| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Pathology Research Laboratory (M.M., N.M., D.A., R.T.M.) and Department of Pathology (A.B.C.), Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts 02129; Department of Endocrinology, University of Pisa (M.M., L.C., F.L., A.P.), 56100 Pisa, Italy; and Department of Pathology, University of Athens (N.M., S.T.B.), 10554 Athens, Greece
Address all correspondence and requests for reprints to: Dr. Michele Marinò, Department of Endocrinology, University of Pisa, Via Paradisa 2, 56100 Pisa, Italy. E-mail: m.marino{at}endoc.med.unipi.it
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
FRTL-5 cells, a rat thyroid cell line, cultured on filters in dual chambers form tight junctions and exhibit features of polarity, with expression of megalin exclusively on the upper (apical) surface. After the addition of unlabeled Tg to the upper chamber and incubation at 37 C, some Tg was transcytosed intact across FRTL-5 cells into the lower chamber. Two antimegalin ectodomain antibodies precipitated transcytosed Tg in fluids collected from the lower chamber. After the addition of Tg to surface-biotinylated FRTL-5 cells, an anti-Tg antibody and the two antimegalin ectodomain antibodies precipitated high molecular mass biotinylated material in fluids collected from the lower chamber, corresponding to much of the megalin ectodomain, as well as smaller amounts of lower molecular mass material. The results indicate that Tg transcytosed across FRTL-5 cells remains complexed with megalin ectodomain components, which we refer to as megalin secretory components.
In aminotriazole-treated rats, which develop increased megalin-mediated Tg transcytosis, antimegalin antibodies precipitated some of the Tg in the serum. Tg was also precipitated by antimegalin antibodies in sera from patients with Graves’ disease, in which we found increased megalin expression on the apical surface of thyrocytes. In contrast, in thyroidectomized patients with metastatic papillary thyroid carcinoma, in whom Tg is directly secreted by neoplastic thyroid cells into the circulation rather than transcytosed, serum Tg was not precipitated by antimegalin antibodies. The detection of Tg-megalin complexes may help identify the source of serum Tg in patients with thyroid diseases.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Transcytosis of Tg from the colloid is thought to be one pathway by which Tg enters the circulation (3, 4, 5). However, Tg may escape from the thyroid in other ways, and the relative importance of transcytosis apparently differs in various thyroid states. Transcytosis of Tg is believed to occur to only a limited extent under physiological conditions, as evidenced by the finding that in healthy subjects living in geographical areas with normal iodine intake, serum Tg levels are usually low or undetectable (6, 7). It is thought, however, that heightened TSH stimulation increases Tg transcytosis (5). In support of this, we previously provided evidence (8) that increased Tg transcytosis accounts for elevated serum Tg levels in rats with aminotriazole goiter, an experimental model in which increased TSH release from the pituitary leads to massive Tg endocytosis (9). Elevated Tg levels are also seen in patients with Graves’ disease (10), in whom TSH-like stimulation results from TSH receptor-stimulating autoantibodies (11). However, in other conditions, different mechanisms are probably responsible for passage of Tg into the bloodstream. In nontoxic goiter and subacute thyroiditis, passage of Tg into the circulation is thought to result mainly from leakage of Tg from the colloid as a consequence of disruption of the follicular epithelium (12, 13). In papillary thyroid carcinoma, Tg is thought to be directly secreted by neoplastic thyroid cells into the circulation (14, 15).
We have recently shown that megalin (gp330) mediates transcytosis of Tg by thyroid cells (8). Megalin is a member of the low density lipoprotein (LDL) receptor family (16, 17), which includes the LDL receptor, the very low density lipoprotein receptor, and the LDL receptor-related protein. The structure of megalin is characterized by a large extracellular domain, with four cysteine-rich ligand binding repeats, a single transmembrane domain, and a relatively short cytoplasmic tail (16, 17), which bears the signal for endocytosis after ligand binding to the ectodomain. Megalin has been shown to mediate endocytosis of multiple, unrelated ligands via coated pits (18, 19, 20, 21, 22). In immunohistochemical studies, megalin has been found on the apical surface of a restricted group of absorptive epithelial cells (23, 24). In the thyroid, megalin is expressed exclusively on the apical surface of thyrocytes, directly facing the follicle lumen (23, 24). Based on the assumption that physiological ligands of megalin are present in fluids to which the receptor is exposed (18, 19, 20, 21, 22), we postulated that megalin in the thyroid is an endocytic receptor for Tg. In support of this hypothesis, we found that 1) Tg binds to megalin in solid phase assays with features of high affinity receptor-ligand interactions (25); 2) megalin expression on thyroid cells is TSH dependent (8, 26); and 3) megalin mediates binding and uptake of Tg in cultured thyroid cells (26). More recently, we found that megalin-mediated endocytosis of Tg by cultured thyroid cells largely results in transcytosis of intact Tg from the apical to the basolateral cell membrane (8). Evidence that megalin mediates Tg transcytosis was also obtained in vivo in rats treated with aminotriazole (8).
