The enantioselective synthesis of APTO and AETD: Polyhydroxylated β-amino acid constituents of the microsclerodermin cyclic peptides

Article abstract of DOI:10.1039/b707891a

The polyhydroxylated β-amino acids (2S,3R,4S,5S,7E)-3-amino-8-phenyl- 2,4,5-trihydroxyoct-7-enoic acid (APTO) and (2S,3R,4S,5S,7E,9E)-3-amino-10-(4- ethoxyphenyl)-2,4,5-trihydroxydeca-7,9-dienoic acid (AETD) are key components of the microsclerodermin fam

Full text of DOI:10.1039/b707891a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
The enantioselective synthesis of APTO and AETD: polyhydroxylated
b-amino acid constituents of the microsclerodermin cyclic peptides†
Emily C. Shuter,a Hung Duong,a Craig A. Huttonb,c and Malcolm D. McLeod*a
Received 24th May 2007, Accepted 26th July 2007
First published as an Advance Article on the web 17th August 2007
DOI: 10.1039/b707891a
The polyhydroxylated b-amino acids (2S,3R,4S,5S,7E)-3-amino-8-phenyl-2,4,5-trihydroxyoct-7-enoic
acid (APTO) and (2S,3R,4S,5S,7E,9E)-3-amino-10-(4-ethoxyphenyl)-2,4,5-trihydroxydeca-7,9-dienoic
acid (AETD) are key components of the microsclerodermin family of anti-fungal cyclic peptides. They
have been synthesised in protected form in twelve steps using a unified strategy, with the introduction of
the unsaturated sidechain in the final step of the synthesis from a common aldehyde intermediate. The
synthesis features the ordered application of asymmetric aminohydroxylation and dihydroxylation
reactions to efficiently introduce the stereochemistry of the targets with high selectivity.
Introduction
The microsclerodermins are a family of cyclic hexapeptides iso-
lated from marine sponges of the Microscleroderma and Theonella
genera in waters off New Caledonia and the Philippines.1 Nine
members of this natural product family, microsclerodermins A–
I, have been isolated to date and all exhibit anti-fungal activity
against the clinically significant fungal pathogen Candida albicans
in a paper disc diffusion assay. In addition, microsclerodermins
F–I display moderate cytotoxicity against the HCT-116 cell line
(1.0–2.4 lg mL−1).1
Structurally, members of the family are typified by micro-
sclerodermin C (1, Fig. 1), which contains a 23-membered
ring constructed of six amino acid residues. Three of these,
glycine, N-methylglycine and (3R)-4-amino-3-hydroxybutyric
acid (GABOB) are common to all members of the microsclero-
dermin family. The remaining three unusual amino acids display
considerable variation across the microsclerodermin family and
for microsclerodermin C consist of (R)-N-carbamoyl-6-chlorotry-
ptophan, [(2R,3S)-3-amino-2-hydroxy-5-oxo-2-pyrrolidin-2-yl]-
acetic acid and (2S,3R,4S,5S,7E)-3-amino-8-phenyl-2,4,5-
trihydroxyoct-7-enoic acid (APTO).
Fig. 1 Microsclerodermin C 1 and AETD.
The synthetic challenge posed by the microsclerodermins com-
alkanoic acid head group but vary in the nature of the aromatic
bined with their significant biological activity has made them
terminated sidechain.
an attractive target for total synthesis. A key issue for any
A number of synthetic studies targeting the microsclerodermins
projected total synthesis is the efficient construction of the unusual
have appeared to date. Foremost among these reports is the
total synthesis of microsclerodermin E by Zhu and Ma in 2003.2
This group developed an enantiospecific 21-step synthesis of a
protected AETD–GABOB dipeptide fragment wherein the stere-
ochemistry of the head group was derived from d-gluconolactone.
Synthetic studies targeting the 3-amino-2,4,5-trihydroxy acid
component of the microsclerodermins have also been reported.
These include the pioneering work of Shioiri and co-workers3
who completed a 31-step enantiospecific synthesis of the 3-
amino-2,4,5-trihydroxy acid component of microsclerodermins
A and B, (2S,3R,4S,5S,6S,11E)-3-amino-12-(4-methoxyphenyl)-
6-methyl-2,4,5-trihydroxydodec-11-enoic acid (AMMTD), from
methyl (R)-(−)-3-hydroxy-2-methylpropionate. Recently, Chan-
drasekhar and Sultana4 have reported the 26-step enantiospe-
cific synthesis of AMMTD from (S)-(−)-citronellal as starting
amino acids and in particular the 3-amino-2,4,5-trihydroxyacid
typified by APTO in microsclerodermin C. Without exception, the
congeners of APTO contain the same 3-amino-2,4,5-trihydroxy-
aSchool of Chemistry, F11, University of Sydney, NSW 2006, Australia.
E-mail: m.mcleod@chem.usyd.edu.au; Fax: +61-2-9351-6650; Tel: +61-2-
9351-5877
bSchool of Chemistry, The University of Melbourne, VIC 3010, Australia.
