Characterization of the ThyroxineBinding Site of Thyroxine-Binding Globulin by Site-Directed Mutagenesis
Christoph Buettner*, Helmut Grasberger, Kristine Hermansdorfer, Bingkun Chen†, Bettina Treske, and Onno E. Janssen Department of Medicine Klinikum Innenstadt Ludwig-Maximilians-University D-80336 Munich, Germany
The principal transport protein for T4 in human blood, thyroxine-binding globulin (TBG), binds T4 with an exceptionally high affinity (Ka 5 1010 M21). Its homology to the superfamily of the serpins has recently been used in the design of chimeric proteins, providing experimental evidence that an eight-stranded b-barrel domain encompasses the ligand-binding site. We have now characterized the T4 binding site by site-directed mutagenesis. Sequence alignment of TBG from several species revealed a phylogenetically highly conserved stretch of amino acids comprising strands 2B and 3B of the b-barrel motif. Mutations within this region (Val228Glu, Cys234Trp, Thr235Trp, Thr235Gln, Lys253Ala, and Lys253Asp), designed to impose steric hindrance or restriction of its mobility, had no significant influence on T4 binding. However, binding affinity was 20-fold reduced by introduction of an N-linked glycosylation site at the turn between strands 2B and 3B (Leu246Thr) without compromising the proper folding of this mutant as assessed by immunological methods. In most other serpins, this glycosylation site is highly conserved and has been shown to be crucial for cortisol binding of corticosteroid-binding globulin, the only other member of the serpins with a transport function. The ligand-binding site could thus be located to a highly aromatic environment deep within the b-barrel. The importance of the binding site’s aromatic character was investigated by exchanging phenylalanines with alanines. Indeed, these experiments revealed that substitution of Phe249 in the middle of strand 3B completely abolished T4 binding, while the substitution of several other phenylalanines had no effect. (Molecular Endocrinology 13: 1864–1872, 1999) 0888-8809/99/$3.00/0 Molecular Endocrinology Copyright © 1999 by The Endocrine Society
INTRODUCTION In human serum, 70% of thyroid hormones is bound to thyroxine-binding globulin (TBG), a 54-kDa glycoprotein of hepatic origin (1, 2). Biochemical interest in TBG stems mainly from its exceptionally high binding affinity for T4 [association constant (Ka) 5 1010 M21] and T3 (Ka 5 109 M21) (3). In contrast to the other T4-binding proteins, transthyretin and albumin, its ligand interaction is undefined at a molecular level as attempts to crystallize TBG have failed (4), probably due to its microheterogenic glycosylation (5). TBG belongs to the superfamily of serine-proteinase inhibitors (serpins) (6), a functionally heterogeneous group of more than 100 proteins, including a1-proteinase inhibitor (a1PI), a1-antichymotrypsin, and corticosteroid-binding globulin (CBG). TBG and CBG are the only serpins with a transport function for small hydrophobic ligands. The crystallographic structures of several serpins have been determined and were found to be highly conserved (reviewed in Refs. 7 and 8). The structure of the archetypal a1PI consists of three large b-sheets (A–C) and eight well defined helices (9). This structure model has been successfully used as template for structure-function correlations of other serpins (7), including the heat-resistant TBG-Chicago variant (10). Since natural variants with binding defects invariably also have low expression levels and increased serum concentrations of the denatured molecule, the correlation of their mutations with the structure model has been of limited value (reviewed in Refs. 11 and 12). Binding studies of T4 analogs support the notion that all parts of the molecule participate in its avid binding to TBG (13). Considering the high binding affinity, T4 would seem to bind deep in a binding pocket, similar to T4 binding to transthyretin (14) rather than binding to the flat pocket in albumin (15). In analogy to the ligand-binding site of transthyretin, a barrel of b-strands, formed by sheets B and C, has 1864
Ligand Binding Site of TBG
been proposed as the binding domain of TBG (16). This is supported by affinity cross-linking of Lys253 (17), which maps to the binding cavity. The b-barrel of TBG has also recently been transferred to the a1PI scaffold, generating a T4-binding chimera, confirming that this structural domain encompasses the ligandbinding site (18). A region corresponding to strands 2B and 3B of the eight-stranded b-barrel is highly conserved when compared with TBG from sheep (19) and rat (20) (Fig. 1), all of which have the same high T4binding affinity (21). In this paper, we present a characterization of this phylogenetically conserved region by site-directed mutagenesis.
