Fleetamine (3-O-α-d-glucopyranosyl-swainsonine): The synthesis of a hypothetical inhibitor of endo-α-mannosidase

Article abstract of DOI:10.1016/j.tetasy.2012.06.011

3-O-α-d-Glucopyranosyl-swainsonine was originally proposed ¹⁷ as a potential inhibitor of the mammalian enzyme endo-α-mannosidase, but its synthesis has not been reported. Herein we report the total synthesis of this enigmatic compound, utilizing a halide-ion catalysed glycosylation of a swainsonine lactam with a glucosyl iodide donor as the key step. The resulting inhibitor was evaluated as an inhibitor of human endo-α-mannosidase, and as a ligand for bacterial orthologs from Bacteroides thetaiotaomicron and Bacteroides xylanisolvens, including active-centre variants, although no evidence for binding or inhibition was observed. The surprising lack of binding was rationalized by using structural alignment with an endo-α-mannosidase inhibitor complex, which identified deleterious interactions with the swainsonine piperidine ring and an essential active site residue. 2012 Elsevier Ltd.

Full text of DOI:10.1016/j.tetasy.2012.06.011

Tetrahedron: Asymmetry 23 (2012) 992–997
Contents lists available at SciVerse ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Fleetamine (3-O-
a
-D-glucopyranosyl-swainsonine): the synthesis
of a hypothetical inhibitor of endo-a-mannosidase
Tim Quach a, Sammi Tsegay a, Andrew J. Thompson b, Nikolay V. Kukushkin c, Dominic S. Alonzi c,
Terry D. Butters c, Gideon J. Davies b, Spencer J. Williams a,
⇑
a School of Chemistry and Bio21 Molecular Science and Biotechnology Institute, University of Melbourne, 30 Flemington Rd., Parkville 3010, Australia
b Department of Chemistry, University of York, Heslington, York YO10 5DD, United Kingdom
c Department of Biochemistry, Glycobiology Institute, University of Oxford, Oxford OX1 3QU, United Kingdom
a r t i c l e i n f o
a b s t r a c t
-Glucopyranosyl-swainsonine was originally proposed17 as a potential inhibitor of the mamma-
-mannosidase, but its synthesis has not been reported. Herein we report the total syn-
Article history:
Received 28 May 2012
Accepted 19 June 2012
3-O-
a-D
lian enzyme endo-
a
thesis of this enigmatic compound, utilizing a halide-ion catalysed glycosylation of a swainsonine lactam
with a glucosyl iodide donor as the key step. The resulting inhibitor was evaluated as an inhibitor of
human endo-a-mannosidase, and as a ligand for bacterial orthologs from Bacteroides thetaiotaomicron
and Bacteroides xylanisolvens, including active-centre variants, although no evidence for binding or inhi-
bition was observed. The surprising lack of binding was rationalized by using structural alignment with
an endo-
a-mannosidase inhibitor complex, which identified deleterious interactions with the swainso-
nine piperidine ring and an essential active site residue.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
of the Golgi apparatus that uniquely cleaves mannoside linkages
internally, within the first branch of an N-glycan chain (Fig. 1).7
Cells in which glucosidase I or II has been inhibited or subjected
to genetic knockout retain 40–80% normal N-glycan processing
The majority of mammalian proteins are modified with N-
linked glycans, which help to control protein folding and stability,
physiological properties and cellular communication.1–3 N-linked
glycan biosynthesis occurs within the secretory pathway, and com-
mences with the co-translational installation of a pre-formed 14-
mer oligosaccharide in the endoplasmic reticulum (Fig. 1).2 The
glycan structures are trimmed by the exo-acting
and II that sequentially remove single glucose residues. These glu-
cosidases can be inhibited by sugar-shaped heterocycles of the
gluco configuration such as deoxynojirimycin 1 (DNJ) and castano-
spermine 2 (CST) (Fig. 2).4 Deglucosylated N-linked glycans are
substrates for exo-acting mannosidases in the endoplasmic reticu-
lum and Golgi apparatus, which ultimately trim these glycans to a
core hexasaccharide, prior to their elaboration to hybrid and com-
plex-type N-glycans. The exo-mannosidases of the ER and Golgi are
function as a result of endo-
a CST-blockade of -glucosidase I and II, HepG2 cells produced
N-linked glycans with a normal structure owing to the rescue by
endo- -mannosidase, as demonstrated by the release of Glc1–3
Man oligosaccharides.11 Specific inhibitors of endo-
-mannosidase
have been developed by modifying the exo-mannosidase specific
inhibitors, DMJ 1 and isofagomine12,13 with an endo-
-mannosi-
dase targeting -gluco-
-1,3-linked glucosyl residue.14–16 Thus,
syl-1,3-deoxymannojirimycin 5 (GlcDMJ) has been widely used
as a specific endo-
-mannosidase inhibitor.15 Recently, we re-
ported the development of -glucosyl-1,3-isofagomine 6 (GlcIFG),
which is a 40-fold stronger binding ligand of Bacteroides thetaiota-
a
-mannosidase.8–10 Similarly, during
a
a
-
a
-glucosidases I
a
D
-
a
a
a
a
a
omicron GH99 endo-a-mannosidase (BtGH99) compared to GlcDMJ
inhibited by
D
-mannose-mimicking nitrogenous heterocycles, such
5.16 Ternary complexes of GlcDMJ 5 or GlcIFG 6 (in the ꢀ2 and ꢀ1
as deoxymannojirimycin 3 (DMJ; inhibitor of ER mannosidase I)
subsites) and a-1,2-mannobiose (in the +1 and +2 subsites) with
and swainsonine 4 (inhibitor of Golgi mannosidase II).4
Bacteroides xylanisolvens GH99 endo-a-mannosidase (BxGH99) al-
endo-a-Mannosidase (classified into the carbohydrate-active
lowed the active site residues to be defined and was suggestive
enzymes family GH99;5,6 www.cazy.org) is a glycoside hydrolase
of an enzyme mechanism involving
intermediate.16
a 1,2-anhydro sugar
A previous report in this journal proposed the structure of a
new potential inhibitor of endo- -mannosidase, based on the glu-
⇑
Corresponding author. Tel.: +61 3 8344 2422; fax: +61 3 9347 5180.
