A Comparison of 30-kDa and 10-kDa Hormone-containing Fragments of Bovine Thyroglobulin*

Studies have been carried out on reduced and alky- lated 19 S bovine thyroglobulin to characterize naturally occurring, iodine-rich fragments. In this report, the purification and properties of a 30-kDa, hormone-enriched polypeptide (TgE) are described and com- pared to that of a previously reported 10-kDa fragment (TgF). The amino acid sequence of TgF was found to overlap with that of TgE. In spite of its larger size, TgE contains only a single hormone bearing site. Both the 10- and 30-kDa fragments are derived from the NHz-terminal end of the bovine thyroglobulin. These fragments contain the principal hormone-forming site at residue 5 of the thyroglobulin sequence and appear to be formed by cleavage of the parent polypeptide chain. The mechanism which generates these cleavages is not clear since the sequences surrounding the cleav- age points which give rise to these peptides are quite different. These two fragments may be precursor and product in such a process. The amino acid sequence contained within TgE includes two putative sites for N-linked glycosylation. Since no glucosamine was ob- served and only small amounts of neutral sugar were detected, it appears that this part of the molecule is not extensively glycosylated.

its large size, complex microheterogeneity and posttranslational modification, and its unique, iodine-containing residues. Recent efforts have supplied new information about the chain of events involved in Tg processing and T, formation.
The naturally occurring peptide patterns of reduced and alkylated Tgs, as seen on denaturing polyacrylamide gel electrophoresis, have been documented in several species (Kim et al., 1984;Chernoff and Rawitch, 1981;Rawitch et al. 1984a). In each species, peptides over a broad range of sizes (-330-10 kDa) were observed. A consistent feature of the peptide patterns is the presence of two discrete, hormone-rich small peptides (Dunn et al., 1981(Dunn et al., , 1982. The presence of the small bands has been correlated with hormone production (Dziadik-Turner et al., 1983, 1985Lejeune et al., 1983;Marriq et al., 1984). The peptide of lowest molecular weight has been isolated and characterized from the Tgs of several species including cow (Chernoff and Rawitch, 1981), sheep and pig (Rawitch et al., 1984a), and human (Marriq et al., 1984).
The amino acid sequence surrounding T4 was first determined by isolating a tryptic fragment from the smallest peptide in bovine Tg . The identical sequence has been found in the Tg of sheep and pigs (Rawitch et al., 1984a), humans (Lejeune et al., 1983;, dogs (Gregg, 1985), and rats? The sequence was determined to be NH,-Asn-Ile-Phe-Glu-T4-Gln-Val-Asp-Ala-Gln-Pro-Leu-Arg-Pro-Cys-Glu-Leu-Gln-Arg-COOH. Examinations of the thyroglobulin cDNA sequences have placed the coding for this peptide at the extreme 5' end of the human, bovine, and rat mRNAs. Mercken et al., 1985a;Di Lauro et ale, 1985).
In the present work, we have isolated and characterized a 30-kDa peptide of bovine Tg (TgE) and compared it with the previously studied 10-kDa fragment, TgF (Chernoff and Rawitch, 1981;Rawitch et al., 1984aRawitch et al., , 1984b.

MATERIALS AND METHODS
of Chernoff and Rawitch (1981). Bovine thyroid glands were obtained Preparation of Tg-Tissue was prepared following the procedure from a slaughterhouse and were kept frozen until use. The thyroid tissue was sliced (1-2 mm) and extracted in 0.15 M saline, pH 7.0, for 12 h. The extract was subjected to ammonium sulfate fractionation, and the protein insoluble a t 45% saturation but soluble at 37% saturation was designated crude Tg. After extensive dialysis against 0.1% ammonium bicarbonate, the preparation was lyophilized and stored at -20 "C. Agarose Chromatography-Approximately 1.0 g of crude thyroglobulin was applied to a 5.0 X 100-cm column of Bio-Gel A-5m (Bio-Rad). The protein was eluted from the column using 0.1% ammonium bicarbonate. Fractions corresponding to 19 S thyroglobulin were lyophilized and stored at -20 "C.
Reduction and Alkylation-Reduction and alkylation was performed as described earlier (Chernoff and Rawitch, 1981). Pure 19 S * R. Di Lauro, personal communication.
