Interactions between thyroglobulin (Tg) and endocytic receptors on the apical membrane of thyroid epithelial cells (TEC) result in Tg uptake and delivery to post-endocytic pathways. Several receptors have been postulated or demonstrated to mediate Tg endocytosis and some are involved in thyroid homeostasis. The known Tg endocytic receptors are here reviewed, preceded by a brief introduction on intrathyrodal metabolism of Tg and mechanisms of Tg uptake.

Intrathyroidal metabolism of Tg
Tg is secreted by TEC into the thyroid folliular lumen, where it is involved in thyroid hormone formation by iodination and coupling of its tyrosine residues.(1). Then, Tg is either stored to form the colloid or is processed further for hormone release. For the latter, Tg is endocytosed by TEC, followed by lysosomal degradation, even though hormones can also be released within the colloid through the action of extracellular proteases (1). Although most of the internalized Tg undergoes lysosomal degradation, some follows other intracellular routes. After uptake, poorly glycosylated Tg molecules undergo a further passage through the trans-Golgi network, where glycosylation is completed, and are then recycled back into the follicular lumen (1-2). In addition, Tg can undergo transepithelial transport from the colloid to the bloodstream (transcytosis), thereby reaching the circulation (1-2). It is unknown how Tg crosses the capillary wall from the basolateral interstitium following transcysosis.

Mechanisms of Tg uptake by TEC: role of nonspecific and receptor-mediated endocytosis
In rodents, TEC can internalize Tg by macropinocytosis, but in most species, including humans, uptake of Tg occurs mainly by micropinocytosis (1-2). The majority of investigators agree that fluid phase micropinocytosis is the major route of uptake leading to hormone release, because of the very high concentrations of Tg within the colloid (1-2). Thus, fluid phase uptake is a constitutive process that occurs by passive gradient diffusion of substances that are highly concentrated in extracellular fluids and it ends invariably in lysosomes (1-2). Because endocytic receptors take up substances that are present in very low concentrations in extracellular fluids, it is unlikely that they mediate uptake of large amounts of Tg (1-2). Receptors may contribute to hormone release under special circumstances, for example iodine deficiency, but under physiological conditions they are more likely to sort Tg molecules to post-endocytic pathways that do not lead to hormone release, namely recycling or transcytosis (2). The contribution of the various means of endocytosis depends on the biochemical features of Tg, especially hormone content and degree of glycosylation (Fig. 1). Ultimately, both fluid phase endocytosis and receptor mediated endocytosis are aimed at the same goal, namely to render hormone release effective, which is achieved not only by targeting of Tg to lysosomes, but also by favoring complete glycosylation of immature Tg molecules through their recycling, or by eliminating hormone-poor Tg molecules from the colloid by transcytosis. The latter process is probably especially important when hormone-poor Tg molecules are present in excess within the colloid.

Tg endocytic receptors
Several molecules expressed on the membrane of TEC have been proposed to function as Tg endocytic receptors. Two of them, the asialoglycoprotein receptor (AGR) and megalin, are well characterized receptors. The characteristics of another receptor that binds to exposed N-acetylglucosamine residues of Tg [N-acetylglucosamine receptor (NAGR)] are unknown. As detailed below, the three identified Tg receptors probably deliver Tg to different, post-endocytic pathways (Fig. ).

Legend: Fig. Schematic representation of the known Tg endocytic pathways and of the receptors involved.

It is likely that other receptors for Tg also exist (2), but their nature and roles are not established.

The thyroid AGR
The existence of a thyroid AGR similar to the liver receptor was postulated because removal of sialic acid units from Tg increases its binding to thyroid membranes (3). Indeed, AGR is expressed by TEC and expression of AGR in thyroid PC C13 cells is TSH-dependent (4-6), suggesting a thyroid-specific function of the receptor. By immunohistochemistry, AGR can be found on the apical membrane of TEC in rat thyroid sections (5), directly facing the follicular lumen, therefore in the ideal position to mediate Tg endocytosis.
Asialo-Tg binds to AGR in solid phase assays and it also binds to the native receptor in PC C13 cells (4-6). In addition, Tg uptake by PC C13 cells can be in part reduced by an antibody against AGR (6), suggesting that, at least in cultured cells, AGR is involved in Tg endocytosis.
AGR binds especially to Tg molecules with a low degree of glycosylation, a feature that makes it an ideal candidate to be involved in Tg recycling. In addition, early studies showed that binding of asialo-Tg to thyroid membranes occurs optimally at low pH (3), which should prevent dissociation of Tg from the receptor due to the acidic pH of endosomes, as it sometimes occurs in ligand recycling or transcytosis. Nevertheless, degradation of Tg in PC C13 cells is reduced by an antibody against AGR (6), suggesting that, at least in part, AGR delivers Tg to lysosomes, which presumably should result in hormone release. It is possible that, also in view of its TSH dependence, AGR may facilitate maximal hormone release under special circumstances, for example iodine deficiency. However, a dual role with AGR mediating to some extent also Tg recycling cannot be completely excluded (Fig.). Unfortunately, no in vivo data are available, and therefore the actual impact of AGR on thyroid function remains to be elucidated.