RESULTS Design of TBG Mutations TBG from sheep (19) and rat (20) share 82% and 76% of their amino acid sequence with human TBG, respectively. However, within the putative ligand-binding domain, a stretch of 26 residues (Val228-Lys253) is almost completely conserved in all known TBG sequences. This region corresponds to two antiparallel b-strands, 2B and 3B, of the b-barrel motif of a1PI. Comparison of TBG with other serpins reveals several distinct differences (Fig. 1). At the entrance to the b-barrel, the Thr235 of TBG corresponds to the bulkier Trp238 in a1PI. This substitution has been speculated to block the access to the b-barrel in a1PI (7, 22, 23), which has no known ligand. To test this hypothesis, Thr235 of TBG was substituted by a tryptophan (TBGmB). This mutation destroys the glycosylation site 233–235, which is unique to and conserved in all known TBG sequences. The corresponding glycosyl-
1865
ation site of other serpins is located at residues 247– 249 (a1PI numbering) and lies at the opposite end of the b-barrel. Mutation mA (Cys2343Trp) introduces a tryptophan at the entrance to the b-barrel but keeps the glycosylation site 233–235 intact. Mutation mC (Thr2353Gln) destroys the glycosylation site, but the small glutamine should not block access to the b-barrel. Mutation mD (Leu2463Thr) generates an intact glycosylation site at 244–246, corresponding to the highly conserved glycosylation site of other serpins. Compared with normal TBG (TBG-N), mutant mD thus has an additional glycosylation site. Access to the b-barrel could also be dependent on the mobility of the involved b-strands. In a1PI, a1-antichymotrypsin, and several other serpins, but not in TBG, the b-strands 2B and 3B are held together by a salt bridge. Substitution of Val228 by glutamic acid (mG) allows the formation of a salt bridge between residues 228 and 253 of TBG and thus might impede T4 binding. Binding studies with T4 analogs (13) suggest that interactions of the negatively charged oxygen and carboxy groups and the positively charged amino group of the T4 molecule are important for the high-affinity binding. An attractive candidate for electrostatic interactions is the positively charged Lys253, which has been shown to be near the ligand-binding domain by affinity labeling (17) and lies at the entrance of the b-barrel. Lys253 was substituted with the negatively charged aspartic acid (mE) and the neutral alanine (mH) to test whether electrostatic interactions at this position play a role in T4-binding. However, interactions within the binding cavity would appear to be more important than those with the entrance of the b-barrel. It has been speculated (22) that the iodines of T4 could interact with aromatic residues, comparable to mercury-iodide binding to sperm whale myoglobin (24). To test this
Fig. 1. Sequence Alignment of TBG from Different Species with Other Serpins The sequence between V228 and E254 of TBG is phylogenetically highly conserved. The only mismatches are at position 230 and 243. In the a1PI structure model this region forms strands 2B and 3B, which contribute to the b-barrel motif. Note that TBG has an Asn-linked glycosylation consensus site at the beginning of s2B (boxed), whereas in other serpins a glycosylation site is found near the turn between s2B and s3B and thus at the opposite side of the b-barrel. hTBG, Human TBG; rTBG, rat TBG; sTBG, sheep TBG; ACh, a1-antichymotrypsin.
MOL ENDO · 1999 1866
hypothesis, phenylalanines thought to contribute to the aromatic interior of the binding cavity were individually substituted by alanine (mutations mF249, mF284, mF288, mF371, and mF315). All mutants were constructed by site-directed mutagenesis and verified by dideoxy sequencing as described in Materials and Methods. Expression of Normal and Mutant TBG in Reticulocyte Lysate The linearized vectors for TBG-N and the mutants were transcribed in vitro and translated in reticulocyte lysate. All mutants were synthesized with comparable efficiency and had identical patterns of nonglycosylated forms on SDS-PAGE (data not shown). After addition of canine microsomal membranes to the translation reactions, all variants had a major band at 44 kDa, which corresponded to unglycosylated TBG after signal peptide processing, and several new bands of higher mol wt, which reflected the number of available N-glycosylation sites (Fig. 2). As expected, mutants mB and mC had one less glycosylation site and mutant mD had an additional glycosylation site when compared with TBG-N. The phenylalanine mutations were not expressed in the reticulocyte lysate system, since no effect on glycosylation was expected.
