Total synthesis of (±) maculalactone A, maculalactone B and maculalactone C and the determination of the absolute configuration of natural (+) maculalactone A by asymmetric synthesis

Article abstract of DOI:10.1016/j.tet.2004.04.060

Maculalactones A, B and C from the marine cyanobacterium Kyrtuthrix maculans are amongst the only compounds based on the tribenzylbutyrolactone skeleton known in nature and (+) maculalactone A from the natural source possesses significant biological activ

Full text of DOI:10.1016/j.tet.2004.04.060

Tetrahedron 60 (2004) 5439–5451
Total synthesis of (6) maculalactone A, maculalactone B and
maculalactone C and the determination of the absolute
configuration of natural (1) maculalactone A by
asymmetric synthesis
b
Geoffrey D. Browna,b and Ho-Fai Wong
,
*
aSchool of Chemistry, The University of Reading, Whiteknights Rd., Reading RG6 6AD, UK
bThe Department of Chemistry, The University of Hong Kong, Pokfulam Rd., Hong Kong, People’s Republic of China
Received 7 November 2003; revised 2 April 2004; accepted 22 April 2004
Abstract—Maculalactones A, B and C from the marine cyanobacterium Kyrtuthrix maculans are amongst the only compounds based on the
tribenzylbutyrolactone skeleton known in nature and (þ) maculalactone A from the natural source possesses significant biological activity
against various marine herbivores and marine settlers. We now report a concise synthesis of racemic maculalactone A in five steps from
inexpensive starting materials. Maculalactones B and C were synthesized by a minor modification to this procedure, and the synthetic design
also permitted an asymmetric synthesis of maculalactone A to be achieved in around 85% ee. The (þ) and (2) enantiomers of maculalactone
A were assigned, respectively, to the S and R configurations on the basis of the chiral selectivity expected for catecholborane reduction of an
unsymmetrical ketone in the presence of Corey’s oxazoborolidine catalyst. Surprisingly, it appeared that natural (þ) maculalactone A was
biosynthesized in K. maculans in a partially racemic form, comprising ca. 90–95% of the (S) enantiomer and 5–10% of its (R) enantiomer.
Coincidentally therefore, the percentage enantiomeric excess of the product obtained from asymmetric synthesis almost exactly matched that
found in nature.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
(3)5 and L,6 as well as the dibenzyldiphenyl-4,5,6,7-
tetrahydrobenzofuranones,5 maculalactones D (4), E, F, G,
H, I, J and K, and the seco-dibenzyldiphenyl-4,5,6,7-
tetrahydrobenzofuranone, maculalactone M,7 as minor
secondary metabolites (Fig. 1). K. maculans is still the
only known natural source for all three classes of secondary
metabolite. Our suspicions that some at least of these
unusual metabolites might be responsible for providing the
chemical defense which K. maculans seemed to enjoy
against other marine organisms had been recently confirmed
by experiments, in which samples of maculalactone A from
the natural source were tested against a generalist marine
herbivore (Chlorostoma argyrostoma) and the larvae of
several barnacle species (Balanus amphitrite, Ibla cumingii
and Tetraclita japonica). These experiments clearly demon-
strated that compound 1 was a potent inhibitor of both
herbivory and barnacle settling.8–10
Kyrtuthrix maculans Umezaki (Brachytrichia maculans
Gomont; Stigonemataceae)1 is a marine cyanobacterium
found growing on moderately exposed rocky shores in Hong
Kong and elsewhere in the world.2,3 We first became
interested in investigating the chemistry of K. maculans
following the observation that this species normally
survived as pure colonies, which were uncontaminated
by other marine bacteria and were also apparently exempt
from settling of barnacles and foraging by marine
herbivores.
Maculalactone A (1),4 a secondary metabolite with the
highly unusual tribenzylbutyrolactone skeleton, was the first
natural product to be reported from a preliminary chemical
investigation of K. maculans made in 1996. Compound 1
was by far the most abundant secondary metabolite to be
found in this species, although—following subsequent more
extensive investigations—we have since described three
further tribenzylbutyrolactones, maculalactones B (2),5 C
Thus, there were now reasonable grounds to suppose that
maculalactone A (1) might have some potential for
development as a novel marine anti-fouling agent;9,10 and
also that the biological activity of some of the other unusual
metabolites from K. maculans might be worthy of further
investigation. Accordingly, we set out to devise a concise
and inexpensive synthesis of compound 1 which would
allow its preparation on a multi-gram scale for future
Keywords: Marine metabolites; Asymmetric synthesis; Grignard
reactions/reagents; NMR.
*
Corresponding author. Tel.: þ44-118-3787418; fax: þ44-118-9316331;
e-mail address: g.d.brown@reading.ac.uk
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.04.060

5440
G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
Figure 1. Maculalactones A, B, C and D (1–4), which have been reported as natural products from Kyrtuthrix maculans (the absolute configuration at the
4-position of compound 1 was not determined in the original report of this secondary metabolite; only relative stereochemistry is shown for compound 4).
evaluation in field trials, when formulated as a paint9,10 (the
extracted yield of maculalactone A from K. maculans is of
the order of 1 mg/g and it is very tedious and time-
consuming to scrape even small amounts of the biological
source material from the rocks on which it grows, making it
almost impossible to obtain sufficiently large amounts of
this compound from the natural source to permit further
biological evaluation). A secondary goal of this project was
to synthesize some of the other tribenzylbutyrolactones
from K. maculans, which have not-as yet-been subjected to
biological evaluation, due to their low natural abundance.
Finally, we also sought to determine the absolute configur-
ation of maculalactone A at the 4-position, an issue which
was not addressed in the original report of this compound as
a natural product from K. maculans.4 The synthetic strategy
described below, which is based on a well known route to
the dibenzylbutyrolactone lignans that are commonly found
in terrestial plants,11 has resulted in the realization of all
three of these goals.
loading of palladium catalyst was reduced in the ethyl
acetate solution, although the rate of reaction also became
slower: a 5% catalyst loading was found to provide the best
compromise for obtaining good yields of 7 (74%) in a
reasonable time (18 h). Fully assigned NMR data for all of
the compounds 5–9 (assigned using the 2D NMR
1
techniques HSQC, HMBC, H–1H COSY and NOESY)
are reported in Table 1; all other NMR data that is reported
in Table 2 was also rigorously assigned by these same 2D
NMR techniques.
The third aromatic substituent in the completed tribenzyl-
butyrolactone skeleton of 1 was introduced by the addition
of a Grignard reagent derived from benzyl bromide (BnBr)
to the symmetrically-substituted anhydride 7. The optimum
conditions for forming the key intermediate, tribenzyl-
substituted compound 10, were somewhat unusual and
considerable experimentation was required in order to arrive
at them. First of all, reflux should be avoided as this leads to
dimerization of the Grignard reagent, forming stilbene (11).
At lower temperatures (0 8C), concentrated solutions of the
anhydride 7 and the ethereal Grignard reagent reacted
rapidly to give the tetra-substituted addition compounds 12
and 13 (Fig. 2) as the favored products, in which 2 equiv. of
Grignard reagent have participated in both 1,2- and 1,4-
additions to the carbonyl group of the anhydride. Approxi-
mately half of the starting material 7 was recovered
unchanged when these reactions were performed under
stoichiometric conditions, which suggested that the desired
compound 10 was perhaps being formed as an intermediate,
but was then preferentially participating in the second
addition of a Grignard reagent to yield the corresponding
tetra-substituted products (this situation is analogous to the
alkylation of an ester by a Grignard reagent, which, as is
well known, commonly results in a di-substituted tertiary
alcohol product).‡
2. Results and discussion
2.1. Synthesis of racemic maculalactone A ((6)-1)
Our synthesis of the tribenzylbutyrolactone skeleton of
maculalactone A (1) (Scheme 1) was based on previously
reported syntheses of the 2,3-aryl di-substituted butyro-
lactone ring system, from the acid anhydride 7.11 2,3-
Dibenzylidenesuccinic acid (5), which was employed as the
starting material for this synthesis, is commercially
available (Section 3.2.1) although this source is rather
costly for a large-scale synthesis of maculalactone A;
alternatively, it can be prepared much more economically
(albeit in low yield) from the Stobbe condensation of
dimethyl succinate with benzaldehyde, as is shown in
Scheme 1.12,13 Treatment of the dicarboxylic acid 5 with
acetyl chloride resulted in cyclization to the symmetrical
acid anhydride 6,14,15 and the conjugated diene functional
group in 6 then underwent preferential 1,4-hydrogenation to
yield the tetra-substituted alkene 7. Compound 8, a minor
side-product from this reaction, in which the diene has been
completely reduced, probably arose via an initial 1,2-
reduction, forming intermediate 9 (Fig. 2).† The ratio of the
desired product 7 to the unwanted product 8 increased as the
The optimum conditions for formation of the desired
addition product, compound 10, required that both reactants
were present as very dilute solutions, and in order to make
compound 10 the dominant product over compound 12, it
‡
The lactol group in the intermediate 10 arising from the first nucleophilic
addition can reversibly interconvert in situ with its ring-opened keto/acid
tautomer; the reactivity of the ketone group in this tautomer towards a
second nucleophilic attack would then be greater than that of the starting
material, acid anhydride 7, thereby favoring a second Grignard addition
reaction.
†
Compound 9 was isolated in substantial amounts when the duration of the
hydrogenation reaction was decreased from 18 to 2 h (see Section 3.2.3).

