Preparation of dicarboxylate analogues of Cerulenin

Article abstract of DOI:10.1002/jhet.5570420511

We have proposed and synthesized several new structural classes of Cerulenin analogues, which have potential as inhibitors of both fatty acid and polyketide synthase multi-enzyme complexes. These analogues contain cis epoxides bearing flanking carboxylate groups. Our syntheses have been designed to allow access to a wide range of fatty acid and polyketide-like side chains from readily available starting materials in convergent fashion in just four to five steps. In total, ～40 potential analogues have been prepared and characterized, covering all the structural sub-classes proposed, the majority of which constitute novel functional groupings.

Full text of DOI:10.1002/jhet.5570420511

Jul-Aug 2005
819
Preparation of Dicarboxylate Analogues of Cerulenin
$
Jonathan D. Moseley* and James Staunton
University Chemical Laboratories, Lensfield Road, Cambridge, CB2 1EW, UK
Present address: AstraZeneca, Process Research and Development, Avlon Works, Hallen, Bristol, BS10 7ZE, UK
E-mail: jonathan.moseley@astrazeneca.com
$
Fax: +44 117 938 5601
Received November 3, 2004
We have proposed and synthesized several new structural classes of Cerulenin analogues, which have
potential as inhibitors of both fatty acid and polyketide synthase multi-enzyme complexes. These
analogues contain cis epoxides bearing flanking carboxylate groups. Our syntheses have been designed to
allow access to a wide range of fatty acid and polyketide-like side chains from readily available starting
materials in convergent fashion in just four to five steps. In total, ~40 potential analogues have been
prepared and characterized, covering all the structural sub-classes proposed, the majority of which consti-
tute novel functional groupings.
J. Heterocyclic Chem., 42, 819 (2005).
Introduction.
of these is on the role of the unusual 1,3-unsaturated
side chain, and particularly its role in inhibiting FAS
and other systems [10-12]. Biochemical studies have
shown that the cis-epoxide flanked by two carbonyl
groups is critical to Cerulenin’s inhibitory activity [2f],
in which a cysteine residue opens the epoxide at C-2
(3) [13], thus irreversibly deactivating the β-ketoacyl
synthase (Figure 2).
Cerulenin (1) is a broad-spectrum antibiotic that was
first detected in the fermentation broth of Cephalosporium
caerulens due to its potent antifungal properties [1]. It was
quickly shown to be active against the fatty acid synthases
(FAS) and polyketide synthases (PKS) of many fungi,
yeast and bacteria [2]. Consequently, it has been greatly
used as a biochemical tool in a wide range of biochemical
studies [3], and not only limited to the biochemical eluci-
dation of fatty acid and polyketide metabolic pathways [4].
Cerulenin’s structure was finally confirmed in 1974 by
Arison [5]. A number of total and formal syntheses, in both
racemic and homochiral form, have been reported and these
were comprehensively reviewed by Yoda in 1993 [6]. Since
then three further homochiral syntheses have been reported
by Hegedus [7], Townsend [8] and Ghosez [9]. An added
complication for the synthesis of Cerulenin is that in protic
solvents the open chain form 1 exists in equilibrium with its
hydroxylactam form 2, which is expected to predominate in
biological systems [10] (Figure 1).
In spite of the successful syntheses surveyed by Yoda
[6], the all-carbon backbone of Cerulenin containing
the dicarbonyl flanked cis epoxide functionality
remains synthetically challenging. Syntheses are often
lengthy, especially where the side chain is carried
through the synthetic route. However, the more conver-
gent syntheses [8-9, 14] could be amenable to the intro-
duction of all-carbon analogues in a more efficient
fashion as the authors propose, by addition of the C /C
8
9
side chains to homochiral C /C “head unit” synthons
4
3
(Scheme 1).
Figure 1. Interconversion of Cerulenin tautomers.
We were interested in Cerulenin as a potential tool
for elucidating the mechanism of polyketide biosynthe-
sis. Our ultimate aim was to substitute the unsaturated
fatty acid-like side chain of Cerulenin with a tailor-
made polyketide-like side chain designed for a specific
biosynthesis, whilst retaining the inhibitory properties
of Cerulenin. A number of studies of Cerulenin ana-
logues have been reported, but the focus of the majority
Figure 2. Cysteine opening of Cerulenin’s epoxide functionality.

820
J. D. Moseley and J. Staunton
Scheme 1
Vol. 42
Retrosynthesis of Cerulenin.
We speculated that the C head unit might be replaced
by a series of similar functional groups based on dicar-
hydroxy analogues, including 14, showed that the lack of a
carbonyl group at C-4 destroyed the inhibitory activity
[10]. This could be attributed to their inability to form the
hydroxylactam tautomers. In contrast, the 4-keto ana-
logues with a dimethyl amide, which also cannot form
hydroxylactams, were active in Tishler’s hands in both the
cerulenin-like (15) and tetrahydrocerulenin (16) cases (but
the cyclic amides pyrrolidine and piperidine were inactive)
(Figure 4). However, a more comprehensive study by
Lichtenhaler [12] of several 4-hydroxy and dimethylamide
analogues with variable chain lengths including 16,
showed no activity against any plant FAS [16]. The only
other analogue that has featured variation in the head unit
is the carbocyclic analogue prepared by Iwasaki [17] (17),
a deliberate attempt to mimic the hydroxylactam form 2.
This molecule, although a direct analogue of racemic
Cerulenin with the 1,3-unsaturated side chain, was effec-
tively inactive against yeast FAS, which tends to discount
2 as the active form. However, the authors also noted that
the nitrogen atom may be important in this case.
It was hoped that our analogues would help to resolve
this question since they retained the cis epoxide with the
flanking dicarbonyl motif, but would be unable to form the
hydroxylactams [18]. Interestingly, whilst many analogues
of 1 have been prepared and investigated [10-12,17], they
have all been carbon analogues (13). Studies in these cases
have usually focused on relatively simple analogues in the
side chain [11,12], and variation in the head unit has been
restricted to the relatively minor modifications of substi-
tuted amides discussed above [10,14]. With the exception
of 17, the dicarboxylate analogues 4-12 are the first signif-
icant examples of analogues of the head unit.
4
boxylates, which would be easier to synthesize. This was
important as we expected that the potential polyketide-like
side chains we required would be synthetically challenging
in themselves. Addition of either simple or complex tailor-
made side chains to these analogous head units would
allow us quickly to explore the structure-activity relation-
ships of Cerulenin analogues in a wide range of biologi-
cally interesting systems.
Therefore, we proposed the structures 4-12, which
retained the cis epoxide functionality and the flanking car-
bonyls groups, but with both carbonyls at the carboxylate
oxidation level i.e. replacing the ketone with an ester (4-6),
amide (7-9) or thioester (10-12) function (Figure 3). The
syntheses of these compounds would be greatly simplified
compared to their all-carbon analogues (13). To further
increase the scope of analogue synthesis, the other car-
boxylate function was to be from a common intermediate.
Thus we decided to prepare the carboxylic acids (4, 7, 10),
[15] the methyl esters (5, 8, 11), and the carboxamides (6,
9, 12) in each series.
There is some debate as to which is the active form of
Cerulenin, 1 or 2, although this may be dependent on the
individual biological system. Tishler’s results on the 4-
Results and Discussion.
The ester-acids (4) were readily prepared starting from
the known epoxymaleic anhydride 19 (Scheme 2). Sodium
tungstate catalysed epoxidation of maleic acid or its anhy-
dride gave the disodium salt in 99% yield on a molar scale,
which after conversion through its barium salt gave the
Figure 3. Proposed dicarboxylate analogues of Cerulenin (target structures
boxed in following schemes).

