Synthesis, biological evaluation and docking studies of casuarine analogues: Effects of structural modifications at ring B on inhibitory activity towards glucoamylase

Article abstract of DOI:10.1002/ejoc.201000632

We report the total synthesis of a series of pyrrolizidine analogues of casuarine (1) and their 6-O-α-glucoside derivatives. The synthetic strategy is based on a totally regio- and stereoselective 1,3-dipolar cycloaddition of suitably substitutedalkenes a

Full text of DOI:10.1002/ejoc.201000632

FULL PAPER
DOI: 10.1002/ejoc.201000632
Synthesis, Biological Evaluation and Docking Studies of Casuarine Analogues:
Effects of Structural Modifications at Ring B on Inhibitory Activity Towards
Glucoamylase
Claudia Bonaccini,*[a] Matteo Chioccioli,[a] Camilla Parmeggiani,[b] Francesca Cardona,*[b]
Daniele Lo Re,[b] Gianluca Soldaini,[b] Pierre Vogel,[c] Claudia Bello,[c] Andrea Goti,[b] and
Paola Gratteri[a]
Keywords: Azasugars / Enzymes / Inhibitors / Molecular modeling / Biological activity
We report the total synthesis of a series of pyrrolizidine ana- lecular modeling study focused on glucoamylase (GA) in
logues of casuarine (1) and their 6-O-α-glucoside derivatives.
The synthetic strategy is based on a totally regio- and stereo-
selective 1,3-dipolar cycloaddition of suitably substituted
alkenes and a carbohydrate-based nitrone. We also report
the evaluation of the biological activity of casuarine and its
derivatives towards a wide range of glycosidases and a mo-
which the binding modes of the newly synthesized com-
pounds within the enzyme cavity are investigated. The re-
sults highlight the prominent structural features of casuarine
and its derivatives that make them selective glucoamylase
inhibitors.
Introduction
tive inhibitors of these classes of enzymes are potential anti-
diabetes, anti-viral and anti-cancer agents.[1,2] Recently,
interesting immunosuppressive activities have been discov-
ered for this class of compounds.[4] In the past 40 years,
more than 100 polyhydroxylated alkaloids have been iso-
lated from plants and microorganisms[5] with structures that
include polyhydroxylated piperidines, pyrrolidines, indolizi-
dines, pyrrolizidines and nortropanes. For instance, the pi-
peridine alkaloid 1-deoxynojirimycin (DNJ, Scheme 1), pre-
pared first by Paulsen et al. in 1967[6a] and then isolated
from a species of Moris (Moraceae),[6b] was found to
strongly inhibit α-glucosidases.[5a] N-Alkylated derivatives
of DNJ have found applications as anti-diabetic drugs (i.e.,
Miglitol, Glyset) or anti-HIV agents (Glycovir, SC
49483).[2a] The indolizidine alkaloid (+)-lentiginosine
(Scheme 1) was isolated in 1990 from the leaves of Astraga-
Iminosugars are very attractive carbohydrate mimics in
which the endocyclic oxygen atom is replaced by the more
basic, trivalent nitrogen atom.[1] In their protonated form,
iminosugars resemble the transition state or intermediate
generated during the hydrolysis reaction catalysed by glyco-
sidases, key hydrolytic enzymes involved in many physiolog-
ical functions. Since the discovery of the inhibitory proper-
ties of iminoalditols towards glycosidases, they have re-
ceived increasing attention as diagnostic compounds as well
as tools for the investigation of the structures, functions and
catalytic mechanisms of carbohydrate-processing en-
zymes.[2,3] Furthermore, given the important role of glyco-
sidases and glycosyltransferases in controlling the structures
and functions of carbohydrates at the cell surface, competi-
[a] Laboratory of Molecular Modeling Cheminformatics & QSAR, lus lentiginosus and was found to inhibit amyloglucosid-
ases.[7] Its non-natural enantiomer (–)-lentiginosine was re-
Department of Pharmaceutical Sciences, Laboratory of Design,
Synthesis and Study of Biologically Active Heterocycles (Het-
cently discovered to possess proapoptotic activity towards
tumoral cells.[8] Castanospermine (Scheme 1), isolated in
1981 from the seeds, leaves and barks of Castanospermum
australe and in 1988 from the seeds, leaves and barks of
Alexa sp.,[5a] and its ester and salt derivatives are able to
inhibit tumour growth and metastasis.[9]
Casuarine (1, Scheme 2) and its 6-O-α-glucoside, casuar-
ine-6-O-α-glucopyranoside (2, Scheme 2), have been iso-
lated from the bark of Casuarina equisetifolia L. (Casua-
rinaceae) and from the leaves of Eugenia jambolana Lam.
(Myrtaceae).[10] We recently reported that casuarine (1) is
able to inhibit a human maltase-glucoamylase (MGAM,
eroBioLab), University of Florence,
Via Ugo Schiff 6, 50019 Sesto Fiorentino, Firenze, Italy
Fax: +39-055-4573780
E-mail: claudia.bonaccini@unifi.it
paola.gratteri@unifi.it
[b] Department of Chemistry “Ugo Schiff”, Laboratory of Design,
Synthesis and Study of Biologically Active Heterocycles (Het-
eroBioLab), University of Florence,
Via della Lastruccia 3-13, 50019 Sesto Fiorentino, Firenze, Italy
Fax: +39-055-4573531
E-mail: francesca.cardona@unifi.it
[c] Laboratory of Glycochemistry and Asymmetric Synthesis
(LGSA), Swiss Institute of Technology (EPFL),
Batochime, 1015 Lausanne, Switzerland
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201000632.
View this journal online at
wileyonlinelibrary.com
5574
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5574–5585

Casuarine Analogues
Scheme 2. Casuarine and its 6-O-α--glucoside.
Our total synthesis of casuarine (1) and its 6-O-α-gluco-
side (2) took advantage of a complete stereoselective
nitrone cycloaddition strategy with Tamao–Fleming oxi-
dation and selective α-glucosylation as the key steps.[11]
Glucoamylase (1,4-α--glucan glucohydrolase, GA; EC
3.2.1.3; glycoside hydrolase family GH15, www.cazy.org) is
an exo-hydrolase that catalyses the removal of glucose units
from the non-reducing end of starch and related oligosac-
Scheme 1. Glycosidase inhibitors.
EC 3.2.1.20) more strongly than the pseudo-tetrasaccharide charides. The hydrolytic reaction, which preferentially oc-
acarbose (Scheme 1) currently on the market as an anti- curs at α-1,4 linkages, proceeds with inversion of configura-
diabetic drug (Glucobay, Precose) and thus has promise for tion at the anomeric carbon atom. Glucoamylases are also
the development of novel anti-diabetic drugs.[11]
able to hydrolyse α-1,6 linkages, but the specific activity is
Table 1. Structures and inhibition activities (IC50) of compounds 1–11 towards glucoamylase from aspergillus niger.
[a] For compounds with an inhibition percentage less than 90% at 1 m concentration, the IC50 values were not determined, and the
percentage inhibition is reported.
Eur. J. Org. Chem. 2010, 5574–5585
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5575

C. Bonaccini, F. Cardona et al.
FULL PAPER
only 0.2% with respect to α-1,4 hydrolysis.[12] The interest
The first key step in the synthesis of 7-deoxycasuarine
in glucoamylase is related to its use in the industrial pro- (4), its lactam derivative 6 and hyacinthacine A2 (11) is the
duction of bioethanol, glucose and fructose syrups.[13] 1,3-dipolar cycloaddition of nitrone 12 to dimethylacrylam-
Furthermore, given the presence of these enzymes in a wide ide, which, after N–O bond cleavage of the isoxazolidine
variety of organisms and their quite simple obtainment in 14a with Zn in acetic acid gave lactam 16 in 68% yield over
a pure form, GA has been extensively studied as a model two steps (Scheme 4). Compound 16 is the key intermediate
for other members of the glycosyl hydrolase family.[14]
for the total synthesis of all three target molecules. Hydro-
In this paper we report the total synthesis of the pyrroliz- genation in EtOH catalysed by Pd/C afforded lactam 6 in
idine analogues of casuarine 4, 6, 7 and 11 (see Table 1), all 88% yield. Reduction of the C=O bond with LiAlH4 in
bearing the same stereochemical pattern at the more substi- THF at reflux gave compound 17 in 75% yield, which, after
tuted five-membered ring A as that of casuarine, and their catalytic hydrogenation in EtOH, afforded 7-deoxycasuar-
6-O-α-glucoside derivatives 3, 5 and 8–10. We also report ine (4) in 88% yield (Scheme 4). Deoxygenation at C6 was
the evaluation of the inhibitory activities of compounds 1– achieved through the mesylation of 16 followed by re-
11 towards a wide range of commercially available glycosid- duction with LiAlH4 in THF at reflux. This gave 18 in 80%
ases and a molecular modelling study on glucoamylase yield over two steps. Finally, catalytic hydrogenation gave
(GA) from Aspergillus awamori as a result of the selective hyacinthacine A2 (11) in 72% yield (Scheme 4).
inhibitory activity towards GA shown by the tested com-
pounds.
Results and Discussion
Synthesis
The general strategy followed for the synthesis of the pyr-
rolizidine alkaloids is outlined in Scheme 3. We took advan-
tage of a stereocontrolled cyclic nitrone cycloaddition strat-
egy[15] employing polyfunctionalized nitrone 12 and suitable
dipolarophiles 13, which afforded regio- and stereoselec-
tively isoxazolidines 14. These were then converted into pyr-
rolizidinone derivatives 15 by reductive ring-opening/cycli-
zation. Intermediates 15 bear a free hydroxy group at C6 of
Scheme 4. Syntheses of 7-deoxycasuarine (4), its lactam derivative
the pyrrolizidine ring, which allows selective glucosylation
6 and hyacinthacine A2 (11): Reagents and conditions: (a) dimeth-
at this position. A good choice of dipolarophile is crucial
for the success of the strategy. For instance, for the synthesis
of casuarine (1) a good regioselectivity of the cycloaddition
was assured by using dipolarophile 13 with Y = SiMe2Ph
and X = OEt.[11]
ylacrylamide, CH2Cl2, room temp., 3 d, 85%; (b) Zn, AcOH/H2O,
50 °C, 4 h, 80%; (c) H2, Pd/C, EtOH, 3 d, 88%; (d) LiAlH4, THF,
reflux, 3 h, 75%; (e) H2, Pd/C, HCl, EtOH, 3 d, 88%; (f) MsCl,
NEt3, CH2Cl2, room temp., 2 h, 100%; (g) LiAlH4, THF, reflux,
1.5 h, 80%; (h) H2, Pd/C, HCl, MeOH, 3 d, 72%.
The lactam intermediate 16 was also employed in the
synthesis of glucoside 5; the initial selective α-glucosylation
provided compound 19.[19] Reduction of its amide moiety
with LiAlH4 followed by catalytic hydrogenolysis gave 5 in
45% yield over two steps (Scheme 5).
Scheme 3. General procedure for the synthesis of the pyrrolizidine
alkaloids.
Scheme 5. Synthesis of 7-deoxycasuarine glucosyl derivative 5: Rea-
gents and conditions: (a) 2,3,4,6-tetra-O-benzylglucopyranosyl tri-
chloroacetimidate, TMSOTf, Et2O, 1 h, 88%; (b) LiAlH4, THF,
room temp., 1 h, 58%; (c) H2, Pd/C, HCl, MeOH, room temp.,
18 h, 77%.
Nitrone 12 was conveniently prepared on a multigram
scale by starting from commercially available tribenzyl -
arabinose.[16] It has the absolute configuration of the ste-
reogenic centres at C1, C2 and C3 required for casuarine
and its analogues such as non-natural 7-deoxycasuarine
For the synthesis of the non-natural 7-homocasuarine (7)
(4),[16a,17] its lactam derivative 6 and hyacinthacine A2 and of its glucosyl derivatives 8–10, dimethyl maleate was
(11).[16a,18]
chosen as the dipolarophile. Treatment of the isoxazolidine
5576
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5574–5585

