J. Endocrinol. Invest. 15: 525-531, 1992

Effects of opioid receptor blockade on luteinizing hormone (LH) pulses and interpulse LH concentrations in normal women during the early phase of the menstrual cycle 1 W.S. Evans, J.Y. Weltman, M.L. Johnson, A. Weltman, J.D. Veldhuis, and A.D. Rogol Departments of Medicine (WSE, MLJ, AW, JDV), Pediatrics (JYW, ADR) and Pharmacology (MLJ, ADR) and the Interdisciplinary Graduate Biophysics Program (MLJ, JDV), Health Sciences Center; Department of Human Services (ALW), Curry School of Education; and National Science Foundation Science and Technology Center for Biological Timing (WSE, MLJ, JDV, ADR), University of Virginia, Charlottesville, Virginia, USA. terpulse valley mean serum LH concentration (lUlL; 6.3±0.4 vs 5.0±0.4; p=0.0013). No difference was noted in the mean incremental LH pulse amplitude (lUlL; 1.9±0.1 vs 2.1±0.1; p=0.13), or peak duration (min; 40±1.8 vs 45.0±2.4; p=0.06). Mean LH peak area (IU/Umin) was greater on the control (45.0±2.4) vs naltrexone (40±1.8) days (p=0.0475). These results suggest that inhibition of endogenous opioid peptide systems with naltrexone in the early follicular phase of the menstrual cycle is associated with an alteration of the gonadotropin-releasing hormone (GnRH) pulse generator firing rate as appraised indirectly by enumeration of distinct LH pulses. The enhanced maximal serum LH pulse amplitude and interpulse valley mean LH concentrations, together with no observed change in the incremental amplitude, suggest a naltrexone-associated augmentation of low levels of basal LH release, a skewing of LH secretory burst waveform, andlor a prolongation of LH half-life.

ABSTRACT. To determine the role of endogenous opioid peptides in regulating pulsatile luteinizing hormone (LH) release in the early follicular phase of the menstrual cycle of eumenorrheic women, we evaluated serum LH concentrations in blood collected every 10 min for 12 h in 27 women each studied during two menstrual cycles: (1) without pretreatment and (2) following oral administration of naltrexone, a mu opiate receptor blocking agent, at a dose of 1.0 mg/kg. Pulsatile LH release was assessed by the CLUSTER algorithm. The mean (±SE) integrated serum LH concentration (IU/Umin) increased following the administration of naltrexone (4715±298) in comparison to the control day (3997±381; p=0.0008). The mean number of LH pulses (/12 h) detected on the naltrexone day (10.3±0.3) was higher than on the control day (8.9±0.4; p=0.0068). Mean maximal LH peak height (lUlL) was greater on the naltrexone (7.8±0.5) vs control (6.7±0.5) days (p=0.0064) as was the in-

INTRODUCTION

cyof LH pulses appears to reflect the frequency with which effective quanta of gonadotropin-releasing hormone (GnRH) are secreted by the hypothalamus (1-3), the amplitude characteristics of the LH pulse may be dictated by the amount of GnRH secreted per hypothalamic discharge and/or the prevailing response characteristics of the LH-secreting gonadotropes. Within this context, substantial evidence has accrued in the rat (4-9), monkey (10), and human (11-16) suggesting that opioid peptides playa key role in modulating gonadotropin secretion, More specifically, considerable investigative attention has focused on the possibility that the brain opioid peptidergic systems mediate the feedback effects of the gonadal hormones (17-26) presumably by alteration of the firing rate of the GnRH pulse generator (12, 15,24,27-29).

Although the release of luteinizing hormone (LH) is episodic and the characteristics of LH pulses vary during the menstrual cycle, the physiological mechanisms that ultimately regulate such pulsatile activity remain to be fully defined. Whereas the frequen-

1This work was supported in part by NIH RR-00847 to the General Clinical Research Center of the University of Virginia; by Research Career Development Awards HD-00711 (to WSE) and HD-00634 (to JDV); by NIH grant HD-20465 (to ADR); and by the NIH-sponsored CLiNFO data reduction systems. Key-words: LH pulses, opioid antagonists, early follicular phase. 1Correspondence: Dr. William S. Evans, Box 511, University of Virginia, Health Sciences Center, Charlottesville, Virginia 22908, USA.

