Mitochondrial pharmacology: its future is now.

Mitochondrial medicine is an evolving discipline whose importance derives from the central function of mitochondria in adenosine triphosphate (ATP) production, generation of reactive oxygen species, and cell death by necrosis or apoptosis. Consequently, mitochondrial dysfunction plays an important role in the progression of aging and the pathophysiology of many common diseases and off-target drug effects. This provides an impetus for the development of mitochondrial pharmacology, and some promising therapeutic targets for mitochondrial protective therapy have been identified.

Serendipity and the discovery of novel compounds that restore mitochondrial plasticity.

The mitochondrial electron transport chain (ETC) plays a central role in energy generation in the cell. Mitochondrial dysfunctions diminish adenosine triphosphate (ATP) production and result in insufficient energy to maintain cell function. As energy output declines, the most energetic tissues are preferentially affected. To satisfy cellular energy demands, the mitochondrial ETC needs to be able to elevate its capacity to produce ATP at times of increased metabolic demand or decreased fuel supply. This mitochondrial plasticity is reduced in many age-associated diseases. In this review, we describe the serendipitous discovery of a novel class of compounds that selectively target cardiolipin on the inner mitochondrial membrane to optimize efficiency of the ETC and thereby restore cellular bioenergetics in aging and diverse disease models, without any effect on the normal healthy organism. The first of these compounds, SS-31, is currently in multiple clinical trials.

Targeting mitochondrial cardiolipin and the cytochrome c/cardiolipin complex to promote electron transport and optimize mitochondrial ATP synthesis.

BACKGROUND AND PURPOSE: Cardiolipin plays an important role in mitochondrial respiration and cardiolipin peroxidation is associated with age-related diseases. Hydrophobic interactions between cytochrome c and cardiolipin converts cytochrome c from an electron carrier to a peroxidase. In addition to cardiolipin peroxidation, this impedes electron flux and inhibits mitochondrial ATP synthesis. SS-31 (D-Arg-dimethylTyr-Lys-Phe-NH2 ) selectively binds to cardiolipin and inhibits cytochrome c peroxidase activity. Here, we examined whether SS-31 also protected the electron carrier function of cytochrome c. EXPERIMENTAL APPROACH: Interactions of SS-31 with cardiolipin were studied using liposomes and bicelles containing phosphatidylcholine alone or with cardiolipin. Structural interactions were assessed by fluorescence spectroscopy, turbidity and nuclear magnetic resonance. Effects of cardiolipin on electron transfer kinetics of cytochrome c were determined by cytochrome c reduction in vitro and oxygen consumption using mitoplasts, frozen and fresh mitochondria. KEY RESULTS: SS-31 interacted only with liposomes and bicelles containing cardiolipin in about 1:1 ratio. NMR studies demonstrated that the aromatic residues of SS-31 penetrated deep into cardiolipin-containing bilayers. SS-31 restored cytochrome c reduction and mitochondrial oxygen consumption in the presence of added cardiolipin. In fresh mitochondria, SS-31 increased state 3 respiration and efficiency of ATP synthesis. CONCLUSIONS AND IMPLICATIONS: SS-31 selectively targeted cardiolipin and modulated its interaction with cytochrome c. SS-31 inhibited the cytochrome c/cardiolipin complex peroxidase activity while protecting its ability to serve as an electron carrier, thus optimizing mitochondrial electron transport and ATP synthesis. This novel class of cardiolipin therapeutics has the potential to restore mitochondrial bioenergetics for treatment of numerous age-related diseases.

[1-deamino-4-cyclohexylalanine] arginine vasopressin: a potent and specific agonist for vasopressin V1b receptors.

