| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Articles |
Department of Life Science and Biotechnology Research Center, Pohang University of Science and Technology, Pohang; and the Department of Internal Medicine, Seoul National University College of Medicine (C.B.Y.), Seoul, Korea
Address all correspondence and requests for reprints to: Dr. Chi-Bom Chae, Department of Life Science and Biotechnology Research Center, Pohang University of Science and Technology, Pohang 790–784, Korea. E-mail: cbchae{at}vision.postech.ac.kr
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Ever since the complementary DNA (cDNA) for TSHR has been available, overexpression of the receptor has been attempted in a variety of systems (9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19). The TSHR produced in Escherichia coli does not interact with either Graves’ autoantibody or TSH (9, 14, 15), suggesting that either a posttranslational modification, such as glycosylation, or correct folding of TSHR is important for its function. However, it was reported that the soluble extracellular domain of TSHR (TSHRE) produced in the insect cells infected with recombinant baculovirus is recognized by Graves’ IgG (12, 13, 16, 18, 19). However, conflicting results were obtained with regard to the binding of TSH to TSHRE (16, 18, 19).
In this study, TSHRE encoding amino acids 1–413 was overexpressed in HeLa cells using recombinant vaccinia virus. This expression system is suitable for the production of soluble, posttranslationally modified proteins in a multitude of animal and human cells (20). The purified receptor was used for development of an ELISA assay for the binding of Graves’ IgGs to TSHRE. Using this assay system, we have identified hexapeptide sequences that interfere with the binding of Graves’ IgG to TSHRE from a vast mixtures of random sequences of hexapeptides. The peptide sequences inhibited not only the binding of Graves’ IgG to TSHRE, but also the cAMP synthesis induced by the IgGs of some of the patients with Graves’ disease. We believe that the peptide sequences most likely mimic the epitopes recognized by a clone of Graves’ IgGs that stimulates the TSHR.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Expression of TSHRE domain protein in E. coli
The extracellular region of hTSHR corresponding to amino acids
1–413 was copied from a plasmid carrying the full-length hTSHR cDNA
(from Dr. Kazuo Tahara, University of Chiba, Chiba, Japan) by PCR with
the following primers; forward primer, 5'-CGggatccATGAGGCCGGCGGAC-3';
and reverse primer, 5'-CGggatccTTAGTAGCCCATTATCTGCTT-3'. The
1.3-kilobase fragment produced by PCR was digested with
BamHI and subcloned into the expression vector pGEX-KG (21).
The protein was produced as a fusion protein with glutathione
S-transferase (GST) by induction with 1 mM
isopropyl-ß-D-thiogalactoside (IPTG). The expressed
GST-TSHRE fusion protein was subjected to SDS-PAGE and immunoblot
analysis for characterization of the anti-TSHR peptide antibody.
Construction of recombinant vaccinia virus transfer vector,
pSC11-TSHRE
pSC11-TSHRE was constructed by inserting the TSHR gene
corresponding to amino acids 1–413 into the SmaI site of
vaccinia recombination vector pSC11 (22). The TSHRE gene fragment was
copied from the full-length hTSHR cDNA by PCR.
Production of recombinant vaccinia virus
African green monkey CV-1 cells were grown to 70% confluency in
six-well plates and infected with 0.05 plaque-forming unit of vaccinia
virus/cell. These cells were grown in DMEM (Life Technologies, Grand
Island, NY) containing 10% FBS (Life Technologies). After 2 h at
37 C, the cells were transfected with 0.5 ml DNA precipitates of
pSC11-TSHRE. The DNA precipitates were prepared as follows. Ten
micrograms of the pSC11-TSHRE DNA and 1 µg wild-type virus DNA were
mixed in 250 µl HEPES-buffered saline, and the DNA mixture was
precipitated by the addition of 250 µl CaCl2 to a final
concentration of 125 mM. Cells were harvested 2 days later,
and viruses were released from the cells by three cycles of freezing
and thawing.
