The effect of a single nucleotide polymorphism of the mu-opioid receptor at nucleotide position 118 (OPRM1:c.118A>G) was investigated on morphine-6-glucuronide (M6G)-induced analgesia and respiratory depression in a group of healthy volunteers.
Sixteen subjects of either sex received 0.4 mg/kg (n = 8) or 0.6 mg/kg M6G (n = 8). At regular time intervals, the isocapnic acute hypoxic ventilatory response, pain tolerance (derived from a transcutaneous electrical acute pain model), and arterial blood samples were obtained. Data acquisition continued for 14 h after drug infusion. Population pharmacokinetic-pharmacodynamic sigmoid Emax models were applied to the respiratory and pain data. All collected data were analyzed using the statistical program NONMEM (San Francisco, CA).
Four of the subjects were OPRM1:c.118GA heterozygotes, and the remainder of the subjects were OPRM1:c.118AA homozygotes. M6G analgesia: In contrast to analgesic responses in OPRM1:c.118AA homozygotes, responses were small and inconsistent in OPRM1:c.118GA heterozygotes and best described by the function Effect(t) = baseline (P < 0.01 vs. OPRM1:c.118AA homozygotes). Emax and C50 values in heterozygotes equaled 0.55 +/- 0.18 (or a 55% increase in current above baseline) and 161 +/- 42 ng/ml, respectively. M6G-induced respiratory depression: For the acute hypoxic response, neither Emax nor C50 (value = 282 +/- 72 ng/ml) differed between genotypes.
The data indicate that the OPRM1:c.118A>G polymorphism affects opioid analgesic and respiratory effects differentially. Despite reduced analgesic responses to M6G the OPRM1:c.118A>G single-nucleotide polymorphism does not protect against the toxic effects of the tested opioid. However, some caution in the interpretation of the data is needed because of the small sample size. Further studies are needed to explore the link between this polymorphism and respiratory/analgesic responses beyond the small human sample. In OPRM1:c.118AA homozygotes, the potency parameters differed by a factor of 2 for analgesic versus respiratory effect. In this respect, M6G differs favorably from morphine.
POTENT opioid analgesics, such as morphine and its metabolite morphine-6-glucuronide (M6G), produce their intended (analgesia) and side effects (such as respiratory depression) by an action at the μ-opioid receptor (OPRM1 ).1–3Recent studies identified several single-nucleotide polymorphisms (SNPs) of the OPRM1 gene.4The most widespread SNP of the OPRM1 gene associated with a change in the amino acid sequence of the gene product is the substitution of the nucleotide adenine (A) with guanine (G) in exon 1 at nucleotide position 118 (OPRM1:c.118A>G SNP, dbSNP1799971:A>G). The result of this substitution at the receptor level is the exchange of amino acid asparagine (Asn) by aspartate (Asp) at the site of amino acid 40. Various studies have addressed the biologic effect of the OPRM1:c.118A>G SNP with respect to (1) opioid affinity to the OPRM1 ,4,5(2) μ-receptor endocytosis/desensitization,5(3) vulnerability to substance abuse (opioid and nonopioid),6,7(4) stress response to μ-receptor blockade,8–10(5) opioid-induced pupil constriction,11and (6) opioid-induced analgesia.12–14The picture that emerges from in vitro studies (1 and 2) is that, in contrast to most opioids (such as morphine and M6G),5β endorphin binds three times more tightly to the Asp40 (OPRM1:c.118G ) variant of the receptor (with three times greater potency) then to the Asn40 (OPRM1:c.118A ) variant.4However, no differences in μ-receptor endocytosis or internalization was observed between receptor types, indicating no marked functional differences.5Studies in humans4–6do point toward differences in opioid response in carriers of the OPRM1:c.118G allele compared with homozygous carriers of the OPRM1:c.118A receptor form. The cortisol response to opioid receptor blockade with naloxone is greater in carriers of the OPRM1:c.118G allele compared with OPRM1:c.118AA homozygotes.8,9The potency of morphine and its metabolite M6G to constrict the pupil is