E-mail: chutton@unimelb.edu.au; Fax: +61-3-9347-5180; Tel: +61-3-8344-
2393
cBio21 Molecular Science and Biotechnology Institute, The University of
Melbourne, VIC 3010, Australia
† Electronic supplementary information (ESI) available: General experi-
mental details together with experimental procedures and spectroscopic
1
data for compounds 4–9, 16–21, 23–27 and copies of H and 13C NMR
spectra for (7E)-2, and (7E,9E)-3. See DOI: 10.1039/b707891a
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3183–3189 | 3183
©

View Article Online
material. Herein we disclose the efficient enantioselective synthesis
of 3-amino-2,4,5-trihydroxy acid components of microscleroder-
mins C and D (APTO) and microsclerodermin E (AETD) in
protected form.
Results and discussion
On consideration of the target structures, it was envisaged that
protected APTO 2 and AETD 3 (Scheme 1) could be derived from
a common, aldehyde intermediate 4 in a single step, by Wittig
olefination with an appropriate phosphonium salt. Aldehyde 4
could be derived from diol 5 in three steps: protection of the
C4–C5 diol as the acetonide, oxidative deprotection of the p-
methoxyphenyl protecting group and oxidation of the resulting
alcohol to the aldehyde. Diol 5 could in turn be formed from
(E)-alkene 6 by Sharpless asymmetric dihydroxylation (AD). The
diastereoselectivity of this key transformation would depend on
effective stereochemical induction from the AD catalyst and any
induction derived from the chiral substrate.
(E)-Alkene 6 could be obtained by modified Julia olefination
of aldehyde 7, which could in turn be prepared from b-amino
alcohol 8. Finally, it was envisaged that b-amino alcohol 8 could
be constructed with high enantioselectivity from (E)-alkene 9 by
Sharpless asymmetric aminohydroxylation (AA), wherein careful
matching of substrate and catalyst controls the enantio- and regio-
selectivity of the reaction.
This strategy allows for the synthesis of both unusual amino
acids APTO 2 and AETD 3 from aldehyde 4 through introduction
of the unsaturated sidechain in the final step of the synthesis.
The synthesis also features a carefully orchestrated sequence
of Sharpless AA and AD reactions to efficiently install the
polyhydroxylated amino acid functionality present in APTO,
AETD and other common amino acid building blocks of the
microsclerodermin family.1
=
Scheme 2 (a) CH2 CHCOOMe, Grubbs’ second generation catalyst
11 (7 mol%), CH2Cl2, D, 36 h, 65%; (b) OsO4 (1 mol%), NaIO4, Et2O,
H2O, rt, 39 h; (c) (EtO)2POCH2COOMe 13, n-BuLi, THF, −78 ◦C
to rt, 5 h, 91% over two steps; (d) t-BuOCONH2, K2OsO2(OH)4
(5 mol%), (DHQD)2PHAL 14 (7 mol%), 1,3-dichloro-5,5-dimethyl-
hydantoin, NaOH, n-PrOH, H2O, rt, 48 h, 84%; (e) Ac2O, Et3N,
4-(dimethylamino)pyridine, CH2Cl2, rt, 1 h, 40%.
quantities of (E)-alkene 9 demanded the development of a more
cost effective alternative. Oxidative cleavage of allyl ester 10 gave
aldehyde 127 in quantitative yield which was used directly in a
Horner–Wadsworth–Emmons reaction with methyl diethylphos-
phonoacetate 13 to give the products with an E : Z ratio of 25 : 1.
Separation of the diastereomers was readily achieved by flash
chromatography to afford pure (E)-alkene 9 in 91% yield over
two steps.
Two synthetic routes were developed for the synthesis of the (E)-
alkene substrate required for the substrate directed AA reaction
(Scheme 2). Cross metathesis of allyl 4-methoxybenzoate 105
and methyl acrylate using Grubbs’ second generation catalyst
116 selectively afforded (E)-alkene 9 in 65% yield. Although this
process was efficient on a small scale, the need for substantial
The (DHQD)2PHAL 14 mediated AA reaction8 of (E)-alkene
9 proceeded smoothly to afford b-amino alcohol 8 as a single
Scheme 1 Retrosynthetic analysis.
This journal is
3184 | Org. Biomol. Chem., 2007, 5, 3183–3189
The Royal Society of Chemistry 2007
©

View Article Online
regioisomer in 84% yield and 97% enantiomeric excess as de-
termined by chiral HPLC of the acetate derivative 15.9 The
(S)-absolute configuration of the secondary alcohol bearing C2
stereocentre was established by the modified Mosher’s method10
and was consistent with the outcome predicted by the Sharpless
mnemonic.8 The high regioselectivity obtained in this AA reaction
derives from the high levels of substrate direction afforded by 4-
methoxybenzoate ester protection of the adjacent allylic alcohol
functionality.5,7
With b-amino alcohol 8 in hand two avenues for the elaboration
to AD substrate 6 were explored. Initially, protection of secondary
alcohol 8 gave silyl ether 16 in 75% yield and transesterification
of the 4-methoxybenzoate ester protecting group gave primary
alcohol 17 in 64% yield (Scheme 3). Despite the success of
this sequence, attempted oxidation of alcohol 17 by a range of
methods11 failed to give the desired aldehyde required for the
Julia olefination and instead gave decomposition to a range of
unidentified products. Alcohol 17 was also observed to undergo
lactonisation on prolonged storage.