G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
5441
Scheme 1. Synthesis of maculalactones A–C (1–3) from 2,3-dibenzylidenesuccinic acid (5).
Figure 2. Minor products (8, 9 and 11–16) obtained during the syntheses of maculalactones A–C which are described in Scheme 1.

5442
G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
Table 1. 13C and 1H NMR assignments (CDCl3) for compounds 5–9
dC
dH
Position
Mult.a
5b
6
7
8
9
5b
6
7
8
9
1
2
3
4
C
C
170.4
129.6
143.9
136.5
130.9
129.8
130.8
170.4
129.6
143.9
136.5
130.9
129.8
130.8
166.1
119.5
139.7
134.0
130.1
127.3
130.8
166.1
119.5
139.7
134.0
130.1
127.3
130.8
165.8
143.0
30.3 (CH2)
135.0
129.1
128.8
127.5
165.8
143.0
30.3 (CH2)
135.0
129.1
128.8
127.5
171.7
165.7
123.5
141.1
132.7
130.4
129.5
131.4
171.7
45.6 (CH)
33.7 (CH2)
134.8
129.3
128.7
127.7
—
—
7.91
—
7.35
7.50
7.35
—
—
7.91
—
7.35
7.50
7.35
—
—
7.94
—
6.83
6.85
7.15
—
—
7.94
—
6.83
6.85
7.15
—
—
—
3.50
3.10, 3.12
—
7.14
7.33
7.28
—
3.50
3.10, 3.12
—
7.14
7.33
7.28
—
—
7.80
—
7.58
7.56
7.53
—
45.7 (CH)
32.0 (CH2)
136.3
129.0
128.5
127.4
171.7
45.7 (CH)
32.0 (CH2)
136.3
CH
C
CH
CH
CH
C
C
CH
C
3.79, 3.79
—
7.13
7.28
7.31
—
—
3.79, 3.79
—
7.13
7.28
7.31
5¼9
6¼8
7
1a
2a
3a
4a
5a¼9a
6a¼8a
7a
4.38
3.35, 3.28
—
6.89
7.19
7.19
CH
CH
CH
129.0
128.5
127.4
a
Multiplicity determined from DEPT.
Values recorded in d4-MeOH solution.
b
was necessary to transfer the Grignard reagent to the
solution of anhydride via a canula, rather than to perform the
addition in the reverse sense (as is conventional and was
practiced in the earlier trials). This procedure resulted in
yields of up to 65% of 10 in the crude product (as judged by
1H NMR analysis); its success may be the result of always
maintaining the concentration of benzyl magnesium
bromide in solution at a low level, thereby allowing
compound 10, which is the initial product of the Grignard
reaction, sufficient time to precipitate out of solution before
the lactol functional group undergoes a second Grignard
reaction. The 1,4-adduct, compound 14, consistently
accounted for around 15% of the reaction product under
all conditions that were assayed, and even under optimized
conditions it was still necessary to separate compounds 10
and 14 chromatographically before proceeding to the next
step.
absolute stereochemistry of 1 that are described at the start
of Section 2.3.
2.2. Synthesis of maculalactones B (2) and C (3)
By making a slight modification to the above procedure, it
was found that the key intermediate 10 could be
quantitatively converted into compound 2, which had
identical physical properties to those of the natural product
maculalactone B from K. maculans.5 Since only the
Z-geometric isomer was obtained from the dehydration of
10 in the presence of sulfuric acid16 it appeared that
maculalactone B must therefore be the thermodynamically
more stable product (perhaps as a result of the decreased
steric interaction between the benzyl and benzylidene
substituents at the 3- and 4-positions in this isomer).
However, synthetic maculalactone B (2) could be simply
converted into the thermodynamically less favored
E-isomer, with physical properties identical to those of
maculalactone C (3) from nature,5 by irradiation with UV
light.17 This reaction proceeded in good yield, always
producing a mixture which consisted predominantly of
compound 3 together with some unreacted starting
material (2), that could only be separated by
preparative normal phase HPLC. The ratio of
The final step in the synthesis of maculalactone A, reduction
of the lactol functionality in 10 by sodium borohydride, was
straightforward, resulting in the preparation of compound 1
as a racemic mixture in almost quantitative yield. It is worth
noting that although sodium borohydride was found to be a
very effective reducing agent for converting compound 10
into compound 1, reducing agents derived from neutral
boranes (such as catechol borane) were completely
unreactive. This may be a consequence of the ability of
sodium borohydride to act also as a base, first removing the
proton of the lactol hydroxy group in 10, and thereby
promoting ring opening to a keto/carboxylate tautomer, in
which the carbonyl group of the ketone is more readily
reduced to a secondary alcohol (this alcohol then condenses
with the carboxylic acid group to recyclize the butyro-
lactone ring of 1). The spectral properties (NMR, IR and
MS) of (^)-1 obtained from synthesis were identical with
those of the natural product, (þ) maculalactone A.4 The two
enantiomers of racemic 1 could be separated by HPLC using
a semi-preparative Daicel chiral OD-column (Fig. 3). This
procedure was then routinely automated so as to yield
several 10 s of milligrams of each optical antipode, which
were then used in making the further investigations into the
maculalactone
C
(3) to maculalactone
B
(2)
(approximately 4:1) that was obtained remained quite
constant when UV irradiation was performed in a variety
of solvents (e.g. benzene, dichloromethane or tetrahydro-
furan) and it seems therefore to represent a photochemical
equilibrium state.
An alternative synthesis of racemic maculalactone A from
maculactone B was also devised by effecting hydrogenation
of the diene functional group in maculalactone B (2) over a
palladium catalyst (Scheme 1). This resulted preferentially
in reduction at the tri-substituted double bond of 2, yielding
racemic 1 together with small amounts of the fully reduced
tribenzylbutyrolactone 15 (trace amounts of a second fully
reduced tetrabenzyl-substituted butyrolactone, compound
16 (Fig. 2), which may have arisen from a very minor

