Synthesis of the cryptophycins

Full text of DOI:10.1016/S0040-4020(02)00238-7

TETRAHEDRON
Pergamon
Tetrahedron 58 :2002) 4343±4367
Tetrahedron report number 608
Synthesis of the cryptophycins
Marcus A. Tiusa,b,p
aDepartment of Chemistry, University of Hawaii, 2545 The Mall, Honolulu, HI 96822, USA
bThe Cancer Research Center of Hawaii, 1236 Lauhala Street, Honolulu, HI 96813, USA
Received 15 January 2002
Contents
1. Introduction
2. The ®rst Hawaii synthesis
3. The Leahy±Gardinier synthesis
4. The Tius±Li synthesis
5. Syntheses of unit A
4343
4344
4347
4349
4350
4350
4351
4352
4354
4355
4355
4356
4356
4358
4359
4360
4361
4362
4364
5.1. The Sih chemoenzymatic synthesis
5.2. The Lilly chemoenzymatic synthesis
5.3. Synthesis of unit A via a [2,3]-Wittig rearrangement
5.4. Asymmetric crotylboration
5.5. Diastereoselective allylstannation
5.6. A chiral auxiliary approach
5.7. Noyori hydrogenationÐFrater alkylation
5.8. :S)-:2)-2-Acetoxysuccinic anhydride as a starting material for unit A
5.9. Vinyl epoxide reduction
Â
5.10. Planar chiral molybdenum cationic complex
5.11. Diastereoselective addition to :R)-mandelaldehyde
6. Dihydroxylation of seco-cryptophycin 52
7. 1-Aza-cryptophycin 1, an unstable cryptophycin
8. Conclusion
1. Introduction
discovered a large number of new cytotoxic compounds.2±4
Many of these had unique and bizarre structures, and many
were potent and selective cytotoxins. The most promising of
all of these was isolated in 1991 after more than 1000 cyano-
phytes had been screened. The crude lipophilic extract from
Nostoc sp. GSV 224was found to be cytotoxic against the
KB cell line :a human nasopharyngeal carcinoma) at
0.24ng/mL, and against the LoVo cell line :a human
colorectal adenocarcinoma) at 6 ng/mL.5 Moreover, the
crude extract showed tumor-selective cytotoxicity. The
major cytotoxin which was isolated through bioassay-
guided fractionation of the crude cyanobacterial extracts
was cryptophycin 1 :cryptophycin A), a cyclic depsipeptide.
Cryptophycin 1 co-occurs with a number of structurally
related cryptophycins. After the Hawaii team had elucidated
the gross structure of cryptophycin 1, it was found that a
group at Merck had isolated the same compound some years
earlier from the extracts of a related cyanophyte, Nostoc sp.
ATCC 53789.6,7 The Merck team published neither the
For many years, the Chemistry Department at the University
of Hawaii housed an interdisciplinary research program
involving the collection and cultivation of cyanophytes
:blue-green algae), and the screening of their extracts for
anti-cancer activity. During the 1980s, my colleagues in the
Chemistry Department, Richard Moore and Gregory
Patterson, examined several hundreds of culture extracts
in a bioassay-guided search for new leads for cancer
chemotherapeutic agents.1 The blue-green algae suggested
themselves as a promising group of organisms for scrutiny
in a search for novel cytotoxins, since there had been
numerous reports of the adverse effects of algal blooms on
human and animal health. The Moore±Patterson team
Keywords: cryptophycins; cyanophytes; anti-cancer activity.
Tel.: 11-808-956-2779; fax: 11-808-956-5908;
e-mail: tius@gold.chem.hawaii.edu
p
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020:02) 00238-7

4344
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Figure 1. Some common cryptophycins.
relative nor the absolute stereochemistry in their patent. It is
surprising that they apparently failed to discover the extra-
ordinary selective cytotoxicity of the natural product,
although they did note strong anti-fungal activity against
Cryptococcus :Fig. 1).
that in a macrocyclic cryptophycin intermediate one dia-
stereoface of the styrene double bond would be shielded
by the bulk of the molecule, leaving the other face open
for a selective reaction with electrophilic epoxidizing
reagents. As will be seen, this assumption was wrong.
Scheme 1 summarizes the ®rst successful approach to unit
A. Enoate 1.1 was prepared in 86% yield from com-
mercially available dihydrocinnamaldehyde and trimethyl-
phosphonoacetate in the presence of tetramethylguanidine
:TMG) in THF. Reduction of the ester function in 1.1 with
DIBAL-H led to the anticipated allylic alcohol in 90%
isolated yield. Sharpless asymmetric epoxidation, using
l-:1)-diethyl tartrate as the chiral inducer, led to epoxy
alcohol 1.2 :94% yield; .95% ee). Regio- and stereo-
speci®c epoxide ring-opening with trimethylaluminum led
to the desired 1,2-diol 1.3 in 95% yield. None of the 1,3-diol
that would have resulted from the alternative mode of attack
by the trimethylaluminum could be detected in the reaction
mixture. Protection of the diol as acetonide 1.4 :97% yield)
was straightforward, setting the stage for the introduction of
the benzylic E double bond. Benzylic bromination of 1.4
with NBS in carbon tetrachloride led to unstable bromide
1.5 which was immediately dehydrobrominated by exposure
to DBU at 708C. This provided 1.6 in 80% overall yield for
the two steps. Although this process worked well on small
scale, upon scaleup we found that the yield of 1.6 was
eroded by the formation of diastereomeric tetrahydrofuryl
alcohols 1.10 as undesired reaction byproducts. The appear-
ance of 1.10 is most easily rationalized as arising from
acetonide hydrolysis by adventitious HBr which is formed
during the radical halogenation reaction, followed by intra-
molecular displacement of bromide to generate more HBr.
Consistent with this mechanistic hypothesis, slow addition
of 1 equiv. of 2,2-dimethoxypropane to the radical halo-
genation reaction mixture as an acid scavenger effectively
suppressed the formation of 1.10. Similar reactivity of the
benzylic carbon atom of unit A is a recurring motif in
cryptophycin chemistry.
2. The ®rst Hawaii synthesis
The Hawaii group was the ®rst to publish the relative and
absolute stereochemistry for the cryptophycins in 1994.8
Our ®rst synthesis, in 1995, corrected the structures of
cryptophycins 1 and 3. The modular nature of the depsi-
peptide structure renders the synthetic approach much
simpler than it would otherwise be. For the purposes of
the retrosynthetic analysis, cryptophycin 1 can be dis-
connected into four units. Unit A, the only polyketide-
derived molecular fragment, provides the greatest synthetic
challenge. Unit B is O-methyl-d-chlorotyrosine, whereas
units C and D correspond to :R)-3-amino-2-methyl-
propanoic acid and :S)-2-hydroxy-4-methylvaleric
:l-leucic) acid, respectively. There are a number of alter-
natives for macrocyclization, all of which appear to be
viable and attractive. One of the most appealing is to form
the unit B to unit C amide bond during the key step. Such an
approach takes advantage of the small steric requirement of
the primary amino group of unit C :Fig. 2).
The epoxide function of unit A poses a problem. The high
reactivity of this styrene oxide suggests that it would be
most prudent to introduce it as late in the synthetic sequence
as possible, preferably in the very last step. We had assumed
Hydrolysis of the acetonide group in 1.6 with aqueous
methanolic HCl led to a diol :93% yield) which was selec-
tively monotosylated at the primary position according to
Figure 2. Retrosynthetic disconnection showing units A, B, C and D.

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4345
Scheme 1. :a) DIBAL-H, THF, 278 to 258C, 90%; :b) l-:1)-DET, Ti:O-iPr)4, t-BuOOH, CH2Cl2, 2208C, 94%; :c) AlMe3, hexane/CH2Cl2, 0±258C, 95%;
:d) :MeO)2CMe2, PPTS, CH2Cl2, 258C, 97%; :e) NBS, CCl4, :MeO)2CMe2, hn, 258C; :f) DBU, 708C, 80% :two steps); :g) 1% aqueous HCl/MeOH, 258C,
93%; :h) Bu2Sn:OMe)2, PhMe, Dean-Stark; TsCl, Et3N, 0±258C, 82%; :i) TBSOSO2CF3, Et3N, CH2Cl2, 258C, 98%; :j) KCN, DMSO, 608C, 92%; :k) DIBAL-
H, CH2Cl2, 278 to 258C, 95%; :l) :MeO)2POCH2CO2Me, TMG, THF, 278 to 258C, 83%.
Ley's procedure.9 Exposure of the diol to dibutyldimethoxy-
stannane in re¯uxing toluene with azeotropic removal of
methanol led to a dibutylstannylene ketal which was
directly converted to the primary tosylate in 82% overall
yield from the diol by treatment with tosyl chloride and
triethylamine. Protection of the secondary alcohol as the
TBS ether led to 1.7 in 98% yield. The remaining carbon
atoms of unit A were introduced as follows. Displacement
of the tosylate by cyanide in DMSO :92% yield) was
followed by reduction with DIBAL-H in dichloromethane
to give aldehyde 1.8 in 95% yield. Horner±Emmons homo-
logation with trimethylphosphonoacetate in the presence of
TMG led to key intermediate 1.9 in 83% yield. This ®rst
synthesis of unit A is fairly long, however, the overall yield
from dihydrocinnamaldehyde is high at 29%. Moreover,
none of the steps makes use of exotic reagents or extreme
conditions, therefore this is an eminently practical
synthesis of unit A. Considerations of practicality informed
our synthetic plan, because it was obvious to us from the
outset that the cryptophycins were destined for the clinic,
and also that it would not be possible to use Nostoc in
culture for the production of material. The reasons for this
last point may not be obvious, and therefore deserve some
comment. Nostoc is an obligate phototroph, so large-scale
fermentation in a closed system, as required by cGMP
standards, presents an insurmountable engineering
problem. Also, it is often the case that the natural product
is not the optimal clinical candidate. For these reasons total
synthesis provides the only means to secure material for the
clinic.
Scheme 2 shows the assembly of unit A1unit B. Methyl
ester 1.9 was hydrolyzed to carboxylic acid 2.1 with lithium
hydroxide in acetone :95% yield). Protected unit B 2.2 was
prepared from commercially available d-tyrosine in ®ve
steps. Chlorination with sulfuryl chloride in glacial acetic
acid10 was followed by protection of the nitrogen atom as
the Boc derivative with di-tert-butyldicarbonate and
triethylamine :94% yield). Methylation of this material in
acetone took place in the presence of potassium carbonate
Scheme 2. :a) LiOH, acetone, 258C, 95%; :b) FDPP, DIEA, DMF, 258C, 73%; :c) MeCN, 50% aqueous HF :95/5), 258C, 90%.

4346
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Scheme 3. :a) NH3, MeOH, 508C, sealed tube, 7 days, 66%; :b) BH3±THF, THF, 08C to re¯ux, 77%; :c) :Boc)2O, Et3N, MeOH, 258C, 100%; :d) RuCl3,
NaIO4, CCl4, MeCN, H2O, 258C, 74%; :e) DMAP, DCC, CH2Cl2, 0±258C, 92%; :f) THF, morpholine, Pd:PPh3)4, 258C, 100%.
and dimethyl sulfate at re¯ux for 4h to produce the methyl
ester of N-:tert-butoxycarbonyl)-3-:3-chloro-4-methoxy-
phenyl)-d-alanine in 84% yield. Hydrolysis back to the
carboxylic acid took place with aqueous NaOH in dioxane
in 86% yield. Esteri®cation of the free acid with 2,2,2-
trichloroethanol in the presence of DCC and pyridine gave
the Troc ester in 65% yield after recrystallization of the
crude product from hexanes. Exposure of this material to
tri¯uoroacetic acid at room temperature led to quantitative
cleavage of the Boc group, leading to the tri¯uoroacetate
salt of 2.2. The tri¯uoroacetate was converted to 2.2 in situ,
during the condensation with 2.1. Mixing a solution of 2.1 in
anhydrous DMF with penta¯uorophenyl diphenylphos-
phinate :FDPP), the tri¯uoroacetate salt of 2.2 and diiso-
propylethylamine :DIEA) at room temperature led to amide
2.3 in 73% yield.11 Cleavage of the silyl ether protecting
group with aqueous HF in acetonitrile produced 2.4 in 90%
yield.
The Hawaii synthesis follows a highly convergent strategy
whereby 2.4 was joined to unit C1D fragment 3.6
:Scheme 3). Commercially available :S)-:1)-3-hydroxy-2-
Scheme 4. :a) DMAP, DCC, CH2Cl2, 0±258C, 94%; :b) Zn, THF, AcOH, sonicate, 258C; :c) TFA :neat), 258C, 89% :two steps); :d) FDPP, DIEA, DMF, 258C,
64%; :e) m-CPBA, CH2Cl2, 08C, ca. 2/1 cryptophycin 1/4.4.