Casuarine Analogues
14b with Zn in acetic acid at 50 °C for 3 h gave lactam 20
We recently presented the total synthesis of casuarine (1)
in 70% yield over two steps (Scheme 6). Reduction of both and its 6-O-α-glucoside 2,[11,21] as well as its epimer at C6,
the ester and lactam moieties with excess LiAlH4 in THF namely uniflorine A,[22] obtained from a derivative of lac-
at reflux provided diol 21 quantitatively. Finally, catalytic tam 15 (Y = SiMe2Ph) by inversion of the configuration at
hydrogenolysis with Pd/C as catalyst in the presence of HCl C6. We have now also synthesized the lactam derivative 3
gave non-natural 7-homocasuarine (7) in 89% yield in 77% yield by hydrogenolysis of the fully protected lactam
(Scheme 6).
24 (Scheme 8).[11]
Scheme 8. Synthesis of casuarine 6-O-α-glucoside derivative 3. Rea-
gents and conditions: (a) H2, Pd/C, MeOH/EtOAc, room temp.,
24 h, 77%.
The lactams 10, 9 and 3 were synthesized to investigate
the importance of the basic nitrogen atom in glycosidase
inhibition. With all these compounds in hand, we investi-
gated their inhibitory activity towards a wide range of com-
mercially available glycosidases.
Scheme 6. Synthesis of 7-homocasuarine (7). Reagents and condi-
tions: (a) dimethyl maleate, CH2Cl2, room temp., 4 d, 78%; (b) Zn,
AcOH/H2O, 50 °C, 3 h, 90%; (c) LiAlH4, THF, reflux, 2 h, 100%;
(d) H2, Pd/C, HCl, EtOH, room temp., 4 d, 89%.
Lactam 20 was also selectively α-glucosylated to give α-
glucoside 22 in 75% yield.[19] Intermediate 22 could be ma-
nipulated in different ways allowing us to obtain the three
glucosyl derivatives 8–10. Catalytic hydrogenolysis of 22 af-
forded compound 10 (96%), which retained both the ester
and lactam moieties. Treatment of 22 with excess LiBH4
and BH3·THF[20] led to the complete reduction of both the
ester and lactam moieties, and subsequent catalytic hydro-
genolysis afforded the target glucosylated 7-homocasuarine
8 in 77% yield over two steps. Selective reduction of the
ester moiety was achieved by treatment of 22 with LiBH4
in THF at room temperature for 18 h. This afforded 23 in
62% yield. Finally, catalytic hydrogenolysis of 23 provided
glucoside 9 in 72% yield (Scheme 7).
Glycosidase Inhibitory Activities
Compounds 1–11 were assayed with respect to a panel
of 13 commercially available glycosidases (Table 2): α--
fucosidase (EC 3.2.1.51) from bovine kidney, α-galactosid-
ase (EC 3.2.1.22) from coffee beans, β-galactosidase (EC
3.2.1.23) from Escherichia coli and Aspergillus orizae, α-glu-
cosidase (EC 3.2.1.20) from yeast and rice, amyloglucosid-
ase (EC 3.2.1.3) from Aspergillus niger, β-glucosidase (EC
3.2.1.21) from almonds, α-mannosidase (EC 3.2.1.24) from
jack beans, β-mannosidase (EC 3.2.1.25) from snails, β-xy-
losidase (EC 3.2.1.37) from Aspergillus niger, β-N-acetylglu-
cosaminidase (EC 3.2.1.30) from jack beans and bovine
kidney with appropriate p-nitrophenyl glycoside sub-
strates.[23] The errors in the measurements were estimated
to be around 20% (statistical study carried out with model
compounds) and in the concentrations to be around 10–
15% (errors in sample weight). Casuarine (1) was found to
be a potent and competitive inhibitor of amyloglucosidase
from Aspergillus niger (IC50
= 1.9Ϯ0.4 µ, Ki =
2.0Ϯ0.4 µ; ref.[24] IC50 = 0.7 µ). It also inhibits α-gluco-
sidase from yeast and rice (91 and 94% at 1 m, respec-
tively) and, to a lesser extent, β-glucosidase from almonds,
α-mannosidase from jack beans, β-xylosidase from Asper-
gillus niger and β-N-acetylglucosaminidase from jack beans
(46, 21, 24 and 16% at 1 m, respectively). Glucoside 2 was
slightly less potent with the IC50 and Ki values in the same
order of magnitude (IC50 = 4.4Ϯ0.9 µ, Ki = 3.9Ϯ0.8 µ,
mixed-type inhibition; ref.[24] IC50 = 1.1 µ), but more se-
lective than casuarine: indeed at 1 m concentration it gave
only 20% inhibition towards α-glucosidase from yeast and
did not inhibit at all α-glucosidase from rice. Moreover, it
showed only 19% inhibition towards α--fucosidase from
bovine kidney (1 m). Glucosides 5 and 8 and 7-homocasu-
Scheme 7. Synthesis of 7-homocasuarine glucosyl derivatives 8–10.
Reagents and conditions: (a) H2, Pd/C, MeOH/EtOAc, room
temp., 4 d, 96%; (b) LiBH4, BH3·THF, THF, room temp., 11 d,
98%; (c) H2, Pd/C, HCl, MeOH, room temp., 12 h, 79%; (d) LiBH4,
THF, room temp., 18 h, 62%; (e) H2, Pd/C, MeOH, 24 h, 72%.
Eur. J. Org. Chem. 2010, 5574–5585
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5577

C. Bonaccini, F. Cardona et al.
FULL PAPER
arine (7) were good and selective competitive inhibitors of tivity towards amyloglucosidase from Aspergillus niger.
amyloglucosidase from Aspergillus niger (IC50 = 7.7Ϯ1.5, Hyacinthacine A2 (6,7-dideoxycasuarine, 11) was also a
8.1Ϯ1.6 and 24Ϯ4.7 µ and Ki = 7.4Ϯ1.2, 11Ϯ2.1 and very strong inhibitor of this enzyme (IC50
=
23Ϯ4.6 µ, respectively). They inhibited α-glucosidase 1.9 µϮ0.4 µ, Ki = 2.6 µϮ0.5 µ, non-competitive in-
from yeast weakly (29, 45 and 55% at 1 m, respectively) hibition; ref.[25] IC50 = 8.6 µ). Among the remaining com-
and did not inhibit α-glucosidase from rice. 7-Deoxycasuar- pounds, 3 was not active towards any of the glycosidases
ine (4) was a very potent and selective competitive inhibitor assayed, 9 and 10 showed weak and very selective inhibitory
of amyloglucosidase from Aspergillus niger (IC50
=
activity (25 and 76% inhibition at 1 m, respectively)
4.5Ϯ0.9 µ, Ki = 3.5Ϯ0.7 µ). Among the other glycosid- towards amyloglucosidase from Aspergillus niger. Gluco-
ases assayed, only α-glucosidase from rice was also in- sides 5 and 8 together with the parent compound 2 have
hibited weakly (36% at 1 m). Lactam 6 exhibited weak also been found to be potent inhibitors of bacterial and
(IC50 = 0.21Ϯ0.04 m) but very selective inhibitory ac- insect trehalases.[19]
Table 2. Inhibitory activity of compounds 1–11 towards commercially available glycosidases.[a]
Enzyme (pH)
1
2
3
4
5
6
α--Fucosidase
n.i.
19
n.i.
n.i.
n.t.
n.i.
Bovine kidney (6)
α-Galactosidase
n.i.
n.i.
n.i.
n.i.
n.t.
n.i.
Coffee beans (6)
β-Galactosidase Escherichia coli (7)
Aspergillus orizae (4)
α-Glucosidase
n.i.
n.i.
91
n.i.
n.i.
20
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.t.
n.t.
29
n.i.
n.i.
n.i.
Yeast (7)
Rice (4)
94
98
n.i.
97
n.i.
n.i.
36
100
n.i.
97
n.i.
91
Amyloglucosidase Aspergillus niger (5)
IC50 = 1.9 µ
Ki = 2.0 µ
46
IC50 = 4.4 
Ki = 3.9 µ
n.i.
IC50 = 4.5 µ IC50 = 7.7 µ IC50 = 0.21 m
Ki = 3.5 µ
Ki = 7.4 µ
β-Glucosidase
Almonds (5)
α-Mannosidase
Jack beans (5)
β-Mannosidase
Snails (4)
n.i.
n.t.
n.t.
n.t.
n.i.
n.i.
n.t.
42
n.i.
n.i.
n.i.
44
21
n.i.
24
n.i.
n.t.
n.t.
n.t.
n.i.
n.i.
n.i.
n.i.
n.t.
n.t.
n.t.
n.t.
β-Xylosidase
Aspergillus niger (5)
β-N-Acetylglucosaminidase
Jack beans (5)
16
Bovine kidney (4)
n.i.
7
n.t.
8
n.i.
9
n.t.
10
n.t.
11
20
Enzyme (pH)
α--Fucosidase
n.i.
n.t.
n.i.
n.i.
n.t.
Bovine kidney (6)
α-Galactosidase
n.i.
n.t.
n.i.
n.i.
n.i.
Coffee beans (6)
β-Galactosidase Escherichia coli (7)
Aspergillus orizae (4)
α-Glucosidase
n.i.
n.i.
45
n.t.
n.t.
55
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
72
Yeast (7)
Rice (4)
n.i.
99
n.i.
92
n.i.
25
n.i.
76
n.i.
97
Amyloglucosidase Aspergillus niger (5)
IC50 = 8.1 µ IC50 = 24 µ
IC50 = 1.9 µ
Ki = 2.6 µ
n.i.
Ki = 11 µ
Ki = 23 µ
β-Glucosidase
Almonds (5)
α-Mannosidase
Jack beans (5)
β-Mannosidase
Snails (4)
n.t.
n.t.
n.i.
n.i.
n.t.
n.t.
n.t.
n.t.
n.i.
n.i.
n.i.
n.i.
n.i.
n.i.
n.t.
n.t.
n.t.
n.t.
n.t.
n.t.
n.t.
n.t.
n.t.
n.t.
n.i.
n.i.
n.i.
n.i.
n.i.
β-Xylosidase
Aspergillus niger (5)
β-N-Acetylglucosaminidase
Jack beans (5)
Bovine kidney (4)
[a] Percentage inhibition at a concentration of 1 m. n.i. = no inhibition, n.t. = test not performed.
5578
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5574–5585

