Convenient syntheses of the in vivo carbohydrate metabolites of mycophenolic acid: reactivity of the acyl glucuronide

Article abstract of DOI:10.1016/j.tetlet.2009.06.060

Following in vivo use of mycophenolic acid, the O-aryl and O-acyl glucuronides, as well as the recently discovered O-aryl glucoside (Scheme 1), are all found as metabolites. We describe convenient preparations of all three derivatives. The phenolic glycos

Full text of DOI:10.1016/j.tetlet.2009.06.060

Tetrahedron Letters 50 (2009) 4973–4977
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Convenient syntheses of the in vivo carbohydrate metabolites
of mycophenolic acid: reactivity of the acyl glucuronide
Amy E. Jones a, Helen K. Wilson a, Paul Meath a, Xiaoli Meng a, David W. Holt b, Atholl Johnston c,
Michael Oellerich d, Victor W. Armstrong d, Andrew V. Stachulski a,
*
a The Robert Robinson Laboratories, Department of Chemistry, University of Liverpool, Liverpool L69 7ZD, UK
b Analytical Unit, St. George’s, University of London, London, UK
c Clinical Pharmacology William Harvey Research Institute, Barts and the London, School of Medicine and Dentistry, Charterhouse Square, London EC1 6BQ, UK
d Department of Clinical Chemistry, University Medicine Goettingen, 37075 Goettingen, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Following in vivo use of mycophenolic acid, the O-aryl and O-acyl glucuronides, as well as the recently
discovered O-aryl glucoside (Scheme 1), are all found as metabolites. We describe convenient prepara-
tions of all three derivatives. The phenolic glycosides are obtained by phase-transfer-catalysed alkylation
of methyl mycophenolate in very high yield, as an excellent alternative to the Königs-Knorr reaction. We
carefully optimised our earlier synthesis of the acyl glucuronide to give a highly pure product in a much
improved yield. Finally, we describe the value of a synthetic acyl glucuronide in demonstrating its reac-
tivity towards a known target protein with superior response to the naturally obtained material.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 11 May 2009
Revised 4 June 2009
Accepted 12 June 2009
Available online 16 June 2009
The mould metabolite mycophenolic acid 1 (Scheme 1) was
first isolated from a Penicillium species in 1896.1 Its potential
as an antibacterial and an antifungal agent was reported in
1946;2,3 later there was considerable interest in its antiviral
and antitumour4,5 activities and many semi-synthetic analogues
were prepared.6 Recently, 1 and its derivatives have become
established as highly effective immunosuppressive7,8 drugs in
the course of transplant surgery. The human metabolic products
of 1 are therefore of great importance. Interestingly, a new anti-
cancer indication has recently been found for 1; its hydroxamic
acid derivatives show significant activity as histone deacetylase
(HDAC) inhibitors.9
Following in vivo administration, the major metabolite of 1 is
the O-aryl (phenolic) glucuronide 2,10 apparently a purely detox-
ifying metabolite. A much smaller part of the dose (ca. 1.25%)11
is converted into the O-acyl glucuronide 3; this compound
appears to retain activity as an inhibitor of the proliferation of
mononuclear leukocytes12 and it has also been implicated in
unfavourable interactions with liver proteins.13–15 Finally, the
aryl glucoside 4 has also been identified as a human metabo-
lite:16 this Letter also gave evidence for the formation of an acyl
glucoside in the kidney, but probably not in therapeutically sig-
nificant amounts.
matic preparation of 3 was achieved using homogenised horse li-
ver18 after removing small amounts of 2 by preparative HPLC; no
preparation of 4 has been reported. We now report efficient chem-
ical syntheses of 2–4, avoiding heavy metals19 for 2 and 4. Further,
the selective acylation method for acyl glucuronide synthesis,20,21
has been optimised to yield highly pure 3 in very good yield. We
also report on the protein reactivity of synthetic 3.
For the preparation of aryl glucuronide 2, an ester of 1 is neces-
sary. It was best to heat 1 with methanol and the acid ion-ex-
change resin IR-120, or treat with tosic acid at 20 °C, to afford
methyl ester 517 in high yield and purity, Scheme 2. The use of
CH2N2 or Me3SiCHN2 invariably led to over-reaction, with forma-
tion of the phenolic O-methyl ether.
