The glucosinolate-myrosinase system. New insights into enzyme-substrate interactions by use of simplified inhibitors

Article abstract of DOI:10.1039/b502990b

Myrosinase, a thioglucoside glucohydrolase, is the only enzyme able to hydrolyse glucosinolates, a unique family of molecules bearing an anomeric O-sulfated thiohydroximate function. Non-hydrolysable myrosinase inhibitors have been devised and studied for

Full text of DOI:10.1039/b502990b

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
The glucosinolate–myrosinase system. New insights into
enzyme–substrate interactions by use of simplified inhibitors
Aure´lie Bourderioux,a Myriam Lefoix,a David Gueyrard,a Arnaud Tatiboue¨t,*a
Sylvain Cottaz,b Steffi Arzt,c Wim P. Burmeisterd and Patrick Rollina
a Institut de Chimie Organique et Analytique (ICOA), UMR 6005, Universite´ d’Orle´ans, BP
6759, F-45067, Orle´ans Cedex 2, France. E-mail: arnaud.tatibouet@univ-orleans.fr
b Centre de Recherche sur les Macromole´cules Ve´ge´tales (CERMAV–CNRS), affiliated with
Universite´ Joseph Fourier, Grenoble, BP 53, F-38041, Grenoble Cedex 9, France
c ESRF, BP 220, F-38043, Grenoble Cedex, France
d Institut Universitaire de France, Laboratoire de Virologie Mole´culaire et Structurale, FRE
2854 CNRS-UJF and EMBL, Grenoble Outstation, B.P. 181, F-38042, Grenoble Cedex 9,
France
Received 28th February 2005, Accepted 1st April 2005
First published as an Advance Article on the web 14th April 2005
Myrosinase, a thioglucoside glucohydrolase, is the only enzyme able to hydrolyse glucosinolates, a unique family of
molecules bearing an anomeric O-sulfated thiohydroximate function. Non-hydrolysable myrosinase inhibitors have
been devised and studied for their biological interaction. Diverse modifications of the O-sulfate moiety did not result
in a significant inhibitory effect, whereas replacing the D-glucopyrano residue by its carba-analogue allowed
inhibition to take place. X-Ray experiments carried out after soaking allowed for the first time inclusion of a
non-hydrolysable inhibitor inside the enzymatic pocket. Structural tuning of the aglycon part in its pocket is being
used as a guide for the development of simplified and more potent inhibitors.
such activation appears necessary nor indeed possible for plant
myrosinase. The glycosyl intermediate formed in the process
Introduction
Glucosinolates 1 are naturally occurring thiosugars mainly
is hydrolysed by a water molecule perfectly positioned by a
found in the botanical order Brassicales. The structural frame-
glutamine residue (and not a glutamate residue as in classical
work of glucosinolates invariably results from a combination of
glycosidases).2 The interaction between enzyme and substrate
three parts: a b-D-glucopyranosyl unit, an unique O-sulfated
has been determined using various analogues of glucosinolates.
anomeric thiohydroximate function and a broad library of
These have been prepared with structural modifications on the
aglycons, whose structure varies in the vegetal kingdom de-
glycosidic moiety,2 on the aglycon chain3 or on the anionic
pending on the species (Fig. 1).1 Myrosinase (thioglucoside
site.4 Most of those modifications proved useful in clearly
glucohydrolase EC 3.2.3.1) is the only enzyme able to hy-
demonstrating the specificity of myrosinase towards the D-
drolyze those unusual thiosaccharidic compounds and one
glucopyranosyl moiety and the flexibility with regard to the
of the few enzymes able to hydrolyse a thioglycosidic bond.
Myrosinase and glucosinolates are stored in different parts of
aglycon moiety. So far, 2-fluoro-2-deoxy-glucotropaeolin 2a was
the only good inhibitor acting through the formation of a
the plant, but especially in seeds. Mixing of the enzyme and
covalent glucosyl–enzyme intermediate. It allowed studies of
the substrate induces glucosinolate hydrolysis with retention
the activity of myrosinase, notwithstanding a tendency to slow-
of the anomeric configuration. Myrosinase belongs to family
rate hydrolysis.2f This slow hydrolysis allowed the trapping of
1 of glycoside hydrolases, albeit an atypical member of this
the glucosyl–enzyme intermediate during structural analysis
family: whereas classical glycosidases activate the glycosidic
by X-ray crystallography. This allowed an understanding of
oxygen by a catalytic acid residue in the glycosylation step, no
the interaction of the D-glucosyl residue with the active site.
However, the fast hydrolysis of the aglycon part of 2-fluoro-2-
deoxy-glucotropaeolin is the reason why an X-ray analysis5 of
the myrosinase–substrate interaction was precluded.2b
In order to improve our knowledge of substrate conformation
and binding inside the active site, a non-hydrolyzable substrate
would be an invaluable tool. But mainly, an improved inhibitor
would be an important tool in the physiological study of
glucosinolate–myrosinase location and interaction in plants. We
chose glucotropaeolin (GTL) 2c as the starting point because
of its current use as an European Union official standard
for the analysis of glucosinolates.3b The first attempt with a
C-glucopyranosyl analogue 2b of GTL 2c3b was unsucessful
inasmuch as it did not interact with myrosinase. In this article we
present two novel approaches: one consists of the variation of the
anionic part of the aglycon, whereas the second, while keeping
the aglycon structure of GTL 2c, consists of the modification of
the carbohydrate moiety into a non-hydrolysable structure.
Fig. 1 Glucosinolates 1, glucotropaeolin 2c and modified analogues.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 7 2 – 1 8 7 9
1 8 7 2
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
©

View Article Online
Results and discussion
Synthesis of inhibitors. Modification of the thiohydroximate
function
We have explored the modification of the thiohydroximate
function and its importance in the interaction with the enzyme:
for example, the imino double-bond system was replaced either
by a flexible 2- or 3-carbon acyclic alkyl chain or by a
more constrained aryl moiety. Furthermore, we replaced the
sulfate group with a sulfonate group (Scheme 1). Two synthetic
pathways were needed to introduce such aglycon modifications
(see retrosynthetic plans in Scheme 1).
Fig. 2
gluco conformation and the thiohydroximate function intact
but preclude hydrolysis by myrosinase. Further exploration
oriented us towards a family of simplified structures, in which
the carbohydrate moiety itself would be replaced by acyclic alkyl
chains (from ethyl to x-hydroxybutyl).
The general approach to build the O-sulfated thiohydroximate
structures followed a two-step sequence (Scheme 3 and Table 1):
a) condensation on selected thiol derivatives of phenylacethy-
droximoyl chloride 13 prepared according to a previously
developed methodology;7 and b) sulfation with sulfur trioxide–
pyridine complex followed, without intermediate purification,
by potassium cation exchange using KHCO3. The functionalized
derivatives 14 required further steps: peracetylation to yield
intermediates 15 then selective deprotection of the thiohydrox-
imate function to form monoacetates 16. Final O-sulfation,
followed by in situ cation exchange with KHCO3 and MeOK-
catalysed transesterification led to sulfates 17.
Most of the starting thiols were commercially available, with
the exception of protected 1-thio-5a-carba-b-D,L-glucopyranose
21. Following Ogawa’s procedure, 2,3,4,6-tetra-O-benzoyl-5a-
carba-a-D,L-glucopyranose 18 was first prepared.8 Sulfur atom
introduction through a pseudo a to b epimerisation at the
pseudo-anomeric center was realised via a two-step sequence
involving formation of triflate 19, then nucleophilic displace-
ment with thiourea to provide the isothiouronium salt 20 with a
60% overall yield. Further hydrolysis yielded thiol 21, which was
condensed with the hydroximoyl chloride 13 as described above
to yield thiohydroximate 14f. After sulfation–cation exchange,
the benzoyl groups were removed through transesterification to
yield the required carba-glucosinolate 12 (Scheme 4).9
Scheme 1 Retrosynthetic scheme to sulfate analogues.
