| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
INSERM U-38, Faculté de Médecine, 13385 Marseille Cedex 5, France
Address all correspondence and requests for reprints to: Annie Giraud, INSERM U-38, Faculté de Médecine, 27 boulevard Jean Moulin, 13385 Marseille Cedex 5, France. E-mail: inserm38{at}newsup.univ-mrs.fr
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Tg binding to the apical surface of thyroid inside-out follicles is
selective relative to the binding of other molecules. It increases
sharply over pH 8.0 and occurs through specific sites of moderately
high affinity (Kd = 
200 nM; 2 x
104 sites/cells). At pH < 8.0 it occurs through numerous
sites of very low affinity considered nonspecific. Endocytosis,
although weak under our conditions, increases at pH 8.0 concomitantly
with binding. Over pH 8.2, however, some inhibition occurs.
Surprisingly, Tg binding and endocytosis are not tissue specific, as they showed the same properties on thyroid inside-out follicles and Madin-Darby canine kidney or Chinese hamster ovary cells. Thus, a selective uptake of Tg mediates its recapture by thyroid cells. This selectivity is an intrinsic Tg property, not dependent on the thyrocyte apical surface, as it was observed with Madin-Darby canine kidney and Chinese hamster ovary cells. Given the pH effect observed, we suggest that Tg binding is a regulated phenomenon and that, through luminal pH control, it can vary from a basal level at neutral pH to a stimulated level over pH 8.0.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell culture
Porcine thyroid cells were isolated from thyroid glands by
discontinuous trypsin-EGTA treatment (7). Inside-out follicles were
obtained as described previously (8, 9); freshly isolated cells were
suspended in DMEM containing antibiotics and fungizone and supplemented
with 10% FBS (106 cells/ml). The cells were cultured as
unstirred suspension in polystyrene dishes not treated for tissue
culture coated with 1% agarose. They were maintained at 36 C in a
water-saturated, 95% air-5% CO2 atmosphere. The medium
was changed every 5–6 days by centrifuging the cells and resuspending
them in an equivalent volume of fresh medium. After 5–8 days,
inside-out follicles began to appear in the cultures, along with
aggregates (9). After 10–12 days, most of the aggregates had evolved
to form inside-out follicles (8), and the cultures were considered
ready for use. Twenty-four hours before each experiment,
10-4 M (Bu)2cAMP was added to the
culture medium.
Madin-Darby canine kidney (MDCK) cells were grown in DMEM supplemented
with 10% FBS and antibiotics (growth medium) on petri dishes (10-cm
diameter) treated for tissue culture. After reaching confluence, the
cells were resuspended by incubation in 1 ml/dish 0.25% trypsin and
0.2 g/liter EDTA in PBS (10 min at 37 C). After a 10-fold dilution
in growth medium, the suspended cells were seeded on petri dishes
(10-cm diameter) not treated for tissue culture, coated with 1%
agarose. The MDCK cells grown for several days in unstirred suspensions
formed large aggregates and occasionally follicle-like structures.
Chinese hamster ovary (CHO) cells were handled as MDCK cells, except that the growth medium was Ham’s F-12 medium supplemented with 10% FBS and antibiotics. When grown in unstirred suspensions, they remained free or associated in small loose aggregates.
Preparation of modified BSA
Polyglucosamine and T4 were coupled to BSA with the
hetero-bifunctional cross-linking reagent S-SMPB using a three-step
procedure described previously (10) for
m-maleimidobenzoyl-N-hydroxysuccinimide ester, another
N-hydroxysuccinimide ester used for coupling.
Briefly, in step 1, S-SMPB reacts by its ester group with primary amino groups of polyglucosamine or T4. Step 2 consists of a reduction of disulfide bonds in BSA by sodium borohydride. In step 3, S-SMPB-T4 or S-SMPB-(GlcNH2)n is coupled to reduced BSA; the cross-linker reacts by its maleimide group with aliphatic thiols (e.g. SH of reduced BSA). The molar ratio used for the coupling reaction was: BSA, 1; S-SMPB, 25; and (GlcNH2)n or T4, 30. (GlcNAc)n-BSA was obtained by acetylation of (GlcNH2)n-BSA with acetic anhydride.
The coupling of T4 was checked by coupling [125I]T4, whereas (GlcNAc)n coupling was checked by wheat-germ agglutinin affinity chromatography of (GlcNAc)n-BSA.
