Pivaloylcodeine, a new codeine derivative, for the inhibition of morphine glucuronidation. An in vitro study in the rat

Article abstract of DOI:10.1016/j.bmc.2013.09.062

We have previously found that phenanthrenic opioids, including codeine, modulate morphine glucuronidation in the rat. Here codeine and five of its derivatives were compared in their effects on the synthesis of morphine-3-glucuronide (M3G) and morphine-6-glucuronide (M6G) from morphine by rat liver microsomal preparations, and by primary cultures of rat hepatocytes previously incubated for 72 h with either codeine or its derivatives. Acetylcodeine and pivaloylcodeine shared the capability of the parent compound of inhibiting the synthesis of M3G by liver microsomes through a noncompetitive mechanism of action. Their IC₅₀ were 3.25, 2.27, and 4.32 μM, respectively. Dihydrocodeine, acetyldihydrocodeine, and lauroylcodeine were ineffective. In all the experimental circumstances M6G was undetectable in the incubation medium. In primary hepatocyte cultures codeine only inhibited M3G formation, but with a lower efficacy than that observed with microsomes (IC ₅₀ 20.91 vs 4.32 μM). Preliminary results show that at micromolar concentrations codeine derivatives exhibit a low rate of affinity for μ opiate receptors. In conclusion, acetyl and pivaloyl derivatives of codeine noncompetitively inhibit liver glucuronidation of morphine interacting with microsomes. This study further strengths the notion that phenanthrenic opioids can modulate morphine glucuronidation independently from their effects on μ opiate receptors.

Full text of DOI:10.1016/j.bmc.2013.09.062

Bioorganic & Medicinal Chemistry 21 (2013) 7955–7963
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Pivaloylcodeine, a new codeine derivative, for the inhibition
of morphine glucuronidation. An in vitro study in the rat
Letizia Antonilli a, Anna Rita Togna a, Giovanna Sabatini a, Alessandro Venditti b, Laura Guarcini b,
Giuseppina I. Togna a, Rosario Nicoletti b, Filomena Sanasi b, Armandodoriano Bianco b, Paolo Nencini a,
⇑
a Department of Physiology and Pharmacology ‘Vittorio Erspamer’, Sapienza University of Rome, Piazzale Aldo Moro 5, 00185 Rome, Italy
b Department of Chemistry, Sapienza University of Rome, Piazzale Aldo Moro 5, 00185 Rome, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 31 July 2013
Revised 25 September 2013
Accepted 26 September 2013
Available online 19 October 2013
We have previously found that phenanthrenic opioids, including codeine, modulate morphine
glucuronidation in the rat. Here codeine and five of its derivatives were compared in their effects on
the synthesis of morphine-3-glucuronide (M3G) and morphine-6-glucuronide (M6G) from morphine
by rat liver microsomal preparations, and by primary cultures of rat hepatocytes previously incubated
for 72 h with either codeine or its derivatives. Acetylcodeine and pivaloylcodeine shared the capability
of the parent compound of inhibiting the synthesis of M3G by liver microsomes through a noncompet-
Keywords:
Codeine
itive mechanism of action. Their IC50 were 3.25, 2.27, and 4.32 lM, respectively. Dihydrocodeine,
acetyldihydrocodeine, and lauroylcodeine were ineffective. In all the experimental circumstances
M6G was undetectable in the incubation medium. In primary hepatocyte cultures codeine only
inhibited M3G formation, but with a lower efficacy than that observed with microsomes (IC50 20.91
Acetylcodeine
Pivaloylcodeine
Morphine-3-glucuronide
Liver microsomes
Glucuronidation
vs 4.32
lM). Preliminary results show that at micromolar concentrations codeine derivatives exhibit
a low rate of affinity for
l
opiate receptors. In conclusion, acetyl and pivaloyl derivatives of codeine
noncompetitively inhibit liver glucuronidation of morphine interacting with microsomes. This study
further strengths the notion that phenanthrenic opioids can modulate morphine glucuronidation inde-
pendently from their effects on
l opiate receptors.
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
morphine either acutely or chronically.8 We replicated this finding
in the rat showing that repeated exposure to heroin, in condition of
either passive administration (ip) or of iv self-administration, was
associated with the presence in the blood of M6G, which is not
usually formed by this species.9,10 When liver microsomes of her-
oin-treated rats were incubated with increasing concentrations of
morphine, Vmax of M3G formation was significantly reduced with
respect to saline-injected rats, but they yielded a saturable curve
of M6G formation, which was not possible with liver microsomes
obtained from both controls and morphine-injected rats. A like-
wise conducted experiment demonstrated that heroin also inhibits
the formation of estradiol-3-glucuronide, an effect that was shared
by naltrexone.11 When primary rat hepatocytes were incubated
with heroin for 72 h, again M3G formation was inhibited in favour
of the formation of M6G, demonstrating that liver is the main site
of this heroin effect.12 In a further study we observed that the shift
from M3G to M6G formation induced by heroin in isolated hepato-
cytes was associated to an enhancement of UGT2B1 expression and
an inhibition of that of UGT1A1.13
Liver glucuronidation is the main metabolic process whereby
morphine is transformed in the body leading to the formation of
morphine-3-glucuronide (M3G) and morphine-6-glucuronide
(M6G), both of which being more polar than the parent compound
are more easily excreted.1 These metabolites are far from being
inactive. Whereas M6G is a potent
l opioid receptor (MOP) ago-
nist, M3G increases neural excitability contributing to the develop-
ment of both hyperalgesia and allodynia consequent to prolonged
exposure to morphine.2–4
M3G formation is the most preferred, in humans accounting for
about three quarter of the total morphine metabolism whereas
M6G for less than the 20%.1 Considering their opposed effects, fac-
tors influencing the balance in the formation of the two metabo-
lites are of considerable interest. Indeed, exposure to some drugs
has been found to alter M3G/M6G ratio in different experimental
conditions.5–7 Following this vein of research we found that in
chronic intravenous users of street heroin M3G/M6G ratio is signif-
icantly reduced with respect to opioid-naïve subjects treated with
Apparently, heroin-induction of M6G formation by rats is not
mediated by MOP, as far as it is neither replicated by methadone
nor antagonized by naltrexone.14 Although we are not yet able to
provide mechanistic explanations for this heroin effect, it is worth
⇑
Corresponding author. Tel.: +39 06 49912497; fax: +39 06 49912597.
E-mail address: paolo.nencini@uniroma1.it (P. Nencini).
