| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Division of Endocrinology (P.S.K., S.M., B.L., S.A.H., J.-H.B.), University of Cincinnati, Cincinnati, Ohio 45267; Department of Biochemistry (P.J., B.P.), Faculty of Science, Mahidol University, Bangkok, Thailand; and Division of Metabolism, Endocrinology & Diabetes (J.L., P.A.), University of Michigan Medical Center, Ann Arbor, Michigan 48109
Address all correspondence and requests for reprints to: Prof. Peter Arvan, M.D., Ph.D., Chief, Division of Metabolism, Endocrinology & Diabetes, University of Michigan Medical Center, MSRB2 Room 5560, Ann Arbor, Michigan 48109. E-mail: parvan{at}umich.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Secretory proteins must achieve an acceptable folding state to be allowed to leave the ER, and the efficiency of Tg secretion is directly related to the ability of the newly synthesized protein to achieve a native or near-native conformation (4). Indeed, Tg export from the ER has been identified as a critical step in the process of thyroid hormone synthesis (5). In the ER Tg is cotranslationally translocated and undergoes initial N-linked glycosylation and sugar trimming, which is linked to the folding of the nascent polypeptide (6). This includes the formation of approximately 60 intrachain disulfide bonds per Tg monomer. Correct formation of these disulfides is rate limiting for Tg monomer folding (7), which in turn is a prelude to Tg homodimerization and export from the ER (8). Deficiency of Tg export is an established cause of congenital hypothyroid goiter in humans and animal models of the disease (9). We have previously shown that the missense mutation encoding L2263P in the acetylcholinesterase homology region of Tg causes congenital hypothyroid goiter in homozygous cog/cog mice (10). A markedly distended ER was shown to result from accumulation of the mutant Tg (11), with induction of the ER stress response that up-regulates molecular chaperones of the ER lumen (12). Chronic TSH stimulation leads to goitrogenesis in the hypothyroid mice, eventually compensating for Tg deficiency as the adult animals gradually achieve normal serum T4 levels at the expense of a grossly enlarged thyroid gland (13).
A number of distinct point mutants have been identified as causing human hypothyroid goiter (14). Specifically, Q851H (residue 870 when counting the signal peptide), C1976S (1995 with signal peptide), and C1244R (1263 with signal peptide) alleles have been reported in human patients with congenital goiter. As far as is known, each of these alleles is partially or completely defective for intracellular transport from the ER (15).
In the current study, we have investigated which missense mutation might be responsible for congenital hypothyroid goiter from among three that coexist in the index patient from an inbred Brazilian kindred (16). By introducing individual nucleotide substitutions into the Tg cDNA, we have been able to assess the relative significance of each mutation, and from this conclude that the primary defect derives from a Tg R19K mutation. This result is quite surprising because of conservation of a positively charged residue at this position, yet we show that Tg-R19K is retained within the ER, bound to ER chaperones, disposed of by ER-associated degradation, and is dependent not on the charge of residue R19 but the packing of the residue side chain within the amino-terminal domain of Tg.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
As shown in Fig. 1A
, with these maneuvers, background signal in the Tg region of the Western blot was eliminated both in cell lysates and media. Moreover, introduction of the C-terminal myc-6xHis tag did not block intracellular expression nor secretion of wild-type Tg. Both Tg-R835Q and Tg-V2452L substitutions yielded significant amounts of secreted Tg (Fig. 1A). However, despite similar levels of intracellular protein, only a small amount of Tg-R19K was secreted (Fig. 1A).
|
). By contrast the Tg-L2263P (cog) appeared maximally defective with virtually no secreted Tg (Fig. 1B), i.e. even more severe that seen for Tg-R19K (Fig. 1A).
Intracellular Disposal of Tg-R19K
COS-7 cells transiently expressing Tg-R19K were pulse-labeled with 35S-amino acids and chased for various times. At each chase time, the cells were lysed and immunoprecipitated with anti-Tg, and analyzed by reducing SDS-PAGE followed by phosphorimaging. Unlike smaller misfolded secretory and membrane proteins that may be degraded very rapidly, a significant fraction of Tg-R19K mutant survived the first 6 h after synthesis, eventually disappearing within cells over 18 h (Fig. 2A, quantified in Fig. 2B). These kinetics appear similar to those reported previously for the degradation of cog Tg [L2273P (17)] and rdw Tg mutants [G2320R (18)].
