Changes in Endocrine Orcadian Rhythms as Markers of Physiological and Pathological Brain Aging

We studied the circadian rhythm of plasma melatonin, growth hormone (GH), prolactin (PRL), adrenocorticotropic hormone (ACTH), and Cortisol in 52 mentally healthy old subjects, 35 old demented patients, and 22 clinically healthy young controls. When compared to young controls, the circadian profile of plasma melatonin of old subjects, both demented or not, was clearly flattened, particularly during the night. The selective impairment of nocturnal melatonin secretion was significantly related to both the age and the severity of mental impairment (Mini Mental State Examination [MMSE] score). The PRL and GH circadian profiles were similar in the three groups during the day, but a significant lowering of the values recorded during the night occurred with aging. The impairment of the nocturnal secretion was related to the subjects' age and, for the GH secretory pattern only, also to the MMSE score. The ACTH circadian profile was similar in the three groups studied, even when old subjects exhibited higher ACTH levels throughout the 24h cycle, compared to young controls. Significantly higher Cortisol values at evening- and nighttime occurred in elderly subjects and particularly in the demented group. Both the mean levels and the nadir values of plasma Cortisol were positively related to age and negatively to MMSE score. In order to verify the sensitivity of the hypothalamo-pituitary-adrenal (HPA) axis to the steroid feedback, the circadian profile of plasma Cortisol was evaluated also after dexamethasone (DXM) administration (1 mg at 23:00h); the sensitivity of the HPA axis was significantly impaired in old subjects and particularly in the demented ones. These findings suggest that the neuroendocrine alterations already present in physiological aging, due to both anatomical damages and unbalanced central neurotransmitters, are enhanced in senile dementia. (Chronobiology International, 14(4), 385–396, 1997)     

Evaluation of hypothalamic-pituitary function in a combination of tests with four hypothalamic releasing hormones and L-dopa in normal subjects and in patients with hypothalamic and/or pituitary disorders

Hypothalamic-pituitary function was evaluated in a combination of tests with four hypothalamic releasing hormones (4RHs) and L-dopa in normal subjects and in patients with hypothalamic and/or pituitary disorders. Plasma concentrations of anterior pituitary hormones (GH, ACTH, TSH, PRL, LH and FSH) were measured before and after simultaneous iv administration of GHRH, CRH, TRH and LHRH. In addition, changes in the plasma levels of GHRH and GH were investigated before and after oral administration of L-dopa. Normal subjects showed appreciable responses to both tests. In five patients with hypothalamic disorders, the response of plasma anterior pituitary hormones varied, but plasma GHRH and GH did not respond to L-dopa. Patients with idiopathic and postpartum hypopituitarism showed low response to 4RHs or none at all, but L-dopa evoked a normal GHRH response in 2 of the 4 cases having no GH response. In the patients with hypopituitarism due to resection of a pituitary tumor, the response of anterior pituitary hormones to 4RHs was low, and L-dopa administration induced a normal GHRH and low GH response in 5 out of the 7 cases. After 4RHs administration, the patients with ACTH deficiency syndrome showed different patterns of impaired ACTH secretion, and isolated, combined or limited ACTH reserve. Seven patients with anorexia nervosa showed exaggerated GH, delayed TSH and FSH, low ACTH and LH, that is, normal PRL response to 4RHs, but no response of plasma GHRH or GH to L-dopa, suggesting the presence of hypothalamic dysfunction. These results indicate that the combination of the 4RHs test and L-dopa test is a simple and useful means for evaluating hypothalamic-pituitary function by measuring the response of plasma GHRH and six anterior pituitary hormones in the patients with endocrine disorders.     

Circadian cortisol secretion and circadian adrenal responses to ACTH are maintained in dexamethasone suppressed capuchin monkeys (Cebus apella)

We tested the hypothesis that the capuchin monkey adrenal (Cebus apella) gland has oscillatory properties that are independent of adrenocorticotropic hormone (ACTH) by exploring under ACTH suppression by dexamethasone: (i) maintenance of a circadian rhythm of plasma cortisol and (ii) clock time dependency of plasma cortisol response to exogenous ACTH. The capuchin monkey had a clear ACTH and plasma cortisol rhythm. Dexamethasone treatment resulted in low non-rhythmic ACTH levels and decreased cortisol to 1/10 of control values; nevertheless, the circadian rhythm of plasma cortisol persisted. We found that cortisol response to exogenous ACTH was clock time-dependent. The maximal response to ACTH occurred at the acrophase of the cortisol rhythm (0800 h). These results suggest that the capuchin monkey adrenal cortex may possess intrinsic oscillatory properties that participate in the circadian rhythm of adrenal cortisol secretion and in the circadian cortisol response to ACTH.     