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4347
methylpropanoate 3.1 was ®rst converted to the amide by
3. The Leahy±Gardinier synthesis
treatment with ammonia in methanol at 508C in a sealed
tube. The amide was isolated in 66% yield, along with
unreacted 3.1 that was separated from the product and was
recycled. Reduction of the amide function with borane in
THF led to amine 3.2 in 77% yield. Protection of the amino
group as the Boc derivative :100% yield), followed by
oxidation of the primary alcohol with ruthenium tetroxide
:74% yield) led to carboxylic acid 3.3. Commercially avail-
able l-leucic acid was converted to allyl ester 3.4 in 93%
yield under phase-transfer conditions, by stirring a two-
phase mixture of the acid and allyl bromide in dichloro-
methane with aqueous sodium bicarbonate containing
tetra-n-butylammonium chloride. The coupling of 3.3 with
3.4 was accomplished with DCC and DMAP in dichloro-
methane to produce 3.5 in 92% yield. Cleavage of the allyl
ester was carried out under neutral conditions, by
exposure of 3.5 to dry morpholine and catalytic Pd:PPh3)4
in THF.12 Unit C1D acid 3.6 was isolated in quantitative
yield.
The ®rst synthesis to address the stereochemical problem
posed by the epoxide is due to Leahy and Gardinier.13 A diol
was used as a convenient synthetic equivalent of the epox-
ide. The stereochemistry at carbon atoms C7 and C8 was
established from the outset thereby avoiding any ambiguity
regarding the epoxide stereochemistry. Leahy and
Gardinier's highly successful synthesis of unit A is sum-
marized in Scheme 5. Two asymmetric fragments were
coupled in the ®rst step, :R)-mandelaldehyde derivative
5.1 and the boron enolate derived from Evans chiral imide
5.2.14 This led to a single crystalline product, 5.3, in 84 %
yield. Three of the four stereocenters were thus established
in the ®rst step. Trimethylaluminum-mediated transamida-
tion of 5.3 gave the corresponding Weinreb amide in 93%
yield. Following treatment with allylmagnesium bromide,
allyl ketone 5.4 was isolated :92% yield). The stereo-
chemistry at C5 was controlled by means of the Evans±
Hoveyda tandem aldol±Tishchenko reaction: exposure of
5.4 to acetaldehyde and samarium iodide reduced the C5
ketone carbonyl group to the a alcohol, and at the same time
converted the C7 hydroxyl to the corresponding acetate
:96% yield).15 Protection of the C5 hydroxyl group as the
p-methoxybenzyl :PMB) ether :70% yield) was followed by
treatment with DIBAL-H to cleave the C7 acetate reduc-
tively :90% yield). Finally, protection as the TIPS derivative
gave 5.5 in 93% yield. The conversion of 5.5 to unit A 5.6
was accomplished by oxidative cleavage of the terminal
alkene to the aldehyde :91% yield), followed by Horner±
Emmons homologation using the Masamune±Roush condi-
tions :90% yield) and oxidative cleavage of the PMB
protecting group with DDQ :95% yield). This represents
the ®rst highly ef®cient synthesis of unit A with control of
C7 and C8 stereochemistry. There is one limitation of this
route, namely that the Evans aldol reaction is sensitive to
scale. On larger scale, increasing quantities of the undesired
isomer 5.7 are generated, therefore this reaction was limited
to a 2 g scale.
The joining together of the two large fragments and the
macrolactamization are summarized in Scheme 4. The
ester linkage between 2.4 and 3.6 was installed through
the use of DCC/DMAP to produce seco compound 4.1 in
94% yield. Sequential removal of the Troc and Boc protect-
ing groups was accomplished in 89% yield following treat-
ment with zinc and acetic acid, followed by neat
tri¯uoroacetic acid. Macrolactamization of 4.2 with FDPP
in the presence of DIEA in DMF gave 4.3 :cryptophycin 3)
in 64% yield. Epoxidation of 4.3 with m-CPBA at 08C led to
a ca. 2/1 mixture of cryptophycin 1 and the corresponding
:7S,8S)-trans-epoxide 4.4. As will be seen from the discus-
sion of the subsequent cryptophycin syntheses, this ratio is
essentially invariant regardless of the oxidant or the condi-
tions that are chosen. A more vexing problem is that the
separation of cryptophycin 1 from 4.4 can only be accom-
plished by HPLC, which greatly complicates the preparative
task. Several solutions to the stereochemical problem posed
by the epoxide have been developed, both by our group and
by others. These will be discussed in some detail in what
follows.
Leahy and Gardinier follow a rather different strategy for
the assembly of units B, C, and D. Protected aminoalcohol
6.1 :Scheme 6) was prepared from the reduction of the
known amide.16 The amine was then coupled with unit B
acid 6.2, which was prepared according to the ®rst Hawaii
synthesis, in 97% yield. Fluorodesilylation with TBAF
The ®rst Hawaii synthesis has been discussed in consider-
able detail because the syntheses that followed invariably
borrowed from it to varying degrees.
Scheme 5. :a) Bu2BOTf, DIEA, 84%; :b) Me3Al, MeONHMe´HCl, 93%; :c) CH2vCHCH2MgBr, 92%; :d) MeCHO, SmI2, 96%; :e) PMBOC:NH)CCl3, cat.
TfOH, 70%; :f) DIBAL-H, 90%; :g) TIPSOTf, DIEA, 93%; :h) OsO4, NMO; NaIO4, 91%; :i) :EtO)2POCH2CO2tBu, DBU, LiCl, 90%; :j) DDQ, 95%.

4348
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Scheme 6. :a) EDC, HOBT, 97%; :b) TBAF, 96%; :c) RuCl3, NaIO4, 83%; :d) DCC, 95%; :e) Ra/Ni, H2, 84%.
:96% yield) was followed by oxidation of the primary
alcohol function with ruthenium tetroxide :83% yield) to
give unit B1C compound 6.3. Oxidation of alcohol to
acid, rather than coupling 6.2 with a b-alanine ester, is
one of several noteworthy features of this total synthesis.
Coupling of 6.3 with the benzyl ester of l-leucic acid 6.4
took place with DCC to give 6.5 in 95% yield. Finally,
reductive cleavage of the benzyl ester with Raney nickel
led to unit B1C1D compound 6.6 in 84% yield.
silyl ether protecting groups were inert under these condi-
tions. Macrolactamization was carried out with O-benzo-
triazol-1-yl-N,N,N0,N0-bis:pentamethylene)uronium hexa-
¯uorophosphate to give the macrocycle in 76% overall
yield for the two steps.18 Cleavage of the two silyl protecting
groups with TBAF gave diol 7.2 in 95% yield. The ®nal
problem in the synthesis was to convert the syn-diol to the
b-epoxide. Although Sharpless had published an excellent
method for the conversion of vicinal diols to epoxides,19 the
strongly basic conditions of the procedure are incompatible
with the macrocycle. In particular, the unit C to unit D ester
link is readily cleaved in the presence of base. This is, in
fact, one of the pathways for the deactivation of the crypto-
phycins in vivo. As a consequence, Leahy and Gardinier
developed a very ingenious modi®cation of the Sharpless
procedure. 4-Azido-1,1,1-trimethoxybutane 7.3, prepared in
The ®nal steps of the Leahy±Gardinier synthesis are
summarized in Scheme 7. Coupling of unit A compound
5.6 with unit B1C1D 6.6 under Yamaguchi conditions
produced 7.1 in 91% yield.17 Cleavage of the Boc and
tert-butyl ester protecting groups took place in a single
operation, by exposure of 7.1 to HCl in ethyl acetate. The
Scheme 7. :a) DIPEA, 2,4,6-trichlorobenzoyl chloride, 91%; :b) HCl, EtOAc; :c) O-benzotriazol-1-yl-N,N,N0,N0-bis:pentamethylene)uronium hexa¯uoro-
phosphate, 76% :two steps); :d) TBAF, 95%; :e) 7.3, TMS±Cl, 63%; :f) PPh3, H2O, 63%; :g) K2CO3, acetone, 98%.

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4349
two steps from commercially available 4-chlorobutyroni-
trile, was allowed to react with diol 7.2 in the presence of
trimethylchlorosilane according to Sharpless' conditions.
This led stereospeci®cally to the anticipated azidobutyrate
7.4 with inversion of con®guration at C7 :63% yield).
Exposure of 7.4 to triphenylphosphine and water led to a
Staudinger reaction with generation of a primary amino
group that underwent an intramolecular lactamization reac-
tion to produce the C7±C8 chlorohydrin. Brief exposure of
the chlorohydrin to potassium carbonate in acetone led to
cryptophycin 1.
is likely to be mechanistically distinct from the conventional
concerted Diels±Alder process, the stereochemistry of the
products can be rationalized as resulting from an exo transi-
tion state in a chelation controlled process.22 Acid catalyzed
cleavage of the ethoxyethyl protecting group was followed
by an intramolecular Michael addition of the hydroxyl to
give bicyclic products 8.6 and 8.7. Stirring a solution of this
product mixture in acetonitrile over KF/alumina23 for 48 h
epimerized 8.6, leading to a 1/4mixture of axial methyl 8.6
to equatorial methyl 8.7 products in 48% overall yield from
8.1. It is noteworthy that diene 8.4 was used as a 9/1 mixture
of :E,Z) to :Z,Z) geometrical isomers. The stereochemistry
of the diene is re¯ected in the ratio of Diels±Alder products,
but since one can epimerize the product mixture, one need
not separate the isomers of the diene.
The Leahy±Gardinier synthesis is exceptionally clever, and
is noteworthy for solving several of the problems that are
posed by the structure in unique and ingenious ways. The
use of the Evans aldol reaction with mandelaldehyde
derivative 5.1, the Tishchenko reduction of the C5 keto
group and the design of azidoester 7.3 so as to overcome
the limitations of the conventional Sharpless procedure
distinguish this work.
Reduction of the 1/4mixture of 8.6 and 8.7 with l-selectride
at 2788C in THF gave alcohol 8.8 as a single :axial) isomer
in 93% yield, based on recovered starting material. The
recovered ketone was determined to be a ca. 6±7/1 mixture
of 8.6 and 8.7. It therefore appears that reduction of axial
isomer 8.6 by l-selectride is very slow, and since the reac-
tion medium is basic, some epimerization of 8.6 to 8.7 can
take place during the course of the reduction. Intermediate
8.8 incorporates all four asymmetric carbon atoms of unit A.
No adjustment of the oxidation state of C3 is necessary,
since it is present in 8.9 as a ketal. Careful exposure of
8.8 to boron tri¯uoride etherate and 1,3-propanedithiol in
dichloromethane at 08C converted the ketal function to the
dithioketal in 85% yield. Protection of the syn C7±C8 diol
as the acetonide gave 8.9 in 93% yield. It was necessary to
exercise care during the formation of the dithioketal. If the
temperature was allowed to rise much above 08C, or if the
reaction was not quenched promptly following the
consumption of 8.8, tetrahydrofuran 8.11 was isolated in
yields up to 30%. This compound is analogous to 1.10,
and its formation provides another example of the reactive
nature of the benzylic carbon atom in unit A. Treatment of
dithioketal 8.9 with iodomethane and calcium carbonate in
aqueous acetonitrile led to the expected b-hydroxyaldehyde.
4. The Tius±Li synthesis
:R)-Mandelic acid is an attractive starting material for unit
A. It occurred to us that it should be possible to assemble
unit A using mandelic acid as the sole source of asymmetry.
Scheme 8 summarizes our successful unit A synthesis.20
:R)-Methyl mandelate 8.1 was converted to O-ethoxyethyl
derivative 8.2. Although the presence of an asymmetric
center on the protecting group complicates the interpretation
of the NMR spectra, we found that this choice of oxygen
protecting group was critical for the success of the subse-
quent reactions in the scheme. Reduction of the ester func-
tion in 8.2 with DIBAL-H gave protected mandelaldehyde
8.3, which was combined with diene 8.4 in THF in a magne-
sium bromide-catalyzed hetero-Diels±Alder reaction.21
Exposure of the crude Diels±Alder product 8.5 to tri¯uoro-
acetic acid, followed by workup, gave a 10/1 mixture of
products 8.6 and 8.7. Although this cycloaddition process
Scheme 8. :a) ethyl vinyl ether, PPTS; :b) DIBAL-H, ether, 2788C; :c) MgBr2, THF, 36±428C; :d) TFA, THF, 8.6/8.7ˆ10/1; :e) KF, alumina, MeCN,
8.6/8.7ˆ1/4, 48% of 8.7 overall from 8.1; :f) l-selectride, THF, 2788C, 93%; :g) BF3´Et2O, CH2Cl2, HS:CH2)3SH, 08C, 85%; :h) Me2C:OMe)2, TsOH,
acetone, 93%; :i) MeI, CaCO3, MeCN, H2O, 708C; :j) LiClO4, DIPEA, MeCN, :EtO)2POCH2CO2CH2CHvCH2, 68% from 8.9.

