Activation of G-Proteins by Morphine and Codeine Congeners: Insights to the Relevance of O- and N-Demethylated Metabolites at μ- and δ-Opioid Receptors

Article abstract of DOI:10.1124/jpet.103.058602

Phenotypic differences in analgesic sensitivity to codeine (3-methoxymorphine) results from polymorphisms in cytochrome P450-2D6, which catalyzes O-demethylation of codeine to morphine. However, O-demethylation reportedly is not required for analgesic activity of the 7,8-saturated codeine congeners dihydrocodeine, hydrocodone, and oxycodone. This study determined the potency and efficacy of these compounds and their demethylated derivatives to stimulate μ- and δ-opioid receptor-mediated G-protein activation using agonist-stimulated guanosine 5′-O-(3-[³⁵S]thio)triphosphate ([³⁵S]GTPγS) binding. Results showed that 7,8-saturated codeine congeners were more efficacious than codeine in activating μ-receptors, but only dihydrocodeine was more efficacious at δ-receptors. Hydrocodone and oxycodone were ～10-fold more potent than codeine and dihydrocodeine at either receptor. Morphine-like compounds with a 3-hydroxy group were ～30- to 100-fold more potent than their 3-methoxy analogs at the μ-receptor, and these compounds generally exhibited greater efficacy (e.g., morphine produced 2-fold greater maximal stimulation than codeine). Removal of the N-methyl group did not affect efficacy or potency of codeine congeners to activate μ-receptors, whereas this modification generally increased efficacy but decreased potency of morphine congeners. At the δ receptor, morphine congeners showed greater potency and structure-dependent differences in efficacy compared with codeine congeners, whereas removal of the N-methyl group had effects similar to those observed at the μ-receptor. These results demonstrate that 7,8-saturated codeine congeners are more efficacious than codeine, which may explain their lack of requirement for 3-O-demethylation in vivo. Nonetheless, because all 7,8-saturated codeine congeners were significantly less potent than their morphine derivatives, further research is needed to understand the relationship between metabolism and in vivo activity of these compounds.

Full text of DOI:10.1124/jpet.103.058602

0022-3565/04/3082-547–554
THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS
U.S. Government work not protected by U.S. copyright
JPET 308:547–554, 2004
Vol. 308, No. 2
58602/1124543
Printed in U.S.A.
Activation of G-Proteins by Morphine and Codeine Congeners:
Insights to the Relevance of O- and N-Demethylated
Metabolites at ␮- and ␦-Opioid Receptors
Chad M. Thompson, Heidi Wojno, Elisabeth Greiner, Everette L. May, Kenner C. Rice,
and Dana E. Selley
Department of Pharmacology and Toxicology and Institute for Drug and Alcohol Studies (C.M.T., H.W., E.L.M., D.E.S.), Virginia
Commonwealth University, Medical College of Virginia, Richmond, Virginia; and Laboratory of Medicinal Chemistry (E.G.,
K.C.R.), National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland
Received August 15, 2003; accepted October 29, 2003
ABSTRACT
Phenotypic differences in analgesic sensitivity to codeine (3- and these compounds generally exhibited greater efficacy (e.g.,
methoxymorphine) results from polymorphisms in cytochrome morphine produced 2-fold greater maximal stimulation than
P450–2D6, which catalyzes O-demethylation of codeine to codeine). Removal of the N-methyl group did not affect efficacy
morphine. However, O-demethylation reportedly is not required or potency of codeine congeners to activate ␮-receptors,
for analgesic activity of the 7,8-saturated codeine congeners whereas this modification generally increased efficacy but de-
dihydrocodeine, hydrocodone, and oxycodone. This study de- creased potency of morphine congeners. At the ␦ receptor,
termined the potency and efficacy of these compounds and morphine congeners showed greater potency and structure-
their demethylated derivatives to stimulate ␮- and ␦-opioid dependent differences in efficacy compared with codeine con-
receptor-mediated G-protein activation using agonist-stimu- geners, whereas removal of the N-methyl group had effects
lated guanosine 5Ј-O-(3-[35S]thio)triphosphate ([35S]GTP␥S) similar to those observed at the ␮-receptor. These results dem-
binding. Results showed that 7,8-saturated codeine congeners onstrate that 7,8-saturated codeine congeners are more effica-
were more efficacious than codeine in activating ␮-receptors, cious than codeine, which may explain their lack of requirement
but only dihydrocodeine was more efficacious at ␦-receptors. for 3-O-demethylation in vivo. Nonetheless, because all 7,8-
Hydrocodone and oxycodone were ϳ10-fold more potent than saturated codeine congeners were significantly less potent
codeine and dihydrocodeine at either receptor. Morphine-like than their morphine derivatives, further research is needed to
compounds with a 3-hydroxy group were ϳ30- to 100-fold understand the relationship between metabolism and in vivo
more potent than their 3-methoxy analogs at the ␮-receptor, activity of these compounds.
The opioid analgesic, codeine (3-methoxymorphine), under- this gene are phenotypically described as poor metabolizers
goes several metabolic routes including N- and O-demethyl-
ation, as well as glucuronidation (Dayer et al., 1988; Mikus et
al., 1991; Caraco et al., 1996; Yu et al., 2002). It is well
established that codeine must be O-demethylated to mor-
phine to produce analgesia in both humans (Chen et al.,
1991) and rats (Mikus et al., 1991; Cleary et al., 1994). In
humans, cytochrome P450–2D6 (CYP-2D6) catalyzes this re-
action, and individuals with dysfunctional allelic variants of
(PMs) and are less sensitive to codeine. Individuals express-
ing multiple copies of active CYP-2D6 are extensive metabo-
lizers (EMs) and are more sensitive to codeine. EMs can
exhibit the PM phenotype, however, when codeine is coad-
ministered along with a CYP-2D6 inhibitor such as quini-
dine. The requirement for O-demethylation is not surprising
given that the reported Ki values for binding to ␮-opioid
receptors is 200 times higher for codeine than for morphine
(Chen et al., 1991; Mignat et al., 1995). Large differences in
binding affinities are also seen between other codeine conge-
ners and their O-demethylated metabolites; for example, the
Ki value for hydrocodone is 30-fold greater than that for
hydromorphone (Chen et al., 1991). These data indicate that
This work was supported by Grants DA-10770, DA-05274, DA-01647, and
DA-07027 from the National Institute on Drug Abuse.
Article, publication date, and citation information can be found at
http://jpet.aspetjournals.org.
DOI: 10.1124/jpet.103.058602.
