Enantioselective synthesis of isotopically labeled homocitric acid lactone

Article abstract of DOI:10.1021/ol402802g

A concise synthesis of homocitric acid lactone was developed to accommodate systematic placement of carbon isotopes (specifically ¹³C) for detailed studies of this cofactor. This new route uses a chiral allylic alcohol, available in multigram quantities from enzymatic resolution, as a starting material, which transposes asymmetry through an Ireland-Claisen rearrangement.

Full text of DOI:10.1021/ol402802g

ORGANIC
LETTERS
2013
Vol. 15, No. 22
5615–5617
Enantioselective Synthesis of Isotopically
Labeled Homocitric Acid Lactone
Jared T. Moore,‡ Nadine V. Hanhan,‡ Maximillian E. Mahoney,‡ Stephen P. Cramer,‡,§
and Jared T. Shaw*,‡
Department of Chemistry, University of California, One Shields Avenue, Davis,
California 95616, United States and Physical Biosciences Division, Lawrence Berkeley
National Laboratory (LBNL), Berkeley, CA 94720
jtshaw@ucdavis.edu
Received September 27, 2013
ABSTRACT
A concise synthesis of homocitric acid lactone was developed to accommodate systematic placement of carbon isotopes (specifically 13C) for
detailed studies of this cofactor. This new route uses a chiral allylic alcohol, available in multigram quantities from enzymatic resolution, as a
starting material, which transposes asymmetry through an IrelandÀClaisen rearrangement.
Homocitric acid lactone (1, Scheme 1) in the hydrolyzed
ring-open form is an essential component of nitrogenase,
the enzyme responsible for fixation of atmospheric nitro-
gen by bacteria and archaea. Homocitrate is incorporated
into the nitrogenase FeÀMo cofactor before its introduc-
tion into the nifDK protein matrix.1 Homocitrate coordi-
nates to the Mo site of the FeÀMo cofactor in a bidentate
fashion through alkoxide and carboxylate functionalities
(atoms 1 and 6, respectively, Scheme 1), creating a 5-mem-
bered ring in one of the most complex metal cofactors
found in nature.2 Replacement of homocitrate by citrate, a
deletion of a single methylene group, reduces N2 reduction
activity to only 7% that of wild-type enzyme.3 Shah and
co-workers studied the ability of a wide range of homo-
citrate analogues to reconstitute nitrogenase activity
in mutants lacking homocitrate and concluded that the
minimal requirements were the hydroxyl group, the
1- and 2-carboxyl groups, and the R configuration of the
stereogenic center.4 Based on these observations, we de-
vised a synthesis of 1 that incorporated 13C-labels at
these positions. Specific labeling of these functionalities
would open up a number of NMR, IR, and EPR/ENDOR
experiments related to nitrogenase biosynthesis and
mechanism.
The difficulty of isolating homocitric acid from natural
sources has prompted many synthetic efforts directed to-
ward this compound to support biological studies. Two
racemic syntheses have been reported, the latter of which
employs a biomimetic approach that efficiently produces
(()-1 in three steps.5 Several approaches using chiral pool
materials as auxiliaries or for semisynthesis offer access to
(R)-1 and (S)-1 with varying degrees of efficiency.6 Finally,
two approaches employ asymmetric catalysis for installa-
tion of the stereogenic center in 1.7 Although these routes
are effective for making enantiomerically enriched 1, they
(4) Imperial, J.; Hoover, T. R.; Madden, M. S.; Ludden, P. W.; Shah,
V. K. Biochemistry 1989, 28, 7796–7799.
(5) (a) Chen, H.-B.; Chen, L.-Y.; Huang, P.-Q.; Zhang, H.-K.; Zhou,
Z.-H.; Tsai, K.-R. Tetrahedron 2007, 63, 2148–2152. (b) Zeng-Chun, Li,
J.-Q. X. Molecules 1998, 3, 31–34.
(6) (a) Pansare, S. V.; Adsool, V. A. Tetrahedron Lett. 2007, 48, 7099–
7101. (b) Rodriguez, R. G.; Biellmann, J. F. J. Org. Chem. 1996, 61,
1822–1824. (c) Xu, P.-F.; Matsumoto, T.; Ohki, Y.; Tatsumi, K.
Tetrahedron Lett. 2005, 46, 3815–3818.
(7) (a) Paju, A.; Kanger, T. n.; Pehk, T. n.; Eek, M.; Lopp, M.
Tetrahedron 2004, 60, 9081–9084. (b) Pansare, S. V.; Adsool, S. V.;
Dyapa, R. Tetrahedron: Asymmetry 2010, 21, 771–773.
‡ University of California.
§ Lawrence Berkeley National Laboratory (LBNL).
(1) (a) Hu, Y.; Ribbe, M. W. J. Biol. Chem. 2013, 288, 13173–13177.
(b) Hu, Y.; Ribbe, M. W. Biochim. Biophys. Acta Bioenerg. 2013, 1827,
1112–1122.
(2) (a) Kim, J.; Rees, D. C. Nature 1992, 360, 553–560. (b) Mayer,
S. M.; Lawson, D. M.; Gormal, C. A.; Roe, S. M.; Smith, B. E. J. Mol.
Biol. 1999, 292, 871–891. (c) Spatzal, T.; Aksoyoglu, M.; Zhang, L.;
Andrade, S. L.; Schleicher, E.; Weber, S.; Rees, D. C.; Einsle, O. Science
2011, 334, 940.
(3) McLean, P. A.; Dixon, R. A. Nature 1981, 292, 655–656.
r
10.1021/ol402802g
2013 American Chemical Society
Published on Web 11/01/2013