Passive immunization with an anti-oCRF41 immune serum inhibits the circadian increase of plasma ACTH in rats

For 24 hrs. after i.v. injection of 1 ml of an undiluted immune serum raised against oCRF41, the diurnal surge of plasma ACTH dropped to a short-lived limited rise above baseline level. On the second day after injection, the ACTH level in treated rats rose to a subnormal level, although both plasma dilution of the immune serum and its binding capacity in the plasma remained unchanged throughout the experiment. Plasma corticosterone, on the contrary, displayed a normal circadian rhythm during the entire experiment. However, in animals given a second injection of 0.5 ml oCRF41 immune serum 32 hrs. after the first, both ACTH and corticosterone titers fell rapidly below their circadian minimal levels in controls. Concomitantly, the concentration of immune serum in the peripheral plasma, and its capacity to bind to oCRF, rose by 50%. The major role of CRF41 as a diurnal trigger of the circadian rhythm of ACTH is discussed, as well as the limits of passive immunization.     

Persistance of phase-coupling between the circadian rhythms of hypophyseal hormones in free-running rats

The circadian rhythms in plasma ACTH, TSH, LH and PRL were explored in sighted or blind, spayed and estrogen-implanted rats. A marked endogenous circadian rhythmicity was shown to persist in the blind animals for the 4 endocrine rhythms. The endogenous rhythms also kept very close reciprocal phase relationship as in the synchronized state, and they were peaking almost simultaneously, after 60 d. of free-running. Finally the endogenous hormonal rhythm maintained their usual phase relationships with the endogenous activity rhythm, so that the circadian phase of increased hormonal secretion coincided with the circadian resting phase of the sleep/wake rhythm. These results are discussed in the light of the alternate theory of one vs multiple but phase-locked circadian pacemakers driving endocrine and behavioral circadian rhythms.     

Evaluation of the inhibitory effect of circulating corticosteroids on circadian stimulated and stress induced secretion of ACTH in rats

Male rats were bilaterally adrenalectomized in order to measure the extent of inhibition exerted by endogenous corticosteroids on both basal ACTH secretion along its circadian rhythm and ether-stress induced ACTH secretion. In intact controls, plasma ACTH levels at the circadian maximum exceeded by 4 times the circadian minimum, and ACTH response 15 min after ether-stress surpassed the circadian minimum by 20 times. In adrenalectomized rats, the daily minimum was 8 times that of the controls. Nevertheless the circadian maximum was 3 times above the rhythm's minimum, while the maximal stress response (15 min) surpassed the circadian minimum by 8 times. In adrenalectomized rats supplemented with a solid source of corticosterone inducing a stable plasma corticosterone level equivalent to the controls' circadian minimum (3 micrograms/100 ml), the ACTH rhythm still fluctuated twice as high as in intact controls. The tonic feed-back inhibition exerted by endogenous corticosteroids on ACTH secretion appeared thus significantly stronger than the GABAergic inhibition to the corticotropic system which was previously studied under similar standard conditions.     

Seasonal, circadian and episodic variations of human immunoreactive beta-MSH, ACTH and cortisol

The 24-hour profile of plasma levels of immunoreactive beta-MSH (IR-beta MSH), ACTH and cortisol was obtained at 15-min intervals in six normal males in summer and winter. In the radioimmunoassay used, dilution curves of human beta-MSH and human beta-LPH were not parallel. A seasonal variation in basal pituitary-adrenal secretion, with higher levels in winter than in summer, was demonstrated. A circadian rhythm was found to be significant for ACTH, IR-beta MSH and cortisol in all investigations. Whereas ACTH and cortisol patterns were largely concordant in all studies, there was a significant desynchronization of the circadian rhythm of IR-beta MSH as compared to ACTH in five cases. Eighty-three percent of the secretory spikes of cortisol but only 68% of the IR-beta MSH spikes were concomitant with an ACTH spike. Correlations between maximal levels of concomitant spikes were higher during the quiescent period of pituitary-adrenal secretion (22.00--04.00) for ACTH-cortisol whereas for ACTH-IR-beta MSH, highest correlations were found during the active early morning secretory phase (04.00--10.00). For the three plasma constituents studied, longer apparent half-lives were found to occur when the basal level before spiking was already elevated, suggesting that ACTH, beta-LPH and cortisol are secreted in bursts superimposed on a continuous basal secretion. Absolute increments of concentration appeared to be relatively independent of the level before spiking. It is suggested that the dissociations between ACTH and IR-beta MSH fluctuations in plasma observed here result from in vitro proteolysis of beta-LPH.     

