Improved synthesis of functionalized molecular platforms related to marine cyclopeptides

Article abstract of DOI:10.1016/S0040-4020(00)01173-X

We report an improved synthesis of new molecular platforms, functionalized for studies in molecular recognition and combinatorial chemistry. Their structures are related to naturally occurring marine cyclopeptides such as the Dolostatins and Dendroamides that feature side chain functionality positioned on the same face of a rigid platform. Incorporation of methyl substituted oxazole and thiazole rings into the platforms enhances the rigidity of structure and significantly improves the yields of the macrocyclization reaction in the synthesis.

Full text of DOI:10.1016/S0040-4020(00)01173-X

TETRAHEDRON
Pergamon
Tetrahedron 57 (2001) 1699±1708
Improved synthesis of functionalized molecular platforms related
to marine cyclopeptides
p
 Â
Laszlo Somogyi, Gebhard Haberhauer and Julius Rebek, Jr.
The Skaggs Institute for Chemical Biology and Department of Chemistry, The Scripps Research Institute, La Jolla, CA 92037, USA
Received 1 September 2000; accepted 21 December 2000
AbstractÐWe report an improved synthesis of new molecular platforms, functionalized for studies in molecular recognition and com-
binatorial chemistry. Their structures are related to naturally occurring marine cyclopeptides such as the Dolostatins and Dendroamides that
feature side chain functionality positioned on the same face of a rigid platform. Incorporation of methyl substituted oxazole and thiazole rings
into the platforms enhances the rigidity of structure and signi®cantly improves the yields of the macrocyclization reaction in the synthesis.
q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
macrocycle the structure becomes rigid and nearly planar.
This rigidity is further stabilized by the intramolecular array
of hydrogen bond donors and acceptorsÐthe hydrogens of
the secondary amides and the lone pairs of the heterocyclic
nitrogensÐall directed to the center of the macrocycle.
There does not appear to be enough space in the center to
accommodate even a single water molecule, and solvation
of the amides must occur at the peripheral oxygens rather
than the inwardly directed hydrogens. This feature may
contribute to the transport properties of these compounds
across biological membranes.
In recent years there has been rapid progress in the ®eld of
isolation and characterization of natural marine cyclo-
peptides,1 especially those incorporating ®ve-membered
heterocyclic rings.2 These compounds are usually metabolic
products of primitive marine organisms, algae, fungi and
often possess signi®cant biological activities, including
cytotoxicity, antibacterial and antiviral activities.3 Examples
have also been found for their ability to overcome multi-
drug resistance or to act as antineoplastic agents.4
Dendroamide C (1),4a Bistatramide D (2),2b Westiellamide
(3)5 (Fig. 1) and numerous other natural cyclopeptides can
be identi®ed as biologically modi®ed hexapaptide deriva-
tives.3a,6 These modi®cations include not only macrocycli-
zation, but also heterocyclization of serine, threonine and
cysteine side chains onto the backbone carbonyl groups to
create ®ve-membered heterocyclic rings (e.g. oxazoles,
thiazoles and oxazolines).3a In cases where three aromatic
heterocycles are linked together with trans-amide bonds in a
The orientation of the side-chain substituents, of course,
depends on the con®guration at the a carbon atoms. All-
syn substituted analogues, especially with functionalized
side-chains [e.g. Dendroamide A, B, C (1)4a and Dolastatine
34b] are rare among the marine cyclopeptides characterized
to date. Accordingly, synthesis provides the best access
to these structures which might also play roles as core
molecules for solution phase combinatorial chemistry and
as templates for synthetic receptors.7 In the former role, the
macrocycle offers sizeable dimensions and three addresses.
These large surface areas are likely to be required for inter-
fering with protein±protein and protein±nucleic acid inter-
actions. In the latter role, the macrocycle's rigidity suggests
its use as a spacer element on which functional groups may
be presented, preorganized for selective molecular recogni-
tion of smaller molecular targets. These roles are by no
means unrelated, but require the ef®cient synthetic proce-
dures that we relate here.
Keywords: platforms; macrocyclization; oxazoles and thiazoles; combina-
torial chemistry.
Abbreviations: Boc: t-butyloxy-carbonyl; tBu: t-butyl; Bzl: benzyl; DAST:
(diethylamino)sulfur tri¯uoride; DBU: 1,8-diazabicyclo[5.4.0]undec-7-
È
ene; DCM: dichloromethane; DIEA (Hunig's base): N,N-diisopropylethyl-
amine; DMAP: 4-dimethylaminopyridine; DMF: N,N-dimethylformamide;
DPPA: diphenylphosphoryl azide; EDCI: 1-ethyl-3-(3-dimethylamino-
propyl) carbodiimide; FDPP: penta¯uorophenyl diphenylphosphinate;
HBTU: 2-(1H-benztriazole-1-yl)-1,1,3,3-tetramethyluronium hexa¯uoro-
phosphate; HOBt: 1-hydroxybenztriazole; PyBOP: benztriazole-1-yl-oxy-
tris-pyrrolidino-phosphonium hexa¯uorophosphate; TBAF: tetrabutyl-
ammonium ¯uoride hydrate; TBC: 2,4,6-trichlorobenzoyl chloride; TEA:
triethylamine; TFA: tri¯uoroacetic acid; THF: tetrahydrofuran; TMSI:
iodotrimethylsilane; Z: benzyloxycarbonyl.
Platform 4 (Fig. 2)Ðsynthesized earlier8Ðwas a promising
candidate for these purposes, but the macrocyclization step
ending the stepwise synthesis proved a limiting factor due to
the low yields (13±15%). This was puzzling because the
same attributes that contribute to the rigidity of the ®nal
p
Corresponding author. Tel.: 11-858-784-2250; fax: 11-858-784-2876;
e-mail: jrebek@scripps.edu
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)01173-X

1700
L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
Figure 1. Natural marine cyclopeptide platforms with all-syn side chains.
Figure 2. Side chain functionalized synthetic analogues with C3 symmetry.
structure should stabilize the transition state conformation
for the macrocyclization. One solution to the problem was
discovered by Wipf and coworkers,2f who fashioned the
®ve-membered heterocycles after the macrocyclization
step. Their improved yields suggested that the simple
oxazole or thiazole derivatives somehow resist macrocycli-
zation with respect to other processes. It seemed possible
that the fraction of the intermolecular coupling products
could be reduced by altering these heterocycles, by limiting
access of nucleophiles to the activated carboxyl of the
macrocyclic precursor. Accordingly, we synthesized new
core molecules with C3 symmetry (see general structure 5
in Fig. 2) featuring methyl substituted oxazole and thiazole
units in the macrocyclic ring. This did enhance the cycliza-
tion step and yields up to 69% were obtained. In addition, a
stepwise synthetic route to orthogonally protected cyclo-
trimers was developed in a way that the side-chain functions
can be addressed selectively.
