A highly toxic morphine-3-glucuronide derivative

Article abstract of DOI:10.1016/j.bmcl.2003.12.016

By the coupling of octylamine to the uronic acid function of morphine-3-glucuronide (M3G) a new glycoconjugate (morphine-3- octylglucuronamide, M3GOAM) was prepared. When assayed in both rats and mice up to ng/kg (ip) doses none of the animals survived. The aliphatic octyl chain may be the lethal factor since a closely related derivative (M3GNH2), was not toxic and showed similar opioid antagonist properties than naloxone.

Full text of DOI:10.1016/j.bmcl.2003.12.016

Bioorganic & Medicinal Chemistry Letters 14 (2004) 905–908
A highly toxic morphine-3-glucuronide derivative
Mariona Salvatella,a Gemma Arsequell,a
Gregorio Valenciaa,* and Raquel E. Rodrıguezb
a
´
Unit of Glycoconjugate Chemistry, Instituto de Investigaciones Quımicas y Ambientales (I.I.Q.A.B.-C.S.I.C.),
Jordi Girona 18-26, 08034-Barcelona, Spain
bDepartment of Biochemistry and Molecular Biology, Faculty of Medicine, University of Salamanca,
Avda. Campo Charro s/n, 37007-Salamanca, Spain
Received 26May 2003; revised 13 October 2003; accepted 3 December 2003
Abstract—By the coupling of octylamine to the uronic acid function of morphine-3-glucuronide (M3G) a new glycoconjugate
(morphine-3-octylglucuronamide, M3GOAM) was prepared. When assayed in both rats and mice up to ng/kg (ip) doses none of
the animals survived. The aliphatic octyl chain may be the lethal factor since a closely related derivative (M3GNH2), was not toxic
and showed similar opioid antagonist properties than naloxone.
# 2004 Elsevier Ltd. All rights reserved.
Owing to its wide and long standing use, the effects of
morphine are well characterized. However, the
mechanisms by which morphine produces these effects
are not clear. In recent years the potential role of mor-
phine metabolites in such mechanisms has also been
addressed.
of both morphine and M6G when assayed in rats.13,14
However, such claims of functional antagonism are in
contradiction with other reported findings.15ꢀ17
Controversial hypothesis and results do also concern
blood–brain barrier (BBB) penetration by morphine
glucuronides. Thus, while molecular modelling stud-
ies18,19 provide support for the observation that these
glycoconjugates apparently cross the BBB more easily
than anticipated, other investigations of transport
across the BBB20,21 are consistent with their high polarity
and poor BBB diffusion.
After administration in man, the major fate of mor-
phine is glucuronidation.1 Indeed, 70% of a morphine
dose is normally converted into morphine glycoconju-
gates. Among these, by far the most abundant, is mor-
phine-3-glucuronide (M3G) accounting for a 45–55% of
the converted morphine, while a small proportion of 10–
15% corresponds to morphine-6-glucuronide (M6G).2,3
With the aim of contributing to a better understanding
of the endogenous mechanisms describing opioid activ-
ity mediated by morphine glucuronides, we have con-
ducted structure–activity relationship studies on
morphine glucuronides to better ascertain the role of the
sugar molecule on opioid activity.22 Focusing on M3G
and in line with a preceeding study of 3-substituted
synthetic opioid antagonists of nalmefene and naltrex-
one glycoconjugates,23 we have synthesized a series of
M3G analogues by replacing the glucuronic acid moiety
on M3G by simple monosaccharide units (i.e., glucose,
galactose). After biological in vivo testing of the com-
pounds biological responses were not different from
those of M3G (see below).
Morphine glucuronides are probably one of the best
examples showing that glucuronidation not always ter-
minates the pharmacological activity of medicinal
drugs.4 It is well documented that M6G is a potent
agonist of a m-opioid receptor subtype.5 Antinocicep-
tion caused by M6G has been reported in animal mod-
els6ꢀ8 and also analgesia has been observed in
humans.9,10 On the other hand, M3G shows nor neu-
roexcitatory nor opioid effects in man11,12 but some
studies have reported to possibly antagonize the effects
Keywords: Morphine glucuronides; M3G; Opioid antagonists; Alkyl
glycosides.
