Endocrinol. Japon. 1983, 30 (6), 737-745

Radioimmunoassay of Human Transcortin

KAORI SUEDA, KATSUHITO OGAWA AND Nosuo MATSUI

The Research Institute of Environmental Medicine, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464

Abstract

A specific, sensitive and simple radioimmunoassay (RIA) system for human transcortin was developed. A highly purified transcortin was prepared from pooled human serum by the following four successive steps ammonium sulfate fractionation, -Sepharose column chromatography, Ultragel AcAA column chromatography and hydroxylapatite column chromatography. Anti-transcortin was raised by immunizing rabbits The RIA employing 125i-labeled transcortin preparation and polyethylene glycol solution for separation of free and bound-form was sensitive to transcortin in concentrations as low as 10 ng/ml. This RIA was reliable in the tests of dilution, reproducibility and recovery. The presence of cortisol does not interfere with the assay. Also a test of cross reactivity revealed that the system was not influenced by human serum in a concentration 10,000 times that of trans- cortin. The transcortin concentrations determined by the RIA (Y) and those by the conventional steroid-binding assay (X) revealed a good correlation (Y=1.01 X + 6.25) in normal serum, and the immunoreactivity and steroid binding activity revealed a good correlation in heat and acid-inactivated transcortin. With some total cortisol concentrations given, the transcortin concentrations were inversely correlated with -unbound cortisol concentrations. The present assay is useful not only for biochemical research but also for clinical studies, in which the determination of transcortin makes it possible to evaluate the concentration of protein-unbound cortisol which is the physiologically active fraction in serum.

Transcortin is a binding protein specific 1966 ; Westphal, 1971). The determination for in plasma (Daughaday, of this fraction is thus important in the 1958 ; Slaunwhite and Sandberg, 1959). In evaluation of adrenocortical activity, but, it normal human plasma, nearly 80 % of the has very seldom been performed either in cortisol is present in a transcortin-bound clinical or experimental studies, probably fraction and of the remaining 20 %, one because of its difficulty. third exists in a protein-unbound from, and The concentration of unbound-cortisol in the other two-thirds are in an albumin- plasma is mainly influenced by two factors, bound fraction. It is widely accepted that one of which is the total cortisol concen- among the three fractions of plasma cortisol tration and the other is the concentration of the sole physiologically active fraction is the transcortin. Although the protein-unbound protein-unbound one (Matsui and Plager, cortisol concentration increases along with the total cortisol concentration, they do not Received July 12, 1983 always change in parallel. For example, Endocrinol. Japon. 738 SUEDA et al. December 1983

during pregnancy when plasma transcortin Materials and Methods increases, the plasma protein-unbound corti- sol concentration remains relatively low in Reagent spite of a significantly high level of Sephadex G-25, Sephadex G-100 and Dextran T 70 were obtained from Pharmacia Fine Chemicals, total cortisol (O'conell and welsh, 1969 ; Uppsala, Sweden. Ultragel AcAA was from LKB, Matsui et al., 1979). In contrast, in certain Bromma, Sweden ; hydroxylapatite Biogel HTA from disease states, protein-unbound cortisol con- Bio-Rad Laboratories, Richmond, U.S.A. ; [1, 2, 6, centrations are elevated even when total 7]-3H-cortisol (3H-cortisol) (91 Ci/rnmole) and radio- active iodine (125I) (170 Ci/mg) from New England cortisol concentrations are within the normal Nuclear Corp., Boston, U.S.A. ; and 125I-cortisol-21- range, and this difference between the total tyrosine-methylester (125I-cortisol) (250 Ci/mmole)from cotisol and protein-unbound cortisol con- Daiichi Radioisotope, Ltd., Tokyo, Japan. Cortisol centrations has been attributed to decreased was from Sigma Chemical Company, St. Louis, U.S.A. ; complete Freund's adjuvant from Difco transcortin concentrations (McCann and Laboratories, Michigan, U.S.A. ; and charcoal Norit Fulton, 1975 ; Yamamoto et al., 1983). extra, polyethylene glycol having a mean molecular Therefore, determination of transcortin con- weight of 6000 and all the other reagents were centrations as well as total cortisol concen- purchased from Wako Pure Chemical Industries Ltd., Osaka, Japan and were of analytical grade. trations may be useful in estimating the Cortisol-Sepharose (21-deoxy-21- (w-aminohexyl) ami- protein-unbound cortisol concentrations. nocortisol agarose) was prepared according to the Until recently, the plasma transcortin method described by Chan et al. (1977). concentrations were estimated by determin- Subjects and blood sampling ing the cortisol-binding capacity. The Serum samples were obtained between 9: 00 and methods included equilibrium dialysis 11 : 00 in the morning from normal adult men (Daughaday et al., 1962 ; Murphy and Pat- (number : 73) and women (60), pregnant women (143) and patients with the following diseases : cir- tee, 1963), the adsorption method (Fiorelli rhosis (4), nephrotic syndrome (7), Cushing's syn- et al., 1972 ; Kawagoe and Hiroi, 1978) drome (7), Addison's disease (2), hyperthyroidism and gel filtration (De Moor et al., 1962 ; (10) and hypothyroidism (3). Doe et al., 1964). Recently immunological The gestational stages were divided into 3, as follows ; first (5-12 weeks), second (13-28 weeks) methods such as radial immunodiffusion (Van and third trimester (more than 28 weeks). Baelen and De Moor, 1979) and electroim- The sera were kept frozen at -20•Ž until de- munodiffusion (Racadot et al., 1976) have termination.