(GlcNH2)n was prepared by partial hydrolysis of chitosan by hydrochloric acid and then by charcoal retention of the (GlcNAc)n molecules liberated by hydrolysis (11). The monomeric forms were extracted from charcoal by 5% ethanol and were discarded. The polymeric forms were then extracted by 60% ethanol and purified on a DE 52 (OH form) column, Whatman Scientific Limited (Maidstone, UK).
Iodination of proteins
The molecules to be labeled [porcine Tg prepared as described
previously (12), IgG, and occasionally ovomucoid or modified BSA] were
iodinated with 125I by the chloramine-T method. Briefly, 4
µl (0.4 µCi) Na125I and 5 µl chloramine-T (1 mg/ml in
0.1 M phosphate buffer, pH 7.0) were added to 15 x
10-12 moles (in 10 µl) of the protein to be iodinated.
After 1-min exposure to chloramine-T at room temperature, 10 µl
sodium metabisulfite (1 mg/ml in the same phosphate buffer) were added,
and the reaction mixture volume was increased to 1 ml by adding PBS
containing 0.1% BSA. Excess reagents were removed by passage over a PD
10 (Sephadex G-25) column. The specific radioactivity obtained was
about 20–40 x 106 cpm/µg (10–20
µCi/µg).
Binding and endocytosis experiments
The suspended follicles (thyroid cells) or aggregates (CHO and
MDCK cells) were washed in Eagle’s Spinner salt solution and
resuspended in DMEM without bicarbonate but with 20 mM
HEPES containing 3% BSA and made 5 mM Ca2+ and
3 mM Mg2+. This medium will be referred to as
binding medium. Depending on the experiments, the pH varied from
7.2–8.5. In some experiments (see Results), the binding
medium was supplemented with molecules to test their effects (unlabeled
Tg and dextran sulfate). The follicle suspensions generally contained
0.5–2.0 x 106 cells/ml (determined by DNA
evaluation). Occasionally, higher cell concentrations (up to 10 x
106 cells/ml) were used with MDCK or CHO only. Binding and
endocytosis experiments were performed at 36 C under gentle agitation
in 25-ml or 100-ml siliconized flasks containing, respectively, 2.5 and
10 ml cell suspension.
Incubations started by the addition of labeled protein (7.6 x 10-14 moles/ml and 0.3 µCi/ml). At the end of the incubation, samples (1 ml) were collected in 5-ml test tubes. The cells were centrifuged (200 x g, 5 min) and washed twice in 2 ml ice-cold binding medium. They were then suspended in 1 ml of the same ice-cold medium, transferred to clean tubes, and centrifuged again. The supernatant was discarded, and pronase treatment of the pellet allowed separation of internalized and cell surface-associated ligand. The pellet was then suspended in 1 ml 2 µg/ml pronase in PBS, pH 8.2, or in 1 ml 5 µg/ml pronase in PBS, pH 7.4, depending on the pH used in the experiments. Occasionally, with the higher cell concentrations, 7 and 15 µg/ml pronase were used at pH 8.2 and 7.4, respectively. After 5 min at 36 C, protease inhibitors were added, and the cell suspension was cooled on ice. Under these conditions no cell lysis was observed, and the cells remained impermeable to erythrosine. Cells were collected by centrifugation and then washed in PBS. We considered that once pooled, the last two supernatants of centrifugation contained the surface-associated labeled ligand, whereas the final pellet contained the internalized ligand. After counting the supernatants and pellets, we used the pellets for DNA determination.
Nonspecific cell-associated ligand was determined by taking samples immediately after starting the incubation (time zero) and was subtracted from the data presented. When binding alone was to be studied (without internalization) incubations were performed at 4 C, and cellular ATP was depleted by the addition of 2 mM NaF and 10 mM NaN3, as described by Schmid and Carter (13). For apparent binding constant determination, labeled Tg (75 pM) was incubated in the presence of increasing amounts of cold Tg. At the end of the incubation, cell surface-associated ligand was recovered as described above. Nonspecific binding was determined in the presence of an excess of unlabeled Tg (30 mg/ml; 45 µM; a concentration leading to maximum inhibition of labeled Tg binding). The binding data were analyzed with the Ligand program (14).
DNA was determined using the fluorimetric method described by Labarca and Paigen (15).