Scheme 4
acetal 19 and product 18. The use of 2-methoxypropene (2-
MP) at lower temperature (entry 2) also failed to give full
conversion. The use of 2-methoxypropene and pyridinium p-
toluenesulfonate (PPTS) at elevated temperature (entry 3) gave
complete conversion and an improved yield of oxazolidine 18 but
also afforded condensation by-product 20 as a single diastereomer
of undetermined configuration. Finally, it was discovered that
reaction of alcohol 8 at room temperature for 2 h gave clean
conversion to intermediate acetal 19, after which, the volatile
components were removed from the reaction mixture on a rotary
evaporator and oxazolidine formation was promoted by heating
the remaining toluene solution. In this manner oxazolidine 18 was
formed in 84% yield with a small amount of amino alcohol 8 (15%)
being recovered.
Transesterification of the 4-methoxybenzoate ester 18 with
caesium carbonate gave an 81% yield of primary alcohol 21
(Scheme 5). This was oxidised to aldehyde 7 in quantitative yield
using Dess–Martin periodinane,13 setting the stage for the Julia
olefination chain extension.
Scheme 3 (a) TBSOTf, 2,6-lutidine, CH2Cl2, −78 ◦C to rt, 22 h, 75%;
(b) Cs2CO3, MeOH, rt, 24 h, 64%.
In an effort to overcome these difficulties the protection of b-
amino alcohol 8 as the oxazolidine derivative 18 was explored
(Scheme 4, Table 1). In addition to protecting both amine
and alcohol functional groups, it was anticipated that the trans
relationship of the ester and alcohol functional groups in this cyclic
derivative would eliminate the propensity for lactone formation.
The protection of alcohol 8 to give oxazolidine 18 proved
difficult under a range of conditions.12 The use of excess 2,2-
dimethoxypropane (2,2-DMP) and p-toluenesulfonic acid (TsOH)
(Table 1, entry 1) gave mixtures of starting alcohol 8, intermediate
Both benzothiazol-2-yl (BT)14 and 1-phenyl-1H-tetrazol-5-yl
(PT)15 sulfones were synthesised in order to evaluate the influence
of the aromatic substituent on the selectivity of the modified Julia
olefination reaction.16 The synthesis started with mono-protected
propanediol 2217 (Scheme 6). In each case, formation of the
thioether was achieved under Mitsunobu conditions18 followed by
molybdenum(VI) mediated oxidation19 to give the corresponding
sulfones 23 and 24.
Julia olefination using BT-sulfone 23 (Table 2, entry 1) or PT-
sulfone 24 (entry 2) and lithium diisopropylamide (LDA) as the
Table 1 Conditions for oxazolidine formation
Yield (%)
Entry
Reagent
Acid
Solvent
Temp./◦C (time/h)
8
18
19
20
1
2
2,2-DMP
2-MP
2-MP
TsOH
TsOH
PPTS
PPTS
Benzene
Benzene
Toluene
Toluene
95 (32)
40
22
—
15
36
60
77
84
2
—
—
20
—
0 (2), rt (5)
rt (1), 110 (3)
rt (2), 110 (3)
—
—
—
3
4a
2-MP
a Low boiling volatile components were removed on a rotary evaporator prior to heating the reaction mixture.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3183–3189 | 3185
©

View Article Online
Table 2 Julia olefination of aldehyde 7 with sulfones 23 and 24
Entry
Sulfone
Solvent
Base
Temp./◦C
Time/h
Yield (%)
E : Za
1
2
3
4
23
24
24
24
THF
THF
1,2-DME
1,2-DME
LDA
LDA
NaHMDS
KHMDS
−78
−78
−60
−60
17
17
0.5
0.75
42
35
58
67
1.2 : 1
1.6 : 1
1.8 : 1
2.7 : 1
a Determined by integration of the C3-H signals in the 200 MHz 1H NMR spectrum of the crude reaction mixture.
using the more polar solvent 1,2-dimethoxyethane (1,2-DME)
in conjunction with sodium bis(trimethylsilyl)amide (NaHMDS,
entry 3) or potassium bis(trimethylsilyl)amide (KHMDS, entry 4).
