ISOLATION, CHARACTERIZATION AND RADIOIMMUNOASSAY OF CORTICOSTEROID-BINDING GLOBULIN (CBG) IN HUMAN SERUM – CLINICAL SIGNIFICANCE AND COMPARISON TO THYROXINE-BINDING GLOBULIN (TBG)

Isolation of the corticosteroid-binding globulin CBG was achieved by 5 chromatographical steps on Cortisol Sepharose, QAE-Sephadex A-50, Con A-Sepharose and hydroxylapatite. The purity of the isolated CBG was demonstrated in Polyacrylamide gel electrophoresis, SDS electrophoresis, immunodiffusion and ultracentrifligation. Microheterogeneity was shown in isoelectric focusing by 5 bands in the pH range of 3.7-4.2, which could be reduced to one major band after neuraminidase treatment. The equi- molar binding of Cortisol to CBG was demonstrated by binding studies. The association constant for Cortisol was 2.8 χ ΙΟ8 Μ -1, for progesterone !» Abteilung für Klinische Chemie und Klinische Biochemie der Chirurgischen Klinik der Universität, Nußbaumstraße 20,. D-8000 München 2, -> Forschergruppe Diabetes, Städtisches Krankenhaus München-Schwabing, Kölner Platz, D-8000 München 40, •J» Fakultät für Chemie der Universität Bielefeld, Univcrsitätsslraße, D-4800 Bielefeld 1 Dedicated to Professor Dr. rer. nat. Dr. med. h.c. Theodor Bücher at the occasion of his 65th birthday. Supported by Deutsche Forschungsgemeinschaft (SFB 51). a> Preliminary results were presented in part at the 22nd Symposium of the Deutsche Gesellschaft für Endokrinologie, Travemünde, 1977. C. Β. was awarded with the "Marius-Tausk"-Prize of the Deutsche Gesellshaft für Endokrinologie 1978 for this work. Abbreviations used in this paper: CBG, corticosteroid-binding globulin; TBG, thyroxine-binding globulin. 370 I.7xlO(5M_1. From analytical ultracentrifugation, the molecular weight was calculated on 50 700; the sedimentation coefficient was 3.6 S, the partial specific volume 0.690 ml/g, the Stokes radius 38 Ä and the fric- tional coefficient ratio 1.5. A sped lie radioimmunoassay for CBG was established using the purified CBG for immunization, radioiodination and for calibration standards. The normal range of CBG levels in human serum was 2.4-4.4 mg/100 ml (mean ± 2 SD). Studies were performed to compare the levels of CBG and thyroxine-binding globulin (TBG). No sex differences but a significant biphasic age dependence were observed for both proteins. In pregnancy and under oestrogen treatment of women and men, CBG was demonstrated to be the more distinct indicator of oestrogenic activity as compared with TBG, whereas the sensitivity of TBG was more pronounced to supposedly antiestrogenic substances like Danazol, and in severe disease. No coin­ cidence of genetic CBG and TBG deficiencies have been found so far. In human serum. Cortisol is bound to a specific transport protein, corticosteroid- binding globulin (CBG) or transcortin, with high affinity and low capacity which migrates on paper electrophoresis as an alpha1-globulin. In addition Cortisol is bound to albumin which has low affinity, but high capacity (Daugha- day I956a,b; Slaunwhite 8c Sandberg 1959; Slaunwhite et al. 1966; Middoon 8c Westphal 1967). The concentration of CBG was until now estimated by measuring the total binding capacity of serum for Cortisol using gel filtration and equilibrium dialysis {De Moor el al. 1962; Murphy 8c Pattee 1963; Westphal 1971; Schwartz 8: Hammerslein 1975; Angelt el al. J 977), and recently more specifically by radial immunodiffusion {Rosner el al. 1973; Racadot el al. 1974; Van Baelen 8c De Moor 1974). There is major evidence that CBG has a buffer function for the biologically active free hormone fraction in blood {Slaunwhite et al. 1962; De Moor et al. 1963; Sandberg 8c Slaunwhite 1963), rather than an active transport function for steroid hormones to the target organ cell. But a possible role of proteins in hormone-receptor interactions is also discussed {Westphal 1971; Werthamer el al. 1973; Wong et al. 1973). The CBG concentrations