4350
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Scheme 9. :a) Pd:PPh3)4, morpholine, 100%; :b) Et3N, EDCI, CH2Cl2, 08C to rt, 81%; :c) DCC, DMAP, CH2Cl2, 81%; :d) TFA, CH2Cl2, 08C;
:e) 2-hydroxypyridine, PhMe, rt, 62% from 9.2; :f) :MeO)3CH, PPTS, CH2Cl2, rt; :g) AcBr, CH2Cl2, rt; :h) KHCO3, DME/EtOH/MeOH :6/4/1), 408C,
70% from 7.2.
Horner±Emmons reaction with allyldiethylphosphono-
acetate led to unit A compound 8.10 in 68% overall yield
for the two steps from 8.9. It is noteworthy that the C5
hydroxyl group was not protected at any time during this
synthesis.
to powdered potassium bicarbonate in a 6/4/1 mixture of
DME/ethanol/methanol at 408C for 24h led to cryptophycin
1 in 70% overall yield from 7.2. The conditions for the
conversion of bromohydrin formate to epoxide were
developed through a painstaking optimization process.
The small amount of methanol in the reaction mixture was
necessary to partially dissolve the bicarbonate. However,
larger proportions of methanol led to methanolysis of the
unit C±unit D ester.
The allyl ester in 8.10 was cleaved in quantitative yield in
the presence of catalytic palladium:0) and morpholine
:Scheme 9).12 Coupling with unit B compound 2.2 took
place in the presence of EDCI and triethylamine to produce
9.1 in 81% yield. Esteri®cation of the C5 hydroxyl group
with unit C1D 3.6 took place in the presence of DCC and
DMAP in 81% yield to give seco compound 9.2. Simul-
taneous removal of the Boc and acetonide protecting groups
took place with tri¯uoroacetic acid in dichloromethane at
08C. The elegant conditions that were developed at Eli Lilly
Co. were applied to the macrolactamization step.18 Simple
treatment in toluene at room temperature with 2-hydroxy-
pyridine led to 7.2, the same intermediate from Leahy's
synthesis, in 62% overall yield for the two steps from 9.2.
The trichloroethyl protecting group in unit B serves to
activate the ester during the macrolactamization, so it serves
two complementary roles during the synthesis. The Lilly
conditions for the macrolactamization do not require exotic
or expensive peptide coupling reagents, and are well suited
for large-scale work.
In summary, this cryptophycin synthesis is both ef®cient
:6.8% overall yield) and stereospeci®c. :R)-Methyl
mandelate serves as the sole source of asymmetry for unit
A. The key feature of this synthesis is the stereoconvergent
hetero-Diels±Alder reaction that is used to assemble unit A.
In Section 5, each of the other approaches to unit A that have
been described to date is discussed.
5. Syntheses of unit A
5.1. The Sih chemoenzymatic synthesis
The Sih synthesis of unit A relies upon an enzymatic
hydrolysis of a racemic ester for the key step :Scheme
10).24 :E)-Methyl styryl acetate 10.1 was ®rst treated with
LDA, followed by dimethyl sulfate to produce :^)-10.2 in
92% yield. Exposure of the racemic ester to Candida rugosa
lipase from Sigma Chemical Co. at room temperature in a
phosphate buffer led to carboxylic acid :S)-10.3 in 45%
yield and methyl ester :R)-10.2 in 48% yield. Both products
were isolated in .96% ee, therefore the kinetic resolution
was highly successful. Separation of methyl ester :R)-10.2
from the carboxylic acid was followed by reduction with
DIBAL-H to give aldehyde 10.4 in 95% yield. The intro-
duction of the remaining four carbon atoms of unit A was
Since one of our goals for this work was to develop as
simple a cryptophycin synthesis as possible, we decided
not to make use of azido orthoformate 7.3 for the critical
diol to epoxide transformation. We were mindful of the
lability of the unit C±unit D ester linkage toward base, so
we used a modi®cation of the Sharpless procedure. Treat-
ment of 7.2 with trimethyl orthoformate and PPTS, followed
by acetyl bromide, led to a sensitive bromohydrin formate
that was not puri®ed. Exposure of this intermediate product

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4351
Scheme 10. :a) LDA, Me2SO4, THF, 92%; :b) 2-propanol-treated Candida rugosa lipase, 0.2 M phosphate buffer, rt, :S)-10.3 45% :ee.96%), :R)-10.2 48%
:ee.96%); :c) DIBAL-H, Et2O, 2708C, 95%; :d) Zn±Pb, Et2O/PhH :1/1), 10.6 22% :ee.96%), 10.7 18% :ee.96%), a-isomers 25%; :e) 2,4-dinitrobenzoic
acid, DEAD, PPh3, 10.8 15%, 10.9 59%; :f) K2CO3, MeOH, 30 min, 92%.
accomplished by means of a Reformatsky reaction with tert-
butyl 4-bromocrotonate 10.5. This is not a straightforward
reaction, since there are a number of pathways that are
available to 10.5. In fact, in spite of the considerable efforts
of the Sih group to optimize the Reformatsky reaction, the
process was not selective, and all possible products were
formed: the undesired syn g-adduct 10.6 was isolated in
22% yield along with the desired anti-g-adduct 10.7 :18%
yield). The a-adducts were also formed in 25% yield. Since
the major product had the wrong stereochemistry it was
necessary to convert 10.6 to 10.7 through a Mitsunobu
inversion. Exposure of 10.6 to 2,4-dinitrobenzoic acid,
diethylazodicarboxylate :DEAD) and triphenylphosphine
led in 59% yield to dinitrobenzoate 10.9 with the correct
C5 stereochemistry, along with 15% of triene 10.8. The
presence of the enoate function renders the protons at C4
relatively acidic, consequently the Mitsunobu process in this
instance was accompanied by unusually large amounts of
elimination product 10.8. Nitrobenzoate 10.9 was
hydrolyzed to 10.7 with methanolic potassium carbonate
in 92% yield. The overall yield of 10.7 from 10.1 was
12.6%. Although Sih's synthesis of unit A provides an
illustration of the synthetic utility of an enzyme-mediated
kinetic resolution, the yield is inferior to that of the ®rst
Hawaii synthesis :29% for unit A). Also, the need to
separate and recycle 10.6 to 10.7 diminishes the preparative
utility of the Sih approach. Nevertheless, Sih notes that the
Hawaii synthesis `¼is relatively lengthy and is not easily
adaptable for the synthesis of cryptophycins in gram quan-
tities.' Sih's statement should be interpreted in light of the
following comments by Moher, Moore and co-workers: `As
part of our initial efforts to generate supplies of 2 and
synthetic intermediates to fuel clinical development and
SAR, respectively, we adopted the Moore±Tius synthesis
of 2, making necessary adjustments for multi-kilo opera-
tions.'25 :Compound 2 is cryptophycin 52; see Scheme 23
for the structure of cryptophycin 52.)
5.2. The Lilly chemoenzymatic synthesis
Chemists at Eli Lilly Company have described a successful
chemoenzymatic approach to unit A :Scheme 11).26
Commercially available :R)-carvone 11.1 was bioreduced
with Trigonopsis variabilis :ATCC 10679) using glucose
as a carbon source to produce saturated alcohol 11.2 in
50±62% isolated yield without the need for chromato-
graphy. The product was isolated as a single diastereomer
:de.98%, as judged by GC). The ®rst step of this synthesis
serves to establish the stereochemistry at C5 and C6 of unit
A. Protection of the secondary hydroxyl group as the TBS
derivative :87% yield) was followed by ozonolysis and
Criegee rearrangement to produce alcohol 11.4 after
hydrolytic cleavage in 73% overall yield from 11.3.27,28
Oxidation to ketone 11.5 took place in a two-phase system
with aqueous sodium hypochlorite and dichloromethane in
the presence of catalytic TEMPO, to give cyclohexanone
11.5.29 At this stage it was necessary to perform the C±C
bond cleavage that converts cyclic intermediate 11.5 to a
linear product that can be elaborated to unit A. Since both a
carbon atoms in 11.5 are unsubstituted, the only control
element that can be used to in¯uence the regiochemistry
of the cleavage is the b-silyloxy group. The electron-with-
drawing effect of the b-silyloxy substituent was expected to
favor rearrangement to 11.6 during a Baeyer±Villiger

4352
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Scheme 11. :a) Trigonopsis variabilis, pH 7 buffer, glucose, 50±62%; :b) TBDMSCl, DBU, DMF, 87%; :c) O3, MeOH, CH2Cl2; :d) Ac2O, DMAP; :e) NaOH,
MeOH, 73% :three steps); :f) TEMPO :cat.), NaOCl, CH2Cl2, 87%; :g) CF3CO3H, CH2Cl2, TFA :1/4), 2158C, 2 h, 81%; :h) DIBAL-H, PhMe, 2788C;
:i) :MeO)2POCH2CO2Me, TMG, THF, 66% :two steps); :j) TEMPO :cat.), NaOCl, CH2Cl2; :k) PhMgBr, 2788C, 67% :two steps); :l) 3 equiv. Ms2O, 9 equiv.
Et3N, DMAP, CH2Cl2, 42 h, 53%.
oxidation.30 The optimized conditions involved the use of
30% aqueous hydrogen peroxide as the oxidant :1.6 equiv.
oxidant) with suf®cient tri¯uoroacetic anhydride to form
peroxytri¯uoroacetic acid and to consume all the water,
leading to a solvent mixture of dichloromethane and
tri¯uoroacetic acid :4/1). At 2158C under these conditions
lactone 11.6 was formed in a 98/2 ratio with its regioisomer
in a combined yield of 83% :81% yield of 11.6). The authors
note that consistently better results for the Baeyer±Villiger
were obtained from the reactions that used 30% hydrogen
peroxide and tri¯uoroacetic anhydride, than from hydrogen
peroxide urea complex and tri¯uoroacetic anhydride. They
postulate that conditions that favor the protonation of the
b-silyloxy group, thereby amplifying its inductive electron-
withdrawing effect, favor the production of 11.6.
methane. This led to styrene 1.9, the intermediate from
the ®rst Hawaii synthesis, in 53% yield :97.3% ee). The
desired 1.9 was formed in a 4/1 ratio with tetrahydrofuran
11.9, the product of intramolecular displacement of the
mesylate by the oxygen atom at C5. During the course of
this study it was found that para substitution on the phenyl
ring by electron-donating substituents :e.g. methyl)
increases the rate of the desired elimination reaction relative
to tetrahydrofuran formation, perhaps by rendering the
intermediate cation more stable, and hence longer lived,
thereby allowing the elimination process to compete with
the cyclization. Alternatively, stabilization of the transition
state for the elimination process would account for the
observation.
The use of methanesulfonic anhydride for the elimination
also deserves comment. When methanesulfonyl chloride
was used instead of the anhydride for the elimination of
water from 11.8, the benzylic chloride was isolated in
25% yield. The chloride did not undergo elimination
under the reaction conditions.
The conversion of 11.6 to unit A follows a conventional
strategy. Reduction of lactone to lactol with DIBAL-H
was followed by
a Horner±Emmons reaction with
trimethylphosphonoacetate to give 11.7 in 66% yield for
the two steps. Oxidation of the primary hydroxyl group in
11.7 to the aldehyde, followed by addition of phenyl-
magnesium bromide gave diastereomeric benzylic alcohols
11.8 in 67% overall yield for the two steps from 11.7. The
route was designed to allow introduction of the phenyl
group toward the end of the unit A synthesis. In this way,
a series of cryptophycin phenyl ring analogs can be prepared
from a common precursor. The elimination of water from
benzylic alcohol 11.8 proved to be surprisingly dif®cult. The
optimized conditions involved treatment of 11.8 with
3 equiv. of methanesulfonic anhydride and 9 equiv. triethyl-
amine in the presence of catalytic DMAP in dichloro-
The Lilly chemoenzymatic synthesis of unit A is interesting
because it relies upon a non-obvious :to this writer) strategy.
The next unit A synthesis to be discussed makes use of a
classical organic reaction in the key step.
5.3. Synthesis of unit A via a [2,3]-Wittigrearrangement
The comprehensive and far-ranging treatment of the anionic
[2,3]-Wittig rearrangement by Nakai and Mikami31 strongly
suggested that the critical C5±C6 stereochemistry within