Using the Königs-Knorr synthesis17 of 2, we found that it was
simplest to react 5 with an excess (1.5 equiv) of glycosyl bromide
622 in quinoline with Ag2CO3 catalysis. In this way, all the methyl
ester 5 was converted to afford the desired conjugate 7 in high
yield and purity after recrystallisation: excess 5, in contrast, could
only be fully removed by chromatography.
In general, it is desirable to avoid the use of heavy metal cata-
lysts in glycosidation. Glycosidation of phenols by the trichloroace-
23
timidate method using, for example, BF3 is normally excellent,
but here cyclisation of 5 occurs, giving the tetrahydropyran deriv-
ative 8.24 This side-reaction also prohibits the glucuronidation of 5
using perester 9 or hemiacetal 10 with a strong Lewis acid.23b We
found that the cyclisation of 5 was much slower with the mild Le-
wis acid ZnCl2; coupling of 5 with 11 using ZnCl2 as catalyst was
therefore attempted, but no conjugate 7 could be obtained and 8
was again formed slowly.
It is important that analytical standards of compounds 2–4
should be available for analysis and toxicology. The synthesis of
2 via the classical Königs-Knorr procedure is known17 and an enzy-
* Corresponding author. Tel.: +44 0 151 794 3542; fax: +44 0 151 794 3588.
E-mail address: stachuls@liv.ac.uk (A.V. Stachulski).
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.06.060

4974
A. E. Jones et al. / Tetrahedron Letters 50 (2009) 4973–4977
O
O
OH
OH
O
Me
OH
Me
HO
HO
O
O
HO
HO
O
O
HO
Me
O
HO
O
O
MeO
HO2C
O
3
MeO
Me
OH
2
O
HO
HO
O
HO
Me
Me
OH
Me
O
O
HO2C
HO2C
O
O
MeO
MeO
4
Me
1
Scheme 1. Carbohydrate phase 2 metabolites of mycophenolic acid 1. It is understood that the same UDPGT isoform, namely UGT1A10, is involved in the biosynthesis of 2
and 3.
effective. PhCH2N+Et3 Clꢀand nBu4N+ Brꢀ are good catalysts, with
MeO C
2
nBu4N+Brꢀ marginally superior, and Cs2CO3 is an excellent base.
Zemplen deprotection of 7 (using an excess of base since two
carboxylates are formed in this step) afforded aryl glucuronide 2
in excellent yield and purity after acidification using IR-120(H+),
filtration, evaporation and trituration with ether. Originally,
acetone was used,17 but MeOH gave a much cleaner product.
Me
O
O
O
OMe
O
AcO
AcO
O
X
MeO
AcO
Me
8
9 X = β-OAc
O
O
OMe
OR
10 X = α/β-OH
O
O
AcO
AcO
HO
HO
OH
11 X = α-OC(=NH)CCl3
AcO
HO
I
13 R = CH2CH=CH2
12
14 R = CH2Ph
An excellent alternative method for the transformation of 5 into
7 involves phase-transfer alkylation: initially we used the in situ
formed Li phenolate of 5.19 Reaction of this salt (Scheme 2) with
6 in a water/(CH2Cl)2 mixture catalysed by nBu4N+ Brꢀ gave satis-
factory conversion into 7, although it was difficult to drive the
reaction to completion: pure 7 was obtained in approximately
40% yield after chromatography. Glycosyl iodide 1225 did not give
a better yield.
We have already outlined the preparation of acyl glucuronide
20,21 in fair yield, using our selective acylation method, from ester
13 or 14, and we now report a careful optimisation of this method.
The use of the benzyl ester 14 is now preferred, Scheme 3; depro-
tection of allyl esters requires soluble Pd(0) catalysts20 and
removal of Pd residues may be troublesome.
