Potential proinsecticides of fluorinated carboxylic acids and β-ethanolamines. IV. Evaluation of the Δ2-oxazoline-1,3 structure by 19F NMR monitoring of the in vitro metabolism in locust tissues

Article abstract of DOI:10.1016/S0022-1139(00)00382-1

The enzymatic effect of locust tissues upon hydrolysis of the fluorinated Δ²-oxazoline-1,3 Ia was elucidated using ¹⁹F[¹H] NMR monitoring. In a phosphate buffer at pH = 7.4 (mean physiological pH of locust tissues), the substrate Ia hydrolyses slowly into the corresponding fluorinated hydroxylamide VIa. If diluted, locust haemolymph (12.5% in phosphate buffer) catalyses slightly this hydrolytic pathway, it overall triggers the unmasking of carboxylate IIIa, corresponding to the expected proinsecticide behaviour of Ia. This behaviour is spectacularly almost the unique reaction observed during in vitro assays in concentrated fat body and mesenteron. Inasmuch as β-hydroxylamide VIa is not hydrolysed into carboxylate IIIa during such conditions, it must be concluded that carboxylate formation exclusively results from hydration and hydrolysis of substrate Ia via the aminoester Va. The formation of this intermediate aminoester is demonstrated by complementary assays. The enzymes supposed to intervene are of the α-chymotrypsine type for the first step (hydration) and of the esterase type for subsequent hydrolysis of intermediate aminoester Va. Thus, this work constitutes the first example of a Δ²-oxazoline-1,3 structure exploited for elaborating proinsecticides of carboxylates III and/or β-ethanolamines II based on enzymatic activation in insects.

Full text of DOI:10.1016/S0022-1139(00)00382-1

Journal of Fluorine Chemistry 107 (2001) 387±395
Potential proinsecticides of ¯uorinated carboxylic acids
and b-ethanolamines
IV. Evaluation of the D2-oxazoline-1,3 structure by F NMR
19
monitoring of the in vitro metabolism in locust tissues$
*
J.-C. Cherton , S. Hamm, J.-C. Halle, L. Menguy, M. Beaufour
Â
Departement de Chimie, Laboratoire SIRCOB (UPRESA CNRS 8086), Universite de Versailles, 45 Avenue des Etats-Unis,
Â
78035 Versailles Cedex, France
Â
Â
Received 3 April 2000; accepted 10 May 2000
Abstract
The enzymatic effect of locust tissues upon hydrolysis of the ¯uorinated D2-oxazoline-1,3 Ia was elucidated using 19F[1H] NMR
monitoring.
In a phosphate buffer at pH ˆ 7:4 (mean physiological pH of locust tissues), the substrate Ia hydrolyses slowly into the corresponding
¯uorinated hydroxylamide VIa.
If diluted, locust haemolymph (12.5% in phosphate buffer) catalyses slightly this hydrolytic pathway, it overall triggers the unmasking of
carboxylate IIIa, corresponding to the expected proinsecticide behaviour of Ia. This behaviour is spectacularly almost the unique reaction
observed during in vitro assays in concentrated fat body and mesenteron. Inasmuch as b-hydroxylamide VIa is not hydrolysed into
carboxylate IIIa during such conditions, it must be concluded that carboxylate formation exclusively results from hydration and hydrolysis
of substrate Ia via the aminoester Va. The formation of this intermediate aminoester is demonstrated by complementary assays. The
enzymes supposed to intervene are of the a-chymotrypsine type for the ®rst step (hydration) and of the esterase type for subsequent
hydrolysis of intermediate aminoester Va.
Thus, this work constitutes the ®rst example of a D2-oxazoline-1,3 structure exploited for elaborating proinsecticides of carboxylates III
and/or b-ethanolamines II based on enzymatic activation in insects. # 2001 Elsevier Science B.V. All rights reserved.
Keywords: D2-oxazoline-1,3; Proinsecticide; Metabolism; 19F NMR monitoring; Insect biological media; African locust Locusta migratoria
1. Introduction
b-ethanolamine II and a carboxylic acid III by dehydration.
So structure I can be considered as reversibly masking both
moities II and III, since by acid hydrolysis, the reversed
reaction is expected (Equation 1).
From a retrosynthetic standpoint the D2-oxazoline-1,3
structure I results from the condensation between a
Meyers and co-workers have intensively exploited this
reversible masking for the development of oxazoline
mediated asymetric synthesis using these versatile chiral
$ See [19,35,36] for parts I±III.
* Corresponding author. Fax: ‡33-139254452.
E-mail address: cherton@chimie.uvsq.fr (J.-C. Cherton).
0022-1139/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ( 0 0 ) 0 0 3 8 2 - 1

388
J.-C. Cherton et al. / Journal of Fluorine Chemistry 107 (2001) 387±395
auxiliaries groups [1,2], particularly for the elaboration of a-
substituted carboxylic acids in high chemical and optical
yields [3,4].
In a ``proinsecticide perspective'' the metabolisation of
simple oxazolines chosen as models has been studied in our
laboratory for about 10 years. For monitoring the expected
demasking of carboxylates [5,6] and b-ethanolamines [7]
selected as active principles, 1H NMR [8,9], HPLC [10] and
19F NMR [11,12] were used as analytical tools during in
vitro assays in locust haemolymph.
