Practical, Catalytic, Asymmetric Synthesis of ?-Lactones via a
Sequential Ketene Dimerization/Hydrogenation Process: Inhibitors
of the Thioesterase Domain of Fatty Acid Synthase
Vikram C. Purohit,†Robyn D. Richardson,‡Jeffrey W. Smith,*,‡and Daniel Romo*,†
Department of Chemistry, Texas A & M UniVersity, P.O. Box 30012, College Station, Texas 77842-3012
and Cell Adhesion and Extra Cellular Matrix Biology, Burnham Institute for Medical Research,
Infectious and Inflammatory Disease Center, La Jolla, California 92037
ReceiVed February 24, 2006
The recent finding that the FDA-approved antiobesity agent orlistat (tetrahydrolipstatin, Xenical) is a
potent inhibitor of the thioesterase domain of fatty acid synthase (FAS) led us to develop a concise and
practical asymmetric route to pseudosymmetric 3,4-dialkyl-cis-?-lactones. The well-documented up-
regulation of FAS in cancer cells makes this enzyme complex an interesting therapeutic target for cancer.
The described route to 3,4-dialkyl-?-lactones is based on a two-step process involving Calter’s catalytic,
asymmetric ketene dimerization of acid chlorides followed by a facial-selective hydrogenation leading
to cis-substituted-?-lactones. Importantly, the ketene dimer intermediates were found to be stable to flash
chromatography, enabling opportunities for subsequent transformations of these optically active, reactive
intermediates. Subsequent R-epimerization and R-alkylation or acylation led to trans-?-lactones and
?-lactones bearing R-quaternary carbons, respectively. Several of the ketene dimers and ?-lactones
displayed antagonistic activity (apparent Kiin the low micromolar range) in competition with a fluorogenic
substrate toward a recombinant form of the thioesterase domain of fatty acid synthase. The best antagonist,
a simple phenyl-substituted cis-?-lactone 3d, displayed an apparent Ki(2.5 ( 0.5 µM) of only ∼10-fold
lower than that of orlistat (0.28 ( 0.06 µM). In addition, mechanistic studies of the ketene dimerization
process by ReactionView infrared spectroscopy support previous findings that ketene formation is rate
The development of synthetic processes that deliver optically
active products from achiral starting materials in a practical and
catalytic manner continues to be a vital area of research in
organic synthesis.1In the area of ?-lactone synthesis, the
Wynberg process continues to be a benchmark for further
developments in this area.2Several groups have recently
developed various catalytic, asymmetric routes to ?-lactones
involving nucleophile-catalyzed aldol-lactonizations and Lewis-
acid-catalyzed [2+2] cycloadditions.3With regard to the former
route, we recently reported an intramolecular, nucleophile-
catalyzed aldol-lactonization process building on the work of
Wynberg that effectively joins catalytic, asymmetric carbocycle
synthesis with ?-lactone synthesis employing organocatalysis.4
This was the first example of this process with highly electro-
philic (e.g., non-R-chlorinated) aldehydes. More recently, this
* To whom correspondence should be addressed.
†Texas A & M University.
‡Burnham Institute for Medical Research.
(1) (a) Noyori, R. Asymmetric Catalysis in Organic Synthesis; John Wiley
& Sons: New York, 1994. (b) Williams, J. M. J. Catalysis in Asymmetric
Synthesis; Sheffield Academic Press: Sheffield, England, 1999.
(2) (a) Wynberg, H.; Staring, E. G. J. J. Am. Chem. Soc. 1982, 104,
166. (b) Wynberg, H.; Staring, E. G. J. Top. Stereochem. 1986, 16, 87.
(3) For reviews describing asymmetric routes to ?-lactones, see: (a)
Yang, H. W.; Zhao, C.; Romo, D. Tetrahedron 1997, 53, 16471. (b) Wang,
Y.; Tennyson, R. L.; Romo, D. Heterocycles 2004, 64, 605.
(4) (a) Cortez, G. S.; Tennyson, R. L.; Romo, D. J. Am. Chem. Soc.
2001, 123, 7945. (b) Cortez, G. S.; Oh, S. H.; Romo, D. Synthesis 2001,
1731. (c) Oh, S.; Cortez, G.; Romo, D. J. Org. Chem. 2005, 70, 2835.
10.1021/jo060392d CCC: $33.50 © 2006 American Chemical Society
Published on Web 05/18/2006
J. Org. Chem. 2006, 71, 4549-4558
limitation was overcome by use of Lewis-acid additives in the
Wynberg process by Nelson and Calter, allowing access to cis-
?-lactones in high enantiopurity.5In a related process, the Calter
group described a novel application of asymmetric organo-
catalysis for the dimerization of methylketene6and more recently
several in situ generated ketenes leading to pseudosymmetric,
chiral 3-alkyl-4-methylene-2-oxetanones 2 (Scheme 1, 1 f 2).7
These homodimers were typically not isolated but directly ring-
opened with lithiated secondary amines to provide enolates that
could undergo subsequent aldol reactions leading to polypro-
pionate fragments and ultimately to polyketide natural products.8
Important to the process described herein, the geometry of the
olefin in dimer 2 was determined to be Z by virtue of the
stereochemical outcome of the subsequent aldol process. In this
article, we describe hydrogenation of isolated, optically active
ketene dimers 2 leading to cis-?-lactones 3 and subsequent
R-epimerization and alkylations leading to trans-?-lactones 4
and R,R-disubstituted-?-lactones, respectively (Scheme 1). This
process provides a practical (two steps from acid chlorides) and
complimentary approach to simplified pseudosymmetric dialkyl-
Our long-standing interest in the development of enantiose-
lective methodologies for ?-lactone synthesis and our recent
discovery that tetrahydrolipstatin (orlistat) potently inhibits fatty
acid synthase (FAS)9prompted us to explore the potential of
optically active ketene dimers as precursors to pseudosymmetric
trans-3,4-dialkyl-?-lactones 4 and study the ability of these
simplified ?-lactones to inhibit the thioesterase domain of fatty
acid synthase (FAS TE) (Figure 1). Human fatty acid synthase
(FAS) is the enzyme responsible for cellular synthesis of
palmitate. FAS contains seven separate enzymatic pockets that
function sequentially to condense acetyl-CoA with malonyl-
CoA to form a four-carbon intermediate. Six additional turns
of the enzyme’s cycle convert this intermediate to palmitate,
which is then liberated from FAS by the action of the
thioesterase domain.10FAS is attracting great interest as a drug
target in oncology because it is up-regulated in most solid
tumors, including those of the breast,11prostate,12and ovary.13
Furthermore, a number of studies show that a pharmacologic
blockade of FAS can be cytostatic and cytotoxic to tumor cells.14
For example, the fungal product cerulenin15and a synthetic
(5) (a) Calter, M. A.; Tretyak, O. A.; Flaschenriem, C. Org. Lett. 2005,
7, 180. (b) Zhu, C.; Shen, X.; Nelson, S. G. J. Am. Chem. Soc. 2004, 126,
(6) Calter, M. A. J. Org. Chem. 1996, 61, 8006.
(7) (a) Calter, M. A.; Orr, R. K.; Song, W. Org. Lett. 2003, 5, 4745. (b)
For a review on the use of ketenes in asymmetric synthesis, see: Orr, R.