Jul-Aug 2005
Preparation of Dicarboxylate Analogues of Cerulenin
821
Figure 4. All-carbon analogues of Cerulenin.
free acid (18) in 75-80% yield overall [19]. The key anhy-
dride 19 [20] was formed by stirring in neat trifluoroacetic
anhydride [21] or, more economically, by refluxing 18
with 2 equivalents of TFAA in dichloromethane. Ring
opening of 19 was achieved by dissolution in the appropri-
ate neat alcohol to give the half ester-acids 4a-j in near
quantitative yields (Scheme 2 and Table 1). These were
generally viscous oils or gums that were difficult to purify,
as typified by the known 4a [21,22]. However, the octyl
(4h) and benzyl (4j) analogues did give crystalline prod-
ucts. Others were successfully used crude to give later
intermediates. For example, both symmetrical (20) and
unsymmetrical (5) diesters [23] could be prepared from
these half ester-acids by activation with PCl via 21, and
reaction with the appropriate alcohol.
The shorter chain symmetrical esters (5a, 20b) could
also be prepared directly from 18 by acid-catalysed esteri-
fication [24]. Attempts to make the longer chain esters (i.e.
5
>iPr) led to ether formation instead, so PCl activation of
5
18 via the diacid chloride (not shown) had to be used to
prepare, for example, 20e and 20j. An alternative attempt
to directly activate diacid 18 with SOCl and quench with
2
methanol led to formation of the chlorohydrin 22 instead.
The carboxamide series (6) was prepared in a similar fash-
ion, activating the ester-acids with PCl and then quench-
5
ing the intermediate acid chlorides (21) with NH to give
3
Scheme 2
Reagents and conditions: (i) TFAA (neat), TFA (cat.); (ii) TFAA (2.0 eq.), CH Cl ; (iii) ROH; (iv) PCl , CH Cl ; (v) MeOH; (vi) NH (g); (vii) ROH,
2
2
5
2
2
3
+
H ; (viii) PCl , py., ROH; (ix) SOCl , MeOH.
5
2

822
J. D. Moseley and J. Staunton
Vol. 42
the ester-amides 6c and 6f-h. Yields were lower for the
amide series, but adequate for our purposes. All three
routes were short and ideal for analogue synthesis with the
potential to add more complex side chains as a last step.
Activation of the amide-acids (7) with PCl , oxalyl
5
chloride or DCC and then quenching of the intermedi-
ates with methanol gave moderate yields of the
aliphatic amide-esters, which were difficult to isolate
and purify (only 8g characterized). The NMR data sug-
gested they might co-exist in their methoxy-lactam
forms (23) (Scheme 4). Furthermore, these methods
failed altogether for the aromatic acids (7k-p), giving
complex mixtures of brown gum; hygroscopicity was
thought to be an issue. A better procedure for the
aliphatic amide-esters 8, and for the PMB example 8q
was to start from the ester-acids 4a or crude 4b, acti-
Table 1
Key for Generic R Group Substituent on all Schemes
a = Me f = nHex
b = Et g = nHep
c = iPr h = nOct
d = nBu i = PhCH CH n = 2,4-diNO C H NH
k = Ph
l = 4-MeOC H
m = 4-NO C H
p = 2 x PhCH
2
q = 4-MeOC H CH
6
4
6
4
2
r = nPen
2
6 4
2
2
2 6 3
e = tBu j = PhCH
o = Ph C
s-w = as noted
on schemes
2
3
vate with DCC or PCl , and quench with the appropri-
5
ate amine (Scheme 4; only 8g, 8q and 24 prepared).
Simple treatment of 8q with methanolic ammonia gave
the PMB-diamide 9q; in the alkyl series 9g was pre-
pared directly from 7g instead. In view of the difficul-
ties encountered with the synthesis of structures 8 and
9, alternative routes to these classes of compounds
were attempted which are discussed below. However,
examples were prepared in each class of target mole-
cule (7-9) as planned.
The third major group of target molecules was that of
the thioesters 10-12. Curiously, dissolution of anhydride
19 in neat thiol to give the thioester-acids 10 did not pro-
ceed as for the alcohols 4, and none could usefully be
prepared in pure form. Instead, ester-acids 4 were acti-
vated with DCC and quenched with stoichiometric thiol
to give the thioester-esters (11c, e, g, r, 25t, 25u) in mod-
erate to good yields (Scheme 5). Symmetrical
dithioesters could be also prepared under these condi-
tions directly from acid 18, of which a single example 26
was prepared.
Our second series of targets were the amide analogues
(7-9). Ring opening of the anhydride 19 with aromatic
amines was a known strategy [20]. Several anilines, a
hydrazine, tritylamine and dibenzylamine all reacted
smoothly and in good yields to give the highly crystalline
aromatic amide-acids (7k-p) (Scheme 3). However, 4-
methoxybenzylamine failed to react cleanly and its ana-
logue 7q could not be prepared. Simple aliphatic amines
required more forcing conditions (overnight reflux in
THF), and gave the desired amide-acids in fair yields but
only moderate purity.
Scheme 3
Since the thioester-acids 10 could not be prepared, the
analogous strategy as that used for the amide-amides 9 (cf.
Reagents and conditions: (i) ArNH , Et O.
2
2
Scheme 4
Reagents and conditions: (i) 4a, PCl , CH Cl , then heptylamine; (ii) 4a, DCC, DMAP, CH Cl then 4-PMB-NH ; (iii) crude 4b, PCl , CH Cl , then 4-
5
2
2
2
2
2
5
2
2
PMB-NH ; (iv) crude 4g, PCl , CH Cl , then NH (g); (v) NH (aq), MeOH.
2
5
2
2
3
3

Jul-Aug 2005
Preparation of Dicarboxylate Analogues of Cerulenin
Scheme 5
823
Reagents and conditions: (i) ROH; (ii) Ester-acid 4, DCC, DMAP, THF, then RSH; (iii) 18, DCC, DMAP, THF, then nPenSH.
Scheme 4) could not be used to make the thioester-amides
12 (attempts to use the crude thioester-acids did not suc-
ceed). Two nitrogen-protected thioester-amides (27v, 27w)
were prepared from dibenzyl amide-acid (7p) under stan-
dard conditions (DCC, RSH), but the benzyl groups could
not be removed to give the respective primary carboxam-
ides (12g, 12r) (Scheme 6) (presumably due to catalyst
poisoning by thiol residues).
Alternatively, we tested the more facile removal of the O-
benzyl ester from diester 28 which gave the desired crude half
ester-acid 4d in high yield without reduction of the epoxide
(Scheme 7). Hydrogenolysis of the analogous thioester-ester
25u would have given access to a thioester-acid 10r from
which a thioester-amide 12r might have been prepared, but
these reductions again failed, even with the use of stoichio-
metric “catalyst”, and the more effective Pearlman’s catalyst.
Scheme 6
Reagents and conditions: (i) DCC, DMAP, THF, then RSH; (ii) Pd/C, H .
2
Scheme 7
Reagents and conditions: (i) DCC, DMAP, THF, then RXH; (ii) 25u or 28, Pd/C, H , THF; (iii) 25u, Pd(OH) , H , EtOH.
2
2
2

824
J. D. Moseley and J. Staunton
Vol. 42
Finally, although we had shown that thioester tertiary-
group, but would almost certainly have reduced the epox-
ide group [26]. However, anhydride 19 successfully
reacted with hexamethyldisilazane on refluxing in THF in
high crude yield. Crystallization from methanol gave a
yellow solid, which was not the expected silyl amide (31),
but rather the primary amide acid (32) [27], an apparently
novel compound [28].
A further effort to find a better route to the amide-acids
7 and amide-esters 8 was investigated by ring-opening
maleic anhydride with amines in toluene to give the known
amides such as 27v and 27w could be prepared, we had
failed to synthesize the primary amide versions (eg. 12g,
12r, Scheme 6), which were of special interest. Since
removal of the dibenzyl groups had been unsuccessful in
the thioester series, we investigated other potential protect-
ing groups for primary amides. These were the phthalim-
ide (29), 2,5-dimethyl pyrrole (30), hydrazide (7n), trityl
(7o), dibenzyl (7p) and hexamethyldisilazane (31) substi-
tuted analogues (Figure 5).
Figure 5
Reaction of anhydride 19 with potassium phthalimide
under a variety of conditions failed to give the desired
amide-acid (29, R = H). Reaction with ester-acid 4a in
place of the anhydride, also failed to give the desired prod-
uct (29, R = Me), although some reaction did take place in
crystalline maleamic-acids 33 in good yields [29].
However, no example of 33 could be epoxidised using
either the classical H O /Na WO catalysed conditions
[19] or alkaline NaOCl, as reported by Tishler [10], pre-
sumably due to the strongly deactivated nature of the dou-
ble bond. Finally, other oxidation states (protected alco-
hols and aldehydes) were investigated as alternative ways
of arriving at the desired carboxylate functionalities, but
these routes were potentially longer and less flexible, and
the studies were inconclusive, and so are not reported here.
2
2
2
4
this case (PCl , py.). A less common choice of primary
5
amide protecting group is 2,5-dimethyl pyrrole [25] but
again, reaction with either 19 or 4a either with neat pyrrole
or its sodium salt failed to give the desired amide
acid/ester products (30, R = H or Me). In all cases, organic
mass recoveries from the resulting black tars were low.
The hydrazido amide-acid (7n) and trityl amide-acid
(7o) had been synthesized but as noted above their analo-
gous amide-esters could not be prepared. Deprotection of
the trityl group in 7o might have hydrolysed the epoxide
and/or ester as well, but hydrogenolysis of the N-N bond in
7n would have been expected to be facile and would have
given access to the ester-amide series 6. Unfortunately,
hydrogenation of 7n was initially unsuccessful and not fur-
ther investigated. Dibenzylamine 7p could not be hydro-
genated under standard conditions; more forcing condi-
Conclusions.
In conclusion, we have synthesized a wide range of
Cerulenin analogues (~40) covering all the structural sub-
classes of the proposed analogues, the majority of which
constitute novel functional groupings. Our syntheses are
short (four to five steps) and are from readily available
starting materials. Most importantly, they are amenable to
the addition of more complex side chains in a convergent
fashion to the key head unit building blocks. These ana-
logues have potential as inhibitors of both FAS and PKS
multi-enzyme complexes. Biological results will be
reported elsewhere.
tions (40-50 psi H ) might well have removed the benzyl
2

