Aminocyclopropanes as precursors of endoperoxides with antimalarial activity

Article abstract of DOI:10.1039/c0ob00308e

This contribution describes the synthesis of several novel bicyclic α-amino endoperoxides, including CF₃-substituted compounds, prepared by the aerobic electrochemical oxidation of a family of bicyclic aminocyclopropanes. These, in turn, are re

Full text of DOI:10.1039/c0ob00308e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Aminocyclopropanes as precursors of endoperoxides with antimalarial
activity†
Claire Madelaine,a Olivier Buriez,*b Benoˆıt Crousse,c Isabelle Florent,d Philippe Grellier,d Pascal Retailleaua
and Yvan Six*a,e
Received 23rd June 2010, Accepted 3rd September 2010
DOI: 10.1039/c0ob00308e
This contribution describes the synthesis of several novel bicyclic a-amino endoperoxides, including
CF3-substituted compounds, prepared by the aerobic electrochemical oxidation of a family of bicyclic
aminocyclopropanes. These, in turn, are readily synthesised by a titanium-mediated intramolecular
cyclopropanation process (Kulinkovich–de Meijere reaction), starting from N-alkenyl amides that
contain a vic-disubstituted double bond, with high diastereoselectivity. An evaluation of the biological
activities of several of the molecules produced, against the parasite Plasmodium falciparum, is also
presented.
1. Introduction
In the context of investigations on aminocyclopropanes prepared
using the Kulinkovich–de Meijere reaction,1 we recently showed
that a-amino endoperoxide diastereoisomers 1a exhibit antimalar-
ial activity.2 There was a distinct possibility that this property
stemmed from the a-heteroatom endoperoxide moiety. Indeed,
this feature is present in established drugs, such as artemisinin and
artemether, as well as in much simpler active compounds, and the
peroxide function has been shown to play a critical role in their
activities against Plasmodium falciparum, the parasite responsible
for the disease (Fig. 1).3
We wanted to establish whether other a-amino endoperoxides
of general structure 1 could also display antiplasmodial activity,
with a view to finding more potent compounds than 1a. Our efforts
to prepare a set of new peroxides, as well as the results obtained
with respect to their antimalarial activities, are described in the
present article.
Fig. 1 Endoperoxides displaying activity against Plasmodium falciparum.
aCentre de Recherche de Gif, Institut de Chimie des Substances Naturelles,
2. Preparation of the starting materials
CNRS, Avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France
b
´
Ecole Normale Supe´rieure, De´partement de Chimie, UMR 8640 CNRS-
As initially reported by Wimalasena, a-amino endoperoxides can
be obtained by the aerobic oxidation of aminocyclopropanes.4 Our
group has recently shown that electrosynthesis under aerobic con-
ditions represents a particularly general and convenient method
for this type of transformation.2 We thus based our synthetic plan
on this process.
The aminocyclopropane precursors could be accessed by the
intramolecular Kulinkovich–de Meijere reaction,1,5 starting from
N-alkenyl amides. We were especially attracted by substrates of
the formula 3 (Scheme 1), featuring both a substituent R2 at the
homoallylic position a to the nitrogen atom, and a substituent
R3 at the olefin moiety. We anticipated that their intramolecu-
lar cyclopropanations would raise interesting diastereoselectivity
problems (vide infra).
ENS-UPMC, F75231 Paris, France
cLaboratoire BioCIS-CNRS, Faculte´ de Pharmacie, Universite´ Paris-Sud
11, rue Jean-Baptiste Cle´ment, F-92296 Chaˆtenay-Malabry, France
dMuse´um National d¢Histoire Naturelle, FRE 3206 CNRS, 61 rue Buffon,
75231 Paris Cedex 05, France
ePresent address: De´partement de Chimie, UMR 7652, CNRS/Ecole
Polytechnique, 91128 Palaiseau, France. E-mail: six@dcso.polytechnique.fr;
Fax: +33(0)1 6933 5972; Tel: +33(0)1 6933 5979
† Electronic supplementary information (ESI) available: Preparation of
the starting amides 3c–j and 3m. Method used for the analysis, by NMR
spectroscopy, of the crude products of the intramolecular Kulinkovich–de
Meijere reactions, as well as the crude oxidation products obtained from
the bicyclic aminocyclopropanes. Mechanistic considerations. Additional
results. NMR spectra of compounds (E)-3c, (Z)-3c, (E)-3d, (Z)-3d, 3e,
(E)-3f, (Z)-3f, (R*,R*)-3g, (R*,S*)-3g, 3h (E/Z 84 : 16), 3i, 3j, 3m, endo–
cis-2c, exo–cis-2c, endo–cis-2d, exo–cis-2d, endo-2f (cis/trans 79 : 21), endo-
2h, endo–cis-1c, exo–cis-1c, endo–cis-1d, exo–cis-1d, exo-1h, endo-1k, 6.
CCDC reference number 782249. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/c0ob00308e
We decided that some of the new molecules would bear a trifluo-
romethyl group as the R2 substituent, because the introduction of
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5591–5601 | 5591
©

View Article Online
Scheme 1 Plan for the synthesis of a-amino endoperoxides 1.
Scheme 2 Preparation of N-alkenyl amides 3c, 3d and 3f. CQ =
2,6-dichloro-1,4-benzoquinone.
a CF3 group in biologically active compounds can deeply influence
their properties,6 and CF3-substituted artemisinin derivatives have
been shown to be much more active and metabolically stable than
the non-fluorinated counterparts.7
The choice of the R3 substituent was limited: intramolecular
Kulinkovich–de Meijere reactions perform generally poorly with
substrates that contain a disubstituted double bond.8 However,
previous work, carried out with simpler substrates (R2 = H; R3
π H), has established that R3 groups fitted with a coordinating
function, such as a benzyloxy or a methoxy group, located at a
suitable distance from the C C double bond, can overcome this
impediment.9
Scheme 3 Preparation of the N-alkenyl amide 3g by homo-cross
metathesis.
With these requirements, the new N-alkenyl amides 3c, 3d
and 3f were synthesised using cross-metathesis reactions from
the corresponding known simpler molecules 3b and 3e,10 for
which the carbon–carbon double bond is monosubstituted (R3 =
H) (Scheme 2). The two diastereoisomeric diamides (R*,R*)-3g
(chiral) and (R*,S*)-3g (not chiral) were produced by homo-cross
metathesis of 3b (Scheme 3). Their configurations were assigned
unambiguously on the basis of an X-ray crystal structure of the
(R*,R*) diastereoisomer (Fig. 2).‡
For comparison purposes, compound 3h, that is identical to 3c
except for the absence of the CF3 group, was prepared as well
(Scheme 4). Finally, acetamides 3i and 3j (Fig. 3) were synthesised
‡ Single crystals of (R*,S*)-3g were glued on the top of a glass pin and
transferred to the goniometer head of an Enraf-Nonius kappa CCD
diffractometer using a graphite-monochromated Mo-Ka radiation (l =
˚
0.7107 A). The structure was solved by conventional direct methods
Fig. 2 X-Ray crystal structure of diamide (R*,R*)-3g. The (S,S) absolute
configuration is displayed arbitrarily. Displacement ellipsoids are shown
at the 30% probability level.‡.
and refined by full-matrix least-squares on F2. All non-hydrogen atoms
were refined anisotropically whereas hydrogen atoms were located from
difference Fourier maps but refined as a riding model with Uiso set to
1.5 (methyl) and 1.2 (other) times the equivalent isotropic displacement
parameter of the parent atoms.Crystal data. C26H28F6N2O4, M = 546.50,
˚
orthorhombic, a = 11.001(2), b = 11.945(2), c = 20.385(5) A, V =
3
-3
-1
˚
2678.7(9) A , T = 293 K, Dc = 1.355 g cm , F(000) = 1136, m = 0.119 mm ,
space group P212121 (no. 19), Z = 4, 19854 reflections measured, (1.98◦ £ q
£ 20.80◦), -10 £ h £ 10, -11 £ k £ 11, -20 £ l £ 20) 1614 unique (Friedel pairs
merged) (Rint = 0.018), which were used in all calculations. The final wR(F2)
was 0.096 (all data). R1 = 0.041 and wR2 = 0.096 for all reflections, R1 =
0.035 and wR2 = 0.091 for 1447 observed reflections [I > 2s(I)], refining
348 parameters and no restraints. Goodness-of-fit on F2: S = 1.044, final
Scheme 4 Preparation of the N-alkenyl amide 3h.
-3
˚
electron density between -0.126 and 0.130 e A . The programs used in
crystallographic study were as follows: data collection, COLLECT;23 cell
refinement, DENZO24 and COLLECT; data reduction, SCALEPACK24
and COLLECT; program used to solve structure, SHELXS97;25 program
used to refine structure, SHELXL97;25 molecular graphics, ORTEP within
PLATON.26
in order to ascertain whether gem-disubstituted alkenes could
be suitable substrates for intramolecular Kulinkovich–de Meijere
reactions.
5592 | Org. Biomol. Chem., 2010, 8, 5591–5601
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Fig. 3 Structures of N-alkenylamides 3i and 3j.
3. Aminocyclopropanation reactions
In principle, the aminocyclopropanation reactions of substrates 3
are expected to provide four diastereoisomers of the products 2,
when R2 and R3 differ from H (Scheme 1). However, preliminary
reports established that simpler substrates lacking the R2 group
(R2 = H; R3 π H) are converted into the corresponding bicyclic
aminocyclopropanes in a diastereospecific fashion.9,11 The relative
configuration of the carbon atoms 4 and 5 is fully controlled by
the configuration (E or Z) of the olefin moiety of the starting
material (the relative configuration of the carbon atoms 3 and 5
is fixed because the ring junction must be cis). This is rationalised
by a mechanism where the first elementary steps (epititanation
of the alkene moiety, and 1,2-insertion of the carbonyl group
in the resulting titanacyclopropane) proceed with retention, and
the final cyclopropane-ring closing event occurs with inversion of
configuration: the rotation of a C–C bond takes place in such a
way that the final zwitterionic intermediate adopts a M-shaped (or
W-shaped) conformation (Scheme 5).11,12
Scheme 6 Stereochemical aspects of the intramolecular Kulinkovich–de
Meijere reactions of substrates 3, with R2 π H.
As far as the relative configuration of the angular methyl group
and the R3 substituent is concerned, this should still be valid in
the cases where an extra substituent R2 is added. It was thus
expected that single diastereoisomers of fully substituted alkenyl
amides 3 would lead to mixtures of only two diastereoisomers of
the corresponding aminocyclopropanes 2, with full control of the
relative configuration at C-4 and C-5 (Scheme 6). In addition, we
were hoping that substantial control of the relative configurations
of centres C-1 and C-5 would operate. Indeed, a preliminary study
conducted with monosubstituted olefin substrates (R2 π H; R3 = H)
had shown the preferred formation of the cis diastereoisomers,13
with the best diastereoselectivities being observed using diethyl
ether as the solvent, rather than THF.10
can be drawn for the substrates fitted with both R2 and R3
substituents.
(1) The relative configuration (endo or exo)13 at the centres C-4
and C-5 of the aminocyclopropane product 2 is dictated by the
configuration (E or Z) of the carbon–carbon double bond of the
reactant 3, regardless of other factors (entries 2–5 and 7).
(2) The control of the relative configuration (cis or trans)13 at the
centres C-1 and C-5 of 2 is analogous to that previously reported
for substrates where R3 = H,10 patently favouring the cis products,
but is modulated by the configuration of the carbon–carbon
double bond of 3. Namely, compared with simple substrates for
which R3 = H, an erosion of the diastereoselectivity is observed
starting from (E) compounds (entry 1 vs. entries 2 and 4; entry 6 vs.
entry 7), and an enhancement of the diastereoselectivity starting
from (Z) alkenes (entry 1 vs. entries 3 and 5).
The new aminocyclopropanation reactions were thus performed
in diethyl ether. The results obtained with compounds 3c, 3d,
3f and 3g are presented in Table 1, along with those already
obtained with the monosubstituted olefins 3b and 3e.10 They are
in agreement with our expectations, and the following conclusions
Scheme 5 Mechanistic aspects of the stereochemical fate of the intramolecular Kulinkovich–de Meijere reactions of substrates 3, with R2 = H.
This journal is The Royal Society of Chemistry 2010 Org. Biomol. Chem., 2010, 8, 5591–5601 | 5593
©

View Article Online
Table 1 Intramolecular Kulinkovich–de Meijere reactions of N-alkenyl amides 3b–ja
Entry
Alkenyl amide
Aminocyclopropane products (diastereoselectivity)13
Yieldb
110
2
3
3b
cis-2b + trans-2b (89 : 11)
57%
58%
49%
54%
45%
71%
53%
(E)-3c
(Z)-3c
(E)-3d
(Z)-3d
3e
endo-cis-2c + endo-trans-2c (76 : 24)
exo-cis-2c + exo-trans-2c (93 : 7)
endo-cis-2d + endo-trans-2d (79 : 21)
exo-cis-2d + exo-trans-2d (90 : 10)
cis-2e + trans-2e (82 : 18)
4
5
610
7
(E)-3f
(R*,R*)-3g
(R*,S*)-3g
3h (E/Z 84 : 16)
3i
endo-cis-2f + endo-trans-2f (79 : 21)
no reaction
complex mixture, containing essentially the starting material
endo-2h + exo-2h (79 : 21)
complex mixture
8
9
10
11
12
66%
3j
no reaction
a Reactions performed in Et2O; reagents: Ti(OiPr)4 (1.5 equiv), cC5H9MgCl (4.0 equiv). b Yield estimated by 1H NMR of the crude product.15
Table 2 Oxidation potentials of bicyclic aminocyclopropanes 2
Entry
Aminocyclopropane
E(Ox1)a/V
E(Ox2)a/V
12
2
3
4
5
6
7
exo-2a
+0.67
+0.77
+0.82
+0.80
+0.81
+0.53
+0.61
+1.08
+1.08
+1.10
+1.07
+1.08
+0.82
+1.05
endo-cis-2c
exo-cis-2c
endo-cis-2d
exo-cis-2d
endo-2h
endo-2k9
a Oxidation potential vs. SCE, measured by cyclic voltammetry at 0.2 V s-1 in MeCN as the solvent.
This effect, as well as the stereochemical preference for the
formation of the cis diastereoisomers is discussed in more detail
in the Supporting Information.
was also observed in the presence of air. The first wave, located
between +0.61 and +0.82 V vs. the saturated calomel electrode
(SCE), was assigned to the oxidation of the substrates 2 into
the corresponding aminium cation radicals. The second oxidation
wave, only observed under aerobic conditions, corresponds to the
oxidation of the endoperoxides 1 formed in the diffusion layer,
after reaction with oxygen of the electrogenerated aminium cation
radicals at the potential value of the wave Ox1. Importantly, the
higher potential value of Ox2 as compared to Ox1 in all cases
supports an autocatalytic process, where the cation radical of
the product 1 can oxidise the starting aminocyclopropane 2.2 As
expected, close Ox1 and Ox2 potential values were obtained for all
the diastereoisomers of the starting compounds 2c and 2d, showing
little influence of the remote OMe and OPMB groups, or of the
stereochemistry of the substrates. Similarly, the distant OMe and
OTBS groups have virtually no influence on the values of both Ox1
and Ox2 (compare 2a with 2k). On the contrary, the presence of
an electron-donating group (OMe) on the aromatic ring attached
to the nitrogen atom facilitates the oxidation of both the starting
aminocyclopropane 2 and the endoperoxide 1 (compare 2h with
2k). Conversely, the presence of an electron-withdrawing group
(CF3) in the neighbourhood of the aminocyclopropane system
renders the oxidation processes more difficult (compare 2c with
2h).
The result obtained with the N-benzyl molecule 3f is remark-
able: so far, the reactions of substrates with a disubstituted alkenyl
moiety had been limited to N-aryl compounds, and all the N-alkyl
derivatives that we had studied previously, with R2 = H, had given
yields lower than 20%.9,11,14 Thus, this limitation is shown to be
overcome with the help of the Thorpe–Ingold effect exercised by
an R2 substituent, such as a phenyl group, and products 2f are
obtained in a comparatively satisfactory 53% yield (entry 7).
To account for the little reactivity displayed by the diamides
(R*,R*)-3g and (R*,S*)-3g (entries 8 and 9), a distinct possibility
is the formation of stable titanacyclopropane complexes from these
substrates.15
Starting from the amides 3i and 3j, the corresponding aminocy-
clopropanes could not be obtained, despite the application of
various reaction conditions.15 This can be readily explained by the
difficulty of the ligand exchange elementary step with such gem-
disubstituted olefins, although a related example, that involves an
intramolecular alkene-nitrile coupling, is known.16
4. Oxidation reactions
Following our previously reported approach,2 the oxidation poten-
tials of the aminocyclopropane compounds were measured using
cyclic voltammetry (Table 2). As we had observed before, the cyclic
voltammograms of the aminocyclopropanes 2 exhibited a single
oxidation wave (Ox1) under argon, whereas a second wave (Ox2)
The cyclic voltammetry results were used for the selective
preparation of the endoperoxides by electrosynthesis (Table 3).
Indeed, the preparative-scale electrolyses were carried out in
divided cells, at constant potential values corresponding roughly
to the peak potentials of Ox1. This control of potential is of high
5594 | Org. Biomol. Chem., 2010, 8, 5591–5601
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Table 3 Electrochemical aerobic oxidation of aminocyclopropanes 2
Entry
Substrate [potential, reaction time]
Endoperoxide product(s) (diastereoselectivity)13
Yielda
12
2
3
4
5
6
7
exo-2a [+0.60 V, 90 min]
endo-1a + exo-1a (55 : 45)
73%
67%
76%
75%
70%
59%
58%
endo-2c (cis/trans 94 : 6) [+0.80 V, 15 min]
exo-2c (cis/trans 92 : 8) [+0.80–0.85 V, 20 min]
endo-2d (cis/trans >95 : 5) [+0.80 V, 15 min]
exo-2d (cis/trans 89 : 11) [+0.80–0.85 V, 20 min]
endo-2h [+0.52–0.57 V, 50 min]
endo-cis-1c + exo-cis-1c (48 : 52)
endo-cis-1c + exo-cis-1c (52 : 48)
endo-cis-1d + exo-cis-1d (60 : 40)
endo-cis-1d + exo-cis-1d (60 : 40)
endo-1h + exo-1h (50 : 50)
endo-2k [+0.65 V, 30 min]
endo-1k + exo-1k (50 : 50)
a Yield estimated by 1H NMR of the crude product.15
Table 4 In vitro activities against chloroquine-resistant FcB1 strains of
importance to avoid the possible oxidation of the endoperoxide
products 1 prepared in situ, the potential values of which are
higher by about 300 mV. Under these conditions, the aminium
cation radicals could thus be generated cleanly and selectively,
to react with the oxygen introduced by bubbling air through
the anodic compartment. The electrolyses were followed by in
situ cyclic voltammetry, which uncovered the consumption of the
aminocyclopropanes 2 and the formation of the endoperoxides 1.
In some cases, the potential values of the electrolyses could be
adjusted to optimise the efficiency of the process. Typically, the
experiments were stopped when the intensity ratio I(Ox2)/I(Ox1)
was maximum i.e., after the quasi-total consumption of the
aminocyclopropanes 2.
Plasmodium falciparum
Entry
Compound
IC50/mMa
IC90/mMa
1
2
3
4
5
6
7
8
9
chloroquine
artemisinin
cis-2b10
0.11
0.012
0.21
0.021
58; 53
13; 15
23; 19
51; 48
29; 24
18; 16
26; 23
14; 13
34; 28
7.4; 8.1
12; 10
29; 27
14; 12
9.8; 9.7
12; 10
7.4; 7.3
endo-cis-2c
endo-cis-2d
exo-cis-2d
endo-2f (cis/trans 80 : 20)
endo-2h
endo-2k
10
2l18
All the desired endoperoxides could be produced in good yields
as attested by NMR spectroscopy of the crude products, but their
purification proved difficult. Pure samples could nonetheless be
obtained, but with considerable material loss whatever the method
used: standard flash chromatography (on silica gel or neutral
alumina gel) or HPLC. Similar problems were met by the group
of Wimalasena with their own peroxide compounds.4
11
cis-2m10
8.1; 7.4
15; 16
122
132
14
15
16
endo-1a
13
4.4
1.1; 1.0
4.7; 4.9
2.3; 1.8
22
8.3
1.8; 1.7
7.4; 9.6
4.1; 3.5
exo-1a
endo-cis-1c
exo-cis-1c
endo-cis-1d
These observations may be rationalised by invoking an equi-
librium, that is likely to take place between the endoperoxides 1
and the corresponding opened forms 4, resulting from carbon–
oxygen bond cleavage assisted by the lone pair of the nitrogen
atom (Scheme 7).17 Such zwitterionic species could then lead
to the formation of hydroperoxide enamines 5 by inter- or
intra-molecular proton displacements. Various decomposition
pathways may then operate. Among them, a plausible dehydration
a In most cases, two independent experiments were performed, and the
results are separated by semicolons.
to produce a vinylogous amide is supported by the formation
of compounds 6 and 7 in the course of the purification of
diastereoisomeric endoperoxides 1k by flash chromatography. It
is likely that these two regioisomeric molecules could equilibrate,
under these conditions, via the aminodiketone 8.
5. Biological assays
The pure samples of endoperoxides were put under scrutiny
in biological assays, aimed at determining their in vitro activ-
ities against chloroquine-resistant FcB1 strains of Plasmodium
falciparum. Purified diastereoisomers of the bicyclic aminocy-
clopropane compounds 2c, 2d, 2f and 2h were also tested, as
well as other aminocyclopropanes 2b, 2k, 2l and 2m, that had
been previously prepared in our laboratory without having been
evaluated (Table 4).
Interestingly, all of the aminocyclopropanes studied displayed
moderate but significant antiplasmodial activity. At first glance,
this observation might be attributed to their oxidation, in vitro, to
the corresponding peroxides. Indeed, the ranking of the biological
activities of several compounds is fully consistent with their
relative propensities to undergo oxidation: 2h > 2k (Table 4, entries
Scheme 7 Proposed mechanism for the formation of the vinylogous
amides 6 and 7.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5591–5601 | 5595
©

View Article Online
8 and 9, and Table 2 entries 6 and 7), 2m > 2b (Table 4, entry 11
vs. entry 3),19 and 2c, 2d > 2b (entries 4–6 vs. entry 3). However,
explaining the biological activity of the aminocyclopropanes is
probably not so straightforward: (i) the antiplasmodial tests were
performed under oxygen-deficient atmosphere (see the experi-
mental part), which obviously should disfavour the formation of
endoperoxides; (ii) endo-cis-2d and exo-cis-2d should be oxidised
to the same mixture of diastereoisomeric endoperoxide products
endo-cis-1d and exo-cis-1d; and yet they display significantly
different activities against Plasmodium falciparum (entries 4 and 5);
(iii) with a N-benzyl substituent, compound 2f should be especially
difficult to oxidise; yet it does not fall amidst the least active
molecules (Table 4, entry 7).
In the case of the three new endoperoxides that were evaluated,
higher activities were observed than those of their aminocyclo-
propane counterparts. Compounds endo-cis-1c and endo-cis-1d
displayed the most potent effect (entries 14 and 16), with IC50s of
about 1 and 2 mM respectively. Both molecules are therefore more
active than the endoperoxides 1a. However and unfortunately,
these activities remain moderate, and none of the molecules tested
so far is sufficiently active to be investigated further.
chased from Alfa Aesar, distilled under reduced pressure and
stored under argon for several months. Cyclo-pentylmagnesium
chloride 2 M solution in diethyl ether was purchased from Sigma-
Aldrich or Fluka and titrated once a month according to a
previously reported method.20
The syntheses of compounds 1a,2 2a,9 2b, 2e,10 2k,9 2l,18 2m,10
3a,9 3b,10 3k9 and 3l18 have already been reported. Detailed
procedures and analytical data for compounds 3c–j and 3m are
included in the Supporting information of the present article.
Ti-mediated aminocyclopropanations, procedure A
Titanium(IV) iso-propoxide (1.50 equiv, 1.50 mmol, 444 mL)
is added to a solution of N-alkenyl amide 3 (1.00 equiv,
1.00 mmol) in freshly distilled Et2O (20.0 mL), followed by cyclo-
pentylmagnesium chloride (2.00 M in Et2O, 4.00 equiv, 4.00 mmol,
2.00 mL), dropwise at 20 ◦C. After 30 min of stirring, water
(1.00 mL) is added to the dark solution, which is exposed to air,
and stirring is continued until decolouration. The mixture is then
filtered through a 1 cm layer of celite on a fritted-disc funnel, that
is then rinsed with Et2O (2 ¥ 10 mL). The combined organic layers
are dried over sodium sulfate, filtered and concentrated under
reduced pressure to afford the crude product.
6. Conclusions
Ti-mediated aminocyclopropanations, procedure B
The Kulinkovich–de Meijere intramolecular reactions of N-alk-
3-enyl amides, that feature a vic-disubstituted C C double bond
and a substituent at the allylic position, a to the nitrogen atom, give
the corresponding bicyclic aminocyclopropanes in satisfactory
yields, even starting from substrates that are not aniline derivatives.
These transformations are highly diastereoselective, and diastere-
ospecific with respect to the configurations of the C C double
bonds of the substrates. Our electrochemical oxidation method
gives good results starting from these aminocyclopropanes, and
the corresponding bicyclic a-aminoendoperoxides, that are found
to be rather unstable, exhibit moderate antimalarial activity, with
IC50s down to 1 mM. The significant antimalarial activities of the
aminocyclopropanes themselves, disclosed in the course of this
study, is intriguing and raises questions with respect to their mode
of action.
Methyltitanium(IV) tri-iso-propoxide (1.50 equiv, 1.50 mmol,
359 mL) is added to a solution of N-alkenyl amide 3 (1.00 equiv,
1.00 mmol) in freshly distilled Et2O (20.0 mL), followed by cyclo-
pentylmagnesium chloride (2.00 M in Et2O, 4.00 equiv, 4.00 mmol,
2.00 mL), dropwise at 20 ◦C. After 30 min of stirring, water
(1.00 mL) is added to the dark solution, which is exposed to air,
and stirring is continued until decolouration. The mixture is then
filtered through a 1 cm layer of celite on a fritted-disc funnel, that
is then rinsed with Et2O (2 ¥ 10 mL). The combined organic layers
are dried over sodium sulfate, filtered and concentrated under
reduced pressure to afford the crude product.
(1R*,5R*,6S*)-6-(2-Methoxyethyl)-2-(4-methoxyphenyl)-1-
methyl-3-(trifluoromethyl)-2-azabicyclo[3.1.0]hexane (endo-2c)
We will devote our next efforts to the design and synthesis
of more stable and more biologically active endoperoxides. We
also intend to investigate their potential as synthons in organic
synthesis, since they are plausible precursors of iminium species.