K.; Calter, M. A. Tetrahedron 2003, 59, 3545.
(8) (a) Calter, M. A.; Guo, X. J. Org. Chem. 1998, 63, 5308. (b) Calter,
M. A.; Zhou, J. G. Tetrahedron Lett. 2004, 45, 4847. (c) Calter, M. A.;
Guo, X. Tetrahedron 2002, 58, 7093. (d) Calter, M. A.; Liao, W. S.; Struss,
A. J. Org. Chem. 2001, 66, 7500.
(9) (a) Kridel, S. J.; Axelrod, F.; Rozenkrantz, N.; Smith, J. W. Cancer
Res. 2004, 64, 2070. (b) Knowles, L. M.; Axelrod, F.; Browne, C. D.; Smith,
J. W. J. Biol. Chem. 2004, 279, 30540.
(10) Mattick, J. S.; Nickless, J.; Mizugaki, M.; Yang, C. Y.; Uchiyama,
S.; Wakil, S. J. J. Biol. Chem. 1983, 258, 15300.
(11) (a) Alo, P. L.; Visca, P.; Marci, A.; Mangoni, A.; Botti, C.; Di
Tondo, U. Cancer 1996, 77, 474. (b) Swinnen, J. V.; Roskams, T.; Joniau,
S.; Van Poppel, H.; Oyen, R.; Baert, L.; Heyns, W.; Verhoeven, G. Int. J.
Cancer 2002, 98, 19.
(12) (a) Rossi, S.; Graner, E.; Febbo, P.; Weinstein, L.; Bhattacharya,
N.; Onody, T.; Bubley, G.; Balk, S.; Loda, M. Mol. Cancer Res. 2003, 1,
707. (b) Pizer, E. S.; Wood, F. D.; Heine, H. S.; Romantsev, F. E.;
Pasternack, G. R.; Kuhajda, F. P. Cancer Res. 1996, 56, 1189.
(13) Gansler, T. S.; Hardman, W., III; Hunt, D. A.; Schaffel, S.; Hennigar,
R. A. Hum. Pathol. 1997, 28, 686.
FIGURE 1. Schematic diagram of FAS showing enzymatic steps involved in fatty acid biosynthesis leading to palmitic acid. FAS is comprised
of six enzymatic domains and an acyl-carrier protein (ACP). The steps in fatty acid biosynthesis are as follows. (i) The malonyl/acetyl transferase
domain (1) transfers an acetyl group onto the ACP. It is then translocated to the active-site cysteine by R-ketoacyl synthase (2). This position,
marked “R”, also serves as the loading position for the growing acyl chain in subsequent iterations. (ii) The malonyl/acetyl transferase domain (1)
then transfers a malonyl group to the ACP, and the two are condensed (2) into a four-carbon product bound to the enzyme through the thiol of the
ACP. (iii) The ketoacyl reductase (3) reduces the ketone at C-3 to an alcohol. (iv) The dehydrase (4) further reduces the alcohol to an alkene. (v)
The enoyl reductase domain (5) further reduces the alkene bond to an alkane, and the ACP-bound chain is translocated back to the active-site
cysteine (2). Steps ii-v are then repeated six times to yield a 16-carbon, fully saturated palmitic acid bound to the ACP. (vi) The palmitate is
released from FAS by the enzyme’s intrinsic thioesterase domain (FAS TE, 6), which is the target of orlistat and the ?-lactones generated in this
cis-3,4-Dialkyl-?-lactones 3 via a Ketene Dimerization/
Hydrogenation Process from Acid Chlorides 1 and
Epimerization to trans-?-Lactones 4
Strategy toward Pseudosymmetric
Purohit et al.
4550 J. Org. Chem., Vol. 71, No. 12, 2006
derivative, c75,16were found to inhibit the ketosynthase (KS)
domain of FAS and thus shut down tumor cell proliferation and
in some cases induce apoptosis.17We recently showed that
orlistat, which bears a ?-lactone and is a drug approved for
treating obesity, is a potent inhibitor of FAS TE and that this
natural product derivative is cytotoxic and cytostatic to tumor
cells in vitro and can inhibit tumor growth in vivo.9However,
orlistat has poor solubility and poor bioavailability, so there is
a great need to develop new ?-lactones that overcome these
problems and that can be deployed as potential antitumor drugs.
In addition, simplified derivatives that are readily prepared
would also ultimately be attractive from the standpoint of
A particularly attractive feature of the described strategy is
its practicality since ?-lactones 4, which are simplified but
structurally analogous to orlistat, could potentially be obtained
in three steps from acid chlorides in optically active form.
Following ketene dimerization and hydrogenation, R-epimer-
ization would allow access to the thermodynamically favored
trans-?-lactones 4 (Scheme 1) corresponding to the relative
stereochemistry found in orlistat (5) and also commonly
observed in ?-lactone-containing natural products such as
nocardiolactone (6), the ebelactones (7a,b), and most recently
the belactosins (e.g., belactosin C (8), (Figure 2)). Furthermore,
given the potential that ?-lactones are showing currently as
inhibitors of several enzymes including proteases,18HMG Co-A
synthase,19esterases, and also the proteasome,20the ketene
dimerization/hydrogenation process described herein could also
find wider application.
Results and Discussion
Development of a One-Pot Ketene Dimerization/Hydro-
genation Sequence. Initial studies focused on the possibility
of developing a one-pot, two-step ketene dimerization/hydro-
genation process employing hexanoyl chloride 1a to access
?-lactone 3a (Scheme 2). Following ketene dimerization by the
method of Calter7with quinidine (9, Figure 3) as nucleophilic
catalyst and simple filtration to remove amine hydrochloride
salts, the reaction mixture was transferred to a hydrogenation
vessel and pressurized to 30 psi of H2. This procedure provided
only modest overall yields of ?-lactone 3a due to apparent
degradation of the ketene dimers in the presence of traces of
dissolved quaternary ammonium salts, a process with precedent
in the literature.21Despite careful filtration of the amine salts
prior to hydrogenation, extensive degradation of the dimer was
always observed. More significantly, the enantiopurity of the
ketene dimer (97% f 78% ee at 40% conversion) and thus the
enantiopurity of the ?-lactone (97% f 74% ee) were found to
erode during the hydrogenation, thus rendering this method
ineffective for obtaining highly enantiopure ?-lactones. The
stereochemistry of ?-lactone 3a was assigned based on coupling
constant analysis (Jcis≈ 6 Hz, Jtrans4-4.5 Hz)22and subse-
quently confirmed by X-ray analysis of the cyclohexyl-contain-
ing ?-lactone 3c (vide infra).
Isolation and Purification of Ketene Dimers. To avoid
erosion of optical purity and degradation during the hydrogena-
tion step, we studied the stability of ketene dimers 2 toward
isolation and purification. Previous large-scale synthesis and
purification of racemic ketene dimers bearing long alkyl chains
(>13 carbons) relied on acidic extraction to remove alkylam-
(14) (a) Kuhajda, F. P.; Pizer, E. S.; Li, J. N.; Mani, N. S.; Frehywot, G.
L.; Townsend, C. A. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 3450. (b)
Kuhajda, F. P.; Jenner, K.; Wood, F. D.; Hennigar, R. A.; Jacobs, L. B.;
Dick, J. D.; Pasternack, G. R. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 6383.