Jul-Aug 2005
Preparation of Dicarboxylate Analogues of Cerulenin
825
EXPERIMENTAL
Monooctyl-cis-2,3-epoxysuccinate (4h).
This compound was isolated in quantitative yield as a white
-1 1
Melting points were determined on a Reichert hot-stage
apparatus or a Buchi 510 melting point apparatus and are uncor-
rected. Elemental analyses (C, H, N and S) were performed by
the “in house” chemistry department analytical service. Infra
red spectra were run on Perkin-Elmer 297 or 1310 spectropho-
syrup; ir (CHCl ): 2920, 2860 (C-H), 1750 (C=O) cm ; H NMR
3
(250 MHz, CDC ): δ 7.5-8.2 (1H, bs, OH), 4.22 (2H, t, J=6.8,
3
CH O), 3.78 (1H, d, J=4.7) and 3.74 (1H, d, J=4.7, epoxide),
2
1.66 (2H, qu, J=6.8, CH CH O), 1.27 (10H, m, (CH ) ), 0.87
2
2
2 5
+
+
(3H, t, J=6.5, CH ); ms: m/z (EI ) 244 (M , 100%), 199 (M-
3
-1
tometers between 4000 and 600 cm and only readily assigned
CO H, 45), 169 (85), 157 (55), 133 (100), 115 (100).
2
-1
1
absorbances above ~1500 cm are listed. H NMR spectra
were recorded on Bruker WM-250 (250 MHz), AC-250 (250
MHz), or AM-400 (400 MHz) spectrometers. The following
additional abbreviations have been used for first order spectra:
qu = quintet, sx = sextet, sp = septet, m = multiplet, b = broad.
Anal. Calcd. for C
H O : C, 58.99; H, 8.27. Found: C,
12 20 5
59.17; H, 8.24.
Monobenzyl-cis-2,3-epoxysuccinate (4j).
This compound was isolated in quantitative yield as a white
13
C NMR spectra were recorded on Bruker AM-400 (100.6
waxy solid; mp 73-74 °C (neat); ir (CHCl ): 3300-3600 (O-H),
3
-1
1
MHz) or AC-250 (62.9 MHz) spectrometers and were proton
2880, 2810 (C-H), 1757 (CO H), 1742 (CO Bz) cm ; H NMR
2
2
+
decoupled. Electron impact (EI ) mass spectra were recorded
(250 MHz d -acetone): δ 7.35 (5H, m, PhO), 5.23 (1H, d, J=12.4)
6
on Kratos MS 902 (ex A.E.I.) or MS 890 spectrometers. Fast
and 5.17 (1H, d, J=12.4, PhCH O), 3.93 (1H, d, J=4.7) and 3.85
2
+
+
+
atom bombardment (FAB ) mass spectra (carrier gas Xe) were
(1H, d, J=4.7, epoxide); ms: m/z (EI ) 222 (M , 25%), 178 (M-
CO , 35), 150 (18), 115(70), 107 (PhCH O, 55), 91 (PhCH , 100).
recorded on Kratos MS 50 or MS 890 spectrometers. Analytical
TLC was carried out on commercially prepared plates coated
2
2
2
Anal. Calcd. for C H O : C, 59.45; H, 4.55. Found: C,
11 10
5
with 0.25 mm of self-indicating Merck Kieselgel 60 F
and
59.25; H, 4.48.
254
were developed using a 5% solution of phosphomolybdic acid
in ethanol. Preparative flash silica chromatography was per-
formed using Merck Kieselgel 60 (230-400 mesh). All other
reagents and solvents were purified and dried as necessary
according to standard procedures.
General Method for Lower Homologue Symmetrical Epoxy-
succinate Diesters 20b and 5a.
The following general method is exemplified by the procedure
described below for the preparation of 20b, starting from diacid
18 and other reagents scaled accordingly.
cis-2,3-Epoxysuccinic anhydride (19) [20, 21].
Diethyl-cis-2,3-epoxysuccinate (20b) [24].
The diacid 18 [19,20] (0.75 g, 5.7 mmol) was stirred
overnight in TFAA (6 ml) at RT. The excess TFAA was
removed by concentration and the pale pink residual solid was
purified by sublimation at 50 °C and 0.1 mmHg for 3-4 hours
to yield the title compound as colourless prisms (0.65 g,
Concentrated H SO (7.8 g, 80 mmol) was added dropwise
2
4
over 1 hour to a chilled suspension of the barium salt of 18 [19]
(16.0 g, 60 mmol) in absolute ethanol (80 ml), keeping the tem-
perature below 10 °C. The mixture was allowed to warm to 20
°C overnight and heated at reflux for 5 hours. On cooling, cal-
cium carbonate (2.0 g) was added and the slurry filtered through
Celite washing copiously with ether. The crude concentrate was
partitioned between water (50 ml) and ether (50 ml), separated
and the aqueous layer extracted with ether (3 x 40 ml). The
combined ethereal extracts were washed sequentially with satu-
100%); mp 64-65 °C (lit.[21] mp 63-64 °C); ir (CHCl ): 1887,
3
–1
1
1875, 1808 and 1756 cm ; H NMR (250 MHz, CDCl ): δ
4.31 (s); ms: m/z (EI ) 115 (M+1, 10%), 102 (35), 87 (25), 71
3
+
(95), 69 (M-CO H, 60), 59 (28), 44 (CO , 100). Anal. Calcd.
2
2
for C H O : C, 42.12; H, 1.77. Found: C, 41.85; H, 1.74.
4
2 4
Alternatively, the diacid 18 (0.99 g, 7.5 mmol) was heated at
reflux with TFAA (2.1 ml, 15.0 mmol) in dichloromethane (8
ml) for 5 hours. The crude concentrate was purified by subli-
mation to yield the title compound as colourless prisms (0.79
g, 93%); mp 63-65 °C. Other data as above.
rated NaHCO solution (40 ml), water (40 ml), dried (Na SO )
3
2
4
and concentrated to a yellow oil (6.1 g, 54%); ir (CHCl ): 2900
3
-1
1
(C-H), 1745 (C=O) cm ; H NMR (250 MHz, CDCl ): δ 4.18
3
(4H, q, J=7.1, 2 x CH O), 3.63 (2H, s, epoxide), 1.24 (6H, t,
2
13
J=7.1, 2 x CH ); C NMR (100.6 MHz, CDCl ): δ 165.62
3
3
General Method for Epoxysuccinic Ester-Acids 4a-j.
(C=O), 61.85 (CH O), 52.41 (epoxide), 14.86 (CH ); ms: m/z
2
3
+
+
The anhydride 19 (114 mg, 1.0 mmol) was stirred the appro-
priate distilled alcohol (3-5 ml) for 3-48 hours. The excess alco-
hol was removed by evaporation at reduced pressure to give
quantitative yields of the desired compounds which were used
without further purification. Full characterisations are given only
for those compounds isolated in their own right (i.e. the known
4a, and 4h and 4j).
(EI ) 188 (M , 40%), 161 (100), 143 (M-OEt, 100), 132 (30),
115 (100), 70 (100).
Dimethyl-cis-2,3-epoxysuccinate (5a) [22].
This compound was prepared according to the method for 20b
and isolated as a yellow oil (4.4 g, 56%); ir (CHCl ): 1745 (C=O)
3
-1
1
cm ; H NMR (250 MHz, CDCl ): δ 3.79 (6H, s, 2, x Me), 3.71
3
+
+
(2H, s, epoxide); ms: m/z (EI ) 160 (M , 75%), 147 (M-Me, 15),
129 (M-OMe, 100), 115 (80), 71 (80).
Monomethyl-cis-2,3-epoxysuccinate (4a) [21, 22].
This compound was isolated in quantitative yield as a viscous
General Method for Higher Homologue Symmetrical Epoxy-
succinate Diesters 20e and 20k.
gum; ir (CHCl ): 1760 (CO H), 1746 (CO Me) and 1711
3
2
2
-1; 1
(CO H) cm
H NMR (250 MHz CDCl ): δ 6.1-6.8 (1H, bs,
OH), 3.85 (3H,s, Me), 3.80 (1H, d, J=4.9) and 3.77 (1H, d, J=4.9,
epoxide); ms: m/z (EI ) 147 (M+1, 2%), 129 (M-OH, 12), 115
2
3
The following general method is exemplified by the procedure
described below for the preparation of 20e, starting from diacid
18 and other reagents scaled accordingly.
+
(M-OMe, 10), 102 (M-CO , 27), 87 (20), 71 (100).
2