For most compounds, a three-step synthetic sequence was
implicated (Scheme 2). S-Alkylation of commercially avail-
able protected 1-thio-b-D-glucopyranose 3 using alkyl iodides
proceeded with moderate to excellent yields under standard
conditions6 and subsequent O-sulfatation and glucose depro-
tection yielded acyclic derivatives 7 and 8 in reasonable overall
yields. The sulfonate analogue 9 was obtained through a very
efficient two-step procedure starting from 3. The constrained
chain was introduced using o-hydroxythiophenol in a standard
Lewis acid-catalysed thioglycosylation on per-O-acetylated b-
D-glucopyranose, yielding intermediate 10, which was first O-
sulfated, then de-O-acylated to give compound 11 in high yield.
Inhibition studies
The activity of Sinapis alba myrosinase towards the different
glucosinolate analogues was determined by titration of re-
leased glucose from sinigrin. The simplest method for assaying
Synthesis of inhibitors. Modification of the carbohydrate subunit
We have previously published the synthesis of
a non-
hydrolyzable C-glucoside analogue 2b (C-GTL) (Fig. 2), which
did not inhibit myrosinase activity.3b This result showed the
importance of the anomeric sulfur atom in the thiohydroximate
function and prompted us to design and synthesize a range
of non-hydrolyzable substrate analogues of myrosinase, which
preserve the anomeric sulfur atom. We first prepared the
carba-glucosinolate 12, which would maintain both the D-
Table 1 Modified anionic subunit, inhibition determination
Substrate
7
8
9
11
Inhibition
30% at
10 mM
IC50 = 5 mM
15% at
10 mM
No inhibition
at 10 mM
Scheme 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 7 2 – 1 8 7 9
1 8 7 3

View Article Online
Scheme 3
Scheme 4
myrosinase is to record the decrease of the UV absorption of
sinigrin at 227.5 nm (DSA assay),10 but this was inappropriate
in our case since all the inhibitors absorbed strongly at the
above wavelength. We selected the glucose oxidase–peroxidase
coupled enzyme assay11 as an alternate method, which was found
to be more sensitive than the pH-stat assay11 and less costly
than the glucose deshydrogenase–hexokinase coupled enzyme
assay.12 The Km (Michaelis constant) of myrosinase with sinigrin
as substrate was found to be 0.3 mM, in agreement with the
reported value of 0.156 mM when using the DSA assay.10
The glucose oxidase–peroxidase system is known to undergo
inhibition by the myrosinase activator ascorbic acid,11,12 but
no inhibition was observed with the compounds tested. The
drawback of this method is that some compounds (7, 8, 9 and
11) may act as poor substrates and their hydrolysis will lead to
the production of glucose which will be detected in the assay.
The IC50 (molar concentration for 50% inhibition of my-
rosinase under the conditions described in the experimental
section) and percentage of inhibition values for the compounds
modified at the thiohydroximate are reported in Table 1. It is
of importance to note that the four compounds showing low
levels of inhibition did not act as substrates. Only the sulfate
derivative 8 shows an interesting IC50 value. Comparison of
sulfate 7 and sulfonate 9, which have the same chain length,
showed no significant differences. The rigidified structure of 11
showed no inhibition, the hydroxy-phenyl moiety seeming to
hamper the interaction with the enzyme probably due to the
displacement of the water molecule which interacts with Arg259
and the hydroximate nitrogen (Fig. 3a).
inhibitor than 17b, indicating a detrimental effect due to the
presence of the hydroxyl group; however, the higher homologues
3-hydroxypropyl and 4-hydroxybutyl derivatives 17d and 17e
respectively, appeared to be stronger inhibitors. We presume
that the presence of a hydroxyl group at an appropriate distance
from the sulfur atom may be involved in stabilizing hydrogen
bonds with the protein and that a longer aliphatic chain may
show less conformational flexibility due to an interaction with
the benzyl ring.
Crystallographic studies
The electron density (Fig. 3a) showed that the carba-
glucosinolate 12 bound in the active site of Sinapis alba my-
rosinase. Electron density for the benzyl ring, the sulfate group
and the glucopyranose ring are clearly visible. We observed the
biologically active D-enantiomer, which was selectively bound
from the D/L-mixture used in the experiment. Despite this, in
the crystal only about half of the active sites are occupied with
the inhibitor and the glucose ring is even more disordered. The
low occupancy of the inhibitor despite a concentration of 20 mM
used in the soak may be due to competition with the high sulfate
ion concentration present in the crystal and binding of glycerol
from the cryoprotectant.
Table 2 Modified glucose subunit, inhibition determination
Starting thiols RSH
Inhibition at 1 mM (%)
IC50/mM
a
We were pleased to observe that the non hydrolysable carba-
glucosinolate 12 (Table 2) shows an inhibition similar to earlier
values determined for classical glycosidase inhibitors.3c
The IC50 values for the ethyl derivative 17b and its hydroxy-
lated homologues 17c–17e are reported in Table 2. Surprisingly,
the ethyl derivative 17b acts as a better inhibitor than the carba-
glucosinolate 12. The 2-hydroxyethyl derivative 17c is a weaker
17a PhCH2SH
17b CH3CH2SH
17c HO(CH2)2SH
17d HO(CH2)3SH
17e HO(CH2)4SH
12
-
—
67
0.58 0.09
nd
0.44 0.02
0.25 0.01
0.99 0.01
23 (at 10 mM)
70
88
50
a Not determined due to solubility problems.
1 8 7 4
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 7 2 – 1 8 7 9

View Article Online
Fig. 3 a) Stereoview of the myrosinase active site. The refined structure of the carba-glucosinolate 12 (yellow) and the residues of the active site
interacting with the inhibitor (white) are shown. Residues involved in hydrogen bonds are labelled in grey, residues forming the hydrophobic pocket
binding the phenyl group are labelled in dark green and residues contributing to the hydrophobic surface next to the sulfate group are shown in
light green. For clarity, Gln187 at the back is not labelled. The averaged electron density (as described in the methods section) of a Finh–Fnat Fourier
synthesis contoured at 2.5 r around the inhibitor is shown in cyan. Hydrogen bonds are shown as dotted lines. b) The active site is represented by its
accessible surface coloured according to atom types, the carba-glucosinolate inhibitor 12 is shown in a stick representation (carbon atoms in green).
The position of the glucopyranosyl group in the 2-F glucosyl enzyme5 is shown in magenta.
The benzyl group is bound in a hydrophobic pocket formed
by residues Phe331, Phe371, Phe473, Ile382 and Tyr330 as
predicted.5 The sulfate group establishes hydrogen bonds to
Arg259(Ng2), and Ser190(Oc) as well as Gln187(Ne2), the
residue replacing the catalytic glutamic acid residues classically
present in glucosidases. The positive charge of Arg194 interacts
more weakly with the sulfate group. Apart from these polar
interactions, the group is located on top of a hydrophobic surface
formed by Tyr189, Phe282 and Ile 257. A water molecule bridges
the nitrogen atom of the hydroximate function to Arg259(Ne)
through hydrogen bonds.
boat conformation will then lead with a high commitment to
catalysis.