Selection of recombinant vaccinia TSHRE virus
Recombinant vaccinia TSHRE virus was selected based on the fact
that homologous recombination would result in an insertional
inactivation of vaccinia virus thymidine kinase (TK) gene. Human
osteosarcoma HuTK- 143B cells (80% confluent) in a
six-well plate were infected with serially 10-fold diluted virus
(10-1–10-3 M). HuTK- 143B cells
were grown in Eagle’s MEM with 10% FBS. After 2 h of infection,
4 ml medium containing 1% low melting point agarose, 2.5% FBS, 1
mM glutamine, and 25 µg of 5-bromodeoxyuridine were
overlaid onto each monolayer. After 2 days, each plate was overlaid
with 0.5 ml of medium containing 1% low melting point agarose, 2.5%
FBS, 1 mM glutamine, 25 µg 5-bromodeoxyuridine, and 330
µg X-gal for the assay of ß-galactosidase activity. Blue plaques
were picked from the plate and suspended in 0.5 ml of medium
supplemented with 10% FBS. After freezing and thawing, the plaque
assay was performed twice. Finally, homogeneous recombinant virus was
amplified in HuTK- cells.
Overexpression and purification of TSHRE
HeLa S3 cells were grown in Spinner’s MEM with 6% horse serum
(Life Technologies). A suspension culture of HeLa cells (5 x
108) was infected at a multiplicity of infection of 30
plaque-forming units/cell with recombinant TSHRE virus and incubated
for 20 h at 37 C. Cells were collected by centrifugation
(1000 x g, 15 min) and suspended in lysis buffer (20
mM Tris, pH 7.4; 1 mM EDTA/EGTA; 0.1
mM dithiothreitol; and 2 mM
phenylmethylsulfonylfluoride) containing 2 M KCl. The cell
suspension was sonicated with a Branson sonifier model 450 (1 in. D
horn; 10–20 strokes). After 2 h at 4 C, the lysates were
centrifuged at 15,000 x g for 30 min. A total of 90 mg
protein in the supernatant fraction was loaded onto a TSK phenyl 5PW
HPLC column (21.5 x 150 mm; Pharmacia Co., Milwaukee, WI)
preequilibrated with lysis buffer containing 2 M KCl. The
bound protein was eluted for 60 min with a linear gradient from 0–2
M KCl at a flow rate of 5 ml/min. The column eluate was
monitored at 280 nm. Each fraction was subjected to immunoblot analysis
to monitor TSHRE protein. Fractions showing positive signal were pooled
and dialyzed against 20 mM Tris-HCl (pH 7.6) containing 1
mM EDTA/EGTA. A total of 8 mg protein of the dialyzed
sample was applied to a Mono-Q column (5 x 50 mm; Pharmacia Co.)
preequilibrated with 20 mM Tris-HCl (pH 7.6) containing 1
mM EDTA/EGTA and 0.1 mM dithiothreitol. The
protein was eluted for 40 min with a gradient from 0–0.4 M
NaCl and then for 20 min with 1 M NaCl at a flow rate of 1
ml/min. The column eluate was monitored at 280 nm. Each fraction was
subjected to immunoblot analysis to monitor TSHRE protein.
SDS-PAGE and immunoblot
SDS-PAGE was performed according to the method of Laemmli (23).
Proteins were separated on 8% SDS-PAGE and stained with Coomassie
brilliant blue G250. For immunoblot analysis, proteins were transferred
to nitrocellulose membrane and incubated for 30 min with blocking
buffer (1% BSA-PBS), then with 1:2000 diluted rabbit anti-TSHR peptide
(amino acids 32–40) antibody at least for 4 h. After washing
three times, blots were incubated for 2 h with 1:2000 diluted goat
antirabbit IgG conjugated with horseradish peroxidase
(Bio-Rad Laboratories, Richmond, CA). After washing five times, the
blots were developed using 4-chloro-1-naphthol as the substrate.
Patients’ sera
The sera of several patients with Graves’ disease were used to
assess binding of IgG to the purified TSHRE. These sera were diagnosed
by established clinical and biochemical criteria, such as the
stimulatory activity for cAMP synthesis (24) and inhibition of TSH
binding to TSHR (25). Igs were prepared by using the protein A
purification method (Bio-Rad). Igs from normal healthy individuals were
used as controls.
Enzyme-linked immunosorbent assay (ELISA)
ELISA with purified TSHRE.