These studies also aimed at possibly improve M3G
diffusion across the BBB by enhancing its hydrophobicity.
* Corresponding author. Tel.: +34-934006113; fax: +34-932045904;
e-mail: gvpqbp@iiqab.csic.es
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.12.016

906
M. Salvatella et al. / Bioorg. Med. Chem. Lett. 14 (2004) 905–908
Thus, following our own previously described strategy
for opioid peptides24 we have synthesized a lipophilic
M3G analogue (Fig. 1). The simplest route was to form
an amide bond between the carboxylic acid group on
M3G and a primary lineal alkyl amine. In doing so,
octyl amine was chosen as the shortest but yet hydro-
phobic amine and the uronium reagent, HBTU, because
avoided the use of hydroxyl protecting groups25
Accordingly, the whole procedure for M3GOAM
synthesis started following a one-pot reaction method to
prepare M3G. This glycosidation reaction departed
from morphine and methyl (2,3,4-tri-O-acetyl-a-d-glu-
copyranosyl bromide) uronate as substrates and used
LiOH as promotor and base for the removal of acetyl
protecting groups.26 In a second step, the coupling
reaction between the glucuronide M3G and octylamine
was accomplished by means of the uronium salt HBTU
[2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium
hexafluorophosphate].27 Similarly, the new analogue
morphine-3-glucuronamide (M3GNH2) was prepared
from morphine and the corresponding acetylated glu-
curonamide bromide derivative also using the LiOH
method. The morphine glucuronide derivatives were
purified by column chromatography and crystallization.
Their purity was assessed by RP-HPLC and their
morphine and saline groups presented the expected
biological activity but M3GOAM treated animals died
within the next 30 min.
It is remarkable that the toxic effect persisted even at a
such low dosis of 1 ng/kg. The effect first appeared as
flaccidity of both fore and hind paws followed by pro-
found sleep and abrupt cardiorespiratory stop. These
results were confirmed by a 10 day study of the beha-
viour of 10 groups of three animals each one receiving
either 1 ng/kg of M3GOAM, 5 mg/kg of morphine or
saline. Everyday one group was examined and similar
death rate and biological effects could be observed.
Experiments were also conducted to also confirm that
the toxic effect was neither due to a particular batch of
product or lot of animals. Results with a new batch
of product matched previous observations in a group of
n=5 rats receiving 1ng/kg of M3GOAM and a control
group (n=4) of morphine. In addition, to investigate if
the time period between administration and death was
always constant a group of n=12 was inoculated with a
1 ng/kg dosis. The survival times recorded for n=5 were
less than 30 min, a group of n=4 did not lasted 45 min
and the rest (n=3) not survived 90 min.
1
structural characterization28 performed by H and 13C
To get some insight about the toxicity mechanism the
experiments outlined in Table 2 involving the use of the
opiod antagonist naloxone were performed. We wanted
to check if by prior receptor ocupancy or by displace-
ment of receptor bound M3GOAM, naloxone could
either prevent or revert the toxic effect.
NMR and ESI-MS.
The potential analgesic properties of M3GNH2 and
M3GOAM, were studied on the tail flick and paw
pressure behavioural tests29 using as reference com-
pounds morphine and both commercial and our own
synthesized M3G.
None of the experiments with naloxone provided
improvement in survival times and allowed us to
conclude that opioid receptors seem not to participate
in the toxicity mechanism of M3GOAM.
1. Biological activity of M3GOAM
Table 1. Effect of M3GOAM doses containing decreasing amounts
of the compound
Knowing that an ip dose of 5 mg/kg of morphine nor-
mally produces 65% of analgesia in rats, the initial
assay of M3GOAM was conducted at this same dose in
a group of n=10. The observation that the animals died
after 10 min let us supose that the dose was too exces-
sive. In a second series of experiments a 0.1 mg/kg dose
was assayed in a group of n=35. All animals seemed to
present 100% of analgesia in both thermal and
mechanical tests but only survived between 35 and 45
min after treatment. Looking for a dose that was
analgesic but not toxic a third set of experiments
including lower doses and control groups as to work
blindly was performed. Table 1 shows that the
Compd
Dose (mg/kg)
n
Effect
M3GOAM
5
2.5
8
4
Death
Death
1
0.5
0.05
0.005
(1 ng/kg)
5
6Death
6Death
6Death
3
6Death
5
Death
Morphine
Saline
60–70%a
Pain
12
a Values of analgesia.