E-mail address: sjwill@unimelb.edu.au (S.J. Williams).
a
0957-4166/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetasy.2012.06.011

T. Quach et al. / Tetrahedron: Asymmetry 23 (2012) 992–997
993
Figure 1. Summary of the major steps of the N-glycan trimming in the endoplasmic reticulum and Golgi apparatus. Red block arrows indicate the sites of endo-
a-
mannosidase cleavage. Boxed abbreviations represent inhibitors for the indicated steps: DNJ = deoxynojirimycin 1, CST = castanospermine 2, DMJ = deoxymannojirimycin 3,
SW = swainsonine 4, GlcDMJ = a-D-glucopyranosyl-1,3-deoxymannojirimycin 5, GlcIFG = a-D-glucopyranosyl-1,3-isofagomine 6.
OH
Thus, it is unclear whether swainsonine or DMJ might represent
the preferred cores for elaboration to an -1,3-glucosylated analogue
a
HO
for endo-a-mannosidase inhibition. However, to the best of our
HO
HO
N
N
NH
OH
HO
HO
HO
HO
knowledge, no synthesis of Fleetamine has been reported.23 Herein
we report the first synthesis of Fleetamine and its evaluation as an
OH
NH
H
H
HO
HO
HO
HO
OH
inhibitor of bacterial and mammalian endo-a-mannosidases.
deoxynojirimycin 1 castanospermine 2
deoxymannojirimycin 3
swainsonine 4
2. Results and discussion
N
HO
HO
OH
HO
HO
Our approach to Fleetamine required the identification of an
efficient route to a selectively-protected swainsonine derivative.
Whilst dozens of syntheses of swainsonine have been reported,24,25
only a handful of these are sufficiently practical enough to allow
the synthesis of adequate quantities to enable elaboration to the
glycoside. Additionally, the tertiary amine of swainsonine is basic
and may interfere with reagents used for glycosylation. With these
concerns in mind, we identified the work of Pearson et al. as a prac-
tical route to swainsonine, which proceeds via a d-lactam interme-
diate that appeared to constitute a promising starting point for our
synthesis.26,27 Accordingly, following Pearson’s route, we synthe-
sized gram quantities of the advanced d-lactam 8 in eight steps
H
HO
O
NH
HO
O
O
NH
O
OH
HO
HO
HO
HO
HO
HO
HO
HO
HO
O
O
HO
HO
GlcDMJ 5
GlcIFG 6
Fleetamine 7
Figure 2. Structures of various glycosidase inhibitors.
cosylation of swainsonine, namely 3-O-
swainsonine 7 (Fleetamine).
a-
D
-glucopyranosyl-
17
These authors noted that whereas
DMJ is a strong inhibitor of ER mannosidase I and a weak inhibitor
of mannosidase II, swainsonine possessed a reciprocal specificity.17
from
D
-ribose.27 Benzylation of alcohol 8 afforded monobenzyl
ether 9, which upon exposure to TsOH in methanol afforded diol
10. Identification of suitable protecting groups for this derivative
was problematic. After some experimentation, we ultimately
elected to selectively protect the cis-diol of 10 as a monobenzyl
ether. Alternative approaches involving monosilylation (with tert-
butylchlorodiphenylsilane) occurred with the desired regioselec-
tivity; however the resultant alcohol was a refractory glycosyl
We have coined the name Fleetamine for this compound based on the original
proposal of this structure by Fleet et al.,17 and in recognition of his seminal
contributions to the development of glycosidase inhibitors,18 including the first total
synthesis of (ꢀ)-swainsonine.19 To describe this compound we use an informal
numbering system that highlights the structural similarity of swainsonine to
mannose and is consistent with the binding orientation of swainsonine in co-
complexes with various glycoside hydrolases.20–22 The numbering used in the
Section 4 reflects the formal numbering used for indolizidines.4

994
T. Quach et al. / Tetrahedron: Asymmetry 23 (2012) 992–997
acceptor. Treatment of 10 with NaH and benzyl bromide in DMF
afforded dibenzyl ether 12, accompanied by tribenzyl ether 11,
which could be recycled (hydrogenolysis, then isopropylidenation
to afford 8). Evidence for the structure of the dibenzyl ether was
obtained by detailed 1H–1H COSY analysis of the acetate derivative
13 (Fig. 3).
up to 1.0 mM, no inhibition was noted. Fleetamine was also evalu-
ated as a ligand for wild-type BtGH99 using isothermal titration
calorimetry (ITC). Using ITC, no binding was observed to either
the wild-type or E154A variant (vide infra) BtGH99. Attempts at
complex formation using both soaking and co-crystallization of
Fleetamine with BxGH99 at ligand concentrations of up to
50 mM also failed to reveal any binding.
O
O
Whilst the rationale for adding an endo-
ing -1,3-linked glucose to traditional exo-mannosidase inhibitors
appears sound in principle, and is effective in converting DMJ 3 and
isofagomine into endo- -mannosidase inhibitors, our results high-
a-mannosidase-target-
O
HO
a
N
N
OH
a
b
BnO
HO
H
H
HO
OH
a
O
N
O
O
O
D-ribose
light the limitations of this approach. In order to gain insight into
the poor inhibitory properties of Fleetamine 7, we performed a
structural alignment using the macromolecular model building
software COOT.33 A single swainsonine molecule 4 was positioned
over the deoxymannojirimycin moiety within the previously
solved BxGH99–GlcDMJ co-complex (PDB entry 4AD3)16 using
Least Squares Fitting of matching atom positions between the
two molecules. Within the solved enzyme–inhibitor complex, the
O4 and O6 atoms of DMJ appear at appropriate H-bonding dis-
8
9
O
O
O
N
N
c
d
BnO
+
BnO
BnO
H
HO
H
H
OH
BnO
11
OBn
RO
OBn
10
12 R = H
13 R = Ac
tances from the Oe1 and Oe2 atoms of glutamate 156 (E156, equiv-
Figure 3. Reagents and conditions: (a) eight steps;27 (b) BnBr, NaH, Bu4NI, DMF,
86%; (c) TsOH, MeOH, 87%; (d) BnBr, NaH, Bu4NI, DMF, 29% dibenzyl ether 12, 35%
tribenzyl ether 11; (f) Ac2O, pyridine, quantitative.
alent to E154 of BtGH99), an interaction required for catalysis as
shown by site-directed mutagenesis studies.16 The structural align-
ment of swainsonine 4 highlights a likely breakdown of this impor-
tant interaction and a rationale for the apparent poor binding of 7
The next step required the challenging
a-glucosylation of the
to endo-
approximately 0.4 Å closer to O
a
-mannosidase. The O1 atom of swainsonine appears
1 (as calculated by COOT), whilst
hindered 3-hydroxy group of 12, which is located on the endo-face,
adjacent to a bridgehead, and flanked by a cis-vicinal benzyl ether.