In this combined study of inhibition and crystal structure
using a carba-glucosinolate, we were, for the first time, able to
observe directly an enzyme–substrate analogue complex and the
interactions of the aglycon with the myrosinase active site. The
previous studies5,13 only exhibited structures of the glucosyl–
enzyme intermediate as well as myrosinase complexed with
several transition state analogues. Attempts to bind substrate
analogues had failed so far.13,14
There are only three interactions of the glucose group with
myrosinase, involving i) the hydroxyl group in position 2 and
Gln187 Oe1, ii) the 4-hydroxyl group and the indole nitrogen
of Trp457 and iii) the 6-hydroxyl and the main chain amide
of Phe465 (Fig. 3a). The weak interaction combined with a
lack of order of this group suggested that the glucose moiety
did not significantly contribute to the affinity of the inhibitor.
Furthermore, its position differs from the one of the 2-deoxy-
2-fluoroglucopyranosyl group in the covalent glucosyl enzyme
complex (Fig. 3b).5,13 These observations led us to design and
synthesize simplified inhibitors of the ethyl derivative series,
which, as expected, proved to be better inhibitors than the carba-
glucosinolate 12.
A crystal structure of the complex of myrosinase with the ethyl
derivative 17b showed similar electron density for the sulfate and
the benzyl ring as observed for 12, but only low electron density
for the ethyl group (data not shown). The interactions of sulfate
and benzyl groups of 17b with the enzyme correspond to those
observed for 12 (Fig. 3a).
Conclusions
Our inhibition studies suggest that the thiohydroximate function
is important for substrate recognition by myrosinase. The
anionic charge by itself was not sufficient to induce inhibition:
even a slight modification of the function dramatically reduces
binding. This is not surprising as all polar atoms of the
function interact directly or indirectly through hydrogen bonds
with the active site (Fig. 3a). A similar loss of binding was
observed using C-GTL 2b, in which the sulfur atom of the
thioglucosidic linkage is replaced by a methylene group.3b In
contrast, when the endo-cyclic oxygen of the carbohydrate
moiety is replaced by a methylene group, the ability to bind
to myrosinase (IC50 = 1 mM) is preserved, pointing the way
to drastic simplification of the structure to an ethyl group,
which even improved the inhibitory effect (IC50 = 0.6 mM).
This shows the unimportance of the glucosyl group during
the initial interaction with the active site. Increasing the chain
length (propyl and butyl) and adding a x-hydroxyl function
further improved inhibition up to an IC50 of 0.2 mM. These
inhibitors may be more conformationally constrained due to a
hydrophobic contact between the alkyl chain and the benzyl ring
and in addition be able to engage one or more hydrogen bonds
with the enzyme. The detailed knowledge on the interaction
The substrate analogue 12 binds to myrosinase in a confor-
mation corresponding probably to a state of prebinding as the
position of the glucosyl group differs from that observed for the
glucosyl–enzyme (Fig. 3a). A transient movement of the glucosyl
group to this position combined with a distortion towards a
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 7 2 – 1 8 7 9
1 8 7 5

View Article Online
of the sulfate group, the hydroximate function and the benzyl
aglycon with the myrosinase active site will facilitate further
design and improvement of myrosinase inhibitors.
General protocol for sequential sulfation/de-O-acylation
to 7, 8, 9
To a solution of 4 or 5 (1 eq.) in dichloromethane, pyridine–
SO3 complex (10 eq.) was added under dry conditions. The
solution was stirred at room temperature for 20 h, then an
aqueous solution of KHCO3 (15 eq.) was added and stirring
was maintained for 1 h. After removal of the solvents, the crude
residue was purified over silica gel. The resulting compounds
were de-O-acylated using catalytic potassium hydroxide in
methanol at room temperature. After overnight reaction, Dowex
H+ resin was added and after filtration and evaporation, the
crude material was purified over silica gel.
Experimental
Chemical synthesis
Melting points were determined on a Ko¨fler hot-stage apparatus
and are uncorrected. 1H and 13C NMR spectra were recorded on
a Bruker Avance DPX250 at 250 MHz and 62.89 MHz respec-
tively. The chemical shifts (d) are reported in ppm downfield from
TMS as internal standard or from the residual solvent peak.
Coupling constants (J) are reported in Hz. Specific rotations
were measured at 20 ◦C using a Perkin-Elmer polarimeter 141.
Low resolution mass spectra (MS) were recorded by the ICOA
Analytical Service on a Perkin-Elmer SCIEX API 300 (ion spray:
IS) and HRMS spectra were recorded on a VG analytical 70
SV. Elemental analysis was performed by the analytical service
of the CNRS, Vernaison. Evaporation was conducted in vacuo
with a Bu¨chi rotary evaporator. Analytical TLC was carried out
on precoated silica gel 60F-254 plates (E. Merck) and spots were
detected by UV light (254 nm) and by heat treatment with a 10 :
85 : 5 mixture of sulfuric acid, ethanol and water. Flash column
chromatography was performed on Kieselgel 60 (230–400 mesh)
silica gel (E. Merck).
(2-Sulfato)ethyl 1-thio-b-D-glucopyranoside 7. Obtained as a
gum in 52% yield from 4; [a]2D5 = −9 (c = 1.0 in H2O); 1H NMR
(D2O): 2.96 (m, 2H, H-1ꢀ), 3.22 (t, 1H, J2–3 = 8.7 Hz, H-2), 3.32–
3.44 (m, 3H, H-3, H-4, H-5), 3.62 (dd, 1H, J6b–5 = 4.9 Hz, H-6b),
3.81 (dd, 1H, J6a–6b = 12.4 Hz, J6a–5 = 1.5 Hz, H-6a), 4.15 (m,
2H, H-2ꢀ), 4.51 (d, 1H, J1–2 = 9.8 Hz, H-1); 13C NMR (D2O):
29.3 (C-1ꢀ), 61.2 (C-6), 68.7 (C-2ꢀ), 69.8 (C-4), 72.6 (C-2), 77.5
(C-3), 80.2 (C-5), 85.8 (C-1); MS (IS-): m/z: 319.0 [M− − K];
HRMS calcd for C8H15KO9S2: 357.9794. Found: 357.9798.
(3-Sulfato)propyl 1-thio-b-D-glucopyranoside 8. Obtained as
a gum in 57% yield from 5; [a]2D5 = −9 (c = 1.0 in H2O); H
1
NMR (D2O): 2.06 (m, 2H, H-2ꢀ), 2.88 (m, 2H, H-1), 3.34 (t, 1H,
J2–3 = 8.7 Hz, H-2), 3.43–3.56 (m, 3H, H-3, H-4, H-5), 3.73 (dd,
1H, J6b–5 = 5.3 Hz, H-6b), 3.92 (dd, 1H, J6a–6b = 12.4 Hz, J6a–5
=
1.7 Hz, H-6a), 4.19 (m, 2H, H-2ꢀ), 4.57 (d, 1H, J1–2 = 9.8 Hz,
H-1); 13C NMR (D2O): 26.6 (C-2ꢀ), 29.6 (C-1ꢀ), 61.2 (C-6), 67.9
(C-2ꢀ), 69.8 (C-4), 72.6 (C-2), 77.5 (C-3), 80.1 (C-5), 85.8 (C-1);
MS (IS-): m/z: 333.0 [M− − K]; HRMS calcd for C9H17KO9S2:
371.9951. Found: 371.9957.
General protocol for S-alkylation of protected
1-thio-b-D-glucopyranose 3 to compounds 4, 5 and 6
The thiol 3 (1 eq.) was dissolved in acetone (DMF in the case
of 6) and alkylated according to Cerny et al.6 using 1.1 eq. of
the iododerivative and 1 eq. of K2CO3. After completion, the
reaction medium was extracted with CH2Cl2, washed with water,
then brine and evaporated to a crude mixture which was purified
on silica gel.