This compound was prepared from (E)-3c (113 mg, 327 mmol) by
applying the general procedure A. Analysis of the crude product
1
(bright yellow oil, 89.4 mg) by H NMR spectroscopy15 revealed
the presence of two diastereoisomeric aminocyclopropane prod-
ucts endo-cis 2c (43.8%) and endo-trans-2c (14.1%), i.e. an overall
estimated yield of 58% and a cis/trans ratio of 76 : 24. Flash
chromatography (neutral alumina, activity II, AcOEt/heptane,
gradient from 0% to 20%) led to the isolation of pure endo-2c
(cis/trans 94 : 6, 32.2 mg, 97.7 mmol, 30%).
7. Experimental
NMR spectra were recorded with AM 300, AVANCE 300 (1H at
300 MHz, 13C at 75.5 MHz) and AVANCE 500 (1H at 500 MHz,
13C at 125.8 MHz) Bruker spectrometers. Chemical shifts are given
in ppm, referenced to the peak of tetramethylsilane, defined at d =
0.00 (1H NMR), or the solvent peak of CDCl3, defined at d =
77.0 (13C NMR). Infrared spectra were recorded with a Perkin-
Elmer BX FT-IR spectrometer. Flash column chromatography
was performed on SDS Chromagel silica gel 60 (35–70 mm).
Preparative HPLC was performed on a Waters SunFireTM 5 mm
C18 (19 ¥ 150 mm) column. All reactions were carried out under
argon, unless stated otherwise. Diethyl ether was purified using
a PureSolv solvent purification system (Innovative Technology
endo-cis-2c (1R*,3R*,5R*,6S*). Colourless oil;
n
max/cm-1
2932, 2831, 1509, 1275, 1261, 1242, 1177, 1154, 1118, 1038, 817;
dH(300 MHz; CDCl3; Me4Si) 0.99 (1 H, ddt, J 14.5, 9.5 and 6.5),
1.26 (1 H, ddd, J 9.5, 8.0 and 4.5), 1.54 (1 H, td, J 8.0 and 2.0),
1.54 (3 H, s), 1.70 (1 H, dtd, J = 14.5, 7.0 and 4.5), 2.03 (1 H, ddd,
J 14.5, 9.0 and 2.0), 2.32 (1 H, ddd, J 14.5, 8.0 and 1.5), 3.29 (3 H,
s), 3.31–3.41 (2 H, m), 3.74 (3 H, s), 4.02 (1 H, dqd, J 8.5, 7.0 and
1.5, CHCF3), 6.74 (4 H, AA¢BB¢ system,21 dA 6.70, dB 6.79, N 9.0,
L 9.0, K 5.5 (M could not be measured accurately)); dC(75.5 MHz;
R
Inc.). Titanium(IV) iso-propoxide (VERTECꢀ TIPT) was pur-
5596 | Org. Biomol. Chem., 2010, 8, 5591–5601
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
CDCl3; Me4Si) 20.8, 24.5, 24.7, 26.4, 31.5, 49.5, 55.7, 58.7, 69.0
(q, J2 30.0), 72.3, 114.1, 114.6, 126.5 (q, J1 284), 140.3, 152.1;
m/z (ES+) 204, 330 (MH+), 384, 420; m/z (ES+) 330.1692 (MH+
C17F3H23NO2 requires 330.1681).
J 9.0, 7.0 and 2.0, CHCF3), 4.38 (2 H, s), 6.64–6.92 (6 H, m),
7.17–7.32 (2 H, m); dC(75.5 MHz; CDCl3; Me4Si) 20.8, 24.5, 24.9,
26.4, 31.6, 49.5, 55.2, 55.7, 68.9 (q, J2 29.5), 69.6, 72.7, 113.8,
114.1, 114.6, 126.6 (q, J1 283), 129.2, 130.5, 140.3, 152.1, 159.2;
m/z (ES+) 436 (MH+), 458 (MNa+), 459, 526; m/z (ES+) 458.1911
(MNa+ C24F3H28NNaO3 requires 458.1919).
endo-trans-2c (1R*,3S*,5R*,6S*). dH(300 MHz; CDCl3;
Me4Si) characteristic signal 4.88 (1 H, ddq, J 9.0, 8.0 and 6.0,
CHCF3).
endo-trans-2d (1R*,3S*,5R*,6S*). dH(300 MHz; CDCl3;
Me4Si) characteristic signal 4.86 (1 H, ddq, J 9.0, 8.5 and 6.0,
CHCF3).
Compound endo-2d was also prepared using the general proce-
dure B from (E)-3d (34.4 mg, 76.2 mmol). Analysis of the crude
product (yellow oil, 30.0 mg) by 1H NMR spectroscopy15 revealed
the presence of the two diastereoisomeric aminocyclopropane
products endo-cis-2d (42.3%) and endo-trans-2d (10.9%), i.e. an
overall estimated yield of 53% and a cis/trans ratio of 80 : 20.
(1R*,5R*,6R*)-6-(2-Methoxyethyl)-2-(4-methoxyphenyl)-1-
methyl-3-(trifluoromethyl)-2-azabicyclo[3.1.0]hexane (exo-2c)
This compound was prepared from (Z)-3c (102 mg, 295 mmol) by
applying the general procedure A. Analysis of the crude product
(yellow oil, 103 mg) by 1H NMR spectroscopy15 revealed the pres-
ence of two diastereoisomeric aminocyclopropane products exo-
cis-2c (45.4%) and exo-trans-2c (3.5%), i.e. an overall estimated
yield of 49% and a cis/trans ratio of 93 : 7. Flash chromatography
(neutral alumina, activity II, AcOEt/heptane, gradient from 0%
to 20%) led to the isolation of pure exo-2c (cis/trans 92 : 8, 38.9 mg,
118 mmol, 40%).
(1R*,5R*,6R*)-6-(2-(4-Methoxybenzyloxy)ethyl)-2-(4-
methoxyphenyl)-1-methyl-3-(trifluoromethyl)-2-
azabicyclo[3.1.0]hexane (exo-2d)
exo-cis-2c (1R*,3R*,5R*,6R*). Colourless oil;
n
max/cm-1
This compound was prepared from (Z)-3d (121 mg, 268 mmol) by
applying the general procedure A. Analysis of the crude product
(orange yellow oil, 126 mg) by 1H NMR spectroscopy15 revealed
the presence of two diastereoisomeric aminocyclopropane prod-
ucts exo-cis-2d (40.9%) and exo-trans-2d (4.6%), i.e. an overall
estimated yield of 45% and a cis/trans ratio of 90 : 10. Flash
chromatography (neutral alumina, activity II, AcOEt/heptane,
gradient from 0% to 5%) led to the isolation of pure exo-2d
(cis/trans 89 : 11, 47.2 mg, 108 mmol, 40%).
2933, 2831, 1509, 1277, 1243, 1180, 1154, 1119, 1038, 817;
dH(300 MHz; CDCl3; Me4Si) 0.86 (1 H, td, J 7.0 and 4.5), 1.19
(1 H, ddd, J 7.0, 4.5 and 2.0), 1.47 (3 H, s), 1.67 (2 H, AB part
of an ABX2Y system, dA 1.61, dB 1.74, JAB 14.5, JAX 6.5, JAY 7.0,
JBX 6.5, JBY 7.0), 2.07 (1 H, ddd, J 14.0, 9.0 and 2.0), 2.43 (1 H,
ddd, J 14.0, 7.0 and 1.5), 3.38 (3 H, s), 3.51 (2 H, t, J 6.5), 3.76
(3 H, s), 3.95 (1 H, dqd, J 9.0, 7.5 and 1.5, CHCF3), 6.79 (4 H,
AA¢BB¢ system,21 dA 6.76, dB 6.83, N 9.0, L 9.0, K 6.0 (M could
not be measured accurately)); dC(75.5 MHz; CDCl3; Me4Si) 14.9,
29.7, 30.3, 31.0, 36.2, 50.5, 55.8, 58.6, 68.2 (q, J2 30.0), 72.2, 114.7,
115.1, 126.7 (q, J1 283), 140.6, 152.4; m/z (ES+) 204, 330 (MH+);
m/z (ES+) 330.1696 (MH+ C17F3H23NO2 requires 330.1681).
exo-cis-2d (1R*,3R*,5R*,6R*). Colourless oil;
n
max/cm-1
2935, 2905, 2857, 2833, 1508, 1277, 1242, 1177, 1154, 1127, 1097,
1034, 817; dH(300 MHz; CDCl3; Me4Si) 0.86 (1 H, td, J 7.0 and
4.5), 1.19 (1 H, ddd, J 7.0, 4.5 and 1.5), 1.47 (3 H, s), 1.70 (2
H, AB part of an ABX2Y system, dA 1.61, dB 1.80, JAB 14.0, JAX
6.5, JAY 7.5, JBX 6.5, JBY 7.5), 2.06 (1 H, ddd, J 14.0, 9.0 and
1.5), 2.42 (1 H, ddd, J 14.0, 7.5 and 1.5), 3.58 (2 H, t, J 6.5),
3.74 (3 H, s), 3.79 (3 H, s), 3.94 (1 H, dqd, J 9.0, 7.5 and 1.5,
CHCF3), 4.48 (2 H, s), 6.71–6.83 (4 H, m), 6.87 (2 H, m), 7.26 (2
H, m); dC(75.5 MHz; CDCl3; Me4Si) 14.9, 29.7, 30.4, 31.0, 36.3,
50.5, 55.2, 55.7, 68.2 (q, J2 29.5), 69.6, 72.8, 113.8, 114.7, 115.0,
126.7 (q, J1 282), 129.1, 130.4, 140.6, 152.3, 159.1; m/z (ES+) 432,
436 (MH+), 458 (MNa+), 459, 526; m/z (ES+) 458.1908 (MNa+
C24F3H28NNaO3requires 458.1919).
exo-trans-2c (1R*,3S*,5R*,6R*). dH(300 MHz; CDCl3;
Me4Si) characteristic signals 2.24 (1 H, ddd, J 14.0, 6.5 and 4.0),
2.60 (1 H, ddd, J 14.0, 10.5 and 7.0), 4.80 (1 H, dqd, J 10.5, 7.5
and 4.0, CHCF3).
(1R*,5R*,6S*)-6-(2-(4-Methoxybenzyloxy)ethyl)-2-(4-
methoxyphenyl)-1-methyl-3-(trifluoromethyl)-2-
azabicyclo[3.1.0]hexane (endo-2d)
This compound was prepared from (E)-3d (210 mg, 465 mmol) by
applying the general procedure A. Analysis of the crude product
(yellow oil, 213 mg) by 1H NMR spectroscopy15 revealed the
presence of two diastereoisomeric aminocyclopropane products
endo-cis-2d (42.6%) and endo-trans-2d (11.4%), i.e. an overall
estimated yield of 54% and a cis/trans ratio of 79 : 21. Flash
chromatography (neutral alumina, activity II, AcOEt/heptane,
gradient from 0% to 30%) led to the isolation of pure endo-2d
(cis/trans >95 : 5, 64.9 mg, 149 mmol, 32%).
exo-trans-2d (1R*,3S*,5R*,6R*). dH(300 MHz; CDCl3;
Me4Si) characteristic signals 2.58 (1 H, ddd, J 14.0, 10.5 and 7.0),
4.79 (1 H, dqd, J 10.5, 7.5 and 4.0, CHCF3).
(1R*,5R*,6S*)-2-Benzyl-6-(2-methoxyethyl)-1-methyl-3-phenyl-2-
azabicyclo[3.1.0]hexane (endo-2f)
endo-cis-2d (1R*,3R*,5R*,6S*). Colourless oil; nmax/cm-1
2934, 2856, 2837, 1509, 1274, 1241, 1177, 1155, 1120, 1098, 1034,
817; dH(300 MHz; CDCl3; Me4Si) 1.03 (1 H, ddt, J 14.5, 8.5 and
6.5), 1.26 (1 H, ddd, J 8.5, 8.0 and 5.0), 1.52 (1 H, td, J 8.0 and
2.0), 1.53 (3 H, s), 1.71 (1 H, dtd, J = 14.5, 7.0 and 5.0), 2.01
(1 H, ddd, J 14.5, 9.0 and 2.0), 2.29 (1 H, ddd, J 14.5, 8.0 and
2.0), 3.41 (2 H, m), 3.74 (3 H, s), 3.80 (3 H, s), 4.00 (1 H, dqd,
This compound was prepared from (E)-3f (145 mg, 430 mmol) by
applying the general procedure A. Analysis of the crude product
1
(orange oil, 133 mg) by H NMR spectroscopy15 revealed the
presence of two diastereoisomeric aminocyclopropane products
endo-cis-2f (42.1%) and endo-trans-2f (11.2%), i.e. an overall
estimated yield of 53% and a cis/trans ratio of 79 : 21. Flash
chromatography (neutral alumina, activity II, AcOEt/heptane,
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5591–5601 | 5597
©

View Article Online
gradient from 0% to 30%) led to the isolation of pure endo-2f
(cis/trans 80 : 20, 66.5 mg, 207 mmol, 48%).
Cyclic voltammetry and electrochemical aerobic oxidation of
aminocyclopropanes, procedure C
Tetrabutylammonium tetrafluoroborate (TBABF4) used as the
supporting electrolyte was prepared from NaBF4 (Acros), and
n-Bu4NHSO4 (Acros), recrystallised from ethyl acetate–hexane
(both from Acros), and dried at 60 ◦C.
endo-2f (cis/trans 80 : 20). Colourless oil; nmax/cm-1 2924,
2859, 2823, 1493, 1453, 1156, 1114, 1027, 965, 747, 697. m/z
(ES+) 322 (MH+), 323, 344 (MNa+); m/z (ES+) 322.2175 (MH+
C22H28NO requires 322.2171).
Cyclic voltammetry experiments were performed in acetonitrile,
at room temperature, under argon or in the presence of air, in a
three-electrode cell using an Autolab potentiostat (PGSTAT 20).
The reference electrode was a saturated calomel electrode (SCE-
Tacussel), which was separated from the solution by a bridge
compartment filled with the same solvent/supporting electrolyte
solution (0.1 M) as in the cell. The counter electrode was a 1 cm
gold wire (Goodfellow). A homemade platinum electrode (0.5 mm
diameter; Goodfellow) was used as the working electrode.
The preparative electrolyses were performed in a divided cell at
a constant potential value corresponding to the peak potential of
the first oxidation wave (Ox1). The working (anodic compartment)
and counter electrodes (cathodic compartment) were gold grids
(2 cm2; Goodfellow). The reference electrode was the same as that
used in the cyclic voltammetry experiments. Oxygen was brought
in by bubbling air into the anodic compartment. Typically, 100 to
200 mmol of aminocyclopropane 2 were introduced in the anodic
compartment containing the solvent MeCN (22 mL). At end of the
electrolyses, the solutions contained in the anodic compartment
were concentrated under vacuum, extracted with Et2O, filtered
and concentrated again to give the crude products, that were then
analysed by NMR spectroscopy.
endo-cis-2f (1R*,3R*,5R*,6S*). dH(300 MHz; CDCl3; Me4Si)
0.67 (1 H, td, J 8.5 and 5.5), 1.19 (1 H, dd, J 8.5 and 6.5), 1.28
(3 H, s), 1.72 (2 H, AB part of an ABXYZ system, dA 1.64, dB
1.81, JAB 14.0, JAX 8.5, JAY 7.5, JAZ 6.0, JBX 5.5, JBY 7.0 JBZ 7.5),
2.16 (2 H, AB part of an ABXY system, dA 2.01, dB 2.31, JAB
13.5, JAX 8.0, JAY 6.5, JBX 8.5, JBY 0.0), 3.37 (3 H, s), 3.40 (2 H,
AB part of an ABXY system, dA 3.38, dB 3.43, JAB 13.5, JAX 7.5,
JAY 7.0, JBX 6.0, JBY 7.5), 3.82 (1 H, dd, J 8.5 and 8.0, CHPh),
3.83 (2 H, AB system, dA 3.77, dB 3.90, JAB 14.0), 6.95–7.53 (10
H, m); dC(75.5 MHz; CDCl3; Me4Si) 23.1, 25.5, 26.1, 27.4, 35.8,
50.8, 54.4, 58.6, 70.8, 73.5, 126.3, 126.5, 127.2, 127.6, 127.9, 128.9,
140.5, 145.8.
endo-trans-2f (1R*,3S*,5R*,6S*). dH(300 MHz; CDCl3;
Me4Si) characteristic signals 0.79 (1 H, m), 1.23 (3 H, s), 2.31
(1 H, ddd, J 14.0, 7.5 and 7.0), 3.35 (3 H, s), 3.68 (2 H, AB
system, dA 3.65, dB 3.71, JAB 15.5), 4.51 (1 H, dd, J 11.0 and 7.0,
CHPh); dC(75.5 MHz; CDCl3; Me4Si) characteristic signals 19.1,
23.5, 25.1, 33.2, 33.9, 48.7, 51.5, 58.6, 73.0, 79.7, 126.2, 128.5,
141.6, 142.9.
(1R*,5R*)-6-(2-Methoxyethyl)-2-(4-methoxyphenyl)-1-methyl-2-
(3aR*,5R*,6aS*)-3-(2-Methoxyethyl)-6-(4-methoxyphenyl)-6a-
methyl-5-(trifluoromethyl)hexahydro-[1,2]dioxolo[3,4-b]pyrrole
(cis-1c)
azabicyclo[3.1.0]hexane (2h)
This compound was prepared from 3h (E/Z 84 : 16, 353 mg,
1.27 mmol) by applying the general procedure A. Analysis of the
crude product (brown oil, 348 mg) by 1H NMR spectroscopy15 re-
vealed the presence of two diastereoisomeric aminocyclopropane
products endo-2h (52.2%) and exo-2h (14.1%), i.e. an overall
estimated yield of 66% and an endo/exo ratio of 79 : 21.
Flash chromatography (neutral alumina, activity II,
AcOEt/heptane, gradient from 0% to 10%) led to the isolation of
pure endo-2h (63.7 mg, 244 mmol, 19%), and an exo/endo mixture
of 2h (exo/endo 70 : 30, 24.0 mg, 91.8 mmol, 7%).