2. Results and discussion
The synthetic approach hardly merits the term `retrosynthetic
disconnection' but Scheme 1 outlines the heterocyclic
modules and the suitably protected acylamino threonine
derivatives that lead ultimately to 5. Many other isosteric
heterocyles such as oxadiazoles can be considered and have
recently been made from appropriate acylamino oxalic acid
derivatives.9 Starting from two amino acid components
(6a,b and 7a,b see Scheme 2) we coupled them to form
dipeptides (8a or b) through conventional peptide proce-
dures. (For the synthesis of compound 7a see Section 4.
The other amino acid residues are commercially available.)
The yields for the amide bond formation reactions were
anticipated and found to be quite high (90±95%).
The heterocyclic ring was fashioned in two different ways.
The ®rst involved the cyclodehydration of dipeptide 8a to
Scheme 1. Retrosynthetic disconnection of general platform 5.

L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
1701
Scheme 2. Synthesis of modules: (i) HBTU, HOBt, Et3N, DMF, 2208C to rt, 20 h, 92% for 6a and 7a; DPPA, DIEA, DMF, 08C to rt, 20 h, 95% for 6b and 7b;
(ii) DAST, CH2Cl2, 2788C, 2 h; (iii) BrCCl3, DBU, CH2Cl2, 08C to rt, 20 h, 23% for (ii) and (iii); (iv) Dess±Martin periodinane, CH2Cl2, 08C to rt, 4±6 h;
(v) PPh3, I2, Et3N, CH2Cl2, 08C to rt, 1 h, 46±65% for (iv) and (v); (vi) Lawesson's reagent, THF, re¯ux, 5 h, 54±65% for (iv) and (vi).
the appropriate oxazoline derivative 9a by (diethylamino)
sulfur tri¯uoride (DAST)10 at 2788C in 2 h. This was
followed by oxidation with BrCCl3 in the presence of
DBU11 to afford subunit 11a in 23% yield.
inane12 to form amidoketone derivatives 10a or b. These
were transformed without puri®cation into either the
oxazole or the thiazole modules (11a±d) using PPh3 in the
presence of I2 and Et3N13 or Lawesson's reagent,14 respec-
tively. The yields for the protected modules, calculated from
the dipeptides, were considerably higher (46±65%) by this
route.
AlternativelyÐand more successfullyÐthe dipeptides
were subjected to oxidation using the Dess±Martin period-
Scheme 3. One-pot cyclization of modules: (i) Zn, 90% aqueous AcOH, rt, 5 h, quant., for 11a and c; 2 M aqueous NaOH, MeOH/dioxane, 08C to rt, 20 h,
quant., for 11b and d; (ii) TFA, CH2Cl2, 1/2±3 h, quant.; (iii): see Table 1; (iv): H2, 10%Pd±C, EtOAc, 20 h, quant. for 14a; H2, Pearlman's catalyst, EtOH/
EtOAc, 2 days, 90% for 14b; (CH3)3SiI, CHCl3, 408C, 12 h, for 14c and d.

1702
L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
Table 1. Reaction conditions for the one-pot cyclization of amino acid modules
Monomer
Coupling reagent
Concentration (mol/l)
Solvent
Reaction time (days)
Yield (%)
Trimer (14a±d)
Tetramer (15a±d)
13a
DPPA
FDPP
0.01
0.05
0.05
0.05
0.02
0.05
0.02
DMF
ACN
DMF
DMF
DMF
DMF
DMF
7
7
7
14
4
14
5
18
25
24
38
35
36
22a
n/i
n/i
n/i
24
n/i
16
n/i
PyBOP
PyBOP
DPPA
PyBOP
DPPA
13b
13c
13d
In a general cyclization procedure 2 equiv. of coupling reagent was added to 1 equiv. of monomer in the presence of 10 equiv. of DIEA; n/i: not isolated.
7% of syn±syn±anti diastereomer was also isolated.
a
Two methods for the macrocyclization of the monomeric
modules (11a±d) were examined: a `one-pot' procedure and
the stepwise route. In Scheme 3 and in Table 1 we summar-
ize our results from one-pot reactions. These involved
deprotection of the carboxyl and the amino terminus of
monomeric compounds 11a±d and afforded the TFA salts
of the appropriate amino acid modules 13a±d. These under-
went cyclooligomerization upon treatment with common
peptide coupling reagents (PyBOP, FDPP, DPPA) in the
presence of HuÈnig's base over a time span of 4±14 days.
These ®ndings are in agreement with the data from other
laboratories for similar, non-functionalized oxazoline5 and
thiazole15 compounds.
Nevertheless, there were some differences worth pointing
out. First, the yields of the enantiomerically pure, func-
tionalized cyclotrimers (14a±d) are usually high enough
(35±38%) to afford these macrocycles on a gram scale.
Second, the cyclotrimers can easily be separated by column
chromatography on silica gel from cyclotetramers (15a and
c) or from any other by-products, including diastereomeric
cyclotrimers (see Table 1). The formation of the syn±syn±
anti diastereomer beside the usually isolated all-syn
compound (14d) is most likely the result of partial epimer-
ization at the chiral center during the methyl ester deprotec-
tion of building block 11d under basic conditions (see
Scheme 3). Bzl protecting groups were removed under
Scheme 4. Stepwise procedure for the cyclization of modules: (i) TFA, CH2Cl2, 1/2±3 h, quant.; (ii) HBTU, HOBt, Et3N, 12a, DMF, 2208C to rt, 20 h, 64%
for 16a; PyBOP, DIEA, 12b, DMF, rt, 24 h, 72% for 16b; (iii) HBTU, HOBt, DIEA, 12a, DMF, 2208C to rt, 20 h, 64% for 17a; PyBOP, DIEA, 12b, DMF, rt,
40 h, 89% for 17b; (iv) Zn, 90% aqueous AcOH, rt, 5 h, quant., for 18a; 2 M aqueous NaOH, MeOH/dioxane, 08C to rt, 20 h, quant., for 18b; (v) PyBOP,
DIEA, rt, 4 days, 69% of 14a for 18a; same conditions but 15% of 14b1enantiomer and 26% of 191enantiomer for 18b; (vi) H2, 10%Pd±C, EtOAc, 20 h;
(vii) H2, Pearlman's catalyst, EtOH/EtOAc, 2 days.

L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
1703
catalytic hydrogenolysis to afford oxazole containing plat-
forms 5a and 5b, and using TMSI for platforms 5c and 5d
incorporating thiazole rings.
Doubtless larger cyclooligomers are formed during the
one-pot reactions,17 but we have not attempted their
isolation.
Third, substantial amounts of cyclotetramers were recovered
from the reaction mixtures only if longer reaction times were
used. (Table 1) The formation of cyclic products most likely
follows a stepwise fashion through linear dimers and trimers;
tetramers can arise by both 311 and 212 condensations.
Cyclotetramers incorporating ®ve-membered heterocycles
into the macrocycle are also well known among natural
cyclopeptides.16 The interactions between the lone pairs
of the oxazole or thiazole nitrogens and the hydrogens of
the secondary amidesÐtheir `intramolecular hydrogen
bonds'Ðare probably weaker than those of cyclotrimers.