are not easily translated to allow for selective incorporation
of carbon isotopes.
studies, they can easily be incorporated into 4 and 6.
For our purposes, carbon atoms 5 and 6 would be labeled
by using (13C)2-diethyl oxalate. Herein we present an effici-
ent asymmetric synthesis of (R)-1 that is scalable, suitable
for preparing labeled analogues, and derives asymmetry
from an enzyme-catalyzed kinetic resolution.
The Claisen rearrangement substrate 3 was assembled
in a short sequence (Scheme 2). Commercially available
allylicalchol(()-4 wasresolvedtoafford(À)-4 withacrylic
bound C. antarctica lipase and vinyl propionate in 94%
yield (based on a 50% conversion) with enantioselectivity
of 99:1 as judged by chiral HPLC analysis of the corre-
sponding 3,5-dinitrobenzoate ester. Diethyl oxalate was
treated with freshly prepared 3-butenylmagnesium bromide
to produce 7 in 86% yield,10 which was then reduced with
NaBH(OAc)3 and O-alkylated with PMB-trichloroacetimi-
date. Saponification and coupling to (À)-4 yielded 3 in 76%
yield as a 50:50 mixture of diastereomers. Similar yields were
We required an efficient, scalable synthesis that would
allow for the systematic placement of 13C isotopic labels at
positions 5 and 6 on 1. We initially attempted the route dis-
closed by Pansare, which assembles (R)-1 from three mod-
ular fragments and employs ephedrine as a chiral auxiliary.
This route was unsuitable for two reasons. First, the assem-
bly of the morpholinedione7b required oxalyl chloride, which
worked as described to make the unlabeled intermediate.
However, it would be difficult to handle oxalyl chloride as a
labeled reagent prepared from oxalic acid, given its air
sensitivity. Second, and more importantly, ephedrine is a
controlled substance that has become difficult to acquire.
Our laboratory was unable to identify a vendor that would
ship useful quantities to meet our synthetic requirements as
a chiral auxiliary. Attempts to circumvent this problem with
other amino alcohols (not shown) were not fruitful.
Mindful of the limited commercial availability of iso-
topically labeled starting materials, a thorough retrosyn-
thetic analysis of 1 (Scheme 1) identified diethyloxalate as
the optimal precursor (Scheme 1). This material is readily
available with both carbonyl carbons as 13C in high
abundance. Moreover, it can be synthesized from per-13C
D-glucose by an established procedure for large-scale pre-
parations. Use of an ester enolate Claisen rearrangement8
would install a fragment in the appropriate oxidation state
and would also derive asymmetry from a chiral allylic
alcohol. In this case, (()-4 is commercially available and
easily resolved with C. antarctica lipase in high yield and
with high selectivity.9Additionally, alkene placement fol-
lowing this rearrangement is ideal for installing carboxylic
acid functionality by oxidative cleavage at a late stage in
the synthesis. It is important for this oxidation to occur in
the last step because small molecules with multiple car-
boxylic acid groups can be difficult to purify and isolate.
Importantly, if additional labels are required for future
Scheme 2. Synthesis of IrelandÀClaisen Precursora
a Yields in brackets refer to 13C-labeled intermediates.
Scheme 1. Retrosynthetic Analysis of (R)-1
Scheme 3. Ireland Ester Enolate Claisen Rearrangementa
a Yields in brackets refer to 13C-labeled intermediates.
(8) Burke, S. D.; Fobare, W. F.; Pacofsky, G. J. J. Org. Chem. 1983,
48, 5221–5228.
(9) Chojnacka, A.; Obara, R.; Wawrzenczyk, C. Tetrahedron: Asym-
metry 2007, 18, 101–107.
(10) Macritchie, J. A.; Silcock, A.; Willis, C. L. Tetrahedron: Asym-
metry 1997, 8, 3895–3902.
5616
Org. Lett., Vol. 15, No. 22, 2013