ABBREVIATIONS: PM, poor metabolizer; EM, extensive metabolizer; [35S]GTP␥S, 5Ј-O-(3-[35S]thio)triphosphate; P450, cytochrome P450;
DMEM, Dulbecco’s modified Eagle’s medium; hMOR-CHO, Chinese hamster ovary K1 cells stably expressing the human ␮-opioid receptor;
DAMGO, [D-Ala2,N-MePhe4,Gly5-ol]enkephalin; ANOVA, analysis of variance; SNC-80, (ϩ)-4-[(␣R)-␣-((2S,5R)-4-allyl-2,5-dimethyl-1-piperazinyl)-
3-methoxy-benzyl]-N,N-diethyl-benzamide; Emax, efficacy.
547

548
Thompson et al.
metabolic conversion of most codeine congeners leads to the hydrocodone to hydromorphone have also shown that the
formation of metabolites with greater affinity for opioid re- parent compound appears to be responsible for analgesic
ceptors than their parent compounds.
effects in rats (Tomkins et al., 1997), monkeys (Lelas et al.,
Although O-demethylation of codeine to morphine is re- 1999), and humans (Otton et al., 1993).
quired for in vivo activity, attempts to demonstrate the im-
Even more surprising is dihydrocodeine, which (like co-
portance of this reaction for other codeine derivatives have deine) is reported to have a Ki for the ␮-receptor over 100-fold
failed. Oxycodone, for instance, is structurally similar (see greater than its O-demethylated metabolite dihydromor-
Fig. 1) to codeine and undergoes similar metabolism (Wein- phine (Schmidt et al., 2002). Studies using human liver mi-
stein and Gaylord, 1979; Ishida et al., 1982; Poyhia et al., crosomes indicate that CYP-2D6 is responsible for the O-
1992); thus, it was hypothesized that it too would require demethylation of dihydrocodeine and that EMs have 10 times
O-demethylation at the 3-position to produce analgesia. the clearance rate of PMs (Kirkwood et al., 1997). Research in
Studies in humans, however, suggest that oxycodone, and not humans (Wilder-Smith et al., 1998) and rats (Jurna et al.,
its O-demethylated derivative, is responsible for analgesia 1997), however, suggests that inhibition of O-demethylation
(Kaiko et al., 1996). A similar conclusion was drawn from has no effect on analgesia. Interestingly, intrathecally in-
studies on psychomotor effects and subject self-reporting of jected dihydrocodeine and dihydromorphine displayed simi-
drug experience following oxycodone administration (Heis- lar initial antinociceptive effects, but the action of dihydro-
kanen et al., 1998). Studies examining O-demethylation of codeine dissipated more rapidly, suggesting that
biotransformation was not occurring during the assay (Jurna
et al., 1997).
Most clinically relevant opiate agonists, such as morphine,
produce biological effects primarily through activation of
␮-opioid receptors (Matthes et al., 1996; Sora et al., 1997; Loh
et al., 1998). However, ␦-opioid receptors may also modulate
the antinociceptive and reinforcing effects of these opiates
(Heyman et al., 1989; Martin et al., 2000) and may play a role
in opiate tolerance (Abdelhamid et al., 1991; Nitsche et al.,
2002). Because ␮- and ␦-opioid receptors activate G-proteins
of the Gi/o subfamily, agonist efficacy at these receptors can
be measured in vitro using agonist-stimulated [35S]GTP␥S
binding (Traynor and Nahorski, 1995; Selley et al., 1997;
Szekeres and Traynor, 1997), which provides a direct mea-
sure of receptor-mediated G-protein activation (Asano and
Ross, 1984; Hilf et al., 1989). Under conditions where there is
no receptor reserve for G-protein activation, such as in rat
thalamic membranes, maximal stimulation of [35S]GTP␥S
binding by agonist-occupied receptor directly correlates with
intrinsic efficacy (Selley et al., 1998). To our knowledge, no
one has used [35S]GTP␥S binding to directly measure G-
protein activation by codeine congeners and their metabolites
that result from P450-mediated biotransformations. The pur-
pose of these studies is to assess whether discrepancies re-
ported between in vitro binding data and in vivo behavioral
studies can be explained at the level of G-protein activation.
In addition, these studies will provide information on struc-
ture-activity relationships at ␮- and ␦-opioid receptors. These
results will elucidate which structural moieties increase ef-
ficacy and potency of codeine-based opioid analgesics and
provide insight into the importance of specific P450-mediated
reactions in the actions of these drugs.
Materials and Methods
Materials. Male Sprague-Dawley rats (150–200 g) were pur-
chased from Harlan (Indianapolis, IN). All chemicals were obtained
from Sigma-Aldrich (St. Louis, MO) except the following: [35S]GTP␥S
(1250 Ci/mmol) was purchased from PerkinElmer Life and Analyti-
cal Sciences (Boston, MA), GTP␥S from Roche Diagnostics (India-
napolis, IN), DMEM/F-12 from Invitrogen (Carlsbad, CA), and Eco-
lite scintillation fluid from ICN Biomedicals Inc. (Costa Mesa, CA).
All agonists, with the exception of noroxycodone, were obtained from
the National Institute on Drug Abuse (Bethesda, MD). Whatman
GF/B glass fiber filters were purchased from Fisher Scientific Co.
(Pittsburgh, PA).
Fig. 1. Structural formulas of codeine and morphine congeners. Struc-
tures on the right are the 3-hydroxy analogs of those on the left. 17-N-
demethylated (nor) metabolites have a methyl group removed from the
nitrogen in position 17 of the morphinan backbone.

G-Protein Activation by Morphine and Codeine Congeners
549
hMOR-Low Cell Line Construction. Transfection of Chinese The product separated upon cooling as colorless solid (0.6 g, 70%;
hamster ovary K1 cells with the human ␮-opioid receptor cDNA in a Currie et al., 1961), m.p. (found): Ͼ295°C (dec), m.p. (Speyer and
pIRES eukaryotic expression vector was performed using Lipo-
Sarre, 1924): Ͼ295°C (dec).
fectAMINE according to the manufacturer’s recommended proce-
Data Analysis. All data are reported as the means Ϯ S.E.M. of at
dure. Stable clones were selected by further incubation in the pres- least three experiments, each performed in triplicate. Nonlinear
ence of 0.25 mg/ml hygromycin B and isolated using sterilized regression analysis was conducted by iterative fitting using JMP
cloning wells. Clones were screened for ␮-opioid receptor expression (SAS for Macintosh; SAS Inc., Cary, NC). Statistical significance was
levels using [3H]naloxone saturation binding, and G-protein coupling
determined by ANOVA followed by the Student’s t test. For groups
with unequal variances due to large differences in potency (i.e.,
was confirmed by [35S]GTP␥S binding assays with DAMGO (see
micromolar range to nanomolar range), unequal variances were con-
firmed, and the means were then compared using the Welch
ANOVA. Percentage of maximal stimulation is expressed as net
stimulated binding by agonist as a percentage of stimulation by
DAMGO (30 ␮M for thalamic and 10 ␮M for hMOR-CHO mem-
branes) or 5 ␮M SNC-80 (for NG108-15 membranes).
below).