been reported. In spite of its various ad- Isolation of transcortin vantages, however, only one report on the Human transcortin was isolated from pooled RIA for human transcortin has appeared serum in four successive steps, i.e., ammonium sulfate fractionation, affinity column chromato- (Bernutz et al., 1979). graphy, gel-filtration column chromatography and In the present report, an RIA method to hydroxylapatite column chromatography. All steps determine human transcortin is described. were performed at 4•Ž. One liter of pooled serum The validity of the assay was ascertained. was utilized. The serum was fractionated by am- monium sulfate and the resulting precipitate from In addition, it was shown that the determi- 40 % to 65 % was dialyzed in water. The fraction nation of plasma transcortin concentrations was subjected to cortisol-Sepharose affinity column is useful in estimating protein-unbound chromatography according to Rosner and Bradlow cortisol concentrations from the inverse cor- (1971). The transcortin-rich fraction was con- centrated in a collodion bag, and applied to an relation between transcortin concentrations Ultragel AcAA column (1.6 x 100 cm) with 0.005 M and protein-unbound cortisol concentrations sodium phosphate buffer, pH 8.6, containing 5x under similar total cortisol concentrations. 10-4M ascorbic acid and employing 3H-cortisol as a tracer. The radioactive peak was dialyzed against 0.003 M sodium phosphate buffer, pH 6.8, and sub-

jected to an hydroxylapatite column (1 x 10 cm) chro- matography. The final procedure was repeated until 739 Vol 30, No 6 RADIOIMMUNOASSAY OF HUMAN TRANSCORTIN the purified protein showed a single band in poly- autowell gamma scintillation counter. Nonspecific acrylamide gel electrophoresis (Orstein, 1964 ; Davis, binding which precipitated without the antiserum 1964). The purity of the transcortin preparation was solution (blank) accounted for 5 % to 10 % of the also tested by sodium dodecyl sulfate (SDS) poly- total radioactivity. B/Bo % was calculated as fol- acrylamide gel electrophoresis (Laemmli, 1970) and lows ; immunoelectrophoresis (Graber and Williams, 1953), followed by protein staining and autoradiography. The transcortin concentration was determined by UV transmission using an molecular absorption coefficient of 0.645 for 1 cm at 279 nm (Muldoon Determination of serum transcortin concentrations by and Westphal, 1967). steroid-binding assay The molecular weight (MW) of the transcortin The cortisol-binding capacity of serum transcortin was estimated from its mobility on SDS polyacryl- was determined by the adsorption technique (Kawa- amide gel electrophoresis utilizing the following molecular weight standards : phosphorylase B (MW : goe and Hiroi, 1978). To begin with, 0.1 ml of serum with 0.5 ml of dextran coated charcoal (DCC) 94,000), bovine (BSA) (68,000), ov- was incubated at 45°C for 30 min. to remove albumin (45,000), carbonic anhydrase (30,000) and endogenous cortisol and the supernatant (stripped soybean trypsine inhibitor (21,000). serum) was separated by centrifugation. To over- The cortisol binding characteristics of the purified saturate the binding site of transcortin, radioactive transcortin were examined by Scatchard plot analy- cortisol (0.12 pCi/0.1 pg cortisol/0.1 ml) was added sis (Scatchard, 1949). to 0.5 ml of the stripped serum and incubated at 37•Ž for 30 min., the protein-unbound cortisol was Preparation of anti-transcortin antibody removed by DCC at 4•Ž and the radioactivity of Adult female New Zealand white rabbits were the supernatant (protein-bound cortisol) was counted. immunized 5 times every 3 weeks with 100 pg of To determine cortisol bound to albumin, the strip-

