Modulation of morphine physical dependence and discriminative stimulus effects by ovarian hormones: Role of estradiol

Abstract

Ovarian hormones influence the activity of endogenous opioids, and exogenous administration of estradiol reduces opioid intake and opioid seeking in animal models of opioid reward and reinforcement. The purpose of this study was to examine the effects of ovarian hormones on the discriminative stimulus effects of morphine and naloxone-precipitated opioid withdrawal. To this end, separate groups of ovariectomized female rats were trained to discriminate the stimulus effects of either 3.0 or 10 mg/kg morphine, and substitution tests were conducted with estradiol or progesterone alone and in combination with morphine. At the conclusion of discrimination testing, rats were treated chronically with estradiol, progesterone, or their combination, and challenged with naloxone to measure opioid-like withdrawal symptoms. Finally, the effects of estradiol, progesterone, and their combination were examined on naloxone-precipitated withdrawal in morphine-dependent rats. Neither estradiol nor progesterone substituted for the morphine discriminative stimulus, but estradiol significantly increased the potency of morphine in rats trained to discriminate 10 mg/kg but not 3 mg/kg morphine. When administered chronically, neither hormone nor their combination produced an opioid-like withdrawal syndrome following a naloxone challenge. Acute administration of estradiol, but not progesterone or a combination of estradiol and progesterone, significantly reduced naloxone-precipitated weight loss in morphine-dependent rats. These data indicate that estradiol influences the behavioral effects of morphine, possibly by increasing endogenous tone at mu opioid receptors.

1. Introduction

A small but growing body of literature suggests that ovarian hormones influence the endogenous opioid system. For instance, ovariectomy decreases mu opioid receptor density, and estradiol treatment restores mu receptor density to levels observed in intact subjects (Brown et al., 1996; Joshi et al., 1993). Furthermore, coadministration of estradiol and morphine in ovariectomized rats increases mu opioid receptor (OPRM1) gene expression (Cruz et al., 2015). Consistent with these findings, estradiol increases concentrations of the endogenous opioid peptide met-enkephalin (Dupont et al., 1980), upregulates preproenkephalin-A mRNA and mu opioid receptor mRNA (Priest et al., 1995; Quiñones-Jenab et al., 1997), and increases mu-mediated stimulation of GTPγS binding (i.e., mu receptor coupling) in the dorsal striatum (Acosta-Martinez and Etgen, 2002). The effects of progesterone on the endogenous opioid system are less clear, but the limited data available suggest that progesterone decreases opioid receptor density in the hypothalamus and media preoptic area, but only in the presence of estradiol (Mateo et al., 1992; Weiland and Wise, 1990).

The effects of ovarian hormones on the endogenous opioid system have functional consequences, particularly regarding opioid-mediated reward and reinforcement. For instance, heroin self-administration decreases during the proestrus phase of the estrus cycle in female rats (Lacy et al., 2016; Schmidt et al., 2021), a phase in which both estradiol and progesterone reach peak concentrations (Freeman, 2006; Smith et al., 1975). Proestrus-associated decreases in heroin intake can be blocked by the estrogen receptor antagonist, raloxifene, but not the progesterone receptor antagonist, mifepristone (Smith et al., 2021). Moreover, acute administration of estradiol significantly decreases heroin intake in ovariectomized rats, but acute administration of progesterone fails to alter heroin intake and does not alter the ability of estradiol to decrease heroin intake (Smith et al., 2020). Ovariectomized rats treated chronically with estradiol self-administer less heroin than ovariectomized rats treated chronically with progesterone (Smith et al., 2021), and intact female rats treated chronically with estradiol self-administer less of the synthetic mu opioid agonist, remifentanil, than intact female rats treated with vehicle (Sharp et al., 2021). Finally, estradiol decreases the reinstatement of heroin seeking following extinction in both ovariectomized (Sedki et al., 2015) and intact (Vazquez et al., 2020) female rats.