Neuro-endocrine correlations of hypothalamic-pituitary-thyroid axis in healthy humans

Neuro-endocrine hormone secretion is characterized by circadian rhythmicity. Melatonin, GRH and GH are secreted during the night, CRH and ACTH secretion peak in the morning, determining the circadian rhythm of cortisol secretion, TRH and TSH show circadian variations with higher levels at night. Thyroxine levels do not change with clear circadian rhythmicity. In this paper we have considered a possible influence of cortisol and melatonin on hypothalamic-pituitary-thyroid axis function in humans. Melatonin, cortisol, TRH, TSH and FT4 serum levels were determined in blood samples obtained every four hours for 24 hours from ten healthy males, aged 36-51 years. We correlated hormone serum levels at each sampling time and evaluated the presence of circadian rhythmicity of hormone secretion. In the activity phase (06:00 h-10:00 h-14:00 h) cortisol correlated negatively with FT4, TSH correlated positively with TRH, TRH correlated positively with FT4 and melatonin correlated positively with TSH. In the resting phase (18:00 h-22:00 h-02:00 h) TRH correlated positively with FT4, melatonin correlated negatively with FT4, TSH correlated negatively with FT4, cortisol correlated positively with FT4 and TSH correlated positively with TRH. A clear circadian rhythm was validated for the time-qualified changes of melatonin and TSH secretion (with acrophase during the night), for cortisol serum levels (with acrophase in the morning), but not for TRH and FT4 serum level changes. In conclusion, the hypothalamic-pituitary-thyroid axis function may be modulated by cortisol and melatonin serum levels and by their circadian rhythmicity of variation.     

Diurnal rhythm of galanin-like immunoreactivity in the paraventricular and suprachiasmatic nuclei and other hypothalamic areas

The peptide galanin (GAL), when injected into the rat hypothalamus, is known to stimulate feeding behavior and affect the secretion of various hormones, including insulin and the adrenal steroid, corticosterone. To determine whether endogenous peptide levels shift in relation to natural rhythms of feeding and circulating hormone levels, rats were sacrificed at different times of the light/dark cycle, and their GAL levels were measured, via radioimmunoassay, in medial hypothalamic dissections and micropunched hypothalamic areas. The results suggest the existence of two distinct diurnal rhythms for hypothalamic GAL. One rhythm, detected exclusively in the area of the SCN, is characterized by bimodal peaks of GAL, threefold higher than basal peptide levels, around the onset of the dark and light periods. The second rhythm shows a single peak of GAL towards the middle of the nocturnal feeding cycle, specifically between the third and sixth hour. This latter rhythm is evident in the dorsal region of the medial hypothalamus, localized specifically to the lateral portion of the PVN. Moreover, it is inversely related to circulating insulin but unrelated to the adrenal steroids, suggesting a possible association between this pancreatic hormone and GAL in the PVN.     

Single exposure to testosterone in adulthood rapidly induces regularity in the growth hormone release process

The neonatal gonadal steroid milieu is known to be important in imprinting the striking sexual dimorphism of growth hormone (GH) secretion; however, the influence of the sex steroids on GH control in adult life and their mechanism/site of action are largely unknown. In the present study, we tested the hypothesis that testosterone (T) subserves the gender-specific regularity of the GH release process in adulthood. The approximate entropy statistic (ApEn) was used to quantify the degree of regularity of GH release patterns over time. Eighteen hours after a single subcutaneous injection of 1 mg T, both sham-operated and ovariectomized (OVX) female adult rats displayed plasma GH profiles that were strikingly similar to the regular male-like ultradian rhythm of GH secretion. The highest ApEn values, denoting greater disorderliness of GH secretion, were observed in the ovary-intact group, and T injection significantly (P < 0.001) reduced this irregularity whether or not the ovaries were present. Serial intravenous injections of GH-releasing hormone (GHRH) caused a similar increase in plasma GH levels in sham-operated females independently of time of administration. In contrast, female rats administered T exhibited a male-like intermittent pattern of GH responsiveness to GHRH, the latter known to be due to the cyclic release of endogenous somatostatin. These results demonstrate that acute exposure to T during adult life can rapidly and profoundly "masculinize" GH pulse-generating circuits in the female rat. Our findings suggest that the enhanced orderliness characteristic of the GH release process in males, compared with females, is regulated by T. We postulate that this T-induced regularity is mediated at the level of the hypothalamus by inducing regularity in somatostatin secretion, which in turn governs overall GH periodicity.     

