From natural product-inspired pyrrolidine scaffolds to the development of new human Golgi α-mannosidase II inhibitors

Article abstract of DOI:10.1002/asia.201300680

A little inspiration: The systematic preparation of sixteen natural product-inspired polyhydroylated pyrrolidine-based isomeric scaffolds is described. Each scaffold possesses four stereogenic centers and one exo-aminomethyl moiety, which allows for rapid

Full text of DOI:10.1002/asia.201300680

COMMUNICATION
DOI: 10.1002/asia.201300680
From Natural Product-Inspired Pyrrolidine Scaffolds to the Development of
New Human Golgi a-Mannosidase II Inhibitors
Ting-Jen R. Cheng,[a] Ting-Hao Chan,[a] En-Lun Tsou,[a] Shang-Yu Chang,[a]
Wen-Yi Yun,[a] Pei-Jung Yang,[a] Ying-Ta Wu,[a] and Wei-Chieh Cheng*[a, b]
Natural products are an important source of inspiration in
drug discovery, medicinal chemistry, and chemical biol-
ogy.[1,2] An ongoing challenge, however, is the efficient prep-
aration of natural product-based molecules that incorporate
a large number of stereogenic centers. Difficult stereoselec-
tive reactions and tedious separation processes are frequent-
ly required.[3]
ple hydroxy groups of pyrrolidine iminosugars and pyra-
nose-based glycans have yet to be comprehensively estab-
lished. Obviously, a systematic correlation study is limited
by difficulties associated with acquiring and modifying all
possible polyhydroxylated pyrrolidines. Indeed, two interest-
ing questions can be posed: 1) Can all sixteen stereoisomeric
ADMDP-based scaffolds with four stereogenic centers
(Figure 1) be prepared? and 2) How can we apply this
unique set of scaffolds as a powerful tool to efficiently
answer questions of biological relevance?
Many polyhydroxylated pyrrolidines and their derivatives,
which bear identical functional groups of different spatial
configurations, are known to exhibit a variety of different
biological properties and inhibitory activities against various
glycoprocessing enzymes, which are involved in diseases
such as diabetes and cancer, viral infections, and lysosomal
storage disorders.[4,5] Because of their broad-spectrum bio-
logical activities, this type of molecules can be considered
a privileged scaffold.[6] For example, synthetic ADMDP
(aminodeoxy-DMDP) with the configuration pattern of
(2R,3R,4R,5R), inspired from naturally occurring DMDP
(2,5-dideoxy-2,5-imino-d-mannitol) or DAB (1,4-dideoxy-
1,4-imino-d-arabinitol),[7,8] was successfully synthesized by
us and others, and its derivatives have been extensively ap-
plied as potent b-hexosaminidase and a-glucosidase inhibi-
tors.[9–11] The C-2 aminomethyl moiety of the pyrrolidine
ring can be considered a diversity position to which various
functional moieties or substituents can be conjugated, which
has the potential to improve inhibitory potency and selectiv-
ity.[12]
According to general mechanisms of enzymatic glycoside
hydrolysis and transglycosylation,[4] the pyrrolidine skeleton
might mimic the sp2-like hybridization of the glycosyl unit;
and the protonated N1-nitrogen can mimic the charge distri-
bution at the anomeric position. However, the correspond-
ing relationships between the stereo-configurations of multi-
Figure 1. Chemical structures of all sixteen ADMDP-based scaffolds.
Human Golgi a-mannosidase II (hGMII) is a key enzyme
in N-glycan processing, and its inhibition is a known anti-
cancer strategy.[13] Several naturally occurring or synthetic
inhibitors toward various GMIIs have been reported
(Figure 2).[14–17] For example, swainsonine is a potent GMII
inhibitor and also reduces certain tumors and hematological
dysfunctions, but its use is associated with side effects such
as co-inhibition of lysosomal a-mannosidases that limit its
clinical study.[13,17] DMJ (1,5-dideoxy-1,5-imino-d-mannitol)
is a common mannosidase inhibitor but is not a potent
dGMII inhibitor (Ki =400 mm).[14] Interestingly, recent re-
search shows several b-glucosidase inhibitors that can also
inhibit dGMII by binding similar locations with mannose-
[a] Dr. T.-J. R. Cheng, T.-H. Chan, E.-L. Tsou, S.-Y. Chang, W.-Y. Yun,
P.-J. Yang, Dr. Y.-T. Wu, Prof. W.-C. Cheng
Genomics Research Center, Academia Sinica
128 Academia Road, Sec. 2, Taipei 115 (Taiwan)
[b] Prof. W.-C. Cheng
Department of Chemistry, National Cheng-Kung University
1, University Road, Tainan (Taiwan)
Fax : (+886)227899931
E-mail: wcheng@gate.sinica.edu.tw
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/asia.201300680.