The results were in agreement with previous studies16 with the
most (E)-selective conditions being 1,2-DME with KHMDS. The
highest yield was also obtained under these conditions. Despite
the modest selectivity of the modified Julia olefination, it was
discovered that alkene mixtures enriched in (Z)-alkene (Z)-6 could
be successfully isomerised to the (E)-alkene isomer (E)-6 using
phenylthiyl radical.20
With quantities of the (E)-alkene (E)-6 in hand, the AD
reaction21 to install the 4,5-diol functionality was investigated
(Scheme 7). It was envisaged that the diastereoselectivity of this
reaction would be governed by both the influence of the resident
stereocentres within the substrate and the control imparted by
the chiral ligands of the dihydroxylation catalyst. The intrinsic
diastereoselectivity imparted by the substrate was determined
by performing the reaction in the absence of chiral ligand
(Table 3, entry 1); this favoured the (4S,5S)-product (4S,5S)-
5 in a modest 2 : 1 ratio of diastereomers as determined by
1H NMR integration of the C2-H and C3-H signals.22 This
modest preference for the 3,4-anti diastereomer was consistent
with previous reports regarding the dihydroxylation of N-tert-
butoxycarbonyl-2,2-dimethyloxazolin-3-yl substituted olefins.23
Scheme 5 (a) Cs2CO3, MeOH, rt, 16 h, 81%; (b) Dess–Martin perio-
dinane, CH2Cl2, rt, 1 h, quant.; (c) 23 or 24; then base (see Table 2);
(d) PhSH, AIBN, benzene, 85 ◦C, 10 days, 98%, E : Z 6.5 : 1.
Scheme 6 (a) ArSH, PPh3, THF, diisopropyl azodicarboxylate, 0 ◦C to
rt, Ar = BT, 15 h, 82%, Ar = PT, 2.5 h, 69%; (b) (NH4)6Mo7O24·4H2O,
H2O2, EtOH, 0 ◦C to rt, Ar = BT, 30 h, 98%, Ar = PT, 20 h, 99%.
base under Barbier conditions16 gave alkene 6 in modest yields and
selectivities. The PT-sulfone was found to most strongly favour
formation of the (E)-alkene. The effect of solvent polarity and
base counter-ion on the selectivity of the reaction was studied
Scheme 7
Table 3 Diastereoselective dihydroxylation of alkene 6
Entry
Temperature/◦C (time/h)
Ligand
Yield (%)
(4S,5S) : (4R,5R)a
1b
2c
3c
4d
5d
6d
7d
8d
28 (2)
none
48
0
0
11
5
42
19
85
2.0 : 1
n/a
n/a
3.3 : 1
1 : 2.5
7.4 : 1
1 : 5.1
7.4 : 1
0 (5), 28 (72)
0 (5), 28 (72)
17 (17)
17 (17)
17 (17)
17 (17)
17 (10), 27 (4)
(DHQ)2PHAL
(DHQD)2PHAL
DHQ–CLB
DHQD–CLB
DHQ–IND
DHQD–IND
DHQ–IND
a Determined by integration of the C2-H and C3-H signals in the 300 MHz 1H NMR spectrum of the crude reaction mixture. b OsO4 (10 mol%),
N-methylmorpholine-N-oxide (3.2 eq.), THF–t-BuOH–H2O. c AD-mix-a (entry 2) or AD-mix-b (entry 3),21 MeSO2NH2 (1.0 eq.), NaHCO3 (3.0 eq.),
t-BuOH–H2O. d K2OsO2(OH)4 (1 mol%), ligand (6 mol%), K3Fe(CN)6 (3.0 eq.), K2CO3 (3.2 eq.), NaHCO3, (3.2 eq.), MeSO2NH2 (3.2 eq.), t-BuOH–H2O.
3186 | Org. Biomol. Chem., 2007, 5, 3183–3189
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Protection of diol 5 to give acetonide 25 proceeded smoothly
(99%) and oxidative deprotection of the PMP ether then afforded
primary alcohol 26 in 95% yield (Scheme 8). Oxidation13 afforded
aldehyde 4, which was used without purification in the subsequent
step.
Fig. 2 Sharpless ligands.
Dihydroxylation with the phthalazine-derived chiral ligands
(DHQ)2PHAL (Fig. 2) and (DHQD)2PHAL 14 failed to afford
the desired product, instead leading to slow formation of a polar
compound consistent with base mediated hydrolysis of the methyl
ester to a carboxylic acid. The low reactivity of the alkene was
presumed to be due to the large steric demands imposed by
the proximal N-(tert-butoxycarbonyl)-2,2-dimethyloxazoline ring
system, which prevented effective interaction with the catalyst.
To overcome this problem sterically less demanding mono-
alkaloid ligands were investigated. Although these have been
found to typically provide lower levels of catalyst derived stere-
ocontrol in the AD reaction,21 it was anticipated that matching
the stereocontrol of the ligand and substrate would increase
the modest preference for the (4S,5S)-diastereomer (4S,5S)-5
observed above.24 Asymmetric dihydroxylation reaction with the
mono-alkaloid ligand DHQ–CLB gave a modest increase in
selectivity for the desired diastereomer (entry 4). Conversely,
the pseudo-enantiomeric DHQD–CLB ligand overturned the
substrate bias to favour the (4R,5R)-diastereomer (entry 5). The
use of the mono-alkaloid ligand DHQ–IND gave a 7.4 : 1 ratio
favouring the desired (4S,5S)-diastereomer in a matched reaction
(entry 6), which was overturned by the DHQD–IND ligand
(entry 7). On a larger scale, it was observed that higher yields could
be obtained through the use of elevated temperature providing
an 85% yield of the diols, favouring the (4S,5S)-diastereomer,
with no loss of diastereoselectivity. The diastereomers produced
by the dihydroxylation reaction were inseparable by column
chromatography so the mixture was carried through the remaining
steps and separated at the conclusion of the synthesis.