Circadian Profile of Serum Melatonin in Cushing's Disease and Acromegaly

We evaluated the circadian profiles of serum melatonin (MT) and Cortisol in 6 patients with Cushing's disease while those of serum MT and GH were evaluated in 8 patients with acromegaly. The control group consisted of 15 healthy subjects in whom MT, Cortisol and GH were determined. The presence of a circadian rhythmicity was validated by the cosinor method, while the diurnal and nocturnal amount of MT secretion were expressed in terms of area under the curve. Gross alterations of MT rhythm were not apparent in Cushing's patients. In acromegalics, we observed a blunted day-night oscillation of MT accounted for by a significant increase of its secretion during the day-time period.     

Ovine corticotropin-releasing factor administration in normal men. Pituitary and adrenal responses in the morning and evening

We administered ovine corticotropin-releasing factor (CRF) as a bolus intravenous injection (1 microgram/kg) at 09.00 and at 20.00 to assess the influence of circadian changes in the hypothalamic-pituitary-adrenal axis on the response to CRF. The increase in plasma ACTH levels after CRF was only slightly lower in the morning than in the evening. The plasma cortisol response to ACTH, however, was significantly greater in the evening than in the morning (p less than 0.005). At both times of day CRF administration had no effect on plasma concentrations of GH, PRL, LH, AVP, insulin, PRA or glucose. No effects were observed on the hematopoietic system, kidneys or liver. In addition, CRF had no effect on heart rate, blood pressure or respiratory rate at the dose employed. Approximately 10% of the subjects complained of a transient upper body and facial hot flush. These observations indicate that the magnitude of the plasma cortisol rise after CRF depends on the time of administration.     

Failure of beta-endorphin antiserum, naloxone, and naltrexone to alter physiologic growth hormone and insulin secretion.

The role of endogenous opiate-like peptides in physiologic regulation of growth hormone (GH) and insulin (IRI) secretion was assessed by passive immunization with β-endorphin antiserum and by administration of the opiate antagonists naloxone and naltrexone. Six-hour secretory profiles were obtained from 5 groups of freely-moving chronically cannulated male rats following the i.v. administration of (I) β-endorphin antiserum, (II) normal rabbit serum, (III) naloxone (1 mg/kg), (IV) naltrexone (1 mg/kg), and (V) normal saline. The typical ultradian rhythm of GH secretion was evident in all groups with most peak GH values >400 ng/ml. No disruption in amplitude of periodicity of the GH rhythm was observed and there was no significant difference in mean 6-hr plasma GH levels. Plasma IRI levels fluctuated minimally over the 6-hr sampling period. There was no significant difference in mean 6-hr IRI levels between groups I and II, or between groups III, IV and V. These data do not support the view that endogenous opiate-like peptides play a physiologically important role in maintaining basal GH and IRI secretion.     

Circadian rhythm parameters of endocrine functions in elderly subjects during the seventh to the ninth decade of life

The circadian rhythm of 17 endocrine parameters (ACTH, aldosterone, cortisol, C-peptide, DHEA-S, FSH, growth hormone, insulin, LH, 17-OH progesterone, prolactin, testosterone, total T3, total T4 and TSH and estradiol and progesterone in women only) were studied in 63 clinically apparently healthy men (124 profiles) and 86 women (154 profiles) during the 7th to 9th decade of life. The subjects lived under very uniform conditions in a home for the aged with their daily schedule standardized by institutional routine with rest at night on the average from 21:30 to 06:30 local time and 3 daily meals at 08:30, 13:00 and 18:30. Blood was drawn over a 24-h span at 4-h intervals. Circadian periodicity was ascertained and the rhythm parameters quantified by cosinor analysis. In clinically healthy elderly subjects, circadian periodicity persisted in most parameters studied well into the 9th decade of life. The timing of the circadian rhythm was comparable between subjects in their 7th decade and 9th decade of life with the exception of cortisol and DHEA-S, which showed a phase advance with advancing age. A decrease in circadian amplitude is limited during this part of the human life span to only a few of the functions investigated and with the exception of prolactin in the women, a decrease in amplitude did accompany a decrease in MESOR.     