Chem. Asian J. 2013, 8, 2600 – 2604
2600
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Wei-Chieh Cheng et al.
Scheme 2. General strategy for the preparation of scaffolds 1a–p from
eight cyclic nitrones 2a–h.
Figure 2. Examples of reported inhibitors against Golgi a-mannosidase II
(dGMII=Drosophila GMII).
containing substrates.[14] Additionally, other molecules pos-
sessing distinct features including scaffolds and configura-
tions of hydroxy groups exhibit potent inhibitory activities
against mannosidases from jack bean or almonds.[18–21] Nev-
ertheless, systematic exploration of a whole series of novel
polyhydroxylated pyrrolidine-based scaffolds with regard to
the development of new and potent hGMII inhibitors has
yet to be reported.
Herein, we describe the preparation of all sixteen
ADMDP isomers as primary scaffolds, followed by efficient-
ly developing potent and selective hGMII inhibitors. The
general strategy is illustrated in Scheme 1.
Scheme 3. Typical synthetic method for the preparation of new poly-
hydroxylated pyrrolidines 1g and 1h from cyclic nitrone 2d. Reagents
and conditions: a) trimethylsilyl cyanide, MeOH, 508C, 10 h; b) 1. Raney
nickel, H2(g), Boc2O, MeOH, RT, 2 h; 2. Pd(OH)2, H2(g), MeOH, RT, 10 h;
3. TFA, DCM, 08C, 10 min; c) MsCl, Et3N, THF, 08C, 1 min; d) NaBH4,
THF/MeOH (4:1), 08C, 18 h.
gave the 2,3-trans isomer 3 as a major product (87%, d.r.ꢀ
20:1), which could be converted directly to 1g in good yield
(86%). On the other hand, the C-2 nitrile group in 3 was in-
verted through an elimination–reduction sequence to afford
5 (83% over two steps). Subsequently, global hydrogenolysis
of 5 gave 1h in 92% yield. Following the similar protocol
with a different cyclic nitrone, all desired scaffolds 1a–p
were obtained in a yield ranging 30–35% from the corre-
sponding cyclic nitrones 2a–h. Notably, 1g, 1j, 1m, 1n, and
1p are novel compounds.
With all sixteen scaffolds in hand, their biological evalua-
tion could commence. Recombinant hGMII was prepared
(see the Supporting Information) and used for inhibition
studies. At a concentration of 1 mm, only 1i showed inhibi-
tory activity against hGMII with a potency of more than
80% (Figure 3). This finding suggests that the (2R,3R,4S,5R)
configuration of pyrrolidine 1i is critical to inhibitory poten-
cy against hGMII. By contrast, its C2 epimer 1j was dramat-
ically less active (approximately 20% inhibition), thus indi-
cating that the orientation of the aminomethyl moiety signif-
icantly influences the inhibitory potency. Notably, though
the four scaffolds 1i, 1j, 1k, and 1l share the same 3,4-diol
configuration (3R,4S), the inhibition activity of 1i was more
potent than that of the other compounds. These observa-
tions emphasize that all four stereogenic centers play an im-
portant role in the hGMII inhibitory activity.
Scheme 1. Strategy for the development of hGMII inhibitors inspired
from naturally occurring pyrrolidine-based alkaloids.
Based on our synthetic design (Scheme 2 and Figure S1 in
the Supporting Information), all sixteen ADMDP-typed iso-
mers 1a–p can be prepared from the corresponding eight
five-membered chiral tri-O-benxyl cyclic nitrones 2a–h as
key intermediates (see the Supporting Information).[9] Each
cyclic nitrone can act as a precursor to two desired products
bearing the 2,3-cis and 2,3-trans configuration, dependent on
the chemical transformations used.[9] For example
(Scheme 3), the highly diastereoselective nucleophilic addi-
tion of TMSCN to cyclic nitrone 2d in methanol at 508C
Chem. Asian J. 2013, 8, 2600 – 2604
2601
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Wei-Chieh Cheng et al.