To date, there are no vasopressin (VP) agonists that exhibit a high affinity and selectivity for the VP V1b receptor with respect to the V1a, V2, and oxytocin receptors. In this study, we describe the synthesis and pharmacological properties of [1-deamino-4-cyclohexylalanine] arginine vasopressin (d[Cha4]AVP). Binding experiments performed on various membrane preparations revealed that d[Cha(4)]AVP exhibits a nanomolar affinity for V1b receptors from various mammalian species (rat, bovine, human). It exhibits high V1b/V1a and V1b/oxytocin selectivity for rat, human, and bovine receptors. Furthermore, it exhibits high V1b/V2 specificity for both bovine and human vasopressin receptors. Functional studies performed on biological models that naturally express V1b receptors indicate that d[Cha4]AVP is an agonist. Like VP, it stimulated basal and corticotropin-releasing factor-stimulated ACTH secretion and basal catecholamine release from rat anterior pituitary and bovine chromaffin cells, respectively. In vivo experiments performed in rat revealed that d[Cha4]AVP was able to stimulate both ACTH and corticosterone secretion and exhibits negligible vasopressor activity. It retains about 30% of the antidiuretic activity of VP. This long-sought selective VP V1b receptor ligand with nanomolar affinity will allow a better understanding of V1b-mediated VP physiological effects and is a promising new tool for V1b receptor structure-function studies.

Resensitization of blood pressure response to mu-opioid peptide agonists after acute desensitization.

IV administration of mu-opioid peptide agonists (DAMGO, DALDA, and [Dmt(1)]DALDA) results in a transient, naloxone-sensitive, increase in blood pressure in awake sheep. Despite significant differences in pharmacokinetics, these blood pressure responses all last < 15 min. The lack of correlation between half-life and duration of action suggested rapid desensitization. When a second dose of the same agonist was repeated 30 min later, the response was completely abolished. An increase in blood pressure and rapid desensitization was also observed with the kappa-opioid agonist (U50488H), whereas delta-agonists (DPDPE and DELT) had no effect on blood pressure. The response to DAMGO was abolished after prior exposure to DAMGO or DALDA, but there was no evidence of cross-desensitization between mu and delta, or mu and kappa, opioid agonists. Full resensitization of the blood pressure response occurred by 4 h for DAMGO (t(1/2) = 15 min) and by 48 h for [Dmt(1)]DALDA (t(1/2) = 1.8 h). These data support our hypothesis that the transient nature of the blood pressure response to mu-opioid agonists is caused by rapid desensitization and suggest that the rate of resensitization is dependent on the pharmacokinetics of the agonist.

In vivo pharmacokinetics of selective mu-opioid peptide agonists.

Recent evidence suggests that highly selective mu-opioid agonists may provide good analgesia with less development of tolerance and dependence. H-Tyr-D-Arg-Phe-Lys-NH2 (DALDA) and H-Dmt-D-Arg-Phe-Lys-NH2 ([Dmt1]DALDA) were found to display high binding affinity and much greater selectivity for the mu-opioid receptor (K(i)delta/K(i)mu) > 10,000) compared with H-Tyr-D-Ala-Gly-MePhe-Gly-ol (DAMGO). In addition, [Dmt1]DALDA was 3000-fold more potent than morphine when administered intrathecally. A potential problem with peptide analogs as therapeutic agents is their susceptibility to enzymatic degradation in vivo and short elimination half-lives. In this study, we compared the stability of DAMGO, DALDA, and [Dmt1]DALDA after systemic administration in sheep. Peptide concentrations were measured using high performance liquid chromatography-mass spectrometry. When incubated in sheep blood at 37 degrees C, DAMGO, DALDA, and [Dmt1]DALDA were stable over 2 h. When given intravenously to sheep, the apparent volume of distribution was 50 to 80 ml/kg for all three peptides, suggesting that distribution was limited to blood volume. Plasma clearance of DAMGO (223 ml/kg/h) was 10-fold faster than DALDA and [Dmt1]DALDA (24 ml/kg/h), and their elimination half-lives were 0.24, 1.5, and 1.8 h, respectively. The half-lives of DALDA and [Dmt1]DALDA are even longer than morphine or meperidine in sheep. These favorable pharmacokinetic properties of DALDA and [Dmt1]DALDA, together with their mu-selectivity, potency, and long duration of action, make them ideal candidates as opioid analgesics.

Antinociceptive and respiratory effects of intrathecal H-Tyr-D-Arg-Phe-Lys-NH2 (DALDA) and [Dmt1] DALDA.