This compound was prepared from endo-2c (cis/trans 94 : 6;
27.8 mg, 84.4 mmol) by applying the general procedure C (+0.80 V
during 15 min). Analysis of the crude product (orange oil, 51.8 mg)
by 1H NMR spectroscopy15 revealed the presence of the two
diastereoisomeric endoperoxide products endo-cis-1c (32.2%) and
exo-cis-1c (34.7%), i.e. an overall estimated yield of 67% and an
endo/exo ratio of 48 : 52. Flash chromatography (neutral alumina,
activity II, AcOEt/heptane, gradient from 0% to 20%) led to
the isolation of moderately pure peroxide endo-cis-1c (8.4 mg,
23.2 mmol, 27%). A pure sample was obtained after purification
by preparative HPLC [(H2O + 0.1% HCO2H)/(MeOH + 0.1%
HCO2H) 40 : 60, flow rate 17 mL min-1, UV detection 254 nm]
(1.0 mg, 2.7 mmol, 3%).
Compounds endo-cis-1c and exo-cis-1c were also obtained from
exo-2c (cis/trans 92 : 8; 40.4 mg, 123 mmol) using the general
procedure C (+0.80 V during 15 min and then +0.85 V during
5 min). Analysis of the crude product (yellow-orange oil, 74.8 mg)
by 1H NMR spectroscopy15 revealed the presence of the two
diastereoisomeric endoperoxide products endo-cis-1c (39.5%) and
exo-cis-1c (36.6%), i.e. an overall estimated yield of 76% and an
endo/exo ratio of 52 : 48. Flash chromatography (neutral alumina,
activity II, AcOEt/heptane, gradient from 10% to 50%) led to
the isolation of a mixture of the pure peroxides cis-1c (endo/exo
ratio of 59 : 41, 16.0 mg, 44.2 mmol, 36%). Further purification
by preparative HPLC [(H2O + 0.1% HCO2H)/(MeOH + 0.1%
endo-2h (1R*,3R*,6S*). Pale yellow oil; nmax/cm-1 2929, 2865,
2828, 1508, 1237, 1178, 1114, 1039, 816; dH(300 MHz; CDCl3;
Me4Si) 1.05–1.35 (2 H, m), 1.41 (1 H, td, J 7.5 and 1.0), 1.49 (3 H,
s), 1.59 (1 H, m), 1.88 (1 H, dddd, J 13.5, 10.0, 5.5 and 1.0), 2.30 (1
H, dddd, J 13.5, 10.0, 7.0 and 5.5), 3.05 (1 H, td, J 10.0 and 5.5),
3.29 (3 H, s), 3.27–3.48 (2 H, m), 3.73 (3 H, s), 4.03 (1 H, td, J 10.0
and 5.5), 6.69 (4 H, AA¢BB¢ system,21 dA 6.60, dB 6.78, N 9.0, L
8.5, K 5.5 (M could not be measured accurately)); dC(75.5 MHz;
CDCl3; Me4Si) 20.5, 22.7, 24.6, 28.3, 30.3, 46.6, 55.7, 56.3, 58.5,
72.7, 114.5, 114.8, 142.7, 151.0; m/z (ES+) 262 (MH+), 294, 330,
332; m/z (ES+) 262.1799 (MH+ C16H24NO2requires 262.1807).
exo-2h (1R*,5R*,6R*). dH(300 MHz; CDCl3; Me4Si) charac-
teristic signals 1.46 (3 H, s), 2.71 (1 H, td, J 9.5 and 8.5), 3.85 (1 H,
td, J 9.5 and 2.0); dC(75.5 MHz; CDCl3; Me4Si) 15.2, 24.2, 26.4,
30.1, 30.2, 46.6, 53.4, 55.6, 58.6, 72.6, 114.3, 118.3, 144.0, 152.6.
5598 | Org. Biomol. Chem., 2010, 8, 5591–5601
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
HCO2H) 40 : 60, flow rate 17 mL min-1, UV detection 254 nm]
yielded pure exo-cis-1c (retention time 23.2 min; 5.0 mg, 13.8 mmol,
11%).
6.5 and 3.0), 3.08 (1 H, ddd, J 9.0, 5.5 and 3.0), 3.50–3.66 (2 H,
m), 3.79 (3 H, s), 3.81 (3 H, s), 4.27 (1 H, ddq, J 8.0, 6.5 and 6.0,
CHCF3), 4.45 (2 H, AB system, dA 4.44, dB 4.47, JAB 11.5), 4.53
(1 H, td, J 6.5 and 5.5), 6.80–6.94 (4 H, m), 7.14–7.32 (4 H, m);
dC(125.8 MHz; CDCl3; Me4Si) 22.3, 24.3, 27.9, 55.3, 55.3, 56.8,
64.5 (q, J2 29.5), 66.8, 72.9, 82.2, 105.3, 113.8, 113.9, 125.3 (q,
J1 280), 129.3, 129.4, 130.1, 134.7, 157.7, 159.3. This compound
could not be characterised more extensively because of the small
amount isolated.
endo-cis-1c (3R*,3aS*,5S*,6aR*). Yellow oil; dH(500 MHz;
CDCl3; Me4Si) 1.49 (3 H, s), 1.88–1.96 (2 H, m), 2.05 (1 H, ddd, J
13.5, 9.5 and 8.0), 2.27 (1 H, ddd, J 13.5, 6.5 and 3.0), 3.11 (1 H,
ddd, J 9.5, 5.0 and 3.0), 3.36 (3 H, s), 3.45–3.56 (2 H, m), 3.79 (3 H,
s), 4.29 (1 H, ddq, J 8.0, 6.5 and 6.0, CHCF3), 4.52 (1 H, ddd, J 7.5,
6.0 and 5.0), 7.02 (4 H, AA¢BB¢ system,21 dA 6.85, dB 7.18, N 9.0,
(K, L and M could not be measured accurately)); dC(75.5 MHz;
CDCl3; Me4Si) characteristic signals 22.2, 27.7, 55.3, 56.8, 58.8,
64.5 (q, J2 30.0), 69.4, 82.2, 105.3, 113.8, 129.5, 134.7, 157.7. This
compound could not be characterised more extensively because of
the small amount isolated.
exo-cis-1d (3R*,3aR*,5R*,6aS*). dH(300 MHz; CDCl3;
Me4Si) 1.45 (3 H, s), 1.93 (2 H, AB part of an ABXYZ system,
dA 1.81, dB 2.05, JAB 14.0, JAX 9.0, JAY 5.5, JAZ 4.5, JBX 4.5, JBY
4.5 JBZ 9.0), 2.17 (2 H, AB part of an ABXY system, dA 2.11,
dB 2.22, JAB 13.0, JAX 6.5, JAY 2.0, JBX 9.5, JBY 8.5), 2.97 (1 H,
ddd, J 8.5, 2.0 and 1.5), 3.57 (2 H, AB part of an ABXY system,
dA 3.54, dB 3.61, JAB 9.5, JAX 5.5, JAY 4.5, JBX 9.0, JBY 4.5), 3.79
(3 H, s), 3.81 (3 H, s), 4.33 (1 H, ddd, J 9.0, 4.5 and 1.5), 4.42
(1 H, ddq, J 9.5, 6.5 and 5.5, CHCF3), 4.44 (2 H, AB system, dA
4.42, dB 4.46, JAB 11.5), 6.82–6.93 (4 H, m), 7.18–7.29 (4 H, m);
dC(125.8 MHz; CDCl3; Me4Si) 21.8, 30.3, 33.9, 55.3, 55.3, 59.3,
64.2 (q, J2 30.0), 66.2, 72.9, 87.3, 104.1, 113.8, 113.9, 125.1 (q,
J1 281), 129.4, 130.2, 130.3, 134.6, 158.0, 159.3. This compound
could not be characterised more extensively because of the small
amount isolated.
exo-cis-1c
(3R*,3aR*,5R*,6aS*). Colourless
crystals;
dH(300 MHz; CDCl3; Me4Si) 1.46 (3 H, s), 1.91 (2 H, AB part
of an ABXYZ system, dA 1.78, dB 2.03, JAB 14.5, JAX 9.0, JAY
5.5, JAZ 4.0, JBX 4.5, JBY 4.5 JBZ 9.5), 2.20 (2 H, AB part of an
ABXY system, dA 2.14, dB 2.25, JAB 13.0, JAX 6.0, JAY 2.0, JBX
9.5, JBY 8.5), 2.99 (1 H, ddd, J 8.5, 2.0 and 1.5), 3.45 (3 H, s),
3.50 (2 H, AB part of an ABXY system, dA 3.46, dB 3.53, JAB 9.5,
JAX 5.5, JAY 4.5, JBX 9.0, JBY 4.5), 3.79 (3 H, s), 4.32 (1 H, ddd,
J 9.5, 4.0 and 1.5), 4.43 (1 H, ddq, J 9.5, 6.0 and 5.5, CHCF3),
7.04 (4 H, AA¢BB¢ system,21 dA 6.86, dB 7.22, N 9.0, L 8.5, K 5.5
(M could not be measured accurately)); dC(125.8 MHz; CDCl3;
Me4Si) 21.8, 30.3, 33.8, 55.3, 58.8, 59.3, 64.2 (q, J2 30.0), 68.7,
87.1, 104.1, 113.8, 125.1 (q, J1 280), 130.3, 134.5, 158.0. This
compound could not be characterised more extensively because
of the small amount isolated.