Although estradiol decreases measures of opioid reward and reinforcement under multiple experimental conditions, the mechanisms by which this occurs are not known. It is possible that estradiol increases endogenous opioid receptor tone, thus mimicking the effects of agonists such as methadone and buprenorphine, but it is also possible that estradiol decreases opioid receptor tone, thus mimicking the effects of antagonists such as naloxone and naltrexone. The purpose of the present study was to examine the effects of estradiol and progesterone on opioid-mediated behavioral endpoints that are relevant to opioid misuse and addiction, but under conditions in which traditional opioid receptor agonists and antagonists produce divergent effects. To this end, the effects of estradiol and progesterone were examined alone and in combination with morphine in female rats trained to discriminate either a low (3.0 mg/kg) or high (10 mg/kg) dose of morphine.

Previous studies report that naloxone precipitates a mild, opioid-like withdrawal syndrome during high estradiol/progesterone states in human females (Roche and King, 2015). Consequently, at the conclusion of drug discrimination testing, rats were chronically treated with estradiol, progesterone, or a combination of estradiol and progesterone, and opioid-like withdrawal symptoms were measured after a naloxone challenge. Finally, because both estradiol and progesterone attenuate morphine withdrawal in mice (Sadeghi et al., 2009), the acute effects of estradiol, progesterone, and a combination of estradiol and progesterone were examined on naloxone-precipitated withdrawal following chronic administration of morphine.

2. Materials and methods

2.1. Subjects

The subjects were 16 ovariectomized female Long-Evans rats obtained from Charles River Laboratories (Raleigh, NC, USA). The subjects received ovariectomies from the vendor on postnatal day (PND) 42 and arrived at our institution on postnatal day 49. Upon arrival, all rats were housed individually in transparent polycarbonate cages in a large colony room and provided with various materials for environmental enrichment (e.g., chew sticks, plastic shelters, bedding materials) that were changed twice per week. Cages were kept in a temperature- (23 °C ± 1 °C) and humidity- (45 % ± 10 %) controlled colony room maintained on a 12-h light-dark cycle (lights on: 0500). Water was freely available in the home cage throughout the study, and food was provided once a day to maintain each subject’s body weight at 80 % of their age-adjusted body weight, as determined via a growth chart provided by Charles River Laboratories. All rats were maintained in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the United States National Institutes of Health, and all procedures were approved the Davidson College Animal Care and Use Committee.

2.2. Drug discrimination

Drug discrimination training and testing was conducted in operant conditioning chambers from Med Associates, Inc. (St. Albans, VT, USA). Each chamber was equipped with a houselight, two response levers, a stimulus light above each response lever, a food pellet dispenser, a food pellet receptacle located between the two response levers, and a speaker to mask extraneous sounds. Scheduling of experimental events was accomplished through software and interfacing provided by Med Associates, Inc.

Beginning six days after arrival, rats were trained to lever press under a fixed ratio (FR1) schedule of food reinforcement (45 mg grain pellets). Over the course of two weeks, the ratio requirement was gradually increased to FR15. Once responding on the FR15 schedule stabilized, discrimination training commenced in which rats received intraperitoneal injections of 3.0 mg/kg morphine or saline (low training dose: n = 8), or 10 mg/kg morphine or saline (high training dose: n = 8), 20 min before the session. During sessions in which morphine was administered prior to the session, responses on the drug-appropriate lever were reinforced on the FR15 schedule of reinforcement, whereas responses on the opposite lever had no programmed consequences. Conversely, during sessions in which saline was administered prior to the session, responses on the saline-appropriate lever were reinforced, whereas response on the opposite lever (i.e., the drug-appropriate lever) had no programmed consequences. Positions of the drug- and saline-appropriate levers were counterbalanced across subjects. All training sessions were 20 min in duration and were conducted at least five days/week until the discrimination was acquired. A pseudorandom sequence was used to determine which injection (morphine vs. saline) was administered, with the restrictions that the same injection was not administered for more than two consecutive sessions and the number of saline and morphine injections were equal over a two-week period. These conditions remained in effect until a rat met the discrimination criterion, which was defined as emitting at least 80 % of responses on the injection-appropriate lever prior to the delivery of the first reinforcer, and over 80 % of the total responses on the injection-appropriate lever for the entire session, for six consecutive sessions.