DALDA (H-Tyr-D-Arg-Phe-Lys-NH(2)) and [Dmt(1)]DALDA (H-Dmt-D-Arg-Phe-Lys-NH(2)) (Dmt = 2',6'-dimethyltyrosine) are potent and highly selective mu-opioid agonists (K(i)(delta)/K(i)(mu) > 10,000 and K(i)(kappa)/K(i)(mu) > 100). Both peptides carry a 3+ charge at physiological pH. Their antinociceptive and respiratory effects were compared with morphine (MOR) after intrathecal administration in rats. Both DALDA and [Dmt1]DALDA produced dose-dependent and naloxone-reversible antinociceptive effects with relative potencies of 14 and 3000x that of MOR. The antinociceptive potency of [Dmt1]DALDA far exceeded its affinity and potency at the mu-opioid receptor and may be explained by its ability to inhibit norepinephrine (NE) uptake in spinal cord synaptosomes. The antinociceptive response to [Dmt1]DALDA was significantly attenuated by the alpha(2)-adrenergic antagonist yohimbine. Thus, [Dmt1]DALDA may be regarded as a drug with dual actions, and its antinociceptive potency is better described by both its affinity and potency at mu-opioid receptors, and its potency at inhibiting NE uptake. The analgesic duration of an equipotent dose of MOR, DALDA, and [Dmt1]DALDA was 3, 7, and 13 h, respectively, and the long duration may be due to the hydrophilic nature of these peptide analogs. Respiratory effects were determined using whole body plethysmography at 3 and 30x the antinociceptive ED(50). A significant depression in minute ventilation was observed with the higher dose of morphine and both doses of DALDA, but not with either dose of [Dmt1]DALDA. Because of its high antinociceptive potency, long duration of action, and low propensity to induce respiratory depression, [Dmt1]DALDA is of interest as a drug candidate for intrathecal analgesia.

Matrix-assisted laser desorption/ionization mass spectrometric quantification of the mu opioid receptor agonist DAMGO in ovine plasma.

The synthetic opioid peptide analog Tyr-D-Ala-Gly-N-methyl-Phe-Gly-ol (DAMGO), which is a mu opioid receptor-selective agonist, was quantified in ovine plasma samples with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOFMS), using delayed extraction and a reflectron. The internal standard was pentadeuterated DAMGO. Timed-ion selection was used to select the precursor ion. The analysis of the post-source decay fragments improved the detection sensitivity, and the use of the precursor-product ion relationship optimized the specificity. For plasma samples, the inter-assay variability of this method was 6.4% (n = 79) and the intra-assay variability was 6.0% (n = 10). The variability for controls was 3.4% (n = 43). The profile of DAMGO amount versus time was determined in sheep plasma, and the corresponding pharmacokinetic data were calculated.

Dynorphin stimulates corticotropin release from mouse anterior pituitary AtT-20 cells through nonopioid mechanisms.

Dynorphin (Dyn) peptides were previously shown to increase plasma corticotropin (ACTH) in the ovine fetus, but the site of its action remains unclear. In the present study, Dyn A(1-17) was found to stimulate ACTH release from mouse anterior pituitary tumor AtT-20 cells in a dose-dependent manner. Naloxone did not block the effect of Dyn A(1-17) and the selective kappa-opioid receptor agonist U50488H did not stimulate ACTH release. Dyn A(2-17), a degradative peptide fragment that does not bind to opioid receptors, also stimulated ACTH release from AtT-20 cells. Although the nonopioid effects of Dyn have previously been attributed to N-methyl-D-aspartate (NMDA) receptors, the ACTH-releasing effects of Dyn A(1-17) in AtT-20 cells were not affected by co-administration of NMDA receptor antagonist LY235959. The ACTH response to Dyn A(1-17) could not be blocked by alpha-helical CRH (CRH antagonist) and was additive with a maximal stimulatory dose of CRH, suggesting different mechanisms of action. These results show that the release of ACTH by Dyn A(1-17) in AtT-20 cells is not mediated by kappa-opioid receptors or by the NMDA receptor.

A peripheral site of action for the attenuation of baroflex-mediated bradycardia by intravenous mu-opioid agonists.