Casuarine Analogues
As 2 is a glucoside we verified whether the amyloglucosi-
After completion of docking calculations, ring A of casu-
dase was able to hydrolyse its glucosidic bond or not under arine (1) was found to be deeply located within the –1 site
our test conditions. Indeed, we had to consider the possibil- where it is involved in an optimal hydrogen-bonding net-
ity that the inhibitory activities observed for 2 were due to work involving C8–OH and C2–OH of the ligand and
the casuarine liberated by the hydrolysis reaction catalysed active-site residues R54 and D55 together with the nucleo-
by amyloglucosidase. Thus, we performed a series of mass philic water Wat501. In addition, C1–OH is positioned at a
spectral analyses using HR-ESI-TOF-MS (high-resolution hydrogen-bond distance from both the carbonyl oxygen
ESI mass spectrometry, positive ionization mode). As atom of residue L177 and Wat501 (Figure 1). All other pyr-
shown in Figure S48 (see the Supporting Information) the rolizidine molecules were oriented in a similar way and con-
peak assigned to the glucoside (A), MW = 367.35, was de- served these interactions, which have also been observed in
tected in solution when the measurement was performed the crystallographic complexes with both DNJ and acar-
immediately after the addition of the enzyme (C) as well as bose.[26a,26b]
after 20 min of incubation at room temperature (D) and
after 20 min of incubation at 37 °C (E). We verified that the
mixture of enzyme and buffer did not give similar signals
(B). The spectra of the different assays performed with the
buffered solution (phosphate) of the inhibitor in the pres-
ence of the enzyme neither showed peaks corresponding to
the aglycon (MW = 205.21) nor to glucose (180.16) alone
(see Figure S48 of the Supporting Information). This indi-
cates that glucoside 2 is not hydrolysed significantly by the
enzyme (amyloglucosidase from Aspergillus niger) under the
conditions of our test.
Computational Studies
The data reported in Table 2 show that the casuarine de-
rivatives presented here, with the exception of glycosylated
lactams 3, 9 and 10, all inhibit more than 90% of Aspergil-
lus niger amyloglucosidase activity with IC50 values ranging
from 1.9 µ for pyrrolizidines 1 and 11 to 24 µ for gluco-
side 8, with lactam 6 showing the weakest activity
(0.21 m). Furthermore, the active compounds showed a
competitive inhibition profile, thus indicating a similar
Figure 1. Docked orientation of molecule 1 within the glucoamyl-
binding mode within the enzyme. Nevertheless, no signifi-
ase active site. Hydrogen-bonds are depicted as magenta dashed
cant differences were found in the inhibition activity of glu-
cosides and their parent compounds, which clearly indicates
a lack of correlation between inhibition and the ability to
occupy a second subsite.
lines.
As a consequence of this orientation, the protonated ni-
trogen atom of pyrrolizidine derivatives is not involved in
In the past years, several crystallographic structures of the strong hydrogen-bond interaction with E179 observed
the proteolytic fragment of glucoamylase G2 from Aspergil- for the nitrogen atom in the acarbose complex,[26b] but is
lus awamori (95% sequence identity with the A. niger pro- oriented towards the region in which the nucleophile
tein) bound to different inhibitors have been reported,[26] Wat501 lies, analogous to what was observed for the imi-
which has made it possible to investigate the nature of the nosugar-type inhibitor DNJ.[26a] We could hypothesize that
interaction between glucoamylase and its ligands in detail. the high affinity of acarbose (Ki = 10–12 )[26b] for glu-
The enzyme active site is characterized by an excess of coamylase is due, at least in part, to the presence of the
negative charge, which has been principally ascribed to resi- charged hydrogen-bond interaction with E179 given that it
dues D55, E179 and E400 from the –1 subsite. Further- has been suggested that charged hydrogen bonds can be re-
more, E179 and E400 have been shown to be the putative sponsible for a change in the binding constant by a factor
catalytic acid and base, respectively, and the hydrolysis reac- of 1000.[27] Indeed, a maltoside hetero-analogue carrying a
tion was hypothesized to proceed through the formation of nitrogen atom at the interglycosidic linkage, which enables
a glucopyranosyl cation intermediate after nucleophilic at- it to establish a charged hydrogen bond with E179, showed
tack by a water molecule. With this information in mind a 1000-fold stronger competitive inhibition than the ana-
and with the aim to interpret the biological data in a struc- logue in which the interglycosidic atom is sulfur.[28] In con-
tural way, we decided to study the docking on the glu- trast, a comparison of the crystal complexes of GA with
coamylase structure to investigate the possible binding acarbose and its weaker -gluco-dihydro derivative (Ki =
mode of the casuarine derivatives presented here.
10–8 ) shows that the two molecules bind in a very similar
Eur. J. Org. Chem. 2010, 5574–5585
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5579

C. Bonaccini, F. Cardona et al.
FULL PAPER
way (including the charged hydrogen bond with E179), and ring A of acarbose. In all the selected poses showing this
the 6 kcal/mol difference in binding energies could be orientation the pyrrolizidine nucleus orients the N–H+
largely attributed to unfavourable steric interactions be- towards the aromatic ring of Y311 in the +1 site with a
tween the hydrogen atoms at C7A of -gluco-dihydroacar- geometry compatible with an NH···π interaction. Further-
bose and the catalytic water,[26b] thus highlighting a com- more, C8–OH is able to donate one hydrogen bond to E180,
parable effect of electrostatic and hydrophobic interactions thus contributing to the stabilization of this binding mode.
in guiding interactions with GA. Molecules 1, 4, 6 and 7 Anyway, by comparison of the GLU-IN orientation of 2
are able to bind E179 through a neutral hydrogen bond in- with the docked orientation of the hydrolysable maltose,
volving C6–OH, whereas the 6-deoxy derivative 11 clearly which perfectly overlaps the glucose moiety of 2 with its
lacks this interaction. Anyway, apart from C6–OH when non-reducing end, it is not clear how the casuarine gluco-
present, pyrrolizidine ligands are oriented towards the nega- side could resist hydrolysis, as we observed, because the nu-
tively charged E179, a very hydrophobic portion of the mo- cleophile Wat501 is perfectly oriented towards the anomeric
lecule, which could contribute negatively to the binding.
carbon atom of the ligand. In contrast, the CAS-IN binding
With respect to the crystallographic orientation of acar- mode of glucosides explains the resistance to hydrolysis that
bose, the ring B atoms of the pyrrolizidines extend towards we observed for molecule 2 because the glucose moiety is
the +1 site, with C6 and C7 of 1 almost perfectly overlapped located in the +1 site, far from the nucleophile Wat501.
with C5B and N4B of the acarbose ring B, respectively. C7– Moreover, given the longer C2–C6 distance of pyrrolizidine
OH of 1 is thus in close proximity to both R305 and the (4.9 Å) with respect to the C1–C4 distance of glucose
carbonyl oxygen atom of W178, but it does not present the (2.9 Å), the CAS-IN orientation allows ring B of pyrrolizid-
correct geometry for hydrogen-bond formation, in contrast ine to extend to the +1 site such that the glucose moiety is
to what has been observed for C3B–OH of acarbose. The projected towards the third GA subsite (see Figure S49 of
elimination of the C7 substituent (molecule 4) led to a slight the Supporting Information). Interestingly, after docking,
decrease in the IC50 value as well as its elongation (molecule the lactam derivatives of glucosides 3, 9 and 10, which did
7), although this latter modification allowed the molecule not show inhibitory activity, all adopted the same GLU-IN
to hydrogen-bond to both W178 and R305. Molecule 6, orientation. The interactions of glucose at the –1 site are
which shows a very weak inhibition, is the only non-glycos- the same as those observed for active glucosides, and at the
ylated lactam of this series, and after docking it is oriented level of the +1 site the results of docking converged to a
like molecule 4. Anyway, given the limited conformational unique solution, which differ from the active glucosides for
flexibility imposed on molecule 6 by the presence of the the aglycon conformation clearly influenced by the pres-
lactam structure, the conformation of ring B is clearly influ- ence/absence of the lactam structure. Furthermore, the ab-
enced. Indeed, molecule 6 is also able to hydrogen-bond sence of charge on the nitrogen atom prevents lactam deriv-
E179 through C6–OH, but the absence of the positive atives in the GLU-IN orientation from reinforcing the inter-
charge on the nitrogen atom eliminates the possibility of action at the +1 site through NH···π interactions. These ob-
molecule 6 compensating the excess negative charge present servations on lactam molecules highlight some interesting
in the glucoamylase active site, which has been hypothe- features of the GA interaction: the fact that the CAS-IN
sized as one of the mechanisms involved in complex stabili- orientation was never found for lactam-glucoside, together
sation.[26a,26b] Finally, the unfavourable effects due to the with the observations we made on the structural features of
positioning of a hydrophobic portion of the ligand close to the interaction with molecule 6, clearly indicate a poor
a charged amino acid (E179) is even more pronounced here binding of the lactam moiety to the –1 site due to both the
given the higher hydrophobicity of 6 relative to 4. All these absence of positive charge on the ligand and the steric ef-
considerations are in agreement with the very low IC50 fects associated with the lactam structure. Given also that
value found for 6.
at the +1 site the lack of charge on the ligand could be
As far as the glucoside derivatives are concerned, the re- detrimental to the binding affinity and considering that the
sults of the docking calculations are comparable for all mo- GLU-IN orientation of glucoside molecules could be asso-
lecules, with ligands showing two possible binding modes ciated with the hydrolysis of the molecule, it is not surpris-
in which either the pyrrolizidine or the glucose moiety is ing that lactam molecules are not able to inhibit glucoamyl-
oriented in the –1 site; we call these two orientations CAS- ase.
IN and GLU-IN, respectively. In the CAS-IN orientation,
In summary, we have analysed the binding features of
the position of the pyrrolizidine nucleus is almost coinci- competitive pyrrolizidine inhibitors of glucoamylase as de-
dent with the docked pose of the corresponding unglucosyl- termined by docking simulations. None of the molecules we
ated compounds, except for the conformation of ring B, considered in this study presented structural variations in
which is influenced by the positioning of the glucose moi- ring A, which in casuarine perfectly mimics the stereochem-
ety, and for the lack of a hydrogen-bond donor for E179 at ical arrangement of glucose. As far as ring B is concerned,
the C6 atom. The glucose moiety is oriented in the outer none of the structural variations introduced at the 6- (in-
part of the +1 site, where it can assume different conforma- cluding glycosylation) and 7-positions seem to significantly
tions that allow it to hydrogen-bond E180 and/or Y311. influence the inhibitory activity. In contrast, the presence
When the molecules adopt the GLU-IN orientation, the of the lactam structure at the 4- and 5-positions has a very
glucose in the –1 site overlaps well with DNJ and with dramatic effect on the activity, and this could be due to
5580
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5574–5585