The present study was undertaken to determine whether a portion of megalin remains complexed with Tg after its transcytosis and release from the basolateral surface of thyroid epithelial cells. We considered this possibility because of findings with the poly Ig receptor, which mediates transcytosis of polymeric IgA across intestinal epithelial cells or hepatocytes from the basolateral to the apical cell surface (27). Like megalin, the poly Ig receptor is an integral membrane protein. After endocytosis, IgA travels with the receptor to the apical membrane where the receptor is cleaved by proteolysis, releasing the ligand complexed with a soluble portion of the receptor (the secretory component), which lacks the cytoplasmic domain. In the present study we provide evidence that Tg transcytosed by thyroid cells is complexed with a large extracellular portion of megalin. This information may be of diagnostic value in certain patients with thyroid diseases by permitting identification of the source of circulating Tg.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Tg was purified from frozen rat thyroids by ammonium sulfate
precipitation and column fractionation, as previously described (8, 25, 26, 28). Tg preparations were analyzed by both nonreducing and reducing
SDS-PAGE, followed by Coomassie staining or Western blotting, as
previously reported (8, 25, 26, 28). Under nonreducing conditions two
bands were seen at about 660 and 330 kDa. The 660-kDa band corresponded
to covalently linked Tg dimers. Size exclusion gel chromatography
showed that almost all (
95%) of the 330-kDa band represented
monomers derived from noncovalently associated Tg dimers that had been
dissociated by SDS-PAGE, with a small fraction (5%) of free Tg
monomers. Under reducing conditions, two bands, one slower and one
faster, were seen, as previously described (14, 29). Other Tg products
with lower molecular masses were present in minimal amounts.
Antibodies
A rabbit antihuman Tg antibody cross-reactive with Tg from other species was purchased from Axle (Westbury, NY). A horseradish peroxidase (HRP)-conjugated mouse monoclonal anti-human Tg antibody cross-reactive with Tg from other species was obtained from DAKO Corp. (Carpinteria, CA).
A rabbit antibody, designated A55, prepared against immunoaffinity-purified rat megalin and a mouse monoclonal antibody, designated 1H2, that reacts with rat megalin ectodomain epitopes in the second cluster of ligand binding repeats were previously described (30). Both of these antibodies cross-react with human megalin (our unpublished observation). A rabbit antibody prepared against a megalin-glutathione-S-transferase (GST) fusion protein, which includes the second cluster of ligand binding repeats, was previously described (23, 30). A rabbit antibody, prepared against a peptide corresponding to a sequence in the cytoplasmic tail of rat megalin, designated Rb3, was previously described (23). None of the antimegalin antibodies reacted with the Tg preparation used in this study (not shown).
Cell cultures
Fisher rat thyroid cells (FRTL-5; American Type Culture Collection, Manassas, VA) were cultured as previously described (31, 32) in Coon’s F-12 medium containing 5% FCS and a mixture of six hormones. We previously showed that the FRTL-5 cells cultured under these conditions synthesize and secrete intact Tg after radiometabolic labeling (8), which indicates that they maintain functions of differentiated thyroid cells.
Transcytosis experiments
As previously described (8), FRTL-5 cells were cultured on high density large pore (3-µm) filters in cell culture inserts (Becton Dickinson and Co., Bedford, MA) placed in 24-well plates. These devices allow polarization of the cells and make it possible to trace transport of molecules across the cell layer, from the upper (insert) to the lower (cell culture well) chamber. Cells were used at complete confluence. The mean number of cells at confluence was 5.2 x 104 cells/well, and the mean amount of protein in cell lysates was 5.03 µg/well. In certain transcytosis experiments, before plating on filters, FRTL-5 cells were surface labeled with biotin, using EZ-Link Sulfo-NHS-LC-biotin from Pierce Chemical Co. (Rockford, IL), according to the manufacturer’s instructions.
In transcytosis experiments FRTL-5 cells were incubated at 37 C with unlabeled Tg (50 µg/mL in Coon’s F-12 medium, 5 mmol/L CaCl2, 0.5 mmol/L MgCl2, and 0.5% ovalbumin). Tg was added in a volume of 200 µL to the upper chamber, and the lower chamber was rinsed with 200 µL buffer without Tg. After 6 h, the medium from the lower chamber was collected and subjected to immunoprecipitation, as described below.