partially purified transcortin preparation (approxi- ped serum was heated for 30 min. at 60•Ž and was mate purity ; 50 %) emulsified with complete processed in the same manner as described above Freund's adjuvant. Blood was collected 15 days (alubumin-bound cortisol). after the last immunization and the serum was Cortisol bound to transcortin was expressed as stored at-20•Ž. This antiserum contained not the difference between the protein-bound and the only the anti-transcortin antibody but also albumin-bound cortisol. Assuming that the molecu- to albumin and other in the gamma- lar weight of cortisol and transcortin are 362 and fractions. 52,000 (Seal and Doe, 1966), the transcortin con- centrations were calculated as follows ; RIA procedure The purified transcortin was labeled with 125I by the chloramine-T method (Greenwood and Hunter, 1963). The labeled transcortin was purified by Sephadex G-25 and by Sephadex G-100 column

chromatography. The specific activity was between where 10: the factor to convert the unit from ƒÊg/ 3.6 and 3.8 pCil pg transcortin and maximal binding 0.1 ml to ƒÊg/ml of the labeled transcortin to antiserum was A : the factor to compensate for the loss of between 90 % and 100%. This tracer was kept sample during assay. frozen at -20•Ž under which condition its ability to bind with the antiserum was well maintained Determination of serum protein-unbound cortisol con- for 2 months. centration All of the solution was diluted with sodium Total cortisol concentrations in the sera were phosphate buffered saline, pH 7.4 containing 1 % determined by RIA. Protein-unbound cortisol con- BSA. A mixture consisting of 0.1 ml standard centrations were measured by the isocolloidosmolar transcortin solution or serum sample (dilution of equilibrium dialysis method (Matsui et al., 1979). 1 : 1,000), 0.1 ml tracer solution (10,000 cpm) and 0.1 ml antiserum solution (dilution of 1 : 100,000) was incubated for 2 days at 4°C. After the incubation, 50 ƒÊl of normal rabbit serum was added, and the Results separation of free and bound 125I-transcortin was performed by adding 1 ml of 25 % polyethylene Evaluation of the purified transcortin glycol solution in 0.1 M borate buffer, pH 8.6. The purified transcortin preparation After centrifugation at 3,000 rpm for 15 min., the radioactivity of the precipitate was counted with an showed a single protein band on SDS poly- Endocrinol. Japon. 740 SUEDA et al. December 1983 acrylamide gel electrophoresis. Ouchterony transcortin with a range between 1 and 100 double diffusion test showed one precipitin ng/tube is possible. Cortisol and progeste- line between transcortin preparation and rone did not affect the binding of transcortin anti-transcortin antiserum. Immunoelectro- to its antibody (B/Bo %), even when their phoresis also demonstrated a single precipitin molar concentration was 100 times that of line. When the gel was overlaid with 125I- the transcortin. also cortisol solution, washed with saline, dried did not interfere with the assay even at a and processed for autoradiography, the molar concentration of 10,000 times that of radioactivity of 125I-cortisol was clearly de- the transcortin. Thus this assay is free from tected only on the precipitin line (Fig. 1). the influence of endogenous steroids and The molecular weight of this preparation albumin in the serum. estimated by the mobility on SDS poly- When the reliability of this assay system acrylamide gel electrophoresis was 53,000. was examined, the dilution curve of normal Scatchard plot analysis indicated that the and pregnant sera were found to be parallel association constant of this protein to corti- to the standard curve within the range be- sol was 3.0 x 108 M-1 and that this protein tween 1 : 2,000 and 1 : 200 dilution. The binds to cortisol in a molar ratio of 1 : 1. rates of recovery for added transcortin in These results agreed well with the properties amounts of 20, 40 and 100 ng per tube were attributed to transcortin (Muldoon and West- 97 %, 100 % and 89 %, respectively. The phal, 1967 ; Seal and Doe, 1966). ranges of intra- and interassay variation co- efficients were between 6.4 % and 8.9 % and Evaluation of the RIA between 7.8 % and 13.9 %, respectively. A typical standard curve is depicted in The above results indicate the reliability of Fig. 2. It appears that the determination of the present RIA.

