Plasma Transcortin Research Articles

Plasma Transcortin Influences Endocrine and Behavioral Stress Responses in Mice

Glucocorticoids are released after hypothalamus-pituitary-adrenal axis stimulation by stress and act both in the periphery and in the brain to bring about adaptive responses that are essential for life. Dysregulation of the stress response can precipitate psychiatric diseases, in particular depression. Recent genetic studies have suggested that the glucocorticoid carrier transcortin, also called corticosteroid-binding globulin (CBG), may have an important role in stress response. We have investigated the effect of partial or total transcortin deficiency using transcortin knockout mice on hypothalamus-pituitary-adrenal axis functioning and regulation as well as on behaviors linked to anxiety and depression traits in animals. We show that CBG deficiency in mice results in markedly reduced total circulating corticosterone at rest and in response to stress. Interestingly, free corticosterone concentrations are normal at rest but present a reduced surge after stress in transcortin-deficient mice. No differences were detected between transcortin-deficient mice for anxiety-related traits. However, transcortin-deficient mice display increased immobility in the forced-swimming test and markedly enhanced learned helplessness after prolonged uncontrollable stress. The latter is associated with an approximately 30% decrease in circulating levels of free corticosterone as well as reduced Egr-1 mRNA expression in hippocampus in CBG-deficient mice. Additionally, transcortin-deficient mice show no sensitization to cocaine-induced locomotor responses, a well described corticosterone-dependent test. Thus, transcortin deficiency leads to insufficient glucocorticoid signaling and altered behavioral responses after stress. These findings uncover the critical role of plasma transcortin in providing an adequate endocrine and behavioral response to stress.

Cortisol and transcortin in human seminal plasma and amniotic fluid as estimated by modern specific assays

Cortisol concentrations in human seminal plasma, as estimated by the very specific Amersham 'Amerlite' luminescence immunoassay, were 176 +/- 43 (85-260) nmol/l, that is, 63.7 +/- 15.5 (31-94) ng/ml (mean +/- SD, n = 21). This is about 60% of random levels in blood serum and is the first description of cortisol in seminal fluid. In human amniotic fluid at 16-22 weeks of gestation, cortisol concentrations were lower, at 72.6 +/- 14.6 (63-124) nmol/l, that is, 29.3 +/- 5.3 (23-45) ng/ml (n = 21). Concentrations were about 15% of random maternal serum levels in the second trimester of pregnancy. The cortisol concentrations in both fluids were considerably higher than those reported for saliva, which has a mean of about 10 nmol/l. Transcortin (corticosteroid binding globulin, CBG), has been found in human seminal plasma and amniotic fluid for the first time. Concentrations were low, with values up to 12 micrograms/ml, with no significant difference between the two fluids, when using the IRE-Megenix monoclonal iodinated radioimmunoassay. Transcortin concentrations were about 10% of levels in non-pregnant blood serum, compared with about 0.1% for saliva. The higher concentrations of transcortin could perhaps account for the greater diffusion of cortisol into seminal plasma and amniotic fluid. The presence of beta-endorphin, ACTH and cortisol in amniotic fluid, seminal fluid, ovarian follicular fluid, endometrial fluid and gastric fluid may possibly, indicate the existence of a small paracrine ACTH-cortisol axis in the relevant secretory tissues.

Is the mineralocorticoid receptor in Brown Norway rats constitutively active?

In a previous study using corticosterone treatment of adrenalectomized rats, we hypothesized that mineralocorticoid receptor (MR)-related mechanisms are constitutively active and that glucocorticoid receptor (GR)-mediated mechanisms are more efficient in Brown Norway rats compared to Fischer 344 (F344) rats. In order to discriminate the mineralocorticoid from the glucocorticoid actions exerted by corticosterone, F344 and Brown Norway adrenalectomized rats were treated with increasing doses (1, 5 and 25 microg/ml of drinking water) of deoxycorticosterone (DOC, MR-specific ligand) or RU 28362 (GR-specific ligand). These rats were compared with long-term adrenalectomized (ADX) untreated rats and sham-ADX rats. This study confirms our previous results, notably the lack of effect of ADX on body weight and fluid intake in Brown Norway rats. Moreover, DOC treatment had no effect in Brown Norway rats whereas the higher dose restored fluid intake of the F344 ADX group to sham values. These results support the hypothesis of a constitutive activation of the MR and therefore the insensitivity of this receptor to its ligand in Brown Norway rats. Alternatively, RU 28362 treatment induced greater weight loss, decrease in food intake, anxiolysis, thymus involution, and decrease in plasma transcortin concentration and pituitary corticosteroid receptor densities in Brown Norway rats than in F344 rats, which is consistent with greater efficiency of GR mechanisms in Brown Norway rats than in F344 rats. Therefore, these strains are of great utility to disentangle MR and GR effects on complex phenotypes.