(3-Sulfonato)propyl 1-thio-b-D-glucopyranoside 9. Obtained
as a gum in quantitative yield from 6; [a]2D5 = −7 (c = 1.0 in H2O);
1H NMR (D2O): 2.00 (m, 2H, H-2ꢀ), 2.78 (m, 2H, H-1ꢀ), 2.94
(m, 2H, H-3ꢀ), 3.21 (dd, 1H, J2–3 = 8.8 Hz, H-2), 3.33–3.47 (m,
3H, H-3, H-4, H-5), 3.60 (dd, 1H, J6b–5 = 4.8 Hz, H-6b), 3.79 (d,
(2-Hydroxy)ethyl
2,3,4,6-tetra-O-acetyl-1-thio-b-D-gluco-
1H, J6a–6b = 12.5 Hz, H-6a), 4.45 (d, 1H, J1–2 = 9.8 Hz, H-1); 13
C
pyranoside 4 was obtained with 94% yield according to the
literature.6
NMR (D2O): 25.3 (C-2ꢀ), 29.2 (C-1ꢀ), 49.9 (C-3ꢀ), 61.2(C-6), 69.8
(C-4), 72.6 (C-2), 77.5 (C-3), 80.2 (C-5), 85.8 (C-1); MS (IS-):
m/z: 317.0 [M− − K]; HRMS calcd for C9H17KO8S2: 356.0002.
Found: 355.9997
(3-Hydroxy)propyl 2,3,4,6-tetra-O-acetyl-1-thio-b-D-gluco-
pyrano◦side 5. Obtained with 60% yield as a white solid; mp:
1
75–77 C, [a]2D5 = −20 (c = 1.0 in CHCl3); H NMR (CDCl3):
(2-Hydroxy)phenyl 2,3,4,6-tetra-O-acetyl-1-thio-b-D-glucopy-
ranoside 10. 1,2,3,4,6-Penta-O-acetyl-b-D-glucopyranose (0.5 g,
1.28 mmol) in acetone (10 mL) was reacted with 2-
hydroxythiophenol (0.139 g, 1.34 mmol) in the presence of
BF3·Et2O (227 lL, 1.8 mmol) first at 0 ◦C, then overnight at room
temperature. After hydrolysis with ice and dichloromethane
extraction, the organic phases were collected, dried over MgSO4
and evaporated. The crude material was purified over silica◦gel
giving a 72% yield; [a]2D5 = −6 (c = 1.0 in CHCl3); mp: 124–126 C;
1H NMR (CDCl3): 1.95 (s, 3H, CH3CO), 1.98 (s, 3H, CH3CO),
2.07 (s, 3H, CH3CO), 2.10 (s, 3H, CH3CO), 3.70 (m, 1H, H-5),
4.14 (d, 2H, J6–5 = 3.9 Hz, H-6), 4.59 (d, 1H, J1–2 = 10.0 Hz,
H-1), 4.90 (t, 1H, J2–3 = 9.2 Hz, H-2), 4.99 (t, 1H, J4–5 = 9.8 Hz,
H-4), 5.20 (t, 1H, J3–4 = 10.0 Hz, H-3), 6.84 (td, 1H, J = 1.3 Hz,
J = 7.5 Hz, Har), 6.96 (dd, 1H, J = 1.3 Hz, J = 8.3 Hz, Har),
7.02 (s, 1H, OH), 7.26–7.38 (m, 2H, Har); 13C NMR (CDCl3):
20.6, 20.7, 20.8 (OAc), 61.7 (C-6), 67.9 (C-4), 69.7 (C-2), 73.7 (C-
3), 76.1 (C-5), 86.4(C-1), 114.3 (Car), 116.3, 120.7, 132.6, 137.3
(CHar), 160.6 (Car), 169.1, 169.3, 170.1, 170.6 (CO); MS (IS+):
m/z: 474 [M+ + NH4], 479 [M+ + Na]; elemental analysis calcd
for C20H24O10S:C, 52.63; H, 5.30. Found: C, 52.32; H, 5.27%.
1.81 (m, 2H, H-2ꢀ), 1.98 (s, 3H, CH3CO), 2.00 (s, 3H, CH3CO),
2.03 (s, 3H, CH3CO), 2.06 (s, 3H, CH3CO), 2.78 (m, 2H, H-1ꢀ),
3.70 (m, 3H, H-5, H-3ꢀ), 4.12 (dd, 1H, J6b–5 = 2.6 Hz, J6b–6a
=
12.4 Hz, H-6b), 4.21 (dd, 1H, J6a–5 = 4.7 Hz, H-6a), 4.47 (d,
1H, J1–2 = 10.0 Hz, H-1), 5.02 (t, 1H, J2–3 = 9.2 Hz, H-2), 5.04
(t, 1H, J4–5 = 10.0 Hz, H-4), 5.20 (t, 1H, J3–4 = 9.4 Hz, H-3);
13C NMR (CDCl3): 21.0, 21.1 (OAc), 26.6 (C-2ꢀ), 32.5 (C-1ꢀ),
60.9 (C-3ꢀ), 62.4 (C-6), 68.6 (C-4), 70.1 (C-2), 74.1 (C-3), 76.3
(C-5), 83.9 (C-1), 169.8, 169.9, 170.5, 171.1 (CO); MS (IS+):
m/z: 423,0 [M+ + H]; elemental analysis calcd for C17H26O10S:
C, 48.33; H, 6.20. Found: C, 48.13; H, 6.18%.
(3-Sulfonato)propyl 2,3,4,6-tetra-O-acetyl-1-thio-b-D-gluco-
pyranoside, potassium salt 6. Obtained quantitatively; mp =
236–238 ◦C, [a]D25 = +15 (c = 1.0 in MeOH), 1H NMR (DMSOd6):
1.81 (m, 2H, H-2ꢀ), 1.92, 1.96, 1.99, 2.01 (4s, 12H, 4 CH3CO),
2.52 (m, 2H, H-3ꢀ), 2.70 (m, 2H, H-1ꢀ), 3.96–4.01 (m, 2H, H-5
and H-6b), 4.12 (dd, 1H, J6a–5 = 5.8 Hz, J6a–6b = 12.6 Hz, H-6a),
4.77 (t, 1H, J2–3 = 10.0 Hz, H-2), 4.88 (t, 1H, J4–5 = 10.0 Hz, H-
4), 4.90 (d, 1H, J1–2 = 10.0 Hz, H-1), 5.23 (t, 1H, J3–4 = 9.4 Hz,
H-3); 13C NMR (DMSOd6): 21.1, 21.2, 21.3, 21.4 (OAc), 26.7
(C-2ꢀ), 30.0 (C-1ꢀ), 50.9 (C-3ꢀ), 62.8 (C-6), 69.0 (C-4), 70.7 (C-
2), 73.9 (C-3), 75.1 (C-5), 83.1(C-1), 170.0, 170.1, 170.4, 171.0
(CO), MS (IS-): m/z: 485 [M+ − H]; elemental analysis calcd for
C17H25O12S2K: C, 38.93; H, 4.81. Found: C, 38.67; H, 4.79%.