Vol 13 No. 11
and incubated in medium with [35S]methionine. The TBG mutants synthesized and secreted into the medium were then submitted to SDS-PAGE (Fig. 3). The broad appearance of the bands is due to the microheterogenic glycosylation, which is faithfully reproduced in the Xenopus system. In addition, variants mB–mH differed by size depending on the number of glycosylation sites. However, no significant differences of the amount of secreted protein could be observed when compared with TBG-N except for mutant mA, which was not produced by the oocytes in several experiments (Fig. 3A). Of the phenylalanine mutants, mF284–mF371 were synthesized by the oocytes in the same amount as TBG-N (data not shown). TBG mF249 was also expressed with a mol wt equal to TBG-N, but in slightly lower amounts (Fig. 3B). Analysis of T4 Binding to TBG The amount of functionally active TBG synthesized in Xenopus oocytes was examined by measurement of the T4 binding characteristics. The TBG mutants were expressed as described above, but without the addi-
Expression of Normal and Mutant TBG in Xenopus Oocytes Oocytes were removed from Xenopus laevis, culled, injected with TBG synthetic messenger RNA (sRNAs),
Fig. 2. Synthesis of Normal and Mutant TBG in Reticulocyte Lysate In vitro transcribed RNA of the TBG mutants was translated in rabbit reticulocyte lysate. In the presence of canine microsomal membranes (CMM), all variants showed a major band (44 kDa) corresponding to unglycosylated TBG with a processed signal peptide. The number of the slower migrating bands reflects the number of N-glycosylation sites available for core glycosylation within the microsomes. Note the introduction of an additional glycosylation site in mutant mD, the deletion of a glycosylation site in mutants mB and mC, and the detrimental effect of mutant mA on the amount of the completely glycosylated product. MWM, Mol wt marker (69K, 46K).
Fig. 3. SDS-PAGE Analysis of TBG Variants Expressed in Xenopus Oocytes Oocytes injected with sRNAs were incubated in the presence of [35S]methionine. A, The mutants mB–mH differed by size depending on the number of glycosylation sites. While mutant mA was not secreted into the medium, no significant differences in the efficiency of synthesis and secretion of the other TBG mutants compared with TBG-N were found in four independent experiments. B, Of the Phe3Ala mutants, mF249 was secreted in slightly reduced amounts compared with TBG-N, while mF284, mF288, mF315, and mF371 differed neither in expression level nor in apparent mol wt from TBG-N (not shown). MWM: 14C-labeled mol wt markers.
Ligand Binding Site of TBG
tion of [35S]methionine. Scatchard analysis of the secreted TBG mutants revealed no significant differences in T4 binding affinity and capacity/oocyte of the variants mB, mC, mG, and mH (Fig. 4A and Table 1). However, the T4 binding affinity of mD was 20 times lower than that of TBG-N. Of the phenylalanine mutations, only mF249 showed a significant reduction of binding affinity, by at least 2 orders of magnitude (Fig. 4B and Table 1).
1867
Table 1. Affinity Constants (Ka) of T4 Binding of the TBG Mutants Variant
Ka (z 1010
21 a
Variant
TBG-N mB mC mD mE mG mH
1.40 6 0.18 1.42 6 0.19 1.42 6 0.29 0.078 6 0.19 1.14 6 0.36 1.41 6 0.20 1.17 6 0.32
TBG-N mF284 mF288 mF315 mF371 mF249
M
)
Ka (z 1010
M
21 b
)
1.63 1.58 1.66 1.55 1.42 (#0.02)c
Mean 6 SD of three experiments. Mean of two experiments. c Estimated upper limits. The markedly reduced T4 binding capacity did not allow Scatchard analysis. a
b
Heat Resistance of TBG Variants To determine the functional stability of the TBG variants, heat denaturation was performed. At 58 C, no significant differences in the rate of heat denaturation of TBG-N and the variants mB, mC, mE, mG, and mH were found (Fig. 5). Immunoprecipitation of TBG-N and Variant mD To test for the immunological integrity of variant mD, samples were immunoprecipitated with either a polyclonal antibody against both nTBG and dnTBG or an antibody specific for dnTBG. Both TBG-N and mD were almost exclusively recognized by anti-nTBG (Fig. 6A). In addition, immunoprecipitation in the presence of excess nTBG or dnTBG, respectively, revealed that both TBG-N and TBG mD were competed only by nTBG, indicating that both molecules were correctly folded (Fig. 6B).