Strain differences in corticosteroid receptor efficiencies and regulation in Brown Norway and Fischer 344 rats.

During the dark phase of the diurnal cycle, and during recovery from restraint stress, Brown Norway (BN) rats secrete less corticosterone than Fischer 344 (F344) rats. These strains also display different levels of corticosteroid receptors in the hippocampus, and of plasma transcortin. Because corticosteroid receptors, plasma transcortin and corticosterone secretion are mutually regulated, we examined brain and pituitary mineralocorticoid receptor (MR) and glucocorticoid receptor (GR) expression and some of the parameters modulated by these receptors (i.e. body and thymus weight, fluid intake, plasma transcortin) in BN and F344 rat strains, by comparing the effects of either hormone deprivation by long-term (21 days) adrenalectomy (ADX), or chronic elevation of corticosterone given in drinking fluid to ADX rats. In BN rats, body weight gain and fluid intake were insensitive to corticosterone deprivation, suggesting that MR-related mechanisms are constitutively active in this strain. Body weight (b.w.) gain, plasma transcortin and thymus weight were reduced to a greater extent by chronic corticosterone in BN rats than in F344 rats, possibly as a consequence of higher free, active fraction of plasma corticosterone due to lower plasma transcortin concentrations and/or a greater efficiency of GR-related mechanisms in BN rats. F344 rats displayed twofold higher brain and pituitary MR levels than BN rats, whereas tissue-and strain-specific regulations were observed for GR levels. The differences in MR levels observed between BN and F344 strains cannot completely explain the differences in corticosterone actions, suggesting that strain differences in response to ADX or corticosterone treatment result from variable receptor efficiencies.

Hypothalamic-pituitary-adrenal axis activity in the inbred brown Norway and Fischer 344 rat strains

We have examined the basal and the stress-induced secretion of corticosterone in relation to the expression of adrenal steroid receptors in the pituitary, hypothalamus and hippocampus of the inbred Brown Norway and Fischer 344 rat strains. Our data indicated that plasma transcortin and integrated plasma corticosterone levels were significantly higher in Fischer 344 compared to Brown Norway rats. Fischer 344 hypersecrete corticosterone during the dark phase of the diurnal cycle and during the phase of recovery following a 20 min period of restraint stress compared to Brown Norway rats. This hypersecretion of corticosterone was negatively correlated with the size of the adrenal gland but might be related to the higher density of mineralocorticoid receptors in the hippocampus of Fischer 344 rats.

Ventromedial hypothalamic lesions inhibit corticosteroid feedback regulation of basal ACTH during the trough of the circadian rhythm.

We have determined the effects of bilateral electrolytic lesions of the ventromedial hypothalamus (VMH) on activity in the hypothalamo-pituitary-adrenal (HPA) system. Acutely, during the first 5 days, lesions of the anterior-medial VMH caused loss of the diurnal rhythms in food intake and plasma corticosterone (B) levels. Plasma B concentrations were elevated during the time of the normal trough of the basal diurnal rhythm in HPA axis activity and the diurnal rhythm in food intake was abolished, in agreement with the results of others. Consistent with hyperactivity in the HPA axis, lesioned rats had increased adrenal weight, decreased thymus and body weights and decreased plasma transcortin concentrations. To determine how lesions of the VMH provoke these increases in activity of the HPA system, the sensitivity of ACTH in adrenalectomized, lesioned rats to replacement with exogenous B was determined under basal conditions during the trough (morning-AM) and peak (evening-PM) of the diurnal rhythm in HPA axis activity. ACTH in lesioned rats in the AM was insensitive to feedback over the very low range of plasma B of 1-4 micrograms/dl, whereas sham-lesioned controls exhibited the normal, high sensitivity of ACTH to B at this time of day. There was no difference between the sensitivity of ACTH to this low range of B in the PM in VMH- and sham-lesioned rats. Two to 5 weeks after VMH lesions, as found by others, mean daily plasma B levels did not differ from sham-lesioned controls; however, plasma B during the AM was still mildly elevated in these rats. Inhibition of plasma B in the PM by dexamethasone was less effective in lesioned rats. Although HPA system responses to hypoglycemia, corticotropin-releasing factor and ACTH were normal, the lesioned rats exhibited obesity, hyperinsulinemia, hyperglycemia, hypertension and tachycardia, all signs consistent with mild hyperactivity of the PHA axis. Occupancy of type I, high-affinity corticosteroid receptors is known to control basal activity of the HPA system during the trough of the diurnal rhythm and to interact with glucocorticoid receptors to affect basal activity during the peak of the diurnal rhythm and during AM stress. We conclude that VMH lesions disrupt transmission of inhibitory signals, mediated by occupancy of type I corticosteroid receptors, that are initiated by a B feed-back site.