|
During a 1-h labeling period, wild-type Tg is likely to have already undergone considerable early folding. Indeed, at the zero chase time, relatively little normal Tg was coprecipitated with BiP, calnexin, or PDI (Fig. 3A). This fraction fell further as a function of chase time (as did the amount of directly immunoprecipitable intracellular Tg) because in addition to folding, labeled wild-type Tg is continually secreted from cells, decreasing the intracellular content. In comparison to wild-type Tg, an increased fraction of Tg-R19K was coprecipitated with BiP, calnexin, and PDI at the zero chase time (Fig. 3B). Furthermore, over the first 6 h of chase, the absolute amount of labeled Tg-R19K coprecipitated with these ER chaperones did not fall, even as the total amount of labeled Tg (directly precipitated with anti-Tg) was decreasing (Fig. 3B). With prolonged chase, as disposal of Tg-R19K was ongoing, the percent of remaining Tg-R19K coprecipitated by these ER chaperones grew further still (Fig. 3B), with PDI coprecipitating approximately half of residual labeled mutant Tg at 24 h (Fig. 3C). Thus, whereas Tg-R19K is not secreted (Fig. 1), the intracellular population of Tg-R19K molecules are extensively engaged with, and retained by, bound ER molecular chaperones before their intracellular disposal.
|
, upper panel). However, the abnormal chaperone binding to Tg-R19K (Fig. 3C) had no apparent effect on the synthesis or secretion of SEAP (Fig. 3D, lower panel), indicating normal function of the secretory pathway in these cells despite impaired intracellular transport of Tg.
Tg-R19K Fails to Undergo Anterograde Transport through the Golgi Complex
Tg has multiple N-linked glycans that are normally processed from endoglycosidase H (endo H) sensitive to endo H resistant after arrival in the Golgi complex (6). To check whether the Tg-R19K mutant could be transported through the Golgi, 35S-amino acid pulse-labeled COS-7 cells that had been transiently transfected with Tg-R19K were chased for various times, lysed, digested with endo H, and analyzed by reducing SDS-PAGE and phosphorimaging. At all chase times, Tg-R19K appeared endo H sensitive even as the newly synthesized protein disappeared from the cells (Fig. 4). These results indicating a failure of the Tg-R19K mutant to reach the site of Golgi carbohydrate modification, strongly imply intracellular disposal of Tg-R19K by ER-associated degradation (ERAD). This result is similar to that observed for cog Tg (12) that is highly defective in secretion (Fig. 1B), whereas thyroid homogenates of patients expressing Tg-C1244R or Tg-C1976S have been found to contain endo H-resistant Tg (20). Taken together, these data suggest that the folding defect of Tg-R19K is relatively more severe than those of a number of other human Tg missense mutations that have been reported previously.
|
). Addition of either MG132 or KIF comparably inhibited disposal of both Tg mutants (Fig. 5) indicating, to a first approximation, that Tg-R19K follows a similar ERAD route to that identified for cog Tg. (Addition of MG132 and KIF together did not further inhibit ERAD of mutant Tg; not shown.)
|
, in cells treated with MG132 or KIF, the total amounts of Tg-R19K coprecipitated with BiP, calnexin, and PDI were increased when compared with the untreated cells. However, as a percent of labeled intracellular Tg remaining at 18 or 24 h, the fraction coprecipitated with these ER chaperones was unchanged. These data are consistent with the idea that inhibiting retrotranslocation or proteasomal-mediated disposal of Tg-R19K causes a backlog of undegraded molecules upstream in the ERAD pathway but does not change the nature of the recognition of mutant Tg by ER chaperones.
Effect of the R19K Mutation on the Secretion of Severely Truncated Tg
Full-length wild-type Tg is thought to undergo extensive monomer folding and homodimerization before export from the ER. Such monomer folding, leading to dimerization, is thought to importantly require the carboxyl-terminal acetylcholinesterase-homology domain comprising approximately the final 500 amino acids of the Tg primary sequence (21). However, van de Graff and colleagues (22) described a patient that was able to synthesize T4 despite a homozygous defect resulting in expression of a Tg protein truncated after residue 276. Whereas homodimerization of such a highly truncated Tg might not be expected, we nevertheless assumed that monomeric truncated Tg might pass ER quality control standards and be allowed export through the secretory pathway. With this in mind, using site-directed mutagenesis, a stop codon was introduced in place of Tg residue Cys175 (called Tg-C175Z, the stop codon would be numbered as residue 194 when including the 19-residue signal peptide). The foreshortened Tg could be specifically detected by Western blotting of COS-7 cell lysates with an antibody raised against the N terminus of Tg (Fig. 6, left). Moreover, Tg-C175Z was secreted as a truncated Tg protein (Fig. 6, right). However, secretion was blocked in the Tg-R19K,C175Z double mutant (Fig. 6). Thus, introduction of the R19K point mutation must induce local misfolding within the amino-terminal 174 residues of Tg. Such a result seems quite surprising because R-to-K represents a substitution conserving the positive charge at Tg residue 19.