DISCUSSION
Fig. 4. Scatchard Analysis of T4-Binding to TBG Variants Expressed in Xenopus Oocytes TBG-N and the TBG mutants were expressed in oocytes. The secreted TBGs were incubated with [125I]T4 and increasing amounts of unlabeled T4. Of all mutants, no significant differences in T4-binding affinity (slope) and binding capacity/ oocyte (intercept) were found, except for TBG mD and mF249. A, The binding affinity (Ka) of mutant mD was almost 20-fold lower than that of TBG-N (108 M21 vs. 1010 M21). (F), TBG-N; E, mB; ‚, mC; f, mD; Œ, mE; M, mG; L, mH. B, No Scatchard analysis could be obtained from mutant mF249 due to its substantially reduced specific T4 binding with an estimated upper limit of its Ka of at least 2 orders of magnitude lower than TBG-N or the other F3A variants, respectively. F, TBG-N; E, mF284; M, mF288; ‚, mF315; L, mF371.
In this paper we describe a rational approach to define the impact of individual residues of TBG on T4 binding by site-directed mutagenesis. Previous attempts to characterize the binding site of TBG have relied on the identification of naturally occurring mutations. So far, all of the six binding-deficient variants characterized on a molecular level exhibit lower expression levels than TBG-N and more or less explicit heat lability (11, 12, 25). Since these natural variants are also characterized by increased dnTBG serum concentrations, they most likely decrease T4 binding by gross structural changes. Thus, the analysis of natural occurring TBG variants has failed to contribute to our understanding of the structural requirements of TBG for T4 binding, consistent with the location of the affected residues throughout the molecule but not within the b-barrel. Apart from mutant mA, all TBG mutants generated in this study were synthesized and secreted by Xenopus
MOL ENDO · 1999 1868
Vol 13 No. 11
Fig. 5. Heat Stability of the TBG Variants TBG variants synthesized by Xenopus oocytes were heated at 58 6 0.1 C for different time intervals, and the residual T4 binding activity was determined. Values are expressed as protein-bound T4 relative to the basal levels. F, TBG-N; E, m; ‚, mC; Œ, mE; M, mG; L, mH.
oocytes in similar amounts as TBG-N. The respective amino acid substitutions can thus be assumed not to interfere with the folding and processing of the molecule, because this has been found to lead to lower expression levels in all known natural mutations (26, 27). Mutants mB, mC, mG, and mH even had a normal heat stability. The secretion defect of mA most likely results from inefficient glycosylation of the neighboring Asn233, as demonstrated by the reduced utilization of this glycosylation site in the cell-free translation system (Fig. 2). This finding is compatible with a previous report on the glycosylation efficiency of N-linked core glycosylation, where glycosylation sites with Trp in the X-position of an Asn-X-Ser sequon were found to be inefficiently processed (28). Neither the introduction of a putative salt bridge between the b-strands, nor the sterical hindrance by introduction of a bulky tryptophan, nor modification of the glycosylation at the entrance to the ligand-binding site had an impact on the T4 binding affinity. A major effect, however, was seen after introduction of a new glycosylation site at position 244 (mutant mD), a site highly conserved in other serpins (Fig. 1). The integrity of mutant mD could not be ascertained by heat denaturation due to its low binding capacity. However, immunological studies showed that it was detected only with antibodies against nTBG and not dnTBG and could only be displaced with nTBG and not dnTBG from polyclonal antibodies (Fig. 6). Therefore, introduction of the additional glycosylation site in mutant mD appears to cause a specific binding defect. As shown in a model of the TBG b-barrel (Fig. 7), Leu246 faces into the solvent rather than into the binding pocket. Its substitution by threonine in TBG-mD should thus not interfere with binding per se, but rather by causing glycosylation of Asn244. The new carbohydrate could cause the loss of binding either by a direct sterical effect or by modulation of
Fig. 6. Immunoprecipitation of TBG-N and TBG mD synthesized in Xenopus Oocytes A, [35S]methionine-labeled samples were immunoprecipitated with either a polyclonal antibody that recognizes both nTBG (n) and dnTBG (dn) or an antibody that recognizes only dnTBG and then submitted to SDS-PAGE and autoradiographed. Both TBG-N and TBG mD were recognized by anti-n/dnTBG (lanes 2 and 4) but only weakly by the antidnTBG antibody (lanes 3 and 5). B, Samples were immunoprecipitated with the anti-n/dnTBG antibody without (2) and with competition; n denotes residual nTBG from samples competed with dnTBG; dn denotes residual dnTBG from samples competed with nTBG. Both TBG-N and TBG mD were competed only by nTBG, indicating that the molecules were correctly folded. MWM, 14C-labeled mol wt markers.
