INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Thyroid hormone synthesis involves a two step modification of tyrosyl residues within the 2750-residue polypeptide chain of bovine Tg¹ (1). First, certain tyrosyls are iodinated to form the hormone precursors MIT and DIT, and then two iodinated tyrosyls couple to form the biologically active iodothyronines T₃ and T₄. In this latter reaction, a tyrosyl donates its iodinated phenyl group to become the outer ring of the iodothyronine residue at an acceptor site, leaving DHA or its derivative at the donor position (2, 3). Both iodination and subsequent coupling are mediated by a thyroperoxidase at the apical membrane of the cell (1).

The most important T₄ forming site in all vertebrate species examined is at Tyr⁵. It appears in a 20-26-kDa N-terminal peptide released by reduction of mature Tg (4-9). Two additional hormonogenic sites occur in the C-terminal region at Tyr²⁵⁵³ and Tyr²⁷⁴⁶ (human numbering) (7-11) as well as at a mid-chain site at Tyr¹²⁹⁰ (7, 8, 12). Priority for hormonogenesis at these sites varies among species (13) and under different physiological conditions, including iodine availability and TSH stimulation (8, 9). Far less is known about the tyrosyls that contribute the outer iodophenyl ring of the iodothyronines. Gavaret et al. (3), using ¹⁴C-labeled Tg from pig thyroid slices, concluded that the residual side chain of the donor existed as DHA while in peptide linkage and was converted to Pyr after its release from Tg by enzymatic digestion, or to acetate after acid hydrolysis. Locating the origins of these donors within Tg has been difficult. Proposed sites, all resting on indirect evidence, have included the Tyr²⁴⁶⁹ or Tyr²⁵²² of bTg, based on the conversion of DHA either to labeled Ala by boro[³H]hydride reduction or to labeled Asp by Na¹⁴CN (14), and Tyr residues 5, 926, 986, 1008, and 1375 of bTg, from 4-aminothiophenol modification of DHA residues (15). Marriq et al. (16) iodinated in vitro a CNBr-derived peptide from low iodine hTg and concluded that Tyr¹³⁰ was converted to DHA with donation of an iodophenyl ring for iodothyronine formation at Tyr⁵. However, Xiao et al. (17), using the same substrate under similar experimental conditions, found that Tyr¹³⁰ served as acceptor rather than donor, suggesting that the CNBr peptide is not a good model for hormonogenesis in intact Tg. Using a different approach based on graded iodination in vitro of low iodine hTg, we identified three potential donors, at Tyr residues 130, 847, and 1487 in hTg (9). These represented sites that were iodinated early but did not form hormone with further iodination. More recently, Gentile et al. (12) suggested from mass spectrometry of a peptide of bTg that Tyr¹³⁷⁵ is an outer ring donor, probably to a neighboring minor hormonogenic site at Tyr¹²⁹⁰.

In the present paper we looked for an outer ring donor in the N-terminal part of Tg, the region of its most important hormonogenic site Tyr⁵, and offer evidence that Tyr¹³⁰ is the dominant donor site.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Preparation of Bovine Tg-- Fresh bovine thyroids were obtained from a local abattoir, transported to the laboratory in iced Earle's medium, and used within 1-3 h of death. We made two Tg preparations: Tg I, from thyroid slices incubated with [¹⁴C]Tyr and subsequently iodinated in vitro; and Tg II, isolated directly from the same beef thyroid without [¹⁴C]Tyr incorporation and without subsequent iodination. For Tg I, 2 g of thyroid slices were incubated in 10 ml of a tyrosine-free medium (Life Technologies, Inc., Selectamine) containing 500 µCi of uniformly labeled [¹⁴C]Tyr (500 µCi/µmol, NEN Life Science Products) at 37 ^oC in a water-saturated atmosphere of air-CO₂ (95:5) for 20 h, then rinsed and briefly homogenized in 0.05 M sodium phosphate buffer, pH 7.4, followed by centrifugation at 105,000 × g for 1 h. Thyroglobulin was isolated from the resultant supernatant fraction at 4 ^oC on a Sephacryl S300 column (2.5 × 100 cm) equilibrated in the same buffer. To obtain Tg II, we processed 5 g of thyroid tissue in the same manner without prior incubation. The protein content of isolated Tg was determined using bovine serum albumin as standard (18) and its iodine content by a modification of the Sandell-Kolthoff method (19).

