Syntheses with a chiral building block from the citric acid cycle: (2R,3S)-isocitric acid by fermentation of sunflower oil

Article abstract of DOI:10.1002/anie.200705000

(Chemical Equation Presented) No longer just analytical: Previously, (2R,3S)-isocitric acid (1), a component of the citric acid cycle, had not been available on a preparative scale. A new route to this acid on a kilogram scale combines a biotechnological formation through fermentation from sunflower oil with a chemical separation process. In a variety of transformations into further chiral derivatives, 1 is established as a valuable new member of the chiral pool (see scheme).

Full text of DOI:10.1002/anie.200705000

Communications
DOI: 10.1002/anie.200705000
Isocitric Acid
Syntheses with a Chiral Building Block from the Citric Acid Cycle:
(2R,3S)-Isocitric Acid by Fermentation of Sunflower Oil**
Philipp Heretsch, Franziska Thomas, Andreas Aurich, Harald Krautscheid, Dieter Sicker, and
Athanassios Giannis*
The citric acid cycle constitutes a main metabolic process.
Since its discovery in 1937 by H. A. Krebs, all of its
intermediates have been prepared in multigramm
amounts—with one exception: (2R,3S)-isocitric acid (1, dS-
threo-isocitric acid). As a new member of the chiral pool it
would be an interesting starting material for organic synthesis.
This chiral a-hydroxy tricarboxylic acid is mainly accompa-
nied by its constitutional isomer, citric acid (2). However,
attempts to separate 1 from 2 have so far been successful only
on an analytical scale.
Though experiments have been carried out to achieve
synthetic access to ent-isocitric acid (ent-1), again only
milligram quantities were obtained.[1] As a result of the
scarce availability of 1 there is virtually no application known
for it in synthesis. In databases only (2R,3S)-isocitric acid
trimethyl ester (3), (2R,3S)-isocitric acid lactone-2,3-dicar-
boxylic acid dimethyl ester (5), (2R,3S)-isocitric acid lactone-
2,3-dicarboxylic acid (6), and (2R,3S)-isocitric acid lactone-
2,3-dicarboxylic acid anhydride (7) are listed superficially.[2]
Surprisingly, no attempts have been made to obtain 1 by
fermentation, although a large number of yeasts are known to
produce and excrete citric acid and (2R,3S)-isocitric acid in
varying ratios when grown on long-chain n-alkanes or
glucose.[3] So far, these fermentations have been optimized
for high levels of citric acid excretion.
excess of thiamine under nitrogen-limited, aerobic condi-
tions.[5] Our aim was to achieve the highest possible ratio of
isocitric to citric acid and concomitant high isocitric acid
concentration.
We succeeded in producing isocitric acid concentrations
of 93 gLꢀ1 and 1/2 ratios of 1.14:1 on the pilot-plant scale in
the cultivation of wild-type Y. lipolytica EH59 on refined sun
flower oil—a hitherto unrivalled achievement especially with
regard to the use of renewable vegetable raw materials. After