the processing of the ligand-binding site during synthesis of the molecule. The respective carbohydrate attached to the corresponding Asn247 of a1PI does extend through the interior of the binding pocket, as shown by its interaction with Trp238 at the opposite end of the b-barrel (29). Molecular modeling of a1PI (9) also reveals that the carbohydrate moiety linked to this site anchors the turn between strands 2B and 3B to helix D. Accordingly, the binding pocket in a1PI is smaller than in ovalbumin, the only other serpin lacking glycosylation of the corresponding site (30). Introduction of the mutation mD in TBG could thus reduce binding by diminishing the available space for T4. In contrast, this glycosylation site has been previously shown to be crucial and sufficient for corticosteroid binding of CBG (31). CBG normally contains six glycosylation sites and inactivation of only the Asn238, which corresponds to the Asn244 of TBG, abolished
Ligand Binding Site of TBG
1869
Fig. 7. Structure Model of the TBG b-Barrel TBG was modeled based on the structure of the archetypical serpin a1PI. The b-barrel motif is depicted in green, strands s2B and s3B are depicted in brown, and residues relevant for T4 binding are highlighted in different colors. Phe249 faces into the binding cavity, where it could interact with one of the T4’s iodines. The Leu246, which is substituted by Thr in TBG-mD, faces away from the binding pocket into the solvent. Its substitution per se should thus not interfere with T4 binding, but rather by causing glycosylation of Asn244. Substitution of Lys253, which can be affinity labeled (17), had no significant effect on T4 binding, which supports the concept that T4 binds deep in the b-barrel. Also shown is Ala191, which is substituted by Thr in the binding deficient variant TBG-Aborigine (39).
steroid binding. A mutant containing only this glycosylation site (the other five were inactivated) had normal steroid binding, which was not altered by deglycosylation of the molecule. This suggests that interaction of the polypeptide with the carbohydrate at the glycosylation site Asn238 is necessary for the folding and creation of the steroid-binding site only during CBG biosynthesis (32) and implicates substantial differences in the structure of the ligand-binding sites of TBG and CBG. The reduced binding of mD taken together with the lack of effect of the mutations at the putative entrance of the b-barrel (7) would be compatible with T4 entering the binding cavity from the opposite side of the molecule, as proposed by Terry and Blake (22) and Jarvis et al. (23). While the latter entrance would be conceivably wider, the entrance at the opposite side would still be wide enough to allow T4 with its dimensions of 6312 Å to enter and would directly be restricted by the additional carbohydrate group in TBG-mD. Binding of T4 by TBG depends on the presence of its iodines, especially at the 59-position (13). It has been speculated that these might interact with the aromatic, rather than just hydrophobic, interior of the binding site, namely phenylalanine, and possibly also tyrosine residues (22). Of the four phenylalanines within the binding cavity of TBG (Phe249,
Phe284, Phe288, and Phe371) and the Phe315 outside the b-barrel, only substitution of Phe249 with alanine (mF249) substantially diminished T4 binding (Table 1). This residue is located in the middle of strand 3B and thus in close spatial proximity to mutation mD, as shown in the model of the interior of the ligandbinding site (Fig. 7). In conclusion, mutations mD and mF249 locate the binding site deep within the b-barrel of TBG. A more precise structural analysis of T4 binding awaits crystallization of TBG, attempts at which have now been intensified.
MATERIALS AND METHODS Site-Directed Mutagenesis The cDNA of normal TBG (TBG-N) had been subcloned previously into the pSELECT expression vector (33). Sitedirected mutagenesis (Altered Sites Kit, Promega Corp., Madison, WI) with oligonucleotides (see Table 2) was performed to obtain vectors coding for the TBG mutants mA–mH and the F3A variants (mF249, mF284, mF288, mF315, and mF371). The coding regions of all TBG mutants were sequenced in its entirety by the dideoxy-nucleotide termination method (34) using the Sequenase-2 kit from United States Biochemical Corp. (Cleveland, OH).
MOL ENDO · 1999 1870
Vol 13 No. 11
Table 2. Oligonucleotide Primers Used for Site-Directed Mutagenesis
a
Mutation
59-Position
Sequencea
mA mB mC mD mE mG mH mF249 mF284 mF288 mF315 mF371
750 751 751 784 805 732 805 790 895 907 988 1156
59-GGAATTGAACTGGACAGTTCTGCAA-39 59-GAATTGAACTGCTGGGTTCTGCAAATG-39 59-GAATTGAACTGCCAAGTTCTGCAAAT-39 59-AGCAAGAATGCTACGGCACTCTTTGT-39 59-TTTGTTCTTCCCGATGAGGGACAGATG-39 59-ACTATATCACCTAGAGGATATGGAATT-39 59-TTTGTTCTTCCCGCGGAGGGACAGATG-39 59-AATGCTCTGGCACTCGCTGTTCTTCCCAAGGA-39 59-GGATGGGTTGACTTGGCTGTTCCAAAGTTTTC-39 59-TTGTTTGTTCCAAAGGCTTCCATTTCTGCCAC-39 59-TCTGAAAATGCTGATGCTTCTGGATCCACAGA-39 59-CAAATTGATAGATCTGCCATGTTGTTGATTTT-39
Mutations are underlined.