Membrane Preparations. NG108-15 cells were cultured in
DMEM supplemented with 100 U/ml penicillin, 100 ␮g/ml strepto-
mycin, and 5% fetal calf serum. Chinese hamster ovary K1 cells
stably expressing the human ␮-opioid receptor (hMOR-CHO cells)
were cultured in DMEM/F-12 supplemented with 100 U/ml penicil-
lin, 100 ␮g/ml streptomycin, 0.25 mg/ml hygromycin B, and 10% fetal
calf serum. Male Sprague-Dawley rats were killed by decapitation,
followed by thalamic dissection on ice. Membranes were prepared by
homogenization of tissue or cells in membrane buffer (50 mM Tris-
Results
Agonist Potency and Efficacy at the Rat ␮-Opioid
HCl, 3 mM MgCl2, 1 mM EGTA, pH 7.4) and centrifugation at Receptor. Several studies have provided evidence that bio-
50,000g for 10 min at 4°C; then, they were resuspended in the same
buffer at 1.5 mg/ml and stored at Ϫ80°C.
[
transformation is required for codeine to exert analgesic
properties. In contrast, other codeine congeners, such as di-
hydrocodeine, hydrocodone, and oxycodone, appear not to
require biotransformation to promote analgesia. To examine
the effects of these codeine congeners at the level of ␮-opioid
receptor-mediated G-protein activation, concentration-effect
curves were performed in rat thalamic membranes, which
contain a relatively pure population of ␮-opioid receptors
(Herkenham and Pert, 1982; Sim et al., 1995). Figure 2A
shows the concentration-effect curves for the two analogs,
codeine and dihydrocodeine, and some of their demethylated
derivatives. Despite evidence that only codeine requires O-
35S]GTP␥S Binding Assays. Before assays, samples were
thawed on ice, centrifuged at 50,000g for 10 min at 4°C, and resus-
pended in assay buffer (50 mM Tris-HCl, pH 7.4, 3 mM MgCl2, 0.2
mM EGTA, and 100 mM NaCl). Reactions containing 10 ␮g of mem-
brane protein were incubated for 1.5 h (NG108-15) or 2 h (thalamus,
hMOR-CHO) at 30°C in assay buffer containing 30 ␮M (thalamus,
NG108-15) or 10 ␮M (hMOR-CHO) GDP, 0.1 nM [35S]GTP␥S, and
various concentrations of agonist. Nonspecific binding was deter-
mined with 20 ␮M unlabeled GTP␥S. For thalamic membranes, an
additional 10-min incubation with 4 mU/ml adenosine deaminase
was conducted at 30°C before adding membranes to the reaction.
Reactions were terminated by rapid vacuum filtration through GF/B
glass fiber filters, and radioactivity was measured by liquid scintil-
lation spectrophotometry at 95% efficiency for 35S.
[3H]Naloxone Binding Assay. For [3H]naloxone binding, 30 to
45 ␮g of hMOR-low cell membrane protein were incubated for 2 h at
30°C in Assay buffer with 0.1 to 15 nM [3H]naloxone. Nonspecific
binding was determined in the presence of 10 ␮M naltrexone. Reac-
tions were terminated by rapid vacuum filtration through GF/B glass
fiber filters, and radioactivity was measured by liquid scintillation
spectrophotometry at 45% efficiency for 3H.
Synthesis of Noroxycodone. Noroxycodone was prepared in a
synthesis, comprising 3 steps, starting from oxycodone. In the first
step, oxycodone was transformed into 14-acetoxy-7,8-dihydronorco-
deinone using a modified procedure published by Cheng et al. (Cheng
et al., 1996): A suspension of 0.9 g (2.9 mmol) oxycodone in acetic
anhydride (3.0 ml) was heated at 90°C for 1 h. The mixture was
evaporated and the resulting brown oil was taken up into CHCl3 (10
ml), washed with NH4OH solution (pH ϭ 9.5), brine (10 ml), dried
over Na2SO4 and evaporated to give 14-acetoxy-7,8-dihydronorco-
deinone as colorless solid (0.9 g, 84%). This crude product was used
without further purification. In the second step, 0.9 g (2.5 mmol)
14-acetoxy-7,8-dihydronorcodeinone was dissolved in CHCl3 (60 ml)
and treated with CNBr (3 g) under reflux conditions for 24 h. Evap-
oration of the solvent provided crude 14-acetoxy-N-cyano-7,8-dihy-
dronorcodeinone as a crystalline solid that was recrystallized from
MeOH (0.6 g, 62%), m.p. (found): 266–269°C, m.p. (Currie et al.,
1961): 260°C (dec). In the third step, noroxycodone was prepared
from 0.5
g (1.4 mmol) of 14-acetoxy-N-cyano-7,8-dihydronorco-
deinone, which was suspended in H2SO4 (25%, 20 ml) and stirred
under reflux conditions for 4 h. During this time, the solid gradually
dissolved. The solution was cooled to 10°C and the pH adjusted to 9.5
by addition of NH4OH. The product was extracted into CHCl3 (five
times, 30 ml). The combined organic layers were dried over Na2SO4,
filtered through celite, and evaporated. The resulting tan foam was
dissolved in ethanol (2 ml) and acidified with hydrogen iodide (57%).
Fig. 2. Agonist-stimulated [35S]GTP␥S binding in rat thalamic mem-
branes. Membranes were incubated with 0.1 nM [35S]GTP␥S, 30 ␮M
GDP, and varying concentrations of agonist (see Materials and Methods).
Data are expressed as the means Ϯ S.E.M. (n Ն 3) of the percentage of
maximal stimulation by 30 ␮M DAMGO.