(estimate conversion).[10,12] The 48-membered primary li-
brary 7 was directly evaluated at 25 mm in an hGMII
enzyme-based assay without further purification (Scheme 4).
Excitingly, five compounds were found to display more than
60% inhibition in this primary assay. The screening results
indicated that the structure of the substitution moiety
strongly affects the inhibition potency (Figure 4). For exam-
Figure 3. The inhibition activity of sixteen scaffolds 1a–p at 1 mm against
hGMII.
Next, scaffold 1i was directly coupled with a randomly se-
lected 48-membered acid library 6 by prior activation of the
carboxylic acids in the presence of HBTU (1.5 equiv),
HOBt (1.5 equiv), and DIEA (3 equiv) in DMSO. After
24 hours, the reactions were analyzed by LC-MS. It was
found that amine 1i had been completely consumed and
a significant amount of desired products had been formed
Figure 4. The inhibition activity of library 7 and library 8 at 25 mm against
hGMII and chemical structures of selected potent inhibitors 9 and 10.
ple, when alkyl moieties were used, no significant inhibitory
activity was observed. By contrast, compounds containing
a fluorine-substituted benzyl moiety such as 7-Ec, 7-Cd, and
7-Dd showed high hGMII inhibitory activities. Notably, nei-
ther the acid library 6 nor the coupling reagents showed the
any significant inhibition. For comparison purposes, library
8, which was prepared from scaffold 1a (the C4-epimer of
1i), was also synthesized following the same procedure. As
expected, no potent inhibitor was found, again emphasizing
the importance of the 2R,3R,4S,5R stereo-configuration in
the scaffolds.
Two potent inhibitors 9 (=7-Ec) and 10 (=7-Cd) were re-
synthesized by coupling of acids 6-Ec and 6-Cd, respectively,
with scaffold 1i. Both 9 and 10 were confirmed to be potent
competitive inhibitors with an inhibition constant (Ki) of
24 nm and 31 nm, respectively (Figure 5 and Figure S4, Sup-
porting Information). Notably, the fluorine atom at the para
position of the benzene moiety should play an important
role in the inhibitory potency of 9 and 10.
We then created the three-dimensional (3D) structure of
human GMII (hGMII) by using homology modeling of the
crystallized structure of Drosophila GMII (PDB code
3DDF).[22] The hydroxy groups at C3 and C4 of the five-
membered ring coordinate the zinc ion together with
Asp289, Asp177, His175, and His569, and also form hydro-
gen bonds with Tyr354, Asp426, and Asp570 (Figure 6). The
hydroxy group at C6 forms hydrogen bonds with Asp570
and Tyr808. The protonated nitrogen on the pyrrolidine ring
has strong electrostatic interactions with Asp289. Notably,
the moiety attached at the C2 position shows significant in-
teractions. For example, the aromatic moiety could dock in
the pocket formed by Tyr352, Arg313, Tyr354, and Try496
through aromatic interactions and cation–p interactions.
Our model shows that bulky substituted groups on the aro-
Scheme 4. Parallel synthesis of the primary libraries 7 and 8, followed by
in situ enzyme-based inhibition evaluation against hGMII.
Chem. Asian J. 2013, 8, 2600 – 2604
2602
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Wei-Chieh Cheng et al.
depend on the configuration at the C2 position. Overall, the
modeling plausibly and completely explains the observations
made in our inhibition study.
The inhibitory activity of compounds 9 and 10 towards
a panel of glycosidases,[9,10] including a-glucosidases, b-glu-
cosidase, a-galactosidase, a-mannosidase, and a-rhamnosi-
dase (Table 1), was also examined. Interestingly, although
Table 1. Inhibititory activity of 9 and 10 against various glycosidases.
Enzyme
IC50 [mm]
9
10
acid a-glucosidase
a-glucosidase[a]
b-glucosylceramidase
b-glucosidase[b]
a-galactosidase[c]
amyloglucosidase[d]
a-rhamnosidase[e]
a-mannosidase[f]
hGMII
n.i.
9
54
93
20
n.i.
>200
29
n.i.[g]
88
11
>200
17
n.i.
>200
>200
0.5
0.3
A
ACHTUNGTERN(NUNG Ki=31 nm)
[a] From Bacillus stearothermophilus. [b] From almonds. [c] From green
coffee beans. [d] From Aspergillus niger. [e] Naringinase from Penicillium
decumbens. [f] From Jack bean. [g] n.i., no inhibition at 250 mm.
