| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Medicine (O.E.J., B.S., K.M.), University of Essen, D-45122 Essen, Germany; Department of Medicine II (H.M.B.G.), Klinikum Grosshadern, Ludwig-Maximilians-University, Munich, Germany; and Howard Hughes Medical Institute (H.G.) and Departments of Medicine and Pediatrics and the Committee on Genetics (S.R.), University of Chicago, Chicago, Illinois 60637-1470
Address all correspondence and requests for reprints to: Onno E. Janssen, M.D., Department of Medicine, University of Essen, Hufeleud str. 55, D-45122 Essen, Germany. E-mail: . onno.janssen{at}uni-essen.de
Abstract
T4-binding globulin (TBG) serves to maintain an important serum pool of thyroid hormones and to prevent their excessive loss in urine. TBG has also been implicated in the tissue distribution and targeted delivery of the hormones, the mechanisms of which remain unclear. By virtue of sequence homology, TBG belongs to the serine proteinase inhibitors superfamily of proteins that are characterized by a reactive site loop serving as a recognition site for serine proteinases. However, both TBG and another serpin with hormone transport function, corticosteroid-binding globulin, are noninhibitory. Cleavage of corticosteroid-binding globulin by human leukocyte elastase results in the reduction of its hormone-binding affinity and capacity. In this communication we confirm previous observations that TBG is also cleaved by elastase and undergoes the characteristic conformational changes. In addition, contrary to a previous report, the present work demonstrates that the cleaved product has reduced T4-binding affinity and, as expected, increased heat stability. Additional fragmentation of the molecule results in the loss of the hormone-binding site that is in agreement with a recent in vivo observation of apparent consumption at sites of inflammation. These data suggest that TBG may play a role in the targeted delivery of thyroid hormones to tissues rich in proteinases.
THYROXINE, WHICH REPRESENTS about 97.5% of the circulating iodothyronines, is 5'-deiodinated in peripheral tissues to form the metabolically active hormone, T3 (1). More than 99% of the serum T4 and T3 are bound to transport proteins. The major fraction of T4 (75%) is bound to T4-binding globulin (TBG), a glycoprotein of hepatic origin (2, 3). Transthyretin carries about 15% of T4 in serum (4) and is the major thyroid-hormone binding protein in cerebrospinal fluid (5). Albumin has a rather low binding affinity but still carries about 7% of T4 because of its high serum concentration, while several lipoproteins transport only a minor fraction (3%) of thyroid hormones (4).
According to the free-hormone hypothesis (4, 6, 7, 8, 9), only thyroid hormones in this form are biologically active (0.03% and 0.3% of the total serum T4 and T3, respectively). This concept is supported by the correlation of thyroid hormone disposal rates with the free, rather than the total, hormone fraction (10, 11, 12) and the normal thyroid state of persons affected by one of the numerous genetic abnormalities associated with increased and, more often, decreased concentrations of any of the thyroid hormone transport proteins (13, 14).
What then is the physiologic function of the thyroid hormone transport proteins? Undisputedly, they provide an extrathyroidal reservoir for thyroid hormones that would otherwise be lost in urine (14, 15). The system of several serum thyroid hormone-binding proteins with different affinities constitutes a multicomponent buffer that keeps the hormone concentration constant over a wider range than a single component system could (5). The transport proteins thus provide a vehicle for the equal tissue distribution of thyroid hormones (6). Transthyretin may have a role in the distribution of thyroid hormones to the central nervous system and other organs because transthyretin receptors have been shown on human astrocytoma cells and rat primary hepatocytes (16, 17). A similar mechanism may apply to T4 bound to lipoproteins because they are internalized by specific receptors (18).
TBG shares considerable sequence homology with
1- proteinase inhibitor (PI) and other members of the serine proteinase inhibitor (serpin) superfamily of proteins (19). The serpins are characterized by an exposed reactive site loop (RSL) that serves as a recognition site for proteinases. Most serpins bind to specific proteinases and inactivate them. This inhibition involves cleavage of the RSL and is accompanied by a characteristic conformational rearrangement of stressed to relaxed (S to R) transition that leads to a considerable increase in thermodynamic stability (20).