in human serum have been estimated mainly under the aspect of oestrogen in­ fluence {Sandberg 8c Slaunwhite 1959; De Moor el al. 1962; Doe el al. 1964; Schwartz 8c Hammer stein 1975) and genetic deficiency variations (Rosner el al. 1973). Recently, characteristic changes of the thyroxine-binding globulin (TBG) have been demonstrated under the influence of different metabolic conditions {Horn et al. 1977; Horn 8c Gärtner 1979). Therefore, it was investigated, whether parallelism in changes of these two different transport proteins, CBG and TBG could be detected. For this purpose CBG had to be isolated from human serum and a specific

for its determination had to be established. 371 24 ·•'•• Μ Α Τ Ε RIALS AND Μ Ε Τ HODS Reagents. - Pure Cortisol was purchased from the Merck AG, Darmstadt, FRG, [·5Η] Cortisol (110 mCi/mg Cortisol) from the Radiochemical Center in Amersham, England, and Cortisol hemisuccinate from Sigma Chemical Company, St. Louis, USA. Sodium1-r> iodide (10 Ci/mg 1) came from the Hoechst AG, Frankfurt, FRG. AH-Sepharose 4B for affinity chromatography, Con A-Sepharose, QAE-Sephadex A-50 for ion exchange chromatography and Sephadex G-10 were obtained from Pharmacia Fine Chemicals, Uppsala, Sweden. The hydroxylapatite Biogel HTP was purchased from Bio-Rad Laboratories, Richmond, USA. Florisil, acrylamide and bovine gamma globulin were obtained from the Serva Biochemica, Heidelberg, FRG. Sodium Lauryl Sulphate (SDS) from Sigma Chemical Company, St. Louis, USA. Bovine albumin and alpha-methyl- D-mannoside from Roth, Karlsruhe, FRG, neuraminidase (Clostridium perfringens) from Boehringer Mannheim GmbH. The following reagents came from the Merck AG, Darmstadt, FRG: N-ethyl-N'-ivS-dimethylaminopropylJ-carbodiimide-hydrochlorid, po- lyethylenglycol 6000, complete Freund's adjuvant and all reagents (pro-analysis grade) for the preparation of the buffer solutions. Human transferrin and pertussis vaccine were obtained from the Behringwerke, Marburg, FRG. The human plasma anti- coagulated with acid-citrate-dextrose (ACD) was freshly obtained from the blood bank and immediately used for preparation. Isoelectric focusing. - Isoelectric focusing studies (Radola Sc Graesslin 1977) of isolated CBG were performed in slab gels (16x0.4 cm), using LKB equipment (LKB produkter, Bromma, Sweden). The concentrations of Polyacrylamide and Ampholine® solutions pH 3.5-5.0 and pH 3.5-10 respectively (LKB produkter) were each 5% (v/v). After isoelectric focusing (500 V, 20 W, 6 hours) gels were either stained using bromo- phenol blue or cut in slices of 0.5 cm. After elution with destilled water by diffusion pH-values were measured with a m.icroelectrode, the CBG-concentrations by radio­ immunoassay and [:5H] Cortisol in a /i-scintillation counter. Isolation of CBG from human plasma Affinity chromatography. - Using a modification of the method described by Cuatrecasas (1970), 2 g Cortisol hemisuccinate (=4.1 mmol) dissolved in 50 °/o dime- thylformamide were covalently bound to 15 g AH-Sepharose 4B (=0.48 mmol amino groups) by activation with 2.17 g carbodiimide (= 10 mmol) for 20 h at room tempera­ ture. The Cortisol Sepharose was then washed alternatively with 1 Μ glycine and 0.2 Μ Tris-HCl buffer pH 8.6. From 5 liters human plasma the endogenous steroids were removed by 50 g llorml. Then, plasma and Cortisol Sepharose were stirred for 30 min at room temperature and for 60 min at 4°C. The plasma was then filtered, the gel was washed with 1000 ml cold 0.2 Μ Tris-HCl buffer pH 8.0 and packed in a water- jacketed column. From this column CBG was eluted by elevation of the temperature up to 30°C with 20 mg Cortisol hemisuccinate in 100 ml 0.16 Μ NaCl in 0.05 Μ Tris- HCl buffer pH 8.6. Purification of CBG by column chromatography techniques The following purification steps were performed at 4°C. The CBG peak from Cortisol Sepharose was given on a QAE-Sephadex A-50 column (4.2 