Thioimidate N-oxides: From nature to synthetic pathways

Article abstract of DOI:10.1055/s-0029-1219382

Inspired by the unexpected reactivity of desulfated naturally occurring glucoraphenin, methods to synthesize thioimidate N-oxides (TIO) have been devised on simple or carbohydrate templates. Either through halocyclization or under Mitsunobu conditions, the starting thiohydroximates cyclized to generate efficiently the corresponding TIO. Georg Thieme Verlag Stuttgart ? New York.

Full text of DOI:10.1055/s-0029-1219382

LETTER
725
Thioimidate N-Oxides: From Nature to Synthetic Pathways
T
hioimida
u
te
N
-Oxide
s
lie Schleiss,a Deimante Cerniauskaite,a David Gueyrard,a Renato Iori,b Patrick Rollin,a Arnaud Tatibouët*a
a
Institut de Chimie Organique et Analytique, UMR 6005, Associé au CNRS, Université d’Orléans, B.P. 6759, 45067 Orléans, France
Fax +33(2)38417281; E-mail: arnaud.tatibouet@univ-orleans.fr
Consiglio per la Ricerca e la Sperimentazione in Agricoltura (C.R.A.-C.I.N.), Via di Corticella 133, 40129 Bologna, Italy
b
Received 23 October 2009
OH
K+
glucoraphenin (GRE) 2
Abstract: Inspired by the unexpected reactivity of desulfated natu-
rally occurring glucoraphenin, methods to synthesize thioimidate
N-oxides (TIO) have been devised on simple or carbohydrate tem-
plates. Either through halocyclization or under Mitsunobu condi-
tions, the starting thiohydroximates cyclized to generate efficiently
the corresponding TIO.
O
HO
HO
S
A =
A
SOMe
*
OH
N
O3SO–
sulfatase
glucosinolates 1
Key words: carbohydrates, cyclization, heterocycles, azasugars,
thio function
A = alkyl, aryl, indolyl,
hydroxyalkenyl,
SOMe
S
β-D-Glc
thiofunctionalized chain
N
3
HO
Along our journey for studying sulfur-containing second-
ary metabolites, glucosinolates have been a cornerstone in
the disclosure and development of original chemical moi-
eties. Glucosinolates 1 (GL) are remarkable for their ho-
mogeneous structure: a hydrophilic β-D-glucopyrano
framework bearing an O-sulfated anomeric (Z)-thiohy-
droximate moiety connected to a fairly hydrophobic agly-
con side chain (Scheme1).1 Those molecules are heavily
involved in human and animal nutrition, having shown
impact on health status either as chemopreventive agents,
antioxidant activities2a or by inducing antinutritional ef-
fects on livestock.2b This latter aspect has led to establish-
ing standard procedures (EU official method ISO-9167-1)
for determining the amount of GL in Brassicaceae, nota-
bly in rapeseed.3 This method is based on preliminary en-
zymatic desulfation of GL, followed by HPLC analysis of
the resulting more stable desulfo-GL. We have recently
reported the unique behavior of glucoraphenin (4-methyl-
sulfinylbut-3-enyl GL) 2, whose desulfo-counterpart 3 is
readily converted into an unprecedented cyclic thioimi-
date N-oxide 4 (Scheme1).4
desulfo-GRE
S
β-D-Glc
N+
–O
4
SOMe
Scheme 1 Structure of glucosinolates and unexpected desulfation
of glucoraphenin 2 resulting in a cyclized TIO
clization or a hydroxyl group prior to nucleophilic
substitution (Scheme2).
thioimidate N-oxide
starting materials
thiohydroximate
nucleophilic
substitution
+
N
O–
N
OH
O
O
RS
OH
RS
halo-
cyclization
+
N
N
X
O
O–
This can be explained by intramolecular concerted Micha-
el addition of the thiohydroximate moiety of 3 onto the vi-
nyl sulfoxide acceptor. The originality of this rare
thioimidate N-oxide function prompted us to design pre-
parative synthetic methods, as a prerequisite to evaluate
the chemical potential and reactivity scope of thioimidate
N-oxides.4b The formation of a cyclic thioimidate N-oxide
could be controlled by avoiding direct attack on an elec-
trophilic carbon. We chose to introduce a thiohydroximate
function at one end of the chain and an activable moiety at
the other end. Two possibilities for the latter were first ex-
plored: using either an alkenyl group to perform halocy-
HO
RS
RS
OH
Scheme 2 Two routes to prepare a thioimidate N-oxide
At first, we used natural alkenyl glucosinolates
(Scheme1), sinigrin (A = ethenyl), gluconapin (A = prop-
2-enyl), and epiprogoitrin (A = 1-hydroxyprop-2-enyl)
which fulfilled our requirements and gave us rapidly the
opportunity to test the terminal alkene approach. To reach
the desired thiohydroximate templates, standard enzymat-
ic desulfation was performed,3 providing the correspond-
ing desulfoglucosinolates 5a–c in quantitative yield
(Scheme3). Prior to the halocyclization step, the OH
groups of the sugar ring were masked in a two-step se-
quence: standard per-O-acetylation of 5a–c followed by
selective deprotection of the N-hydroxyimino group using
hydrazinium acetate. All reactions proceeded smoothly,
SYNLETT 2010, No. 5, pp 0725–0728
1
6.
3
.2
0
1
0
Advanced online publication: 10.02.2010
DOI: 10.1055/s-0029-1219382; Art ID: D30109ST
© Georg Thieme Verlag Stuttgart · New York

726
J. Schleiss et al.
LETTER
OH
OAc
O
O
1) Ac2O,
pyridine, DMAP
HO
HO
AcO
AcO
S
R
S
R
n
n
OH
OAc
2) AcONH3NH2
N
N
HO
HO
5a n = 0, R = H, desulfosinigrin
5b n = 1, R = H desulfogluconapin
5c n = 1, R = OH desulfoepiprogoitrin
6a n = 0, R = H, 89%
6b n = 1, R = H, 84%
6c n = 1, R = OAc, 70%
NBS, CH2Cl2
OAc
O
AcO
AcO
S
n
OAc
R
N+
–O
7b n = 1, R = H, 57%
7c n = 1, R = OAc, 43%
Br
Scheme 3 TIO derived from desulfoglucosinolates through halocyclization
O–
N+
giving (70–94% yield) thiohydroximates 6a–c, which
were further submitted to the halocyclization conditions
previously established by Grigg5 and Jäger.6 On reaction
with NBS in dichloromethane, protected desulfosinigrin
6a only led to degradation products. In contrast, desulfo-
GL 6b and 6c readily underwent 5-exo-trig bromocycliza-
tion to give glucosyl thioimidate-N-oxides 7b and 7c in
57% and 43% yield, respectively. Those compounds ap-
peared rather unstable and degradation occurred at room
temperature.
Br
SEt
O
O
D-ribose
10
77%,
1) H2SO4 1 M
acetone, MeOH
3) BuLi, THF
4) NH2OH⋅HCl
dr: L-lyxo/D-ribo = 60:40
2) Ph3P, imidazole
I2, THF
pyridine, EtOH
NBS, CH2Cl2
Nevertheless, the above results strengthened our resolve
to develop more accessible models based on carbohydrate
frameworks.
OH
N
OH
N
SEt
a) NCS, DMF
b) EtSH, Et3N
O
With a view to improving the stability of our structures,
we turned to a carbohydrate template (Scheme4). Starting
with D-ribose, the acyclic aldoxime 8 could be obtained in
four steps with a 34% overall yield.6b,7 The thiohydroxi-
mate function was more smoothly introduced, compound
9 being isolated in 71% yield.8 NBS-induced halocycliza-
tion afforded the ethyl thioimidate N-oxide 10 in 77%
yield. After standing for a few days at room temperature,
compound 10 also revealed some instability, whereas no
major degradation was observed when kept at –18 °C:
thus a carbohydrate template approach is favorable in the
synthesis of TIO.
O
O
O
9
71%
8
54% from D-ribose
Scheme 4 Halocyclization on more simple thiohydroximate tem-
plates
thiohydroximate 13a albeit in relatively poor yield
(Scheme5).10
We thus moved to a new sequence in order to shorten the
process. A range of four lactones, γ-butyrolactone (14a),
γ-valerolactone (14b), δ-valerolactone (14c), and e-capro-
lactone (14d), was selected for conversion into hydroxam-
ic acids, which were readily protected in the form of bis-
O-silylated derivatives 15 (Scheme6). The overall yields
(44–57%, Table1) are in a reasonable range, taking into
Our second path to generate TIO involved nucleophilic
substitution of an activated terminal alcohol. Mono-O-
silylation of butane-1,4-diol (11) followed by Swern oxi-
dation then aldoxime formation afforded in 42% yield
precursor 12,9 which could be converted into the ethyl
OH
OTIPS
OH
1) TIPSCl, imidazole
DMF, 0 °C
1) NCS, CHCl3
2) EtSH, Et3N, DMF
3) TBAF, THF
2) (COCl)2, DMSO,
Et3N, –78 °C, CH2Cl2
3) NH2OH⋅HCl,
EtOH–pyridine
OH
N
N
OH
OH
11, X = OH
H
12
SEt
13a 16%
Scheme 5 The nucleophilic displacement approach (pathway 1)
Synlett 2010, No. 5, 725–728 © Thieme Stuttgart · New York

LETTER
Thioimidate N-Oxides
727
account the sensitivity of O-TBDMS-protected hydrox- closure yield and stability of the TIO. We have explored
amic acids to silica gel chromatography. The following in parallel both synthetic sequences previously examined
critical synthetic step to deliver ethyl thiohydroximates 13 to access a suitable template, carrying a thiohydroximate
was performed using a protocol developed by Carreira et at one end and a free alcohol at the other end. The lactone
al.11 to generate nitrile oxides. Condensation was effected approach starting with 2,3-O-isopropylidene-D-erythro-
in a two-step sequence: formation of the triflated hydrox- nolactone proved unsuccessful, as we were unable to iso-
imates, then condensation with ethanethiol under basic late the hydroxamic acid whatever the conditions used.
conditions. Obtained in ca. 80% yield, the intermediate O- On the contrary, easy one-pot conversion of 2,3-O-isopro-
silylated thiohydroximates were then deprotected using pylidene-L-erythrose (17)13 into the aldoxime 186b,14 al-
standard TBAF conditions, and the thiohydroximate pre- lowed a four-step, efficient access (67% overall yield) to
cursors 13 were obtained in 65–75% yields. It can be not- the thiohydroximate 19 (Scheme7).15 Final de-O-silyla-
ed from example 13a how much shorter and more tion proved more efficient using tetrabutylammonium tet-
efficient this last sequence is.
rafluoroborate (TBAT) instead of standard TBAF. Two
different ring-closing procedures were then applied.
Method A involved mesyl activation of the primary alco-
hol prior to application of basic conditions to induce in-
tramolecular cyclization. Sodium hydrogencarbonate
proved inefficient, leading to rapid hydrolysis to afford
the hydroxylactam 20 in 70% yield. Sodium hydroxide al-
lowed formation of TIO 21 as the minor product (20%
yield) together with 20 (50% yield). Method B used the
Mitsunobu procedure: under those conditions, the expect-
ed TIO 21 was obtained in nearly quantitative yield.16
OTBDMS
HN
TBDMSO
R
1) KOH, HONH3Cl,
R
O
MeOH
O
O
2) TBDMSCl,
imidazole, DMF
15a–nd
n
14a–d
1) Tf2O, Et3N,
CH2Cl2, 0 °C
2) EtSH, Et3N,
CH2Cl2
3) TBAF, THF
O–
OH
+
Ph3P, DIAD
OH
R
N
N
O
R
THF
S
S
OH
OTBDMS
OH
HONH3Cl, pyridine
then tBDMSCl
n
N
n
O
O
13a–d
O
16a–d
18
67%
O
17
Scheme 6 The nucleophilic displacement approach (pathway 2)
Table 1 From Lactone Opening to Thiohydroximates and TIO
a) NCS, DMF
b) EtSH, Et3N
c) TBAT, THF
Lactone
Hydroxamic Thiohy-
TIO
Overall
acid
(yield, %)
droximate (yield, %) yield (%)
(yield, %)
OH
O–
N+
OH
OH
N
14a n = 1, R = H
15a 45
13a 65
13b 74
13c 75
13d 71
16a 35 10
16b 79 33
N
SEt
+
O
SEt
14b n = 1, R = Me 15b 57
O
O
O
19
84%
O
O
O
14c n = 2, R = H
14d n = 3, R = H
15c 51
15d 44
16c
16d
–
–
20
21
20%
method A: MsCl, pyridine–CH2Cl2
then NaOH, 2 M, 0 °C
98% method B: Ph3P, DIAD, THF
50%
–
Our major concern being to prevent hydrolysis of the cy-
clic TIO into an N-hydroxylactam, we postulated that a
mild neutral reaction protocol might be appropriate. Mit-
sunobu conditions were thus applied to thiohydroximates
13 and the internal nucleophilic displacement produced
interesting results. While six- or seven-membered rings
did not form (neither 16c nor 16d were detected in the re-
action medium), the five-membered ring TIO 16a and 16b
were obtained in 35% and 79% yield, respectively. This
dramatic improvement in reactivity might be attributed to
a stabilization effect related to the ring size; in addition,
the considerable yield increase observed from 16a to 16b
is likely to derive from some transition-state stabilization
effect in the methyl-branched precursor 13b.12
Scheme 7 The nucleophilic displacement approach (pathway 3)
In conclusion, initially inspired by a standard analytical
protocol used for determining the glucosinolate content in
plants, we have explored two approaches for the synthesis
of thioimidate N-oxides. The halocyclization studied re-
vealed some stability problems, when using complex (glu-
cosyl 6b, 6c) precursors. This could be overcome by using
a D-ribose-derived template, from which a fairly stable
TIO 10 could be readily prepared.
The intramolecular nucleophilic substitution approach
also proved inadequate when applied to simple frame-
works, whereas the use of a carbohydrate scaffold com-
bined with an application of the Mitsunobu methodology
proved most efficient in yielding TIO 21. We have thus
The above synthetic approach to TIO was further extend-
ed to carbohydrate-type templates, to increase the ring-
Synlett 2010, No. 5, 725–728 © Thieme Stuttgart · New York

728
J. Schleiss et al.
LETTER
at r.t. for 4 h. After cooling to –78 °C, ethanethiol (1.33 mL,
18 mmol) then Et3N (2.5 mL, 18 mmol) were added
dropwise. The reaction mixture was slowly allowed to reach
r.t. and stirred for a further 12 h. After hydrolysis (H2O, 100
mL) and extraction with CH2Cl2 (3 × 100 mL), the combined
organic phases were washed with brine (3 × 50 mL), dried
over MgSO4, filtered, and then evaporated. Purification of
the residue over silica gel (PE–EtOAc, 7:3) afforded the 4-
O-silylated derivative of 24 as an oil. This intermediate was
dissolved in THF (20 mL) and reacted with TBAT (1.2
equiv) at r.t. for 14 h. H2O (20 mL) was then added and the
mixture was extracted with CH2Cl2 (3 × 30 mL). The
combined organic phases were washed with brine, dried over
MgSO4, then evaporated under reduced pressure. S-Ethyl
2,3-O-isopropylidene-L-erythronimidothioate (19) was
isolated as a colorless solid (1.2 g, 84% yield) by silica gel
flash chromatography (PE–EtOAc, 1:1). Mp 40–41 °C;
Rf = 0.18 (PE–EtOAc, 1:1); [α]D20 –54 (c 1.0, CHCl3). IR
(KBr): 3356, 2992, 1688, 1451, 1375, 1269, 1223, 1208,
1056, 998, 895, 859, 806, 740 cm–1. 1H NMR (400 MHz,
CDCl3): δ = 1.33 (t, 3 H, J = 7.6 Hz, CH3CH2), 1.40 and 1.53
[2 s, 6 H, C(CH3)2], 3.11 (m, 2 H, CH3CH2), 3.68 (m, 1 H,
H-4b), 3.79 (m, 1 H, H-4a), 4.40 (dd, 1 H, J2,3 = 6.0 Hz,
J3,4 = 4.4 Hz, H-3), 4.89 (d, 1 H, J2,3 = 6.0 Hz, H-2), 9.40 (s,
1 H, NOH). 13C NMR (100 MHz, CDCl3) : δ = 15.1
(CH3CH2), 25.6 (CH3CH2), 25.4, 27.3 [C(CH3)2], 61.7 (C-4),
76.9 (C-2), 78.6 (C-3), 109.5 (C(CH3)2), 150.8 (C=N). MS
(IS): m/z = 236.0 [M + H]+. ESI-HRMS: m/z [M + H]+ calcd
for C9H18NO4S: 236.0957; found: 236.0968.
disclosed two original approaches to introduce the rare
thioimidate N-oxide function, which is under current
study in our laboratory.17
Acknowledgment
J.S. thanks the MENRT for a fellowship; D.C. thanks the French
Embassy in Lithuania for a fellowship.
References and Notes
(1) Fahey, J. W.; Zalcmann, A. T.; Talalay, P. Phytochemistry
2001, 56, 5.
(2) (a) Valgimigli, L.; Iori, R. Environ. Mol. Mutagen. 2009, 50,
222. (b) Mawson, R.; Heaney, R. K.; Zdunczyk, Z.;
Kozłowska, H. Nahrung 1995, 39, 21.
(3) Purified arylsulfatase (E.C.3.1.6.1) from Helix pomatia is
currently used: (a) EEC Regulation No. 1864/90, Enclosure
VIII ; Offic. J. Eur. Commun.; 1990, L170: 27 (b)Wathelet,
J.-P.; Iori, R.; Leoni, O.; Rollin, P.; Quinsac, A.; Palmieri, S.
Agroindustria 2004, 3, 257.
(4) (a) Iori, R.; Barillari, J.; Gallienne, E.; Bilardo, C.;
Tatibouët, A.; Rollin, P. Tetrahedron Lett. 2009, 50, 3302.
(b) Coates, R. M.; Firsan, S. J. J. Org. Chem. 1986, 51, 5198.
(5) (a) Grigg, R.; Hadjisoteriou, M.; Kennewell, P.; Markandu,
J.; Thornton-Pett, M. J. Chem. Soc., Chem. Commun. 1992,
1388. (b) Grigg, R.; Hadjisoteriou, M.; Kennewell, P.;
Markandu, J.; Thornton-Pett, M. J. Chem. Soc., Chem.
Commun. 1993, 1340.
(16) Preparation of the Thioimidate N-Oxide 21
Ph3P (55 mg, 0.21 mmol) was added to a solution of DEAD
(40% in toluene, 95 μL, 0.21 mmol) in THF (10 mL). After
10 min of stirring, the thiohydroximate 19 (50 mg, 0.21
mmol) was added, and the reaction mixture was kept at
reflux overnight. After hydrolysis (H2O, 20 mL) and
extraction with CH2Cl2 (3 × 30 mL), the combined organic
phases were dried over MgSO4, filtered, and then
(6) (a) Jäger, V.; Bierer, L.; Dong, H.-Q.; Palmer, A. M.; Shaw,
D.; Frey, W. J. Heterocycl. Chem. 2000, 37, 455. (b) Gulla,
M.; Bierer, L.; Schmidt, S.; Redcliffe, L.; Jäger, V. Z.
Naturforsch., B: Chem. Sci. 2006, 61, 471.
(7) Revuelta, J.; Cicchi, S.; Goti, A.; Brandi, A. Synthesis 2007,
485; and references cited therein.
(8) Bourderioux, A.; Lefoix, M.; Gueyrard, D.; Tatibouët, A.;
Cottaz, S.; Arzt, S.; Burmeister, W. P.; Rollin, P. Org.
Biomol. Chem. 2005, 3, 1872.
(9) (a) Lohse-Fraedel, N.; Carreira, E. M. Org. Lett. 2005, 7,
2011. (b) Becker, N.; Carreira, E. M. Org. Lett. 2007, 9,
3857.
(10) (a) Rowley, M.; Leeson, P. D.; Williams, B. J.; Moore, K.
W.; Baker, R. Tetrahedron 1992, 48, 3357. (b) Elworthy,
ST. R.; Roepel, M. G.; Smith, D. B. WO 03/007941, 2003.
(c) Cicchi, S.; Corsi, M.; Brandi, A.; Goti, A. J. Org. Chem.
2002, 67, 1678.
(11) Muri, D.; Bode, J. W.; Carreira, E. M. Org. Lett. 2000, 2,
539.
(12) Jung, M. E.; Gervay, J. J. Am. Chem. Soc. 1991, 113, 224.
(13) Argyropoulos, N. G.; Panagiotidis, T. D.; Gallos, J. K.
Tetrahedron: Asymmetry 2006, 17, 829.
(14) Cicchi, S.; Marradi, M.; Vogel, P.; Goti, A. J. Org. Chem.
2006, 71, 1614.
(15) Preparation of the Thiohydroximate 19
NCS (0.88 g, 6.6 mmol) was added to a solution of aldoxime
18 (1.75 g, 6 mmol) in DMF (5 mL), and the mixture was left
evaporated. (3S,4S)-2-Ethylsulfanyl-3,4-isopropylidene-
dioxy-3,4-dihydro-5H-pyrrole-1-oxide (21) was isolated as
a colorless solid (45 mg, 98% yield) after silica gel flash
chromatography (EtOAc). Mp 145–150 °C; Rf = 0.5
(EtOAc); [α]D20 –136.5 (c 1.0, MeOH). IR (neat): 1570,
1422, 1383, 1261, 1234, 1204, 1154, 1076, 1021, 864, 836,
706, 663 cm–1. 1H NMR (400 MHz, CDCl3): δ = 1.37 (t, 3 H,
J = 7.6 Hz, CH3CH2), 1.39 and 1.44 [2 s, 6 H, C(CH3)2], 3.13
(q, 2 H, J = 7.6 Hz, CH3CH2), 4.06 (dt, 1 H, 2J5b,5a = 14.7 Hz,
J5b,4 = 5J5b,3 = 1.2 Hz, H-5b), 4.13 (dd, 1 H, 2J5b,5a = 14.7 Hz,
J5a,4 = 5.3 Hz, H-5a), 4.90 (ddd, 1 H, J4,3 = 6.5 Hz, J4,5a = 5.3
Hz, J4,5b = 1.4 Hz, H-4), 5.34 (d, 1 H, J4,3 = 6.5 Hz, H-3).
13C NMR (100 MHz, CDCl3): δ = 15.6 (CH3CH2), 23.4
(CH3CH2), 26.0, 27.2 [C(CH3)2], 66.4 (C-5), 73,1 (C-4), 81.7
(C-3); 112.8 [C(CH3)2], 144.0 (C-2). MS (IS): m/z = 218.0
[M + H]+. ESI-HRMS: m/z [M + H]+ calcd for C9H16NO3S:
218.0851; found: 218.0841.
(17) Schleiss, J.; Rollin, P.; Tatibouët, A. Angew. Chem. Int. Ed.
2010, 49, 577.
Synlett 2010, No. 5, 725–728 © Thieme Stuttgart · New York

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not
be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.