826
J. D. Moseley and J. Staunton
Vol. 42
Di-tert-butyl-cis-2,3-epoxysuccinate (20e).
dichloromethane (6 ml). The ice-bath was removed and the solu-
tion allowed to warm to 20 °C over 3 hours. After this time the
ice-bath was replaced and a dry ice-acetone condenser added.
Liquid ammonia was condensed into the reaction mixture for 5-
10 minutes, and then stirred for a further 30 minutes. After warm-
ing to 20°C, the reaction was quenched by the addition of water
(3 ml), separated and extracted with chloroform (4 x 3 ml). The
PCl (2.50 g, 12.0 mmol) was added on one portion to a chilled
5
(0 °C) suspension of diacid 18 (0.66 g, 5.0 mmol) in
dichloromethane (12 ml). The cooling bath was removed and
solution stirred at 20 °C for 3 hours. The resulting homogeneous
solution was cooled again to 0 °C and t-butyl alcohol (1.48 g,
20.0 mmol) dissolved in dichloromethane (1 ml) was added drop-
wise, followed by pyridine (2.1 ml, 30.0 mmol) after 5 minutes.
The solution was warmed to 20 °C over 2 hours, diluted with
combined organic extracts were dried (Na SO ) and concentrated
2
4
to a crude oil, which was purified by flash silica gel chromatogra-
phy in 1:1 hexane:ethyl acetate (R 0.24) to yield the title com-
f
chloroform (20 ml) and washed sequentially with 1 M H SO (3
2
4
pound as a white solid (118 mg, 55%); mp 136-138 °C; ir
x 30 ml), saturated NaHCO (2 x 30 ml), dried (Na SO ), and
3
2
4
(CHCl ): 3495, 3380 (N-H), 2920 (C-H), 1735 (CO R), 1685
3
2
concentrated to a brown oil. The crude product was purified by
flash silica gel chromatography in CHCl (R 0.28) to yield the
-1
1
(CONH) cm ; H NMR (250 MHz, CDCl ): δ 6.42 (1H, bs) and
5.66 (1H, bs, NH ), 5.08 (1H, sp, J=6.3, CH(CH ) ), 3.68 (3H, d,
3
3
f
2
3 2
title compound as a white solid (0.54 g, 44%); mp 101.5-102 °C
J=5.0) and 3.65 (1H, d, J=5.0, epoxide), 1.29 (3H, d, J=6.3) and
-1
1
(CHCl ); ir (CHCl ): 2900 (C-H), 1745 (C=O) cm ; H NMR
3
3
+
1.24 (3H, d, J=6.3, CH(CH ) ); ms m/z (EI ) 174 (M+1, 6%),
3 2
(400 MHz, CDCl ): δ 3.54 (2H, s, epoxide), 1.49 (9H, s, (CH ) );
3
3 3
+
173 (M , 5), 158 (M-NH , 2), 132 (100), 130 (100), 114 (110),
13
2
C NMR (100.6 MHz, CDCl ): δ 165.19 (C=O), 61.07, 52.89
(epoxide), 27.88 (CH ); ms: m/z (EI ) 244 (M , 22%), 187 (M-
3
88(100), 70(100).
+
+
3
Anal. Calcd. for C H NO : C, 48.54; H, 6.41; N, 8.09. Found:
7
11
4
tBu, 32), 171 (M-O tBu, 100), 115 (100), 70 (75).
C, 48.55; H, 6.40; N, 8.16.
Anal. Calcd. for C H O : C, 58.99; H 8.27. Found: C, 58.84;
12 20
5
H, 8.41.
4-Hexyl-cis-2,3-epoxysuccinamate (6f).
Dibenzyl-cis-2,3-epoxysuccinate (20j).
This compound was isolated as a white waxy solid (64%); mp
136-137 °C; ir (CHCl ): 3490, 3375 (N-H), 2905, 2840 (C-H),
3
This compound was isolated after flash silica gel chromatogra-
-1
1
1735 (CO R), 1680 (CONH) cm ; H NMR (250 MHz, CDCl ):
2
3
phy in 4:1 hexane:ether (R 0.18) as a yellow waxy solid (89%);
f
δ 6.4 (1H, bs) and 5.6 (1H, bs NH ), 4.18 (2H, t, J=6.8, CH O),
-1
2
2
mp 45-48 °C (neat). ir (CHCl ): 2890 (C-H), 1745 (C=O) cm ;
3
3.72 (1H, d, J=5.0) and 3.66 (1H, d, J=5.0, epoxide), 1.62 (2H,
qu, J=7.0, CH CH O), 1.3 (6H, m, (CH ) ), 0.88 (3H, t, J=6.7,
1
H NMR (400 MHz, CDCl ): δ 7.29-7.36 (10H, m, Ph), 5.07
3
13
2
2
2 3
(4H, s, 2 x CH O), 3.70 (2H, s, epoxide); C NMR (100.6 MHz,
2
+
+
CH ); ms: m/z (EI ) 216 (M+1, 5%), 215 (M , 2), 199 (M-NH ,
3
2
CDCl ): δ 165.45 (C=O), 134.71, 128.65, 128.61, 67.54 (CH O),
3
2
1), 172 (M-CONH, 3), 159 (6), 132 (100), 114 (100), 85 (100).
+
+
52.60 (epoxide); ms: m/z (EI ) 312 (M , 55%), 223 (M-Bz, 17),
Anal. Calcd. for C
H NO : C, 55.79; H, 7.98; N, 6.51.
10 17 4
207 (M-BzO, 75), 107 (PhCH O, 45), 91 (PhCH , 100).
2
2
Found: C, 55.95; H, 7.93; N, 6.40.
Anal. Calcd. for C
H O : C, 69.21; H, 5.17. Found: C,
18 16 5
69.48; H, 5.15.
4-Heptyl-cis-2,3-epoxysuccinamate (6g).
2-Chloro-1,4-dimethyl-3-hydroxysuccinate (22).
This compound was isolated as a white solid (67%); mp 42.5-
43.5 °C; ir (CHCl ): 3490, 3380m (N-H), 2910, 2850 (C-H),
3
Thionyl chloride (3.7 ml, 51.0 mmol) was added dropwise to
methanol (20 ml) cooled to –15 °C. Diacid 18 [19,20] (3.3g, 25.0
mmol) was added on one portion and the solution stirred at –10
°C for 1 hour before warming to 20 °C overnight. The crude con-
-1
1
1740 (CO R), 1685 (CONH) cm ; H NMR (250 MHz, CDCl ):
2
3
δ 6.42 (1H, bs) and 5.57 (1H, bs, NH ), 4.18 (2H, t, J=6.8,
2
CH O), 3.72 (1H, d, J=5.0) and 3.66 (1H, d, J=5.0, epoxide),
2
1.61 (2H, m, CH CH O), 1.27 (8H, m, (CH ) ), 0.87 (3H, t,
centrate was partitioned between saturated NaHCO solution (30
2
2
2 4
3
+
J=6.6, CH ); ms: m/z (EI ) 230 (M+1, 4%), 172 (2), 159 (3), 132
ml) and chloroform (30 ml), separated and the aqueous layer
extracted with chloroform (2 x 30 ml). The combined organic
3
(100), 114 (100), 105 (28), 98 (40), 86 (97), 70 (99).
Anal. Calcd. for C H NO : C, 57.61; H, 8.37; N, 6.11.
extracts were dried (MgSO ) and concentrated to yield the title
11 19
4
4
Found: C, 57.80; H, 8.32; N, 6.28.
compound as a colourless oil (4.4 g, 86%); ir (film): 3500 (O-H),
-1
1
2900 (C-H), 1753 (C=O) cm ; H NMR (250 MHz, CDCl ): δ
3
4-Octyl-cis-2,3-epoxysuccinamate (6h).
4.89 (1H, dd, J=7.5, 2.4, CHOH), 4.80 (1H, d, J=2.4, CHCl),
3.85 (3H, s, Me), 3.84 (3H, s Me), 3.31 (1H, d, J=7.5, OH);
This compound was isolated as a white solid (68%); mp 42-43
13
C
°C; ir (CHCl ): 3495, 3380 (N-H), 2910, 2840 (C-H), 1740
NMR (100.6 MHz, CDCl ): δ 170.60, 166.80 (2 x C=O), 71.80
3
3
-1
1
(CO R), 1685 (CONH) cm ; H NMR (250 MHz, CDCl ): δ
6.42 (1H, bs) and 5.62 (1H, bs NH ), 4.16 (2H, t, J=6.8, CH O),
(CHOH), 59.66 (CHCl), 53.62 and 53.33 (2 x MeO); ms: m/z
2
3
+
(FAB ) 197/199 (M+1, 100), 137/139 (M-CO Me, 80).
2 2
2
3.72 (1H, d, J=5.0) and 3.66 (1H, d, J=5.0, epoxide), 1.64 (2H,
qu, J=6.9, CH CH O), 1.25 (10H, m, (CH ) ), 0.87 (3H, t, J=6.6,
Anal. Calcd. for C H ClO : C, 36.65; H, 4.62. Found: C,
6
9
5
36.90; H, 4.36.
2
2
2 5
+
CH ); ms: m/z (EI ) 244 (M+1, 15%), 200 (M-CONH, 5), 159
3
General Method for Epoxysuccinamate Esters, 6c and 6f-h.
(12), 147 (86), 133 (95), 132 (100).
Anal. Calcd. for C
Found: C, 59.20; H, 8.97; N, 5.58.
H NO : C, 59.23; H, 8.72; N, 5.76.
The following general method is exemplified by the procedure
described below for the preparation of 6c, using the appropriate
crude ester-acid 4c and other reagents scaled accordingly.
12 21 4
General Method for Epoxysuccinamic Acids 7k-p.
4-iso-Propyl-cis-2,3-epoxysuccinamate (6c).
Anhydride 19 (generally 228 mg, 2.00 mmol) was heated at
reflux in diethyl ether with the appropriate aniline (2.05 mmol)
for 2-6 hours, or at reflux with the appropriate amine in THF for
PCl (312 mg, 1.5 mmol) was added in one portion to a chilled
5
(-5 °C) solution of crude ester-acid 4c (220 mg, 1.25 mmol) in

Jul-Aug 2005
Preparation of Dicarboxylate Analogues of Cerulenin
827
18 hours. The title compounds were isolated by filtration from
the cooled reaction mixture and recrystallised if necessary from
acetone-hexane.
Anal. Calcd. for C H NO : C, 73.97; H, 5.14; N, 3.75.
23 19 4
Found: C, 73.57; H, 5.17; N, 3.83.
4-(N,N-Dibenzyl)-cis-2,3-epoxysuccinamic acid (7p).
4-N-Phenyl-cis-2,3-epoxysuccinamic acid (7k) [20].
This compound was isolated as white prisms (448 mg, 72%);
This compound was isolated as fine white needles (374 mg,
90%); mp 179-181°C (dec.) (lit.[20] 179-180 °C); ir (KBr): 3360
(N-H), 2500 (O-H), 1735 (C=O.O), 1640 (C=O.N), 1605, 1595
mp 147-148 °C (acetone-hexane); ir (KBr): 2500 (O-H), 1730
1
(C=O.O), 1620 (C=O.N); H NMR (250 MHz, d -acetone): δ
6
7.24-7.39 (10H, m, Ph N), 4.77 (1H, d, J=16.4) and 4.65 (1H, d,
2
-1
1
(C=C) cm ; H NMR (80 MHz, d -acetone) 7.57-7.74 (2H, m)
J=4.7, PhCH N), 4.63 (1H, d, J=14.9) and 4.39 (1H, d, J=14.9,
6
2
and 7.00-7.44 (3H, m, Ph), 3.94 (1H, d, J=5.0) and 3.87 (1H, d,
Ph’CH N), 4.15 (1H, d, J= 4.7) and 3.83 (1H, d, J=4.7, epoxide),
2
+
+
13
J=5.0, epoxide); ms: m/z (EI ) 207 (M , 60%), 163 (M-CO , 8),
3.77 (3H, s, MeO); C NMR (100.6 MHz, d -acetone): δ 168.38
2
6
+
119 (25), 105 (22), 93 (PhNH , 100), 77 (Ph , 40).
and 166.31 (2 x C=O), 137.53, 137.34, 129.50, 129.25, 128.94,
2
Anal. Calcd. for C H NO : C, 57.96; H, 4.39; N, 6.76. Found:
C, 58.11; H, 4.36; N, 6.76.
128.41 and 128.06 (Ph N), 55.04 and 53.85 (epoxide), 50.00 and
10
9
4
2
+
48.14 (CH NCH ); ms: m/z (FAB ) 312 (M+1, 69%), 289 (8),
2
2
198 (Ph NH , 46), 154 (78), 136 (69), 107 (PhCH NH , 26), 91
2
2
2
2
4-N-(4-Methoxyphenyl)-cis-2,3-epoxysuccinamic acid (7l).
+
(PhCH , 100).
2
This compound was isolated as a purple coloured solid (391
mg, 82%); mp 160-162 °C (dec.) (acetone-hexane); ir (KBr):
3305 (N-H), 2510 (O-H), 1720 (C=O.O), 1640 (C=O.N), 1605
Anal. Calcd. for C
Found: C, 69.46; H, 5.53; N, 4.58.
H NO : C, 69.43; H, 5.51; N, 4.50.
18 17 4
4-N-(4-Methoxybenzyl)-cis-2,3-epoxysuccinamide (9q).
-1
1
(C=C) cm ; H NMR (250 MHz, d -acetone): δ 8.86 (1H, bs,
6
NH), 7.57 (2H, dd) and 6.89 (2H, dd, J=6.9, 2.2, ArH), 3.90
(1H, d, J=5.0) and 3.86 (1H, d, J=5.0, epoxide), 3.77 (3H, s,
DCC (413 mg, 2.00 mmol) and DMAP (12 mg, 0.10 mmol)
were added to a solution of ester-acid 4a (146 mg, 1.00 mmol) in
dichloromethane (5ml) cooled to -5°C. After stirring for 5 minutes,
4-methoxybenzylamine (143 µL, 1.10 mmol) was added by
syringe in one portion and the reaction mixture stirrred at 20 °C for
24 hours. The solution was filtered through a pad of Celite, washed
with dichloromethane and concentrated to dryness (600 mg). The
crude concentrate was purified by flash silica gel chromatography
+
+
MeO); ms: m/z (EI ) 237 (M , 38%), 219 (M-H O, 3), 192 (M-
2
+
CO H, 8), 148 (22), 133 (63), 121 (MeOC H N , 86), 107
2
6 4
+
(MeOC H , 72).
6
4
Anal. Calcd. for C H NO : C, 55.69; H, 4.68; N, 5.90.
11 11
5
Found: C, 55.56; H, 4.68; N, 5.98.
4-N-(4-Nitrophenyl)-cis-2,3-epoxysuccinamic acid (7m) [20].
in 1:1 hexane:ether (R 0.28) to yield the semi-pure 8q as a buff
f
This compound was isolated as a light cream coloured solid
(360 mg, 71%); mp 167-169 °C (dec.) (lit.[20] 178-179 °C); ir
(KBr): 3355 (N-H), 2500 (O-H), 1730 (C=O.O), 1655 (C=O.N),
solid (100 mg, 38%) which was used without further purification.
The amide-ester 8q (59 mg, 0.22 mmol) was stirred with 0.88
aqueous ammonia solution (40 µl, 0.80 mmol) and methanol (1.0
ml) in a sealed flask for 5 days. The resulting precipitate was col-
lected by filtration and dried in vacuo on an oil pump to yield the
title compound as a buff coloured solid (24 mg, 43%); mp 177-178
-1
1
1615, 1595 (C=C) cm ; H NMR (80 MHz, d -acetone): δ 9.5
6
(∼1H, bs, OH), 8.25 (2H, dt, J=9.5, 2.4) and 7.95 (2H, dt, J=9.5,
2.4, ArH), 4.7-5.2 (∼1H, vbs, NH), 4.00 (1H, d, J=5.0) and 3.93
+
+
(1H, d, J=5.0, epoxide); ms: m/z (EI ) 252 (M , 100%), 208 (M-
°C; ir (KBr): 3460 (N-H), 3290, 3180 (NH ), 1670, 1645 (CONH
2
+
-1
1
CO , 20), 138 (100), 115 (M-ArNH , 30).
and CONH ) cm ; H NMR (250 MHz, d -DMSO): δ 8.37 (1H,
2
2
6
Anal. Calcd. for C H N O : C, 47.62; H, 3.20; N, 11.11.
bt J=6.0, ArNHCO), 7.52 (1H, bs) and 7.37 (1H, bs, CONH ), 7.17
10
8 2 6
2
Found: C, 47.41; H, 3.10; N, 11.23.
(2H, d, J=8.6) and 6.86 (2H, d, J=8.7, Ar-H), 4.28 (1H, dd, J=14.9,
6.2) and 4.13 (1H, dd, J=14.9, 5.8, ArCH N), 3.73 (3H, s, MeO),
3.72 (1H, d, J=5.3) and 3.63 (1H, d, J=5.2, epoxide); C NMR
2
4-N-(2,4-Dinitrophenylhydrazino)-cis-2,3-epoxysuccinamic acid
(7n).
13
(100.6 MHz, d -DMSO): δ 167.33 and 165.13 (2 x C=O), 158.35,
6
This compound was isolated as a bright yellow solid (555 mg,
89%); mp 153-155 °C (dec.) (acetone-hexane); ir (KBr): 3340
130.73, 128.75 and 113.78 (C=C), 55.18 (MeO), 54.47 and 54.41
+
(epoxide), 41.53 (ArCH N); ms: m/z (EI ) 206 (M-CONH , 72%),
2
2
-1
+
+
(N-H), 2550 (O-H), 1720 (C=O.O), 1610, 1590 (C=O.N) cm ;
163 (44), 136 (ArCH NH , 100), 121 (ArCH , 77).
2
2
1
H NMR (250 MHz, CDC1 ): δ 9.6-11.5 (2H, vbs) and 9.80 (1H,
Anal. Calcd. for C H N O : C, 57.58; H, 5.65; N, 11.20.
3
12 14
2 4
bs, NHNH and OH), 8.98 (1H, d, J=2.6, Ar H-3), 8.28 (1H, dd,
J=9.5, 2.6, Ar H-5), 7.56 (1H, d, J=9.5, Ar H-6), 4.01 (1H, d,
J=4.9) and 3.95 (1H, d, J=4.9, epoxide); ms: m/z (FAB ) 313
(M+1, 89%), 312(M , 36), 165 (16), 136 (92), 107 (PhNHNH ,
33), 89 (34).
Found: C, 57.20; H, 5.77: N, 10.93.
1-Ethyl-4-N-(4-methoxybenzyl)-cis-2,3-epoxysuccinamate (24).
+
+
+
The method of 6c was used starting from crude ester-acid 4b
and 4-methoxybenzylamine to yield the title compound after
flash silica gel chromatography in 2% methanol:dichloromethane
Anal. Calcd. for C H N O : C, 38.47; H, 2.59; N, 17.95.
10
8 4 8
(R 0.48) as a white solid (376 mg, 67%); mp 102.5-103 °C; ir
Found: C, 38.12; H, 2.65; N, 18.11.
f
(CHCl ): 3395 (N-H), 1738 (C=O.O), 1661 (CONH), 1605
3
4-N-Trityl-cis-2,3-epoxysuccinamic acid (7o).
-1
1
(C=C) cm ; H NMR (250 MHz, CDCl ): δ 7.19 (2H, d, J=8.6)
and 6.84 (2H, d, J=8.6, Ar-H), 6.65-6.75 (1H, bs, NH), 4.39 (1H,
3
This compound was isolated as a white solid (255 mg, 67%); mp
175-179 °C (dec.) (acetone-hexane); ir (KBr): 2500 (O-H), 1730
dd, J=14.5, 6.2) and 4.29 (1H, dd, J=14.5, 5.6, ArCH N), 4.12
2
1
(C=O.O), 1630 (C=O.N); H NMR (250 MHz, d -acetone): δ 7.89
(2H, q, J=7.1, OCH ), 3.78 (3H, s, OMe), 3.71 (1H, d, J=5.0) and
6
2
(1H, s, NH), 7.20-7.39 (15H, m, Ph CN), 3.94 (1H, d, J=5.1) and
3.68(1H, d, J=5.0, epoxide), 1.22 (3H, t, J=7.1, CH CH ); ms:
3
2 3
+
+
+
3.73 (1H, d, J=5.1, epoxide); ms: m/z (EI ) 374 (M+1, 75%), 356
(M-OH, 44), 340 (63), 311 (100), 243 (100), 192 (100), 104 (100).
m/z (EI ) 279 (M , 100%), 206 (M-CO Et, 100), 177 (80), 164
2
(50), 136 (ArCH NH , 100), 121 (ArCH , 100), 109 (60).
+
+
2 2

