Search

A method is described by which tritium-labeled all-trans retinoic acid of high specific activity (up to approximately 51 Ci/mmol corresponding to 85% of theoretical) is converted photolytically within a fraction of a second to a mixture of retinoic acid stereoisomers. One of these isomers, 9-cis-retinoic acid, was obtained in high radiochemical purity by reverse-phase HPLC of the stereoisomer mixture. This fast photolysis was obtained by using a high-pressure 100-W mercury lamp operated at 86 +/- 2 W. A copper sulfate solution was used as a light filter to eliminate short-wave ultraviolet radiation as well as much of the infrared radiation. The geometry of the experimental set-up allowed a maximal amount of the light output of the lamp to reach the retinoic acid solution. Reverse-phase HPLC of the photolytically generated retinoic acid stereoisomer mixture provided pure 9-cis-retinoic acid in 4.5% yield after irradiation for 0.6 s. A steady-state equilibrium of retinoic acid stereoisomers was reached when the irradiation time was extended to a total of 4-6 s (10-11% yield of 9-cis retinoic acid).

Apolipoprotein-E (apoE) has been shown by noncovalent binding and photoaffinity labeling with [125I]T4 to possess a single L-T4 binding site with a K5 of about 3 x 10(7) M-1 and a relative affinity for analogs of L-T4 = D-T4 = rT3 = triiodothyroacetic acid > L-T3. T4 binding was not affected by the flavonoid EMD 21388 or heparin, but was inhibited by diclofenac = mefenamic acid > furosemide. Localization of the T4 site to the N-terminal 62-amino acid region of the mature peptide coded by exon 3 was deduced from the following evidence. 1) The N-terminal 15- to 26-kilodalton (kDa) fragments (within residues 1-160 to 210), but not the approximately 10- to 11-kDa fragments (within residues approximately 220-299), were labeled by [125I]T4. 2) Variants apoE2 and apoE4, with nonconservative mutations at positions 112 and 158 (the latter unable to interact with the apoB/E receptor), maintained the ability to bind T4. 3) Monoclonal antibodies MAb 1D7 and 3H1 (epitopes at positions 139-169 and 243-272, respectively) failed to inhibit T4 binding, but MAb 6H7 (epitope at 1-125) decreased labeling by about 24%. 4) Polymers of apoE were specifically labeled despite the interaction between amphipathic alpha-helices of the exon 4-encoded region (63-299). We conclude that apoE, as previously observed with apoA-I and apoB, possesses a T4-binding domain separate from the lipid-binding domain and distinct from both the heparin- and the cell receptor-binding sites. Thyroid hormone binding by apoE may facilitate uptake of the hormone by cells through apoB/E receptors, which are widely distributed in tissues.

Enzymatic mechanisms for deiodination of 3,5,3'-triiodothyronine or thyroxine in the phenolic ring (5'-deiodinase) or tyrosyl ring (5-deiodinase) are found in cells of many organs, including the intestinal wall. Deiodinases are highly active in intestinal tissue of developing rat fetuses and relatively inactive in adult intestinal cells, but little is known about these systems in the luminal contents of intestines. We have found both 5- and 5'-deiodinase activities in adult rat intestinal contents and have shown that their expression is inhibited by resident intestinal microflora, which are normally present in the adult but not in the fetus, possibly because they are bound by intestinal bacteria in the adult.