filtration of the biomass, electrodialysis was performed to
convert the obtained trisodium salts into the free acids, before
the removal of water was accomplished under reduced
pressure. Then, it was time to search for an adequate process
to separate the two isomers.
Esterification of the highly viscous concentrated solution
yielded the corresponding triesters of both tricarboxylic acids.
From this mixture citric acid trimethyl ester 4 crystallized as a
colorless solid, while (2R,3S)-isocitric acid trimethyl ester 3
did not, as it is a liquid under standard conditions. Utilizing
this formerly unknown fact, separation of the isomeric esters
3 and 4 could be carried out simply by filtration of 4 from 3
(Scheme 1). In view of the intended application of 1 in
Herein we describe a combination of ecologically desir-
able biotechnological and chemical methods yielding enan-
tiopure (2R,3S)-isocitric acid (1) and its derivatives in kilo-
gram amounts, thus representing an unadulterated applica-
tion of the “white biotechnology for green chemistry”
concept.[4] We discovered that the thiamine auxotrophic
yeast Yarrowia lipolytica excretes organic acids in high
percentage when it is grown on vegetable oils with an
Scheme 1. Esterification of the concentrated fermentation broth:
a) MeOH, 2,2-dimethoxypropane, 10 mol% trimethylsilyl chloride,
3 days, RT, 80–88%.
stereoselective synthesis, a variety of different building blocks
should become accessible by facile and efficient transforma-
tions. Therefore, we first searched for possible differentiation
methods of the three carboxylic acid moieties. The formation
of a five-membered lactone structure in 5 and 6 was easily
accomplished starting from 3. Both 5 and 6 could be
converted into 1 as the parent compound (Scheme 2).
[*] P. Heretsch, F. Thomas, Prof. Dr. D. Sicker, Prof. Dr. A. Giannis
Institut für Organische Chemie, Universität Leipzig
Johannisallee 29, 04103 Leipzig (Germany)
Fax: ( +49)341-973-6599
E-mail: giannis@uni-leipzig.de
Prof. Dr. H. Krautscheid
Institut für Anorganische Chemie, Universität Leipzig
Johannisallee 29, 04103 Leipzig (Germany)
The anhydride 7 of 6 turned out to be the key element in
most of the subsequent transformations. Selective ring open-
ing yielded exclusively the monoester derivatives 8 and 9 with
the ester groups at the C2 position (Scheme 3) and thus led to
compounds with three differentiated carboxylic acid moieties.
Crystalline (ꢀ)-menthyl ester 9 was also used to confirm the
absolute stereochemistry of the fermentation product by X-
ray crystallographic analysis (Figure 1).[6] Reduction of the
Dr. A. Aurich
Helmholtz-Zentrum für Umweltforschung—UFZ
Umwelt- und Biotechnologisches Zentrum (UBZ)
Permoserstrasse 15, 04318 Leipzig (Germany)
[**] P.H. is grateful for a PhD scholarship from the Fonds der
Chemischen Industrie (VCI).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
1958
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1958 –1960