χ 24 cm) which was equilibrated with 0.18 Μ NaCl in 0.05 Μ Tris-HCl buffer pH 8.6. After washing the gel with one column volume of starting buffer, CBG was eluted by the elevation of the NaCl concentration up to 0.22 Μ in 0.05 Μ Tris-HCl buffer ρ Η 7.4 in a volume of 372 90 ml. This CBG peak was transferred on a Con A-Sepharosc column (3.2x22 cm). The column was washed with one column volume of 0.05 Μ sodium phosphate bulier pH 7.4 and CBG eluted with 0.06 Μ alpha-methyl-D-mannoside in the same buffer. This CBG peak of 70 ml was dialyzed and concentrated in a collodion bag. The following chromatography was performed on a hyclroxylapatite column (1.2x45 cm) equilibrated with 0.001 Μ sodium phosphate buffer pH 6.8. The same buffer was used for the elution step. The CBG peak was concentrated and transferred to a QAE- Sephadex A-50 column (2.2x45 cm), equilibrated with 0.19 Μ NaCl in 0.05 Μ Tris- HCl buffer pH 8.6. The final elution was performed with a linear gradient (400 ml) from 0.19 to 0.24 Μ NaCl in 0.05 Μ Tris-HCl buffer 8.6 in a volume of 75 ml. The CBG solution was repeatedly dialyzed in a collodion bag against aqua bidest., frozen, lyophilized and stored at - 20°C. The CBG peaks in the eluates from each column were identified using a modification of the [:*H]cortisol-uptake-test on 5 ml Sephadex G-50 {Horn et al 1975). Radioimmunoassay of CBG Immunization. - Rabbits were immunized with 120 ug CBG in complete Freund's adjuvant and 0.5 ml pertussis vaccine antigen using a multiple intradermal injection technique (50 sites) on the back flanks. They were boosted every three weeks with the same technique. Radioiodinalion of CBG was done with the chloramine-T-method {Greenwood et al 1963). The separation of the CBG tracer was performed only on a 1 ml Con A- Sepharose column. The CBG tracer was eluted with 0.06 Μ alpha-methyl-D-m.annoside and diluted in 2 g/1 bovine albumin in 0.05 Μ sodium phosphate buffer pH 7.4. The tracer could be used for 8 weeks without further purification. For the radioimmunoassay 100 μ\ CBG standard solution or 1:100 diluted serum were incubated with 100 μ] CBG tracer and 100 μ\ diluted CBG antiserum. The bound/free-separation could be performed by the polyethylene glycol precipitation {Desbuquois Sc Aurbac/i 1971) or the double antibody technique. R Ε S U L Τ S Criteria of purity and characterization Elerlrophorclical methods. - The isolated CBG migrated on Polyacrylamide gel electrophoresis and SDS electrophoresis {Maurer 1971) each performed in two different buffer systems pH 8.9 and pH 7.0 and three different gel con­ centrations (10, 7.5 and 5%) in one homogeneous band. Immunoelectrophoresis and Ouchterlony double diffusion test demonstrated a single precipitation line between CBG antiserum and CBG solution and normal human serum, respec­ tively. The overlapping of both precipitation lines indicated identity between endogenous and isolated CBG. Isoelectric focusing studies revealed a microheterogeneity of 5 different bands in the pH range of pH 3.7-4.2. In the same pH area CBG was detected in the radioimmunoassay as well as the radioactivity of [3H] Cortisol pre- incubated with CBG. Neuraminidase treatment of CBG reduced the micro- 373 heterogeneity to one major and one minor band in the pH range of pH 6.0 (Fig. 1). Analytical ultracenlrif ligation. — The purity and the homogeneity of the isolated CBG could be demonstrated by a homogeneous curve in sedimentation velocity runs and by a straight line in the high speed sedimentation equilibrium runs. The sedimentation coefficient corrected to 20° C and water {Schachmann 1957; Schachmann 8c Edelstein 1966) was calculated to be 3.6 S. Considering a partial specific volume of 0.690 ml/g derived from the amino acid and carbohydrate composition (Table 1) a molecular weight (Yphantis 1964) of 50 700 ± 2500 (n = 4) for CBG was determined. The diffusion coefficient was calculated to be 5.6 D, the Stokes radius 38 A and the frictional coefficient ratio 1.5. Quantitative amino acid analysis in two different CBG preparations is sum­ marized in Table 1. The high content of aspartic acid and glutamic acid explains the low isoelectric point of CBG. The carbohydrate composition is shown in the lower part of Table 1. pH 5.0 10.0 f pH 3.5 3.5 Fig. L Isoelectric focusing of isolated CBG. Left: untreated CBG, pH gradient 3.5-5.0. Right: Desialylated CBG (0.03 U/ml neuraminidase, acetate buffer pH 5.6. 