We previously reported that i.v. DAMGO (Tyr-D-Ala-Gly-NMePhe-Gly-ol), a selective mu-opioid agonist, causes an increase in blood pressure with no change in heart rate in unanesthetized sheep and subsequently demonstrated that DAMGO attenuates baroreflex-mediated bradycardia. To determine the site and mechanism by which mu-agonists inhibit baroreflex sensitivity, we have carried out further investigations by using DAMGO and another mu-agonist, DALDA (Tyr-D-Arg-Phe-Lys-NH2). The bradycardic response to norepinephrine (NE) was significantly blunted after i.v. DAMGO or DALDA in both nonpregnant and pregnant sheep. In contrast, the tachycardic response to sodium nitroprusside (SNP) remained unchanged in the presence of DAMGO or DALDA. In view of the highly restricted distribution of DALDA across the blood-brain barrier (BBB), we hypothesized that the blunting of reflex-mediated bradycardia by mu-opioid agonists can occur peripherally. Pretreatment with the quaternary opioid antagonist, naloxone methiodide (NM), completely blocked the attenuation of baroreflex sensitivity by DAMGO and DALDA in both nonpregnant and pregnant animals. These data suggest that in addition to central mechanisms, mu-opioid agonists can inhibit baroreflex sensitivity at a peripheral site, most likely by inhibiting vagal influence on heart-rate control rather than by acting directly at baroreceptors.

Nociceptin/orphanin FQ increases blood pressure and heart rate via sympathetic activation in sheep.

This study was undertaken to examine the cardiovascular effects of nociceptin/Orphanin FQ (OFQ). Nociceptin/OFQ (10-300 nmol/kg, IV) stimulates an increase in mean blood pressure (MBP) and heart rate (HR) in chronically catheterized sheep. Pretreatment with phenoxybenzamine (5 mg/kg) attenuated the pressor response, consistent with sympathetically mediated vasoconstriction. Furthermore, the lack of a reflex bradycardia suggests either blunting of the baroreflex by nociceptin/OFQ or direct beta-adrenergic activation. The bradycardic response to norepinephrine (0.6 microg/kg, IV) remained intact after nociceptin/OFQ administration, demonstrating that nociceptin/OFQ does not blunt the baroreflex. Additionally, the increase in HR was completely reversed by pretreatment with propranolol. These data suggest that nociceptin/OFQ plays a role in cardiovascular regulation via sympathetic activation.

Cardiovascular effects of a mu-selective opioid agonist (tyrosine-D-arginine-phenylalanine-lysine-NH2) in fetal sheep: sites and mechanisms of action.

OBJECTIVE: We investigated the effects of DALDA (tyrosine-D-arginine-phenylalanine-lysine-NH2), a mu-selective opioid peptide, on heart rate and blood pressure in fetal sheep with long-term instrument implantation. STUDY DESIGN: DALDA was given as an intravenous bolus in doses ranging from 0.15 to 0.5 mg/kg. A 0.5 mg/kg dose of DALDA was given in the presence of the opioid antagonist naloxone and its quaternary analog naloxone methiodide (6 mg/h); it was also given in conjunction with the beta-adrenergic antagonist propranolol (2 mg/h). Statistical analyses were performed by 1-way and 2-way analysis of variance. RESULTS: The fetus responded to DALDA with an increase in heart rate with all doses (P <.01) but without any change in blood pressure. This response was abolished by naloxone (P <.001), naloxone methiodide (P =.003), and propranolol (P <.001). CONCLUSIONS: In the fetus intravenous DALDA increases heart rate without any change in blood pressure by way of the mu receptor and through central sympathetic activation. These effects of DALDA are different from those seen in the adult, suggesting different sites and mechanisms of action.

Respiratory depression after intravenous administration of delta-selective opioid peptide analogs.

We compared the effects of three micro-(DAMGO, DALDA, TNPO) and three delta-(DPDPE, DELT, SNC-80) opioid agonists on arterial blood gas after IV administration in awake sheep. None of the mu agonists altered pO2, pCO2 or pH. All three mu agonists decreased pO2 increased pCO2 and decreased pO2, and this effect was not sensitive to naloxone or TIPPpsi, a delta-antagonist, suggesting that it is not mediated by beta-opioid receptors. When administered to pregnant animals, there were significant changes in fetal pCO2 and pH. It may be possible to develop delta-selective opioid agonists which do not produce respiratory depression.

The role of N-methyl-D-aspartate receptors in the release of adrenocorticotropin by dynorphin A1-13.

