J. Endocrinol. Invest. 15: 741-747,1992
Nocturnal prolactin pulses in relation to luteinizing hormone and thyrotropin J. Saini, C. Simon, G. Brandenberger, G. Wittersheim, and M. Follenius Laboratoire de Physiologie et de Psychologie Environnementales UMR 32, CNRS/INRS Strasbourg Cedex, France plasma and secretory hormone pulses were identified by a peak detection computer program. For statistical analysis the night studies of each subject were concatenated. Concomitance between the plasma pulses of both TSH and LH with PRL was insufficient to reject the null hypothesis of random coincidence. An increase in the number of subjects demonstrating significant coincidence between the hormone pulses was obtained when secretory pulses were analysed. Seven of the 12 and 10 of the 12 subjects showed significant concomitance between PRL and respectively TSH and LH. This proportion was sufficient to confirm copulsatility between PRL and LH. These results suggest that LH regulatory mechanisms are involved in the generation of the nocturnal pulsatile PRL profile, TRH mayaiso play a role in the secretion of PRL at a centrallevel, but was not reflected in the plasma or secretory profiles because of other overriding regulatory factors.
ABSTRACT. The two hypothalamic releasing factors, luteinizing hormone releasing hormone (LHRH) and thyrotropin releasing hormone (TRH), have been shown to stimulate pituitary prolactin (PRL) release as weil as their respective pituitary hormones, luteinizing hormone (LH) and thyrotropin (TSH). In this study the influence of LH and TSH regulatory mechanisms on nocturnal PRL secretion was investigated by evaluating whether the coincidence of PRL with LH and TSH pulses occurred more frequently than would be expected if the hormone generators were not coupled. Thirty night studies were conducted in twelve healthy male subjects. Six subjects underwent 3 studies and 6 subjects 2 studies. Blood was collected into aliquots at 10 min intervals throughout the night and plasma concentrations of PRL, TSH, and LH were determined. From the plasma profiles, hormone secretory rates were calculated using a method of deconvolution. Significant INTRODUCTION
receiving a LHRH agonist (5, 8) and a number of studies have found an association between LH and PRL pulses at certain times during the menstrual cycle (9-12). This synchrony is also reported to occur after the administration of progesterone to estrogen-primed women (13). The influence of LH and TSH stimulatory pathways in generating the basal pulsatile PRL plasma profile has not been established. All of the three pituitary hormones (TSH, LH and PRL) show a nocturnal rise (14), although for TSH an inhibitory effect of sleep is superimposed on the general nocturnal trend (15). For LH the nocturnal increase is only slight (16) and in women is modulated by the menstrual cycle (17). Various studies have tried to associate the specific nocturnal plasma fluctuations to specific sleep stages but no direct relationship has been identified. Early reports related PRL to sleep cycles (18) but have since been criticized (19) and a relationship between LH and sleep stages has been described in pubertal boys (20) but not in adults (21). We have recently shown the presence of a temporal relationship between PRL, TSH, LH and sleep stages in that plas-
Prolactin (PRL) secretion is largely regulated by tonic hypothalamic inhibition and dopamine has been identified as one of the major inhibitory factors (1, 2). Two of the hypothalamic releasing factors, thyrotropin releasing hormone (TRH) and luteinizing hormone re leasing hormone (LHRH) have, however, been shown to stimulate PRL release as weil as their respective pituitary hormones thyrotropin (TSH) and luteinizing hormone (LH) (3-5). In vitra studies suggest that TRH has an identical mode of action on pituitary PRL and TSH cells (6) and TRH stimulation in viva provokes both TSH and PRL responses in hypothyroid patients which show similar time courses (7). Concerning LHRH, simultaneous pulsatile release of both LH and PRL has been demonstrated in normal and hypogo.nadal women
Key-words.· Prolactin, luteinizing hormone. thyrotropin, pulsatility. Correspondence. Dr. J. Saml, Laboratoire de Physiologie et de Psychologie Environnementales 21, rue Becquerel, 67087 Strasbourg Cedex France.
Received June 10, 1991; accepted June 11. 1992.