After testing several glucosylation methods, we obtained some
success with Lemieux’s halide-ion catalysed glycosylation,28 using
glycosyl iodide 14.29 Thus, treatment of 12 with iodide 14 (pre-
pared from the corresponding anomeric acetate by treatment with
iodotrimethylsilane)30 afforded the target glycoside 15 in low yield
(27%; unreacted 12 was recovered in 51%) but in sufficient
amounts to proceed (Fig. 4).
e
there is no equivalent of the H-bonding of E156 and the O6 posi-
tion of GlcDMJ. Carbon 7 of swainsonine approaches the side chain
of E156 to within a distance of approximately 2.2 Å, which whilst
perhaps being insufficient to be judged as a serious steric clash,
this close interaction coupled with a loss of hydrogen-bonding ex-
plains the poor inhibitory properties observed for 7. In order to
probe whether removal of the equivalent residue in BtGH99 could
reinstate inhibitor binding, we generated the E154A BtGH99 mu-
tant, which was designed to create additional space in the active
site. The ITC of Fleetamine and E154A BtGH99 still failed to display
any evidence of binding. It thus appears that even with the extra
space created within the ꢀ1 subsite in this mutant, the bulky nat-
ure of the swainsonine moiety present within Fleetamine may pre-
vent access into the enzyme active site (Fig. 5).
O
O
N
AcO
BnO
O
a
BnO
BnO
N
H
BnO
O
+
BnO
O
OBn
H
BnO
I
HO
OBn
BnO
BnO
15
14
12
AcO
O
HCl
N
N
HO
AcO
b
c
H
H
O
OH
7.HCl
O
OAc
HO
O
AcO
O
HO
HO
AcO
16
AcO
HO
AcO
Figure 4. Reagents and conditions: (a) Bu4NI, 2,4,6-tri-tert-butylpyrimidine, tolu-
ene, 4d, 27%; (b) (i) H2, Pd(OH)2-C, MeOH/EtOAc/AcOH; (ii) Ac2O, pyr, 77% over two
steps; (c) (i) BH3.Me2S, THF; (ii) NaOMe, MeOH then HCl, 67% over two steps.
Figure 5. Superposition of swainsonine 4 with GlcDMJ 5 in a complex with BxGH99
(PDB ID: 4AD3)16 in a divergent stereoview prepared using CCP4Mg.34 Shown is a
Least Squares Fit (LSQ) alignment of swainsonine with the deoxymannojirimycin
fragment of 5 prepared using COOT.35 GlcDMJ is shown as a disaccharide with
carbons in grey and oxygen atoms in red. Swainsonine is shown with carbon atoms
in yellow and oxygens in dark red. Amino acid residues required for catalysis/
binding are labelled; H-bonding interactions between E156 and GlcDMJ are
illustrated as a dashed line.
Deprotection of disaccharide 15 required considerable experi-
mentation to determine the optimum order and choice of reagents.
Unexpectedly, the free base of Fleetamine proved to be unstable.
Ultimately, a sequence involving hydrogenolysis of disaccharide
lactam 15, followed by acetylation to afford hexaacetate 16, and
then borane dimethyl sulfide reduction, deacetylation and imme-
diate conversion to the hydrochloride afforded FleetamineꢁHCl
7ꢁHCl.
3. Conclusion
The inhibitory activity of Fleetamine was explored using
Herein we have reported the synthesis of Fleetamine 7, a puta-
human recombinant endo-
a
-mannosidase with
a
fluorescent
tive inhibitor of endo-
a-mannosidase that has remained an enigma
Glc3Man7GlcNAc2 substrate.31,32 However, at concentrations of
since its original proposal by Fleet et al.,17 Despite the enticing

T. Quach et al. / Tetrahedron: Asymmetry 23 (2012) 992–997
995
rationale for its proposal as an inhibitor of endo-
we found no evidence that it acts as an inhibitor of either the hu-
man or bacterial forms of this enzyme. Fleet has suggested that the
a
-mannosidase,17
was heated at reflux. After 21 h, additional p-TsOH (50 mg) was
added, and the resulting mixture was heated at reflux for 5 h.
The reaction was cooled to rt, then Et3N (1 mL) was added and
the mixture stirred at rt for 15 min. Removal of the volatiles under
reduced pressure gave a crude residue, which was purified by flash
chromatography (10% MeOH/EtOAc) to give 10 (295 mg, 87%) as a
transition state for exo-acting
a-mannosidase cleavage may in
some cases be better inhibited by pyranose mimics, and in others
by furanose mimics, in which cases swainsonine provides more
effective inhibition than DMJ.4,17 In this context, the transition
white fluffy material.
½
a 2D4
ꢃ
¼ ꢀ30:9 (c 1.1, CHCl3). 1H NMR
states of the bacterial and mammalian endo-a-mannosidases can
(500 MHz, CDCl3): d 7.39–7.30 (m, 5H, Ph), 4.64 (ABq, J = 12.0 Hz,
2H, CH2Ph), 4.34 (m, 1H, H2), 4.20 (m, 1H, H1), 3.86 (ddd,
J = 11.0, 9.0, 4.0 Hz, 1H, H8), 3.69 (ddd, J = 12.0, 8.0, 3.0 Hz, 1H,
H3), 3.45 (d, J = 8.5 Hz, 1H, H8a), 3.34 (dd, J = 12.0, 10.0 Hz, 1H,
H3), 2.50 (m, 1H, H6), 2.32 (m, 1H, H6), 2.14 (m, 1H, H7), 1.76
(m, 1H, H7). 13C NMR (125 MHz, CDCl3): d 169.4 (C@O), 138.3,
128.6, 128.0, 127.8 (Ph), 71.6, 71.2, 71.0, 69.9, 65.6, 49.5, 29.5,
be considered to be ‘pyranose-like’ given that GlcIFG and GlcDMJ
are better inhibitors than Fleetamine.
