Synthesis of a cell-permeable analogue of a glycosylphosphatidylinositol (GPI) intermediate that is toxic to the living bloodstream form of Trypanosoma brucei

Article abstract of DOI:10.1016/j.tetlet.2005.08.112

The synthesis of two cell-permeable analogues related to a GPI intermediate is described for studies with living trypanosomes and human (HeLa) cells. One of the analogues is metabolised by the former GPI pathway and is toxic to the parasite Trypanosoma br

Full text of DOI:10.1016/j.tetlet.2005.08.112

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 7419–7421
Synthesis of a cell-permeable analogue of a
glycosylphosphatidylinositol (GPI) intermediate that is toxic to
the living bloodstream form of Trypanosoma brucei
Arthur Crossman, Jr., Terry K. Smith, Michael A. J. Ferguson* and John S. Brimacombe
School of Life Sciences, Division of Biological Chemistry and Molecular Microbiology, University of Dundee,
The Wellcome Trust Biocentre, Dundee DD1 5EH, Scotland, UK
Received 1 June 2005; accepted 19 August 2005
Available online 9 September 2005
Abstract—The synthesis of two cell-permeable analogues related to a GPI intermediate is described for studies with living trypano-
somes and human (HeLa) cells. One of the analogues is metabolised by the former GPI pathway and is toxic to the parasite Try-
panosoma brucei but not to human cells.
Ó 2005 Elsevier Ltd. All rights reserved.
The World Health Organisation (WHO)1 estimates that
sleeping sickness threatens over 60 million people in 36
countries of sub-Saharan Africa. African sleeping sick-
ness is a debilitating and often fatal disease caused by
trypanosomes transmitted by the tsetse fly. These extra-
cellular protozoan parasites are able to survive in the
human bloodstream by virtue of a dense cell-surface
coat composed of variant surface glycoprotein (VSG).2
This dense palisade of VSG is attached to the plasma
membrane by means of glycosylphosphatidylinositol
(GPI) anchors.3 It is our contention that disruption of
GPI biosynthesis would seriously impair the parasiteÕs
ability to survive in the host.4,5 However, it must be
stressed that parasite-specific inhibitors of GPI biosyn-
thesis are required since essential GPI-anchored glyco-
proteins are found in mammals.6 The GPI biosynthetic
pathway in mammalian cells appears to be broadly simi-
lar to that of the bloodstream form of Trypanosoma
brucei, the causative agent of African sleeping sickness.
One major difference between the parasite and mamma-
lian pathways is the relative order of D-mannosylation
and inositol acylation of a-D-GlcpN-(1!6)-PI (1,
Fig. 1). T. brucei strictly D-mannosylates a-D-GlcpN-
(1!6)-PI (1) to a-D-Manp-(1!4)-a-D-GlcpN-(1!6)-PI
OR2
O
OH
OCOCnH2n+1
OCOCnH2n+1
O
HO
O
HO
P
O
O
NH2
HO
R1O
O
OH
R1
H
R2
1
H
H
2 α-D-Manp
3 α-D-Manp
COC15H31
Figure 1.
(2) prior to inositol acylation (!3), whereas the reverse
holds for the mammalian pathway.3
We have already synthesised an extensive library of var-
ious analogues of a-D-GlcpN-(1!6)-PI (1)7 in order to
ascertain the substrate specificities of T. brucei and
human (HeLa) GPI biosynthetic enzymes in cell-free
systems.5,8 For the purpose of this Letter, the following
results are relevant: (a) a-D-GalpN-(1!6)-PI (4, Fig. 2)
was not metabolised by either the trypanosomal or the
mammalian system, presumably due to epimerisation
of the hydroxyl group at the site of D-mannosylation;
(b) the inositol-methylated analogue 5 is D-mannosyl-
ated by the T. brucei GPI pathway but not by the
Keywords: Trypanosoma brucei; Cell-permeable glycosylphosphatidyl-
inositol (GPI) analogues; African sleeping sickness; Bioactivatable
acetoxymethyl ester.