A few years ago VorbruÈggen and Krolikiewicz developed
non steroidal anti-in¯ammatory drugs (NSAID) based upon
D2-oxazolines-1,3 containing such masked carboxylic acids
[13]. These derivatives, designed as potential prodrugs [14],
hydrolyse gradually in vivo under physiological pH and
enzymatic action, into the starting NSAID.
This paper deals with the ¯uorinated D2-oxazoline-1,3 Ia,
designed both as a contact insecticide due to its pronounced
lipophilicity: log P ˆ 2:9 according to the Rekker indexes
[15], and as a proinsecticide of the corresponding carbox-
ylate IIIa and/or b-ethanolamine IIa (cf. Scheme 1). In fact,
oxazoline Ia exhibited noticeable biological activities with
respect to some insect or pests species [12]. Thus, it
appeared necessary to study its metabolism in insects to
appreciate the extent of the unmasking of the active prin-
ciple(s) and their possible subsequent catabolisation(s).
Assuming that the penetration in insects was effective when
Ia was applied on their integument, we developed in vitro
assays in the tissues of the locust Locusta migratoria for
modelling the true metabolic processes in living insect. The
selection of those tissues is connected to their known
potentialities to trigger hydrolysis of xenobiotics due to
their hydrolases contents: haemolymph [9±12,16,17]; fat
body [11,16,17] and mesenteron [18,19].
For monitoring the behaviour of ¯uorinated xenobiotics in
biological tissues, 19F NMR is a very convenient tool, having
inherent advantages. First, most biological systems are
devoid of NMR-detectable ¯uorinated endogenous com-
pounds. This implies that all resonances observed can be
unambiguously ascribed to the studied xenobiotic and
its biodegradation products, without interference from
endogenous or solvent nuclei. Second, ¯uorine 19 has a
Scheme 1. Hydrolysis pathways for D2-oxazoline-1,3 Ia. Entities labelled F1 and F2 in the 19F spectrum presented in Fig. 1 were, respectively, identified to
intermediate aminoesters Va and VIa.

J.-C. Cherton et al. / Journal of Fluorine Chemistry 107 (2001) 387±395
389
sensitivity to NMR detection that is 83% that of the sensi-
1
tivity of H, which sensitivity is obviously an important
2. Results and discussion
factor because of the usually low concentrations of xeno-
biotic compounds and their biodegradation products. Third,
due to a chemical shift range of about 500 ppm, the 19F
nucleus is highly sensitive to its molecular surroundings,
resulting in relatively large changes in chemical shifts as a
result of biochemical modi®cations of a xenobiotic with few
chances of peak overlap.
2.1. 19F NMR monitoring of the behaviour of the D2-
oxazoline-1,3 Ia in a phosphate buffer at different pH
2.1.1. Feasibility of the monitoring
The feasibility of the 19F monitoring was clear from the
recording (cf. following protocol) of the 19F spectra of the
substrate Ia synthesised according to the Vorbru
Èggen
Numerous advantages of the ¯uorine spin labelling for
NMR studies of biological systems have been recognised for
more than 30 years [20], and related reviews have appeared
[21±23]. Brie¯y, 19F NMR spectroscopy has been applied to:
monitoring of drugs in mammalian tissues developed by
Malet-Martino and Martino [24,25]; pH measurement using
¯uoropyridoxol [26]; complexation study of carbonic anhy-
drase with a ¯uorosulfonamide [27] and also for agrochem-
ical and environmental research [28±31], for the
determination of ¯uorinated pesticides in foods [32], study
of a fungicide in a pathogenic fungus [33], and as structural
probe for internal amino acid packing in proteins [34]. In the
``proinsecticide perspective'', 19F NMR was also used in our
laboratory, to clarify the behaviour of ¯uorinated enol esters
[35], esters [36] and recently, N-acylaziridines in insect
tissues [12,19].
This work highlights the convenience of direct 19F[1H]
monitoring to study the behaviour of the D2-oxazoline-1,3 Ia
during in vitro assays in locust haemolymph, fat body and
mesenteron, and demonstrates that this nitrogen heterocycle
is an effective structure for the elaboration of proinsecticides
of carboxylic acids. As a comparison, the effects of com-
mercial enzymes, a-chymotrypsine and a supported esterase
were also studied.
method [13], and IIIa in a phosphate buffer at pH 7.4.
Indeed, the chemical shifts of the two compounds were,
respectively, 40.06 and 41.49 ppm relative to external
reference CF3COOH (cf. Table 1), i.e. suf®ciently different
to allow their quantitative measurement.
2.1.2. Detailed 19F NMR protocol
The lipophilicity of the substrate (log P ꢀ 2:9) required
the use of a water-soluble organic co-solvent for its solu-
bilisation in the phosphate buffer.
We chose 5% of DMSO-d6, since in such conditions the
hydrolases enzymatic activity of locust tissues is well pre-
served [9,11,19]. Due to the high sensitivity of 19F chemical
shifts to the medium conditions, we have been working
under controlled temperature (258C) and pH (buffered solu-
tion) and with comparable substrate concentration. p-Fluor-
obenzylalcohol was selected as the internal ¯uorinated
standard IS for its structural analogy with the substrate.
Effectively, we observed similar chemical shifts for Ia and
IS and similar relaxation times T1 (cf. Table 1). This
authorises the use of peak integration for the quantitative
following of the balance-sheet of the ¯uorinated entities. In
addition, the internal standard facilitated signal identi®ca-
tion when varying media conditions.