Experimental animals
Serum samples were obtained in a previous study (8) from 12 female Lewis rats, weighing 100–120 g (Charles River Laboratories, Inc., Wilmington, MA). Six rats received aminotriazole (3-amino-1,2,4-triazole, Sigma, St. Louis, MO) in drinking water (0.04%) for 12 days, and six rats that had not received aminotriazole were used as controls. Animal care and sacrifice procedures were in accordance with institutional guidelines. Serum samples were obtained from each rat at death and stored at -20 C until used for immunoprecipitation experiments described below.
Patients
Serum samples from 15 patients with thyroid diseases were collected at the Department of Endocrinology, University of Pisa (Pisa, Italy). Seven patients had Graves’ disease (1 man and 6 women; mean age, 40.2 ± 18.4 yr; age range, 26–69 yr); 3 of them were untreated hyperthyroid, and 4 were euthyroid under methimazole therapy. Eight patients had metastatic papillary thyroid carcinoma (1 man and 7 women; mean age, 57.6 ± 13.2 yr; age range, 29–72 yr). They had all been treated with total thyroidectomy and ablation of residual thyroid tissue with radioiodine therapy and were hypothyroid due to L-T4 therapy withdrawal for whole body scan. In addition, we used serum samples from 5 normal subjects with no evidence of thyroid disease.
All serum samples were tested for the presence of anti-Tg autoantibodies, using a commercial kit (anti-Tg MELISA, Byk Gulden SpA, Milan, Italy) and for the presence of antimegalin autoantibodies, as previously described (33). None of the sera used here had anti-Tg or antimegalin autoantibodies.
The diagnosis of Graves’ disease was based on the presence of hyperthyroidism associated with diffuse goiter and circulating anti-TSH receptor autoantibodies. The diagnosis of papillary thyroid carcinoma was based on histological findings.
Immunohistochemistry
Archival formalin-fixed, paraffin-embedded thyroid specimens from three patients with Graves’ disease and from three patients with papillary thyroid carcinoma were obtained from the Department of Pathology of the University of Athens (Athens, Greece). The diagnosis of Graves’ disease and papillary thyroid carcinoma was based on clinical, histological, and laboratory findings. Surgery was performed in all cases for therapeutic purposes. At the time of surgery, the three patients with Graves’ disease were euthyroid under carbimazole treatment, and the three patients with papillary thyroid carcinoma were untreated. Normal thyroid tissue was obtained from the contralateral lobe of two thyroid glands of patients subjected to total thyroid thyroidectomy for papillary thyroid carcinoma. These two patients were untreated at the time of surgery. As a positive control for megalin staining, we used an archival formalin-fixed, paraffin-embedded kidney specimen obtained from the Department of Pathology of the Massachusetts General Hospital (Boston, MA), which was collected at autopsy from a patient without kidney disease.
Immunohistochemistry was performed with 5-µm sections from paraffin-embedded specimens, using a previously described procedure (34, 35, 36). Briefly, sections were deparaffinized, dehydrated, microwaved, and treated with methanol. Sections were then incubated with the monoclonal antimegalin antibody 1H2 (20 µg/mL), followed by biotin-labeled horse antimouse IgG (Vector Laboratories, Inc., Burlingame, CA) and HRP-conjugated streptavidin (Biogenix, San Ramon, CA). The chromogen 3-amino-9-ethylcarbazole (Aldrich Chemical Co., Inc., Milwaukee, WI) was used as a substrate for HRP. Sections were counterstained with Gill’s hematoxylin (Fisher Scientific, Fairlawn, NJ).
Immunoprecipitation experiments
Samples were incubated overnight at 4 C with 50 µL protein A agarose beads (Pharmacia Biotech, Piscataway, NJ) coupled at saturation with the rabbit anti-Tg antibody, with A55, with the antimegalin-GST antibody, with Rb3, or, as a control, with normal rabbit IgG. Beads were washed extensively and resuspended in nonreducing Laemmli buffer. Samples were boiled and spun down, and supernatants were subjected to SDS-PAGE and Western blotting, as described below.
Western blotting
Samples were subjected to nonreducing 5–16% SDS-PAGE and blotted onto nitrocellulose membranes. To detect Tg, membranes were incubated with the HRP-conjugated mouse anti-Tg antibody (1:400). To detect biotinylated proteins, membranes were incubated with HRP-conjugated streptavidin (Bio-Rad Laboratories, Inc., Hercules, CA; 1:1000). Bands were detected using a chemiluminescent substrate kit (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD). The pixel density of the bands obtained was measured in scanned images using personal computer software (NIH Imager 2.1).
Tg assay in human sera
Tg was measured in human sera by RIA, using a kit from Diagnostic Products (Los Angeles, CA). Before Tg measurement, serum samples were incubated overnight at 4 C with protein A beads coupled at saturation with A55 or, as a control, with normal rabbit IgG. Beads were spun down, and Tg was measured in the supernatants.
