Highly terminal-selective epoxidation of linolenic acid with an amphiphilic iron porphyrin catalyst casted in bilayer membranes

Article abstract of DOI:10.1246/cl.2002.162

An amphiphilic iron porphyrin having four hexadecanoic acid chains on each face of the porphyrin was prepared and the catalytic epoxidation of linolenic acid with the heme-casted bilayer membranes was achieved in high regioselectivity at the terminal doub

Full text of DOI:10.1246/cl.2002.162

1
62
Chemistry Letters 2002
Highly Terminal-Selective Epoxidation of Linolenic Acid with an Amphiphilic Iron Porphyrin
Catalyst Casted in Bilayer Membranes
ꢀ
y
Yoshinori Naruta, Masa-oki Goto, Toshifumi Tawara, and Fumito Tani
Institute for Fundamental Research of Organic Chemistry, Kyushu University, Higashi-ku, Fukuoka 812-8581
y
Department of Chemistry, Faculty of Science, Kyoto University, Sakyo-ku, Kyoto 606-8501
(Received November 7, 2001; CL-011114)
An amphiphilic iron porphyrin having four hexadecanoic
acid chains on each face of the porphyrin was prepared and the
catalytic epoxidation of linolenic acid with the heme-casted
bilayer membranes was achieved in high regioselectivity at the
terminal double bond of the substrate. The orientation of the
porphyrin ring in the membrane was determined to be parallel to
the aqueous-bilayer interface by means of ESR.
In organic synthesis, the discrimination of similar functional
groups in one molecule both in electronical and sterical
equivalence is difficult task. However, enzymes accomplish such
selective molecular conversion with sophisticated molecular
recognition system in them. For example, cytochrome P450BM-3
is capable to hydroxidize long normal saturated fatty acids (C12 –
1
C18) at their terminal positions in good selectivity. From X-ray
2
a
crystallographic studies of substrate-bound and -unbound
P450BM-3, the polar interaction between a Tyr group (Tyr
1) in the substrate access channel and the fatty acid carboxylate
2b
5
plays a major role to fix the terminal position of the substrate in
close proximity to the reaction center. Though a functional model
system of such substrate recognition was applied to the
epoxidation of steroids and linoleic acid (two double bonds) with
use of a steroid-linked porphyhrin in bilayer membranes, the
discrimination of two double bonds in linoleic acid was
3
Figure 1. Schematic drawing of amphiphilic iron porphyrin
FeTP(C )P and linolenic acid casted in DPDAB bilayer.
DPDAB molecules around the substrate are omitted for clarity.
incomplete.
1
6
The key to accomplish good substrate recognition in an
enzyme-mimic system is the fixation of the relative arrangement
between the specific position of a substrate and the catalyst
reaction center. The combination of a cationic quaternary
ammonium amphiphile and an anionic carboxylate in a basic
media could promote the fixation of their relative positions in
close proximity by electrostatic interaction.
afford meso-tetrakis[2,6-di-(methoxycarbonylhexadecanyloxy)-
5
phenyl]porphyrin 1 in 17% yield. After iron insertion to 1,
5
FeTP(C )P was obtained by hydrolysis of the resultant iron
1
6
porphyrin under basic conditions (KOH-MeOH) and succeeding
dialysis of the mixture with a cellulose membrane.
We report here the excellent regioselective epoxidation of
linolenic acid by discrimination between its three cis double
bonds with use of a new amphiphilic porphyrin catalyst in a
bilayer membrane system.
To inspect the applicability of the amphiphilic heme to the
substrate recognition, epoxidation of linolenic acid having three
consecutive cis double bonds was performed in a cationic bilayer
membrane consisted of quaternary ammonium salt, dimethyldi-
We designed amphiphilic porphyrin FeTP(C )P as a
6
16
parmitylammonium bromide (DPDAB). Reaction was per-
ꢁ
catalyst, which has four fatty-acid chains on its both faces. In
bilayer membranes, the eight polar anchors could stretch out the
aliphatic chains on both of the bilayer surface. Consequently, the
heme could be fixed in the middle of the bilayer keeping the heme
plane parallel to the membrane surface, irrespective of the high
flexibility of the long fatty acid chains. Electrostatic interaction is
also expected between substrate linolenic acid and the cationic
amphiphile (Figure 1).
formed with use of PhIO as an oxidant at 34 C slight above the
ꢁ
15;16
Tc (28 C) of the bilayer in order to keep the sufficient fluidity of
the membrane. The regioselectivity of the terminal olefin (Á
in the epoxidation reached to 82% (Table 1). In comparison with
)
the statistical distribution of the isomers in the homogeneous
6
system, the observed selectivity in bilayer is very high. The
3
present results exceeded those of the preceding examples.
Furthermore, the total yields of the monoepoxide and recovered
starting material reached 93% in the bilayer system. Such high
material balance suggests the suppression of the product
decomposition and formation of the di- and triepoxides.
The synthesis of the porphyrins was performed as follows:
meso-Tetrakis(2,6-dihydroxyphenyl)porphyrin4 was alkylated
with methyl 16-iodohexadecanoate and K2CO3 under Ar to
Copyright Ó 2002 The Chemical Society of Japan