550
Thompson et al.
demethylation to promote analgesia in vivo, dihydrocodeine analogs, although they showed a trend toward lower potency.
exhibits potency characteristics similar to those of codeine. The presence or absence of an O- or N-methyl group generally
This finding was confirmed by nonlinear curve fitting anal- had no significant effects on the relative efficacies of the
ysis of the data, as shown in Table 1. Codeine and dihydro- codone compounds, with the exception that oxymorphone
codeine exhibited 55- and 107-fold higher EC50 values com- exhibited a greater relative Emax value compared with its
pared with morphine and dihydromorphine, respectively. 3-methoxy derivative, oxycodone. Thus, the results of studies
Interestingly, the relative efficacy (Emax value) of dihydroco- in rat thalamus indicate that only codeine and oxycodone are
deine was significantly greater than that of codeine (66% rendered considerably more efficacious by O-demethylation,
versus 36%, respectively). Both codeine and dihydrocodeine whereas the potency of all codeine congeners is substantially
exhibited significantly lower relative efficacy values com- increased (Ͼ10-fold) by O-demethylation.
pared with their 3-hydroxy derivatives, morphine and dihy-
Agonist Potency and Efficacy at the Human ␮-Opi-
dromorphine. This difference was greatest with morphine, oid Receptor. To examine whether the findings in rat thal-
which exhibited a 1.8-fold greater Emax value than codeine.
amus are species-specific, we studied several of these com-
Although the affinity of the N-demethylated metabolite of pounds in hMOR-CHO cells. Using [3H]naloxone binding to
codeine (norcodeine) for opioid receptors is not different from quantify receptor expression (see Materials and Methods), a
that of codeine (Chen et al., 1991), it is possible that N- clone stably expressing a low level of receptor (Bmax
ϭ
demethylation of dihydrocodeine might increase the potency 0.393 Ϯ 0.0237 pmol/mg, KD ϭ 1.61 Ϯ 0.40 nM, n ϭ 6) was
of this agonist. Because such an effect could explain the lack selected to avoid receptor reserve and maintain the ability to
of requirement for O-demethylation for in vivo activity of discriminate the relative efficacies of partial agonists. Figure
dihydrocodeine, the effect of nordihydrocodeine on ␮-recep- 3 shows the concentration-effect curves for several of these
tor-mediated G-protein activation was examined. However, opioids at the human ␮-opioid receptor, and Table 1 lists the
results (Fig. 2A; Table 1) showed that both the efficacy and Emax and EC50 values of these compounds adjacent to the
potency of this derivative were essentially identical to those values observed at the rat ␮-opioid receptor. Again, signifi-
of dihydrocodeine.
cantly greater potencies of all 3-hydroxy relative to 3-me-
Oxycodone and hydrocodone are two clinically relevant thoxy analogs were observed, as well as a significantly
opioids that also undergo P450-mediated biotransformation greater relative efficacy of morphine compared with codeine.
yet do not appear to require this metabolism to exert their in Interestingly, comparison of several compounds in hMOR-
vivo effects. Figure 2B shows the concentration-effect curves CHO cells and rat thalamus revealed that only codeine and
for these two compounds as well as their O- and N-demeth- oxycodone exhibited significantly different relative efficacies
ylated derivatives. The results of curve-fitting analysis (Ta- between the two systems, with approximately 1.5-fold
ble 1) show a 30- to 40-fold increase in the potency of the greater relative Emax values in hMOR-CHO cells. Addition-
3-hydroxy derivatives, oxymorphone and hydromorphone, ally, only dihydrocodeine and dihydromorphine exhibited sig-
compared with oxycodone and hydrocodone, respectively. In nificant differences in potency between the two systems, with
contrast, norhydrocodone and noroxycodone were not signif- 2- to 2.5-fold greater EC50 values in thalamus compared with
icantly different in potency compared with their N-methyl hMOR-CHO cells.
TABLE 1
EC50 and Emax values for ␮-opioid receptor-mediated stimulation of ͓35S͔GTP␥S binding in rat thalamus and hMOR-CHO cells
Data are means Ϯ S.E.M. (n Ն 3) obtained from nonlinear regression analysis of agonist concentration-effect curves. Emax and EC50 values with common letter designations
were not significantly different from other congeners within the subgroup, whereas those without common letter designations were significantly different (P Ͻ 0.05, Student’s
t test).
Rat Thalamus
hMOR-CHO
Agonist
EC50
Emax
EC50
Emax
␮M
%Max
␮M
%Max
Codeine Congeners
Codeine
10.64 Ϯ 0.84a
20.35 Ϯ 2.99b
2.23 Ϯ 0.28c
1.46 Ϯ 0.17c
36 Ϯ 1c
66 Ϯ 1a
51 Ϯ 3b
48 Ϯ 2b
6.59 Ϯ 1.82
55 Ϯ 3*b
73 Ϯ 1a
54 Ϯ 6b
Dihydrocodeine
9.35 Ϯ 2.75*
1.50 Ϯ 0.18
1.40 Ϯ 0.24
Hydrocodone
Oxycodone
67 Ϯ 4*ab
Morphine Congeners (O-demethylated
codeine congeners)
Morphine
Dihydromorphine
Hydromorphone
Oxymorphone
a
a
b
b
b
0.194 Ϯ 0.034†
0.190 Ϯ 0.035†
0.051 Ϯ 0.003†
0.049 Ϯ 0.006†
64 Ϯ 1†
75 Ϯ 3†
54 Ϯ 2c
63 Ϯ 3†
0.19 Ϯ 0.035†
0.071 Ϯ 0.012*†
0.030 Ϯ 0.006†
0.048 Ϯ 0.010†
70 Ϯ 4*ab
77 Ϯ 4a
58 Ϯ 4b
64 Ϯ 4b
a
b
N-Demethylated Codeine Congeners
Nordihydrocodeine
Norhydrocodone
18.5 Ϯ 0.63a
4.4 Ϯ 0.25b
5.2 Ϯ 1.3b
62 Ϯ 2b
54 Ϯ 4ab
48 Ϯ 3a
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
Noroxycodone
N-Demethylated Morphine Congeners
Normorphine
Norhydromorphone
Noroxymorphone
a
a
0.98 Ϯ 0.24¶
81 Ϯ 5¶
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
b
a
0.21 Ϯ 0.03¶
69 Ϯ 4¶
53 Ϯ 4b
b
0.20 Ϯ 0.02¶
%Max, percentage of maximal stimulation; N.D., not determined.
* P Ͻ 0.05, significantly different between thalamic and hMOR-CHO membranes.
† P Ͻ 0.05, significantly different from cognate codeine congener.
¶ P Ͻ 0.05, significantly different from cognate N-methylated congener.

G-Protein Activation by Morphine and Codeine Congeners
551
Fig. 3. Agonist-stimulated [35S]GTP␥S binding in hMOR-CHO cell mem-
branes. Membranes were incubated with 0.1 nM [35S]GTP␥S, 10 ␮M
GDP, and varying concentrations of agonist (see Materials and Methods).
Data are expressed as the means Ϯ S.E.M. (n Ն 3) of the percentage of
maximal stimulation by 10 ␮M DAMGO.