In Vitro Transcription
Measurement of T4 Binding to TBG
sRNA was prepared with the Gemini-II in vitro transcription kit and T7 RNA polymerase according to the recommendations of the supplier (Promega Corp.).
Parameters of T4-binding to TBG were measured by a method previously described in detail (39). Briefly, TBG preparations were incubated with [125I]T4 (DuPont-NEN) in the presence of increasing amounts of unlabeled T4. TBG-bound T4 was separated from free T4 with anion exchange resin beads (Mallinckrodt, Inc., St. Louis, MO) and the proteinbound [125I] activity was determined. The affinity constants (Ka) of TBG preparations were determined by the method of Scatchard (40).
Cell-Free Translation Translations were carried out in rabbit reticulocyte lysate in a final volume of 50 ml according to the protocol supplied by the manufacturer (Promega Corp.). Samples were metabolically labeled with [35S]methionine (DuPont-NEN, Boston, MA). For the analysis of early cotranslational events, eight equivalents of canine microsomal membranes were added. Preparation of Oocytes and RNA Injection Ovaries were removed from mature Xenopus laevis (H. Ko¨hler, Hamburg, Germany) (35) and suspended in OR-IIa medium (83 mM NaCl, 2.5 mM KCl, 1 mM MgCl2, 1 mM Na2HPO4, and 5 mM HEPES, pH 7.6) (36). After manual dissection of the follicles, oocytes were dissociated from the surrounding connective tissue by incubation in OR-IIa containing 0.2% collagenase type IA (Sigma Chemical Co., St. Louis, MO) for 2 h with shaking at room temperature. The liberated oocytes were then rinsed extensively in OR-IIb medium (OR-IIa with 1 mM CaCl2 and 100 mg/ml gentamycin), and stage VI oocytes (37) were separated and kept up to 3 days in OR-IIb with daily medium changes. After injection with 100 nl of sRNA (0.5 mg/ml), oocytes were kept on ice for 1 h and then for 2–6 h at 19 C in OR-IIb. Intact oocytes (routinely .95%) were transferred to 24-well plates (Costar) and kept in OR-IIc (OR-IIb with 1 mM sodium pyruvate), 5 ml/oocyte, at 19 C for up to 4 days, with daily exchange of medium. Typically, 100 oocytes were injected with each sRNA preparation. Control oocytes were either injected with water or noninjected, with identical results. In some experiments, proteins synthesized by the oocytes were metabolically labeled by addition of 250 mCi [35S]methionine per 500 ml medium. SDS-Gel Electrophoresis Products of cell-free translation or proteins synthesized in Xenopus oocytes were analyzed by the method of Laemmli (38), using 10% polyacrylamide gels. Gels were dried and autoradiographed at 290 C on X-AR5 film (Eastman Kodak Co.) with an intensifying screen.
Immunoprecipitation Immunoprecipitation was performed according to Kessler (41), with minor modifications as described (33). The polyclonal antibodies used in this study (against both nTBG and dnTBG and against dnTBG only) have been described (42). Heat Denaturation The functional stability of the TBG variants was quantified by thermal denaturation in a water bath at 58 6 0.1 C for various periods of time. The samples were then cooled on ice and centrifuged for 15 min at 13,000 3 g to remove precipitated protein. Residual specific T4 binding capacity was expressed relative to controls kept at 4 C. The half-lives (t1/2) of heat denaturation were calculated by least square analysis of semilogarithmic plots of remaining specific T4 binding vs. time of incubation. Experimental Animals All animal studies were conducted in accord with the principles and procedures outlined in the “Guidelines for Care and Use of Experimental Animals.”
Acknowledgments We thank S. Refetoff (University of Chicago, Chicago, IL) for kindly providing us with anti-TBG antisera and preparations of nTBG and dnTBG. We also thank P. Gardner (Howard Hughes Medical Institute, University of Chicago, Chicago, IL) for the synthesis of oligonucleotide primers and R. Huber and R. Engh (Max Planck Institute for Biochemistry, Martinsried, Germany) for help with the structural data. Received April 30, 1999. Re-revision received July 16, 1999. Accepted July 19, 1999.