The fourth observation concerns the salubrious effect of the
additional methyl substituent on the oxazole or thiazole
heterocycles. A similar example from the literature is the
one-pot synthesis of the natural trioxazoline derivative
Westiellamide (3)5. The authors were able to isolate not
only the cyclotrimer, but also the cyclotetramer in fairly
high yields, 20 and 25%, respectively. The extra methyl
group attached to the oxazoline ring may have contributed
to these yields. We tested the effect of the methyl group on
cyclization through a module made from a serine derivative
instead of threonine. The protective groups were the same as
of 11a. After the deprotection of the carboxyl and the amino
functions (see Scheme 3) the cyclooligomerization was
performed using PyBOP in the presence of HuÈnig's base
in DMF (0.05 M) for 14 days. The Bzl-esters of both the
appropriate cyclic trimer 4 and cyclic tetramer were each
1
The H NMR spectra show the doublets of the amide NH
resonances shifted about 0.5 ppm up®eld of those of the
cyclotrimers. The structures of cyclotetramers also appear
to be more ¯exible, so these molecules are less desirable
targets to serve as e.g. templates for synthetic receptors.
Scheme 5. Synthesis of orthogonally protected platform: (i) H2, 10% Pd±C, EtOAc, 20 h; (ii) TBC, Et3N, THF, CH3C(O)CH2OH, DMAP, CH2Cl2, 83% for
20a; TBC, Et3N, THF, tBu±OH, DMAP, CH2Cl2, 52% for 20b; (iii) Zn, 90% aqueous AcOH, rt, 5 h, quant., (iv) TFA, CH2Cl2, 1/2±3 h, quant.; (v) HBTU,
HOBt, DIEA, 21a, DMF, 2208C to rt, 4days, 68%; (vi) HBTU, HOBt, DIEA, 21b, DMF, 2208C to rt, 4 days, 32%; (vii) PyBOP, DIEA, rt, 10 days, 24%
calculated from linear trimer 22; (viii) TBAF, THF, rt, 80%.

1704
L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
isolated in 12% yield. These yields are substantially lower
than those from 13a (38% for cyclic trimer 14a and 24% for
cyclic tetramer 15a). The cyclization-inducing effect of the
`extra' methyl substituent is apparent.
3. Conclusions
The synthetic molecular platforms described above have
features that can make them ideal candidates for studies in
molecular recognition and combinatorial chemistry. Besides
the fairly rigid, almost planar heterocyclic ring-systemÐ
found also in natural analoguesÐthey possess functional-
ized side-chains on the same face of the molecule. Investi-
gations revealing their behavior as core molecules are
currently underway in our laboratories. Our experiments
have shown that both the one-pot and the stepwise synthetic
routes can afford these functionalized molecular platforms
on a gram scale. The incorporation of methyl substituted
heterocycles into the macrocyclic ring was proven to
enhance the yield of the macrocyclization step. The addi-
tional methyl group presumably helps to preorganize the
structure of the linear oligomer compound into a sickle
shape better suited for cyclization. The selectively addres-
sable, side-chain functionality opens the way to diversi®ca-
tion through combinatorial synthesis.
Further to this point, we also synthesized triacid platform 5a
and trialcohol platform 5b by the longer stepwise route.
(Scheme 4) The oxazole modules 11a and 11b were de-
protected separately on the carboxyl and the amino terminus
to give free acid monomers 12a and 12b and TFA salts of
16a and 16b, respectively. These monomers were coupled
to form protected dimers 17a and 17b again using conven-
tional peptide coupling procedures. Boc cleavage and
coupling with acid modules (12a and b) were repeated to
afford enantiomerically pure linear trimers 18a and 18b.
Protecting groups were removed from the carboxyl and
amino termini and the linear trimer modules were subjected
to cyclization with PyBOP and HuÈnig's base in DMF.18 The
reactions were worked up after 4 days and the crude
mixtures were puri®ed on silica gel. The enantiomerically
pure tri-benzyl ester platform 14a was isolated in 69% yield
from 18a. From 18b we found 26% of the diastereomeric
compounds (19 and its enantiomer) beside 15% of the
symmetric macrocycles (14b and its enantiomer). The
probable cause of diastereomerization again is the
epimerization at the chiral centers during methyl ester
deprotection under basic conditions. This epimerization
could be eliminated by only using protective groups
removable under neutral or acidic conditions (see module
11a) or by replacing the hydrogen at the chiral center with a
methyl group.19
4. Experimental
4.1. Spectroscopy, puri®cations and abbreviations
All reactions were performed under N2 atmosphere and
NMR spectra were recorded on a Bruker DRX 600 spectro-
meter; the chemical shifts are given in ppm relative to TMS.
The spectra were referenced to deuterated solvents indicated
in brackets in the analytical data. High resolution MALDI-
FTMS measurements were performed on an IonSpec FTMS
mass spectrometer using 2,5-dihydroxybenzoic acid matrix.
The electrospray ionization (ESI) mass spectrometry experi-
ments were performed on an API 100 Perkin±Elmer SCIEX
single quadrupole mass spectrometer. TLC: silica gel plates
IB2-F; J. T. Baker Column chromatography: silica gel 60;
230±400 mesh; E. Merck KGaA. Preparative chromato-
graphic plates: silica gel 60 F254 (20£20 cm2, 0.5 mm).
Anhydrous solvents and chemicals were used as purchased
from commercial suppliers.
There is a striking difference between the yields of cycliza-
tion for the oxazole platform 5a (69%) and platform 4 (13±
15%)8 lacking the extra methyl groups. Whether this effect
involves preorganization to a sickle shape appropriate for
cyclization or prevention of higher oligomers through steric
hindrance of external nucleophiles is unknown.
Addressable side chains are available through the methods
outlined in Scheme 5. From module 11a the orthogonally
protected triester platform 23 was assembled. First, modules
20a and 20b were prepared by removing the Bzl ester group
under catalytic hydrogenolysis and re-protecting the free
side-chain carboxyl with acetol20 and tert-butyl groups,
respectively, using Yamaguchi's procedure.21 These new
modules were transformed to the free acid derivatives
(21a and 21b) then coupled to the amino building block
16a. The protected linear trimer compound 22 (Scheme 5)
was obtained through coupling procedures described
previously (Scheme 4). Deprotection of the trichloroethyl
ester afforded the free acid derivative as usual, but cleavage
4.2. General procedures
(1) Cleavage of the trichloroethyl-ester group. To a stirred
solution of protected compound (1 equiv.) in glacial acetic
acid/waterˆ9:1 (0.01 M) Zn powder (20±50 equiv.) was
added slowly at rt. The mixture was stirred until TLC
showed the consumption of all starting material (usually
5 h), then ®ltered and concentrated in vacuo. The residue
was taken into EtOAc, extracted with 5% HCl solution,
water and brine, then dried over MgSO4 and concentrated
to give the pure acid residue. (2) Cleavage of the methyl-
ester group. Protected compound (0.05 M) was dissolved in
a mixture of methanol/dioxaneˆ10:7 followed by the slow
addition of 2 M NaOH solution (10 equiv.) at 08C. Stirring
was continued for 20 h and the reaction mixture was
allowed to warm slowly to rt. The organic solvents were
removed in vacuo followed by addition of water and 1 M
HCl solution. The aqueous phase was repeatedly extracted
with DCM, the organic layers were combined and dried over
MgSO4, then concentrated in vacuo to give the pure acid
compound. (3) Cleavage of the Boc group. Protected
t
of the Boc group under acidic conditions caused the Bu-
ester to be partly deprotected as well. (According to the
1
integration of the H NMR amide NH signals the ratio
between the desired compound and the over-deprotected
one was 2:1.) Cyclization in DMF with PyBOP activation
in the presence of DIEA for 10 days yielded the ortho-
gonally protected triester platform 23 in 24% from protected
linear trimer 22. The carboxyl-protecting groups were
removed step-by-step using TBAF, TFA and catalytic
hydrogenolysis to give mono- (24), di- (25) and triacid
(5a) platforms, respectively.

