| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Division of Endocrinology and Metabolism, Department of Medicine, Mount Sinai School of Medicine, New York, New York 10029-6574
Address correspondence and requests for reprints to: Dr. P. Graves, Department of Medicine, Box 1055, Mount Sinai Medical Center, One Gustave L. Levy Place, New York, New York 10029-6574. E-mail: PGraves{at}mssm.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
and TSHR-ß subunits. TSHR-
subunits were detected by antibodies to epitopes localized within the
amino terminal end of the TSHR ectodomain and migrated diffusely
between 45–55 kDa, reflecting a differentially glycosylated status.
TSHR-ß subunits were detected by antibodies to epitopes within the
carboxyl end of the TSHR ectodomain. Several species of TSHR-ß
subunit were present, the most abundant having apparent molecular
masses of 50, 40, and 30 kDa. These data demonstrated that
post-translational processing of the TSHR in human thyroid tissue
involved multiple cleavage sites. |
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
subunit) contains several noncontiguous sites involved in TSH
binding (4). The carboxyl-terminal, nonglycosylated ß subunit
contains seven transmembrane helices and a cytoplasmic tail, and is
involved with G protein coupling and signal transduction. Support for
this two-subunit model, in which and ß subunits remain associated
by disulfide bonds, has strengthened in recent years. Proteolysis has
now been ascribed to a matrix metalloproteinase acting at the plasma
membrane (5). After cleavage, the majority of ectodomain fragments (
subunits) were shown to be shed from cultured thyrocytes and
TSHR-transfected cells into the medium as a result of disulfide bond
reduction (6). As these events were independent of cell disruption,
they provided further evidence for the physiological relevance of the
subunits. Further, an excess of TSHR-ß compared to subunits (up
to 3:1) in solubilized thyroid tissue plasma membrane has also
suggested in vivo cleavage and shedding (3). The cleavage issue is but one example of how the paucity and instability of TSHR proteins derived from thyroid tissue have retarded progress on the native structure. With the cloning of TSHR complementary DNA, these issues were addressed by expressing the receptor in higher yielding nonthyroidal systems (bacterial, insect, and mammalian cells). For example, functioning TSHRs on transfected CHO and COS cells have been invaluable for analyzing structure-function mutations identified in thyroid disease (7). However, such systems have proven to be suboptimal models for examining post-translational processing as it applied to the thyroid gland. As shown in comparative studies, the advantage of over-expression in nonthyroidal expression systems was accompanied by both quantitative (decreased processing efficiency) and qualitative (system-specific) differences (8). Thus, conclusions about TSHR structure and processing derived from such models require tissue validation. Accordingly, our strategy has involved using epitope-mapped TSHR antibodies and sensitive immunodetection protocols to catalog TSHR protein species present in thyroid tissue plasma membranes.
Using this approach, we have previously shown that porcine thyroid
contained both uncleaved TSH holoreceptors and cleavage-derived
subunits, as well as dimeric and higher order complexes of these
species (9). We suggested that -ß cleavage occurred between TSHR
residues 366–397, based on the observation that our TSHR 397–415
peptide antibody bound TSHR-ß subunits on immunoblots and the earlier
detection of TSHR-s by TSHR 352–366 antibody (10). Because the
primary disease of TSHR is Graves’ disease, a uniquely human disorder
with autoantibodies to TSHR, we have now extended our analyses to human
thyroid tissue, using a broader panel of TSHR antibodies to further
explore the issue of human-specific post-translational processing.