reduced by a factor of approximately 2 in OPRM1:c.118GA heterozygotes and by a factor of 3–4 in OPRM1:c.118GG homozygotes compared with OPRM1:c.118AA homozygotes.11Despite the relatively high frequency of the mutated allele in the population (10–30%),4,14,15few studies have addressed the issue of OPRM1:c.118A>G SNP and opioid-induced analgesia. Klepstad et al. 12showed that patients with cancer who are homozygous for the OPRM1:c.118G allele require twice as much morphine to achieve adequate pain control compared with heterozygous OPRM1:c.118G patients and homozygous OPRM1:c.118AA patients. In healthy volunteers, we recently observed a threefold reduction in M6G analgesic potency in OPRM1:c.118GA heterozygotes compared with OPRM1:c.118AA homozygotes.14Interestingly, the OPRM1:c.118A>G polymorphism has not been studied in relation to one of the most important side effect of opioids, respiratory depression. This is of importance because anecdotal data suggest that the OPRM1:c.118G allele protects against opioid toxicity.16,17In this study, we assessed the analgesic and respiratory effects of M6G in 16 healthy volunteers. We chose to test M6G in this study because we already have evidence for a reduced M6G-induced analgesic response in OPRM1:c.118GA heterozygotes.14
Note that we use the official SNP notation for the OPRM1 SNP at nucleotide position 118: OPRM1:c.118A>G , with alleles OPRM1:c.118A and OPRM1:c.118G .††Homozygotes are noted as OPRM1:c.118AA or OPRM1:c.118GG ; heterozygotes are noted as OPRM1:c.118GA . At the protein level, phenotypes are OPRM1:p.40Asn for OPRM1:c.118AA homozygotes and OPRM1:p.40Asp and OPRM1:p.40Asn for OPRM1:c.118GA heterozygotes.
Materials and Methods
Sixteen healthy volunteers (eight men, eight women; aged 18–30 yr) participated in the protocol after approval was obtained from the local human ethics committee (Commissie Medische Ethiek, Leiden, The Netherlands). Oral and written consent was obtained from all volunteers. None of the subjects had a history of illicit drug use; all women were taking oral contraceptives. The subjects were asked to have a normal night of sleep and not to eat or drink for at least 6 h before the study. During the studies, the subjects were allowed to drink water and to eat a light meal.
After arrival in the research unit, electrocardiographic electrodes were placed and an arterial line (for blood sampling) inserted in the left or right radial artery during local anesthesia. In the contralateral arm, an intravenous line was inserted for drug infusion.
The subjects were grouped randomly into four groups, receiving (1) 0.4 mg/kg M6G at 09:00 h, (2) 0.6 mg/kg M6G at 09:00 h, (3) 0.4 mg/kg M6G at 18:00 h, and (4) 0.6 mg/kg at 18:00 h. M6G was donated by CeNeS Ltd. (Cambridge, United Kingdom). The local pharmacy performed randomization and prepared the syringes on the day before the experiment. M6G (dissolved in 5 ml normal saline) was infused intravenously over 90 s. Before and after M6G infusion, we obtained analgesic and respiratory responses. These measurements continued for 14 h after the M6G infusion at regular intervals, with the exception of the sleep period in groups 3 and 4 (from approximately 24:00 to 07:00 h). This design enabled us to obtain measurements evenly spread out over the 14-h time period without the need to wake up subjects during their sleep period.
Blood sampling took place at times t =−10, 2, 5, 10, 15, 20, 25, 50, 80, 90, 100, 105, 150, 200, 210, and 240 min and every next hour until 12 h after the bolus infusion. In instances where blood sampling coincided with pain assessment, the pain test preceded the sampling. Plasma was separated within 15 min after blood collection and centrifuged for 10 min at 3,500 min−1. Plasma samples were immediately stored at −25°C until analysis. Plasma M6G concentrations were determined with liquid chromatography tandem mass spectrometry. The lower limits of quantification were set at 2.0 ng/ml. The coefficient of variation varied from 4 to 8% over the calibration range of 2–10.000 ng/ml.
