SEROTONIN precursors, antagonists, and reuptake

Transcription

1 X/83/ O$02.0O/O Journal of Clinical Endocrinology and Metabolism Copyright 1983 by The Endocrine Society Vol. 56, No. 1 Printed in U.S.A. Transient Effect of L-5-Hydroxytryptophan on Pituitary Function in Men and Women* C. ANN MASHCHAK,f OSCAR A. KLETZKY, CAROLE SPENCER, AND RAUL ARTAL Departments of Obstetrics and Gynecology and Medicine, University of Southern California School of Medicine, Los Angeles, California ABSTRACT. To further elucidate the role of serotonin in the secretion of pituitary hormones, a loading dose of 0.8 mg/kg for 1 h, followed by a maintenance dose of 0.1 mg/kg-h of L-5- hydroxytryptophan (5-HTP) for 23 h were administered iv to five normal men and five normal women. Five additional men received repeated loading doses of 5-HTP 4 and 6 h after the initial one, with or without the maintenance dose. The initial studies demonstrated a significant, but transient, increase in plasma PRL, GH, cortisol (P < 0.01) and TSH (P < 0.05) in the five men and a consistent and significant transient increase only in PRL and cortisol in the five women. Plasma LH and FSH values were not affected by 5-HTP administration. The constant administration of 5-HTP revealed a blunting effect on the nocturnal rise of TSH in men. The continuous administration of 5- HTP failed to maintain the rise induced by the loading dose. Individuals receiving additional loading doses of 5-HTP demonstrated a subsequent increase in GH and cortisol, but not in serum PRL. This study suggests that endogenous serotonin may exert its stimulatory effect on pituitary hormone secretion primarily by sporadic release, rather than by continuous secretion. (J Clin Endocrinol Metab 56: 170, 1983) SEROTONIN precursors, antagonists, and reuptake blockers, rather than serotonin itself, have been used in human research because orally or iv administered serotonin does not cross the blood-brain barrier and produces excessive gastrointestinal side effects (1-4). Some studies suggest that acute elevation of serotonin stimulates the release of PRL, GH, and ACTH; however, the data have not been entirely consistent (2, 5-8). The evidence for the role of serotonin in TSH, LH, and FSH secretion is even less conclusive (5, 8-10). The purpose of this study was to further elucidate the role of serotonin in the control of pituitary hormone secretion by determining the immediate and long term effect of a 24-h infusion of L-5-hydroxytryptophan (5- HTP), the immediate precursor of serotonin (11). Frequent determinations of serum PRL, GH, TSH, FSH, LH, and cortisol (as an indirect measurement of ACTH) before, during, and after a 24-h infusion of 5-HTP were used to characterize the serotonin hypothalamic pituitary interaction. Received February 23, Address requests for reprints to: Oscar A. Kletzky, M.D., Women's Hospital, 1240 North Mission Road, Room L-1013, Los Angeles, California * This work was supported in part by USPHS Grant GCRC-RR-43. f Present address: Department of Obstetrics and Gynecology, Loma Linda University, Loma Linda, California Materials and Methods Ten men and five women (three in the early and two in the late follicular phase) voluntarily participated in the study. The men ranged from yr of age, and the women ranged from yr of age; their weights ranged from and from lb respectively (within 20% of ideal body weight). All were healthy subjects without a history of alcoholism or any drug intake. Five of the men and the five women were admitted to the Clinical Research Center at 0700 h for 3 consecutive days. They were allowed partial ambulation and received a low protein (40 g/day) low fat diet, with cal/kg BW. A continuous iv infusion of 0.9% sodium chloride (NaCl) was maintained at a rate of 60 ml/hour throughout the 72-h study to insure uniform hydration. To diminish the peripheral conversion of 5-HTP to serotonin and to reduce the side effects, oral carbidopa at a dose of 100 mg every 8 h was administered during the first 48 h. The 24-h 5-HTP infusion through the NaCl iv line was begun at 0900 h on the second day. 5-HTP was administered as a loading dose at a rate of 0.8 mg/kg over the first hour. For the remaining 23 h, the drug was administered at the maintenance rate of 0.1 mg/kg-h. On each of the 3 days, plasma samples were obtained at 15-min intervals for 2 h, at 0.5-h intervals for an additional 2 h, and every 2 h for the rest of the day. Three additional men received the same maintenance dose for only 8 h, with the loading dose administered initially and repeated 4 and 6 h later. The second and third iv doses of 5-HTP were given over 15 min at the same rate as the first dose. The remaining two received only the loading doses of 5-HTP without the maintenance dose. Samples were ob- 170

