Selective inhibition of glycosidases by feedback prodrugs

Article abstract of DOI:10.1002/anie.200600808

(Figure Presented) In the loop: A novel concept for the selective inhibition of glycosidases has been developed whereby a glycosidase inhibitor is released by glycosidase-mediated hydrolysis of a prodrug, which subsequently inhibits the glycosidase that i

Full text of DOI:10.1002/anie.200600808

Angewandte
Chemie
Enzyme Inhibition
ition prodrugs described herein are composed of an inhibitor
that is modified by a linker that renders them inactive
(Scheme 1a). The linker, in turn, is attached to a glycoside
that can be hydrolyzed by a targeted glycosidase. The
DOI: 10.1002/anie.200600808
Selective Inhibition of Glycosidases by Feedback
Prodrugs**
Jun Guo, Jinkeng Asong, and Geert-Jan Boons*
Protein- and lipid-linked oligosaccharides at the surface of
eukaryotic cells are responsible for a wide range of biological
processes impacting health and disease.[1–4] Examples of such
processes include fertilization, embryogenesis, neuronal
development, hormone activities, and the proliferation of
cells and their organization into specific tissues. Remarkable
changes in cell-surface carbohydrates occur with tumor
progression, which appears to be intimately associated with
the dreaded state of metastasis.[5–7] Furthermore, carbohy-
drates of host cells are often employed by pathogens for cell
entry. Not surprisingly, compounds that can interfere in the
biosynthesis of oligosaccharides are regarded as attractive
leads for drug discovery for a wide range of diseases.[8–14] For
example, inhibitors of pancreatic a-amylase, such as arcar-
bose, voglibose, and miglitol, have been introduced for the
treatment of diabetes.[15] Furthermore, compounds such as
zamamivir and oseltamivir (tamiflu) have been developed as
inhibitors of neuramidase for the treatment of the flu.[16–20]
Despite these successes, it has been difficult to develop safe
and efficacious glycosidase inhibitors for the treatment of
many other diseases. A major problem of many natural and
synthetic glycosidase inhibitors is that they inhibit other
glycosidases, which may lead to serious side effects. More-
selective inhibitors may be obtained by carefully designed
structure—activity-relationship studies[21–29] in combination
with a better understanding of the mechanism of action of
glycosidases. This approach is, however, complicated by the
fact that of the approximate 100 mammalian glycosidases,
only a small number have been cloned and overexpressed.
We report herein a general approach for the design and
synthesis of more-selective glycosidase inhibitors based on
feedback inhibition. Feedback inhibition is a mechanism in
which a biosynthetic pathway regulates itself through inhib-
ition of the first committed step in the pathway by a down-
stream or final product.[30] Although this type of inhibition is
widely used in nature, it has not been exploited for the design
and synthesis of selective therapeutics. The feedback-inhib-
Scheme 1. a) General strategy for the inhibition of enzymes by a feedback
mechanism. b) A feedbackinhibitor for mannosidases.
resulting intermediate is designed in such a way that it will
undergo self-fragmentation to release the active inhibitor.
Subsequently, the inhibitor inhibits the enzyme that initiated
its release. The attraction of this approach is that the inhibitor
will only be released at the site of the enzyme. Selectivity is
achieved owing to the exquisite selectivity of glycosidases for
their substrates. Furthermore, the release of the inhibitor is
terminated when a sufficient quantity of the inhibitor is
released. The remaining pool of prodrug will, however, be
activated when the active enzyme reappears, thereby approx-
imating conditions of continuous infusion.
Compound 1 is a prototypic feedback inhibitor that is
designed to selectively target a-mannosidases. It is composed
of mannostatin A,[10] which is a potent inhibitor of a-
mannosidases and has garnered attention as a lead compound
for cancer-drug development (Scheme 1b). The amino group
of mannostatin A is modified by a 4-hydroxybenzyl carba-
mate, which in turn is linked to an a-mannoside. The acylation
of the amino group of mannostatin A renders the compound
inactive for inhibition of mannosidases.[31] However, an a-
mannosidase can hydrolyze the glycosidic linkage resulting in
the formation of intermediate 3, which will self-fragment[32] to
give mannostatin A (5). The active mannostatin A will then
inhibit the a-mannosidase that initiated its release, therefore
terminating the enzymatic reaction. On the other hand, the
enzymatic hydrolysis of compound 2 will release amino-
cyclopentitetrol 6, which is not an inhibitor of an a-
mannosidase,[33] and therefore this reaction is expected to
proceed to completion.