1
Table 2. 13C and H NMR assignments (CDCl3) for compounds 10 and 12–17
dc
dH
Position
Mult.a
10
12
13
176.9
14
15
177.3
48.1 (CH)
42.8 (CH)
83.7 (CH)
31.0
138.6
128.5
128.7
126.4
29.6
139.4
128.7
128.6
126.7
37.5
137.4
129.0
128.4
126.6
—
—
—
—
—
16
177.6
50.3 (CH)
51.8
17
10
12
13
14
15
16
17
1
2
3
4
1a
C
C
C
C
CH2
C
171.0
130.4
159.7
106.4
29.1
137.0
128.4
128.6
126.2
32.3
135.9
129.1
128.8
127.1
43.3
133.1
130.5
128.2
127.6
—
172.7
130.2
162.9
90.1
171.8
47.6 (CH)
56.9
173.1
30.0
137.7
129.2
128.9
128.1
39.4
135.1
130.5
128.7
127.2
42.1
167.6
153.1
129.3
204.5
33.9
138.2
128.2
128.7
126.5
37.3
135.9
129.0
128.9
127.1
49.6
133.4
130.0
128.3
126.9
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
52.4
3.16
—
—
3.03, 3.35
—
7.28
7.37
7.35
2.54, 3.42
—
6.78
7.18
7.18
3.11, 3.27
—
3.16
2.96
4.59
2.80, 3.28
—
7.09
7.27
7.19
2.82, 2.94
—
7.02
7.25
7.19
2.55, 2.96
—
7.06
7.25
7.19
—
2.75
—
4.24
2.65, 3.11
—
49.0 (CH)
104.8
41.2
136.1
131.0
128.5
126.8
30.4
139.4
129.7
128.7
128.8
45.8
132.9
130.8
128.0
127.5
40.1
2.48
—
2.97, 3.56
—
7.22
84.7 (CH)
30.7
29.3
3.29, 3.48
—
3.23, 3.23
—
6.25
6.98
7.04
3.68, 3.68
—
6.99
7.23
7.23
2.90, 3.09
—
7.16
7.23
7.23
2.90, 3.09
—
7.16
7.23
7.23
3.86, 3.86
—
2a
137.0
127.8
128.9
125.8
33.2
135.3
128.9
128.2
127.2
42.8
134.5
130.3
128.4
127.2
42.8
134.5
130.3
128.4
127.2
—
139.9
129.0
128.6b
126.5
39.6
135.5
130.8
128.6b
127.1
35.3
137.9
128.9
128.5b
126.8
37.6
137.1
131.5
128.4b
127.2
—
3a¼7a
4a¼6a
5a
1b
2b
3b¼7b
4b¼6b
5b
1c
2c
3c¼7c
4c¼6c
5c
1d
2d
3d¼7d
4d¼6d
5d
CH
CH
CH
CH2
C
CH
CH
CH
CH2
C
CH
CH
CH
CH2
C
CH
CH
CH
CH3
6.69
7.11
7.11
3.65, 3.81
—
7.11
7.26
7.26
3.06, 3.23
—
7.21
7.26
7.14
7.30
7.22
3.70, 3.70
—
7.07
7.28
7.24
3.52, 3.52
—
7.31
c
c
2.91, 3.10
—
7.25
7.35
7.29
2.10, 2.25
—
6.85
7.18
7.23
2.65, 3.26
—
7.07
7.24
c
3.07, 3.07
—
6.84
7.18
c
2.86, 3.08
—
7.23
7.28
c
133.6
130.3
128.9
127.6
—
7.17
7.26
7.35
—
7.22
7.35
7.31
—
7.03
7.26
7.20
—
3.11, 3.11
—
7.45
7.37
7.32
—
—
—
—
—
136.9
130.2
128.5
127.1
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
–OMe
—
—
—
52.4
—
—
—
—
—
3.67
a
Multiplicity determined from DEPT.
Interchangeable within column.
Not assigned.
b
c

5444
G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
Figure 3. Analytical chiral HPLC chromatograms (Chiracel OD–H column) of: a) racemic (^) maculalactone A from synthesis; and b) natural (þ)
maculalactone A from K. maculans.
product carried through the synthesis from the preceding
Grignard reaction, were also isolated).
2.3. Asymmetric synthesis of 1
Having developed two routes to racemic 1—firstly, by
borohydride reduction of the key intermediate 10
(Section 2.1) and, secondly, by hydrogenation of the diene
2 (Section 2.2)—we next attempted to perform an
enantioselective synthesis of the 4R and 4S forms of 1. As
noted in the Introduction, the absolute stereochemistry of
natural maculalactone A had yet to be determined, and the
need to achieve an asymmetric synthesis became particu-
larly pressing following on from the failure of all of the
three most commonly used physical techniques: CD,18
X-ray (Fig. 4) and NMR19–21 (Mosher esterification of the
secondary alcohol obtained from reductive opening of the
butyrolactone ring of 1) to define the absolute stereo-
chemistry of the (þ) and (2) enantiomorphs of 1 which had
been obtained from semi-preparative chiral HPLC of the
synthetic racemate (Section 2.1).
Accordingly, it was decided to modify the strategy for
preparing 1 that has been reported in Sections 2.1 and 2.2 in
order to effect an enantioselective synthesis of macula-
lacatone A. The high yield that had been recorded for the
preparation of maculalactone A (1) from hydrogenation of
maculalactone B (2) (Section 2.2) suggested that it might be
possible to achieve an efficient enantioselective synthesis of
Figure 4. ORTEP diagram of the X-ray structure for synthetic (þ)
maculalctone A (which was separated from its enantiomer by semi-
preparative chiral HPLC). The stereochemistry shown at the 4-position was
maculalactone A by employing a chiral rhodium(I)-DuPhos
suggested, but not conclusively established, by X-ray crystallographic
catalyst22–24 in place of the palladium-on-charcoal hydro-
genation catalyst. In the event, although the overall yield of
analysis (similarly, the absolute stereochemistry could not be definitively
established for synthetic (2) maculalactone A which was also obtained
maculalactone A (1) from the reduction of 2 remained high
from semi-preparative chiral HPLC).

G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
5445
in the presence of this catalyst, this strategy yielded
2.4. The absolute stereochemistry of natural
maculalactone A from K. maculans
maculalactone A with an enantiomeric excess (% ee) very
close to zero. However, it did prove possible to achieve a
much more enantioselective synthesis of 1 by modifying the
synthetic procedure that was described in Section 2.1 in
order to enable reduction by a neutral borane-derivative
rather than by a borohydride reducing agent. As noted in
Section 2.1, the key lactol intermediate 10 was unaffected
by borane reduction and it was therefore necessary to first
convert this compound to its methyl ester derivative,25
compound 17 (Scheme 1). Catecholborane reduction of the
ketone group in this ring-opened substrate, in the presence
of either the R- or S- enantiomer of the chiral oxazo-
borolidine catalyst 18 which has been developed by
Corey,26–31 then resulted in either (þ) or (2) maculalactone
A, respectively, in high chemical yield (85–90%) and with
a reproducibly good-to-high enantiomeric excess (80–85%
ee), as determined by analytical chiral HPLC (see Section
3.2.6 in the Experimental).
Intriguingly, although all the samples of natural macula-
lactone A that we had ever isolated4,8–10 were dextroratory,
none had ever exhibited such a large positive optical
rotation as that recorded from the second fraction which had
been obtained from semi-preparative chiral HPLC of the
synthetic racemate, which should correspond to (þ)-(4S)-1.
Rather, samples of natural maculalactone A collected at
different times from K. maculans growing in different
localities seemed to vary quite substantially in their specific
optical rotation, with an [a]D averaging around þ120–130.
When we subjected samples of natural maculalactone A
isolated from K. maculans to analytical chiral HPLC (as in
Section 3.2.6), the chromatograms consistently demon-
strated that the natural form of 1 was a mixture of two
compounds, with retention times that were consistent with
the contamination of the more predominant (4S) enantiomer
by a little of the (4R) form of the metabolite (see Fig. 3(b))
for a representative chromatogram). These analytical HPLC
studies typically indicated an ‘enantiomeric excess’ for
naturally occurring (þ) maculalactone A of the order of
85–95% ee.
Based on the mechanism which has been proposed for the
enantiomeric induction by the Corey catalyst (Fig. 5),32 it is
possible to infer that the (þ) form of 1-which was obtained
when using the R- enantiomer of the oxazoborolidine
catalyst (R)-18—should correspond to the (4S) enantiomer
of maculalactone A, and that the (2)-form of maculalactone
A must therefore be the (4R) enantiomer. Hence, it appeared
that the first fraction obtained from chiral HPLC in Figure 3,
i.e. (2) maculalactone A with [a]D¼2156.7 [c 1.0], was
the enantiomeric form of 1 with the R configuration at the
4-position, while the second fraction, (þ) maculalactone A
([a]D¼þ153.1 [c 1.4]), corresponded to the 4S absolute
configuration.
Final confirmation that natural maculalactone A was, in
fact, a partially racemic mixture of the two enantiomorphs
of 1, was obtained by performing 1H NMR experiments with
maculalactone A isolated from K. maculans in the presence
of a chiral shift reagent (europium tris[3-heptafluoropropyl-
hydroxymethylene]-(þ)-camphorate). The effects of
making successive additions of this shift reagent to a
CDCl3 solution of maculalactone A (1) from the natural
Figure 5. (a) The transition state assembly of the chiral oxazoborolidine (R)-18 with a generalized prochiral ketone, and the resulting absolute configuration of
the secondary alcohol product from catecholborane reduction, that has been proposed in the literature to account for the enantioselectivity of the Corey
catalyst.32 (b) The predicted absolute configuration of (þ) maculalactone A (1) from the enantioselective reduction of the ketone 17 in the presence of (R)-18,
based on this proposed mechanism.