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4353
Scheme 12. :a) propargyl bromide, 50% aqueous NaOH, n-Bu4NHSO4 :cat.), 86%; :b) n-BuLi, THF, 2908C to rt, 71%; :c) TBDPSCl, DMF, imidazole, 92%;
:d) 2-methyl-2-butene, BH3´THF, 08C; H2O2, aqueous KH2PO4/K2HPO4 :2.2 M), 08C, 83%; :e) :MeO)2POCH2CO2Me, TMG, THF, 2788C, 92%; :f) O3,
CH2Cl2, pyridine, Sudan 7B, 2788C; Zn, HOAc, 85%; :g) PhCHvPPh3, THF, 2788C to rt; PhSH, VAZO 88, PhMe, re¯ux, 82%; :h) aqueous HF, MeCN, rt,
93%.
unit A would be accessible through the rearrangement of a
suitably substituted, non-racemic allylpropargyl ether.32
The optically enriched starting material for the synthesis
was prepared by incubating commercially available racemic
:E)-3-penten-2-ol with porcine pancreatic lipase :PPL) and
tri¯uoroethyl laurate in anhydrous ether for 80 h.33 This
resulted in a kinetic resolution in which the slow-reacting
:S) enantiomer 12.1 :Scheme 12) could be distilled from the
reaction mixture in 40% yield. This material was .95% ee,
conversion to unit A. Protection of the C5 alcohol as the
TBS ether took place in 92% yield to give silyl ether 12.4.
Selective hydroboration of the terminal alkyne in the
presence of the 1,2-disubstituted alkene with disamylborane
:or catecholborane), followed by oxidation of the inter-
mediate dialkylvinyl boron species, led to aldehyde 12.5
in 83% yield. Horner±Emmons homologation of the
aldehyde to a,b-unsaturated ester 12.6 took place in 92%
yield.
1
as judged by analysis of the H NMR spectrum of the
Mosher ester.34 Exposure of 12.1 to propargyl bromide
and strong base under phase transfer reaction conditions
led to allyl propargyl ether 12.2 in 86% isolated yield.
The conditions for the subsequent anionic [2,3]-Wittig
rearrangement were developed with care. An excess of
n-butyllithium in hexanes was transferred to the reaction
¯ask that was cooled to 2908C. Evaporation of most of
the hexanes under vacuum produced a residue to which
was slowly added a solution of 12.2 in anhydrous THF.
The reaction mixture was allowed to warm from 2908C
to room temperature overnight, and was quenched with
ammonium chloride and worked up to give a 9/1 mixture
of diastereomers in 71% yield :the major diastereomer 12.3,
:3R,4R)-4-methylhept-5-:E)-en-1-yn-3-ol, is shown in
Scheme 12). The solvent exchange was found to be critical
to the success of the Wittig rearrangement. The diastereo-
mers were separated by column chromatography on silica
gel to produce 12.3 in .95% ee. This intermediate requires
elaboration at both ends of the molecule for its eventual
It was now necessary to differentiate between the electron-
poor and electron-rich alkene groups of 12.6. Ozonolysis of
12.6 at 2788C in dichloromethane containing pyridine, and
monitoring the disappearance of the red color of a small
amount of Sudan 7B which was added as an indicator
made it possible to terminate the ozonolysis before signi®-
cant cleavage of the C2±C3 double bond had taken place.35
Reductive workup with zinc powder and acetic acid led to
aldehyde 12.7 in 85% yield.
Wittig reaction with benzylidene triphenylphosphorane led
to an E/Z mixture :ca. 4/1) of styrenes that were isomerized
to the more stable E geometrical isomer 12.8 by re¯uxing
the mixture in toluene in the presence of thiophenol and
1,10-azobis:cyclohexanecarbonitrile) :VAZO 88). The over-
all yield of 12.8 for the two steps from 12.7 was 82%.
Cleavage of the TBDPS protecting group from 12.7 took
place with aqueous HF in acetonitrile, to give unit A
compound 12.9 in 93% yield.
Scheme 13. :a) n-BuLi, PhLi or t-BuLi, 13.2/13.3 ca. 1/2.5.

4354
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Scheme 14. :a) NaH, TBSCl, THF, 08C, 95%; :b) TEMPO :cat.), NaOCl, NaHCO3, CH2Cl2, :pH 9.0±9.2), 91%; :c) :E)-crotyldiisopinocampheylborane,
Et2O, 2788C; :d) H2O2, NaOAc, 67% from 14.3; :e) TBAF, 100%; :f) TEMPO :cat.), NaOCl, NaHCO3, CH2Cl2, :pH 9.0±9.2), 100%; :g) :MeO)2POCH2-
CO2Me, base; :h) KOH, 85% overall from 14.4 :four steps).
This synthesis of unit A, like the Lilly chemoenzymatic
synthesis, follows a `phenyl last' strategy. While this is
well-suited for the preparation of diverse phenyl ring
analogs of unit A from a common precursor, in the present
case one might wonder whether a more ef®cient synthesis of
12.9 might have been possible by incorporating the phenyl
group at the beginning of the sequence. In fact, this strategy
was examined :Scheme 13). Racemic propargyl ether 13.1
was prepared in a straightforward way from phenyl
propenyl ketone in two steps. In this case, all attempts to
perform the anionic [2,3]-Wittig rearrangement failed.
Exposure of 13.1 either to n-butyllithium, phenyllithium
or tert-butyllithium led to a mixture of tertiary alcohol
13.2 and phenone 13.3. None of the desired [2,3]-Wittig
product 13.4 was obtained. Tertiary alcohol 13.2 is
apparently formed by means of a [1,2]-Wittig rearrange-
ment of a benzyl carbanion, whereas rearrangement of the
chemistry at C5 and C6 of unit A.36±38 Michael Martinelli
and coworkers in the Chemical Process Group at Eli Lilly
Company have described a very brief synthesis of unit A
based on this strategy.39 The starting point for the synthesis
is 1,3-propanediol 14.1 :Scheme 14) which was selectively
mono protected to give TBS ether 14.2 in 95% yield. Oxida-
tion of the free hydroxyl group in 14.2 with sodium hypo-
chlorite in the presence of catalytic TEMPO led to aldehyde
14.3 in 91% yield.29 The enantiomerically pure Brown
asymmetric crotylboron reagent was prepared in situ from
:2)-a-pinene of 81% ee, according to the published proto-
cols.36±38 Addition of the crotylboron reagent to 14.3 at
2788C, followed by oxidation with hydrogen peroxide in
the presence of sodium acetate, led to homoallylic alcohol
14.4 in 67% yield and $99% ee and de. Quantitative
removal of the silyl ether protecting group with
¯uoride gave diol 14.5. Oxidation of this material with
hypochlorite and TEMPO, as described previously,
produced b-hydroxyaldehyde 14.6 in quantitative yield.
Homologation via the phosphonate led to methyl ester
14.7 which was hydrolyzed with base to produce carboxylic
acid 14.8 in 85% overall yield for the four steps from 14.4.
The phenyl ring of cryptophycin 1 was introduced by means
of a Heck reaction with iodobenzene at the end of a
synthetic sequence that led to cryptophycin 52, a synthetic
analog of the natural product. Whereas the overall yield of
14.8 was excellent, the Heck process on the macrocyclic
intermediate proceeded in only 31% yield, underscoring
same carbanion through
a vinylogous process can
rationalize the occurrence of phenone 13.3. These results
underscore the idiosyncratic nature of the anionic [2,3]-
Wittig rearrangement. The sensitivity of this process to
solvent, base, and substitution pattern add to the challenge
of incorporating it into synthetic planning.
5.4. Asymmetric crotylboration
The Brown asymmetric crotylboration reaction provides an
excellent means for controlling relative and absolute stereo-
Scheme 15. :a) TBSCl, imidazole, DMF, rt, 93%; :b) O3, 2788C; Me2S, warm to rt, 62%; :c) PhCH2PO:OEt)2, n-BuLi, THF, 2788C to rt, 79%; :d) HF±
pyridine, THF, 99%; :e) Dess±Martin, CH2Cl2, 98%; :f) Ph3PCHCO2tBu, CH2Cl2, 85%.

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4355
once again the dif®culties that are encountered when
attempting to functionalize C7 or C8 once the macrocycle
is in place. The goal of this synthesis was to prepare an
advanced intermediate that could be converted to a series
of phenyl ring analogs, and in this it was highly successful.
stereomeric homoallylic alcohols 16.2 according to Keck's
conditions.45 The ratio of diastereoisomers was improved to
20/1 by conducting the allylstannation at 21008C, accord-
ing to Linderman's conditions, rather than at 2788C.46 The
major isomer 16.2 was isolated in 76% yield. Protection of
the hydroxyl group as the TBS ether led to 16.3 which
was oxidatively cleaved to produce aldehyde 16.4 in 76%
yield. Homologation with :tert-butoxycarbonylmethylene)-
triphenylphosphorane,44 as in White's previously discussed
synthesis of unit A, led to 16.5. Introduction of the styryl
portion of unit A was accomplished through aldehyde 16.7,
which was prepared in the conventional way from 16.5. The
Takai reaction of aldehyde 16.7 with iodoform led to
:E)-vinyl iodide 16.8 in 92% yield.47 This was followed
by a Stille reaction with phenyltrimethylstannane to give
unit A compound 15.5 in 67% yield.48
Georg40 and White41,42 and their respective co-workers have
applied a similar strategy to the synthesis of unit A. White
and co-workers have prepared 14.4 in the same way as the
Lilly group, however, their synthesis diverges thereafter
:Scheme 15). Protection of the free hydroxyl group in
14.4 as the TBS ether led to bis-TBS ether 15.1 in 93%
yield. Ozonolytic cleavage of the terminal ole®n in 15.1
gave aldehyde 15.2 :62% yield). The styryl group was intro-
duced through the Horner±Emmons reaction of the alde-
hyde with diethylbenzylphosphonate.43 The E isomer 15.3
was the exclusive reaction product, isolated in 79% yield.
The benzyl Horner±Emmons reagent is therefore to be
preferred over the corresponding phosphorane :see
12.7!12.8, Scheme 12) which requires an additional
isomerization step. Selective removal of the primary silyl
ether in 15.3 took place in nearly quantitative yield with
HF±pyridine complex. The Dess±Martin periodinane
oxidized alcohol to aldehyde in 98% yield to give 1.8, the
intermediate from the ®rst Hawaii synthesis. Homologation
of 1.8 with :tert-butoxycarbonylmethylene)triphenyl-
phosphorane44 in dichloromethane led to protected unit A
compound 15.5 in 85% yield.
Although the allylation approach to unit A was successful
and ef®cient :14% overall yield from 16.1) the modest dia-
stereoselectivity of the key step, and the attendant need to
separate the minor diastereoisomer from the mixture,
diminished the appeal of this route, and prompted White
and coworkers to develop the synthesis of Scheme 15.
5.6. A chiral auxiliary approach
A conceptually distinct strategy for the synthesis of unit A
was developed by Kobayashi and co-workers in the context
of their improved synthesis of arenastatin A.49 The arena-
statins were isolated from the Okinawan sponge Dysidea
arenaria, and they are structurally similar to the crypto-
phycins.50,51 Unit B of the arenastatins is d-O-methyltyro-
sine, which is easily prepared from the commercially
available unnatural amino acid, whereas unit C of the
arenastatins is achiral b-alanine. For these reasons arena-
statin synthesis is less challenging than cryptophycin
synthesis. Although the arenastatins have good in vitro
anti-tumor activity, they are essentially inactive in vivo.
This is almost certainly in large part a consequence of the
much greater propensity for hydrolytic cleavage of the unit
C±unit D ester link in the arenastatins relative to the crypto-
phycins. As a result of the attenuated activity of the arena-
statins, whatever interest their synthesis has generated has
been related to their utility as simpler cryptophycin analogs.
The White synthesis is noteworthy for several reasons. The
overall yield is high: 20% from 14.3. Also, the use of
diethylbenzylphosphonate de®nes a useful alternative to
the conventional Wittig reagent for installing the styrene
function. In addition to the enantioselective crotylboration
approach, White has also disclosed a related strategy, based
on a diastereoselective allylation of a chiral, non-racemic
aldehyde. The allylstannation approach will be discussed in
Section 5.5.
5.5. Diastereoselective allylstannation
The allylstannation synthesis of unit A by White and
co-workers makes use of :R) aldehyde 16.1 :Scheme 16).41
This compound was converted to a 11/1 mixture of dia-
Scheme 16. :a) Bu3SnCH2CHvCH2, SnCl4, CH2Cl2, 21008C, 76% :20/1); :b) TBSCl, DMF, imidazole, 91%; :c) cat. OsO4, NaIO4, aqueous THF, 76%;
:d) Ph3PvCHCO2tBu, CH2Cl2, 95%; :e) DDQ, CH2Cl2, H2O, 08C, 92%; :f) Dess±Martin, CH2Cl2, 92%; :g) CHI3, CrCl2, THF, 08C, 92%; :h) PhSnMe3,
PdCl2:MeCN)2, DMF, 67%.