Statistical analysis
Unpaired t test was performed using personal computer software (StatView, Abacus Concepts, Berkeley, CA).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
FRTL-5 cells were cultured on permeable filters in the upper chamber of dual chambered devices. We previously showed (8) that under these culture conditions FRTL-5 cells exhibit features of polarity, notably megalin expression exclusively at the upper surface of the cell layer and the presence of microvilli and coated pits on the upper surface, as seen by electron microscopy. We also showed (8) that with cells cultured under these conditions there is no paracellular leakage of Tg from the upper to the lower chamber.
We performed transcytosis assays by adding preparations of unlabeled rat Tg (comprising both the 660- and 330-kDa forms) to the upper chamber of dual chambered devices containing FRTL-5 cells, followed by incubation at 37 C. After 6 h, fluids were collected from the lower chamber, which, as in our previous study (8), were found by Western blotting to contain exclusively 330-kDa Tg (not shown).
To investigate whether transcytosed Tg is complexed with megalin, we performed immunoprecipitation experiments in fluids collected from the lower chamber, using three previously described antimegalin antibodies (23, 30). To detect megalin ectodomain epitopes, we used two rabbit antibodies: 1) A55, an antibody reactive with multiple ectodomain epitopes; and 2) antimegalin-GST, an antibody against a megalin-GST fusion protein, which includes a sequence corresponding to the second cluster of ligand binding repeats. To detect the megalin cytoplasmic tail, we used Rb3, a rabbit antibody against a synthetic peptide corresponding to a sequence in the cytoplasmic tail of rat megalin. Normal rabbit IgG was used as a control.
Fluids collected from the lower chamber were incubated with protein A
beads coupled with a rabbit polyclonal anti-Tg antibody or with each of
the three antimegalin antibodies. After immunoprecipitation, samples
were subjected to nonreducing SDS-PAGE, followed by Western blot
analysis with a monoclonal anti-Tg antibody. Normal rabbit IgG did not
precipitate transcytosed Tg (not shown). The rabbit anti-Tg antibody
precipitated intact 330-kDa Tg (Fig. 1
, lane 1). Furthermore, both antimegalin ectodomain antibody, A55 (Fig. 1, lane 2), and the antimegalin-GST fusion protein (Fig. 1, lane 3)
precipitated 330-kDa Tg, whereas the antimegalin cytoplasmic tail
antibody Rb3 did not (Fig. 1, lane 4). The results indicate that Tg
transcytosed by FRTL-5 cells is complexed with a portion of megalin
devoid of the cytoplasmic tail.
|
. The pixel density of the
band precipitated by the anti-Tg antibody was 38.34
pixels/cm2 (Fig. 1, lane 1). The pixel densities
of the bands precipitated by A55 and the anti-megalin-GST fusion
protein antibody were, respectively, 24.55
pixels/cm2 (Fig. 1, lane 2) and 26.58
pixels/cm2 (Fig. 1, lane 3). We estimated that
65–70% of the transcytosed Tg was precipitated by the two antimegalin
antibodies, representing the proportion complexed with megalin. Megalin secretory components complexed with Tg transcytosed by FRTL-5 cells
We made several attempts to characterize further the megalin
secretory component in Tg transcytosis experiments with FRTL-5 cells.
After immunoprecipitation of fluids collected from the lower chamber
with the antimegalin ectodomain antibody A55, we performed Western blot
analysis with 1H2, a mouse monoclonal antibody that reacts with the
second cluster of megalin ligand binding repeats, in the extracellular
portion of the molecule (30). A faint band of high molecular mass was
found, but only in some experiments (not shown). Because of the low
sensitivity of this method for the detection of megalin secretory
components, we tried another approach, namely biotin labeling of
surface proteins on FRTL-5 cells. After labeling, cells were plated on
permeable filters in the dual chambered devices; 24 h later,
unlabeled Tg was added to the upper chamber, and the cells were
incubated for 6 h at 37 C. The fluid from the lower chamber was
subjected to immunoprecipitation with the rabbit anti-Tg antibody or
with the three rabbit antimegalin antibodies (A55, anti-megalin-GST
fusion protein, or Rb3), followed by Western blotting and detection of
biotinylated proteins with streptavidin. As shown in Fig. 2, both the anti-Tg (lane 2) and the two
antimegalin ectodomain antibodies (A55 and antimegalin-GST fusion
protein; lanes 3 and 4) precipitated an intense band of very high
molecular mass (>500 kDa). In addition, other bands of lower molecular
mass and lesser intensity were precipitated by anti-Tg, A55, or
antimegalin-GST antibodies (Fig. 2, lanes 2–4). Among these bands, two
with molecular masses of more than 200 kDa (300 and 350 kDa),
were particularly intense. Rb3, the rabbit antibody against the megalin
cytoplasmic tail, did not precipitate any band (Fig. 2, lane 5),
indicating that megalin transcytosed with Tg across FRTL-5 cells had
lost its cytoplasmic domain. No band was precipitated by normal rabbit
IgG, which was used as a control (not shown).