|
, top): R13, R19, and K25. (A fourth position, K21 of mouse Tg is not conserved in humans, in which T21 is expressed.) To determine whether the specific charged residues in this region are essential to the folding of Tg, we examined mutants encoding Tg-R13K and Tg-K25R in comparison to Tg-R19K. However, the R13K and K25R mutants were entirely benign with respect to the ability of Tg to be secreted, whereas Tg-R19K secretion was fully blocked (Fig. 7, bottom). A Tg-R19Q was also prepared because, although the carboxamide side chain cannot mimic the guanidinium side chain for charge, Q is a better mimic for R in terms of the shape of the side chain than is the linear four-carbon side chain of K (23). Interestingly, a Tg-R19Q substitution was well tolerated, leading to an efficiently secreted Tg (Fig. 7, bottom). These data underscore that residue R19 plays a structural role in the amino-terminal region of Tg, and disruption of local folding in this region is recognized by ER chaperones, blocking protein trafficking from the ER and leading to congenital hypothyroid goiter.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
At an initial glance, Tg-R19K would appear unlikely to be a pathogenic mutation because it appears to be a conservative substitution, maintaining a positive charge. Indeed, similar substitutions within the first 25 residues of Tg, including R13K and K25R, have no discernible secretory defect (Fig. 7). However, the Tg-R19K mutant fails to be secreted (Fig. 1) and instead is disposed of intracellularly (Fig. 2). For the entire duration of intracellular existence of Tg-R19K, an abnormally increased amount of the protein is associated with ER molecular chaperones BiP, calnexin, and PDI (Fig. 3), whereas proteasome inhibition or inhibition of ER mannosidase-I inhibits Tg disposal and increases the ER chaperone association still further (Fig. 5). Given that the Tg-R19K protein disappears via proteasomal digestion without ever detectably arriving in the Golgi complex (Fig. 4), these findings meet all the criteria of ERAD of misfolded secretory glycoproteins (24, 25) indicating a cell biologically (and clinically) relevant perturbation in the final tertiary structure of Tg. In cells expressing Tg-R19K, the secretory pathway continues to work properly as judged by ongoing synthesis and secretion of SEAP (Fig. 3D), indicating that ER quality control is inhibiting transport of only those substrates deemed to be improperly folded. Interestingly, an R19Q substitution did not block Tg secretion (Fig. 7), suggesting a Tg structural requirement that is more related to the shape of the side chain at this position. Finally, the complete block of secretion of the truncated Tg-C175Z upon introduction of the R19K mutation strongly suggests that the structural defect involves local misfolding (Fig. 6).
The first approximately 1,000 residues of Tg include multiple type-1 repeat domains that are roughly 60–70 residues in length (26), with the so-called type-1A repeats containing three disulfide bonds (1). With homology to GA733-2 in which precise disulfide pairs have been mapped, the Tg type-1A repeat involves sequential disulfides in the pattern Cys1-Cys2, Cys3-Cys4, Cys5-Cys6; in which Cys4 and Cys5 are separated by only one residue forming the CWC motif, yet Cys4 and Cys5 of this motif are not covalently engaged with each other but rather to upstream Cys3 and downstream Cys6 residues, respectively. This leaves Cys1 to directly partner with Cys2 (27). In the very first type-1A repeat domain of authentic Tg, the CWC motif involves residues 51 and 53. By structural homology, these residues are predicted to engage upstream Cys44 and downstream Cys73 residues, respectively, leaving Cys15 to partner with Cys33. Therefore, one interesting hypothesis that merits further testing is that Tg-R19K may adopt a conformation that disrupts local structure and inhibits the ability of Cys15 to engage in a disulfide bond with its normal partner.