30 min) pH gradient 3.5-10. 374 Table I. Amino acid and carbohydrate composition of CBG. The amino acid analysis was performed with two different preparations of CBG, the carbohydrate analysis only with one preparation. Assuming a molecular weight of 50 700 the left column indicates the numbers of amino acid residues per mol CBG, the right column the percentage of the single amino acids of the polypeptid residue (molecular weight 32 950). The lower part of the table indicates the carbohydrate composition of the second preparation of CBG. amino acids: No. of residues/mol CBG δ­of residue/100 g of polypeptid preparation I II ι II Lysine 15 15 6.7 6.7 Histidine 11 12 5.2 5.7 Arginine 10 10 5.3 5.3 Aspartic acid 35 37 14.1 14.9 Threonine 18 17 6.5 6.1 Serine 21 16 6.7 5.1 Glutamic acid 33 34 14.7 15.2 Proline - 8 - 2.7 Glycine 11 10 2.5 2.2 Alanine 14 15 3.7 4.0 Cystine - - - - Valine 18 21 6.4 7.4 Methionine 12 11 5.4 5.0 Isoleucine 15 18 5.9 7.1 Leucine 35 39 13.9 15.5 Tyrosine 13 9 7.1 4.9 Phenylalanine 23 25 11.8 12.8 carbohydrate compos tion in % by weight II Mannose 9.5 Galactose 5.3 Glucosamine 10.3 Sialic acid 10.0 375 Bijiding of steroid hormones to the isolated CBG The binding of several steroids to CBG was investigated by two different methods, firstly equilibrium dialysis in micro-cells (Dianorm Apparatus, Dia- chemica AG, Switzerland) as the reference method and secondly, gel filtration on small columns with 2 ml Sephadex G-10 as a very simple method for estimation of the relative affinity constants of several steroids. The binding studies were performed in phosphate buffer pH 7.4 at 4° C. The incubation time for gel filtration was 15 min, the dialyzing time 8 h. Equilibrium dialysis was performed for four different steroids using firstly the corresponding tracer and secondly [3H] Cortisol in order to test for cross reactivity between the different steroids and Cortisol tracer in CBG binding. There was a large measure of agreement between the calculated binding constants of each method (Table 2). Gel filtration resulted in lower association constants than equili­ brium dialysis. This was due to the disturbance of equilibrium during the filtration on Sephadex G-10. The binding constant for Cortisol was found to be approximately 8 times lower in gel filtration than in equilibrium dialysis. Utilizing this factor for the correction of the association constants of 17a- hydroxy-progesterone, 11-deoxyCortisol, progesterone and testosterone, each calculated by displacement of [8H] Cortisol using gel filtration, the corrected 0.4- 0,2- Scatchard Plot b= 2.2x108 r = 0.97 ι 1,5 2,5 bound Cortisol [ nM ] Fig. 2. Determination of the association constant of Cortisol and isolated CBG by means of Scatchard plot. Ordinate: ratio of CBG bound and free Cortisol tracer. Abscissa: amount of bound Cortisol, calculated as product of the percentage of [3H] Cortisol (B/T) and the molecular concentration. 