4356
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Scheme 17. :a) Bu2BOTf, Et3N, CH2Cl2; MeOCH2O:CH2)2CHO, 95% :de.99%); :b) Bu3B, AcOH, LiBH4, THF, 80%; :c) TsCl, pyr; :d) LiBEt3H, THF, 88%
:two steps).
The synthesis of Kobayashi and co-workers is summarized
in Scheme 17. The diastereoselective aldol reaction through
the dibutylboryl enolate of 17.1 with b-:methoxy)methoxy-
propanal led to 17.2 in 95% yield :de.99%). Reduction of
imide to primary alcohol 17.3 took place in 80% yield with
lithium borohydride in the presence of dibutylacetoxy-
borane. Selective tosylation of the primary hydroxyl group
in 17.3, followed by reduction with lithium triethylboro-
hydride led to 17.4 in 88% yield for the two steps. This
compound was not homologated to unit A, because
Kobayashi and co-workers used a strategy that does not
proceed through a discreet unit A compound. Although
this unit A synthesis is highly stereoselective, the need to
adjust the oxidation state of the group at C6 adds to the
number of steps.
alkylation, led to anti-product in 74% yield.54 The diastereo-
meric excess of the product was determined to be 92% by
integration of the signals for the C3 methine proton in the 1H
NMR spectrum of the reaction mixture. Hydrogenolytic
cleavage of the benzyl ether protecting group, followed by
protection of the primary alcohol as the TBS derivative, led
to 18.4 in 93% yield for the two steps. The reduction of the
ester group in 18.4 with DIBAL-H in THF led to the antici-
pated primary alcohol in 90% yield. This was followed by
oxidation with TPAP/NMO55 to produce aldehyde 15.2 in
88% yield. This is the same intermediate that White uses,
and from this point on, Georg's synthesis parallels White's.
Since the development of catalytic methods for performing
diastereo- and enantioselective aldol reactions, as well as
Â
ones that make use of chiral auxiliaries, the Frater alkylation
A closely related synthesis of unit A has been described by
Ghosh and Bischoff, the main point of difference between
this work and the route that is summarized in Scheme 17
being the use of :1)-:1R,2S)-1-:N-tosylamino)-2-indanol as
the chiral auxiliary group.52
has not seen much use. Georg's synthesis provides a nice
illustration of this method's enduring utility in synthesis.
5.8. 8S)-82)-2-Acetoxysuccinic anhydride as a starting
material for unit A
Â
5.7. Noyori hydrogenationÐFrater alkylation
A very unconventional unit A synthesis has been described
Â
by Lavalee and co-workers at BioChem Therapeutics, Inc.
Georg and co-workers40 have made clever use of the Noyori
asymmetric hydrogenation in order to control the absolute
stereochemistry at C5 of unit A :Scheme 18).53 Methyl
5-benzyloxy-3-oxopentanoate 18.1 was hydrogenated at
50 psi and 508C for 5 h in methanol in the presence of
:S)-BINAP/RuBr2 to give the :R) alcohol 18.2 in 97%
yield and 97% ee, as determined by chiral HPLC analysis.
Exposure of 18.2 to LDA in THF, followed by a solution of
:Scheme 19).56 The sole source of asymmetry for unit A was
:S)-2-acetoxysuccinic anhydride 19.1. Regioselective ring
opening at C1 of 19.1 with lithium phenylacetylide was
followed by non-stereoselective reduction of the inter-
mediate ynone by sodium borohydride. Base-mediated
hydrolysis of the acetate led to a 1/1 mixture of diastereo-
meric diols 19.2 in 73% yield. Exposure of this mixture to
tosic acid in benzene at 508C led to diastereomeric lactones
19.3 :43% yield) and 19.4 :45% yield). The overall yield of
Â
iodomethane in HMPA, the conditions for the Frater
Scheme 18. :a) H2, :S)-BINAP-RuBr2, 97% :97% ee); :b) LDA, HMPA, MeI, THF, 74% :92% de); :c) H2, Pd/C, THF; :d) TBSCl, DMF, imidazole, 93% :two
Ê
steps); :e) DIBAL-H, THF, 278 to 2108C, 90%; :f) TPAP, NMO, CH2Cl2/MeCN :10/1), 4A sieves, 88%.

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4357
Scheme 19. :a) Li±phenylacetylide, THF, 2788C; :b) NaBH4, EtOH, 2788C; :c) NaOH, 08C, 73% :three steps); :d) TsOH, PhH, 508C, 63% overall from
phenylacetylene; 43% for 19.3, 45% for 19.4; :e) DHP, TsOH, THF, rt, 92%; :f) LAH, Et2O, rt, 100%; :g) PivCl, pyr, DMAP, 08C to rt; Ac2O, 59%; :h) HOAc,
H2O, rt, 3 h, 100%; :i) Me2CuLi, Et2O, 08C, 34%; :j) DHP, TsOH, THF, rt, 98%; :k) LAH, THF, 08C, 97%; :l) DMSO, TEA, :COCl)2, CH2Cl2, 278 to 08C;
:m) Ph3PvCHCO2tBu, CH2Cl2, 73% :two steps); :n) HOAc, H2O, 408C, 80%; :o) H2, Lindlar, quinoline, EtOAc, MeOH, 08C, 85%; :p) MeLi, CuBr´Me2S,
Et2O, 2358C; :q) DHP, TsOH, THF, rt; :r) LAH, 08C; :s) DMSO, TEA, :COCl)2, 278 to 08C; :t) Ph3PvCHCO2tBu, CH2Cl2; :u) HOAc, H2O, 408C, 23%
overall from 19.9.
lactones from phenylacetylene was 63%. It is somewhat
unusual that the homochiral building block was used in
excess, whereas phenylacetylene was the limiting species.
Conversion of 19.8 to unit A compound 10.6, the same one
that Sih subsequently reported in 1996,24 took place in ®ve
steps. The C5 alcohol was protected once again as the THP
derivative :98% yield), and the pivaloate was cleaved reduc-
tively with lithium aluminum hydride :97% yield). Swern
oxidation of the primary alcohol to the aldehyde, followed
by exposure to :tert-butoxycarbonylmethylene)triphenyl-
phosphorane44 led to the tetrahydropyranyloxy tert-butyl
ester in 73% yield for the two steps. Hydrolytic cleavage
of the THP protecting group with aqueous acetic acid led to
unit A compound 10.6 in 80% yield.
Each of the diastereomeric lactones 19.3 and 19.4 was inde-
pendently converted to the same unit A fragment. The route
that proceeds from 19.3 will be discussed ®rst. Protection of
the free hydroxyl group in 19.3 as the tetrahydropyranyloxy
ether :92% yield) was followed by quantitative reduction of
the butyrolactone with lithium aluminum hydride to
produce diol 19.5. Sequential pivaloylation of the primary
hydroxyl group, followed by acetylation of the secondary
hydroxyl in situ, led to fully protected triol 19.6 in 59%
yield. Selective hydrolysis of the THP protecting group
with aqueous acetic acid took place in quantitative yield
to give 19.7. Exposure of this allylic acetate to lithium
dimethyl cuprate in ether at 08C led to 19.8 in 34% yield.
Unfortunately, the key cuprate step was not regioselective,
and led to a 1/1 mixture of SN2 :desired) and SN20
:undesired) products. The cuprate reaction of the fully
protected intermediate 19.6 failed to provide a pathway to
19.8, leading instead to a mixture of conjugated dienes from
a reduction/elimination process. This made it necessary to
deprotect the hydroxyl group at C5 :unit A numbering) only
to reprotect it following the cuprate step :vide infra). This all
contributes to the length of the synthesis.
A slightly shorter synthesis of 10.6 is possible from 19.4.
Semihydrogenation of the alkyne group over Lindlar's cata-
lyst led to Z styryl butyrolactone 19.9 in 85% yield. Methyl
cuprate addition to 19.9 took place with retention of stereo-
chemistry at the reacting carbon atom, but complete conver-
sion to the E geometry of the styrene was observed in
product 19.10. Protection of the C5 hydroxyl group as the
THP ether, followed by reduction of the carboxylic acid to
the primary alcohol with lithium aluminum hydride led to
19.11. This material was elaborated to 10.6 through the
same three steps :Swern oxidation; Wittig homologation;
hydrolysis) as has been described for the synthesis that
proceeds through intermediate 19.8. The overall yield of
10.6 from 19.9 was 23%.

4358
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Ê
Scheme 20. :a) Ph3PvCMeCO2Et, THF, 08C, .74%; :b) DIBAL-H, Et2O, 08C, 97%; :c) Ti:OiPr)4, :1)-DET, t-BuOOH, CH2Cl2, 2258C, 4A sieves, 95%;
:d) DMSO, TEA, :COCl)2, CH2Cl2, 2788C, 1.5 h; Et3N, rt, 1.5 h, 80%; :e) PhCHvPPh3, 21008C, THF, 10 h, rt, 67%; :f) Pd2dba3´CHCl3, nBu3P, HCO2H,
TEA, 1,4-dioxane, rt, 10 h, 91%; :g) AlCl3, CH2Cl2, rt, 30 min, 75%; :h) TBSCl, DMAP, imidazole, DMF, rt, 8 h, 94%; :i) HOAc, H2O, THF :1/1/2), 72 h, rt,
85%; :j) DMSO, TEA, :COCl)2, CH2Cl2, 2788C, 1 h; Et3N, rt, 1 h; :k) :MeO)2POCH2CO2Me, NaH, THF, rt, 4h, 72% :two steps).
Some total syntheses are noteworthy because of their
brevity and/or their ef®ciency, whereas others are distin-
guished by a non-obvious strategy, or by their use of unusual
[1-:ethoxycarbonyl)ethyl]triphenylphosphorane in THF at
08C :.74% yield), followed by reduction of the ester func-
tion with DIBAL-H in ether to give primary alcohol 20.2
:97% yield). This material was converted to the non-
racemic epoxide under the Sharpless conditions, using
:1)-diethyl tartrate as the chiral inducer :95% yield).
Swern oxidation to epoxyaldehyde 20.3 took place in 80%
yield. Exposure of 20.3 to benzylidenetriphenylphosphor-
ane at 21008C, followed by warming to room temperature
led to Z styryl compound 20.4 in 67% yield. The key step
in the sequence is the palladium mediated reduction±
isomerization step that converts 20.4 into 20.5, and which
sets the relative and absolute stereochemistry at C5 and C6
:unit A numbering). The conversion to 20.6 took place in
91% yield upon exposure to catalytic Pd:0), tri-n-butyl-
phosphine and triethylammonium formate. The stereo-
chemistry requires some comment. The process is initiated
by the :conventional) epoxide ring opening with formation
of a p-allyl palladium species. Interconversion between this
p-allyl intermediate and a s-bonded palladium species is
reversible, allowing for rotation about C7±C8 to take place.
Relief of non-bonded interactions between the phenyl and
methyl groups favors one of the two rotamers of the
s-bonded intermediate. Conversion to another p-allyl
species, followed by the reductive elimination of
palladium, completes the catalytic cycle :Fig. 3). This is
very clever.
Â
reagents or reaction conditions. The Lavalee synthesis falls
in the second category. The use of non-racemic acetoxy-
succinic anhydride for unit A synthesis is unique to this
research group. One suspects that there are other syntheses
of unit A waiting to be discovered which make use of this
chiral pool starting material. Finally, both pathways to unit
A which have been summarized in Scheme 19 could
probably have been made a bit shorter by forgoing the
protection±deprotection of the C5 hydroxyl :Scheme 19,
protection steps :j) and :q), and deprotection steps :n) and
:u)). To wit, selective oxidation of the primary hydroxyl
group in diol 14.5 :Scheme 14) by the hypochlorite/
TEMPO reagent proceeds in good yield, as does the subse-
quent Horner±Emmons reaction that leads to 14.7. These
results suggest that protection at C5 in 19.8 and 19.10 may
not have been necessary.
5.9. Vinyl epoxide reduction
Like the original Hawaii work, the unit A synthesis of
Furuyama and Shimizu makes use of the Sharpless
asymmetric epoxidation for control of the absolute stereo-
chemistry.57 The para-methoxybenzyl derivative 20.1 of
3-hydroxypropanal :Scheme 20) was treated ®rst with
Figure 3. The proposed mechanism for the conversion of 20.4 to 20.5.