Table 2. Experiments attempting to revert the toxic effect by pre- or
post-administration of the opioid antagonist naloxonea
First dose
Second dose
Interval (min)
n
Naloxone
Naloxone
Naloxone
M3GOAM
M3GOAM
M3GOAM
M3GOAM
M3GOAM
M3GOAM
M3GOAM
Naloxone
Naloxone
Naloxone
Naloxone
10
15
20
10
15
20
30
5
5
5
5
5
5
5
a Naloxone was assayed at three different doses (1, 5, 10 ng/kg) while
M3GOAM at only one (1 ng/kg).
Figure 1. Chemical structures of morphine glycoconjugates: M3G,
M3GNH2 and M3GOAM.

M. Salvatella et al. / Bioorg. Med. Chem. Lett. 14 (2004) 905–908
907
Possible interspecies differences were investigated by
administering ip doses of 1 ng/kg of the compound to
mice. No analgesia was observed as measured by both
tail flick and paw pressure tests. All animals (n=5) died,
however, survival times (2 h) were longer than with rats.
These results were confirmed with a second group
(n=4) showing survival times on the range of 1 to 4 h.
Table 3. Opioid activity of in house prepared and commercial M3G
Compd
Dose (mg/kg)
n
Effect
Synthetic M3G
1
5
10
1
15
20
20
15
5
None
None
None
None
None
Commercial M3G
5
10
5
5
7
20
None
60–70%a
None
Morphine
Saline
2.. Biological activity of M3G
a Values of analgesia.
Validation of the synthetic methods and the biological
tests was carried out by comparing the activity of com-
mercially available M3G and our own prepared pro-
duct. Because reported activities for M3G are of opioid
antagonist, it was first check that the products were
devoid of agonist activity by working blindly and using
the protocol of Table 3.
Table 4. Opioid antagonist activity of M3G and naloxone by per-
ipheral administrationa
First dose
Second dose
Interval (min)
n
M3G
Naloxone
Morphine
Morphine
20
20
5
5
a In all cases M3G and naloxone were administered at three different
doses (1, 5, 30 mg/kg) while morphine at a reference dose of 5 mg/kg.
The opioid antagonist activity of M3G was assessed
against morphine and its potency compared to nalox-
one. Three doses were studied using the protocol of
Table 4.
produced a highly toxic compound. The result is more
striking because similar strategies (addition of octyl
moieties) when applied to other sugar derivatives have
provided promising drug candidates.30 Due to such high
toxicity it is, up to now, unclear if the compound is able
to induce analgesia. However, since the observation that
the lethal effect can not be reverted or prevented by
naloxone it may seem to indicate that the toxicity
mechanism does not involve interactions at the opioid
receptor level. This fact may also seem supported by the
lack of significative biological activity differences among
the two animal species tested.31
Morphine was able to elicit analgesia only in the group
of n=15 which received M3G probing that M3G does
not act as opioid antagonist by peripheral administra-
tion. However, iv injection of 1 mL of a 5 mg/mL solu-
tion of M3G in the IV ventricle in a group of n=10 rats
and after 20 min of ip dosis of 5 mg/kg of morphine,
masked the analgesic effect of morphine. The action
lasted for at least 120 min indicating that M3G acts as
opioid antagonist only after central administration.