Table 1
19F monitoring of the behaviour of D2-oxazoline-1,3 Ia in phosphate buffera
d ppm [T1] (s)
Iab [4.94]
IIIa [4.97]
ISc [6.25]
F1 ˆ Va
F2 ˆ VIa [4.76]
At pH 7.4
t0 (%)
1 h
40.06 (89)
(88)
41.49 (0)
39.40 (100)
(100)
39.81 (4)
(5)
40.27 (7)
(7)
(0)
(0)
(0)
(0)
(3)
2 h
(81)
(100)
(9)
(10)
3 h
4 h
(78)
(73)
(100)
(100)
(10)
(11)
(12)
(16)
10, 12 h
(35)
(100)
(18)
(44)
At pH 6.3
t0
(81)
(60)
(41)
(30)
(23)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(100)
(100)
(100)
(100)
(100)
(100)
(10)
(31)
(50)
(61)
(68)
(90)
(9)
(9)
(9)
(9)
(9)
(10)
1 h
2 h
3 h
4 h
12 h
a 19F chemical shifts (d are reported relative to the resonance peak of CFCOO (5% v/v aqueous solution) used as an external reference. See Section 4 for
NMR conditions and T1 of Ia, carboxylate IIIa, internal standard IS and fluorinated entities F1 and F2. Evolution with time within the limits of the integration
precision, the balance ([Ia] ‡ [IIIa] ‡ [Va ˆ F1] ‡ [VIa ˆ F2]) ˆ [IS]) ˆ 100%.
b [Ia] ˆ 10 3 M.
c [IS] ˆ 10 3 M.

390
J.-C. Cherton et al. / Journal of Fluorine Chemistry 107 (2001) 387±395
Fig. 1. 19F NMR spectrum of a solution of D2-oxazoline-1,3 Ia in phosphate buffer at pH 7.4 after 10 h (cf. also Table 1).
Optimised NMR conditions from the standpoints of sen-
sitivity and duration resulted in spectrum acquisition over
30 min (cf. Section 4).
about 10-fold more acidic solution than previously) in order
to facilitate F1structure elucidation.
2.1.4. Structure of the fluorinated intermediate F1
2.1.3. Behaviour of Ia in phosphate buffer at pH 7.4
(physiological) and 6.3
2.1.4.1. 13C NMR monitoring of the evolution of Ia (c ˆ
0.08 M) at pH 6.3. Comparison of the 19F NMR monitorings
of the evolution of Ia in phosphate buffer at pH 7.4 and 6.3
(cf. Table 1) indicates an important increase in the formation
of F1 at the lower pH after t0 spectrum, with a very
significant difference after 2 h: 50% at pH 6.3 compared
to only 9% at pH 7.4. Thus, after 2 h standing at pH about
ꢀ6.3 (the buffer pH was doubtless modified by the high ratio
of DMSO-d6), the 13C[1H] spectrum of a 8 Â 10 2 M
solution of Ia in phosphate buffer±DMSO-d6 (80±20%)
solution was recorded in 12 000 scans. Beside the signals
of Ia and VIa the spectrum showed a series of signals
assignable to F1 (cf. Table 2). In the expanded region at
about 175 ppm, it is of great importance to notice the
presence of two Cq resonances both corresponding
clearly to carbonyls, of amidic or ester character, which
were assigned to the hydroxylamide VIa (ˆ F2) and the
aminoester Va (ˆ F1).
To appreciate subsequently the enzymatic effect of the
biological media on the hydrolysis of the substrate Ia, we
have ®rst studied a possible pH effect for these media, using
the same NMR conditions, in the phosphate buffer at
pH ˆ 7:4.
As early as the ®rst spectrum at ``t0'' time, which repre-
sents in fact an averaged evolution over a 30 min period,
there is formation of two ¯uorinated entities labelled F1 at
39.81 ppm and F2 at 40.27 ppm, differing from Ia, IS
and IIIa. The intensities of F1 and F2 increase slowly with
time, cf. Table 1. On the ``10 h'' spectrum (cf. Fig. 1), there
is also to notice the appearance of a very small signal at
41.4 ppm corresponding to the carboxylate IIIa. F2 was
assigned to the b-hydroxylamide VIa by overloading with
an authentic sample. Its formation can be rationalised
according to pathway 1,a in the Scheme 1, which implies
the orthoimidate intermediate 2-hydroxy-oxazolidine-1,3
IVa, as generally assumed in the literature [37±42]. Both
the formation of F1, which appears as a transient species, and
of carboxylate IIIa, can be explained by the other hydrolytic
cleavage possibility of the heterocycle Ia described in path-
way 1,b (Scheme 1). Thus, F1 would be the aminoester Va.
