A simple synthesis of morphine-3,6-di-β-D-glucuronide

Article abstract of DOI:10.1016/S0040-4020(00)00664-5

A convenient two-step synthesis of morphine-3,6-di-β-D-glucuronide in fair yield from morphine and methyl 2,3,4-tri-O-isobutyryl-1-O-trichloroacetimidoyl-α-D-glucopyranuronate is reported. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00664-5

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 7591–7594
A Simple Synthesis of Morphine-3,6-di-b-d-glucuronide
a
a
b
Richard T. Brown,a, Neil E. Carter, Stephen P. Mayalarp and Feodor Scheinmann
*
aDepartment of Chemistry, University of Manchester, Manchester M13 9PL, UK
bUltrafine (UFC) Ltd, Synergy House, Guildhall Close, Manchester Science Park, Manchester M15 6SY, UK
Received 5 June 2000; revised 6 July 2000; accepted 20 July 2000
Abstract—A convenient two-step synthesis of morphine-3,6-di-b-d-glucuronide in fair yield from morphine and methyl 2,3,4-tri-O-
isobutyryl-1-O-trichloroacetimidoyl-a-d-glucopyranuronate is reported. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Introduction
The discovery of the analgesic properties of morphine-6-b-
d-glucuronide (M6G) 3, a metabolite of morphine 1, was of
considerable medical interest,1 and created a demand for
appreciable quantities of M6G and related glucuronides
for pharmacological and clinical testing. However, syn-
thesis from morphine by standard Koenigs–Knorr coupling
or related heterogeneous methods was effectively precluded
by both limitation of scale and contamination of the
products by heavy-metal ions. As a consequence, we
devised and described in a 1993 patent and a subsequent
paper, homogeneous procedures suitable for large-scale
preparation of various glucuronides of morphine and
derivatives by Lewis acid catalysed condensation with
glucuronate esters.2,3 In particular, we gave specific experi-
mental details for a short synthesis (two steps) of morphine-
3,6-di-b-d-glucuronide (M3,6diG) 4 from morphine and
methyl 2,3,4-tri-O-isobutyryl-1-O-trichloroacetimidoyl-a-
d-glucopyranuronate 5, which we used to prepare multi-
gram quantities of M3,6diG in 35–50% yield for further
research.2,4 Nevertheless, in a publication of their own
synthesis (six steps) of M3,6diG, Berrang, Brine and Carroll
apparently found our procedure difficult to reproduce.5
Their failure was hardly surprising, since, although they
mentioned the specified isobutyrate 5 briefly in passing,
they actually used the corresponding acetate derivative 6
and hence only achieved largely transacetylation, which
confirmed our exploratory work6 on b-glucuronidation of
morphine using Schmidt’s imidate procedure.7
We had found that attempted Lewis acid catalysed coupling
of methyl 2,3,4-tri-O-acetyl-1-O-trichloroacetimidoyl-a-d-
glucopyranuronate 6 with 3-acetylmorphine 2 afforded
mostly 3,6-diacetylmorphine and little of the desired
6-glucuronate, and similar poor results with a varied
mixture of products were obtained with morphine.6 It was
this experience which led us to propose replacing the methyl
of the sugar acetates by a larger group to increase steric
hindrance. By this simple means, the rate of nucleophilic
attack at the carbonyl, and hence transacylation, would be
reduced, whereas the rate of glycosylation should be rela-
tively unaffected. To this end, the t-butyl group of pivalate
was effective but subsequent hydrolysis proved problematic
with formation of unwanted by-products. In the event, the
slightly less hindered isobutyrate was found to be the ester
Keywords: alkaloids; coupling reactions; glycosidation; stereoselection.
