| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Biologym, University Roma Tre, 00146 Rome, Italy
Address all correspondence and requests for reprints to: Dr. Sandra Incerpi, Department of Biology, University of Rome, Roma Tre, Viale Marconi, 446, 00146 Rome, Italy. E-mail: incerpi{at}uniroma3.it.
T4 is usually thought to function as a prohormone, yielding T3 by deiodination at target tissues. T3, in turn, gains access to the cell interior and the cell nucleus, binding to a nuclear receptor protein that acts as heterodimeric or homodimeric transactivator for thyroid hormone-regulated genes. This takes time, which is quite all right for a hormone traditionally considered to be a regulator of protein synthesis, cell metabolism, and the slow business of differentiation and development. Actually, it is also true that T3 and T4 sometimes, when the experimental design permits, induce very fast responses in cells, called nongenomic or extranuclear effects because they appear within minutes or even seconds; although such effects have been known for many years, it is not clear how they are brought about or what their physiological roles are. As so often happens when how and why of actions are left unanswered, these nongenomic aspects of thyroid hormone action have been tacitly ignored.
That may be about to change now. In this issue of Endocrinology, Bergh et al. (1) identified the cell surface T4 receptor, and surprisingly it turns out that the long-searched-for plasma membrane receptor is a well-known membrane protein, 
Vß3 integrin. It is likely that this protein, or similar membrane receptors, will turn out to be the trigger for most of the nongenomic actions of thyroid hormone reported in the literature, which include alterations in solute transport (Ca2+, Na+, H+, glucose), changes in activity of signal transducing kinases like protein kinase A, protein kinase C, pyruvate kinase M2, phosphatidylinositol 3-kinase and the MAPK pathway (2, 3, 4). We can see these actions to be rapid onset when we make cells thyroprival and reexpose them to thyroid hormone. Some of these effects rapidly lead to posttranslational modifications of nucleoproteins, e.g. serine phosphorylation of p53, the estrogen receptor, and even the nuclear thyroid hormone receptor TRß1 (5, 6, 7). The effects on nucleoproteins are mediated by MAPK (ERK1/2), and protein kinase C and the phosphatidylinositol pathway are probably activated upstream of MAPK (8). Recently an association between thyroid hormone receptor-ß1 and phosphatidylinositol 3-kinase-Akt/protein kinase B cascade has also been reported (9). In any case, the mechanisms by which thyroid hormones nongenomically affect the activity of plasma membrane ion channels and ion pumps are not well understood (3, 10). Membrane binding sites for thyroid hormones were identified many years ago in cell membranes from erythrocytes and hepatocytes (11, 12 ; for a review, see Ref. 13), but the linkage between binding sites and hormone actions has not been established until now. Although we can call these actions rapid in the experimental paradigms we design, the ambient levels of thyroid hormone in intact organisms are constant, and thus we assume that these nongenomic actions contribute to basal levels of activities of ion pumps and channels or to rates of transcription.
Bergh et al. (1) demonstrates that isolated Vß3 integrin binds T4 with very high affinity. The bound ligand can be displaced by the thyroid hormone metabolite tetraiodothyroacetic acid, which previously has been shown to be able to inhibit nongenomic effects of T4, and binding is also inhibited by specific anti-Vß3 antibodies and a peptide ligand that binds to the integrin Arg-Gly-Asp (RGD) recognition site (1). Using a fibroblast cell line expressing plasma membrane Vß3 but lacking the nuclear thyroid hormone receptor, they show that exposure to T4 leads to rapid activation of MAPK and induction of angiogenesis; these effects were prevented by tetraiodothyroacetic acid, the anti-Vß3 antibodies, and the inhibitory peptide. In addition, cells devoid of the V- or the ß3-protein did not respond to the hormone treatment, leaving little doubt that the integrin is the receptor.
Integrins are a family of transmembrane glycoproteins that form noncovalent heterodimers. The extracellular domains of the integrins interact with a variety of ligands including matrix glycoproteins (14), whereas the intracellular domains are linked to the cytoskeleton (15). Integrin Vß3 has a large number of extracellular protein ligands, including growth factors, and in most cases ligand binding can activate the MAPK pathway (16, 17). Many integrins contain the RGD recognition site that is important for the interaction with peptide ligands containing an Arg-Gly-Asp sequence (14). Recently Hoffman et al. (18) have shown that blocking the Vß3 integrin RGD site prevented T4-induced bone resorption. Another novelty of the paper by Bergh et al. (1) is that a small molecule like T4 can bind to an integrin; usually integrins bind polypeptides. The location of the ligand binding site for T4 is not known, but the results obtained with inhibitory peptides show it must be located at or near the RGD binding groove (19, 20). It is curious that the crystal structure of the membrane T4 receptor actually was known before the protein was identified as a receptor.