(c) Pizer, E. S.; Chrest, F. J.; DiGiuseppe, J. A.; Han, W. F. Cancer Res.
1945, 5, 4611.
(15) Funabashi, H.; Kawaguchi, A.; Tomoda, H.; Omura, S.; Okudas,
S.; Iwasaki, S. J. Biochem. 1989, 105, 751.
(16) Kuhajda, F. P.; Pizer, E. S.; Li, J. N.; Mani, N. S.; Frehywot, G.
L.; Townsend, C. A. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 3450.
(17) (a) Pizer, E. S.; Jackisch, C.; Wood, F. D.; Pasternack, G. R.;
Davidsonf, N. E.; Kuhajda, P. Cancer Res. 1996, 56, 2745-2747. (b) Pizer,
E. S.; Chrest, F. J.; Digiuseppew, J. A.; Han, F. Cancer Res. 1998, 58,
(18) (a) Lall, M. S.; Ramtohul, Y. K.; James, M. N. G.; Vederas, J. C.
J. Org. Chem. 2002, 67, 1536.
(19) (a) Mayer, R. J.; Louis-Flamberg, P.; Elliott, J. D.; Fisher, M.; Leber,
J. Biochem. Biophys. Res. Commun. 1990, 169, 610. (b) Thompson, K. L.;
Chang, M. N.; Chiang, Y. C. P.; Yang, S. S.; Chabala, J. C.; Arison, B. H.;
Greenspan, M. D.; Hanf, D. P.; Yudkovitz, J. Tetrahedron Lett. 1991, 32,
3337. (c) Romo, D.; Harrison, P. H. M.; Jenkins, S. I.; Riddoch, R. W.;
Park, K.; Yang, H. W.; Zhao, C.; Wright, G. D. Bioorg. Med. Chem. 1998,
6, 1255. (d) Tomoda, H.; Kumagai, H.; Ogawa, Y.; Sunazuka, T.;
Hashizume, H.; Nagashima, H.; Omura, S. J. Org. Chem. 1997, 62, 2161.
(e) Tomoda, H.; Ohbayashi, N.; Kumagai, H.; Hashizume, H.; Sunazuka,
T.; Omura, S. Biochem. Biophys. Res. Commun. 1999, 265, 536.
(20) (a) Tomoda, H.; Omura, S. Yakugaku Zasshi-J. Pharm. Soc. Jpn.
2000, 120, 935. (b) Masse, C. E.; Morgan, A. J.; Adams, J.; Panek, J. S.
Eur. J. Org. 2000, 14, 2513.
(21) Brady, W. T.; Waters, O. H. J. Org. Chem. 1967, 32, 3703.
(22) Mulzer, J.; Zippel, M.; Bruntrup, G.; Sengner, J.; Finke, J. Liebigs
Ann. Chem. 1980, 1108.
FIGURE 2. Structures of representative, bioactive natural products
FIGURE 3. Cinchona alkaloids and derivatives 9-11 employed in
the ketene dimerization.
Asymmetric ?-Lactone Synthesis via a Ketene Dimerization/
Attempted One-Pot, Two-Step Catalytic
?-Lactone Inhibitors of Fatty Acid Synthase
J. Org. Chem, Vol. 71, No. 12, 2006 4551
monium salts followed by vacuum distillation, and this could
potentially be applied to the asymmetric process described
herein.23However, in efforts to develop a procedure that would
be tolerant of acid-sensitive functionality and avoid extensive
heating, we studied the stability of ketene dimers to silica gel
chromatography. Indeed, we were able to purify all ketene
dimers by normal silica gel flash chromatography. Ketene dimer
2a possessed sufficient stability to allow determination of its
enantiomeric purity by gas chromatography using cyclodextrin
bis-OTBS as chiral stationary phase.24?-Lactones (R)-2a and
(S)-2a were obtained in high optical purity by use of either QND
(9, >98% ee) or O-TMS QUIN (12, >96% ee), respectively
(Table 1, entries 1 and 5). The enantiopurity of other dimers
was determined following hydrogenation to the cis-?-lactones
(vide infra). Silylated alkaloids gave comparable yields of ketene
dimer 3a compared to in situ generated, acetylated derivatives
(Table 1, entry 1 vs 2).25A slight improvement in yield was
realized when shorter reaction times were employed (6 vs 24
h), consistent with our finding that the ketene dimer is unstable
to prolonged exposure to trialkylammonium salts (Table 1,
entries 1-3 vs 4). The yield of ketene dimer could be improved
further by use of doubly-distilled acid chloride (75% vs 62%
yield, Table 1, entry 4 vs 5). While our work was in progress,
Calter reported that ketene dimers derived from propanoyl
chloride and malonyl chloride half-ester could be purified by
rapid filtration through silica gel.7The ability to isolate and
purify optically active ketene dimers 2 greatly extends the
potential utility of these intermediates for ?-lactone synthesis,
allowing for subsequent transformations including epoxidation
as reported by us recently, leading to novel spiro-epoxy-?-
Hydrogenation of Ketene Dimers. We next studied the
hydrogenation of the isolated ketene dimers. Prior hydrogenation
studies of diketenes have primarily focused on the parent
diketene, 4-methylene-2-oxetanone, in both racemic and asym-
metric fashion, as a means to obtain the corresponding 4-methyl-
2-oxetanone, a commodity chemical utilized on ton scale for
polymer applications.27Several catalysts have been utilized for
hydrogenation of enol ethers;28however, for simplicity and
practicality, palladium on carbon was studied initially with
racemic ketene dimer 2a. At the outset, we expected high facial
selectivity for the hydrogenation due to the proximity of the
alkene to the R-stereogenic carbon at C3 of the ?-lactone.
However, hydrogenation of dimer 2a employing 5 mol % Pd/C
(5 wt %) resulted in high yield but low diastereoselectivity,
providing a mixture of cis- and trans-?-lactones 3a (4:1 favoring
cis) after a 24 h reaction time (entry 1, Table 2). Addition of
100 mol % triethylamine relative to Pd catalyst, in an attempt
to reduce the activity of the catalyst, improved the diastereo-
selectivity to 17:1 with the same reaction time (entry 2, Table
2). However, repeating these conditions with optically active
dimer 2a (98% ee) indicated that racemization was occurring
under these conditions necessarily at the dimer stage, providing
cis-?-lactone 3a with reduced enantiopurity (71% ee). Shorten-
ing the reaction time using 5 mol % catalyst without added Et3N
also gave high diastereoselectivity (entry 3, Table 2). Taken
together, these results suggest that both longer reaction times
and the absence of catalyst poison leads to erosion in diaste-
reoselectivity. Optimal conditions that prevented racemization
and maintained high diastereoselectivity were eventually realized
by decreasing the amount of Pd/C to 1 mol % and reducing the
reaction time to 30 min under 30 psi of H2pressure without
Et3N. Under these conditions, no racemization or epimerization
was observed for either the ketene dimer or the major diaste-
reomer isolated, cis-?-lactone 3a. The latter could be isolated
in 90% yield as a >19:1 mixture of cis/trans diastereomers 3a
as determined by coupling constant analysis (3a, JHa,Hb) 6.3
Hz) in 96% ee (Table 2, entry 4).29In this manner, hydrogena-
tion of a series of ketene dimers 2a-f gave consistently high
yields of the corresponding ?-lactones 3a-f with high enan-
tiomeric purities (Table 3). Enantiomer ratios were determined
by chiral GC analysis following hydrogenation. Racemic
?-lactone 3c was crystalline, and thus, X-ray analysis verified
the cis stereochemistry obtained during hydrogenation and
suggested by coupling constant analysis for the entire series
(23) For a lead reference to processes for ketene dimerization, see:
Zhang, J. J. U.S. Patent 5525738, 1996.