(3aR*,6aS*)-3-(2-Methoxyethyl)-6-(4-methoxyphenyl)-6a-
methylhexahydro-[1,2]dioxolo[3,4-b]pyrrole (1h)
This compound was prepared from endo-2h (49.2 mg, 188 mmol)
by applying the general procedure C (+0.52 V during 20 min,
then + 0.57 V for 30 min). Analysis of the crude product (orange
oil, 49.0 mg) by 1H NMR spectroscopy15 revealed the presence of
the two diastereoisomeric endoperoxide products endo-1h (29.1%)
and exo-1h (29.6%), i.e. an overall estimated yield of 59% and
an endo/exo ratio of 50 : 50. Flash chromatography (silica gel,
AcOEt/heptane, gradient from 10% to 50%) led to the isolation
of pure exo-1h (9.1 mg, 31.0 mmol, 16%).
(3aR*,5R*,6aS*)-3-(2-(4-Methoxybenzyloxy)ethyl)-6-(4-
methoxyphenyl)-6a-methyl-5-(trifluoromethyl)hexahydro-
[1,2]dioxolo[3,4-b]pyrrole (cis-1d)
This compound was prepared from endo-2d (cis/trans >95 : 5;
23.0 mg, 52.8 mmol) by applying the general procedure C (+0.80 V
during 15 min). Analysis of the crude product (yellow oil)
by 1H NMR spectroscopy15 revealed the presence of the two
diastereoisomeric endoperoxide products endo-cis-1d (45.2%) and
exo-cis-1d (29.7%), i.e. an overall estimated yield of 75% and an
endo/exo ratio of 60 : 40.
endo-1h (3R*,3aS*,6aR*). dH(300 MHz; CDCl3; Me4Si) char-
acteristic signals 1.46 (3 H, s), 3.06 (1 H, td, J 8.0 and 5.5),
3.35 (3 H, s), 4.45 (1 H, ddd, J 7.5, 6.0 and 5.5), 6.88 (4
H, AA¢BB¢ system,50dA 6.81, dB 6.96, N 9.0, K 5.5 (L and M
could not be measured accurately)); dC(75.5 MHz; CDCl3; Me4Si)
characteristic signals 19.6, 23.8, 27.1, 50.0, 61.4, 69.7, 81.5, 103.7,
119.3, 138.5, 153.9.
Compounds endo-cis-1d and exo-cis-1d were also obtained from
exo-2d (cis/trans 89 : 11; 33.3 mg, 76.5 mmol) using the general
procedure C (+0.80 V during 15 min, and then +0.85 V during
1
5 min). Analysis of the crude product (green oil) by H NMR
spectroscopy15 revealed the presence of the two diastereoiso-
meric endoperoxide products endo-cis-1d (42.0%) and exo-cis-1d
(28.0%), i.e. an overall estimated yield of 70% and an endo/exo
ratio of 60 : 40.
exo-1h (3R*,3aR*,6aS*). Yellow oil; dH(300 MHz; CDCl3;
Me4Si) 1.52 (3 H, s), 1.92 (2 H, AB part of an ABXYZ system,
dA 1.79, dB 2.05, JAB 14.5, JAX 8.5, JAY 6.0, JAZ 4.5, JBX 5.0, JBY
4.5, JBZ 9.0), 2.09 (2 H, AB part of an ABXYZ system, dA 1.95,
dB 2.21, JAB 12.5, JAX 6.5, JAY 6.5, JAZ 4.5, JBX 7.0, JBY 6.5, JBZ
8.5), 2.94 (1 H, ddd, J 8.5, 4.5 and 2.5), 3.27–3.38 (1 H, m), 3.34 (3
H, s), 3.43–3.59 (3 H, m), 3.76 (3 H, s), 4.21 (1 H, ddd, J 9.0, 4.5
and 2.5), 6.91 (4 H, AA¢BB¢ system,21 dA 6.81, dB 7.01, N 9.0, L
8.5, K 5.5 (M could not be measured accurately)); dC(75.5 MHz;
CDCl3; Me4Si) 20.3, 28.2, 33.5, 50.3, 55.5, 58.8, 64.2, 69.1, 85.7,
103.0, 114.1, 120.1, 138.3, 154.1.
Preparative HPLC [(H2O + 0.1% HCO2H)/(MeOH + 0.1%
HCO2H) 32 : 68, flow rate 17 mL min-1, UV detection 254 nm] of
the combined crude products of the two experiments (69.1 mg)
led to the isolation of pure endo-cis-1d (retention time 27.5 min;
6.0 mg, 13 mmol, 10%) and exo-cis-1d (retention time 30.0 min;
6.0 mg, 13 mmol, 10%).
endo-cis-1d
(3R*,3aS*,5S*,6aR*). Colourless
oil;
dH(300 MHz; CDCl3; Me4Si) 1.48 (3 H, s), 1.93 (2 H, q, J
6.5), 2.01 (1 H, ddd, J 13.5, 9.0 and 8.0), 2.26 (1 H, ddd, J 13.5,
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5591–5601 | 5599
©

View Article Online
(3aR*,6aS*)-3-(2-Methoxyethyl)-6a-methyl-6-phenylhexahydro-
[1,2]dioxolo[3,4-b]pyrrole (1k)
Acknowledgements
We are grateful to the Centre National de la Recherche Scientifique
and to the Muse´um National d¢Histoire Naturelle for financial
support. We also wish to thank the people of the HPLC, mass
spectrometry and elemental analysis services of the Institut de
Chimie des Substances Naturelles for their contribution to this
work.
This compound was prepared from endo-2k (30.0 mg, 130 mmol) by
applying the general procedure C (+0.65 V during 30 min). Anal-
ysis of the crude product (yellow oil) by 1H NMR spectroscopy15
revealed the presence of the two diastereoisomeric endoperoxide
products endo-1k (29.1%) and exo-1k (29.3%), i.e. an overall
estimated yield of 58% and an endo/exo ratio of 50 : 50. Flash
chromatography (neutral alumina, activity II, AcOEt/heptane,
gradient from 0% to 10%) led to the isolation of partially purified
endo-1k (4.1 mg), contaminated with two vinylogous amides 6 and
7 (endo-1k/6/7 ratio 37 : 40 : 23), that were not present in the crude
Notes and references
1 Reviews: (a) O. G. Kulinkovich and A. de Meijere, Chem. Rev.,
2000, 100, 2789–2834; (b) A. de Meijere, S. I. Kozhushkov, and
A. I. Savchenko, in Titanium and Zirconium in Organic Synthesis, ed.
I. Marek, Wiley-VCH, Weinheim, 2002, pp 390-434; (c) A. de Meijere,
S. I. Kozhushkov and A. I. Savchenko, J. Organomet. Chem., 2004, 689,
2033–2055; (d) A. Wolan and Y. Six, Tetrahedron, 2010, 66, 15–61.
2 C. Madelaine, Y. Six and O. Buriez, Angew. Chem., 2007, 119, 8192–
8195; C. Madelaine, Y. Six and O. Buriez, Angew. Chem., Int. Ed., 2007,
46, 8046–8049.
1
product as verified by H NMR. Attempted further purification
by preparative HPLC [(H2O + 0.1% HCO2H)/(MeOH + 0.1%
HCO2H) 50 : 50, flow rate 17 mL min-1, UV detection 254 nm]
only led to the isolation of pure 6 (1.5 mg, 6.1 mmol, 5%).
3 Selected papers and reviews: (a) J. Wiesner, R. Ortmann, H. Jomaa
and M. Schlitzer, Angew. Chem., 2003, 115, 5432–5451; J. Wiesner, R.
Ortmann, H. Jomaa and M. Schlitzer, Angew. Chem., Int. Ed., 2003,
42, 5274–5293; (b) A. Robert, F. Benoit-Vical and B. Meunier, Coord.
Chem. Rev., 2005, 249, 1927–1936; (c) P. A. Stocks, P. G. Bray, V. E.
Barton, M. Al-Helal, M. Jones, N. C. Araujo, P. Gibbons, S. A. Ward,
R. H. Hughes, G. A. Biagini, J. Davies, R. Amewu, A. E. Mercer, G.
Ellis and P. M. O’Neill, Angew. Chem., 2007, 119, 6394–6399; P. A.
Stocks, P. G. Bray, V. E. Barton, M. Al-Helal, M. Jones, N. C. Araujo,
P. Gibbons, S. A. Ward, R. H. Hughes, G. A. Biagini, J. Davies, R.
Amewu, A. E. Mercer, G. Ellis and P. M. O’Neill, Angew. Chem., Int.