Thus reaction of 1 with 14 using HATU and either N-methyl-
morpholine (NMM)20 or N-ethylmorpholine (NEM) in acetonitrile
afforded benzyl ester conjugate 15 in 55–60% yield after chroma-
tography. It proved best to maintain a 1:1 ratio of 1 to 14 so as
to minimise the formation of close-running materials that were
difficult to separate fully from 15. We suspected that these
by-products might be due to the free phenolic OH group, and to
investigate this, 1 was converted into its O-benzyl ether 16.
3
Use of solid phase-transfer catalysis26 [K2CO3 as base in CHCl3
solvent, a minimum amount of water (1:1 molar ratio with the
bromosugar) and 2 equiv of 6] gave a much cleaner reaction:
hydrolysis of 5 was avoided and 7 was obtained in 82% yield. We
believe that this is the first application of these conditions for
glucuronidation: the use of traces of H2O was mentioned only in
a footnote in the original paper26 but appears crucial and very
O
OMe
O
O
AcO
AcO
OMe
O
O
Me
OH
Me
AcO
AcO
O
AcO
Me
RO2C
O
AcO
O
Br
(iii)
6
MeO2C
MeO
1 R = H
5 R = Me
O
2
(ii)
MeO
Me
7
(i)
Scheme 2. Synthesis of O-aryl glucuronide 2. Reagents and conditions: (i) MeOH, Amberlite IR-120 (H+), heat, 93%; (ii) quinoline, Ag2CO3, 0–20 °C or LiOH, nBu4N+ Brꢀ,
(CH2Cl)2–H2O, heat, or K2CO3, CHCl3, PhCH2N+Et3 Clꢀ, trace H2O, 82%; (iii) aq NaOH, MeOH, then IR-120 (H+), 98%.

A. E. Jones et al. / Tetrahedron Letters 50 (2009) 4973–4977
4975
O
Me
OH
Me
O
OBn
Me
OH
Me
O
O
(i)
HO
HO
(ii)
HO2C
O
O
3
O
HO
MeO
O
MeO
1
15
Scheme 3. Synthesis of O-acyl glucuronide 3. Reagents and conditions: (i) 14, HATU, NMM or NEM, MeCN, 3 Å MS, 20 °C, 82%; (ii) H2, 10% Pd-C, THF–iPrOH, or catalytic
transfer hydrogenation, 95%.
Bis-benzylation of 1 using K2CO3 and benzyl bromide in DMF gave
the bis-benzyl derivative 17, then hydrolysis of the benzyl ester
afforded the protected acid 16, Scheme 4.
1 is the active metabolite of the immunosuppressive drugs myco-
phenolate mofetil (MMF) and mycophenolate sodium (MPS), both
of which are widely used in immunosuppressant protocols after
Both steps proceeded in excellent yield,27 but 16 did not give a
better yield than 1 in the conjugation step with 14, and there were
difficulties in the deprotection of the intermediate. Instead
(Scheme 3), when the solution of 1, 14 (1.1 equiv) and NMM
(4 equiv) in MeCN was stirred and dried with 3 Å sieves under N2
for 2 h before addition of HATU (1.1 equiv), ester 15 was obtained
in a much improved yield (82%) as the single 1b-anomer (dca. 5.5,
d, J = 8 Hz for H-1). For debenzylation of 15, either conventional
hydrogenation or catalytic transfer hydrogenation (1,4-cyclohexa-
Table 1
Affinity labelling of human r-IMPDH with two different sources of acyl glucuronide 3
Gel 2 Lane
Concentration and source of 3
Incubation time (h)
Response
(a) No NaCNBH3 (Gel 2)
1
2
3
4
5
6
7
8
9
10
0 mM Sa
24
24
24
24
24
24
24
24
24
24
ꢀ
ꢀ
ꢀ
ꢀ
+
ꢀ
++
+
0 mM Nb
10 mM S
10 mM N
50 mM S
50 mM N
100 mM S
100 mM N
200 mM S
200 mM N
i
diene, PrOH, 60 °C) could be used but it was crucial to use only
iPrOH as solvent, to avoid reduction of the trisubstituted double
bond: using THF as co-solvent led to C@C reduction (LC–MS evi-
i
dence), in proportion to the THF: PrOH ratio. Filtration, evapora-
tion and trituration with ether then afforded highly pure product
++
+
3; see the 1H NMR spectrum (Supplementary data).