Scheme 8 (a) 2,2-DMP, CH2Cl2, (1S)-(+)-10-camphorsulfonic acid,
35 min, 99%; (b) (NH4)2Ce(NO3)6, CH3CN, H2O, 0 ◦C, 18 min, 95%;
(c) Dess–Martin periodinane, CH2Cl2, rt, 45 min, quant.; (d) benzyltri-
phenylphosphonium chloride, KHMDS, benzene, 0 ◦C to rt, 15 min; then
aldehyde 4, benzene, 0 ◦C to rt, 1 h, 78%.
Wittig reaction of aldehyde 4 using the ylide derived from
treatment of benzyltriphenylphosphonium chloride with potas-
sium bis(trimethylsilyl)amide gave alkene 2 in 78% yield, in a
1
3.5 : 1 ratio of (7E) : (7Z)-diastereomers as determined by H
NMR integration of the C7-H and C8-H signals. The isomers
were inseparable by flash chromatography. Preparative HPLC was
used to separate the two major isomers and a mixed fraction
containing trace by-products. The by-products were not analysed
further but were assumed to be alkene products derived from the
minor diastereomer formed in the AD reaction which were carried
through the subsequent steps. The assigned (E)-stereochemistry of
the major isomer (E)-2 was consistent with the magnitude of the
C7-H vicinal coupling constant (3J7,8 = 15.9 Hz). In contrast,
a smaller C7-H vicinal coupling constant (3J7,8 = 11.7 Hz) was
observed for the minor (7Z)-isomer (Z)-2. Protected APTO (E)-
2 was synthesised in twelve linear steps from allyl ester 10 in an
overall yield of 12% and in 97% ee. With this target in hand the
project turned to focus on the synthesis of protected AETD which
contains the same polyhydroxylated amino acid core but differs in
the unsaturated sidechain.
The phosphonium salt 27 required for the synthesis of protected
AETD 3 was prepared in two steps from the known (E)-allylic
alcohol 28 (Scheme 9).25 Bromination using phosphorus tribro-
mide afforded unstable allylic bromide 29 which was immediately
treated with triphenylphosphine to give phosphonium salt 27 as a
colourless hygroscopic solid.
The Wittig reaction with aldehyde 4 was conducted under
analogous conditions to those developed for the synthesis of
APTO derivative 2 and gave diene product 3 in 63% yield, with
a 3.2 : 1 ratio of (7E,9E) : (7Z,9E) diastereomers as determined
The successful application of the AA followed by AD sequence
for the synthesis of the protected 3-amino-2,4,5-trihydroxy acid
head group of APTO and AETD is noteworthy. In the synthesis
of AMMTD reported by Chandrasekhar and Sultana4 the alter-
native sequence of AD followed by AA reaction was attempted to
construct a closely related stereochemical array. In this case the
AA reaction failed to provide the desired product and necessitated
a lengthy synthetic detour to install the required functionality. The
failure of the AA reaction in this context can be explained given
the generally lower substrate tolerance of the AA reaction when
compared to its AD counterpart.8,21
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3183–3189 | 3187
©

View Article Online
10 mL) and the crude product was extracted into ethyl acetate
(3 × 30 mL). The combined organic layers were dried (Na2SO4),
filtered and concentrated under reduced pressure to give an orange
residue (0.24 g). Flash chromatographic purification (20% ethyl
acetate : hexanes) of the crude product gave an inseparable mixture
of (7E)-alkene (7E)-2 and (7Z)-alkene (7Z)-2 as a viscous, clear
colourless liquid (0.096 g, 78%); [a]1D9 −2.2 (c 1.7, CH2Cl2); mmax
(thin film)/cm−1 1763 (m), 1740 (m), 1701 (s, C O); HRMS (+ESI)
=
calc. for C26H37O7N + Na 498.2468, found 498.2458; m/z (+ESI)
498 ([M + Na]+, 100%), 442 (24). Preparative HPLC purification
(RTI Zorbax Sil, 7.5% ethyl acetate : hexanes) of a portion