2 EFFECT OF 5-HTP ON PITUITARY FUNCTION IN MAN 171 tained from a separate indwelling scalp vein needle which was kept open with a heparin lock. Blood samples were immediately cooled on ice and centrifuged at 5 C within 1 h of collection, and plasma samples were stored at -20 C until assayed. Selected samples of seven subjects were collected in EGTA-GSH and stored at 70 C for the measurement of epinephrine and norepinephrine by radioenzymatic procedure (12). The interand intraassay coefficients of variation were below 10%, and the assay's sensitivity was 10 pg/ml. Urine was collected at 8-h intervals throughout the study for the measurement of the serotonin metabolite, 5-hydroxyindoleacetic acid (5-HIAA) (13). Plasma concentrations of PRL, GH, FSH, LH, TSH, and cortisol were measured by a previously reported RIA in all samples (14). The inter- and intraassay coefficients of variation for all of these assays were less than 15%. The assay's sensitivity were as follows: LH and FSH, 1.5 mlu/ml; PRL, 2 ng/ml; GH, 0.4 ng/ml; TSH, 0.5 juu/ml; and cortisol, 1 jug/dl. For statistical analysis, the paired or Student's t test was used, and the area under the curve was calculated by the trapezoidal rule (15). A P value of <0.05 was considered statistically significant. Results Two men and three women had transient nausea and vomiting; however, all subjects completed the study. The urinary excretion of 5-HIAA increased 2- to 3-fold during the day of the infusion, reaching a mean maximum of 11 ± 1 (±SEM) mg/8-h aliquot. The measurement in the first aliquot of the third day remained elevated (10 ± 2), returning slowly to baseline levels at the end of the study. The PRL response to 5-HTP in five normal men is shown in the upper panel of Fig. 1. The mean (±SEM) baseline plasma PRL level at 0900 h on the admitting day was 5.8 ± 2.1 ng/ml. The same pattern with similar levels was seen on the second and third days. In response to the loading dose of 5-HTP, a significant increase in plasma PRL levels occurred in all five men. The PRL increase centered around the peak is demonstrated in the inset of Fig. 1. The mean (±SEM) peak PRL (23.4 ± 3.2 ng/ml) occurred 80 ± 10 min after the beginning of the infusion. In spite of the maintenance infusion, PRL levels returned rapidly (75 min) to baseline. From this point on, the constant infusion of 5-HTP had no detectable effect on the levels of PRL, including its nocturnal rise. In addition, plasma PRL concentrations during the day after the infusion were similar to those measured during the first day. The PRL response to 5-HTP in the three women in the early follicular phase was similar to that of the two in the late follicular phase; therefore, data of all five were combined. The women showed the same PRL pattern as the men, including a significant response to the loading dose of 5-HTP (Fig. 1, lower panel). Thereafter (75 ± 5 min), the PRL concentrations returned to pretreatment levels. The mean peak PRL response for the women was higher than but not significantly different from that for the men (38.1 ± 9.8 vs ± 3.2 ng/ml). A significant difference in the magnitude of the nocturnally related PRL rise was detected between men and women on all 3 days (P < 0.02). A transient acute response of GH to the loading dose of 5-HTP, which was significantly different from baseline (P < 0.05), was seen in all five men, but was found in only two of the women. Baseline plasma GH levels between 0700 and 0900 h on days 1, 2, and 3 were not significantly different from each other. The centered mean peak plasma GH level of 19.8 ±7.7 ng/ml occurred 81 ± 9 min after the beginning of the infusion and returned to baseline 4 h later. No significant differences were seen in the nocturnally related GH rise between men and women. A cortisol response to 5-HTP administration which was significantly different (P < 0.05) from