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.09.062

7956
L. Antonilli et al. / Bioorg. Med. Chem. 21 (2013) 7955–7963
MOP can inhibits microglia priming induced by different inflam-
matory stimuli.4 Accordingly, we found that codeine, which is a
very weak opioid agonist, inhibited in vitro M3G formation, along
with reduced UGT2B1 expression at 6 h and a trend toward inhibi-
tion of UGT1A1 expression at 72 h. Summing up, it seems that the
presence of the phenanthrenic backbone structure is required to
make opiates modulators of morphine glucuronidation. This possi-
bility is made plausible by two related findings. First, planar phen-
anthrene inhibits 1-OH-pyrene glucuronidation performed by liver
CYP1A isozymes;15 second, UTG activity towards morphine, 4-
nitrophenol, and 1-naphthol was increased by some phenanthrenic
derivatives with nitrogen-containing heterocycles.16 Hence, bidi-
rectional changes in the glucuronidation process can be obtained
varying phenanthrenic substituents. On the basis of this notion
we argued that differences in the effects of phenanthrenic opiates
on morphine metabolism depend upon different substitutions on
Figure 1. Structures of codeine and dihydrocodeine derivatives.
noting that nonMOP-mediated effects of opioids are not rare
events; for instance, the (+)-naloxone stereoisomer inactive at
Figure 2. (A) Kinetics of M3G formation by rat liver microsomes. The solid line represents the best fit of Michaelis–Menten equation to the data by nonleast-square
regression. Values represent mean S.E. for three separate incubations. Eadie–Hofstee plot is shown as inset on graph. (B) M3G formation by rat liver microsomes in presence
of codeine, acetylcodeine, pivaloylcodeine and dihydrocodeine. Data are represented as% of the basal M3G formation (13.12 0.26 nmol/min/mg) from morphine (conc). Each
point represents the mean S.E of triplicate determinations. (C) Effect of laurylcodeine and acetyldihydrocodeine on M3G formation by rat liver microsomes. ⁄P < 0.05, as
compared with control incubations (no inhibitor added) according to analysis of variance with Bonferroni’s test.

L. Antonilli et al. / Bioorg. Med. Chem. 21 (2013) 7955–7963
7957
the backbone structure of morphine. As known, codeine differs
from morphine in that the 3-OH function is methylated. This
change causes a remarkable reduction in the affinity for MOP17
which is not restored by the insertion of substituents into 6-
OH,18 whereas the 7,8-saturated congener dihydrocodeine main-
tains the same affinity for MOP19 but increases the efficacy in acti-
vating G-proteins.20 Hence we reasoned that any difference that
codeine derivatives have upon morphine glucuronidation would
lend further support to the hypothesis that modulation of glucu-
ronidation exerted by phenanthrenic opioids does not require
MOP binding.
pivaloyl derivative differed from codeine and acetylcodeine in that
it significantly reduced the Vmax in all the range of concentrations
here tested but only increased the Km at the highest concentrations
(10–50 lM). As a result, all three compounds caused a concentra-
tion-related monotonic decrease of the enzymatic clearance of
morphine. The kinetic profile of M3G formation in presence of
increasing concentrations of the three compounds is shown in Fig-
ures 3–5, respectively. Representing the data according to the Ea-
die–Hofstee plot clearly shows that codeine changed the kinetic
of M3G formation from hyperbolic (Michaelis–Menten) to sigmoi-
dal (Fig. 3B), as further confirmed by the Hill co-efficient that at co-
Here we report the results of a study comparing the effects on
morphine glucuronidation of codeine with those produced by five
different derivatives of this naturally occurring opiate (Fig 1).
The study was entirely conducted in vitro. In a first experiment
morphine glucuronidation by rat liver microsomes was studied in
presence of each of the five codeine derivatives. In the second, the
effects on morphine metabolism was studied in primary cultures of
rat hepatocytes.
deine concentrations of 5 and 50 lM positively deviates from the
unity, indicating a cooperative enzymatic reaction (see Table 1).
In order to further characterize the type of inhibition exerted by
codeine on M3G formation, data were represented according to
the Lineweaver–Burk double-reciprocal plot (1/V vs 1/S) (Fig 3C).
When inhibitory concentrations lower or equal to 5.0 lM were
introduced, plots increased their slope according to codeine con-
centrations, converging behind the y axis, hence excluding a com-
petitive mechanism of inhibition (Fig. 3C insert). At codeine
concentrations of 10 and 50 l
M, plots lost linearity showing R2 val-
2. Results
ues of 0.956 and 0.947, respectively.
Figures 4 and 5 show the rates of M3G formation from mor-
phine obtained in presence of different concentrations of acetylco-
deine and pivaloylcodeine, respectively. The acetyl derivative
appeared to preserve the inhibitory pattern shown by the parent
compound, namely, it transformed the kinetics of M3G formation
from hyperbolic to sigmoidal (Fig. 4B) and significantly increased
Hill co-efficient. Again, the Lineweaver–Burk double-reciprocal
plot showed a lack of linearity at the two highest concentrations
2.1. Liver microsomes
When microsomes were incubated with concentrations of mor-
phine varying from 0.1 to 5 mM, M3G formation strictly followed
Michaelis–Menten kinetics (Fig. 2A) with an apparent Km and Vmax
of 0.22 0.02 mM and 13.12 0.26 nmol/min/mg protein, respec-
tively. The inhibitory effects of codeine and of its derivatives on
M3G formation are shown in Figures 2B and C. Codeine, acetylco-
deine and pivaloylcodeine produced a concentration-related inhi-
(10 and 50
l
M), whereas, when plots of concentrations below or
M were only considered, they again converged in a
equal to 5.0
l
bition of M3G formation with IC50 values in the
lM range;
point left to the y axis, excluding a competitive mechanism of inhi-
bition (Fig. 4C insert). Pivaloylcodeine also shifted the kinetics of
M3G formation from hyperbolic to sigmoidal (Fig. 5B) and signifi-
cantly increased the Hill co-efficient. The Lineweaver–Burk double-
reciprocal plot showed a lack of linearity at the two highest con-
dihydrocodeine preserved an inhibitory effect that however was
rather weak plateauing at some 40%, inhibition of M3G synthesis,
whereas lauroylcodeine and acetyldihydrocodeine caused variable
and concentration-independent effects (Fig. 2C). It is worth noting
that M6G was never detectable in the incubation medium either
when microsomes were incubated with morphine alone or in com-
bination with codeine or with its derivatives.