Preparation for electron microscopy
Inverted follicles were fixed in 3.75% glutaraldehyde in 0.2
M cacodylate buffer, pH 7.4, at room temperature, rinsed
twice in same buffer containing 0.25 M saccharose, and
postfixed in 1% OsO4. After dehydration in graded acetone,
follicles were embedded in araldite. Sections were obtained with a
Reichert Ultracut E ultramicrotome (Leica, Paris, France), stained with
uranyl acetate and lead citrate, and examined with a JEOL 1220 electron
microscope (JEOL France, Roissy, France).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). This polarity was verified by
electron microscopy. The apical surface of thyroid cells, characterized
by microvilli, faces the culture medium (Fig. 1B). Another indicator of
correctly inverted polarity, the (Bu)2cAMP-induced swelling
of follicles suggesting water transport from culture medium toward
lumen, was regularly observed.
|
),
which indicates a good tightness of the epithelial cell wall and
accessibility limited to the part of the cell membrane facing the
culture medium (apical membrane).
Binding and internalization of Tg and IgG
Time-course experiments were performed with inside-out follicles
(Fig. 2). Measurements of total cell-associated
125I-labeled Tg and IgG showed that there was more
cell-associated Tg than IgG. This was clear at pH 7.2 (Fig. 2A) and was
even more pronounced at pH 8.2 (Fig. 2B). In fact, raising the pH from
7.2 to 8.2 enhanced the amount of cell-associated Tg, whereas the
amount of cell-associated IgG remained the same. Other molecules,
i.e. ovomucoid and modified albumins to which cross-linked
GlcNAc or T4, or cross-linker alone was added behaved like IgG and did
not bind to the cells as Tg did (data not shown). To discriminate
between surface-associated label and internalized label, we treated
cells with pronase (see Materials and Methods). At pH 7.4,
the cell surface association and internalization of
125I-labeled Tg were more efficient than IgG binding and
internalization (Fig. 2C). At pH 8.1, the binding of Tg was strongly
enhanced relative to that at pH 7.4, whereas only a moderate, if any,
enhancement of internalization was observed. IgG binding and
internalization were still very low and comparable to those observed at
pH 7.4 (Fig. 2D). Light microscopic controls showed that after 60–90
min of incubation at 37 C, follicles were still intact whatever the pH
from 7–8.3.
|
, A and B) or CHO cells (Fig. 3, C and D).
With the two cell lines, at both pH 7.4 (Fig. 3, A and C) and pH 8.2
(Fig. 3, B and D), we observed preferential Tg binding and
internalization comparable to that previously seen with inside-out
follicles, and we found an enhancement of Tg binding at pH 8.2.
|
shows a strong dependence on pH: a sharp increase in
surface-associated label occurred at pH 7.9/8.0, followed by a slow
decrease as the pH rose. Internalization was also affected by the pH
changes, but to a lesser extent. After reaching a maximum around pH
8.0, it decreased to its pH 7.4 value as the pH kept rising. We
observed no effect of pH on IgG binding and internalization (data not
shown).
|
, B (MDCK) and C (CHO), Tg binding on nonthyroid
cells was also dependent on pH. IgG binding on nonthyroid cells was not
pH dependent (not shown).
Effect of dextran sulfate
The effect of pH on [125I]Tg binding prompted us to
study the effect of negatively charged molecules used as potential
binding inhibitors. Binding and internalization of
[125I]Tg were performed without or with dextran sulfate
(150 µg/ml) added to the culture medium.
At pH 8.2, Tg binding to inside-out follicles was strongly inhibited by
dextran sulfate, whereas IgG binding was not. Internalization of Tg was
also inhibited, but to a lesser extent (Fig. 5B).
|
), but in other experiments, we
found no inhibition (not shown); with cells from four different
primocultures we twice observed an inhibition of binding and twice
observed no effect. Internalization was slightly or not inhibited in
three of four cases and was clearly inhibited in one case. In fact, we
found difficult to obtain clear reproducible results when working at pH
7.4.
At pH 8.2, Tg binding to MDCK (Fig. 5D) or to CHO (Fig. 5F) was
strongly inhibited by dextran sulfate. Internalization was also
inhibited, but less.
At pH 7.4, Tg binding to MDCK (Fig. 5C) or to CHO (Fig. 5E) was
slightly or not inhibited in most cases (four of five). However, at
this pH, we are still confronted with a dispersion of results that was
difficult to overcome. Internalization was inhibited to various extents
in all experiments.
Binding of 125I-labeled Tg: apparent
binding constants and number of receptor sites
Bound [125I]Tg was determined after overnight
incubation at 4 C in the presence of NaF and NaN3 (see
Materials and Methods) or after 2-h incubation at 4 C (long
enough to reach steady state in these conditions).
At pH 8.1, two sets of sites were found on inside-out follicles. One
set (2 x 104 sites/cell) bound Tg with moderately
high affinity (Kd = 200 nM). The
other set (>106 sites/cell) bound Tg with low affinity
(30 µM; Fig. 6).
|
MDCK and CHO cells exhibited the same two sets of sites at pH 8.1
(Kd = 200–400 nM and 4–6 x
104 sites/cell for the first; Kd =
60–120 µM and 2–3 x 106 sites/cell
for the second; data not shown).