Finally the previously unsynthesised O-aryl glucoside 4 was
prepared, Scheme 5. Here again the Königs-Knorr procedure17 or
the two-phase alkylation method19 was viable, but the above
two-phase method using minimal added water26 was best. Thus
reaction of glucosyl bromide 18 with methyl ester 5 gave an excel-
lent (92%) yield. The resulting fully protected intermediate 19
finally crystallised from ether-hexane and was subjected to depro-
tection as described for 2. Glucoside 4 was obtained in excellent
yield and purity.
Gel 1 Lane
(b) With NaCNBH3 (Gel 1)
1
2
3
4
5
6
7
8
10 mM S
10 mM N
50 mM S
50 mM N
100 mM S
100 mM N
200 mM S
200 mM N
24
24
24
24
24
24
24
24
+
+
++
++
+++
+++
+++
+++
We now demonstrate the great value of pure synthetic acyl
glucuronides in studying protein interactions. Mycophenolic acid
a
S = synthetic 3 prepared as in text.
N = naturally obtained material (biochemically synthesised).
b
Me
O
Me
OH
Me
O
O
(i)
RO C
2
HO C
2
O
O
MeO
MeO
Me
1
17 R = CH2Ph
(ii)
16 R = H
Scheme 4. Synthesis of mycophenolic acid O-benzyl ether. Reagents and conditions: (i) PhCH2Br, K2CO3, DMF, 96%; (ii) NaOH, aq EtOH, then H+, 98%.
OAc
O
AcO
AcO
OAc
O
O
AcO
AcO
Me
OH
Me
O
AcO
Me
MeO2C
AcO
18
O
O
Br
(ii)
MeO2C
MeO
4
O
(i)
MeO
19
Me
5
Scheme 5. Synthesis of mycophenolic acid O-glucoside. Reagents and conditions: (i) quinoline, Ag2CO3, 0–20 °C or LiOH, nBu4N+ Brꢀ, (CH2Cl)2–H2O, heat, or K2CO3, CHCl3,
PhCH2N+Et3 Clꢀ, trace H2O, 92%; (ii) aq NaOH, MeOH, then IR-120 (H+), 98%.

4976
A. E. Jones et al. / Tetrahedron Letters 50 (2009) 4973–4977
CO H
2
XHN
R
O
HO
HO
+
X-NH
Transacylation
2
OH
O
HO
CO H
CO H
2
O
O
4
2
O
HO
2
X-NH
2
HO
HO
Acyl
migration
3
R
RCO
OH
2
Glycation
HO
(Also 2- and
4-O-acyl isomers)
HO
O
CO H
2
OH
CO H
OH
NH-X
2
Rearrangement
HO
2
HO
2
RCO
NX
RCO
HO
O
NaCNBH
3
CO H
2
OH
NH-X
HO
2
RCO
HO
Scheme 6. Rearrangement and reduction of acyl glucuronide–protein adducts. Here X–NH2 is a protein Lys side-chain.
solid organ transplantation.28 The immunosuppressant action of 1
resides in the uncompetitive, selective and reversible inhibition of
inosine-50-monophosphate dehydrogenase (IMPDH), resulting in a
decreased de novo synthesis of guanine nucleotides with conse-
quent impairment of nucleic acid synthesis.28 MPA 1 is primarily
metabolised in the liver to the inactive metabolite 2, but also to a
lesser extent to the chemically reactive metabolite 3.29 Covalent
binding of acyl glucuronides to proteins is considered an initiat-
ing event for the organ toxicity of drugs containing a carboxylic
acid group.14,15 In a previous study we showed that 3 is an inhib-
itor of IMPDH,12 a key enzyme in the de novo synthesis of purine
nucleotides, and is also capable of forming protein adducts in
vivo.13,30,31
pounds 2–4. The value of pure synthetic 3 in studying protein
reactivity is clearly shown. These procedures will be most valuable
in providing analytical standards for all who are engaged in
research on the therapeutic and toxic effects of mycophenolic acid
1 and add to our understanding of the important topic of acyl
glucuronide–protein reactivity and potential toxicity.15
Acknowledgements
We are most grateful to Analytical Services International Ltd.
for financial support to A.E.J., H.K.W. and P.M.; to Novartis plc for
a generous gift of mycophenolic acid; and to Roche Palo Alto LLC
for a biochemical sample of mycophenolic acid acyl glucuronide.