Acute Pain Model
Acute pain was induced by an electrical current through two surface electrodes (Red Dot; 3M, Neuss, Germany) placed on the skin overlaying the tibial bone (shin bone) of the left leg. The electrodes were attached to a computer-interfaced current stimulator, which was locally designed and constructed. This pain model has been validated previously.14The intensity of the noxious stimulation was increased from 0 mA in steps of 0.5 mA/s with a pulse duration of 0.2 ms at 10 Hz (cutoff = 128 mA). The subjects were instructed to press a button on a control box when no further increase in stimulus intensity was acceptable (i.e. , pain tolerance). When the subjects pressed the button, the stimulus train ended, and the current was collected and stored on the hard disc of a computer for further analysis. Before drug infusion, the subjects were trained on both sessions for approximately 1 h during which several stimulus trains were applied. These data were discarded. After a subsequent resting period, baseline tolerance was assessed in triplicate. The intensity of antinociceptive measurements was every 10 min during the first 4 h and every 30 min afterward until the 14 h time point was reached. No measurements were made from 23:00 to 07:00 h.
End-tidal gas forcing and data acquisition were performed using the dynamic end-tidal forcing technique (see Dahan et al. 18for an explanation of the technique). In brief, a personal computer provided control signals to a set of mass-flow controllers (Bronkhorts, Veenendaal, The Netherlands) so that the composition of the inspired gas mixtures could be adjusted to force end-tidal oxygen and carbon dioxide concentrations to follow a specified pattern in time, independent of the ventilatory response. The inspired and expired oxygen and carbon dioxide concentrations and the arterial hemoglobin–oxygen saturation were measured with a Datex Multicap gas monitor (near the mouth) and Datex Satellite Plus pulse oximeter, respectively (Datex-Engstrom, Helsinki, Finland). End-tidal concentrations of oxygen and carbon dioxide, inspired minute ventilation (Vi), and oxygen saturation were collected and stored on disc for further analysis.
In this study, we performed steps from normoxia (end-tidal oxygen tension 110 mmHg for 8 min, end-tidal carbon dioxide tension = 50 mmHg) into hypoxia (end-tidal oxygen tension = 45 mmHg—values reached within four to six breaths, duration of hypoxia = 3 min, end-tidal carbon dioxide tension = 50 mmHg). Before drug infusion, control or baseline hypoxic responses were obtained. Next, the drugs were infused. Breathing responses were initially obtained at 30-min intervals (at t = 30 and 60 min after the bolus drug infusion) followed by 60-min intervals until the end of the study (no studies performed from 23:00 to 07:00 h).
The breath-to-breath data of the last 10 breaths of normoxia (Vi(normoxia) and the last 10 breaths of hypoxia (Vi(hypoxia) were averaged. Because the relation between ventilation and arterial oxygen saturation is linear, we calculated the difference between the hypoxic and normoxic minute ventilation and the oxygen saturation measured by pulse oximetry (Spo2) data points and expressed the acute hypoxic ventilatory response (AHR) or sensitivity as follows:
(units L/min per % desaturation).