In 1988 Pemberton et al. (21) presented evidence that TBG and another member of the serpin superfamily, corticosteroid-binding globulin (CBG), are cleaved by human leukocyte elastase (HLE) producing the characteristic conformational changes from S to R. Although this cleavage reduced CBG-binding affinity and capacity for cortisol, it did not have a similar effect on TBG (21). In 1996, we presented data showing that TBG cleaved by HLE underwent a similar reduction in the binding affinity for T4 as well as capacity (22). This work, quoted by publications from two other laboratories that come to the same conclusion (23, 24), are the subject of the current communication. Thus, the mechanism implicated in the targeted delivery of cortisol to HLE-rich tissues, such as sites of inflammation, may also apply to TBG.
Materials and Methods
Materials
TBG, purified from human serum, was a kind gift from R. Gärtner (Ludwig-Maximilians-University, Munich, Germany). It had a single band on SDS-PAGE. HLE (E.C.3.4.21.37, mol wt 29 kDa) was from Calbiochem-Novabiochem (La Jolla, CA). It had a single band on SDS-PAGE and an activity of more than 20 U/mg protein as assessed by the digestion of the specific chromogen methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide (MeO-Suc-A-A-P-V-pNA). [125I]T4 (specific activity 46 MBq/mg) was obtained from NEN Life Science Products (Boston, MA). All other reagents were of analytical grade from Sigma (Deisenhofen, Germany) or Merck \|[amp ]\| Co., Inc. (Darmstadt, Germany).
Isolation and activation of leukocytes
Leukocytes were isolated by a hypaque-ficoll technique (25). The viability of the cells was checked by trypan-blue staining. 3.5 · 105 cells were diluted in Hanks buffered saline solution to a final volume of 200 ml and activated with 50 mM N-formyl-L-methionyl-L-leucyl-L-phenylalanine (FMLP), 50 mM phorbol-12-myristate-13-acetate (PMA) or 0.5 mg/ml TNF-
by incubation for 4 h at 37 C. The activated leukocytes were then chilled on ice and centrifuged at 5,000x g for 5 min. The supernatants were used to cleave the purified TBG preparation. Inhibition of serine proteinases was accomplished by the addition of 1 mM phenylmethylsulfonyl fluoride (PMSF).
Cleavage of TBG by HLE
Purified TBG, at a concentration of 4 µg/ml was incubated with various amounts of HLE in 0.5 M NaCl, 100 mM TrisHCl (pH 7.5) at 37 C. At various time intervals, aliquots were removed from the reaction mixture and treated with 1 mM PMSF to inhibit further digestion and stored at -20 C.
Amino acid sequence analysis
Partial amino acid sequence at the HLE cleavage site was determined by Dr. J. Kellermann at Max Planck Institute of Biochemistry (Martinsried, Germany). For this purpose, 1 µg purified TBG, digested with HLE (see above), was submitted to SDS-PAGE (see below) and blotted onto a polyvinylidene fluoride membrane. The two fragments of approximately 50 kDa and 4 kDa were excised and submitted to automated Edman degradation with an amino acid sequencer type 492 (PE Applied Biosystems, Langen, Germany).
Electrophoresis
SDS-PAGE was performed according to Laemmli (26), using 10% and 17% discontinuous Tris/glycine or precast 10% tricine gels (Novex, San Diego, CA). The buffer contained 5% ß-mercaptoethanol and 0.5% SDS. Protein bands were stained with Sypro-Orange (Mobitec, Göttingen, Germany). For nondenaturing PAGE and subsequent autoradiography, ß-mercaptoethanol and SDS were omitted from all solutions and the samples were preincubated with [125I]T4 for 1 h at room temperature followed by 1 h on ice. The dried gels were exposed to XR4 film (Kodak, Stuttgart, Germany).
