Configurational analysis of cyclopropyl fatty acids isolated from Escherichia coli

Article abstract of DOI:10.1021/ol052550d

(Chemical Equation Presented) The absolute configuration of methyl lactobacillate and its 9,10 homologue, both isolated from Escherichia coli B-ATCC 11303, was found to be 11R,12S and 9R,10S, respectively.

Full text of DOI:10.1021/ol052550d

ORGANIC
LETTERS
2
006
Vol. 8, No. 1
9-81
Configurational Analysis of Cyclopropyl
Fatty Acids Isolated from Escherichia
coli
7
Laura J. Stuart, James P. Buck, Amy E. Tremblay, and Peter H. Buist*
Department of Chemistry, Carleton UniVersity, Ottawa, Ontario K1S 5B6
pbuist@ccs.carleton.ca
Received October 20, 2005
ABSTRACT
The absolute configuration of methyl lactobacillate and its 9,10 homologue, both isolated from Escherichia coli B
−ATCC 11303, was found to
be 11R,12S and 9R,10S, respectively.
Lipids containing cyclopropyl fatty acids such as 1-3 occur
targets for mechanism-based inhibitors, and an X-ray crystal-
lographic study has been published recently. Several in vitro
1,2
5
widely in microorganisms and in the seed oils of various
3
subtropical plants. Interest in these natural products has
studies on a related Escherichia coli enzyme utilizing simpler
6
-11
grown with the discovery that the pathogenicity of Myco-
bacterium tuberculosis is highly dependent on the presence
olefinic substrates have also been undertaken.
hypotheses
Earlier
1
2-14
featuring rate-limiting methyl transfer from
4
of cyclopropyl moieties in their membrane lipids. Thus,
S-adenosyl-L-methionine (SAM) to olefin followed by rapid
proton loss have gained support (Scheme 1). A metal-
Scheme 1. Proposed Mechanism for Biochemical
Cyclopropane Ring Formation via Methylation of the
Corresponding Unsaturated Substrate
mycobacterial cyclopropane synthases constitute promising
15
assisted, sulfonium ylid-carbenoid-type process has been
effectively ruled out on the basis of observed fluorine
*
To whom correspondence should be addressed. Phone: 613-520-2600
ext 3643. Fax: 613-520-3749.
1) Grogan, D. W.; Cronan, J. E., Jr. Microbiol. Mol. Biol. 1997, 61,
29.
(
4
(5) Huang, C.; Smith, C. V.; Glickman, M. S.; Jacobs, W. R.; Sacchettini,
J. C. J. Biol. Chem. 2002, 277, 11559.
(6) Molitor, E. J.; Paschal, B. M.; Liu, H.-w. ChemBioChem 2003, 4,
1352.
(
(
2) Cronan, J. E., Jr. Curr. Opin. Microbiol. 2002, 5, 202.
3) Bao, X.; Katz, S.; Pollard, M.; Ohlrogge, J. Proc. Natl. Acad. Sci.
U.S.A. 2002, 99, 7172.
4) Glickman, M. S.; Cox, J. S.; Jacobs, W. R. Mol. Cell. 2000, 5, 717.
(
(7) Iwig, D. F.; Booker, S. J. Biochemistry 2004, 43, 13496.
1
0.1021/ol052550d CCC: $33.50
© 2006 American Chemical Society
Published on Web 12/06/2005