We used two primers to amplify part of exon 1 of the OPRM1 gene containing the OPRM1:c.118A>G SNP (NM_000914.1:c.118A>G, dbSNP1799971:A>G): primer Oprm1F (5′-GGTCAACTTGTCCCACTTAGATCGC-3′) with a single nucleotide substitution (underlined in sequence), which creates a restriction site for the enzyme Bst UI when the G118 allele is present and Oprm1R (5′-AATCACATACATGACCAGGAAGTTT-3′). Polymerase chain reaction was performed on 100 ng genomic DNA isolated from blood samples in a total volume of 25 μl at a final concentration of 10 mm Tris-HCl (pH 8.8), 75 mm KCl, 1.5 mm MgCl2, 100 μm each dNTP, and 0.025 U/μl E-Taq polymerase (Eurogentec, Liège, Belgium) in the presence of 7.5 pmol of primers. Denaturation was 3 min at 94°C, followed by 38 cycles of amplification with denaturation for 30 s at 94°C, annealing for 1 min at 62°C, and extension for 1 min at 72°C, with a final extension for 10 min. A 20-μl polymerase chain reaction sample was analyzed on a 2% agarose gel. Amplified OPRM1 products were digested with the restriction enzyme Bst UI (New England Biolabs, Beverly, MA) according to the recommendation of the manufacturer. Each sample was analyzed on a 2% agarose gel, stained with ethidium bromide, and visualized by an ultraviolet transilluminator. To confirm the results from the restriction fragment length polymorphism analysis, amplified OPRM1 products were purified by use of the Qiaquick polymerase chain reaction purification kit (Qiagen, Valencia, CA) and sequenced on an ABI 377 sequencer using the same primers and the Big Dye Terminator cycle sequencing kit (Perkin Elmer, Shelton, CT) (see also figure 1).
Pharmacokinetic–Pharmacodynamic Data Analysis
The pharmacokinetic–pharmacodynamic analysis was performed with NONMEM version V, level 1.1 (a data analysis program for nonlinear mixed effects modeling; San Francisco, CA) using a population approach.19First, a pharmacokinetic analysis was performed. Two-and three compartment models were fitted to the data. Next, the pharmacodynamic analysis was performed on the analgesic and respiratory data with fixed individual pharmacokinetic parameters.
To eliminate a possible hysteresis between opioid plasma concentrations, as described by the pharmacokinetic model, and analgesic effect, an effect compartment was postulated. This effect compartment equilibrates with the plasma compartment with a time constant t1/2ke0(blood–effect site equilibration half-life).
Acute hypoxic responses were analyzed using the following inhibitory sigmoid Emax model:
where AHR(t) is AHR at time t, AHR0is baseline (= predrug) AHR, Ce(t) is the effect site concentration at time t, C50is the effect site or steady state concentration causing a 50% depression of AHR, Emax is the maximum possible effect, and γ is a dimensionless shape parameter.
We assume that M6G attenuates the response to the applied noxious stimuli by inhibition of signal propagation or central signal processing or both. As a consequence, stronger stimuli are needed before a subject presses the pain tolerance button. The attenuation (A) was described by an inhibitory sigmoid Emax model20
where Emax is the maximum attenuation and AC50is the effect site concentration causing 50% of the maximal attenuation effect. Because a response of the subject occurs when his or her pain sensation exceeds the response threshold (for pain tolerance), we may rewrite this as
where BLN is baseline (= predrug) current.
Likelihood ratio tests were performed to determine whether γ and Emax equaled 1. The interindividual variability of each model parameter was assumed to be log-normally distributed and was characterized by percent coefficient of variation. The improvement of the model fit by inclusion of covariates (time of infusion, sex, and genotype) was tested using the likelihood ratio criterion. Separate analyses were performed on pain tolerance, Vi(normoxia), and AHR. P values less than 0.01 were considered significant (e.g. , a decrease of > 6.63 in the NONMEM objective function). Values are reported as population value (median) ± SE.
The mean age of the subjects was 21.5 yr (range, 19–23 yr), and the mean weight was 71.1 kg (range, 55–91 kg). The sexes were evenly spread over the two M6G doses. All 16 subjects completed the protocol without major side effects. Nausea or vomiting did not occur.