Four preparations of high density lipoprotein HDL2, five of HDL3, and purified apolipoproteins apoA-I, apoA-IV, and apoE were photoaffinity labeled with [125I]T4 and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Gels were also immunoblotted with antiserum against apoA-I, apoA-II, apoA-IV, apoE, or apo(a), and the immunostained membrane was then autoradiographed. In HDL2, the two major radioactive bands migrated near the origin of the resolving gel and at 28-31 kilodaltons (kDa). The first band, stained by anti-apo(a) and anti-apoB-100, accounted for 40-96% of the total radioactivity and was attributed to lipoprotein(a), which is isolated in the same density range as HDL2. The second band, stained by anti-apoA-I, accounted for 1-57% [41-95% after correction for contaminating lipoprotein(a)] of the [125I]T4 in the resolving gel. In HDL3, the major radioactive band was identified as apoA-I and contained 93-94% of the [125I]T4 in the resolving gel. Minor radioactive bands in both HDL2 and HDL3 were identified as apoA-II (17-18 kDa), apoA-II monomer (7-10 kDa), apoE (36-38 kDa), and apoAII-apoE heterodimer (46 kDa). In addition, HDL3 contained apoA-IV (43 kDa). Photoaffinity labeling of isolated apoA-IV and apoE showed that each protein interacted with [125I]T4. In both HDL2 and HDL3, photoaffinity labeling in the presence of unlabeled L-T4 (1-10 microM) showed inhibition, suggesting a Kd in the micromolar range. This inhibition varied among different apo bands of the same HDL2 or HDL3 preparation and among the same bands of different preparations. Labeling in the presence of heparin or other inhibitors of T4 binding to plasma proteins (furosemide, diclofenac, and mefenamic acid) showed that HDL2-associated apoA-I was more sensitive to inhibition than HDL3-associated apoA-I. In conclusion, 1) HDL2 and HDL3 carry T4 mainly through apoA-I and secondarily through apoA-II and apoE. The inter- and intrasubclass variations in T4 binding and sensitivity to inhibitors can be explained by the known heterogeneity of HDL particles and possible differences in conformation of the apo. The findings reported here, that apo other than apoA-I and apoB exhibit saturable binding of T4, suggest that thyroid hormone-lipoprotein interactions may have even wider physiological implications than previously appreciated.

Counter-current chromatography (CCC) of the bromoacetylation product of 3,3',5-triiodo-L-thyronine (T3) produced an unusually sharp peak for the desired product, N-bromoacetyl T3 (BrAcT3). A series of experiments revealed that bromoacetic acid, probably present as a side reaction product in the sample solution, was responsible. This compound repressed the ionization of the carboxyl group of BrAcT3 forcing it into the less polar stationary phase until the bromoacetic acid had eluted completely from the apparatus. At this point, the sudden increase of pH and consequent ionization of the BrAcT3 allowed the ammonium salt of the latter to enter the more polar moving phase where it eluted rapidly from the column as a sharp peak. The same phenomenon was observed in the CCC fractionation of a series of indole auxins where addition of trifluoroacetic acid to the sample caused peak sharpening by the same process. The phenomenon recalls pH gradient elution and isoelectric focussing except that the substance responsible for the pH range here is added along with the sample in one bolus forming a sharp pH gradient at its trailing edge. As with gradient elution, the technique is of practical interest since it permits collection of the eluting compounds with increased detectability in fewer fractions. The technique can also enhance separation of compounds whose partition coefficients differ with a change in pH.

A one-step bromoacetylation of L-thyroxine (T4) produces N-bromoacetyl-L-thyroxine (BrAcT4) in good yield. The reaction product is best purified by high-speed countercurrent chromatography. While HPLC is satisfactory only for purification of microgram and submicrogram quantities, amounts ranging from about 1 ng to 1 g of BrAcT4 can be processed by high-speed countercurrent chromatography (HSCCC), a method which we have previously used for the purification of N-bromoacetyl-3,3',5-triiodo-L-thyronine (BrAcT3). Operating conditions for the one-step synthesis of BrAcT4 and BrAcT3 differ due to differences in solubility and reactivity of the two hormones. BrAcT4 purified by HSCCC and shown to be pure by analytical HPLC has been characterized by alpha max and epsilon max in the near and far uv in several solvents, mass spectrum, 1H NMR spectrum, TLC in three solvent systems, retention time in reverse-phase HPLC (C18) in relation to the retention times of two internal standards, 3,3',5-triiodo-L-thyronine and T4, and melting point. Corresponding data for BrAcT3, not previously reported, have also been determined. The described procedure can provide not only substantial amounts of highly purified BrAcT4 for competition studies, but also 125I-labeled BrAcT4 of high specific activity for affinity labeling. Since solutions of BrAcT4 and of BrAcT3 undergo partial decomposition on evaporation to dryness, suitable procedures for the preparation of these hormones in solid form and for storage in solutions have been devised.