Table 1 shows the kinetic parameters of the inhibitory effects of
codeine, acetylcodeine and pivaloylcodeine, as resulted by adopt-
ing the nonlinear regression model.
centrations (10 and 50 lM).
Table 2 shows values of the Akaike information criterion (AIC)
generated fitting the experimental data of the rate versus [S] at
each fixed inhibitor concentrations into the equations for compet-
itive, uncompetitive, noncompetitive and mixed inhibition.
Accordingly, for pivaloylcodeine the noncompetitive model should
be considered more plausible than the mixed model as indicated
At concentrations varying from 2.5 to 50 lM, both codeine and
acetylcodeine statistically reduced Vmax and increased Km. The
Table 1
Kinetic parameters for M3G formation in rat liver microsomes in presence of codeine and its derivatives
Vmax (nmol/min/mg prot)
Km
(lM)
n
Cl (
ll/min/mg prot)
IC50
(lM)
Ki (lM)
Control
13.12 0.26
0.22 0.02
1.01 0.01
60.69 3.93
Codeine (lM)
1.0
2.5
5.0
10
12.51 0.13
12.40 0.16
8.04 0.12⁄
4.41 0.19⁄
4.06 0.15⁄
0.26 0.01
0.41 0.05⁄
0.43 0.02⁄
0.37 0.02⁄
0.48 0.11⁄
0.97 0.03
0.75 0.02⁄
1.42 0.07⁄
1.17 0.06
1.75 0.16⁄
47.97 0.18
31.11 3.61⁄
18.65 1.05⁄
11.95 0.03⁄
8.35 0.09⁄
4.32
6.90
5.17
2.56
50
Acetylcodeine (lM)
1.0
2.5
5.0
10
9.74 0.23
10.12 0.35⁄
7.92 0.30⁄
3.31 0.17⁄
4.01 0.65⁄
0.26 0.02
0.41 0.01
0.56 0.06⁄
0.35 0.012
0.69 0.07⁄
0.79 0.02
0.91 0.08
0.99 0.01
1.43 0.07⁄
1.49 0.12⁄
37.77 2.05⁄
24.73 0.75⁄
25.01 2.26⁄
9.38 0.75⁄
5.97 0.46⁄
3.25
2.27
50
Pivaloylcodeine (lM)
1.0
2.5
5.0
10
7.03 0.14⁄
6.28 0.10⁄
4.53 0.04⁄
4.63 0.13⁄
3.95 0.18⁄
0.13 0.01
0.22 0.01
0.36 0.02
0.72 0.12⁄
0.62 0.04⁄
1.09 0.03
1.69 0.09⁄
1.02 0.04
1.37 0.08⁄
1.34 0.04⁄
52.74 0.57
29.05 0.75⁄
12.65 0.73⁄
6.88 1.22⁄
6.54 0.23⁄
50

7958
L. Antonilli et al. / Bioorg. Med. Chem. 21 (2013) 7955–7963
Figure 4. Kinetic profile (A) and Eadie-Hofstee plots (B) for M3G formation by rat
liver microsomes in presence of increased acetylcodeine concentrations: 0 (s), 1.0
(d), 2.5 (N), 5.0 ( ), 10 (h) and 50 M (j). (C) Lineweaver–Burk plot (1/V against 1/
morphine) in absence (s) or presence of codeine 1.0 (d), 2.5 (N), 5.0 ( ), 10 (h) and
50 M (j). Each point represents the mean S.E. of triplicate determinations. In the
inset acetylcodeine concentrations up to 5 M are used.
Figure 3. Kinetic profile (A) and Eadie–Hofstee plots (B) for M3G formation by rat
liver microsomes in the presence of increased codeine concentrations: 0 (s), 1.0
D
l
D
(d), 2.5 (N), 5.0 (
morphine) in the absence (s) or presence of codeine 1.0 (d), 2.5 (N), 5.0 (
and 50 M (j). Each point represents the mean S.E. of triplicate determinations.
In the inset codeine concentrations up to 5 M are used.
D
), 10 (h) and 50
l
M (j). (C) Lineweaver–Burk plot (1/V vs 1/
l
D
), 10 (h),
l
l
l
codeine was however somewhat lower than that observed with
microsomes, as assessed by the higher value of IC50 (20.91 vs
4.32 lM). Acetylcodeine and pivaloylcodeine showed a weak con-
centration independent inhibition of M3G formation, whereas
dehydrocodeine and its acetyl congener were almost completely
ineffective.
by the ratio of probability. In the case of both codeine and acetylco-
deine the noncompetitive and the mixed type mechanisms ap-
peared equally plausible (see ratio of probability). However,
when data were submitted to the nested-model method of statis-
tical discrimination,21 the noncompetitive inhibition mechanism
appeared the most preferred kinetic model for all the three
compounds.
2.3. Binding assays
At micromolar concentrations codeine derivatives showed a
2.2. Primary rat hepatocyte cultures
variable yet always weak rate of binding to both
l (MOP) and kap-
pa (KOP) opioid receptor (Table 3). Likewise they showed low affin-
ity for the delta opioid receptor (DOP) with the exception of
pivaloylcodeine which affinity for DOP was quite similar to that
of morphine.
As shown in Figure 6, exposing primary rat hepatocyte cultures
to codeine for 72 h inhibited M3G formation when thereafter cells
were incubated with morphine for 2 h. The inhibitory effect of

L. Antonilli et al. / Bioorg. Med. Chem. 21 (2013) 7955–7963
7959
3. Discussion
Besides confirming our recent observation that codeine inhibits
morphine glucuronidation by both liver microsomes and intact pri-
mary rat hepatocytes, the present study provides the following
new findings: (i) hydrogenation of C-7/C-8 double bond of codeine
or of its derivative acetylcodeine completely or almost completely
extinguished the capability of inhibiting M3G formation; (ii) the
inhibitory efficacy on M3G formation by microsomes is slightly in-
creased by esterifying hydroxyl residue in position 6 with acetyl or
pivaloyl, but not lauroyl, moieties; (iii) codeine as well as both its
acetyl and pivaloyl derivatives shared a common noncompetitive
kinetics of inhibition of M3G formation; (iv) none of the codeine
derivatives was as active as the parent compound in inhibiting
morphine glucuronidation in intact hepatocytes; (v) at micromolar
concentrations codeine derivatives show a low level of binding to
MOP.
Altogether these results provide further support to the hypoth-
esis that the phenanthrenic backbone is required to make opioids
capable to modulate morphine glucuronidation, yet this activity
is influenced by changes in the chemical structure of the phen-
anthrenic compounds. In particular, the presence of the 7,8-double
bond in the phenanthrenic backbone of codeine seems particularly
important, since its saturated congeners dihydrocodeine and acet-
yldihydrocodeine were almost completely devoid of effects on
morphine glucuronidation. Moreover, 6-OH acylation with acetyl
or pivaloyl residues slightly increased the inhibition of M3G forma-
tion, whereas enlarging the acyl substituent to the dimension of
lauroyl chain completely suppressed the inhibitory effect of the
parent compound. This finding is consistent with our previous
observation that the acetylation of both the 3 and 6 hydroxyl func-
tions of morphine (i.e., heroin) increases the inhibitory activity on
M3G formation.9,12,14 Notably, codeine and its acetyl and pivaloyl
derivatives shared the same noncompetitive kinetics of inhibition,
suggesting a common microsomial mechanism of action. Since in a
previous study11 we found that both heroin and naltrexone inhibit
estradiol-3-glucuronide formation, a marker of UGT1A1 activity22
also involved in the morphine glucuronidation,23 we may suggest
that codeine, as well as its acetyl and pivaloyl derivatives, binds
to this UGT isozyme in a noncompetitive way.
The present results also support the hypothesis that the inhibi-
tion of morphine glucuronidation does not depend on MOP binding
which in fact was quite low for all the compounds here studied.
This hypothesis was based on our previous findings that the pip-
erydil opiate methadone is devoid of effects on morphine glucu-
ronidation;14 that naltrexone does not inhibit heroin-induced
switch from M3G to M6G formation but converges with heroin
in inhibiting estradiol-3-glucuronide formation.11 Here we found
that dihydrocodeine was devoid of the inhibitory effects of the
unsaturated congener; yet dihydrocodeine is more efficient than
codeine in stimulating MOP-mediated G-protein activation.20
Moreover, we found that acetylcodeine was somewhat more po-
tent than codeine in inhibiting M3G formation, whereas it is some
Figure 5. Kinetic profile (A) and Eadie–Hofstee plots (B) for M3G formation by rat
liver microsomes in presence of increased pivaloylcodeine concentrations: 0 (s),
1.0 (d), 2.5 (N), 5.0 (
against 1/morphine) in the absence (s) or presence of codeine 1.0 (d), 2.5 (N), 5.0
), 10 (h) and 50 M (j). Each point represents the mean S.E. of triplicate
D), 10 (h) and 50 lM (j). (C) Lineweaver-Burk plot (1/V
(D
l
determinations. In the inset pivaloylcodeine concentrations up to 5 lM are used.
Table 2
Model discrimination analysis for inhibition of M3G formation. K is the number of adjustable model parameters in each model, SS indicates the absolute sum of square,
D
AICc is
the Akaike information criterion differences and aindicates the probability (%) that the given model is correct
SS
D
AICc
Probabilitya (%)
Model
K
cod
ACcod
PIVcod
cod
ACcod
PIVcod
cod
ACcod
PIVcod
Competitive
4
4
4
5
86.45
60.60
48.63
45.72
70.29
50.04
39.11
36.11
96.74
60.86
53.08
52.41
27.43
10.73
0.38
0
29.45
13.14
1.31
0
26.92
4.67
ꢀ1.89
0
<0.01
0.47
45.21
<0.01
0.14
34.19
<0.01
8.80
72.01
Uncompetitive
Noncompetitive
Mixed type

7960
L. Antonilli et al. / Bioorg. Med. Chem. 21 (2013) 7955–7963
Figure 6. M3G formation in primary rat hepatocyte cultures pre-exposed to codeine, acetylcodeine, pivaloylcodeine, dihydrocodeine, and acetyldihydrocodeine at different
concentrations. Each bar represents the mean S.E. of three experiments. ⁄P < 0.05 versus control conditions.
Table 3
glucuronidation by hepatocytes was due to the impossibility of
codeine derivatives to penetrate into the cell is unlikely since both
acetylcodeine and pivaloylcodeine are less polar than the parent
compound. If we compare the effects of repeated exposure of hepa-
tocytes to heroin, morphine, naltrexone and codeine, it appears that
all these phenanthrenic opiates affect morphine glucuronidation
but with substantial differences. In particular, the sole heroin shifts
glucuronidation from M3G to M6G in a way that is not affected by
naltrexone, whereas both morphine and codeine share the inhibi-
tion of M3G formation.9,13,14 We may then argue that acetyl- and
pivaloylcodeine preserve the capability of the parent compound
of inhibiting morphine glucuronidation acting on the enzyme but
not on the expression of UGT genes. As far as the physiological rel-
evance of our findings concern, we have previously observed that
in vivo exposure to analgesic doses of codeine modulates liver
microsomal M3G formation eliciting de novo synthesis of M6G
when microsomes were incubated with morphine.13 Further stud-
ies will be aimed at assessing the effects of codeine derivatives in
the same experimental conditions.
Results of binding assays for Mu Opioid Receptors (MOP), Delta Opioid Receptors
(DOP), and Kappa Opioid Receptors (KOP). Results are expressed as the mean% of
inhibition of basal binding of two different experiments
Compound
Morphine 1
MOP
DOP
KOP
l
lM
M
81
98
19
43
17
40
ꢀ2
5
72
95
18
60
7
95
100
25
66
4
51
3
13
6
Morphine 10
Acetylcodeine 1
Acetylcodeine 10
Dihydrocodeine 1
Dihydrocodeine 10
Acetyldihydrocodeine 1
Acetyldihydrocodeine 10
Pivaloylcodeine 1
Pivaloylcodeine 10
l
l
M
M
l
M
l
M
24
l
l
M
M
ꢀ10
ꢀ14
57
l
M
20
38
0
l
M
l
M
90
42
8
40
Lauroylcodeine 1
Lauroylcodeine 10
l
ꢀ7
M
12
13
four times less active as an analgesic with respect to the parent
compound.24 Hence, again there is no correlation between opioid-
ergic effects and the capability of inhibiting morphine
glucuronidation.
4. Conclusions
Surprisingly enough, codeine only showed full inhibitory effects
on the synthesis of M3G in intact hepatocytes when these cells
were exposed to the drug for 72 h. In a previous study,13 we have
found that codeine significantly reduced hepatocyte expression of
UGT2B1 and caused a trend toward the inhibition of UGT1A1
expression, whereas M3G formation from morphine was abated
in the same extent we found in the present study. Hence, prolonged
exposure of cells to codeine produces some sort of down regulation
of morphine glucuronidation. Why acetyl and pivaloyl derivatives
do not share this effect of the parent compound is a matter of
conjectures. The possibility that their lack of effect on morphine
The present study shows that codeine is an effective inhibitor of
M3G formation from morphine by both liver microsomes and pri-
mary hepatocytes, whereas its pivaloyl and acetyl derivatives de-
void of MOP activity preserve the inhibitory effect of the parent
compound on liver microsomes whereas this effect was rather
weak in the intact hepatocytes. These results strength the notion
that phenanthrenic structure is required to make opioids capable
to modulate morphine glucuronidation. Taking into account that
M3G is a potent activator of microglia,4 inhibition of its synthesis
when morphine is administered may help to control a key step
in the development of hyperlagesia and allodynia. Phenanthrenic

L. Antonilli et al. / Bioorg. Med. Chem. 21 (2013) 7955–7963
7961
derivatives of codeine devoid of MOP activity but inhibitors of M3G
synthesis may be candidate to play a role in this strategy.
The final product was crystallised from ethanol and compared with
an authentic sample.
Acetyldihydrocodeine: was prepared following the Klemenc’s
procedure29 with slight modification. The final product was crys-
tallised from ethanol and spectral data were consistent with those
reported in literature.28
5. Experimental section
5.1. Materials
5.2.2. Synthesis of pivaloylcodeine
Substances and reagents used were of analytical grade, pur-
chased from Sigma-Aldrich (St. Louis, MO, USA) or Carlo Erba Rea-
genti (Milano, Italy), codeine phosphate hemihydrate was
purchased from Salars spa (Como, Italy). Hepatocytes maintenance
medium (William’s E), gentamicin, penicillin/streptomycin and fe-
tal calf serum (FCS) were purchased from Invitrogen (Paisley,
Scotland).
Two hundred mg of codeine phosphate hemihydrate (153.8 mg
of free base, 0.630 10ꢀ3 mol) were poured in a 25 mL reaction flask
with 5 mL of dry pyridine and 0.31 mL of pivaloyl chloride were
added. The reaction flask was purged with argon and the mixture
was heated to 60 °C under magnetic stirring for 17 h. The reaction
was controlled by TLC revealing the spots by UV fluorescence and
by Dragendorf spray reagent.
When reaction was completed, pyridine was eliminated under
N2 flux. Then 10 mL of satd NaHCO3 solution was added and the
mixture was stirred for 30 min. The resulting mixture was ex-
tracted 3 times with 15 mL of diethyl ether. The organic solution
was washed with Brine solution and dried on anhydrous Na2SO4.
After evaporation of organic solvent and crystallization from etha-
nol we recovered 227.0 mg (0.592 10ꢀ3 mol) of pivaloylcodeine (fi-
nal yield 94%).
5.2. Chemistry
5.2.1. Synthesis of acetylcodeine, lauroylcodeine,
dihydrocodeine, acetyldihydrocodeine
The synthesis, purification and structural identification of sev-
eral codeine derivatives yet known as acetylcodeine, lauroylco-
deine, dihydrocodeine, acetyldihydrocodeine, were performed as
reported in the literature and details are below reported.
Acetylcodeine: 619.3 mg of codeine phosphate hemihydrate
(476.4 mg of free base, 1.59 10ꢀ3 mol) were poured in a 10 mL
reaction flask and 1.5 mL of (CH3CO)2O were added. The mixture
was heated to 80 °C under magnetic stirring for 20 h. The reaction
was controlled by TLC revealing the spots by UV fluorescence and
by Dragendorf spray reagent. At completed reaction, the mixture
was added with MeOH to eliminate the excess of acetic anhydride
and then concentrated under N2 flux. 5 mL of satd NaHCO3 solution
was added to form a buffer solution that avoided drastic changes in
pH when a solution 0.8 N of NaOH was added to yield acetyl-co-
deine as free base. The resulting suspension was extracted with
ethyl acetate (3 ꢁ 15 mL). The resulting organic solution, washed
with Brine solution and dried on sodium sulphate was evaporated
recovering 525.5 mg of crude reaction product, which was re-crys-
tallized from ethanol as off-white powder, to gave 522.1 mg (1.53
10ꢀ3 mol) of acetylcodeine (final yield 96.2%). The final product
was analyzed by a spectroscopic method, as well as by 1H NMR,
13C NMR (Supplementary Fig. S1 and S2) and ESI-MS and was con-
sistent with data available in literature.25,26
The spectroscopic data of final product were as follow (Supple-
mentary Fig. S3–S5).
Pivaloylcodeine (2): 1H NMR, CDCl3, 300 MHz: d 6.65 (1H, d
J = 8.1 Hz, H-1),
d 6.52 (1H, d J = 8.1 Hz, H-2), d 5.65 (1H,
dm J = 11.1 Hz, H-7), d 5.42 (1H, dm J = 11.1 Hz, H-8), d 5.11 (2H,
overlapped signals, H-5, H-6), d 3.80 (3H, s, –OCH3), d 3.43 (1H,
m, H-9), d 3.02 (1H, d J = 18.1 Hz, H-16a), d 2.80 (1H, bs, H14), d
2.66 (1H, dd J = 12.4; 4.2 Hz, H-10a), d 2.44 (3H, s, –NCH3), d 2.38
(2H, m, overlapped signals, H-10b, H15a), d 2.06 (1H, td J = 12.6;
5.1 Hz, H-16b), d 1.84 (1H, bd, J = 14.1 Hz, H-15b), d 1.26 (9H, s, –
C(CH3)3).
13C NMR, APT, CDCl3, 75 MHz: d 177.8 (C@O, p iv), d 146.9 (C-4),
d 142.5 (C-3), d 130.2 (C-12), d 129.4 (C-7, CH), d 128.1 (C-8, CH), d
125.6 (C-11), d 119.1 (C-1, CH), d 114.6 (C-2, CH), d 87.6 (C-5, CH), d
67.6 (C-6, CH), d 59.5 (C-9, CH), d 56.8 (–OCH3), d 46.9 (C-16, CH2), d
42.6 (C-13), d 42.1 (–NCH3), d 39.5 (C-14, CH), d 38.7 (C(CH3) 3 piv.),
d 34.4 (C-15, CH2), d 27.1 (C(CH3)3 piv.), d 20.8 (C-10, CH2).
ESI-MS: m/z 384.17 [M+H]+; m/z 406.15 [M+Na]+.
5.2.3. Hydrochloride derivative of pivaloylcodeine
Acetylcodeine: 1H NMR, CDCl3, 300 MHz: d 6.66 (1H, d, J 8.2 Hz,
H-1), d 6.55 (1H, d, J 8.2 Hz, H-2), d 5.64 (1H, dm, J 10.1 Hz, H-7), d
5.43 (1H, dm, J 10.1 Hz, H-8), d 5.18 (1H, m, H-6), d 5.08 (1H, dd, J
6.6–0.9 Hz, H-5), d 3.85 (3H, s, –OCH3), d 3.43 (1H, m, H-9), d 3.05
(1H, d, J 18.6 Hz, H-16a), d 2.86 (1H, bs, H-14), d 2.68 (1H, dd, J 12.6
- 4 Hz, H-10a), d 2.49 (3H, s, N-CH3), d 2.42 (1H, m overlapped, H-
16b), d 2.41 (1H, m overlapped, H-10b), d 2.15 (3H, s, CH3CO–), d
1.88 (1H, dm, J 10.8 Hz, H-15).
The pivaloylcodeine as free base was solubilized in diethylether
and gaseous anhydrous HCl was fluxed into the solution with
hydrochloride formation as off white precipitate. The ethereal
solution was filtered on black ribbon filter paper. The filtrate was
washed with diethylether until neutral reaction. The product was
dried under reduced pressure at room temperature and stored in
sealed vials in the dark at 4 °C. The yield of the reaction was
93.6%. The final product was analyzed by a spectroscopic method,
as well as by 1H NMR, 13C NMR (Fig. S6 and S7)
13C NMR, APT, CDCl3, 75 MHz: d 170.5 (C@O, Ac), d 146.8 (C-4), d
142.2 (C-3), d 130.7 (C-12), d 129.6 (C-7, CH), d 128.4 (C-8, CH), d
127.1 (C-11), d 119.1 (C-1, CH), d 114.1 (C-2, CH), d 88.1 (C-5,
CH), d 68.3 (C-6, CH), d 59.1 (C-9, –CH), d 56.7 (C-18, –OCH3), d
46.7 (C-16, CH2), d 43.0 (C17, CH3), d 42.8 (C-13), d 40.7 (C-14,
CH), d 35.4 (C-15, CH2), d 20.8 (C-10, CH2), d 20.4 (C-20, CH3).
ESI-MS: m/z 342.16 [M+H]+; m/z 364.13 [M+Na]+; m/z 380.14
[M+K]+.
Lauroylcodeine: was prepared following the procedure reported
on the patent by LRBL (Laboratoires de Recherches Biologiques
Laborec)27 with slight modifications. The final product was crystal-
lised from ethanol.
1H NMR (400 MHz, D2O) d 6.94 (1H, d, J = 8.3 Hz, H-1), 6.84 (1H,
d, J = 8.1 Hz, H-2), 5.80 (1H, d, J = 9.7 Hz, H-7), 5.62 (1H, d,
J = 10.0 Hz, H-8), 5.32 (1H, d, J = 6.4 Hz, H-6), 5.24 (1H, d,
J = 6.3 Hz, H-5) 4.26 (1H, bs, H-16a), 3.83 (3H, s, –OCH3), 3.47–
3.11 (4H, m, overlapped signals, H-16b; H-14; H-10a; H-9), 3.02
(3H, s, N-CH3), 2.95 (1H, d, J = 6.5 Hz, H-15a), 2.34 (1H, m, H-
15b), 2.16 (1H, m, H-10b), 1.24 (9H, s, H–), 1.21 (s, 3H), 1.12 (d,
J = 9.1 Hz, C(CH3)3).
13C NMR (100 MHz, D2O) d 180.69 (C@O, piv), 145.33 (C-4),
142.31 (C-3), 129.78 (C-8), 128.42 (C-12), 125.96 (C-7), 123.61
(C-11), 120.70 (C-1), 114.73 (C-2), 86.45 (C-5), 67.05 (C-6), 60.85
(C-9), 56.34 (C-18), 47.72 (C-16), 40.91 (C-17), 40.43 (C-13),
38.51 (C-20), 37.94 (C-14), 32.29 (C-11), 26.30 (C-21), 20.61 (C-15).
Dihydrocodeine: was prepared by reduction of codeine with Ra-
ney Ni, following the procedure described by Ramanathan et al.28

7962
L. Antonilli et al. / Bioorg. Med. Chem. 21 (2013) 7955–7963
5.3. Biological evaluation
cytes.12,32 At the end of the 72 h incubation, medium was removed
and substituted with fresh medium. After 10 min, hepatocytes were
incubated in medium containing morphine, as substrate, at the final
concentrations of 0.25 mM for 120 min. Control samples consisted
in cells incubated with morphine (as substrate) but not previously
exposed to codeine derivative and in cells incubated with codeine
derivatives but without morphine as substrate.
All procedures were in accordance with the Italian Law on Ani-
mal Research (DLGS 116/92) and with the guidelines for the care
and use of laboratory animals issued by Italian Ministry of Health.
5.3.1. Microsomes preparation
Liver microsomes were prepared from male Sprague Dawley
rats (200 g) as previously described.9 Briefly, tissues were minced
and rinsed in ice-cold 1.15% KCl and homogenized in 3 volumes
of 100 mM phosphate buffer (pH 7.4) containing 0.25 M sucrose.
The homogenate was centrifuged for 20 min at 9000g. The super-
natant was further centrifuged for 60 min at 105,000g. The result-
ing microsomal pellet was resuspended in 100 mM phosphate
buffer containing 0.25 M sucrose, to obtain a final protein concen-
tration of 1 mg/mL.
Incubation was then stopped adding 0.5 mL of ice-cold acetoni-
trile. The samples were then centrifuged for 5 min at 1400g and the
resulting supernatants were injected into HPLC system.
5.3.3. Kinetic studies
The apparent Michaelis–Menten values (Km and Vmax) of M3G
formation in basal conditions were determined using rat liver
microsomes as the enzyme source. The activity was measured at
seven morphine concentrations ranging from 0.1 to 5 mM. Poten-
tial inhibitors (codeine, acetylcodeine, dihydrocodeine, acety-
ldihydrocodeine, pivaloylcodeine and lauroylcodeine) were tested
at a fixed morphine concentration of 0.2 mM. The inhibitory effects
on M3G formation of codeine and its derivatives were investigated
5.3.1.1. Glucuronidation assays.
The morphine glucuronida-
tion assay was performed as previously described.14 In detail,
microsomal preparations were suspended in 100 mM phosphate
buffer (pH 7.4) to a final protein concentration of 0.5 mg/mL.