Angewandte
Chemie
Scheme 5. Transformation of 6: a) 2-methylpropene, cat. H2SO4,
CH2Cl2, 5 d, 88%; b) Ca(BH4)2, THF/iPrOH, 84%.
compound 15 was used to find mild conditions to invert the
conformation at the C3 stereocenter without concomitant loss
of stereoinformation at C2 and hence complete racemization.
These optimum conditions were applied to invert the readily
available di-tert-butyl ester 13, giving rise to derivatives of the
non-naturally occurring (ꢀ)-allo-isocitric acid (Scheme 6).
Scheme 2. Conversion of 3: a) 10 mol% para-toluenesulfonic acid,
toluene, reflux, 80%; b) 4.0m HCl, reflux, 8 h, quant.; c) 1.0m HCl,
reflux, 4 h, quant.; d) 3 equiv NaOH, amberlite IR-120, quant.
Scheme 3. Formation of 7 and its regioselective opening: a) Ac2O,
1608C, 15 min, 85%; b) for 8: anhydrous tBuOH, reflux, 15 h, quant.;
for 9: (1R,2S,5R)-(ꢀ)-menthol, 1008C, 36 h, 27%.
Scheme 6. a) 2-Methylpropene, cat. H2SO4, CH2Cl2, 48 h, quant.;
b) inversion of the conformation at C3 for 16: DBU, CH2Cl2, 508C,
30 min, 80% (based on recovered starting material); for 17: DBU,
CH2Cl2, 508C, 1 h, 89% (based on recovered starting material).
DBU=1,8-diazabicyclo[5.4.0]undec-7-ene.
In summary, we succeeded in the production and isolation
of (2R,3S)-isocitric acid, a substance from the citric acid cycle
that was of only analytical interest until now. For the first
time, (2R,3S)-isocitric acid is now available in kilogram
amounts from the fermentation of vegetable oils such as
sunflower oil. Its further transformation to new chiral
building blocks points at its promising applications in natural
product synthesis and its use as a starting material in the
pharmaceutical industry.
Figure 1. Crystal structure of 9.[6]
Received: October 29, 2007
Published online: January 31, 2008
mono-tert-butyl ester 8 yielded alcohol 10, which rearranged
to give 11. An entry to the class of amino acids was opened
with the non-natural lactone–amino acid 12, which was
synthesized via 11 from 10 (Scheme 4). Selective reduction
of the lactone moiety was successful when starting from 13
and yielded the succinic acid derivative 14 (Scheme 5).
Keywords: chiral pool · citric acid cycle · green chemistry ·
isocitric acid
.
[1] a) J. D. Aebi, M. A. Sutter, D. Wasmuth, D. Seebach, Liebigs Ann.
Chem. 1983, 12, 2114 – 2126; b) C. Schmitz, A.-C. Rouanet-
Dreyfuss, M. Tueni, J.-F. Biellmann, J. Org. Chem. 1996, 61, 1817 –
1821.
[2] a) M. Tori, Y. Sono, Y. Nakashiba, N. Hamada, M. Sono, Y.
Asakawa, M. Suganuma, S. Okabe, H. Fujiki, Tetrahedron Lett.
2000, 41, 3095 – 3098; b) J. P. Greenstein, N. Izumiya, M. Winitz,
S. M. Birnbaum, J. Am. Chem. Soc. 1955, 77, 707 – 716; c) R.
Carrington, G. Halek, Ger. Offen. DE226473, 1973.
[3] a) G. Barth, C. Gaillardin in Nonconventional Yeasts in Biotech-
nology (Ed.: K. Wolf), Springer, Berlin, 1996, pp. 313 – 388;
b) T. V. Finogenova, I. G. Morgunov, S. V. Kamzolova, O. G.
Chernyavskaya, Appl. Biochem. Microbiol. 2005, 41, 418 – 486.
[4] U. Stottmeister, A. Aurich, H. Wilde, J. Andersch, S. Schmidt, D.
Sicker, J. Ind. Microbiol. Biotechnol. 2005, 32, 651 – 664.
[5] A. Aurich, A. Förster, S. Mauersberger, G. Barth, U. Stottmeister,
Biotechnol. Adv. 2003, 21, 454 – 455.
Scheme 4. Transformation and rearrangement of 8: a) BH3/THF, THF,
08C!RT, 5 h, 88%; b) H+, 78%; c) methanesulfonyl chloride, pyridine,
CH2Cl2, 24 h; d) NaN3, DMF, 47% over two steps; e) Pd/C, H2, EtOAc,
8 h, 90%.
Finally, we searched for methods to synthesize other
stereoisomers of isocitric acid. The unsymmetrically esterified
Angew. Chem. Int. Ed. 2008, 47, 1958 –1960
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1959

Communications
[6] Crystal data of C16H24O6: Stoe IPDS-2T diffractometer, Mm =
residual electron density peaks 0.21/ꢀ0.20 e ” 106 pm3; structure
solution and refinement: SHELX-97 (G. M. Sheldrick, SHELXS-
97, Program for the Solution of Crystal Structures, Universität
Göttingen, 1997). CCDC-665724 contains the supplementary
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
312.35 gmolꢀ1, crystal size 0.1 ” 0.4 ” 0.9 mm3, monoclinic, space
group P21 (no. 4), a = 821.95(9), b = 561.50(4), c = 1760.2(2) pm,
b = 97.141(9)8, V= 806.07(14) ” 106 pm3, Z = 2, T= 180(2) K,
1calcd = 1.287 gcmꢀ3
,
m = 0.098 mmꢀ1
,
l = 71.073 pm (MoKa),
10207 measured, 2805 independent reflections, Rint = 0.073, 2608
with I > 2s(I), 207 parameters, H atoms in idealized positions,
R1 (observed reflections) = 0.051, wR2 (all data) = 0.15, max/min
1960 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1958 –1960