Fig. 1. Immunoelectrophoresis of transcortin preparation. The upper well (A) con- tained transcortin prepara- tion (30ƒÊg) and the lower well (B), undiluted human serum. Electrophoresis was

processed for 90 min. at 150 vol.with Tris-glycine buffer,

pH 8.6. Migration towards the right is to the anode. The upper trough (a) con- tained 30 ƒÊl of anti-human transcortin antiserum (rab- bit), the lower (b), anti- human whole serum anti- serum (horse). After dif- fusion for 24 h to permit formation of precipitin arcs, 25I-cortisol solution (105cpm/1

0.1 ml) was painted on the surface of the gel. The gel was washed, dried and pre-

pared for autoradiography using Kodak ( •~ -0 mat) AR film for 3 days, after which the gel was stained with Commassie Brilliant Blue R. Vol. 30, No. 6 RADIOIMMUNOASSAY OF HUMAN TRANSCORTIN 741

Sterod-biniding assay (ƒÊg/m1)

Fig. 3. Comparison between transcortin concentra- tions determined by the present RIA and those by steroid-binding assay. Cortisol-binding capacity of serum transcortin was determined by the adsorption technique (Kawa- goe and Hiroi, 1978).

Fig. 2. Standard curve of transcortin. Incubation volume was 0.3 ml ; 0.1 ml standard whether there were any similar effects on transcortin solution, 0.1 ml tracer solution and the RIA. Pooled serum was incubated at 0.1 ml antiserum solution. After incubation for 60•Ž for different time intervals, and pooled 2 days at 4°C, 50 pl of normal rabbit serum was added, and the separation of free and bound serum was adjusted to the pH with 1N tracer was performed by adding polyethylene NaOH or 1N HC1 for 1 h at 37•Ž and re- glycol solution. adjusted with 1 M Tris-HCl buffer, pH 7.4. After these treatments, transcortin concen- trations was measured by both RIA and Comparison of transcortin concentrations de- steroid-binding assay. As will be seen from termined by the present RIA with those by Table 1, transcortin values obtained by the steroid-binding assay RIA decreased with increased duration of Transcortin concentrations in several heat treatment as did those by steroid- serum samples were determined by both binding assay. The acid treatment also in- methods. The concentrations thus determin- duced a similar diminution in the transcortin ed correlated well with each other, as is values as detected by RIA and steroid- shown in Fig. 3. binding assay. It is well known that exposure to acid These results indicate that the antibody or heat destroys the steroid binding ability employed in this assay system recognizes of transcortin (Westphal, 1971 ; Daughaday transcortin which has the ability to bind et al., 1962). A test was run to determine with a steroid. Endoerinol. Japon. 742 SUEDA et al. December 1983

Table 1. Effects of heat and pH treatment on Table 2. Transcortin concentrations transcortin determination by RIA and steroid-binding assay

Values are mean•}SD.

In heat treatment, pooled serum was incubated at # statistically different from normal control women 60•Ž for different time intervals. In pH treatment, (p<0.001). the pooled serum was adjusted to the pH with 1N @ statistically different from pregnant women in NaOH or 1N HCl for 1 h at 37•Ž and readjusted the first trimester (p<0.001). with 1M Tris-HCl buffer pH 7.4. After each of * statistically different from pregnant women in these treatments, residual activity was measured by the second trimester (p<0.001). RIA and by steroid-binding assay. & statistically different from normal controls (p< 0.05).