Ligand Binding Site of TBG
Address requests for reprints to: Onno E. Janssen, M.D., Molecular Thyroid Study Unit, Department of Medicine, Klinikum Innenstadt, Ludwig-Maximilians-University, Ziemssenstrasse 1, D-80336 Munich, Germany. E-mail: Onno.E.
[email protected]. * Current address: Thyroid Division, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts 02115. † Current address: Department of Pathophysiology, Jiamusi Medical College, Jiamusi, Heilongjiang, 154002 China.
REFERENCES 1. Oppenheimer JH 1968 Role of plasma proteins in the binding, distribution, and metabolism of the thyroid hormones. N Engl J Med 278:1153–1162 2. Refetoff S, Larsen PR 1989 Transport, cellular uptake, metabolism of thyroid hormone. In: DeGroot LJ (ed) Endocrinology. WB Saunders, Philadelphia, chapter 38, pp 541–561 3. Korcek L, Tabachnick M 1976 Thyroxine-protein interactions. Interaction of thyroxine and triiodothyronine with the human thyroxine-binding globulin. J Biol Chem 251: 3558–3562 4. Marshall J, Pensky J, Williams S 1973 Studies on human thyroxine-binding globulin. 8. Isoelectric focusing evidence for microheterogeneity of thyroxine-binding globulin. Arch Biochem Biophys 156:456–462 5. Ga¨rtner R, Henze R, Horn K, Pickardt CR, Scriba PC 1981 Thyroxine-binding globulin: investigation of microheterogeneity. J Clin Endocrinol Metab 52:657–664 6. Hunt LT, Dayhoff MO 1980 A surprising new protein superfamily containing ovalbumin, antithrombin-III, and a1-proteinase inhibitor. Biochem Biophys Res Commun 95:864–871 7. Huber R, Carrell RW 1989 Implications of the threedimensional structure of a1-antitrypsin for structure and function of serpins. Biochemistry 28:8951–8966 8. Whisstock J, Skinner R, Lesk AM 1998 An atlas of serpin conformations. Trends Biochem Sci 23:63–67 9. Loebermann H, Tokuoka R, Deisenhofer J, Huber R 1984 Human a1-proteinase inhibitor. Crystal structure analysis of two crystal modifications, molecular model and preliminary analysis of the implications for function. J Mol Biol 177:531–556 10. Janssen OE, Chen B, Bu¨ttner C, Refetoff S, Scriba PC 1995 Molecular and structural characterization of the heat-resistant thyroxine-binding globulin-Chicago. J Biol Chem 270:28234–28238 11. Refetoff S 1989 Inherited thyroxine-binding globulin abnormalities in man. Endocr Rev 10:275–293 12. Janssen OE, Bertenshaw R, Takeda K, Weiss R, Refetoff S 1992 Molecular basis of inherited thyroxine-binding globulin defects. Trends Endocrinol Metabol 3:49–53 13. Snyder SM, Cavalieri RR, Goldfine ID, Ingbar SH, Jorgensen EC 1976 Binding of thyroid hormones and their analogues to thyroxine-binding globulin in human serum. J Biol Chem 251:6489–6494 14. Blake CCF, Geisow MJ, Swan ID, Rerat C, Rerat B 1974 Structure of human plasma prealbumin at 2.5 A resolution. A preliminary report on the polypeptide chain conformation, quaternary structure and thyroxine binding. J Mol Biol 88:1–12 15. He XM, Carter DC 1992 Atomic structure and chemistry of human serum albumin. Nature 358:209–215 16. Flink IL, Bailey TJ, Gustefson TA, Markham BE, Morkin E 1986 Complete amino acid sequence of human thyroxine-binding globulin deduced from cloned DNA: close homology to the serine antiproteases. Proc Natl Acad Sci USA 83:7708–7712
1871
17. Tabachnik M, Perret V 1987 Specific labeling of the thyroxine binding site in thyroxine-binding globulin: determination of the amino acid composition of a labeled peptide fragment isolated from a proteolytic digest of the derivatized protein. Biochem Int 15:409–417 18. Grasberger H, Buettner C, Janssen OE 1999 Modularity of serpins: a bifunctional chimera possessing a1-proteinase inhibitor and thyroxine-binding globulin properties. J Biol Chem 274:15046–15051 19. Tsykin A, Schreiber G 1993 Sheep thyroxine-binding globulin: cDNA sequence and expression. Mol Cell Endocrinol 98:91–97 20. Imamura S, Mori