We observed the OPRM1:c.118G allele with a frequency of 12.5%. Four of the subjects were heterozygous for the OPRM1:c.118G allele (genotype: OPRM1:c.118GA ; on protein level: OPRM1:p.40Asp + OPRM1:p.40Asn ), the remaining subjects were all homozygous for the OPRM1:c.118A allele (genotype: OPRM1:c.118AA ; on protein level: OPRM1:p.40Asn ). Examples of analgesic and hypoxic responses of homozygous OPRM1:c.118AA and heterozygous OPRM1:c.118GA individuals are given in figure 2. It illustrates the general observation that M6G produces consistent analgesic and respiratory responses in homozygous OPRM1:c.118AA individuals. OPRM1:c.118GA subjects showed M6G-induced respiratory depression, just like the OPRM1:c.118AA subjects, however, coinciding with small and inconsistent analgesic responses.
M6G-induced Analgesia in OPRM1:c.118GA versus OPRM1:c.118AA Subjects
Analysis of the data of the 16 subjects indicated an improvement in data fits (at the P < 0.01 level) when genotype (but not time of infusion or sex) was included as a covariate. However, we were unable to obtain a reliable estimate of the potency of M6G in the four OPRM1:c.118GA subjects (because of their small and inconsistent responses). The OPRM1:c.118GA analgesic data were best described by the function Effect(t) = baseline. In contrast, OPRM1:c.118AA subjects increased their current tolerance by 55% above baseline. The parameter estimates of the NONMEM analysis are given in table 1. In figure 3, some examples of the data fits in OPRM1:c.118AA and OPRM1:c.118GA subjects are given.
M6G-induced Respiratory Depression in OPRM1:c.118GA versus OPRM1:c.118AA Subjects
All subjects showed respiratory depression in response to M6G infusion. To get an indication of the response of OPRM1:c.118GA subjects relative to the OPRM1:c.118AA subjects, we plotted the individual responses of the OPRM1:c.118GA subjects against the mean OPRM1:c.118AA responses ± 95% confidence intervals for the 0.4 and 0.6 mg/kg M6G groups in figure 4. Apart from showing a dose-dependent effect of M6G on the AHR, it shows that M6G produces respiratory depression in both genotypes, with little difference in time course or magnitude of effect. The results of the population pharmacodynamic analysis indicate that inclusion of covariates time of infusion, sex, and genotype did not improve the model fits. The pharmacodynamic parameter values are given in table 1. Emax did not differ from 1, indicating that AHR = 0 was the maximum effect. In figure 5, we give some examples of the data fits in OPRM1:c.118AA and OPRM1:c.118GA subjects.
The observed differences in analgesic effect between genotypes were unrelated to differences in the pharmacokinetics of M6G. Plasma concentrations did not differ between genotypes. The population pharmacokinetic analysis indicated that the pharmacokinetic data were best described by a three-compartment model, with parameter values very similar to those observed previously (data not shown).14
A population pharmacokinetic–pharmacodynamic model was developed to assess the effect of the OPRM1:c.118A>G SNP on M6G-induced analgesic and respiratory responses. We observed that the OPRM1:c.118G allele had no effect on M6G-induced respiratory responses, whereas it caused a severe reduction in the analgesic efficacy of M6G.
In contrast to previous studies, we analyzed the respiratory data using an inhibitory sigmoid Emax model. We previously used a power model to analyze the respiratory effect of M6G. Because the power model does have our preference considering the flexibility of the model (see Dahan et al. 21for a discussion on the model), we also analyzed the current data set with the power model. Because the analysis with the sigmoid Emax model resulted in significantly better data fits (NONMEM objective function differed by 17 points), we present the analysis of the sigmoid model.