Iodination of Tg I-- Iodination was carried out by exposure of Tg to iodine at a ratio of 20 atoms of iodine/molecule of 660-kDa Tg. We incubated 3.5 mg (5.3 nmol) of ¹⁴C-labeled Tg in 4 ml of 0.025 M sodium phosphate buffer, pH 7.0, containing: 1.6 mg of glucose (Calbiochem); 15 µg of glucose oxidase, 296 units/mg (Calbiochem); 17.6 µg (106 nmol) of KI; and 162 µg of lactoperoxidase (Sigma). The reaction was started by the addition of glucose oxidase. Carrier-free K¹²⁵I (0.05 µCi) was added to an aliquot (50 µl) of this mixture at the beginning of the incubation. Iodination, with and without added ¹²⁵I, was carried out at 37 ^oC for 1 h and the reaction stopped by the addition of 0.02% sodium azide. The ¹²⁵I-labeled sample was used directly to determine protein bound ¹²⁵I and ¹²⁵I iodoamino acid distribution after Pronase digestion by paper chromatography (4). Excess reagents and unreacted iodine in the non-¹²⁵I-labeled sample were removed by chromatography on a column (1.5 × 100 cm) of Sephacryl S300 in sodium phosphate buffer, pH 7.4, containing 0.02% sodium azide. The peptide composition of ¹⁴C-labeled Tg (Tg I) following iodination and of nonlabeled Tg (Tg II) was analyzed by SDS-PAGE on 4-20% gradient gels under reducing conditions (20). Gels were stained with Coomassie Brilliant Blue and ¹⁴C-labeled gels subjected to autoradiography after treatment with EN³HANCE^TM (NEN Life Science Products). Molecular mass markers were myosin (200 kDa), galactosidase (116 kDa), phosphorylase b (97 kDa), bovine serum albumin (66 kDa), ovalbumin (43 kDa), carbonic anhydrase (31 kDa), soybean trypsin inhibitor (21 kDa), and lysozyme (14 kDa).

Separation of Peptides of Reduced Tg-- Samples of Tg I and Tg II were reduced with 2-mercaptoethanol at a 40-fold molar excess in 8 M urea at pH 8.5 and alkylated with acrylonitrile as described previously (21). The resultant peptides were separated on a Sephacryl S200 column in 0.1 M sodium phosphate buffer, pH 7.4, with 6 M urea and 0.02% sodium azide. Isolation of the 22-kDa N-terminal peptide is described under "Results."

Selective Proteolysis of the N-terminal 22-kDa Peptide of Tg and Isolation of the Products-- The 22-kDa peptide fractions from Tg I and from Tg II were each digested with L-1-tosylamide 2-phenylethyl chloromethyl ketone-treated trypsin (Sigma) at an enzyme to substrate ratio of 1:50 (w/w) in 0.1 M NH₄HCO₃, 0.1 mM CaCl₂ for 1 h at 37 ^oC and the reaction stopped with soybean trypsin inhibitor (Sigma). Tryptic peptides were first fractionated on a column (1.5 × 60 cm) of Bio-Gel P-6 (Bio-Rad), and selected fractions were further separated by reverse-phase HPLC on a C₈ column in a 0.1 M NH₄HCO₃:acetonitrile gradient.