Y, Murata Y, Yamamori I, Miura Y, Oiso Y, Seo H, Matsui N, Refetoff S 1991 Molecular cloning and primary structure of rat thyroxine-binding globulin. Biochemistry 30:5406–5411 21. Refetoff S, Robin NI, Fang VS 1970 Parameters of thyroid function in serum of 16 selected vertebrate species: a study of PBI, serum T4, free T4, and the pattern of T4 and T3 binding to serum proteins. Endocrinology 86:793–805 22. Terry CJ, Blake CC 1992 Comparison of the modelled thyroxine binding site in TBG with the experimentally determined site in transthyretin. Protein Eng 5:505–510 23. Jarvis JA, Munro SLA, Craik DJ 1992 Homology model of thyroxine binding globulin and elucidation of the thyroid hormone binding site. Protein Eng 5:61–67 24. Kretsinger RH, Watson HC, Kendrew JC 1968 Binding of mercuri-iodide and related ions to crystals of sperm whale myoglobin. J Mol Biol 31:305–314 25. Refetoff S, Murata Y, Mori Y, Janssen OE, Takeda K, Hayashi Y 1996 Thyroxine-binding globulin: organization of the gene and variants. Horm Res 45:128–138 26. Marquardt T, Helenius A 1992 Misfolding and aggregation of newly synthesized proteins in the endoplasmatic reticulum. J Cell Biol 117:505–513 27. Gething M-J, McCammon K, Sambrook J 1986 Expression of wild-type and mutant forms of influenza hemagglutinin: the role of folding in intracellular transport. Cell 46:939–950 28. Shakin-Eshleman SH, Spitalnik SL, Kasturi L 1996 The amino acid at the position of an Asn-X-Ser Sequon is an important determinant of N-linked core-glycosylation efficiency. J Biol Chem 271:6363–6366 29. Powell LM, Pain RH 1992 Effects of glycosylation on the folding and stability of human, recombinant and cleaved a1-antitrypsin. J Mol Biol 224:241–252 30. Stein PE, Leslie AGW, Finch JT, Turnell WG, McLaughlin PJ, Carrell R 1990 Crystal structure of ovalbumin as a model for the reactive center of serpins. Nature 347: 99–102 31. Avvakumov GV, Warmels-Rodenhiser S, Hammond GL 1993 Glycosylation of human corticosteroid-binding globulin at asparagine 238 is necessary for steroid binding. J Biol Chem 268:862–866 32. Avvakumov GV, Hammond GL 1994 Glycosylation of human corticosteroid-binding globulin. Differential processing and significance of carbohydrate chains at individual sites. Biochemistry 33:5759–5765 33. Janssen OE, Refetoff S 1992 In vitro expression of thyroxine-binding globulin (TBG) variants: impaired secretion of TBGPro-227 but not TBGPro-113. J Biol Chem 267: 13998–14004 34. Sanger F, Nicklen S, Coulson AR 1977 DNA sequencing with chain terminating inhibitors. Proc Natl Acad Sci USA 74:5463–5467 35. Colman A 1984 Translation of eukaryotic messenger RNA in Xenopus oocytes. In: Hames BD, Higgins SJ (eds) Transcription, Translation: A Practical Approach. IRL Press, Washington DC, pp 271–302 36. Maller JL 1983 Use of microinjection techniques to study protein kinases and protein phosphorylation in amphibian oocytes. Methods Enzymol 99:219–226
MOL ENDO · 1999 1872
37. Dumont JN 1972 Oogenesis in Xenopus laevis (Daudin): stages of oocyte development in laboratory maintained animals. J Morphol 136:153–180 38. La¨mmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685 39. Murata Y, Refetoff S, Sarne DH, Dick M, Watson F 1985 Variant thyroxine-binding globulin in serum of Australian Aborigines: its physical, chemical and biological properties. J Endocrinol Invest 8:225–232
Vol 13 No. 11
40. Scatchard G 1949 The attractions of proteins for small molecules and ions. Ann NY Acad Sci 51:660–672 41. Kessler SW 1976 Cell membrane antigen isolation with the Staphylococcal Protein A-antibody adsorbent. J Immunol 117:1482–1490 42. Refetoff S, Murata Y, Vassart G, Chandramouli V, Marshall JS 1984 Radioimmunoassays specific for the tertiary and primary structures of thyroxine-binding globulin (TBG): measurement of denatured TBG in serum. J Clin Endocrinol Metab 59:269–277