A potential drawback of our study is the small number of OPRM1:c.118GA subjects in our sample. The small sample size may have caused the overestimation and/or underestimation of the genotype effect for analgesia and respiration, respectively. To increase the power of our study, we performed a post hoc analysis on an extended data set. We increased the number of OPRM1:c.118GA subjects in the analysis of the analgesic data to 10 by adding data from a previous study from our laboratory on M6G analgesia.14In that study, subjects received 0.3 mg/kg intravenous M6G (two thirds given as a bolus over 90 s, the remainder given as a continuous infusion over 1 h). Analgesic measurements, identical to those of the current study, were obtained for 7 h. Reanalysis of the total data set yielded the following population parameter estimates for the OPRM1:c.118GA subjects: Emax = 0.30 ± 0.07, AC50= 59 ± 12 ng/ml, t1/2ke0= 2.0 ± 0.8 h, and a value of γ not different from 1 (see also fig. 6for the steady state dose–response relation of the 10 OPRM1:c.118GA and 12 OPRM1:c.118AA subjects). This indicates a sharp reduction in opioid efficacy in heterozygous OPRM1:c.118GA subjects with only a small analgesic effect, which occurred rapidly after M6G was infused. The magnitude and rapid onset of effect suggests that the analgesic response in the 10 heterozygous 118A>G subjects bears characteristics of a placebo response rather than a pure M6G-induced analgesic effect. The placebo response (related to phenomena such as anticipation, memory and suggestion) may be apparent because of the lack of significant M6G effect in this group. Studies on the development of placebo analgesia in OPRM1:c.118GA versus OPRM1:c.118AA subjects are needed to increase our insight in this matter.
For the respiratory data we were unable to extend our data set. Despite the fact that all four OPRM1:c.118GA subjects displayed overt respiratory depression (fig. 4), which contrasted sharply with their small and inconsistent analgesic responses (fig. 3), we believe that our respiratory data must be interpreted with care. Although our data clearly indicate that the OPRM1:c.118G allele is linked to a significant decrease in M6G analgesic efficacy, the absence of such an association with respect to M6G-induced respiratory effect should be considered preliminary. It may be that the effect of the OPRM1:c.118G allele is much smaller than its effect on antinociception and hence could not be unearthed from the small sample size. However, we believe that our results are more than coincidental (fig. 4) and suggest that the OPRM1:c.118G allele will not protect for M6G (or any other opioid)–induced respiratory depression. We therefore hypothesize at this point that the OPRM1:c.118G allele affects opioid analgesic and respiratory effect differentially.
Although our study was not intended to explain the putative mechanism of the differential effect of the OPRM1:c.118G allele on analgesia and respiration, some speculation in this respect is of interest. The OPRM1:c.118A>G substitution occurs within exon 1 of the OPRM1 gene and is expressed in the N-terminal region of the seven-transmembrane extracellular structure of the μ-opioid receptor.4The resultant exchange of amino acid asparagine by aspartate results in the loss of a putative site for N-glycosylation.4,In vitro studies did not show a difference in M6G binding between OPRM1:p.Asp40 and OPRM1:p.Asn40 receptors.5However, some SNPs of the OPRM1 gene (other than OPRM1:c.118A>G ) have been associated with alteration in G-protein coupling,22suggesting an important change in the functionality of μ-opioid receptor variants. In vivo differences in μ-opioid receptor functionality induced by the OPRM1:c.118A>G SNP remain unknown. Animal studies using antisense directed at exon 1 of Oprm1 indicate that M6G (but not morphine) antinociceptive effects remain unaltered when blocking exon 1. For example, in rats, treatment with antisense probes targeting exon 1 did not block M6G analgesia but significantly reduced morphine analgesia, whereas probes targeting exons 2 or 3 decreased M6G but not morphine analgesia.23Similarly, Oprm exon 1 knockout mice displayed analgesic responses to M6G (but not to morphine),24whereas exon 2 knockout mice displayed no analgesia in response to morphine but small (non–opioid-related) hyperalgesic responses to M6G.3The data from exon 1 knockout mice must be interpreted with care because it is unknown whether the observed analgesic responses were blocked by μ-opioid receptor antagonists. Our current findings are partially in contrast with the animal work. Our data suggest that exon 1 of the human OPRM1 gene is an essential requirement for at least part of M6G-induced analgesia (in contrast to the data in exon 1 knockout mice) but not for M6G respiratory effect. Our data then suggest the existence of distinct μ-opioid receptor variants (e.g. , splice variants) expressed in neurons involved in pain processing and those involved in the control of breathing. The receptors involved in pain processing are critically dependent on the functionality of exon 1 of the receptor gene, whereas receptors expressed in respiratory pathways are not. Another possibility is that the variation at nucleotide 118 of exon 1 of the OPRM1 gene (which alters a glycosylation site in the receptor) impacts on receptor targeting to the cell surface and that in some neurons (nociceptive but not respiratory) receptors are not properly located on the neuron surface to be fully functional. Evidently, further investigations aimed at studying the molecular effects of the OPRM1:c.118A>G variation in respiratory and nonrespiratory neurons are needed to clarify our insight in the above.