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4359
Scheme 21. :a) BH3´Me2S, THF, rt; NaBH4, THF, rt, 89%; :b) TIPSCl, imidazole, DMF, 2308C, 71%; :c) K2CO3, MeOH, heat, 70%; :d) H2CvCMeOMe,
PPTS, CH2Cl2, 84%; :e) DIBAL-H, CH2Cl2, 2788C; pyr, DMAP, Ac2O, 84%; :f) PhSH, ZnCl2, CH2Cl2, 2308C, 87%; :g) LDBB, Li, THF, 2788C;
:h) Bu3SnCl, THF, 2788C, 71% from 21.4; :i) PhCH:OMe)2, TsOH, CH2Cl2, rt, 89%; :j) nBuLi, THF, 2788C; :k) CuBr´Me2S, 2788C; add 21.7; :l) O2,
CHCl3, rt, 71% from 21.6; 21.8/21.9ˆ1.2/1; :m) TsOH, MeOH, rt, 63%; :n) chromatographic separation; :o) NaIO4, MeOH, H2O :3/1), rt; :p) :MeO)2POCH2-
CO2Me, TMG, THF, 2788C to rt, 83% :two steps), :E/Z$25/1).
From 20.6, the synthesis follows familiar ground. Protection
of both hydroxyl groups as the TBS ethers :94% yield),
followed by selective hydrolysis of the primary TBS
group with aqueous acetic acid :85% yield) led to 15.4,
the same intermediate that White reported. Swern oxidation
of 15.4 to the aldehyde, followed by Horner±Emmons
homologation, led to 1.9, the unit A compound of the ®rst
Hawaii synthesis, in 72% yield for the two steps from 15.4.
For the reason discussed in the preceding section, it was
probably not necessary to protect the C5 alcohol.
molybdenum complex, and does not appear in unit A. The
synthesis is outlined in Scheme 21.
Selective reduction of dimethyl malate 21.1 according to
Moriwake's procedure was followed by protection of the
primary hydroxyl group as the TIPS derivative and saponi-
®cation of the C4carboxylate. 59 This led to b-hydroxyacid
21.2 in 53% overall yield from 21.1. This compound was
converted to a dioxanone :84% yield) that was reduced with
DIBAL-H and immediately acetylated according to
Rychnovsky's conditions :84% yield) to produce a dia-
stereomeric mixture of O-acetyl acetals 21.3.60 Trans-
acetalization with benzenethiol was catalyzed by zinc
chloride to give a 9/1 mixture of diastereomeric thioacetals
21.4 in 87% yield. Reductive lithiation with 4,40-di-tert-
butylbiphenyl :LDBB) and lithium led to an intermediate
a-alkoxylithium species which was intercepted with tri-n-
butyltin chloride to produce axial tributylstannane 21.6 in
71% yield, epimerization having taken place at the stage of
5.10. Planar chiral molybdenum cationic complex
One of the more unusual syntheses of unit A is due to
Kocienski and co-workers.58 It makes use of methyl
:S)-malate as the starting material, however, the stereogenic
center in the malate is used as a control element for a dia-
stereoselective alkylation step of a planar chiral cationic

4360
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
the lithium species.61 Transacetalization of the acetonide
with benzaldehyde dimethyl acetal in the presence of
catalytic tosic acid led to benzylidene acetal 21.6 as a chro-
matographically separable 7/1 mixture of diastereomers at
the acetal carbon atom in 89% yield.
de®ne the limits and to test the scope of applicability of
new methodology. The unit A synthesis by Kocienski and
co-workers should be judged in this light. The lack of regio-
selectivity in the key step diminishes the preparative utility
of this route. However, had preparative utility been the over-
riding concern, the desired C5±C6 stereochemistry could
have been secured in one step, through stereospeci®c
methylation of the commercially available enantiomer of
21.1 according to Seebach's published procedure.65,66
The key steps in the synthesis of unit A by Kocienski and
co-workers took place through a series of transmetalations
of 21.6. Exposure to n-butyllithium at 2788C led to the
lithio species, which was treated with cuprous bromide to
produce the axial copper species. To this was added cationic
molybdenum complex 21.7. The axially chiral molybdenum
complex was prepared as a diastereomeric mixture from :S)-
4-phenyl-:E)-3-buten-2-ol.62,63 Oxidative decomplexation
of the molybdenum led to a 1.2/1 mixture of products
21.8 and 21.9 in 71% yield. The organocopper species
that was derived from 21.6 evidently did not discriminate
between the methyl- and phenyl-substituted termini of
complex 21.7.64 Unfortunately, 21.8 and 21.9 were not
separable by chromatography, so they were carried through
to the next step as a mixture. Exposure of this mixture to
tosic acid in methanol at room temperature led to a mixture
of triol 21.10 and 21.11 in 63% yield. These were separated,
and 21.10 was treated sequentially with sodium periodate to
generate the C3 :unit A numbering) aldehyde. The aldehyde
was elaborated to unit A compound 12.9 in 83% yield for
the two steps from 21.10. Note that the stereogenic center
that was derived from dimethyl malate is lost during the
oxidative cleavage step.
5.11. Diastereoselective addition to 8R)-mandelaldehyde
A synthesis of unit A that controls the stereochemistry at C7
and C8 is due to Larcheveque and co-workers.67 The starting
Ã
material was the methoxymethyl ether 22.1 of :R)-methyl
mandelate :Scheme 22). Reduction of the ester function
with DIBAL-H gave the protected mandelaldehyde which
underwent chelation-controlled addition of the bromo-
magnesium derivative of protected propargyl alcohol to
give 22.2 in 87% overall yield for the two steps from
22.1. Careful reduction of the propargyl alcohol with
RedAlw in ether at 2258C led to E allylic alcohol 22.3 in
72% yield. The ®rst strategy for introduction of the C6
methyl group failed. Epoxidation of 22.3 under the
Sharpless conditions failed to take place, whereas vanadyl
acetoacetate-catalyzed epoxidation with tert-butyl hydro-
peroxide led to the anti-epoxide diastereomer with very
low diastereomeric excess. Moreover, epoxide ring-opening
by lithium dimethylcuprate was dif®cult, consequently the
synthetic strategy was modi®ed.
One of the uses of natural products total synthesis is to
Scheme 22. :a) DIBAL-H, Et2O/pentane :1/1), 2788C; :b) BrMgCCCH2OTBDPS, Et2O, 87% :two steps); :c) Red-Al, Et2O, 2258C, 72%; :d) 2 M HCl,
Ê
MeOH, 408C, 2 d, 92%; :e) CuSO4 :anh), TsOH, acetone, rt, 2 d, 91%; :f) Ti:OiPr)4, :1)-DIPT, 2208C, 4A sieves, CH2Cl2, tBuOOH, 4d, 76% :95/5);
:g) hexane/CH2Cl2, :5/1), Me3Al, 2108C, 1 h; rt, 20 h, 44118% 22.6; :h) TMSCl, MeC:OMe)3, CH2Cl2, 08C, 1H, MeOH, K2CO3, 2 h, 65%; :i) MeCN :anh),
LiClO4, KCN, 708C, 6 h, 70%; :j) TBSOTf, 2,6-lutidine, CH2Cl2, 08C, 92%; :k) PhMe, DIBAL-H, 2788C, 58%; :l) :MeO)2POCH2CO2Me, TMG, THF,
2788C to rt, 78%.