2ꢀ-Sulfatophenyl-1-thio-b-D-glucopyranoside 11. Pyridine·
SO3 complex (0.991 g, 6.2 mmol) was added to
a
dichloromethane solution of the phenolic compound 10
(284 mg, 0.62 mmol). The resulting mixture was heated for
1 h, then, after cooling to room temperature, quenched with
1 8 7 6
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 7 2 – 1 8 7 9

View Article Online
aqueous KHCO3 (1.3 g, 13 mmol) for 1 h. The resulting slurry
was suspended in MeOH, filtrated on celite, then evaporated
to dryness. The crude residue was dissolved again in MeOH
(5 mL) and MeOK (1M solution in MeOH) was added until
pH 8 was reached. After 2 h, Dowex 50H was added, then
eliminated through filtration. The crude material obtained after
evaporation was purified over silica gel, yielding 11 (0.233 g,
0.57 mmol) 92%; [a]D25 = −27 (c = 1.0 in H2O); 1H NMR (D2O):
(Z) 1-S-[2,3,4,6-Tetra-O-benzoyl-5a-carba-b-D-glucopyrano
syl] phenylacetothiohydroximate 14f. (70%);1H NMR (CDCl3):
1.69 (dd, 1H, J5 b–1 = 11.0 Hz, J5 b–5 a = 13.5 Hz, H-5ꢀb), 1.89 (ddd,
ꢀ
ꢀ
ꢀ
1H, J5 a–1 = J5 a–5 = 3.8 Hz, H-5ꢀa), 2.13–2.4 (m, 1H, H-5), 3.74
(d, 1H, J = 15.7 Hz, CH2Ph), 3.82 (d, 1H, CH2Ph), 3.82–3.93
(m, 1H, H-1), 4.15 (dd, 1H, J6b–5 = 4.7 Hz, J6a–6b = 11.6 Hz,
ꢀ
ꢀ
H-6b), 4.28 (dd, 1H, J6a–5 = 4.1 Hz, H-6b), 5.46 (dd, 1H, J3–2
=
9.4 Hz, J3–4 = 10.6 Hz, H-3), 5.52 (dd, 1H, J2–1 = 11.4 Hz, H-2),
5.64 (dd, 1H, J4–5 = 9.5 Hz,H-4), 7.17–7.56 (m, 17H, Har), 7.70–
7.98 (m, 8H, Har); 13C NMR (CDCl3): 33.0 (C-5ꢀ), 40.0 (C-5),
40.7 (CH2Ph), 43.7 (C-1), 63.7 (C-6), 72.0 (C-2), 74.0 (C-3), 75.2
(C-4), 128.3, 128.4, 128.5, 8.6, 128.7, 128.8, 128.9, 129.1, 129.2,
129.7, 129.8, 129.9, 130.0, 133.1, 133.2, 135.8 (CHar and Car),
3.38–3.57 (m, 4H, H-2, H-3, H-4, H-5), 3.69 (dd, 1H, J6b–5
=
5.1 Hz, H-6b), 3.88 (dd, 1H, J6a–6b = 12.4 Hz, J6a–5 = 1.9 Hz,
H-6a), 4.97 (d, 1H, J1–2 = 9.6 Hz, H-1), 7.27 (m, 3H, Har), 7.64
(dd, 1H, J = 1.7 Hz, J = 7.5 Hz, Har); 13C NMR (D2O): 61.0
(C-6), 69.6, 72.3, 77.4, 80.1 (C-2, C-3, C-4, C-5), 86.4 (C-1),
122.2, 127.3, 129.1, 131.7 (CHar), 127.1 (Car), 158.9 (Car);
MS (IS-): m/z: 367 [M− − K]; HMRS calcd for C12H15KO9S2:
405.9795. Found: 405.9793.
=
151.8 (C N), 165.5, 165.6, 166.0, 166.3 (CO); MS (IS+): m/z:
744.5 [M+ + H], 766.5 [M+ + Na]; elemental analysis calcd for
C43H37NO9S, C, 69.43; H, 5.01. Found: C, 69.81; H, 5.02%.
General protocol for synthesizing thiohydroximates 14
Preparation of di-O-acetates 15
A dichloromethane solution of b-nitrostyrene and Et3SiH
(2.1 eq.) was reacted with TiCl4 (2.2 eq.) under Ar for 16 h
according to Kulkarni.7 The mixture was hydrolyzed with
ice, extracted with dichloromethane, dried over MgSO4 and
evaporated to dryness. The crude residue was dissolved in
dichloromethane and Et3N (3 eq.) before the thiol (1.2 eq.) was
added. At the optimum time (TLC monitoring), the resulting
mixture was directly purified over a silica gel column.
The thiohydroximates 14 were acetylated in dry pyridine with
Ac2O added dropwise. After 12 h at room temperature, the
mixtures were co-evaporated with toluene and purified by silica
gel column chromatography to obtain di-O-acetates 15
(Z) O-Acetyl-S-(2-acetoxyethyl) phenylacetothiohydroximate
15c. (71%); oil; 1H NMR (CDCl3): 1.76 (s, 3H, N–OAc), 1.93
(s, 3H, CH2OAc), 2.75 (t, 2H, J = 6.2 Hz, CH2S), 3.80–3.82 (m,
4H, CH2O, CH2Ph), 7.01–7.10 (m, 5H, Har); 13C NMR (CDCl3):
18.6, 20.0 (COCH3), 27.8 (CH2S), 38.0 (PhCH2), 62.0 (CH2O),
(Z) S-◦Benzyl phenylacetothiohydroximate 14a. (41%); mp:
1
=
134–135 C; H NMR (CDCl3): 3.78 (s, 2H, CH2C N), 3.92 (s,
=
126.8, 127.6, 128.3, 134.3 (CHar and Car), 162.7 (C N), 167.0,
2H, CH2S), 7.23–7.31 (m, 10H, Har), 13C NMR (CDCl3): 33.9
169.7 (CO); MS (IS+): m/z: 296 [M+ + H]; elemental analysis
calcd for C14H17NO4S C, 56.96; H, 5.80; N, 4.74. Found: C,
56.76; H, 5.78; N, 4.73%.
=
(SCH2), 39.2 (PhCH2C N), 127.1, 127.5, 128.3, 128.7, 128.8,
135.9, 136.3 (CHar and Car), 154.7 (C N); MS (IS+): m/z: 258
=
[M+ + H]; elemental analysis calcd for C15H15NOS C, 70.01; H,
(Z) O-Acetyl-S-(3-acetoxypropyl) phenylacetothio hydroxi-
5.87; N, 5.44. Found: C, 69.76; H, 5.85; N, 5.42%.
1
mate 15d. (90%): oil; H NMR (CDCl3): 1.75 (qt, 2H, J =
(Z) S-Ethyl phenylacetothiohydroximate 14b. (42%); oil;1H
NMR (CDCl3): 1.18 (t, 3H, J = 7.5 Hz, CH3), 2.73 (q, 2H,
CH2S), 3.82 (s, 2H, CH2Ph), 7.21–7.35 (m, 5H, Har), 8.9 (ls,
1H, OH); 13C NMR (CDCl3): 14.7 (CH3), 23.8 (CH2S), 39.1
6.6 Hz, CH2); 1.98, 2.19 (2s, 2 × 3H, OAc); 2.78 (t, 2H, J =
7.4 Hz, CH2S); 3.98–4.03 (m, 4H, OCH2, CH2Ph); 7.23–7.32
(m, 5H, Har); 13C NMR (CDCl3): 19.0, 20.4 (COCH3); 26.1
(CH2); 28.2 (CH2S); 38.6 (PhCH2); 61.9 (CH2O); 127.0, 127.8,
(CH2Ph), 127.1, 128.4, 128.8, 136.2 (CHar and Car), 155.3
=
128.6, 134.7 (CHar and Car); 163.6 (C N); 167.3, 170.3 (CO);
+
MS (IS+): m/z: 310 [M+ + H]; elemental analysis calcd for
C15H19NO4S C 58.23; H, 6.19; N, 4.53. Found: C 58.11; H, 6.30;
N, 4.37%.
=
(C N); MS (IS+): m/z: 196 [M + H]; elemental analysis calcd
for C10H13NOS C, 61.50; H, 6.71; N, 7.17. Found: C, 61.22; H,
6.68; N, 7.14%.
(Z) S-(2-Hydroxyethyl) phenylacetothiohydroximate 14c.
(71%); oil; 1H NMR (CD3OD): 2.91 (t, 2H, J = 6.4 Hz,
CH2S), 3.60 (t, 2H, CH2O), 3.93 (s, 2H, CH2Ph), 7.27–7.36 (m,
5H, Har); 13C NMR (CD3OD): 32.2 (CH2S), 39.1 (CH2Ph),
(Z) O-Acetyl-S-(4-acetoxybutyl) phenylacetothiohydroximate
15e. (94%); oil; 1H NMR (CDCl3): 1.47–1.63 (m, 4H,
CH2CH2), 2.00, 2.21 (2s, 2 × 3H, OAc), 2.74 (t, 2H, J =
7.0 Hz, CH2S), 3.93–3.98 (m, 4H, OCH2, CH2Ph), 7.25–7.33
(m, 5H, Har); 13C NMR (CDCl3): 19.2, 20.8 (COCH3), 25.7,
27.3 (CH2CH2), 29.3 (CH2S), 39.0 (PhCH2), 63.3 (CH2O), 127.2,
62.0 (CH2O), 127.6, 128.9, 129.4, 137.6 (CHar and Car), 154.8
+
=
(C N); MS (IS+): m/z: 212 [M + H]; elemental analysis calcd
for C10H13NO2S C 56.85; H, 6.20; N, 6.63. Found: C 56.70; H,
6.19; N, 6.61%.
=
128.0, 128.8, 134.9 (CHar and Car), 164.0 (C N), 167.7, 170.8
(CO); MS (IS+): m/z: 324 [M+ + H]; elemental analysis calcd
for C16H21NO4S C, 59.42; H, 6.54; N, 4.33. Found: C, 59.11; H,
6.71; N, 4.05%.
(Z) S-(3-Hydroxypropyl) phenylacetothiohydroximate 14d.
(62%); oil; 1H NMR (CDCl3): 1.70 (qt, 2H, J = 6.6 Hz, CH2),
2.32 (ls, 1H, OH), 2.81 (t, 2H, J = 7.2 Hz, CH2S), 3.59 (t, 2H,
J = 6.6 Hz, CH2O), 3.82 (s, 2H, CH2Ph), 7.22–7.34 (m, 5H,
Har), 9.67 (ls, 1H, OH); 13C NMR (CDCl3): 25.9 (CH2), 32.3
Protocol for selective de-O-acetylation to monoacetates 16.
Under dry conditions, di-O-acetates 15 were dissolved in DMF
and stirred at 60 ◦C for 4 h in the presence of hydrazinium
monoacetate. After completion, the reaction medium was
extracted with ethyl acetate, washed with water to remove
DMF and salts, then washed with brine, dried over MgSO4
and evaporated to a crude residue which could be used without
purification for the next step. Structure checking was effected by
MS and 1H NMR.
(CH2S), 39.1 CH2Ph), 60.7 (CH2O), 127.1, 128.4, 128.9, 136.2
+
=
(CHar and Car), 155.1 (C N); MS (IS+): m/z: 226 [M + H];
elemental analysis calcd for C11H15NO2S C, 58.64; H, 6.71; N,
6.22. Found: C, 58.45; H, 6.82; N, 5.93%.
(Z) S-(4-Hydroxybutyl) phenylacetothiohydroximate 14e.
1
(59%); oil; H NMR (CDCl3): 1.48–1.52 (m, 4H, CH2), 2.67
(t, 2H, J = 6.6 Hz, CH2S), 2.99 (ls, 1H, OH), 3.48 (t, 2H, J =
5.6 Hz, CH2O), 3.79 (s, 2H, CH2Ph), 7.18–7.31 (m, 5H, Har),
10.27 (ls, 1H, OH); 13C NMR (CDCl3): 26.0, 29.0 (CH2), 31.2
(Z) S-(2-Acetoxyethyl) phenylacetothiohydroximate 16c.
(66%);1H NMR (CD3OD): 1.97 (s, 3H, OAc), 2.94 (t, 2H, J =
6.6 Hz, CH2S), 3.87 (s, 2H, CH2Ph), 4.00 (t, 2H, J = 6.6 Hz,
CH2O), 7.20–7.30 (m, 5H, Har); MS (IS+): m/z: 254 [M+ + H].
(CH2S), 38.9 (CH2Ph), 61.8 (CH2O), 127.0, 128.2, 128.7, 136.2
+
=
(CHar and Car), 155.1 (C N); MS (IS+): m/z: 240 [M + H];
elemental analysis calcd for C12H17NO2S C, 60.22; H, 7.16; N,
5.85. Found: C, 59.95; H, 7.13; N, 5.83%.
(Z) S-(3-Acetoxypropyl) phenylacetothiohydroximate 16d.
(77%);1H NMR (CDCl3): 1.75 (qt, 2H, J = 6.8 Hz, CH2), 1.99
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 7 2 – 1 8 7 9
1 8 7 7

View Article Online
(s, 3H, OAc), 2.76 (t, 2H, J = 7.4 Hz, CH2S), 3.82 (s, 2H, CH2Ph),
4.01 (t, 1H, J = 6.2 Hz, CH2O), 7.22–7.33 (m, 5H, Har), 9.98
(bs, 1H, OH); MS (IS+): m/z: 268 [M+ + H].
overnight. Work-up using ice and 10% K2SO4 followed by
dichloromethane extraction gave an organic phase which was
washed with sat. NaHCO3, water, dried over MgSO4 and
1
evaporated. Crude 19 was pure enough for the next step. H
(Z) S-(4-Acetoxybutyl) phenylacetothiohydroximate 16e.
ꢀ
ꢀ
NMR (CDCl3): 2.18 (ddd, 1H, J5 b,5 = 13.6 Hz, J5 b,1 = 2.0 Hz,
1
(83%); oil; H NMR (CDCl3): 1.50–1.59 (m, 4H, CH2CH2),
H-5ꢀb), 2.54 (ddd, 1H, J5 a,5 b = 15.8 Hz, J5 a,1 = 3.6 Hz, J5 a,5
=
=
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
2.00 (s, 3H, OAc), 2.71 (t, 2H, J = 7.0 Hz, CH2S), 3.82 (s, 2H,
CH2Ph), 3.95 (t, 2H, CH2O), 7.22–7.29 (m, 5H, Har); MS (IS+):
m/z: 282 [M+ + H].
4.0 Hz, H-5ꢀa), 2.78–2.89 (m, 1H, H-5), 4.38 (dd, 1H, J6b,5
4.5 Hz, H-6b), 4.51 (dd, 1H, J6a,6b = 11.7 Hz, J6a,5 = 3.5 Hz,
H-6a), 5.49 (dd, J2,1 = 2.7 Hz, H-2), 5.60 (sl, 1H, H-1), 5.72 (dd,
J3,2 = 10.4 Hz, H-3), 6.11 (dd, 1H, J3,4 = J4,5 = 10.1 Hz, H-4),
7.21–7.61 (m, 12H, Har), 7.78–8.03 (m, 8H, Har); MS (IS+):
m/z: 599 [M+ − TfOH + Na], 594 [M+ − TfOH + NH4], 577
[M+ − TfOH + H].
Sulfation protocol
Thiohydroximates 16 were dissolved in CH3CN under dry Ar
then reacted with pyridine–SO3 complex in excess (5 eq.). After
1 h heating at 50 ◦C, the mixture was cooled and an aquous
solution of KHCO3 (10 eq.) was added. The crude mixture
was evaporated to dryness and purified by chromatography
to yield the final compound 17. In the case of acetylated
intermediates 16c-e, de-O-acylation was performed through
MeOK-catalysed methanolysis. Compounds 17 were purified
by silica gel chromatography. The same protocol was applied to
prepare the carbaglucosinolate 12.9
1-S-(2,3,4,6-Tetra-O-benzoyl-5a-carba-b-DL-glucopyranosyl)
isothiouronium triflate 20. Under dry conditions, the crude
triflate 19 and thiourea (46 mg, 0.6 mmol) were dissolved in
butanone. The solution was heated at 80 ◦C for 5 h, then cooled
and evaporated. The crude mixture was directly purified over
1
silica gel, to produce 20 (0.17 g, 0.02 mmol, 67% yield). H
NMR (CD3OD): 2.08–2.23 (m, 1H, H-5ꢀb), 2.55 (ddd, 1H,
J5 a,5 b = 13.5 Hz, J5 a,1 = J5 a,5 = 4.0 Hz, H-5ꢀa), 2.65–2.81 (m,
1H, H-5), 4.26–4.35 (m, 1H, H-1), 4.41 (dd, 1H, J6b,5 = 4.7 Hz,
H-6b), 4.48 (dd, 1H, J6a,6b = 11.5 Hz, J6a,5 = 4.4 Hz, H-6a),
5.71–5.88 (m, 3H, H-2, H-3, H-4), 7.21–7.59 (m, 12H, Har),
7.69–7.70 (m, 2H, Har), 7.83–803 (m, 6H, Har); 13C NMR
(CD3OD): 31.4 (C-5ꢀ), 40.8 (C-5), 47.4 (C-1), 65.4 (C-6), 73.4,
75.7, 76.2 (C-2, C-3, C-4), 129.3, 129.7, 129.8, 129.9, 130.2,
130.4, 130.6, 130.7, 134.3, 134.5, 134.6, 134.9 (CHar and Car),
ꢀ
ꢀ
ꢀ
ꢀ
(Z) S-Benzyl phenylacetothiohydroximate-O-sulfate 17a.
(67%);1H NMR (CD3OD): 3.78 (s, 4H, 2 × CH2Ph), 6.94–
7.22 (m, 10H, Har); 13C NMR (CD3OD): 35.0 (CH2S), 39.6
(CH2Ph), 128.2, 128.6, 129.2, 129.3, 129.7, 129.8, 130.0, 137.2,
−
=
137.3 (CHar and Car), 162.0 (C N); MS (IS-): m/z: 336 [M −
K]; HRMS calcd for C15H14KNO4S2: 375.001. Found 374.9997.
(Z) S-Ethyl phenylacetothiohydroximate-O-sulfate 17b.
(99%);1H NMR (CD3OD): 1.1 (t, 3H, J = 7.6 Hz, CH3), 2.74
=
166.7, 167.0, 167.1, 167.7 (CO), 185.4 (C N); MS (IS+): m/z:
(q, 2H, CH2S), 3.95 (s, 2H, CH2Ph), 7.13–7.35 (m, 5H, Har); 13
C
653 [M+ − TfO]; HRMS calcd for C37H33F3N2O11S2 802.1447.
NMR (CD3OD): 14.5 (CH3), 24.8 (CH2S), 39.4 (CH2Ph), 128.0,
Found 802.1453
=
129.2, 129.7, 137.1 (CHar and Car), 162.9 (C N); MS (IS-):
m/z: 274 [M− − K]; HRMS calcd for C10H12KNO4S2 312.9845.
2,3,4,6-Tetra-O-benzoyl-1-thio-carba-b-DL-glucopyranose 21.
The isothiouronium salt 20 (0.69 g, 0.86 mmol) was suspended in
water (8 mL) containing Na2S2O5 (0.33 g, 1.72 mmol) and heated
to 85 ◦C before CHCl3 (7 mL) was added. The resulting solution
was heated for another 3 h in order to complete hydrolysis. After
usual work-up, extraction with CHCl3 followed by washes with
water and brine, silica gel column purification gave a 80% yield
of thiol 21 (0.5 g, 0.82 mmol). MS (IS+): m/z: 633 [M+ + Na];
1H NMR (CDCl3) 1.74–1.87 (m, 2H, H-5ꢀb, SH), 2.38–2.47 (m,
Found 313.9852.
(Z) S-(2-Hydroxyethyl) phenylacetothiohydroximate-O-
sulfate 17c. (48%);1H NMR (CD3OD): 2.87 (t, 2H, J =
6.6 Hz, CH2S), 3.53 (t, 2H, J = 6.6 Hz, CH2O), 3.99 (s, 2H,
CH2Ph), 7.24–7.40 (m, 5H, Har); 13C NMR (CD3OD): 33.2
(CH2S), 39.5 (CH2Ph), 62.0 (CH2O), 128.1, 129.3, 129.8, 137.2
−
=
(CHar and Car), 162.4 (C N); MS (IS-): m/z: 290 [M − K];
HRMS calcd for C10H12KNO5S2 328.9794. Found 328.9788.
2H, H-5ꢀa, H-5), 3.14–3.19 (m, 1H, H-1), 4.23 (dd, 1H, J6b,5
=
(Z) S-(3-Hydroxypropyl) phenylacetothiohydroximate-O-
sulfate 17d. (68%);1H NMR (CD3OD): 1.64 (qt, 2H, J =
6.8 Hz, CH2), 2.82 (t, 2H, J = 7.3 Hz, CH2S), 3.50 (t, 2H,
J = 6.2 Hz, CH2O), 3.97(s, 2H, CH2Ph), 7.21–7.40 (m, 5H,
Har); 13C NMR (CD3OD): 27.1 (CH2), 33.6 (CH2S), 39.4
5.9 Hz, H-6b), 4.39 (dd, 1H, J6a,6b = 11.4 Hz, J6a,5 = 3.7 Hz, H-
6a), 5.40 (dd, J2,1 = 9.7 Hz, J2,3 = 10.6 Hz, H-2), 5.54–5.67 (m,
2H, H-3, H-4), 7.10–7.67 (m, 12H, Har), 7.38–7.95 (m, 8H, Har);
13C NMR (CDCl3): 35.2 (C-5ꢀ), 39.7 (C-1), 40.5 (C-5), 64.0 (C-
6), 72.3, 74.95 (C-3, C-4), 76.9 (C-2), 128.3, 128.4, 128.5, 128.6,
129.0, 129.1, 129.3, 129.7, 129.8, 129.9, 133.1, 133.2, 133.3, 133.4
(CHar and Car), 165.7, 165.8, 166.0, 166.4 (CO); MS (IS+): m/z:
633 [M+ + Na]; elemental analysis calcd for C35H30O8S C, 68.83;
H, 4.96. Found: C, 67.83; H, 4.88%.
(CH2Ph), 60.9 (CH2O), 128.1, 129.3, 129.7, 137.2 (CHar and
−
=
Car), 162.6 (C N); MS (IS-): m/z: 304 [M − K]; HRMS calcd
for C11H14KNO5S2 342.9951. Found 342.9947.
(Z) S-(4-Hydroxybutyl) phenylacetothiohydroximate-O-
sulfate 17e. (63%);1H NMR (CD3OD): 1.46–1.51 (m, 4H,
CH2), 2.75 (t, 2H, J = 7.1 Hz, CH2S), 3.44 (t, 2H, J = 6.0 Hz,
CH2O), 3.96 (s, 2H, CH2Ph), 7.25–7.40 (m, 5H, Har); 13C
NMR (CD3OD): 27.0, 30.3 (CH2), 32.3 (CH2S), 39.5 (CH2Ph),
Enzyme assay
All reagents were purchased from Sigma. The buffer was 0.1 M
sodium phosphate at pH 7. Myrosinase from Sinapis alba10,15
was a gift from Dr R. Iori (CRA, ISCI Bologna, Italy). The
enzyme kinetics were carried out using a method adapted from
the literature.11 Myrosinase (10 lL, 36 U mL−1) was incubated
at 37 ◦C in the absence or presence of inhibitor (38 lL at final
concentrations of 10−3 to 10 mM in H2O) with sinigrin (302 lL at
final concentration of 2.3 mM in buffer), glucose oxidase (10 lL
at concentration of 55U mL−1 in buffer), peroxidase (10 lL at
concentration of 131U mL−1 in buffer), and o-dianisidine (2.5 mg
mL−1 in H2O). The release of glucose was measured by following
the variation of the absorbance at 436 nm. Classical Michaelis–
Menten kinetics were observed under the conditions used, and
62.0 (CH2O), 128.1, 129.3, 129.8, 137.3 (CHar and Car),
−
=
162.6 (C N); MS (IS-): m/z: 318 [M − K]; HRMS calcd for
C12H16KNO5S2: 357.0107. Found 357.0112.
Preparation of the 1-thio-carbaglucose derivative 21
2,3,4,6-Tetra-O-benzoyl-1-O-trifluoromethanesulfonyl-5a-carba-
a-DL-glucopyranose 19: under dry conditions, 2,3,4,6-tetra-O-
benzoyl-5a-carba-a-DL-glucopyranose 18 (0.15 g, 0.25 mmol)
was dissolved in dichloromethane containing pyridine
(0.124 mL, 1.51 mmol) and DMAP (6 mg, 0.05 mmol). After
cooling to −78 ◦C, triflic anhydride (0.124 mL, 0.76 mmol)
was added and the mixture was stirred for 30 min at −78 ◦C
and then allowed to return to room temperature and stirred
R
the IC50 determined using the GRAFITꢀ software.
1 8 7 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 7 2 – 1 8 7 9

View Article Online
Crystallographic studies
Biochemistry, 1996, 35, 15256–15259; (c) M. Blanc-Muesser, H.
Driguez, B. Joseph, M. C. Viaud and P. Rollin, Tetrahedron Lett.,
1990, 31, 3867–3868; (d) C. Gardrat, A. Quinsac, B. Joseph and P.
Rollin, Heterocycles, 1996, 35, 1015–1027; (e) B. Joseph and P. Rollin,
J. Carbohydr. Chem., 1993, 12, 719–729; (f) S. Cottaz, P. Rollin and
H. Driguez, Carbohydr. Res., 1997, 298, 127–130.
3 (a) A. Kjaer and T. Skrydstrup, Acta Chem. Scand., 1987, 41, 29–33;
(b) V. Aucagne, D. Gueyrard, A. Tatiboue¨t, A. Quinsac and P. Rollin,
Tetrahedron, 2000, 56, 2647–2654.
4 S. Lazar and P. Rollin, Tetrahedron Lett., 1994, 35, 2173–2174.
5 W. P. Burmeister, S. Cottaz, H. Driguez, R. Iori, S. Palmieri and B.
Henrissat, Structure, 1997, 5, 663–675.
6 M. Cerny, T. Trnka and M. Budesinsky, Collect. Czech. Chem.
Commun., 1996, 61, 1489–1500.
7 S. Cassel, B. Casenave, G. De´le´ris, L. Latxague and P. Rollin,
Tetrahedron, 1998, 54, 8515–8524.
8 (a) S. Ogawa, K. Nakamura and T. Takagaki, Bull. Chem. Soc.
Jpn., 1986, 59, 2956–2958; (b) S. Ogawa, Y. Yato, K. Nakamura
and T. Takagaki, Carbohydr. Res., 1986, 148, 249–255; (c) S. Ogawa,
I. Kasahara and T. Suami, Bull. Chem. Soc. Jpn., 1979, 52, 118–
123.
9 M. Lefoix, A. Tatiboue¨t, S. Cottaz, H. Driguez and P. Rollin,
Tetrahedron Lett., 2002, 43, 2889–2890.
10 S. Palmieri, O. Leoni and R. Iori, Anal. Biochem., 1982, 123, 320–
324.
Sinapis alba myrosinase crystals were grown as described.5
Crystals were soaked for some hours in their mother liquor (66%
sat. ammonium sulfate, 100 mM HEPES pH 7.5) with 20 mM
of inhibitor 12 or 10 mM of 17b. Data (resolution between
˚
1.5 and 1.8 A, Rcryst typically 0.07, completeness 99%) were
collected at cryogenic temperature at the ESRF synchrotron
(Grenoble, France) on beamlines ID14-1 and ID14-2 and
reduced using standard crystallography software as described.13
The experiments were interpreted with initial Finh–Fnat difference
Fourier maps where Finh are the structure factor amplitudes of a
dataset collected on a crystal soaked with inhibitor, whereas Fnat
and the phases were obtained from a native myrosinase structure
(pdb entry 1e4m13). The signal in these maps is reduced due
to the presence of a well ordered glycerol molecule and several
water molecules in the active site. In the case of the glucosinolate
inhibitor, difference maps of three independent experiments were
averaged in order to improve the signal to noise ratio of the
electron density for the inhibitor (software MAPMAN,16). The
inhibitor structures (compound 13 and 17b) have been built into
the electron density and subjected to refinement using CNS17 at
˚
a resolution of 1.7 and 1.6 A. The crystallographic R-factors are
11 R. Bjo¨rkman and B. Lo¨nnerdal, Biochim. Biophys. Acta, 1973, 327,
121–131.
0.175 and 0.179 (Rfree 0.191 and 0.181). Due to an occupancy
of only about 0.5 of the carba-glucosinolate inhibitor, alternate
structures in presence and in absence of the inhibitor have been
modeled into the active site. Coordinates and structure factors
have been submitted to the PDB (entries 1w9b and 1w9d) where
further details about data collection and refinement can be
found. Illustrations were made using BOBSCRIPT18 and PyMol
(Delano Scientific).
12 A. P. Wilkinson, M. J. C. Rhodes and G. R. Fenwick, Anal. Biochem.,
1984, 139, 284–291.
13 W. P. Burmeister, S. Cottaz, P. Rollin, A. Vasella and B. Henrissat,
J. Biol. Chem., 2000, 275, 39385–39393.
14 A. M. Scofield, J. T. Rossiter, P. Witham, G. C. Kite, R. J. Nash and
L. E. Fellows, Phytochemistry, 1990, 29, 107–109.
15 A. Pessina, R. M. Thomas, S. Palmieri and P. L. Luisi, Arch. Biochem.
Biophys., 1990, 280, 383–389.
16 G. J. Kleywegt and T. A. Jones, Acta Crystallogr., Sect. D: Biol.
Crystallogr., 1996, D52, 826–828.
17 A. T. Bru¨nger, P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros,
R. W. Grosse-Kunstleve, J. S. Jiang, J. Kuszewski, M. Nilges, N. S.
Pannu, R. J. Read, L. M. Rice, T. Simonson and G. L. Warren, Acta
Crystallogr., Sect. D: Biol. Crystallogr., 1994, D54, 905–921.
18 R. M. Esnouf, Acta Crystallogr., Sect. D: Biol. Crystallogr., 1999,
D55, 938–940.
References
1 J. W. Fahey, A. T. Zalcmann and P. Talalay, Phytochemistry, 2001,
56, 5–51.
2 (a) R. Iori, P. Rollin, H. Streicher, J. Thiem and S. Palmieri, FEBS
Letters, 1996, 385, 87–90; (b) S. Cottaz, B. Henrissat and H. Driguez,
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 8 7 2 – 1 8 7 9
1 8 7 9