At pH 7.4, we again observed only low affinity binding whose constants could not be correctly determined by Scatchard transformation due to the dispersion of the experimental points.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
As our first aim was to study Tg binding to thyroid apical membranes, we needed cells whose apical side was accessible. We chose suspensions of inside-out follicles rather than monolayers of cells on culture dishes or on culture plate inserts, to suppress the high nonspecific binding of Tg on cell supports. Although (Bu)2cAMP is not necessary to maintain the follicular structure, it was added 24 h before each experiment on inside-out follicles to stimulate thyroid cell metabolism and any synthesis of putative binding sites. CHO and MDCK cells were also used as cell suspensions to prevent nonspecific binding.
At equal molar concentrations, IgG binding to inside-out follicles is lower than Tg binding. Similarly, the binding of ovomucoid or of modified BSA (i.e. BSA bearing T4 or GlcNAc) is lower than Tg binding (not shown).
Tg binding also differs from IgG binding in its striking enhancement at basic pH; binding increases sharply at pH 8.0, whereas IgG binding remains unchanged. The possible significance of this finding (physiological phenomenon or mere artifact) will be discussed later. As the pH dependence of binding suggested the implication of some charges, polyanions such as dextran sulfate (see Results) or fucosan (not shown) were introduced in the binding medium; both inhibited the pH-dependent binding of Tg (the binding at pH 7.4 was only slightly inhibited). In contrast, IgG binding was not inhibited. We do not yet know whether the negative charges involved in Tg binding are borne by Tg or by the cell surfaces. Anionic charges are involved in the binding of ligands to macrophage scavenger receptors (18), but are not involved in the nonspecific binding of advanced glycosylation end products to macrophages (19). All known ligands of macrophage scavenger receptors are polyanionic molecules or macromolecular complexes, but the exact determinant(s) that confers binding ability has not been identified (18).
To further characterize Tg binding, we determined the apparent binding
constants of Tg on inside-out follicles. As shown in
Results, at pH 8.1 Tg binds to cell surfaces through two
sets of sites: one of moderately high affinity
(Kd = 200 nM) and one of low
affinity (Kd = 30 µM). At pH
7.4, despite experimental difficulties, the results suggest the
presence of low affinity sites only. The low affinity binding observed
at pH 7.4 and 8.1 can hardly be considered specific, and we think that
only the moderately high affinity binding found over pH 8.0 is
specific.
Surprisingly, when the experiments performed on inside-out follicles were reproduced on mammalian cell lines, we found that Tg binding to MDCK and CHO cell surfaces was also selective and had the same properties (affinity, pH dependence, and anion competition) as Tg binding to apical surfaces. Tg binding is not tissue specific. Its selectivity probably depends on intrinsic Tg properties. In this context, the selective Tg binding and endocytosis at the basolateral side described recently by Gire et al. (17) are in agreement with our observations.
Some tentatives, not shown here, to purify a Tg receptor by affinity chromatography on Tg-Sepharose of solubilized thyroid membranes or by cross-linking of Tg to cell surfaces and immunoprecipitation provided several proteins (including Tg itself) with no obvious enrichment for any of them. Indeed, these results may only reflect the difficulties encountered when attempting to purify a receptor that displays such a moderate affinity toward its ligand, but they may also reflect the fact that different molecules act as Tg ligands. The moderate affinity of the specific binding sites is comparable to that observed for lectin binding to human erythrocytes (20) and human lymphocytes (21) and to that observed for maleylated albumin binding to scavenger receptors of rat sinusoidal cells (22). In all of these cases a single molecule is able to bind different ligands that share a common binding determinant (sugar in the case of lectin, and polyanion in the case of scavenger receptor). We suggest that at pH 8, Tg undergoes charge and/or conformation modifications that transform it into a multiligand molecule. If so, it is likely that the different Tg ligands share a common binding determinant. A confirmation of this hypothesis will have to await identification of this determinant and identification of the Tg parts involved in the binding.
Shifrin and Kohn (23) studied the binding, on purified thyroid
membranes, of bovine Tg whose lysine or tyrosine residues had been
chemically modified. They found that removal of the positive charges of
lysine increased binding, whereas modification of the phenolic hydroxyl
groups of tyrosine residues by O-acetylation inhibited it in
a reversible way. If we assume that porcine Tg behaves in the same way,
some of our results could be explained by a change in the ionization of
lysine (
NH2) or tyrosine (phenolic OH) residues at pH
8.0. Indeed the pK values of free lysine and tyrosine are higher than
8.0, but the pK of an amino acid included in a macromolecule changes
under the influence of its environment.