L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
1705
compound was dissolved in anhydrous DCM (0.1 M) and
TFA (50±100 equiv.) was added slowly at rt. Stirring was
continued until TLC showed the consumption of all starting
material (usually 0.5±3 h). The mixture was concentrated in
vacuo to yield the TFA salt of the pure deprotected amino
compound. (4) Peptide coupling conditions. To a stirred
solution of acid compound (1 equiv.) and the TFA salt of
the amino compound (1.2 equiv.) in DMF (0.2 M) were
added HBTU (1.2 q) and HOBt (1.2 equiv.) at 2208C,
followed by the slow addition of TEA or DIEA
(3.6 equiv.). Stirring was continued for 20 h while the
mixture was allowed to warm to rt. Solvent was evaporated
and the residue was dissolved in EtOAc, then extracted with
5% aqueous HCl, saturated NaHCO3 solution and brine. The
organic layer was dried over MgSO4 or Na2SO4 and concen-
trated in vacuo. Puri®cation was accomplished by chromato-
graphy on silica gel with an EtOAc/hexanes mixture as
eluent.
128.6; 128.4; 79.7; 73.7; 71.1; 68.3; 58.6; 55.4; 52.4;
28.6; 20.6. MALDI-FTMS [M1Na]1: calculated:
433.1945; observed: 433.1961.
4.2.4. Synthesis of oxazole building block 11a. (1) To a
stirred solution of dipeptide 8a (1 equiv.) in DCM (0.05 M),
DAST reagent (1.1 equiv.) was added slowly at 2788C. The
mixture was stirred for 2 h and poured into saturated K2CO3
solution at 08C and the aqueous phase was extracted with
CHCl3. The organic layer was dried over Na2SO4 and
concentrated in vacuo to give oxazoline compound 9a.
The oxazoline 9a was dissolved in anhydrous DCM (at
0.1 M) at 08C and DBU (,1.2 equiv.) was added followed
by the slow addition of BrCCl3 (,1.2 equiv.). The mixture
was stirred for 20 h and slowly warmed to rt, then extracted
with saturated NH4Cl solution and brine. The organic layer
was dried over MgSO4 and concentrated in vacuo. Pure
oxazole building block 11a was obtained in 23% after
chromatography on silica gel with EtOAc/hexanesˆ1:4.
(2) To a stirred solution of dipeptide 8a (1 equiv.) in
DCM (0.1 M) Dess±Martin periodinane (1.5 equiv.) was
added at 08C. The ice bath was removed after 30 min and
the mixture was stirred for four more hours at rt. To this
mixture saturated Na2S2O3 and saturated NaHCO3 solutions
were added and stirring was continued for 45 min. The
phases were separated, the organic layer was washed with
brine, dried over Na2SO4 then concentrated in vacuo to
afford amidoketon 10a which was used without further puri-
®cation. To the stirred solution of 10a (0.05 M) in DCM,
triphenylphosphine (2 equiv.), iodine (2 equiv.) and TEA
(4 equiv.) were added at 08C. The cooling bath was removed
after 20 min and stirring was continued at rt for 1 h. To this
mixture saturated Na2S2O3 solution was added and stirring
was continued overnight. The phases were separated and the
organic layer was diluted with Et2O. After ®ltration of the
precipitate, the ®ltrate was concentrated in vacuo. Pure
oxazole 11a was obtained in 46% after chromatography
4.2.1. Synthesis of the TFA salt of H-(l)-Thr-OCH2CCl3
(7a). To a stirred solution of Boc-(l)-Thr-OH (1 equiv.) in
CH2Cl2 (0.3 M) were added CCl3CH2OH (1.5 equiv.), EDCI
(1.2 equiv.) and HOBt (1.2 equiv.), followed by the slow
addition of DIEA (3.6 equiv.) at 2208C. The mixture was
stirred overnight and allowed it to warm up to rt. Solvent
was evaporated and the residue was partitioned between
EtOAc and 5% aqueous HCl solution. The organic layer
was further extracted with saturated NaHCO3 solution and
brine, then dried over MgSO4 and concentrated in vacuo to
afford Boc-(l)-Thr-OCH2CCl3 in 90%. The Boc group was
removed as described above to give the TFA salt of
compound 7a.
4.2.2. Synthesis of Boc-[(l)-Asp(OBzl)-(l)-Thr]-
OCH2CCl3 (8a). General procedure for peptide coupling
was used. Acid compound: Boc-(l)-Asp(OBzl)-OH (6a),
amino compound: H-(l)-Thr-OCH2CCl3 (7a), puri®cation:
EtOAc/hexanesˆ1:2 to yield pure dipeptide 8a as a yellow
1
on silica gel with EtOAc/hexanesˆ1:4. H NMR (acetone-
1
oil in 92%. H NMR (acetone-d6): 7.48 (d, Jˆ8.8 Hz, 1H,
d6): 7.38±7.31 (m, 5H); 6.69 (d, Jˆ8.6 Hz, 1H, NH); 5.28
2
NH); 7.41±7.32 (m, 5H); 6.57 (d, Jˆ8.2 Hz, 1H, NH); 5.15
(AB, Jˆ13.3 Hz, 1H); 5.13 (AB, Jˆ13.3 Hz, 1H); 4.91
(AB, Jˆ12.1 Hz, 1H); 4.85 (AB, Jˆ12.1 Hz, 1H); 4.66
(m, 1H); 5.14 (s, 2H); 5.08 (s, 2H); 3.17 (dd, Jˆ16.3 Hz,
2
3
3Jˆ6.8 Hz, 1H); 3.05 (dd, Jˆ16.3 Hz, Jˆ6.8 Hz, 1H);
2.64 (s, 3H); 1.42 (s, 9H). 13C NMR (acetone-d6): 170.6;
162.6; 160.6; 158.7; 155.9; 137.3; 129.3; 128.9; 128.8;
127.3; 96.1; 79.9; 74.4; 66.9; 46.5; 38.2; 28.6; 12.6.
MALDI-FTMS [M1Na]1: calculated: 557.0620; observed:
557.0634.