baseline was seen in all males and females (Fig. 2). The response of cortisol followed a pattern similar to those of PRL and GH. The baseline plasma cortisol levels at 0900 h were similar throughout the 3 days. The mean peak plasma cortisol level in men (23.3 ± 3.7 jug/dl) and that in women (19.2 ± 5.2 jug/dl) were not significantly different. The mean peak occurred 90 ± 15 min after the beginning of the 5-HTP infusion. No significant difference in the diurnal variation in plasma cortisol was seen between men and women. Daily baseline plasma TSH levels at 0900 h were not significantly different between men and women. The five men showed a small but significant increase in mean TSH levels after the onset of the infusion (P < 0.05; Fig. 3). The transient increase in TSH occurred 72 ± 15 min after the onset of the loading dose and was significantly higher (P < 0.05) than the corresponding values in the five women. Although a similar pattern appeared to be present in the women, none of the values was significantly different from baseline. TSH was the only pituitary hormone that showed a significant suppression of the nocturnally related pattern. This suppressive effect was seen only in the male subjects (Fig. 4). The nocturnal increment in TSH was significantly lower during the study day whether the rise was calculated as the peak increase or the area under the curve. The nocturnal rise of TSH was significantly higher in men than in women only during day 3 (P < 0.05). None of the men or women showed any significant change in plasma levels of LH and FSH in response to acute or chronic administration of 5-HTP. Plasma concentrations of epinephrine and norepinephrine after the infusion of 5-HTP were not significantly different from baseline values (Table 1). Serum glucose levels were determined in three male subjects during the initial 3 h of the 5-HTP infusion to determine if the acute hormone responses may have been associated with induced hypoglycemia. No significant

3 172 MASHCHAK ET AL. JCE & M 1983 Vol 56«No! Lr '0.8 mg/ kg/hr 5-HTP O.lmg/kg/hr J*""-*B * * *{ 0700HRS MINUTES I500HRS FIG. 1. Plasma PRL response to loading and maintenance iv infusions of 5-HTP in five normal men (upper panel) and five normal women (lower panel). Values are the mean ± SEM. The insets represent the mean PRL responses centered around the peak increases change in the plasma glucose concentration was seen in these patients. Plasma PRL, GH, and cortisol levels in the three subjects receiving the three loading doses as well as the maintenance dose are shown in Fig. 5. Plasma PRL did increase to 18.8 ± 4.5 ng/ml after the initial dose of 5- HTP, but there was no response to the following two doses of 5-HTP. In contrast, plasma levels of GH and cortisol not only showed a significant initial response, but increased significantly after the third dose of 5-HTP (P < 0.05). The two subjects who received no maintenance dose of 5-HTP (Fig. 6) had a single PRL response, while their GH and cortisol responses were variable. One subject had a GH response to each of the three doses of 5-HTP and a cortisol response to the first and third doses. The other subject showed a GH rise after the first two doses and a cortisol increase in response to all three doses of 5-HTP. Discussion To clarify the role of serotonin in the control of pituitary hormones, its immediate precursor, 5-HTP, was administered iv in this study. That compound has been

4 EFFECT OF 5-HTP ON PITUITARY FUNCTION IN MAN 173 5MEN 5.0 = 4.0 I I 30 -DAY I- -QAY2 DAY 3 N'5MEN (MEAN! SEM CSS TOTAL AREA p< x TO l0 i TO 0900 FIG. 2. Plasma cortisol response to loading and maintenance iv infusions of 5-HTP in five normal men {upper panel) and five normal women (lower panel). Values are the mean ± SEM. 5MEN FIG. 4. Blunted nocturnal rise of plasma TSH during continuous iv administration of 5-HTP (day 2) compared to TSH levels on control days 1 and 3 in five normal men. Data are centered around the peak increase (0 h), with samples obtained 6 h before ( ) and after (+) the peak. TABLE 1. Effect of loading iv infusion of 5-HTP on norepinephrine and epinephrine plasma levels in seven normal men and women FIG. 3. Plasma TSH response to loading and maintenance iv infusions of 5-HTP in five normal men (upper panel) and five normal women (lowerpanel). Values are