The clinical implications of our data are that the OPRM1:c.118G allele does not protect for the toxic effects of M6G, which may occur after morphine or M6G administration in patients with renal impairment. This contrasts a hypothesis of Lötsch et al. 25that the OPRM1:c.118G allele is among the protective factors against M6G-related toxicity such as severe sleepiness and drowsiness in renal patients. However, before definite conclusions from our study can be drawn regarding any association between the OPRM1:c.118G allele and M6G respiratory responses, studies with larger sample sizes are required. Our data are in close accord with a recent clinical study.12Patients with cancer who were heterozygous for the OPRM1:c.118A>G polymorphism had significantly more pain at equal steady state morphine and M6G concentrations. Furthermore, patients who were homozygous for the OPRM1:c.118A>G polymorphism required significantly more morphine for adequate pain relief (homozygous OPRM1:c.118AA and OPRM1:c.118GG subjects required 97 mg/24 h and 225 mg/24 h, respectively) with corresponding greater morphine and M6G steady state plasma concentrations. Interestingly, side effects such as fatigue, nausea and vomiting, dyspnea, and constipation did not differ among the three genotypes.12
Finally, some comments on the pharmacodynamic parameter values are needed. (1) Although the pharmacodynamic parameter values (and their variability) are in accord with previous findings on the analgesic properties of M6G for both genotypes,14the onset/offset times and potency of M6G respiratory effect are different from those reported previously. A value for t1/2ke0of 1 h and a relatively low potency (C50> 450 ng/ml) for the acute hypoxic ventilatory response after 0.2 mg/kg M6G was reported by us.26We relate the longer onset/offset half-life and increased potency in the current study to the greater M6G doses given (0.4 and 0.6 mg/kg), which caused greater concentrations of M6G in the brain. M6G is known to pass the blood–brain barrier slowly and is a substrate of P-glycoprotein,27an adenosine triphosphate–dependent drug efflux pump expressed in brain capillary endothelial cells. Saturation of the P-glycoprotein efflux pump occurring at high M6G brain concentrations may have caused trapping of the M6G molecule within the brain compartment causing prolonged central effects at a relatively low potency compared with effects observed at much lower M6G brain concentrations. (2) In homozygous OPRM1:c.118AA subjects, we observed greater M6G plasma concentrations needed for 50% respiratory effect relative to 50% analgesic effect. This indicates that less M6G is needed for its intended effect than for its unintended effect (potency ratio AC50:C50∼1:2). In this respect, M6G differs favorably from morphine, which has an AC50:C50ratio of 1:1 (i.e. , similar morphine concentrations are needed to cause 50% respiratory and analgesic effects).28Our observations (this study and the data from Dahan et al. 28) are in agreement with earlier statements and strengthen our belief that M6G has an increased margin of safety relative to that of morphine.26,28,29
The authors thank Ivonne J. H. M. van Minderhout, B.Sc. (Laboratory Technician, Department of Human Genetics, Leiden University Medical Center, Leiden, The Netherlands), for her technical assistance in the genetic analysis. The authors thank Jeff Mogil, Ph.D. (Professor, McGill University, Montreal, Canada), and Brigitte Kieffer, Ph.D. (Professor, Université Louis Pasteur, Illkirch, France), for reading the manuscript and their valuable comments.