828
J. D. Moseley and J. Staunton
Vol. 42
Anal. Calcd. for C
Found: C, 60.05; H, 6.09; N, 4.93.
H
NO : C, 60.20; H, 6.15; N, 5.02.
NMR (100.6 MHz, CDCl ): δ 193.30 (C=O.S), 165.70 (C=O.O),
14 17
5
3
58.14 (C-2), 53.72 (C-3), 52.68 (MeO), 30.79, 28.80, 28.71 and
+
22.11 (4x CH ), 13.87 (CH ); ms: m/z (FAB ) 233 (M+1, 100%),
2
3
General Method for Epoxysuccinic Thioester-Esters, 11, 25, 26,
and 28.
231 (M-1, 14), 201 (M-MeO, 28).
Anal. Calcd. for C O S: C, 51.70; H, 6.96; S, 13.80.
H
10 16
4
The following general method is exemplified by the procedure
described below for the preparation of 11c, using the appropriate
ester-acid 4a and other reagents scaled accordingly.
Found: C, 51.86; H, 6.86; S, 13.64.
4-Butyl-1-S-heptyl-cis-2,3-epoxysuccin-1-thioate (25t).
This compound was isolated as a yellow oil (32%); ir (CHC ):
2920, 2860 (C-H), 1745 (C=O.O), 1670 (C=O.S) cm ; H NMR
3
4-Methyl-1-S-(iso-propyl)-cis-2,3-epoxysuccin-1-thioate (11c).
-1
1
DCC (413 mg, 2.00 mmol) and DMAP (12 mg, 0.10 mmol)
were added to a solution of ester-acid 4a (146 mg, 1.00 mmol) in
THF (5 ml) cooled to –5 °C. After stirring for 5 minutes, iso-
propane thiol (93 µL, 1.0 mmol) was added by syringe in one
portion and stirring continued at this temperature for 45 minutes
or until TLC indicated the reaction was completed. The crude
reaction mixture was concentrated to dryness and purified by
(250 MHz, CDCl ): δ 4.16 (2H, q, J=6.7, OCH ), 3.78 (1H, d,
3
2
J=4.7) and 3.71 (1H, d, J=4.7, epoxide), 2.91 (2H, t J=7.3,
CH S), 1.50-1.67 (4H, m, CH CH O and CH CH S), 1.26-1.43
2
2
2
2
2
(10H, m, (CH ) ), 0.91 (3H, t, J=7.2, CH ), 0.86 (3H, t, J=6.7,
2 5
3
+
CH ); ms: m/z (EI ) 303 (M+1, 19%), 242 (27), 229 (M-nBuO,
3
44), 201 (M-nBuOCO, 42), 183 (25), 171 (M-HepS, 100), 159
+
(100), 131 (HepS , 100), 115 (100).
flash silica gel chromatography in 4:1 hexane:ether (R 0.17) to
Anal. Calcd. for C
Found: C, 59.73; H, 8.69; S, 10.77.
H O S: C, 59.56; H, 8.68; S, 10.60.
f
15 26 4
yield the title compound as a yellow oil (90 mg, 44%); ir
-1
1
(CHCl ): 2855 (C-H), 1748 (C=O.O), 1665 (C=O.S) cm ; H
3
4-Benzyl-1-S-pentyl-cis-2,3-epoxysuccin-1-thioate (25u).
NMR (250 MHz, CDCl ): δ 3.76 (3H, s, OMe), 3.76 (1H, d,
3
J=4.8) and 3.72 (1H, d, J=4.8, epoxide), 3.71 (1H, sp, J=6.9,
This compound was isolated as a yellow oil (58%); ir (CHCl ):
2932, 2860 (C-H), 1757 (C=O.O), 1672 (C=O.S) cm ; H NMR
3
13
-1
1
CH(CH ) ), 1.29 (6H, d, J=6.9, CH (CH ) ); C NMR (100.6
3 2
3 2
MHz; CDCl ): δ 193.13 (C=O.S), 165.74 (C=O.O), 58.01 (C-2),
(250MHz, CDCl ): δ 7.34 (5H, m, Ph), 5.19 (2H, s, PhCH ), 3.80
3
3
2
53.67 (C-3), 52.63 (MeO), 35.01 (SCH), 22.63 and 22.51
(1H, d, J=4.7) and 3.76 (1H, d, J=4.7, epoxide), 2.83 (2H, m,
CH S), 1.52 (2H, qu, J=7.2, CH CH S), 1.31 (2H, m, CH CH ),
+
+
(CH(CH ) ); ms: m/z (EI ) 204 (M , 48%), 173 (M-MeO, 31)
3 2
2
2
2
2
3
+
+
+
162 (M-Me C , 18), 129 (M-SiPr, 100), 101 (M-COSiPr, 94).
0.87 (3H, t, J=6.9, CH ); ms: m/z (EI ) 308 (M , 23%), 217 (M-
2
3
+
Anal. Calcd. for C H O S: C, 47.04; H, 5.93; S 15.70. Found:
Bz, 51), 202 (M-PhCHO , 100), 194 (100), 173 (M-BzO C,
8
12
4
2
100), 131 (nPenSCO, 84), 115 (100), 107 (BzO, 100), 103 (100),
91 (Bz , 100).
C, 47.13; H, 5.71; S, 15.51.
+
1-S-(tert-Butyl)-4-methyl-cis-2,3-epoxysuccin-1-thioate (11e).
Anal. Calcd. for C
H O S: C, 62.30; H, 6.55; S, 10.39.
16 20 4
This compound was isolated as a yellow oil (56%); ir (CHCl ):
2900, 2855 (C-H), 1726 (C=O.O), 1658 (C=O.S) cm ; H NMR
Found: C, 62.45; H, 6.68; S, 10.57.
3
-1
1
1,4-Dipentyl-cis-2,3-epoxysuccin-1,4-dithioate (26).
(250 MHz, CDCl ): δ 3.76 (3H, s OMe), 3.69 (2H, s, epoxide),
3
13
1.46 (9H, s, C(CH ) ); C NMR (100.6 MHz, CDCl ): δ 193.02
This compound was prepared according to the method of 11c
starting from diacid 18 and pentane thiol to yield the title com-
pound after purification by flash silica gel chromatography in
3 3
3
(C=O.S), 165.89 (C=O.O), 57.96 (C-2), 53.50 (C-3), 52.57
+
(MeO), 49.30 (C(CH ) ), 29.63 (C(CH ) ); ms: m/z (FAB ) 219
3 3
3 3
(M+1, 73%), 185 (12), 163 (100).
14:1 hexane:ether (R 0.21) as a yellow oil (75%); ir (CHCl ):
2900, 2840 (C-H), 1658 (C=O.S) cm ; H NMR (250MHz,
f
3
-1
1
Anal. Calcd. for C
H O S: C, 49.52; H, 6.48; S, 14.69.
19 14 4
CDCl ): δ 3.84 (2H, s, epoxide), 2.92 (4H, t, J=7.3, 2 x CH S),
Found: C, 49.59; H, 6.39; S, 14.86.
3
2
1.55 (4H, m, 2 x CH CH S), 1.33 (8H, m, 2 x (CH ) ), 0.87 (3H,
2
2
2 2
1-S-Heptyl-4-methyl-cis-2,3-epoxysuccin-1-thioate (11g).
+
+
t, J=7.0, 2 x CH ); ms: m/z (EI ) 304 (M , 45%), 303 (M-1, 15),
201 (M-SnPen, 100), 71 (nPen , 57).
3
+
This compound was isolated as a yellow oil (79%); ir (CHCl ):
2910, 2850 (C-H), 1750 (C=O.O), 1670 (C=O.S) cm ; H NMR
3
-1
1
Anal. Calcd. for C
H O S: C, 55.22; H, 7.96; S, 21.06.
16 20 4
(250 MHz, CDCl ): δ 3.79 (1H, d, J=4.7) and 3.74 (1H, d, J=4.7,
Found: C, 55.56; H, 7.92; S, not determined as high S content
outside range of method used.
3
epoxide), 3.77 (3H, s, OMe), 2.92 (2H, t, J=7.3, CH S), 1.56 (2H,
2
qu, J=7.0, CH CH S), 1.26 (8H, m, (CH ) ), 0.86 (3H, t, J=6.4,
2
2
2 4
1-Benzyl-4-butyl-cis-2,3-epoxysuccinate (28).
13
CH ); C NMR (100.6 MHz, CDCl ): 193.25 (C=O.S), 165.67
3
3
(C=O.O), 58.13 (C-2), 53.71 (C-3), 52.63 (MeO), 31.60, 29.12,
28.74, 28.68, 28.63, 27.74 and 22.52 (7 x CH ), 14.00 (CH ); ms:
This compound was prepared according to the method of 11c
starting from ester-acid 4j and nbutanol to yield the title compound
after purification by flash silica gel chromatography in 14:1
2
3
+
+
m/z (EI ) 261 (M+1, 10%), 260 (M , 4), 229 (M-MeO, 85), 201
(M-MeO C,49), 159 (100), 129 (100), 101 (100).
hexane:ether (R 0.21) as a clear oil (62%); ir (CHCl ): 2945, 2882
2
f
3
-1
1
(C-H), 1750 (C=O.O) cm ; H NMR (250MHz, CDCl ): δ 7.35
(5H, m, Ph), 5.20 (2H, s, PhCH O), 4.06 (3H, t, J=6.8, CH O), 3.72
Anal. Calcd. for C
H O S: C, 55.35; H, 7.76; S, 12.31.
3
12 20 4
Found: C, 55.58; H, 7.87; S, 12.45.
2
2
(1H, d, J=4.6) and 3.68 (1H, d, J=4.6, epoxide), 1.54 (2H, m,
CH CH O), 1.32 (2H, sx, J=7.4, CH CH ), 0.90 (3H, t, J=7.2, CH );
4-Methyl-1-S-pentyl-cis-2,3-epoxysuccin-1-thioate (11r).
2
2
2
3
3
13
This compound was isolated as a yellow oil (75%); ir (CHCl ):
C NMR (100.6 MHz, CDCl ): 165.65 and 165.56 (2 x C=O.O),
3
3
-1
1
2905, 2840 (C-H), 1735 (C=O.O), 1660 (C=O.S) cm ; H NMR
134.70, 128.66 and 128.63 (Ph), 67.62 and 65.87 (2 x CH O), 52.67
2
+
+
(250MHz, CDCl ): δ 3.79 (1H, d, J=4.7) and 3.73 (1H, d, J=4.7,
and 52.55 (epoxide), 30.30, 18.89, 13.61; ms: m/z (EI ) 278 (M ,
100%), 221 (M-Bu , 83), 180 (100), 177 (M-BuO C , 100), 172 (M-
3
+
+
epoxide), 3.76 (3H, s, OMe), 2.91 (2H, t, J=7.2, CH S), 1.56 (2H,
2
2
13
+
+
qu, J=7.1), 1.33 (4H, m, CH CH ), 0.87 (3H, t, J=7.3, CH );
C
PhCHO , 100), 107 (PhCH O, 100), 91 (PhCH , 100).
2
2
3
2
2