|
, lane 2). To obtain information concerning the relative amounts of the megalin band with the highest molecular mass precipitated by the anti-Tg antibody or by antimegalin antibodies, we measured the pixel densities of the bands obtained in immunoprecipitation experiments. The pixel density of the band precipitated by the anti-Tg antibody was 42.65 pixels/cm2; the pixel densities of the bands precipitated by A55 and the antimegalin-GST fusion protein antibody were, respectively, 54.98 and 46.13 pixels/cm2. The results suggest that 75–90% of the major megalin secretory component was complexed with transcytosed Tg.
To investigate whether release of megalin into the lower chamber occurs
in the absence of Tg transcytosis, we performed experiments with
biotinylated FRTL-5 cells in which buffer lacking Tg was added to the
upper chamber. The material in the lower chamber was subjected to
immunoprecipitation with A55, followed by Western blotting with
streptavidin. As shown in Fig. 2 (lane 1), only extremely faint bands
were precipitated by A55. The results indicate that transcytosis of
megalin occurs only to a minimal extent in the absence of added Tg. The
small amount of megalin released into the lower chamber in the absence
of exogenously added Tg may have been complexed with endogenous Tg,
which is known to be synthesized and secreted by FRTL-5 cells (31, 32),
as we have found in the cells used here (8). The results suggest that
release of the megalin secretory component does not occur
constitutively in the absence of ligand binding.
Detection of circulating Tg-megalin complexes in vivo
Model of aminotriazole goiter in rats. To investigate megalin-mediated transcytosis of Tg in vivo, we previously used the model of aminotriazole goiter in rats (8). Aminotriazole inhibits iodination of newly synthesized Tg, resulting in increased TSH release from the pituitary (9). After several days, progressive changes occur in the thyroid due to the stimulatory effects of TSH, characterized by enlargement and proliferation of thyroid cells with massive endocytosis of Tg from the colloid (9). By 10–12 days the colloid is almost completely depleted (9). We found that the rats treated with aminotriazole had a striking increase in megalin expression on thyrocytes as well as elevated Tg levels and reduced T3 levels in the serum, supporting the conclusion that Tg internalized by megalin is transcytosed (8).
To investigate whether circulating Tg in aminotriazole-treated rats was
complexed with megalin, we used serum samples obtained in our previous
study (8) from six rats treated with aminotriazole for 12 days. Serum
Tg levels in these rats were markedly and significantly
(P = 0.0021) higher than those in normal untreated rats
(aminotriazole-treated rats, 1037 ± 42 U/mL; normal untreated
rats, 272 ± 225 U/mL) (8). Serum samples were subjected to
immunoprecipitation with the rabbit anti-Tg antibody or with the
antimegalin ectodomain antibody A55, followed by Western blotting with
the mouse anti-Tg antibody. As shown in Fig. 3A, 330-kDa Tg was precipitated by both
the anti-Tg antibody (lane 1) and A55 (lane 2) in sera from
aminotriazole-treated rats, whereas in normal rats Tg was precipitated
by the anti-Tg antibody (Fig. 3B, lane 1), but not by A55 (Fig. 3B, lane 2). No bands were precipitated by normal rabbit IgG, which was
used as a control (not shown). The pixel density of the band
precipitated by the anti-Tg antibody in aminotriazole-treated rats was
930 pixels/cm2, and the pixel density of the band
precipitated by A55 was 37.5 pixels/cm2. The
results indicate that about 4% of circulating Tg in aminotriazole rats
was complexed with megalin.
|
Patients with Graves’ disease. In humans, elevated serum Tg levels are seen in association with intense TSH or TSH-like stimulation, as occurs in Graves’ disease due to TSH receptor-stimulating autoantibodies (11). In this condition there is increased Tg endocytosis with varying degrees of colloid depletion (37). Although in Graves’ disease there are elevated thyroid hormone levels in the serum, indicating lysosomal degradation of Tg, high levels of serum Tg are also seen, which are thought to result from transcytosis (14, 15). In contrast, in papillary thyroid carcinoma, in which increased serum levels of Tg are seen, mechanisms other than transcytosis are thought to be responsible, because the neoplastic thyroid cells are not organized in follicles and are therefore not exposed to colloid. The cells, however, are sufficiently differentiated to synthesize Tg, which is apparently directly secreted into the circulation (14, 15).
To investigate megalin-mediated transcytosis of Tg in humans, we
studied patients with Graves’ disease and papillary thyroid carcinoma.