* Corresponding author. Tel.: ϩ44-161-2754632; fax: ϩ44-161-2754939;
e-mail: r.t.brown@man.ac.uk
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00664-5

7592
R. T. Brown et al. / Tetrahedron 56 (2000) 7591–7594
Scheme 1. Reagents: (i) NaOMe/MeOH; (ii) Me2CHCOCl/py; (iii) aq. NH3/CH2Cl2; (iv) CCl3CN/CH2Cl2/Na2CO3; (v) BF3·Et2O/CH2Cl2; (vi) aq. NaOH/
MeOH; (vii) AcOH.
of choice, combining minimal transacylation with ease of
hydrolysis. Subsequent use of the tri-isobutyrate 5 led to
effective preparations of M6G and related derivatives,2,3,6,8,9
with essentially complete stereoselection for the b-anomers
due to participation of the neighbouring C-2 acyl group.7
We now give a full report of our convenient and expeditious
synthesis of M3,6diG (Scheme 1).
pure b isomer 8, mp 126–8ЊC, could be obtained by recrys-
tallisation from petrol. Control of the temperature during
acylation maximised the proportion of the b isomer,
which was desirable since it was more readily hydrolysed
selectively to the hemi-acetal 9 in the next step. Initially,
this was carried out by bubbling gaseous ammonia through a
solution of 8 in dichloromethane and methanol, a successful
but rather lengthy procedure, but a two-phase system with
dichloromethane and saturated aqueous (880) ammonia
proved to be more rapid and effective on a large scale,
affording methyl 2,3,4-tri-O-isobutyrylglucuronate 9 as a
mixture of C-1 epimers in 73% yield. The formation of a
mixture was not of any material consequence since the
epimers interconverted, and hence base-catalysed reaction
with trichloro-acetonitrile gave in high yield (85–89%) only
the pure a-imidate 5, mp 80–1ЊC, characterised by an NMR
Results and Discussion
The readily available d-glucuronolactone 7 was converted
with sodium methoxide into a methyl ester, which was then
reacted with isobutyryl chloride in pyridine at below 10ЊC to
give methyl tetra-O-isobutyrylglucuronate in ca. 50% yield
as a 9:1 mixture of the b and a anomers, from which the

R. T. Brown et al. / Tetrahedron 56 (2000) 7591–7594
7593
singlet for the NH at dH 8.71 and a doublet at dH 6.66 with
Jˆ3.7 Hz for H-1. We found that alkaline metal carbonates
could be used as the base instead of the standard sodium
hydride,7 to obvious advantage on a large scale, and that
sodium carbonate was more effective than lithium,
potassium or caesium carbonates.
measurement. Merck silica plates F254 were used for TLC
and Kieselgel 60 for column chromatography. Organic
solutions were routinely dried over anhydrous sodium
sulfate.
Methyl 1,2,3,4-tetra-O-isobutyryl-b-d-glucopyranuro-
nate 8. To a solution of NaOMe (0.13 g, 2.4 mmol) in
MeOH (750 mL) was added glucurono-6,3-lactone (103 g,
0.77 mol) in portions with stirring until dissolved. The
solvent was then removed in vacuo, the residue taken up
in pyridine (425 mL, 5.4 mol) and the solution cooled to
0ЊC. Isobutyryl chloride (550 mL, 5.3 mol) in CH2Cl2
(350 mL) was then added with mechanical stirring at a
rate that kept the temperature below 10ЊC, and the reaction
mixture was left at room temperature overnight. More
CH2Cl2 (450 mL) was then added and the solution washed
with water (400 mL), 2 M HCl (3×200 mL), saturated aq.