Figure 5. a) Chemical structures of potent hGMII inhibitors 9 and 10 and
b) Lineweaver–Burk plot of inhibitor 9.
matic ring could decrease the overall binding affinity due to
steric restrictions in the described pocket. By contrast, the
fluorine atom on the aromatic ring would increase the inter-
action by forming weak hydrogen bonds with Tyr496 and
multipolar interaction with the backbone carboxylate of
Tyr352. It can be seen that all interactions of the moiety
both 9 and 10 were potent inhibitors against hGMII, they
displayed only moderate to weak inhibitory activities against
other glycosidases including a-mannosidase from Jack bean.
This result suggests that both 9 and 10 could be highly selec-
tive and potent inhibitors against hGMII.
In summary, the systematic
preparation of all possible six-
teen enantiopure ADMDP-
based scaffolds has been effi-
ciently achieved from eight
five-membered cyclic nitrones
through general and flexible
transformations. Each scaffold
possesses four stereocenters
and
the
exo-aminomethyl
group, which allows rapid sub-
stituent diversity to be accom-
plished by conjugation with
structurally diverse acids. To ex-
emplify the biological utility of
these scaffolds, two novel
hGMII inhibitors were devel-
oped through our concise
method. Because these mole-
cules are easily prepared and
modified, they are suitable as
initial molecules for further in-
vestigation of the inhibition se-
lectivity towards other manno-
sidases including lysosomal and
endoplasmic reticulum (ER)
Figure 6. A computational modelling of the complex between hGMII (blue) and inhibitor 9 (green).
Chem. Asian J. 2013, 8, 2600 – 2604
2603
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemasianj.org
Wei-Chieh Cheng et al.
T. R. Cheng, W. C. Cheng, Bioorg. Med. Chem. 2008, 16, 10198–
10204.
mannosidases in the future. It is hoped that this versatile set
of scaffolds as well as the practical development strategy
used will become a powerful chemical tool for the explora-
tion of other interesting or disease-associated glycosidases
or glycotransferase inhibitors.
[10] a) H. W. Shih, K. T. Chen, S. K. Chen, C. Y. Huang, T. J. R. Cheng,
C. Ma, C. H. Wong, W. C. Cheng, Org. Biomol. Chem. 2010, 8,
2586–2593; b) P. H. Liang, W. C. Cheng, Y. L. Lee, H. P. Yu, Y. T.
Wu, Y. L. Lin, C. H. Wong, ChemBioChem 2006, 7, 165–173.
[11] T. M. Wrodnigg, F. Diness, C. Gruber, H. Hꢅusler, I. Lundt, K.
Rupitz, A. J. Steiner, A. E. Stꢄtz, C. A. Tarling, S. G. Withers, H.
Wçlfler, Bioorg. Med. Chem. 2004, 12, 3485–3495.
[12] a) C. Y. Wu, C. F. Chang, J. S. Y. Chen, C. H. Wong, C. H. Lin,
Angew. Chem. 2003, 115, 4809–4812; Angew. Chem. Int. Ed. 2003,
42, 4661–4664; b) A. Brik, Y. C. Lin, J. Elder, C. H. Wong, Chem.
Biol. 2002, 9, 891–896.
Acknowledgements
This work was supported by the National Science Council and Academic
Sinica.
[13] a) D. R. Rose, Curr. Opin. Struct. Biol. 2012, 22, 558–562; b) J. M.
van den Elsen, D. A. Kuntz, D. R. Rose, EMBO J. 2001, 20, 3008–
3017.
[14] D. A. Kuntz, C. A. Tarling, S. G. Withers, D. R. Rose, Biochemistry
2008, 47, 10058–10068, and references therein.
Keywords: alkaloids · combinatorial chemistry · natural
products · pyrrolidine-based scaffold · synthetic inhibitors
[15] a) D. A. Kuntz, A. Ghavami, B. D. Johnston, B. M. Pinto, D. R.
Rose, Tetrahedron: Asymmetry 2005, 16, 25–32; b) H. Fiaux, D. A.
Kuntz, D. Hoffman, R. C. Janzer, S. G. Lemaire, D. R. Rose, L. J.
Jeanneret, Bioorg. Med. Chem. 2008, 16, 7337–7346, and references
therein.
[16] D. A. Kuntz, W. Zhong, J. Guo, D. R. Rose, G. J. Boons, ChemBio-
Chem 2009, 10, 268–277.