Circadian rhythms of electrolyte and 17-hydroxycorticosteroid excretion in hyperthyroidism and hypothyroidism.

The circadian rhythms of excretion of sodium, potassium, calcium, magnesium, phosphorus and 17-hydroxycorticosteroid (17-ohcs) were determined in five normal subjects, in six patients with hyperthyroidism and five with hypothyroidism. Constant diets with identical 3-hourly feedings were employed, and urine collections were made every 3 hrs during a 3-day study period. The circadian patterns of urinary excretion of sodium, potassium and 17-OHCS were similar in all three groups with distinct daytime peaks and nighttime nadirs. The total quantities of the ions and 17-OHCS excreted were greater in hyperthyroid than in hypothyroid patients with the greatest difference noted with the 17-OHCS. The rhythms for calcium, magnesium and phosphorus excretion were accentuated in hyperthyroid patients but similar to those in normal subjects with early morning calcium and magnesium peaks and a phosphorus peak approximately 12 hrs later. While a similar although blunted circadian pattern for calcium and perhaps magnesium excretion was noted in hypothyroid patients, their phosphorus rhythms were distorted and rather flat. These latter results confirm the observation of MINTZ et al. and are compatible with their interpretation that thyroid hormone is permissibly necessary for the expression of a normal phosphaturic rhythm and that the circulating level of thyroid hormone influences the amplitude of the phosphaturic rhythm.     

Interaction of the circadian rhythms of the functional activity of the hypophyseal-adrenal system and of the mitotic activity in the esophageal epithelium of rats with experimental hypo- and hypercorticism

In experimental hypercorticism the average daily concentration of glucocorticoids in blood plasma is 4 times higher than in control one, the amplitude of the circadian rhythm increases almost 3 times, the time of maximum concentration does not change. The adrenocorticotrophic hormone (ACTH) average circadian concentration decreases twofold. The adrenalectomy results in abrupt smoothing down of the circadian rhythm of glucocorticoids concentration and in increasing (in 2,4 times) of the average circadian concentration of ACTH. The pattern of circadian rhythm remains, however the acrophase of hormone secretion is shifted. The circadian rhythm of cells devision in esophagus epithelium changes. The maximum of mitotic index is absent in adrenalectomized rats and the amplitude of its circadian variations decreases. In experimental hypercorticism the biphase rhythm of mitoses in induced.     

Daily profiles of plasma prolactin (PRL), growth hormone (GH), insulin-like growth factor-1 (IGF-1), luteinizing hormone (LH), testosterone, and melatonin, and of pituitary PRL mRNA and GH mRNA in male Long Evans rats in acute phase of adjuvant arthritis

We studied the effects of adjuvant arthritis (AA) on the endocrine circadian rhythms of plasma prolactin (PRL), growth hormone (GH), insulin-like growth factor-1 (IGF-1), luteinizing hormone (LH), testosterone, and melatonin and of pituitary PRL and GH mRNA in male Long Evans rats. Groups of control and AA rats (studied 23 days after AA induction) that were housed under a 12/12 h light/dark cycle (light on at 06:00 h) were killed at 4 h intervals starting at 14:00 h. Cosinor analysis revealed a significant 12 h rhythm in PRL and PRL mRNA (p < 0.001) in controls with peaks at 14:00 h and 02:00 h, respectively. The peak at 02:00 h was abolished in the AA group resulting in a significant 24 h rhythm in parallel with that of PRL (p < 0.05) and PRL mRNA (p < 0.0001). Growth hormone showed no rhythm, but a significant rhythm of GH mRNA was present in both groups (p < 0.0001). Insulin-like growth factor-1 showed a 24 h rhythm in control but not in AA rats. The mean values of GH, GH mRNA, and IGF-1 were significantly reduced in AA. Luteinizing hormone displayed a significant 24 h rhythm (p < 0.01) peaking in the dark period in the control but not AA group. Testosterone showed in phase temporal changes of LH levels with AA abolishing the 02:00 h peak. Melatonin exhibited a significant 24 h rhythm in control (p < 0.001) and AA (p < 0.01) rats with maximum levels during the dark phase; the mesor value was higher in the AA males. These results demonstrate that AA interferes with the rhythms of all the studied hormones except the non-24 h (arrhythmic) GH secretion pattern and the rhythm in melatonin. The persistence of a distinct melatonin rhythm in AA suggests the observed disturbances of hormonal rhythms in this condition do not occur at the level of the pineal gland.     