Besides, pH 8.0 is rather high relative to the more neutral seric or cytoplasmic pH values. The pH limits inside the follicle are not yet fully established. To our knowledge, the only determination in thyroid compartments (extracellular fluid, follicular cells, and luminal fluid) was performed by Chow et al. (24) on turtle thyroid glands. They found values of 7.66, 7.26, and 7.32, respectively, for the three compartments considered. Interestingly, they showed that TSH increases the pH values of cells (up to 7.70) and lumen (up to 7.54), but not that of extracellular fluid. TSH-induced luminal pH augmentation is likely to occur in other species as well as turtles. In fact, an apical, amiloride-sensitive sodium uptake system, probably an amiloride-sensitive Na+/H+ exchanger, present on porcine thyroid cells (25) would be a likely candidate (among others) for the role of follicular pH regulator.
Under our experimental conditions (low concentrations of labeled Tg), endocytosis was weak, yet it was related to the amount of bound Tg (except at the highest pH tested, which may reflect some hindrance in vesicle formation). This is not surprising because a minimum Tg concentration is required to induce efficient endocytosis (Chabaud, O., unpublished observations). However, even with the high follicular Tg concentrations found in vivo (up to 250 mg/ml for rat thyroid) (26), fluid phase endocytosis is probably not the predominant mechanism in Tg uptake (6); adsorptive pinocytosis must contribute actively to Tg internalization.
In conclusion, Tg binds in a preferential way to cell membranes, not exclusively to thyroid cell membranes, through nonspecific binding sites and through specific ones that appeared at pH 8.0. Tg could be seen as a binding molecule able to recognize different membrane proteins. In view of the different results (ours and others) discussed above, we propose that in an unstimulated thyroid, Tg binds to apical surfaces through numerous nonspecific binding sites, whereas under TSH stimulation, specific binding sites appeared consecutive to a modification of Tg charge and/or conformation due to a rise in pH. Although the above hypothesis fits with most of the experimental data (ours and others), far more experimental studies are needed for it to be validated.
Whether regulation of the luminal pH exists and is involved in the binding of Tg is currently under investigation. In addition, binding assays of chemically modified Tg and tentative inhibition of the binding of normal and chemically modified Tg by polyanions and polycations will be required to specify the molecular mechanisms involved.
|
Acknowledgments |
|---|
Received September 26, 1996.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. Wangemann, H.-M. Kim, S. Billings, K. Nakaya, X. Li, R. Singh, D. S. Sharlin, D. Forrest, D. C. Marcus, and P. Fong Developmental delays consistent with cochlear hypothyroidism contribute to failure to develop hearing in mice lacking Slc26a4/pendrin expression Am J Physiol Renal Physiol, November 1, 2009; 297(5): F1435 - F1447. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Haberkorn and A. Altmann Imaging Techniques for Gene Therapy: SPECT, PET, and MRI Journal of Pharmacy Practice, October 1, 2001; 14(5): 383 - 396. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Marino and R. T. McCluskey Role of thyroglobulin endocytic pathways in the control of thyroid hormone release Am J Physiol Cell Physiol, November 1, 2000; 279(5): C1295 - C1306. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Marino, G. Zheng, L. Chiovato, A. Pinchera, D. Brown, D. Andrews, and R. T. McCluskey Role of Megalin (gp330) in Transcytosis of Thyroglobulin by Thyroid Cells. A NOVEL FUNCTION IN THE CONTROL OF THYROID HORMONE RELEASE J. Biol. Chem., March 15, 2000; 275(10): 7125 - 7137. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Marino, J. A. Friedlander, R. T. McCluskey, and D. Andrews Identification of a Heparin-binding Region of Rat Thyroglobulin Involved in Megalin Binding J. Biol. Chem., October 22, 1999; 274(43): 30377 - 30386. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Marino, G. Zheng, and R. T. McCluskey Megalin (gp330) Is an Endocytic Receptor for Thyroglobulin on Cultured Fisher Rat Thyroid Cells J. Biol. Chem., April 30, 1999; 274(18): 12898 - 12904. [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 |
, [125I]Tg; 
and 
,
[125I]IgG. This figure illustrates representative
experiments in which duplicate samples were assayed for surface-bound
and internalized radioactivity as described in Materials and
Methods. Duplicate samples are connected by a vertical
bar.