Casuarine Analogues
cotton. The solution was cooled to 0 °C, and, under vigorous stir-
ring, a saturated aqueous solution of NaHCO3 (100 mL) was added
until a basic pH was reached. The aqueous phase was extracted
with EtOAc (3ϫ20 mL), and the combined organic phases were
dried with Na2SO4. After filtration and concentration under re-
duced pressure, 16 was obtained as a yellow oil, pure enough to be
used in the next step (598 mg, 80%). An analytically pure sample
was obtained through purification by flash column chromatog-
raphy on silica gel (eluent: petroleum ether/EtOAc, 1:2, Rf = 0.4).
both the lack of a positive charge, whose role in stabilizing
the complex has been already highlighted, and the limited
conformational flexibility, which could determine unfavour-
able steric contacts.
Conclusions
We have reported a novel and efficient strategy for the
synthesis of casuarine-like pyrrolizidines and their 6-O-α-
glucoside derivatives. Our methodology was based on a tot-
ally regio- and stereoselective 1,3-dipolar cycloaddition of
suitably substituted alkenes with a carbohydrate-based
nitrone. After N–O bond cleavage of the cycloadducts thus
obtained, the lactams were used as key intermediates in the
synthesis of all the target compounds, including the gluco-
1
[α]2D0 = +3.05 (c = 0.9, CHCl3). H NMR (400 MHz, CDCl3): δ =
7.38–7.26 (m, 15 H, Ar), 4.60–4.45 (m, 7 H, Bn, 6-H), 4.34 (t, J =
4.3 Hz, 1 H, 2-H), 4.11 (q, J = 4.5 Hz, 1 H, 3-H), 3.87 (br. s, 1 H,
OH), 3.79 (dd, J = 6.9, 5.0 Hz 1 H, 1-H), 3.73 (dt, J = 8.6, 6.5 Hz,
1 H, 7a-H), 3.63 (dd, J = 9.8, 5.5 Hz, 1 H, 8-Ha), 3.52 (dd, J =
9.8, 4.1 Hz, 1 H, 8-Hb), 2.70 (ddd, J = 12.3, 7.8, 6.1 Hz, 1 H, 7-
Ha), 1.80 (ddd, J = 12.3, 10.4, 8.6 Hz, 1 H, 7-Hb) ppm. 13C NMR
(50 MHz, CDCl3): δ = 174.5 (s, C=O), 137.8, 137.7, 137.4 (s, Ar),
syl derivatives that were obtained by selective α-glucosyl- 128.5–127.6 (d, 15 C, Ar), 89.1 (d, C-1), 85.7 (d, C-2), 73.3, 72.6,
72.4 (t, Bn), 72.0 (d, C-6), 68.8 (t, C-8), 59.9 (d, C-7a), 58.6 (d, C-
ation. Evaluation of the inhibitory activity of casuarine and
its derivatives towards a wide range of commercially avail-
able glycosidases was undertaken, and several new inhibi-
tors of glucoamylase from Aspergillus niger were discov-
ered. Docking experiments performed on pyrrolizidine de-
rivatives allowed us to investigate the binding mode of the
competitive inhibitors. It is evident that an optimal network
of hydrogen-bonding interactions at the inner –1 site has to
be achieved for a ligand to bind. The presence of a positive
charge on the ligand is helpful for the stabilization of the
complex, independently of the possibility that a ligand has
to hydrogen-bond E179 through the charged atom. Finally,
interactions at the +1 site also seem to have an important
3), 37.2 (t, C-7) ppm. IR (CDCl ): ν = 3671, 3373, 3012, 2867,
˜
3
1697, 1454, 1100 cm–1. MS (EI): m/z (%) = 381 (12) [M – Bn]+, 336
(20), 275 (32), 180 (98), 153 (100), 88 (100). C29H31NO5 (473.56):
calcd. C 73.55, H 6.60, N 2.96; found C 73.33, H 6.80, N 2.87.
(1R,2R,3R,6R,7aR)-1,2,6-Trihydroxy-3-(hydroxymethyl)hexahydro-
5H-pyrrolizin-5-one (6): Pd (10% on C, 300 mg) was added to a
stirred solution of 16 (150 mg, 0.32 mmol) in EtOH (15 mL) under
nitrogen. The suspension was stirred under hydrogen at room temp.
for 3 d, then filtered through Celite® and washed with MeOH.
Concentration under reduced pressure afforded a viscous oil that
was purified by flash column chromatography on silica gel (eluent:
MeOH/EtOAc, 1:3, Rf = 0.14) to afford pure 6 as a transparent oil
(57 mg, 88% yield). [α]2D0 = –1.2 (c = 0.25, MeOH). 1H NMR
role in modulating the affinity of more extended molecules, (400 MHz, CD3OD): δ = 4.54 (dd, J = 10.5, 7.8 Hz, 1 H, 6-H),
4.16 (t, J = 6.3 Hz, 1 H, 2-H), 3.84 (dd, J = 3.0, 8.0 Hz, 1 H, 8-
Ha), 3.67–3.53 (m, 4 H, 8-Hb, 1-H, 3-H, 7a-H), 2.78 (ddd, J =
12.2, 7.8, 5.9 Hz, 1 H, 7-Ha), 1.74 (ddd, J = 12.0, 10.7, 8.3 Hz, 1
H, 7-Hb) ppm. 13C NMR (50 MHz, CD3OD): δ = 175.3 (s, C-5),
81.8 (d, C-1), 78.3 (d, C-2), 71.7 (d, C-6), 62.0 (d, C-3), 60.2 (t, C-
8), 59.8 (d, C-7a), 36.8 (t, C-7) ppm. MS: m/z (%) = 204 (5) [M +
H]+, 203 (3) [M]+, 185 (27) [M – H2O]+, 172 (52), 144 (74), 126
(51), 100 (87), 86 (100), 72 (59), 57 (38). C8H13NO5 (203.19): calcd.
C 47.29, H 6.45, N 6.89; found C 47.54, H 6.36, N 6.95.
but the comparison of 2 with acarbose clearly shows that
to increase the affinity for GA it is necessary to bind sub-
sites outside of the +1 site.
Experimental Section
General: Commercial reagents were used as received. All reactions
were carried with magnetic stirring and were monitored by TLC
on 0.25 mm silica gel plates (Merck F254). Column chromatography
was carried out on silica gel 60 (32–63 mm). Yields refer to spectro-
scopically and analytically pure compounds unless otherwise
stated. 1H NMR spectra were recorded with a Varian Mercury-
400 spectrometer. 13C NMR spectra were recorded with a Varian
Gemini-200 spectrometer. Infrared spectra were recorded with a
Perkin–Elmer Spectrum BX FT-IR System spectrophotometer.
Mass spectra were recorded with a QMD 1000 Carlo Erba instru-
ment by direct inlet injection; relative percentages are shown in
parentheses. ESI full mass spectra were recorded with a Thermo
LTQ instrument by direct inlet injection; relative percentages are
shown in parentheses. HR-ESI-TOF-MS experiments were per-
formed with a Q-Tof Ultima mass spectrometer (Waters) fitted with
a standard Z-spray ion source and operated in the positive ioniza-
tion mode. Elemental analyses were performed with a Perkin–El-
mer 2400 analyser. Optical rotation measurements were performed
with a JASCO DIP-370 polarimeter.
(1R,2R,3R,6R,7aR)-1,2-Bis(benzyloxy)-3-[(benzyloxy)methyl]hexa-
hydro-1H-pyrrolizin-6-ol (17): A 1  solution of LiAlH4 in THF
(1.6 mL, 1.61 mmol) was added to a cooled (0 °C) solution of 16
(255 mg, 0.54 mmol) in dry THF (6 mL) under nitrogen. The mix-
ture was then heated at reflux for 1.5 h. Then, after cooling to
0 °C, an aqueous saturated solution of Na2SO4 (560 µL) was added
dropwise. The suspension was then filtered through Celite® and
washed with EtOAc. Concentration under reduced pressure af-
forded 17 as a yellow oil pure enough for the next step (185 mg,
75% yield). An analytically pure sample was obtained through pu-
rification by flash column chromatography on silica gel (eluent:
petroleum ether/EtOAc, 1:4, Rf = 0.3). [α]2D0 = +9.1 (c = 0.83,
CHCl3). 1H NMR (400 MHz, CDCl3): δ = 7.36-7.24 (m, 15 H, Ar),
4.67–4.46 (m, 6 H, Bn), 4.35–4.31 (m, 1 H, 6-H), 4.11 (t, J = 4.6 Hz,
1 H, 1-H), 4.07 (t, J = 4.7 Hz, 1 H, 2-H), 3.60-3.44 (m, 4 H, 7a-H,
3-H, 8-Ha, 8-Hb), 3.21 (dd, J = 12.3, 4.5 Hz 1 H, 5-Ha), 2.98 (dm,
J = 12.3 Hz, 1 H, 5-Hb), 2.21 (ddd, J = 13.8, 9.0, 5.7 Hz, 1 H, 7-
Ha), 1.84 (dm, J = 13.8 Hz, 1 H, 7-Hb) ppm. 13C NMR (50 MHz,
CDCl3): δ = 138.1, 137.8, 137.4 (s, Ar), 128.1-127.2 (d, 15 C, Ar),
88.9 (d, C-1), 85.2 (d, C-2), 73.6 (d, C-6), 72.9, 72.1, 71.8, 71.6 (t,
Bn, C-8), 69.9 (d, C-3), 67.4 (d, C-7a), 63.3 (t, C-5), 40.0 (t, C-7)
(1R,2R,3R,6R,7aR)-1,2-Bis(benzyloxy)-3-[(benzyloxy)methyl]-6-hy-
droxyhexahydro-5H-pyrrolizin-5-one (16): A mixture of 14a
(816 mg, 1.58 mmol) and Zn dust (407 mg) in CH3COOH/H2O
(9:1, 12.5 mL) was heated at 50 °C for 4 h and then filtered through
Eur. J. Org. Chem. 2010, 5574–5585
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5581