Since this pathway 1,b is known to be rather favoured by
acidic conditions [39,41], where Va would be the kinetic
product, we performed other assays with Ia at pH 6.3 (i.e. in
2.1.4.2. Esterase and a-chymotrypsine effects on Va
observed by 19F monitoring. The hypothetical ester
structure Va previously considered for F1, was confirmed
by adding a commercial esterase to the diluted solution of Ia
in a phosphate buffer at pH ꢀ 6:3, since after passage of 12 h
transformation of Ia into Va (ˆ F1) (90%) and VIa (ˆ F2)
(10%) was achieved, (cf. Table 1). The rapid increase in
carboxylate IIIa observed (cf. Fig. 2C) correlates with the
Table 2
13C chemical shifts of D2-oxazoline-1,3 Ia, VIa (ˆ F2) and Va (ˆ F1) observed in the 13C[1H] spectrum of the mixture resulting from standing of Ia for 2 h in
a phosphate buffer at pH ꢀ 6:3, DMSO-d6 (80/20)a
CH3
CH2±O
CH±N
CH2Ar
C2
C01
C02
C03
C04
Ia
21.8
17.5
16.1
76.0
65.9
67.4
61.7
48.8
47.9
34.5
40.6
43.1
169.1
175.2
175.1
132.7
132.7
131.1
132.2
132.3
132.9
117.1
117.0
117.0
163.2
163.3
163.2
VIa ˆ F2
Va ˆ F1
a The chemical shifts d (see Scheme 1 for carbon numbering) are reported relative to TMS as internal reference. N:S: ˆ 12 000 scans, see Section 4 for
other details.

J.-C. Cherton et al. / Journal of Fluorine Chemistry 107 (2001) 387±395
391
Fig. 2. 19F NMR monitoring of carboxylate IIIa formation from Ia and from the intermediate F1 ˆ Va in the presence of an esterase or of a-chymotrypsine:
A (^) esterase on Ia (at pH 7.4); B (~) a-chymotrypsine on Ia (at pH 7.8); C (&) esterase on Va (obtained from Ia after 12 h standing at pH 6.3); D (*) a-
chymotrypsine on Va (at pH 6.3).
decrease of Va since the hydroxylamide VIa ratio remained
essentially constant with time at this pH (cf. Table 1). This
result thus confirmed the aminoester Va structure for F1. a-
Chymotrypsine exercised an even more pronounced
hydrolytic effect on Va (cf. Fig. 2D).
To summarise, under non-enzymatic hydrolytic condi-
tions, the substrate Ia is slowly converted at the physiolo-
gical pH 7.4 into the hydroxylamide VIa (pathway 1,a)
without very signi®cant evolution of intermediate Va
(formed according the pathway 1,b) towards the carboxylate
IIIa (pathway c). This pathway is made ef®cient by the
presence of hydrolase: esterase or a-chymotrypsine.
NMR signal broadening, we put forward the explanation
previously proposed for other ¯uorinated substrates inter-
acting with locust [19,35,36] and caterpillar tissues [19]
that is substrate association with some endogenous compo-
nent of the biological media. It was also often observed that
substrate association with macromolecules resulted in var-
iation of the chemical shifts Dd and enlargement d1/2 of the
signals. For instance, the association of human carbonic
anhydrase with 3,5-di¯uorobenzenesulfonamide showed a
dramatic decrease of longitudinal relaxation time T1 [27]. In
the present case, the T1 determination was impossible
because of substrate Ia reaction in the presence of biological
media, thus the signal-enlargements where de®ned by the d1/
2.2. 19F NMR monitoring of the in vitro behaviour of D2-
oxazoline-1,3 Ia in the presence of biological media: locust
tissues, esterase and a-chymotrypsine
(Hz), but we did not observe very signi®cant Dd (cf.
Table 3).
2
The main effects of the biological media are catalysis of
the hydrolytic pathways 1,a; 1,b and c to varying extents.
It was previously shown that locust tissues such as
haemolymph, fat body and mesenteron are free of detectable
19F resonance over the range covering the signals of Ia, IIIa
and IS [8,9,19,35,36] and in this work, we observed the same
result with esterase and a-chymotrypsine samples.
As early as the ``t0'' spectra evident effects of the biolo-
gical media are observed, contrary to the phosphate buffer
media with: (i) noisy baselines and (ii) peak broadening of
some of the signals (particularly for Ia and Va) in caterpillar
and locust diluted haemolymphs, and mostly in concentrated
locust fat body and mesenteron (cf. Fig. 3A±C). For such 19F
2.2.1. In diluted locust haemolymph (12.5%) at 258C and
pH 7.4
A clear effect of this medium is catalysis of pathway 1,a,
since the ratio of VIa is 44% instead of 16% in a phosphate
buffer after 4 h and also catalysis of pathway c, since IIIa is
formed in 10% after 10 h (cf. Fig. 3A) instead of 3%
(Table 1). The behaviour of Ia in diluted caterpillar hae-
molymph is almost similar for both catalysis and the peak
broadening. At this stage, it can be concluded that Ia acts
partly as a proinsecticide of the carboxylate IIIa.