Circadian variations and extraadrenal effect of ACTH on insulinemia in rabbit.

The present experiment was designed to study the action of ACTH1-24 on insulin secretion during the circadian cycle in normal rabbits and to provide evidence that ACTH1-24 has an extra-adrenal effect on this secretion. In normal rabbits intravenous administration of three doses of ACTH1-24 (1, 10, 100 micrograms/kg) at 10 a. m. increased plasma insulin levels. Hyperglycemia only occurred with doses of 10 and 100 micrograms/kg. A maximum insulin response was already obtained at 1 micrograms/kg. The same experiment performed at 12 p. m. also induced hyperinsulinemia which was only noted at 10 and 100 micrograms/kg; hyperglycemia was only observed after stimulation by the highest dose (100 micrograms/kg). ACTH was therefore more effective during the day; however, at 12 p. m. plasma insulin levels were the highest, but only with the maximum dose of ACTH (100 micrograms/kg). The effect of ACTH1-24 was evaluated throughout the day on normal and adrenalectomized rabbits. In normal animals injection of ACTH1-24 increased plasma glucose and insulin levels both together. In the contrary, in rabbits deprived of adrenal glands, ACTH1-24 induced high insulinemia along with hypoglycemia. We could, therefore, reasonably conclude that ACTH stimulates directly the pancreatic secretion of insulin.     

Synthetic adrenocorticotropin for optimizing murine circadian chronotolerance for adriamycin

An attempt to pre-set the circadian rhythm in murine chronotolerance for adriamycin (ADR) given i.p. or i.v. with ACTH was performed in three studies. In CDF1 mice standardized in LD12:12, it was demonstrated that 1) the circadian rhythm in murine chronotolerance for ADR exhibits a different timing depending upon whether the intravenous or intraperitoneal route is used for the administration of this anticancer agent; 2) ACTH or saline pretreatment does not enhance optimal circadian-stage-qualified ADR tolerance, whatever its route of injection, with any of the circadian stages and schedules explored; 3) near-optimal tolerance can be achieved by a fixed 'best' interval (among those investigated) between ACTH and ADR, irrespective of circadian stage. Tolerance equivalent to optimal circadian-stage-qualified ADR tolerance results from the administration of ACTH 1-17 (HOE433 = Synchrodyn) 24 hours before ADR injection; 4) and acrophase advance of over 6 hours of the tolerance rhythm results from ACTH 1-17 administration at 6 HALO. The acrophase changes do not directly account for an optimal ADR tolerance at a fixed interval of 24 hours after ACTH 1-17. Thus, ACTH may be considered a potential relative chronizer of murine chronotolerance for ADR.     

Effects of a kappa-opioid agonist on adrenocorticotropic and diuretic function in man

Although kappa-opiate receptors represent an important fraction of the total opiate receptor capacity in human brain their endocrine function is unknown. We determined the effects of a kappa-opiate receptor agonist on the secretion of vasopressin, ACTH and cortisol and on diuresis. The racemic benzomorphan kappa agonist MR 2033 or its opiate active (-)-isomer, MR 2034, inhibited the release of cortisol and ACTH in 12 trials in a naloxone reversible manner; plasma levels of vasopressin were not altered. The (+)-isomer, MR 2035, did not affect the secretion of cortisol or ACTH. Surprisingly, in five other subjects large increases were observed in vasopressin, ACTH and cortisol following the kappa-agonist, which were probably elicited indirectly by aversive effects of the opioid. The subjects in whom vasopressin release was not altered by MR 2033 and MR 2034 displayed large decreases in urine osmolality which were not antagonized by naloxone. The opiate inactive (+)-isomer, MR 2035, caused no diuretic response. Subjects in whom vasopressin release was stimulated did not show decreases in urine osmolality indicating that vasopressin is capable of antagonizing the diuretic action of the kappa-agonist. Our data show that a kappa-agonist inhibits secretion of cortisol and ACTH by acting at stereospecific opiate receptors and elicits diuresis by acting at stereospecific, but naloxone-insensitive non-classical opioid receptors. These data support the concept that different types of kappa-receptors can be distinguished in man.