Hormonal function of the adrenal cortex in «medical castration» induced by prolonged infusion of LH releasing factor agonists

Time course of the major corticosteroid hormones and the precursors in the biosynthesis chain was studied, as was the type of corticosteroid complex formation with transcortin in the peripheral blood plasma in male papua hamadryas exposed to a 4-16 week infusions of LH releasing factor agonists (busereline, Hoechst A. G., Germany, and surfagon, Cardiology Research Center of the Russian Academy of Medical Sciences) with the use of osmotic mini-pumps (Alza Corp., Palo Alto, USA). Suppression of the gonadal function resultant from prolonged infusion of LH releasing factor agonists was not associated with essential shifts in the adrenal steroidogenesis and in the specific corticosteroid transport. Such exposure induced a somewhat more marked activation of glucocorticoid and adrenal androgen secretion in response to acute stressor exposure (a single injection of insulin in'a dose of 0.2 U/kg b. m.). One should bear in mind this fact when prescribing therapy with LH releasing factor agonists to patients with prostatic carcinoma, for the adrenal androgens are transformed into active androgens in the prostatic tissue.

A comparison between the composition of maternal peripheral plasma and plasma collected from the retroplacental compartment at caesarean section

In the pregnant uterus unique immunological conditions must prevail. Since compounds of possible fetoplacental origin may modulate the local maternal immunological response, higher concentrations of such compounds in the vicinity of the conceptus might be expected. We have measured the concentrations of several compounds in peripheral and retroplacental plasma, mainly by immunological methods, from women delivered by caesarean section in pregnancy week 38-40. Significantly higher (P less than 0.001) retroplacental concentrations were found for human placental lactogen, prolactin, alpha-fetoprotein, tissue polypeptide antigen, non-conjugated estriol and progesterone. Peripheral plasma levels were significantly higher for cortisol (P less than 0.01), transcortin (P less than 0.001) and sex-hormone binding globulin (P less than 0.002). No difference was found for albumin, alpha 2-macroglobulin, estradiol, placental alkaline phosphatase and human chorionic gonadotrophin. Significant correlations between retroplacental and maternal peripheral plasma levels were found for progesterone, cortisol, estradiol, estriol, transcortin, sex-hormone binding globulin, prolactin, human placental lactogen, human chorionic gonadotrophin, placental alkaline phosphatase, albumin and alpha 2-macroglobulin. There were also significant correlations between estradiol and sex-hormone binding globulin as well as transcortin in peripheral maternal plasma.

Characterization of Corticosterone Feedback Regulation of ACTH Secretiona

Adrenalectomy-induced increases in ACTH secretion in rats are returned to normal by an action of corticosterone on the brain, not on the pituitary. Five days after adrenalectomy with constant steroid replacement, the concentration of free corticosterone in plasma which reduces plasma ACTH by 50% is approximately 0.8 nM. By contrast, the concentration of free plasma corticosterone required for 50% reduction of thymus wet weight or plasma transcortin concentration (both targets for glucocorticoid action) is about 4.5 nM. These results suggested that the inhibition of ACTH by corticosterone might be mediated by association of the steroid with high affinity, type I corticosteroid receptors, whereas the inhibition of thymus weight and transcortin might be mediated by association of the steroid with lower affinity, type II receptors. The results of studies comparing the ability of corticosterone, dexamethasone and aldosterone to inhibit adrenalectomy-induced ACTH secretion support the hypothesis that basal ACTH secretion in rats is mediated by association of corticosterone with type I receptors.

Circadian rhythm in the membrane of circulating human blood cells: microviscosity and number of benzodiazepine binding sites. A search for regulation by plasma ions, nucleosides, proteins or hormones.