is significantly lower in energy than the diastereomeric
conformation in which the allylic n-butyl group is in an
axial position. Although acid 12 could be isolated in high
puritybycolumnchromatography, itwastaken through to
methyl ester 2 in 96% overall yield from 3.
Scheme 4. RCM, Late-Stage Oxidation, and Analysis of
Enantiopuritya
Conversion of 2 to homocitric acid lactone was straight-
forward (Scheme 4). Initial attempts at direct oxidative
cleavage of 2 followed by removal of the PMB group were
low yielding and required difficult separation of the triacid
from the aqueous reaction medium. We reasoned that
cyclohexene 13 would require fewer oxidizing equivalents
and might be cleaved under milder conditions. Diene 2
was smoothly cyclized using Grubbs’s second-generation
catalyst and the PMB group was removed in high yield. At
this stage, the enantiomeric purity was measured at 95:5 er
(90% ee) by conversion to 3,5-dinitrobenzoate ester 14,
which was analyzed by chiral HPLC. Cyclohexene 14
was oxidized, hydrolyzed and dehydrated to provide 1 in
82% yield. The optical rotation of1 was À21.4°, which is in
agreement with the literature value.
a Yields in brackets refer to 13C-labeled intermediates and products.
In conclusion, a scalable enantioselective synthesis of
homocitric acid lactone, (À)-1, in nine steps including
an IrelandÀClaisen rearrangement and a ruthenium-
catalyzed oxidative cleavage is presented. This sequence
allows for the incorporation of carbon isotopes at the
carbon atoms proximal to molybdenum in the MoÀFe
complex. Isotopically labeled homocitrate analogs made
by this route will aid efforts aimed at further elucidating
the mechanism and activity of bacterial and archeal
nitrogenase.
observed when isotopically labeled diethyl oxalate was used
as the starting material (Scheme 2, shown in brackets).
Claisen rearrangement of 4 proceeded in high yield and
with a high degree of stereochemical transposition. After
initial attempts with various amide bases including LDA
and LiHMDS, treatment of 3 with KHMDS followed by
silylation with TMSCl and warming to room temperature
proceeded with high conversion. This was followed by
methylation with CH3I/K2CO3 to produce methyl ester 2
in high yield with no detectable trace of the Z-isomer. The
enantiomeric ratio was determined to be 95:5 by analysis
of a subsequent intermediate (Scheme 4, vide infra), de-
monstrating only a slight loss of enantiomeric purity from
(À)-4. The sense of stereochemistry of this reaction is
explained by a chair transition state emanating from the
Z-enolate (Scheme 3). The slight loss of enantiomeric
purity can potentially be explained by small quantities
of the E-enolate of 9. The high selectivity for formation
of the E-alkene isomer suggests that transition state 10
Acknowledgment. This work was supported by the NIH/
NIGMS (NIH Grant No. GM-65440) and DOE OBER.
Supporting Information Available. Experimental proce-
dures, compound characterization data for all new com-
pounds, and HPLC analysis of 14. This material is available
free of charge via the Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 22, 2013
5617