5446
G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
Figure 6. Expanded 1H NMR spectra of natural (þ)-maculalactone A from K. maculans (expts. 1–7); and racemic (^) maculalactone A from synthesis
(Section 2.1; expts. 8–14) following successive additions (10 mg) of a europium tris[3-(heptafluoropropyl-hydroxymethylene)-(þ)-camphorate] chiral shift
reagent.
1
source are shown in the upper seven H NMR spectra in
Figure 6 (experiments 1–7). Several peaks (most obviously
the H-4 proton at dH 4.94 ppm)§ appeared to resolve into
two components in an approximately 9:1 ratio in this
experiment (this is particularly evident in the 1H NMR
spectrum shown in entry 7 of Figure 6, which contained the
largest amount of added chiral shift reagent). As demon-
strated by experiments 8–14 in Figure 6, the H-4 proton of
synthetic (^)-1 (Section 2.1) was also observed to separate
into two components, in the same manner as for experiments
1–7, although in this case equal intensities were observed
for the two sets of peaks due to the two enantiomers of 1, as
the sample was, of course, racemic. The inescapable
conclusion is therefore that natural samples of (þ)-1 are,
in fact, partially racemic at the 4-position.
entirely enzymic) or as completely racemic mixtures (if
non-enzymic chemical processes are also involved in their
biogenesis). It is rare to find intermediate cases, involving
partially racemic metabolites, in natural product chemistry.
Although this phenomenon has been reported on a few
occasions for terpenoids (e.g. (þ)-a-pinene, which is very
often found as an admixture with lesser amounts of its
(2)-enantiomer),34 it seems to be extremely rare for
lignans,35 to which biogenetic class maculalactone A
presumably belongs.4,5 We were therefore rather concerned
that the apparent partial racemization of 1 might be an
artifact of the extraction and chromatographic procedures
which were employed in the isolation of maculalactone A
from K. maculans, especially as it is quite easy to propose a
mechanism which would account for such racemization: if a
trace of acid (or base) were present at either or both of the
extraction and purification stages for the isolation of
maculalactone A, then an extended enol could be generated
by tautomerization, involving a movement of the H-4 proton
to the carbonyl group; reprotonation from either side of the
resulting 2-hydroxyfuran p-system would then result in
racemization.
To summarize, the results from measurements of optical
1
rotation, chiral HPLC and H NMR in the presence of a
chiral shift reagent all confirmed that natural maculalactone
A (1) was present in K. maculans as a mixture of
approximately 85–95% of the (4S) enantiomer with
5–15% of the (4R) enantiomer. This is a very unusual
situation, as most natural products are normally isolated
either in enantiomerically pure form (if their biosynthesis is
In order to test this possibility, the length of the extraction
procedure that was routinely used for obtaining macula-
lactone A from K. maculanswas varied from 20 h to 10 days.
The effect of including a second column chromatographic
step in the purification of 1 was also investigated. However,
neither procedure appeared to affect the % ee of natural 1, as
determined by chiral analytical HPLC, to any significant
extent. Additionally, when a sample of partially racemic 1
(from synthesis) was dissolved in organic solvent (1 mg/
10 mL EtOAc) and shaken with HCl (1 M, 10 mL) for
15 min-in an attempt to induce racemization via the
proposed enolic form of 1-there was, once again, no
§
The shifts induced by the chiral lanthanide for the proton resonances at H-
1a (dH 3.56 and 3.64 ppm) in (^)-1 showed the most significant changes
in chemical shift as the amount of chiral shift reagent added to the CDCl3
solution was gradually increased. These two protons are closest to the
carbonyl functional group in 1, suggesting that it is the non-bonding pairs
of electrons in the carbonyl group which are donated to the europium ion
of the shift reagent,33 thereby producing the greatest downfield shifts at
the 1a-position. The proton resonance at the 4-position (dH 4.94 ppm)
exhibited the largest separation in chemical shifts between the 4R- and
4S- forms of 1, as the amount of chiral shift reagent was increased,
perhaps because it is this proton which is directly attached to the chiral
center.

G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
5447
significant effect. It therefore appeared that the variation in
the % ee found for (þ)-1 from the natural source was not an
artifact that had been introduced by either the extraction or
purification procedures. Hence, we are forced to conclude
that maculalactone A is indeed naturally present in
K. maculans in a partially racemized form.
Et2O (3£20 mL). The combined organic layers were dried
(MgSO4) and rotary evaporated to give a residue consisting
predominantly of compound 5, which was purified by
washing with Et2O.
2E,3E-Dibenzylidenesuccinic acid (5). Solid, mp 213–
215 8C, (1.87 g, 6.4 mmol, 19%). nmax/cm21 (KBr disc):
3435 (v br), 3069, 1680, 1638, 1611; dH (CD3OD)-7.91 (2H,
s), 7.50 (4H, m), 7.35 (6H, m)—see also Table 1; 13C
NMR—see Table 1; ESI-MS m/z (rel. int.): 317 (Mþ
[C18H14O4]þNaþ) (100).
In closing, it is worth pointing out that-quite by chance-the
ratio of the two enantiomers found for natural macula-
lactone A (1) was almost identical with that which was
obtained from the asymmetric synthesis described in
Section 2.3. Perhaps uniquely in natural product chemistry,
it appears that the asymmetric synthesis of 1 via the Corey
procedure has yielded a target compound of similar
enantiomeric composition to that of the natural material,
and that even though the enantioselectivity of this synthetic
procedure is perhaps relatively modest by the standards of
the current day-at 85% ee-there is little incentive to attempt
to improve the chiral selectivity of the enantiomeric step, as
the product obtained from asymmetric synthesis is already
almost identical with the natural product!
Note that compound 5 was also commercially available
from the Aldrich Chemical Company (Cat. No. S-300187)
but that this source was judged to be too costly as a starting
material for performing a synthesis of maculalactone A on a
gram scale, such as that described in Sections 3.2.1–3.2.5.
3.2.2. Conversion of 5 to 2E,3E-dibenzylidenesuccinic
anhydride (6). To 2E,3E-dibenzylidene succinic acid (5)
(1.5 g, 5.1 mmol) was added excess acetyl chloride
(1.5 mL) and the reaction was refluxed for 2 h. After
completion, as determined by TLC, the reaction mixture
was cooled in an ice bath and the anhydride 6 was separated
by filtration as yellow crystals, which were washed with a
small amount of cold Et2O to yield the pure product.
3. Experimental
3.1. General methods
2E,3E-Dibenzylidenesuccinic anhydride (6). Solid, mp
202–204 8C (1.18 g, 4.3 mmol, 84%). nmax/cm21: 3030,
1823, 1771, 1618, 1595; dH-7.94 (2H, s), 7.15 (2H, tt,
J¼7.1, 1.6 Hz), 6.85 (4H, dd, J¼7.1, 7.1 Hz), 6.83 (4H, dd,
J¼7.1, 1.6 Hz)—see also Table 1; 13C NMR—see Table 1;
HREIMS m/z (rel. int.): 276.0786 (Mþ, calcd 276.0786 for
C18H12O3) (60), 232 (40), 203 (80), 199 (100).
General experimental procedures were similar to those
described recently in: Brown, G.D.; Sy, L.-K., Tetrahedron
2004, 60, 1125–1138.
3.1.1. Isolation of natural (1) maculalactone A (2,3,4-
tribenzyl-g-butyro-2,3-en-lactone) (1) from Kyrtuthrix
maculans. K. maculans (132.8 g wet wt., 78.6 g after freeze
drying overnight) was collected from the shores at Shek Mei
Tau, New Territories, Hong Kong in March. Taxonomic
identification was made by Dr Williams of the Department
of Ecology and Biodiversity, The University of Hong Kong.
Following lyophilization, the sample was pulverized to a
fine powder under liq. N2, repeatedly extracted with
CH2Cl2, dried (MgSO4) and solvent was removed under
reduced pressure to yield a dark brown gum (1.09 g; 0.82
w/w % [fresh wt]). The extract was subjected to gradient CC
(n-hexane!EtOAc) followed by HPLC to obtain compound
1 (115.2 mg, 0.09% w/w [fresh wt.]; Rf 0.35 in 5% EtOAc/
n-hexane) which was recrystallized from 10% i-PrOH/n-
hexane, to obtain maculalactone A (1) as a solid, mp
88–89 8C; [a]D¼þ130.2 (c¼2.4, CHCl3)-see Ref. 4 for
other physical properties.
3.2.3. Reduction of 6 to 2,3-dibenzylbut-2-en-dioic
anhydride (7) and 2R/S,3S/R-dibenzylbutanedioic
anhydride (8). A solution of 2E,3E-dibenzylidene succinic
anhydride (6) (1.0 g, 3.62 mmol) in EtOAc (10 mL) was
charged with 5% (w/w) Pd/C (50 mg) and stirred under H2
(1 atm) at room temperature for 18 h. The resultant colorless
solution was filtered and rotary evaporated to give a crude
product consisting of the anhydrides 7 and 8 in an
approximately 3:1 ratio (as determined by 1H NMR
spectroscopy). Anhydride 8 precipitated out of solution
after thoroughly mixing the crude product with 10% EtOAc/
n-hexane, while compound 7 remained largely in the
supernatant and was further purified by semi-preparative
HPLC (10% EtOAc/n-hexane).
2,3-Dibenzylbut-2-en-dioic anhydride (7). Oil (Rt 19.1 min,
756 mg, 2.71 mmol, 74%). nmax/cm21: 3032, 2928, 1825,
1767, 1603, 1497, 1456; dH 7.31 (2H, t, J¼7.5 Hz), 7.28
(4H, dd, J¼7.5, 7.5 Hz), 7.13 (4H, d, J¼7.5 Hz) 3.79 (4H,
s)—see also Table 1; 13C NMR—see Table 1; HREIMS m/z
(rel. int.): 278.0948 (Mþ, calcd 278.0943 for C18H14O3)
(100), 260 (14), 233 (49), 205 (21).
3.2. Synthesis of (6) maculalactone A
3.2.1. Preparation of 2E,3E-Dibenzylidenesuccinic acid
(5). A solution of dimethyl succinate (5.0 g, 34.2 mmol),
benzaldehyde (7.3 g, 68.4 mmol) and Na (1.7 g, 75.3 mmol)
in anhyd. Et2O (50 mL) was stirred under N2 at 0 8C and
allowed to warm to room temperature over 12 h. MeOH
(10 mL) was added to destroy any remaining excess Na,
followed by H2O (10 mL) and the mixture was then
acidified with HCl (6 M, 10 mL), and extracted with Et2O
(2£10 mL). The combined organic layers were re-extracted
with sat. NaHCO3 (aq.) (3£20 mL) and the combined sat.
NaHCO3 (aq.) extracts were acidified and extracted with
2(R/S),3-Dibenzylbutanedioic anhydride (8). Solid, mp
105–1068 C (Rt 30.0 min, 254 mg, 0.95 mmol, 24%).
n
max/cm21: 3032, 2926, 1863, 1786, 1605, 1496, 1456;
dH-7.33 (4H, dd, J¼7.3, 7.3 Hz), 7.28 (2H, t, J¼7.3 Hz),
7.14 (4H, d, J¼7.3 Hz), 3.50 (2H, m), 3.12 (2H, m), 3.10