Once a rat met the acquisition criterion, substitution tests were conducted. Tests were generally conducted on Tuesdays and Fridays, with training sessions continuing at least three days/week. Responses were reinforced on the FR15 schedule on both levers during test sessions to not bias responding. Tests were conducted with saline, peanut oil (vehicle for estradiol and progesterone), morphine (0.3–10 mg/kg), estradiol (0.0003–0.3 mg/kg), and progesterone (0.003–3.0 mg/kg). All drugs were administered 20 min before the test session. Morphine and saline were administered via intraperitoneal injection, whereas peanut oil, estradiol, and progesterone were administered via subcutaneous injection. Doses and drugs were tested in an irregular order that varied across rats.

Following these initial substitution tests, additional tests were conducted in which estradiol and progesterone were administered in combination with morphine. During these tests, peanut oil, 0.03 mg/kg estradiol or 0.3 mg/kg progesterone was administered concurrently with various doses of morphine (0.3–10 mg/kg), 20 min before a test session. In addition, a combination of 0.03 mg/kg estradiol and 0.3 mg/kg progesterone was tested in combination with morphine to determine if the two hormones enhanced or attenuated the effects of the other. Finally, 0.03 mg/kg estradiol was administered 160 min before various doses of morphine (180 min before test sessions) to determine if the effects of estradiol differed across time points. Only estradiol was tested at this time point because only estradiol produced significant effects in combination with morphine in the primary analysis.

Two rats in the high training dose group died early in training for reasons unrelated to drug administration. One rat in the low training dose group died abruptly following administration of an injection of 0.3 mg/kg morphine. Data from that rat was excluded from the study.

2.3. Opioid dependence and naloxone-precipitated withdrawal

The first 10 rats that completed all drug discrimination tests ceased discrimination training and advanced to opioid dependence and withdrawal testing. At least three days separated the final drug discrimination test and the beginning of this phase of testing.

Throughout dependence and withdrawal testing, the same schedule of drug administration was followed. In this schedule, rats received chronic administration of a test drug twice daily for five days (Monday - Friday) and the morning of the sixth day (Saturday). These injections were administered at 08:00 and 20:00 Monday through Friday, and at 08:00 on Saturdays, for a total of 11 injections. A naloxone challenge was administered 4 h following the final injection (Saturdays at 12:00). No tests or injections were scheduled the following day (Sunday).

In the absence of morphine, naloxone-precipitated withdrawal testing was conducted following chronic administration of peanut oil (0.1 ml, sc, b.i.d., x 5.5 days), estradiol (0.03 mg, sc, b.i.d., x 5.5 days), progesterone (0.3 mg, sc, b.i.d., x 5.5 days), and estradiol + progesterone (0.03 mg and 0.3 mg, respectively, sc, b.i.d., x 5.5 days). The full testing schedule is shown in Fig. 1A, and the order of testing was counterbalanced across rats.

Fig. 1.

Timeline of events in the naloxone-precipitated withdrawal studies. Upper panel (A) depicts schedule of chronic hormone administration prior to naloxone challenge. Lower panel (B) depicts schedule of chronic morphine administration and acute hormone treatment prior to naloxone challenge. Tick marks represent 4-h intervals.

Immediately before withdrawal testing, rats were weighed to the nearest gram, administered naloxone (10 mg/kg, ip), and placed individually in clear transparent cages with no bedding. At 10, 20, and 30 min after the naloxone injection, each rat was viewed for 2 min, and somatic withdrawal symptoms were scored using the method of Gellert and Holtzman (1978), modified to exclude ejaculation (all subjects were female) and to analyze weight loss separately; otherwise, all other symptoms (i.e., escape attempts, wet-dog shakes, abdominal constrictions, diarrhea, teeth chattering, swallowing, salivation, chromodacryorrhea, ptosis, posture, and irritability) were measured as described in the original method. At 40 min post naloxone administration, each rat was weighed a second time (to determine naloxone-induced weight loss) and returned to its home cage.