3.. Biological activity of M3GNH2
Acknowledgements
Three ip doses of 1 mg/kg (n=10), 5 mg/kg (n=8) and
10 mg/kg (n=5) of the M3GNH2 analogue were exam-
ined for analgesia by the tail flick and paw pressure tests
in rats using morphine, 5 mg/kg (n=5) and saline (n=6)
as reference. No analgesia was observed. On the other
hand, ip doses of 5 mg/kg of M3GNH2 administered 10
min after morphine (5 mg/kg) (n=15) fully inhibited its
analgesic effect in a similar pattern as the same doses of
naloxone. Accordingly, M3GNH2 was characterized as
a new opioid antagonist acting by peripheral adminis-
tration. Its lack of toxicity allows us to also conclude
that the amide function seems not to be the structural
feature responsible for the acute toxicity observed in
M3GOAM.
This work was supported by grants EUREKA-391
(Glycotrans) and 96-0175 from Centro para el Desar-
rollo Tecnologico Industrial (C.D.T.I.). Collaboration
of Sr. Isidro Linan Castellet in coordinating many
finantial, administrative and legal aspects of the work is
greatly acknowledged.
References and notes
1. Boerner, U.; Abbott, S.; Roe, R. L. Drug Metab. Rev.
1975, 112, 39.
2. Oguri, K.; Ida, S.; Youshimura, H.; Tsukamoto, H.
Chem. Pharm. Bull. 1970, 18, 2414.
3. Yeh, S. Y.; Gorodetzky, C. W.; Krebs, H. A. J. Pharm.
Sci. 1977, 66, 1288.
4. Mulder, G. J. Trends Pharm. Sci. 1992, 13, 302.
5. Brown, G. P.; Yang, K.; Ouerfelli, O.; Standifer, K. M.;
Bird, D.; Pasternak, G. W. J. Pharm. Exp. Ther. 1997,
282, 1291.
6. Shimomura, K.; Kamata, O.; Ueki, S.; Ida, S.; Oguri, K.;
Yoshimura, H.; Tsukamoto, K. Tohuku J. Exp. Med.
1973, 105, 45.
7. Paul, D.; Standifer, K. M.; Inturrisi, C. E.; Pasternak,
G. W. J. Pharmacol. Exp. Ther. 1989, 251, 477.
8. Gong, Q. L.; Hedner, T.; Hedner, J.; Bjorkman, R.;
Nordberg, G. Eur. J. Pharmacol. 1991, 193, 47.
In summary, we have found out that two relatively
minor modifications on the M3G molecule induce
remarkable biological effects. We here report positive
results showing that replacement of the carboxylic acid
of M3G by a carboxamide function it is a way to mod-
ulate BBB transport properties. Thus, the new molecule
acted as opioid antagonist by peripheral administration
while its parent compound (M3G) showed opioid
antagonist properties only after central administration.
On the other hand, we disclose an unprecedented nega-
tive finding showing that rising M3G lipophilicity

908
M. Salvatella et al. / Bioorg. Med. Chem. Lett. 14 (2004) 905–908
9. Osborne, R.; Joel, S.; Trew, D.; Slevin, M. Lancet 1988, 1,
828.
10. Hanna, M. H.; Peat, S. J.; Woodham, M.; Knibb, A.;
Fung, C. Br. J. Anaesth. 1990, 64, 547.
11. Penson, R. T.; Joel, S. P.; Bakhashi, K.; Clark, S. J.;
Langford, R. M.; Slevin, M. L. Clin. Pharmacol. Ther.
2000, 68, 667.
12. Penson, R. T.; Joel, S. P.; Clark, S.; Gloyne, A.; Slevin,
M. L. J. Pharm. Sci. 2001, 90, 1810.
13. Smith, M. T.; Watt, J. A.; Cramond, T. Life Sci. 1990, 47,
579.
14. Gong, Q. L.; Hedner, J.; Bjorkman, R.; Hedner, T. Pain
1992, 48, 249.
15. Suzuki, N.; Kalso, E.; Rosenberg, P. H. Eur. J. Pharmacol.
1993, 249, 247.
16. Hewett, K.; Dickenson, A. H.; McQuay, H. J. Pain 1993,
53, 59.
