig. 3A). In the presence of DTT during the chase, the
p2 disappeared rapidly, whereas it was rather stable in
the absence of DTT (Table I). The absence
of immunoreactive degradation products and the increase in half-life of
p0 indicate that p2 is most probably
converted into p0 in the presence of DTT. This implicates that the existing disulfide bridges in p2 are
reduced in vivo by DTT because it was also observed in
studies with influenza hemagglutinin (11). This is in contrast to other
cases where DTT did not disrupt existing disulfide bridges in rotavirus
proteins (42).
View this table:
[in this window]
[in a new window]
|
Table I
Half-lives of disappearance of hCG subunit precursors in JEG-3 cells:
influence of in vivo reduction of disulfide bridges with DTT
|
|
Precursor Forms of the -Subunit--
In the cell lysates
the -subunit could be detected as part of an -dimer
(coprecipitated with anti- antibody G10) and as free -subunit as
precipitated by specific anti- antibodies (Fig. 4). The apparent
molecular weight of the -subunit was reduction-insensitive. Reduction-sensitive intermediate forms, like in the case of the -subunit, were not found with the protocol used. Interestingly, a
certain amount of the monoglycosylated -subunit (pre-) was present in the unperturbed cells, whereas it was missing in the presence of DTT (Fig. 5). There is a distinct connection between N-glycosylation and protein folding (43-47). Extensive
investigation have been carried out to characterize the effect of
N-glycosylation on the folding of the hCG--subunit. Feng
et al. (48) have shown that the N-glycans
facilitate the correct disulfide bridge bond pairing. Association of a
mutant -subunit lacking the N-glycans with the
-subunit accelerates the formation of certain disulfide bridges.
This was interpreted as a hint on the chaperone-like function of the
-subunit for the folding pathway of the -subunit (46). The
-subunit might also act as a chaperone facilitating the
N-glycosylation of the -subunit because the
pre--subunit was only present in the free fraction and was not
observed in the -dimer fraction (Fig. 4). However, presently we
cannot rule out the possibility that only a completely
N-glycosylated -subunit is associated with the
-subunit.
Formation of -Dimer--
Our gel filtration
experiments clearly show the presence of -dimers that were
coprecipitated with the specific anti- antibody. The free
-subunit was not precipitated by this antibody. It is remarkable to
note that after a short pulse of 2 min, a distinct amount of
-dimer has already formed (Fig. 1). Unexpectedly, we found an
-complex present in JEG-3 cells even if the DTT has already been
present during the pulse (Fig. 3) or even if it was added together with
the Met/Cys-deficient medium prior to the pulse (not shown).
Based on these results, it seems that DTT acts rapidly and
quantitatively so that during the pulse (in the presence of DTT) the
subunits remain completely in a reduced form. In vitro
translation experiments have indicated that the -dimer formation
is a late event when all p0 was already converted into
other intermediates. It was shown that the p2-uncombined
intermediate acquires an association competent form giving rise to an
p2 complex after having evolved from
p0 through p1-early and
p1-late (31, 40, 41, 46). All these intermediate forms
are characterized by the sequence of disulfide bridges formed.
Moreover, it was demonstrated that no association takes place in
Chinese hamster ovary cells transfected with a -subunit mutant
(Cys100 replaced by alanine). This was interpreted to mean
that the formation of individual disulfide bridges is the prerequisite
for subunit association (32).
Our experiments clearly suggest that the subunit precursors may acquire
an association-competent conformation much earlier to give rise to the
-dimer in the JEG-3 cells, even if no disulfide bridges have
formed. The post-translational formation of the disulfide bridges after
removal of DTT seems to yield the same subunit precursors as in the
untreated cells. It should be emphasized that the dimer formed
in the presence of DTT represents an immature form that has not yet
expressed typical epitopes of the native hCG. This may be concluded
from the fact that this -dimer was precipitated by the polyclonal
antiserum G10 but not by three different monoclonal antibodies directed
against epitopes that are present only on hCG and not on the free
subunits (not shown).
Recombination of mature subunits in vitro does not require
the addition of reducing agents (49, 50). Association of the subunits
in the cell and in the test tube is at least initiated very differently
because it begins in vivo from partially folded subunits,
whereas in vitro the subunits are completely folded and
oxidized. Recombination of mature hCG subunits in vitro
proceeds slowly and appears to be more complex than a second order
process (51), whereas in vivo association occurs rapidly
(31, 51). Several studies have shown that during recombination in
vitro, -intermediates are rapidly formed and only partially
share the physical, biologic, and immunologic properties of the native
hormone (52, 53). These intermediates regain the full properties of the
hormone in a slower reaction (53-55). This process is also probably
responsible for the inability of the -dimer formed in the
presence of DTT to react with quaternary structure-specific anti-hCG
monoclonal antibodies (see above). A detailed study on the folding of
the bovine pancreatic trypsin inhibitor revealed that the kinetics of
folding are highly dependent on the presence of unstable folding
intermediates that are mostly undetectable because of their short
half-lives. In the course of folding, disulfide bridges are formed
randomly and are later rearranged spontaneously by intramolecular
thiol-disulfide exchange (56). This may be the reason for the marked
increase in the association rate of hCG subunits in vitro in
response to the addition of protein disulfide isomerase (31). In the
case of -dimerization, the situation is complicated by the
presence of cystine knots (57) and the "seat belt" structure that
is formed by a 21-amino acid arm of the -subunit that "embraces"
the -subunit (58, 59). Native hCG can be readily dissociated into
its subunits by reduction and alkylation (60), but the usual mode of
subunit preparation is by dissociation at low pH in the presence of
urea (60, 61). Prevention of disulfide bridge formation in
vivo does not interfere with heterodimer formation, as
demonstrated above. However, in the cell, association-competent folding
states of the reduced subunits (in the presence of DTT) seem to
prevail, whereas unfolding of the mature subunits by reduction and
alkylation probably results in a random structure that does not allow recombination.
A transient formation of aggregates linked by disulfide bridges, as in
the case of vesicular stomatitis virus G-protein (14), was not
observed. In the case of the -subunit, the N-linked
carbohydrate residues seem to be attached much faster when disulfide
bridge formation was prevented by DTT. It was also shown that in the further pathway (trimming of the carbohydrate residues, terminal glycosylation in the Golgi apparatus) the Asn78-linked
oligosaccharide of the -subunit is processed much more slowly than
the Asn52-linked residue (62). Our experiments seem to
indicate that already the attachment of at least one of the two
carbohydrates of the -subunit is sterically hindered by the folding
of the polypeptide chain. Moreover, this part of glycosylation is
possibly facilitated by the presence of the -subunit that acts in a
chaperone-like fashion.
and
TOP
ABSTRACT
INTRODUCTION
Present address: Hormone Biochemistry Laboratory, Inst. of Self
Organizing Systems and Biophysics, North-Eastern Hill University, Permanent Campus, Shillong-793022, Meghalaya, India.