Using a previously described method32 for the analysis and
characterisation of 3, the above synthetic material was compared
with a biochemically synthesised sample regarding their modifi-
cation of IMPDH. By HPLC analysis, the two samples of 3 had
virtually identical purity (1–1.5% aglycone present). We have
previously shown8 that the acyl glucuronide isolated from trans-
plant recipients treated with mycophenolate mofetil, although
apparently homogeneous according to HPLC, is in fact a mixture
of acyl isomers by ester group migration to O-2, 3 and 4. It is
feasible to improve the stability of 3 by acidification,33 in com-
mon with other O-acyl glucuronides.15 The synthetic material
showed a superior ability to conjugate with protein in the
absence of NaCNBH3 trapping, but the two showed very similar,
concentration-dependent responses with NaCNBH3 present
[Table 1; see Supplementary data for the Western blot analysis
(figure)].
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tetlet.2009.06.060.
References and notes
1. Gosio, B. Rivista Igiene Sanita Pubblica Ann. 1896, 7, 825.
2. Florey, H. W.; Gilliver, K.; Jennings, M. A.; Sanders, A. G. Lancet 1946, 1, 46.
3. Gilliver, K. Ann. Botany 1946, 10, 271.
4. Williams, R. H.; Lively, D. H.; De Long, D. C.; Cline, J. C.; Sweeney, M. J.; Poore, G.
A.; Larsen, S. H. J. Antibiot. 1968, 21, 463.
5. Carter, S. B.; Franklin, T. J.; Jones, D. F.; Leonard, B. J.; Mills, S. D.; Turner, R. W.;
Turner, W. B. Nature 1969, 233, 848.
6. Jones, D. F.; Mills, S. D. J. Med. Chem. 1971, 14, 305; none of the semi-synthetic
analogues had superior activity.
7. Mitsui, A.; Suzuki, S. J. Antibiot. 1969, 22, 358.
This confirms the value of the synthetic material. The significant
difference in response between the two batches of 3 and IMPDH in
the absence of reducing agent is interesting, and may reflect the
two possible modes of reaction (in the absence of NaCNBH3) shown
in Scheme 6. Thus both direct transacylation of the acyl residue
and glycation [reaction of an amine at C(1)], forming initially an
imine, then a rearranged (Amadori) product are possible. If the
initial purity of synthetic 3 is higher (i.e., less acyl migration has
occurred), it could more readily undergo transacylation as this
pathway is known to occur much more readily with an anomeric
ester rather than a 2, 3 or 4-ester group after migration.15 On the
other hand, the glycation pathway requires acyl migration to occur
first.
8. Shipkova, M.; Armstrong, V. W.; Wieland, E.; Niedmann, P. D.; Schütz, E.;
Brenner-Weiss, G.; Voihsel, M.; Braun, F.; Oellerich, M. Brit. J. Pharmacol. 1999,
126, 1075.
9. Chen, L.; Wilson, D.; Jayaram, H. N.; Pankiewicz, K. W. J. Med. Chem. 2007, 50,
6685.
10. (a) Franclin, T.; Jacobs, V.; Jones, G.; Ple, D.; Bruneau, P. Cancer Res. 1996, 56,
984; (b) Sweeney, M. J.; Hoffmann, D.; Esterman, M. A. Cancer Res. 1972, 32,
1803.
11. Shipkova, M.; Wieland, E.; Schütz, E.; Wiese, C.; Niedmann, P.; Oellerich, M.;
Armstrong, V. W. Transplant Proc. 2001, 33, 1080.
12. Schütz, E.; Shipkova, M.; Armstrong, V. W.; Wieland, E.; Oellerich, M. Clin.
Chem. 1999, 45, 419.