Polystyrene 96-well microtiter
plates (Nunc immunoplate Maxisorb; GIBCO-BRL, Paisley, Stratchclyde,
UK) were coated with 50 µl purified TSHRE at 2 µg/ml in PBS (pH
7.4) overnight at 4 C. The wells were blocked for nonspecific binding
sites with 5% milk in PBS at room temperature for 1 h. The plates
were washed with PBS containing 0.05% Tween-20 (pH 7.4). The following
were then added in succession, with washing between each step: 50 µl
protein A-purified Ig from patients with Graves’ disease (50 µg/ml
in blocking buffer) at room temperature for 2 h, 50 µl goat
antihuman Ig(G+M+A) conjugated to horseradish peroxidase (diluted
1:2000 in 1% BSA/PBS) at room temperature for 2 h, and 50 µl
substrate including o-phenyldiamine dihydrochloride (4
mg/ml) and 0.003% H2O2. The reaction was
stopped by adding 50 µl 3 N hydrochloric acid. Absorbance
was measured at 492 nm in a microplate reader (Titertek Multiskan,
Flow Laboratory, Rockville, MD).
Competitive ELISA: the screening method used for the
identification of inhibitory peptides.
ELISA assay was performed
in 96-well polystyrene microtiter plates, as described above, in the
presence of peptides. After blocking each well with milk, 5 µl of
each peptide mixture of the Positional Scanning Synthetic Peptide
Combinatorial Library (PS-SPCL; see below) at a fixed concentration
were added, followed by 5 µl protein A-purified Ig from patients with
Graves’ disease (1 mg/ml) and 40 µl 5% milk in PBS. The remaining
steps were the same as those for the ELISA described above.
For determination of the percent inhibition of binding of IgG to TSHRE by peptides the following formula was used: % inhibition = 100 x [(antibodies in the absence of peptides - antibodies in the presence of peptides)/antibodies in the absence of peptides]. The absorbance was corrected for the background absorbance.
Preparation of PS-SPCL
Peptide libraries were synthesized according to the protocol of
Houghton et al. (26) and Pinilla et al. (27, 28).
Briefly, PS-SPCLs, consisting of six residues with free N-terminals and
amidated C-terminals, were synthesized. A single position in each
peptide mixture was individually and specifically defined with 1 of 18
natural L-amino acids (cysteine and tryptophan excluded), whereas the 5
remaining positions consisted of mixtures of the same 18 amino acids.
Defined positions are represented by O and the positions with mixed
amino acids are represented by X. The 6 sets of PS-SPCLs (total of 108
pools) are represented by the formula: OXXXXX-NH2,
XOXXXX-NH2, XXOXXX-NH2, XXXOXX-NH2,
XXXXOX-NH2, XXXXXO-NH2. Libraries of peptides
were constructed on Rapidamide (DuPont, Boston, MA) resin beads as
previously described (29). The amount of each amino acid used to yield
approximately equimolar coupling was determined empirically.
Side-chains were deprotected with a mixture of trifluoroacetic
acid-ethanedithiol-water-thioanisole (90:5:4:1, vol/vol/vol/vol). Each
of the 108 peptide mixtures was individually extracted with water,
lyophilized, and dissolved in water at a final concentration of 16
nM for a peptide sequence in each pool.
Purification of individual peptides
Peptide pool X1X2FDDA was loaded on an
anion exchange AX 300 analytical column (Synchrom, Lafayette, IN)
preequilibrated with 10 mM potassium phosphate (pH 7.4).
The bound peptides were eluted for 30 min with a gradient from 0–0.15
M KCl and then for 10 min with 0.3 M KCl at a
flow rate of 1 ml/min. The column eluates were monitored at 215 nm. For
further purification, each fraction was applied to a C18
reverse phase preparative column (Vydac, Hesperia, CA). This column was
preequilibrated with deionized water containing 0.05% trifluoroacetic
acid. The peptide was eluted for 60 min with a gradient from 0–50%
acetonitrile (ACN) and then for 10 min with 100% ACN. The ACN also
contained 0.05% trifluoroacetic acid. The purity and identity of the
peptide were determined by amino acid composition analysis and peptide
sequencing. Sequence analysis was carried out according to the
manufacturer’s procedure using an Applied Biosystems 471A
protein/peptide microsequencer (Applied Biosystems, Foster City, CA).
Amino acid analysis was performed according to the Pico-Tag system
(Millipore Corp., Milford, MA). The data were also used for
determination of the concentrations of peptides.