The finding that thyroid hormone uses an integrin as a surface receptor may explain several of the nongenomic and genomic effects reported in the literature. Angiogenesis, the growth of new blood vessels, plays a key role in development, wound repair, inflammation, and tumor growth and is a process known to be mediated by thyroid hormones in vivo (21, 22). Recently the role of nongenomic effects of T4 in angiogenesis has also been shown in vitro by the group of Davis (23) with the classical chick chorioallantoic membrane assay, a system that has a high level of expression of the Vß3-integrin (24). The mechanism by which thyroid hormone causes angiogenesis in the chorioallantoic membrane model is complex but is clearly initiated at the plasma membrane by the hormone and involves transduction of the hormone signal via the MAPK pathway into a fibroblast growth factor-2-dependent angiogenic response (23). Thyroid hormones have been shown to increase intracellular pH through nongenomic activation of the Na+/H+ exchanger, a plasma membrane protein that exchanges extracellular Na+ with internal protons, and this is also a proliferative and proangiogenic factor (25, 26, 27). In fact, the higher intracellular pH is an important requisite for cell spreading and appears to be regulated by several integrins (28, 29). It is known that both human and chick blood vessels involved in angiogenesis have enhanced expression of Vß3, and consistently the expression of Vß3 in cultured endothelial cells can be induced by various cytokines in vitro (24). Thyroid hormones enhance the actions of several cytokines and growth factors, such as interferon (IFN)-
and epidermal growth factor. Davis and colleagues have shown that there are two mechanisms by which thyroid hormone can potentiate the IFN effect (30): the first is a protein synthesis-dependent mechanism evidenced by enhancement of IFN antiviral action upon incubation with T3, T4, or the analog rT3 and inhibition of this enhancement by tetraiodothyroacetic acid or cycloheximide; the second is a protein synthesis-independent (posttranslational) mechanism induced by incubation of T4 or T3, but not reverse T3, with IFN and is not inhibited by tetraiodothyroacetic acid or cycloheximide (30).
Thyroid hormones are required for the normal development and differentiation of the cells of the central nervous system, in particular the oligodendrocyte precursor cells, the most important source of axons remyelination in the adult. Furthermore the cell-cycle blocking mechanism, terminal differentiation, and myelin production all depend on thyroid hormones (31). Deficiency of thyroid hormones during the perinatal period results in severe mental and physical retardation, known in humans as cretinism. Beside the known genomic action of thyroid hormones mediated by classical nuclear receptors, also nongenomic actions of thyroid hormone, such as actin polymerization and extracellular organization of laminin, play a role in brain development and in particular in neuronal migration (31). T4 is converted to T3 by a 5'-deiodinase in astrocytes, and T3 is then transferred to neurons where it binds to nuclear receptors to regulate gene expression. Probably T3 in these cells is directly regulating gene expression of proteins such as myelin basic protein, neurotropins, and reelin (32). Interestingly, Farwell and Dubord-Tomasetti (33) have shown that thyroid hormones can modulate the arrangement of laminin, an extracellular matrix protein that plays a pivotal role in the nerve cell migration during central nervous system morphogenesis through the interaction between integrins and components of the cytoskeleton.
One particular property of the Vß3 integrin is that T4 binds with a very high affinity, whereas T3 is a weak ligand, whereas exactly the opposite is true for the nuclear receptors. This demonstrates that T4 is not just a precursor of T3 but has a life of its own. However, also T3 gives rise to nongenomic effects after binding to a receptor on the cell surface, as shown by the use of membrane-impermeant T3 derivatives (S. Incerpi, unpublished results). Does this mean that there is another integrin receptor in the plasma membrane? The association of integrins with thyroid hormones appears even more appealing in the light of recent findings concerning the structure-activity relationship of these proteins: through conformation changes they can transmit both outside-in and inside-out signaling (34). Either intracellular (cytoplasmic/nuclear) or extracellular interaction can give rise to active or inactive states of the integrin, i.e. the integrin itself or the integrin-ligand complex can shuttle between different conformations. This could enable cells to expose either high- or low-affinity binding sites on the surface and perhaps provide an explanation for the interaction between nongenomic and genomic effects of thyroid hormones. In fact, thyroid hormones display such a wide variety of effects that the integrin mechanism appears to be a sort of deus ex machina able to account for all the effects and mechanisms that cannot be explained at present. The delivery is going to take some time, though.
 |
Footnotes |
|---|
Received April 4, 2005.
Accepted for publication April 4, 2005.
|
References |
|---|
 Top References |
|---|
Vß3 contains a cell surface receptor site for thyroid hormone that is linked to activation of mitogen-activated protein kinase and induction of angiogenesis. Endocrinology 146:2864–2871V integrin-mediated Ras-ERK signaling during two pathways of angiogenesis. J Cell Biol 162:933–943Vß3 negatively modulates IGF-1-mediated migration of tumor growth. Cancer Res 64:977–984Vß3 in complex with an Arg-Gly-Asp ligand. Science 296:151–155Vß3 for angiogenesis. Science 264:569–5715ß1 integrin expression and function. Atherosclerosis 154:377–385[CrossRef][Medline]
by two mechanisms. J Cell Physiol 167:269–276[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  O A Sukocheva and D O Carpenter Anti-apoptotic effects of 3,5,3'-tri-iodothyronine in mouse hepatocytes. J. Endocrinol., November 1, 2006; 191(2): 447 - 458. [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 |
