An improved procedure for isolation of thyroxine-binding globulin from human pregnancy serum

The International Standard for Thyroxine Binding Globulin

OBJECTIVE

Thyroxine binding globulin is the major thyroid hormone binding and transport protein of the plasma, and its quantitative estimation is therefore of clinical importance. The objectives of the present study were to prepare and ampoule a preparation of thyroxine binding globulin suitable to serve as an international standard, to confirm the suitability of the preparation in a multi-centre collaborative study, and to assign an ampoule content.

DESIGN

The collaborative study was designed with the following aims: (1) to compare the candidate preparation with local standards; (2) to calibrate the preparation with local standards and to confirm the assigned ampoule content; (3) to assess the stability of the ampouled preparation.

MEASUREMENTS AND RESULTS

Thyroxine binding globulin was purified by a combination of affinity and conventional chromatography. Analysis by gel electrophoresis, N-terminal sequencing and electrospray mass spectroscopy showed the preparation to be > 99% pure and to have a structure consistent with the known structure of thyroxine binding globulin. Thyroxine binding globulin assays performed in the multi-centre collaborative study indicated that the preparation behaved as thyroxine binding globulin when compared to local standards and that the ampoule content was consistent with that assigned on the basis of physicochemical measurements of protein concentration.

CONCLUSIONS

The thyroxine binding globulin preparation coded 88/638 was established by the Expert Committee on Biological Standardization of the World Health Organization as the International Standard for Thyroxine Binding Globulin, 30.0 International Units per ampoule, with the additional information that for the preparation in 88/638, 1 IU is equivalent to 1 microgram. The preparation is available from the National Institute for Biological Standards and Control.

Identification and purification of a high-affinity thyroxine binding protein that is distinct from albumin and prealbumin in the blood of a turtle, Trachemys scripta

Fractionation of plasma proteins in the turtle, Trachemys scripta, confirmed the presence of a high-affinity thyroxine (T4) binding protein (TBP) that was distinct from albumin (ALB) and prealbumin (PA). The TBP was isolated by adsorption on a T4-affinity column and a high degree of purification was achieved by gel filtration and preparative electrophoresis. Analysis by reversed-phase HPLC showed a single peak of protein with T4 binding activity. The electrophoretic mobility of the TBP, based on staining and binding to [125I]T4 on nondenaturing polyacrylamide gels (PAGE), corresponded to that of the major T4 binding activity previously identified in plasma (ca. 60 kDa). PA was fractionated as a complex with retinol binding protein (PA-RBP) based on retinol associated fluorescence using ion exchange chromatography on DEAE-Sephacel and gel filtration. This complex behaved as a larger and more highly charged molecule than TBP; it was partially dissociated in low ionic strength basic solution. SDS-PAGE of the PA-RBP-enriched fraction revealed a major component of about 48 kDa (possibly free PA), with smaller components corresponding to those expected for free RBP (ca. 22 kDa) and subunits of PA (e.g., 14 and 28 kDa). ALB was purified by ion exchange chromatography on DEAE and gel filtration; it behaved as less basic than PA with MW approximately 67 kDa. TBP accounted for virtually all the T4 binding activity of whole plasma: TBP was about 100 times as active and PA and ALB were less than 1% as active as plasma. The binding affinity of purified TBP was similar to that of whole plasma from turtle and human (e.g., approx. 10(9) M-1 on Sephadex G-25).