We first assessed megalin expression in the thyroid by
immunohistochemistry, using archival formalin-fixed specimens. To
detect megalin, we used the mouse monoclonal antibody 1H2, which
cross-reacts with human megalin. As positive controls, we used sections
of normal human kidney prepared from formalin-fixed specimens, which,
as expected, showed staining restricted to the brush border region of
proximal tubules (not shown). However, the staining was of only
moderate intensity, indicating that the sensitivity of megalin
detection was low under the conditions used here, which explains why
megalin was not detected on normal thyroid cells in two specimens
studied (Fig. 4A) even though megalin
expression has been demonstrated in normal thyrocytes in other studies
(23, 24). In thyroid sections of three patients with Graves’ disease,
apical staining for megalin was found on many thyrocytes and was
especially intense in follicles depleted of colloid (Fig. 4B). This
finding is consistent with the conclusion that TSH-like stimulation
up-regulates megalin expression on human thyrocytes, as previously
shown in rat thyrocytes (8). In sections from three patients with
papillary thyroid carcinoma, moderate diffuse intracellular staining of
thyroid cells was found, with no distinct surface staining (Fig. 4C).
This pattern suggests that megalin is not normally transported to the
surface of the neoplastic cells, perhaps as the result of defective
chaperoning or assembly.
|
Serum samples were incubated with protein A beads coupled with the
rabbit anti-Tg antibody or with the antimegalin ectodomain antibody
A55, which cross-reacts with human megalin, followed by Western blot
analysis with the mouse anti-Tg antibody. As shown in Fig. 5A, in sera from patients with Graves’
disease 330-kDa Tg was precipitated by the anti-Tg antibody (lane 1)
and by A55 (lane 2). In contrast, in serum samples from patients with
metastatic papillary thyroid carcinoma, Tg was precipitated by the
anti-Tg antibody (Fig. 5B, lane 1), but not by A55 (Fig. 5B, lane 2).
No bands were precipitated by normal rabbit IgG, which was used as a
control (not shown). We also performed immunoprecipitation experiments
in serum samples from five normal subjects. However, Tg levels in these
samples were undetectable by RIA, and serum Tg was not precipitated by
either the anti-Tg antibody or A55 (not shown).
|
We then evaluated the effect of preadsorption of serum samples with
antimegalin antibodies on serum Tg levels in patients with Graves’
disease and in patients with metastatic papillary thyroid carcinoma.
Serum samples were incubated with protein A beads coupled with A55 or,
as a control, with normal rabbit IgG, followed by Tg measurement by
RIA. After preadsorption with normal rabbit IgG beads, the mean serum
Tg level in patients with Graves’ disease was 90.7 ± 61.4 ng/mL
(range, 25–215 ng/mL). Preadsorption with A55 beads resulted in a
reduction of serum Tg levels to 46.4 ± 32.2 ng/mL (range, 5–100
ng/mL), indicating that about 50% of detectable Tg in the serum of
patients with Graves’ disease was complexed with megalin. After
preadsorption with normal rabbit IgG beads, the mean serum Tg level in
patients with metastatic papillary thyroid carcinoma was 454 ±
485 ng/mL (range, 25–215 ng/mL). The mean serum Tg level after
preadsorption with A55 was 468 ± 462 ng/mL (range, 90–1505
ng/mL), indicating that virtually all of the detectable Tg in the serum
from patients with metastatic papillary thyroid carcinoma was recovered
after adsorption with A55. As shown in Fig. 6, the mean proportion of serum Tg levels
recovered after preadsorption with A55 beads was significantly lower
(P = 0.006) in patients with Graves’ disease
(51.1 ± 32.1%) than in patients with metastatic papillary
thyroid carcinoma (103.2 ± 32.8%).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As noted previously, megalin is a member of the LDL receptor family (16, 17) that is expressed on the apical surface of thyrocytes (22, 23). In previous studies we have shown that megalin is a high affinity receptor for Tg (25), capable of mediating Tg uptake by cultured thyroid cells (26). More recently, we demonstrated that after megalin-mediated endocytosis, Tg is transcytosed intact from the apical to the basolateral surface of thyroid epithelial cells (8). Our findings support and extend studies by Herzog and associates (4, 5), who demonstrated Tg transcytosis across pig thyroid cells and postulated that Tg transcytosis is a major mechanism for its passage into the bloodstream (8).
In our previous (8) and present study we used FRTL-5 cells cultured on permeable filters in the upper chamber of dual chambered devices. We found that when Tg was added to the upper chamber, it was transported intact into the lower chamber by transcytosis (8). We concluded that passage of Tg across FRTL-5 cells was from the apical to the basolateral surface, because under the conditions used, FRTL-5 cells express megalin (an apical membrane receptor) exclusively at the upper surface, where microvilli and clathrin-coated pits are seen by electron microscopy (8). Evidence that passage of Tg occurred by transcytosis and not by paracellular leakage was provided by several findings, notably the inhibitory effects of low temperature and of the microtubule-disruptive agent colchicine, the inhibitory effects of specific megalin competitors, and the low passage across the cell layer of [3H]mannitol, a molecule of very low mass (8). Furthermore, FRTL-5 cells cultured under the conditions used here form tight junctions, as indicated by the presence of junctional complexes by electron microscopy and by the expression of occludin, a tight junction-associated protein (8, 38).