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4361
Global deprotection of 22.3 by heating to 408C in methanol
with 2 M hydrochloric acid for two days led to triol 22.4 in
92% yield. The syn diol portion of the molecule was
protected as acetonide 22.5 :91% yield). Sharpless asym-
metric epoxidation, using :1)-diisopropyl tartrate as the
chiral inducer, led to a 95/5 mixture of diastereomeric
epoxy alcohols from which 22.6 was isolated in 76%
yield. Introduction of the C6 methyl group was accom-
plished by exposing 22.6 to trimethylaluminum over a
prolonged period. After 20 h at room temperature, 22.7
was isolated in 44% yield, along with 18% of the starting
material 22.6. The low reactivity of 22.6 is most likely due
to competition for the Lewis acidic reagent by the ketal
solution to the problem is to oxidize 23.2 to the syn-b-diol,
and then to convert diol to epoxide. Unfortunately, all
efforts to convert styrene 23.2 to diol 23.5 failed. Osmium
tetroxide, catalytic or stoichiometric, gave roughly equi-
molar mixtures of the four diastereomers resulting from
indiscriminate attack at each of the two carbon±carbon
double bonds in 23.2. Ruthenium-catalyzed dihydroxylation
also failed, and led to destruction of 23.2 with the formation
of aldehydes. The results from the Sharpless AD which were
obtained with the dihydroquinidine based phthalazine
ligand :DHQD)2PHAL were disappointing because of the
insolubility of 23.2 in the reaction solvent, aqueous tert-
butanol.68
Ã
oxygen atoms. Although Larcheveque and co-workers do
not provide an explanation for the low material balance
for the epoxide ring-opening reaction, it seems likely that
in the presence of trimethylaluminum some cyclization of
22.7 to a tetrahydrofuran byproduct takes place. This ®ts the
reactivity pattern for the benzylic carbon atom of unit A that
has been described by others :see 1.10, 8.10 and 11.9).
The sum of these negative results suggested that the styryl
carbon±carbon double bond in cyclic compound 23.2 would
be very dif®cult to functionalize, and that it was unlikely
that this could be done either regio- or stereoselectively. The
obvious way to circumvent this problem would be to
perform the oxidation on a ring-open compound, such as
23.1. Scheme 23 summarizes the successful route that is
based on this premise. The Sharpless AD of 23.1 with
:DHQD)2PHAL in the presence of catalytic K2OsO2:OH)2
led to the desired syn diol 23.4 in 61% yield. The conditions
for the AD had to be optimized carefully, and differ a little
from the conventional conditions. Even with 23.1, the reac-
tion is slow and 1±2 mol% osmium must be used, rather
than the 0.2 mol% osmium that suf®ces for many alkenes.
One must use a correspondingly larger amount of the chiral
ligand, at least 2 mol%. Decreasing the concentration of
osmium and/or the ligand results in slower kinetics and an
inferior yield of 23.4, because of exposure of starting
material and product to the basic reaction conditions over
a prolonged period of time. Buffering the reaction mixture,
adding the base slowly, or adding more than 1 equiv. of
methanesulfonamide as an accelerant failed to improve
the yield.
Diol 22.7 was converted to epoxide 22.8 through Sharpless'
conditions in 65% yield.19 One-carbon homologation of
22.8 with potassium cyanide in the presence of lithium
perchlorate led to nitrile 22.9 :70% yield). Protection of
the free C5 hydroxyl group as the TBS ether :92% yield),
followed by reduction with DIBAL-H in toluene :58%
yield) led to aldehyde 22.11. Immediate reaction with
trimethylphosphonoacetate with TMG in THF produced
unit A compound 22.12 in 78% yield. The overall yield of
the sequence of reactions leading from 22.1 to 22.12 was
3.3%, the largest loss of material having taken place at the
point of introduction of the C6 methyl group. Although this
synthesis has the advantage of being able to control C7±C8
stereochemistry from the outset, the low overall yield
renders it an unattractive choice for preparative scale
work. An alternative approach to the control of C7±C8
stereochemistry is given in Section 6.
The conversion of seco diol 23.4 to cryptophycin 52 was
accomplished in excellent yield, once the optimal reaction
conditions were de®ned. Acidolysis of the Boc protecting
group with tri¯uoroacetic acid in dichloromethane took
place during 1.5 h at 08C. It is noteworthy that no side
reactions involving a benzylic carbocation took place during
the treatment with acid. Macrolactamization was accom-
plished with 2-hydroxypyridine to give 23.5 in 82% yield
for the two steps from 23.4. Diol to epoxide conversion took
place through a modi®ed Sharpless procedure.69 Trimethyl
orthoformate in the presence of PPTS converted the diol to
the cyclic orthoformate, which upon exposure to iodo-
trimethylsilane at 08C and reductive workup furnished iodo-
formate 23.6 in 93% overall yield from diol 23.5. Potassium
carbonate in THF/methanol at 08C gave cryptophycin 52 in
98% yield. The diol to epoxide conversion was also accom-
plished through the use of acetyl bromide in place of iodo-
trimethylsilane. This led to a bromohydrin formate, which
was converted in high yield to cryptophycin 52. Trimethyl
orthoformate was used in all successful cases, rather than
trimethyl orthoacetate: if one uses orthoacetate, and
performs the next step with chlorotrimethylsilane, the
product is a chlorohydrin acetate, hydrolytic cleavage of
which with base requires more vigorous conditions than
does cleavage of the formate. Inevitably, some base
6. Dihydroxylation of seco-cryptophycin 52
A solution to the stereochemical problem posed by the
epoxide in unit A was solved in the context of the total
synthesis of cryptophycin 52, a hydrolytically stable analog
of the natural product :Scheme 23).25 The only difference
between cryptophycins 1 and 52 is in unit C. In crypto-
phycin 52, unit C bears an additional methyl group on the
a carbon atom. The small increase in steric hindrance
engendered by the quaternary carbon atom of the modi®ed
unit C diminishes the rate of hydrolytic cleavage of the unit
C±unit D ester link. Fray18 has shown that brief exposure of
seco-cryptophycin 52 23.1 :Scheme 23) to neat tri¯uoro-
acetic acid at room temperature selectively removes the
Boc protecting group. Macrocyclization is mediated by
2-hydroxypyridine, yielding 23.2 in 80% yield for the two
steps from 23.1. Epoxidation of 23.2 with m-CPBA gave at
best a 2/1 mixture of cryptophycin 52 along with the
a-epoxy diastereomer 23.3. Following separation of the
two diastereoisomers by HPLC, cryptophycin 52 could be
isolated in 51% overall yield from 23.1. The need for a
separation by HPLC in the last step, and the loss of material
from the epoxidation are major shortcomings. An obvious

4362
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Scheme 23. :a) TFA :neat), rt, 15 min, 0.5 M NaOH; :b) 2-hydroxypyridine :2.0 equiv.), PhMe, 0.02 M, 208C, 80% from 23.1; :c) m-CPBA, CH2Cl2, 23.3/
cryptophycin 52ˆ1/2, 51% of cryptophycin 52 overall from 23.1; :d) K2OsO2:OH)2, :DHQD)2PHAL, K2CO3, K3Fe:CN)6, MeSO2NH2, tBuOH/H2O :1/1), rt,
20.5 h, 61%; :e) TFA, CH2Cl2, 08C, 1.5 h; aqueous K2CO3; :f) 2-hydroxypyridine :2.0 equiv.), MeCN/PhMe :1/1), 408C, 21 h, 82% :two steps); :g) PPTS,
CH2Cl2, :MeO)3CH, rt; aqueous NaHCO3; :h) CH2Cl2, 08C, TMSI, 45 min; 10% aqueous Na2S2O3, 93% :two steps) :i) THF/MeOH :1/1), K2CO3, 08C, 6 h,
98%.
catalyzed ring-opening of the macrocycle also takes place
under these conditions, hence the preference for the
orthoformate.
deactivation through hydrolytic cleavage of the ester
bonds prompted the design and synthesis of analogs that
would retain the high activity of the natural product while
being resistant to this decomposition pathway. Crypto-
phycin 52 accomplished this goal by introducing steric
hindrance into unit C, thereby stabilizing the unit C±unit
D link. An alternative strategy for stabilizing the crypto-
phycins against hydrolytic cleavage is by substituting
amide links for ester links. In this section, the synthesis by
the Hawaii group and aspects of the reactivity of 1-aza-
cryptophycin 1 is be discussed.70
The protocols that are summarized in Scheme 23 underscore
an unanticipated feature of cryptophycin reactivity. The
styrene function is held in close proximity to unit D in the
cyclic compounds, and is not easily accessible to
electrophilic species. Molecular modeling con®rms this.
Section 7 serves to illustrate a consequence of the confor-
mational rigidity of the cyclic cryptophycins in a different
context.
Scheme 24summarizes the synthesis of 24.11, the aza
analog of unit A. This synthesis controls the epoxide stereo-
chemistry from the outset, by using :R)-methyl mandelate
24.1 as the starting material and as the sole source of
asymmetry for unit A. Conversion of methyl ester 24.1 to
Weinreb amide 24.2 was mediated by trimethylaluminum
7. 1-Aza-cryptophycin 1, an unstable cryptophycin
The importance of the cryptophycins as potential cancer
chemotherapeutic agents and the propensity for in vivo

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4363
Scheme 24. :a) HNMe:OMe)´HCl, Me3Al, PhH, 0±58C, 2 h; add 24.1, rt, 12 h, 93%; :b) TBSOTf, Et3N, CH2Cl2, 08C, 3 h, 93%; :c) EtMgBr, THF, 5 h, 94%;
:d) Bu2BOTf, DIPEA, CH2Cl2, 2788C, 15 min; 08C, 2 h; add 3,3-diethoxypropanal, 2788C, 1 h, 08C, 2 h, MeOH, H2O2, 08C, 2 h, 65%; :e) Cl3CCO2H,
CH2Cl2/H2O :20/1), rt, 1 h, 87%; :f) :EtO)2POCH2CO2CH2CHvCH2, LiCl, DIPEA, MeCN, rt, ca. 4h, 63%; :g) TsCl, KOH :s), THF, 0 8C to rt, 86%;
:h) NaBH4, CeCl3´7H2O, EtOH, 08C, 96%; :i) tetramethylguanidinium azide, MeNO2, 70±808C, 5 h, 69%; :j) PPh3, THF, H2O, 50±608C, 12 h.
:93% yield). Protection of the hydroxyl group in 24.2 as the
TBS ether :93% yield), followed by exposure of the product
to ethylmagnesium bromide led to ethyl ketone 24.4 :94%
yield). Condensation of the di-n-butylboryl enolate71,72
derived from 24.4 with 3,3-diethoxypropanal,73,74 followed
by oxidative workup, led exclusively to syn aldol 24.5 in
65% yield following puri®cation by ¯ash column chromato-
graphy. Condensation of 24.4 with 3,3-diethoxypropanal
through the lithium enolate, generated with LDA, led to a
ca. 10/1 mixture of syn and anti-aldol products. Acid
hydrolysis of the diethyl acetal function proceeded to give
aldehyde 24.6 in 87% yield. Elimination of the C5 :unit A
numbering) hydroxyl, which is b to both the ketone and
aldehyde functions, did not take place to any appreciable
extent. Horner±Emmons homologation of 24.6 under
Masamune±Roush conditions gave allyl ester 24.7 in 63%
yield. There are two remaining tasks that must be performed
on this intermediate to transform it to unit A. First, the C5
nitrogen atom must be introduced with inversion of stereo-
chemistry. Second, stereoselective reduction of the C7 keto
group to either a or b alcohol can lead to the desired a
epoxide. Introduction of the nitrogen atom was done
classically: conversion of alcohol to tosylate 24.8 :86%
yield) was followed by Luche reduction of the ketone
with 20/1 stereoselectivity :96% yield).75 Exposure of
alcohol 24.9 to tetramethylguanidinium azide :69% yield)
led to 24.10. Reduction of azide 24.10 to primary amine
24.11 took place with triphenylphosphine in aqueous THF.
overall yield from 24.10. Palladium catalyzed cleavage of
the allyl ester produced carboxylic acid 25.2 in 91% yield.
Carboxylate 25.2 was coupled with protected unit B 2.2 to
give protected seco compound 25.3 in 69% yield. Exposure
of this material to tri¯uoroacetic acid at 08C, followed by
evaporation to dryness and resuspension in toluene contain-
ing 2-hydroxypyridine, led to 25.4 in 50% overall yield.
Treatment with tri¯uoroacetic acid cleaved both Boc and
TBS protecting groups, however, reactions involving the
intermediacy of a benzylic carbocation did not take place
to any appreciable extent. The trichloroethyl ester group in
25.3 served two functions, as before: it masked the carboxy
group in unit B during the coupling step, and it activated the
same carbonyl group for the macrolactamization.
Conversion of the syn diol unit in 25.4 to the epoxide was
accomplished through the orthoformate, which was treated
with acetyl bromide to produce the anticipated formyloxy
bromide.25 Exposure of this compound to potassium
bicarbonate led to an approximately equimolar mixture of
two compounds, one of which was the desired epoxide, and
another was identi®ed as 25.5. The conversion of 1-aza-1-
cryptophycin 1 to 25.5 took place spontaneously, even upon
storage at 258C. Chromatographic separation on basic
alumina, or reverse phase HPLC produced only mixtures
of the two compounds. Although a pure sample of 1-aza-
1-cryptophycin could not be obtained, its presence could be
inferred from the characteristic signals at 3.0 :H7) and 3.7
1
:H8) ppm in the H NMR spectrum of the mixture.
Assembly of the units of 1-aza-cryptophycin 1 is shown in
Scheme 25. The condensation of 24.11 with unit C±unit D
fragment 3.6 was mediated by FDPP to give 25.1 in 74%
The rearrangement of 1-aza-cryptophycin 1 to 25.5 was
certainly not anticipated, however, the mechanism probably

4364
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Scheme 25. :a) 3.6, DMF, FDPP, DIPEA, rt, 5 min, add 24.11, rt, 4h, 74% from 24.10; :b) Pd:PPh3)4, THF, morpholine, rt, 4h, 91%; :c) FDPP, DMF, add 2.2,
Et3N, rt, 5 h, 69%; :d) TFA :neat), 08C to rt, 1 h; evaporate to dryness; 2-hydroxypyridine, PhMe, rt, 20 h, 50%; :e) PPTS, :MeO)3CH, CH2Cl2, rt, 2 h;
:f) MeCOBr, CH2Cl2, 6 h, rt; :g) aqueous NaHCO3; DME, MeOH, EtOH, KHCO3, 4 08C, 15 h; :h) chromatography on silica gel, 62% from 25.4.
follows the pathway indicated in Scheme 25. Activation of
the epoxide is followed by nucleophilic ring opening by the
amide carbonyl oxygen atom of unit D, leading to the seven-
membered ring imino ether 25.5. Release of ring strain of
the epoxide presumably provides suf®cient thermodynamic
driving force. No such rearrangement takes place with
cryptophycin 1. For example, treatment of cryptophycin 1
with ferric chloride leads to the C7 ketone. It therefore
appears that the enforced proximity of the unit D amide
carbonyl oxygen atom to the benzylic carbon atom in
1-aza-cryptophycin 1, as well as the greater electron density
of the amide oxygen atom relative to the ester oxygen atom
of cryptophycin 1, are responsible for the marked difference
in reactivity of these two cryptophycins. This unanticipated
reactivity precludes 1-aza-cryptophycin 1 from considera-
tion as a clinical candidate.
ted at partial or total syntheses that are the subject of this
report. The molecular complexity of the cryptophycins is
modest by contemporary standards, therefore it seems likely
that most of the research activity has been driven by a
realization of the potential that these compounds have as a
new class of chemotherapeutic agents against solid tumors.
Although the intellectual challenges of devising and execut-
ing a laboratory scale synthesis of the cryptophycins are also
modest, the same cannot be said for the design of a synthesis
ef®cient enough to supply clinical trials. Such a route must
be able to produce multi-kilogram quantities of material,
and clearly this is at or near the upper limit of what is
currently feasible in pharmaceutical development. That
the art and practice of organic synthesis has reached
maturity is re¯ected in the diversity of methods and
strategies that have been brought to bear on unit A. These
span the spectrum from classical aldol and Diels±Alder
methods to chemoenzymatic approaches as well as methods
based on carbon±carbon bond forming reactions by means
of transition metal complexes. Given the clinical potential
of this group of natural products, it seems likely that new
and perhaps better cryptophycin syntheses will continue to
be developed for some time to come.
8. Conclusion
The discovery of the cryptophycins and the recognition of
their exceptional potency as anti-microtubule agents has led
to the relatively large number of independent efforts direc-