Measurement of cAMP
Chinese hamster ovary (CHO) cells producing a high level
of hTSHR (6) were used within 10 subcultures. The cells were maintained
under 5% CO2 at 37 C and cultured in Ham’s F-12 mixture
medium supplemented with 10% FBS and 1 mg/ml Geneticin (G418; Life
Technologies). The cells were passaged every 3 days and provided with
fresh medium every day (6). The hTSHR-producing CHO cells were seeded
at 2 x 105 cells/well in 24-well plates for 24 h
before use. After the culture medium was removed, cells were washed
with NaCl-free isotonic Hanks’ Balanced Salt Solution (5.4
mM KCl, 1.3 mM CaCl2, 0.8
mM MgSO4.7H2O, 0.3 mM
Na2HPO4, 0.4 mM
KH2PO4, 0.1% glucose, 20 mM HEPES,
1.5% BSA, and 274 mM sucrose, pH 7.5). A peptide was
preincubated with either the Graves’ IgG (0.4 mg) or TSH (20 µU;
Sigma Chemical Co., St. Louis, MO) in 200 µl NaCl-free isotonic HBSS
containing 0.5 mM 3-isobutyl-L-methylxanthine
for 1 h at 37 C, and the mixtures were added to the washed CHO
cells. After 3 h of incubation at 37 C, 300 µl absolute ethanol
were added to the plate to extract the cAMP from the cells. The plate
was incubated for 3 h at -20 C. The cells in ethanol were scraped
into a 1.5-ml microcentrifuge tube and centrifuged for 10 min at 4 C at
2500 x g. The supernatant was evaporated to dryness in
a Speed-Vac (Savant Instruments, Farmingdale, NY). After evaporation,
the residues were dissolved in 0.15 ml 50 mM Tris-HCl (pH
7.5) and 4 mM EDTA. After centrifugation to clarify the
samples, the amount of cAMP formed was determined by a commercial RIA
kit (TRK432, Amersham, Aylesbury, UK).
TSH-binding inhibitory Ig (TBII) assay
Graves’ IgG inhibits the binding of labeled TSH to solubilized
thyroid membrane. The TBII assay (TBII Assay Kit, RSR Limited, Pentwyn,
Cardiff, UK) mixture contained 50 µl solubilized porcine thyroid
membrane fraction, 100 µl [125I]TSH (0.36
kilobecquerels), and 50 µl serum or IgG from Graves’ patients. The
tube contents were mixed vigorously by vortexing for about 5 sec. After
2-h incubation at room temperature, 1 ml 16.5% polyethylene glycol-1.6
M NaCl were added to each tube and incubated for 30 min at
room temperature. Each tube was centrifuged at 1500 x
g for 30 min at 4 C, and the pellet was counted for
radioactivity in a 
-scintillation counter.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Characterization of the anti-TSHR peptide antibody using the fusion
protein of GST-TSHRE
To identify the TSHR molecules during purification, we produced
rabbit antibodies against amino acids 32–40 of hTSHR. To characterize
the antibodies, hTSHRE was produced in E. coli as a fusion
protein with GST. The fusion gene was under the control of the
lac promoter, and the predicted size of the fusion protein
was 72,000 daltons. When the cells were treated with 1 mM
IPTG, a polypeptide of the predicted size appeared (see lane 4 of Fig. 1
). No such protein was present in the cells carrying
the cloning vectors or the cells not treated with IPTG. When the gel
was subjected to immunoblot analysis with the rabbit antiserum for the
TSHR peptides, the 72,000-dalton polypeptide overproduced in the
presence of IPTG was the principal peptide recognized by the antibody
(Fig. 1B, lane 4). A small amount of polypeptide of 55,000 daltons was
also detected by the antibody. This is most likely the receptor peptide
released by proteolysis of the fusion protein. It is known that the
fusion protein prepared with GST is prone to proteolysis. The antibody
also specifically reacted with the TSHR protein produced as a fusion
protein with E. coli maltose-binding protein (data not
shown). Thus, it appears that the rabbit antibody raised against the
TSHR peptide specifically recognizes the TSHR. In the cells transformed
with the cloning vector, IPTG treatment increased the background level
of proteins larger than 60,000 daltons recognized by the antiserum
(Fig. 1B, lane 2). However, the antiserum was of sufficient quality to
allow us to identify TSHRE peptides unambiguously during purification
from HeLa cell extracts.
|
). The extracellular region of TSHR encoding amino
acids 1–413 is 46,000 daltons. Therefore, the 63,000-dalton
polypeptides appear to be the TSHRE with a different glycosylation
pattern (data not shown). The identity of the polypeptide as TSHR was
further confirmed using a hTSHR monoclonal antibody provided by J. P.