of the purified product (0.048 g) gave pure (7E)-alkene (7E)-2
(tR ≈31.5 min) as a viscous, clear, colourless liquid (0.038 g, 31%);
[a]2D1 −3.1 (c 3.8, CH2Cl2); dH (200 MHz; 340 K; CDCl3) 7.36–7.12
(5H, m, Ph), 6.49 (1H, d, J 15.9 Hz, C8-H), 6.26 (1H, dt, J 15.9,
6.8 Hz, C7-H), 4.71 (1H, d, J 2.1 Hz, C2-H), 4.47 (1H, dd, J 5.2,
2.1 Hz, C3-H), 4.19–4.10 (1H, m, C5-H), 4.01 (1H, dd, J 7.9, 5.2 Hz,
C4-H), 3.73 (3H, s, C1–OCH3), 2.65–2.39 (2H, m, C6-H), 1.61
(3H, s, NC(CH3)A(CH3)BO), 1.56 (3H, s, NC(CH3)A(CH3)BO),
1.49 (9H, s, OC(CH3)3), 1.41 (6H, s, OC(CH3)2O); dC (50 MHz;
340 K; CDCl3) 172.1, 151.9, 137.7, 133.0, 128.5, 127.1, 126.3,
125.6, 109.0, 96.8, 80.9, 80.1, 78.4, 75.2, 61.4, 52.2, 36.6, 28.5, 27.7,
27.4, 27.2 (2C). A second fraction gave pure (7Z)-alkene (7Z)-2
(tR ≈28.5 min) as a viscous, clear, colourless liquid (0.0080 g,
6.5%); [a]2D1 +3.7 (c 0.8, CH2Cl2); dH (200 MHz; 340 K; CDCl3)
7.30–7.16 (5H, m, Ph), 6.58–6.52 (1H, m, C8-H), 5.79 (1H, dt,
J 11.7, 7.0 Hz, C7-H), 4.68 (1H, d, J 2.1 Hz, C2-H), 4.44 (1H,
m, C3-H), 4.22–4.10 (1H, m, C5-H), 3.97 (1H, dd, J 7.8, 5.5 Hz,
C4-H), 3.73 (3H, s, C1–OCH3), 2.66–2.59 (2H, m, C6-H), 1.58
(3H, s, NC(CH3)A(CH3)BO), 1.48 (3H, s, NC(CH3)A(CH3)BO),
1.42 (12H, s, OC(CH3)3, OC(CH3)2O), 1.40 (3H, s, OC(CH3)2O);
dC (50 MHz; 340 K; CDCl3) 172.1, 151.8, 137.5, 131.4, 128.9,
128.2, 127.2, 126.8, 109.1, 96.8, 80.9, 80.4, 78.3, 75.3, 61.5, 52.2,
32.5, 28.4, 27.6, 27.4, 27.2, 27.0.
Scheme 9 (a) PBr3, diethyl ether, 0 ◦C to rt, 80 min, 89%; (b) PPh3,
benzene, rt, 15 h, 87%; (c) KHMDS, benzene, 0 ◦C to rt, 10 min; then
aldehyde 4, benzene, 0 ◦C to rt, 25 min, 63%.
1
by H NMR integration of the C7-H signals. Preparative HPLC
was used to separate the two major isomers and a mixed fraction
containing trace by-products. The assigned (7E)-stereochemistry
of the major isomer (7E,9E)-3 was consistent with the magnitude
of the C7-H vicinal coupling constant (3J7,8 = 14.8 Hz). In contrast,
a smaller C7-H vicinal coupling constant (3J7,8 = 10.9 Hz) was
observed for the minor (7Z)-isomer (7Z,9E)-3.
Conclusions
Protected APTO (E)-2 and AETD (7E,9E)-3 were synthesised in
twelve linear steps from allyl ester 10 in an overall yield of 12%
and 9% respectively and in 97% ee. The strategy allowed for the
synthesis of both unusual amino acids from aldehyde 4 through
the introduction of the unsaturated sidechain in the final step. This
synthesis features a carefully orchestrated sequence of Sharpless
AA and AD reactions to install the polyhydroxylated amino acid
functionalitypresent inthese targets and which is common toother
amino acid building blocks of the microsclerodermin family.
(4R,5S)-3-tert-Butyl-5-methyl-4-{(4S,5S)-5-[(2E,4E)-5-(4-
ethoxyphenyl)penta-2,4-dienyl]-2,2-dimethyl-1,3-dioxolan-4-yl}-
2,2-dimethyloxazolidine-3,5-dicarboxylate (7E,9E)-3 and
(4R,5S)-3-tert-butyl-5-methyl-4-{(4S,5S)-5-[(2Z,4E)-5-(4-
ethoxyphenyl)penta-2,4-dienyl]-2,2-dimethyl-1,3-dioxolan-4-yl}-
2,2-dimethyloxazolidine-3,5-dicarboxylate (7Z,9E)-3
Experimental
To a suspension of phosphonium salt 27 (0.083 g, 0.16 mmol)
in benzene (4.0 mL) at 0 ◦C was added drop-wise a solution of
potassium bis(trimethylsilyl)amide (0.50 M in toluene, 0.32 mL,
0.16 mmol). The crimson solution was allowed to warm to room
temperature for 10 min and was cooled to 0 ◦C. A solution of
a crude mixture of aldehyde (4S,5S)-4 and aldehyde (4R,5R)-4
[(4S,5S)-4 : (4R,5R)-4 = 7 : 1, 0.065 g, 0.16 mmol] in benzene
(2.0 mL) was added. The brown solution was allowed to warm
to room temperature and was stirred for 25 min. Dry methanol
(3.0 mL) was added followed by ammonium chloride solution (sat.