Agonist Potency and Efficacy at the ␦-Opioid recep-
tor. Another potential explanation for the discrepancies ob-
served between reported in vivo effectiveness of dihydroco-
deine, hydrocodone, and oxycodone relative to their low
potency to activate ␮-opioid receptors in vitro is that these
compounds may posses high potency and efficacy to activate
␦-opioid receptors. For this reason, efficacy and potency pro-
files were determined for these codeine-based opiates at the
␦-receptor using NG108-15 cells as model system, and com-
pared with their O- and N-demethylated derivatives. Much of
the reported receptor binding data indicate that these opioids
have much lower affinities for the ␦-receptor than for the
␮-receptor (Mignat et al., 1995; Schmidt et al., 2002). The
results shown in Fig. 4 and Table 2 agree with these binding
studies, as evidenced by the higher EC50 values of all com-
pounds tested in NG108-15 cells compared with rat thalamus
or hMOR-CHO cells. Despite substantially lower potencies of
these compounds at the ␦-receptor, certain effects of O- and
N-methylation status are evident. Morphine exhibited 1.4-
fold greater relative efficacy and over 30-fold greater potency
compared with codeine, whereas dihydromorphine showed
greater potency than dihydrocodeine with no difference in
relative efficacy (Table 2). Hydromorphone had approxi-
mately 6-fold higher potency than hydrocodone, but hydroc-
odone exhibited greater relative efficacy. In contrast, oxy-
morphone showed a 1.6-fold greater relative efficacy than
Fig. 4. Agonist-stimulated [35S]GTP␥S binding in NG108-15 cell mem-
oxycodone, and there was also a trend toward greater po- branes. Membranes were incubated with 0.1 nM [35S]GTP␥S, 30 ␮M
GDP, and varying concentrations of agonist (see Materials and Methods).
tency of the 3-hydroxy analog, but this was not statistically
significant. Interestingly, the presence or absence of an N-
methyl group had no significant effect on the potency or
Data are expressed as the means Ϯ S.E.M. (n Ն 3) of the percentage of
maximal stimulation by 5 ␮M SNC-80.
relative efficacy of dihydrocodeine, hydrocodone, or oxyc- relative efficacy and potency at either the ␮- or the ␦-opioid
odone at the ␦-opioid receptor. These results indicate that receptor. The presence or absence of an N-methyl group on
O-demethylation generally increases the potency of codeine morphine congeners, on the other hand, significantly affected
congeners at the ␦-receptor, whereas N-demethylation does the Emax of these compounds. Normorphine and norhydro-
not.
morphone were significantly more efficacious at the ␮-recep-
Structure-Activity Relationships for Codeine and tor than morphine or hydromorphone, respectively (Table 1).
Morphine Congeners. In addition to investigating meta- At the ␦-receptor, normorphine, norhydromorphone, and no-
bolic effects on G-protein activation by these opioids, the roxymorphone all exhibited greater relative Emax values than
results reported herein can also be examined from a struc- did their N-methyl analogs (Table 2). The potencies of these
ture-activity relationship perspective. For this purpose, nor- compounds at the ␮-receptor were also significantly affected
morphine, norhydromorphone, and noroxymorphone were by N-methylation status. Increases of 4- to 5-fold were ob-
also included in the analysis. The data shown above indicate served in EC50 values of the nor-derivatives of morphine,
that N-demethylation of codeine congeners has little effect on hydromorphone, and oxymorphone. At the ␦-opioid receptor,

552
Thompson et al.
TABLE 2
and potency (EC50) data herein vis-a`-vis studies involving
P450 metabolism is complicated by the fact that concomitant
changes in relative efficacy and potency do not occur across
metabolic biotransformations or structural moieties. For ex-
ample, these data indicate that the potency of hydrocodone
and oxycodone to activate ␮-receptors is increased by O-
demethylation, yet this reaction only significantly affects the
relative efficacy of oxycodone. Similarly, these data indicate
that at the ␦-receptor, O-demethylation of hydrocodone sig-
nificantly decreases efficacy and increases potency, whereas
O-demethylation of oxycodone increases efficacy without sig-
nificantly altering potency (although there was a trend to-
ward an increase). Despite these issues, these results dem-
onstrate that O-demethylation of these codeine congeners
generally produces metabolites that are significantly more
potent than their precursors. These findings suggest that
inhibition of P450 enzymes responsible for O-demethylation
would have discernible effects in vivo.
One possible explanation for why inhibition of P450-medi-
ated O-demethylation of hydrocodone and oxycodone does not
block analgesia is that these two compounds may be potent
enough to function without biotransformation. This seems
unlikely, however, because if the 5-fold difference in EC50
values between codeine and hydrocodone results in an ob-
servable nonrequirement for O-demethylation, then one
would expect that a 30-fold difference in EC50 value (e.g.,
oxycodone versus oxymorphone) would result in measurable
differences in potency in vivo (see Table 1, Rat Thalamus).
An alternative explanation for why codeine requires O-de-
methylation for in vivo activity, whereas hydrocodone and
oxycodone do not, is that both of these codones exhibit
greater efficacy to activate G-proteins through the ␮-receptor
relative to codeine. This finding, in addition to their greater
potency at the ␮-receptor compared with codeine, may negate
the requirement for in vivo biotransformation.
Another potential explanation is that different metabolites
could be responsible for the in vivo activity of hydrocodone
and oxycodone. In fact, whereas it has been shown that
inhibition of CYP-2D leads to increased formation of noroxy-
codone (Heiskanen et al., 1998), Leow and Smith (1994)
showed that noroxycodone provides less antinociception than
oxycodone. These authors also noted that seizures induced by
noroxycodone in rodents are not known to occur in humans,
suggesting that this metabolite may not be formed in signif-
icant quantities in humans. In the present study, N-methyl-
ation status of either hydrocodone or oxycodone did not affect
potency or efficacy to activate G-proteins through ␮- or ␦-opi-
oid receptors, suggesting that this metabolic pathway is not
responsible for their in vivo effects. This conclusion is in
agreement with studies showing that the binding affinity of
norcodeine is not greater than that of codeine (Chen et al.,
1991). Although there was no rationale to examine norco-
deine in the present study due to the well established re-
quirement for O-demethylation of codeine for in vivo effec-
EC50 and Emax values for ␦-opioid receptor-mediated stimulation of
͓
35S͔GTP␥S binding in NG108-15 cells
Data are means Ϯ S.E.M. (n Ն 3) obtained from nonlinear regression analysis of
agonist concentration-effect curves. Emax and EC50 values with common letter des-
ignations were not significantly different from other congeners within the subgroup,
whereas those without common letter designations were significantly different (P Ͻ
0.05, Student’s t test).