376 Table 2. Association constants of different steroids, determined by equilibrium dialysis using the corresponding tracer for each steroid (first column), equilibrium dialysis using [3H] Cortisol tracer only (second column), and gel filtration on Sephadex G-10 using [3H]Cortisol tracer only (third column). For methodological details see text. In the right part of the table the molecular dif­ ferences of the investigated steroids compared with Cortisol are indicated. association constants CM*1] molecular differences compared with Cortisol — pregnane structure *- steroid equilibrium dialysis gel filtration Δ 1 dehydro ο ο ο m 5 £ η?. α ο ο ο " 5 τ. •Ι 5' _. > :ι α Andro- stene structure Fluorine in 9 oc Est rogene Structure Cortisol Correspond irx Tracer 2.2 Χ108 ; Cortisol Tracer 2.2x10ö 2.2χ10β high affinity Prednisolone 5.4χ107 • high affinity Corticosterone 5.0Χ107 • high affinity 17a -Medroxy­ progesterone 5.3 χ 107 4.9χ107 28χ107 • • high affinity 11 -Deoxy - Cortisol 1.3 χ 107 1 6 χΙΟ7 2.8χ107 • high affinity Deoxy cortico­ sterone 2 2χ107 • • high affinity Cortisol - 21 - succinate 1.6 χ 107 • moderate affinity Cortisone 7.2 χ 10* • moderate affinity Methyl - prednisolone Αθχ 106 • • moderate affinity Progesterone 17 χ 10β 5.7>. 106 Α.1 χ 106 • • • moderate affinity Prednisone 3.4χ106 • • moderate affinity Aldosterone 2 6χ 106 • • moderate affinity Testosterone 2.5 χ 10β 2.9χ10β 16 χ 106 • low affinity 9 a - Fluoro - hydrocortisone 5.6χ104 • low affinity Estriol 2Λχ104 • low affinity Estradiol 2 3χ104 • low affinity Triamcinolone - acetonid 1.6 χ 10* • no affinity Dexamethasone, Ethinylestradiol, Carbenoxolone values were in the same order of magnitude as found in the equilibrium dia­ lysis (Table 2). The maximal binding capacity calculated by Scatchard analysis was found to be 715 tug Cortisol per 100 mg CBG indicating an equimolar binding of CBG and Cortisol (Fig. 2). By comparing the association constants with the molecular differences of several steroids to Cortisol (Table 2) it might be supposed that the binding affinity decreased in dependence on the electron attraction and the size of the substitute. 377 Radioimmunologlial determination of CBG. - The antiserum with the highest titer of 1:160 000, determined by 50% tracer binding, was obtained after the third booster. The maximal tracer binding (specific activity: 31 //Ci///g CBG) was nearly 100%. The nonspecific binding without antiserum in the reaction mixture was in the range between 5 to 10%. Using an antiserum dilution of 1:3000 the limit of detection (3 SD from the zero standard) was 2 ng CBG per tube. The 50% intercept was 24 ng CBG per tube, the recovery of added CBG standard in serum was 100%. dilution curves of normal and pregnancy sera were found to be exactly on the calibration curve. The interassay variation coefficient was 7.4%) (mean 3.3 mg/100 ml; η = 23). There was no evidence for cross reactivity of CBG antiserum with alpha.>- macroglobulin, alphao-haptoglobin, alpha!-antitrypsin and albumin in the Ouchterlony double diffusion test, cross reactivity of TBC* was excluded in the radioimmunoassay. Cortisol in serum had no effect on the results of CBG determination. Comparison of CBG and TBG levels in human serum CBG levels in human serum. - In 40 control persons between the ages of 15-50 years the range of serum CBG was 2.4-4.4 mg/100 ml. CBG levels of patients with Cushing's syndrome (n = 4) and Addison's disease (n = 4) were CBG (mg /100ml) A TBG [mg/100 ml] <5 50 years age dependence Fig. 3. Age dependence of CBG and TBG levels. Hatched columns: CBG levels in mg/100 ml. Open columns: TBG levels in mg/100 ml. Hatched lines indicate the normal range of CBG and TBG in controls in the age of 15 to 50 years. I mean db so. 378 CBG [mg/100ml) 1 TBG [mg/100ml]r CBG and TBG levels of 4 women subsequently treated with 4 different oestrogen preparations. The hatched columns: CBG levels, the open columns: TBG levels. The hatched lines represent the normal ranges of CBG and TBG for the age from 15 to 50 years. found in this range. In 134 healthy controls the CBG levels showed a biphasic age dependence of CBG levels (Fig. 3). Significant sex differences of CBG and TBG levels could not be ascertained in any period of life. Oestrogen dependence of CBG and TBG. - In pregnancy CBG as well as TBG levels increased continuously and reached a plateau during the third trimenon (6.9 ± 0.24 mg/100 ml, η - 16). The effect of exogenous oestrogens on CBG and TBG levels was investigated in four ovarectomized women (Fig. 4). The patients were subsequently treated with mestranol (80 //g/day), ethinyl oestradiol (80 //g/day), oestradiol valeri­ anate (40 mg/14 days), and oestriol (2 mg/day), each preparation for four weeks. Between the phases of oestrogen therapy, a period of placebo administration was inserted for 4 weeks. The response of CBG to oestrogens was more pronounced than the TBG increase. The maximal increase for both was ob­ served after mestranol, whereas oestriol had no effect (Fig. 4). In 10 male 379 patients undergoing fosfestrol treatment (Honvan®) for a prostatic carcinoma, a dose-related elevation of CBG levels was observed. Both, CBG and TBG levels reached values of females treated with oestrogens, and again CBG was shown to be the more sensitive indicator of oestrogenic activity. CBG levels in TBG deficiency stales. - TBG deficiency was induced by treatment with Danazol (2,3-isoxazol-derivative of ethinyl-testosterone, Winthrop, Gießen, FRG). Fifteen women were treated with 400 mg Danazol/ day for endometriosis, since Danazol is known to induce endometrium atrophy, and to suppress ovulation and midcycle peaks of gonadotrophins and oestradiol, whereas the mean basal values of LH and oestradiol remain constant (Goebel & Rjosk 1978). The TBG levels were decreased to 50 °/o of the initial values after four weeks of therapy, whereas the CBG levels did not change signi­ ficantly even after 12 weeks (Fig. 5). In 7 patients with severe chronic diseases such as decompensated liver cir­ rhosis and chronic heart failure, TBG levels were found to be decreased to 0.72 ± 0.23 mg/100 ml. This was interpreted as symptomatic TBG deficiency. In these patients the mean CBG levels were decreased as well to 2.6 ±1.0 mg/ 100 ml. But the decrease of CBG was less pronounced than that of TBG. In nine patients (8 men and 1 female) with genetic TBG deficiency (0.4 ± 0.3 mg/100 ml, ± SD), CBG levels were found to be in the normal range. CBG 1 mg/100 ml] i TBG [mg/100ml) mg before A 8 12 weeks after Danazol - therapy Fig. 5. CBG and TBG levels during Danazol therapy. Hatched columns: CBG levels, open columns: TBG levels. The hatched lines represent the normal range of CBG and TBG for the age from 15 to 50 years. 380

The principle of affinity chromatography as described by Cualrecasas (1970) and used at first by Rosncr 8c Bradlow (1971) for the isolation of CBG was modified in our study. Although Rosncr 8c Bradlow (1971) performed the coupling of Cortisol hemisuccinate in pure dioxane and Le Gaillard el al. (1974) thought the coup­ ling in 50% dimethylformamide not to be practicable, we found that the best

were obtained with the latter method. For displacement of CBG from the Cortisol Sepharose, Cortisol hemisuccinate addition to the elution buffer was preferred in order to avoid possible denaturation of CBG which may occur if more aggressive eluents are used. The observation of Rosner 8c Bradlow (1971) who found only CBG and gamma globulins in the eluate of the affinity chromatography column could not be confirmed. Therefore, several additional chromatographic purification steps were required. The overall yield of 20% after 5 different preparatory steps was satisfactory. The purity of the isolated CBG was shown by a single band in overloaded Polyacrylamide and SDS electrophoreses and by ultra- centrifugation studies. The properties of CBG ascertained in our laboratory agreed for the most part with the results of other authors. The molecular weight is reported to be in the range of 49 500 to 58 500 {Slaunwhite el al. 1966; Muldoon 8c Westphal 1967; Le Gaillard et al. 1975). By ultracentrifugation, we determined the value of 50 700, the sedimentation coefficient of 3.6 S and the partial specific volume of 0.690 ml/g, which corresponds to the values