4. Experimental
4.1. General
26.5. FTIR:
m 2961, 2913, 2852, 2233, 1599, 1490, 1452, 1442,
1377, 1358, 1323, 1112, 1069, 1025, 911, 829, 754, 690 cmꢀ1
.
HR-ESIMS m/z 278.13867 [M+H]+, calcd 278.13868 for C15H20NO4.
All reactions were performed under a nitrogen atmosphere and
glassware was flame-dried prior to use when necessary. Dichloro-
methane and THF were dried over alumina according to the meth-
od reported by Pangborn et al.,36 Reactions were monitored by TLC
4.1.3. (1S,2R,8R,8aR)-1,2,8-Tris(benzyloxy)hexahydroindolizin-
5(1H)-one 11 and (1S,2R,8R,8aS)-2,8-bis(benzyloxy)-1-hydroxy-
hexahydroindolizin-5(1H)-one 12
analysis (pre-coated silica gel 60 F254 plates, 250 lm layer thick-
ness on aluminium) and visualized with a 254 nm UV light and/
or by staining with a CAM solution (5 g of cerium sulfate, 25 g of
ammonium molybdate, 50 mL of concd H2SO4 and 450 mL of
H2O). Flash chromatography was performed according to the
method of Still et al.37 1H NMR spectra were obtained at 400 or
500 MHz in CDCl3 or otherwise as noted. Chemical shifts are re-
ported in parts per million with the residual solvent peak used as
an internal standard. Multiplicity is denoted by: s = singlet,
d = doublet, t = triplet, q = quartet, quint = quintet, m = multiplet,
br s = broad singlet, dd = doublet of doublet, dt = doublet of triplet,
dq = doublet of quartet, tt = triplet of triplet. 13C NMR spectra were
measured at 100 or 125 MHz using a proton-decoupled pulse se-
quence with a delay of 3 s. IR spectra were obtained on a Perkin–
Elmer Spectrum One FTIR spectrometer with a zinc selenide/dia-
mond Universal ATR sampling accessory as a thin film.
NaH (60% w/w, 57.8 mg, 1.45 mmol), BnBr (107 lL,
0.903 mmol) and Bu4NI (44.5 mg, 0.120 mmol) were added to a
solution of diol 10 (334 mg, 0.836 mmol) in THF (25 mL) at 0 °C
and the reaction stirred at 4 °C for 18 h, and then at rt for
30 min. The reaction mixture was diluted with EtOAc (25 mL),
quenched with water (1 mL) and the aqueous layer extracted with
EtOAc (2 ꢂ 25 mL). The combined organic extracts were washed
with water (2 ꢂ 20 mL) and brine (20 mL), dried (MgSO4) and con-
centrated under reduced pressure to give a crude residue. Flash
chromatography (65% EtOAc/toluene) gave tribenzyl ether 11
(191 mg, 35%), followed by dibenzyl ether 12 (130 mg, 29%), both
as yellow oils. The column was eluted with 40% MeOH/H2O and
the concentrated filtrate purified by flash chromatography (EtOAc
then 4–25% MeOH/EtOAc) to give unreacted diol 10 (96.4 mg).
Data for tribenzyl ether 11: 1H NMR (500 MHz, CDCl3): d 7.37–
7.24 (m, 10H, Ph), 4.97 (d, J = 11.5 Hz, 1H), 4.59 (dd, J = 11.5,
1.5 Hz, 2H), 4.40 (d, J = 11.5 Hz, 1H), 4.22 (t, J = 3.0 Hz, 1H), 4.10
(ddd, J = 9.5, 8.0, 1.5 Hz, 1H), 3.93 (ddd, J = 10.5, 8.5, 4.0 Hz, 1H),
3.77 (dd, J = 11.5, 8.0 Hz, 1H), 3.56 (dd, J = 10.5, 10.0 Hz, 1H), 3.49
(dd, J = 8.5, 3.0 Hz, 1H), 2.51 (ddd, J = 18.0, 6.5, 2.5 Hz, 1H), 2.35
(ddd, J = 18.0, 11.5, 7.0 Hz, 1H), 2.24 (m, 1H), 1.73 (m, 1H). 13C
NMR (125 MHz, CDCl3): d 168.7 (C@O), 138.5, 138.1, 137.7,
128.7, 128.6, 128.1, 128.00, 127.98, 127.81, 127.76, 127.6 (Ph),
78.3, 76.3, 73.7, 72.4, 71.7, 71.1, 64.9, 47.4, 29.6, 26.3. Data for
4.1.1. (3aR,9R,9aR,9bS)-9-Benzyloxy-2,2-dimethylhexahydro-
[1,3]dioxolo[4,5-a]indolizin-6(9bH)-one 9
NaH (60% w/w, 44.7 mg, 1.12 mmol), BnBr (222 lL, 1.86 mmol)
and Bu4NI (27.5 mg, 0.0745 mmol) were added to a solution of
alcohol 827 (169 mg, 0.745 mmol) in DMF (15 mL) at 0 °C and the
resultant mixture was warmed to rt and stirred for 19 h. The reac-