C. Bonaccini, F. Cardona et al.
FULL PAPER
ppm. IR (CDCl ): ν = 3392, 3010, 2927, 2858, 1748, 1710, 1454, transferred to a column of DOWEX 50WX8 and then washed with
˜
3
1262 cm–1. MS (EI): m/z (%) = 366 (6), 336 (6), 216 (61), 160 (100), MeOH (10 mL), H2O (10 mL) to remove non-amine-containing
90 (100). C29H33NO4 (459.58): calcd. C 75.79, H 7.24, N 3.05;
found C 75.99, H 7.16, N 3.02.
products and then with 6% NH4OH (15 mL) to elute hyacinthac-
ine A2 (11) as a white solid (7 mg, 72% yield). [α]2D4 = +12.4 (c =
0.2, H2O); {ref.[16a] [α]2D4 = +12.7 (c = 0.13, H2O); ref.[18a] [α]D
=
(1R,2R,3R,6R,7aR)-3-(Hydroxymethyl)hexahydro-1H-pyrrolizine-
1,2,6-triol (7-Deoxycasuarine, 4): Concentrated HCl (4–5 drops)
and Pd (10% on C, 250 mg) were added to a stirred solution of
17 (120 mg, 0.26 mmol) in EtOH (10.5 mL) under nitrogen. The
suspension was stirred at room temp. under hydrogen for 3 d, then
filtered through Celite® and washed with MeOH. Concentration
under reduced pressure afforded a viscous yellow oil (66 mg) that
was transferred to a column of DOWEX 50WX8 and then washed
with MeOH (10 mL), H2O (10 mL) to remove non-amine-contain-
ing products and then with 6% NH4OH (15 mL) to elute 7-deoxy-
casuarine (4) as a white solid (43 mg, 88% yield), m.p. 205–208 °C.
[α]2D0 = +19.8 (c = 0.4, H2O) {ref.[17a] [α]2D0 = +10.9 (c = 0.11, H2O);
+12.5 (c = 0.4, H2O); ref.[16c] [α]2D0 = +19.9 (c = 0.97, MeOH);
ref.[18b] [α]2D5 = +10.5 (c = 0.6, H2O); ref.[18f] [α]D = +12.1 (c = 0.3,
H2O); ref.[18g] [α]2D6 = +12 (c = 0.4, H2O); ref.[25] [α]D = +20.1 (c =
1
0.44, H2O)}. H NMR (200 MHz, D2O): δ = 3.72–3.61 (m, 3 H),
3.53 (dd, J = 11.7, 6.2 Hz, 1 H), 3.10–3.00 (m, 1 H), 2.86–2.75 (m,
1 H), 2.70–2.56 (m, 2 H), 1.90–1.58 (m, 4 H) ppm. 13C NMR
(50 MHz, D2O): δ = 82.6 (d, C-1), 79.6 (d, C-2), 71.7 (d, C-3), 68.6
(d, C-7a), 65.3 (t, C-8), 57.4 (t, C-5), 32.2 (t, C-7), 27.0 (t, C-6)
ppm. C8H15NO3 (173.21): calcd. C 55.47, H 8.73, N 8.09; found C
55.49, H 8.61, N 8.10.
Methyl (1S,2R,6R,7R,7aR)-6,7-Bis(benzyloxy)-5-[(benzyloxy)meth-
yl]-2-hydroxy-3-oxohexahydro-1H-pyrrolizine-1-carboxylate (20): A
mixture of 14b (870 mg, 1.55 mmol) and Zn dust (400 mg,
6.2 mmol) in CH3COOH/H2O (9:1, 12.5 mL) was heated at 50 °C
for 3 h and then filtered through cotton. The solution was cooled
to 0 °C, and, under vigorous stirring, a saturated aqueous solution
of NaHCO3 (100 mL) was added until a basic pH was reached.
The aqueous phase was extracted with EtOAc (3ϫ60 mL), and the
combined organic phases were dried with Na2SO4. After filtration
and concentration under reduced pressure, 20 was obtained pure
as a white solid (743 mg, 90% yield), m.p. 111–113 °C. [α]2D0 = –28.7
(c = 0.64, CHCl3). 1H NMR (400 MHz, CDCl3): δ = 7.37–7.23 (m,
15 H, Ar), 4.76 (dd, J = 9.8, 2.9 Hz, 1 H, 2-H), 4.57–4.44 (m, 6 H,
Bn), 4.31–4.28 (m, 1 H, 5-H), 4.23–4.22 (m, 1 H, 6-H), 3.98–3.95
(m, 1 H, 7a-H), 3.92–3.90 (m, 1 H, 7-H), 3.79 (s, 3 H, Me), 3.58–
3.50 (m, 2 H, 8-H), 3.44 (d, J = 3.1 Hz, 1 H, OH), 3.03 (t, J =
9.3 Hz, 1 H, 1-H) ppm. 13C NMR (50 MHz, CDCl3): δ = 173.2,
171.1 (s, C=O), 137.7–137.2 (s, 3 C, Ar), 128.4–127.5 (d, 15 C, Ar),
87.3 (d, C-7), 84.6 (d, C-6), 74.2 (d, C-2), 73.1, 72.1, 71.8 (t, Bn),
68.1 (t, C-8), 62.7 (d, C-7a), 59.3 (d, C-5), 54.8 (d, C-1), 52.5 (q,
1
ref.[16b] [α]2D5 = +23 (c = 0.3, MeOH)}. H NMR (400 MHz, D2O):
δ = 4.44–4.38 (m, 1 H, 6-H), 4.05 (t, J = 8.2 Hz, 1 H, 1-H), 3.74–
3.69 (m, 2 H), 3.56 (dd, J = 11.9, 6.5 Hz, 1 H), 3.24 (ddd, J = 12.6,
8.2, 4.2 Hz, 1 H), 3.09–3.02 (m, 2 H), 2.86 (m, 1 H, 5-Hb), 2.12
(ddd, J = 13.6, 8.5, 3.4 Hz, 1 H, 7-Ha), 1.89 (dm, J = 13.6 Hz, 1
H, 7-Hb) ppm. 13C NMR (50 MHz, D2O): δ = 82.6 (d, C-1), 79.0
(d, C-2), 75.1 (d, C-6), 72.5 (d, C-7a), 67.9 (d, C-3), 64.6 (t, C-8),
63.3 (t, C-5), 39.3 (t, C-7) ppm. MS (EI): m/z (%) = 190 (3) [M +
H]+, 189 (1) [M]+, 176 (66), 158 (65) [M – CH2OH]+, 132 (64), 112
(23), 85 (62), 58 (100). C8H15NO4 (189.21): calcd. C 50.78, H 7.99,
N 7.40; found C 50.37, H 7.63, N 7.74.
(1R,2R,3R,7aR)-1,2-Bis(benzyloxy)-3-(benzyloxymethyl)hexahydro-
1H-pyrrolizine (18): NEt3 (75 µL, 0.54 mmol) was added to a
stirred solution of 16 (95 mg, 0.20 mmol) in dry CH2Cl2 (0.45 mL)
under nitrogen, and, at 0 °C, MsCl (20 µL, 0.26 mmol) was added
dropwise. The solution was stirred at 0 °C for 30 min and at room
temp. for 2 h. The mixture was filtered through Celite® and washed
with EtOAc. Concentration under reduced pressure afforded the
mesylated derivative as a white oil (quantitative yield), which was
dissolved in dry THF (2.5 mL). A 1  solution of LiAlH4 in THF
(0.8 mL, 0.8 mmol) was added dropwise at 0 °C under nitrogen.
The mixture was heated at reflux for 3 h. An aqueous saturated
solution of Na2SO4 (280 µL) was added dropwise, and the mixture
was stirred at room temp. for 10 min. After filtration through Ce-
lite®, a crude residue (109 mg) was obtained that was purified by
flash column chromatography on silica gel (petroleum ether/ethyl
acetate, 1:2, Rf = 0.28) to afford pure 18 (71 mg, 80% yield) as an
Me) ppm. IR (CHCl ): ν = 3690, 3600–3500 (br), 3027, 2920, 1708,
˜
3
1601, 1155, 1070 cm–1. MS (EI): m/z (%) = 513 (0.6) [M – H2O]+,
212 (4), 91 (100), 69 (14). C31H33NO7 (531.6): calcd. C 70.04, H
2.63, N 6.26; found C 70.02, H 2.56, N 6.28.
(1R,2R,3R,6R,7R,7aR)-1,2-Bis(benzyloxy)-3-[(benzyloxy)methyl]-7-
(hydroxymethyl)hexahydro-1H-pyrrolizin-6-ol (21): A 1  solution
of LiAlH4 in THF (1.1 mL, 1.1 mmol) was added to a cooled (0 °C)
solution of 20 (115 mg, 0.22 mmol) in dry THF (3 mL) under nitro-
gen. The mixture was then heated at reflux for 2 h. Then, after
cooling at 0 °C, an aqueous saturated solution of Na2SO4 (700 µL)
was added dropwise. The suspension was then filtered through Ce-
lite® and washed with EtOAc. Concentration under reduced pres-
sure afforded solid 21 pure enough to be used in the next step
(104 mg, quantitative yield). An analytically pure sample was ob-
tained by filtration through a short pad of silica gel (eluent: EtOAc
then EtOAc/MeOH, 5:1), m.p. 85–87 °C. [α]2D0 = +3.58 (c = 1.18,
CHCl3). 1H NMR (400 MHz, CDCl3): δ = 7.40–7.25 (m, 15 H,
Ar), 4.64–4.56 (m, 6 H, Bn), 4.22 (q, J = 5.6 Hz, 1 H, 6-H), 4.13–
4.09 (m, 2 H, 1-H, 2-H), 3.68 (d, J = 6.8 Hz, 2 H, 8-Ha, 8-Hb),
3.53 (d, J = 6.4 Hz, 2 H, 9-Ha, 9-Hb) 3.43–3.37 (m, 2 H, 3-H, 5-
Ha) 3.27 (dd, J = 7.0, 3.9 Hz, 1 H, 7a-H), 2.95 (dd, J = 10.8,
5.6 Hz, 1 H, 5-Hb), 2.24 (quint., J = 6.6 Hz, 1 H, 7-H) ppm. 13C
NMR (50 MHz, CDCl3): δ = 138.2–137.7 (s, 3 C, Ar), 128.5–127.5
1
oil. [α]2D4 = –5.1 (c = 0.6, CHCl3). H NMR (400 MHz, CDCl3): δ
= 7.37–7.25 (m, 15 H, Ar), 4.73–4.46 (m, 6 H, Bn), 4.08 (dd, J =
7.4, 5.9 Hz, 1 H, 2-H), 3.81 (t, J = 5.9 Hz, 1 H, 1-H), 3.60 (dd, J
= 9.6, 4.7 Hz, 1 H, 8-Ha), 3.54 (dd, J = 9.6, 5.7 Hz, 1 H, 7a-H),
3.51–3.46 (m, 1 H, 8-Hb), 3.07 (dt, J = 10.5, 6.1 Hz, 1 H, 5-Ha),
2.98–2.94 (m, 1 H, 3-H), 2.78 (dt, J = 10.5, 6.6 Hz, 1 H, 5-Hb),
2.03–1.95 (m, 1 H, 7-Ha), 1.91–1.82 (m, 1 H, 6-Ha), 1.81–1.74 (m,
1 H, 6-Hb), 1.72–1.62 (m, 1 H, 7-Hb) ppm. 13C NMR (50 MHz,
CDCl3): δ = 138.2, 138.1, 137.9 (s, Ar), 128.0–127.1 (d, 15 C, Ar),
88.6 (d, C-1), 85.5 (d, C-2), 73.0, 72.3, 71.8, 71.6 (t, C-8, Bn), 68.0
(d, C-3), 67.2 (d, C-7a), 54.8 (t, C-5), 31.4 (t, C-6), 25.5 (t, C-7)
ppm.
(1R,2R,3R,7aR)-3-(Hydroxymethyl)hexahydro-1H-pyrrolizine-1,2-
diol (Hyacinthacine A2, 11): Concentrated HCl (3 drops) and Pd
(10% on C, 45 mg) were added to a stirred solution of 18 (25 mg, (d, 15 C, Ar), 88.0 (d), 86.1 (d), 76.6 (d), 73.3 (t), 72.4 (t), 72.0 (t),
0.056 mmol) in MeOH (2.5 mL). The mixture was stirred at room
temp. under hydrogen for 3 d. The mixture was then filtered
through Celite® and washed with MeOH. The solvent was evapo-
rated under reduced pressure to afford a viscous white oil that was
71.5 (t), 70.4 (d), 70.4 (d), 64.2 (t), 62.4 (t), 54.3 (d) ppm. IR
(CDCl ): ν = 3412, 3031, 3010, 2866, 1496, 1454, 1363, 1216, 1212,
˜
3
1211, 1097 cm–1. MS (EI): m/z (%) = 398 (11) [M – C7H7]+, 368
(85), 248 (11), 186 (27), 160 (30), 142 (25), 116 (22), 91 (100), 64
5582
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5574–5585