Banga (16). The level of expression was maximal 20 h after
infection. The time-course expression of the receptor proteins in the
cells infected with the recombinant virus was similar to that observed
for the production of E. coli chloramphenicol
acetyltransferase under the control of the same viral promoter as that
used in this study (32). A substantial amount of TSHRE was also present
in the insoluble fraction when the cells were lysed in a hypotonic
medium (Fig. 2, lane 1). A small proportion of this protein could be
extracted with 2 M KCl (data not shown).
|
, lane 3).
The major band reacted with the TSHR antibody (data not shown). To
assess whether the 63-kDa form of the TSHRE was a glycosylated form,
the purified receptor was treated with N-glycosidase F. This
resulted in the appearance of a protein band of about 50 kDa,
suggesting that the 63-kDa form is most likely the glycosylated TSHRE
(data not shown). It appears that we purified one of several
glycosylated forms of TSHRE. The detailed characterization of TSHRE
will be the subject of a future report.
|
). To determine whether the binding truly represents
the interaction of Graves’ IgG with TSHR, an IgG preparation was
allowed to bind to the immobilized TSHRE, and the unbound supernatant
was removed and used for TBII assay, as shown in Fig. 5.
Graves’ IgG interferes with the binding of TSH to thyroid membrane,
and this assay is routinely used to measure the activity of Graves’
IgG. The results in Fig. 5 show that the IgG remaining after
preincubation with the immobilized TSHRE does not inhibit the binding
of TSH to thyroid membrane as effectively as the Graves’ IgG added to
the well coated with blocking agent (BSA) only. The modest reduction of
the inhibition of TSH binding by preincubated Graves’ IgG is most
likely due to the presence of other Graves’ IgG clones that do not
bind to the immobilized TSHRE. The Graves’ IgG that does not bind to
the immobilized TSHRE inhibits the binding of [125I]TSH
to thyroid membrane even after prior binding to the immobilized TSHRE.
This suggests that the IgG that binds to the immobilized TSHRE
preparation indeed interacts with TSHR.
|
|
Identification of the peptides that inhibit the binding of Graves’
IgG to TSHR
One of the 2 IgG preparations that showed the highest
binding activity (Fig. 4
) was used for identification of the peptide(s)
that inhibits the interaction of Graves’ autoantibodies to the
purified TSHRE. For identification of such peptides, we used mixtures
of random sequences of hexapeptides. For ease of identification of the
peptide sequences, a PS-SPCL originally developed by Houghten et
al. (26) was adopted for our studies. In this scheme, each
position of a hexapeptide is fixed with a known amino acid, and the
other 5 positions are fixed with random mixtures of 20 amino acids. For
each position, 20 peptide pools are possible, and a total of 120
peptide pools can be made. By assaying 120 pools, the amino acids
important at each position of active peptides can be determined. Based
on this information, reiterative synthesis of peptides can be conducted
to identify the sequence of the most active peptides.
In our studies, we prepared a total of 108 peptide pools of PS-SPCL
lacking cysteine and tryptophan. The presence of cysteine results in
inter- and intracross-linking of peptides, and tryptophan-containing
peptides tend to have low solubility. The PS-SPCL was screened by
competitive ELISA for inhibition of the binding of autoantibody to
TSHRE. A graphical representation of the screening results is shown in
Fig. 6. For the peptide pools with the first position
known (OXXXXX), glutamic acid inhibited antibody binding most
effectively. Aspartic acid and glutamine were the second active amino
acids at this position. Methionine, asparagine, and tyrosine also
showed considerable inhibitory activity (50%). For the second
position, SPCL (XOXXXX), many amino acids were found to be active. The
assay was repeated with diluted peptide mixtures to resolve the
differences in inhibitory activities among peptides. In this case,
aspartic acid was the most effective amino acid, followed by glutamic
acid, histidine, asparagine, arginine, threonine, and tyrosine. With
the peptide pools with the third position known (XXOXXX), the peptides
containing phenylalanine, tyrosine, and, to a lesser extent, isoleucine
were found to inhibit the binding of the antibody. The result obtained
for the fourth position (XXXOXX) showed that several amino acids, such
as aspartic acid, glutamic acid, and asparagine, are the candidate
amino acids. For the fifth position, SPCL (XXXXOX), the peptide pool
containing aspartic acid showed the highest activity. However, alanine,
glutamic acid, methionine, and serine also effectively inhibited the
antibody binding. The peptide pools containing asparagine, glutamine,
threonine, and valine showed inhibitory activity greater than 50%. The
profile of the sixth position SPCL (XXXXXO) was similar to that of the
second position SPCL: that is, no amino acid with outstanding
inhibitory activity. Therefore, the assay was also repeated with
10-fold diluted peptide mixtures (Fig. 6). Alanine was the most active
amino acid. Phenylalanine and valine also showed high inhibitory
activity. In theory, the peptide sequence (EEFDDA) composed of the
amino acids showing highest inhibitory activity at each of the six
positions would yield the most effective inhibitory peptide. However,
in the case of searching the ligands for the opioid receptor, this was
not found to be true (33). If more than one amino acid is effective for
a given position, it is impossible to define the sequence of a certain
individual peptide. Moreover, the number of all possible combinations
of the amino acids active at each position is too large to be
synthesized. For this reason, the peptide pools with the amino acids
showing considerable inhibitory activity at each position were
reexamined at different concentrations (data not shown). On the basis
of these results, several peptides were synthesized to identify highly
active peptides. Six peptide mixtures were constructed based on
the following criteria; for each position, amino acids showing 40–60%
inhibition at 0.16 nM were chosen. For the first position,
glutamic acid, methionine, and tyrosine were selected for their
side-chain differences compared with other candidate residues. For the
second position, glutamic acid and threonine were chosen. Histidine was
also chosen for its chemically differing side-chains at this position.