In the present study we show that Tg transcytosed by FRTL-5 cells is complexed with megalin secretory components. Thus, two polyclonal antimegalin ectodomain antibodies, one against multiple ectodomain epitopes (A55) and one prepared against the second cluster of ligand binding repeats (anti-megalin-GST), precipitated 330-kDa Tg transcytosed by FRTL-5 cells. In contrast, a rabbit antibody against a peptide sequence in the megalin cytoplasmic domain (Rb3) failed to precipitate transcytosed Tg, indicating that megalin released as a complex with Tg is devoid of the cytoplasmic tail. Further evidence in support of this conclusion was provided by experiments in which Tg transcytosis across surface biotinylated FRTL-5 cells resulted in the release of biotinylated material into the lower chamber that was precipitated by the two antimegalin ectodomain antibodies, but not by the antibody against the megalin cytoplasmic domain. The main megalin secretory component was a band of very high molecular mass (>500 kDa), which must represent a large portion of the megalin ectodomain. However, several other lower molecular mass bands were also precipitated (although in lower amounts). All megalin components were precipitated by a rabbit anti-Tg antibody, indicating that they were complexed with transcytosed Tg.
Herzog and associates (4, 5) have provided evidence that transcytosis of Tg across pig thyroid cells occurs in microvesicles. Based on their findings (4, 5) and on our observations we postulate the following events in Tg transcytosis. Megalin mediates Tg endocytosis at the apical surface of thyrocytes via clathrin-coated pits (where megalin is concentrated). Somewhere in the endosomal compartment vesicles containing Tg-megalin complexes are formed, which travel to basolateral surfaces, thereby bypassing the lysosomal pathway. At some stage during passage across the cells, megalin is cleaved, probably close to the transmembrane domain, leaving a large portion of the ectodomain complexed in soluble form with Tg. Additional cleavage could account for the smaller megalin fragments that are transcytosed. At basolateral surfaces, the vesicles fuse with the cell membrane, releasing their soluble content into the extracellular compartment. Clearly, future studies aimed at tracing the intracellular route of Tg and megalin using confocal and electron microscopy and cell fractionation techniques will be required to provide direct documentation of this process. Furthermore, the biochemical mechanisms responsible for targeting of Tg-megalin complexes toward the transcytosis pathway and for cleavage of megalin are unknown and require further investigations. These investigations will be aimed also at identifying the main cleavage site of megalin in the ectodomain, which should permit determination of the size of the megalin secretory component that is released complexed with Tg. Based on the results obtained in our previous studies with intact megalin (25, 26), it is likely that Tg binds to the megalin secretory component with high affinity and in a saturable manner. However, we do not have direct evidence for this, and further studies are needed to investigate this issue.
To investigate Tg transcytosis in vivo, we studied an experimental model in rats as well as in patients with thyroid diseases. We previously presented evidence (8) that rats treated with aminotriazole have increased megalin-mediated transcytosis of Tg associated with markedly increased serum levels of intact 330-kDa Tg. Here we found that some of the serum Tg in aminotriazole-treated rats is complexed with megalin, as shown by immunoprecipitation experiments with the rabbit antimegalin ectodomain antibody A55. Similar results were obtained with sera from patients with Graves’ disease, where A55 precipitated intact Tg. Furthermore, by immunohistochemistry we found increased megalin expression on the apical surface of thyrocytes in Graves’ patients, especially in small follicles with little colloid, indicative of intense endocytosis. The results support the idea that there is increased Tg transcytosis in Graves’ disease (14, 15) and provide new evidence that the process is mediated by megalin. In contrast, in thyroidectomized patients with metastatic papillary thyroid carcinoma serum Tg was not precipitated by A55, even though serum Tg levels in these patients were 4.5-fold higher than those in patients with Graves’ disease. This finding as well as the lack of follicles containing colloid and the absence of cell surface staining for megalin support the conclusion that Tg is not transcytosed across neoplastic thyroid cells, but is directly secreted by the cells into the circulation, as previously suggested (14, 15).