Circadian rhythms in both the number of peripheral type binding sites for benzodiazepines in platelet membranes and the microviscosity of the erythrocyte membrane were demonstrated in 7 healthy men. Neither variable appeared to be linked to each other, or regulated by the plasma concentrations of total or free cortisol, testosterone, potassium, magnesium, calcium, cAMP, cGMP or proteins or by the erythrocytic concentration of magnesium or potassium or by the plasma cAMP:cGMP ratio or by the ratio of intra-erythrocyte:plasma concentrations of potassium or magnesium. A highly significant negative correlation was found between the microviscosity of the erythrocyte membrane and the activity of the membrane-bound enzyme, methyltransferase I. Such a correlation was validated both on raw data and on 24 hr-means (r = 0.84; P less than 0.01). A circadian rhythm in the activity of this enzyme was also demonstrated. Moreover, a highly significant correlation was also found between plasma transcortin concentration (TRC) and microviscosity (r = 0.50, P less than 0.01), and between TRC and methyltransferase I activity (r = 0.61, P less than 0.01). Such findings may constitute clues towards the understanding of the regulation of the circadian rhythm in the fluidity of the red blood cell membrane in man and guide future steps with regard to the role of this rhythm upon the availability of drug binding sites at the cell surface.

Tissue Differences in the Up-Regulation of Glucocorticoid-Binding Proteins in the Rat*

The magnitude of increase in glucocorticoid receptor concentration and transcortin-like binding was examined in a variety of peripheral tissues and brain structures after adrenalectomy. Glucocorticoid binding was assayed in liver, heart, kidney, pituitary, hippocampus, cerebral cortex, amygdala-entorhinal area, and hypothalamus. Glucocorticoid receptor concentration, measured using [3H]dexamethasone as ligand, increased in all eight tissues, but the magnitude of this increase varied 30-fold among tissues. The largest increase was shown by kidney cytosol, followed by amygdala-entorhinal cortex, hippocampus, liver, cerebral cortex, hypothalamus, pituitary, and heart. These increases were not due to a selective enhancement of mineralocorticoid receptors. The increase in transcortin in peripheral tissues was variable and exceeded the increase in plasma transcortin by an order of magnitude. It was concluded that up-regulation of the glucocorticoid receptor after adrenalectomy is a response common to most, if not all, glucocorticoid target tissues. However, the magnitude of this response was tissue specific and was not directly related to initial receptor density. The marked increase in tissue transcortin ([3H]corticosterone binding in the presence of excess dexamethasone) suggested that plasma transcortin is sequestered by peripheral tissues in substantial amounts in the acutely adrenalectomized rat. The increase in transcortin uptake by tissues and the increases in cytosolic receptor number are apparently subject to different regulatory control.

Distribution of 18-hydroxycorticosterone between red blood cells and plasma.

Plasma 18-hydroxycorticosterone (18-OH-B) to aldosterone (aldo) concentration ratios reflect adrenal corticosterone methyloxidase type II activity. This ratio is determined not only by the relative secretion rates of the two steroids but also by differences in binding, distribution, and metabolism. Plasma cortisol alters the distribution of aldo between red blood cells (RBC) and plasma. We postulated that the distribution of 18-OH-B, like that of aldo, is determined by the availability of high affinity binding sites on plasma transcortin. Double equilibrium dialyses demonstrated that 18-OH-B, aldo, and cortisol compete for binding sites on transcortin. Increasing amounts of each of the three unlabeled steroids produced progressive decrements in the binding of all three labeled steroids to transcortin. The affinity of 18-OH-B (2 X 10(6) M-1) for transcortin was intermediate between those of cortisol (3 X 10(7) M-1) and aldo (0.9 X 10(6) M-1). Heat treatment of plasma decreased the binding of 18-OH-B and cortisol to transcortin by 82% and 75%, respectively. Gel filtration of plasma revealed that protein-bound [3H]18-OH-B and [14C]cortisol eluted in the same fractions. The addition of increasing quantities of unlabeled cortisol to whole blood in vitro produced similar increments in RBC to plasma concentration ratios of [3H]18-OH-B and [14C]aldo. The ratio of the percentage of circulating 18-OH-B in plasma to the percentage of circulating aldo in plasma was constant in blood containing low and high cortisol concentrations. Therefore, changes in plasma cortisol have similar effects on the distribution of 18-OH-B and aldo between RBC and plasma.