Only phenylalanine was selected for the third position. Considering the
charge of side-chains, aspartic acid and glutamic acid were chosen for
the fourth position. At the fifth position, aspartic acid was the best
choice. Methionine and serine were also chosen, because they showed
40% inhibitory activity. For the sixth position, only alanine was
chosen because of the highest inhibitory effect. In summary, peptides
were synthesized as follows. For positions 3–6, the amino acids were
defined. The remaining two positions, positions 1–2, were composed of
mixtures of the selected amino acids. This generated six hexapeptides
with mixtures of three amino acids at positions 1 and 2: X1 X2 F D D A,
X1 X2 F E D A, X1 X2 F D M A, X1 X2 F E M A, X1 X2 F D S A, and X1 X2 F
E S A, where X1 was a mixture of E, M, and Y, and X2 was a mixture of
E, H, and T.
|
). The results showed that XXFEDA and XXFDDA were
significantly more effective than the other four peptide mixtures. Each
peptide in a mixture was purified by HPLC on an anion exchange AX 300
column and a C18 reverse phase preparative column. Only the
XXFDDA pool was fractionated for further study, because there was
little difference in the inhibitory activities of XXFEDA and XXFDDA.
When the peptide in each column fraction was assayed, EEFDDA, EHFDDA,
and ETFDDA were found to be the most effective for inhibition of the
binding of Graves’ antibody to TSHRE (Fig. 7B).
|
). At this concentration, the
peptides show a low inhibitory effect (
10%) on the action of TSH.
As all three peptides show similar effects on the synthesis of cAMP in
response to Graves’ IgG, the second position of the hexapeptide
appears to play a minor role. To investigate the possible specificity
of the peptides for the IgG of different Graves’ patients, we
investigated the effects of the peptide on the synthesis of cAMP
induced by the IgG taken from eight different Graves’ patients. The
IgGs used in Fig. 9, except for the IgG (Fig. 9A), are
different from the IgG used for the selection of the IgGs for use in
the screening of peptide libraries (Fig. 4). The IgG used in Fig. 4
were exhausted during the course of this work. The results in Fig. 9
show that the peptide affected the action of the IgGs of only four
patients (Fig. 9, A type). Furthermore, the maximal inhibitory activity
was between 50–70% for most IgG.
|
|
). However, no apparent
correlation was found for the three parameters.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The major aim of our work was identification of the inhibitors of Graves’ stimulating autoantibodies from peptide libraries using the TSHRE. We produced hTSHRE in a human cell line (HeLa) with the aid of recombinant vaccinia virus and found that some of the Graves’ stimulating autoantibodies interact with the receptor. Preliminary experiments suggest that TSHRE is heterogeneous in terms of the interaction with Graves’ IgG. Depending on the purification scheme, TSHRE preparations show different specificity toward the IgG of different Graves’ patients (our unpublished results). This could be due to differences in folding or glycosylation patterns among TSHRE produced in the cells. Therefore, the results presented in this report are most likely based on one form of TSHRE recognized by one type of Graves’ stimulating autoantibodies. For this reason, no attempt has been made to fully characterize the purified TSHRE and compare it with the TSHRE produced in the baculovirus system. This will be the subject of a future report.