2
3
(m, 1H); 4.60 (dd, Jˆ8.8 Hz, Jˆ2.5 Hz, 1H); 4.47 (m,
2
1H); 4.33 (d, Jˆ4.6 Hz, 1H, OH); 2.98 (dd, Jˆ16.5 Hz,
2
3
3Jˆ5.7 Hz, 1H); 2.84 (dd, Jˆ16.5 Hz, Jˆ7.4 Hz, 1H);
1.42 (s, 9H); 1.22 (d, Jˆ6.3 Hz, 3H). 13C NMR (acetone-
d6): 172.2; 171.5; 170.0; 156.5; 137.3; 129.2; 128.9;
95.9; 80.0; 75.0; 67.9; 66.9; 58.6; 52.1; 36.8; 28.6; 20.7.
MALDI-FTMS [M1Na]1: calculated: 577.0882; observed:
577.0886.
4.2.5. Synthesis of oxazole building block 11b. As
described for 11a. Puri®cation of 11b was accomplished
1
on silica gel with EtOAc/hexanesˆ1:5. Yield: 65%. H
NMR (acetone-d6): 7.34±7.24 (m, 5H); 6.49 (d, Jˆ8.3 Hz,
1H); 5.05 (m, 1H); 4.56 (m, 2H); 3.87 (m, 2H); 3.82 (s, 3H);
2.57 (s, 3H); 1.42 (s, 9H). 13C NMR (acetone-d6): 163.2;
161.4; 157.2; 156.1; 139.3; 129.1; 128.5; 128.4; 128.3; 79.7;
73.5; 71.1; 51.8; 49.9; 28.6; 12.1. MALDI-FTMS
[M1Na]1: calculated: 413.1683; observed: 413.1681.
4.2.3. Synthesis of Boc-[(l)-Ser(Bzl)-(l)-Thr]-OCH3 (8b).
To
a stirred solution of Boc-(l)-Ser(Bzl)-OH (6b)
(1 equiv.), the HCl salt of H-(l)-Thr-OCH3 (7b)
(1.5 equiv.) and DIEA (1.5 equiv.) in DMF (0.3 M) was
added DPPA (1.5 equiv.) at 08C followed by slow addition
of 3.8 equiv. DIEA. The work-up was as described in
Section 4.2. Puri®cation: EtOAc/hexanesˆ1:2. Yield of
4.2.6. Synthesis of thiazole building block 11c. As
described above. Amidoketone compound 10a was
dissolved in anhydrous THF (0.1 M) and Lawesson's
reagent (1.5 equiv.) was added. The mixture was stirred
under re¯ux for 5 h. The solvent was removed and the resi-
due was dissolved in EtOAc, extracted with water, 5% HCl
1
8b as a yellow oil: 95%. H NMR (acetone-d6): 7.44±7.24
(m, 6H); 6.25 (s, 1H); 4.58 (m, 2H); 4.50 (m, 1H); 4.42 (m,
1H); 4.31 (m, 1H); 4.17 (m, 1H); 3.84 (m, 1H); 3.74 (m,
1H); 3.68 (s, 3H); 1.43 (s, 9H); 1.15 (d, Jˆ6.6 Hz, 3H). 13C
NMR (acetone-d6): 171.8; 171.3; 156.4; 139.3; 129.1;

1706
L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
solution, saturated NaHCO3 solution and brine, then dried
over MgSO4 and concentrated in vacuo. Thiazole containing
building block 11c was puri®ed by chromatography on
154.8; 129.28; 46.6; 38.0; 11.6. MALDI-FTMS [M1Na]1:
calculated: 611.1344; observed: 611.1338.
1
silica gel with EtOAc/hexanesˆ1:5. Yield: 54%. H NMR
4.2.9. Synthesis of trioxazole trialcohol platform 5b. (1)
The one-pot procedure. Module 11b was successively
subjected to methyl and Boc deprotection (see Section
4.2) to give amino acid compound 13b. To the stirred solu-
tion of this module at 0.02 M in anhydrous DMF (0.02 M)
was added DIEA (1 equiv.), DPPA (3 equiv.) followed by
the slow addition of DIEA (7.5 equiv.) at 08C. The mixture
was stirred for 4 days slowly warmed to rt. It was worked-up
as in the case of general peptide couplings. Puri®cation was
accomplished on silica gel using EtOAc/hexanesˆ1:1 to
give 14b in 35%. Platform 14b was dissolved in a mixture
of EtOH/EtOAcˆ1:1 (0.01 M) and palladium hydroxide
(0.45 g/mmol starting material) was added. The mixture
was stirred under an H2 atmosphere for 2 days, ®ltered
through celite and concentrated in vacuo to give the pure
triol 5b in 90% yield. (2) The stepwise procedure. Oxazole
11b was ®rst subjected to Boc deprotection (see Section 4.2)
to give the TFA salt of amino compound 16b. Peptide
coupling of 16b (1 equiv.) with acid monomer 12b
(1 equiv.) was carried out using PyBOP (1.2 equiv.) activa-
tion in the presence of DIEA (3.8 equiv.) in anhydrous DMF
(0.06 M) at rt for 24 h. Work-up as for general peptide
coupling. Pure protected oxazole dimer 17b was isolated
in 72% yield after column chromatography on silica gel
with EtOAc/hexanesˆ1:2. Boc deprotection and coupling
with acid monomer 12b were repeated to give protected
trimer compound 18b (89%). After methyl and Boc depro-
tections (see Section 4.2) the residue was dissolved in anhy-
drous DMF (0.01 M), PyBOP (2.2 equiv.) and DIEA
(7.6 equiv.) were added at rt and the mixture was stirred
for 4 days. Work-up and puri®cation was the same as in
the one-pot reaction. Symmetric, protected cyclotrimer plat-
forms (14b and its enantiomer) were isolated in 15% yield;
diastereomeric asymmetric platforms (19 and its enantio-
mer) were isolated in 26%. Deprotection of benzyl esters
on platform 14b was carried out as described above to yield
platform 5b. 1H NMR (DMSO-d6): 8.34 (d, Jˆ6.1 Hz, 3H);
5.20 (t, Jˆ5.9 Hz, 3H); 5.07 (m, 3H); 3.93 (m, 3H); 3.86 (m,
3H); 2.61 (s, 9H). 13C NMR (DMSO-d6): 159.7; 159.6;
153.4; 127.9; 61.4; 50.6; 11.3. MALDI-FTMS [M1Na]1:
calculated: 527.1497; observed: 527.1473.
(acetone-d6): 7.37±7.34 (m, 5H); 6.91 (d, Jˆ8.3 Hz, 1H,
NH); 5.35 (m, 1H); 5.15 (m, 2H); 5.09 (m, 2H); 3.25 (dd,
2Jˆ16.2 Hz, 3Jˆ6.1 Hz, 1H); 3.08 (dd, 2Jˆ16.2 Hz,
3Jˆ7.4 Hz, 1H); 2.78 (s, 3H); 1.43 (s, 9H). 13C NMR (ace-
tone-d6): 170.9; 169.8; 160.9; 156.0; 147.8; 140.6; 137.3;
129.3; 128.9; 96.3; 80.1; 74.7; 67.0; 50.6; 39.0; 28.6; 13.6.
MALDI-FTMS [M1Na]1: calculated: 573.0391; observed:
573.0364.
4.2.7. Synthesis of thiazole building block 11d. As
described for 11c, with puri®cation on silica gel with
1
EtOAc/hexanesˆ1:4. Yield: 65%. H NMR (acetone-d6):
7.36±7.24 (m, 5H); 6.72 (d, 1H); 5.11 (m, 1H); 4.58 (m,
2H); 3.91 (m, 2H); 3.83 (s, 3H); 2.71 (s, 3H); 1.43 (s, 9H).
13C NMR (acetone-d6): 168.3; 163.5; 156.3; 145.5; 141.9;
139.3; 129.1; 128.5; 128.4; 79.8; 73.6; 72.2; 54.1; 51.9;
28.6; 13.1. MALDI-FTMS [M1Na]1: calculated:
429.1460; observed: 429.1449.
4.2.8. Synthesis of trioxazole triacid platform 5a. (1) The
one-pot procedure. Module 11a was successively subjected
to trichloroethyl and Boc deprotection (Section 4.2) to give
amino acid compound 13a. To the stirred solution of this
module at 0.05 M in anhydrous DMF was added DIEA
(1 equiv.) and PyBOP (2 equiv.), followed by the slow addi-
tion of DIEA (9 equiv.) at rt. The mixture was stirred for
14 days, and was worked-up as described for general
peptide coupling. Puri®cation was accomplished on silica
gel using EtOAc/hexanesˆ2:3. The isolated benzyl-
protected 14a was further puri®ed by crystallization from
EtOAc upon addition of hexanes to form a white powder in
38% yield. Cyclotetramer compound 15a was also separated
by column chromatography in 24% as a yellow oil.
Protected platform 14a was dissolved in EtOAc (0.01 M),
Pd±C (10%) (0.1 g/mmol starting material) was added, and
the mixture was stirred for 20 h at rt under H2. The mixture
was ®ltered through Celitew521 (Aldrich), and concen-
trated in vacuo to give pure trioxazole triacid platform 5a
quantitatively. (2) The stepwise procedure. The oxazole
building block 11a was ®rst subjected to Boc deprotection
(see Section 4.2) to give the TFA salt of amino compound
16a. Peptide coupling with acid monomer 12a was carried
out as usual (see Section 4.2). Pure protected oxazole dimer
17a was isolated in 64% after column chromatography on
silica gel with EtOAc/hexanesˆ1:2. Boc deprotection,
coupling with acid monomer 12a and puri®cation were
repeated to give protected trimer compound 18a again in
64% yield. After trichloroethyl and Boc deprotections of
compound 18a (see Section 4.2) the residue was dissolved
in anhydrous DMF (0.005 M), PyBOP (2 equiv.) and DIEA
(10 equiv.) were added at rt and the mixture was stirred for
4 days. Work-up and puri®cation procedures were the same
as in the case of the one-pot reaction. Pure, protected cyclo-
trimer platform 14a was isolated in 69%. Deprotection of
benzyl esters was carried out as described above to yield
4.2.10. Synthesis of trithiazole triacid platform 5c. The
applied one-pot procedure using module 11c was the same
as in the case of building block 11a and gave protected
cyclotrimer 14c in 36% and cyclotetramer 15c in 16%.
For Bzl-ester deprotection, compound 14c was dissolved
in anhydrous CHCl3 (0.01 M) and (CH3)3SiI (20 equiv.)
was added at 408C. The mixture was stirred for 20 h then
quenched with water. The organic solvent was removed in
vacuo and the residue was diluted with 10% NaHSO3 solu-
tion. This mixture was extracted with EtOAc several times
and the combined organic layers were washed with satu-
rated NaHCO3 solution. The basic aqueous phase was acid-
i®ed with 5% KHSO4 solution to pH,3±4 and re-extracted
with EtOAc. The organic layer was washed with brine
1
1
pure triacid platform 5a. H NMR (acetone-d6): 8.45 (d,
then concentrated in vacuo to give pure trithiazole 5c. H
2
Jˆ5.1 Hz, 3H, NH); 5.33 (m, 3H); 3.19 (dd, Jˆ16.1 Hz,
NMR (acetone-d6): 8.77 (d, Jˆ8.6 Hz, 3H, NH); 5.84 (m,
3H); 3.10 (dd, 2Jˆ16.4 Hz, 3Jˆ4.8 Hz, 3H); 2.95 (dd,
2Jˆ16.4 Hz, 3Jˆ8.7 Hz, 3H); 2.79 (s, 9H). 13C NMR
2
3
3Jˆ5.9 Hz, 3H); 3.16 (dd, Jˆ16.1 Hz, Jˆ4.3 Hz, 3H);
2.61 (s, 9H). 13C NMR (acetone-d6): 171.0; 161.4; 161.2;

L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
1707
(acetone-d6): 171.7; 165.1; 161.7; 142.7; 48.0; 42.1; 12.6.
MALDI-FTMS [M1Na]1: calculated: 637.084; observed:
637.0833.
cleavage (see Section 4.2) to give acids 21a and 21b, respec-
tively. These were coupled to amino compound 16a in a
stepwise fashion as in the synthesis of platform 5a. The
amounts of reagents used were the following: 21a
(1 equiv.), 16a (1.2 equiv.), HBTU (2 equiv.), HOBt
(2 equiv.), DIEA (6 equiv.) in anhydrous DMF (0.1 M) for
4 days. The work-up procedure was the same as for general
peptide coupling reactions. Puri®cation was accomplished
on silica gel with EtOAc/hexanesˆ1:2 to give the protected
dimer in 68%. After Boc deprotection of the dimer
compound (1 equiv.) (see Section 4.2) acid building block
21b was coupled using the same procedure as previously.