We have now found that antibodies that bound to human TSHR (hTSHR) residues 397–415 detected distinct TSHR-ß subunits of estimated molecular masses, approximately 50 kDa, 40 kDa, and 30 kDa in human thyroid tissue. These data suggested that a two-site cleavage model recently reported in hTSHR-transfected CHO cells (11) may not apply in the human thyroid gland. Rather, the variability in size of the cleaved proteins opened the possibility of multiple cleavage sites in this region.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The source of hTSHR in all experiments was the plasma membrane fraction prepared from frozen normal human thyroid tissues or cultured CHO-hTSHR cells. Tissue specimens were frozen in liquid nitrogen within 10 min of removal and stored at -80 C. Tissue samples were weighed and fragmented under liquid nitrogen in a mortar and pestle and the frozen powder mixed with 10 vol HB (250 mM sucrose, 1.25 mM EGTA, 50 mM Tris-HCl, pH 7.6) on ice containing 3 mg/mL of freshly dissolved proteinase inhibitor cocktail tablet (Boehringer Mannheim, Indianapolis, IN). Upon thawing, samples were homogenized on ice for 1 min at medium speed (Polytron, Brinkman Instruments, Inc., Westburg, NY) and clarified by centrifugation (760 g for 10 min). Membranes were pelleted by ultracentrifugation (250,000 g for 1 h). For electrophoresis, pellets were solubilized in electrophoresis (PAGE-SDS) sample buffer containing 2% sodium dodecyl sulfate (± 2% ß-mercaptoethanol) and incubated 30 min at 50 C. CHO-hTSHR cells (JPO9) were kindly provided by Dr. G. Vassart, University of Bruxelles, Belgium, and cultured as previously described (12).
Immunoblot analysis
Protocols for SDS-PAGE fractionation of membrane proteins and immunoblot detection of TSHR species were as previously described (9). Exceptions were that, in this study polyvinylidine fluoride membranes were used for the transblots, and the horseradish peroxidase-complexed secondary antibodies were detected using an upgraded version of enhanced chemiluminescence reagents (ECL-Plus, Amersham, Arlington Heights, IL). For peptide inhibition studies, the ability of a peptide to block detection of specific antigens on immunoblots was assessed by preincubating the peptide (10 ug/mL) with appropriately diluted antibody (in phosphate buffered saline of 0.1% Tween-20/10% dry milk) for 1 h at room temperature before application to the membrane.
TSHR antibodies
The source, type, and TSHR epitope targets for the antibodies
used in our studies are listed in Table 1
, along with appropriate references.
|
Monoclonal antibodies and serum from rabbits immunized with recombinant TSHR ectodomain protein were tested for the presence of antibodies binding TSHR ectodomain peptides in enzyme linked immunosorbent assay systems as described (13). Twenty-six peptides spanning the hTSHR ectodomain (residues 22–415) were provided courtesy of Dr. John Morris, Mayo Clinic (Rochester, MN). The peptides were 20 residues long and overlapped adjacent peptides by 5 residues. Peptides were dissolved at 2 ug/mL in carbonate-bicarbonate buffer, pH 9.5, and absorbed on 96-well plates. Serum samples were diluted 10-4 for the assay.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Figure 1 shows a linear
representation of the TSH holoreceptor depicting the relative locations
of induced TSHR antibodies used in this study. Immunoblots of CHO-TSHR
membranes using a monoclonal antibody to the amino terminus (residues
21–35) of the receptor detected several uncleaved holoreceptor species
of 85–120 kDa and glycosylated TSHR- subunits of 45–60 kDa (Fig. 2, panel A, lane 1). Detection was
inhibited by preincubation of antibody with hTSHR peptide 22–41 (lane
2), a peptide containing the reactive epitope. It was not inhibited by
control peptide hTSHR 397–415 (lane 3). TSHR-ß antibodies to the
peptide hTSHR 397–415 (6) detected similar holoreceptor species and
several TSHR-ß subunit size variants (panel B, lane 1). These ß
fragments included major species of 42 kDa and 55 kDa, as well as
several minor species migrating between these two. Detection of all
these variants was inhibited by hTSHR peptide 397–415 (lane 2) but not
by hTSHR 22–41 (lane 3).