Tg was dissolved in buffer consisting of 8 M urea, 0.4 M Tris-HC1, and 0.2% EDTA sodium salt. The sample was reduced with Pmercaptoethanol for 4 h at room temperature. S-Carboxymethylation was accomplished by adding iodoacetic acid and incubating for 20 min in the dark. The sample was either directly applied to a CL-4B column or was extensively dialyzed against 0.1% ammonium bicarbonate and lyophilized.
Urea CL-4B Chromatography-Reduced and alkylated 19 S thyroglobulin (200 mg) was applied to a 5 X 100-cm column of Sepharose CL-4B (Bio-Rad) equilibrated with 0.1 M sodium phosphate, 6 M urea, pH 7.0. The sample was eluted with the same buffer, and 12-ml fractions were collected. The column profile was monitored at 280 nm and was divided into six pools designated TgA, TgB, TgC, TgE, and TgF. These pools were extensively dialyzed against 0.1% ammonium bicarbonate, lyophilized, and stored at -20 "C.
Sephacryl S-200 Chromatography-A 1.5 X 150-cm column of Sephacryl S-200 (Pharmacia LKB Biotechnology Inc.) was used to further purify pools of bovine TgE and TgF. The column was equilibrated with 0.1% ammonium bicarbonate. Approximately 40 mg of either crude TgE or TgF was dissolved in the same buffer, applied to the column, and eluted at 4 'C. Fractions (2-ml) were collected, monitored for absorbance at 280 nm, pooled, and lyophilized. Samples were stored at -20 "C.
Tryptic Digestion-Protein samples (5 mg) were dissolved in 1 ml of 1.0% ammonium bicarbonate. Tosylphenylalanine chloromethyl ketone-treated trypsin (Sigma and Worthington) corresponding to 5% by weight of the protein sample was dissolved in 0.001 N HCl. One-half of the trypsin was added, and the digest was held at 37 "C. After 2.5 h, the second half of the enzyme was added. The digestion was stopped at 5 h by freezing and lyophilizing the sample.
Peptide Mapping by Reverse-phase HPLC-Peptide maps using reverse-phase HPLC were prepared from tryptic digests of reduced and alkylated 19 S thyroglobulin, TgE and TgF. A Beckman HPLC chromatograph Model 342 with a Beckman 165 detector and Vydac C-18 column (4.6 mm X 25 cm, 300-A pore size) was used. A linear gradient from 0 to 35% acetonitrile in 70 min with a flow rate of 1 ml/min was used to separate the tryptic fragments. The starting buffer consisted of 95 parts of 0.1% ammonium bicarbonate and 5 parts of acetonitrile (v:v). HPLC profiles were monitored at 230,325, and 350 nm to detect peptide bonds, iodinated peptides, and thyroxine-containing peptides, respectively.
Amino Acid Analysis-Protein samples were hydrolyzed in constant boiling HCl(5.7 N) for 22 h at 110 "C under vacuum. The amino acid analyses were carried out on a Beckman Model 121 MB automated amino acid analyzer.
Peptide Sequencing-Automated Edman degradation was performed using a Beckman 890C protein-peptide sequenator with the 0.1 M Quadrol program (Beckman 121078). The butyl chloride fraction were converted to PTH-derivates by treatment with HCI for 10 min a t 80 "C. The fractions were then extracted twice with ethyl acetate and dried. PTH-derivatives were identified by HPLC using a reverse-phase C-18 column (Altex) and a methanol-sodium acetate gradient (Downing and Mann, 1976). PTH-derivatives were backhydrolyzed in HCl and the free amino acids identified by automated analysis to confirm residue assignments.
Carboxypeptidose Y Digestion-Purified protein (I mg) was dissolved in sodium acetate buffer (0.05 M, pH 6.3), and norleucine (9 pg/time point) was added as an internal standard. A substrate-toenzyme ratio of 401 was used for digestion. The reaction was initiated by the addition of carboxypeptidase (Sigma); the reaction mixture was incubated at 25 "C. A t each time point, 40 p1 of the digest was removed and added to 60 r l of glacial acetic acid to quench the reaction. Each sample was dried under a stream of nitrogen and was subjected to amino acid analysis.
Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis with 6 M Urea-Urea-sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis was performed in a vertical gel apparatus (Protean@ or Mini-Protean@, Bio-Rad). The separating gel was 10% acrylamide (acrylamide:bisacrylamide, 300.8), 0.23 M sodium phosphate, pH 7.2, 6 M urea, and 0.1% SDS. The stacking gel was 4% acrylamide (same stock solution as above), 0.16 M sodium phosphate, pH 7.2, 6 M urea, and 0.1% SDS. Protein samples were dissolved in a buffer consisting of 10 mM sodium phosphate, pH 7.2, 7 M urea, and 1% SDS. The samples were incubated at 90 "C for 10 min before loading on the gel. The gel was run at constant voltage until the bromphenol blue tracking dye had migrated to approximately 1 cm from the bottom of the gel. The electrode vessel buffer consisted of 0.1 M sodium &os-methanol and 10% acetic acid. The slabs were stained overnight in a solution of 0.1% Coomassie Brilliant Blue (Bio-Rad), 0.1% cupric sulfate, 50% methanol, and 10% acetic acid. The fixing solution was also used to destain the slabs.

RESULTS
Crude Tg was obtained from bovine thyroids by saline extraction and ammonium sulfate precipitation. Chromatography on a 4% agarose column yielded pure 19 S Tg. After reduction and alkylation of the 19 S Tg, the resulting peptides were fractionated on a CL-4B column equilibrated with 6 M urea, 0.1 M sodium phosphate, pH 7.0 ( Fig. 1). The pools were designated TgA, TgB, TgC, TgD, TgE, and TgF in order of decreasing molecular weight. Each pool was dialyzed against 0.1% ammonium bicarbonate, lyophilized, and stored frozen. Typical CL-4B profiles showed peaks of approximately equal size for TgE and TgF. Yields in dry weight were also similar for TgE and TgF, at about 3-4 mg of each for every 200-mg sample of crude Tg processed. Since the mass of TgE (-30 kDa) is three times that of TgF, on a molar basis more TgF was recovered than TgE.
Samples of TgE and TgF were passed over a Sephacryl S-200 column for further purification. A profile of the TgE pool from the CL-4B column run on an S-200 column is shown in Fig. 2. The peaks were pooled as shown in the figure and lyophilized. Pool 1 contained higher molecular weight material and was not further processed. Pool 2 contained TgE. Fragment TgF, when present, eluted in fractions following TgE. Before use in peptide mapping, TgE and TgF were verified as homogenous using 10% polyacrylamide gel electrophoresis in the presence of 0.1% SDS and 6 M urea (Fig. 3). The S-200 column chromatography successfully separated fragments TgE and TgF from each other. Special care was taken to be sure all detectable TgE was absent in samples designated "pure TgF," and vice versa.
The amino acid sequence of nascent bovine thyroglobulin has been determined by sequencing its cDNA (Mercken et al., 1985a). It can be concluded from several pieces of data that the NH,-terminal region (residues 1-80) predicted from the cDNA corresponds to fragment TgF. Previous characterization of TgF showed it to be a 10-kDa fragment enriched in Samples of reduced and alkylated 19 S thyroglobulin (200 mg in 25 ml) were applied to a Sepharose CL-4B column (100 X 5 cm) equilibrated with 0.1 M sodium phosphate, 6 M urea, pH 7.0. The sample was eluted with the same buffer, and 12-ml fractions were collected, pooled as shown, and designated TgA, TgB, TgC, TgD, TgE, and TgF in order phate, 0.25% SDS, pH 7.2. Gels were fixed in a solution o i 50% of elution. both iodine and thyroid hormone . TgF contains a single T,-containing sequence near its NH, terminus. The amino acid composition of TgF corresponds very well to that of the first 80 residues predicted by the bovine cDNA for Tg (Table I). Moreover, the NH,-terminal sequences established experimentally for both intact TgE and TgF were identical to that predicted for the NH,-terminal 19 residues of Tg. A comparison of the amino acid composition of TgE and the composition of the NH,-terminal234 residues from the predicted sequence showed good agreement ( Table  I). The composition of TgE further suggested that the complete TgF sequence could be contained within TgE. The relationship of these two hormone-containing fragments was studied further using the purified bovine TgE and TgF.