Agonist
EC50
Emax
␮M
%Max
Codeine Congeners
Codeine
102.8 Ϯ 13.3a
96.2 Ϯ 37.6a
9.1 Ϯ 1.0b
31 Ϯ 4bc
69 Ϯ 4a
39 Ϯ 4c
21 Ϯ 2b
Dihydrocodeine
Hydrocodone
Oxycodone
8.3 Ϯ 1.7b
Morphine Congeners (O-demethylated
codeine congeners)
Morphine
3.0 Ϯ 1.2*a
14.3 Ϯ 7.7a
1.6 Ϯ 0.37*a
2.2 Ϯ 1.0a
43 Ϯ 2*b
73 Ϯ 14a
29 Ϯ 1*b
33 Ϯ 4*b
Dihydromorphine
Hydromorphone
Oxymorphone
N-Demethylated Codeine Congeners
Nordihydrocodeine
Norhydrocodone
Noroxycodone
N-Demethylated Morphine Congeners
Normorphine
Norhydromorphone
Noroxymorphone
153.5 Ϯ 33.6a
31.3 Ϯ 11.4b
18.2 Ϯ 4.0b
80 Ϯ 11a
41 Ϯ 3b
27 Ϯ 2b
a
a
9.7 Ϯ 1.0†
87 Ϯ 3†
b
1.3 Ϯ 0.33b
3.8 Ϯ 1.1b
56 Ϯ 4†
b
53 Ϯ 2†
%Max, percentage of maximal stimulation.
* P Ͻ 0.05, significantly different from cognate codeine congener.
† P Ͻ 0.05, significantly different from cognate N-methylated congener.
however, only normorphine exhibited a significantly lower
potency (3.2-fold increase in EC50 value compared with mor-
phine).
Finally, the effect on relative efficacy of the basic structure
of each compound (e.g., 6-hydroxy versus 6-keto and 7,8-
saturated versus 7,8-unsaturated) and of O- or N-methyl-
ation status was examined by two-way ANOVA. At the ␮-re-
ceptor, this analysis revealed a significant interaction (F-
ratio ϭ 8.9) for Emax between the basic structure and the
presence or absence of an O- or N-methyl group on each
metabolite. At the ␦-receptor, however, only the basic struc-
ture of the compound significantly affected relative efficacy.
At the ␮-receptor, compounds with the basic codeine struc-
ture (i.e., 7,8-double bond with no keto group) exhibit the
widest range in relative efficacy, with the 3-hydroxy-N-nor
structures reaching near-full agonist levels. For 7,8-satu-
rated compounds with a 6-hydroxy group, the N-methylation
state had less effect, inasmuch as all such congeners clus-
tered above 60% of maximal stimulation. For 7,8-saturated
compounds with a 6-keto group, the methylation state had
little effect on relative efficacy unless both the O- and N-
methyl groups were absent, although the 14-hydroxy-6-keto
structures exhibited their greatest relative efficacy when the
O-methyl group was absent.
Discussion
This study examined ␮- and ␦-opioid receptor-mediated tiveness, the lack of effect of N-methylation status on the
G-protein activation by several codeine congeners and their efficacy of other codeine congeners in this study suggest that
O- and N-demethylated derivatives. Apparent discrepancies codeine and norcodeine would possess similar efficacy, and
between in vivo studies and in vitro binding data on the this will be tested in future studies.
requirement for metabolic activation of dihydrocodeine, hy-
Whereas the apparent lack of requirement for O-demeth-
drocodone, and oxycodone suggested a need to examine ylation of hydrocodone and oxycodone may be explained in
whether receptor binding correlates with G-protein activa- part by their potency, the same explanation seems implausi-
tion by these compounds. Interpretation of the efficacy (Emax
)
ble for dihydrocodeine. Figure 1 shows the structural simi-

G-Protein Activation by Morphine and Codeine Congeners
553
larity between dihydrocodeine and its 7,8-unsaturated form, tributes to the differences in analgesic potency between co-
codeine. Table 1 shows that dihydrocodeine was not more deine and the 7,8-saturated codeine congeners, and this
potent than codeine, which supports binding data indicating question will be addressed in future studies.
that, like codeine, dihydrocodeine has about 120-fold less
In Tables 1 and 2, the potency of morphine congeners at the
affinity for the ␮-receptor compared with its 3-hydroxy ana- ␦-receptor was similar to that of codeine congeners at the
log (Mignat et al., 1995; Schmidt et al., 2002). Again, a ␮-receptor. Although this suggests that metabolism of co-
possible explanation is that in vivo biotransformation such as deine congeners may be even more important for their action
N-demethylation or glucuronidation alters the potency of at ␦- than at ␮-opioid receptors, behavioral studies concluded
dihydrocodeine. In a study by Schmidt et al. (2002), however, that blockage of O-demethylation has no measurable effect
neither nordihydrocodeine nor dihydrocodeine 6-glucuronide on analgesia by dihydrocodeine, hydrocodone, or oxycodone.
exhibited an increase in affinity for ␮-, ␦-, or ␬-receptors. Thus, the data herein suggest that ␦-opioid receptors may not
Similarly, Table 1 shows that nordihydrocodeine exhibited no play an important role in analgesia by these opioids. This
change in potency or efficacy at the ␮-receptor relative to interpretation is in agreement with ␮-receptor knockout
dihydrocodeine. Similar results were obtained at the ␦-opioid studies indicating an absolute requirement for this receptor
receptor, although the relative efficacy of nordihydrocodeine in analgesia and other in vivo effects of morphine (Matthes et
was slightly higher than that for dihydrocodeine. Taken to- al., 1996; Sora et al., 1997; Loh et al., 1998).
gether, these results suggest that the apparent lack of re-
The basic structure and methylation state of the opioids
quirement for O-demethylation of dihydrocodeine in vivo examined in the present study have significant interactions
may be due to the higher efficacy of this compound to activate at the ␮-opioid receptor but not the ␦-receptor, according to
␮- and possibly ␦-opioid receptors relative to that of codeine. two-way ANOVA. These analyses suggest that a 7,8-satu-
Another possible explanation for the ineffectiveness of rated 14-hydroxy derivative of morphine might exhibit in-
P450 inhibitors against dihydrocodeine, hydrocodone, or oxy- creased efficacy at the ␮-receptor, for the following reasons.
codone may be that sufficient quantities of their O-demeth- First, the 7,8-saturated compounds generally exhibit high
ylated metabolites are formed within the central nervous relative efficacy regardless of methylation state. Comparison
system, where P450 inhibitors are less able to block trans- of morphine and dihydromorphine in Table 1 suggests that
formation. P450-mediated reactions within neurons and/or removal of the double bond between C7 and C8 increases
glial cells in ␮-receptor-rich regions might lead to formation relative efficacy without affecting potency. Second, relative to
of demethylated metabolites. Furthermore, such hydrophilic dihydrocodeine and dihydromorphine, the 6-keto compounds
metabolites might increase local bioavailability due to se- have lower relative efficacy, but the 14-hydroxy group of
questration (via the blood-brain barrier) within the central oxymorphone confers a greater Emax value relative to hydro-
nervous system. Jurna et al. (1997), however, showed that morphone. Thus, the possibility that 14-hydroxy-dihydro-
intrathecal infusion of either dihydrocodeine or dihydromor- morphine (Garadnay et al., 2001) is the most efficacious
phine rapidly and naloxone-reversibly inhibits C-fiber firing ␮-agonist of this series will be examined in future studies.
within 5 min postinfusion, suggesting no difference between
the parent compound and metabolite.