Casuarine Analogues
(21). C30H35NO5 (489.6): calcd. C 73.59, H 7.21, N 2.86; found C
73.58, H 7.09, N 3.18.
= 9.8, 7.6 Hz, 1 H, 6-Hb), 3.26 (dd, J = 10.8, 9.2 Hz, 1 H, 4-H),
3.13 (t, 1 H, OH), 2.44 (m, 1 H, 7Ј-H) ppm. 13C NMR (50 MHz,
CDCl3): δ = 171.4 (s, C=O), 138.4–136.7 (s, 7 C, Ar), 128.7–127.3
(d, 35 C, Ar), 94.9 (d, C-1), 87.7 (d, C-1Ј), 86.1 (d, C-2Ј), 81.4 (d,
C-3), 78.2 (d, C-2), 77.4 (d, C-4), 75.6 (d, C-6Ј), 75.6, 74.9, 73.3,
73.0, 71.9, 71.7, 71.2 (t, Bn), 70.5 (d, C-5), 68.8 (t, C-6), 68.7 (t, C-
9Ј), 60.5 (d, C-7aЈ), 58.4 (t, C-8Ј), 58.4 (d, C-3Ј), 50.4 (d, C-7Ј)
(1R,2R,3R,6R,7R,7aR)-3,7-Bis(hydroxymethyl)hexahydro-1H-pyr-
rolizine-1,2,6-triol (7-Homocasuarine, 7): Concentrated HCl (4–
5 drops) and Pd (10% on C, 230 mg) were added to a stirred solu-
tion of 21 (106 mg, 0.22 mmol) in EtOH (14 mL). The suspension
was stirred at room temp. under hydrogen for 4 d, then filtered
through Celite® and washed with EtOH. Evaporation under re-
duced pressure afforded a viscous oil that was transferred to a col-
umn of DOWEX 50WX8 and then washed with MeOH (10 mL),
H2O (10 mL) to remove non-amine-containing products and then
with 6% NH4OH (15 mL) to elute 7-homocasuarine (7). Evapora-
tion of the solvent afforded 7-homocasuarine as a yellow viscous
oil (38.5 mg, 89%). [α]2D0 = +30.8 (c = 0.7, MeOH). 1H NMR
(400 MHz, D2O): δ = 4.39 (q, J = 10.0 Hz, 1 H, 6-H), 4.04 (t, J =
7.8 Hz, 1 H, 1-H), 3.71 (t, J = 8.5 Hz, 1 H, 2-H), 3.67 (dd, J =
11.7, 3.5 Hz, 1 H, 8-Ha), 3.55-3.44 (m, 3 H, 8-Hb, 9-Ha, 9-Hb),
3.19 (dd, J = 10.0, 5.2 Hz, 1 H, 5-Ha), 3.02–2.95 (m, 2 H, 3-H, 7a-
H), 2.80 (dd, J = 10.0, 5.2 Hz, 1 H, 5-Hb), 2.21 (q, J = 5.9 Hz, 1
H, 7-H) ppm. 13C NMR (50 MHz, D2O): δ = 80.2 (d, C-1), 77.5
(d, C-2), 74.5 (d, C-6), 70.6, 68.5 (d, C-3, C-7), 62.7 (t, C-8), 61.3
(t, C-9), 60.4 (t, C-5), 52.9 (d, C-7a) ppm. MS (EI): m/z (%) = 188
(100), 170 (10), 159 (13), 142 (14), 128 (83), 116 (26), 68 (38), 55
(24). C9H17NO5 (219.23): calcd. C 49.31, H 7.82, N 6.39; found C
49.06, H 7.43, N 6.54.
ppm. IR (CDCl ): ν = 3463 (OH), 3032, 2925, 2870, 1703 (C=O),
˜
3
1454, 1078 cm–1. HRMS (ESI): calcd. for C64H67NO11Na [M +
Na]+ 1048.4606; found 1048.4602. C64H67NO11 (1026.22): calcd. C
74.90, H 6.58, N 1.36; found C 74.65, H 6.69, N 1.53.
6-O-(α-D-Glucopyranosyl)-5-oxo-7-homocasuarine (9): Pd (10% on
C, 230 mg) was added to a stirred solution of 23 (134 mg,
0.13 mmol) in MeOH (13 mL). The suspension was stirred at room
temp. under hydrogen for 24 h, then filtered through Celite® and
washed with MeOH. Concentration under reduced pressure af-
forded pure 9 as a waxy solid (37 mg, 72%). [α]2D6 = +59.3 (c =
1
0.75, MeOH). H NMR (400 MHz, D2O): δ = 5.18 (d, J = 4 Hz,
1 H, 1-H), 4.57 (d, J = 9.6 Hz, 1 H, 6Ј-H), 4.09 (dd, J = 6.8, 6.4 Hz,
1 H, 2Ј-H), 3.82–3.58 (m, 10 H, 1Ј-H, 3Ј-H, 8Ј-Ha,b, 9Ј-Ha, 3-H,
4-H, 5-H, 6-Ha,b), 3.50–3.45 (m, 2 H, 7aЈ-H, 2-H), 3.34 (dd, J =
9.6, 9.2 Hz, 1 H, 9Ј-Hb), 2.50–2.44 (m, 1 H, 7Ј-H) ppm. 13C NMR
(50 MHz, D2O): δ = 173.9 (s, C=O), 98.6 (d, C-1), 79.9 (d, C-6Ј),
77.9, 77.3, 72.7, 72.5, 71.2, 69.3, 62.1, 61.2 (d, 1 C), 60.3, 59.8, 59.7
(t, 1 C), 50.0 (d, 1 C) ppm. IR (KBr): ν = 3378 (OH), 1689 (C=O)
˜
cm–1. HRMS (ESI): calcd. for C65H67NO12Na [M
+
Na]+
7-Deoxy-6-O-(α-D-glucopyranosyl)-7-(methoxycarbonyl)-5-oxocasu-
418.1320; found 418.1312. C15H25NO11 (395.36): calcd. C 45.57, H
6.37, N 3.54; found C 45.25, H 6.28, N 3.04.
arine (10): Pd (10% on C, 150 mg) was added to a stirred solution
of 22 (125 mg, 0.118 mmol) in MeOH/AcOEt (3:1, 12 mL). The
suspension was stirred at room temp. under hydrogen for 4 d, then
filtered through Celite® and washed with MeOH. Concentration
under reduced pressure afforded pure 10 as a waxy solid (48 mg,
96%). [α]2D1 = +75.3 (c = 0.15, MeOH). 1H NMR (400 MHz, D2O):
δ = 5.18 (d, J = 3.6 Hz, 1 H, 1-H), 4.97 (d, J = 9.6 Hz, 1 H, 6Ј-H),
4.08 (dd, J = 6.8, 6.4 Hz, 1 H, 2Ј-H), 3.88 (dd, J = 7.6, 7.2 Hz, 1
H, 1Ј-H), 3.80–3.55 (m, 11 H, 5-H, OCH3, 4-H, 3-H, 6-Ha, 6-Hb,
8-HaЈ, 7aЈ-H, 3Ј-H), 3.41 (dd, J = 9.6, 3.7 Hz, 1 H, 2-H), 3.37 (t,
J = 9.6 Hz, 1 H, 8-HbЈ), 3.28 (dd, J = 9.4, 8.2 Hz, 1 H, 7Ј-H) ppm.
13C NMR (50 MHz, D2O): δ = 171.9, 171.5 (s, C=O), 98.8 (d, C-
1), 79.2 (d, C-6Ј), 78.3, 77.3, 72.2, 72.1, 70.9, 68.6, 61.4, 60.9 (d, 1
C), 59.5, 52.2 (t, 1 C), 52.8 (q, OMe), 52.2 (d, 1 C) ppm. IR (KBr):
6-O-(α-D-Glucopyranosyl)-5-oxocasuarine (3): Pd (10% on C,
180 mg) was added to a stirred solution of 24 (159 mg, 0.144 mmol)
in MeOH/EtOAc (7:1, 12 mL). The suspension was stirred at room
temp. under hydrogen for 24 h, then filtered through Celite® and
washed with MeOH. Concentration under reduced pressure af-
forded pure 3 (42 mg, 0.110 mmol, 77% yield) as a hygroscopic
pale-yellow oil. [α]2D4 = +37.6 (c = 0.28, MeOH). 1H NMR
(400 MHz, D2O): δ = 5.14 (d, J = 3.7 Hz, 1 H, 1-H), 4.53 (d, J =
8.4 Hz, 1 H, 6Ј-H), 4.27 (dd, J = 8.4, 7.0 Hz, 1 H, 7Ј-H), 4.08 (t, J
= 6.1 Hz, 1 H, 2Ј-H), 3.90 (t, J = 6.8 Hz, 1 H, 1Ј-H), 3.78–3.58 (m,
7 H, 3-H, 3Ј-H, 5-H, 6-Ha,b, 8Ј-Ha,b), 3.50–3.44 (m, 2 H, 2-H,
7aЈ-H), 3.34 (t, J = 9.5 Hz, 1 H, 4-H) ppm. 13C NMR (50 MHz,
D2O): δ = 170.8 (s, C-5Ј), 98.0 (d, C-1), 81.6 (d, C-6Ј), 78.2 (d, 1
C), 77.3 (d, C-7Ј), 76.9 (d, C-2Ј), 71.9 (d), 71.5 (d), 70.5 (d), 68.6
(d), 64.7 (d), 60.7 (d), 59.6 (t), 58.8 (t) ppm. MS (ESI): m/z (%) =
404 (100) [M + Na]+. C14H23NO11 (381.33): calcd. C 44.10, H 6.08,
N 3.67; found C 43.96, H 6.19, N 3.42.
ν = 3420 (OH), 1710 (C=O), 1684 (C=O), 1205, 1143, 1024 cm–1.
˜
HRMS (ESI): calcd. for C16H25NO12Na [M + Na]+ 446.1269;
found 446.1266. C16H25NO12 (423.37): calcd. C 45.39, H 5.95, N
3.31; found C 44,93, H 6.27, N 3,37.
6-O-α-D-Tris(benzyloxy)-5-oxo-6-O-(2,3,4,6-tetra-O-benzyl-α-D-gluco-
Enzymatic Assays: The experiments were performed essentially as
follows: 0.01–0.5 unit/mL of enzyme (1 unit = 1 mol of glycoside
hydrolysed/min), preincubated at 20 °C with the inhibitor for
5 min, and increasing concentrations of an aqueous solution of the
appropriate p-nitrophenyl glycoside substrates (buffered at the opti-
mum pH of the enzyme) were incubated at 37 °C for 20 min. The
reactions were stopped by the addition of 0.3  sodium borate
buffer (100 µL, pH = 9.8). The p-nitrophenolate formed was quan-
tified at 405 nm, and IC50 values were calculated. Double-recipro-
cal (Lineweaver–Burk) plots were used to determine the inhibition
characteristics and the Ki values for each compound.
pyranosyl)-7-homocasuarine (23): A 2  solution of LiBH4 in THF
(0.42 mL) was added dropwise to a cooled (0 °C) solution of 22
(221 mg, 0.21 mmol) in dry THF (1.5 mL). The reaction mixture
was stirred at room temp. overnight, and then, after cooling to
0 °C, H2O was added dropwise. The mixture was then filtered
through Celite®, washed with CHCl3 and concentrated under re-
duced pressure. The residue was purified by flash column
chromatography on silica gel (EtOAc/hexane, 1:4) to afford pure
23 (Rf = 0.33, EtOAc/petroleum ether, 1:3) as a colourless oil
(134 mg, 62%). [α]2D3 = +44.4 (c = 0.4, CHCl3). 1H NMR
(400 MHz, CDCl3): δ = 7.47–7.08 (m, 35 H, Ar), 5.68 (d, J =
3.6 Hz, 1 H, 1-H), 5.09 (d, J = 11.7 Hz, 1 H, Bn), 5.02 (d, J = Molecular Modelling: The ligand structures (Table 1) were con-
10.9 Hz, 1 H, Bn), 4.81–4.75 (m, 3 H, Bn), 4.72 (d, J = 7.6 Hz, 1 structed by using Maestro v8.5.[29] All the molecules were subjected
H, 6Ј-H), 4.60–4.40 (m, 9 H, Bn), 4.26 (dd, J = 3.6, 4 Hz, 1 H, 2Ј- to conformational search and clusterization with Macromodel
H), 4.21 (m, 1 H, 3Ј-H), 3.94 (t, J = 9.2 Hz, 1 H, 3-H), 3.87–3.66
(m, 6 H, 5-H, 6-Ha, 7aЈ-H, 8Ј-Ha,b, 1Ј-H), 3.60 (dd, J = 9.2,
9.6[30] in order to sample the most accessible conformations of both
the aglyconic and glucose moieties. The bridgehead nitrogen atoms
3.6 Hz, 1 H, 2-H), 3.45 (d, J = 5.2 Hz, 2 H, 9Ј-Ha,b), 3.37 (dd, J were treated as ionized to better simulate the physiological condi-
Eur. J. Org. Chem. 2010, 5574–5585
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5583