[*] Dr. J. Guo, J. Asong, Prof. G.-J. Boons
Complex Carbohydrate Research Center
University of Georgia
315 Riverbend Road, Athens, GA 30602 (USA)
Fax: (+1)706-542-4412
E-mail: gjboons@ccrc.uga.edu
[**] This research was supported by the Research Resource Center for
Biomedical Complex Carbohydrates (P41-RR-5351) and NCI of NIH
(5UO1A91295).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2006, 45, 5345 –5348
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5345

Communications
The synthesis of compounds 1 and 2 is summarized in
Scheme 2. Coupling of 1-bromo-2,3,4,6-tetra-O-acetyl-a-d-
mannose (7) with p-hydroxybenzaldehyde (8) by using
silver(I)oxide as the promoter in refluxing acetonitrile gave
Figure 1. Release of prodrugs 1 and 2 by human lysosomal mannosi-
dase. HLM (20 mU) was added to a solution of 2 (a) and 1 (b)
(80 nmol) in deuterated sodium acetate buffer solution (100 mm,
Scheme 2. Reagents and conditions: a) Ag2O, CH3CN, D, 1 h;
b) NaBH4, silica, CHCl3, iPrOH; c) p-NO2PhOCOCl, pyridine, DCM,
18 h; d) Et3N, DMF, 18 h; e) NH3 in MeOH; f) NH2NH2·H2O, MeOH,
08C, 5 h; g) Ac2O, pyridine. DCM=dichloromethane, DMF=N,N-
dimethylformamide.
*
^
pH 5.6, 0.2 mL). The enzymatic cleavage of 1 ( ) and 2 ( ) and the
&
formation of the intermediates 3 (&) and 4 ( ) and the final products
1
!
~
5 ( ) and 6 ( ), respectively, were monitored by H NMR spectrosco-
py (600 MHz).
9 in a modest yield. The aldehyde moiety of 9 was reduced
with NaBH4 to give 10, which was converted into activated
intermediate 11 by reaction with 4-nitrophenyl chlorofor-
mate. This compound was immediately reacted with 5[34] and
6[35] to afford compounds 12 and 13, respectively. The acetyl
esters of 12 and 13 were cleaved by treatment with ammonia
or hydrazine in methanol to give the targeted compounds 1
and 2, respectively.
The enzymatic cleavage of 1 and 2 by human lysosomal
mannosidase (HLM) was monitored by 1H NMR spectrosco-
py (600 MHz).[36] HLM (20 mU) was added to a solution of 1
or 2 (80 nmol), which were dissolved in a deuterated sodium
acetate buffer (100 mm, pH 5.6, 0.2 mL). The progress of the
reaction was monitored by the integration of key signals of
the starting material, the intermediate, and final product. As
can be seen in Figure 1a, HLM hydrolyzed 2 producing
intermediate 4, which self-fragmented to aminocyclopentite-
trol 6. Compound 6 is not an inhibitor of HLM and as a result
the enzymatic conversion progressed to completion. On the
other hand, virtually no hydrolysis was observed when 1 was
treated with HLM (Figure 1b). In this case, the released
mannostatin A (1) inhibits HLM and terminates further
hydrolysis of 1. Mannostatin A is a potent inhibitor of
retaining a-mannosidases (Ki = 90 nm, HLM)[37] and thus,
the release of low concentrations of the active compound is
sufficient to abolish enzyme activity. Indeed, the addition of 4-
methylumbelliferyl-a-d-mannopyranoside did not result in
the release of fluorescence methylumbelliferone, demonstrat-
ing that the mannosidase was inhibited by the released
mannostatin A.