NAGR
Asialo-agalacto-Tg, obtained by digestion of asialo-Tg with galactosidase, bears exposed N-acetylglucosamine residues and it binds to thyroid membranes to a greater extent than undigested Tg (7). In addition, asialo-agalacto-bovine serum albumin (BSA) binds to thyroid membranes with high affinity (saturation point 13 nM), and binding can be inhibited by unlabeled native Tg and to an even greater extent by asialo-Tg and asialo-agalacto-Tg (8). These findings suggest the existence of a NAGR capable of interacting with Tg, which may be responsible for recycling of poorly glycosylated Tg molecules (Fig.). In support of this, asialo-agalacto-BSA is released undegraded following endocytosis by cultured TEC, and ovomucoid, a glycoprotein with exposed N-acetylglucosamine residues, accumulates in the Golgi following endocytosis by cultured TEC (9). Although NAGR is very likely to mediate recycling, the exact identity of the receptor is unknown, as previous attempts to identify it failed. In addition, because studies in vivo are not available, the impact of NAGR on thyroid function remains to be established.

Megalin
Megalin is a member of the LDL receptor family expressed by a restricted group of absorptive cells, including TEC, where it can be found on the apical surface (2), thus in the ideal position to mediate Tg endocytosis. Megalin expression in TEC is up-regulated by TSH (2, 10), suggesting a thyroid-specific function.
Tg binds to purified megalin with high affinity (Kd ~9-11 nM), both in solid phase assays and to the native receptor in FRTL-5 cells (10). In the latter, megalin competitors reduce Tg uptake by ~50% (10), suggesting the receptor is involved in Tg endocytosis.
In most instances, megalin-mediated uptake of ligands results in their delivery to lysosomes (10). However, certain ligands undergo a different intracellular fate, namely transcytosis, which is the case for Tg, representing one of the mechanisms by which Tg enters the bloodstream. Thus, transport of Tg across FRTL-5 cell layers is reduced by megalin competitors (2, 10), and portions of the megalin ectodomain (secretory components) remain complexed with transcytosed Tg (11). Furthermore, in conditions associated with increased megalin expression, due to TSH (hypothyroid rats) or TSH-like (patients with Graves’ disease) stimulation, a relatively high proportion of serum Tg is complexed with megalin secretory components (11) and serum Tg levels in megalin KO mice are reduced (12).
The molecular mechanisms responsible for targeting of the Tg-megalin complex to transcytosis are known only in part. Although binding of Tg to megalin is optimal at low pH, transcytosis is only minimally affected by increasing intracellular pH, suggesting that pH resistance is not a major factor (13). The calcium-calmodulin pathway and phosphoinositide 3-kinase (PI3-K) affect Tg transcytosis. Calmodulin antagonists reduce transcytosis and increase T3 release, indicating calmodulin favors transcytosis prior to Tg sorting (10). In contrast, a PI3-K inhibitor increases Tg transcytosis but does not affect T3 release, suggesting that PI3-K exerts an inhibitory effect at a post-sorting stage (14).
A major role in determining targeting of Tg to transcytosis is related to the ability of Tg to bind to cell surface heparinoids. As other megalin ligands, Tg is a heparin-binding protein and heparin and megalin binding sites are functionally related (10). Occupation of a major heparin-binding sequence of rat Tg (2489-2503) abolishes Tg binding to megalin. In addition, transcytosis of rat Tg in FRTL-5 cells is reduced by enzymatic removal of cell surface heparan sulfate proteoglycans (HSPGs) as well as by an antibody against Tg2489-2503 (12). In this regard, optimal exposure of Tg2489-2503 is crucial, and a greater exposure of Tg2489-2503 in hormone-poor rat Tg is responsible for its preferential transcytosis compared with hormone-rich Tg (12). The role of Tg binding to heparinoids appears to be more important in rodents than in humans, because Tg2489-2583 is identical in mice and rats, but differs in humans by 6 residues, resulting in a weaker Tg binding to heparin (15).
As mentioned above, transcytosis is preferential for hormone-poor Tg molecules. Thus, Tg transport across FRTL-5 cells is greater for hormone-poor than for hormone-rich Tg (12). In addition, megalin-mediated transcytosis in vivo, estimated by the proportion of serum Tg complexed with megalin secretory components, is enhanced by inhibition of hormone formation within Tg due to thionamide treatment (12). This selective mode of transcytosis renders hormone release more effective, by preventing hormone-poor Tg to enter the lysosomal pathway, thereby avoiding competition with hormone-rich Tg, as well as wasteful transcytosis of hormone-rich Tg. Thus, megalin KO mice are hypothyroid (Lisi et al., manuscript in preparation). Whether megalin deficiency exists in humans and it causes thyroid dysfunction remains to be established.

REFERENCES