A fluorescence study of isofagomine protonation in β-glucosidase

Article abstract of DOI:10.1039/c5ob00624d

N-(10-Chloro-9-anthracenemethyl)isofagomine 5 and N-(10-chloro-9-anthracenemethyl)-1-deoxynojirimycin 6 were prepared, and their inhibition of almond β-glucosidase was measured. The isofagomine derivative 5 was found to be a potent inhibitor, while the 1-deoxynojirimycin derivative 6 displayed no inhibition at the concentrations investigated. Fluorescence spectroscopy of 5 with almond β-glucosidase at different pH values showed that the inhibitor nitrogen is not protonated when bound to the enzyme. Analysis of pH inhibition data confirmed that 5 binds as the amine to the enzyme's unprotonated dicarboxylate form. This is a radically different binding mode than has been observed with isofagomine and other iminosugars in the literature.

Full text of DOI:10.1039/c5ob00624d

View Article Online
View Journal
Organic &
Biomol ec ular
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: E. Lindbäck, B. W.
Laursen, J. N. Poulsen, K. Kilså, C. M. Pedersen and M. Bols, Org. Biomol. Chem., 2015, DOI:
1
0.1039/C5OB00624D.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/obc

View Article Online
DOI: 10.1039/C5OB00624D
Page 1 of 5
Organic & Biomolecular Chemistry
Journal Name
RSCPublishing
ARTICLE
A FLUORESCENCE STUDY OF ISOFAGOMINE
PROTONATION IN -GLUCOSIDASE
Cite this: DOI: 10.1039/x0xx00000x
Emil Lindbäck, Bo Wegge Laursen, Jens Christian Navarro Poulsen,
†
Kristine Kilså, Christian Marcus Pedersen and Mikael Bols*
Received 00th January 2012,
Accepted 00th January 2012
N-(10-chloro-9-anthracenemethyl)-isofagomine
deoxynojirimycin were prepared, and their inhibition of almond
The isofagomine derivative was found to be a potent inhibitor, while the 1-deoxynojirimycin
derivative displayed no inhibition at the concentrations investigated. Fluorescence
spectroscopy of with almond -glucosidase at different pH values showed that the inhibitor
nitrogen is not protonated when bound to the enzyme. Analysis of pH-inhibition data confirmed
that binds as the amine to the enzyme’s unprotonated dicarboxylate form. This is a radically
5
and N-(10-chloro-9-anthracenemethyl)-1-
6

-glucosidase was measured.
DOI: 10.1039/x0xx00000x
5
www.rsc.org/
6
5

5
different binding mode than has been observed with isofagomine and other iminosugars in the
literature.
protonation relative to pH because the inhibition of
glycosidases by these compounds is heavily pH dependant.
Introduction
4,9
Sugar mimics containing a nitrogen atom in place of the ring-
oxygen atom or anomeric carbon are some of the most potent
1
inhibitors of glycosidases. Many of these molecules are natural
products most commonly being secondary metabolites in
2
plants, which suggest that they play a role protecting the plant
from carbohydrate degrading pests. Examples of such
molecules are 1-deoxynojirimycin (
mulberries and obviously mimics -glucose in the pyranose
form and pyrrolidine from tropical beans, which mimics
1), which is present in
D
2
D-
3
fructose in the furanose form (Figure 1). The nitrogen atom is
very important for the inhibitory activity as the equivalent O-
analogue is many orders of magnitude less potent. The
4
iminosugar does not necessarily have to be a strong base to be a
glycosidase inhibitor, as pK ’s in the range from 3 to 10 are
a
known among these compounds – yet basicity is normally a
5
prerequisite for inhibition. It is therefore obviously important
to understand the function of the amine and especially whether
it is actually the conjugate acid that is the inhibitor. Inhibition-
pH data clearly show that iminosugars bind to -glucosidase
with one proton but whether that proton is bound at nitrogen or
6
at one of the active site carboxylates is unknown. In one case
crystallography has shown that the inhibitor 4-O-
glycopyranosyl isofagomine
3
(Figure 1) binds to cellulase as
the conjugate acid – yet it could be argued that is a rather
strong base (pK 8.4) making protonation particularly facile.
7
3
a
Solid state NMR studies of labelled versions of azafagomine
4
Figure 1. Iminosugar inhibitors
8
(
pK 5.3) led to the conclusion that
4
bound to -glucosidase as
a
the conjugate acid and to isomaltase and glycoamylase as the
base.
Photo-induced electrontransfer has been used as a principle in
fluorescent pH indicators where an aminogroup attached to a
fluorescent chloroanthracene quench the emission; however
9
These studies suggest that iminosugar glycosidase
inhibitors may or may not be protonated in the active site
depending on the inhibitor’s base-strength and the enzyme.
However it could be interesting to observe the inhibitor
11
when it is protonated it does not. The idea behind this study
This journal is © The Royal Society of Chemistry 2013
J. Name., 2013, 00, 1-3 | 1