T4-binding analysis
Parameters of T4 binding to TBG were determined as described previously (27). Briefly, 8 ng purified TBG (untreated or HLE treated) were incubated with [125I]T4 in the presence of increasing amounts of unlabeled T4. TBG-bound T4 was separated from free T4 with anion exchange resin beads (M400, Sigmz, Deisenhofen, Germany). The affinity constants (Ka) and maximal binding capacities of TBG-containing samples were determined by the method of Scatchard (28).
RIA of TBG
TBG concentration was measured with a commercial RIA kit (Cis Bio International, Gif-Sur-Yvette, France) using sheep antihuman-TBG-serum.
Elastase inhibitor assay
Elastase inhibitory activity was determined from the decrease in HLE activity, measured by the degradation of the synthetic substrate MeO-Suc-A-A-P-V-pNA. A recording spectrophotometer at 410 nm was used to monitor the reaction for at least 2 min without and for 5 min with the inhibitor at 25 C. Elastase inhibitory activity was expressed as the relative residual activity after addition of the inhibitor.
Heat stability assay
Samples of cleaved TBG and native controls were incubated at various temperatures and residual binding activity was assessed by T4-binding analysis.
Results
Cleavage of TBG by supernatants from activated leukocytes
Purified TBG was incubated with supernatants from leukocytes activated with the inflammatory cytokine TNF-
, the PKC activator, PMA, (Fig. 1
) or the chemotactic peptide, FMLP (Fig. 2
). These reactions generated two fragments of TBG, of approximately 50 kDa and 4 kDa, compared with the apparent molecular weight of 54 kDa of the intact molecule. Partial amino acid sequencing showed that the N terminus of the larger fragment had the sequence of Ala-Ser-Pro-Glu-Gly-Lys, corresponding to the first six residues of the mature protein and the smaller peptide had the sequence Phe-Leu-His-Pro-Ile-Ile, corresponding to residues 360 to 365. This result is in agreement with the cleavage site previously determined by Pemberton (21) and localized the exact cleavage site between Thr359 and Phe360, corresponding to the RSL of PI and CBG.
|
|
, PMA, and FMLP predominantly induce the release of the serine proteinases HLE and cathepsin G from neutrophils (29). HLE activity in the supernatants was ascertained by the digestion of the specific chromogen MeO-Suc-A-A-P-V-pNA. As expected, cleavage of TBG by the supernatants of activated leukocytes could be abolished by the addition of the specific serine proteinase inhibitor, PMSF (Fig. 2Cleavage of TBG by HLE
To better control the cleavage reaction kinetics, purified TBG was incubated with purified HLE. As shown on Fig. 3
, at a low HLE:TBG ratio of 1:500, only a small fraction of TBG was cleaved within 15 min. At a ratio of 1:150 (1.6 µg TBG with 0.0057 µg HLE), cleavage was complete after 15 min. The reaction products were similar to those obtained after exposure of TBG to supernatants of activated leukocytes (Figs. 1
and 2
). At a high HLE concentration (1:15), digestion was complete after 2 min and generated several new fragments. These most likely represent products of further degradation because the 50-kDa band disappeared with longer incubation periods.
|
Based on the finding described above, an HLE:TBG ratio of 1:150 was used for further studies. This ratio corresponded to that of HLE (complexed with PI and thus inactivated) to TBG in human serum (30). Under these conditions, the cleaved TBG molecule was stable to overnight incubation with HLE and to the addition of new HLE at low concentrations up to a ratio of 1:50 (data not shown). TBG samples of 0.04 µg treated with HLE were incubated with [125I]T4 and subjected to nondenaturing PAGE. Autoradiography revealed that T4 bound to the uncleaved as well as to the cleaved TBG (Fig. 4
). We have previously shown in an elastase inhibitor assay (31) that TBG reacted as a substrate without significant inhibition of HLE, even at a ratio as high as 1:50 and PI completely abolished elastase activity at a ratio of 1:1.5.