the mean ± SEM. Asterisk indicates P < shown in animals to cross the blood-brain barrier and to be converted by L-aromatic amino acid decarboxylase, increasing the brain concentration of serotonin (16). Since the same enzyme converts L-dopa to dopamine, it is possible that the administration of 5-HTP may alter the secretion of dopamine. Animal studies have demonstrated catecholamine displacement from neuronal tissue at times the dose per kg used in the study (17, 18). In this study, the dose of 5-HTP was selected following the pharmacokinetic data of Magnussen and Nielsen- Kudsk (19). They reported that adverse gastrointestinal side effects occurred consistently when plasma levels of 5-HTP reached a concentration of 6 /xmol/liter. Considering the 6-h biological half-life, it was calculated that a Time (min) Norepinephrine (pg/ ml) 465 ± ± ± ± ± ± ± ± ± ± 99 Values given are the mean ± SEM. Epinephrine (pg/ml) 53 ±7 67 ± ±22 55 ±12 53 ± ±30 94 ± ± ± ± 24 loading dose of 0.8 mg/kg in 1 h, followed by a constant infusion of 0.1 mg/kg «h, would achieve and maintain a constant concentration of about 4 /xmol/liter (19). The transient gastrointestinal side effects experienced by some of the subjects suggest that the plasma concentration of 5-HTP achieved in the study was near 6 /rniol/ liter. A low protein diet was chosen to maximize the entry of 5-HTP into the brain by reducing the concentration of competing neutral amino acids (4, 20). Although the plasma concentration of serotonin was not determined, it can be assumed that 5-HTP was converted to serotonin by L-aminodicarboxylase based on the significant increase in the urinary concentration of 5-HIAA, which is a serotonin metabolite. The sustained elevation of 5-HIAA after the discontinuation of 5-HTP infusion in the first 8-h aliquot of the third day reflects the long half-life (6 h) of 5-HTP. The levels of 5-HIAA diminished thereafter, reaching baseline values by the end of the

5 174 MASHCHAK ET AL. JCE & M ' 19B3 Vol 56 < No 1 FIG. 5. Plasma PRL, GH, and cortisol responses to three successive doses of 5-HTP (0.8 mg/kg-h, iv) given over 60 min at 0900 h, and over 15 min at 1300 and 1500 h, respectively (indicated by arrows), and to the maintenance dose (0.1 mg/kg-h, iv) of 5-HTP in three normal men (mean ± SEM). study. The peripheral decarboxylase inhibitor carbidopa was administered to diminish peripheral conversion to serotonin, thereby decreasing the side effects and increasing the concentration of 5-HTP available to cross the blood-brain barrier (21). Furthermore, carbidopa did not affect the plasma concentration of any of the studied hormones, in contradistinction to the report of Brown et al. (22) and Bansal et al. (20) of a carbidopa-related increase in plasma PRL levels. The role of stress in the release of PRL, GH, and cortisol cannot be excluded. However, stress is considered to have played a negligible role in these results, since patients with and without side effects demonstrated the same hormone response. Also, plasma concentrations of epinephrine and norepinephrine did not change significantly from baseline values. Therefore, it can be postulated that the hormonal changes observed in this study are the result of a hypothalamic site of action of serotonin; however, the possibility of a direct action on the pituitary cannot be eliminated. The iv administration of the loading dose of 0.8 mg/h 5-HTP induced an acute and significant response in plasma concentrations of PRL, GH, TSH, and cortisol in all five men initially studied. Women had a consistent FIG. 6. Plasma PRL, GH, and cortisol responses to three successive doses of 5-HTP (0.8/kg-h, iv) given over 60 min at 0900 h and over 15 min at 1300 and 1500 h, respectively (indicated by arrows), in two normal men (one designated and the other O). and significant response only in plasma levels of PRL and cortisol, with an inconsistent GH response. These results are in agreement with those of other reports, supporting the concept that serotonin is an important neurotransmitter controlling the secretion of PRL and GH (7, 8). The cortisol results are consistent with those reported by Imura et al. (7) using 5-HTP, but contrary to those of Woolf and Lee (8) using L-tryptophan. The fact that women had a different TSH response and an inconsistent GH response demonstrates