C. Bonaccini, F. Cardona et al.
FULL PAPER
Prod. 2004, 67, 846–850; d) A. Kato, N. Kato, I. Adachi, J.
Hollinshead, G. W. J. Fleet, C. Kuriyama, K. Ikeda, N. Asano,
R. J. Nash, J. Nat. Prod. 2007, 70, 993–997; e) J. P. Saludes,
S. C. Lievens, T. F. Molinski, J. Nat. Prod. 2007, 70, 436–438.
a) H. Paulsen, I. Sangster, K. Heyns, Chem. Ber. 1967, 100,
802–815; b) M. Yagi, T. Kouno, Y. Aoyagi, H. Murai, Nippon
Nogei Kagaku Kaishi 1976, 50, 571–572.
a) I. Pastuszak, R. J. Molyneux, L. F. James, A. D. Elbein, Bio-
chemistry 1990, 29, 1886–1891; b) F. Cardona, A. Goti, A.
Brandi, Eur. J. Org. Chem. 2007, 1551–1565; c) F. Cardona, G.
Moreno, F. Guarna, P. Vogel, C. Schuetz, P. Merino, A. Goti,
J. Org. Chem. 2005, 70, 6552–6555.
B. Macchi, A. Minutolo, S. Grelli, A. Mastino, F. Cardona,
F. M. Cordero, A. Brandi, Glycobiology 2010, 20, 500–506.
a) M. J. Humphries, K. Matsumoto, L. S. White, K. Olden,
Cancer Res. 1986, 46, 5215–5222; b) G. K. Ostrander, N. K.
Scribner, L. R. Rohrschneider, Cancer Res. 1988, 48, 1091–
1094; c) S. P. Sunkara, P. S. Liu, Eur. Pat. Appl. EP 373663,
1990; d) M. A. Spearman, J. M. Ballon, J. M. Gerrard, A. H.
Greenberg, J. A. Wright, Cancer Lett. 1991, 60, 185–191; e)
R. Pili, J. Chang, R. A. Partis, R. A. Mueller, F. J. Chrest, A.
Passaniti, Cancer Res. 1995, 55, 2920–2926; f) C. S. Yee, E. D.
Schwab, J. E. Lehr, M. Quigley, K. J. Pienta, Anticancer Res.
1997, 17, 3659–3663.
a) R. J. Nash, P. I. Thomas, R. D. Waigh, G. W. J. Fleet, M. R.
Wormald, P. M. Q. Lilley, D. J. Watkin, Tetrahedron Lett. 1994,
35, 7849–7852; b) M. R. Wormald, R. J. Nash, A. A. Watson,
B. K. Bhadoria, R. Langford, A. Sims, G. W. J. Fleet, Car-
bohydr. Lett. 1996, 2, 169–174.
tions, except for the lactam intermediates. All the docking calcula-
tions were performed by using Glide 5.0.[31] The crystal structure
of glucoamylase-471 from Aspergillus awamori complexed with 1-
deoxynojirimycin (PDB ID: 1DOG)[26a] was downloaded from the
PDB and prepared according to the recommended Protein Prepa-
ration module in Maestro 8.5 by using default input parameters
(no scaling factors for the van der Waals radii of non-polar protein
atoms, 0.8 scaling factor for non-polar ligand atoms). This pro-
cedure was used to remove water molecules (except for molecule
Wat501, which is considered as part of the target structure), to
assign missing hydrogen atoms, to optimize hydrogen-bonding in-
teractions and to reduce structural problems. The grids were pre-
pared with the centre of the site defined by the centre of the com-
plexed ligand. All the relevant conformations for the ligands in
Table 1 were docked in the binding site by using the SP scoring
function to score the ligand poses. The docking calculations were
performed in the presence of Wat501. After completion of each
docking run, one pose per ligand conformation was saved. Finally,
for each ligand, the poses (conformations) with the lowest (best)
value of either the model energy score (Emodel)[32] or the Glide
score were chosen.
[6]
[7]
[8]
[9]
Supporting Information (see footnote on the first page of this arti-
cle): Syntheses of 5, 8, 8a, 10a, 16a, 19, 19a,b, 22 and 22a, copies
of the NMR spectra of 3–11, 8a, 10a, 16–23, 16a, 19a,b and 22a,
HR-ESI-TOF-MS spectra and the docking results for 2.
[10]
[11]
[12]
F. Cardona, C. Parmeggiani, E. Faggi, C. Bonaccini, P. Grat-
teri, L. Sim, T. M. Gloster, S. Roberts, G. J. Davies, D. R. Rose,
A. Goti, Chem. Eur. J. 2009, 15, 1627–1636.
K. Hiromi, Z. I. Hamauzu, K. Takahashi, S. Ono, J. Biochem.
1966, 59, 411–418.
B. C. Saha, J. G. Zeikus, Starch/Staerke 1989, 41, 57–64.
J. Sauer, B. W. Sigurskjold, U. Christensen, T. P. Frandsen, E.
Mirgrorodskaya, M. Harrison, P. Roepstorff, B. Svensson, Bi-
ochim. Biophys. Acta 2000, 1543, 275–293.
Acknowledgments
We thank the Ministero dell’Istruzione, dell’Università e della
Ricerca (PRIN 2007 and 2008) and the Ente Cassa di Risparmio
di Firenze, Italy, for financial support. Ente Cassa di Risparmio di
Firenze, Italy, is also gratefully acknowledged for a grant to C. B.
and for granting a 400 MHz NMR spectrometer. B. Innocenti and
M. Passaponti (Dipartimento di Chimica “Ugo Schiff”) are ac-
knowledged for technical assistance. Consorzio Interuniversitario
Nazionale “Metodologie e Processi Innovativi di Sintesi” is grate-
fully acknowledged for a grant to C. P. We also thank the Swiss
National Science Foundation for financial support and Dr. L.
Menin and Dr. A. Razaname for the HRMS measurements.
[13]
[14]
[15]
[16]
A. Brandi, F. Cardona, S. Cicchi, F. M. Cordero, A. Goti,
Chem. Eur. J. 2009, 15, 7808–7821.
a) F. Cardona, E. Faggi, F. Liguori, M. Cacciarini, A. Goti,
Tetrahedron Lett. 2003, 44, 2315–2318; b) A. T. Carmona,
R. H. Wightman, I. Robina, P. Vogel, Helv. Chim. Acta 2003,
86, 3066–3073; c) S. Desvergnes, S. Py, Y. Vallée, J. Org. Chem.
2005, 70, 1459–1462; d) J. Revuelta, S. Cicchi, A. Goti, A.
Brandi, Synthesis 2007, 485–504; e) E.-L. Tsou, Y.-T. Yeh, P.-
H. Liang, W.-C. Cheng, Tetrahedron 2009, 65, 93–100; f) X.-G.
Hu, Y.-M. Jia, J. Xiang, C.-Y. Yu, Synlett 2010, 982–986.
For other total syntheses of 7-deoxycasuarine, see: a) J.-B.
Behr, A. Erard, G. A. Guillerm, Eur. J. Org. Chem. 2002, 1256–
1262; b) ref.[16b]; c) J.-B. Behr, A. Gainvors-Claisse, A. Belrabe,
Nat. Prod. Res. 2006, 20, 1308–1314.
For other total syntheses of hyacinthacine A2, see: a) L. Ram-
baud, P. Compain, O. R. Martin, Tetrahedron: Asymmetry
2001, 12, 1807–1809; b) I. Izquierdo, M. T. Plaza, F. Franco,
Tetrahedron: Asymmetry 2003, 14, 3933–3935; c) ref. 16c; d) P.
Dewi-Wuelfing, S. Blechert, Eur. J. Org. Chem. 2006, 1852–
1856; e) J. Calveras, J. Casas, T. Parella, J. Joglar, P. Clapes, Adv.
Synth. Catal. 2007, 349, 1661–1666; f) C. Ribes, E. Falomir, M.
Carda, J. A. Marco, Tetrahedron 2009, 65, 6965–6971; g) I.
Delso, T. Tejero, A. Goti, P. Merino, Tetrahedron 2010, 66,
1220–1227.
[1] a) P. Compain, O. R. Martin (Eds.), Iminosugars: from synthe-
sis to therapeutic applications, Wiley-VCH, Weinheim, 2007; b)
N. Asano, Curr. Top. Med. Chem. 2003, 3, 471–484, and refer-
ences cited therein.
[2] a) N. Asano, Glycobiology 2003, 13, 93R–104R, and references
cited therein; b) N. Asano, R. J. Nash, R. J. Molyneux, G. W. J.
Fleet, Tetrahedron: Asymmetry 2000, 11, 1645–1680; c) I. Rob-
ina, A. J. Moreno-Vargas, A. T. Carmona, Curr. Drug Metab.
2004, 5, 329–361.
[3] a) A. Vasella, T. D. Heightman, Angew. Chem. Int. Ed. 1999,
38, 750–770; b) T. M. Gloster, P. Meloncelli, R. V. Stick, D.
Zechel, A. Vasella, G. J. Davies, J. Am. Chem. Soc. 2007, 129,
2345–2354; c) M. E. C. Caines, S. M. Hancock, C. A. Tarling,
T. M. Wrodnigg, R. V. Stick, A. E. Stütz, A. Vasella, S. J. With-
ers, N. C. J. Strynadka, Angew. Chem. Int. Ed. 2007, 46, 4474–
4476; d) T. M. Gloster, G. J. Davies, Org. Biomol. Chem. 2010,
8, 305–320.
[4] a) A. A. Watson, R. J. Nash, E. L. Evinson, PCT Int. Appl.
WO2004064715, 2004; b) R. J. Nash, A. A. Watson, E. L. Evin-
son, H. St. P. Parry, PCT Int. Appl. WO2005070418, 2005.
[5] a) A. A. Watson, G. W. J. Fleet, N. Asano, R. J. Molyneux,
R. J. Nash, Phytochemistry 2001, 56, 265–295; b) T. Yamashita,
K. Yasuda, H. Kizu, Y. Kameda, A. A. Watson, R. J. Nash,
G. W. J. Fleet, N. Asano, J. Nat. Prod. 2002, 65, 1875–1881; c)
N. Asano, K. Ikeda, M. Kasahara, Y. Arai, H. Kizu, J. Nat.
[17]
[18]
[19]
F. Cardona, A. Goti, C. Parmeggiani, P. Parenti, M. Forcella,
P. Fusi, L. Cipolla, S. M. Roberts, G. J. Davis, T. M. Gloster,
Chem. Commun. 2010, 46, 2629–2631.
S. Nukui, M. Sodeoka, H. Sasai, M. Shibasaki, J. Org. Chem.
1995, 60, 398–404.
For other total syntheses of casuarine, see: a) S. E. Denmark,
A. R. Hurd, Org. Lett. 1999, 1, 1311–1314; b) S. Denmark,
A. R. Hurd, J. Org. Chem. 2000, 65, 2875–2886; c) I. Izquierdo,
[20]
[21]
5584
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5574–5585