tion mixture was diluted with EtOAc (15 mL), quenched with
water (5 mL) and the aqueous layer extracted with EtOAc
(2 ꢂ 25 mL). The combined organic extracts were washed with
water (3 ꢂ 20 mL) and brine (20 mL), dried (MgSO4) and concen-
trated under reduced pressure to give the crude product. Flash
chromatography (2% MeOH/EtOAc) gave benzyl ether 9 (204 mg,
dibenzyl ether 12:
(500 MHz, CDCl3):
½
a 2D4
ꢃ
¼ ꢀ20:2 (c 0.845, CHCl3). 1H NMR
d
7.40–7.29 (m, 10H, Ph), 4.66 (ABq,
J = 11.5 Hz, 2H), 4.61 (s, 2H), 4.32 (dd, J = 4.0, 3.0 Hz, 1H), 4.11
(td, J = 8.5, 4.0 Hz, 1H), 3.95 (ddd, J = 11.5, 8.5, 4.0 Hz, 1H), 3.72
(dd, J = 10.0, 8.5 Hz, 1H), 3.50 (dd, J = 12.0, 8.5 Hz, 1H), 3.42 (dd,
J = 8.5, 2.5 Hz, 1H), 2.51 (ddd, J = 18.0, 6.5, 2.5 Hz, 1H), 2.33 (ddd,
J = 18.0, 12.0, 6.5 Hz, 1H), 2.19 (m, 1H), 1.76 (m, 1H). 13C NMR
(125 MHz, CDCl3): d 168.8 (C@O), 138.4, 137.0, 128.8, 128.6,
128.5, 128.04, 127.97, 127.9 (Ph), 76.5, 72.5, 71.8, 71.4, 69.1,
86%) as a colourless solid. ½a D23
ꢃ
¼ ꢀ21:2 (c 1.62, CHCl3). 1H NMR
(500 MHz, CDCl3): d 7.37–7.28 (m, 5H, Ph), 4.74 (m, 1H), 4.70 (m,
1H), 4.65 (ABq, J = 12.0 Hz, 2H, CH2Ph), 4.18 (d, J = 14.0 Hz, 1H),
3.97 (ddd, J = 11.0, 7.5, 3.5 Hz, 1H), 3.38 (dd, J = 7.5, 4.0 Hz, 1H),
3.05 (dd, J = 13.5, 5.0 Hz, 1H), 2.48 (dt, J = 18.0, 5.0 Hz, 1H), 2.32
(ddd, J = 17.0, 11.5, 5.5 Hz, 1H), 2.15 (m, 1H), 1.85 (m, 1H), 1.38
(s, 3H, CH3), 1.32 (s, 3H, CH3). 13C NMR (125 MHz, CDCl3): d
168.7 (C@O), 138.3, 128.5, 127.9, 127.8 (Ph), 111.9 (CMe2), 80.0,
77.6, 72.0, 71.7, 65.4, 50.6, 29.4, 26.8, 26.6, 24.9. HR-ESIMS m/z
65.7, 47.7, 29.7, 26.9. FTIR:
m 3349, 2931, 1739, 1620, 1454, 1096,
1068, 737, 698 cmꢀ1. HR-ESIMS m/z 368.18569 [M+H]+, calcd
368.18563 for C22H26NO4.
318.16998 [M+H]+, calcd 318.16998 for C18H24NO4. FTIR:
2913, 2852, 2233, 1599, 1490, 1452, 1442, 1377, 1358, 1323,
1112, 1069, 1025, 911, 829, 754, 690 cmꢀ1
m 2961,
4.1.4. (1S,2R,8R,8aR)-2,8-Bis(benzyloxy)-1-acetoxy-hexahydro-
indolizin-5(1H)-one 13
.
A solution of alcohol 12 (3.2 mg, 8.7
l
mol) in Ac2O (400
lL) and
pyridine (400
l
L) was stirred at rt for 4 h, and then concentrated
under reduced pressure to give crude acetate 13 (3.5 mg). 1H
NMR (500 MHz, CDCl3): d 7.38–7.28 (m, 10H, Ph), 5.70 (dd,
J = 3.5, 2.5 Hz, 1H, H1), 4.55 (ABq, J = 11.5 Hz, 2H, CH2Ph), 4.54
(ABq, J = 11.5 Hz, 2H, CH2Ph), 4.10 (td, J = 9.0, 3.5 Hz, 1H, H2),
4.1.2. (1S,2R,8R,8aS)-8-Benzyloxy-1,2-dihydroxyhexahydroindo-
lizin-5(1H)-one 10
At first, p-TsOH (450 mg, 2.61 mmol) was added to acetal 9
(393 mg, 1.24 mmol) in MeOH (15 mL) and the reaction mixture

996
T. Quach et al. / Tetrahedron: Asymmetry 23 (2012) 992–997
3.72 (dd, J = 12.0, 8.5 Hz, 1H, H3), 3.67–3.61 (m, 2H, H8 and H8a),
3.41 (dd, J = 12.0, 9.5 Hz, 1H, H3), 2.55 (ddd, J = 18.0, 6.5, 2.5 Hz,
1H, H6), 2.36 (ddd, J = 18.0, 11.5, 6.5 Hz, 1H, H6), 2.30 (m, 1H,
H7), 1.98 (s, 3H, OAc), 1.74 (m, 1H, H7). Assignments were con-
firmed by COSY analysis.
170.7, 170.3, 170.1, 170.0, 169.8, 169.6, 168.9 (7C, C@O), 97.5
(C10), 70.8, 70.0, 69.8 (2C), 68.3, 66.7, 64.0, 61.6, 46.3, 28.9, 26.5,
21.1, 20.9, 20.8, 20.76, 20.73, 20.72. FTIR:
1228, 1043 cmꢀ1 HR-ESIMS m/z 624.18978 [M+Na]+, calcd
624.18989 for C26H35NNaO15
m 1747, 1647, 1454,
.
.
4.1.5. (1S,2R,8R,8aR)-2,8-Bis(benzyloxy)-1-(6-O-acetyl-2,3,4-tri-
O-benzyl-a-D-glucopyranosyloxy)-hexahydroindolizin-5(1H)-
one 15
4.1.7. (1S,2R,8R,8aR)-2,8-Bis(hydroxy)-1-(
octahydroindolizine (FleetamineꢁHCl) 7ꢁHCl
4.1.7.1. (1S,2R,8R,8aR)-2,8-Bis(acetoxy)-1-(2,3,4,6-tetra-O-acetyl-
-glucopyranosyloxy)-octahydroindolizine.