the complexity of the neurotransmitter mechanism that may be affected or modulated by the estrogen or androgen milieu. The possibility that the PRL, GH, and cortisol increases were due to hypoglycemia induced by serotonin was eliminated because hypoglycemia did not occur in the three men studied. From this study, it can be concluded that serotonin is not a neurotransmitter controlling the secretion of gonadotropin, since neither the men nor the women demonstrated any significant change in plasma concentrations of LH and FSH in response to the loading or maintenance administration of 5-HTP (8, 9). The chronic administration of 5-HTP had no stimulatory effect on any of the hormones studied. Only the nocturnal rise of TSH was significantly blunted in the five men. This blunted nocturnal TSH was not accompanied by any discernible change in the nocturnal rise of

6 EFFECT OF 5-HTP ON PITUITARY FUNCTION IN MAN 175 any other hormone, including plasma cortisol (23-25). Since dopamine has been demonstrated to have a suppressive effect on the secretion of TSH, it is possible that the chronic administration of 5-HTP, centrally converted to serotonin, had interacted with the catecholamine pathway, resulting in an initial release of dopamine with subsequent inhibition of the nocturnal rise of TSH. After this initial increase, dopamine synthesis may have been impaired because of the neuronal utilization of the enzyme aromatic acid decarboxylase for the synthesis of serotonin instead of dopamine (17). It is also possible that in men, an increase in T3 resulted in a compensatory decrease in plasma TSH (26). The cause(s) of the lower nocturnal TSH increase observed in women is difficult to ascertain, and it is possible that if a larger number of women are studied, a significant change could be demonstrated. Such a sex difference was also seen in the GH response to acute administration of 5-HTP. Only two women had an increase in the plasma GH concentration, which peaked at 81 min, similar to the response seen in the five men. Again, the reason for this sex difference is unclear. The question of why the continuous infusion of 5-HTP did not elicit a constant or prolonged hormonal stimulation is intriguing. It is possible that the achieved concentration of serotonin was insufficient to induce a further stimulation or that the serotonin receptors became refractory to a constant administration of 5-HTP. To investigate these possibilities, five men received additional loading doses of 5-HTP with and without the administration of a maintenance dose. From the dissociation of PRL, GH, and cortisol results, it seems clear that the original protocol provided sufficiently high levels of serotonin in the brain. The change in the availability of 5- HTP per se did not result in a further increase in PRL, while GH and cortisol responded again after serum levels returned to baseline. It is possible that the persistent exposure instead of the interrupted administration of 5- HTP could have caused a change in the affinity or density of the receptors, leading to the absence of hormonal response (26, 27). Thus, this receptor sensitivity to 5-HTP may depend as much on the time of administration as on the brain concentration of serotonin, as demonstrated by the subjects who responded to the second and third doses of 5-HTP. Dichotomy of responses between PRL, GH, and cortisol may be indicative of a different control mechanism of serotonin for these hormones. It has been reported that two populations of serotonin receptors, 5-HTi and 5-HT2, are present in different areas of rat, guinea pig, and bovine brain (28). Therefore, it is possible that the receptors of the campgenerating system responsible for the PRL stimulation remained down-regulated for a longer period of time and were of a different type than the receptors for GH and cortisol stimulation. If this is the case, doses of 5-HTP given at less frequent intervals than those used in this study will result in repetitive PRL responses. Acknowledgments We thank the NIH (Bethesda, MD) for the generous gift of the reagents used in this study. We thank Ms. Louise Hayes for typing the manuscript and Ms. L. Goedert, Mrs. Veryl Moreth, Ms. G. Agilore, and Ms. Y. Cervantes for their technical assistance. References 1. Wirz-Justice A, Puhringer W, Lacoste V, Craw