Compounds 1 and 2 were subjected to a-glucosidase, a-
fucosidase, and a-galactosidase. As expected, these enzymes
did not initiate the release of the compounds 5 and 6,
demonstrating that they have exquisite selectivity for a-
mannosidases (data not shown).[41]
Next, the pH dependence of the self-fragmentation of
intermediates 3 and 4 was investigated. It is well known that
different organelles have different pH values. For example,
lysosomes are significantly more acidic than the Golgi
apparatus. Golgi mannosidase II has been identified as a
target for antimetastatic therapy,[8,38] whereas inhibition of
lysosomal mannosidase leads to the phenotype of lysosomal
storage disease. It was expected that intermediates 3 and 4
would fragment significantly faster in neutral as opposed to
acidic conditions, providing a basis for some selectivity for
Golgi mannosidase over the lysosomal mannosidase. Thus,
compound 2 (0.4 mm), in deuterated phosphate/citrate buffer
solutions (0.2:0.1m) of different pH values, was treated with a
large quantity (7 UnitsmLÀ1) of Jack Bean a-mannosidase
(JBM) to completely convert the prodrug into intermediate 4.
5346
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5345 –5348

Angewandte
Chemie
Subsequently, the decomposition of 4 was monitored by
1H NMR spectroscopy by integration of key signals of the
intermediate and final product. As can be seen in Table 1, the
decomposition of 4 is significantly faster at more-basic pH
values and elevated temperatures. These results can easily be
rationalized by assuming that at a higher pH value, a large
proportion of the intermediate is in the phenoxide form,
which decomposes significantly faster than the analogous
phenol.
cultured in the presence of different concentrations of 14
and 15 for two days. Subsequently, the presence of polylactos-
amine linked to proteins was determined by analyzing cell
lysates by Western blotting with a fluorescein-labeled PHA
lectin.[40] Interestingly, at a concentration of 1 mm of 14, a 50%
reduction in fluorescein labeling was observed, whereas
compound 15 did not display any inhibitory activity
(Figure 2).
Table 1: pH dependence of half-lives upon the self-fragmentation of 2.[a]
pH
4.0
4.5
5.0
5.5
6.0
6.5
378C (min)
258C (min)
480
2800
270
–
100
–
45
–
16
–
9
45
[a] Determined by incubating prodrug 2 (80 nmol) and JackBean
mannosidase (1.4 U) in deuterated 0.2m phosphate/0.1m citrate buffer
(0.2 mL) at 378C or 258C in a NMR tube and measuring 1H NMR spectra
over time on a 600 MHz spectrometer.
For many applications, compounds such as 1 need to be
cell permeable to inhibit intracellular glycosidases. Analysis
of fetal calf serum (FCS), which is often used for culturing
cells, revealed that it contains a significant amount of
mannosidase activity. Thus, these mannosidases can unmask
mannostatin A, which can then pass through the cell mem-
brane to inhibit intracellular mannosidases. It was anticipated
that O-acetylation of 1 and 2 to give 14 and 15, respectively,
would considerably improve the metabolic stability of the
prodrug. Furthermore, it has been shown that O-acetylation
of saccharides can facilitate cellular uptake.[39] Subsequently,
intracellular esterases can cleave the acetyl esters revealing
the saccharide, which can then interfere in the biosynthesis of
oligosaccharides. FCS exhibits only very low levels of esterase
activity therefore improving extracellular stability of acety-
lated carbohydrates. Thus, it was anticipated that compound
14 would be easily taken up by cells, and subsequent
deacetylation by intracellular esterases would reveal 1,
which can then be hydrolyzed by an a-manmosidase to
initiate the release of mannostatin A. To investigate whether
14 possesses intracellular activity, it was subjected to MDAY-
D2 cells and the disappearance of polylactosamine moni-
tored. It is known that the metastatic MDAY-D2 lymphoma
cell line overexpresses the enzyme N-acetyl glucosaminyl-
transferase V (GnTV).[8] This enzyme adds an additional b1-
6-linked N-acetylglucosamine moiety to Asn-linked oligosac-
charides, which are subsequently extended by a polylactos-
amine chain. This increased branching has been observed in
human breast, colon, and skin carcinomas and has been
correlated with invasive and metastatic potential. Inhibition
of the mannose-trimming enzyme, human Golgi a-mannosi-
dase II (HGMII), which acts late in the N-glycan processing
pathway, provides one route to blocking additional branching
of N-linked oligosaccharides.[8] In this case, the biosynthetic
precursor for GnTV is not formed and hence the polylactos-
amine chain cannot be appended. Thus, to investigate
whether the acetylated prodrugs can inhibit intracellular
Golgi mannosidase II, MDAY-D2 lymphoma cells were
Figure 2. Inhibition of polylactosamine formation by compounds 14
and 15. MDAY-D2 cells were cultured for 2 days in the absence
(control) or presence of different concentrations of compounds 14 ( )
&
~
or 15 ( ). a) Cellular proteins (10 mg) were subjected to SDS-PAGE
and polylactosamine was visualized by using a fluorescein-labeled L-
PHA lectin. b) Fluorescence bands were scanned by laser beam
densitometry, and the relative intensity depicted.