*
Corresponding author. Tel.: +44 (0)1382 344219; fax: +44 (0)1382
34889; e-mail: m.a.j.ferguson@dundee.ac.uk
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.08.112

7420
A. Crossman, Jr. et al. / Tetrahedron Letters 46 (2005) 7419–7421
OR3
R4
diester prevents these analogues from diffusing across
the plasma membrane of living trypanosomes.
HO
HO
HO
O
Bearing in mind these results in designing potential
inhibitors of the T. brucei GPI pathway, we synthesised
the neutral phosphoric triesters 6 and 7 (Fig. 2) possess-
ing a bioactivatable acetoxymethyl (AM) protecting
group.10 It was anticipated that these neutral phosphoric
triesters would diffuse across the plasma membrane of
living trypanosomes and revert to the charged phospho-
ric diesters through the action of resident cytosolic ester-
ase(s).11 The closely related analogues 6 and 7 contain a
C18 alkyl chain in place of the phospholipase-sensitive
diacylglycerol of the natural substrate 1, and are inosi-
tol-methylated (cf. 5) to prevent substrate recognition
by human GPI D-mannosyltransferases. Of the two,
only the D-glucosamine-containing analogue 7 should
be metabolised by living trypanosomes, whereas the D-
galactosamine-containing analogue 6 functions as a
non-metabolisable control.
NH2
HO
R2
R1
O
OH
R3
H
R1
R2
H
R4
O
OCOC15H31
OCOC15H31
4
5
6
7
OH
O
O
O
O
P
O
O
OCOC15H31
OCOC15H31
H
OH
H
CH3
CH3
CH3
P
O
OCH2OAc
OH
H
O
O
OC18H37
P
O
OH
OCH2OAc
OC18H37
P
O
The synthesis of the D-galactosamine-containing ana-
logue 6 began by coupling the known trichloroacetimi-
date 812 with the inositol acceptor 97c in the presence
of a catalytic amount of TMSOTf to furnish the pseudo-
disaccharide 10 as a roughly 1:1 mixture of anomers
(Scheme 1). The methoxybenzyl protecting group was
next removed with trifluoroacetic acid to give, after
radial-band chromatography, the desired a anomer 11
in 34% yield. Compounds 11 and 127e reacted in the
Figure 2.
human (HeLa) GPI pathway; (c) only those GPI ana-
logues bearing a negatively charged phosphoric diester
linkage are metabolised by the T. brucei cell-free sys-
tem.9 However, the negative charge on the phosphoric
OCH3
OMBn
OCH3
OH
BnO
BnO
BnO
BnO
O
BnO
OCH3
OBn
BnO
OBn
BnO
BnO
BnO
O
i
O
O
ii
BnO
BnO
N3
OMBn
BnO
BnO
BnO
BnO
N3
HO
N3
8
O
O
O
CCl3
NH
BnO
9
10
11
BnO
OCH3
OCH3
OH
OCH3
O
BnO
BnO
O
OCH2OAc
NHEt3
OC18H37
BnO
BnO
BnO
BnO
BnO
BnO
O
OC18H37
P
P
O
O
NHEt3
O
N3
iii, iv
v
O
N3
N3
O
OC18H37
H
BnO
R2
BnO
R2
BnO
R2
P
O
12
O
O
O
R1
R1
R1
OBn
OBn
OBn
11 R1 = OBn, R2 = H
13 R1 = OBn, R2 = H
14 R1 = OBn, R2 = H
15 R1 = H, R2 = OBn
16 R1 = H, R2 = OBn
17 R1 = H, R2 = OBn
vi
OCH3
HO
OCH3
HO
OCH2OAc
O
HO
HO
O
OC18H37
O
OC18H37
HO
HO
P
P
vii
-AcOH, -HCHO
O
O
NH2
O
NH2
O
HO
R2
R1
HO
R2
R1
O
O
OH
OH
6 R1 = OH, R2 = H
18 R1 = OH, R2 = H
7 R1 = H, R2 = OH
19 R1 = H, R2 = OH
˚
Scheme 1. Reagents and conditions: (i) TMSOTf, MS 4 A, Et2O, rt, 45 min, 80%; (ii) TFA, CH2Cl2, rt, 1 h, 34%; (iii) PivCl, pyridine, rt, 1 h; (iv) I2,
pyridine/H2O {19:1}, rt, 45 min; 43% for 13, 80% for 16; (v) DIEA, acetoxymethyl bromide, THF, rt, overnight; 65% for 14, 94% for 17; (vi) H2, 20%
Pd(OH)2/C, 1:1 THF–MeOH, 3 atm, rt, 2 h; 95% for 6, 96% for 7; (vii) cytosolic esterase(s).