Peptide mapping was performed to determine whether isolated proteins TgE and TgF had sequences in common. The proteins were digested with trypsin and the resulting fragments mapped on reverse-phase HPLC. When the column eluate from the digest of TgE was monitored at 230 nm, maps containing approximately 20 distinct peaks were obtained (Fig. 4a). The same experiment was performed on TgF (Fig.  4b). Careful study of the profiles showed that all of the peaks found on the map of TgF can also be located on the map of TgE. The individual peaks from the HPLC runs were collected, dried, hydrolyzed, and analyzed for amino acids. The t t

43K I8K
TABLE I

Amino acid compositions of TgF and TgE
Amino acid compositions were from purified fragments TgE and TgF. These were then compared to the composition predicted from the cDNA sequence and known COOH termini. *, Trp. was not assaved in these analvses.  amino acid composition data enabled us to identify the elution positions of most of the tryptic peptides predicted from the bovine gene (see Table 11). Peptides smaller than 6 residues were difficult to detect by their absorbance at 230 nm unless they contained an aromatic residue. All of the predicted peptides in TgF were identified except dipeptide T-2 and tetrapeptide T-3. The amino acid analysis of individual peaks confirmed that all of the tryptic peptides in TgF are also found in TgE. Peak 11 contained predicted peptide T-7 when collected from fragment TgE; when collected from fragment TgF, residue 81, lysine, was absent. This confirms earlier work showing the COOH terminus of TgF to be glutamine, residue 81 in the sequence. Interestingly, the presence of the lysine residue does not significantly alter the retention time of this peptide.
The reverse-phase system was capable of resolving peptides differing by only an arginine residue. Peptide T-4b was the same sequence as T-4a but with an additional arginine residue at its NH, terminus. This single sequence change caused a 2min difference in elution position.
The HPLC maps of TgE and TgF were also monitored at 325 nm to detect peptides containing diiodotymsine or TI (Fig. 5). The 325-nm maps showed a prominent peak eluting at 50 min and several smaller peaks eluting from 44 to 49 min. The dominant peak in the TgE profile, retention time approximately 51 min, contained the 19-residue, hormonogenic fragment previously isolated from TgF. As in the 230nm profiles (Fig. 4), all of the peaks seen in TgF were observed in TgE and confirmed by amino acid composition. When HPLC maps of TgE and TgF were monitored at 350 nm to detect T,, the hormonogenic fragment was the only peak showing significant absorbance. Intact peptide TgE was subjected to 30 rounds of Edman degradation and the PTH-derivatives identified by HPLC.
The sequence determined corresponded exactly to that predicted by the cDNA as the amino terminus of intact, mature thyroglobulin. Peptide TgE was also subjected to carboxypep- tidase Y digestion to determine the sequence of its COOH terminus. During digestion the leucine level rose very rapidly, followed by a rise in glutamic acid. The high level of leucine suggests that 2 or more leucine residues are present close to the COOH terminus. The predicted amino acid sequence shows 3 leucine residues at positions 233, 234, and 235 (Fig.  6). A fourth leucine is at position 231, separated from the 3 tandem residues by glutamic acid at position 232. We have thus placed the carboxyl terminus of TgE at residue 234 or 235. The amino acid composition and molecular weight constructed for TgE from this predicted cleavage point agree very well with the experimentally determined values (Table I).

DISCUSSION
The results presented here characterize bovine TgE, a hormone-containing 30-kDa polypeptide from thyroglobulin. TgE was purified from bovine Tg by sequential gel filtration steps after reduction and alkylation. Its mass was determined by SDS-polyacrylamide gel electrophoresis to be -30 kDa. The amino acid composition of TgE was found to be very similar to that found in TgF. No amino sugar was detected on standard amino acid analysis, but some neutral sugar was found (-5%). Like TgF, TgE was enriched in both iodine and thyroxine. The HPLC tyrptic peptide maps of peptides TgE and TgF were compared (Figs. 4 and 5). It was apparent that all of the peaks on the map of TgF were also contained on the maps of TgE. Individual peaks were collected and amino acid analysis performed. Peaks from TgE and TgF with the same elution time were compared to each other and to the composition of tryptic fragments predicted from cDNA sequence data (Mercken et al., 1985a). Each of the correspond-Elution Timeminutes ing peaks from TgE and TgF had similar or identical compositions. The C-18 HPLC column was able to separate tryptic peptides T-4a and T-4b, sequences differing by only an Arg residue at the NH2 terminus. Tryptic peptide T-17 eluted as HPLC peak 18 with a retention of approximately 49 min. Peptide T-17 contains 12 amino acids, four of which are phenylalanine residues. The hydrophobic nature of this peptide probably contributes to its long retention time. This is consistent with the observation that iodinated tyrosines and thyronines elute from reverse-phase columns with acetonitrile in order of increasing iodination, i.e., increasing hydrophobicity?
When HPLC peaks 16-19 were collected for amino acid analysis, their recovery was consistently low in relation to the magnitude of the detected peaks. Although HPLC peak 18 contained no iodinated residues, apparent absorbance at 325 nm was observed. HPLC peaks 17 and 18 were identified as tryptic fragments derived from TgE, but they were also detected, in reduced amounts, in TgF samples. We conclude that there were traces of TgE in our TgF samples. This led to those peptides with strong absorbance signals from the contaminating TgE being detected in the TgF peptide maps.