5448
G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
(2H, m)—see also Table 1; 13C NMR—see Table 1;
HREIMS m/z (rel. int.): 280.1102 (Mþ, calcd 208.1099 for
C18H16O3) (70), 125 (100).
(13). Solid, mp 196–198 8C (7 mg, 0.016 mmol, 1%; Rf
0.31). nmax/cm21: 3649, 3034, 2926, 1749, 1601, 1497,
1454; dH-7.35 (2H, dd, J¼7.5, 7.5 Hz), 7.31 (2H, m), 7.29–
7.14 (12H, m), 7.07 (2H, d, J¼6.9 Hz), 6.85 (2H, dd, J¼6.6,
1.6 Hz), 3.56 (1H, d, J¼13.7 Hz), 3.26 (1H, d, J¼13.9 Hz),
3.10 (1H, dd, J¼13.9, 11.0 Hz), 3.02 (1H, s,–OH), 2.97
(1H, d, J¼13.7 Hz), 2.91 (1H, dd, J¼13.9, 4.2 Hz), 2.65
(1H, d, J¼13.9 Hz), 2.48 (1H, dd, J¼11.0, 4.2 Hz), 2.25
(1H, d, J¼13.9 Hz), 2.10 (1H, d, J¼13.9 Hz)—see also
Table 2; 13C NMR—see Table 2; HREIMS m/z (rel. int.):
444.2092 (Mþ2H2O, calcd 444.2089 for C32H28O2) (45),
353 (100).
When incomplete reduction of 2E,3E-dibenzylidene-
succinic anhydride (6) (20.0 mg, 0.07 mmol) was performed
in an EtOAc (1 mL) solution charged with 10% Pd/C (1 mg,
5% w/w) and stirred under H2 (1 atm) at room temperature
for 2 h, the resulting crude product consisted of the three
anhydrides 7–9, which were separated by semi-preparative
HPLC (10% EtOAc/n-hexane): 7 (4.1 mg, 0.015 mmol,
20%; Rt 19.1 min); 8 (3.3 mg, 0.012 mmol, 16%; Rt
30.0 min); and 9 (5.5 mg, 0.020 mmol, 27%; Rt 32.3 min).
2,2,3(R/S)-Tribenzylbutanedioic anhydride (14). Solid, mp
145–147 8C (85 mg, 0.23 mmol, 13%; Rf 0.49). nmax/cm21
:
2-Benzylidene,3(R/S)-benzyl-butanedioic anhydride (9).
Solid, mp 138–140 8C. nmax/cm21: 3030, 2930, 1771,
1709, 1636, 1603, 1496, 1454, 1447 cm21; dH-7.80 (1H, d,
J¼2.4 Hz), 7.58–7.53 (5H, m), 7.19 (3H, m), 6.89 (2H, m),
4.38 (1H, ddd, J¼5.7, 4.1, 2.4 Hz), 3.35 (1H, dd, J¼13.7,
5.7 Hz), 3.25 (1H, dd, J¼13.7, 4.1 Hz)—see also Table 1;
13C NMR—see Table 1; HREIMS m/z (rel. int.): 278.0944
(Mþ, calcd 278.0943 for C18H14O3) (50), 233 (20), 205 (25),
178 (15), 153 (100).
3028, 2928, 2856, 1782, 1718, 1497, 1454; dH-7.37 (2H, dd,
J¼7.5, 7.5 Hz), 7.35 (3H, m) 7.31 (1H, t, J¼7.5 Hz), 7.28
(2H, d, J¼7.5 Hz), 7.22 (2H, d, J¼7.5 Hz), 7.18 (3H, m),
6.78 (2H, dd, J¼7.5, 1.8 Hz), 3.42 (1H, d, J¼14.4 Hz), 3.35
(1H, dd, J¼14.2, 6.6 Hz), 3.27 (1H, d, J¼13.9 Hz), 3.16
(1H, t, J¼6.4 Hz), 3.11 (1H, d, J¼13.9 Hz), 3.03 (1H, dd,
J¼14.2, 6.6 Hz), 2.54 (1H, d, J¼14.4 Hz)—see also Table 2;
13C NMR—see Table 2; HREIMS m/z (rel. int.): 370.1572
(Mþ, calcd 370.1569 for C25H22O3) (25), 352 (8), 279 (44),
251 (16), 222 (100).
3.2.4. Synthesis of 2,3,4-tribenzyl-4(R/S)-hydroxy-g-
butyro-2-en-lactone (10). A Grignard reagent, freshly
prepared from the addition of Mg (52.4 mg, 2.16 mmol) to
benzyl bromide (0.26 mL, 2.16 mmol) in anhyd. Et2O
(50 mL) at room temperature for 1 h, was transferred to a
solution of anhydride 7 (500 mg, 1.80 mmol) in anhyd. Et2O
(50 mL) at 0 8C via a cannula. The reaction mixture was
stirred under N2 at 0 8C for 30 min, quenched by H2O
(15 mL) and then acidified with HCl (1 M, 15 mL). The
aqueous layer was extracted with Et2O (2£50 mL) and the
combined organic layers were washed with brine (50 mL),
dried (MgSO4) and rotary evaporated to give a crude
product consisting predominantly of the butyrolactone 10
together with small amounts of 11–14, which was purified
by CC (10% EtOAc/n-hexane).
3.2.5. Sodium borohydride reduction of 10 to racemic
maculalactone A (4(R/S)-2,3,4-tribenzyl-g-butyro-2-en-
lactone) (6)-(1). To a solution of the g-hydroxy-butyro-
lactone 10 (250 mg, 0.68 mmol) in THF/H2O (10 mL; 24:1)
was added NaBH4 (128 mg, 3.37 mmol) in portions at 0 8C
with stirring for 2 h. The reaction mixture was quenched by
HCl (1 M, 10 mL), concentrated by rotary evaporation and
extracted with EtOAc (3£10 mL). The combined organic
layers were dried (MgSO4) and rotary evaporated to yield
the butyrolactone 1 without the need for further purification.
4(R/S)-2,3,4-Tribenzyl-g-butyro-2-en-lactone (1). Solid, mp
88–89 8C (237 mg, 0.67 mmol, 99%). nmax/cm21: 3030,
3013, 2926, 1751, 1668, 1603, 1497, 1454; dH-7.30 (2H, m),
7.27 (1H, m), 7.25 (3H, m), 7.17 (1H, m), 7.15 (4H, m), 7.03
(2H, dd, J¼6.6, 1.9 Hz), 6.88 (2H, dd, J¼6.1, 2.3 Hz), 4.94
(1H, dd, J¼6.1, 4.0 Hz), 3.92 (1H, d, J¼15.6 Hz), 3.64 (1H,
d, J¼15.4 Hz), 3.56 (1H, d, J¼15.4 Hz), 3.47 (1H, d,
J¼15.6 Hz), 3.23 (1H, dd, J¼14.6, 4.0 Hz), 2.81 (1H, dd,
J¼14.6, 6.1 Hz)—see also Ref. 4; dC2173.5 C, 161.7 C,
137.7 C, 135.9 C, 134.8 C, 129.5 CH£2, 129.1 CH£2, 128.7
CH£2, 128.6 CH£4, 128.6 C, 128.2 CH£2, 127.3 CH, 127.2
CH, 126.4 CH, 81.6 CH, 38.0 CH2, 33.2 CH2, 29.4 CH2—
see also Ref. 4; HREIMS m/z (rel. int.): 354.1616 (Mþ,
calcd 354.1620 for C25H22O2) (60), 263 (100).
2,3,4-Tribenzyl-4(R/S)-hydroxy-g-butyro-2-en-lactone (10).
Solid, mp 107–109 8C (394 mg, 1.06 mmol, 59%; Rf 0.20).
n
max/cm21: 3527, 3392 (br), 3026, 2928, 1759, 1602, 1497,
1454; dH-7.35 (1H, m), 7.26 (5H, m), 7.17 (2H, m), 7.11
(5H, m), 6.69 (2H, dd, J¼7.1, 1.8 Hz), 3.81 (1H, d,
J¼14.7 Hz), 3.65 (1H, d, J¼14.7 Hz), 3.48 (1H, d,
J¼15.3 Hz), 3.29 (1H, d, J¼15.3 Hz), 3.26 (s, –OH), 3.23
(1H, d, J¼13.9 Hz), 3.06 (1H, d, J¼13.9 Hz)—see also
Table 2; 13C NMR—see Table 2; HREIMS m/z (rel. int.):
352.1461 (Mþ2H2O, calcd 352.1463 for C25H20O2) (100),
279 (55).
2,3,4,4-Tetrabenzyl-g-butyro-2-en-lactone (12). Solid, mp
135–137 8C (19 mg, 0.043 mmol, 3%; Rf 0.40). nmax/cm21
3.2.6. Separation of racemic (6) maculalactone A
(4(R/S)-2,3,4-tribenzyl-g-butyro-2-en-lactone) (1) into
its (1) and (2) enantiomers. Semi-preparative chiral
HPLC separation of a sample of (^)-1 (60 mg) was
performed using a Waters chromatograph equipped with
:
3030, 3013, 2926, 1747, 1496, 1454; dH-7.23 (9H, m), 7.16
(4H, dd, J¼7.5, 1.8 Hz), 7.04 (1H, t, J¼7.0 Hz), 6.99 (2H, d,
J¼7.0 Hz), 6.98 (2H, dd, J¼7.0, 7.0 Hz), 6.25 (2H, d,
J¼7.0 Hz), 3.68 (2H, s), 3.23 (2H, s), 3.09 (2H, d,
J¼14.3 Hz), 2.90 (2H, d, J¼14.3 Hz)—see also Table 2;
13C NMR—see Table 2; HREIMS m/z (rel. int.): 444.2089
(Mþ, calcd 444.2089 for C32H28O2) (30), 353 (100).
RI 410 detector and
a Daicel chiral OD-column
(10 mm£25 cm; 10 mm particle size; Cat. no. 14045; Chiral
Technologies Inc., Exton, PA, USA) operating isocratically
.
with 10% i-PrOH/ n-hexane at a flow rate of 8 mL min21
Typical recoveries per injection (1 mg of (^)-1 in 100 mL)
were ca. 0.35 mg of each enantiomer, yielding overall:
2,2,3(R/S),4(R/S)-Tetrabenzyl-4-hydroxy-g-butyrolactone