View Article Online
DOI: 10.1039/C5OB00624D
Organic & Biomolecular Chemistry
Page 2 of 5
ARTICLE
Journal Name
was to use this principle to study iminosugar protonation excitation at 358 nm at 5 different pH values of A) -
whenbound to the enzymes. We have prepared compounds
, C) a blank solution
and (Figure 1), iminosugar analogues of the above mentioned and D)
alone. The substraction spectra B-A and D-C, which
5
glucosidase alone, B) -glucosidase and
5
6
5
pH indicators, and studied their inhibition of -glucosidase.
represents the net effect of the inhibitor 5 with or without the
enzyme present, are shown in Figure 3. The enzyme used was
chromatographically purified -glucosidase that essentially
showed one protein band in the SDS-page and its concentration
Results
Compounds
inhibitors isofagomine
respectively (Figure 1). Isofagomine (
5
and
6
were prepared from the corresponding was 40 M on assumption it was completely active.
and 1-deoxynojirimycin ), Concentration of
was 0.8 M. This means that, with the Ki
) was prepared as values given in Table 1, most of the inhibitor is binding the
was obtained as the tetra-O- enzyme active site even if 75% of the enzyme is an inactive
(
7
)
(
1
5
7
1
2
previously described, while
1
1
3
benzyl derivative according literature methods and then form (denatured).
debenzylated by hydrogenolysis with Pd/C catalyst in dilute
60
50
40
30
20
10
0
pH = 5.0
D-C
B-A
1
00
80
60
40
20
0
pKa 5.1
data
400
450
500
550
600
2
4
6
8
pH

(nm)
Figure 2. Fluorescence titration of 5. Shown is the percentage
pH = 5.5
D-C
B-A
of fluorescence emission as function of pH.
40
HCl solution. Attachment of the 10-chloroantracene group was
performed by reductive amination of the amines
0-chloro-9-anthralaldehyde using sodium cyanoborohydride.
As predicted antracenyl compounds and displayed pH
1 and 7 with
1
5
6
dependant fluorescence with a large fluorescence emission
when protonated, and essentially no fluorescence in basic
20
solution. The pK of
5
and
6 was measured by fluorescence
a
titration (Figure 2), which gave a value 5.1 for both
1
0
compounds. This may seem low compared to
1 (pK 6.7) and
a
8
0
7
(pK 8.4) , however it is perfectly in line with previous
a
400
450
500
550
600
chloroantracenemethyl derivatives, which also were 2-3 orders
1
1
(nm)
of magnitude less basic than the parent amine. The reason is
possibly partially electronic effects and steric hindrance
towards protonation.
20
15
10
5
0
pH 6.06
D-C
B-A
pH
5.00 5.85 6.80 6.85 7.58 8.20
K (M) 11.3 1.35 1.00 1.78 0.64 0.49
i
Table 1. Inhibition constants for the inhibition of -glucosidase
by
5 at different pH values. K values were measure in
i
phosphate buffer containing 10% DMSO to ensure complete
solubility of
5
.
In a nitrophenol assay the compounds were tested for
inhibition of almond -glucosidase. The isofagomine derivative
5
was found to be a competitive inhibitor with K values
i
between 0.49 and 11 M in the pH range 5 to 8.2 (Table 1),
while the 1-deoxynojirimycin analogue on other hand did not
display appreciable competitive inhibition at a concentration of
00 M and was therefore not further studied.
Fluorescence of upon binding to almond -glucosidase
was studied by measuring fluorescence emission upon
400
450
500
550
600
6

(nm)
2
5
2
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 2012

View Article Online
DOI: 10.1039/C5OB00624D
Page 3 of 5
Journal Name
Organic & Biomolecular Chemistry
ARTICLE
as we see the K tends to rise at low pH to a value of 11.3 M at
i
1
0
8
6
4
2
0
pH 5.
pH 6.8
D-C
Inhibitors such as
curve when 1/K is plotted versus pH, which means that
inhibition is highest at a certain pH which depend on the
inhibitor’s and the enzyme’s pK
appears widespread if not the norm among iminosugar
inhibitors and detailed analysis of the data show that the
inhibitor and glycosidase in essentially all cases binds as an
EHI complex i.e. with just one proton. The K data of Figure 1
1 and 7 are known to give a bell-shaped
B-A
i
4,14
values.
This behaviour
a
1
4
6
i
does however not appear to conform to this norm. A plot of
1/K versus pH (Figure 4) shows that the data does not conform
i
4
00
450
500
550
600
to the curve one would expect from an inhibitor of pK 5.1
a

(nm)
binding to -glucosidase as the EHI-complex. However
surprisingly the data fit well with the inhibitor binding as an EI-
complex i.e. without any protons. In other words this means
that binding follow the deionisation of the least acidic
carboxylate in the enzyme (pK 6.7) and that unprotonated
binds exclusively to dicarboxylate form of the enzyme.
pH = 7.5
4
3
2
1
0
D-C
B-A
5
a
1
/K
i
-
1
(
M )
data
2
1
EH
binding
400
450
500
550
600

(nm)
E binding
2
7
Figure 3. Subtracted fluorescence spectra of
almond -glucosidase at 5 different pH. (from top to bottom
.0, 5.5, 6.1, 6.8, 7.5). Solid lines (D-C) and dashed lines (B-A)
represents the net effect of without or with enzyme present.
5 with and without
pH
5
Figure 4. 1/Ki versus pH for inhibition of almond pH
glucosidase by in phosphate buffer containing 10% DMSO.
The data is compared with predicted curves for inhibition if
5
5
All solutions contain 90% phosphate buffer and 10% DMSO.
5
binds to the enzyme with one proton (EH binding) or none (E
binding). It is seen that the E binding curve can fit the data,
while the EH curve cannot.
Discussion
The fluorescence spectra show that at pH 6.8 and 7.5 the
The pH independent binding constant is K (4) = [I][E]/[EI]
i
fluorescence of
5
is essentially unaffected by the presence of = 0.45 M. This is in accordance with the fluorescence data
binds to the
these pH (Table 1) most if not all of the inhibitor must be enzyme and none would be seen from an EI complex.
bound to the active site but since there is no fluorescence the
These results show that in compound , unlike its parent
nitrogen of
is not protonated or the fluorescence is quenched isofagomine and other iminosugars, the nitrogen atom has no
by another functional group in the enzyme. At pH values 6.06, importance for binding as it is can neither be involved in salt
.5 and 5.0 we see an increasing fluorescence from
in the bridge or hydrogen bonding interactions with the active site
the enzyme (Figure 3). Due to the low K of the inhibitor at discussed above as no fluorescence is seen when
5
i
5
5
5
5
presence (B-A) and absence of enzyme (D-C) with latter carboxylate groups (Figure 5). The attachment of the anthracyl
consistently being twice as high. This means that binding of the group has apparently made protonation of any of the basic
inhibitor to the enzyme decrease it fluorescence if not groups in the active site unfeasible presumably due to steric
completely then partially. The simplest interpretation of these reasons. Perhaps the negatively charged carboxylate groups
data is that
5
is not protonated in the active site of - repel each other and thus provide enough space for
5 to enter
glucosidase. The fact that significant fluorescence is seen in the the active site. Compound
5
is still a strong and competitive
enzyme containing samples (B-A, Figure 3) at low pH could inhibitor and this must be due to lipophilic and/or aromatic
simply be because of the presence of some unbound inhibitor, interactions of the anthracyl group as the nitrogen can play no
role and simple monosaccharides are poor inhibitors. To our
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1-3 | 3

View Article Online
DOI: 10.1039/C5OB00624D
Organic & Biomolecular Chemistry
Page 4 of 5
ARTICLE
Journal Name
knowledge compound
EI complex formation has been observed.
5 is the first glycosidase inhibitor where
N-(10-chloro-9-anthracenemethyl)
1-deoxynojirimycin
yellow crystals, H-NMR (500 MHz, d6-DMSO):  8.6 (d, 1H),
.5 (d, 1H,), 8.2 (dd, 1H), 7.9 (dd, 2H), 7.7 (m, 4H), 5.5 (m, 1H,
H-1’a), 5.4 (m, 1H, H-1’b), 3.8 (bd, 1H, H-6a), 3.7 (dd, 1H, H-
(6).
1
8
6
3
b), 3.6 (m, 1H, H-2), 3.4 (m, 5H, H-4, 4OH), 3.2 (t, 1H, H-3),
.1 (dd, 1H, H-1eq), 2.8 (bs, 1H, H-5), 2.7 (dd, 1H, H-1ax).
C-NMR (500 MHz, d6-DMSO): 132.9, 130.2 (2C), 127.8,
27.2, 126.7, 126.2 (8C), 125.5 (2C), 124.3 (2C), 76.7 (C-3),
7.7, 66.8 (C-2 & C-4), 60.1 (C-5), 57.6 (C-6), 55.2 (C-1’),
1
3
1
6
4
3
+
6.3 (C-1) ppm. HRMS(ES) calc. for M+H 388.1310, found
88.1391.
Procedure for fluorescence measurements. The inhibitor
5 was
dissolved at 0.012 mg/mL in DMSO and diluted 4 times with
DMSO to create a stock solution of concentration 8.05 M. -
glucosidase from almonds was obtained from Sigma (G4511,
1
0-30 units/mg).
Figure 5. Depiction of the nonfluorescent form of
5 binding to
Four solutions were made:
A. 3 mg -glucosidase G4511 in 0.45 mL phosphate buffer and
dicarboxylate form of the -glucosidase active site. Binding is
presumably primarily caused by lipophilic interaction with the
aromatic system and hydrogen bonding to the 3 hydroxyl
groups.
0.05 mL DMSO. Protein-concentration was 40 M.
B. 3 mg -glucosidase G4511 in 0.45 mL phosphat buffer and
0
4
.05 mL inhibitor stock solution. Protein-concentration was
0 M and [5] was 0.8M
C. 0.45 mL phosphat buffer and 0.05 mL DMSO
D. 0.45 mL phosphat buffer, and 0.05 mL inhibitor stock
Conclusions
This work has shown that 10-chloro-9-anthracyl-isofagomine
solution. [5] was 0.8M
(
5
) is a competitive inhibitor of -glucosidase and that it binds
as the unprotonated amine form. This is perhaps not so
surprising as has a comparatively low basicity (pK = 5.1). It
is more remarkable that
dicarboxylate form of -glucosidase. This is in contrast to
isofagomine ( ) and other iminosugars that have so far been
Fluorescence spectra was taken of samples A-D irradiating with
light of wavelength 358 nm on a Perkin Elmer LS50
instrument. The subtracted spectra (B-A) and (D-C) are given
in Figure 3.
5
a
5
exclusively bind to the uncatalytic
7
6
,7,10,12
analysed.
The fact that a relatively simple alkylation of 7
Procedure for measuring -glucosidase inhibition. These
experiments were performed at 25 C in an aqueous phosphate buffer
changed its binding behaviour profoundly, suggest that similar
behaviour could be widespread among iminosugar analogues.
o
(0.1M) containing 10% DMSO to make sure the inhibitor was
completely dissolved. As substrate 4-nitrophenyl -
D
-
glucopyranoside was used, while the enzyme was almond -
glucosidase of the quality described above. In a multisample
spectrophotometer 5-8 reactions with variant substrate
concentration (1-20 mM) was simultaneous started by addition
of enzyme (0.3 nM) and the formation of 4-nitrophenol
followed by measuring absorbance at 400 nM in the first 2
minutes. This was performed with and without the presence of
Experimental
Procedure for synthesis of 5 or 6. A mixture of iminosugar (0.15
mmol) and 10-chloro-9-anthralaldehyde (36 mg, 0.15 mmol) in
methanol (15 ml) and CHCl (1 ml) was heated to reflux until a
homogeneous solution was formed. Then NaCNBH (18 mg,
3
3
0.29 mmol) was added. After 10 min HOAc (0.1 ml) was
the inhibitor in a concentration close to the expected K . From
i
o
added, and the mixture heated to about 60 C for 4 h and
furthermore stirred at room temperature for 24 h. The solution
was evaporated, and the residue chromatographed with CHCl3
until the lipophilic compounds were eluted (10-chloro-9-
anthralaldehyde and 10-chloro-9-anthracenemethanol, ~65%)
these data the Km could be determined with and without
inhibition and from those values K was calculated (using K =
i
i
[I]/(K ’/K - 1).
m
m
Analysis of inhibition-pH data. The method from reference 6
was used. 1/Ki have the following dependency of H
concentration (for details see ref 6):
and then changed to CHCl /MeOH (8:1) giving the product in
about 30% yield.
3
1
1
=
ꢄ
+1ꢃ
퐻
ꢁ
퐴퐸ꢂ∗ꢁ퐴퐸2∗ꢁ퐴퐼 ꢁ퐴퐸ꢂ∗ꢁ퐴퐸2 ꢁ퐴퐸ꢂ∗ꢁ퐴퐼 ꢁ퐴퐸ꢂ ꢁ퐴퐼
i(obs)
퐾 (1)∗ꢀ
+
+
+
+
N-(10-chloro-9-anthracenemethyl) isofagomine (5). yellow
퐾
푖
3
2
2
퐻
퐻
퐻
퐻
1
1
crystals, H-NMR (500 MHz, D O):  8.08 (d, 2H, J 8.8 Hz),
2
ꢄ
ꢁ
퐴퐸2∗ꢁ퐴퐼 ꢁ퐴퐸2 ꢁ퐴퐼
ꢁ퐴퐼
ꢁ퐴퐸ꢂ ꢁ퐴퐸ꢂ
퐻
퐾
푖
(ꢅ)∗(
+
+
+1+
퐻
+
)
)
7
3
2
.90 (d, 2H, J 8.8 Hz), 7.46 (m, 4H), 4.95 (s, 2H, H-1’), 3.33-
.71 (m, 6H, H-2eq, H-6eq, H-3, H-4, H-5’a, H-5’b), 3.07 (m,
H, H-2ax, H-6ax), 1.74 (m, 1H, H-5). C-NMR (500 MHz,
2
퐻
퐻
퐻
1
2
ꢄ
1
3
ꢁ퐴퐸2 ꢁ퐴퐸2
퐻
퐻
퐾
푖
(ꢆ)∗(
+
+1+
+
+
+
퐻
ꢁ퐴퐼
ꢁ퐴퐼 ꢁ퐴퐸ꢂ ꢁ퐴퐼∗ꢁ퐴퐸ꢂ
1
퐻
d6-DMSO): 134.5 (1C), 132.0, 131.3 (4C), 127.8, 127.6, 125.6,
2
2
3
퐻
퐻
퐻
퐻
1
5
3
24.8 (8C), 121.6 (1C), 71.0, 68.0 (C-3 & C-4), 58.8 (C-5’), 퐾 (4)∗(1+ꢁ퐴퐼+
+
+
)
푖
ꢁ퐴퐸2 ꢁ퐴퐼∗ꢁ퐴퐸2 ꢁ퐴퐸ꢂ∗ꢁ퐴퐸2 ꢁ퐴퐼∗ꢁ퐴퐸ꢂ∗ꢁ퐴퐸2
5.1, 54.0, 51.2 (C-1’, C-2 & C-6), 40.8 (C-5) ppm. UV (nm):
99 (3.93), 378 (3.96), 360 (3.76), 344 (3.46), 327 (3.14), 251
The EHI and EI binding curves shown in Figure 4 was made in
+
(
>4.32), 220 (4.13). HRMS(ES) calc. for M+H 372.1366,
-5.1
a spreadsheet by entering the relevant acid constants (
5
: 10
,
found 372.1364.
-4.4
-6.7