sodium bicarbonate (5×400 mL) and brine (400 mL). After
drying, filtering and evaporating in vacuo, a gum was
obtained which crystallised on trituration with petrol (40–
60Њ). Filtration and drying at 40ЊC in a vacuum oven yielded
the title product (150 g, 54%) containing ϳ10% of the a
epimer. Recrystallisation from MeOH or petrol afforded
the pure b isomer 8 as needles, mp 126–8ЊC, [a]Dˆϩ10
(c 1.7 CHCl3); nmax(film): 1740 cmϪ1; m/z (CI): 506
(MϩNHϩ4 ), 418, 401; dH (300 MHz, CDCl3): 5.78 (d,
Jˆ8 Hz, H-1), 5.39 (t, Jˆ9.5 Hz, H-3), 5.25 (t, Jˆ9.5 Hz,
H-4), 5.23 (dd, Jˆ9.5, 8 Hz, H-2), 4.19 (d, Jˆ9.5 Hz, H-5),
3.75 (s, OMe), 2.65–2.45 (m, 4×CHMe2), 1.17–1.07 (m,
4×CHMe2). Found: C, 56.7; H, 7.6. C23H36O11 requires: C,
56.6; H, 7.4.
A molar excess of the imidate 5 was then reacted with dry
morphine in dichloromethane in presence of BF3 catalyst to
give the M3,6diG derivative 10 with exclusive b-stereo-
chemistry at C-1 for both glycosides. Although morphine
was virtually insoluble in dichloromethane, it slowly
dissolved as it reacted, and after two days essentially only
10 could be detected by TLC, together with excess sugars
and a trace of an M3G derivative. Chromatography removed
the last and the bulk of the sugars, and recrystallisation of
the crude product from ethanol afforded pure morphine-3,
6-di(methyl tri-isobutyrylglucuronate) mp 229–230ЊC
[a]DˆϪ110 (CHCl3). The isolated yield of crystalline
material was 60%, but analysis of the reaction mixture
indicated Ͼ90% conversion of morphine. Alternatively,
the reaction mixture could be directly triturated with hot
methanol to give crystalline product in ca. 35% yield.
The structure was established from its spectroscopic data:
inter alia, the UV maxima at 282 and 235 nm did not shift on
addition of alkali, hence no free phenol was present, and an
ion at m/z 1086 (Mϩ1) in the FAB mass spectrum corre-
sponded to the expected morphine diglycoside, which was
corroborated by combustion analysis for C55H75NO21.
Assignment of the 1H NMR spectrum confirmed the
structure as 10, and in particular, the b-linkages of both
glucuronate moieties from doublets for H-10 and 100 at d
5.81 and 4.71 with trans-diaxial couplings of 8 Hz for
both. Subsequent careful hydrolysis of 10 with alkali and
precipitation by acidifying with acetic acid afforded in 89%
yield morphine-3,6-di-b-d-glucuronide 4, which could be
recrystallised from aqueous methanol, mp 243–4ЊC (dec.).
It was homogeneous from its 500 MHz 1H NMR spectrum,
which again confirmed the structure and showed both
glucuronides to be b-anomers.
For reference purposes, an alternative procedure with iso-
butyric anhydride and perchloric acid catalyst, working up
with sat. aq. (880) ammonia to destroy excess anhydride,
was used to prepare in 13% yield the microcrystalline
a-isomer, mp 89–90ЊC, [a]Dˆϩ103 (c 0.03 CHCl3);
nmax(film): 1750 cmϪ1; m/z (CI): 506 (MϩNHϩ4 ), 401,
314; dH (300 MHz, CDCl3): 6.41 (d, Jˆ3.5 Hz, H-1), 5.57
(t, Jˆ10 Hz, H-3), 5.23 (t, Jˆ10 Hz, H-4), 5.14 (dd, Jˆ10,
3.5 Hz, H-2), 4.39 (d, Jˆ10 Hz, H-5), 3.72 (s, OMe), 2.72–
2.37 (m, 4×CHMe2), 1.25–1.04 (m, 4×CHMe2). Found: C,
56.5; H, 7.4. C23H36O11 requires: C, 56.6; H, 7.4.
We had thus achieved the first synthesis of morphine-3,6-di-
b-d-glucuronide in a minimum overall yield from morphine
of ca. 50%. However, this could obviously be considerably
increased by direct hydrolysis of the crude 10, rather than
recrystallised material, since the only significant impurity is
excess sugar, which would remain in solution.
Methyl 2,3,4-tri-O-isobutyryl-d-glucopyranuronate 9.