Table 3
Variations of 19F NMR chemical shifts Dd and peak width (d1/2) of the fluorinated entities Ia, IIIa, IS, F1 ˆ Va and F2 ˆ VIa according to the media: insect
tissues, a-chymotrypsine and phosphate buffer
d (d1/2) ppm (Hz)
Ia
40.06 (2)
IIIa
41.49 (2)
IS
39.40 (2)
VIa ˆ F2
Va ˆ F1
Phosphate buffer 7.4
Locust haemolymph
Caterpillar haemolymph
Locust fat body
40.27 (2)
39.81 (2)
40.00 (6)
41.48 (4.5)
41.48 (2)
41.46 (3)
41.46 (3)
41.51 (2)
39.40 (4.5)
39.39 (2)
39.40 (3)
39.40 (3)
39.40 (2)
40.24 (4.5)
40.25 (2.5)
40.28 (3)
39.78 (4.5)
39.77 (9)
40.03 (7.2)
40.06 (30)
39.99 (130)
40.06 (2)
±
±
±
Locust mesenteron
a-Chymotrypsine
40.21 (2.5)
40.28 (2)

392
J.-C. Cherton et al. / Journal of Fluorine Chemistry 107 (2001) 387±395
Fig. 3. Behaviour of D2-oxazoline-1,3 Ia in insect tissues or in the presence of a-chymotrypsine. (A) [Ia] ˆ 10 3 M in diluted locust haemolymph (12.5% in
phosphate buffer); balance-sheet ˆ 100% at t ˆ t0, 4 and 10 h. (B) [Ia] ˆ 5  10 4 M in locust fat body; balance-sheet ˆ 100% at t ˆ t0, 1 and 3 h. (C) [Ia]
ˆ 5  10 4 M in locust mesenteron; balance-sheet ˆ 63, 55 and 100% at t ˆ t0, 1 and 3 h, respectively. (D) [Ia] ˆ 5  10 4 M in the presence of a-
chymotrypsine; balance-sheet ˆ 100% at t ˆ t0, 4 and 10 h.
2.2.2. In concentrated locust fat body and mesenteron at
258C
The unmasking of carboxylate IIIa (pathways 1,b and c) is
``3 h'' in Fig. 3B, where IIIa integrates to ꢀ60%, VIa and
unchanged Ia for ꢀ20%, and spectacularly IIIa is almost the
unique product of evolution in the mesenteron, since in 3 h
Ia has completely disappeared (cf. Fig. 3C: 87% IIIa and
the preponderant development in the fat body, cf. spectrum

J.-C. Cherton et al. / Journal of Fluorine Chemistry 107 (2001) 387±395
393
13% VIa). In both tissues, intermediate Va (ˆ F1) was not
detected, which result con®rms a metabolic correlation
between Va and IIIa, the only difference between the
two media being the spectacularly more pronounced enlar-
gement of the resonance of Ia in the mesenteron sample.
We have previously demonstrated that pathway d is
inef®cient in locust tissues [19], which result is now con-
®rmed by the nearly constant amount of b-hydroxylamide
VIa observed in Fig. 3C.
From these independant in vitro assays into concentrated
locust tissues it can be concluded that the D2-oxazoline-1,3
Ia acts mainly as a proinsecticide of the carboxylate IIIa.
The hydrolases content of these tissues are likely to be
different from the haemolymph media which, even diluted,
catalysed mainly the formation of b-hydroxylamide VIa. Fat
body and mesenteron are known to contain hydrolytic
enzymes of esterase [35,36] and a-chymotrypsine [18,19]
types, likely to trigger catalysis of the pathways 1,b and c.
In this work, during assays using the aminoester Va, we
have made evident the role of a commercial esterase and of
a-chymotrypsine in pathway c catalysis. Thus, a possible
effect of these commercial enzymes was tested directly
against Ia.
show signi®cant Dd nor increase in d1/2 (cf. Table 3). It must
be concluded that a-chymotrypsine catalyses both pathways
1,b and c and thus presents some analogies with the locust
fat body and mesenteron actions (cf. Table 4).
3. Conclusion
Direct in vitro 19F[1H] monitoring in insect tissues pro-
vides a very suitable and ef®cient tool for evaluating the
potentiality of chemical structures in the masking of ¯uori-
nated active principles.
This work demonstrates that D2-oxazoline-1,3 is a good
model nitrogen heterocyclic structure for developing proin-
secticides based upon carboxylic acids and/or b-ethanola-
mines, as active principles, ¯uorinated or not. In fact, their
hydrolytic unmasking is nearly the only reaction observed
during in vitro assays in concentrated insect tissues. The
different assays performed with isolated locust tissues or
with commercial enzyme showed both catalysis of the
pathway 1,b and the hydrolysis of intermediate F1 identi®ed
as aminoester Va via enzymes of the a-chymotrypsine and/
or of the esterase types. There is no doubt that in a living
insect the unmasking would be even more pronounced due
to the simultaneous action of all the undiluted tissues
susceptible to trigger enzymatic hydrolysis. This study is
now developed in order: to con®rm this hypothesis, per-
forming ex vivo assays in locust [19,35], and to determine
possible chiral catalytic effect of insect biological media on
the pathways 1,a, 1,b and c. Moreover, due to the spectacular
activation of Ia in the locust mesenteron, it would be
certainly useful to carry out tests by ingestion against
insects, using baits [43] treated with Ia and more polar
oxazolines, i.e. those not designed as contact insecticides.
Finally, taking into account that intermediate Va resulting
from Ia is ef®ciently activated by esterases, it would be also
interesting to test the effects of Ia against insecticide-
resistant aphids, where metabolic resistance results from
the increased production of esterases [6].
2.2.3. In the presence of esterase
This diluted biological medium (see Section 4) did not
show a signi®cant catalytic effect on the hydrolysis of Ia at
pH 7.4, which clearly means that this esterase has no effect
on pathway 1,b (cf. Fig. 2A and Table 4).
2.2.4. In the presence of a-chymotrypsine
By contrast with the esterase, this dilute biological med-
ium (see Section 4) exercised a noticeable catalytic effect in
the hydrolysis of Ia leading to IIIa (cf. Fig. 2B), but did not
Table 4
Summary of the different catalytic effects exercised by pH, and biological
media on the evolution of Ia, Va (ˆ F1) VIa (ˆ F2) (cf. Figs. 2 and 3)a
Pathway/catalysis
(cf. Scheme 1)
Proinsecticide
behaviour
Hydroxyamide
formation
4. Experimental
1,b
c
1,a
d
pH < 7/Ia
pH > 7/Ia
yes
no
±
e
yes
±
±
4.1. Biological samples
yes
Esterase/Ia
Esterase/Va
no
±
no
no
±
±
±
The caterpillar (Mamestra brassicae) and the african
migratory locust Locusta migratoria were reared according
to the crowded conditions described in [19].
The a-chymotrypsin from bovine pancreas was supplied
by Sigma (Sigma Aldrich Chimie L'Isle d'Abeau Chesnes
B.P. 701 38297, Saint-Quentin Fallavier, France). One unit
of the enzyme will hydrolyse 1.0 mmole of N-benzoyl-L-
tyrosine ethylester per minute at pH 7.8 at 258C.
yes
a-Chymotrypsine/Ia
a- Chymotrypsine/Va
a- Chymotrypsine/Via
yes
±
yes
yes
±
no
±
±
±
±
±
no
Haemolymph diluted/Ia
Haemolymph diluted/VIa
yes
±
yes
±
yes
±
±
no
Mesenteron*/Ia
Mesenteron*/VIa
yes
±
yes
±
no
±
±
no
The esterase from hog liver immobilised on Eupergit was
supplied by Fluka (L'Isle d'Abeau Chesnes B.P. 701, 38297,
Saint-Quentin Fallavier, France). One unit of the enzyme
a Important, weak, very weak and non-significant effects are indicated,
respectively, by yes, yes, e and no.

394
J.-C. Cherton et al. / Journal of Fluorine Chemistry 107 (2001) 387±395
will hydrolyse 1.0 mmole of ethylvalerate per minute at
pH 8 and 25.
(5% in D2O, v/v) for solutions in DMSO-d6/phosphate
buffer (5/95, v/v). Spectra were run with 1H decoupling in
the following instrumental conditions: sweep width,
2000 Hz; pulse angle, 908; delay, 15 s; computer resolu-
tion, 0.5 Hz/point; number of scans, 104; memory size,
8 K; line exponential multiplication of FID, 1 Hz.
19F relaxation times measured by the inversion recov-
ery method are 4.94 s for D2-oxazoline-1,3 Ia in solution
in phosphate buffer, 4.76 s for b-hydroxylamide VIa [19],
4.97 s for carboxylate IIIa and 6.25 s for p-fluorobenzy-
lalcohol IS (cf. Table 1).
4.2. In vitro assays
Haemolymph sampled from adults of locusts (male and
female, pH 7.4 were centrifuged for 5 min at 3000 g, then
diluted eight times with phosphate buffer (0.1 M, pH 7.4)
and incubated with the substrate Ia (‰substrateŠ ˆ 10 3 M).
Five percent of DMSO-d6 was added in order to dissolve the
lipophilic substrate and for locking the NMR apparatus.
Locust fat body: a mixture of 260 mg of this tissue and
400 ml of phosphate buffer (0.1 M, pH 7.4) was ground and
centrifuged resulting in three fractions differing in density:
the bottom consisting of tissue fragments, the aqueous
fraction and the lipidic supernatant. To 475 ml of the inter-
mediate aqueous fraction, 25 ml of a 10 2 M solution of D2-
oxazoline-1,3 Ia in DMSO-d6 was added, giving a
5 Â 10 4 M substrate solution with 5% proportion (v/v)
of DMSO-d6.
ꢁ High-resolution mass spectrometry (HRMS) was per-
Â
formed at the ``Laboratoire de Spectrometrie de Masse,
Universite de Mont-Saint-Aignan'', Rouen, France using
Â
a Jeol AX 500 spectrometer (resolution 500) equipped
with a PDP11 data system.
4.4. Reagents and chemicals
Locust mesenteron: the sample of locust caeca (pH 6.4±7)
was ground without dilution and centrifuged. To 475 ml of
the supernatant tissues, 25 ml a stock solution of a 10 2 M
D2-oxazoline-1,3 Ia in DMSO-d6 was added to obtain the
same ®nal substrate solutions as previously.
a-Chymotrypsin and esterase effects onIa: a 500 ml sam-
ple of a 5 Â 10 4 M solution of the substrate Ia in DMSO-d6
and phosphate buffer (5/95) was incubated with a-chymo-
trypsin (62.5 units), or with 60 mg of immobilised esterase,
and the evolution of the mixtures were monitored according
the classic 19F NMR protocol.
a-Chymotrypsin and esterase effects on intermediate Va: a
10 3 M solution of Ia in DMSO-d6 and phosphate buffer
0.1 M and pH 6.3 (5/95) was left to develop at room
temperature over 12 h. Then a-chymotrypsin (62.5 units)
or immobilised esterase (60 mg) were added to an aliquot of
600 ml, and the evolution of the mixtures were monitored
with time according the classic 19F NMR protocol.
4-Fluorophenylacetic acid IIIa and 2-amino-1-propanol
IIa were supplied by Aldrich (Sigma Aldrich Chimie L'Isle
d'Abeau Chesnes B.P. 701, 38297, Saint-Quentin Fallavier,
France).
2-(4-Fluorobenzyl)-4-methyl-D2-oxazoline-1,3 Ia was
obtained according to the VorbruÈggen protocol [13]. An
equimolecular mixture (1.623 mmoles) of IIa and IIIa
dissolved in 3.57 mmoles of hexachloroethane and 5 ml
Ê
of dichloromethane (dried over 3 A molecular sieves) was
cooled at 28C under nitrogen atmosphere. Then a mixture
of 3.57 mmoles of triphenylphosphine and 7.14 mmoles of
triethylamine in 4 ml of dry dichloromethane was slowly
added. After re¯ux (3 h) and standing at room temperature
(24 h), the mixture was washed with dilute NaOH and
extracted with dichloromethane. After precipitation of
Ph3PO with hexane, ®ltration and evaporation of the solvent,
the resulting crude D2-oxazoline-1,3 Ia was distilled
Eb0.5 mmHg, 848C; yield, 35%.
IR (KBr): 2971 (s, M, n=C±H), 2927, 2883, (s, M, nC±H),
1666 (s, S, nC=N), 1510 (s, S, nC=C), 1222 (s, S, nC±F), 826 (s,
M, nC±H, p-disubstituted aromatic).
4.3. Structural characterisation
The following instruments and abbreviations were used:
1H NMR (CDCl3) d/TMS: 7.27 (2H, m, meta F in C6H4F);
7.00 (2H, m, ortho F in C6H4F); 3.58 (2H, s, CH2Ar); the
heterocyclic protons (H4, H5, H05) constitute an ABX system
at the limits of an AMX system, complicated by the
C4H=CH2Ar coupling: 4.33 (1H, dd, H5), 4.18 (1H, m,
H4), 3.78 (1H, ꢀ t, H50 ); J H4±H5 ˆ 9:4 Hz, J H5±H05 ˆ
7:9 Hz, J H4±H05 ˆ 7:4 Hz.
ꢁ IR spectra were taken on Bruker IFS 25 IR-FT spectro-
meter and are described with the following abbreviations:
for the vibrations n, d and g, the wavenumbers are
expressed in cm 1, S: strong, M: medium, W: weak, s:
sharp, b: broad.
ꢁ
1H NMR, 13C[1H] NMR, 19F[1H] NMR spectra were
recorded on Bruker AC300 spectrometer, respectively
at 300, 75 and 282 MHz. The 1H and 13C chemical shifts
d are reported in ppm relative to TMS for solutions in
CDCl3 The symbols for the proton multiplicities are as
follows: t: triplet; d: doublet; m: multiplet and s: singlet.
For 13C NMR, Cq is the symbol for quaternary carbon. 19F
chemical shifts are reported in ppm relative to CF3COO
13C NMR (CDCl3) d/TMS: 165.5 (C=N), 161.9
(1JCF ˆ 243:4 Hz, Cq of C6H4F), 130.4 (3JCF ˆ 7:9 Hz,
CH of C6H4F), 130.9 (4JCF ˆ 3:2 Hz, Cq of C6H4F),
115.4 (2JCF ˆ 21:2 Hz, CH of C6H4F), heterocyclic car-
bons: 74.2 (CH2), 61.4 (CH); 34.0 (CH2±Ar); 21.4 (CH3).
HRMS was performed in the electron impact mode (EI)
‡
using direct introduction. Accurate and observed M/e for M
ion were, respectively, 193.0903 and 193.0901 Da which

J.-C. Cherton et al. / Journal of Fluorine Chemistry 107 (2001) 387±395
395
[14] N. Bodor, J.H. Miller, Drugs of the future VI, Prodrugs, topical and -
ocular drug delivery, in: K.B. Sloan (Ed.), Drugs and the
Pharmaceutical Science, Vol. 5, Marcel Dekker, New York, 1992,
p. 49.
concordance (DM ˆ 0:0002) agrees with a molecular
weight of 193 for the compound Ia.
N-(1-methyl-2-hydroxyethyl)-(4-¯uorophenyl)-acetamide
VIa was prepared as previously described [19].
[15] R.F. Rekker, The Hydrophobic Fragmental Constant, Elsevier, New
York, 1977.
19F monitorings were performed with 5 Â 10 4 or 10 3 M
solutions of Ia (or Va, or VIa) in a phosphate buffer 0.1 M at
pH 7.4 (or at pH 6.3, cf. Table 1, Figs. 1 and 2), or in the
presence of biological media with 5% DMSO-d6, using the
optimised 19F NMR conditions described before. The IS
concentration was 10 3 M. For monitoring at pH 6.3, the
concentrations of Ia and IS were both 10 3 M.
13C spectrum of the reaction mixture resulting from the
evolution of Ia at pH 6.3: a 8:8 Â 10 2 M solution of
substrate in DMSO-d6 phosphate buffer 0.1 M at
pH ꢀ 6:3 (20/80) was left to develop for 2 h, then the
13C[1H] spectrum recorded in 12 000 scans (cf. Table 2).
[16] C. Loutelier, C. Lange, P. Cassier, J.-C. Cherton, J. Chromatogr. B
281 (1994) 656.
[17] C. Loutelier, J.-C. Cherton, C. Lange, M. Traris, A. Vey, J.
Chromatogr. A 738 (1996) 181.