Our present and previous (8) observations provide compelling evidence
that megalin can mediate Tg transcytosis both in cultured thyroid cells
and in vivo. Furthermore, our findings indicate that
megalin-mediated transcytosis is the major mechanism for Tg passage
across cultured thyroid cells. Thus, we previously showed that megalin
accounts for 65–80% of Tg transcytosis across FRTL-5 cells (8), and
here we found that roughly 65–70% of Tg transcytosed by FRTL-5 cells
was complexed with megalin. However, our estimates of the proportion of
circulating Tg complexed with megalin were substantially lower in
patients with Graves’ disease (25–50%) and were drastically lower in
aminotriazole-treated rats (4%). The results suggest that
additional transport mechanisms may contribute to the passage of Tg
into the circulation in these conditions. However, other explanations
are worth considering. Thus, Tg-megalin complexes may dissociate more
extensively in vivo than in the cell culture chamber from
which transcytosed Tg is recovered in experiments with FRTL-5 cells.
Furthermore, Tg-megalin complexes may be cleared more quickly from the
circulation than unbound Tg. Further studies are needed to determine
the fate of Tg-megalin complexes in vivo and to investigate
the possibility that mechanisms other than megalin-mediated
transcytosis contribute to Tg transport into the bloodstream.
Nevertheless, the finding that in these conditions some serum Tg is
complexed with megalin provides compelling evidence that megalin
mediates its transcytosis at least in part.
Recently, Druetta et al. (15) developed a method to identify the origin of serum Tg based on the measurement of its hormone content. Because Tg is iodinated exclusively at the apical surface of thyroid cells within the follicle lumen (1), the finding of a high thyroid hormone content is considered to provide evidence that Tg has been derived from the colloid, rather than secreted directly from thyrocytes into the bloodstream (15). In their study Druetta et al. (15) found no difference in the hormone content of serum Tg between patients with Graves’ disease and patients with metastatic papillary thyroid carcinoma (15). However, although the method used by Druetta et al. (15) appears to be sensitive and accurate, the finding of serum Tg with low hormone content does not exclude the possibility that Tg is derived from the colloid, because some poorly iodinated forms of Tg are present there (1, 2, 3). We propose that measurement of the proportion of serum Tg complexed with megalin may help determine whether Tg entered the circulation from the colloid. This approach may be used in patients suspected of having papillary thyroid carcinoma before they are subjected to thyroid surgery. Serum Tg in such patients can result either from direct secretion of Tg by tumor cells or from transcytosis of Tg in normal thyroid tissue. The absence of serum Tg bound to megalin would suggest Tg secretion into the blood by tumor cells. Clearly, further studies are needed to standardize the method for measuring the proportion of serum Tg bound to megalin and to evaluate its usefulness in a large series of thyroid patients.
The importance of megalin-mediated transcytosis of Tg under physiological and pathological conditions probably depends in large part on the selectivity of the process and the regulation of megalin expression on thyrocytes. In this regard, we have obtained preliminary data suggesting that immature forms of Tg with a low hormone content are preferentially transcytosed by megalin, even though Tg forms with a higher hormone content are transcytosed to some extent (Marinò, M., et al., manuscript in preparation). Thus, under physiological conditions, where megalin expression on thyrocytes is low, only very small amounts of mature Tg would be lost from the colloid as the result of megalin-mediated transcytosis. However, under conditions of massive TSH or TSH-like stimulation, when megalin expression is increased, as shown previously (8, 26) and here in patients with Graves’ disease, mature as well as immature forms of Tg may be transcytosed, which may serve to reduce the extent of hormone release by diverting Tg from the lysosomal pathway. Further studies are needed to investigate these hypotheses.
|
Footnotes |
|---|
Received February 21, 2000.
Revised June 5, 2000.
Accepted June 8, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-macroglobulin receptor. Ann NY Acad Sci. 737:114–123.[Medline]
This article has been cited by other articles:
 |
|
 |
|
  C. R. Morales, J. Zeng, M. El Alfy, J. L. Barth, M. R. Chintalapudi, R. A. McCarthy, J. P. Incardona, and W. S. Argraves Epithelial Trafficking of Sonic Hedgehog By Megalin J. Histochem. Cytochem., October 1, 2006; 54(10): 1115 - 1127. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Lisi, A. Pinchera, R. T. McCluskey, T. E. Willnow, S. Refetoff, C. Marcocci, P. Vitti, F. Menconi, L. Grasso, F. Luchetti, et al. Preferential megalin-mediated transcytosis of low-hormonogenic thyroglobulin: A control mechanism for thyroid hormone release PNAS, December 9, 2003; 100(25): 14858 - 14863. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. L. TUMA and A. L. HUBBARD Transcytosis: Crossing Cellular Barriers Physiol Rev, July 1, 2003; 83(3): 871 - 932. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Marino, L. Chiovato, S. Lisi, A. Pinchera, and R. T. McCluskey Binding of the Low Density Lipoprotein Receptor-Associated Protein (RAP) to Thyroglobulin (Tg): Putative Role of RAP in the Tg Secretory Pathway Mol. Endocrinol., October 1, 2001; 15(10): 1829 - 1837. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