Several pieces of data were used to place the peptides TgE and TgF at the NHz terminus of thyroglobulin. The tryptic peptides mapped on HPLC gave excellent agreement with the predicted composition. Using the COOH-terminal determined in this study the overall amino acid compositions of TgF and TgE are very similar to those predicted from the gene. The COOH-terminal sequences determined by carboxypeptidase digestion were found to be Glu-Leu-Leu for TgE and Gln-C. Dziadik-Turner and A. Rawitch, unpublished data. Six homologous repeats centered around this sequence have been predicted in the amino half of bovine Tg (Merken et  al., 1985b). A consistently appearing, sharp peak that precedes a tall peak 4. Very difficult to collect free of contaminating peak 4. Same as HPLC peak 5 but without an Arg residue. Differs from HPLC peak 4 by one additional amino acid Could not be clearly identified as a cDNA predicted pep-Contains 17 amino acids; no Tyr.
Contains Cys-Trp-Cys-Val-Asp-Ala sequence also found in Composition almost identical to peak 10. Sequence contains 2 Tyr residues. When isolated as the C-terminal peptide in TgF, residue 81 (Lys) is absent. The presence of Lys does not alter the retention time. (Arg).
tide. Not present in profiles of TgF. T-5.
Composition very similar to T-7, peak 11. The amino acids contained in peptide T-20 coelute with other amino acids. T-20 is the COOH-terminal peptide of TgE, as determined by carboxypeptidase Y digestion. Wide peak probably due to the close elution of three different peptides, each differing by 1 Arg residue (i.e. alternate trypsin cleavage sites). Not a clear match for any predicted tryptic peptide. Not a clear match for any predicted tryptic peptide. Same composition as HPLC peak 14, but with low tyrosine.
Long retention time probably due to 4 Phe residues. Hormone-containing peak; sequence conserved in Tg from T-8 with at least one Tyr iodinated. several sDecies.

~~ ~~~
Tryptic peptides are numbered from the NH, terminus of thyroglobulin. See Fig. 6. the absence of aromatic residues, peptides of less than six amino acids could not be detected using this method.
Leu-Gln for TgF. These sequences are located at residues 77-80 for TgF and 232-234 for TgE. It should be noted that the COOH-terminal sequences of TgF and TgE, as well as the additional neighboring amino acids near the cleavage sites which result in TgE and TgF, are quite different from each other (see Fig. 6). The sequences do not correspond to tryptic cleavages or to the cleavage pattern(s) of any of the common processing enzymes. Our data indicate that the TgF peptide is contained in the TgE peptide but cannot answer the question of whether TgF is formed through proteolytic processing of TgE or in a more direct process from the parent Tg chain. The complete sequencing of the bovine cDNA has revealed that the hormone-containing sequence found at the NH2 terminus is not repeated (Mercken et al., 1985a). Thus, TgE and TgF are derived uniquely from the amino terminus of the thyroglobulin molecule, overlap within that sequence, and contain the primary T4 synthesis site.
Whether TgE and TgF are formed from alternate cleavages of the initial Tg gene product or whether TgE is the precursor of TgF is not completely clear. Peptides TgE and TgF together account for a large fraction of the T4 produced in bovine Tg.
The very different COOH-terminal sequences of TgE and TgF imply that the same enzyme or chemical mechanism is probably not responsible for both cleavages. It has been suggested that clipping of the Tg protein may take place during iodination (Deme et al., 1975;Dunn et al., 1983). In addition, a CNBr fragment of Tg found only when T4 is produced decreases in molecular weight after incubation with thyroid peroxidase (Dziadik-Turner et al., 1983). In neither case has a specific molecular mechanism for breaking peptides bonds been proposed. If TgE and TgF are precursor and product then one should be able to follow the incorporation of radiolabeled iodine during a pulse-chase experiment. Although this experiment has not been done using bovine Tg, it has been carried out with rabbit and human thyroglobulins (Dunn et aL, 1981;Marriq et al., 1984). In both of these species, two low molecular weight, iodine-rich Tg fragments have been characterized. In both cases it was concluded that the iodine found initially in the larger of the two fragments was detected later in the smaller fragment.