[17] M. Venkatesan, D. A. Kuntz, D. R. Rose, Protein Sci. 2009, 18,
2242–2251.
[18] a) H. Fiaux, F. Popowycz, S. Favre, C. Schꢄtz, P. Vogel, S. G. Lem-
aire, L. J. Jeanneret, J. Med. Chem. 2005, 48, 4237–4246; b) F. Popo-
wycz, S. G. Lemaire, E. R. Garcꢁa, C. Schꢄtz, P. Vogel, Helv. Chim.
Acta 2003, 86, 1914–1923; c) F. Popowycz, S. G. Lemaire, R. De-
mange, E. R. Garcꢁa, A. T. Asenjo, I. Robina, P. Vogel, Bioorg.
Med. Chem. Lett. 2001, 11, 2489–2493.
[19] C. Bello, M. Cea, G. Dal Bello, A. Garuti, I. Rocco, G. Cirmena, E.
Moran, A. Nahimana, M. A. Duchosal, F. Fruscione, P. Pronzato, F.
Grossi, F. Patrone, A. Ballestrero, M. Dupuis, B. Sordat, A. Necioni,
P. Vogel, Bioorg. Med. Chem. 2010, 18, 3320–3334.
[20] R. Łysek, C. Schꢄtz, P. Vogel, Helv. Chim. Acta 2005, 88, 2788–
2811.
[21] C. Saotome, C. H. Wong, O. Kanie, Chem. Biol. 2001, 8, 1061–1070.
[22] P. Englebienne, H. Fiaux, D. A. Kuntz, C. R. Corbeil, S. G. Lemaire,
D. R. Rose, N. Moitessier, Proteins 2007, 69, 160–176.
[1] a) A. L. Harvey, Drug Discovery Today 2008, 13, 894–901; b) F. E.
Koehn, G. T. Carter, Nat. Rev. Drug Discovery 2005, 4, 206–220.
[2] a) S. Wetzel, K. Klein, S. Renner, D. Rauh, T. I. Oprea, P. Mutzel,
H. Waldmann, Nat. Chem. Biol. 2009, 5, 581–583; b) J. Y. Ortho-
land, A. Ganesan, Curr. Opin. Chem. Biol. 2004, 8, 271–280; c) K.
Grabowski, K. H. Baringhaus, G. Schneider, Nat. Prod. Rep. 2008,
25, 892–904; d) D. Morton, S. Leach, C. Cordier, S. Warriner, A.
Nelson, Angew. Chem. 2009, 121, 110–115; Angew. Chem. Int. Ed.
2009, 48, 104–109; e) M. Dꢁaz-Gavilꢂn, W. R. Galloway, K. M.
OꢃConnell, J. T. Hodkingson, D. R. Spring, Chem. Commun. 2010,
46, 776–778.
[3] E. Comer, H. Liu, A. Joliton, A. Clabaut, C. Johnson, L. B. Akella,
L. A. Marcaurelle, Proc. Natl. Acad. Sci. USA 2011, 108, 6751–6756.
[4] A. E. Stꢄtz, Iminosugars as glycosidase inhibitors: Nojirimycin and
beyond, Wiley-VCH, Weinheim, 1999.
[5] W. Zou, Curr. Top. Med. Chem. 2005, 5, 1363–1391.
[6] a) W. Meutermans, G. T. Le, B. Becker, ChemMedChem 2006, 1,
1164–1194; b) M. E. Welsch, S. A. Snyder, B. R. Stockwell, Curr.
Opin. Chem. Biol. 2010, 14, 347–361; c) Y. F. Chang, C. W. Guo,
T. H. Chan, Y. W. Pan, E. L. Tsou, W. C. Cheng, Mol. Diversity
2011, 15, 203–214.
[7] T. M. Wrodnigg, Monatsh. Chem. 2002, 133, 393–426.
[8] G. Brahmachari, Bioactive natural products: Opportunities and Chal-
lenges in Medicinal Chemistry, World Scientific Publishing Co., Sin-
gapore, 2011.<.
Received: May 18, 2013
Published online: August 13, 2013
[9] a) E. L. Tsou, Y. T. Yeh, P. H. Liang, W. C. Cheng, Tetrahedron
2009, 65, 93–100; b) E. L. Tsou, S. Y. Chen, M. H. Yang, S. C. Wang,
Chem. Asian J. 2013, 8, 2600 – 2604
2604
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim