Synthesis of the antibiotic (R)-reutericyclin via Dieckmann condensation

Article abstract of DOI:10.1002/hlca.200590226

(R)-Reutericyclin ((R)-1), a bactericidal, amphiphilic natural product with a trisubstituted tetramic acid moiety, was prepared in four steps from D-leucine in anoverall yieldof 24%. The chiral heterocyclic portion of 1 was synthesized by Dieckmann cyclization of ethyl N-(acetoacetyl)leucinate (7), and the resulting pyrrole derivative 8 was N-acylated with (E)-dec-2-enoyl chloride in the presence of BuLi at-70° (Scheme2). This new procedure is straightforward and allows the synthesis of both antipodes of reutericyclin in anenantiomeric excess (ee) of ca. 80%.

Full text of DOI:10.1002/hlca.200590226

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2837
Synthesis of the Antibiotic (R)-Reutericyclin via Dieckmann Condensation
by Roswitha Böhmea), Günther Jung*b), and Eberhard Breitmaiera)
a) Kekulé Institute of Organic Chemistry and Biochemistry, University of Bonn, Gerhard-Domagk-Strasse 1,
D-53121 Bonn (e-mail: info@ebmaier.de)
b) EMC microcollections GmbH, Sindelfinger Strasse 3, D-72072 Tübingen
(e-mail: emc@microcollections.de)
(R)-Reutericyclin ((R)-1), a bactericidal, amphiphilic natural product with a trisubstituted tetramic acid
moiety, was prepared in four steps from D-leucine in an overall yield of 24%. The chiral heterocyclic portion
of 1 was synthesized by Dieckmann cyclization of ethyl N-(acetoacetyl)leucinate (7), and the resulting pyrrole
derivative 8 was N-acylated with (E)-dec-2-enoyl chloride in the presence of BuLi at ꢀ708 (Scheme 2). This new
procedure is straightforward and allows the synthesis of both antipodes of reutericyclin in an enantiomeric
excess (ee) of ca. 80%.
Introduction. – Reutericyclin (=(2R)-4-acetyl-1,2-dihydro-5-hydroxy-2-(2-methyl-
propyl)-1-[(2E)-1-oxodec-2-enyl]-3H-pyrrol-3-one; (R)-1)1) is a tetramic acid that
acts against a broad spectrum of Gram-positive bacteria. The amphiphilic metabolite
was the first low-molecular-weight antibiotic isolated from a lactic acid bacterium,
i.e., Lactobacillus reuteri LTH2584 [1][2]. Reutericyclin is produced during food indus-
trial sourdough fermentation by Lactobacillus reuteri, and contributes to the resistance
of its producer strain [3]. The mode of action is based on its proton-ionophoric proper-
ties, selectively dissipating the transmembrane proton potential (DpH) [4][5].
There are two general pathways to synthesize reutericyclin (1). In the first published
synthesis [6], leucine tert-butyl ester is N-acylated with (E)-dec-2-enoyl chloride to the
amide 2 (Scheme 1). After ester hydrolysis, the resulting N-acylamino acid is reacted
with Meldrum’s acid to the intermediate condensation product 3, which thermally
cyclizes to the tetramic acid 4. The Ac group is introduced during the last of the
seven steps (starting from leucine) to yield racemic 1. Racemization is due to the last
1
)
Systematic name of the enolic form drawn (tautomerism; see Figure in the Exper. Part).
© 2005 Verlag Helvetica Chimica Acta AG, Zürich

2838
HELVETICA CHIMICA ACTA – Vol. 88 (2005)
step in which TiCl4, a strong Lewis acid, is used as a catalyst. The separation of the enan-
tiomers of racemic reutericyclin is expected to be difficult for large-scale preparations.
Furthermore, the reaction conditions for the thermal cyclization with Meldrum’s acid
have to be strictly followed to obtain reproducible results. For these reasons, we devel-
oped a second synthetic pathway aiming at the synthesis of the two antipodes of 1.
Scheme 1. Published Synthesis of Racemic 1 [6]
Results and Discussion. – To avoid the vigorous conditions typically required for C-
acylation, we designed a synthetic route that includes N-acylation with (E)-dec-2-enoyl
chloride in the last step. As outlined in Scheme 2, the Ac group was introduced first by
coupling -leucine ethyl ester (5) with N-hydroxysuccinimidyl acetoacetate (6) [7]. The
D
resulting leucinate 7 was cyclized to the 1,2-dihydro-3H-pyrrol-3-one 8 via Dieckmann
condensation in the presence of EtONa/EtOH, following a modified procedure to
cyclize pyrrole derivatives [8]. Alternative attempts to introduce the acetoacetyl
group and to carry out the Dieckmann cyclization failed.
The last, crucial step involved N-acylation of tautomeric 8 (cf. Figure) with (E)-dec-
2-enoyl chloride. Since 8 is a cyclic amide, several bases such as DMAP, Et3N/N,N-
dimethylaniline (via an Me3SiN group), pyridine (Schotten–Baumann conditions),
and LiNH2 were tested as acylation catalysts. However, no coupling product with the
acid chloride was obtained. Similarly, when (E)-dec-2-enoic acid was activated with
1-hydroxy-1H-benzotriazole (HOBt) or transformed into a mixed anhydride with iso-
butyl chlorocarbonate (using Et(i-Pr)2N or DBU as base), no reaction to 1 was
observed. Attempted condensations by fusion of the neat compounds or in the presence
of 18-crown-6 in dimethoxyethane (DME) were also unsuccessful. Finally, BuLi (2.2
equiv.)2) was found to remove the amide proton of 8 at ꢀ708 under salt formation,
and, after 20 min of vigorous stirring, the resulting strong nucleophile was reacted
with dec-2-enoyl chloride. After aqueous workup and column chromatographic purifi-
cation, (R)-1 was obtained as an amorphous powder.
Further purification was achieved by recrystallization. Several solvents were tested
to this end, especially EtOH, Me2CO, AcOEt, Et2O, MeOH, and petroleum ether.
2
)
Only few cases are known in which unbranched BuLi was used in combination with NH compounds [9].

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2839
Scheme 2. Synthesis of (R)-1 from Ethyl -Valinate
D
a) CH2Cl2, r.t., 3 h. b) EtONa/EtOH, D, 3 h. c) 1. THF, BuLi (2.2. equiv.), ꢀ708; 2. (E)-dec-2-enoyl chloride;
3. H3O+.
Since reutericyclin (1) possesses pronounced amphiphilic properties due to its lipo-
philic alkyl chain and the hydrophilic tetramic acid part, it is soluble in many solvents.
In most of the above solvents, it formed a gel upon cooling, rather than crystals; and
slow evaporation of the solvent rendered an oily residue. Nevertheless, we finally ach-
ieved to obtain a solid material from a cyclohexane/acetone mixture upon slow evapo-
ration.
We observed that the NMR spectra of reutericyclin (1) seem to depend on the
workup procedure and NMR solvent, which may give rise to different tautomeric
forms and H-bonding patterns. Due to staple effects, the 1H NMR spectra of some reu-
tericyclin preparations showed only poorly resolved signals, with almost no fine struc-
1
tures. However, it was possible to record acceptable H- and 13C-NMR spectra and to
perform a HMQC experiment with a dilute (D6)DMSO solution. The complete NMR
data of 1 have been previously published [1][6].
After total hydrolysis and derivatization of the samples of synthetic 1, GC/MS anal-
ysis on a chiral capillary column showed an enantiomeric purity of ca. 90% (80% ee) of
the corresponding
L- or
D
-leucine, depending on the configuration of the original sub-
strate 5.
Experimental Part
General. M.p.: Büchi SMP-20 apparatus; uncorrected. UV Spectra: Beckmann DU660 spectrometer; lmax
(e) in nm. Optical rotation: Perkin-Elmer 341 polarimeter. IR Spectra: Perkin-Elmer 1600 spectrometer; in
cmꢀ1. NMR Spectra: Bruker DPX-400 and DRX-500 spectrometers; at 400 or 500 (1H), or at 100 or 125
MHz (13C), resp. HR-EI-MS: Kratos MS-50 spectrometer (A.E.I., Manchester).