Jul-Aug 2005
Preparation of Dicarboxylate Analogues of Cerulenin
829
Omura and T. Hata, J. Biochem., 71, 783 (1972); [e] D. Vance, I.
Goldberg, O. Mitsuhashi, K. Bloch, S. Omura and S. Nomura, Biochim.
Biophys. Res. Commun., 48, 649 (1972); [f] G. D’Agnolo, I. S.
Rosenfeld, J. Awaya, S. Omura and P. R. Vagelos, Biochim. Biophys.
Acta, 326, 155 (1973).
[3] Cerulenin was mentioned by name in the chemical literature
in 6 papers per month on average during 2003, predominantly in bio-
chemical journals.
[4] For example, a selection of some recent non-FAS/PKS studies
are as follows:- [a] Effect on steroid biosynthesis; N. D. Ridgway and
T. A. Lagace, Biochemical Journal, 372, 811 (2003); [b] Effect on
melanin biosynthesis; C. Fleet and C. Breuil, Mycological Research, 106,
1331 (2002); [c] Activation of hypothalamic POMC neurons; I-W. Shu,
D. L. Lindenberg, T. M. Mizuno, J. L. Roberts and C. V. Mobbs, Brain
Research, 985, 1 (2003); [d] Role in apoptosis; S. J. Heiligtag, R.
Bredehorst and K. A. David, Cell Death and Differentiation, 9, 1017
(2002); [e] Potential role in cancer therapy; R. F. Slade, D. A. Hunt, M.
M. Pochet, V. J. Venema and R. A. Hennigar, Anticancer Research, 23,
1235 (2003); [f] Potential treatment for obesity; C. V. Mobbs and H.
Mackimura, Nature Medicine, 8, 335 (2002); J. Hirsch, Proc. Natl. Am.
Soc., 99, 9096 (2002); [g] Potential treatment for reduction in mammalian
hair growth; C. S. Hwang, J. P. Henry, G. S. Ahluwalia and D. Shander,
PCT Int. Appl., WO 2003063810 (2003); Chem. Abstr., 139, 159970
(2003).
Anal. Calcd. for C
64.64; H, 6.46.
H O : C, 64.73; H, 6.53. Found: C,
15 18 5
Hydrogenation of 28 to give 4d.
Benzyl ester (28) (72 mg, 0.26 mmol) was stirred with 10%
Pd/C (10 mg) in THF (2.0 ml) under an atmosphere of H for 75
mins. The reaction mixture was filtered through a pad of Celite,
washed copiously with ether and concentrated to dryness to yield
crude 4d as a clear oil (41 mg, 81%).
2
Unsuccessful Hydrogenations of 25u to give 10r.
Benzyl thioester (25u) (81 mg, 0.26 mmol) was stirred with
10% Pd/C (10 mg) in ether (3.0 ml) under an atmosphere of H
2
for 2.5 hours, after which time, TLC indicated no reaction had
occurred. Further catalyst (70 mg) was added and the reaction
continued for 20 hours. Still no reaction occurred so the recov-
ered 25u (80 mg) was heated to reflux in fresh ether with a large
excess of wood charcoal for 1 hour, before being filtered and con-
centrated to dryness. The concentrate was dissolved in THF (5.0
ml) and stirred under an atmosphere of H with 10% Pd/C (120
mg) for 3 days. TLC and H NMR analysis indicated no reaction
and 25u was recovered quantitatively. Lastly, 25u (40 mg, 0.13
2
1
[5] B. H. Arison and S. Omura, J. Antibiotics, 27, 28 (1974).
[6] H. Yoda, Recent Prog. Chem. Synth. Antibiot. Relat. Microb.
Prod., 939 (1993).
[7] T. E. Kedar, M. W. Miller and L. S. Hegedus, J. Org. Chem.,
61, 6121 (1996).
mmol) was stirred with Pearlman’s catalyst (Pd(OH )/C) (9 mg)
2
in ethanol (2.0 ml) under an atmosphere of H for 6 days. TLC
2
1
and H NMR analysis again indicated no reaction and 25u was
recovered quantitatively.
cis-2,3-Epoxysuccinamic acid (32).
[8] N. S. Mani and C. A. Townsend, J. Org. Chem., 62, 636
(1997).
[9] M. Renard and L. A. Ghosez, Tetrahedron, 57, 2597 (2001).
[10] A. A. Jakubowski, F. S. Guziec, M. Sugiura, C. C. Tam, M.
Tishler and S. Omura, J. Org. Chem., 47, 1221 (1982).
Anhydride 19 (228 mg, 2.0 mmol) and hexamethyldisilazane
(420 µl, 2.0 mmol) were stirred in ether (3.5 ml) for 4 days at RT
(or alternatively heated at reflux in THF for 6 hours). The solu-
tion was concentrated to dryness to yield the crude title com-
pound as an orange gum that solidified on standing. One crys-
tallisation from methanol gave a pale yellow solid (164 mg,
[11a] N. Morisaki, H. Funabashi, J. Furukawa, R. Shimazawa, A.
Kanematsu, T. Ando, S. Okuda and S. Iwasaki, Chem. Pharm. Bull., 40,
2945 (1992); [b] N. Morisaki, H. Funabashi, R. Shimazawa, J. Furukawa,
A. Kawaguchi, S. Okuda and S. Iwasaki, Eur. J. Biochem., 211, 111
(1993); [c] F. Schneider and C. Cassagne, Eur. J. Biochem., 228, 704
(1995); [d] D. S. Lawrence, J. T. Zilfou and C. D. Smith, J. Med. Chem.,
42, 4932 (1999); [e] M. L. De Vos, D. S. Lawrence and C. D. Smith,
Biochemical Pharmacology, 62, 985 (2001); [f] G. P. Royer and C. A.
Townsend, PCT Int. Appl., WO 9519765, (1995); Chem. Abstr., 123,
218396 (1995); [g] J. J. Blumenstein, T. D. Copeland, S. Oroszlan and C.
J. Michejda, Biochem. Biophys. Res. Commun., 163, 980 (1989); [h] T.
Ohno, J. Awaya, T. Kesado, S. Nomura and S. Omura, Antimicrob.
Agents Chemother., 6, 387 (1974).
63%); mp 158-160 °C (dec.); ir (KBr): 3435, 3200 (NH ), 2500-
3150 (O-H), 1735 (C=O.O), 1680, 1630 (C=O.N) cm ; H NMR
2
-1
1
(250MHz, d -DMSO): δ 13.3 (1H, vbs, OH), 7.52 (1H, bs) and
6
7.26 (1H, bs, NH ), 3.75 (1H, d, J=5.1) and 3.63 (1H, d, J=5.1,
2
13
epoxide); C NMR (100.6 MHz, d -DMSO): δ 167.75 and
167.28 (C=O.OH and C=O.NH), 53.63 and 52.81 (epoxide); ms:
m/z (CI ) (%) 130 (M-1, 100%), 113 (M-OH, 88), 112 (M-H O,
6
-
2
64), 86 (14), 81 (38), 79 (39).
Anal. Calcd. for C H NO : C, 36.64; H, 3.85; N, 10.69.
4
5
4
Found: C, 36.26; H, 3.71; N, 10.60.
[12] B. List, A. Golz, W. Boland and H. K. Lichtenthaler, Z.
Naturforsch., C., Biosci., 47, 382 (1992).
[13] H. Funabashi, A. Kawaguchi, H. Tomoda, S. Omura, S.
Okuda and S. Iwasaki, J. Biochem., 105, 751 (1989).
[14] H. Yoda, T. Katagiri and K. Takabe, Tetrahedron Lett., 32,
6771 (1991).
[15] We did not expect the acids to demonstrate strong activity and
recognised they would be prone to decarboxylation, especially in a FAS
or PKS multi-enzyme complex, but decided to isolate samples as they
would be natural precursors to the ester, amide and thioester series.
[16] Lichtenhaler suggested that the activity seen by Tishler for
certain analogues, including 16, may be due to the use of particularly sen-
sitive organisms in his assays. Alternatively, there may be a general
species-specific dependence (i.e. plant vs. bacteria).
Acknowledgements.
We thank the SERC for generous finance (to JDM); John
Gilday (Avlon) and David Harrowven (Southampton) for helpful
comments on the preparation of this manuscript; and Anne Dyke
and Penny Goodland (Avlon) for contributing to the typing of the
experimental section.
REFERENCES AND NOTES
[1a] T. Hata, Y. Sano, A. Matsumae, Y. Kamio, S. Nomura and R.
Sugawara, Nippon Saikingaku Zasshi, 15, 1075 (1960); [b] S. Omura, M.
Katagiri, A. Nakagawa, Y. Sano, S. Nomura and T. Hata, J. Antibiot. Ser.
A., 20, 344 (1967).
[2a] S. Omura, Bacteriol. Rev., 40, 681 (1976); [b] S. Omura,
Methods in Enzymology, 72, 520 (1981); [c] A. Matsumae, S. Nomura
and T. Hata, J. Antibiotics, 25, 365 (1972); [d] S. Nomura, T. Horiuchi, S.
[17] R. Shimazawa, Y. Ogawa, N. Morisaki, H. Funabashi, A.
Kawaguchi and S. Iwasaki, Chem. Pharm. Bull., 40, 2954 (1992).
[18] Except possibly in the case of the thioester-amides 12, for
which we expected the C-4 carbonyl to display partial ketone-like proper-
13
ties (supported by a C NMR shift of ~193 ppm), hence potentially lead-
ing to intramolecular cyclisation.