In summary, the G-protein activation data herein gener-
ally support previous receptor binding data regarding differ-
An additional potential explanation is agonist-dependent ences in ␮-receptor binding affinities between codeine and
differential activation of G-protein subtypes by the 7,8-satu- morphine congeners. The observation that dihydrocodeine
rated codeine congeners versus codeine. In such a system, the and dihydromorphine exhibit greater potency differences
level of total G-protein stimulation may be irrelevant if spe- than codeine and morphine is surprising given that in vivo
cific G-protein subtypes and analgesia-relevant effector sys- data suggest that only codeine requires O-demethylation by
tems are activated more effectively by certain agonists. This P450 to promote analgesia. The nonrequirement for O-de-
“stimulus trafficking” (Kenakin, 1995) may explain the dif- methylation of dihydrocodeine, hydrocodone, and oxycodone
ferences between in vitro and in vivo studies of codeine ver- indicates that current understanding of the action of these
sus dihydrocodeine, hydrocodone, and oxycodone. Table 1 drugs is incomplete. The present study demonstrates that
indicates that only two of the eight congeners studied in rat one factor these three compounds have in common is greater
thalamic membranes and hMOR-CHO cells exhibited signif- efficacy for ␮-receptor-mediated G-protein activation relative
icant differences in relative efficacy between the two sys- to codeine, which may explain the differential requirement
tems. It is not clear whether these differences are due to for P450 metabolism. Nonetheless, these data also suggest
species-dependent differences in ␮-receptor structure or to that O-demethylation of these 7,8-saturated codeine conge-
the G␣-protein subtypes present in these two tissues. Al- ners would greatly increase their potency for ␮-receptor ac-
though human and rat ␮-opioid receptors share significant tivation, and therefore, blockade of this reaction should in-
sequence identity, G␣o is the predominant G␣ subtype in fluence their potency in vivo. Thus, further research is
brain, whereas G␣i2 is the predominant subtype in CHO cells needed to fully elucidate the relationship between pharma-
(Gettys et al., 1994). The fact that greater relative efficacies cokinetic influences and pharmacodynamic activity of these
of codeine and oxycodone were observed in hMOR-CHO cells clinically relevant drugs.
suggests that these two compounds might preferentially ac-
tivate G␣i2. Interestingly, neither receptor structure nor G␣
References
Abdelhamid EE, Sultana M, Portoghese PS, and Takemori AE (1991) Selective
subtype appeared to affect hydrocodone or dihydrocodeine,
although it is possible that these compounds signal prefer-
entially through other G␣ subtypes not present in CHO cells,
such as G␣i1. In any case, there remains a possibility that
agonist-dependent differential activation of G␣ subtypes con-
blockage of delta opioid receptors prevents the development of morphine tolerance
and dependence in mice. J Pharmacol Exp Ther 258:299–303.
Asano
T and Ross EM (1984) Catecholamine-stimulated guanosine 5Ј-O-(3-
thiotriphosphate) binding to the stimulatory GTP-binding protein of adenylyl
cyclase: kinetic analysis in reconstituted phospholipid vesicles. Biochemistry 23:
5467–5471.

554
Thompson et al.
Caraco Y, Tateishi T, Guengerich FP, and Wood AJ (1996) Microsomal codeine
N-demethylation: cosegregation with cytochrome P4503A4 activity. Drug Metab
Dispos 24:761–764.
reward effect and withdrawal symptoms in mice lacking the mu-opioid-receptor
gene. Nature (Lond) 383:819–823.
Mignat C, Wille U, and Ziegler A (1995) Affinity profiles of morphine, codeine,
dihydrocodeine and their glucuronides at opioid receptor subtypes. Life Sci 56:
793–799.
Chen ZR, Irvine RJ, Somogyi AA, and Bochner F (1991) Mu receptor binding of some
commonly used opioids and their metabolites. Life Sci 48:2165–2171.
Cheng CY, Hsin L, Lin Y, Tao P, and Jong T (1996) N-Cubylmethyl substituted
morphinoids as novel narcotic antagonists. Bioorg Med Chem 4:73–80.
Cleary J, Mikus G, Somogyi A, and Bochner F (1994) The influence of pharmacoge-
netics on opioid analgesia: studies with codeine and oxycodone in the Sprague-
Dawley/Dark Agouti rat model. J Pharmacol Exp Ther 271:1528–1534.
Currie AC, Newbold GT, and Spring FS (1961) 14-Hydroxynorcodeine and its deriv-
atives. J Chem Soc Perkin Trans I:4693–4700.
Dayer P, Desmeules J, Leemann T, and Striberni R (1988) Bioactivation of the
narcotic drug codeine in human liver is mediated by the polymorphic monooxy-
genase catalyzing debrisoquine 4-hydroxylation (cytochrome P-450 dbl/bufI). Bio-
chem Biophys Res Commun 152:411–416.
Garadnay S, Herczegh S, Makleit T, Friedmann P, Riba P, and Fu¨rst S (2001)
Reactions of morphine derivatives with phenyliodo(III)diacetate (PIDA): synthesis
of new morphine analogs. Curr Med Chem 8:621–626.
Gettys TW, Sheriff-Carter K, Moomaw J, Taylor IL, and Raymond JR (1994) Char-
acterization and use of crude ␣-subunit preparations for quantitative immunoblot-
ting of G-proteins. Anal Biochem 220:82–91.
Mikus G, Somogyi AA, Bochner F, and Eichelbaum
M (1991) Codeine O-
demethylation: rat strain differences and the effects of inhibitors. Biochem Phar-
macol 41:757–762.
Nitsche JF, Schuller AG, King MA, Zengh M, Pasternak GW, and Pintar JE (2002)
Genetic dissociation of opiate tolerance and physical dependence in delta-opioid
receptor-1 and preproenkephalin knock-out mice. J Neurosci 22:10906–10913.
Otton SV, Schadel M, Cheung SW, Kaplan HL, Busto UE, and Sellers EM (1993)
CYP2D6 phenotype determines the metabolic conversion of hydrocodone to hydro-
morphone. Clin Pharmacol Ther 54:463–472.
Poyhia R, Seppala T, Olkkola KT, and Kalso E (1992) The pharmacokinetics and
metabolism of oxycodone after intramuscular and oral administration to healthy
subjects. Br J Clin Pharmacol 33:617–621.
Schmidt H, Vormfelde S, Klinder K, Gundert-Remy U, Gleiter CH, Skopp G, Aderjan
R, and Fuhr U (2002) Affinities of dihydrocodeine and its metabolites to opioid
receptors. Pharmacol Toxicol 91:57–63.