2840
HELVETICA CHIMICA ACTA – Vol. 88 (2005)
Figure. 1H- and 13C-NMR (italics) chemical shifts and multiplicities for compound 8 present in two major
tautomeric forms. The observed ratio of A/B was ca. 4 :1 in CDCl3 solution at 208. The two Me groups of
the i-Bu side chain were tentatively assigned.
(E)-Dec-2-enoic Acid. This compound was prepared through Knoevenagel condensation of octanal and
malonic acid [6][10]. Variations of temp., base, and reaction time were tested, but always a certain amount of
b-hydroxydecanoic acid was detected, which could not be separated completely from the desired product. A
most-useful old variant [11] proved to be stirring at r.t. for 36 h. Yield: 70% (for 0.2 mol).
(E)-Dec-2-enoyl Chloride. This compound was prepared by heating the acid with SOCl2 followed by distil-
lation [6]. Yield 76% (for 0.12 mol).
Ethyl D-Leucinate Hydrochloride (5). D-Leucine was heated under stirring in EtOH soln. saturated with
HCl gas [12]. Yield: 87% (for 38 mmol). [a]2D0 =ꢀ18 (c=5, EtOH; [a]2D0 of commercially available
L-enan-
tiomer: +18).
Ethyl N-(Acetoacetyl)-D-leucinate (7) [7]. To a soln. of 5 (4.892 g, 25 mmol) in CH2Cl2 (40 ml) was added 2
equiv. of Et(i-Pr)2N (6.463 g, 8.56 ml, 50 mmol). Under Ar gas, a previously prepared suspension of commercial
N-(hydroxysuccinimidyl)acetoacetate (6; 4.979 g, 25 mmol) in CH2Cl2 (40 ml) was added through a dropping
funnel within 15 min, and the resulting mixture was stirred at r.t. for 3 h. The soln. was washed with 5% aq.
HCl (5×40 ml), and the aq. layers were re-extracted with CH2Cl2 (3×50 ml). The combined org. layers were
washed with sat. NaHCO3 soln. (3×) and H2O (1×). After drying (Na2SO4) and evaporation of the solvent, a
pale yellow oil was obtained, which was used without further purification. Yield of 7: 5.71 g (94%). Mr 243.30
1
g/mol. H-NMR (CDCl3): 0.93, 0.94 (2d, 2 Me of i-Bu); 1.27 (t, MeCH2O); 1.55–1.66 (m, CH2 of i-Bu)); 1.68
(m, Me2CH); 2.27 (s, Ac); 3.43 (s, AcCH2); 4.18 (q, OCH2); 4.59 (td, NCH); 7.17 (d, NH). 13C-NMR
(CDCl3): 14.2 (MeCH2O); 22.0, 22.9 (2 Me of i-Bu); 25.0 (Me2CH); 31.0 (Me of Ac); 41.4 (CH2 of i-Bu));
49.7 (AcCH2); 51.1 (NCH); 61.4 (OCH2); 165.4 (COO); 172.7 (NCO); 204.1 (C=O of Ac).
(2R)-4-Acetyl-1,2-dihydro-5-hydroxy-2-(2-methylpropyl)-3H-pyrrol-3-one (8)1). To a soln. of EtONa, pre-
pared by dissolving elemental Na (0.58 g, 1.07 equiv.) in anh. EtOH (18 ml) under Ar, 7 (5.71 g, 23.5 mmol)
in EtOH (30 ml) was added through a dropping funnel within 25 min at r.t., and the resulting mixture was
refluxed for 3 h. After cooling down, the mixture was neutralized with 10% AcOH in EtOH (4.57 ml), and
the solvent was evaporated. CH2Cl2 (25 ml) was added to the pale-yellow, gel-like residue, which was washed
with H2O (4×) to remove AcONa. The aq. layers were extracted with CH2Cl2 (3×), the combined org. layers
were dried (Na2SO4), and the solvent was removed on a rotary evaporator. The remaining solid was crystallized
from EtOH in several fractions to afford colorless 8. Yield: 3.152 g (68%). Mr 197.23 g/mol. M.p. 133–1348. UV
(EtOH): 215 (60130), 275 (146540). [a]2D0 =+1178 (c=0.1, EtOH)3). IR (KBr): 3226 (NH), 2960 (CH), 1669
(C=O), 1449 (NH), 1367, 1319, 1277, 1229. 1H- and 13C-NMR: see the Figure. EI-MS: 197.1 (1.2, M+), 182.1 (5.5,
[MꢀCH3]+), 154.1 (25, [MꢀAc]+), 141.1 (100, [MꢀC4H8]+), 123.1 (19, [MꢀC4H8ꢀH2O]+), 86.1 (30), 70.1 (20).
(R)-Reutericyclin (= (2R)-4-Acetyl-1,2-dihydro-5-hydroxy-2-(2-methylpropyl)-1-[(2E)-1-oxodec-2-enyl]-
3H-pyrrol-3-one; (R)-1). A 250-ml three-neck round-bottom flask equipped with a strong mechanical stirrer
was dried and evacuated several times, flooded with Ar gas, and then filled with a soln. of 8 (2.958 g, 15.00
-Enantiomer: [a]2D0 =ꢀ1288 (EtOH).
3
)
L