830
J. D. Moseley and J. Staunton
Vol. 42
[19] G. B. Payne and P. H. Williams, J. Org. Chem., 24, 54 (1959).
[20] S. M. Creighton and D. L. Mitchell, Can. J. Chem., 45, 1304
(1967).
[21] G. G. Allan and A. N. Neogi, Chem. Ind. (London), 545
(1971).
[22a] D. Habich and W. Hartwig, Tetrahedron Lett., 28, 781 (1987);
[b] G. Sabbioni and J. B. Jones, J. Org. Chem., 52, 4565 (1987).
[23] Without giving each individual structure a unique number,
and in an effort to make it easier for the reader to follow the substituents
across multiple series of compounds by retaining the same alphabetical
designator for the generic R groups (see Table 1), compounds 5a and
20a are actually the same (i.e. R, R’ = Me). As 5a was a target structure,
this designation has been used through-out. No other compounds are
affected.
[25] R. Baker and J. L. Castro, J. Chem. Soc., Perkin Trans. I, 47
(1990).
[26] P. N. Rylander, Hydrogenation Methods, Academic Press,
London, 1985.
[27] All data, including micro-analysis and HRMS, were consis-
tent with this structure. The TMS groups were presumably lost by proto-
desilylation on the glassware.
+
[28] The (2S, 3S)-diastereomer of 32 has been isolated as its K
salt:- R. Bihovsky, J. C. Powers, C. M. Kam, R. Walton and R. C. Loewi,
+
J. Enzyme Inhibition, 7, 15 (1993); The K salt of an unspecified
stereoisomer or mixture of 32 has been reported in a Japanese patent, JP
55136205, (1980); Chem. Abstr., 94, 59781 (1981);
[29a] L. E. Coleman, J. F. Bork and H. Dunn, J. Org. Chem., 24, 135
(1959); [b] N. B. Mehta, A. Phillips, F. F. Lui and R. E. Brooks, J. Org.
Chem., 25, 1012 (1960).
[24] E. Boyland and W. H. Down, Eur. J. Cancer, 5, 495 (1971).