Additional tests were then conducted in morphine-dependent rats to determine the effects of acute administration of peanut oil, estradiol, progesterone, and estradiol + progesterone on naloxone-precipitated withdrawal. In these tests, the same schedule described above was followed in which injections were given twice daily for five days and once on the sixth day. According to this schedule, rats were given three consecutive injections of 1 mg, sc, morphine (Monday morning - Tuesday morning), followed by three consecutive injections of 5 mg, sc morphine (Tuesday evening - Wednesday evening), followed by five consecutive injections of 20 mg, sc morphine (Thursday morning - Saturday morning). Rats were then administered either peanut oil (0.1 ml, sc), estradiol (0.03 mg, sc), progesterone (0.3 mg, sc), or estradiol + progesterone (0.03 mg and 0.3 mg, respectively, sc) exactly 20 min prior to the naloxone challenge. The full testing schedule is shown in Fig. 1B, and the order of testing was counterbalanced across rats. All other conditions were identical to those described above.

2.4. Chemicals

Morphine was generously supplied by the National Institute on Drug Abuse (Research Triangle Institute, Research Triangle Park, NC, USA) and dissolved in 0.9 % saline. Naloxone was purchased from MilliporeSigma (St. Louis, MO, USA) and dissolved in 0.9 % saline. Estradiol and progesterone were purchased from MilliporeSigma and dissolved in peanut oil.

Doses of estradiol and progesterone tested in the drug discrimination procedure spanned a dose range estimated to produce plasma concentrations spanning levels equivalent to and above those found in normally cycling female rats (Pfaus and Pfaff, 1992; Hu et al., 2004; White and Uphouse, 2004). Only one dose of estradiol (30 μg) and progesterone (0.3 mg) were evaluated on naloxone-precipitated withdrawal.

2.5. Data analysis

Drug discrimination data were expressed as a percentage of drug-appropriate responding (%DAR), which was calculated by dividing the number of responses on the drug-appropriate lever by the summation of responses on both levers and multiplying by 100. The primary measure was determined a priori as %DAR before the first reinforcer was obtained. The secondary measure was %DAR for the full session. Full substitution was defined as ≥80 %DAR. Under conditions in which full substitution was obtained, ED₅₀ values and 95 % confidence limits (95% CL) were calculated using log/linear interpolation from all doses of the curve. Potency differences were considered significant if the 95%CL did not overlap. Where appropriate, significant differences were confirmed via two-way ANOVA. Response rates were measured as responses per second.

Body weight loss during tests of withdrawal were calculated by subtracting body weight 40 min after naloxone challenge from body weight immediately before naloxone challenge. Weight loss was expressed to the nearest gram. All other somatic signs of withdrawal (with the exception of ejaculation) were marked as present or not present, quantified, and summated according to the method of Gellert and Holtzman (1978). All withdrawal data were analyzed via repeated-measures ANOVA using drug condition as the factor. Significant main effects were followed by paired t-tests using the Holms-Bonferroni correction for multiple comparisons. All tests were two-tailed and used an alpha value of 0.05.

3. Results

3.1. Drug discrimination

Rats trained to discriminate 3.0 mg/kg (low dose) morphine acquired the discrimination in a mean of 45 sessions (range: 27–73 sessions); whereas rats trained to discriminate 10 mg/kg (high dose) acquired the discrimination in a mean of 37 sessions (range: 26–47 sessions). During testing, morphine generated a linear dose-effect curve with full substitution in both training groups (Fig. 2). In the primary analysis (%DAR before the first reinforcer), morphine was significantly more potent in rats trained at the low training dose than at the high training dose (Table 1), and this effect was confirmed by an ANOVA (main effect of training dose: F[1,11] = 8.911; p = .012). Morphine dose-dependently decreased rate of responding, but no dose of morphine reduced responding to <50 % of saline control values.

Fig. 2.

Substitution dose-effect curves for morphine (A, D), estradiol (B, E), and progesterone (C, F). Dotted lines indicate criterion for full (80%DAR) substitution. Upper panels (A, B, C) represent %DAR before the first reinforcer; middle panels (D, E, F) represent %DAR for the full session. Response rates during morphine (G), estradiol (H), and progesterone (I) substitution tests. For all graphs, open circles indicate data from the group trained with 3 mg/kg morphine and filled circles indicate data from the group trained with 10 mg/kg morphine. All points indicate the mean (±SEM). n = 7 for the 3 mg/kg training dose; n = 6 for the 10 mg/kg training dose.

Table 1.

ED₅₀ (95 % Confidence Intervals) of morphine.