Peptide TgE contains 7 tyrosine residues, in addition to the residue in tryptic fragment T-4 (residue 28). These occur at positions 89, 97, 107, 130, 192, 215, and 217. HPLC maps made at 325 nm of tryptic digests of TgE show several peaks eluting from about 45 to 50 min, just prior to the hormonecontaining peptide peak. When HPLC peak 17 was analyzed, its composition closely matched that predicted for T-8, except for the tyrosine value which was low. Peptide T-8 has 3 tyrosine residues; the composition of HPLC peak 17 showed only 2 tyrosine residues. The amino acid data, 325-nm absorbance, and long column retention time are all consistent with peptide T-8 containing at least one site for diiodotyrosine formation. As tyrosine residues are iodinated the peptide's retention time would be expected to increase and its tyrosine content on amino acid analysis decrease. It is unlikely that monoiodotyrosine alone is formed in this peptide since the extinction coefficient of monoiodotyrosine at 325 nm is too low to account for the observed absorbance (Dziadik-Turner plea and conditions were the same as in Fig. 4, except that the effluent was monitored at 325 nm to detect peptides containing iodinated tyrosine derivates. Elution Timeminutes . HPLC peak 16 also had a composition similar to T-8, but it was not as good a match as peak 17.

Panel a,
These observations illustrate the necessity of obtaining protein data along with DNA data in locating sites of iodination and/or modification in thyroglobulin. Although the cDNA data can yield a predicted amino acid sequence, they give no information relating to the amount or location of iodinated residues. Thus, both kinds of information are required to establish a clear picture of thyroglobulin's mature structure and the nature of its posttranslational changes.
Three hormone-containing sequences in addition to the 19residue fragment discussed here have been reported in porcine thyroglobulin (Marriq et al., 1983). All three are located near the carboxy terminus of the intact Tg structure. However, hormone formation in the carboxy sites appears to contribute relatively little to the total T, production in normal animals in the absence of thyrotropin stimulation. This is consistent with the fact that only the amino-terminal T,-containing peptide can be detected in goiter Tg at low iodine levels (Dziadik-Turner and Rawitch 1984). The other sites are most probably iodinated and coupled to some degree at higher iodine levels, after the primary site is filled. The location of the principal and secondary hormone-producing sites near the ends of the protein has implications for both hormone formation and proteolytic processing. The tyrosine used to form hormone must be accessible to thyroid peroxidase for both iodination and coupling. Furthermore, thyroglobulin containing T, can react with antibody specific for a T,-containing epitope (Byfield et al., 1982(Byfield et al., , 1984. Clearly, hormone contained within the Tg sequence is exposed on the surface of the protein. Thus, a location near the chain terminus may be highly preferred as the coupling acceptor site. Thyroid hormone must be proteolytically cleaved from the Tg structure to become available as free, circulating triiodothyronine or T4. Recent studies have shown that in hog thyroid lysosome extracts there are thiol proteases that can release T, from T g (Yoshinari and . The rate of T, release was found to increase as the size of the thyroxine-containing polypeptide substrate decreased (Nakagawa and Ohtaki, 1985). In light of this, the formation of T4 near a sequence terminus could be seen as facilitating its release. Thus, only one peptide bond near either terminus of the molecule need be broken to produce a small fragment.
The very similar amino acid compositions of fragments TgE and TgF invite examinations of their amino acid sequence for internal homology. A study of the cDNA sequence has outlined a 60-residue sequence that is repeated seven times within Tg (Mercken et aL, 1985b). Data presented here indicates that three of these homologous duplications occur within the sequences of TgE.
Both TgE and TgF are naturally-occurring fragments of Tg, derived from the 330-kDa gene product via posttranslational processing. The in vivo production of these fragments may be a required step in the chain of events releasing free T4 from the protein backbone. In this way hormone release can occur without requiring the complete degradation of thyroglobulin. The hormone contained in bovine TgE and TgF represents the major portion of the hormone synthesized in bovine thyroglobulin. While the specific cleavage mechanism which results in these polypeptide fragments in not yet known, further studies of thyroid protease specificity and in vivo iodination of thyroglobulin should help to clarify this question. Both the 10-and 30-kDa hormone-rich polypeptides seen in reduced and alkylated thyroglobulins are derived from the amino-terminal region of thyroglobulin. Both contain a common hormone bearing amino acid sequence. It is likely, based on in vitro iodination experiments and on radiolabeling experiments, that the 10-kDa fragment is derived from the 30-kDa fragment. However, alternate cleavages from the parent molecule generating TgE in one case and TgF in another cannot be completely ruled out at this time.