Identification of [¹⁴C]Tyrosine and Its Derivatives-- Samples from Tg I were digested with Pronase, substrate:enzyme ratio 1:10, in 50 µl of 0.1 M NH₄HCO₃ for 16 h at 37 °C, and the products were passed through a Centricon 10 membrane with several rinses of 0.1% trifluoroacetic acid. ¹⁴C-labeled compounds from the digestion were separated by reverse-phase HPLC using a Sephasil C₈ column (Amersham Pharmacia Biotech) and a gradient of 0-90% acetonitrile in 0.1% trifluoroacetic acid at a flow rate of 0.5 ml/min. Standards used were Tyr, MIT, DIT, T₃, T₄, and Pyr or ¹⁴C-labeled Pyr (16 µCi/µmol, NEN Life Science Products). The fraction corresponding to Pyr was then eluted isocratically on an ion exclusion column (Aminex HPX-87H, Bio-Rad) with 0.0025 N H₂SO₄ at a flow rate of 0.6 ml/min. Oxalic, citric, malic, succinic, formic, and acetic acids (Bio-Rad) were used as standards in addition to Pyr and ¹⁴C-labeled Pyr. Detection was by ¹⁴C-label content or A₂₁₀. Pyruvate eluted with a retention time of 7.70 min, between citrate (7.50 min) and malate (9.04 min).

Analysis of Peptides by Edman Degradation and Mass Spectrometry-- Tg I was used to identify [¹⁴C]Pyr-containing peptides. The same fractions isolated in parallel from Tg II were then used for further analysis by mass spectrometry. The molecular masses of selected peptides were determined in a Finnigan-MAT TSQ7000 system with an electron spray ion source interfaced to a 10 cm x 75 µm id POROS 10RC reversed-phase capillary column. Peptides were eluted from the column by 0.1 M acetic acid-acetonitrile gradient at a flow rate of 0.6 µl/min. Sequences of peptides were determined by CAD using electron spray-tandem mass spectrometry with argon as the collision gas and in some cases by automated Edman degradation on an Applied Biosystems 470A Protein Sequencer as described previously (7).

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

We obtained 9 mg of Tg I (1.2 million dpm/mg of protein) from thyroid slices incubated with [¹⁴C]Tyr. Twenty-three percent of ¹⁴C in the soluble fraction of labeled slices was recovered in Tg (Fig. 1A). The yield of Tg II was 130 mg (Fig. 1B).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Separation of Tgs I and II on Sephacryl S300. A, Tg I, thyroglobulin-containing fractions were identified by SDS-PAGE and pooled as indicated. The late peak (tube 110) corresponds to free [14C]Tyr. Inset, autoradiogram of isolated Tg I after iodination. Molecular mass markers are indicated in kDa. B, Tg II; inset, SDS-PAGE of isolated Tg fraction, Coomassie Blue stained.

We iodinated 8.1 mg of Tg I using 20 atoms of iodine/molecule 660-kDa Tg. The insets in Fig. 1 show the pattern on gel of Tg I (Fig. 1A) after iodination and of Tg II (1B). A small aliquot of Tg I, iodinated in the same way but with added carrier-free Na¹²⁵I, showed 91% incorporation of ¹²⁵I into protein, including 14% as T₄ and 8% as T₃. Tg I contained 7.9 ng of iodine/µg of protein before in vitro iodination and 9.2 ng of ¹²⁷I/µg of protein afterward, compared with 9.8 ng of ¹²⁷I/µg of protein in Tg II. Fig. 2 and Table I (top two lines) compare the distribution after Pronase digestion of ¹⁴C in Tg I before iodination with that after iodination. Two peaks are apparent in noniodinated Tg I (Fig. 2A), the major one corresponding to Tyr and a second minor unknown peak labeled Unk in Fig. 2. Digestion of iodinated Tg I (Fig. 2B) released MIT, DIT, and thyroid hormones as well as an additional ¹⁴C-labeled peak eluting early on HPLC with a retention time corresponding to that of Pyr. This peak was absent in noniodinated Tg I. We established the identity of the compound in this peak by showing in a subsequent separation step by ion exclusion chromatography that it co-eluted with authentic Pyr.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Distribution of 14C-labeled compounds following Pronase digestion from Tg I before (A) and after (B) iodination. Separation was by reverse phase-HPLC using a trifluoroacetic acid/acetonitrile gradient.