In conclusion, a feedback prodrug approach in which a
glycosidase inhibitor is released by the targeted glycosidase
offers an attractive approach for enhancing the selectivity of a
glycosidase inhibitor. Another advantage of this type of
prodrug is that only the amount of compound required for
complete inhibition of the targeted enzyme is released. It is to
be expected that chemical modification of a lead compound in
combination with a feedback prodrug formulation may make
it easier to develop glycosidase inhibitors as safe therapeutics.
Furthermore, efforts are underway to modify the linker
region of the feedback prodrug, to increase the catalytic
efficiency of enzymatic hydrolysis, and to increase the rate of
self-fragmentation. Also, the possibility of using feedback
prodrugs for other glycosidases, proteases, and kinases is
being explored.
Received: March 1, 2006
Revised: May 12, 2006
Published online: July 17, 2006
Keywords: glycosidases · inhibitors · mannosidases · prodrugs ·
.
selectivity
Angew. Chem. Int. Ed. 2006, 45, 5345 –5348
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5347

Communications
[34] S. Ogawa, Y. M. Yu, J.Chem.Soc.Chem.Commun.
[35] T. Suami, Y. Sakota, K. Tadano, S. Nishiyam, Bull.Chem.Soc.
Jpn. 1971, 44, 2222.
1991, 890.
[1] A. Varki, Glycobiology 1993, 3, 97.
[2] R. A. Dwek, Chem.Rev. 1996, 96, 683.
[3] T. Kolter, K. Sandhoff, Angew.Chem. 1999, 111, 1632; Angew.
Chem.Int.Ed. 1999, 38, 1532.
[36] J. C. Wilson, M. J. Kiefel, D. I. Angus, M. von Itzstein, Org.Lett.
1999, 1, 443.
[4] J. W. Dennis, M. Granovsky, C. E. Warren, BioEssays 1999, 21,
412.
[5] M. M. Fuster, J. D. Esko, Nat.Rev.Cancer 2005, 5, 526.
[6] D. H. Dube, C. R. Bertozzi, Nat.Rev.Drug Discovery 2005, 4,
477.
[7] J. W. Dennis, M. Granovsky, C. E. Warren, Biochim.Biophys.
Acta 1999, 1473, 21.
[8] P. E. Goss, M. A. Baker, J. P. Carver, J. W. Dennis, Clin.Cancer
Res. 1995, 1, 935.
[9] V. H. Lillelund, H. H. Jensen, X. F. Liang, M. Bols, Chem.Rev.
2002, 102, 515.
[10] A. Berecibar, C. Grandjean, A. Siriwardena, Chem.Rev. 1999,
99, 779.
[37] B. Li, S. P. Kawatkar, S. George, H. Strachan, R. J. Woods, A.
Siriwardena, K. W. Moremen, G. J. Boons, ChemBioChem 2004,
5, 1220.
[38] A. D. Elbein, FASEB J. 1991, 5, 3055.
[39] A. K. Sarkar, T. A. Fritz, W. H. Taylor, J. D. Esko, Proc.Natl.
Acad.Sci.USA 1995, 92, 3323.
[40] J. W. Dennis, K. Koch, S. Yousefi, I. Vanderelst, Cancer Res.
1990, 50, 1867.
[41] Treatment of compound 1 with cell lysate in the presence of the
mannosidase inhibitors swainsonine and kifunensine did not
lead to significant decomposition, therefore indicating that
proteases or other enzymes do not decompose the prodrug.
[11] N. Asano, R. J. Nash, R. J. Molyneux, G. W. J. Fleet, Tetrahe-
dron: Asymmetry 2000, 11, 1645.
[12] T. D. Heightman, A. T. Vasella, Angew.Chem. 1999, 111, 794;
Angew.Chem.Int.Ed. 1999, 38, 750.