Microsomes were pre-incubated for 20 min in 0.05% deoxycholic
acid at 5 °C to achieve full enzymatic activity. Morphine concentra-
tions ranged from 0.1 to 6 mM for the calculation of M3G kinetics,
and from 0.1 to 20 mM for the calculation of M6G kinetics. The
incubation mixture consisted of 5 mM UDP-glucuronic acid (UDP-
GA), 100 mM phosphate buffer (pH 7.4), microsomes, codeine
at concentrations of 1.0, 2.5, 5.0, 10 and 50 lM. Control incubations
were carried out by omitting the substrate. Incubation was per-
formed in triplicate for IC50 and Ki determinations.
5.3.3.1. Data analysis.
Individual data points represent the
mean of triplicate estimates. Kinetic data for M3G formation were
analyzed by fitting untransformed experimental data to the
Michaelis–Menten (Eq.1) and Hill equation (Eq. (2)):
derivatives (0 to 40 lM) and substrate (morphine) to a final vol-
ume of 1 ml. The reaction was started adding UDPGA. Samples
and blanks (without UDPGA) were incubated in triplicates at
37 °C for 30 min and the reaction was stopped with 0.5 mL of ace-
tonitrile and centrifuged. All sample were kept at 4 °C for 15 min;
then they were centrifuged for 10 min at 5800g. Supernatants were
kept at 4 °C until HPLC analysis.
v
v
¼ Vmax½Sꢂ=Km þ ½Sꢂ
ð1Þ
ð2Þ
n
n
¼ Vmax½Sꢂ =Knm þ ½Sꢂ
Vmax, Km and n in Eq.1 and 2 were defined as the maximum
velocity, the substrate concentration at which velocity is equal to
half of the maximum and the Hill co-efficient, respectively.
The appropriate model was selected by (i) visual inspection of
data plots (Michaelis–Menten and Eadie–Hofstee), (ii) distribution
of the residuals, (iii) size of the sum of the squared residuals and
(iv) the standard error of the estimates. Kinetic parameters were
estimated by nonlinear regression analysis with GraphPad Prism
5 (GraphPad Inc., software. San Diego, CA, USA). The concentration
of the inhibitor that is required to produce 50% inhibition of the
enzymatic activity (IC50) was determined by linear interpolation
from the curves obtained by plotting enzymatic activity versus
inhibitor concentrations.
To determine both the model of inhibition produced by either
codeine or its derivatives and the enzyme–inhibitor dissociation
constant (Ki), the experimental data of the rate versus [S] at each
fixed inhibitor concentrations were fit into the equations for com-
petitive (Eq. (3)), uncompetitive (Eq. (4)), noncompetitive (Eq. (5))
and mixed inhibition (Eq. (6)), respectively.33
5.3.2. Hepatocyte isolation
Hepatocytes were obtained from male Sprague-Dawley rats,
according to the procedure described by Seglen30 with some mod-
ifications.31 Briefly, rats were anesthetized with ZoletilÒ (tileta-
mine plus zolazepam ip 40 mg/kg) and their livers were perfused
in situ with warm Ca++ free Hanks buffer modified, supplemented
with albumin (2%) and EGTA (0.6 mM), at a flow rate of 20 mL/
min for 7 min at least. Then the livers were perfused with Hanks
buffer containing collagenase (130 U/mL) and CaCl2 (5.2 mM) for
approximately 6–7 min at a flow rate of 25 mL/min. The hepato-
cytes were released from the digested liver and suspended in
Krebs–Henseleit buffer containing glucose 0.1%. Within 2 min, at
room temperature, the dispersed cells were filtered and centri-
fuged (45g) at 6 °C for 3 min and then resuspended in Krebs–
Henseleit buffer without glucose. The cell viability was determined
by Trypan blue exclusion: preparations that exhibit less than 80%
viability were excluded from experimental studies.
v
v
v
v
¼ Vmax½Sꢂ=½Sꢂ þ Kmð1 þ ½Iꢂ=KiÞ
ð3Þ
ð4Þ
ð5Þ
ð6Þ
5.3.2.1. Hepatocyte culture.
William’s Medium E containing 10% FCS, insulin (0.02 U/mL), gen-
Hepatocytes, suspended in the
¼ Vmax½Sꢂ=½Sꢂð1 þ ½Iꢂ=KiÞ þ Km
tamicin (0.065 mg/mL), penicillin/streptomycin (100 U/100 lg/
mL), were distributed in 24 well plates at the density of 6.5 105
cells/mL and maintained at 37 °C in a humidified atmosphere of
95% air/5% CO2. Two hr after plating and every day thereafter, med-
ium was changed.
¼ Vmax½Sꢂ=½Sꢂð1 þ ½Iꢂ=KiÞ þ Kmð1 þ ½Iꢂ=KiÞ
¼ Vmax½Sꢂ=½Sꢂð1 þ ½Iꢂ= KiÞ þ Kmð1 þ ½Iꢂ=KiÞ
a
The most appropriated model was selected considering both the
After the initial 2 h incubation with medium alone, cells were
cultured in triplicate for 72 h with medium containing one of the fol-
lowing: codeine, acetylcodeine and pivaloylcodeine (5, 10, 20 or
40 lM); dihydrocodeine and acetyldihydrocodeine (5.0 or 10 lM).
Seventy-two hours of incubation have been found to allow the max-
F-test (P < 0.05) and the Akaike Information Criterion (AIC). Gener-
ated by fitting all data sets simultaneously, AIC is a metric that
ranks different models in such a way that lower is the value better
is the model.34 It is assumed that AICc values among 0 and 2 pro-
vide the best support to the model.
imal induction of glucuronidation activity in cultured hepato-

L. Antonilli et al. / Bioorg. Med. Chem. 21 (2013) 7955–7963
7963
4. Lewis, S. S.; Hutchinson, M. R.; Rezvani, N.; Loram, L. C.; Zhang, Y.; Maier, S. F.;
Rice, K. C.; Watkins, L. R. Neuroscience 2010, 165, 569.
5. Aasmundstad, T. A.; Ripel, A.; Bodd, E.; Bjørneboe, A.; Mørland, J. Biochem.
Pharmacol. 1993, 46, 961.
6. Aasmundstad, T. A.; Størset, P. Hum. Exp. Toxicol. 1998, 17, 347.
7. Morrish, G. A.; Foster, D. J.; Somogyi, A. A. Br. J. Clin. Pharmacol. 2006, 61, 326.
8. Antonilli, L.; Semeraro, F.; Suriano, C.; Signore, L.; Nencini, P.
Psychopharmacology 2003, 170, 200.