17. Lipkowski, A. W.; Carr, D. B.; Langlade, A.; Osgood,
P. F.; Szyfelbein, S. K. Life Sci. 1994, 55, 149.
18. Carrupt, P. A.; Testa, B.; Bechalany, A.; El Tayar, N.;
Descas, P.; Perrissoud, D. J. Med. Chem. 1991, 34, 1272.
19. Gaillart, P.; Carrupt, P. A.; Testa, B. Bioorg. Med. Chem.
Lett. 1994, 4, 737.
20. Wu, D.; Kang, Y. S.; Bickel, U.; Pardridge, W. M. Drug
Metab. Dispos. 1997, 25, 768.
21. Bickel, U.; Schumacher, O. P.; Kang, Y. S.; Voigt, K. J.
Pharm. Exp. Ther. 1996, 278, 107.
22. Valencia, G.; Rodrıguez, R.E. PCT/GB98/01578, 1998.
23. Tuttle, R. R.; Dixon, R. EP 0324212A1, 1988.
24. Rodrıguez, R. E.; Reig, F.; Valencia, G.; Herrero, J. F.;
Garcıa Anton, J. Neuropeptides 1986, 8, 335.
25. (a) Anisfeld, S. T.; Landsbury, P. T. J. Org. Chem. 1990,
55, 5560. (b) Anisfeld, S. T.; Landsbury, P. T. J. Am.
Chem. Soc. 1993, 115, 10531.
2.62 (m, 1H), 2.53 (s, 3H, N-CH3), 1.99 (m, 1H), 1.71 (d,
1H, J=14.0 Hz). 13C NMR (75 MHz, D2O) d 176.2,
147.9, 139.9, 133.7, 131.1, 128.3, 127.5, 126.8, 121.2,
118.2, 101.6, 92.2, 77.2, 76.4, 73.8, 72.7, 66.9, 60.7, 43.0,
41.8, 38.7, 33.4, 24.2; HPLC: RT=21.22 min (LiChro-
spher 100 RP-18, 5 mm, ACN-water gradient elution: 80%
to 70% of water in 5 min and 70% to 20% of water in 10
min). Morphine 3-ꢀ-D-octylglucuronamide (M3GOAM):
mp=89–91 ꢁC (from acetone–water); ESI-MS: 572.4
[M+H+]; [a]D=ꢀ124.5 (c=1, MeOH); 1H NMR
(200 MHz, DMSO-d6): d 6.20 (d, 1H, J=8.2 Hz, 2-H),
5.90 (d, 1H, J=8.2 Hz, 1-H), 5.45 (d, 1H, J=15.0 Hz, 7-
H), 5.25 (d, 1H, J=14.8 Hz, 8-H), 5.05 (d, 1H, J=3.0 Hz,
10-H), 4.70 (d, 1H, J=5.5 Hz, 5-H), 4.15 (m, 1H), 3.71 (m,
1H), 3.42 (m, 1H), 3.05 (m, 3H), 2.30 (s, 3H, N-CH3),
1.81–2.62 (m, 6H), 1.23–1.35 (m, alkyl). 13C NMR
(50 MHz, DMSO-d6) d 167.9, 147.3, 138.1, 131.1, 128.7,
118.4, 116.6, 100.6, 91.9, 76.3, 75.5, 73.0, 70.9, 66.2, 57.8,
45.8, 42.7, 42.6, 35.3, 31.2, 28.8, 28.7, 28.6, 26.3, 22.0;
HPLC: RT=12.80 min (LiChrospher 100 RP-18, 5 mm,
ACN-water gradient elution, same programme as above).