(24) Shitangkoon, A.; Vigh, G. J. Chromatogr., A. 1996, 738, 31-42.
(25) Calter previously reported increased rates with silylated quinine
derivatives, see ref 7a.
(26) Duffy, R. J.; Morris, K. A.; Romo, D. J. Am. Chem. Soc. 2005,
(27) For lead references to hydrogenation of diketene, see: (a) Sixt, J.
U.S. Patent 2,763,664, 1956. For lead references to asymmetric hydrogena-
tion of diketene using chiral BINAP-Ru(II)catalysts, see: (b) Ohta, T.;
Miyake, T.; Seido, N.; Kumobayashi, H.; Takaya, H. J. Org. Chem. 1995,
60, 357. (c) Takaya, H.; Ohta, T.; Kumobayashi, H.; Okeda, Y.; Gonda, Y.
U.S. Patent 5,412,109, 1995.
(28) For a lead reference to hydrogenation of enol ethers, see: Branda ¨nge,
S.; Fa ¨rnba ¨ck, M.; Leijonmarck, H.; Sundin, A. J. Am. Chem. Soc. 2003,
(29) Samples taken at 2, 4, 6, 12, and 18 h revealed no erosion of
enantiomeric purity by chiral GC analysis for both ketene dimer 2a and
on Efficiency of Ketene Dimerization Leading to Dimer 2a
Effect of Catalyst Structure and Reaction Conditions
entrycatalyst time (h)% yielda
O-TBS QND (11)
O-TMS QND (10)
O-TMS QUIN (12)
O-TMS QUIN (12)
aYields refer to isolated, purified dimer.bFreshly, doubly-distilled acid
chloride was used.cND ) not determined.dAbsolute configurations are
based on the precedent of Calter (ref 5).
Diastereoselectivity and Enantioselectivitya
Effect of Catalyst Loading, Base, and Reaction Time on
entry mol % Pd/Cmol % Et3N
time (h) % yielddrb
aReactions were conducted at 0.1 M in CH2Cl2.bRatios estimated by
1H NMR (500 MHz) integration of Haand Hbin crude reaction mixtures.
cEnantiomeric excess of cis-3a was determined by GC analysis to be 98%
Purohit et al.
4552 J. Org. Chem., Vol. 71, No. 12, 2006
R-Epimerization of cis- to trans-?-Lactones. As described
above, the majority of naturally occurring ?-lactones possess
trans-?-lactone stereochemistry including orlistat (see Figure
2). Thus, we explored conditions to epimerize to the thermo-
dynamically preferred trans-?-lactones. Deprotonation leading
to ?-lactone enolates is possible since ?-elimination in these
systems is a symmetry forbidden process.30Competing inter-
molecular Claisen condensations can be precluded provided
there is an R-substituent; however, in some cases, low yields
of R-monoalkylated systems have been obtained using inverse
addition with highly reactive electrophiles.31Numerous ther-
modynamic conditions to effect epimerization were studied
including (a) triethylamine/ammonium acetate buffer system in
dichloromethane (∼10-15% conversion by1H NMR analysis
after prolonged stirring for 48 h at 23 °C), (b) t-BuOK in
t-BuOH, (c) DABCO, (d) DBU, 1,2-dichlorobenzene, or aceto-
nitrile, (e) HMDS, THF, (f) catalytic KHMDS/18-crown-6/THF,
(g) t-BuOK/HMDS/THF, (h) t-BuOK/HMDS/18-crown-6/THF.
However, all these conditions gave low conversion to the trans-
isomers along with decomposition. Ultimately, an unsatisfying
but functional procedure was developed involving kinetic
deprotonation and quenching to provide a mixture of cis- and
trans-?-lactones, which could then be separated by silica gel
chromatography to deliver pure trans-isomers 3a and 3c
Alkylation/Acylation of cis-?-Lactones. Enolization with
lithium hexamethyldisilazide (LiHMDS) followed by addition
of various electrophiles proved more fruitful and allowed access
to ?-lactones bearing R-quaternary centers. Alkylation proceeded
with high diastereoselectivity when benzyl bromide was em-
ployed (entries 2 and 3, Table 4), whereas the smaller electro-
phile, methyl iodide, provided only moderate diastereoselectivity
(entry 1, Table 4), as previously observed for ?-lactones with
moderately sized ?-substituents.32The relative stereochemistry
of the major diastereomer of benzyl-?-lactone 8b was deter-
mined by nOe analysis.33In the case of acylation with benzyl
(30) Mulzer, J.; Zippel, M.; Bruentrup, G.; Segner, J.; Finke, J. Liebigs
Ann. Chem. 1980, 1108.
(31) (a) Parsons, P. J.; Cowell, J. Synlett 2000, 107. (b) Hanessian, S.;
Tehim, A.; Chen, P. J. Org. Chem. 1993, 58, 7768.
Catalytic Asymmetric ?-Lactone Synthesis via a Sequential, Two-Step Ketene Dimerization/Hydrogenation Sequencea
aAll reactions were carried out at 0.1 M (final concentration) with freshly distilled acid chloride.bRefers to isolated, purified yields.cEnantiomeric
excess was determined by chiral GC analysis.dAbsolute configuration of the major enantiomer is depicted and based on the previously determined absolute
configuration of ketene dimers (ref 7).eReaction time was 3 h at 0 °C.fReaction was performed with 5 mol % Pd/C for 3.5 h.gND ) not determined.
FIGURE 4. X-ray crystal structure (POV Chem rendering) of (()-
Epimerization of cis-?-Lactones to the
?-Lactone Inhibitors of Fatty Acid Synthase
J. Org. Chem, Vol. 71, No. 12, 2006 4553
chloroformate, the use of sodium hexamethyldisilazide (Na-
HMDS) gave improved yields and diastereoselectivities relative
to LiHMDS (entry 3, Table 4).
Monitoring Ketene Formation and Dimerization of Hy-
drocinnamoyl Chloride by ReactionView in Situ Mid-
Infrared (IR) Spectroscopy. The formation and dimerization
of 3-phenyl-1-propen-2-one derived from hydrocinnamoyl
chloride by the action of Hu ¨nig’s base was monitored in efforts
to detect intermediates by mid-IR spectroscopy and follow the
overall dimerization process. Figures 5 and 6 show results of
ReactionView IR monitoring for the overall dimerization
reaction of hydrocinnamoyl chloride in CH2Cl2(0.1 M) leading
to ketene dimer 2d. Under Reaction Conditions I, the nucleo-
philic catalyst, TMS-QND (10), is added 5 min after the addition
of EtNi-Pr2(Hu ¨nig’s, DIPEA) base at 30 min (Figure 5). Under
these conditions, three distinct species (i) hydrocinnamoyl
chloride (1793 cm-1), (ii) 3-phenyl prop-1-en-1-one (2118
cm-1), and (iii) 3-benzyl-4-(2-phenylethylidene)oxetan-2-one
(2d, 1710 cm-1) were detected and monitored over a period of
540 min. The data are plotted as absorbance versus time in a
two-dimensional format (Figure 5a) and three-dimensional
format (Figure 5b) following peak fitting (ReactionView
software). Upon addition of 1.0 equiv of Hu ¨nig’s base at t )
30 min to the hydrocinnamoyl chloride, ketene generation is
immediately initiated, and after 5 min, a 50% decrease in
absorbance for hydrocinnamoyl chloride is observed. Separate
experiments revealed that ketene was stable at this temperature
for at least 15 min. Subsequent addition of TMS-QND (10) at
t ) 35 min completely consumes the ketene after 44 min.
However, a brief induction period is observed wherein ketene
dimer formation is retarded until hydrocinnamoyl chloride is
almost consumed at t )160 min. It is clear that under these
conditions the formed ketene is rapidly consumed, and following
this time, presumably an undetectable amount of ketene is
produced slowly from acid chloride, leading to complete
conversion to ketene dimer 2d.
In a second experiment the order of addition of reagents was
altered (Reaction Conditions II, Figure 6). Addition of hydro-
cinnamoyl chloride at t ) 18.3 min to a mixture of Hu ¨nig’s
base and TMS-QND immediately produced ketene dimer along
with a trace amount of ketene. Under these reaction conditions
(32) For a review on the utility of ?-lactones and their application as
intermediates in natural product synthesis, see: Wang, Y.; Tennyson, R.
L.; Romo, D. Heterocycles 2004, 64, 605.
(33) See Supporting Information for details.
cis-?-Lactone 4a Leading to Quaternary Carbon Bearing
Diastereoselective Alkylations and Acylations of
entryR cmpd no. base% yielda
aRefers to isolated, purified yield.bDiastereomeric ratio was determined
by integration (1H NMR 500 MHz) of crude reaction mixtures.
FIGURE 5. Reaction View IR monitoring (1600-2200 cm-1, t )
0-540 min) of the ketene dimerization process leading to ketene dimer
2d under Reaction Condition I from hydrocinnamoyl chloride: (a)
normalized Lorentz curve fit for absorbance vs time (min) for three
distinct species (acid chloride, ketene, and ketene dimer 2d) for the
dimerization reaction of hydrocinnamoyl chloride. (b) Mid-IR Reac-
tionView spectrum for dimerization of 3-phenyl prop-1-en-1-one.
FIGURE 6. ReactionView IR monitoring (1600-2200 cm-1, t )
0-232 min) of the ketene dimerization process leading to ketene dimer
2d under Reaction Condition II from hydrocinnamoyl chloride: (a)
normalized Lorentz curve fit for absorbance vs time (min) for three
distinct species. (b) Mid-IR ReactionView spectrum for dimerization
of 3-phenyl prop-1-en-1-one.
Purohit et al.
4554 J. Org. Chem., Vol. 71, No. 12, 2006
it appears that the rate of ketene consumption exceeds that of
ketene generation in accordance with the results above and also
those observed by Calter,7and thus, the ketene resonance is
fleeting on the time scale of mid-IR measurements. These
ReactionView studies provide further evidence for a rate-
determining ketene generation step followed by rapid ketene
dimerization as found previously by Calter during rate studies
of the ketene dimerization process.7a
Enzyme Inhibition Studies. A recombinant form of the
thioesterase domain of fatty acid synthase was used in a
substrate-based screen to measure the apparent Ki’s of ketene
dimers 2a-d and both cis-?-lactone and trans-isomers of the
derived simplified orlistat derivatives. Importantly, orlistat is
presumed to be a covalent inhibitor of the FAS TE domain, in
analogy to the adduct formed with the active site serine of
pancreatic lipase.34We presume that all of the ?-lactones
examined in this study function via the same mechanism.
Therefore, the results are reported as apparent inhibition
constants (app Ki), as the term Kiis usually used for reversible
inhibitors. 4-Methylumbelliferyl heptanoate (4-MUH) was
utilized as a substrate for FAS TE as it was found to provide
the best signal-to-noise ratio and an acceptable turnover rate.
The product of this substrate, 4-methylumbelliferone, fluoresces
at 450 nM (excitation at 350 nM) and provides a convenient
readout of thioesterase activity. In this assay, the hydrolysis of
4-MUH is blocked by orlistat, our lead antagonist with an
apparent Kiof 0.21 µM. The ability of the various ?-lactones
and some ketene dimer precursors described above to act as
antagonists of 4-MUH toward recombinant FAS TE was
(34) (a) Borgstrom, B. Biochim. Biophys. Acta 1988, 962, 308. (b)
Hadvary, P.; Lengsfeld, H.; Wolfer, H. J. Biol. Chem. 1988, 256, 357. (c)
Hadvary, P.; Sidler, W.; Meister, W.; Vetter, W.; Wolfer, H. Biochem. J.
1991, 266, 2021.
Antagonistic Activity of Ketene Dimers 2 and ?-Lactones 3 and 8 toward Recombinant FAS TE Compared to Orlistat (1)
FIGURE 7. Dose-response curve illustrating the inhibition of FAS
TE by trans-?-lactone 3a in a fluorogenic assay.
?-Lactone Inhibitors of Fatty Acid Synthase
J. Org. Chem, Vol. 71, No. 12, 2006 4555
measured by this assay (Table 5). The ?-lactones were tested
across a concentration range of 1-100 µM, and the dose-
response curve for trans-?-lactone 3a is shown as a representa-
tive example (Figure 7).
Ketene dimers were tested and showed moderate activity (5-
16 µM) with the exception of dimer 2c which showed no activity
(Table 5, entries 1-4). The most potent ketene dimer tested
was phenyl derivative 2d with an app Kiof ∼5.0 µM, which is
approximately 25-fold less potent than orlistat.
In the cis-isomeric series (Table 5, entries 9-15), addition
of an ester substituent in proximity to the ?-lactone nucleus
led to a drastic reduction in activity (Table 5, entry 14).
However, a remote methoxy substituent as in ?-lactone cis-3f
(entry 15) restored activity comparable to that observed for the
n-butyl-substituted ?-lactone, cis-3a (entry 10).35As expected,
the trans-isomers, in general, were found to possess greater
inhibitory activity compared to the cis-isomers with the same
substituents (entries 5-8 vs 10-13). ?-Lactones 8a-c bearing
R-quaternary carbons showed moderate activity (10-19 µM)
but presumably also lead to covalent adducts suggestive of the
possibility that the thioesterase domain is not selective for only
In the case of the phenyl-substituted ketene dimer 2d (5.0
µM), a ∼2-fold increase in activity to 2.5 µM was observed
following hydrogenation to give the phenyl-substituted cis-?-
lactone 3d (Table 5, entry13). Surprisingly, little variation in
activity is observed (3.3 µM, entry 8) when this derivative is
epimerized to trans-?-lactone 3d possessing the relative and
absolute stereochemistry corresponding to that found in orlistat.