Procedure A. A solution of the above crude tetra-isobutyrate
(9.0 g, 18 mmol) in CH2Cl2 (360 mL) and MeOH (36 mL)
was cooled to Ϫ4ЊC, dry ammonia gas was then bubbled
through for 2 h, and the solution left to warm up to room
temperature overnight. TLC indicated ϳ50% conversion, so
the procedure was repeated until no starting material was
observed by TLC. Then after bubbling nitrogen gas through
the solution for an hour, it was evaporated, the residue taken
up in CH2Cl2, loaded on to a pad of silica and eluted with
CHCl3/MeOH 9:1. Evaporation of the eluate gave the title
product 9 as a glass (6.96 g, 90%) containing a 4:1 mixture
of a/b epimers (dH 5.54 (d, Jˆ3.5 Hz, H-1a)/ 5.78 (d,
Jˆ8 Hz, H-1b). Recrystallisation from CHCl3/petrol (40–
60Њ) afforded the pure microcrystalline a-epimer, mp 89–
90ЊC, [a]Dˆϩ76 (c 2.2 CHCl3); nmax(film): 3460,
1750 cmϪ1; m/z (FAB): 418 (Mϩ); dH (300 MHz, CDCl3):
5.65 (t, Jˆ10 Hz, H-3), 5.54 (d, Jˆ3.5 Hz, H-1), 4.92 (dd,
Jˆ10,3.5 Hz, H-2), 4.60 (d, Jˆ10 Hz, H-5), 3.75 (s, OMe),
Experimental
General
Melting points were obtained on a Kofler block and are
uncorrected. An Optical Activity AA-100 polarimeter was
used for optical rotations [a]D. Ultra-violet spectra (lmax
)
were recorded on a Shimadzu UV-260 spectrometer, infra-
red spectra (nmax) on a Perkin–Elmer 1710 FT-IR spectro-
meter, proton magnetic resonance spectra (dH) on Varian
500 or 300 MHz spectrometers and mass spectra (m/z EI,
CI, FAB) on a Kratos Concept IS spectrometer. Molecular
formulae were determined by combustion analysis and mass

7594
R. T. Brown et al. / Tetrahedron 56 (2000) 7591–7594
2.61–2.43 (m, 4×CHMe2), 1.20–1.05 (m, 4×CHMe2).
Procedure B. A suspension of dried morphine (4.0 g,
14 mmol) and the imidate (36.1 g, 64 mmol) in dry
CH2Cl2 (100 mL) was stirred under argon at room tempera-
ture, while BF3·Et2O (7.8 mL, 64 mmol) was added. After
90 min, the solution was evaporated in vacuo to a gum,
which was crystallised from MeOH to afford the
morphine-3,6-diglucuronate (5.4 g, 35%) as needles, mp
229–230ЊC.
Found: C, 54.7; H, 7.4. C19H30O10 requires: C, 54.6; H, 7.2.
Procedure B. The tetra-isobutyrate (102 g, 0.21 mol) in
CH2Cl2 (100 mL) was stirred in a sealed vessel with sat.
aq. (880) ammonia (40 mL) overnight. The solution was
washed with water (50 mL) and then brine (50 mL), dried,
filtered and evaporated to give the gummy hemi-acetal
(64 g, 73%) as an a/b mixture similar to that above.