Tg I and Tg II were individually reduced and alkylated, and then each was separated into four fractions (FI-FIV) by size exclusion chromatography (shown in Fig. 3 for Tg I). Fractions I and II contained primarily peptides of >200 kDa (Fig. 3, inset) and several intermediate bands of 200-40 kDa. Fraction III contained a major band of ~22 kDa and several smaller bands. Our designation of ~22 kDa is based on its migration on gel; it appears the same as the 30 kDa described by Gregg et al. (22), who perhaps calculated that mass from its constituents. We identified the ~22 kDa as an N-terminal peptide normally released from Tg by reduction, on the basis of its size, its high hormone content (Table I), and from sequencing some of its tryptic peptides (see below). Fraction IV contained a major band of ~10 kDa, together with bands of ~22 and 20 kDa. Its high TH content suggested this fraction also contained N-terminal peptides of Tg. Both fractions III and IV had a high Pyr content after Pronase digestion (Table I).

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Separation of reduced and alkylated Tg I on Sephacryl S200. Fractions pooled as indicated. Inset, autoradiogram from SDS-PAGE of 14C-labeled fractions FI-FIV.

Fractions III from both Tg I and Tg II were selected for further study, separately but in parallel. Each was digested with trypsin, and the resultant peptides were separated into three fractions on a size-exclusion column (Fig. 4 for Tg I). Of these, fraction IIIc (from Tg I) had the highest [¹⁴C]Pyr content (2%, compared with 1% in fractions IIIa and IIIb), and on HPLC it resolved into at least 20 peptides (Fig. 5A) and 8 ¹⁴C-labeled peaks (Fig. 5B, A-H). Fractions corresponding to each of the ¹⁴C-labeled peaks were analyzed by one or more methods, including the distribution of ¹⁴C in Pronase-digested material (from Tg I) and sequencing by Edman degradation (from Tg I) or by mass spectrometry (from Tg II) (Table II), and were related to the theoretical tryptic peptides of N-terminal bTg, shown in Fig. 6. The high [¹⁴C]Pyr content of Peak B suggested that it included a donor-containing peptide. It was well separated from a free Pyr standard in this HPLC system. Edman degradation identified several peptides in this peak but none had Tyr in the cDNA-derived sequence of bTg, suggesting that the putative donor peptide may have been blocked at its N terminus and therefore unreadable. Peptide T8 (Fig. 6) was identified in truncated form ending immediately before residue 130 (Tyr) by Edman degradation of material from Tg I and by mass spectrometry of this peptide from Tg II. The measured molecular mass of the latter was 1459.6 ± 0.5 Da, identical to the theoretical mass of residues 118-129 with the N-terminal Gln residue converted to a pyrolline carboxylic acid, an expected consequence of the acidic conditions used during the analysis. The CAD spectrum for this peptide from Tg II confirmed its sequence and the absence of residues after Val¹²⁹.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Separation of tryptic peptides of FIII from Tg I (Fig. 3) on Biogel P6. Fractions were pooled as indicated.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Separation of tryptic peptides of FIIIc from Tg I (Fig. 3) on HPLC using a C8 column and an ammonium bicarbonate:acetonitrile gradient. Flow rate was 1 ml/min with 1-min collections. A, peptide distribution, A214; B, 14C distribution.

View larger version (33K): [in this window] [in a new window]   Fig. 6.   The cDNA-derived sequence of N-terminal bTg (23). The C termini of 22 and 10 kDa are indicated by arrows (21). Theoretical tryptic peptides are indicated as T1-T19.