|
|
subunits in
normal human thyroid tissue
Human TSH holoreceptors and cleavage-derived hTSHR- subunits
were also detected via immunoblotting of human thyroid membrane
proteins (Fig. 3). An antibody specific
for the amino-terminus (residues 21–35) of the receptor identified a
broad band of proteins of apparent molecular mass 45–55 kDa (lane 1)
characteristic of glycosylated TSHR- subunits previously described
in porcine thyroid (9). This antibody also detected two major species
of 110 and 120 kDa characteristic of glycosylated holoreceptors (2). An
antibody specific for an epitope residing between hTSHR residues
147–229 also detected the glycosylated TSHR- subunits but, in
contrast, did not readily detect the hTSH holoreceptor forms (Fig. 3, lane 2). Hence, this antibody had a low affinity for the holoreceptors,
most likely due to the epitope remaining buried despite reduction.
|
Immunoblots with a variety of antibodies also detected hTSH
holoreceptors and hTSHR-ß subunits in normal human thyroid. Using
antibody to hTSHR peptide 397–415, which readily detected recombinant
hTSHR ectodomain (residues 1–415) made in bacteria (Fig. 4, lane 1), we detected a major TSHR-ß
species of apparent molecular mass 30 kDa (lane 2). In addition,
holoreceptors of more than 120 kDa were most easily seen in the soluble
fraction of this preparation (lane 3). This fraction also contained two
species of TSHR-ß subunits: a major species of apparent molecular
mass 30 kDa and lesser amounts of an 50 kDa species.
|
The detection of at least two TSHR-ß species in thyroid tissue
by an antibody to hTSHR residues 397–415 (Fig. 4
), including one
species larger (50 kDa) than previously reported (3, 9), was evidence
for more than one holoreceptor cleavage site. Furthermore, the
different apparent sizes of TSHR-ß subunits in CHO-TSHR cells (Fig. 2) and human tissue (Fig. 3) suggested different cleavage sites in the
two systems. This was further investigated using additional TSHR
antibodies. An antibody raised against hTSHR fragment 350–416 detected
the 50 kDa species of TSHR-ß in human thyroid tissue (Fig. 5, lane 1). However, an antibody raised
against the entire hTSHR ectodomain and shown to bind peptides
primarily within the C-terminal portion of the sequence detected both a
40 kDa and a 50 kDa TSHR-ß species (lane 2) in a different human
thyroid preparation. In summary, antibodies shown to bind TSHR-ß
epitopes recognized several different TSHR-ß size variants in
immunoblots of human thyroid tissue membranes, indicative of multiple
holoreceptor cleavage sites or post-cleavage proteolysis, The clear
banding, however, indicated that such proteolysis was nonrandom. The
sizes of the largest variants detected also exceeded previous estimates
of TSHR-ß subunit lengths and were the least likely to have been
degraded after holoreceptor cleavage. These larger species are,
therefore, most informative regarding the initial cleavage sites in the
holoreceptor.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
subunits of 45–55 kDa and uncleaved
holoreceptors. The latter were present as both nonglycosylated (85 kDa)
and glycosylated (
110 and 120 kDa) species. Antibodies recognizing
epitopes near the carboxyl terminus of the hTSHR ectodomain detected at
least 3 nonglycosylated subunits of approximately 30, 40, and 50
kDa.
The first report using well-characterized TSHR antibodies for
immunoblot analysis of TSHR species present on human thyroid membranes
was that of Loosefelt et al. (3). In those studies,
monoclonal antibodies generated by immunization with an N-terminal
fragment (residues 19–243) of recombinant hTSHR detected glycosylated
TSHR- subunits of approximately 50 kDa (35 kDa fter deglycosylation
with N-glycosidase F). Conversely, monoclonals to a C-terminal fragment
(residues 604–764) detected nonglycosylated TSHR-ß subunits that
migrated diffusely between 30–40 kDa. The subunits were generated from
holoreceptors by reduction of samples before electrophoresis,
supporting a two-subunit model of the receptor in which the subunits
were disulfide-linked. This confirmed the original two-subunit model of
Rees Smith et al. (1), which was based on cross-linking of
radiolabeled TSH to the receptor.