Completely protected linear trimer 22 was isolated in 32%
yield after column chromatography on silica gel with
EtOAc/hexanesˆ1:1. The trichloroethyl ester group was
removed using the general procedure, whereas the following
Boc cleavage was carried out in DCM (0.03 M) using only
20 equiv. of TFA for 1 h at rt. The solvent was evaporated
and the residue was dissolved in anhydrous DMF (0.01 M),
PyBOP (2 equiv.) and DIEA (10 equiv.) were added at rt
and the mixture was stirred for 10 days. The work-up was
the same as in the general case for peptide couplings and
puri®cation was accomplished on a preparative chromato-
graphic plate using EtOAc/hexanesˆ2:1. Pure orthogonally
protected platform 23 was isolated in 24% (from linear
4.2.11. Synthesis of trithiazole trialcohol platform 5d.
The one-pot procedure was applied on module 11d as in
the case of 11b. The pure protected cyclotrimer 14d was
obtained in 22% yield. (The diastereomeric platform with
syn±syn±anti side-chain con®guration was also separated in
7% yield.) For Bzl-ester deprotection, compound 14d was
dissolved in anhydrous CHCl3 (0.01 M) and (CH3)3SiI
(20 equiv.) was added at 408C. The mixture was stirred
for 30 h then quenched with water. The organic solvent
was removed in vacuo and the residue was diluted with
10% NaHSO3 solution. This mixture was extracted with
EtOAc several times and the combined organic layers
were washed with saturated NaHCO3 solution, brine and
concentrated in vacuo to give pure trithiazole trialcohol
1
platform 5d in quantitative yield. H NMR (CDCl3/ace-
tone-d6ˆ2:1): 8.60 (d, Jˆ8.0 Hz, 3H, NH); 5.44 (m, 3H);
3.96 (dd, 2Jˆ10.7 Hz, 3Jˆ4.8 Hz, 3H); 3.77 (dd, 2Jˆ
3
10.7 Hz, Jˆ6.9 Hz, 3H); 2.79 (s, 9H). 13C NMR (CDCl3/
acetone-d6ˆ2:1): 163.2; 161.7; 142.2; 141.9; 65.7; 52.9;
12.7. MALDI-FTMS [M1Na]1: calculated: 575.0812;
observed: 575.0797.
1
trimer 22). H NMR (acetone-d6): 8.51 (d, Jˆ5.3 Hz, 1H,
NH); 8.47 (d, Jˆ5.3 Hz, 1H, NH); 8.45 (d, Jˆ5.6 Hz, 1H,
NH); 7.36±7.30 (m, 5H); 5.38 (m, 2H); 5.33 (m, 1H); 5.10
(AB, Jˆ12.3 Hz, 1H); 5.06 (AB, Jˆ12.3 Hz, 1H); 4.71 (s,
4.3. Synthesis of acetol- (20a) and tert-butyl (20b)
protected modules
2
2H); 3.27 (m, 2H); 3.20 (m, 2H); 3.08 (dd, Jˆ15.3 Hz,
Compound 11a was subjected to Bzl ester deprotection
under hydrogenolytic conditions (as in the one-pot synthesis
of 5a). The residue (1 equiv.) was dissolved in anhydrous
THF (0.02 M) and TBC (1.2 equiv.) was added, followed by
the slow addition of TEA (1.2 equiv.) at rt. The mixture was
stirred for 20 min, ®ltered and the solvent was removed in
vacuo to give the active ester. To the stirred solution of
acetol (2 equiv.) or tBu±OH (2 equiv.) and DMAP
(2 equiv.) in anhydrous DCM (0.02 M) was added the
DCM solution of the active ester compound at 08C. The
mixture was stirred for 2 h and was allowed to warm to rt,
then concentrated in vacuo and puri®ed on silica gel using
EtOAc/hexanesˆ1:2. Yield of acetol-ester (20a) and tert-
butyl ester (20b) were 83 and 52%, respectively.
2
3
3Jˆ4.0 Hz, 1H); 2.98 (dd, Jˆ15.3 Hz, Jˆ6.6 Hz, 1H);
2.63 (s, 3H); 2.61 (s, 3H); 2.53 (s, 3H); 2.09 (s, 3H); 1.37
(s, 9H). 13C NMR (acetone-d6): 201.5; 169.8; 169.5; 169.1;
161.4; 161.3; 161.2; 161.1; 160.8; 160.8; 155.0; 154.9;
154.8; 137.0; 129.4; 129.3; 129.2; 129.2; 129.0; 81.7;
69.4; 67.1; 46.7; 46.6; 46.4; 40.2; 38.9; 38.2; 28.2;
26.0; 11.6; 11.5. MALDI-FTMS [M1Na]1: calculated:
813.2702; observed: 813.2719.
4.3.2. Synthesis of monoacid platform 24 and diacid plat-
form 25. To the stirred solution of orthogonally protected
platform 23 (1 equiv.) in anhydrous THF (0.01 M) was
added TBAF (8 equiv., as a 1 M solution in THF) at rt
and stirring was continued for 4 h. The mixture was diluted
with water (to 1.5 times the original volume) and concen-
trated in vacuo. The residue was dissolved in EtOAc,
extracted with 5% aqueous KHSO4 solution and brine,
and dried over Na2SO4. The residue contained monoacid
platform 24 in 80%. 1H NMR (acetone-d6): 8.51 (d,
Jˆ5.3 Hz, 1H, NH); 8.48 (d, Jˆ5.4 Hz, 1H, NH); 8.47 (d,
Jˆ5.7 Hz, 1H, NH); 7.36±7.32 (m, 5H); 5.40±5.34 (m, 3H);
5.10 (AB, Jˆ12.3 Hz, 1H); 5.07 (AB, Jˆ12.3 Hz, 1H); 3.21
1
20a. H NMR (acetone-d6): 6.70 (d, Jˆ8.6 Hz, 1H, NH);
5.26 (m, 1H); 5.08 (s, 2H); 4.73 (s, 2H); 3.22 (dd,
2Jˆ16.6 Hz, 3Jˆ6.6 Hz, 1H); 3.10 (dd, 2Jˆ16.6 Hz,
3Jˆ6.9 Hz, 1H); 2.68 (s, 3H); 2.13 (s, 3H); 1.42 (s, 9H).
13C NMR (acetone-d6): 202.0; 170.3; 162.6; 160.7; 158.9;
155.9; 127.3; 96.2; 79.9; 74.4; 69.4; 46.4; 37.8; 28.6; 26.1;
12.6. MALDI-FTMS [M1Na]1: calculated: 523.0412;
observed: 523.0424.
2
3
(m, 2H); 3.18 (m, 2H); 3.08 (dd, Jˆ15.3 Hz, Jˆ4.2 Hz,
2
3
1
1H); 3.00 (dd, Jˆ15.3 Hz, Jˆ6.4 Hz, 1H); 2.64 (s, 3H);
2.61 (s, 3H); 2.54 (s, 3H); 1.38 (s, 9H). 13C NMR (acetone-
d6): 170.9; 169.8; 169.0; 161.5; 161.3; 161.2; 161.1; 160.9;
155.9; 154.8; 137.0; 129.4; 129.3; 129.2; 129.2; 129.0; 81.7;
67.1; 46.7; 46.6; 46.5; 40.2; 38.9; 38.1; 28.2; 11.6; 11.5.
MALDI-FTMS [M1Na]1: calcd: 757.244; observed:
757.2445.
20b. H NMR (acetone-d6): 6.63 (d, Jˆ8.7 Hz, 1H, NH);
5.21 (m, 1H); 5.07 (m, 2H); 2.98 (dd, 2Jˆ15.8 Hz,
2
3
3Jˆ7.0 Hz, 1H); 2.83 (dd, Jˆ15.8 Hz, Jˆ7.8 Hz, 1H);
2.67 (s, 3H); 1.42 (s, 18H). 13C NMR (acetone-d6): 169.9;
162.8; 160.8; 158.8; 155.9; 127.2; 96.2; 81.4; 79.8; 74.4;
46.7; 39.6; 28.6; 28.2; 12.6. ESI-MS [M1Na]1: calculated:
523.0; observed: 523.0.