BH3ꢁSMe2
(10 M, 12 L, 120 mol) was added to a solution of lactam 16
(14.0 mg, 23.3 mol) in THF (1.5 mL) at 0 °C and the solution was
a-D-glucopyranosyloxy)-
A solution of alcohol 12 (77.3 mg, 0.210 mmol) in dry toluene
(2 mL) was added to a mixture of flame-dried powdered 4 Å molec-
ular sieves (80 mg), Bu4NI (427 mg, 1.16 mmol) and 2,4,6-tri-tert-
butylpyrimidine (144 mg, 0.578 mmol) after which the suspension
was stirred at rt for 10 min. Iodide 14 [prepared by the treatment
of the corresponding anomeric acetate (557 mg, 1.04 mmol) in
a-D
l
l
l
stirred at 0 °C for 30 min and then at rt for 1.5 h. The reaction mix-
ture was then cooled to 0 °C and quenched carefully with EtOH
(0.5 mL). The solvent was removed under reduced pressure and
the residue dissolved in EtOH (1.5 mL) and then heated at reflux
for 2 h. The solution was cooled to rt and concentrated under re-
duced pressure to give a crude residue of the title compound
(14.9 mg), which was used without purification in the next step.
1H NMR (400 MHz, CDCl3): d 5.56 (t, J = 12.0 Hz, 1H), 5.22 (td,
J = 7.2, 2.4 Hz, 1H), 5.10 (dd, J = 10.4, 4.0 Hz, 1H), 5.03 (app t,
J = 10.0 Hz, 1H), 4.82 (d, J = 3.6 Hz, 1H), 4.76 (dd, J = 10.0, 4.4 Hz,
1H), 4.43 (dd, J = 6.8, 4.0 Hz, 1H), 4.27 (ddd, J = 12.8, 4.8, 2.0 Hz,
1H), 4.17 (dd, J = 12.4, 4.8 Hz, 1H), 4.09 (dd, J = 12.4, 2.0 Hz, 1H),
3.12 (dd, J = 11.2, 2.4 Hz, 1H), 3.04–3.02 (m, J = 11.0 Hz, 1H), 2.57
(dd, J = 10.8, 7.6 Hz, 1H), 2.45–2.39 (m, 1H), 2.17 (s, 3H), 2.14 (s,
3H), 2.08 (s, 3H), 2.00 (s, 3H), 1.99 (s, 3H), 1.97 (s, 3H), 1.92 (td,
J = 11.2, 3.6 Hz, 1H), 1.11 (dq, J = 4.8 Hz, 1H). HR-ESIMS m/z
CH2Cl2 (12 mL) with TMSI (658 l
L, 4.63 mmol) at 0 °C]30 in toluene
(3 mL) was added via cannula and the mixture was stirred at 55 °C
for 4 d. The reaction was quenched by the addition of water (1 mL),
stirred at rt for 1 h and filtered through Celite. The aqueous phase
was extracted with EtOAc (5 mL) and the combined organic ex-
tracts washed with satd aq NaHCO3 (5 mL) and brine (5 mL), dried
(MgSO4) and concentrated under reduced pressure to give a crude
residue. Flash chromatography (50–85% EtOAc/petrol) gave glyco-
side 15 (47.0 mg, 27%) as a yellow oil, in addition to unreacted
alcohol 12 (39.7 mg). ½a D22
ꢃ
¼ þ24:2 (c 0.21, CHCl3). 1H NMR
(500 MHz, CDCl3): d 7.36–7.24 (m, 23H, Ph), 7.16–7.14 (m, 2H,
Ph), 5.07 (d, J = 11.0 Hz, 1H), 5.02 (d, J1,2 = 3.5 Hz, 1H, H10), 4.96
(ABq, J = 11.0 Hz, 2H), 4.84 (d, J = 11.0 Hz, 1H), 4.57 (d,
J = 11.5 Hz, 1H), 4.55 (s, 2H), 4.52 (d, J = 10.5 Hz, 1H), 4.48 (dt,
J = 10.5, 2.5 Hz, 1H, H2), 4.35 (dd, J = 3.5, 2.5 Hz, 1H, H1), 4.18
(ddd, J = 11.0, 9.0, 4.0 Hz, 1H), 4.14–4.10 (m, 2H), 4.06 (t,
J = 11.0 Hz, 1H), 4.00 (ddd, J = 10.0, 8.5, 3.5 Hz, 1H), 3.96 (dd,
J = 2.5, 1.0 Hz, 2H), 3.76 (dd, J = 11.5, 8.0 Hz, 1H), 3.58 (dd,
J = 10.0, 4.0 Hz, 1H), 3.53 (dt, J = 10.0, 9.0 Hz, 2H), 3.41 (dd,
J = 8.5, 2.0 Hz, 1H), 2.41 (ddd, J = 18.0, 6.0, 2.0 Hz, 1H, H6), 2.17
(ddd, J = 18.0, 12.0, 6.5 Hz, 1H, H6), 1.99 (s, 3H, Ac), 1.89 (m, 1H,
H7), 1.50 (m, 1H, H7). 13C NMR (100 MHz, CDCl3): d 170.8, 168.9
(2C, C@O), 138.5, 138.2, 138.1, 137.2, 128.7, 128.62, 128.60,
128.56, 128.54, 128.21, 128.18, 127.99, 127.95, 127.89, 127.78,
127.6. 127.1 (Ph), 99.0 (C10), 82.4, 80.6, 77.8, 77.4, 76.4, 75.9,
75.7, 75.0, 72.3, 70.2, 69.1, 68.9, 65.1, 62.8, 47.4, 29.4, 25.4, 21.0.
588.22851 [M+H]+, calcd 588.22868 for C26H38NO14
.
4.1.7.2. (1S,2R,8R,8aR)-2,8-Bis(hydroxy)-1-(
oxy)-octahydroindolizine (FleetamineꢁHCl) 7ꢁHCl.
NaOMe (0.5 M in MeOH, 45.0 L, 22.5 mol) was added to a
solution of the hexaacetate amine (13.0 mg, 22.1 mol) in dry
a-
D-glucopyranosyl-
At first,
l
l
l
MeOH (1.5 mL) and the reaction was stirred at rt for 2 h. The
pH was adjusted to 4 by adding 0.1 M HCl in MeOH (0.5 mL).