aq., 10 mL) and the crude product was extracted into ethyl acetate
(3 × 30 mL). The combined organic layers were dried (Na2SO4),
filtered and concentrated under reduced pressure to give an orange
residue (0.14 g). Flash chromatographic purification (15% ethyl
acetate : hexanes) of the crude product gave an inseparable
mixture of dienes 3 as a viscous, clear colourless liquid (0.056 g,
63%); Rf 0.30 (20% ethyl acetate : hexanes), 0.11 (10% ethyl
(4R,5S)-3-tert-Butyl-5-methyl-4-{(4S,5S)-5-[(E)-3-phenylprop-2-
enyl]-2,2-dimethyl-1,3-dioxolan-4-yl}-2,2-dimethyloxazolidine-
3,5-dicarboxylate (7E)-2 and (4R,5S)-3-tert-butyl-5-methyl-4-
{(4S,5S)-2,2-dimethyl-5-[(Z)-3-phenylprop-2-enyl]-1,3-dioxolan-
4-yl}-2,2-dimethyloxazolidine-3,5-dicarboxylate (7Z)-2
To a suspension of benzyltriphenylphosphonium chloride (0.10 g,
0.26 mmol) in benzene (6.0 mL) at 0 ◦C was added a solution of
potassium bis(trimethylsilyl)amide (0.50 M in toluene, 0.53 mL,
0.27 mmol). The bright orange solution was allowed to warm to
room temperature for 15 min and was cooled to 0 ◦C. A solution
containing a mixture of aldehyde (4S,5S)-4 and aldehyde (4R,5R)-
4 [(4S,5S)-4 : (4R,5R)-4 = 7 : 1, 0.11 g, 0.27 mmol] in benzene
(2.5 mL) was added. The brown solution was allowed to warm to
room temperature and was stirred for 1 h. Dry methanol (4.0 mL)
was added followed by ammonium chloride solution (sat. aq.,
3188 | Org. Biomol. Chem., 2007, 5, 3183–3189
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
acetate : hexanes); [a]1D7 +4.7 (c 1.4, CH2Cl2); mmax (thin film)/cm−1
3 S. Sasaki, Y. Hamada and T. Shioiri, Tetrahedron Lett., 1999, 40, 3187;
S. Sasaki, Y. Hamada and T. Shioiri, Tetrahedron Lett., 1997, 38, 3013.
4 S. Chandrasekhar and S. S. Sultana, Tetrahedron Lett., 2006, 47, 7255.
5 E. J. Corey, A. Guzman-Perez and M. C. Noe, J. Am. Chem. Soc., 1995,
117, 10805.
=
=
1763 (m), 1740 (m), 1701 (s, C O), 1605, 1510 (s, Ar(C C)),
1250, 1175, 1089 (C–O–C); HRMS (+ESI) calc. for C30H43O8N +
Na 568.2886, found 568.2871; m/z (+ESI) 584 ([M + K]+, 28%),
568 ([M + Na]+, 100), 512 (43). Preparative HPLC purification
(RTI Zorbax Sil, 10% ethyl acetate : hexanes) of a portion of
the purified product (0.029 g) gave pure (7E,9E)-diene (7E,9E)-3
(tR ≈33.0 min) as a viscous, clear, colourless liquid (0.020 g, 23%);
[a]1D8 −1.8 (c 1.8, CH2Cl2); dH(200 MHz; 340 K; CDCl3) 7.29–7.21
(2H, m, Ar-H), 6.85–6.78 (2H, m, Ar-H), 6.61 (1H, dd, J 15.5,
10.0 Hz, C9-H), 6.39 (1H, d, J 15.6 Hz, C10-H), 6.26 (1H, dd, J
15.1, 10.0 Hz, C8-H), 5.79 (1H, dt, J 14.8, 7.3 Hz, C7-H), 4.70
(1H, d, J 2.1 Hz, C2-H), 4.45 (1H, dd, J 5.1, 2.0 Hz, C3-H), 4.11–
3.94 (4H, m, C5-H, C4-H, OCH2CH3), 3.76 (3H, s, C1–OCH3),
2.52–2.31 (2H, m, C6-H), 1.61 (3H, s, NC(CH3)A(CH3)BO), 1.57
(3H, s, NC(CH3)A(CH3)BO), 1.50 (9H, s, OC(CH3)3), 1.42–1.35
(9H, m, OC(CH3)2O, OCH2CH3); dC (50 MHz; 340 K; CDCl3)
172.1, 158.7, 151.9, 133.7, 130.9, 130.5, 128.4, 127.5, 127.1, 115.0,
109.0, 96.8, 80.9, 80.1, 78.4, 75.3, 63.7, 61.4, 52.2, 36.5, 28.5, 27.7,
27.4, 27.2, 27.1, 14.8. A second fraction gave pure (7Z,9E)-diene
(7Z,9E)-3 (tR ≈29.0 min) as a viscous, clear, colourless liquid
(0.0067 g, 7.5%); [a]D18 +10.3 (c 0.57, CH2Cl2); dH (200 MHz; 340 K;
CDCl3) 7.36–7.32 (2H, m, Ar-H), 6.97–6.81 (3H, m, C9-H, Ar-
H), 6.49 (1H, d, J 15.9 Hz, C10-H), 6.26 (1H, t, J 11.0 Hz, C8-
H), 5.56 (1H, dt, J 10.9, 7.5 Hz, C7-H), 4.72 (1H, d, J 2.2 Hz,
C2-H), 4.49 (1H, dd, J 4.6, 1.9 Hz, C3-H), 4.15–4.00 (4H, m,
C5-H, C4-H, OCH2CH3), 3.75 (3H, s, C1–OCH3), 2.61 (2H, t,
J 6.2 Hz, C6-H), 1.63 (3H, s, NC(CH3)A(CH3)BO), 1.58 (3H, s,
NC(CH3)A(CH3)BO), 1.50 (9H, s, OC(CH3)3), 1.44–1.37 (9H, m,
OC(CH3)2O, OCH2CH3); dC (50 MHz; 340 K; CDCl3) 172.2,
158.1, 151.9, 132.9, 131.5, 130.5, 127.8, 125.6, 122.5, 115.0, 109.1,
96.9, 80.9, 80.3, 78.4, 75.2, 63.7, 61.5, 52.3, 31.8, 28.5, 27.8, 27.5,
27.2 (2C), 14.8.
6 M. Scholl, S. Ding, C. W. Lee and R. H. Grubbs, Org. Lett., 1999, 1,
953.
7 C.-Y. Chuang, V. C. Vassar, Z. Ma, R. Geney and I. Ojima, Chirality,
2002, 14, 151.
8 J. A. Bodkin and M. D. McLeod, J. Chem. Soc., Perkin Trans. 1, 2002,
2733.
9 Determined by chiral HPLC of the acetate derivative 15 using a
Chiralcel OD-H column (250 × 4 mm, Daicel), eluent iPrOH : hexane,
30 : 70, flow rate 0.5 mL min−1, UV detection at 270 nm; 15 tR
=
13.1 min, ent-15 tR = 44.2 min, 98.5 : 1.5 er. The identity of the
minor stereoisomer was confirmed by independent synthesis of the
enantiomer ent-15 from alkene 9 using (DHQ)2PHAL ligands.
10 The (S)-configuration of the C2 hydroxyl-bearing stereocentre of
alcohol 6 was confirmed by the formation of (R)- and (S)-Mosher’s
esters and application of the modified Mosher’s method; I. Ohtani, T.
Kusumi, Y. Kashman and H. Kakisawa, J. Am. Chem. Soc., 1991, 113,
4092.
11 Oxidation methods surveyed included: Dess–Martin periodinane,
Swern reagent with aqueous workup, Parikh–Doering reagent and
Swern reagent with anhydrous workup.
12 T. W. Greene and P. G. M. Wuts, Protective Groups in Organic Synthesis,
Wiley & Sons Inc., New York, 2nd edn, 1991.
13 D. B. Dess and J. C. Martin, J. Am. Chem. Soc., 1991, 113, 7277.
14 J. B. Baudin, G. Hareau, S. A. Julia and O. Ruel, Tetrahedron Lett.,
1991, 32, 1175.
15 P. R. Blakemore, W. J. Cole, P. J. Kocienski and A. Morley, Synlett,
1998, 26.
16 P. R. Blakemore, J. Chem. Soc., Perkin Trans. 1, 2002, 2563.
17 A. J. Schvartzapel, L. Zhong, R. Docampo, J. B. Rodriguez and E. G.
Gros, J. Med. Chem., 1997, 40, 2314.
18 D. L. Hughes, Org. React., 1992, 42, 335.
19 D. R. Williams, G. S. Cortez, S. L. Bogen and C. M. Rojas, Angew.
Chem., Int. Ed., 2000, 39, 4612.
20 M. Schwarz, G. F. Graminski and R. M. Waters, J. Org. Chem., 1986,
51, 260.
21 H. C. Kolb, M. S. VanNieuwenhze and K. B. Sharpless, Chem. Rev.,
1994, 94, 2483.
22 The (S)-configuration of the C5 hydroxyl-bearing stereocentre of
the major diol (4S,5S)-5 was confirmed by the selective formation
of the (R)- and (S)-Mosher’s esters at C5 and application of the
modified Mosher’s method (ref. 10). The signals derived from the major
diastereomer were used in each case allowing assignment of the absolute
stereochemistry at C5.
23 A. Dondoni, P. Merino and D. Perrone, Tetrahedron, 1993, 49, 2939.
24 K. Morikawa and K. B. Sharpless, Tetrahedron Lett., 1993, 34, 5575.
25 G. Cignarella, E. Occelli and E. Testa, J. Med. Chem., 1965, 8, 326.
Notes and references
1 C. A. Bewley, C. Debitus and D. J. Faulkner, J. Am. Chem. Soc., 1994,
116, 7631; E. W. Schmidt and D. J. Faulkner, Tetrahedron, 1998, 54,
3043; A. Qureshi, P. L. Colin and D. J. Faulkner, Tetrahedron, 2000,
56, 3679.
2 J. Zhu and D. Ma, Angew. Chem., Int. Ed., 2003, 42, 5348.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3183–3189 | 3189
©