Although none of the simple 3,4-dialkyl-?-lactones described
herein were found to be as potent as orlistat, the findings are
significant since these are highly simplified ?-lactones devoid
of the N-formyl aminoester side chain. These polar substituents
are expected to serve as additional recognition sites via potential
hydrogen-bonding and polar-polar interactions to FAS TE. In
addition, they should also contribute to water solubility.
However, a promising lead compound for further optimization
is the phenyl-substituted derivative trans-3d.
In summary, we developed a scaleable process to prepare
cis-3,4-disubstituted ?-lactones using a ketene dimerization/
hydrogenation sequence from readily available acid chlorides
in good overall yields and high enantioselectivity. This reaction
could be run as a single-pot, two-step process, but higher overall
yields and optical purities were obtained upon isolation of the
ketene dimer by silica gel chromatography and subsequent
hydrogenation of the purified ketene dimer at moderate pres-
sures. Several ketene dimers were isolable, and their enantio-
meric purity could be determined by chiral GC analysis. A
current limitation in this process is the accessibility of only
pseudosymmetric ketene dimers 2 via the Calter procedure
leading to pseudosymmetric ?-lactones 3. Enolization followed
by alkylation and acylation of the cis-?-lactones provided ready
access to R,R-disubstituted ?-lactones with high diastereose-
lectivities. trans-?-Lactones could also be obtained by R-epimer-
ization in low yield; however, a general, practical solution for
this process remains elusive. These highly simplified dialkyl-
?-lactones are analogous to the natural products norcardiolac-
tone, valilactone, and orlistat and were found to exhibit moderate
to good inhibitory activity toward recombinant FAS TE as
measured by enzymatic activity using recombinant protein with
a fluorogenic substrate. The best antagonist, cis-?-lactone 3d,
displayed an app Kiof only ∼10-fold less than that of orlistat.
In addition, analysis of the ketene dimerization process by
ReactionView IR spectroscopy further substantiates findings by
Calter that ketene formation is rate determining in the catalyzed
ketene dimerization process. Further transformations of these
optically pure ketene homodimers are being explored as a means
to provide practical routes to ?-lactones more structurally
analogous to orlistat and expected to have higher affinity for
FAS TE. In addition, the use of hetero-ketene dimers will
expand the ?-lactone structures accessible by this strategy.
General Experimental Procedure for Dimerization As De-
scribed for (R,Z)-3-Butyl-4-pentylideneoxetan-2-one (2a). This
procedure is slightly modified from the method of Calter.7aTo a
flame-dried 1 L round-bottomed flask was added 764 mg (5 mol
%, 1.926 mmol) of TMS-quinine, 385 mL of CH2Cl2(0.1 M), and
6.86 mL (1.0 equiv, 38.53 mmol) of Hu ¨nig’s base under nitrogen
atmosphere at 22 °C. To this colorless solution, 5 mL (5.18 g, 38.53
mmol) of freshly double-distilled, hexanoyl chloride was added over
15 min via syringe. After 6 h the dark yellow solution was
concentrated down to 100 mL (one-fifth original volume) in vacuo
and 250 mL of pentanes was added to precipitate the ammonium
salts. Filtration through Whatmann filter paper (#1, qualitative
grade), concentration in vacuo, and purification by flash column
chromatography on deactivated SiO2(10% H2O) (2.5 cm × 35.0
cm column, 15 cm pad) eluting with 0-20% Et2O:hexanes gave
2.83 g (75%) of 2a as a colorless oil (96% ee, chiral GC analysis).
Rf0.54 (15% Et2O:hexanes); [R]D26+5.3 (c ) 0.51, CHCl3); IR
(thin film) 1865 cm-1;1H NMR (300 MHz, CDCl3) δ 4.70 (dt,
J ) 1.3, 6.3 Hz, 1H), 3.94 (dt, J ) 1.0, 7.0 Hz, 1H), 2.13 (app q,
2H), 1.75-1.83 (m, 2H), 1.28-1.53 (m, 8H), 0.88-0.96 (m, 6H);
13C NMR (300 MHz, CDCl3) δ 14.5, 14.6, 22.9, 23.1, 25.1, 28.0,
29.2, 32.3, 54.45, 102.4, 146.4, 170.6; ESI LRMS calcd for
C12H20O2[M + Li], 202; found, 202.
Racemic ketene dimers were initially prepared in a similar
manner using Hu ¨nig’s base. Subsequently, 5 mol % quinuclidine
hydrochloride with 1.0 equiv of Hu ¨nig’s base gave optimal results
in terms of reaction rate and yield; therefore, this method was
subsequently used for preparation of racemic ketene dimers.
Representative Procedure for Hydrogenation As Described
for cis-(3R,4S)-3-Butyl-4-pentyloxetan-2-one (3a). The purified
n-butyl ketene dimer 2a (5.1 mmol, 1.0 g) was dissolved in 40 mL
of dichloromethane (0.1 M) and transferred to a Parr bomb
apparatus under a N2atmosphere using an additional 10 mL of
dichloromethane as wash solvent. Pd/C (107 mg, 1 mol %, 5 wt
%) was then added at one time under N2atmosphere. The Parr
bomb was then subjected to three consecutive evacuation-saturation
cycles of hydrogen gas and then pressurized to 30 psi hydrogen
gas pressure. Hydrogenation with shaking (Parr shaker) was
continued for 30 min at this pressure, and then the heterogeneous
slurry was vacuum filtered through a plug of Celite and concen-
trated, yielding a colorless oil. Flash chromatography with gradient
elution (5-15% diethyl ether/hexanes; 2.5 × 35.0 × 5 cm pad)
gave 3-butyl-4-pentyl-oxetan-2-one (3a, 897 mg, 90%) as a
(35) For previous studies of the impact of ?-lactone stereochemistry and
substitution on enzyme inhibitory activity, see: (HMG-CoA synthase) (a)
Tomoda, H.; Ohbayashi, N.; Kumagai, H.; Hashizume, H.; Sunazuka, T.;
Omura, S. Biochem. Biophys. Res. Commun. 1999, 265, 536. (b) Romo,
D.; Harrison, P. H. M.; Jenkins, S. I.; Riddoch, R. W.; Park, K.; Yang, H.
W.; Zhao, C.; Wright, G. D. Bioorg. Med. Chem. 1998, 6, 1255. (c) Tomoda,
H.; Kumagai, H.; Ogawa, Y.; Sunazuka, T.; Hashizume, H.; Nagashima,
H.; Omura, S. J. Org. Chem. 1997, 62, 2161. (d) Mayer, R. J.; Louis-
Flamberg, P.; Elliott, J. D.; Fisher, M.; Leber, J. Biochem. Biophys. Res.
Commun. 1990, 169, 610. (e) Romo, D.; Harrison, P. H. M.; Jenkins, S. I.;
Riddoch, R. W.; Park, K.; Yang, H. W.; Zhao, C.; Wright, G. D. Bioorg.
Med. Chem. 1998, 6, 1255. (proteasome) (f) Macherla, V. R.; Mitchell, S.
S.; Manam, R. R.; Reed, K. A.; Chao, T. H.; Nicholson, B.; Deyanat-Yazdi,
G.; Mai, B.; Jensen, P. R.; Fenical, W. F.; Neuteboom, S. T. C.; Lam, K.