Received October 24. 1991; accepted May 21, 1992.

525

WS. Evans, J Y. Weltman, M.L. Johnson, et al.

Whereas studies in the rat (6) and chimpanzee (30) have suggested that opioid peptides may influence gonadotropin release during the time of initial follicular development, investigations in early follicular phase women have both confirmed (12) and refuted (19, 26) an effect of opioid peptide receptor blocking agents on LH release. We are unaware, however, of any studies in which quantitative characteristics of pulsatile LH release have been appraised via application of an objective, validated pulse detection algorithm to LH time-concentration series obtained with a relatively intensive and extended blood sampling protocol. We have now undertaken such a study to test the hypotheses that administration of the opioid-receptor blocking agent naltrexone to women in the normal early follicular phase will: (i) effect changes in the frequency of the LH pulse signal; (ii) alter particular amplitude and/or duration characteristics encoding the LH pulse; and/or (iii) modify the serum concentration of LH measured between the identified LH pulses.

followed by a mid luteal progesterone concentration of more than 16 nmol/L. Assays

LH concentrations were determined by radioimmunoassay (Clinetics, Inc., Tustin, CA). The minimal detectable concentration was 2.0 IU/L. Intraand interassay coefficients of variation were 3 to 5% at LH concentrations from 2.5 to 100 IU/L. All samples from an individual woman were analyzed in a single assay run to eliminate interassay variance. Serum estradiol was assayed by radioimmunoassay (Diagnostic Products Corporation, Los Angeles, CA) and progesterone was measured using the method described by Veldhuis et al. (31). Intra- and interassay variations for estradiol were 7.0% and 8.1 %, respectively, at values from 1 to 69 nmol/L. Pulse analysis

Pulsatile LH release was assessed using the CLUSTER, as previously described (32). This objective algorithm defines a pulse as a statistically significant increase in a consecutive group or "cluster" of hormone values followed by a statistically significant decrease in a second cluster of values. The increase or decrease is judged in relation to the dose-dependent experimental error (variance) expressed by a power function relationship of all the intrasample standard deviations and the mean sample concentrations (doses). The program permits the operator to specify the threshold t statistics and the cluster sizes of the test peaks and preand post peak nadirs. For the present analysis a 2 x 1 cluster size was used with a t statistic of 2.0 for the upstroke and 2.0 for the downstroke, as suggested by in vivo biological validation and computer simulation testing of sensitivity and positive accuracy (33). The CLUSTER program estimated the number of significant LH pulses/12 h, the mean maximal peak amplitude (maximal peak height), incremental amplitude (algebraic difference between nadir and peak maximum), peak duration, peak area, and interpulse "valley" mean serum LH concentration, as defined earlier (34).

MATERIALS AND METHODS Subjects

Twenty-seven women with a history of entirely normal menstrual cycles (mean age 32.4 yr; range 25 to 39) volunteered for and participated in this study. All subjects underwent a detailed history and physical examination and provided written informed consent in accordance with the guidelines established by the Human Investigation Committee of the University of Virginia. Baseline measures of hepatic, renal, metabolic, hematological and endocrine function were normal. Subjects were admitted on two occasions to the General Clinical Research Center at the University of Virginia during the early follicular phase (days 3 or 4 from the onset of menses) of two consecutive menstrual cycles. An indwelling heparin lock cannula was inserted in a forearm vein at 07:00 hand venous blood samples were withdrawn at 10 min intervals for 12 h (08:00h to 20:00h) on each day. During one cycle (control) no drug was given. During the other cycle naltrexone (1.0 mg/kg) was given orally at 07:30h, 30 min prior to the onset of sampling. The order of the control and naltrexone cycles was randomized such that half the subjects were studied under control circumstances first and the other half administered naltrexone first. Single daily blood samples were withdrawn from day 9 of the cycle until the onset of menses during both the control and naltrexone cycles and assayed for LH, FSH, estradiol and progesterone. Ovulation was considered likely by documentation of a LH surge

Statistical analysis The number of LH pulses, their characteristics and the interpulse concentrations determined for profiles obtained during baseline sampling and opiate antagonist administration were compared within subjects by paired Student's two tailed ttesting. An alpha level of p=0.05 was chosen a priori. Statistical power was evaluated assuming paired comparisons with a putative difference between control

526

Opioid receptor antagonist-associated LH release

Table 1 - Mean (± SE) serum concentrations of LH, FSH, 17-beta estradiol and progesterone measured in women during the early fol-

licular (EF), late follicular (LF) and midluteal (ML) phases of a control cycle (control) and a cycle with administration of an opioid-receptor antagonist (naltrexone). EF

LF

ML

Control

Naltrexone

Control

Naltrexone

Control

Naltrexone

LH (lUlL)

5.37±0.44

5.97±0.53

19.0±2.3 1

24.0±7.12

8.24±0.76

9 .93±1.35

FSH (lUl L)

11.3± 0.53

11 .8±0.72

15.1± 1.78

16.4±2.7

10.2±0.58

11.4±0.8 1

Estradiol (pmol/L)

91.3±14.4

90.5±3.9

913.8±64.6

912.7±61 .8

358±35.2

344.0±37

Progesterone (nmoI/L)

1.28±0.27

1. 00±0. 11

1.16±0 .16

100±0 15

25 .24± 2.78

2 1.9± 1.68

versus naltrexone treatments of 30% at an alpha level of O.OS and with a pooled variance estimated from the 27 subjects studied.

number was slightly increased in response to naltrexone , the maximal amplitude of the LH pulses did appear higher on the day of naltrexone administration . In addition , these profiles raised the possibility that the interpulse serum LH concentrations were augmented on the day of naltrexone administration. The mean integrated LH concentration (lUlL/min) in response to naltrexone (471S±298) was higher (p=0.0008) than that documented on the control day (3997±381). The results of the identification and characterization of LH pulses and estimation of LH concentrations between pulses are shown in Table 2. The number of LH pulses identified with CLUSTER on the day of naltrexone administration was slightly but significantly (p=0.0068) higher than on the control day. Mean maximal peak amplitude

RESULTS Temporal and hormonal characteristics of the control and naltrexone cycles

All women were studied on comparable days during the early follicular phase of two menstrual cycles (control cycle day 3.6±0.17 vs naltrexone cycle day 3.2±0.17; mean±SE; p>O .OS) which were of similar lengths (27.4±0.48 days vs 27 .3±0.S7 days ; control vs naltrexone; p=0 .87) . All women exhibited a midcycle LH surge during the control and naltrexone cycles on statistically indistinguishable days (day 1S.1±0.68 vs 14.8±0.4S; control vs naltrexone ; p=0 .74). Table 1 summarizes the mean (±SE) serum concentrations of LH , FSH , estradiol and progesterone obtained in the control and naltrexone cycles during the early follicular phase (the day of the 12 hour sampling), during the late follicular phase (one day prior to the preovulatory LH surge) and during the mid luteal phase (6 days after the LH surge). As can be seen, serum estradiol rose appropriately in the late follicular phase and remained elevated in the mid luteal phase during both the control and naltrexone months. Serum progesterone concentrations were low during the early and late follicular phases but were elevated during the midluteal phase of both cycles. For each phase of the cycle, there were no differences in the values of LH , FSH , E2 or progesterone measured during the control and naltrexone months.