published by Westphal (1971). The hydrodynamic parameters particularly the frictional coefficient ratio sug­ gest that CBG can still be regarded as a globular protein. The Stokes radius of the molecule is about 38 Ä which corresponded well with other proteins with a molecular weight in this range. The association constants for Cortisol and for other steroids partially obtained by two different methods agreed well with the literature {Westphal 1977; Stroupe el al. 1978). Only the affinity of pro­ gesterone to CBG was found to be lower {Westphal 1971). The carbohydrate content of 35% by weight and the mean N-acetylneura- minic acid content of approximately 16 residues per mol isolated CBG was surprisingly high as compared with the literature {Slaunwhite el al. 1966; Le Gaillard et al. 1975; Rosner 1976). After treatment with neuraminidase the microheterogeneity of CBG consist­ ing of 5 single bands was focused to one major band into the alkaline direction. The residual more acidic minor band may be due to incomplete desialylation. Therefore the microheterogeneity of CBG may be due only to the different N-acetylneuraminic acid content as is already established for other glyco­ proteins and as recently shown for TBG {Horn 8c Gärtner 1979). 381 Radioimniunological quantitation of CBG in scrum. - Although Rosner el al. (1973) and Van Baelen Sc De Moor (1974) preferred the subcutaneous or intra­ muscular application technique using a ten-times higher amount of CBG, a monospecific antiserum was obtained by the intracutaneous injection of ap­ proximately 120 //g CBG. As the precision and practicability of the radioim­ munoassay was satisfactory, the method was preferable to the radial immuno­ diffusion technique (Rosner el al. 1973; Racadol el al. 1974: Van Baelen Sc De Moor 1974) for measuring CBG concentrations. The CBG levels in serum of healthy adults corresponded well with the values estimated by the earlier published methods (Westphal 1971; Rosner et al. J973; Racadol ct al. 1974: Van Baelen Sc De Moor 1974; Rosner 1976). CBG levels in normal controls. - Elevated binding capacities of CBG in serum have been observed in newborns and infants by several authors (De Moor et al. 1962; Angeli el al. 1977; Wagner 1978). These findings could now be ascertained by the direct CBG radioimmunoassay and in addition a further increase of CBG was seen in elder subjects. This biphasic age dependence was parallel with the TBG levels (Horn el cd. 1977). Sex differences of CBG or TBG levels could not be ascertained in any period of life. These results were surprising with regard to the known oestrogen influence on the levels of both proteins, therefore apart from the well known oestrogen induced increase of both transport proteins supposedly other factors have an influence on CBG and TBG. Oestrogen dependence of CBG and TBG levels. - The oestrogen induced in­ crease of CBG and TBG levels, well known from earlier investigations (Doe et al. 1964; Sandberg Sc Slaunwhile 1959; Jngbar 1971; Horn el al. 1977; Wagner 1978). was now ascertained by the quantitative and specific radioimmunological determination. The investigation of sera in pregnancy and during oestrogen therapy of women and men demonstrated a more pronounced increase of CBG as compared with TBG. The increase is probably due to an augmented syn­ thesis of this protein as the half-life is identical in controls and oestrogen treated persons (Sandberg el al. 1964). After therapy of four women with sup­ posedly equivalent doses of four different oestrogen preparations, no increase of CBG and TBG was found after oestriol. As expected, the increase was significantly higher after ethinyloestradiol as compared with oestradiol valerianate. Surprisingly the most pronounced increase was induced by me- stranol. which may be due to its hepatic metabolism (Bird Sc Clark 1973). CBG levels in TBG deficiency stales. - During Danazol therapy, only a decrease of the TBG levels was observed, while the CBG levels did not change. During the Danazol therapy, clinical signs of peripheral oestrogen deficiency were observed despite of normal oestradiol levels. There is possibly an inter- 382 ference of Danazol and oestrogens on liver cell receptors, which effects only TBG decreases. Likewise, the decrease of TBG levels in catabolic states was more pronounced than that of CBG. These differences in the behaviour of CBG and TBG under various hormonal and metabolic influences cannot yet be explained. No coincident genetic defects of the two different transport proteins have been observed. R Ε F ΕR ENCES Angeli Α., Frajria R.7 Richiardi L., Agrimonti F. Sc Gaidano G.: Clin. chim. Acta 77 (1977) 1. Bird C. E. Sc Clark A. F.: J. clin. Endocr. S(i (1973) 296. Citatrccasas P.: J. biol. Chem. 245 (1970) 3059. Daughaday W. H.: J. clin. Invest. 35 (\956a) 1428. Daughaday W. H.: J. clin. Invest. 35 (1956/;) 1434. De Moor P., Deckx R., Raas J. Sc Denef C: Metabolism 12 (1963) 592. De Moor P., Heirwegh K.. Heremans J. F. Sc Declerck-Raskin M.: J. clin. Invest. 41 (1962) 816. Desbuquois B. Sc Arnbach G. D.: J. clin. Endocr. 33 (1971) 732. Doe R. P., Fernandez R. N, Sc Seal U. S.: J. clin. Endocr. 24 (1964) 1029. Goebel R. Sc Rjosk FL Κ.: Geburtsh. u. Frauenheilk. 38 (197S) 932. Greenwood F. C, Hunter W. M. Sc Glover J. S.: Biochem. J. 89 (1963) 114. Horn K. Sc Gärtner R.: Acta endocr. (Kbh.) Suppl. 225 (1979) 433. Horn K.. Henner /.. Müller Ο. Α. Sc Scriba Ρ C: Ζ. klin. Chem. 13 (1975) 173. Horn Ä'., Kubiczek ΓΛ., Pickardt C. R. &· Scriba P. C: Klin. Wschr. 55 (1977) 881. lngbar S. H. In: Werner S. C. and Ingbar S. H.. Eds. The Thyroid. Harper and Row, New York (1971) 243. Le Gaillard F.y Han Κ. K. Sc Dautrevaux M.: Biochimie 57 (1975) 559. Le Gaillard F.. Racadot Α.. Racadol-Leroy N. Sc Dautrevaux M.: Biochimie 56 (1974) 99. Maurer IL R.: Disc Electrophoresis and Related Techniques of Polyacrylamide Gel Electrophoresis. De Gruyter, Berlin-New York (1971). Muldoon T. B. 8c Westphal U.: j. biol. Chem. 242 (1967) 5636. Murphy Β. Ε. P. Sc Pat tee C. J.: J. clin. Endocr. 23 (1963) 459. Racadot /l.. Racadol-Leroy N. 8: Dautrevaux M.: Lille Med. 19/9 (1974) 938. Radola B. J. Sc Gracssl'ui D.: Electrofocusing and Isotachophoresis, Dc Gruyter, Berlin- New York (1977). Rosner W. In: Jamieson G. A. and Grecnwalt 1 . J., Eds. The binding of steroid hor­ mones in human serum. Progress in Clinical and Biological Research, Vol. 5. Elsevier. Amsterdam (1976) 377. Rosner W. 8c Bradlow H. L.: J. clin. Endocr. S3 (1971) 193. Rosner W., Darmstadt R. A. Sc Toppel S.: J. clin. Endocr. 37 (1973) 983. Sandberg A. A. Sc Slaunwhite W. R. jr.: J. clin. Invest. 38 (1959) 1290. Sandberg A. A. Sc Slaunwhite W. R. jr.: j. clin. Invest. 42 (1963) 51. Sandberg Α. Α., Woodruff M. Sc Rosenthal H.: J. clin. Invest. 43 (1964) 461. Schachmann Η. Κ.: Methods in Enzymology 4 (1957) 32. Schachmann Η. Κ. Sc Edelstein S. ].: Biochemistry 5 (1966) 2681. 383 Schwartz U. Sc Hammerstein J.: Ζ. klin. Chem. 13 (1975) 291. Slaunwhile W. R. jr., Lockte G. N., Back N. Sc Sandberg Α. Α.: Science 135 (1962) 1062. Slaunwhile W. R. Sc Sandbcrz Α. Α.: .]. clin. Invest. 3S (1959) 384. Slaunwhile W. R. jr., Schneider S., Wissler F. C. Sc Sandberg Α. Α.: Biochemistry ο (1966) 3527. Stroupe S. D., Gray R. D. Sc Westphal U.: FEBS Letters 86 (1978) 61. Van Baelen H. Sc De Moor P.: J. clin. Endocr. 39 (1974) 160. Wagner R. K.: Acta endocr. (Kbh.) Suppl. 218 (1978) 5. Werthamer S.. Samuels A. J. Sc Amoral /.; J. biol. Chem. 248 (1973) 6398. Westphal U.: Steroid-Protein Interactions. Monographs on Endocrinology. Springer- Verlag, Berlin-Heidelberg-New York (1971). Westphal U.: Klin. Wschr. 55 (1977) 877. Wong K. C., Kornel L., Bezkorovainy A. Sc Murphy Β. E. P.: Biochim. biophys. Acta (Amst.) 328 (1973) 133. Yphcmtis D. Α.: Biochemistry 3 (1964) 297. Received on January 18th, 1979. 384