S.; Palladino, M. A.; Potts, B. C. M. J. Med. Chem. 2005, 48, 3684.
Purohit et al.
4556 J. Org. Chem., Vol. 71, No. 12, 2006
colorless oil (>96% ee, chiral GC): [R]D26-27.1 (c ) 0.5, CHCl3);
Rf0.47 (15% Et2O:hexanes); IR (thin film) 1824 cm-1;1H NMR
(300 MHz, CDCl3) δ 4.54 (ddd, J ) 2.1, 3.6, 5.7 Hz, 1H), 3.59
(ddd, J ) 4.5, 5.4, 8.4 Hz, 1H), 1.46-1.84 (m, 6H), 1.31-1.44
(m, 8H), 0.88-0.94 (m, 6H);13C NMR (500 MHz, CDCl3) δ 14.5,
14.7, 23.2, 23.2, 24.4, 26.0, 30.5, 30.9, 32.2, 53.3, 76.5, 173.1;
ESI LRMS calcd for C12H22O2[M + Li], 205; found, 205.
The enantiomeric purity of ketene dimer 2a and ?-lactone 3a
was determined to be >96% ee by chiral GC analysis. Column
type, chiral bis-OTBS-cyclodextrin; retention time, tdimer 16.96
(major) and 17.16 (minor), t?-lactone26.23 (major) and 26.44 (minor).
Conditions: make up flow, 25 mL/min; H2flow, 30 mL/min; air
flow, 300 mL/min; injector temperature, 200 °C, pressure, 5 psi
(hold time 30 min); oven temperature gradient, 100 f140 °C (hold
time 30 min); detector temperature, 250 °C.
trans-(3R,4R)-3-Butyl-4-pentyloxetan-2-one (trans-3a). To a
-78 °C solution of 100 mg (0.51 mmol) of 3a in 5 mL of THF
was added 760 µL of LiHMDS (1.5 equiv, 1.0 M in THF), and
this was allowed to stir for 1 h. Tetramethylenediamine (TMEDA,
120 µL, 1.5 equiv, 0.76 mmol) was then added at -78 °C and
allowed to stir for an additional 30 min, after which the solution
was quenched with glacial acetic acid (130 µL, 3.0 equiv, 2.27
mmol) and warmed to 22 °C. After extraction with diethyl ether
(2 × 6 mL), the combined organics were washed with 2 mL of pH
7.0 buffer and 2 mL of brine and then dried over Na2SO4.
Concentration in vacuo gave a colorless oil, which upon purification
by flash chromatography on SiO2(15% Et2O:hexanes) gave cis-3a
and trans-3a (72 mg, 72% yield) as a 1:1 mixture of diastereomers.
Further purification by gravity column chromatography (2×, 5%
Et2O:hexanes) delivered 34 mg (34%) of trans-3a and 38 mg of
cis-3a (38%): Rf0.38 (10% Et2O:hexanes; cis-3a Rf0.29); IR (thin
film) 1824 cm-1;1H NMR (500 MHz, CDCl3) δ 4.22 (ddd, J )
4.2, 6.0, 7.5 Hz, 1H), 3.16 (ddd, J ) 3.9, 6.6, 9.0 Hz, 1H), 1.67-
1.93 (m, 4H), 1.26-1.49 (m, 10H), 0.91 (bs, 6H);13C NMR (300
MHz, CDCl3) 14.5, 14.7, 23.1, 23.2, 25.4, 28.3, 29.9, 32.1, 35.2,
56.9, 78.9, 172.5; ESI LRMS calcd for C12H22O2[M + Li], 205;
tan-2-one (trans-3c). To a -78 °C solution of 30 mg (0.108 mmol)
of cis-?-lactone 3c in 3 mL of THF was added 162 µL of LiHMDS
(1.5 equiv, 1.0 M in THF), and this was stirred for 1 h.
Tetramethylenediamine (TMEDA, 25 µL, 1.5 equiv, 0.162 mmol)
was then added at -78 °C and allowed to stir for an additional 30
min, after which the solution was quenched with glacial acetic acid
(18 µL, 3.0 equiv, 0.323 mmol) and warmed to 22 °C. After
extraction with diethyl ether (2 × 3 mL), the combined organics
were washed with 2 mL of pH 7.0 buffer and 2 mL of brine and
then dried over Na2SO4. Concentration in vacuo gave a colorless
oil that was purified by flash chromatography on SiO2(10% Et2O:
hexanes) to give a mixture of cis- and trans-3c (19.8 mg, 66%
yield). Further purification by gravity column chromatography (10%
Et2O:hexanes) delivered 12 mg (40%) of trans-3c and 7.6 mg of
cis-3c (25%): Rf0.47 (10% Et2O:hexanes; cis-3c, Rf0.40); IR (thin
film) 1824 cm-1;1H NMR (500 MHz, CDCl3) δ 4.17 (ddd, J )
4.2, 6.0, 7.4 Hz, 1H), 3.24 (ddd, J ) 4.0, 6.5, 9.0 Hz, 1H), 1.79-
1.89 (m, 1H), 1.59-1.78 (m, 12H), 1.30-1.39 (m, 1H), 1.07-
1.30 (m, 10H), 0.82-0.99 (m, 4H);13C NMR (300 MHz, CDCl3)
15.4, 26.1, 26.2, 26.3, 26.4, 26.6, 29.8, 31.9, 32.6, 32.9, 33.2, 33.3,
35.9, 37.4, 54.2, 66.0, 79.4, 172.3; ESI LRMS calcd for C18H30O2
[M + Li], 285; found, 285.
(3R,4S)-3-Butyl-3-methyl-4-pentyloxetan-2-one (8a). A solu-
tion of ?-lactone 3a (36.3 mg, 0.1835 mmol) in 1.9 mL of THF
was cooled to -78 °C, and 370 µL of LiHMDS (0.367 mmol, 2.0
equiv, 1.0 M solution in THF) was added under a nitrogen
atmosphere. After 1.5 h, 23 µL (0.367 mmol, 2.0 equiv) of
iodomethane was added and the reaction warmed to -40 °C and
stirred for an additional 45 min. The reaction mixture was
concentrated in vacuo and purified by flash chromatography (0-
15% Et2O:hexanes) to give ?-lactone 8a (28 mg, 73%) as a colorless
oil and a mixture of cis/trans diastereomers (dr, 6:1). Data provided
for major diastereomer: Rf(15% Et2O:hexanes) 0.64; IR (thin film)
1824 cm-1;1H NMR (500 MHz, CDCl3) δ 4.18 (dd (major diast.),
J ) 6.0, 8.5 Hz, 1H), 1.63-1.78 (m, 2H), 1.45-1.53 (m, 2H), 1.39
(s, 3H), 1.24-1.39 (m, 10H), 0.89-0.94 (m, 6H);13C NMR (500
MHz, CDCl3) 14.1, 14.2, 20.0, 22.1, 25.5, 25.6, 26.5, 30.3, 31.7,
35.9, 56.8, 84.5, 175.5; ESI LRMS calcd for C13H24O2Li [M +
Li], 219; found, 219.