[13] M. vonItzstein, R. J. Thomson, Curr.Med.Chem. 1997, 4, 185.
[14] N. Asano, Glycobiology 2003, 13, 93R.
[15] R. Perfetti, P. S. Barnett, R. Mathur, J. M. Egan, Diabetes/Metab.
Rev. 1998, 14, 207.
[16] M. Von Itzstein, W. Y. Wu, G. B. Kok, M. S. Pegg, J. C. Dyason,
B. Jin, T. V. Phan, M. L. Smythe, H. F. White, S. W. Oliver, P. M.
Colman, J. N. Varghese, D. M. Ryan, J. M. Woods, R. C. Bethell,
V. J. Hotham, J. M. Cameron, C. R. Penn, Nature 1993, 363, 418.
[17] M. vonItzstein, J. C. Dyason, S. W. Oliver, H. F. White, W. Y. Wu,
G. B. Kok, M. S. Pegg, J.Med.Chem. 1996, 39, 388.
[18] C. U. Kim, W. Lew, M. A. Williams, H. T. Liu, L. J. Zhang, S.
Swaminathan, N. Bischofberger, M. S. Chen, D. B. Mendel, C. Y.
Tai, W. G. Laver, R. C. Stevens, J.Am.Chem.Soc. 1997, 119, 681.
[19] W. X. Li, P. A. Escarpe, E. J. Eisenberg, K. C. Cundy, C. Sweet,
K. J. Jakeman, J. Merson, W. Lew, M. Williams, L. J. Zhang, C. U.
Kim, N. Bischofberger, M. S. Chen, D. B. Mendel, Antimicrob.
Agents Chemother. 1998, 42, 647.
[20] P. Chand, P. L. Kotian, A. Dehghani, Y. El-Kattan, T. H. Lin,
T. L. Hutchison, Y. S. Babu, S. Bantia, A. J. Elliott, J. A.
Montgomery, J.Med.Chem. 2001, 44, 4379.
[21] G. Godin, P. Compain, O. R. Martin, K. Ikeda, L. Yu, N. Asano,
Bioorg.Med.Chem.Lett. 2004, 14, 5991.
[22] F. M. Platt, G. R. Neises, R. A. Dwek, T. D. Butters, J.Biol.
Chem. 1994, 269, 8362.
[23] F. M. Platt, G. R. Neises, G. B. Karlsson, R. A. Dwek, T. D.
Butters, J.Biol.Chem. 1994, 269, 27108.
[24] H. S. Overkleeft, G. H. Renkema, J. Neele, P. Vianello, I. O.
Hung, A. Strijland, A. M. van der Burg, G. J. Koomen, U. K.
Pandit, J. Aerts, J.Biol.Chem. 1998, 273, 26522.
[25] C. Y. Wu, C. F. Chang, J. S. Y. Chen, C. H. Wong, C. H. Lin,
Angew.Chem. 2003, 115, 5119; Angew.Chem.Int.Ed. 2003, 42,
4661.
[26] L. L. Johnson, T. A. Houston, Tetrahedron Lett. 2002, 43, 8905.
[27] I. Robina, A. J. Moreno-Vargas, J. G. Fernandez-Bolanos, J.
Fuentes, R. Demange, P. Vogel, Bioorg.Med.Chem.Lett. 2001,
11, 2555.
[28] R. Lysek, C. Schutz, P. Vogel, Helv.Chim.Acta 2005, 88, 2788.
[29] T. D. Butters, R. A. Dwek, F. M. Platt, Curr.Top.Med.Chem.
2003, 3, 561.
[30] A. B. Pardee, G. P. V. Reddy, Gene 2003, 321, 17.
[31] S. B. King, B. Ganem, J.Am.Chem.Soc. 1994, 116, 562.
[32] F. M. H. de Groot, W. J. Loos, R. Koekkoek, L. W. A. van Ber-
kom, G. F. Busscher, A. E. Seelen, C. Albrecht, P. de Bruijn,
H. W. Scheeren, J.Org.Chem. 2001, 66, 8815.
[33] S. Ogawa, T. Morikawa, Bioorg.Med.Chem.Lett. 2000, 10, 1047.
5348
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5345 –5348