Condition	3 mg/kg training dose	10 mg/kg training dose
Morphine alone
Before 1st SR	0.89 (0.48–1.61)	3.43 (2.12–5.61)*
Full session	0.41 (0.18–0.82)	1.25 (0.69–2.22)
Morphine +0.03 mg/kg estradiol
Before 1st SR	1.09 (0.58–1.99)	0.63 (0.19–1.72)†
Full session	0.38 (0.19–0.69)	0.26 (0.12–0.52)†
Morphine +0.3 mg/kg Progesterone
Before 1st SR	0.77 (0.45–1.30)	1.03 (0.37–2.68)
Full session	0.38 (0.20–0.69)	0.20 (0.05–0.52)†
Morphine +0.03 mg /kg estradiol (180 min pretreatment)
Before 1st SR	0.78 (0.49–1.23)	0.57 (0.26–1.21)†
Full session	0.19 (0.10–0.32)	0.27 (0.13–0.50)†
Morphine +0.03 mg /kg estradiol +0.3 mg/kg Progesterone
Before 1st SR	1.39 (0.65–3.01)	0.50 (0.22–1.03)†
Full session	0.69 (0.29–1.55)	0.35 (0.19–0.59)†

Notes.

SR = reinforcing stimulus.

3 mg/kg training dose (n = 7); 10 mg/kg training dose (n = 6).

Asterisk (*) indicates significant difference from 3 mg/kg training dose.

Crosses (^†) indicate significant difference from morphine alone.

Neither estradiol nor progesterone produced >24%DAR in the primary analysis across a dose range estimated to produce plasma concentrations spanning levels equivalent to and above those found in normally cycling female rats (Fig. 2). Response rates approximated those obtained during vehicle testing even at the highest doses tested. Solubility limitations prevented the testing of higher doses.

Neither estradiol nor progesterone altered the discriminative stimulus effects of morphine in the 3 mg/kg training dose group in either the primary or secondary analysis when co-administered with morphine 20 min before the session, and ANOVA did not reveal any significant effects of condition (i.e., morphine alone versus combination). In the 10 mg/kg training dose group, 0.03 mg/kg estradiol significantly shifted the morphine dose-effect curve 5.4-fold to the left in the primary analysis (Fig. 3; Table 1), and this effect was confirmed by an ANOVA (main effect of condition: F[1, 5] = 14.322; p = .013). A dose of 0.3 mg/kg progesterone shifted the morphine dose-effect curve 3.3-fold to the left in the primary analysis, but this effect was not significant as determined via both potency ratios (Table 1) and ANOVA (condition: F[1, 5] = 3.494; p < .121). In the secondary analysis (%DAR for the full session), this dose of progesterone significantly shifted the morphine dose-effect curve 6.3-fold to the left; however, stimulus control was minimal under these conditions as indicated by the high level of drug-appropriate responding engendered by saline (~50%DAR). Neither hormone altered the rate-decreasing effects of morphine in either training group.

Fig. 3.

Substitution dose-effect curves for morphine (MOR; open circles), morphine +0.03 mg/kg estradiol (MOR + EST; downward-pointing triangles), and morphine +0.3 mg/kg progesterone (MOR + PRO; upward-pointing triangles) tested in the 3 mg/kg training dose (A, C) and 10 mg/kg training dose (B, D) groups. Dotted lines indicate criterion for full (80%DAR) substitution. Upper panels (A, B) represent %DAR before the first reinforcer; middle panels (C, D) represent %DAR for the full session. Response rates during morphine (MOR; open circles), morphine +0.03 mg/kg estradiol (MOR + EST; downward-pointing triangles), morphine +0.3 mg/kg progesterone (MOR + PRO; upward-pointing triangles) tested in the 3 mg/kg training dose (C) and 10 mg/kg training dose (D) groups. All points indicate the mean (±SEM). n = 7 for the 3 mg/kg training dose; n = 6 for the 10 mg/kg training dose.

In the primary analysis, estradiol significantly shifted the morphine dose-effect curve 6.0-fold to the left when administered 180 min before testing (Fig. 4; Table 1; main effect of condition: F[1, 5] = 92.258; p < .001). Coadministration of 0.03 mg /kg estradiol and 0.3 mg/kg progesterone significantly shifted the morphine dose-effect curve 6.9-fold to the left (Table 1; main effect of condition: F[1, 5] = 16.454; p = .010), but this shift did not differ significantly from the shift produced by estradiol alone. Neither pretreatment altered the effects of morphine on response rates.