-glucosidase: 10
and 10 and reasonable values for the
4
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 2012

View Article Online
DOI: 10.1039/C5OB00624D
Page 5 of 5
Journal Name
Organic & Biomolecular Chemistry
ARTICLE
-
7
constants K (1-4). For the EHI binding curve K (3) = 3 x 10
i
i
while K (1,2 and 4) was set to 1. For the EI binding curve K (4)
i
i
-
7
=
4.5 x 10 while K (1-3) was set to 1.
i
Acknowledgements
We thank Birgitta Kegel for technical assistance and the
Lundbeck foundation for generous financial support.
Supporting information
NMR spectra of compounds
5 and 6 (6 pages) are available.
Notes and references
*
Department
Universitetsparken 5, DK-2100 Copenhagen Ø, Denmark. Tel:
4535320160. Email: bols@chem.ku.dk.
Faculty of Landscape Architecture, Horticulture, and Crop Production
of
Chemistry,
University
of
Copenhagen,
+
†
Science, Swedish University of Agricultural Sciences, SE-23053 Alnarp,
Sweden.
1
A. E. Stuetz and T. M. Wrodnigg, Carbohydr. Chem. 2013, 39
20-149.
R. Nash, Chemistry World 2004,
L. Fellows, New Scientist, 1989, 26 august, 45-48.
,
1
2
3
4
1, 42-44.
M. P. Dale, H. E. Ensley, K. Kern, K. A. R. Sastry, L. D. Byers
Biochemistry 1985, 24, 3530–3539.
5
6
7
A. E. Stüzt (Ed.), Iminosugars as glycosidase inhibitors:
Nojirimycin and beyond, John Wiley & sons, 1999.
Ó. López, F. –L. Qing, C. M. Pedersen, M. Bols Bioorg. Med.
Chem. 2013, 21, 4755–4761.
A. Varrot, C. A. Tarling, J. M. MacDonald, R. V. Stick, D. L.
Zechel, S. G. Withers, G. J. Davies, J. Am. Chem. Soc. 2003,
1
25, 7496–7497.
H. H. Jensen, L. Lyngbye, A. Jensen, M. Bols Chem. Eur. J.
002, , 1218-1226.
8
9
2
8
A. C. Sivertsen, M. Gasior, M. Bjerring, S. U. Hansen, O.
Lopez Lopez, N. C. Nielsen, and M. Bols, Eur. J. Org. Chem.
2
007, 1735–1742.
1
1
1
1
0
1
2
3
M. P. Dale, H. E. Ensley, K. Kern, K. A. R. Sastry, L. D.
Byers, Biochemistry 1985, 24, 3530-3539.
A. P. da Silva, R. A. D. D. Rupasinghe, Chem. Commun. 1985
669–1670.
,
1
T. M. Jespersen, M. Bols, M. R Sierks, T. Skrydstrup,
Tetrahedron 1994, 50, 13449-13460.
T. Wennekes, B. Lang, M. Leemann, G. A. van der Marel, E.
Smits, M. Weber, J. van Wiltenburg, M. Wolberg, J. M. F. G.
Aerts, H. S. Overkleeft, Org. Process Res. Dev. 2008, 12, 414-
4
23.
1
4
T. M. Gloster, P. Meloncelli, R. V. Stick, D. Zechel, A.
Vasella, G. J. Davies J. Am. Chem. Soc. 2007, 129, 2345-
2
354.
This journal is © The Royal Society of Chemistry 2012
J. Name., 2012, 00, 1-3 | 5