Serum transcortin concentrations in normal

subjects, pregnant women and patients with ml) in Addison's disease were observed,

various diseases though the differences were not significant . Serum transcortin concentrations were The value of hyperthyroidism (41.8 ƒÊg/ml) determined by the present RIA in normal and of hypothyroidism (49.7 ƒÊg/ml) did not subjects, pregnant women and patients with differ appreciably, although levels in the several kinds of dysproteinemia as well as former were lower than those in the latter. several endocrinopathies, and the results are shown in Table 2. Transcortin concentra- Relation between transcortin concentrations tions in normal men and women were 43.2 and protein-unbound cortisol concentrations ± 10.3 ƒÊg/ml and 43.6 •}12.3 ƒÊg/ml respec- The correlation between transcortin and tively, and no sex difference was observed. protein-unbound cortisol concentrations was In pregnant women, the value was higher examined. The serum samples from normal than in non-pregnant women and as preg- controls, pregnant women and the patients nancy progressed, the mean concentrations with liver cirrhosis, nephrotic syndrome and increased gradually and reached its highest Cushing's syndrome were divided into four level in the third trimester which was groups according to transcortin concentra- two times that of non-pregnant women. tions, i. e., first group (below 30 ƒÊg/ml), The transcortin level was found to be signifi- second group (30-50 ƒÊg/ml), third group (50- cantly low in liver cirrhosis (Mean ; 30.4 70 ƒÊg/ml) and fourth group (above 70 ƒÊg/ μg/ml) and nephrotic syndrome (34.0 ƒÊg/ml). ml). In Fig. 4, protein-unbound cortisol Among adrenal diseases, a slightly low concentrations in each sample were plotted transcortin level (35.5 ,ƒÊg/ml) in Cushing' s against total cortisol concentrations. It is syndrome and slightly high level (50.8 ƒÊg/ apparent that protein-unbound cortisol con- Vol. 30, No. 6 RADIOIMMUNOASSAY OF HUMAN TRANSCORTIN 743

Fig. 4. The correlation be- tween transcortin and pro- tein-unbound cortisol con- centrations. Transcortin and total corti- sol concentrations were det- ermined by RIA and protein -unbound cortisol concent- rations were determined by the isocolloidosmolar equilibrium dialysis method (Matsui et al., 1979). The serum samples were divided into four groups according to their transcortin concen- trations. The protein-unbo- und cortisol concentrations were plotted against the tortal cortisol concentrations.

centrations increased with the increase in be processed by a single person. In addition, total cortisol concentrations. But for a given this method is so sensitive that it can de- level of total cortisol concentration the termine the transcortin concentration at protein-unbound cortisol concentration was levels as low as 10 ng/ml. In other words, it inversely related to the transcortin concen- requires only 0.1 ƒÊl of serum for transcortin tration. These results indicate that the determination. A high concentration of change in transcortin concentrations is close- cortisol does not affect the determination, ly correlated with the protein-unbound co- which means that this system is available rtisol concentrations. to measure transcortin whether it binds with cortisol or not. Therefore, the stripping of transcortin is not needed. This method is Discussion not influenced by the presence of albumin, providing much easier measurement than the Since the concentration of protein-un- steroid-binding assay. The reliability of the bound cortisol correlates inversely with the present method examined by dilution test, transcortin concentration under a given total recovery and precision tests is sufficient for cortisol concentration as shown in Fig. 4, a both clinical and laboratory use. simple method to determine transcortin is The transcortin concentrations determin- useful in estimating adrenocortical activity, ed by the RIA and the conventional steroid- especially in patients whose total cortisol binding assay showed a good correlation. concentration and clinical findings are not It is interesting to note that the affinity of in agreement. transcortin to its antibody and to steroids The RIA method described in the present showed a similar pattern of decline when study has many advantages over the con- subjected to increasing degrees of heat and ventional steroid-binding assay. This method acidity, as shown in Table 1. This also is very simple and hundreds of samples can indicates that the present RIA could tell Endocrinol. Japon. 744 SUEDA et al. December 1983 biologically active transcortin from the in- sidering the small number of patients studied, active form. further study is necessary before conclusive The transcortin levels in normal subjects results can be obtained. determined in this study are 43.2 and 43.6

μg/ml for normal men and women,respec- t ively. These values are slightly higher than Acknowledgement the 34.5± 5 pg/ml given by Bernutz et al.