Fig. 4.

Substitution dose-effect curves for morphine (MOR; open circles), morphine +180 min pretreatment with 0.03 mg/kg estradiol (MOR +180 min pretreatment EST; downward-pointing triangles), and morphine +0.03 mg/kg estradiol +0.3 mg/kg progesterone (MOR + EST + PRO; upward-pointing triangles) tested in the 3 mg/kg training dose (A) and 10 mg/kg training dose (B) groups. Dotted lines indicate criterion for full (80%DAR) substitution. Upper panels (A, B) represent %DAR before the first reinforcer; middle panels (C, D) represent %DAR for the full session. Response rates during morphine (MOR; open circles), morphine +180 min pretreatment with 0.03 mg/kg estradiol (MOR +180 min pretreatment EST; downward-pointing triangles), and morphine +0.03 mg/kg estradiol +0.3 mg/kg progesterone (MOR + EST + PRO; upward-pointing triangles) tested in the 3 mg/kg training dose (C) and 10 mg/kg training dose (D) groups. All points indicate the mean (±SEM). n = 7 for the 3 mg/kg training dose; n = 6 for the 10 mg/kg training dose.

3.2. Naloxone challenge

Ten rats (3 mg/kg training dose group: n = 5; 10 mg/kg training dose group: n = 5) advanced to naloxone challenge testing at the conclusion of drug discrimination testing and training.

In the absence of morphine, chronic administration of estradiol (0.03 mg, sc, b.i.d. x 5.5 days), progesterone (0.3 mg/kg, sc, b.i.d. x 5.5 days), and their combination did not produce opioid-like withdrawal effects following a naloxone (10 mg/kg, ip) challenge (Fig. 5). In all three tests, weight loss and somatic signs of withdrawal were minimal and did not differentiate from that produced by vehicle (peanut oil, 0.1 ml, sc, b.i.d. x 5.5 days).

Fig. 5.

Naloxone-induced loss of body weight (A) and withdrawal scores (B) following chronic treatment with vehicle (VEH), 0.03 mg/kg estradiol (EST), 0.3 mg/kg progesterone (PRO), and 0.03 mg/kg estradiol +0.3 mg/kg progesterone (E + P) in the absence of morphine. All bars indicate the mean (±SEM) of 10 rats.

Chronic administration of morphine (1–20 mg/kg, sc, b.i.d. x 5.5 days) produced physical dependence as evidence by naloxone precipitated weight loss and somatic withdrawal symptoms (Fig. 6). Withdrawal symptoms were present in all rats and did not differ between rats originating from the low versus high training dose groups in the drug discrimination experiments. Acute administration of estradiol (0.03 mg, sc) 20 min before a naloxone challenge significantly attenuated naloxone-precipitated body weight loss (Fig. 6). In contrast, progesterone (0.3 mg/kg, sc) failed to alter weight loss during withdrawal and prevented estradiol from reducing weight loss when administered in combination. Neither estradiol, progesterone, nor their combination significantly altered somatic signs of naloxone-precipitated morphine withdrawal. No order effects were observed.

Fig. 6.

Naloxone-induced loss of body weight (A) and withdrawal scores (B) in morphine-dependent rats (1–20 mg/kg, sc, b.i.d. x 5.5 days) following acute administration of vehicle (VEH), 0.03 mg/kg estradiol (EST), 0.3 mg/kg progesterone (PRO), and 0.03 mg/kg estradiol +0.3 mg/kg progesterone (E + P). All bars indicate the mean (±SEM) of 10 rats. Asterisk (*) indicates significant difference.

4. Discussion

The purpose of this study was to determine the effects of ovarian hormones on mu opioid-mediated drug discrimination and physical dependence in ovariectomized female rats. The principal finding of this study is that estradiol increases the potency of morphine’s discriminative stimulus at a high (but not low) training dose and attenuates naloxone-precipitated weight loss in morphine-dependent subjects. These findings are consistent with the hypothesis proposed previously that estradiol increases mu receptor tone (e.g., Acosta-Martinez and Etgen, 2002; Allen et al., 2014; Brown et al., 1996; Dupont et al., 1980; Joshi et al., 1993; Priest et al., 1995; Quiñones-Jenab et al., 1997; Roche and King, 2015; Smith, 2006).