741
J. Saini, C. Simon, G. Brandenberger, et al.
Hormone secretion rate For each night study PRL, TSH and LH secretion rates during each 10 min interval were calculated using the respective plasma values which had been smoothed by the Robust Least Square Method of Cleveland (24). The secretion rates were derived by a mathematical analysis, termed "deconvolution" using either an open one-compartment (for PRL and TSH) or two compartment (for LH) model as has been described by Eaton et al. for C-peptide (25). The hormone half-lives, which were taken from the literature, were for PRL, 25 min, for TSH, 35 min and for LH, 18 min for the first component and 90 min for the second component, with a fractional amplitude of the second component of 0.37 (26) and the volume of distribution of each hormone was taken to be that of albumin. The term secretory profile is used to describe the pattern obtained from the consecutive secretion rates, calculated from the plasma values obtained at 10 min intervals.
ma oscillations of all three hormones are rarely in the ascending phase at the onset of rapid eye movement (REM) sleep (22). In light of the data suggesting that common regulatory mechanisms are involved in the stimulation of LH and PRL and of TSH and PRL, also the similar temporal relationship plasma PRL, TSH and LH fluctuations have with REM sleep onset, we have investigated the possible influence of LH and TSH regulatory mechanisms on nocturnal PRL profiles by examining the incidence of concomitant nocturnal plasma and secretory PRL pulses with respectively the plasma and secretory pulses of LH and TSH. MATERIALS AND METHODS Subjects and procedures
Twelve healthy male subjects aged between 20-29 years with a mean weight of 67±8 kg and height 1.79±0.03 m were recruited for the experiment after receiving their informed consent. They all had conventional life-styles and sleep-wake rhythms and were taking no medication. The experiments were carried out in sound-proof air-conditioned sleep chambers. Six of the subjects (1-6) underwent three night studies from 23:0008:00 and six (7-12) two night studies from 23:0007:00 giving a total of thirty night studies, in addition to a night of habituation. Lights were switched off at 23:00 and the subjects were woken either at 07:00 or 08:00. A catheter was inserted into an antecubital vein and blood was taken continuously throughout the night study using a peristaltic pump. Sampies were collected into EOTA (1 mg.ml- 1 blood) tubes containing aprotinin (500 kalikrein inhibitor units ml- 1 ) over 10 min periods. The sampies were centrifuged at 4 C and the plasma immediately stored at -25 C until analysis. Sieep was recorded throughout the night and scored according to established criteria (23).
Analysis of hormone pulses For both the plasma and secretory hormone profiles significant pulses were identified using an objective pulse detection program (ULTRA) at a threshold of 3 and 4 times the coefficient of variation for plasma concentrations and secretion rates respectively (27). The data given on each hormone pulse, including the time of peak levels and the length of the ascending phase was used to examine the relationship between the PRL and TSH and LH pulses. To assess the coincidence between PRL and TSH and PRL and LH pulses the first sampie point of each pulse detected was taken to represent the time of pulse occurrence (28). A PRL pulse was considered coincident to a LH or TSH pulse if the delay between the occurrence of the two pulses was less than 20 min (i.e. within 2 sampie intervals of one another). Initially Monte Carlo simulations were carried out using 100,000 simulations on each conditional pair, with the constraints that no pulse onset was located within the first and last sampie of the series and that no pulse onset occurred in immediately successive sampies (29). For a large number of se ries the probability distributions for the number of coincident peaks was calculated using the hypergeometrie probability density function as suggested by Veldhuis et al (30). The cumulative probability distributions obtained by means of these two methods were then compared using KolmogorovSmirnov tests. No statistical difference between the two methods was found; so all further calculations, were carried out using the hypergeometric method. For each subject the null hypothesis of
Hormone analysis PRL, TSH and LH were measured by radioimmunoassay using kits from Corning Medical (Le Vesinet, France) for PRL, INCSTAR Corporation (Stillwater USA) for TSH and from Biomerieux (Lyon, France) for LH. Intraassay coefficient of variation for duplicate sampies was for PRL 3% for levels 9 ng/ml, for TSH 6% for levels 1.6 uU/ml and for LH 4.4% for levels 8.4 mU/ml. The detection limit of the hormones was 1 ng/ml PRL, 0.05 uU/ml TSH and 0.25 mU/ml LH.