substituent effects,6 SAM analogue7 and KIE studies,8
inductively coupled plasma-atomic emission spectrometry
analysis (ICP-AES),9 and X-ray crystallographic data.
Despite this progress, the facial selectivity of cyclopropa-
nation (methylenation) as it occurs in E. coli has not been
elucidated. Because cyclopropanation is catalyzed by a single
gene product in this organism, the enantioselectivity of initial
methyl transfer can be probed by determining the absolute
configuration of the two major cyclopropane fatty acids
found in E. coli lipids. Herein, we report on the results of
our stereochemical analysis.
fatty acid methyl ester did not affect the subsequent stere-
1
ochemical analysis. The diagnostic GC-MS, H NMR, and
5
13
C NMR data of biosynthetic 2 matched those of an
authentic standard in all respects (see Supporting Informa-
tion).
Quasisymmetrical cyclopropyl fatty acids such as 1 and 2
are only weakly optically active, which renders comparison
1
9-21
with chiral reference standards
problematic. However,
long-chain cyclopropyl fatty acids are readily oxidized to a
pair of separable, regioisomeric, keto derivatives and these
compounds can be easily correlated with the appropriate
reference compounds on the basis of their distinctive
The lipid fraction (1 g) of E. coli B-ATCC 11303 (Avanti
Polar Lipids, Inc., Alabaster, Alabama) was hydrolyzed
2
2
chiroptical properties. Thus, mild CrO
mg) yielded ketones 4 (7.9 mg, R ) 0.08, [SiO
Et O (10:1)]) and 5 (6.2 mg, R ) 0.11); in a similar manner,
6 (8.0 mg, R ) 0.11) and 7 (4.4 mg, R ) 0.13) were
3
oxidation of 1 (72
f
2
, Hexane/
(
refluxing 2 N KOH, 50% ethanol), and the free fatty acids
were isolated and methylated (BF /MeOH) essentially as
previously described.16 The fatty acid methyl ester fraction
FAME, 729 mg) was analyzed by GC-MS; the presence of
2
f
3
f
f
obtained from 2 (112 mg) (see Figure 1). The keto derivatives
(
two cyclopropyl fatty acids, methyl 9,10-methanohexade-
canoate 1 (20%) and its C-19 homologue commonly known
as methyl lactobacillate 2 (12%), was detected. The remain-
ing FAMEs were identified as methyl tetradecanoate (1%),
methyl hexadecanoate (36%), methyl octadecanoate (1%),
methyl (Z)-11-octadecenoate (28%), and methyl (Z)-9-
hexadecenoate (2%). This profile is typical of E. coli
17
FAME. The identity of each analyte was initially confirmed
through a comparison of retention time and mass spectral
characteristics of authentic standards. (Synthetic cyclopropyl
fatty acid methyl esters 1 and 2 were prepared from the
corresponding, commercially available, olefinic precursors
18
by a modified Simmons-Smith reaction. ) To isolate each
individual biosynthetic cyclopropyl fatty acid, the E. coli lipid
extract was chromatographed using reversed-phase HPLC
(
Whatman Partisil Magnum 9 10/50 ODS-2 column, 25%
EtOAc/ACN), and fractions enriched in 1 (238 mg) and 2
112 mg) were obtained from a total of 70 chromatographic
runs. Crude 1 was treated with meta-chloroperbenzoic acid
55% pure, 165 mg, 0.5 mmol) to convert coeluting olefinic
fatty acids to the more polar epoxides which were subse-
quently removed by flash chromatography (SiO , 10%
Figure 1. Comparison of [Φ] values obtained for ketones 4-7
D
(
derived from biosynthetic 1 and 2 with synthetic standards 8 and
22
9.
(
were separated by flash chromatography (SiO
2 2
, Hexane/Et O
2
[
10:1]) and identified on the basis of diagnostic mass spectral
EtOAc/hexanes). In this manner, 72 mg of purified biosyn-
fragmentation patterns which are typical for this class of
compounds. All analytical data (R
NMR, and C NMR data) matched those for authentic
standards obtained upon oxidation of synthetic 1 and 2 (see
Supporting Information). The optical rotation of each ketone
1
thetic 1 was obtained as a colorless oil; the GC-MS, H
22
1
f
values and MS, H
1
3
NMR, and C NMR data of this material correlated well
with those of a synthetic reference standard (see Supporting
Information). Crude 2 was not purified further to remove
methyl hexadecanoate because the presence of this saturated
13
21
21
was obtained (4, [R]
25.7 (c 0.62, Et O); 6, [R]
) +24.9 (c 0.44, Et
molecular rotations [Φ] were compared to the values
D
) -20.1 (c 0.70, Et
2
O); 5 [R]
) -17.4 (c 0.62, Et
2
O)), and the corresponding
D
)
2
1
+
[
2
D
2
O); 7
(
8) Iwig, D. F.; Grippe, A. T.; McIntyre, T. A.; Booker, S. J. Biochemistry
004, 43, 13510.
9) Courtois, F.; Guerard, C.; Thomas, X.; Ploux, O. Eur. J. Biochem.
004, 271, 4769.
10) Iwig, D. F.; Uchida, A.; Stromberg, J. A.; Booker, S. J. J. Am. Chem.
Soc. 2005, 127, 11612.
21
R]
D
2
2
D
(
2
2
obtained by Tocanne for related compounds 8 and 9, as
displayed in Figure 1.
(
(
(
(
(
(
11) Courtois, F.; Ploux, O. Biochemistry 2005, 44, 13583.
12) Lederer, E. Q. ReV. Chem. Soc. 1969, 23, 453.
13) Buist, P. H.; Maclean, D. B. Can. J. Chem. 1982, 60, 371.
14) Arigoni, D. Chimia 1987, 41, 9.
(19) Kobayashi, S.; Tokunoh, R.; Shibasaki, M.; Shinagawa, R.; Mu-
rakami-Murofushi, K. Tetrahedron Lett. 1993, 34, 4047. Note that the
specific rotations reported for the two enantiomers of synthetic 1 reported
in this paper are reversed in sign compared to those determined for analogous
enantiomer(s) of synthetic 220 and synthetic 3.
15) Cohen, T.; Herman, G.; Chapman, T. M.; Kuhn, D. J. Am. Chem.
21
Soc. 1974, 96, 5627.
(
(
16) Buist, P. H.; Behrouzian, B. J. Am. Chem. Soc. 1998, 120, 871.
17) Law, J. H.; Zalkin, H.; Kaneshiro, T. Biochim. Biophys. Acta 1963,
(20) Coxon, G. D.; Al-Dulayymi, J. R.; Baird, M. S.; Knobl, S.; Roberts,
E.; Minnikin, D. E. Tetrahedron: Asymmetry 2003, 14, 1211.
(21) Lou, L.; Horikawa, M.; Kloster, R. A.; Hawryluk, N. A.; Corey, E.
J. J. Am. Chem. Soc. 2004, 126, 8916.
7
0, 143.
18) Imai, N.; Sakamoto, K.; Takahashi, H.; Kobayashi, S. Tetrahedron
Lett. 1994, 35, 7045.
(
(22) Tocanne, J. F. Tetrahedron 1972, 28, 363.
80
Org. Lett., Vol. 8, No. 1, 2006