We do not know of any studies that have examined the effects of ovarian hormones on the discriminative stimulus effects of morphine; however, there is evidence that females may be more sensitive than males to the discriminative stimulus effects of mu opioids. For instance, female rats acquire a morphine vs. saline discrimination at a faster rate than male rats, and mu opioid agonists are more potent in females than males (Craft et al., 1996). Females are less sensitive to the rate-decreasing effects of mu opioids than males, which complicates interpretation of drug discrimination data due to biases in reinforcement frequency between vehicle and drug sessions (i.e., the rate-decreasing effects of mu opioids leads to fewer response-reinforcer exposures when morphine is the discriminative stimulus than when saline is the discriminative stimulus). Indeed, when experimental parameters are modified to eliminate reinforcement bias, sex differences in the discriminative stimulus effects of mu opioids are no longer observed (Craft et al., 1999).

Two notable findings from the drug discrimination experiments are (1) the effects of estradiol and progesterone are limited to the discriminative-stimulus effects of morphine and do not extend to its rate-decreasing effects, and (2) the influence of estradiol and progesterone on the discriminative-stimulus effect of morphine are limited to the high training dose. Both the discriminative-stimulus and rate-decreasing effects of morphine are mediated by the mu receptor (Walker et al., 1994); however, a greater fractional receptor occupancy is required for mu opioids to produce rate-decreasing effects than discriminative-stimulus effects (Picker et al., 1990b). Differences in receptor occupancy may explain why estradiol and progesterone were able to influence morphine’s discriminative-stimulus effects (requiring lower receptor occupancy) than to influence morphine’s rate-decreasing effects (requiring higher receptor occupancy). Complicating this explanation is the finding that estradiol increased morphine discriminative-stimulus effects at a high training dose only, which requires higher receptor occupancy than a low training dose (Grabus et al., 1999; Picker et al., 1993). It must be noted that a high training dose of morphine increases the selectivity of the drug discrimination procedure to detect mu-mediated activity (Shannon and Holtzman, 1979; Picker et al., 1990a). Consequently, one explanation for the effects of estradiol at the high training dose is the greater selectivity of the high training dose for mu-mediated discriminative-stimulus effects.

Females are less sensitive than males to mu opioid-mediated physical dependence and withdrawal. For instance, female rodents undergoing spontaneous withdrawal from morphine exhibit less weight loss, lower withdrawal scores, and delayed withdrawal symptoms compared to males receiving equivalent doses of morphine for equivalent periods of time (Cicero et al., 2002; Papaleo and Contarino, 2006). Female rodents also display less severe naloxone-precipitated withdrawal symptoms than male rodents receiving the same regimen of agonist administration (Craft et al., 1999; Diaz et al., 2001, 2005; Nayebi and Rezazadeh, 2008; Sadeghi et al., 2009; but see Cicero et al., 2002). The role of ovarian hormones in these sex differences are not clear, but it is notable that both estradiol and progesterone decrease naloxone-precipitated weight loss in morphine-dependent mice (Sadeghi et al., 2009), and similar effects were observed with estradiol in morphine-dependent rats in the present study. Only one dose of estradiol and one dose of progesterone were tested in the present study, and higher doses may have produced more robust effects. The doses tested are estimated to produce plasma concentrations that mimic endogenous levels; however, plasma concentrations were not directly measured, which may be considered a limitation of the study. An additional limitation was the lack of a negative control in which estradiol was given in combination with naloxone in nondependent subjects, but we note that naloxone failed to induce withdrawal in the absence of morphine dependency under the present conditions (see Fig. 5).

The effects of estradiol on the discriminative stimulus effects of morphine and on naloxone-precipitated withdrawal are generally consistent with the limited data on estradiol’s effects on the endogenous opioid system. As noted above (see Introduction), estradiol increases met-enkephalin concentrations (Dupont et al., 1980), increases mu opioid receptor density (Brown et al., 1996; Joshi et al., 1993), increases preproenkephalin-A and mu opioid receptor mRNA (Priest et al., 1995; Quiñones-Jenab et al., 1997), and increases mu-opioid stimulated GTPγS binding in the dorsal striatum (Acosta-Martinez and Etgen, 2002). Data obtained in the present study suggest these types of estradiol-induced increases in mu opioid receptor tone have functional consequences.

Unlike estradiol, a physiological dose of progesterone did not significantly alter the discriminative effects of morphine in the primary analysis (%DAR before the first reinforcer). In the secondary analysis (% DAR for the full session), progesterone significantly increased the potency of morphine, but stimulus control was minimal under these conditions as indicated by high levels of %DAR for vehicle. The reason stimulus control was weak under these conditions is not known. A dose of progesterone 10-fold higher than that tested in combination with morphine did not significantly impact response rates when administered alone, so it is possible that a higher (supraphysiological) dose of progesterone may have produced a significant shift in the morphine dose-effect curve using our primary measure, which was associated with significantly greater stimulus control. Finally, it is notable that a combination of estradiol and progesterone did not produce further leftward shifts in the morphine dose-effect curve relative to estradiol alone, indicating that progesterone does not enhance the effects of estradiol on the morphine discriminative stimulus.

Several recent studies report that estradiol decreases behavioral measures relevant to opioid abuse and addiction. These effects are robust, observed following both acute and chronic estradiol administration, generalizable across rat strains, apparent on measures of both opioid intake and opioid seeking, and extend across different types of opioids (Freeman, 2006; Lacy et al., 2016; Schmidt et al., 2021; Sedki et al., 2015; Smith et al., 1975, 2020, 2021; Vazquez et al., 2020). Both opioid agonists (e.g., methadone, buprenorphine) and antagonists (e.g., naltrexone) decrease measures of opioid reinforcement and are clinically used to treat opioid use disorder (Lobmaier et al., 2011; Mello and Mendelson, 1985; Negus, 2006). Opioid agonists increase mu receptor activity, thereby reducing craving and withdrawal associated with abstinence, whereas opioid antagonists block mu receptor activity in the presence of exogenously administered opioids, thus reducing the behavioral and neurochemical effects that contribute to their abuse. Data obtained in this study suggest that estradiol increases mu opioid receptor tone, at least under some conditions, and it is possible that estradiol reduces measures of opioid intake and opioid seeking by a similar mechanism.

Studies have reported that estradiol increases mu opioid receptor tone in humans. For instance, in normally cycling women, high-estrogen states are associated with greater mu opioid receptor availability and greater activation of endogenous opioid neurotransmission during a pain stressor, whereas low-estrogen states are associated with reductions in endogenous opioid tone in the thalamus, amygdala, and nucleus accumbens, as well as hyperalgesic responses to nociceptive stimuli (Smith, 2006). Additionally, post-menopausal women treated with estrogen experience greater increases in systolic blood pressure following an acute stressor than women treated with placebo, and this effect is attenuated by naltrexone, suggesting estrogen increases endogenous opioid activity in this population (Allen et al., 2014). Finally, women in the mid-luteal phase of the menstrual cycle, when estrogen and progesterone concentrations are high, scored approximately 2 points higher than women in the early follicular phase on an opioid-receptor-antagonist-specific, adverse-effect scale after a naltrexone challenge (Roche and King, 2015). High scores on this scale are characteristic of naltrexone-precipitated opioid withdrawal and indicative of elevated concentrations of endogenous opioid activity. Data from these studies indicate that estradiol-associated increases in endogenous mu opioid receptor tone have functional consequences, which has translational implications for clinical conditions such as pain and addiction. It is not known whether similar effects are seen in males, and future studies must determine whether estradiol has similar effects in both sexes.

5. Conclusions

In ovariectomized rats, estradiol increased the potency of morphine’s discriminative stimulus effects at a high training dose and attenuated naloxone-precipitated weight loss in morphine-dependent rats. These data suggest that estradiol increases endogenous mu opioid receptor tone under some conditions, which is consistent with data from previous studies examining the effects of estradiol on mu opioid receptors and peptides. Increasing endogenous mu opioid receptor tone may be a mechanism by which estradiol decreases measures of opioid intake and opioid seeking.