Purification and partial characterization of thyroid hormone binding proteins in canine serum

Thyroxine-binding prealbumin (TBPA) and thyroxine-binding globulin (TBG) were isolated from canine serum and partially characterized. TBPA was isolated by retinol-binding protein (RBP) affinity chromatography and further purified by preparative agarose gel electrophoresis or FPLC ion exchange chromatography. TBG was purified by thyroxine (T4)-Sepharose chromatography followed by gel filtration on Sephacryl S-300 and preparative electrofocusing in a granulated dextran gel. Molecular weights were estimated by SDS-polyacrylamide gradient gel electrophoresis. Canine TBPA had a tetramer molecular weight of 56,000, an extinction coefficient of 12.8 cm2cg.1, an isoelectric point of 5.26-5.70 and a microheterogeneity pattern similar to that of human TBPA. Partial immunochemical identity with human TBPA was also found. Plasma concentrations of TBPA were measured by rocket immunoelectrophoresis in 43 normal and 35 hypothyroid dogs. Reference levels for TBPA ranged between 205 and 474 mg/l. Hypothyroid dogs had a mean TBPA level of 315.0 mg/l (SD: 91.1 mg/l). TBG had a molecular weight of 75,000 and an isoelectric point of 5.0. No immunochemical identity with human TBG was found. Gel filtration of serum on Sephacryl S-200, identification of T4-binding proteins with 125I-T4, and protein- and lipoprotein staining of fractions was performed. Thyroxine-binding was found to TBG in the beta-globulin region, TBPA in the alpha 2-region, albumin, and to the high density lipoprotein (HDL2) in the alpha 1-region and the very low density lipoprotein (VLDL) in the pre-beta region. A corresponding band to the latter protein in serum was masked by TBG and TBPA, and T4-binding in the alpha 1-region was not always seen in serum. Many similarities were found between man and dog regarding TBPA, but not TBG. The differences in structure of TBG may in part be responsible for the low serum T4 levels and rapid T4 metabolism seen in dogs.

Partition of purified human thyroxine-binding globulin in aqueous two-phase systems in response to reactive dyes

Thyroxine-binding globulin was isolated from pooled human serum by a two-step method involving affinity chromatography on thyroxine-Sepharose and preparative disc-electrophoresis. The final product was found to be homogeneous by polyacrylamide gel electrophoresis and has a molecular weight of 59,000. Isoelectric focusing resolved the protein into seven bands with isoelectric points ranging from 3.9 to 4.3. The isolated protein showed affinity to a number of different dyes as recognized by affinity phase partitioning. The interaction of the protein with the dye Cibacron Blue F3G-A was found to be strongly competitive with the natural ligand thyroxine.

Thyroid hormone binding in serum of 15 vertebrate species: isolation of thyroxine-binding globulin and prealbumin analogs

The binding of [125I]T4 to serum proteins was studied in human, monkey, cattle, sheep, goat, water buffalo, horse, swine, dog, cat, rabbit, rat, chicken, frog, and salmon. Attempts were made to isolate thyroxine-binding globulin (TBG) and thyroxine-binding prealbumin (TBPA) from serum of all species, utilizing purification methods based on the specific properties of these proteins. TBPA was found to exist in all species examined. The protein was found anodal to albumin only in human, monkey, horse, and chicken. In cattle, swine, dog, cat, rabbit, frog, and salmon, TBPA was found cathodal to albumin, while sheep, goat, water buffalo, and rat had identical mobility of albumin and TBPA. The presence of TBG was demonstrated in larger mammals. In cat, rabbit, rat, chicken, frog, and salmon, TBG could not be demonstrated. The thyroxine-binding capacity of TBPA in serum varied from 1000 to greater than 6000 nmol/l and that of TBG between 150 and 600 nmol/l. TBPA from all species except salmon showed affinity to human retinol-binding protein. The presence of TBPA in all vertebrates suggests prealbumin to be a far more important thyroxine carrier than earlier anticipated.