Ultraviolet irradiation of the [3H]colchicine-tubulin complex leads to direct photolabeling of tubulin with low but practicable efficiency. The bulk (70% to greater than 90%) of the labeling occurs on beta-tubulin and appears early after irradiation, whereas alpha-tubulin is labeled later. The labeling ratio of beta-tubulin to alpha-tubulin (beta/alpha ratio) is reduced by prolonged incubation, prolonged irradiation, urea, high ionic strength, the use of aged tubulin, dilution of tubulin, or large concentrations of colchicine or podophyllotoxin. Glycerol increases the beta/alpha ratio. Limited data with [3H]podophyllotoxin show that it covalently bound with a similar beta/alpha distribution. Vinblastine, on the other hand, exhibits preferential attachment to alpha-tubulin. The possibilities that colchicine binds at the interface between alpha-tubulin and beta-tubulin, that the drug spans this interface, and that both subunits may contribute to the binding site are suggested.

N-Bromoacetyl-3,3',5-triiodo-L-thyronine and carrier-free [3'-125I]-N-bromoacetyl-3,3',5-triiodo-L-thyronine, to be used for affinity labeling of thyroid hormone receptors, were synthesized using a one-step procedure: a solution of the thyroid hormone 3,3',5-triiodo-L-thyronine and bromoacetyl bromide in ethyl acetate was refluxed for an optimal period of time which depends on the amount of hormone processed. The bromoacetylated hormone thus obtained was then fractionated by high-speed counter-current chromatography which yielded N-bromoacetyl-3,3',5-triiodo-L-thyronine that was pure by the criteria of high-performance liquid chromatography and thin-layer chromatography with different solvent systems. The pure product was well separated from all contaminants including one which in high-performance liquid chromatography was not easily separated from N-bromoacetyl-3,3',5-triiodo-L-thyronine. The latter was characterized by 1H nuclear magnetic resonance, plasma desorption mass spectrometry, thin-layer chromatography, high-performance liquid chromatography, UV spectrophotometry, and melting point. Amounts of 3,3',5-triiodo-L-thyronine ranging from picograms, including carrier-free 125I-labeled triiodothyronine, to 200 to 300 mg can be processed with the equipment used in the present investigation.

We tested the ability of nine monoclonal antibodies (MAb) against human apolipoprotein-A-I (apoA-I), the 28.3-kDa major apoprotein of high density lipoproteins (HDL), to inhibit its photoaffinity labeling with [125I]T4. Two forms were evaluated: isolated lipid-free apoA-I (Sigma or Calbiochem) and lipid-complexed apoA-I [HDL2, (density, 1.063-1.125 g/ml) and HDL3 (density, 1.125-1.210 g/ml)]. After labeling with 0.5 nM [125I]T4 in the presence of MAb or normal mouse IgG, the products were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and subsequent densitometric quantitation of radioactivity associated with the 28.3-kDa band. Group I MAbs, namely those having epitopes in the N-terminal portion of apoA-I, include MAb 16 (epitopes at residues 1-16), 4 and 14 (residues 1-86), and 18 (residues 98-105); group II includes MAbs 7,10, 15, and 17 (epitopes at residues 87-148); group III includes MAb 9 (residues 149-243). All group I MAbs inhibited [125I]T4 binding to isolated apoA-I with this order of potency: MAb 16 (-50% to -61%) greater than MAb 14 (-37% to -41%) greater than MAb 4 (-27% to -33%) greater than MAb 18 (-19% to -27%). In the case of lipid-associated apoA-I, the pattern of hierarchy was variable, presumably related to the known markedly polydisperse nature of HDL, but a constant feature, in contrast to the case of isolated apoA-I, was that MAb 4 was more potent than MAb 14. Group II MAbs gave less than 3% inhibition in both isolated and lipid-complexed apoA-I. Group III MAb 9 either failed to inhibit or gave 18-27% inhibition (one preparation each of HDL2 and HDL3). We conclude that the T4 site of apoA-I is in the N-terminal domain of apoA-I, closer to the epitope for MAb 16 than to that for MAb 18, and that conformational changes occurring when apoA-I is associated with lipids in the HDL particle alter the spatial relationship between some epitopes and the T4 site. Our definition of the T4 site of apoA-I is consistent with another set of data showing that heparin failed to inhibit [125I]T4 binding to isolated apoA-I. Heparin is known to interact with clusters of basic residues, and these residues are concentrated in the midregion of apoA-I.