The solution was concentrated under reduced pressure to give
a residue, which was purified by flash chromatography (100%
MeOH) to afford FleetamineꢁHCl 7ꢁHCl (5.5 mg, 67%) as a white
solid. ½a 2D2
ꢃ
¼ þ44:5 (c 0.28, MeOH). 1H NMR (400 MHz, CD3OD):
d 5.14 (d, J = 4.0 Hz, 1H), 4.35–4.28 (m, 2H), 3.96 (ddd, J = 10.0,
5.6, 2.4 Hz, 1H), 3.88–3.83 (m, 1H), 3.79 (dd, J = 12.0, 2.4 Hz,
1H), 3.74 (t, J = 7.2 Hz, 1H), 3.66 (dd, J = 12.0, 5.6 Hz, 1H), 3.37
(dd, J = 9.6, 3.6 Hz, 1H), 3.35 (s, 1H), 2.99–2.95 (m, 2H), 2.55
(dd, J = 10.4, 7.6 Hz, 1H), 2.06–2.03 (m, 1H), 1.96 (dd, J = 11.6,
3.2 Hz, 1H), 1.91 (dd, J = 9.2, 3.6 Hz, 1H), 1.74–1.61 (m, 2H),
FTIR:
738, 698 cmꢀ1
864.37182 for C51H55NNaO10
m
3379, 2925, 1739, 1645, 1454, 1364, 1236, 1072, 1028,
.
HR-ESIMS m/z 864.37169 [M+Na]+, calcd
.
4.1.6. (1S,2R,8R,8aR)-2,8-Bis(acetoxy)-1-(2,3,4,6-tetra-O-acetyl-
-glucopyranosyloxy)-hexahydroindolizin-5(1H)-one 16
A mixture of Pd(OH)2 on carbon (10%, 255 mg), pentabenzyl
ether 15 (51.0 mg, 60.6 mol) in MeOH (5 mL), EtOAc (1.5 mL)
a-
D
1.18 (dq, J = 4.8 Hz, 1H). FTIR:
1043 cmꢀ1. HR-ESIMS m/z 336.1659 [M+H]+, calcd 336.1658 for
14H26NO8.
m 1747, 1647, 1454, 1228,
l
C
and AcOH (0.5 mL) in a thick-walled flask was shaken under
70 psi of H2 for 21 h. The flask was evacuated and the contents fil-
tered through Celite. The filtrate was concentrated and the residue
dried twice azeotropically using toluene to give the crude product,
which was dissolved in pyridine (3.5 mL) and Ac2O (3.5 mL) and
heated at 50 °C for 15 h. The reaction mixture was concentrated
to give a crude residue, which was purified by flash chromatogra-
phy (100% EtOAc) to give hexaacetate 16 (28.0 mg, 77%) as a clear
4.2. Binding studies
The ability of Fleetamine to bind bacterial endo-a-man-
nosidases was assessed through isothermal titration calorimetry
(ITC) with both wild-type and an E154A variant of BtGH99, and
through attempted complex formation by soaking and co-crystalli-
zation with wild-type BxGH99. Inhibition studies were performed
using Glc3Man7GlcNAc2 as the substrate with recombinant human
oil. ½a 2D3
ꢃ
¼ þ46:7 (c 0.50, CHCl3). 1H NMR (500 MHz, CDCl3): d 5.36
(dd, J = 10.0, 10.0 Hz, 1H), 5.19–5.12 (m, 2H), 5.11–5.07 (m, 2H),
4.86 (d, J = 3.5 Hz, 1H), 4.47 (dd, J = 3.5, 3.0 Hz, 1H), 4.33 (ddd,
J = 10.0, 3.5, 2.5 Hz, 1H), 4.25 (dd, J = 12.5, 4.0 Hz, 1H), 4.06 (dd,
J = 12.5, 2.0 Hz, 1H), 3.91 (dd, J = 12.0, 8.5 Hz, 1H), 3.71 (dd,
J = 7.0, 2.5 Hz, 1H), 3.54 (dd, J = 12.0, 10.0 Hz, 1H), 2.53 (dt,
J = 17.5, 6.0 Hz, 1H), 2.43 (ddd, J = 17.5, 9.0, 6.0 Hz, 1H), 2.22 (m,
1H), 2.11 (s, 3H), 2.10 (s, 3H), 2.09 (s, 3H), 2.05 (s, 3H), 2.03 (s,
3H), 1.99 (s, 3H), 1.89 (m, 1H). 13C NMR (125 MHz, CDCl3): d
endo-a-mannosidase.
4.2.1. ITC
The ligand affinity of BtGH99 wild-type and BtGH99 E154A for
Fleetamine 7 was studied by ITC using an iTC200 calorimeter (Mic-
roCal). Assays were carried out at 25 °C with Fleetamine (5 mM) ti-
trated into the ITC cell containing 315 lM (wild-type) and 429 lM

T. Quach et al. / Tetrahedron: Asymmetry 23 (2012) 992–997
997
3. Skropeta, D. Bioorg. Med. Chem. 2009, 17, 2645–2653.
4. Winchester, B.; Fleet, G. W. J. Glycobiology 1992, 2, 199–210.
5. Henrissat, B. Biochem. J. 1991, 280, 309–316.
(E154A BtGH99) enzyme. Data were analysed using the Origin 7
software package (MicroCal).
6. Henrissat, B.; Davies, G. Curr. Opin. Struct. Biol. 1997, 7, 637–644.
7. Roth, J.; Ziak, M.; Zuber, C.; Roth, J.; Ziak, M.; Zuber, C. Biochimie 2003, 85, 287–
294.
8. Fujimoto, K.; Kornfeld, R. J. Biol. Chem. 1991, 266, 3571–3578.
9. Rabouille, C.; Spiro, R. G. J. Biol. Chem. 1992, 267, 11573–11578.
10. Volker, C.; De Praeter, C. M.; Hardt, B.; Breuer, W.; Kalz-Fuller, B.; Van Coster, R.
N.; Bause, E. Glycobiology 2002, 12, 473–483.
11. Moore, S. E.; Spiro, R. G. J. Biol. Chem. 1990, 265, 13104–13112.
12. Jespersen, T. M.; Dong, W.; Sierks, M. R.; Skrydstrup, T.; Lundt, I.; Bols, M.
Angew. Chem., Int. Ed. Engl. 1994, 33, 1778–1779.
13. Dong, W.; Jespersen, T.; Bols, M.; Skrydstrup, T.; Sierks, M. R. Biochemistry 1996,
35, 2788–2795.