Morphine 3-ꢀ-D-glucuronamide (M3GNH2): ESI-MS:
460.2 [M+H+]; [a]D=ꢀ52.4 (c=1, MeOH); 1H NMR
(200 MHz, CD3OD): d 6.84 (d, 1H, J=8.3 Hz, 2-H), 6.58
(d, 1H, J=8.3 Hz, 1-H), 5.65 (d, 1H, J=14.7 Hz, 7-H),
5.34 (d, 1H, J=15 Hz, 8-H), 4.25 (m, 1H), 4.18 (d, 1H,
J=9.5 Hz), 3.90 (m, 1H), 3.53 (m, 2H), 3.31 (m, 1H), 3.17
(m, 1H), 3.08 (m, 1H), 2.77 (m, 2H), 2.55 (s, 3H, N-CH3),
2.49 (m, 1H), 2.46(m, 1H), 2.21 (m, 1H), 2.17 (m, 1H),
1.97 (d, 1H, J=13.5 Hz); 13C NMR (50 MHz, CD3OD) d
180.4, 174.2, 149.3, 140.3, 134.2, 132.5, 130.5, 128.9,
120.6, 119.3, 103.2, 93.4, 77.2, 75.6, 74.5, 73.2, 67.7, 60.2,
44.2, 42.8, 40.9, 35.8, 24.1; HPLC: RT=2.19 min
(LiChrospher 100 RP-18, 5 mm, ACN-water gradient
elution, same programme as above).
26. Berrang, B.; Twine, C. E.; Hennessee, G. L.; Carroll, F. I.
Synth. Commun. 1975, 5, 231.
29. Pharmacological tests. Animals were housed in a room
with controlled temperature at 21ꢂ1 ꢁC under 12 h light–
dark cycle and given free access to food and water. Both,
male Sprague–Dawley rats and Swiss-Webster mice were
tested for analgesia after 30 min of the administration of
different ip doses of the compounds in 500 mL of saline
solutions. The tail flick test, was conducted following
described protocols.32 Mechanical nociceptive thresholds
were determined by the paw pressure test on the left hind
paw as reported in the literature.33 Pain thresholds were
measured with an analgesemeter. The apparatus was set up
to apply a force of 0-1000 g increasing from zero at a rate
of 64 g/s. The nociceptive threshold was taken as the point
at which the animal vocalized or struggled vigorously.
30. Fernandez-Mayoralas, A.; De La Figuera, N.; Zurita, M.;
Vaquero, J.; Abraham, G. A.; San Roman, J.; Nieto-
Sampedro, M. J. Med. Chem. 2003, 46, 1286.
27. Synthesis
of
morphin-3-yl-b-d-octylglucuronamide
(M3GOAM). To a solution of 60 mg (0.13 mmol) of
M3G, 35 mg (0.26mmol) of HOBt and 125.2 mg (0.39
mmol) of 2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyl-
uronium hexafluorophosphate (HBTU) in 20 mL of
DMF, octylamine (23 mL, 0.26mmol) was added drop-
wise at room temperature under stirring. The reaction was
left to proceed for 20 h. Solvent evaporation and column
chromatography of the residue on silica gel (230–400
mesh) using a MeOH/H2O/ethyl acetate (4:3:8, v:v:v)
mixture as eluent, yielded the title compound in 40% yield
after crystallization from acetone/water.
28. Physicochemical data for final compounds. Morphine 3-ꢀ-
D-glucuronide (M3G): ESI-MS: 462.2 [M+H+];
[a]D=ꢀ117,8 (c=0.5, H2O); 1H NMR (300 MHz, D2O): d
6.68 (d, 1H, J=8.0 Hz, 2-H), 6.42 (d, 1H, J=8.0 Hz, 1-
H), 5.41 (d, 1H, J=9.5 Hz, 7-H), 5.07 (d, 1H, J=9.5 Hz,
8-H), 4.78 (d, 1H, J=5.5 Hz, 5-H), 4.74 (d, 1H, J=2.7
Hz, 10-H), 4.06(d, 1H, J=9.0 Hz), 3.67 (m, 1H), 3.54 (m,
1H), 3.40 (dd, 1H, J=2.5, 9.0 Hz, 20-H), 3.21 (m, 1H),
3.03 (m, 1H), 2.97 (m, 1H), 2.89 (m, 1H), 2.81 (m, 1H),
31. Bian, J.-T.; Bhargava, H. N. Peptides 1996, 17, 1415.
32. D’Amour, F. E.; Smith, D. L. J. Pharmacol. Exp. Ther.
1941, 72, 74.
33. Randall, L. O.; Selito, J. J. Arch. Int. Pharmacodyn. Ther.
1957, 11, 409.