The hTSHRE produced in HeLa cells corresponds to the region encoding amino acids 1–413 (TSHRE) and the protein has additional Gly-Gly-Ser-Thr-Ser-Ser-Ser at the C-terminus. However, this additional sequence may not affect the structure of the extracellular domain of the receptor, as it is located at the C-terminus. The TSHRE had an apparent mass of 63,000 daltons, which is larger than the expected size of 46,000 daltons. The same extracellular region expressed in insect cells via recombinant baculovirus had the same 63,000-dalton size due to glycosylation (16, 17). Our preparation also appears to be glycosylated because the treatment with N-glycosidase F reduced the apparent mass to about 50,000 daltons (data not shown). The receptor protein preparation used in this study was approximately 70% pure, as determined from the staining pattern of SDS-PAGE and the reactivity with antibodies.
As is the case with TSHRE expressed in insect cells (16, 18, 19), the protein expressed in HeLa cells is also retained within the cells despite the presence of the natural signal sequence at its N-terminus, and a substantial amount remained in the insoluble form, perhaps due to incomplete or incorrect glycosylation or incorrect folding.
The purified TSHRE protein was used for an ELISA assay to investigate
the binding of Graves’ IgG to the receptor. As shown in Fig. 4, the
IgG from 10 Graves’ patients showed a wide variability in the degree
of binding to the TSHRE. Two preparations showed 3-fold higher binding
activities compared to the normal background. To ascertain that the
apparent binding truly reflects binding of Graves’ IgG to the
receptor, the IgG sample taken after preincubation with the immobilized
TSHRE was tested for the inhibition of binding of labeled TSH to
thyroid membrane (TBII assay). The results showed that the IgG that
binds to the immobilized TSHRE showed a reduced inhibition of TSH
binding to thyroid membrane after preincubation with TSHRE. On the
other hand, the Graves’ IgG that does not bind to the immobilized
TSHRE inhibited TSH binding to the same degree as the nonincubated IgG
after preincubation with TSHRE. This strongly suggests that the
apparent binding of Graves’ IgG to the immobilized TSHRE in ELISA
assay is due to the binding to TSHRE and not to contaminating proteins.
The result also suggests that the immobilized TSHRE assumes a structure
recognizable by a Graves’ IgG clone. The inhibition of TSH binding to
thyroid membrane by the preincubation of the Graves’ IgG with the
immobilized TSHRE was modest. This is most likely due to the presence
of polyclonal TSHR autoantibodies in each patient, as suggested by
others (3, 4, 5, 6, 7, 8). TSH apparently does not bind to the immobilized TSHRE. A
similar conclusion was drawn for some of the TSHRE produced in insect
cells (16, 18). We selected one of the two Graves’ IgG preparations
that showed the highest binding to TSHRE for screening of peptide
libraries. We have not investigated the statistical significance of the
variable degree of binding of Graves’ IgG to TSHRE or the possibility
that some of the Graves’ IgG completely fails to bind to the
extracellular domain. These will be investigated in future studies.
A peptide library approach was adopted for identification of the
peptide sequences that inhibit the binding of Graves’ IgG to TSHRE.
The positional scanning synthetic peptide combinatorial library
originally described by Houghten et al. (26) is particularly
suited for identification of the inhibitory peptides in an ELISA
format. After a reiterative process of synthesis and screening, EEFDDA,
EHFDDA, and ETFDDA were found to be the most effective for inhibition
of binding of Graves’ IgGs to TSHR and for inhibition of the cAMP
synthesis induced by Graves’ IgGs. The second position of the three
peptides appears not to be important. The peptide sequence we have
identified most likely mimics a class of epitopes for the Graves’
IgGs, as the antibody that interacts with TSHR could be polyclonal, and
among these the clones that activate the receptor are most likely
oligoclonal (3, 4, 5, 6, 7, 8). It is possible that the thyroid-stimulating
antibodies present in each Graves’ patient may be mono- or
oligoclonal. Furthermore, it has been suggested that Graves’ IgGs
recognize the tertiary structure of TSHR (3, 4, 5, 6, 7, 8). Therefore, the peptide
sequence we have identified most likely does not resemble any part of
the linear sequence of TSHR. Instead, the sequence may mimic a spatial
arrangement of key amino acids brought into close proximity by protein
folding. In this sense, it is not surprising that the hexapeptide
sequence we have discovered has low inhibitory activity
(10-5 M for half-maximal effect) toward the
binding of the antibody to the extracellular domain as well as for the
action of the antibodies in the synthesis of cAMP. Introduction of
physical constraint as well as lengthening of the peptide sequence
could improve the binding affinity of the peptide. We recently
discovered a peptide that stimulates the phosphoinositide hydrolysis in
immune cells from peptide libraries (29). The activity of the peptide
increased more than 100-fold from 10-6 M by
simple replacement of an amino acid with a D-form (our unpublished
data). During the initial screening of peptide libraries, the
concentration of each peptide sequence is in the range of
10-9 M. However, the final identified peptide
sequences are active at 10-5 or 10-6
M. This apparent paradox is due to the cooperative effect
of more than one active peptide present in peptide libraries, as
discussed by Houghten and his group (26, 27, 33).