Comparison of human thyroxine-binding globulin purification by affinity chromatography procedures

Three procedures for the isolation of thyroxine-binding globulin from human serum, using affinity chromatography on triiodothyronine (T3) linked to Sepharose (A), thyroxine (T4) linked to Sepharose (B) or T3 linked to epoxy-Sepharose (C) as the first purification step, were compared. With the use of additional purification steps, the three procedures yielded pure thyroxine-binding globulin without desialylation. With procedure A, the initial binding of T4-binding globulin to T3-Sepharose was very low, yielding a poor final recovery (17%). Procedure B gave the highest yield (35%) after a three-step purification, with a low T4 content (0.15-0.30 mol/mol). Procedure C also gave a high yield (28%) after only two purification steps, with a T4 content greater than 0.7 mol/mol. The microheterogeneity of T4-binding globulin obtained with these three procedures was demonstrated by isoelectric focusing: five major bands were observed between pH 4.1 and 4.6, and intermediate faint bands (often doublets) in the same pH range. However, with procedures A and C, the most acidic bands (pH 4.10-4.20) were always absent. Thyroxine-binding globulin was preincubated with radioactively labelled T3 or T4 and the hormone-protein complex was analyzed by isoelectric focusing. The binding of T3--compared to that of T4--was reduced in the most acidic protein subspecies. These results suggest differences in the thyroid hormone binding properties of the various subspecies of human T4-binding globulin.

The nude mouse. A possible experimental model for investigation of human thyroid tissue

The effect of transplanting hyperfunctioning human thyroid tissue to athymic, nude mice was explored. The thyroid tissue was obtained from a patient with diffuse, thyrotoxic goiter preoperatively treated with a beta-adrenoreceptor blocker. One transplant was placed in each groin of 4 nude mice. Light microscopy after 4 weeks showed that transplants consisted mainly of typical follicles; capillaries were common at the periphery of the transplant but not in between the follicles. The ultrastructure of the follicle cells in the transplants was similar to that of follicle cells in general. All 8 transplants accumulated 125I as revealed by external counting; the uptake in each transplant 24 h after administration of radioiodine was 4-25% of that in the thyroid of the transplanted mouse. The release of radioiodine, measured during a period of 14 days, from the thyroid of a transplanted mouse was delayed as compared to a nontransplanted control; this indicates that the mouse thyroid activity was suppressed due to hormone production in the transplants. Electron microscopic autoradiography 24 h after injection of 125I showed that protein-bound label in transplants was located mainly in follicle lumens. The radioiodine was incorporated in thyroglobulin as demonstrated by separation of soluble proteins from transplants by sucrose density gradient centrifugation. The present study shows that human thyroid tissue can be transplanted to nude mice with maintained structural and functional properties. Transplantation of human thyroid tissue might be a useful model for studies of thyroid diseases.

A method for the preparation of human thyroxine-binding globulin; its importance in the establishment of an accurate and specific radioimmunoassay

Pure thyroxine-binding globulin (TBG) is required in radioimmunoassay to prepare monospecific antisera, [125I]TBG and as primary standard. Homogeneous TBG was prepared by a three-stage affinity chromatography procedure; it could not be dissociated into subunits and its molecular weight by SDS polyacrylamide gel electrophoresis was 59 000. The amino acid composition was in agreement with two earlier reports. The secondary structure determined by circular dichroism in the far U.V. showed it to contain 24% each of alpha-helix and beta-pleated sheet. Serum TBG was determined by a 24-h radioimmunoassay using polyethyleneglycol to separate bound and free TBG. Serum TBG (mg/l, mean +/- S.D.) was: normal men 15.3 +/- 2.11 (n = 34), normal women 18.4 +/- 2.72 (n = 32) (P less than 0.005), in women on oral oestrogens 24.0 +/- 5.0 (n = 23), in normal pregnancy 38.6 +/- 3.0 (n =37), in cord blood 21.7 +/- 3.5 (n = 25) (P less than 0.001) and in euthyroid subjects aged over 60 years 17.8 +/- 4.5 (P n.s.). In women with thyroid disease TBG was reduced in hyperthyroidism: 15.5 +/- 2.5 (n = 28) and elevated in hypothyroidism: 21.0 +/- 4.0 (n = 25). Wider use in TBG-assay of non-denatured TBG of proven composition and structure should decrease the scatter in reference ranges and improve its value as a routine thyroid function test as both a primary measurement and as the T4 : TBG ratio.