Ed., 2007, 46, 6278–6283; (d) R. K. Haynes, W. C. Chan, C.-M. Lung,
A.-C. Uhlemann, U. Eckstein, D. Taramelli, S. Parapini, D. Monti and
S. Krishna, ChemMedChem, 2007, 2, 1480–1497; (e) M. H. Gelb, Curr.
Opin. Chem. Biol., 2007, 11, 440–445; (f) C. W. Jefford, Drug Discovery
Today, 2007, 12, 487–495; (g) K. M. Muraleedharan and M. A. Avery,
Drug Discovery Today, 2009, 14, 793–803.
endo-1k (3R*,3aS*,6aR*). Yellow oil; dH(300 MHz; CDCl3;
Me4Si) 1.57 (3 H, s), 1.89 (2 H, q, J 6.5), 1.93 (1 H, ddd, J 12.5,
8.5, 6.5 and 3.5), 2.10 (1 H, dddd, J 12.5, 9.0, 8.0 and 7.5), 3.11
(1 H, ddd, J 8.5, 7.5 and 5.0), 3.25–3.42 (1 H, m), 3.36 (3 H,
s), 3.45–3.55 (3 H, m), 4.47 (1 H, td, J 6.5 and 5.0), 6.84 (1 H,
t, J 7.5), 6.96 (2 H, d, J 8.5), 7.22 (2 H, dd, J 8.5 and 7.5);
dC(75.5 MHz; CDCl3; Me4Si) characteristic signals 49.2, 58.7,
62.0, 69.7, 81.3, 103.4, 116.8, 119.4, 128.7. This compound could
not be characterised more extensively because of the lack of purity
and the small amount obtained.
exo-1k (3R*,3aR*,6aS*). dH(300 MHz; CDCl3; Me4Si) char-
acteristic signals 1.63 (3 H, s), 2.97 (1 H, ddd, J 8.5, 5.0 and 2.0),
4.20 (1 H, ddd, J 9.0, 4.5 and 2.5); dC(75.5 MHz; CDCl3; Me4Si)
characteristic signals 20.1, 33.6, 49.2, 58.8, 65.0, 69.1, 85.3, 102.7,
116.8, 119.3, 128.7.
6. dH(300 MHz; CDCl3; Me4Si) 1.82 (3 H, s), 2.50 (2 H, t, J 7.5),
2.64 (2 H, t, J 6.0), 3.31 (3 H, s), 3.61 (2 H, t, J 6.0), 3.70 (2 H, t,
J 7.5), 7.17 (2 H, dd, J 7.5 and 1.5), 7.35 (1 H, tt, J 7.5 and 1.5),
7.42 (2 H, t, J 7.5).
7. dH(300 MHz; CDCl3; Me4Si) characteristic signals 2.30 (3 H,
t, J 1.0), 2.69 (2 H, t, J 6.5), 2.95 (2 H, tq, J 9.5 and 1.0), 3.38 (3
H, s), 3.75 (2 H, t, J 6.5), 3.93 (2 H, t, J 9.5), 7.17 (2 H, dd, J 7.5
and 1.5).
4 K. Wimalasena, H. B. Wickman and M. P. D. Mahindaratne,
Eur. J. Org. Chem., 2001, 3811–3817.
5 For recent examples of alternative methods for the synthesis of fused
bicyclic aminocyclopropane derivatives, see: (a) S. Couty, C. Meyer and
J. Cossy, Synlett, 2007, 2819–2822; (b) S. Couty, C. Meyer and J. Cossy,
Tetrahedron, 2009, 65, 1809–1832; (c) P. Bertus and J. Szymoniak, Org.
Lett., 2007, 9, 659–662; (d) A. Joosten, J.-L. Vasse, P. Bertus and J.
Szymoniak, Synlett, 2008, 2455–2458; (e) D. Astashko, H. G. Lee,
D. N. Bobrov and J. K. Cha, J. Org. Chem., 2009, 74, 5528–5532;
(f) P. K. Mykhailiuk, S. Afonin, G. V. Palamarchuk, O. V. Shishkin,
A. S. Ulrich and I. V. Komarov, Angew. Chem., 2008, 120, 5849–5851;
P. K. Mykhailiuk, S. Afonin, G. V. Palamarchuk, O. V. Shishkin, A. S.
Ulrich and I. V. Komarov, Angew. Chem., Int. Ed., 2008, 47, 5765–5767;
(g) S. Ouizem, The`se de l’Universite´ Pierre et Marie Curie, Paris, 2010.
6 Reviews: (a) J.-P. Be´gue´ and D. Bonnet-Delpon, J. Fluorine Chem.,
2006, 127, 992–1012; (b) K. L. Kirk, J. Fluorine Chem., 2006, 127, 1013–
1029; (c) J.-P. Be´gue´, and D. Bonnet-Delpon, In Chimie Bioorganique
et Me´dicinale du Fluor, EDP Sciences and CNRS Editions: Paris, 2005;
(d) J.-P. Be´gue´, and D. Bonnet-Delpon, (Transl. J. Legros), Bioorganic
and Medicinal Chemistry of Fluorine, John Wiley & Sons: Hoboken,
2008; (e) K. Mu¨ller, C. Faeh and F. Diederich, Science, 2007, 317,
1881–1886; (f) W. K. Hagmann, J. Med. Chem., 2008, 51, 4359–4369;
(g) K. L. Kirk, Org. Process Res. Dev., 2008, 12, 305–321.
7 (a) T. T. T. Nga, C. Me´nage, J.-P. Be´gue´, D. Bonnet-Delpon and J.-
C. Gantier, J. Med. Chem., 1998, 41, 4101–4108; (b) F. Grellepois,
F. Chorki, M. Oure´vitch, S. Charneau, P. Grellier, K. A. McIntosh,
W. N. Charman, B. Pradines, B. Crousse, D. Bonnet-Delpon and J.-P.
Be´gue´, J. Med. Chem., 2004, 47, 1423–1433; (c) G. Magueur, B. Crousse,
S. Charneau, P. Grellier, J.-P. Be´gue´ and D. Bonnet-Delpon, J. Med.
Chem., 2004, 47, 2694–2699; (d) C. Chollet, B. Crousse, M. Oure´vitch
and D. Bonnet-Delpon, J. Org. Chem., 2006, 71, 3082–3085; (e) For a
review, see:J.-P. Be´gue´ and D. Bonnet-Delpon, ChemMedChem, 2007,
2, 608–624.
Evaluation of the activity against Plasmodium falciparum
In vitro antimalarial assays were performed on human erythro-
cytes with the chloroquine-resistant FcB1 strain of Plasmodium
falciparum as described.22 Briefly, stock solutions of compounds,
prepared in DMSO, were serially diluted with culture medium and
introduced to asynchronous parasite cultures (1% parasitaemia
and 1% final hematocrite) on 96-well plates for 24 h at 37 ◦C
prior to the addition of 0.5 mCi of [3H]hypoxanthine per well,
for 24 h. The growth inhibition for each drug concentration was
determined by comparison of the radioactivity incorporated into
the treated culture with that in the control culture (without drug)
maintained on the same plate. The concentrations causing 50%
inhibition (IC50) or 90% inhibition (IC90) were obtained from the
drug concentration-response curves of triplicate experiments.
8 A. de Meijere, C. M. Williams, A. Kourdioukov, S. V. Sviridov,
V. Chaplinski, M. Kordes, A. I. Savchenko, C. Stratmann and M.
Noltemeyer, Chem.–Eur. J., 2002, 8, 3789–3801.
5600 | Org. Biomol. Chem., 2010, 8, 5591–5601
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
9 C. Madelaine, N. Ouhamou, A. Chiaroni, E. Vedrenne, L. Grimaud
and Y. Six, Tetrahedron, 2008, 64, 8878–8898.
10 C. Madelaine, A. K. Buzas, J. A. Kowalska-Six, Y. Six and B. Crousse,
Tetrahedron Lett., 2009, 50, 5367–5371.
18 L. Larquetoux, J. A. Kowalska and Y. Six, Eur. J. Org. Chem., 2004,
3517–3525.
19 The structures of the two aminocyclopropanes 2b and 2m are almost
identical, except for the substituent at the a position relative to the
nitrogen atom: a CF3 group in 2b and a phenyl group in 2m. The
ionisation potential of 2m is thus expected to be lower than that of 2b.
20 Y. Six, Eur. J. Org. Chem., 2003, 1157–1171.
21 H. Gu¨nther, Angew. Chem., 1972, 84, 907–920; H. Gu¨nther, Angew.
Chem., Int. Ed. Engl., 1972, 11, 861–874.
22 J. Schre´vel, V. Sinou, P. Grellier, F. Frappier, D. Gue´nard and P. Potier,
Proc. Natl. Acad. Sci. U. S. A., 1994, 91, 8472–8476.
23 B. V. Nonius, COLLECT, data collection software, 1999.
24 Z. Otwinowski, W. Minor, Methods in Enzymology, Academic Press,
New York, 1997, Vol. 276, pp 307–326.
11 N. Ouhamou and Y. Six, Org. Biomol. Chem., 2003, 1, 3007–3009.
12 C. P. Casey and N. A. Strotman, J. Am. Chem. Soc., 2004, 126, 1699–
1704.
13 Endo/exo refer to the relative configuration of the R3 substituent
with respect to the bicyclic system; cis/trans refer to the relative
configuration of the angular methyl group and the R2 substituent.
14 N. Ouhamou, The`se de l¢Universite´ Paris-Sud 11, 2006.
15 See the Supplementary Information for detail.
16 C. Laroche, P. Bertus and J. Szymoniak, Tetrahedron Lett., 2003, 44,
2485–2487.
17 Indeed, this carbon–oxygen bond is in good geometrical alignment
with the lone pair of the nitrogen atom, as pointed out by Professor
Willie Motherwell during a private discussion.
25 G. M. Sheldrick, Acta Crystallogr., Sect. A, 2008, 64, 112–
122.
26 A. L. Spek, J. Appl. Crystallogr., 2003, 36, 7–13.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5591–5601 | 5601
©