Chemistry Letters 2002
Table 1. Comparison of catalytic epoxidation of linolenic acid under different conditions
163
b
Catalyst
Reaction
conditions
Reaction
ꢁ
Isomeric ratio of epoxides/%
12;13
Terminal
Epoxide
Recovery
/%
temp./ C cis-Á9;10
cis-Á
cis-Á15;16 selectivityc total yield/%
FeTPPCla
FeTP(C )PCl
homogeneous, CH2Cl2
DPDAB, pH 9
0
34
22
7
34
11
44
82
0.5
4.6
12
26
46
67
16
a
b
TPP: meso-tetrakisphenylporphyrin. Stereochemistry of the epoxides was determined by the comparison of their authentic samples.
15;16
c
9;10
12;13
Terminal selectivity ¼ ½Á
]/[Á þÁ
].
In order to verify our hypothesis of the catalyst location in
Research ‘‘Design and Control of Advanced Molecular Assembly
Systems’’ from the Ministry of Education, Science, and Culture of
Japan (#08CE2005).
membrane, we determined the orientation of the heme in the lipid
bilayer by ESR. A Cu porphyrin is a good probe for the inspection
of its orientation in the bilayer membrane, because of the clear
differences between the isotropic and anisotropic spectra.
This paper is dedicated to Professor Teruaki Mukaiyama on
the occasion of his 75th birthday.
5
CuTP(C )P, prepared in the similar way as the iron derivative,
1
6
was incorporated into DPDAB bilayer and was deposited on
7
Myler films. The ESR of the film exhibited signals characteristic
of a square planer complex of Cu(II) ion (I ¼ 3=2) as shown in
References and Notes
1Y. Miura and A. J. Fulco, Biochim. Biophys. Acta, 388, 305
(
1975); R. T. Ruettinger and A. J. Fulco, J. Biol. Chem., 256,
5728 (1981).
Figure 2. These anisotropic g (g? ¼ 2:062 and g== ¼ 2:180) and
À1
hyperfine splitting values (A== ¼ 201 cm ) are resemble to the
2
a) H. Li and T. L. Poulos, Nature Struc. Biol., 4, 140 (1997). b)
K. G. Ravichandran, S. S. Boddupalli, C. A. Hasemann, J. A.
Peterson, and J. Deisenhofer, Science, 261, 731(1993).
J. T. Groves and R. Neumann, J. Am. Chem. Soc., 111, 2900
reported one of CuTPP (single crystal, g? ¼ 2:045, g== ¼ 2:190,
À1
8
A== ¼ 209 cm ). Thus, the simple saturated aliphatic acid
chains linked to the heme are sufficient to fix the catalyst in the
center of the bilayer with a desired orientation.
3
4
5
(
1989).
F. Tani, M. Matsu-ura, S. Nakayama, M. Ichimura, N.
Nakamura, and Y. Naruta, J. Am. Chem. Soc., 123, 1133 (2001).
All new compounds gave satisfactorily analytical data. Selected
ꢁ
1
data: 1, mp 110–114 C; H NMR (CDCl3) ꢀ 8.64 (s, 8H), 7.62
(
t, 4H, J ¼ 8:3 Hz), 6.94 (d, 8H, J ¼ 8:3 Hz), 3.73 (t, 166H,
J ¼ 6:8 Hz), 3.65 (s, 24H), 2.28 (t, 16H, J ¼ 6:8 Hz), À2:58 (s,
À1
2
H); IR (KBr) 2918, 2851, 1736, and 1464 cm ; MS (FAB,
NBA) m=z ¼ 2890; UV-vis (CH2Cl2); ꢁ ¼ 420, 515, 548, and
À
5
91nm. FeTP(C )P, ESI-MS m=z ¼ 2884 [M-2H-Cl] ; UV-
1
6
vis (MeOH) ꢁ ¼ 420 and 548 nm. CuTP(C )P, ESI-MS