Chick oviduct glucocorticosteroid receptor. Specific binding of the synthetic steroid RU 486 and immunological studies with antibodies to chick oviduct progesterone receptor.

The glucocorticosteroid receptor (GR) has been studied in oviduct cytosol prepared from estrogen-primed, 4-week-withdrawn chicken. The equilibrium dissociation constant was 6 nM for dexamethasone, and 18 300 receptor sites/cell were measured assuming that all cells contain identical concentrations of GR. Dexamethasone, used in most studies investigating glucocorticosteroid action, was found not to be the best GR ligand. The affinities of several natural and synthetic glucocorticosteroids for GR increased in the following order: cortisol less than deoxycorticosterone less than dexamethasone less than corticosterone less than triamcinolone acetonide. The synthetic steroid RU 486 was the most specific ligand of GR (its affinity was approximately equal to 10-fold higher than that of triamcinolone acetonide), while it did not bind either to plasma transcortin (which binds dexamethasone nor, surprisingly, to progesterone receptor (PR), contrary to what occurs in mammalian species. The molybdate-stabilized, 8-S form of GR was prepared from withdrawn chick oviduct, whole chick embryo or cultured chick embryo fibroblasts (which do not contain PR), and was labeled with either [3H]dexamethasone or [3H]RU 486. The sedimentation coefficient of radioactive ligand--8-S GR complexes was shifted towards heavier forms after incubation with polyclonal (IgG-G3) or monoclonal (BF4) antibodies generated against the molybdate-stabilized, 8-S form of the chick oviduct PR. Since neither IgG-G3 nor BF4 interacted with the steroid binding 4-S form of GR, it is suggested that these antibodies recognized a non-steroid binding protein common to molybdate-stabilized, 8-S forms of GR and PR.

Human transcortin synthesis by a cell-free translation of hepatic mRNA.

To evaluate the site of synthesis and to characterize the translated transcortin, poly (A)-containing RNA (mRNA) from human liver was translated in a cell-free system derived from rabbit reticulocyte lysate. The in vitro synthesized product was identified as transcortin by immuno-precipitation with its specific antiserum. This translated transcortin could be displaced from the antibody by unlabeled purified transcortin obtained from plasma. Furthermore, when the translation mixture was applied to a cortisol-Sepharose column, the translated transcortin was bound to the matrix in a specific manner, indicating that this product binds to cortisol. The molecular weight of the translated transcortin was estimated to be 45,700 by its mobility in sodium dodecyl sulfate polyacrylamide gel electrophoresis, while that of plasma transcortin was 53,800. The difference in molecular weight between the translated transcortin and plasma transcortin was probably due to the presence of pre-sequence (signal peptide) in addition to the absence of carbohydrate moiety in the former. In conclusion, human liver mRNA directed the synthesis of transcortin, and the translated transcortin binds to cortisol in spite of the absence of carbohydrate moiety.

Endometrial and pituitary responses to the steroidal antiprogestin RU 486 in postmenopausal women.

The effects of the antiprogestin RU 486 on the human endometrium were investigated. Seventeen postmenopausal women were injected with estradiol (E2) benzoate (0.625 mg/day) for 15 days. Progesterone (P) (25 mg/day) and/or RU 486 (100 or 200 mg/day) were given to groups of 2-3 women during the last 6 days of E2 benzoate treatment. Serial blood samples were drawn for the measurement of plasma E2, P, and LH and FSH. An endometrial biopsy was performed on the last day of treatment, and processed for histology or for assays of DNA polymerase alpha, E2-dehydrogenase (E2DH), and P receptor (PR). Treatment with E2 benzoate alone resulted in a marked decrease of plasma gonadotropins; in those patients who received either P, RU 486, or both, in addition to E2 benzoate, the concentrations of plasma LH and FSH were further decreased to premenopausal levels. In absence of glycerol, the affinity of RU 486 for the endometrial PR (Kd = 0.8 nM) was higher than that of P (Kd = 1.2 nM). Glycerol decreased markedly the affinity of RU 486, whereas the affinity of P for the PR was unchanged. RU 486 had negligible affinity for plasma transcortin. Either P or RU 486, but not both together, induced secretory changes in the endometrium as determined from histologic sections of tissue biopsies. Either P or RU 486 decreased DNA polymerase alpha and increased E2-DH activities in the endometrium. Unexpectedly, when P and RU 486 were given together. E2-DH activity remained at the level found in E2-treated women. In vitro cultures of proliferative endometrium treated with the synthetic progestagen R 5020 or with RU 486 also had increased E2-DH activity; RU 486 counteracted R 5020 effects. We conclude that, contrary to previous results with experimental animals, the anti-P RU 486 has some progestomimetic activity in humans under specific conditions. Paradoxically, when given together with P, RU 486 lost most of its progestomimetic activity in the endometrium and behaved as a pure antagonist.