The binding of thyroid hormones to isolated plasma membranes was studied in NB41A3 neuroblasts. Saturable binding of L-T3, D-T3 and L-T4 was observed. Binding was time-dependent, with equilibrium reached in less than 60 min and maximal binding occurring between pH 7.4 and 7. Saturation experiments demonstrated two classes of sites for L-T3: a high-affinity site with Ka 8.4 X 10(9) M-1 and a low-affinity site with Ka 7.3 X 10(6) M-1.L-T3 and D-T3 inhibited each other's binding, L-T3 being several-times more potent. Affinity labeling of isolated membranes with bromoacetylated thyroid hormones disclosed stereospecific binding to SDS-PAGE bands with approximate molecular masses of 27 kDa (preferentially labeled by BrAc-L-T3), 32 kDa (preferentially labeled by BrAc-D-T3), and 48 and 87 kDa (preferentially labeled by BrAc-L-T4). Binding of BrAc-L-T3 to the 27 kDa band accounted for 3.4% of total binding, was selectively inhibited by excess L-T3, and may be involved in intracellular transport of L-T3.

By photoaffinity labeling human low density lipoproteins (LDL) with [125I]T4 we confirmed our previous observation that of the three T4 binding sites of apolipoprotein B-100 (apoB-100) one is in its 26% NH2-terminal portion [apoB-26, the 140 kDa fragment, residues 1-1297] and two in the remaining 74% COOH portion (apoB-74, 410 kDa, residues 1298-4536). We now show that of these two sites one is in the NH2 portion of apoB-74 (apoB-44, 240 kDa, residues 1298-3249) and the other in the nonoverlapping COOH portion (apoB-30, 170 kDa, residues 3250-4536). ApoB-100 contains 13 binding sites for heparin, a known inhibitor of T4 binding to the major T4 carrier plasma proteins; however, heparin failed to inhibit T4 binding to apoB-100 and fragments thereof. The same failure was seen with three monoclonal antibodies (MAbs), 4G3, 5E11, and 43 that block totally or partially LDL binding to the LDL receptor [respective epitopes at residues 2980-3084 (apoB-44), 3441-3569, and 4027-4081 (apoB-30)]. Of the other 3 MAbs, all without effect on LDL binding to the LDL receptor, [1D1, 2D8, and 16, respective epitopes at residues 474-539 (apoB-26), 1438-1481 (apoB-44), and 4154-4189 (apoB-30)], only two (MAbs 1D1 and 2D8) inhibited T4 binding (21 to 39%). We conclude that the three T4-binding sites of apoB-100 are outside the LDL receptor binding domain, distant from the heparin binding sites and, assuming no allosteric effects, in the vicinity of residues 474-539 (T4 site of apoB-26), 1438-1481 (T4 site of apoB-44), and in the C terminal quarter of apoB-30.

Among the previously reported putative nuclear thyroid hormone receptor forms having molecular masses of 56-59 kDa and 45-49 kDa, respectively, only the former can be the endogenous receptor. The latter must be a degradation product because it is virtually absent in rat liver nuclear extracts prepared in the presence of 20% glycerol and 5 mM Mg2+, which inhibit degradation. In the absence of glycerol, the receptor form of lower mass was present in large amounts in nuclear extracts. Sucrose could not replace glycerol as a protective agent, even in the presence of Mg2+. Thus, the endogenous nuclear thyroid hormone receptor appears to be labile under the experimental conditions used in preparing nuclear extracts. The molecular mass of the nuclear receptor was determined to be 57 kDa on the basis of SDS-polyacrylamide gel electrophoresis after photoaffinity labeling of nuclear proteins with (3,5-125I)-labeled thyroxine.