On the basis of these considerations, it is clear that 1 and
isolated from E. coli B-ATCC 11303 bear the (R)
include 10 (PHYLPA) isolated from the slime mold, Phys-
1
9
27
2
arum polycephalum, 11 (plakoside A) found in the
Caribbean sponge, Plakortis simplex, and methyl dihy-
configuration at the cyclopropyl methine carbon closest to
the carboxyl group (C-9 for 1 and C-11 for 2) and the (S)
configuration at the cyclopropyl methine carbon proximal
to the methyl terminus (C-10 for 1 and C-12 for 2). This
implies that (Z)-9-hexadecenoate and (Z)-11-octadecenoate
are attacked by the methylating agent, as shown in Scheme
2
5
drosterculate 3 isolated from L. plantarum. That 2 and 3
are produced as quasienantiomers in L. plantarum is of
particular interest and raises intriguing questions regarding
binding of regioisomeric substrates to cyclopropane syn-
2
5,28
thases.
These issues are relevant to the case of E. coli
2
3
2.
These results match those obtained for methyl lactoba-
Scheme 2. Enantioselectivity of Methyl Transfer to
(Z)-9-Hexadecenoyl and (Z)-11-Octadecenoyl Substrates in E.
coli
cyclopropane synthase in that this enzyme also methylenates
Z)-9-octadecenoate in addition to (Z)-11-octadecenoate.29
Interestingly, other (Z)-C-18 monoene positional isomers are
(
2
9
relatively poor substrates for this enzyme. It would be of
interest to compare the facial selectivity of (Z)-9-octade-
cenoate methylenation by E. coli cyclopropane synthase with
that found in the present work for the (Z)-11 isomer. In this
manner, one might gain new insights into the topology of
the active site of the E. coli enzyme, the details of which
could be correlated with new protein structural information
as this becomes available. Experiments designed to address
this issue are being planned.
cillate 2 isolated from two other microorganisms, Brucella
Acknowledgment. Financial support provided by NSERC
to P.H.B. (operating grant), L.J.S. (graduate scholarship), and
A.T. (undergraduate scholarship) is gratefully acknowledged.
24
25
milletensis and Lactobacillus plantarum , as well as those
for methyl dihydrosterculate 3 obtained from a phytochemical
26
source, Litchi chinensis. However, there have been several
reports of cases where the absolute configuration of long-
chain cyclopropyl fatty acid derivatives is reversed. These
Supporting Information Available: Experimental pro-
cedures and characterization data for racemic 1-7. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(
23) E. coli, typically cyclopropanates, preexisting (Z)-9-C-16 and (Z)-
1
1-C-18 olefinic fatty acyl chains at late exponential or early stationary
phase of growth. Both olefinic fatty acid derivatives are known to be
substrates of E. coli cyclopropane synthase. It is considered less likely
2
OL052550D
2
that 2 is a chain elongation product derived from 1 or that 1 is a â-oxidation
product of 2.
(27) Mori, K.; Tashiro, T.; Akasaki, K.; Ohrui, H.; Fattorusso, E.
Tetrahedron Lett. 2002, 43, 3719.
(
(
(
24) Tocanne, J. F.; Bergmann, R. G. Tetrahedron 1972, 28, 373.
25) Rasonyi, S. Diss. ETH # 11318, 1995.
26) Stuart, L. J.; Buist, P. H. Tetrahedron: Asymmetry 2004, 15, 401.
(28) Buist, P. H.; Pon, R. A. J. Org. Chem. 1990, 55, 6240.
(29) Ohlrogge, J. B.; Gunstone, F. D.; Ismail, I. A.; Lands, W. E. M.
Biochim. Biophys. Acta 1976, 431, 257.
Org. Lett., Vol. 8, No. 1, 2006
81