In any event, the increased binding of BiP to Tg-R19K suggests abnormal exposure of one or more hydrophobic patches (28, 29), and such binding increases Tg retention within the ER (30). Calnexin is not normally a major posttranslational binder of wild-type Tg (Fig. 3A) (19) [although calreticulin is (7)], but calnexin and presumably the ERp57 oxidoreductase (31) coprecipitates a detectable quantity of the Tg-R19K mutant. Perhaps the most intriguing Tg-R19K association is with PDI. Normally, wild-type Tg association with PDI is quantitatively insignificant (Fig. 3A), comprising only a subfraction of newly synthesized Tg adduct B that itself is a subfraction of all adducts that comprise a small minority of labeled Tg recovered at the zero chase time after a short pulse-labeling (31). This contrasts strikingly with the large fractional coprecipitation of Tg-R19K with PDI, which has recently been strongly implicated to play a selective role in retrotranslocation of misfolded ERAD substrates including Tg (18, 32). Certainly, the present data do not exclude (and are indeed consistent with) the idea that PDI might act as an unfoldase to prepare retained Tg-R19K for retrotranslocation (33).
Most human studies of Tg defects have been limited to immunohistology of tissue sections and Western blotting of thyroid homogenates. In this report, the fate of the Tg-R19K protein has been characterized in comparison to the expression of a number of established missense mutants (Q851H, C1244R, C1976S, and L2263P) in COS-7 cells. Near-normal secretion of Tg-Q851H seems to correlate with a milder clinical phenotype (goiter without biochemical hypothyroidism). Similarly, expression of Tg-C1244R and Tg-C1976S exhibit moderate secretory defects that seem to correlate with a moderate clinical phenotype (20, 34). However, a preliminary report has suggested that expression of Tg-R19K is associated with clinically severe congenital goiter [requiring surgery in the index patient (16)]. The data herein are consistent with such a conclusion, strongly suggesting that Tg-R19K and not R835Q or V2452L substitutions (Fig. 1) is primarily responsible for the Tg deficiency in affected patients. (Indeed, the R835Q substitution is part of the normal primary structure of bovine Tg, rendering it unlikely to be a pathogenic mutation.) The protein trafficking defect for Tg-R19K is comparable in severity to that observed for the Tg-L2263P mutant, which is responsible for the severe phenotype observed in cog/cog mice (10). We propose that, for the most part, the behavior of mutant Tg molecules in heterologous cells phenocopies their behavior in the thyroid in vivo, allowing experiments to identify and distinguish the potential pathogenic impact of clinically relevant Tg mutations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Site-Directed Mutagenesis of Mouse Tg cDNA
Mouse Tg cDNA was cloned into pcDNA3.1(–)myc-his, appending single myc and 6-His tags to the carboxyl terminus of Tg. Tg mutations were introduced with the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA according to the manufacturer’s instructions). Each mutation was confirmed by direct cDNA sequencing before cDNA expression. The following mutagenic primers were employed: R19K (5'-CCCTGTGAGCTACAAAAAGAGAAAGCCTTTCTG-3' and 5'-CAGAAAGGCTTTCTCTTTTTTTGTAGCTCACAGGG-3'); R835Q (5'-GTGTTGGAGGGAGCCCAAACTCCGCCTGGG-3' and 5'-CCCAGGCGGAGTTTGGGCTCCCTCCAACAC-3'); V2452L (5'-CCAAGCTCCTGGCTCTGAGTGGCCCTTTCC-3' and 5'-CCAAGCTCCTGGCTCTGAGTGGCCCTTTCC-3'); C175Z 5'-GAGTTTATGCCTGTCCAGTGAAAGTTTGTCAATACC-3' and 5'-GGTATTGACAAACTTTCACTGGACAGGCATAAACTC-3'); C1244R (5'-CCAGGGCCACTGATTCGTAGCCTGGAGAGTC-3' and 5'-GACTCTCCAGGCTACGAATCAGTGGCCCTGG-3'); C1976S (5'-GAGTGTGAGCGGCTCAGTGACAGGGACCC-3' and 5'-GGGTCCCTGTCACTGAGCCGCTCACACTC-3'); R19Q (5'-CCCTGTGAGCTACAACAAGAGAAAGCCTTTCTG-3' and 5'-CAGAAAGGC T T T C T C T T T T G T T G T AGC T CACAGGG-3'); K25R 5'-GA GAG AAG GCC TTT CTG AGA CAG GCT GAA TAT GTT CCC-3' and 5'-CT CTC TTC CGG AAA GAC TCT GTC CGA CTT ATA CAA GGG-3'; R13K 5'-G GTA GAT GCA CAG CCA CTC AAG CCC TGT GAG CTA CAA AG-3' and 5'-C GAT CTA CGT GTC GGT GAG TTC GGG ACA CTC GAT GTT TC-3'.