Selley DE, Liu Q, and Childers SR (1998) Signal transduction correlates of mu opioid
agonist intrinsic efficacy: receptor-stimulated [
35S]GTPgammaS binding in
Heiskanen T, Olkkola KT, and Kalso E (1998) Effects of blocking CYP2D6 on the
pharmacokinetics and pharmacodynamics of oxycodone. Clin Pharmacol Ther
64:603–611.
Herkenham M and Pert CB (1982) Light microscopic localization of brain opiate
receptors: a general autoradiographic method which preserves tissue quality.
J Neurosci 2:1129–1149.
mMOR-CHO cells and rat thalamus. J Pharmacol Exp Ther 285:496–505.
Selley DE, Sim LJ, Xiao R, Liu Q, and Childers SR (1997) mu-Opioid receptor-
stimulated guanosine-5Ј-O-(gamma-thio)-triphosphate binding in rat thalamus
and cultured cell lines: signal transduction mechanisms underlying agonist effi-
cacy. Mol Pharmacol 51:87–96.
Sim LJ, Selley DE, and Childers SR (1995) In vitro autoradiography of receptor-
activated G proteins in rat brain by agonist-stimulated guanylyl 5Ј-[gamma-
[35S]thio]-triphosphate binding. Proc Natl Acad Sci USA 92:7242–7246.
Sora I, Takahashi N, Funada M, Ujike H, Revay RS, Donovan DM, Miner LL, and
Uhl GR (1997) Opiate receptor knockout mice define mu receptor roles in endog-
enous nociceptive responses and morphine-induced analgesia. Proc Natl Acad Sci
USA 94:1544–1549.
Heyman JS, Vaught JL, Mosberg HI, Haaseth RC, and Porreca F (1989) Modulation
of ␮-mediated antinociception by d agonists in the mouse: selective potentiation of
morphine and normorphine by [D-Pen2, D-Pen5]enkephalin. Eur J Pharmacol
165:1–10.
Hilf G, Gierschik P, and Jakobs KH (1989) Muscarinic acetylcholine receptor-
stimulated binding of guanosine 5Ј-O-(3-thiotriphosphate) to guanine-nucleotide-
binding proteins in cardiac membranes. Eur J Biochem 186:725–731.
Ishida T, Oguri K, and Yoshimura H (1982) A novel metabolite of oxycodone in the
urine of rabbits. J Pharmacobiodyn 5:134–136.
Speyer E and Sarre K (1924) Action of bromocyanogen on hydroxycodeinone and
dihydrohydroxycodeinone. Chem Ber 57:1427–1430.
Jurna I, Komen W, Baldauf J, and Fleischer W (1997) Analgesia by dihydrocodeine
is not due to formation of dihydromorphine: evidence from nociceptive activity in
rat thalamus. J Pharmacol Exp Ther 281:1164–1170.
Kaiko RF, Benziger DP, Fitzmartin RD, Burke BE, Reder RF, and Goldenheim PD
(1996) Pharmacokinetic-pharmacodynamic relationships of controlled-release oxy-
codone. Clin Pharmacol Ther 59:52–61.
Kenakin T (1995) Agonist-receptor efficacy. II. Agonist trafficking of receptor sig-
nals. Trends Pharmacol Sci 16:232–238.
Kirkwood LC, Nation RL, and Somogyi AA (1997) Characterization of the human
cytochrome P450 enzymes involved in the metabolism of dihydrocodeine. Br J Clin
Pharmacol 44:549–555.
Lelas S, Wegert S, Otton SV, Sellers EM, and France CP (1999) Inhibitors of
cytochrome P450 differentially modify discriminative-stimulus and antinocicep-
tive effects of hydrocodone and hydromorphone in rhesus monkeys. Drug Alcohol
Depend 54:239–249.
Leow KP and Smith MT (1994) The antinociceptive potencies of oxycodone, noroxy-
codone and morphine after intracerebroventricular administration to rats. Life Sci
54:1229–1236.
Loh HH, Liu HC, Cavalli A, Yang W, Chen YF, and Wei LN (1998) mu Opioid
receptor knockout in mice: effects on ligand-induced analgesia and morphine
lethality. Brain Res Mol Brain Res 54:321–326.
Szekeres PG and Traynor JR (1997) Delta opioid modulation of the binding of
guanosine-5Ј-O-(3-[35S]thio)triphosphate to NG108-15 cell membranes: character-
ization of agonist and inverse agonist effects. J Pharmacol Exp Ther 283:1276–
1284.
Tomkins DM, Otton SV, Joharchi N, Li NY, Balster RF, Tyndale RF, and Sellers EM
(1997) Effect of cytochrome P450 2D1 inhibition on hydrocodone metabolism and
its behavioral consequences in rats. J Pharmacol Exp Ther 280:1374–1382.
Traynor JR and Nahorski SR (1995) Modulation by ␮-opioid agonists of guanosine-
5Ј-O-(3-[35S]thio)triphosphate binding to membranes from human neuroblastoma
SH-SY5Y cells. Mol Pharmacol 47:848–854.
Weinstein SH and Gaylord JC (1979) Determination of oxycodone in plasma and
identification of a major metabolite. J Pharm Sci 68:527–528.
Wilder-Smith CH, Hufschmid E, and Thormann W (1998) The visceral and somatic
antinociceptive effects of dihydrocodeine and its metabolite, dihydromorphine. A
cross-over study with extensive and quinidine-induced poor metabolizers. Br J
Clin Pharmacol 45:575–581.
Yu A, Kneller BM, Rettie AE, and Haining RL (2002) Expression, purification,
biochemical characterization, and comparative function of human cytochrome
P450 2D6.1, 2D6.2, 2D6.10 and 2D6.17 allelic isoforms. J Pharmacol Exp Ther
303:1291–1300.
Martin TJ, Kim SA, Cannon DG, Sizemore GM, Bian D, Porreca F, and Smith JE
(2000) Antagonism of delta(2)-opioid receptors by naltrindole-5Ј-isothiocyanate
attenuates heroin self-administration but not antinociception in rats. J Pharmacol
Exp Ther 294:975–982.
Matthes HW, Maldonado R, Simonin F, Valverde O, Slowe S, Kitchen I, Befort K,
Dierich A, Le Meur M, Dolle P, et al. (1996) Loss of morphine-induced analgesia,
Address correspondence to: Dr. Dana E. Selley, Department of Pharma-
cology and Toxicology and Institute for Drug and Alcohol Studies, Virginia
Commonwealth University, Medical College of Virginia, Box 980524, 1112
East Clay Street, Richmond, VA 23298-0524. E-mail: deselley@hsc.vcu.edu