Possible influence of melatonin on the human prolactin circadian secretory pattern

In humans, prolactin (PRL) shows a circadian secretory pattern characterized by a sleep-dependent rising plasma level. In particular, PRL rises are related to non-rapid eye movement sleep (NREMS), even though there is a progressive rise during the night with the highest plasma levels in the second half of nocturnal sleep just when NREMS is less prevalent. In previous work we found that PRL rises were linked to slow wave sleep (SWS). Furthermore, we observed a statistically significant PRL circadian secretory pattern (i.e., the presence of a circadian rhythm) only in "good sleepers" (i.e., subjects having an "effìcient" sleep). In such subjects, the most commonly used "sleep efficiency index" (i.e., the ratio: (SWS + REMS)/(total sleep time); normal range, 0.35-0.60) showed values greater than 0.35. On the contrary, "poor sleepers", with an efficiency index inferior to 0.35, did not demonstrate the presence of a PRL rhythm. A circadian secretory pattern has also been found for melatonin (MT) , to date, the most investigated substance produced by the pineal gland. Even though there are some discordant data, in the case of a normal recurrence of light-dark and wake-sleep cycles, a MT plasma level increase would usually be detectable in the first half of nocturnal sleep and therefore would occur before the PRL rise. The aim of this study was to evaluate whether MT is involved in the regulation of the PRL circadian secretory rhythm. Subjects and methods: Eight healthy volunteers, considered as controls (five males and three females; mean age, 34 years; range 20-41 years), and eight patients with neuroradiological evidence of pineal tumour (six males and two females; mean age, 33 years; range, 18- 38 years) were examined. The latter were collected in a time interval of 5 years. In such patients, the histological diagnosis was one of a well-differentiated, low-grade fibrillary astrocytoma, noninvasive and, in five cases, sharply circumscribed (benign astrocytomas). All the patients presented no neurological or psychopathological symptoms during the testing period. After two adaptation nights, each subject underwent an all-night polysomnogram (EEG, EOG, EMG) from 22:00 to 07:00 with the lights off. Sleep was scored according to standard criteria. Blood samples were drawn every 15 min. In all subjects PRL and MT plasma levels were measured by a homologous double antibody radioimmunoassay (RIA). A week before and during the 3 days spent in the sleep laboratory, the duration of light-dark and wake-sleep cycles, physical activity, and diet were standardized by a pre-established schedule. All the subjects underwent testing either in spring or autumn and the women during the follicular phase of their menstrual cycle. PRL and MT chronobiological characteristics were determined by the single and population mean-cosinor methods: i.e., the 24 h cosine function of a continuous curve best fitted by least-squares analysis in the case of an individual or group, respectively. Such methods provide the possible statistical significance of a circadian secretory curve: i.e., the presence of a rhythm according to the Fvalue for zero amplitude test. Both the individual and group circadian curves are characterized by the following parameters: (a) mesor - 24 h average obtained by the best fitting cosine function used to approximate the circadian curve; (b) amplitude - difference between the maximum value measured at acrophase and the mesor in the cosine curve; (c) acrophase - crest timing of the best fitted curve. Results: All healthy volunteers showed PRL and MT circadian secretory rhythms. The presence of a PRL circadian rhythm is likely due to the presence of efficient sleep. On the contrary, in all the patients with a pineal tumour MT circadian plasma levels were inferior to the RIA normal minimum value and therefore the cosinor methods failed to show significant MT curves. In addition, such patients did not exhibit a PRL circadian rhythm even though their sleep efficiency index was within the normal range. Furthermore, they allow the detection ofa possible PRL circadian rhythm. Our data demonstrate the need for a polygraphic study of sleep to show the presence of a PRL circadian rhythm that in fact is detectable only in subjects having efficient sleep (sleep efficiency index greater than 0.35). Furthermore, the results concerning patients with pineal tumours permit one to nypothesize that, among possible circadian systems (regulating PRL rhythm, an important role is probably played by the pineal gland through MT. Since the PRL rhythm is sleep-dependent, it seems reasonable to infer that the nocturnal entrainment process is mediated by the pineal gland. Of course, an action by other substances (together with or independent from MT) produced by the pineal gland cannot be excIuded. Such a hypothesis awaits further examination by future research. In order to explain the possible influence of MT on PRL circadian rhythm we could suggest that in humans the PRL acrophase while it has been pointed out that in some animals pinealectomy abolishes the PRL circadian rhythm and MT enhances the PRL nocturnal secretion. On the other hand, in addition to normal PRL plasma levels, the group of patients with pineal tumors presented a normal behavioral pattern as far as the recurrence of the wake-sleep cycle and the organization of sleep cycles were concerned. Such results suggest that MT does not influence the hypothalamic- pituitary mechanisms involved in overall PRL circadian secretion and, according to others, either the wake - sleep cycle or the systems controlling the quantity and quality of sleep.