14. Spohr, U.; Bach, M.; Spiro, R. G. Can. J. Chem. 1993, 71, 1919–1927.
15. Hiraizumi, S.; Spohr, U.; Spiro, R. G. J. Biol. Chem. 1993, 268, 9927–9935.
16. Thompson, A. J.; Williams, R. J.; Hakki, Z.; Alonzi, D. S.; Wennekes, T.; Gloster, T.
M.; Songsrirote, K.; Thomas-Oates, J. E.; Wrodnigg, T. M.; Spreitz, J.; Stutz, A. E.;
Butters, T. D.; Williams, S. J.; Davies, G. J. Proc. Natl. Acad. Sci. U.S.A. 2012, 109,
781–786.
4.2.2. Crystallization and data collection
BxGH99 was crystallized in both native form and as attempted
protein–ligand co-complexes from 0.1 M sodium acetate pH 4.9,
20% w/v PEG 2000 MME (mono-methyl ether), 1.6% poly-c-glu-
tamic acid (low molecular weight). Soaking of native protein crys-
tals was conducted by the addition of approximately 50 mM ligand
over time periods ranging from 30 min to 12 h, prior to cryo-pro-
tection. All crystals were cryo-protected by the stepwise addition
of ethylene glycol to a final concentration of 20% v/v. Diffraction
data were collected at York using a Rigaku RUH3R rotating anode
X-ray generator.
4.2.3. Enzyme inhibition assay
17. Ardron, H.; Butters, T. D.; Platt, F. M.; Wormald, M. R.; Dwek, R. A.; Fleet, G. W.
J.; Jacob, G. S Tetrahedron: Asymmetry 1993, 4, 2011–2024.
18. Davis, B. G Tetrahedron: Asymmetry 2009, 20, 652–671.
19. Fleet, G. W. J.; Gough, M. J.; Smith, P. W. Tetrahedron 1984, 25, 1853–1856.
20. Suits, M. D.; Zhu, Y.; Taylor, E. J.; Walton, J.; Zechel, D. L.; Gilbert, H. J.; Davies,
G. J. PLoS One 2010, 5, e9006.
21. van den Elsen, J. M. H.; Kuntz, D. A.; Rose, D. R. EMBO J. 2001, 20, 3008–3017.
22. Zhu, Y.; Suits, M. D. L.; Thompson, A. J.; Chavan, S.; Dinev, Z.; Dumon, C.; Smith,
N.; Moremen, K.; Xiang, Y.; Siriwardena, A.; Williams, S. J.; Gilbert, H. J.; Davies,
G. J. Nat. Chem. Biol. 2010, 6, 125–132.
Glucosylated substrate (Glc3Man7GlcNAc2) was obtained from
tissue-cultured CHO cells treated with N-butyl-deoxynojirimycin
(NB-DNJ) as described previously.32 An aliquot was labelled with
anthranilic acid and analysed by normal phase (NP)-HPLC. The
amount of fluorescence under the curve was determined and con-
verted into molar amounts using an experimentally-derived con-
version factor. Oligosaccharide substrate, 5 pmole, was dried and
resuspended in 10
lL (0.01 lg protein) of human recombinant
23. Fleet, G. W. J. personal communication.
24. Nemr, A. E. Tetrahedron 2000, 56, 8579–8629.
endo-a
-mannosidase enzyme,38 in 0.1 M sodium/MES buffer, pH
6.5. Following incubation at 37 °C for 45 min in the presence or ab-
sence of an inhibitor, the reaction was stopped with glacial AcOH
(4 lL), diluted with water and an equal volume of acetonitrile
added. An aliquot was analysed by NP-HPLC to separate the reac-
tion products and the amount of hydrolysis quantified using peak
area analysis.
25. Pyne, S. G. Curr. Org. Synth. 2005, 2, 39–57.
26. Pearson, W. H.; Hembre, E. J. J. Org. Chem. 1996, 61, 7217–7221.
27. Pearson, W. H.; Ren, Y.; Powers, J. D. Heterocycles 2002, 58, 421–430.
28. Lemieux, R. U.; Hendriks, K. B.; Stick, R. V.; James, K. J. Am. Chem. Soc. 1975, 97,
4056–4062.
29. Meloncelli, P. J.; Martin, A. D.; Lowary, T. L. Carbohydr. Res. 2009, 344, 1110–
1122.
30. Thiem, J.; Meyer, B. Chem. Ber./Recl. 1980, 113, 3075–3085.
31. Alonzi, D. S.; Neville, D. C.; Lachmann, R. H.; Dwek, R. A.; Butters, T. D. Biochem.
J. 2008, 409, 571–580.
32. Petrescu, A. J.; Butters, T. D.; Reinkensmeier, G.; Petrescu, S.; Platt, F. M.; Dwek,
R. A.; Wormald, M. R. EMBO J. 1997, 16, 4302–4310.
Acknowledgements
We gratefully acknowledge support from the Australian Re-
search Council and the School of Chemistry, University of Mel-
bourne. A.J.T. is supported by a Biotechnology and Biological
Sciences Research Council (BBSRC) studentship and work on the
endo-a-mannosidase in York supported by BBSRC grant BB/
G016127/1.
33. Emsley, P.; Cowtan, K. Acta Crystallogr. 2004, D60, 2126–2132.
34. Potterton, L.; McNicholas, S.; Krissinel, E.; Gruber, J.; Cowtan, K.; Emsley, P.;
Murshudov, G. N.; Cohen, S.; Perrakis, A.; Noble, M. Acta Crystallogr. 2004, D60,
2288–2294.
35. Potterton, L.; McNicholas, S.; Krissinel, E.; Gruber, J.; Cowtan, K.; Emsley, P.;
Murshudov, G. N.; Cohen, S.; Perrakis, A.; Noble, M. Acta Crystallogr. 2004, D60,
2288–2294.
36. Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.; Timmers, F. J.
Organometallics 1996, 15, 1518–1520.
References
37. Still, W. C.; Kahn, M.; Mitra, A. M. J. Org. Chem. 1978, 43, 2923–2925.
38. Kukushkin, N. V.; Alonzi, D. S.; Dwek, R. A.; Butters, T. D. Biochem. J. 2011, 438,
133–142.
1. Kornfield, R.; Kornfield, S. Annu. Rev. Biochem. 1985, 54, 631–664.
2. Helenius, A.; Aebi, M. Annu. Rev. Biochem. 2004, 73, 1019–1049.