742
Concomitant PRL, LH and TSH pulses
random pulse coincidence was rejected whenever p
Nocturnal prolactin pulses in relation to luteinizing hormone and thyrotropin J. Saini, C. Simon, G. Brandenberger, G. Wittersheim, and M. Follenius Laboratoire de Physiologie et de Psychologie Environnementales UMR 32, CNRS/INRS Strasbourg Cedex, France plasma and secretory hormone pulses were identified by a peak detection computer program. For statistical analysis the night studies of each subject were concatenated. Concomitance between the plasma pulses of both TSH and LH with PRL was insufficient to reject the null hypothesis of random coincidence. An increase in the number of subjects demonstrating significant coincidence between the hormone pulses was obtained when secretory pulses were analysed. Seven of the 12 and 10 of the 12 subjects showed significant concomitance between PRL and respectively TSH and LH. This proportion was sufficient to confirm copulsatility between PRL and LH. These results suggest that LH regulatory mechanisms are involved in the generation of the nocturnal pulsatile PRL profile, TRH mayaiso play a role in the secretion of PRL at a centrallevel, but was not reflected in the plasma or secretory profiles because of other overriding regulatory factors.
ABSTRACT. The two hypothalamic releasing factors, luteinizing hormone releasing hormone (LHRH) and thyrotropin releasing hormone (TRH), have been shown to stimulate pituitary prolactin (PRL) release as weil as their respective pituitary hormones, luteinizing hormone (LH) and thyrotropin (TSH). In this study the influence of LH and TSH regulatory mechanisms on nocturnal PRL secretion was investigated by evaluating whether the coincidence of PRL with LH and TSH pulses occurred more frequently than would be expected if the hormone generators were not coupled. Thirty night studies were conducted in twelve healthy male subjects. Six subjects underwent 3 studies and 6 subjects 2 studies. Blood was collected into aliquots at 10 min intervals throughout the night and plasma concentrations of PRL, TSH, and LH were determined. From the plasma profiles, hormone secretory rates were calculated using a method of deconvolution. Significant INTRODUCTION
receiving a LHRH agonist (5, 8) and a number of studies have found an association between LH and PRL pulses at certain times during the menstrual cycle (9-12). This synchrony is also reported to occur after the administration of progesterone to estrogen-primed women (13). The influence of LH and TSH stimulatory pathways in generating the basal pulsatile PRL plasma profile has not been established. All of the three pituitary hormones (TSH, LH and PRL) show a nocturnal rise (14), although for TSH an inhibitory effect of sleep is superimposed on the general nocturnal trend (15). For LH the nocturnal increase is only slight (16) and in women is modulated by the menstrual cycle (17). Various studies have tried to associate the specific nocturnal plasma fluctuations to specific sleep stages but no direct relationship has been identified. Early reports related PRL to sleep cycles (18) but have since been criticized (19) and a relationship between LH and sleep stages has been described in pubertal boys (20) but not in adults (21). We have recently shown the presence of a temporal relationship between PRL, TSH, LH and sleep stages in that plas-
Prolactin (PRL) secretion is largely regulated by tonic hypothalamic inhibition and dopamine has been identified as one of the major inhibitory factors (1, 2). Two of the hypothalamic releasing factors, thyrotropin releasing hormone (TRH) and luteinizing hormone re leasing hormone (LHRH) have, however, been shown to stimulate PRL release as weil as their respective pituitary hormones thyrotropin (TSH) and luteinizing hormone (LH) (3-5). In vitra studies suggest that TRH has an identical mode of action on pituitary PRL and TSH cells (6) and TRH stimulation in viva provokes both TSH and PRL responses in hypothyroid patients which show similar time courses (7). Concerning LHRH, simultaneous pulsatile release of both LH and PRL has been demonstrated in normal and hypogo.nadal women
Key-words.· Prolactin, luteinizing hormone. thyrotropin, pulsatility. Correspondence. Dr. J. Saml, Laboratoire de Physiologie et de Psychologie Environnementales 21, rue Becquerel, 67087 Strasbourg Cedex France.
Received June 10, 1991; accepted June 11. 1992.