Photolysis of thyroxine and its analogs in the near UV permitted synthesis in good yield of picogram to gram quantities of thyroid hormone metabolites. Preparation of the same metabolites by classical chemical synthesis requires multistep procedures. Specifically labeled metabolites of high specific activity (e.g., those carrying the label in the nonphenolic ring) were obtained by photolysis of appropriately labeled thyroxine or 3',3',5'-triiodothyronine (reverse triiodothyronine). Some of these labeled metabolites, which are required for metabolic studies (3-iodothyronine and 3,3'-diiodothyronine, labeled in the nonphenolic ring), had not previously been obtained by other methods. Irradiation of thyroxine and reverse triiodothyronine in 150 mM methanolic ammonium hydroxide with greater than 340-nm light caused removal of one iodine atom from the phenolic ring with formation of 3,5,3'-triiodothyronine and 3,3'-diiodothyronine, respectively. Irradiation with higher-energy light (greater than 300 nm) led to stepwise removal of additional iodine atoms. Those in the phenolic ring were removed preferentially, so that 3,5-diiodothyronine and 3-iodothyronine, respectively, were formed. The iodine atoms in the nonphenolic ring were lost more slowly. Tetraiodothyroacetic acid followed a similar photodeiodination pattern. Photolysis with light in the near UV is a simple method for the synthesis of thyroid hormone metabolites.

The solube iodoproteins in a transplantable rat throid tumor (Wollman Line 1-8) were studied after in vivo labeling with 125 I and were partially purified by affinity chromatography on anti-thyroglobulin-AGAROSE. A major fraction ('Peak A') was excluded from gels of large pore size, but had a low sedimentation rate (approximately8S) and did not appear to contain aggregates. It had a high density (approximately1.4) which was possibly due to a high content of carbohydrate, since treatment with a crude glycosidase mixture lowered the density to approximately1.3. A second fraction ('Peak B') had a similar sedimentation coefficient (6-9S) but penetrated the same gels and had a lower density (approximately 1.3). Both proteins formed soluble complexes with antibodies against normal rat thyroglobulin, and had other properties somewhat similar to those of thyroglobulin. After hydrolysis, mono- and diiodotyrosine were the only iodoamino acids liberated. These iodoproteins appears to represent abnormal forms of thyroglobulin synthesized by the tumor.

The transfer of sulfate to thyroid hormones and their analogs by a monkey hepatocarcinoma cell extract was compared to this reaction as catalyzed by two homogeneous aryl sulfotransferases from rat liver. The thyroid hormones 3,5,3'-triiodo-L-thyronine and L-T4 as well as analogs of these hormones, were used as substrates. While these compounds vary widely in their capacity as sulfate acceptors, the catalytic abilities of the three enzyme preparations generally reflect a broad pattern of overlapping specificity. The finding that T4 is a very poor substrate, that 3'-iodothyronine and 3,3'-diiodothyronine are very good substrates, and that T3 is intermediate in its acceptor activity probably explains the relative sulfation of these compounds in vivo and by living hepatocytes. (Endocrinology 108: 454, 1981)

Irradiation of the thyroid hormones thyroxine and 3,5,3'-triiodothyronine in the near UV (greater than 300 nm) causes homolytic fission of C--I bonds in both rings. In the presence of hormone-binding proteins, the phenyl radical thus formed, and possibly also the iodine radical, can establish a covalent bond with certain amino acid residues in the binding site. Most if not all of the iodine radicals, however, appear to be reduced to iodide. Incubation of purified carrier proteins for the thyroid hormones in human serum as well as of an extract of rat liver nuclei or of whole nuclei with trace amounts of 125I- or 14C-labeled hormone, followed by irradiation, resulted in covalent binding. This was proven by gel filtration after boiling with guanidine:HCl and by sodium dodecyl sulfate/polyacrylamide gel electrophoresis of the irradiated solutions or of the excluded-peak material obtained after gel filtration. Sodium dodecyl sulfate/polyacrylamide gel electrophoresis of 125I-labeled irradiated nuclear extracts showed a prominent peak (Mr approximately equal to 45,000), sometimes with a shoulder or small peak at Mr approximately equal to 56,000, and a fast-moving peak (Mr approximately equal to 12,000). Similar patterns were obtained with N-bromoacetylthyroxine or N-bromoacetyltriiodothyronine without irradiation. When a suspension of whole nuclei was irradiated instead of nuclear extracts, the shoulder also became a prominent peak.