We previously reported that dynorphin A1-13 evokes a significant increase in plasma adrenocorticotropin (ACTH) after intravenous administration in the ovine fetus. This response was not sensitive to naloxone and was regulated differently from the response to U50, 488H, a selective kappa-opioid agonist. NMDA appears to play a role in many of the nonopioid actions of dynorphin. We therefore hypothesized that dynorphin A1-13 may release ACTH via N-methyl-D-aspartate (NMDA) receptors. To test this hypothesis, we have compared the ACTH response to dynorphin A1-13 and NMDA in the chronically-instrumented ovine fetus. Our data show that both dynorphin A1-13 (0.5 mg/kg) and NMDA (4 mg/kg) induced a significant release of immunoreactive ACTH in the late-term ovine fetus. The ACTH response to NMDA was of a smaller magnitude, but of longer duration, when compared to dynorphin A1-13. The response to both dynorphin A1-13 and NMDA was significantly attenuated by pretreatment with the noncompetitive NMDA antagonist, MK-801, but was not affected by antagonists of corticotropin-releasing hormone and arginine vasopressin. Finally, the ACTH response to both dynorphin A1-13 and NMDA were inhibited by dexamethasone. The results of this study indicate a role for NMDA receptors in the action of dynorphin A1-13, and suggest that NMDA may act directly at the level of the pituitary to release ACTH without the involvement of hypothalamic secretagogues.

Implications of the kinetics of zidovudine in the pregnant baboon following oral administration.

Zidovudine (ZDV) therapy in pregnancy reduces mother-to-child transmission of HIV. The action of ZDV in the fetus is thought to be an important contributor to efficacy. Previous research in primates has demonstrated that continuous infusion of ZDV to the mother leads to sustained plasma concentrations in the fetus; however, it has not been determined what concentrations of ZDV are achieved in the fetus following oral administration. The pharmacokinetics of drug distribution to the fetus following oral administration of a 100-mg dose of ZDV to the mother are reported from 6 chronically catheterized baboons. The first order elimination half-life of ZDV from both the mother and fetus was approximately 1.2 hours. The area under the concentration-time curve for the fetus was 77% (r2 = 0.98; p < .001) that of the mother and the estimated peak drug levels in the fetus were 52% (r2 = 0.83; p < .01) those in the mother. The rapid transfer and short half-life of ZDV leads to a drug concentration-time profile that would not sustain levels in the fetus with dosing every 4 hours. After comparing these findings with existing data from pregnant and nonpregnant humans, it seems likely that current dose recommendations for ZDV in pregnancy would not maintain levels of the active intracellular metabolite of ZDV in all fetuses. This may explain in part the 8% failure rate of ZDV prophylaxis. The correlation between fetal and maternal plasma concentrations of ZDV would allow titration of dose based on maternal drug levels to achieve fetal levels within the therapeutic range.

Placental transfer and fetal metabolism of zidovudine in the baboon.

Zidovudine (azidothymidine, AZT) is used in pregnancy to reduce mother to infant transmission of HIV. Understanding the disposition of AZT in the fetus is necessary to optimize therapeutic regimens directed toward the fetus. Recent studies in primates found similar steady-state levels of the glucuronide metabolite of AZT (AZT-glu) in the fetus to those in the mother, raising the question of whether the metabolite was of fetal or maternal origin. The objective of this study was to determine whether glucuronidation occurred in the fetal compartment and to quantify the placental and fetal clearances of AZT using the two-compartment model at steady state. Steady-state concentrations were obtained after paired maternal and fetal infusions of AZT in chronically catheterized pregnant baboons. During maternal infusion, the mean (+/-SE) fetal to maternal ratio of AZT was < 1 (0.84 +/- 0.06, p < 0.02), suggesting clearance of AZT in the fetus. Mean total maternal clearance of AZT was 725 +/- 49 mL/min and placental clearance was 36 +/- 4 mL/min, or approximately 5% of maternal clearance. Fetal clearance of AZT was estimated at approximately 15% of placental clearance. This suggests fetal nonplacental clearance is minimal compared with that in the mother, but does not preclude the fetus from actively contributing to the metabolite in the fetal circulation. During infusion of AZT to the fetus, the concentration of AZT-glu in the fetus was 7.0 +/- 0.8 times that in the mother. This is compelling evidence that glucuronide can be formed in the fetal compartment. Thus, fetal metabolism has an impact on the concentration of both AZT and AZT-glu in the fetal circulation.

Baroreflex-mediated bradycardia but not tachycardia is blunted peripherally by intravenous mu-opioid agonists.