|
|
A polyclonal anti-TBG serum that recognizes only native TBG in a RIA with native radioiodide-labeled TBG as tracer (32) was used to measure TBG incubated for various time intervals with HLE. The loss of native TBG was similar to the amounts determined by Scatchard analysis of T4 binding (compare Fig. 6
with Fig. 5C
).
|
Serpins undergo a typical S to R transition if cleaved by a proteinase that results in increased stability of the molecule. Cleaved TBG retained its lower affinity T4 binding even after exposure to temperatures of up to 90 C, but native TBG did not bind T4 after exposure to 60 C for 20 min (Fig. 7
). Thus, cleaved TBG undergoes the typical S to R transition of the serpins both structurally, as shown previously by Pemberton et al. (21), as well as functionally.
|
There is strong evidence in support of the concept that homeostatic mechanisms are designed to maintain normal concentrations of the nonprotein-bound thyroid hormones (free hormone hypothesis) and that this concentration determines a persons thyroid state (4, 7, 8, 9). However, a constant supply of T4 and T3 to thyroid hormone-dependent tissues requires a storage system that is provided by the thyroid hormone transport proteins (6, 14).
The release of hormones by the transport proteins has been the subject of considerable controversy. Owing to the more rapid dissociation rates of the thyroid hormones from transthyretin and albumin, they most likely supply the bulk of hormones that enter the cells in physiologic situations. Perhaps a small fraction of T4 entering cells is mediated by the interaction of transthyretin (16) or lipoprotein (18) with cell surface receptors (7).
Although it represents only a minute fraction of the total serum proteins (0.025%), TBG is the main carrier for thyroid hormones because of its extremely high binding affinity (Ka = 1010 M-1). In this study, we characterized the product of TBG digestion by HLE, the major serine proteinase of leukocytes (29, 33). Even at a relatively low HLE:TBG ratio, corresponding to their normal serum concentrations (1:150), TBG was rapidly cleaved at a site corresponding to the RSL of other serpins, such as PI or CBG (20, 21). Under physiological conditions, cleavage of TBG in serum is prevented by the large excess of PI, which effectively inhibits elastase (HLE:PI ratio
1:30,000 [(34)]. In trauma or sepsis, systemic HLE concentrations reach 3 to 10 times the upper limit of normal (30). Local HLE concentrations at sites of inflammation are several orders of magnitude higher, overrunning the inhibition by PI, which is further inactivated by the concomitant release of free oxygen metabolites from activated neutrophils (33, 35, 36). Serine proteinases were in fact responsible for TBG cleavage by supernatants from activated leukocytes because it could be prevented by the addition of the specific serine proteinase inhibitor PMSF (Fig. 2
).
Cleaved TBG has reduced affinity for T4 and lost most of its T4 binding capacity because of a reduction in the number of binding sites. The decrease in the affinity of T4 binding to TBG by a factor of 3 was less profound than the previously described decrease in cortisol-binding affinity of CBG by a factor of 10 (21) but in agreement with the recently reported increased free-to-bound T4 ratio of TBG exposed to elastase (23). Pemberton et al. (21) were first to observe cleavage of CBG and TBG by HLE resulting in conformational changes from the stressed to their respective relaxed forms (S to R transition). However, they were unable to detect changes in the affinity of cleaved TBG for T4 when such changes were found for the affinity of CBG for cortisol. The reason for this discrepancy is unclear but possibly because of the sensitivity of the method used to detect T4 release from TBG. This is most likely the cause because our results have been confirmed as recently reviewed by Schussler (37). Indeed, work from Schusslers laboratory showed that the proteolytic cleavage of TBG by leukocytes lead to the release of T4 as determined by equilibrium dialysis (23). In another communication (38), the group of Schussler showed a rapid disappearance of TBG from serum of subjects undergoing cardiac bypass surgery that was not accompanied by a loss of other serum proteins. This finding, attributed to TBG consumption because proteinase cleavage at inflammatory sites, is supported by our finding of loss of T4-binding capacity because of further degradation of the primary cleaved fragment during exposure to HLE (see Fig. 3
). Finally, Suda et al. (24) probed the ligand-binding site of TBG with a specific fluorophore, 1,8-anilinonaphthalene sulfonic acid, and concluded that T3 binding decreases after serine proteinase cleavage of the TBG molecule.