Cell Culture, Transient Transfection, and Cell Lysis
COS-7 cells seeded in six-well plates were grown in complete DMEM/10% FBS to 80% confluency and transfected with 1 µg plasmid DNA in 6 µl Lipofectamine per well. After 5 h incubation at 37 C, the transfection mixture was removed and replaced with fresh DMEM containing 10% FBS. For pulse-chase experiments, metabolic labeling were started at 36 h after transfection. For Western blotting experiments, at 48 h after transfection, the cells were washed and replaced with serum-free DMEM and incubated for additional 24 h. The cells were lysed under nondenaturing conditions in buffer containing 1% Triton X-100, 0.1 M NaCl, 2.5 mM Tris-HCl (pH 6.8), and protease inhibitor cocktail. In some experiments, 1% sodium dodecyl sulfate (SDS) with or without 25 mM dithiothreiotol was added to make a denaturing cell lysis buffer. The media were collected, cleared by centrifugation (13,000 x g for 10 min), and precipitated with tricholoracetic acid (10% final). The tricholoracetic acid precipitates were resuspended in SDS sample buffer and boiled for 5 min before SDS-PAGE.
SDS-PAGE and Western Blotting
Equal fractions of media and cell lysates (
60 µg lysate protein per sample) were subjected to reducing SDS-4%-PAGE and transferred to nitrocellulose for immunoblotting with rabbit polyclonal anti-Tg or anti-myc antibodies, and a goat antirabbit secondary antibody coupled to peroxidase. Blots were developed using enhanced chemiluminescence. For Fig. 6, antibody generated against the amino terminus of Tg was used for Western blotting.
Metabolic Labeling and Immunoprecipitation
Transfected COS-7 cells were starved for 30 min in met/cys-free DMEM, then labeled for 1 h at 37 C with 80 µCi/ml 35S-cysteine/methionine mixture. After pulse-labeling, the cells were washed twice with PBS and chased with DMEM supplemented with an excess of cold cys/met in the presence or absence of 20 µM MG132 for various times.
For BiP-Tg coimmunoprecipitation, nondenaturing buffer contained apyrase 5 U/ml for 30 min on ice, followed by 10 mM iodoacetamide for 1 h further on ice. Nuclei were cleared at 10,000 rpm for 10 min at 4 C and the postnuclear supernates were immunoprecipitated overnight at 4 C with rabbit antimouse Tg, anti-BiP, anti-calnexin, or anti-PDI (each affinity-purified against intact or synthetic antigen conjugated to Sulfo-link agarose beads; Pierce, Rockford, IL). Each immunoprecipitate was recovered by incubation with zysorbin-A. After centrifugation, the immunoprecipitates were washed three times with lysis buffer, boiled in SDS sample buffer, and resolved by SDS-4% PAGE and analyzed by phosphorimaging and quantified using ImageQuant 5.0 software (Molecular Dynamics, Sunnyvale, CA).
Endo H Digestion
Aliquots of cell lysates of transfected COS-7 cells were boiled for 5 min in 0.5% SDS plus 1% 2-mercaptoethanol in 20 mM Tris (pH 7.4). The lysates were then diluted to 0.1% SDS and digested with 250 U endo H in 50 mM sodium citrate (pH 5.5) for 1 h at 37 C. Samples were resolved by reducing SDS-4%-PAGE and analyzed by Western blotting as described above. (For radiolabeled Tg-R19K, cell lysates were digested with endo H and followed by immunoprecipitation.)
|
FOOTNOTES |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online October 4, 2007
1 P.S.K., J.L., and P.J. are co-first authors. 
Abbreviations: endo H, Endoglycosidase H; ER, endoplasmic reticulum; ERAD, ER-associated degradation; FBS, fetal bovine serum; KIF, kifunensine; PDI, protein disulfide isomerase; SDS, sodium dodecyl sulfate; SEAP, secreted alkaline phosphatase; Tg, thyroglobulin.
Received for publication April 12, 2007. Accepted for publication September 28, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. Lee, X. Wang, B. Di Jeso, and P. Arvan The Cholinesterase-like Domain, Essential in Thyroglobulin Trafficking for Thyroid Hormone Synthesis, Is Required for Protein Dimerization J. Biol. Chem., May 8, 2009; 284(19): 12752 - 12761. [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 |