OBJECTIVE: We sought to test the hypothesis that an intravenous dose of H-Tyr-D-Arg-Phe-Lys-NH2, a highly mu-receptor selective opioid peptide, suppresses baroreflex sensitivity through a peripheral mechanism. STUDY DESIGN: A transient change in mean arterial pressure was produced in chronically instrumented pregnant ewes by norepinephrine or sodium nitroprusside in the absence or in the presence of H-Tyr-D-Arg-Phe-Lys-NH2, a highly mu-selective opioid peptide. In some studies naloxone methiodide, a peripheral opioid antagonist, was infused starting 60 minutes before the administration of H-Tyr-D-Arg-Phe-Lys-NH2 and maintained for a total of 90 minutes. Linear plots were obtained when the changes in mean arterial pressure during the pressure rise were plotted against the changes in heart rate and the sensitivity of the baroreflex was derived as the slope of the linear regression line. RESULTS: We observed (1) lower baroreflex sensitivity after H-Tyr-D-Arg-Phe-Lys-NH2 administration with a hypertensive stimulus; (2) unchanged baroreflex sensitivity after H-Tyr-D-Arg-Phe-Lys-NH2 administration with a hypotensive stimulus; and (3) unchanged baroreflex sensitivity after H-Tyr-D-Arg-Phe-Lys-NH2 administration with a hypertensive stimulus in the presence of naloxone methiodide. CONCLUSION: H-Tyr-D-Arg-Phe-Lys-NH2 suppresses the hypertensive but not the hypotensive arm of the baroreflex through peripheral opioid receptors. These results suggest that mu-opioid receptors are present in the vagus nerves and that the activation of these opioid receptors inhibits reflex bradycardia in pregnant sheep.

Baroreflex-mediated bradycardia is blunted by intravenous mu- but not kappa-opioid agonists.

To assess the cardiovascular effects of systemically administered opioid agonists, changes in blood pressure and heart rate were observed after intravenous (i.v.) administration of U50,488H (trans-3,4-dichloro-N-[2-(1-pyrrolidinyl) cyclohexyl]benzeneacetamide), a selective kappa-opioid receptor agonist, and DAMGO (D-Ala2, N-Me-Phe4, Gly5-ol), a selective mu-opioid-receptor agonist. Intravenous administration of U50,488H (1.2 mg/kg) and DAMGO (0.3 mg/kg) to the awake sheep resulted in an immediate increase in blood pressure. The pressor response to U50,488H was accompanied by an increase in heart rate. In contrast, there was no accompanying change in heart rate in response to DAMGO. We hypothesized that the lack of a reflex bradycardia to the pressor responses of both the mu- and kappa-opioid-receptor agonists was due to a blunting of baroreflex-mediated bradycardia. The reflex bradycardia to norepinephrine (0.6 microg/kg/min) was significantly reduced in the presence of DAMGO but not U50,488H. In view of the lack of effect of U50,488H on the baroreflex, we further hypothesized that the tachycardia it elicited was due to an increase in sympathetic activity. Pretreatment with propranolol (0.1 mg/kg) completely blocked the tachycardia elicited by U50,488H. These data suggest that the lack of a reflex bradycardia to the pressor response of DAMGO is due to a blunting of baroreflex-mediated bradycardia. In contrast, the increase in heart rate caused by U50,488H is mediated by sympathetic activation of the heart.

Cardiovascular and metabolic responses to two receptor-selective opioid agonists in pregnant sheep.

OBJECTIVE: Our purpose was to assess the effects of an intravenous dose of a highly selective mu-(DALDA) and delta-(DPDPE) opioid peptide to determine which class of peptide has the best clinical potential. STUDY DESIGN: Chronically instrumented pregnant ewes received a 0.3 mg/kg intravenous bolus of each peptide with and without opioid receptor blockade by means of a randomized prospective design. RESULTS: Intravenous DALDA produced only mild hypertension and a loss of heart rate variability, whereas DPDPE produced respiratory depression, maternal hypertension, and a fall in heart rate in both mother and fetus. Uterine blood flow, oxygen uptake, and glucose uptake were unchanged with both drugs. The effects of DALDA but not DPDPE were reversed by opioid receptor blockade. CONCLUSION: The delta-selective agonist had multiple nonopioid adverse effects, whereas the mu-selective agonist was well tolerated by the pregnant ewe, suggesting that mu-selective agonists have better potential for clinical use as an analgesic in pregnancy.