Relative binding affinity of steroids for the corticosterone receptor system in rat hippocampus

In cytosol of the hippocampus corticosterone displays highest affinity for the sites that remain available for binding in the presence of excess RU 26988, which is shown to be a “pure” glucocorticoid. A rather high affinity (≧25%) was found for 11β-hydroxyprogesterone, 21-hydroxyprogesterone, 5α-corticosterone, 19-nor-deoxycorticosterone, 11-deoxycorticosterone and cortisol. A moderate affinity (>5% and <25%) was displayed by about 14 steroids among which progesterone, aldosterone, 9α-fluorocortisol and dexamethasone. Corticosterone also shows highest affinity to plasma transcortin and thymus cytosol in the presence of RU 26988. However, the rank-order in affinity by the competing steroids was distinctly different from that observed in the hippocampus; cf. aldosterone and dexamethasone displaced [ 3H]corticosterone from sites unoccupied by RU 26988 in the hippocampus but not from transcortin or sites in thymus cytosol. In thymus cytosol some potent glucocorticoids have higher affinity for the [ 3H]dexamethasone labeled sites than dexamethasone. The binding of [ 3H]dexamethasone in thymus cytosol is completely abolished in the presence of a 100-fold excess of RU 26988. We conclude that our data support the evidence for RU 26988 as a selective ligand for glucocorticoid receptors. RU 26988 leaves binding sites available with highest affinity for corticosterone in hippocampus cytosol that are distinct from transcortin-like sites as found in thymus cytosol or from plasma transcortin.

Polymerization of human transcortin in plasma

Using radial immunodiffusion and antiserum raised against purified transcortin, polymers of native transcortin have been identified in plasma. They are characterized by multiple precipitin bands, the size of which diminishes on dilution, consistent with reversibility of this form of polymer. Spectrophotometric scans have been used to study the time course of polymerization in purified transcortin, in which dilution reversibility can be demonstrated also, and in which disaggregation may be produced by addition of sodium dodecyl sulphate and dithiothreitol, but apparently not by cortisol.

Physico-chemical properties and evidence for electrophoretic variants of rat transcortin

Isolation of rat plasma transcortin was carried out by affinity chromatography, as previously described for human. The protein was shown to be pure by PAGE and one single N-terminal amino acid was identified (Ser), which suggested that the protein molecule has a single polypeptide chain. This assumption is supported by SDS-PAGE. The amino acid composition was reported and compared with the one of human transcortin. The purified protein always migrated in PAGE (with or without SDS) as a double band; the faster component being more intense than the slower one. Whether transcortin was free or bound to corticosterone, the same aspect was observed. Molecular weight of these two variants were determined by SDS-PAGE as 65,900 and 75,800. Ploymers only appeared after irreversible denaturation of the protein, as previously described for human transcortin. Various other physical parameters were determined: a sedementation coefficient of 3.71 S ± 0.18 was calculated by ultracentrifugation in sucrose gardient, association constants at 4°C for corticosterone and cortisol (2.7 × 10 9 M − and 4.2 × 10 8 M −, respectively).

Androgen control of transcortin binding capacity in adult male ducks

In adult male ducks submitted to marked variations in plasma testosterone concentration, plasma transcortin (CBG) levels were shown to be closely related to the level of plasma testosterone. In connection with previous data on female ducks, the results strongly support the evidence that at least in this species, CBG is under a stimulatory control by testosterone.

Intracellular CBG-like molecules in the rat pituitary

The localization of transcortin (CBG) in pituitary cells of the rat was investigated using the peroxydase-antiperoxydase (PAP) technique. A rabbit antiserum against purified rat plasma transcortin was used as the primary antiserum. Transcortin-like (CBG-like) immunoreactive products were found in the cytoplasma of certain cells in the anterior pituitary, but not in the intermediate lobe and weakly in the posterior pituitary. It is postulated that the CBG-like molecules participate in the cellular uptake process of corticosterone, thereby modulating the feedback signal of this steroid on pituitary function.