Photoaffinity labeling with underivatized thyroxine (T4) was used to identify and compare the T4 binding proteins in rat liver cytosol, nuclear extract, and purified plasma membrane. When these subcellular fractions were incubated with a tracer concentration of [125I]T4, irradiated with light above 300 nm, and individually analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the radioactivity profiles revealed the presence of T4 binding proteins of molecular masses of 70, 52, 43, 37, 30, and 26 kilodaltons (kDa) in cytosol, of 96, 56, 45, and 35 kDa in nuclear extract, and of 70, 44, and 30 kDa in plasma membrane. Competition experiments performed in the presence of a 1000-fold excess of unlabeled T4 demonstrated that these binding proteins display different hormone binding activities. The similar electrophoretic mobilities of some binding proteins present in the different subcellular fractions, i.e., the 70-, 43-45-, and 30-kDa proteins, suggested that these proteins might be identical. However, double-labeling experiments in which plasma membrane, nuclear extract, and cytosol were photolabeled with either [125I] or [131I]T4 and mixed, two at a time, in all possible combinations showed that from one cellular fraction to another, the radioactivity peaks corresponding to the approximately 70-, 43-45-, and 30-kDa proteins were not superimposed. Their relative positions on the gel differed by one or two slices, which indicated differences in molecular mass of 1.9-3.6 kDa. Moreover, enzymatic digestion with Staphylococcus aureus V8 protease of these three proteins, prepared from each subcellular fraction, yielded dissimilar peptide patterns.(ABSTRACT TRUNCATED AT 250 WORDS)

Cultured monkey hepatocarcinoma cells were incubated with [3',5'-125I] diiodo-L--thyronine and with [3,5-125I] diiodo-L-thyronine. In both instances monodeiodination as well as sulfoconjugation took place. [3.-125I] iodothyronine and [3',5'-125I] diiodothyronine were identified as metabolites of [3'-5'-125I]-L-thyroxine in the cells, but neither [3-125I]-iodothyronine nor [3,5-125I] diiodothyronine was detected after incubation of the cells with ]3,5-125I]-L-thyroxine. No products of diphenyl ether splitting were observed in the medium after incubation of the cells with either [3,5-125I] diiodo-L-thyronine or [3,5-125I]-L-thyroxine.

Affinity labeling of human serum prealbumin with N-bromoacetyl-L-thyroxine (BrAcT4) was used to investigate the binding domain for L-thyroxine (T4) on prealbumin. Fluorescence titration with 8-anilinonaphthalene-1-sulfonate revealed a strong and a weak binding site for BrAcT4 (K1 = 1 X 10(8) M-1; K2 = 1 X 10(6) M-1). The reaction of BrAcT4 with prealbumin to form a covalent bond was inhibited in the presence of T4 and binding of T4 to prealbumin was nearly abolished after affinity labeling with BrAcT4. Affinity labeling with 2 mol of BrAcT4/mol of prealbumin resulted in covalent binding of 1 mol of ligand. Acid hydrolysis of affinity-labeled prealbumin gave Nepsilon-carboxymethyllysine and iminodiacetic acid, the latter being derived from the NH2-terminal glycine. A combination of analytical procedures, including tryptic digestion after maleylation, cyanogen bromide cleavage, digestion with yeast protease C, and sequential Edman degradations, revealed that the Nepsilon-carboxymethyllysine was derived from lysine-9 and lysine-15 and that the affinity label had distributed itself among glycine-1, lysine-9, and lysine-15 in a ratio of 29:63:8.