G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
5449
(2)-2,3,4-tribenzyl-g-butyro-2-en-lactone ((2)-1) (15 mg,
Oil (15.8 mg, 0.045 mol, 79%; Rt 18.7 min). nmax/cm21
:
25%; Rt¼16.9 min; [a]D¼2156.7 (c¼1.0, CHCl3); and
(þ)-2,3,4-tribenzyl-g-butyro-2-en-lactone ((þ)-1) (12 mg,
20%, Rt¼18.3 min; [a]D¼þ153.1 (c¼1.2, CHCl3).
3032, 3013, 2928, 1756, 1603, 1494, 1454; dH-7.24–7.10
(11H, m), 7.01 (2H, d, J¼7.3 Hz), 6.84 (1H, s), 6.60 (2H,
dd, J¼7.8, 1.7 Hz), 3.69 (2H, s), 3.66 (2H, s)—see also Ref.
5; dC2169.7 C, 149.3 C, 148.3 C, 137.2 C, 136.2 C, 133.0
C, 132.6 C, 129.3 CH£2, 128.7 CH£4, 128.4 CH£2, 128.2
CH£2, 128.2 CH, 127.7 CH£2, 126.7 CH, 126.4 CH, 115.3
CH, 31.5 CH2, 29.8 CH2—see also Ref. 5; HREIMS m/z
(rel. int.): 352.1465 (Mþ, calcd 352.1463 for C25H20O2)
(100), 261 (17).
The analytical chiral separation shown in Figure 3 was made
using a chiral stationary phase with comparable properties
(Chiracel OD-H column, 4.6 mm£25 cm; 5 mm particle
size; Cat. no. 14325; Chiral Technologies Inc., Exton, PA,
USA). Experiments were performed with either an analyti-
cal HP 1100 series chromatogram or with an HP 2200
HPLC instrument equipped with HP 2210 UV detector
(l¼245 nm), using a 20 mL injection loop (sample
concentration typically 1 mg/mL) and operating isocrati-
cally with 10% i-PrOH/n-hexane at a flow rate of
1 mL min21. Under these conditions, the Rt for (2)-1 was
23.7 min and that for (þ)-1 was 26.6 min.
3.3.3. Hydrogenation of maculalactone B (2) to macula-
lactone A (1). A solution of the butyrolactone 2 (30 mg,
0.09 mmol) in EtOAc (3 mL) was charged with 10% Pd/C
(8 mg) and stirred under H2 (1 atm) at room temperature for
18 h. The resultant mixture was filtered and rotary
evaporated to give a crude product consisting predomi-
nantly of the butyrolactone 1, which was purified from the
alternative hydrogenation products 15 and 16 by preparative
HPLC (10% EtOAc/n-hexane).
3.3. Syntheses of other tribenzylbutyrolactone natural
products
3.3.1. Dehydration of 10 to maculalactone B (2,3-dibenzyl-
4Z-benzylidene-g-butyro-2-en-lactone) (2). To a solution
of the g-hydroxy-butyrolactone 10 (100 mg, 0.27 mmol) in
toluene (4 mL) was added a powder of silica gel onto which
had been adsorbed H2SO4 (30 mg; prepared by suspending
silica [100 g] in acetone [100 mL] containing H2SO4 [conc.;
10 g] and removing the solvent on a rotary evaporator). The
mixture was heated at reflux for 5 h, and after completion of
the reaction, as determined by TLC, the mixture was filtered
and the solvent was removed on a rotary evaporator to yield
the butyrolactone 2 without the need for further purification.
2,3,4(R/S)-Tribenzyl-g-butyro-2-en-lactone (1). Solid mp
88–89 8C (23.7 mg, 0.07 mmol, 79%, Rt 31.4 min)-see
Section 3.2.5 for physical properties.
2(R/S),3(S/R),4(R/S)-Tribenzyl-g-butyrolactone (15). Solid,
mp 145–147 8C (3.7 mg, 0.01 mmol, 12%, Rt 26.9 min).
n
max/cm21: 3028, 2957, 1767, 1603, 1497, 1454; dH-7.27
(2H, m), 7.25 (4H, m), 7.19 (3H, m), 7.09 (2H, d,
J¼7.3 Hz), 7.06 (2H, d, J¼7.0 Hz), 7.02 (2H, d,
J¼7.3 Hz), 4.59 (1H, ddd, J¼9.6, 4.0, 3.9 Hz), 3.28 (1H,
dd, J¼15.0, 4.8 Hz), 3.16 (1H, ddd, J¼10.7, 6.4, 4.6 Hz),
2.96 (2H, m), 2.94 (1H, m), 2.82 (1H, dd, J¼15.5, 4.8 Hz),
2.80 (1H, dd, J¼15.0, 6.4 Hz), 2.55 (1H, dd, J¼14.6,
3.9 Hz)—see also Table 2; 13C NMR—see Table 2;
HREIMS m/z (rel. int.): 356.1779 (Mþ, calcd 356.1776
for C25H24O2) (15), 265 (25), 247 (30), 229 (18), 219 (20),
208 (100).
2,3-Dibenzyl-4Z-benzylidene-g-butyro-2-en-lactone (2).
Solid, mp 114–115 8C (78 mg, 0.21 mmol, 82%). nmax
/
cm21: 3030, 3015, 2920, 1755, 1651, 1603, 1494, 1454;
dH-7.70 (2H, d, J¼7.5 Hz), 7.33 (2H, dd, J¼7.5, 7.5 Hz),
7.28–7.20 (7H, m), 7.17 (2H, d, J¼7.1 Hz), 7.10 (2H, d,
J¼7.4 Hz), 5.96 (1H, s), 3.92 (2H, s), 3.73 (2H, s)—see also
Ref. 5; dC2170.3 C, 150.8 C, 148.2 C, 137.4 C, 136.6 C,
133.0 C, 130.5 CH£2, 128.9 CH£2, 128.8 CH, 128.7 CH£2,
128.6 CH£4, 128.2 CH£2, 127.8 C, 127.0 CH, 126.7 CH,
110.5 CH, 30.6 CH2, 29.8 CH2—see also Ref. 5; HREIMS
m/z (rel. int.): 352.1457 (Mþ, calcd 352.1463 for C25H20O2)
(100), 261 (15).
2(R/S),3,3,4(R/S)-Tetrabenzyl-g-butyrolactone (16). Solid,
mp 240–242 8C (0.5 mg, 0.001 mmol, 1%; Rt 24.8 min).
n
max/cm21: 3026, 2926, 1773, 1497, 1454; dH-7.45 (2H, d,
J¼7.5 Hz), 7.37 (2H, dd, J¼7.5 Hz), 7.32 (1H, t, J¼7.5 Hz),
7.30–7.15 (13H, m), 6.84 (2H, m), 4.24 (1H, dd, J¼10.7,
1.8 Hz), 3.11 (1H, dd, J¼14.9, 10.0 Hz), 3.11 (2H, s), 3.08
(1H, m), 3.07 (2H, s), 2.86 (1H, dd, J¼13.5, 1.8 Hz), 2.75
(1H, dd, J¼10.0, 4.1 Hz), 2.65 (1H, dd, J¼14.9, 4.1 Hz)—
see also Table 2; 13C NMR—see Table 2; HREIMS m/z (rel.
int.): 446.2245 (Mþ, calcd 446.2246 for C32H30O2) (20),
355 (30), 337 (50), 319 (10), 309 (15), 298 (35), 207 (100).
3.3.2. Photo-isomerization of maculalactone B (2) to
maculalactone C (2,3-dibenzyl-4E-benzylidene-g-
butyro-2-en-lactone) (3). A degassed solution of the
butyrolactone 2 (20 mg, 0.06 mmol) in anhyd. C6H6
(10 mL) was irradiated by UV light (450 W medium-
pressure Hg arc lamp) at room temperature under N2 for 1 h.
Solvent was removed by rotary evaporation to give a
mixture containing the butyrolactones 3 and 2, in an
approximately 4:1 ratio (by 1H NMR of the crude product)
which were separated by preparative HPLC (10% EtOAc/n-
hexane).
3.4. Asymmetric synthesis of (1) maculalactone A
((1)-1)
3.4.1. Attempted enantioselective hydrogenation of the
trisubstituted olefin group in maculalactone B (2) in the
presence of (1)-1,2-bis[(2S,5S)-2,5-diethylphospholano]-
benzene-(cyclooctadiene)rhodium (I) trifluoromethane-
sulfonate (Et-DuPhos) catalyst. A 25 mL Schlenk tube
was charged with compound 2 (25 mg, 0.071 mmol) in
anhyd. MeOH (5 mL), and the solution was degassed by
three freeze-thaw cycles under an atmosphere of Ar. The
solution was then transferred to another 25 mL Schlenk
2,3-Dibenzyl-4Z-benzylidene-g-butyro-2-en-lactone (2).
(4.0 mg, 0.011 mol, 20%; Rt 19.8 min). See Section 3.3.1
for physical properties.
2,3-Dibenzyl-4E-benzylidene-g-butyro-2-en-lactone (3).