741
J. Saini, C. Simon, G. Brandenberger, et al.
Hormone secretion rate For each night study PRL, TSH and LH secretion rates during each 10 min interval were calculated using the respective plasma values which had been smoothed by the Robust Least Square Method of Cleveland (24). The secretion rates were derived by a mathematical analysis, termed "deconvolution" using either an open one-compartment (for PRL and TSH) or two compartment (for LH) model as has been described by Eaton et al. for C-peptide (25). The hormone half-lives, which were taken from the literature, were for PRL, 25 min, for TSH, 35 min and for LH, 18 min for the first component and 90 min for the second component, with a fractional amplitude of the second component of 0.37 (26) and the volume of distribution of each hormone was taken to be that of albumin. The term secretory profile is used to describe the pattern obtained from the consecutive secretion rates, calculated from the plasma values obtained at 10 min intervals.
ma oscillations of all three hormones are rarely in the ascending phase at the onset of rapid eye movement (REM) sleep (22). In light of the data suggesting that common regulatory mechanisms are involved in the stimulation of LH and PRL and of TSH and PRL, also the similar temporal relationship plasma PRL, TSH and LH fluctuations have with REM sleep onset, we have investigated the possible influence of LH and TSH regulatory mechanisms on nocturnal PRL profiles by examining the incidence of concomitant nocturnal plasma and secretory PRL pulses with respectively the plasma and secretory pulses of LH and TSH. MATERIALS AND METHODS Subjects and procedures
Twelve healthy male subjects aged between 20-29 years with a mean weight of 67±8 kg and height 1.79±0.03 m were recruited for the experiment after receiving their informed consent. They all had conventional life-styles and sleep-wake rhythms and were taking no medication. The experiments were carried out in sound-proof air-conditioned sleep chambers. Six of the subjects (1-6) underwent three night studies from 23:0008:00 and six (7-12) two night studies from 23:0007:00 giving a total of thirty night studies, in addition to a night of habituation. Lights were switched off at 23:00 and the subjects were woken either at 07:00 or 08:00. A catheter was inserted into an antecubital vein and blood was taken continuously throughout the night study using a peristaltic pump. Sampies were collected into EOTA (1 mg.ml- 1 blood) tubes containing aprotinin (500 kalikrein inhibitor units ml- 1 ) over 10 min periods. The sampies were centrifuged at 4 C and the plasma immediately stored at -25 C until analysis. Sieep was recorded throughout the night and scored according to established criteria (23).
Analysis of hormone pulses For both the plasma and secretory hormone profiles significant pulses were identified using an objective pulse detection program (ULTRA) at a threshold of 3 and 4 times the coefficient of variation for plasma concentrations and secretion rates respectively (27). The data given on each hormone pulse, including the time of peak levels and the length of the ascending phase was used to examine the relationship between the PRL and TSH and LH pulses. To assess the coincidence between PRL and TSH and PRL and LH pulses the first sampie point of each pulse detected was taken to represent the time of pulse occurrence (28). A PRL pulse was considered coincident to a LH or TSH pulse if the delay between the occurrence of the two pulses was less than 20 min (i.e. within 2 sampie intervals of one another). Initially Monte Carlo simulations were carried out using 100,000 simulations on each conditional pair, with the constraints that no pulse onset was located within the first and last sampie of the series and that no pulse onset occurred in immediately successive sampies (29). For a large number of se ries the probability distributions for the number of coincident peaks was calculated using the hypergeometrie probability density function as suggested by Veldhuis et al (30). The cumulative probability distributions obtained by means of these two methods were then compared using KolmogorovSmirnov tests. No statistical difference between the two methods was found; so all further calculations, were carried out using the hypergeometric method. For each subject the null hypothesis of
Hormone analysis PRL, TSH and LH were measured by radioimmunoassay using kits from Corning Medical (Le Vesinet, France) for PRL, INCSTAR Corporation (Stillwater USA) for TSH and from Biomerieux (Lyon, France) for LH. Intraassay coefficient of variation for duplicate sampies was for PRL 3% for levels 9 ng/ml, for TSH 6% for levels 1.6 uU/ml and for LH 4.4% for levels 8.4 mU/ml. The detection limit of the hormones was 1 ng/ml PRL, 0.05 uU/ml TSH and 0.25 mU/ml LH.
742
Concomitant PRL, LH and TSH pulses
random pulse coincidence was rejected whenever p