Photoaffinity labeling of rat liver nuclear extract with underivatized thyroid hormones was performed after incubation with 1 nM [3',5'-125I]thyroxine ([125I]T4) or [3'-125I]triiodothyronine [( 125I]T3) by irradiation with light above 300 nm. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the covalently photolabeled nuclear extract revealed four distinct hormone binding proteins of molecular masses 96, 56, 45, and 35 kilodaltons (kDa), respectively. Distribution of the hormone among these proteins was similar for T4 and T3. The 56- and 45-kDa proteins were the most prominently labeled. The specificity of the photoattachment of thyroid hormones to these nuclear proteins was verified by the irradiation of eight randomly chosen proteins and two proteins known to have thyroid hormone binding sites, human thyroxine binding globulin and bovine serum albumin. Only the latter two were photolabeled with [125I]T4. Competition studies performed by incubating nuclear extracts with [125I]T4 or [125I]T3 in the presence of increasing amounts of the corresponding unlabeled hormone (10-, 100-, and 1000-fold molar excess) demonstrated that (1) photoattachment of labeled T3 or T4 to the 56- and 45-kDa proteins was inhibited by 67-78% and 73-85%, respectively, after incubation with a 1000-fold molar excess of unlabeled hormone, (2) in the presence of lower molar excesses of the corresponding competitor (10- and 100-fold), photoattachment of labeled T3 or T4 to the 56- and 45-kDa receptors was gradually inhibited to a similar extent on both proteins, and (3) the 35- and 96-kDa proteins, although having thyroid hormone binding sites, display lower binding activities since the inhibition of photoattachment of labeled T3 or T4 by a 1000-fold molar excess of unlabeled hormone did not exceed 30-42% and 26-49%, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

[U-14C]Tyrosine-labeled noniodinated hog thyroglobulin was iodinated enzymatically and nonenzymatically (iodine, iodide-chloramine-T, pH 7.4, or iodine monochloride, pH 8.1). This led to similar levels of iodine incorporation as well as of thyroid hormone synthesis. Iodine monochloride at pH 5.5 formed "hormonogenic" iodotyrosine residues, but no hormone residues. The latter were formed when the iodinated thyroglobulin was brought to pH 8.5 and then treated with horseradish peroxidase and glucose-glucose oxidase in the absence of iodide and iodine monochloride. Enzymatic hydrolysates contained labeled hormone and pyruvic acid; acid hydrolysates labeled thyronine and acetic acid. (Treatment with acid converts hormone to thyronine and pyruvic to acetic acid.) After borohydride treatment, labeled alanine was present instead of pyruvic or acetic acid. The pyruvic acid/hormone, acetic acid/thyronine, alanine/hormone, and alanine/thyronine molar ratios always were 1, independently of the method of iodination. The "coupling reaction" consists of an oxidation step and nonoxidative coupling and decomposition steps. The oxidation step may be either enzymatic or nonenzymatic. The decomposition step always leads to 1 dehydroalanine residue for each hormone residue synthesized. (Dehydroalanine residues appear in the various hydrolysates as acetic acid, pyruvic acid, and alanine, respectively.) Since proper alignment of 2 iodotyrosine residues is a prerequisite for coupling, a model is proposed according to which oxidation of hormonogenic iodotyrosine residues leads to a charge transfer complex which is the same zwitterion-biradical resonance hybrid no matter whether it resulted from a free radical (enzymatic) or an ionic (nonenzymatic) oxidation.

The customary methods for labeling T4 And its analogs, iodination or exchange-labeling, are not applicable to those iodothyronines whose iodine atoms are exclusively in the nonphenolic ring. Therefore, [3,5-125I]Diiodo-L-thyronine had to be synthesized by a different metnod. This synthesis involves the coupling of [125I]diiodo-L-tyrosine with 4-hydroxy-3,5-diiodophenylpyruvic acid to form L-thyroxine labeled in the nonphenolic ring, followed by removal of the two unlabeled iodine atoms in the phenolic ring. High specific activity, limited only by that of the [125I]diiodo-L-tyrosine used as starting material, can be achieved. Radioactivity yields are approximately 36% in the coupling reaction and approximately 86% in the deiodination reaction, amounting to an overall yield of greater than 30%. L-Thyroxine and 3,5-diiodo-L-thyronine were purified by ion-exchange chromatography. Their radiochemical purity was checked by radiochromatography.

Glutathione and cysteine bind to the heme of lactoperoxidase, thereby causing a red shift of the Soret band which is reversed upon addition of iodide or guaiacol, two substrates for lactoperoxidase. The rate of formation of the enzyme-thiol complex is enhanced by diiodotyrosine. Binding of diiodotyrosine to lactoperoxidase does not cause a shift of the Soret band which indicates binding to the protein of the enzyme. At neutral pH and low ionic strength, lactoperoxidase is adsorbed on insolubilized diiodotyrosine (diiodotyrosine-agarose). It can be eluted at slightly increased ionic strength which shows that the binding is weak. In the presence of 5 X 10(-4) M glutathione, however, the binding of the enzyme to diiodotyrosine-agarose becomes much stronger so that a high salt concentration is required for elution. Lactoperoxidase is also adsorbed on insolubilized thiols (thiol-agarose). The presence of diiodotyrosine is not required for strong binding. A simple method for the preparation of lactoperoxidase from milk by affinity chromatography is based on the interactions of the enzyme with the two ligands, thiols and diiodotyrosine.