P, Gaspar M 1976 Intravenous L-5-hydroxytryptophan in normal subjects. III. Neuroendocrinological and biochemical changes. Pharmakopsychiatry 9: Ferrari C, Caldara R, Romussi M, Rampini P, Telloli P, Zaatar S, Curtarelli G 1978 Prolactin suppression by serotonin antagonists in man: further evidence for serotoninergic control of prolactin secretion. Neuroendocrinology 25: Masala A, Delitala G, Devilla L, Alagna S, Rovasio P 1979 Enhancement of insulin-induced prolactin secretion by fluoxetine in man. J Clin Endocrinol Metab 49: Schanberg SM 1963 A study of the transport of 5-hydroxytryptophan and 5-hydroxytryptamine (serotonin) into the brain. J Pharmacol Exp Ther 139: Maclndoe JH, Turkington RW 1973 Stimulation of human prolactin secretion by intravenous infusion of L-tryptophan. J Clin Invest 52: Kato Y, Nakai Y, Imura H, Chihara K, Ohgo S 1974 Effect of 5- hydroxytryptophan (5-HTP) on plasma prolactin levels in man. J Clin Endocrinol Metab 38: Imura H, Nakai Y, Yoshimi T 1973 Effect of 5-hydroxytryptophan (5-HTP) on growth hormone and ACTH release in man. J Clin Endocrinol Metab 36: Woolf PD, Lee L 1977 Effect of the serotonin precursor, tryptophan on pituitary hormone secretion. J Clin Endocrinol Metab 45: Kapen S, Vagenakis A, Braverman L 1980 Failure of a serotonergic receptor-blocking drug to change the 24-hour LH secretory pattern in women. J Clin Endocrinol Metab 51: Yoshimura M, Ochi Y, Miyazaki T, Shiomi K, Hachiya T 1973 Effect of L-5-HTP on the release of growth hormone, TSH and insulin. Endocrinol Jpn 20: Wurtman RJ, Fernstrom JD 1972 L-Tryptophan, L-tyrosine and the control of brain monoamine biosynthesis. In: Snyder SH (ed) Perspectives in Neuropharmacology. Oxford University Press, New York, p Artal RS, Glatz TH, Lam R, Nathanielsz PW, Hobel CJ 1979 The effect of acute maternal hemorrhage on the release of catecholamines in the pregnant ewe and the fetus. Am J Obstet Gynecol 135: Goldenberg H 1973 Specific photometric determination of 5-HIAA in urine. Clin Chem 19: Kletzky OA, Davajan V, Mishell Jr DR, Nicoloff JT, Mims R, March CM, Nakamura RM 1977 A sequential pituitary stimulation test in normal subjects and in patients with amenorrhea galactorrhea with pituitary tumors. J Clin Endocrinol Metab 45: Franklin P 1944 Methods of Advanced Calculus. McGraw-Hill, New York, p Davidson J, Sjoerdsma A, Loomis LN, Udenfriend S 1957 Studies with the serotonin precursor, 5-hydroxytryptophan, in experimental animals and man. J Clin Invest 36: Ng LKY, Chase TN, Colburn RW, Kopin IJ 1972 Release of [ 3 H] dopamine by L-5-hydroxytryptophan. Brain Res 45: Butcher LL, Engel J, Fuxe K 1972 Behavioral biochemical and histochemical analysis of the central effects of monoamine precursors after peripheral decarboxylase inhibition. Brain Res 41: Magnussen I, Nielsen-Kudsk F 1979 Pharmacokinetics of intravenously administered L-5-hydroxytryptophan in man. Acta Phar-

7 176 MASHCHAK ET AL. JCE & M 1983 Vol 56 No 1 macol Toxicol 44: Bansal S, Lee LA, Woolf PD 1981 Dopamine regulation of GH secretion in normal man: correlation of L-DOPA and dopamine levels with the GH response. J Clin Endocrinol Metab 53: Magnussen I, Engbak F 1978 The effects of aromatic acid decarboxylase inhibitors on plasma concentration of 5-hydroxytryptophan in man. Acta Pharmacol Toxicol 43: Brown GM, Garfinkel PE, Warsh JJ, Stancer HC 1976 Effect of carbidopa on prolactin, growth hormone and cortisol secretion in man. J Clin Endocrinol Metab 43: Fernstrom JD, Wurtman RJ 1974 Control of brain serotonin levels by the diet. Adv Biochem Psychopharmacol 11: Chan V, Jones A, Liendo P, McNeilly A, Landon J, Besser GM 1978 The relationship between circadian variations in circulating thyrotropin, thyroid hormones and prolactin. Clin Endocrinol (Oxf) 9: Nicoloff JT, Fisher DA, Appleman Jr MD 1970 The role of glucocorticoids in the regulation of thyroid function in man. J Clin Invest 49: Patel YC, Baker HWG, Burger HG, Johns MW, Lediner JE 1974 Suppression of the thyrotropin circadian rhythm by glucocorticoids. J Endocrinol 62: Morley JE, Raleigh MJ, Brammer GL, Yuwiler A, Geller E, Flannery J, Hershman JM 1980 Serotonergic and catecholaminergic influence on thyroid function in the vervet monkey. Eur J Pharmacol 67: Sulser F 1979 New perspectives on the mode of action of antidepressant drugs. Trends Pharmacol Sci 1:92