Casuarine Analogues
M. T. Plaza, J. A. Tamayo, Tetrahedron 2005, 61, 6527–6533;
d) J. Van Ameijde, G. Horne, M. R. Wormald, R. A. Dwek,
R. J. Nash, P. W. Jones, E. L. Evinson, G. W. J. Fleet, Tetrahe-
dron: Asymmetry 2006, 17, 2702–2712; e) T. Ritthiwigrom,
A. C. Willis, S. G. Pyne, J. Org. Chem. 2010, 75, 815–824.
[22] C. Parmeggiani, D. Martella, F. Cardona, A. Goti, J. Nat.
Prod. 2009, 72, 2058–2060.
A. E. Aleshin, C. Hoffman, L. M. Firsov, R. B. Honzatko, J.
Mol. Biol. 1994, 238, 575–591; e) A. E. Aleshin, L. M. Firsov,
R. B. Honzatko, J. Biol. Chem. 1994, 269, 15631–15639.
A. M. Davis, S. J. Teague, Angew. Chem. Int. Ed. 1999, 38, 736–
749.
T. Weimar, B. Stoffer, B. Svensson, B. M. Pinto, Biochemistry
2000, 39, 300–306.
Maestro, v. 8.5, Schrödinger, L.L.C., New York, 2008, available
at http://www.schrodinger.com.
[27]
[28]
[29]
[23] A. Brandi, S. Cicchi, F. M. Cordero, R. Frignoli, A. Goti, S.
Picasso, P. Vogel, J. Org. Chem. 1995, 60, 6806–6812.
[24] A. Kato, E. Kano, I. Adachi, R. J. Molyneux, A. A. Watson,
R. J. Nash, G. W. J. Fleet, M. R. Wormald, H. Kizu, K. Ikeda,
N. Asano, Tetrahedron: Asymmetry 2003, 14, 325–331.
[25] N. Asano, H. Kuroi, K. Ikeda, H. Kizu, Y. Kameda, A. Kato,
I. Adachi, A. A. Watson, R. J. Nash, G. W. J. Fleet, Tetrahe-
dron: Asymmetry 2000, 11, 1–8.
[26] a) E. M. S. Harris, A. E. Aleshin, L. M. Firsov, R. B. Hon-
zatko, Biochemistry 1993, 32, 1618–1626; b) A. E. Aleshin, B.
Stoffer, L. M. Firsov, B. Svensson, R. B. Honzatko, Biochemis-
try 1996, 35, 8319–8328; c) A. Aleshin, A. Golubev, L. M. Fir-
sov, R. B. Honzatko, J. Biol. Chem. 1992, 267, 19291–19298; d)
[30] Macromodel, v. 9.6, Schrödinger, L.L.C., New York, 2008,
available at http://www.schrodinger.com.
[31] Glide, v. 5.0, Schrödinger, L.L.C., New York, 2008, available at
http://www.schrodinger.com.
[32] R. A. Friesner, J. L. Banks, R. B. Murphy, T. A. Halgren, J. J.
Klicic, D. Mainz, M. P. Repasky, E. H. Knoll, M. Shelley, J. K.
Perry, D. E. Shaw, P. Francis, P. S. Shenkin, J. Med. Chem.
2004, 47, 1739–1749.
Received: May 5, 2010
Published Online: August 25, 2010
Eur. J. Org. Chem. 2010, 5574–5585
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5585