While in general agreement with these earlier reports, our data
contained some differences. For example, in both this study and our
previous report on the porcine TSHR (9), we consistently detected
uncleaved TSH holoreceptors in reduced thyroid tissue membranes,
whereas Loosefelt et al. (3) reported nearly complete
conversion of holoreceptors into subunits upon reduction. This apparent
discrepency could be due to differences in the relative ability of
different TSHR antibodies to bind holoreceptors vs.
subunits, as illustrated in Fig. 3. A second difference was our
detection of a small number of electrophoretically discrete TSHR-ß
species, rather than a continuum of diffusely migrating species. The
latter suggests a "clip and nibble" model of processing rather than
a more limited number of discrete sites.
The most recent model of hTSHR post-translational cleavage was based on data from hTSHR expressed in CHO cells and suggested that the receptor ectodomain contained not one, but two, cleavage sites (11). This two-site model rested on the observation that a c-myc eptope replacing hTSHR residues 338–349 was not a component of the subunits, since the c-myc antibody detected holoreceptor forms but not subunits (as did control antibodies against other regions). Cleavage in the region of the presumptive downstream site (site 2) was not abrogated by single residue substitutions, but was abrogated by replacing TSHR residues 367–369 by residues from this locus in the noncleaving LH receptor (14). Because the LH receptor residues introduced an N-linked glycosylation motif, it was possible that cleavage abrogation was secondary to glycosylation and concomitant steric hindrance of cleavage. As to the presumptive proximal cleavage site (site 1), comprehensive analysis of single residue substitutions blanketing the entire region did not reveal any that abrogated cleavage, suggesting that this cleavage required the presence of the TSHR-unique 50 residue insertion spanned by sites 1 and 2, rather than a specific amino acid motif (15).
In the present study, multiple hTSHR-ß size variants were detected,
strongly indicative of more than two cleavage sites. The largest of
these fragments (50 kDa) was larger than TSHR-ß species previously
reported in membrane preparations from human thyroids, porcine
thyroids, and CHO-hTSHR cells, and therefore less likely to be a
nonspecific degradation product. Assuming an intact C-terminus at
residue 764, the 50 kDa estimate suggested a cleavage site near residue
300 and a TSHR-ß species of about 464 residues. While the earlier
c-myc epitope replacement data (11) suggested removal of
this segment in CHO-TSHR cells, the detection of both 50 kDa (with this
region) and 40 and 30 kDa (lacking the region) TSHR-ß subunits in
this study suggested that cleavage may be more random than previously
thought and may vary from cell type to cell type and from preparation
to preparation. This raises new questions as to the physiological role
of post-translational cleavage in TSHR processing as opposed to other
closely related glycoprotein hormone receptors that do not undergo such
cleavage.
|
Footnotes |
|---|
2 T.F.D. is the Theodore and Florence Baumritter Professor of
Medicine.
Received December 1, 1998.
Revised March 1, 1999.
Accepted March 19, 1999.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. J. Costa, Y. Song, P. Macours, C. Massart, M. C. Many, S. Costagliola, J. E. Dumont, J. Van Sande, and V. Vanvooren Sphingolipid-Cholesterol Domains (Lipid Rafts) in Normal Human and Dog Thyroid Follicular Cells Are Not Involved in Thyrotropin Receptor Signaling Endocrinology, March 1, 2004; 145(3): 1464 - 1472. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Ciullo, R. Latif, P. Graves, and T. F. Davies Functional Assessment of the Thyrotropin Receptor-{beta} Subunit Endocrinology, July 1, 2003; 144(7): 3176 - 3181. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Tanaka, G. D. Chazenbalk, S. M. McLachlan, and B. Rapoport Subunit Structure of Thyrotropin Receptors Expressed on the Cell Surface J. Biol. Chem., November 26, 1999; 274(48): 33979 - 33984. [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 |