[18] S.W. Applebaum, in: G.A. Kerkut, L.I. Gilbert (Eds.), Comprehen-
sive Insect Physiology, Biochemistry and Pharmacology, Pergamon
Press, Oxford, 1985, p. 282.
[19] S. Hamm, J.-C. Cherton, L. Menguy, R. Delorme, A. Louveaux, New
J. Chem. 23 (1999) 1239.
[20] T.M. Spotwood, J.M. Evans, J.H. Richards, J. Am. Chem. Soc. 89
(1967) 5052.
[21] J.T. Gerig, Biol. Magn. Reson. 1 (1978) 139.
[22] J.T. Gerig, Prog. Nucl. Magn. Reson. Spectrosc. 26 (1994) 293.
[23] M.A. Danielson, J.J. Falke, Ann. Rev. Biophys. Biomol. Struct. 25
(1996) 163.
[24] M.-C. Malet-Martino, R. Martino, Xenobiotica 19 (1989) 583.
[25] M.-C. Malet-Martino, R. Martino, Biochimie 74 (1992) 785.
[26] S. Hunjan, R.P. Mason, V.D. Metha, P.V. Kulkarni, S. Aravind, V.
Arora, P.P. Antich, Magn. Reson. Med. 39 (1998) 551.
[27] D.L. Veenstra, J.T. Gerig, Magn. Reson. Chem. 36 (1998) 169.
[28] S.A. Mabury, D.G. Crosby, J. Agric. Food Chem. 43 (1995) 1845.
[29] A. Rollins, J. Barber, R. Elliott, B. Wood, Plant Phys. 91 (1989)
1243.
Acknowledgements
C. Robert-Labarre and M.-J. Pouet are warmly thanked
for NMR discussions, and A. Louveaux for the biological
locust tissues.
References
[30] R.D. Mortimer, D.B. Black, B.A. Dawson, J. Agric. Food Chem. 39
(1991) 1781.
[1] D. Haidukewych, A.I. Meyers, Tetrahedron Lett. 30 (1972) 3031.
[2] A.I. Meyers, G. Knaus, K. Kamata, J. Am. Chem. Soc. 96 (1974)
268.
[31] M.G. Boersma, T.Y. Dinarieva, W.J. Middelhoven, W.J.H. Van Berkel
Doran, J.J. Vervoort, I.M.C.M. Rietjens, Appl. Environ. Microbiol.
64 (1998) 1256.
[3] P. Le Perchec, Janssen Chim. Acta 3 (1986) 3.
[4] M. Reuman, A.I. Meyers, Tetrahedron report 181, Tetrahedron 41
(1985) 837.
[32] E.P. Mazzola, A.P. Borsetti, S.W. Page, D.W. Bristol, J. Agric. Food
Chem. 32 (1984) 1102.
[33] A.M. Serre, C. Roby, A. Roscher, F. Nurit, M. Euvrard, M. Tissut, J.
Agric. Food Chem. 45 (1977) 242.
[5] F.L.M. Pattisson, W.C. Howell, A.J. Mac Namara, J.-C. Schneider,
J.F. Walker, J. Org. Chem. 21 (1956) 739.
[34] M.D. Vaughan, P. Cleve, V. Robinson, H.S. Duewel, J.F. Honek, J.
Am. Chem. Soc. 121 (1999) 8475.
[6] D. Hedley, B.P.S. Khambay, A.M. Hooper, R.D. Thomas, A.L.
Devonshire, Pestic. Sci. 53 (1998) 201.
[35] A. Cavellec, J.-C. Cherton, I. Lepotier, P. Cassier, Analusis 24 (1996)
240.
[7] A. Hirashima, K. Shinkai, C. Pan, E. Kuwano, E. Taniguchi, M. Eto,
Pestic. Sci. 55 (1999) 119.
Á Â
[8] R. Oumeddour, These de Doctorat, Universite P. et M. Curie, Paris,
[36] L. Menguy, S. Hamm, J.-C. Cherton, Spectroscopy 13 (1997) 137.
[37] E.M. Fry, J. Org. Chem. 15 (1950) 802.
1988.
[38] R.B. Martin, R.E. Hedrick, A. Parcell, J. Org. Chem. 29 (1964) 3197.
[39] R. Greenhalgh, R.M. Heggie, M.A. Weinberger, Can. J. Chem. 41
(1963) 1662.
[9] J.-C. Cherton, M. Amm, R. Oumeddour, D. Ladjama, O. Convert,
Analusis 18 (1990) 37i.
[10] J.-C. Cherton, D. Ladjama, D. Soyez, Analusis 18 (1990) 71.
Á Â
[11] A. Cavellec, These de Doctorat, Vol. VI, Universite Pierre et Marie
[40] P. Deslongchamps, Tetrahedron 31 (1975) 2463.
[41] R. Jha, J.T. Davis, Carbohyd. Res. 277 (1995) 125.
[42] C.R. Porter, H.N. Rydon, J.A. Schoffield, J. Chem. Soc. (1960) 2686.
[43] G.D. Prestwich, J.K. Mauldin, J.B. Engstrom, J.F. Carvalho, D.Y.
Cupo, J. Econ. Entomol. 76 (1983) 690.
Curie, Paris, 1995.
Á Â
[12] S. Hamm, These de Doctorat, Universite de Versailles, 1999.
[13] H. VorbruÈggen, K. Krolikiewicz, Tetrahedron 49 (1993) 9353.