Morphine-3,6-di-b-d-glucuronide 4. Aq. NaOH (1.5 M,
12 mL) was added to a stirred suspension of the above
diglucuronate (2.3 g, 2.1 mmol) in MeOH (68 mL), and
the mixture left overnight. The solution was then acidified
with glacial acetic acid, the precipitate filtered and washed
with MeOH (ϳ15 mL). Drying at 60Њ under high vacuum
afforded morphine-3,6-diglucuronide (1.2 g, 89%) as a
white solid, which formed microcrystals from aq. MeOH,
mp 243–244ЊC (dec.) [a]DˆϪ188 (c 10 H2O); lmax(H2O):
282, 236, 208 nm; dH (500 MHz, D2O): 6.99 (d, Jˆ8 Hz,
H-2), 6.75 (d, Jˆ8 Hz, H-1), 5.84 (bd, Jˆ10 Hz, H-8), 5.38
(bd, Jˆ10 Hz, H-7), 5.28 (d, Jˆ6 Hz, H-5), 5.07 (d,
Jˆ7 Hz, H-10), 4.70 (d, Jˆ8 Hz, H-100), 4.56 (m, H-6),
4.22 (m, H-13), 3.84 (d, Jˆ8 Hz, H-50), 3.73 (d, Jˆ9 Hz,
H-500), 3.61–3.56 (m, H-40, 30, 20), 3.54 (t, Jˆ9 Hz, H-300),
3.51 (t, Jˆ9 Hz, H-400), 3.40 (dd, Jˆ9,8 Hz, H-200), 3.31 (d,
Jˆ20 Hz, H-12eq), 3.06 (ddd, Jˆ13,12,4.5 Hz, H-16ax), 2.97
(s, NMe), 2.26 (dd, Jˆ19,6 Hz, H-12ax), 2.01 (ddd,
Jˆ13.5,12,4 Hz, H-15ax), 1.84 (bd, Jˆ12 Hz, H-15eq).
Found: C, 50.5; H, 6.4; N, 1.9. C29H35NO15·3H2O requires:
C, 50.4; H, 6.0; N, 2.0.
Methyl 2,3,4-tri-O-isobutyryl-1-O-trichloroacetimidoyl-
a-d-glucopyranuronate 5. To a stirred solution of the
above a/b mixture of hemiacetals (82.4 g, 0.20 mol) in
CH2Cl2 (100 mL) was added trichloroacetonitrile (74 mL,
0.73 mol), followed by anhydrous sodium carbonate
(12.4 g, 0.12 mol), and the mixture left to stir overnight.
Filtration and evaporation in vacuo then yielded the title
product 5 as a semicrystalline gum (95.0 g, 85%), which
crystallised with dry isopropanol as prisms, mp 80–1ЊC
[a]Dˆϩ67 (c
5
CHCl3); nmax(film): 3320, 1750,
1680 cmϪ1; m/z (CI): 579 (MϩϩNH3), 546, 506, 401, 348;
dH (300 MHz, CDCl3): 8.72 (s, NH), 6.66 (d, Jˆ3.5 Hz,
H-1), 5.70 (t, Jˆ10 Hz, H-3), 5.30 70 (t, Jˆ10 Hz, H-4),
5.20 (dd, Jˆ10,3.5 Hz, H-2), 4.51 (d, Jˆ10 Hz, H-5), 3.75
(s, OMe), 2.60–2.43 (m, 3×CHMe2), 1.17–1.06 (m,
3×CHMe2). Found: C, 44.8; H, 5.3; Cl, 18.4; N, 2.6.
C21H30Cl3NO10 requires: C, 44.8; H, 5.3; Cl, 18.9; N, 2.5.
Morphine-3,6-di(methyl 2,3,4-tri-O-isobutyryl-b-d-gluco-
pyranuronate) 10. Procedure A. A suspension of dried
morphine (2.0 g, 7.0 mmol) and the above imidate (15.8 g,
28.1 mmol) in dry CH2Cl2 (40 mL) containing 4A molecu-
lar sieves was stirred under argon at room temperature,
while BF3·Et2O (355 mL, 4.0 g, 28.1 mmol) was added.
After 30 min, virtually all of the starting materials had
gone into solution and stirring was continued for 2 days.