(1979), 37•}3 ƒÊg/m1 by Schiller and Petra This study was supported in part by a Grant-in- (1976) and 23 to 45 ƒÊg/ml by Doe et al. Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan. (1964), but they are almost identical with the 41.5 ƒÊg/ml given by Van Baelen and De Moor (1974) and 39.7•}3.6 ƒÊg/m1 (men) and 42.1 •}3. 9 itig/m1 (women) by Racadot et References al. (1976). Transcortin concentrations in Bernutz, C., W. O. Hansle, K. Horn, C. R. Pickardt, various physiological and pathological states P. C. Scriba, E. Fink, H. Kolb and H. Tschesche also revealed similar patterns of variation as (1979). Isolation, characterization and radioim- reported earlier : a gradual increase through munoassay of -binding globulin (CBG) in human serum-clinical significance and com- pregnancy, lower than normal in nephrotic parison to thyroxine binding globulin (TBG). Acta syndrome and liver cirrhosis (Murphy and Endocr. 92, 370-382. Pattee, 1963 ; Doe et al., 1964 ; McCann Chan, D. W., M. Sharma and W. R. Slaunwhite Jr., and Fulton, 1975 ; Racadot et al., 1976). (1977). The chemistry of human transcortin. Im- These results also indicate the validity of proved affinity matrices for the purification of transcortin. Arch. Biochem. Biophys. 182, 197- the present method. 202. It is generally known that serum trans- Daughaday, W. H. (1958). Binding of corticosteroids cortin concentration increase following estro- by plasma protein. III. The binding of human plasma and plasma protein fractions. J. Clin. gen treatment or during pregnancy. The Invest. 37, 511-518. influence of other hormonal factors on serum Daughaday, W. H., R. E. Adler, I. K. Mariz and transcortin in humans has not been studied. D. C. Rasinski (1962). Measurement of the bind- ing capacity of corticosteroid-binding globulin in yet. However, Feldman et al. (1979) demon- human plasma. J. Clin. Endocr. 22, 704-710. strated that inhibited trans- Davis, B. J. (1964). Disc electrophoresis II. Method cortin synthesis in rabbit liver, while and application to human serum proteins. Ann. stimulated it. On the other hand, Labrie New York. Acad. Sci. 121, 404-427. et al. (1968) reported that thyroid hormone De Moor, P., K. Heirwegh, J. F. Heremans and M. Declerck-Raskin (1962). Protein binding of corti- is important in increasing transcortin syn- coids studied by gel filtration. J. Clin. Invest. 41, thesis. In the present study, serum trans- 816-827. cortin concentrations were slightly low in Doe, R. P., R. Fernandez and U. S. Seal (1964). Measurement of corticosteroid-binding globulin in patients with Cushing's syndrome and slight- man. J. Clin. Endocr. 24, 1029-1039. ly high in those with Addison's disease, Feldman, D., C. E. Mondon, J. A. Horner and J. suggesting an inhibitory action of gluco- N. Weiser (1979). and estrogen corticoids. These results are in agreement regulation of corticosteroid-binding globulin pro- duction by rat liver. Am. J. Physiol. 237 (6), E with Feldman's study. However, slightly high 493-499. transcortin concentrations were observed in Fiorelli, G., P. Piolanti, G. Forti and M. Serio hypothyroid patients and the opposite results (1972). Determination of plasma corticosteroids were observed in hyperthyroid patients. and urinary cortisol by a competitive protein- binding method using dextran-coated charcoal. Therefore, thyroid hormone might act in Clin. Chim. Acta 37, 179-187. increasing not only transcortin synthesis but Grabor, P. and C. A. Williams (1953). Methods also transcortin degradation. However, con- permettant letude conjugee des proprietes electro- Vol. 30, No. 6 RADIOIMMUNOASSAY OF HUMAN TRANSCORTIN 745