We investigated the effect of the peptides on the synthesis of cAMP induced by IgG of eight different Graves’ patients. The peptides inhibited the action of IgG of only four patients. However, the inhibition was not complete even at a high concentration of the peptide. Even for the IgG used for screening the peptide libraries, the peptide showed only 70% inhibition at 10-3 M. The incomplete inhibition of the action of some IgGs by the peptides as well as the lack of effect on other IgGs suggest that each Graves’ patient may contain one or more clones of stimulating antibodies, and roughly half of the population of Graves’ patients may have the clone of the thyroid-stimulating antibodies that recognizes the peptide we have identified. This may be unrelated to the presence of other types of antibodies interacting with TSHR in the same patients. There appears to be no apparent relationship among the degree of stimulation of cAMP synthesis by IgG, inhibition of TSH binding to thyroid membrane by IgG, and inhibition of IgG-induced cAMP synthesis by the peptides. The peptides also inhibited the action of TSH to some extent. The significance is not clear at the present time.
If the stimulating antibodies recognize the tertiary structure of TSHR as previously suggested (3, 4, 5, 6, 7, 8), it is possible that some clones of the antibodies may only recognize a region of the whole TSHR bound to the cell membrane. For these autoantibodies, either thyroid membrane or whole cells will have to be used for screening of peptide libraries. We are currently screening peptide libraries with whole cells expressing hTSHR as well.
Once the mimotopes of all of the Graves’ stimulating antibodies are determined, these can be used as the lead peptides for the development of more effective sequences, and these can be used for diagnosis and prognosis as well as for development of immunosuppressive agents for Graves’ disease.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received June 19, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  C. H. Na, M. H. Lee, B. Y. Cho, and C.-B. Chae A Method for Identification of the Peptides That Bind to a Clone of Thyroid-Stimulating Antibodies in the Serum of Graves' Disease Patients J. Clin. Endocrinol. Metab., April 1, 2003; 88(4): 1570 - 1576. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B.-C. Suh, H. Lee, D.-J. Jun, J.-S. Chun, J.-H. Lee, and K.-T. Kim Inhibition of H2 Histamine Receptor-Mediated Cation Channel Opening by Protein Kinase C in Human Promyelocytic Cells J. Immunol., August 1, 2001; 167(3): 1663 - 1671. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-S. Bae, H. Bae, Y. Kim, T. G. Lee, P.-G. Suh, and S. H. Ryu Identification of novel chemoattractant peptides for human leukocytes Blood, May 1, 2001; 97(9): 2854 - 2862. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D.-G. Bae, Y.-S. Gho, W.-H. Yoon, and C.-B. Chae Arginine-rich Anti-vascular Endothelial Growth Factor Peptides Inhibit Tumor Growth and Metastasis by Blocking Angiogenesis J. Biol. Chem., April 28, 2000; 275(18): 13588 - 13596. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Lee, J. Y. Park, B. Y. Cho, and C.-B. Chae Expression of the Functional Extracellular Domain of Human Thyrotropin Receptor Using a Vaccinia Virus System: Its Purification and Analysis of Autoantibody Binding J. Clin. Endocrinol. Metab., April 1, 1999; 84(4): 1391 - 1397. [Abstract] [Full Text]
|
|
|
|
 |
|
 B. Rapoport, G. D. Chazenbalk, J. C. Jaume, and S. M. McLachlan The Thyrotropin (TSH)-Releasing Hormone Receptor: Interaction with TSH and Autoantibodies Endocr. Rev., December 1, 1998; 19(6): 673 - 716. [Abstract] [Full Text]
|
|
| ||||||||||||||