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2841
mmol) in anh. THF (100 ml). At ꢀ708, 2.2 equiv. of BuLi (1.6
M
soln. in hexane; 21 ml, 33 mmol) was added
through a dropping funnel within 20 min under vigorous stirring. The soln. became turbid as a salt precipitated,
first yellowish, then light orange, and stirring was continued for 20 min. Then, 0.95 equiv. of dec-2-enoyl chloride
(2.83 g, 2.79 ml) in THF (20 ml) was added dropwise within 15 min at ꢀ708. The precipitate gradually disap-
peared, and the soln. was stirred for another 30 min at this temp. After removing the cooling bath, the mixture
was stirred for another 2 h, during which time the soln. reached r.t. Ice water (150 ml) with conc. H2SO4 (3.234 g,
33 mmol) was slowly added, and the aq. layer was separated and extracted with CH2Cl2 (3×50 ml). The org. lay-
ers containing THF were diluted with CH2Cl2 (70 ml), and then washed neutral with H2O (4×70 ml). After dry-
ing (Na2SO4) and evaporation of the solvents, an oily, slightly yellow residue was obtained (4.94 g), which was
purified by column chromatography (CC) on silica gel 60 PF254 (Merck 7747) with cyclohexane/Me2CO 3 :1
(column: 12 cm×6 cm (i.d.)). During chromatography, the yellow substance developed a red ring (probably
due to decomposition of a byproduct). Of the three major fractions, the first one was resubjected to CC (con-
ditions as above; column: 15 cm×3.5 cm (i.d.)); the other two fractions contained the pure product (TLC,
HPLC). Yield of 1: 1.985 g (38%). Mr 349.46 g/mol. M.p. 161–1648 (dec.). UV (EtOH): 228 (17870), 266
20
589
(23875), 278 (25165). Optical rotation (EtOH, c=0.29): [a]24036 =+69, [a]25046 =+19, [a]52708 =+15; [a]
([a]D)=+138. IR (KBr): 2927 (Me, CH2), 2856 (CH), 1718 (ketone C=O), 1622 (amide C=O), 1470, 1355,
1319, 1242, 1155, 980 (HC=CH), 761. 1H- and 13C-NMR: see [1][6]. EI-MS: 349.2 (7.5, M+), 293.1 (42,
[MꢀC4H8]+), 250.1 (18, [MꢀAc]+), 198.0 (20, [M+2ꢀC10H17O]+), 183.1 (25), 153.2 (71, C10H17O+), 141.1
(100), 69.1 (33) 55.1 (29). HR-EI-MS: 349.2244 (M+, C20H31NO4þ ; calc. 349.2235).
(S)-Reutericyclin. This enantiomer was prepared as described for (R)-1, but from (less-expensive) ethyl
leucinate. The enantiomeric purity of 1 was determined as follows: a sample of synthetic (S)-(1) from a batch of
ca. 2 g of product was hydrolyzed with 6 DCl/D2O at 1108 for 24 h. The dried hydrolysate was esterified with
DCl/EtOD at 1108 for 30 min, and the resulting ester was N-protected with trifluoroacetic anhydride to afford
N-trifluoroacetyl- -leucine ethyl ester. GC/MS Analysis of this derivative on a capillary coated with Chirasil-Val
[13] showed two fully separated peaks (DtR 2 min) corresponding to a mixture of 9.7 :90.3 (80.6% ee).
L-
N
D
D/L
Financial support from EMC microcollections GmbH, Tübingen, Germany, is gratefully acknowledged. We
thank Graeme Nicholson for chiral analysis of reutericyclin.
REFERENCES
[1] A. Höltzel, M. G. Gänzle, G. J. Nicholson, W. P. Hammes, G. Jung, Angew. Chem., Int. Ed. 2000, 39, 2766.
[2] M. G. Gänzle, A. Höltzel, J. Walter, G. Jung, W. P. Hammes, Appl. Environ. Microbiol. 2000, 66, 4325.
[3] M. G. Gänzle, R. F. Vogel, Int. J. Food Microbiol. 2002, 80, 31.
[4] M. G. Gänzle, R. F. Vogel, Appl. Environ. Microbiol. 2003, 69, 1305.
[5] M. G. Gänzle, Appl. Microbiol. Biotechnol. 2004, 64, 326.
[6] U. Marquardt, D. Schmid, G. Jung, Synlett 2000, 8, 1131.
[7] A. W. Trautwein, R. D. Süßmuth, G. Jung, Biorg. Med. Chem. Lett. 1998, 8, 2381.
[8] A. Treibs, A. Ohorodnik, Liebigs Ann. Chem. 1958, 611, 139; H. Bauer, Liebigs Ann. Chem. 1970, 736, 1;
G. Pfeiffer, H. Bauer, Liebigs Ann. Chem. 1976, 383.
[9] C. Casnati, M. R. Langella, A. Ricca, A. Umani-Ronchi, Tetrahedron Lett. 1964, 24, 1597.
[10] F. Bohlmann, H. Bornowski, Chem. Ber. 1961, 94, 3189.
[11] V. J. Harding, C. Weizmann, J. Chem. Soc. 1910, 97, 299.
[12] R. D. Gillard, R. Wootton, J. Chem. Soc. B 1970, 364.
[13] H. Frank, G. J. Nicholson, E. Bayer, J. Chromatogr. 1978, 167, 187.
Received July 11, 2005