Samples of fraction IIIc (Fig. 4) from both Tg I and Tg II were also digested with endo-Glu-C, and the resultant peptides were separated on HPLC into 18 or more fractions (Fig. 7A for Tg II) and 5 ¹⁴C-labeled peaks (Fig. 7B, for Tg I). The distribution of radioactivity after Pronase digestion of each peak from Tg I is summarized in Table III. Peak C₂ had a high [¹⁴C]Pyr content. On capillary HPLC, peak C₂ from Tg II separated into five peptide peaks (Fig. 8A), and the molecular masses of peaks 1-4 in Fig. 8A were determined by mass spectrometry. The measured molecular mass of peptide 1 was 1981.6 ± 1.3 daltons, which corresponds closely to the mass of a modified endo-Glu-C peptide (1980.02 daltons) of bTg (Fig. 9A) with the expected loss of Val¹²⁹ from the cleavage described above and the substitution of Pyr or DHA at Tyr¹³⁰ (Fig. 9B). Pyr is more likely than DHA because the mass spectrum shows a fragmented form of the peptide (Fig. 8B) corresponding to a loss of 43 daltons, consistent with removal of the acetyl group of Pyr. A Pyr at residue 130 would explain our inability to sequence the donor peptide in peak B of the tryptic digest, because loss of the amino group at the N-terminal would block Edman degradation. The CAD spectrum for peptide 1 (Fig. 8A) was not readable because of its highly charged ions, expected from its high Arg content (four residues). To confirm the identification of this peptide (peak 1, Fig. 8A), we digested it further with trypsin. The resultant fragment had a measured mass of 1608.3 ± 0.9 daltons, which corresponds to the loss of the C-terminal triplet of the modified endo-Glu-C peptide, an expected product of trypsin digestion (Fig. 9C) with a theoretical mass of 1607.7 daltons. The mass spectrum for this peptide also showed it in fragmented form with a 43 dalton loss. Peak 3 (Fig. 8A) was identified by molecular mass and CAD/MS, as Tg residues 212-225, an expected product of endo-Glu-C digestion. Peaks 2 and 4 did not contain identifiable peptides and peak 5 was not analyzed.

View larger version (22K): [in this window] [in a new window]   Fig. 7.   Separation of endoproteinase Glu-C peptides of FIIIc (Fig. 4) on HPLC using a C8 column and an ammonium bicarbonate:acetonitrile gradient similar to that used in Fig. 5. Flow rate was 1 ml/min with 1-ml collections. A, peptide distribution from Tg II, A214 with indicated retention times. B, 14C distribution, from Tg I.

View larger version (12K): [in this window] [in a new window]   Fig. 8.   Analysis of Pyr-containing fraction (Peak C2, Fig. 7), from Tg II. A, capillary HPLC separation; B, Electron spray-mass spectrometry spectrum of peptide 1 from panel A. The measured molecular mass of the intact peptide is 1981.6 ± 1.3 daltons. A fragment of this peptide is also present, representing a loss of 43 daltons.

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Endo-Glu-C peptide residues 129-146 and its fragments. A, theoretical peptide (22); B, 1981.6 dalton peptide; C, 1608.7 dalton peptide obtained after trypsin digestion, and its lost C-terminal tripeptide.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

We report a new approach to the problem of locating a few iodothyronine outer ring donor sites within the 2750-residue polypeptide chain of Tg. First we isolated peptide fractions containing [¹⁴C]Pyr after in vitro iodination of Tg labeled with [¹⁴C]Tyr. This led to the identification of Pyr-containing peptides from naturally iodinated Tg isolated in parallel to the ¹⁴C-labeled Tg, using mass spectrometry and referring to the cDNA-derived primary structure (23). The Pyr is presumably a derivative of DHA (2, 3). Our results identify the Tyr¹³⁰ of bTg as donor of an outer iodothyronine ring. We base this conclusion on the following: (a) a truncated tryptic peptide ending immediately before residue 130, as identified in Tg I by Edman degradation and in Tg II by CAD/MS, suggested an unstable modification of this residue; (b) an endo-Glu-C peptide from Tg II had the predicted molecular mass of a peptide containing residues 130-146, with Tyr¹³⁰ replaced by Pyr; (c) tryptic digestion shortened this peptide to a size predicted from the loss of a cleaved tripeptide consisting of residues 144-146; and (d) the same HPLC fraction (Fig. 7B, peak C2) from endo-Glu-C digested FIII of Tg I contained [¹⁴C]Pyr. This direct identification of Tyr¹³⁰ as donor agrees with predictions made from indirect evidence by Marriq et al. (16) and by us (9). Support for the physiological importance of this donor site comes from a goitrous hypothyroid family described by Ieiri et al. (24); its members lacked exon 4 of Tg, which codes for a 70-residue segment including Tyr¹³⁰, suggesting that Tyr¹³⁰ was necessary for adequate production of thyroid hormone.