The binding of thyroxine (T4) and 8-anilino-1-naphthalenesulfonic acid (ANS) to human serum prealbumin was measured by equilibrium dialysis at pH 7.4 in 0.05 M phosphate-0.10 M NaCl at 25 degrees. The data were analyzed for the binding constants based on equations for (1) two independent sites and (2) two identical sites with negative interaction. Evaluation by the independent site model gave the following association constants: for T4 binding, KT1 = 1.0 x 10-8 M-1, KT2 = 9.5 x 10-5 M-1; for ANS binding, KA1 = 9.5 x 10-5 M-1, KA2 = 2.1 x 10-5 M-1. The interactive model gave constants kT = 5.5 x 10-7 M-1 and kA = 5.5 x 10-5 M-1. Interaction factors, alpha, defined such that -RT in alpha is the energy of interaction, were: alpha T = 0.041 AND ALPHA A = 0.62 for T4 and ANS, respectively. The "best fit" values for the number of sites were 2.0 and 1.6 for T4 and ANS, respectively. The binding of T4 to human prealbumin was competitive with ANS, and the binding constants evaluated from competition experiments were in agreement with those found for each ligand when studied separately. On the basis of analysis of X-ray data of human prealbumin (Blake et al.) there appear to be two identical T4 sites. It is therefore evident that the binding of T4 represents a case of negative cooperativity which is presumably due to interaction between ligands.

We studied binding of T4 to the lipid-complexed apolipoproteins (apo) of high density lipoproteins (HDL), the major lipoprotein carrier of thyroid hormones in human plasma, and to lipid-free apoA-I. HDL isolated from fresh normal plasma by ultracentrifugation (density, 1.063-1.210 g/mL) was photoaffinity labeled with [3,5-(125)I]T4 and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Two bands corresponding to apoA-I (28.3K) and apoC-II or apoC-III (8.6-9.2K) were seen, and their radioactivity decreased by 50-60% when labeled in the presence of 1 mumol/L T4. Photoaffinity labeling of isolated apoA-I also was demonstrated and was decreased 74% by 1 mumol/L T4, suggesting a higher affinity of the lipid-free protein for T4. T4 binding of isolated apoA-I was optimal at pH 7-8, reached a maximum after 1 h at 23 C, and decreased after incubation at 37 C. Scatchard analysis revealed a single T4-binding site with a Ka of 7.5 x 10(7) L/mol at 23 C, pH 8.2. The potency of T4 analogs as inhibitors of T4 binding to isolated apoA-I was L-T4 = D-T4 = triiodothyroacetic acid = L-rT3 much greater than L-T3 much greater than L-thyronine. The binding of T4 to apoA-I was reduced by known inhibitors of T4 binding to serum proteins (diclofenac = mefenamic acid = furosemide = 8-anilinonaphthalene sulfonic acid much greater than dilantin greater than heparin greater than barbital) and by lipids (unsaturated fatty acids greater than cholesterol = cholesterol esters = phospholipids greater than saturated fatty acids = diglycerides = triglycerides). We conclude that the binding of T4 to HDL is mediated by a specific interaction of the hormone with apoA-I and with apoC-II and/or apoC-III. Since the lipid constituents of HDL inhibit T4 binding to apoA-I, the HDL subfraction in plasma that carries most of the HDL-bound T4 should be one with a low lipid content.

A linear random copolymer of tyrosine and lysine and two synthetic oligopeptides containing two tyrosine residues in addition to lysine residues give thyroid hormone (thyroxine and triodothyronine) residues in good yield upon enzymatic iodination with thyroid peroxidase. These synthetic peptides may serve as simple models for thyroglobulin, the protein in which biosynthesis of the thyroid hormone takes place. For the formation of significant amounts of hormone, such model compounds must contain at least two properly spaced tyrosine residues.