TIME (MINUTES)

Fig. 1 - Serum LH concentration vs time curves for two representative women (A and B) sampled every 10 min for 12 h with-

Serum LH time-concentration series

out drug pretreatment (top pane/) and following naltrexone, 1.0 mg/kg orally, 30 min before the start of venous blood sampling (lower panel). These subjects were chosen to show the range in alterations in the serum LH concentration vs time series caused by mu opiate receptor blockade with naltrexone. The deflections at the top of each graph indicate LH pulses identified by CLUSTER.

Figure 1 illustrates representati ve 12-h serum LH concentration profiles from two women sampled on two consecutive days during the early follicular phase. Visual inspection of these results suggested that, in addition to the possibility that LH pulse

S27

W.S. Evans, J. Y. Weitman, M.L. Johnson, et al.

when Circulating concentrations of estradiol or estradiol and progesterone are increased from endogenous sources during the late follicular (19, 24) and luteal (22, 23) phases of the normal menstrual cycle, respectively, but may (12) or may not (19, 26) affect LH release in the early follicular phase of the cycle. The thesis that opioid peptides bring about inhibition of gonadotropin release by decreasing the firing frequency of the GnRH pulse generator has been suggested by investigations in the monkey (18) and human (12,15,24,28,29). Of particular interest is that, in contrast to observations in the chimpanzee (30) and rat (6), several studies in early follicular phase women have failed to demonstrate an effect of opioid antagonists (11, 19, 26) even though serum estradiol concentrations are considerably higher in such individuals than in the postmenopausal female. One possible explanation for these negative studies relates to the experimental approaches used previously to examine this question; e.g., the sampling protocols utilized and the methods of analysis employed. Earlier protocols may have been adequate to detect major alterations in LH release, but inadequate to reveal more subtle changes in pulsatile profiles. Here we have evaluated the effect of abrogating endogenous opioid tone on pulsatile LH release in early follicular phase women using a blood sampling protocol believed to capture the majority of LH pulses (35) and an objective computer-assisted pulse detection program (32), which minimizes the false positive and false negative error rates inherent to pulse detection (33). Our findings confirm the data of Moult et al. (12) suggesting an opioid-blocking agent-associated increase in the firing rate of the GnRH pulse generator, as assessed indirectly by the number of LH pulses. Moreover, our results indicate that the maximal LH pulse amplitude and the interpulse valley concentrations are enhanced in response to opioid-receptor blockade. Quite intriguing are the possible mechanisms through which the maximal amplitude of LH pulses is increased. In principle, naltrexone may elicit larger LH pulses per se, or LH pulses of unchanged size might be superimposed on a higher baseline. Both interpretations are consistent with the findings of an increase in overall LH concentrations including between the discrete pulses (the interpulse valley mean). Although the mechanism through which the administration of naltrexone to early follicular phase women results in augmentation of serum LH concentrations within and between discrete pulses is unclear, at least four potential explanations deserve consideration. First, it is possible that the interpulse concentrations of

Table 2 - Mean (±SE) LH peak frequency, peak characteristics and interpulse concentrations as assessed by the CLUSTER algorithm in normal early follicular phase women sampled every ten minutes for 12 hours during a control cycle (CONTROL) and during a cycle with administration of an opioid-receptor antagonist (NAL TREXONE). CONTROL

NALTREXONE

Peak Frequency (/12 h)

8.9±OA

10.3±O.31

Maximal Peak Amplitude (lUlL)

6.7±O.5

7.8±O.5 1

Incremental Peak Amplitude (lUlL)

2.1±O.1

1.9±O.1

Peak Area (lUlL/min)

62±6.1

50±4A1

Peak Width (min)

45±2A

40±1.8

Interpulse Valley Mean (lUlL)

5.0±OA

6.3±OA1

Nadir Concentration (lUlL)

4.4±OA

5.5±OA1

1P

Effects of opioid receptor blockade on luteinizing hormone (LH) pulses and interpulse LH concentrations in normal women during the early phase of the menstrual cycle.

To determine the role of endogenous opioid peptides in regulating pulsatile luteinizing hormone (LH) release in the early follicular phase of the mens...
1MB Sizes 0 Downloads 0 Views

Recommend Documents