Mass spectrometry and Edman degradation sequencing of tryptic peptide T8 of bTg (Fig. 6) showed the expected sequence only through residue 129, suggesting peptide cleavage between it and residue 130. The mechanism for this cleavage and its relation to iodination are not clear. Reduction of iodinated Tg has produced hormone-rich N-terminal peptides of 10-30 kDa in all species studied. Gregg et al. (22) have localized these cleavages in reduced bTg to precede residues 81 and either 234 or 235. The cleavage we report here following Val¹²⁹ appears different in type and location. Marriq et al. (25) also reported a cleavage between residues 129 and 130 in a CNBr-generated fragment of hTg. Although we do not discount the possibility that the break we describe occurs simultaneously with iodotyrosyl coupling at that site, we think it more likely that the formation of DHA with the loss of the phenyl ring at the time of coupling makes the 129-130 bond susceptible to breakage during subsequent experimental manipulation. We do not have evidence that this cleavage occurs in the thyroid in vivo. An N-terminal peptide comprised of residues 1-129 has not been reported in reduced Tg after various degrees of iodination. Gavaret et al. (2) proposed that instability of the peptide bond associated with the DHA residue could lead to its hydrolytic cleavage and the transformation of DHA to Pyr at the N terminus of the resultant peptide. Our mass spectral data favor the presence of Pyr rather than DHA in the donor peptides, at least by the time it was isolated and studied. Our inability to sequence the tryptic peptide containing residues 130-133, with the putative donor at 130, is consistent with Pyr blocking the Edman degradation. The work of Gentile et al. (12) suggests that cleavage of the peptide bond and the accompanying transformation of DHA to Pyr at the donor site are not universal consequences of the iodotyrosyl coupling reaction because they reported intact DHA at residue 1375 in bTg within a peptide comprising residues 1366-1381, as identified by mass spectrometry.

In this study, we concentrated on the N-terminal region of bTg because it contains the most important hormonogenic site (Tyr⁵) of Tg and can be isolated by chemical reduction after iodination. Most of the N terminus was in fractions III and IV (Fig. 3 and Table I), which together contained similar portions of the total [¹⁴C]Pyr of Tg and of its total labeled hormone. We did not locate the [¹⁴C]Pyr in fraction IV, but suspect at least some was at residue 130 because this fraction contained appreciable amounts of 22 kDa on autoradioautography (Fig. 3, insert); also FIV would be expected to include a residual peptide encompassing residues 81-234 after cleavage of the 22 kDa to produce an ~10 kDa, as shown in Fig. 6 (22).

Our previous studies showed Tyr⁵ and Tyr¹³⁰ to be the only important early iodination sites in the first 240 residues of hTg (9), which is strongly homologous to bTg, although Xiao, et al. (26) did not find early iodination at Tyr¹³⁰ in their in vitro system. Several reports suggest that the N terminus is sufficient by itself to form T₄; for example, goitrous Dutch goats, with a hereditary defect leading to termination of Tg message transcription at residue 296, were still able to synthesize sufficient hormone for euthyroidism (27). In addition, Tyr¹³⁰ and Tyr⁵ are prominently involved in hormone synthesis in several experimental models (16, 17, 28) although the applicability to in vivo conditions is uncertain. We believe these considerations make Tyr⁵ the most likely acceptor for the donated iodotyrosyl from position 130. Further study of hormone formation in the remainder of the Tg molecule is necessary to match donors and acceptors definitively.