13. Shipkova, M.; Beck, H.; Voland, A.; Armstrong, V. W.; Grone, H. J.; Oellerich, M.;
Wieland, E. Proteomics 2004, 4, 2728.
14. Spahn-Langguth, H.; Benet, L. Z. Drug Metab. Rev. 1992, 24, 5.
15. Stachulski, A. V.; Harding, J. R.; Lindon, J. C.; Maggs, J. L.; Park, B. K.; Wilson, I. D.
J. Med. Chem. 2006, 49, 6931.
In conclusion, we have achieved convenient high-yielding prep-
arations of the phase 2 metabolites of mycophenolic acid, com-
16. Shipkova, M.; Strassburg, C. P.; Braun, F.; Streit, F.; Grone, H. J.; Armstrong, V.
W.; Tukey, R. H.; Oellerich, M.; Wieland, E. Brit. J. Pharmacol. 2001, 132, 1027.

A. E. Jones et al. / Tetrahedron Letters 50 (2009) 4973–4977
4977
17. Ando, K.; Suzuki, S.; Arita, M. J. Antibiot. 1970, 23, 408.
18. Kittleman, M.; Rheinegger, U.; Espigat, A.; Oberer, L.; Aichholz, R.; Francotte, E.;
Ghisalba, O. Adv. Synth. Catal. 2003, 345, 825.
19. Walsh, J. S.; Patanella, J. E.; Halm, K. A.; Facchine, K. L. Drug Metab. Dispos. 1995,
23, 869.
26. Hongu, R. M.; Saito, K.; Tsujihara, K. Synth. Commun. 1999, 29, 2775.
27. Jones, A. E., M. Phil. Thesis, University of Liverpool, in preparation. A similar
procedure was used to prepare the phenolic allyl ether of 1: Meath, P.;
Stachulski, A. V., unpublished observations.
28. Sollinger, H. W. Clin. Transplant. 2004, 18, 485.
20. Perrie, J. A.; Harding, J. R.; Holt, D. W.; Johnston, A.; Meath, P.; Stachulski, A. V.
Org. Lett. 2005, 7, 2591.
29. Shipkova, M.; Armstrong, V. W.; Oellerich, M.; Wieland, E. Expert Opin. Drug
Metab. Toxicol. 2005, 1, 505.
21. Bowkett, E. R.; Harding, J. H.; Maggs, J. L.; Perrie, J. A.; Park, B. K.; Stachulski, A.
V. Tetrahedron 2007, 63, 7596.
22. Bollenback, G. N.; Long, J. W.; Benjamin, D. G.; Lindquist, J. A. J. Am. Chem. Soc.
1955, 77, 3310.
30. Shipkova, M.; Armstrong, V. W.; Weber, L.; Niedmann, P. D.; Wieland, E.; Haley,
J.; Tönshoff, B.; Oellerich, M. Ther. Drug Monit. 2002, 24, 390.
31. Asif, A. R.; Armstrong, V. W.; Voland, A.; Wieland, E.; Oellerich, M.; Shipkova,
M. Biochimie 2007, 89, 393.
23. (a) Schmidt, R. R. Angew. Chem., Int. Ed. Engl. 1986, 25, 212; For examples, in the
glucuronic acid series, see: (b) Fischer, B.; Nudelman, A.; Ruse, M.; Herzig, J.;
Gottlieb, H. E.; Keinan, E. J. Org. Chem. 1984, 49, 4988.
24. McCorkindale, N. J.; Baxter, R. L.; Turner, W. B. Tetrahedron 1981, 37, 2131.
25. Brown, R. T.; Scheinmann, F.; Stachulski, A. V. J. Chem. Res. (S) 1997, 370.
32. Shipkova, M.; Schutz, E.; Armstrong, V. W.; Niedmann, P. D.; Oellerich, M.;
Wieland, E. Clin. Chem. 2000, 46, 365.
33. Mino, Y.; Naito, T.; Matsushita, T.; Kagawa, Y.; Kawakami, J. J. Pharm. Biomed.
Anal. 2008, 46, 603.