1
6
À
m=z ¼ 2892 [M-H] ; UV-vis (MeOH); ꢁ ¼ 418 and 543 nm.
6
In a homogeneous solution: To a CH2Cl2 solution of FeTPPCl
(
0.2 ꢂmol) and linolenic acid (2 ꢂmol), PhIO (2 ꢂmol) was
ꢁ
added and stirred for 3 h at 0 C under N2 atmosphere. After the
evaporation of the solvent, the addition of stearic acid (2 ꢂmol)
as an internal standard for GC analysis, and succeeding
treatment of the mixture with an ethereal solution of CH2N2,
the resultant solution was analyzed by GC. In lipid bilayer: A
CHCl3 solution of linolenic acid (2 ꢂmol), DPDAB (20 ꢂmol),
and FeTP(C )PCl (0.2 ꢂmol) in a round-bottom flask was
1
6
evaporated to make a thin layer on the wall of the flask. Borate
buffer (8 ml, pH 9.14, Na2B4O710H2O 10 mmol in 1 kg water)
was added into the flask, followed by N2 substitution for 15 min.
Figure 2. ESR spectra of FeTP(C )P in an oriented bilayer
16
assembly. (top) Magnetic field is parallel to the Myler film,
ꢁ
ꢃ ¼ 0 . Spectral conditions: microwave pw 1.0 mW, microwave
ꢁ
ꢁ
Ultrasonication of this mixture for 5 min at 50 C (about 20 C
higher than TcÞ with a probe-type sonicator (40–60 W output)
under N2 gave the transparent liquid of vesicles. A solution of
PhIO (2 ꢂmol) in MeOH/water (100 ꢂl, 1 : 1 v/v) was added to
freq. 9.447 GHz, modulation freq. 100 MHz, temp. 18 K. The
angle ꢃ is defined as the angle between the magnetic field and the
normal to the plane of Myler strips.
ꢁ
the mixture at 34 C under slow stirring of the mixture. After 3 h,
In conclusion, we realized the regioselective epoxidation of
linolenic acid and established that the electrostatic interaction
between the heme, the substrate, and bilayer membrane worked as
shown in Figure 1. When one applies this system to other
substrates having different chain lengths, one could optimize the
system just by choosing appropriate chain length for the catalyst
and the amphiphile. In this way, the present system has enough
flexibility to adopt substrates and can be regarded as the prototype
of ‘tailor-made’ catalysts for terminal oxidation.
the solvent was evaporated under reduced pressure followed by
the same workup as that of the homogeneous system.
DPDAB vesicles containing 1mol% CuTP(C16)P were
prepared in borate buffer in the similar way as shown above.
The resultant turbid vesicle suspension was spread on Mylar
films. After drying in air, the homogeneously coated region of
the film was cut into 3-mm strips and ten of these strips were put
between two Teflon plates and all of them were banded together.
This piece was placed in an ESR tube in N2.
7
8
P. T. Monoharan and M. T. Rogers, in ‘‘Electron Spin
Resonance of Metal Complexes,’’ ed. by T. F. Yen, Plenum,
New York (1969), p 143.
This work is supported by the Grant-in-Aid for COE