A. Crossman, Jr. et al. / Tetrahedron Letters 46 (2005) 7419–7421
7421
presence of 3 equiv of pivaloyl chloride13,14 to give the
phosphoric diester 13 (isolated as the triethylammonium
salt) following oxidation of the intermediate phosphoric
diester with iodine in wet pyridine. Treatment of 13 with
acetoxymethyl bromide in THF in the presence of DIEA
afforded the AM derivative 14 as a mixture of two dia-
stereomers. Finally, hydrogenolysis of the benzyl
groups, with concurrent reduction of the azido group,
provided 6 as a mixture of diastereomers.
2. Cross, G. A. Parasitology 1975, 71, 393–417.
3. Ferguson, M. A. J.; Brimacombe, J. S.; Brown, J. R.;
Crossman, A.; Dix, A.; Field, R. A.; Guther, M. L. S.;
¨
Milne, K. G.; Sharma, D. K.; Smith, T. K. Biochim.
Biophys. Acta 1999, 1455, 327–340.
4. (a) Ferguson, M. A. J. J. Cell Sci. 1999, 112, 2799–2808;
(b) McConville, M. J.; Menon, A. K. Mol. Membr. Biol.
2000, 17, 1–16; Morita, Y. S.; Acosta-Serrano, A.;
Englund, P. T. In Oligosaccharides in Chemistry and
Biology—A Comprehensive Handbook; Ernst, P., Sinay, P.,
Hart, G. W., Eds.; Wiley-VCH: Weinheim, Germany,
2000; pp 417–433; (c) Kinoshita, T.; Inoue, N. Curr. Opin.
Chem. Biol. 2000, 4, 632–638.
An identical sequence of reactions was used to
synthesise the D-glucosamine-containing analogue 7
(157c!16!17!7).
5. Smith, T. K.; Crossman, A.; Paterson, M. J.; Borissow, C.
N.; Brimacombe, J. S.; Ferguson, M. A. J. J. Biol. Chem.
2002, 277, 37147–37153.
Details of in vivo studies with the AM-protected GPI
analogues 6 and 7 have been published elsewhere.9 It
suffices here to note that both 6 and 7 fulfilled our expec-
tations and, after diffusing across the plasma membrane,
were hydrolysed by the cytosolic esterase(s) of living try-
panosomes to the charged phosphoric diesters 18 and
19, respectively. Only the D-glucosamine-containing
analogue 19 is subsequently metabolised and this
resulted in killing of the trypanosomes in a concentra-
tion- and time-dependant manner.15 The mechanism
for killing the parasite is not clear, but there are strong
indications9 that a metabolite of 19 that accumulates
within the endoplasmic reticulum, possibly D-Manp3-a-
D-GlcpN-(1!6)-(2-O-methylinositol)-P-C18, is respon-
sible. We also showed that the deprotected form 19 is
parasite specific since it is neither a substrate nor an
inhibitor in a HeLa cell-free system. These results and
the complete lack of toxicity of the non-metabolisable
D-galactosamine-containing analogue 18 provide the
first unambiguous chemical validation of the T. brucei
GPI pathway as a drug target.