Monoacid platform 24 was subjected to tBu ester deprotec-
tion using general Boc cleavage conditions to give a
4.3.1. Synthesis of orthogonally protected platform 23.
Modules 20a and 20b were subjected to trichloroethyl ester

1708
L. Somogyi et al. / Tetrahedron 57 (2001) 1699±1708
1
Nat. Prod. 1996, 59, 581±586. (b) Holzapfel, C. N.; Zyl,
W. J. v. Tetrahedron 1990, 46, 649±660.
quantitative yield of diacid 25. H NMR (acetone-d6): 8.48
(m, 3H, NH); 7.36±7.30 (m, 5H); 5.40±5.35 (m, 3H); 5.08
(m, 2H); 3.22 (m, 3H); 3.19 (m, 3H); 2.63 (s, 3H); 2.61 (s,
3H); 2.53 (s, 3H). 13C NMR (acetone-d6): 171.0; 169.9;
161.5; 161.4; 161.4; 161.3; 161.2; 160.9; 154.9; 154.9;
154.8; 137.0; 129.4; 129.3; 129.3; 129.2; 129.2; 129.1;
67.2; 46.7; 46.6; 46.5; 38.9; 38.2; 38.1; 11.6; 11.5.
MALDI-FTMS [M1Na]1: calculated: 701.1814; observed:
701.1830.
5. Wipf, P.; Miller, C. P. J. Am. Chem. Soc. 1992, 114, 10975±
10977.
6. Wipf, P. Chem. Rev. 1995, 95, 2115±2134.
7. Kocis, P.; Issakova, O.; Sepetov, N. F.; Lebl, M. Tetrahedron
Lett. 1995, 36, 6623±6626.
8. Mink, D.; Mecozzi, S.; Rebek, Jr., J. Tetrahedron Lett. 1998,
39, 5709±5712.
9. Borg, S.; Vollinga, R. C.; Labarre, M.; Payza, K.; Terenius, L.;
Luthman, K. J. Med. Chem. 1999, 42, 4331±4342.
10. (a) Williams, D. R.; Brooks, D. A.; Berliner, M. A. J. Am.
Chem. Soc. 1999, 121, 4924±4925. (b) Phillips, A. J.; Uto, Y.;
Wipf, P.; Reno, M. J. Org. Lett. 2000, 2, 1165±1168.
11. Williams, D. R.; Lowder, P. D.; Gu, Y.-G.; Brooks, D. A.
Tetrahedron Lett. 1997, 38, 331±334.
Acknowledgements
Financial support from the Skaggs Foundation and the
National Institutes of Health is acknowledged with grati-
tude. Gebhard Haberhauer thanks the Deutsche Forschungs-
gemeinschaft for a postdoctoral fellowship. The authors also
express their special thanks to Professors Takeharu Haino of
Hiroshima University and Peter Wipf of the University of
Pittsburgh for advice and encouragement.
12. Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277±
7287.
13. Wipf, P.; Miller, C. P. J. Org. Chem. 1993, 58, 3604±3606.
14. Clausen, K.; Thorsen, M.; Lawesson, S.-O. J. Chem, Soc.,
Perkin Trans. 1 1984, 785±798.
15. Bertram, A.; Hannam, J. S.; Jolliffe, K. A.; de Turismo,
F. G.-L.; Pattenden, G. Synlett 1999, 11, 1723±1726.
16. Boden, C.; Pattenden, G. Tetrahedron Lett. 1994, 35, 8271±
8274.
References
1. (a) Faulkner, D. J. Nat. Prod. Rep. 1999, 16, 155±198. (b)
Pettit, G. R. Pure Appl. Chem. 1994, 66, 2271±2281. (c)
Garson, M. J. Chem. Rev. 1993, 93, 1699±1730. (d) Davidson,
B. S. Chem. Rev. 1993, 93, 1771±1791. (e) Fusetani, N.;
Matsunoga, S. Chem. Rev. 1993, 93, 1793±1806.
17. Sokolenko, N.; Abbenante, G.; Scanlon, M. J.; Jones, A.;
Gahan, L. R.; Hanson, G. R.; Fairlie, G. P. J. Am. Chem.
Soc. 1999, 121, 2603±2604.
18. The methyl ester deprotection of compound 18b resulted in
partial epimerization of its chiral centers. This lead to a
mixture of diastereomers which were not separated but
subjected to Boc cleavage and cyclization. Two cyclic
compounds were isolated after silica gel chromatography: a
symmetric and an asymmetric one, along with their enantio-
mers. If there is equal probability for all three chiral centers to
epimerize, the theoretical ratio between the four possible
diastereomers (syn±syn±syn: syn±syn±anti: syn±anti±anti:
anti±anti±anti) is 8:12:6:1. This has been found experimen-
tally by coupling the symmetrical and asymmetrical triol plat-
forms with Boc-(l)-Asp(OBzl)-OH. In both cases two
diastereomers were separated in the ratio calculated above.
19. Haberhauer, G.; Somogyi, L.; Rebek, Jr., J. Tetrahedron Lett.
2000, 41, 5013±5016.
2. (a) Kigoshi, H.; Yamada, S. Tetrahedron 1999, 55, 12301±
12308. (b) Downing, S. V.; Aguilar, E.; Meyers, A. I. J. Org.
Chem. 1999, 64, 826±831. (c) Freeman, D. J.; Pattenden, G.;
Drake, A. F.; Siligardi, G. J. Chem, Soc., Perkin Trans. 2
1998, 129±135. (d) Moody, C. J.; Bagley, M. C. J. Chem,
Soc., Perkin Trans. 1 1998, 601±607. (e) Sone, H.; Kigoshi,
H.; Yamada, K. Tetrahedron 1997, 53, 8149±8154. (f) Wipf,
P.; Fritch, P. C. J. Am. Chem. Soc. 1996, 118, 12358±12367.
3. (a) Roy, R. S.; Gehring, A. M.; Milne, J. C.; Belshaw, P. J.;
Walsh, C. T. Nat. Prod. Rep. 1999, 16, 249±263. (b) Wipf, P.;
Fritch, P. C.; Geib, S. J.; Se¯er, A. M. J. Am. Chem. Soc. 1998,
120, 4105±4112. (c) Todorova, A. K.; JuÈttner, F.; Linden, A.;
PluÈss, T.; Philipsborn, W. v. J. Org. Chem. 1995, 60, 7891±
7895. (d) Li, Y.-M.; Milne, J. C.; Madison, L. L.; Koller, R.;
Walsh, C. T. Science 1996, 274, 1188±1193. (e) Foster, M. P.;
20. Kundu, B. Tetrahedron Lett. 1992, 33, 3193±3196.
21. Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M.
Bull. Chem. Soc. Jpn 1979, 52, 1989±1993.
Â
Concepcion, G. P.; Caraan, G. B.; Ireland, C. M. J. Org. Chem.
1992, 57, 6671±6675.
4. (a) Ogino, J.; Moore, R. E.; Patterson, G. M. L.; Smith, C. D. J.