5450
G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
tube, which was pre-charged with the catalyst Rh(I)-(S,S)-
Et-DuPhos (0.05 mg, 0.1 mol %; Strem chemical company,
Cat. no. 45–0151), via a cannula and further degassed by
two more freeze-thaw cycles. H2 (1 atm) was introduced to
the system and the reaction mixture was allowed to stir at
room temperature for 3 days, after which complete
conversion to product was indicated by TLC. The resulting
mixture was concentrated to yield an extract consisting
mostly of product 1 (22.5 mg, 0.063 mmol, 89%; Rf 0.22),
following purification by CC (10% EtOAc/n-hexane). The
enantiomeric excess (% ee) was determined as 0.5 by
measurement of the optical rotation ([a]D¼þ0.70 [c¼2.2,
CHCl3]), assuming [a]D (þ)-1¼þ153.1 and [a]D (2)-
1¼2153.1.
0.01 mmol, 0.5 M in toluene) and compound 17 (19.2 mg,
0.05 mmol) in anhyd. toluene (0.5 mL) at 278 8C was
added a solution of catecholborane (21.3 mL, 0.2 mmol)
dropwise under an atmosphere of N2. The reaction mixture
was stirred at 278 8C for 6 h and then kept in the freezer
(218 8C) for 15 h. H2O (3 mL) was added, and the mixture
was extracted with EtOAc (3£5 mL). The combined organic
layers were washed with sat. NaHCO3 (aq.) (10 mL), dried
(MgSO4) and rotary evaporated to obtain a crude product,
which was purified by CC (5% EtOAc/n-hexane) to yield 1
(15.5 mg, 0.043 mmol, 88%; Rf 0.35).
The enantiomeric composition of the product was estimated
at 90.7% of (þ)-1 and 9.3% of (2)-1 by chiral HPLC
analysis (Section 3.2.6), and the % ee of the reaction was
therefore 81.4%.
When the same experimental procedure was employed
using the catalyst Rh(I)-(R,R)-Et-DuPhos (Strem Chemical
Company, Cat. no. 45–0150) in place of its (S) enantiomer,
compound 1 (18.4 mg, 0.052 mmol, 73%; Rf 0.22) was
obtained with 0.4% ee ([a]D¼20.67, [c¼1.8, CHCl3]).
3.4.5. Asymmetric reduction of 17 by catecholborane in
the presence of the chiral oxazaborolidine catalyst (S)-
18. The oxazoborolidine catalyst (S)-18 was prepared in
four steps from (S)-proline, according to Corey’s pro-
cedure.26,30 The same experimental procedure was
employed as for the enantioselective reduction of 17 in
the previous section, except that (S)-18 was substituted for
(R)-18. The product (15.7 mg, 0.044 mmol, 89%) consisted
of 92.2% of (2)-1, and 7.8% of (þ)-1, and the % ee of the
reaction was therefore 84.4%.
3.4.2. Methyl esterification of compound 10 to 2,3-
dibenzyl-4-oxo-5-phenyl-pent-2-en-oic acid methyl ester
(17). Compound 10 (500 mg, 1.35 mmol) was dissolved in
Et2O (20 mL) and then treated with excess of a solution of
CH2N2 in Et2O (100 mL) at room temperature for 4 h. After
the reaction was complete, excess CH2N2 and Et2O were
removed by gently passing N2 gas through the solution,
leaving a product consisting entirely of compound 17
(477 mg, 1.24 mmol, 92%), which could be used without
the need for any further purification.
3.5. Investigation of the extent of partial racemization in
(1) maculalactone A isolated from K. maculans
3.5.1. Determination of the enantiomeric purity of
naturally-occurring (1) maculalactone A (1) by analyti-
cal HPLC using a chiral column. See Section 3.1.1 for
details of the extraction and isolation procedure for
obtaining maculalactone A from K. maculans and Section
3.2.6 for details of the analytical chiral HPLC analysis.
2,3-Dibenzyl-4-oxo-5-phenyl-pent-2-en-oic acid methyl
ester (17). Solid, mp 65–678 C; nmax/cm21: 3030, 3019,
2954, 1710, 1603, 1496, 1454 cm21; dH-7.30 (2H, dd,
J¼6.9, 6.9 Hz), 7.28 (2H, dd, J¼7.9, 7.9 Hz), 7.26–7.20
(5H, m), 7.14 (2H, dd, J¼8.2, 1.1 Hz), 7.07 (2H, dd, J¼7.9,
1.1 Hz), 7.03 (2H, dd, J¼8.1, 1.5 Hz), 3.86 (2H, s), 3.70
(2H, s), 3.67 (3H, s), 3.52 (2H, s)—see also Table 2; 13C
NMR—see Table 2; HREIMS m/z (rel. int.): 352.1471
(Mþ2MeOH, calcd 352.1463 for C25H20O2) (100), 293
(25), 261 (20).
3.5.2. Determination of the enantiomeric purity of
naturally-occurring (1) maculalactone A (1) by 1H
NMR spectroscopy in the presence of a chiral shift
reagent. The enantiomeric composition of 1 which had
been isolated from K. maculans was determined by 1H NMR
spectroscopy in the presence of the chiral shift reagent,
3.4.3. Reduction of compound 17 to racemic macula-
lactone A (6)-1 by catecholborane. To a solution of 17
(10 mg, 0.026 mmol) in anhyd. toluene (0.26 mL) was
added a solution of catecholborane (5.5 mL, 0.052 mmol)
under an atmosphere of N2 at 278 8C. The mixture was
stirred for 6 h and then left in the freezer (at 218 8C) for
15 h. After the reaction was complete, as indicated by TLC,
H2O (1 mL) was added and the resulting mixture was
extracted with EtOAc (3£5 mL). The combined organic
layers were washed with sat. NaHCO3 (aq.) (5 mL), dried
(MgSO4) and rotary evaporated to obtain a crude product,
which was purified by CC (5% EtOAc/n-hexane), to yield
racemic (^)-1 (7.7 mg, 0.022 mmol, 84%; Rf 0.35).
europium
tris[3-(heptafluoropropylhydroxy-methylene)-
(þ)-camphorate]. A sample of 1 (10 mg) from the natural
source was dissolved in CDCl3 (0.5 mL, containing 0.03%
v/v TMS) and 1H NMR spectra were recorded after
successive additions of the chiral shift reagent (10 mg per
addition) into the sample (expts. 1–7 in Fig. 6). The
experiment was stopped when sufficient chiral shift reagent
1
had been added to obtain a clear baseline separation of H
resonances which were considered characteristic of the two
enantiomeric forms of 1. Expts. 8–14 in Figure 6 were
recorded in a similar way using racemic (^)-1 which had
been obtained from synthesis (see Section 3.2.5) in place of
natural maculalactone A.
3.4.4. Asymmetric reduction of 17 by catecholborane in
the presence of the chiral oxazaborolidine catalyst (R)-
18. The oxazoborolidine catalyst (R)-18 was prepared in
four steps from (R)-proline (the more expensive unnatural
enantiomer of this amino acid), according to Corey’s
Acknowledgements
We thank the Chemistry Department of the University of
Hong Kong for providing a postgraduate studentship to Mr.
procedure.26,30 To
a
solution of (R)-18 (20 mL,