6. Tarutani, M.; Itami, S.; Okabe, M.; Ikawa, M.; Tezuka,
T.; Yoshikawa, K.; Kinoshita, T.; Takeda, J. Proc. Natl.
Acad. Sci. U.S.A. 1997, 94, 7400–7405.
7. (a) Cottaz, S.; Brimacombe, J. S.; Ferguson, M. A. J. J.
Chem. Soc., Perkin Trans. 1 1993, 2945–2951; (b) Cottaz,
S.; Brimacombe, J. S.; Ferguson, M. A. J. J. Chem. Soc.,
Perkin Trans. 1 1995, 1673–1678; (c) Crossman, A., Jr.;
Brimacombe, J. S.; Ferguson, M. A. J. J. Chem. Soc.,
Perkin Trans. 1 1997, 2769–2774; (d) Crossman, A., Jr.;
Brimacombe, J. S.; Ferguson, M. A. J.; Smith, T. K.
Carbohydr. Res. 1999, 321, 42–51; (e) Crossman, A., Jr.;
Paterson, M. J.; Ferguson, M. A. J.; Smith, T. K.;
Brimacombe, J. S. Carbohydr. Res. 2002, 337, 2049–2059;
(f) Dix, A. P.; Borissow, C. N.; Ferguson, M. A. J.;
Brimacombe, J. S. Carbohydr. Res. 2004, 339, 1263–
1277.
8. (a) Smith, T. K.; Cottaz, S.; Brimacombe, J. S.; Ferguson,
M. A. J. J. Biol. Chem. 1996, 271, 6476–6482; (b) Smith, T.
K.; Sharma, D. K.; Crossman, A.; Dix, A.; Brimacombe,
J. S.; Ferguson, M. A. J. EMBO J. 1997, 16, 6667–
6675; (c) Smith, T. K.; Sharma, D. K.; Crossman, A.;
Brimacombe, J. S.; Ferguson, M. A. J. EMBO J. 1999, 18,
5922–5930; (d) Smith, T. K.; Paterson, M. J.; Crossman,
A.; Brimacombe, J. S.; Ferguson, M. A. J. Biochemistry
2000, 39, 11801–11807; (e) Smith, T. K.; Crossman, A.;
Borissow, C. N.; Paterson, M. J.; Dix, A.; Brimacombe,
J. S.; Ferguson, M. A. J. EMBO J. 2001, 20, 3322–
3332.
Acknowledgements
This work was supported by a programme grant from
The Wellcome Trust (071463).
9. Smith, T. K.; Crossman, A.; Brimacombe, J. S.; Ferguson,
M. A. J. EMBO J. 2004, 23, 4701–4708.
10. Schultz, C. Bioorg. Med. Chem. 2003, 11, 885–898.
11. In this prodrug approach, the enzymatic hydrolysis of
acetate liberates a labile hemiacetal intermediate that
spontaneously releases formaldehyde and the phosphoric
diester.
12. Grundler, G.; Schmidt, R. R. Liebigs Ann. Chem. 1984,
1826–1847.
13. Slaghek, T. M.; Maas, A. M.; Kamerling, J. P.; Kochet-
kov, N. K. Carbohydr. Res. 1991, 211, 25–39.
Supplementary data
Experimental procedures and partial characterisation
data for all the compounds synthesised in this Letter
are available online. Supplementary data associated
with this article can be found, in the online version, at
doi:10.1016/j.tetlet.2005.08.112.
14. Nikolaev, A. V.; Ivanova, I. A.; Shibaev, N. K.; Kochet-
kov, N. K. Carbohydr. Res. 1990, 204, 65–78.
15. Identical results were obtained with the N-acetylated form
of 7, which is converted into 19 in vivo by the combined
actions of cytosolic esterase(s) and a de-N-acetylase.
References and notes
1. http://www.who.int/topics/trypanosomiasis_african/en/.