Although the release of T4 at sites of inflammation by cleavage of TBG is likely to occur, its pathophysiologic significance is less certain. It has been known for a long time that wound healing is impaired in hypothyroidism (39, 40). It has also been shown that the metabolism of thyroid hormones is increased during bacterial infections and fever (41), compatible with an increased release of T4 and T3 from the transport proteins. Thyroid hormones are in fact concentrated at sites of inflammation, as shown by the correlation of pulmonary scintigrams with [131I]T4 in patients with pneumococcal disease and x-ray-proven lobar pneumonia (42). There was no such localized T4 uptake in patients who had received antibiotics for more than a week or did not have clinical signs of bacterial infection. Furthermore, Meinhold et al. (43) have shown that T4 is degraded to diiodotyrosine (DIT) in phagocytosing and therefore activated leukocytes in patients with severe infections or sepsis. DIT and its metabolites are bactericidal (44). Thus, cleavage of TBG by elastase could provide a means to supply T4 to sites of inflammation to support healing by generation of the bactericidal DIT. Because T4 uptake by the cells is a high-affinity process with an association constant of 109 M-1 (45), a small, 50% reduction of the binding affinity of TBG for T4 might be sufficient to induce the release of a significant amount of T4. Although subjects with partial or even complete TBG deficiency have no obvious deficit in wound healing, this matter has not been investigated in detail. Furthermore, subjects lacking TBG may not need such a releasing mechanism because they already have a higher fractional T4 turnover rate (11).
In summary, our findings suggest a possible role of TBG in the targeted delivery of thyroid hormones to sites of inflammation or other sites rich in proteinases capable of cleaving the TBG molecule. Further studies to clarify the pathophysiologic importance of this mechanism are warranted.
Acknowledgments
We thank the following individuals from the Ludwig-Maximilians-University (Munich, Germany): B. Treske for expert technical assistance; R. Gärtner for his generous gift of purified TBG; M. Jochum and H. Fritz for their instructions with regard to the HLE digestion and activity assays; J. Kellermann (Max Planck Institute of Biochemistry, Martinsried, Germany) for N-terminal amino acid sequencing of cleaved TBG; and J. Köhrle (University of Würzburg, Germany) for helpful discussions of the manuscript.
Footnotes
This work was supported in part by Grants DFG Ja 671/1-3 and SFB 469/B8 from the Deutsche Forschungsgemeinschaft (to O.E.J.) and DK-15070 from the NIH (Bethesda, MD) (to S.R.).
Abbreviations: CBG, Corticosteroid-binding globulin; DIT, diiodotyrosine; FMLP, N-formyl-L-methionyl-L-leucyl-L-phenylalanine; HLE, human leukocyte elastase; MeO-Suc-A-A-P-V-pNA, methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide; PI,
1-proteinase inhibitor; PMA, phorbol-12-myristate-13-acetate; PMSF, phenylmethylsulfonyl fluoride; RSL, reactive site loop; S to R, stressed to relaxed; serpin, serine proteinase inhibitor; TBG, T4-binding globulin.
Received July 27, 2001.
Accepted December 7, 2001.
References
This article has been cited by other articles:
![]() |
A. Zhou, Z. Wei, R. J. Read, and R. W. Carrell Structural mechanism for the carriage and release of thyroxine in the blood PNAS, September 5, 2006; 103(36): 13321 - 13326. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. den Brinker, K. F. M. Joosten, T. J. Visser, W. C. J. Hop, Y. B. de Rijke, J. A. Hazelzet, V. H. Boonstra, and A. C. S. Hokken-Koelega Euthyroid Sick Syndrome in Meningococcal Sepsis: The Impact of Peripheral Thyroid Hormone Metabolism and Binding Proteins J. Clin. Endocrinol. Metab., October 1, 2005; 90(10): 5613 - 5620. [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 |