phoretiques et immunochimiques dim melange de bound plasma cortisol in pregnant and Enovid-E proteines. Application an sanguin. Biochim. Bio- treated women as determined by ultrafiltration. phys. Acta (Amst) 10, 193-215. J. Clin. Endocr. 29, 563-568. Greenwood, F. C., W. M. Hunter (1963). The Ornstein, L. (1964). Disc electrophoresis-I. Back preparation of 131I-labelled human growth ground and theory. Ann. New York Acad. Sci. hormone of high specific radioactivity. Biochem. 121, 321-349. J. 89, 114-123. Racadot, A., N. Racadot-Leroy, F. Le Gaillard and Kawagoe, S. and M. Hiroi (1978). A double ad- M. Dautrevaux (1976). Dosage de la transcortine sorption technique for the determination of corti- serique par electro-immuno diffusion. Clin. Chim. costeroid-binding globulin in human serum. Folia. Acta 66, 171-180. Endocrinol. Jap. 54, 1-10. Rosner, W. and H. L. Bradlow (1971). Purification Labrie, F., G. Pelletier, R. Labrie, M. A. Ho-kim, of corticosteroid-binding globulin from human A. Delgado, B. MacIntosh and C. Fortier (1968). plasma by affinity chromatography. J. Clin. Endo- Liaison transcortine- et contrOle de cr. 32, 193-198. l'activite hypophyso-surrenalienne chez le rat ; In- Scatchard, G. (1949). The attraction of proteins for teractions hypophyse-thyro1de-surrenales-gonades. small molecules and ions. Ann. New York Acad. Ann. Endocrinol. 29, 29-43. Sci. 51, 660-672. Laemmli, U. K. (1970). Cleavage of structural pro- Schiller, H. S. and P. H. Petra (1976). A filter assay teins during the assembly of the head of bacterio- for the corticosteroid binding globulin of human phage T4. Nature (London) 227, 680-685. serum. J. Steroid Biochem. 7, 55-59. Matsui, N. and J. E. Plager (1966). In vitro physio- Seal, U. S. and R. P. Doe (1966). In : Steroid dyna- logical activity of protein-bound and unbound mics (G. Pincus, T. Nakano and J. F. Taint eds.). cortisol. Endocrinology 78, 1159-1164. Academic Press, New York, pp. 63-90 Matsui, N., M. Yamamoto, H. Seo, M. Niinomi Slaunwhite, W.R. and A.A. Sandberg (1959). Trans- and Y. Ariyoshi (1979). Determination of serum cortin, a corticosteroid-binding protein of plasma. cortisol fractions by isocolloidosmolar equilibrium J. Clin. Invest. 38, 384-391. dialysis. Changes during pregnancy. Endocrinol. Van Baelen, H. and P. De Moor (1974). Immuno- Japon. 26, 263-273. chemical quantitation of human transcortin. J. McCann, V. J. and T. T. Fulton (1975). Cortisol Clin. Endocrinol. Metab. 39, 160-163. metabolism in chronic liver disease. J. Clin. Westphal, U. (1971). In : Steroid-Protein inter- Endocrinol. Metab. 40, 1038-1044. actions. (Gross F., A. Labhart, T. Mann and L.T. Muldoon, T. G. and U. Westphal (1967). Steroid- Samuels eds.) Springer-Verlag, Berlin-Heidelberg, New York. protein interactions. XV. Isolation and characteri- zation of corticosteroid-binding globulin from Yamamoto, M., Y. Ariyoshi and N. Matsui (1983). human plasma. J. Biol. Chem. 242, 5636-5643. The serum concentrations of unbound, transcortin Murphy, B. P. and C. J. Pattee (1963). A study of bound and albumin bound cortisol in patients the capacity of corticosteroid-binding globulin in with dysproteinemia. Endocrinol. Japon. 29 (5), 639-646. plasma. J. Clin. Endocrinol. Metab. 23, 459-464. O'connell, M. and G. W. Welsh, 3rd. (1969). Un-