(3S,4S)-3-Benzyl-3-butyl-4-pentyloxetan-2-one (8b). To a -78
°C solution of ?-lactone 3a (153 mg, 0.76 mmol) dissolved in 7.5
mL of THF was added 1.52 mL of LiHMDS (1.52 mmol, 2.0 equiv,
1.0 M solution in THF) under nitrogen atmosphere. After 1.5 h,
180 µL (1.52 mmol, 2.0 equiv) of benzyl bromide was added and
the reaction warmed to -40 °C and stirred for an additional 45
min. The reaction mixture was concentrated in vacuo and purified
by flash chromatography (0-5% Et2O:hexanes) to give ?-lactone
8b (192 mg, 88%) as a mixture of cis/trans diastereomers (>19:1).
Data provided for major diastereomer: Rf0.55(5% Et2O:hexanes);
IR (thin film) 1813 cm-1;1H NMR (500 MHz, CDCl3) δ 7.30-
7.34 (m, 2H), 7.25-7.28 (m, 1H),7.16-7.17 (m, 2H), 4.34 (dd,
J ) 4.5, 9.5 Hz, 1H), 3.13, 2.88 (AB q, J ) 14.5 Hz, 2H), 1.68-
1.78 (m, 2H), 1.45-1.61 (m, 4H), 1.18-1.39 (m, 8H), 0.94 (t, J )
3.5, 3H), 0.87 (t, J ) 3.5, 3H);13C NMR (500 MHz, CDCl3) 13.87,
13.89, 22.4, 23.2, 25.2, 26.3, 28.7, 29.7, 31.4, 38.2, 61.3, 80.3,
127.1(2C), 128.7(2C), 129.8, 135.8, 174.2; ESI LRMS calcd for
C19H28O2[M + Li], 295; found, 295.
(8c). To a -78 °C solution of NaHMDS (1.2 equiv, 0.91 mmol,
45 µL, 2M in THF) in 5 mL of THF was added 150 mg (0.73
mmol) of ?-lactone 3a dissolved in 2.5 mL of THF. After 1.5 h,
benzylchloroformate (1.1 equiv, 0.833 mmol, 120 µL) was added
at one time and stirred for an additional 3 h. This solution was
warmed to 23 °C over 1 h and worked up as described above for
?-lactone 8b. Flash chromatography on SiO2(5% Et2O:hexanes)
gave ?-lactone 8c (185 mg, 74% yield) as a colorless oil: Rf0.48
(15% Et2O:hexanes); IR (thin film) 1824, 1757, 1716 cm-1;1H
NMR (300 MHz, CDCl3) δ 7.40 (m, 5H), 5.27, 5.22 (AB q, J )
12.0, 2H), 4.52 (dd, J ) 4.8, 8.4 Hz, 1H), 2.50-2.56 (m, 1H),
2.22-2.30 (m, 2H), 1.71-1.85 (m, 1H), 1.44-1.60 (m, 4H), 1.15-
1.39 (m, 8H), 0.81-0.92 (m, 6H);13C NMR (500 MHz, CDCl3)
14, 14.6, 14.7, 23.1, 23.3, 23.6, 24.9, 25.8, 26.9, 29.4, 30.4, 31.1,
31.9, 32.3, 40.4, 71.6, 79.8, 121.7, 129.2, 129.4, 129.5, 129.8, 133.1,
152.7, 163.7, 167.3; ESI LRMS calcd for C20H28LiO4+[M + Li],
339; found, 338.
General Experimental Procedure for in Situ Mid-IR Spec-
troscopy with a RemSpec ReactionView System. A RemSpec
Reaction View system was fitted with a double-pass liquid
transmission head that was placed into a flame-dried two-necked
20 × 2.3 cm reaction tube equipped with a spin vane stir bar, and
the reaction flask was placed under a nitrogen atmosphere.
Following addition of 30 mL of CH2Cl2, a background spectrum
was obtained for 15 min and then automatic data collection was
initiated and provided 45 scans/min for both experiments at 23 °C.
The data collected for hydrocinnamoyl chloride, ketene, and ketene
dimer for both experiments were peak fitted with a Lorentzian
function on Grams/AI software and normalized on Microsoft Excel
to obtain absorbance versus time curves.
Reaction Condition I. Sequential addition of hydrocinnamoyl
chloride (0.45 mL, 2.97 mmol; t ) 14.5 min) and 565 µL of Hu ¨nig’s
base (2.97 mmol; t ) 30 min) gave rise to expected acid chloride
and ketene absorbances. After complete consumption of acid
chloride, 5 mol % TMS-QN (22 mg) was added as a solution in 1
mL of CH2Cl2in one portion at t ) 35 min, and data was collected
for an additional 540 min.
Reaction Condition II. In this procedure 5 mol % TMS-QN
(22 mg) and 565 µL of Hu ¨nig’s base was added to 30 mL of CH2-
Cl2. At t ) 18.3 min, hydrocinnamoyl chloride (0.45 mL, 2.97
mmol) was added over a period of 5 min and data was collected
for an additional 230 min.
?-Lactone Inhibitors of Fatty Acid Synthase
J. Org. Chem, Vol. 71, No. 12, 2006 4557
Fluorogenic Assay for Detection of Enzyme Inhibition. Download full-text
Expression of the recombinant thioesterase domain of FAS was
performed as described previously,9and large-scale expression was
performed by Invitrogen Corp. (Madison, WI). The synthetic
fluorogenic substrate, 4-methylumbelliferyl heptanoate (4-MUH),
was purchased from Sigma (St. Louis, MO). The reaction mixture
consisted of 45 µL of 500 nM FAS TE in buffer A (100 mM Tris-
HCl, 50 mM NaCl at pH 7.4) which was preincubated with 2.5 µL
of stock solutions of test ?-lactones dissolved in DMSO at final
concentrations of 0.32-100 µM at 37 °C for 30 min. The reaction
was initiated by addition of 5 µL of 1.25 mM 4-MUH in 1:1 DMSO:
buffer A. The resulting fluorescence from liberated 4-methylum-
belliferone was measured every 5 min at 350/450 nm for 40-60
min. Results are the average of triplicate time points in which the
typical standard deviation was <5%. Each compound was tested
at least twice, yielding essentially identical results.
Acknowledgment. We thank the NIH (CA10658, D.R. and
J.W.S.; CA 81713, J.W.S.; NIAID 055789, J.W.S.; U54
RR020843-01, J.W.S.), the NSF (CHE 9624532, D.R.), the
D.o.D Prostate Cancer Research Program (W81XWH-04-1-
0515, J.W.S.), the Welch Foundation (A-1280, D.R.), and Pfizer
(D.R.) for support of these investigations. We thank Dr. Huda
Henry-Riyad for help with the in situ IR spectroscopy. We thank
Prof. Michael Calter (Wesleyan) for helpful discussions and Ms.
Amanda S. Barela for assistance (summer NIH Bridges student).
We thank Dr. Joe Reibenspies (TAMU) for the X-ray analysis.
Supporting Information Available: General procedures for
dimerization, hydrogenation, and subsequent transformations with
characterization data (including1H and13C NMR spectra) for ketene
dimers 2a-d, 2f, ?-lactones cis-3a-f, trans-3a, 3c, and 8a-8c
and nOe data for ?-lactone 8b. This material is available free of
charge via the Internet at http://pubs.acs.org.
Purohit et al.
4558 J. Org. Chem., Vol. 71, No. 12, 2006