G. D. Brown, H.-F. Wong / Tetrahedron 60 (2004) 5439–5451
5451
16. Chavez, F.; Suarez, S.; Diaz, M. A. Synth. Commun. 1994, 24,
2325–2339.
Wong. This project was funded by a grant from the
Committee on Research and Conference Grants (CRCG) of
The University of Hong Kong.
17. Bushan, K. M.; Rao, G. V.; Soujanya, T.; Rao, V. J.; Saha, S.;
Samanta, A. J. Org. Chem. 2001, 66, 681–688.
18. Allenmark, S. G. Nat. Prod. Rep. 2000, 17, 145–155.
19. Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95,
512–519.
References and notes
20. Seco, J. M.; Quinoa, E.; Riguera, R. Chem. Rev. 2004, 104,
17–117.
1. Kaehler, S. Asian Mar. Biol. 1994, 11, 41–54.
2. Kamura, S.; Choohabandit, S. Galaxea 1986, 5, 97–114.
3. Kaehler, S.; Williams, G. A. Mar. Biol. 1996, 125, 177–187.
4. Tsui, W.-Y.; Williams, G. A.; Brown, G. D. Phytochemistry
1996, 43, 1083–1085.
21. Seco, J. M.; Quin˜oa, E.; Riguera, R. Tetrahedron: Asymmetry
2000, 2781–2791.
22. Burk, M. J. J. Am. Chem. Soc. 1991, 113, 8518–8519.
23. Burk, M. J.; Feaster, J. E.; Harlow, R. L. Tetrahedron:
Asymmetry 1991, 2, 569–592.
5. Lee, S.-C.; Brown, G. D. J. Nat. Prod. 1998, 61, 29–33.
6. Lee, S.-C.; Williams, G. A.; Brown, G. D. Phytochemistry
1999, 52, 537–540.
24. Burk, M. J.; Feaster, J. E.; Harlow, R. L. Organometallics
1990, 9, 2653–2655.
7. Wong, H.-F.; Williams, G. A.; Brown, G. D. Phytochemistry
2002, 60, 425–429.
25. Thompson, H. W.; Wong, J. K.; Lalancette, R. A.; Boyko,
J. A.; Robertiello, A. M. J. Org. Chem. 1985, 50, 2115–2121.
26. Corey, E. J.; Shibata, S.; Bakshi, R. K. J. Org. Chem. 1988, 53,
2861–2863.
8. Lee, S.-C. Chemical ecology of algae and the cyanobacterium
Kyrtuthrix maculans on Hong Kong rocky shores. Ph.D.
Thesis, The University of Hong Kong, 2001.
9. Brown, G. D.; Wong, H.-F.; Hutchinson, N.; Lee, S.-C.;
Williams, G. A. Chemistry and biology of maculalactone A
from the marine cyanobacterium Kyrtuthrix maculans,
International Symposium Chemistry and Biology of Marine
Organisms, Kolympari, Crete., 21–26 September 2003;
pp. 60–61.
27. Corey, E. J.; Da Silva Jardine, P.; Mohri, T. Tetrahedron Lett.
1988, 29, 6409–6412.
28. Corey, E. J.; Da Silva Jardine, P.; Rohloff, J. C. J. Am. Chem.
Soc. 1988, 110, 3672–3673.
29. Corey, E. J.; Reichard, G. A. Tetrahedron Lett. 1989, 30,
5207–5210.
10. Wong, H.-F. Applications of high field nuclear magnetic
resonance spectroscopy to the structure elucidation, confor-
mational analysis and asymmetric synthesis of natural
products. Ph.D. Thesis, The University of Hong Kong, 2002.
11. Ward, R. S.; Pelter, A.; Edwards, M. I.; Gilmore, J.
Tetrahedron 1996, 52, 12799–12814.
30. Corey, E. J.; Bakshi, R. K. Tetrahedron Lett. 1990, 31,
611–614.
31. Corey, E. J.; Bakshi, R. K.; Shibata, S.; Chen, C.-P.; Singh,
V. K. J. Am. Chem. Soc. 1987, 109, 7925–7926.
32. Corey, E. J.; Helal, C. J. Angew. Chem. Int. Ed. 1998, 37,
1986–2012.
12. Johnson, W. S.; Daub, G. H. Organic Reactions; Wiley: New
York, 1951; Vol. 6. pp. 1–73.
33. Kemp, W. Organic Spectroscopy; MacMillan Press: New
York, 1991; pp. 169–170.
13. Andersson, J. Acta Chem. Scand. B 1977, B31, 340–341.
14. Fieser, L. F.; Martin, E. L. Succinic Anhydride. Organic
Synthesis; Blatt, A. H., Ed.; Wiley: London, 1948; Collective
Vol. 2, pp 560–562.
34. Charlwood, B. V.; Banthorpe, D. V. Methods in Plant
Biochemistry. In Terpenoids; Dey, P. M., Harborne, J. B.,
Eds.; Academic: New York; Vol. 7, pp. 33–34.
35. Sy, L.-K.; Brown, G. D. J. Nat. Prod. 1998, 61, 987–992.
15. Davidse, P. A.; Dillen, J. L. M.; Heyns, A. M.; Modro, T. A.;
van Rooyen, P. H. Can. J. Chem. 1990, 68, 741–746.