More CH2Cl2 (20 mL) was added, the solution washed
with saturated aq. sodium bicarbonate (50 mL), water and
brine before being dried. Filtration and evaporation in vacuo
afforded a semisolid residue, which was taken up in CHCl3/
MeOH 40:1 and flash chromatographed on silica. Elution
with CHCl3/MeOH 9:1 gave a crude product (8.5 g) consist-
ing (TLC/NMR) of ϳ80% diglucuronate and ϳ20% excess
sugar. Trituration with EtOH yielded pure morphine-3,6-
diglucuronate 10 (4.6 g, 60%) as colourless needles, mp
229–230ЊC [a]DˆϪ110 (c 1.3 CHCl3); nmax(EtOH): 282,
235, 210 nm; nmax(film): 1750 cmϪ1; m/z (FAB): 1086
(MHϩ), 686, 401; dH (300 MHz, CDCl3): 6.76 (bd, Jˆ
8 Hz, H-2), 6.47 (d, Jˆ8 Hz, H-1), 5.97 (t, Jˆ9.5 Hz,
H-30), 5.81 (d, Jˆ8 Hz, H-1), 5.57 (d, Jˆ10 Hz, H-8),
5.40 (t, Jˆ9.5 Hz, H-300), 5.27–5.21 (dd, Jˆ9.5,8 Hz,
H-20; dd, Jˆ9.5,8 Hz, H-200; d, Jˆ10 Hz, H-7), 5.20 (t,
Jˆ9.5 Hz, H-40), 5.13 (t, Jˆ9.5 Hz, H-400), 4.72 (d, Jˆ
5 Hz, H-5), 4.71 (d, Jˆ8 Hz, H-100), 4.64 (d, Jˆ9.5 Hz,
H-50), 4.05 (d, Jˆ9.5 Hz, H-500), 3.98 (dd, Jˆ5,1.5 Hz,
H-6), 3.75 (s, OMe), 3.69 (s, OMe), 3.34 (dd, Jˆ6,3 Hz,
H-13), 3.02 (d, Jˆ19 Hz, H-12eq), 2.60–2.37 (m, H-16eq,
6×CHMe2), 2.43 (s, NMe), 2.41 (ddd, Jˆ13.5,12,4 Hz,
H-16ax), 2.26 (dd, Jˆ19,6 Hz, H-12ax), 2.01 (ddd, Jˆ
13.5,12,4 Hz, H-15ax), 1.84 (bd, Jˆ12 Hz, H-15eq), 1.34–
1.08 (m, 6×CHMe2). Found: C, 60.6; H, 6.9; N, 1.3.
C55H75NO21 requires: C, 60.8; H, 6.9; N, 1.3.
Acknowledgements
We acknowledge the receipt of EPSRC TT (SPM) and
CASE (NEC) Studentships, a DTI SMART Award to
Ultrafine, and thank Dr S. Joel (St. Bartholemew’s Hospital,
London) for his collaboration.
References
1. Osbourne, R.; Joel, S.; Trew, D.; Slevin, N. The Lancet 1988,
828.
2. Scheinmann, F.; Lumbard, K. W.; Brown, R. T.; Mayalarp,
S. P.; Carter, N. E. Int. Patent WO 93/03051 (PCT/GB92/01449;
EP0597915; US5977326) 1993.
3. Brown, R. T.; Carter, N. E.; Lumbard, K. W.; Scheinmann, F.
Tetrahedron Lett. 1995, 36, 8661–8664.
4. Brown, R. T.; Carter, N. E.; Scheinmann, F.; Turner, N. J.
Tetrahedron Lett. 1995, 1117–1120 (EP0733120).
5. Berrang, B.; Brine, G. A.; Carroll, F. I. Synthesis 1997, 1165–
1168.
6. Mayalarp, S. P. PhD Thesis, University of Manchester, 1993.
7. Schmidt, R. R. Angew. Chem., Int. Ed. Engl. 1986, 25, 212–
235.
8. Brown, R. T.; Mayalarp, S. P.; McGown, A. T.; Hadfield, J. A.
J. Chem. Res. Synop. 1993, 496–497.
9. Jenkins, G. N.; Stachulski, A. V.; Scheinmann, F.; Turner, N. J.
Tetrahedron: Asymmetry 2000, 11, 413–416.