5.3.4. Binding studies
In vitro binding assays were performed using CHO cells for both
DOP and KOP receptors, and HEK-293 cells for MOP. [3H]DADLE
0.5 nM was the agonist radioligand for DOP, [3H]U-69593 1 nM
was the agonist radioligand for KOP and [3H]diprenorphine
0.4 nM was the antagonist radioligand for MOP. Nonspecific bind-
9. Antonilli, L.; Suriano, C.; Paolone, G.; Badiani, A.; Nencini, P. J. Pharmacol. Exp.
Ther. 2003, 307, 651.
10. Meringolo, M.; Brusadin, V.; De Luca, M. T.; Montanari, C. L.; Antonilli, L.;
Nencini, P.; Badiani, A. Psychopharmacology 2012, 221, 195.
11. Antonilli, L.; Brusadin, V.; Milella, M. S.; Sobrero, F.; Badiani, A.; Nencini, P.
Biochem. Pharmacol. 2008, 76, 672.
12. Graziani, M.; Antonilli, L.; Togna, A. R.; Brusadin, V.; Viola, S.; Togna, G.;
Badiani, A.; Nencini, P. Drug Alcohol Depend. 2008, 98, 179.
13. Antonilli, L.; De Carolis, L.; Brusadin, V.; Togna, A. R.; Dovizio, M.; Togna, G. I.;
Patrignani, P.; Nencini, P. Eur. J. Pharmacol. 2012, 693, 7.
14. Antonilli, L.; Petecchia, E.; Caprioli, D.; Badiani, A.; Nencini, P.
Psychopharmacology 2005, 182, 58.
15. Elovaara, E.; Mikkola, J.; Stockmann-Juvala, H.; Luukkanen, L.; Keski-Hynnilä,
H.; Kostiainen, R.; Pasanen, M.; Pelkonen, O.; Vainio, H. Arch. Toxicol. 2007, 81,
169.
16. Franklin, M. R.; Slawson, M. H.; Moody, D. E. Xenobiotica 1993, 23, 267.
17. Mignat, C.; Wille, U.; Ziegler, A. Life Sci. 1995, 56, 793.
18. Crooks, P. A.; Kottayil, S. G.; Al-Ghananeem, A. M.; Byrn, S. R.; Butterfield, D. A.
Bioorg. Med. Chem. Lett. 2006, 16, 4291.
19. Schmidt, H.; Vormfelde, S.; Klinder, K.; Gundert-Remy, U.; Gleiter, C. H.; Skopp,
G.; Aderjan, R.; Fuhr, U. Pharmacol. Toxicol. 2002, 91, 57.
20. Thompson, C. M.; Wojno, H.; Greiner, E.; May, E. L.; Rice, K. C.; Selley, D. E. J.
Pharmacol. Exp. Ther. 2004, 308, 547.
ing was determined by using naltrexone 10 or 1
lM for DOP or
MOP, respectively, and naloxone (10 M) for KOP. Incubation
l
was performed at 22 °C and lasted 120 min. The specific ligand
binding to the receptors was defined as the difference between
the total binding and the nonspecific binding determined in the
presence of an excess of unlabelled ligand. Since the study was
exploratory, only two different concentrations of the codeine
derivatives or morphine as the reference compound were tested.
The experiment was replicated once. In these conditions it was
not possible calculate Ki values and the results were hence ex-
pressed as a percent inhibition of control specific binding obtained
in the presence of the test compounds.
5.3.5. Statistical analysis
One-way ANOVA was used to assess differences in Vmax, Km and
Hill co-efficient (n) for M3G formation between data obtained in
absence and in presence of the inhibitors. A multiple comparison
corrected by Bonferroni’s test was used. The criterion for statistical
significance was P < 0.05.
21. Mannervik, M. Methods Enzymol. 1982, 87, 370.
22. Williams, J. A.; Ring, B. J.; Cantrell, V. E.; Campanale, K.; Jones, D. R.; Hall, S. D.;
Wrighton, S. A. Drug Metab. Dispos. 2002, 30, 1266.
23. King, C. D.; Rios, G. R.; Green, M. D.; MacKenzie, P. I.; Tephly, T. R. Drug Metab.
Dispos. 1997, 25, 251.
Acknowledgments
24. Milo, S.; Ansonoff, M.; King, M.; Rossi, G. C.; Zuckerman, A.; Pintar, J.; Pasternak,
G. W. Cell. Mol. Neurobiol. 2006, 26, 1011.
This study has been supported by an intramural fund of Sapien-
za University of Rome (Finanziamento progetti d’Ateneo).
25. Batterham, T. J.; Bell, K. H.; Weiss, U. Aust. J. Chem. 1965, 18, 1799.
26. Steiger, C.; Finnema, S. J.; Raus, L.; Schou, M.; Nakao, R.; Suzuki, K.; Pike, V. W.;
Wikström, H. V.; Halldin, C. J. Labelled Compd. Rad. 2009, 52, 58.
27. Laboratoires de Recherches Biologiques Laborec, 1962, Hydroxylated alkaloid
derivatives, FR 77201 19620514.
Supplementary data
28. Ramanathan, V. S.; Rao, S. N.; Ramachandran, C. Indian J. Technol. 1964, 2, 350.
29. Klemenc, S. Forensic Sci. Inter. 2002, 129, 194.
30. Seglen, P. O. Methods Cell Biol. 1976, 13, 29.
31. Moldéus, P.; Högberg, J.; Orrenius, S. Methods Enzymol. 1978, 52, 60.
32. Soars, M. G.; Petullo, D. M.; Eckstein, J. A.; Kasper, S. C.; Wrighton, S. A. Drug
Metab. Dispos. 2004, 32, 140.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2013.09.062.
References and notes
33. Copeland, R. A. Enzymes: A Practical Introduction to Structure, Mechanism, and
Data Analysis; Wiley: New York, 2000.
[34]. Burnham, K. P.; Anderson, D. R. Model Selection and Multimodel Inference: A
Practical Information-Theoretic Approach; Springer-Verlag: New York, 2002.
1. Milne, R. W.; Nation, R. L.; Somogyi, A. A. Drug Metab. Rev. 1996, 28, 345.
2. Watkins, L. R.; Hutchinson, M. R.; Johnston, I. N.; Maier, S. F. Trends Neurosci.
2005, 28, 661.
3. Komatsu, T.; Sakurada, S.; Katsuyama, S.; Sanai, K.; Sakurada, T. Int. Rev.
Neurobiol. 2009, 85, 207.