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4365
22. Danishefsky, S. J.; Pearson, W. H.; Harvey, D. F.; Maring,
C. J.; Springer, J. P. J. Am. Chem. Soc. 1985, 107, 1256.
23. Tius, M. A.; Busch-Petersen, J. Synlett 1997, 531.
24. Salamonczyk, G. M.; Han, K.; Guo, Z.-W.; Sih, C. J. J. Org.
Chem. 1996, 61, 6893.
Acknowledgements
I would like to thank my friend and colleague Dick Moore
for suggesting that I design and direct the ®rst synthesis of
cryptophycin 1. This gave me the opportunity to work with
Thomas Hemscheidt, Russell Barrow, Lianhai Li and other
members of the Hawaii cryptophycin synthesis team. My
involvement with the cryptophycin project also led to the
very enjoyable and stimulating interaction with Michael
Martinelli and the process group at Eli Lilly Company. I
would like to thank the NIH :GM57873) and the National
Oceanic and atmospheric administration, project #R/MP-
16, sponsored by the University of Hawaii Sea Grant
College Program, Institutional Grant No. NA86RG0041,
for support.
25. Liang, J.; Moher, E. D.; Moore, R. E.; Hoard, D. W. J. Org.
Chem. 2000, 65, 3143.
26. Varie, D. L.; Brennan, J.; Briggs, B.; Cronin, J. S.; Hay, D. A.;
Rieck, J. A.; Zmijewski, M. J. Tetrahedron Lett. 1998, 39,
8405.
27. Schreiber, S. L.; Liew, W.-F. Tetrahedron Lett. 1983, 24,
2363.
28. Swarts, H. J.; Verstegen-Hacksma, A. A.; Jansen, B. J. M.;
deGroot, A. Tetrahedron 1994, 43, 251.
29. Anelli, P.; Montanari, F.; Quici, S. Org. Synth. 1990, 69, 212.
30. Krow, G. R. Org. React. 1993, 43, 251.
31. Nakai, T.; Mikami, K. Org. React. 1994, 46, 105.
32. Liang, J.; Hoard, D. W.; Khau, V. V.; Martinelli, M. J.; Moher,
E. D.; Moore, R. E.; Tius, M. A. J. Org. Chem. 1999, 64, 1459.
33. Morgan, B.; Oelschlager, A. C.; Stokes, T. M. J. Org. Chem.
1992, 57, 3231.
References
1. Moore, R. E.; Corbett, T. H.; Patterson, G. M. L.; Valeriote,
F. A. Curr. Pharm. Des. 1996, 2, 317. A review of crypto-
phycin synthesis and anticancer activity has been recently
published: Eggen, M.; Georg, G. Med. Res. Revs. 2002, 22, 85.
2. Tolytoxin: Carmeli, S.; Moore, R. E.; Patterson, G. M. L.
J. Nat. Prod. 1990, 53, 1533.
34. Dale, J. A.; Dull, D. L.; Mosher, H. S. J. Org. Chem. 1969, 34,
2543.
35. Veysoglu, Y.; Mitscher, L. A.; Swayze, J. K. Synthesis 1980,
807.
36. Brown, H. C.; Bhat, K. S.; Randad, R. S. J. Org. Chem. 1989,
54, 1570.
3. Tantazoles: Carmeli, S.; Moore, R. E.; Patterson, G. M. L.;
Corbett, T. H.; Valeriote, F. A. J. Am. Chem. Soc. 1990, 112,
8195.
37. Brown, H. C.; Bhat, K. S. J. Am. Chem. Soc. 1986, 108, 5919.
38. Brown, H. C.; Joshi, N. N. J. Org. Chem. 1988, 53, 4059.
39. Dhokte, U. P.; Khau, V. V.; Hutchison, D. R.; Martinelli, M. J.
Tetrahedron Lett. 1998, 39, 8771.
4. Ambiguines: Smitka, T. A.; Bonjouklian, R.; Doolin, L.;
Jones, N. D.; Deeter, J. B.; Prinsep, M. R.; Yoshida, W.;
Moore, R. E.; Patterson, G. M. L. J. Org. Chem. 1992, 57, 857.
5. Trimurtulu, G.; Ohtani, I.; Patterson, G. M. L.; Moore, R. E.;
Corbett, T. H.; Valeriote, F. A.; Demchik, L. J. Am. Chem.
Soc. 1994, 116, 4729.
40. Eggen, M.; Mossman, C. J.; Buck, S. B.; Nair, S. K.; Bhat, L.;
Ali, S. M.; Reiff, E. A.; Boge, T. C.; Georg, G. I. J. Org. Chem.
2000, 65, 7792.
41. White, J. D.; Hong, J.; Robarge, L. A. J. Org. Chem. 1999, 64,
6206.
6. Schwartz, R. E.; Hirsch, C. F.; Sesin, D. F.; Flor, J. E.;
Chartrain, M.; Fromtling, R. E.; Harris, G. H.; Salvatore,
M. J.; Liesch, J. M.; Yudin, K. J. Ind. Microbiol. 1990, 5, 113.
7. Hirsch, C. F.; Liesch, J. M.; Salvatore, M. J.; Schwartz, R. E.;
Sesin, D. F. US Patent 4,946,835, issued August 7, 1990.
8. Barrow, R.; Hemscheidt, T.; Paik, S.; Liang, J.; Moore, R. E.;
Tius, M. A. J. Am. Chem. Soc. 1995, 117, 2479.
42. White, J. D.; Hong, J.; Robarge, L. A. Tetrahedron Lett. 1998,
39, 8779.
43. Johns, A.; Murphy, J. A.; Sherburn, M. S. Tetrahedron 1989,
45, 7835.
44. Chuang, C.-P.; Hart, D. J. J. Org. Chem. 1983, 48, 1782.
45. Keck, G. E.; Park, M.; Krishnamurthy, D. J. Org. Chem. 1993,
58, 3787.
9. Boons, G.-J.; Castle, G. H.; Clase, J. A.; Grice, P.; Ley, S. V.;
Pinel, C. Synlett 1993, 913.
46. Linderman, R. J.; Cusack, K. P.; Taber, M. R. Tetrahedron
Lett. 1996, 37, 6649.
10. Zeynek, R.; Hoppe-Seyler's, Z. Physiol. Chem. 1926, 144,
247.
Â
11. Dudash, Jr., J.; Jiang, J.; Mayer, S. C.; Joullie, M. M. Synth.
Commun. 1993, 23, 349.
47. Takai, K.; Nitta, K.; Utimoto, K. J. Am. Chem. Soc. 1986, 108,
7408.
48. Stille, J. K.; Groh, B. L. J. Am. Chem. Soc. 1987, 109, 813.
49. Kobayashi, M.; Wang, W.; Ohyabu, N.; Kurosu, M.;
Kitagawa, I. Chem. Pharm. Bull. 1995, 43, 1598.
50. Kobayashi, M.; Kurosu, M.; Ohyabu, N.; Wang, W.; Fujii, S.;
Kitagawa, I. Chem. Pharm. Bull. 1994, 42, 2196.
51. Kobayashi, M.; Aoki, S.; Ohyabu, N.; Kurosu, M.; Wang, W.;
Kitagawa, I. Tetrahedron Lett. 1994, 35, 7969.
52. Ghosh, A. K.; Bischoff, A. Org. Lett. 2000, 2, 1573.
53. Noyori, R. Asymmetric Catalysis in Organic Synthesis; Wiley:
New York, 1994.
12. Jefferey, P. D.; McCombie, S. W. J. Org. Chem. 1982, 47, 587.
13. Gardinier, K. M.; Leahy, J. W. J. Org. Chem. 1997, 62, 7098.
14. Evans, D. A.; Kaldor, S. W.; Jones, T. K.; Clardy, J.; Stout,
T. J. J. Am. Chem. Soc. 1990, 112, 7001.
15. Evans, D. A.; Hoveyda, A. H. J. Am. Chem. Soc. 1990, 112,
6447.
16. Hioki, H.; Okuda, M.; Miyagi, W.; Ito, S. Tetrahedron Lett.
1993, 34, 6131.
17. Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M.
Bull. Chem. Soc. Jpn 1979, 52, 1989.
Â
54. Frater, G. Helv. Chim. Acta 1979, 62, 2825.
55. Ley, S. V.; Norman, J.; Grif®th, W. P. Synthesis 1994, 639.
Â
56. Rej, R.; Nguyen, D.; Go, B.; Fortin, S.; Lavalee, J.-F. J. Org.
18. The macrolactamization is easily carried out with cheap
2-hydroxypyridine: Fray, A. H. Tetrahedron: Asymmetry
1998, 9, 2777.
19. Kolb, H. C.; Sharpless, K. B. Tetrahedron 1992, 48, 10515.
20. Li, L.-H.; Tius, M. A. Org. Lett. in press.
21. Danishefsky, S. J. Chemtracts 1989, 273.
Chem. 1996, 61, 6289.
57. Furuyama, M.; Shimizu, I. Tetrahedron: Asymmetry 1998, 9,
1351.

4366
M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
Ã
67. Pousset, C.; Haddad, M.; Larcheveque, M. Tetrahedron 2001,
58. Christopher, J. A.; Kocienski, P. J.; Kuhl, A.; Bell, R. Synlett
2000, 463.
57, 7163.
68. Kolb, H. C.; Van Nieuwenhze, M. S.; Sharpless, K. B. Chem.
Rev. 1994, 94, 2483.
59. Saito, S.; Hasegawa, T.; Inaba, M.; Nishida, R.; Fujii, T.;
Nomizu, S.; Moriwake, T. Chem. Lett. 1984, 1389.
60. Dahanukar, V. H.; Rychnovsky, S. D. J. Org. Chem. 1996, 61,
8317.
69. Chang, H.-T.; Sharpless, K. B. J. Org. Chem. 1996, 61, 6456.
70. Barrow, R.; Moore, R. E.; Li, L.-H.; Tius, M. A. Tetrahedron
2000, 56, 3339.
61. Rychnovsky, S. D.; Buckmelter, A. J.; Dahanukar, V. H.;
Skalitzky, D. J. J. Org. Chem. 1999, 64, 6849.
62. Tate, D. P.; Knipple, W. R.; Augl, J. M. Inorg. Chem. 1962, 1,
433.
71. Masamune, S.; Choy, W.; Kerdesky, F. A. J.; Imperiali, B.
J. Am. Chem. Soc. 1981, 103, 1566.
72. Evans, D. A.; Nelson, J. V.; Vogel, E.; Taber, T. R. J. Am.
Chem. Soc. 1981, 103, 3099.
63. Faller, J. W.; Linebarrier, D. Organometallics 1988, 7, 1670.
64. Faller, J. W.; Lambert, C.; Mazzieri, M. R. J. Organomet.
Chem. 1990, 383, 161.
73. Makin, S. M.; Raifel'd, Y. E.; Limanova, O. V.; Shavrygina,
O. A.; Kosheleva, L. M. J. Org. Chem. USSR 1979, 1665.
74. Trost, B. M.; Dumas, J.; Villa, M. J. Am. Chem. Soc. 1992,
114, 9836.
65. Seebach, D.; Wasmuth, D. Helv. Chim. Acta 1980, 63, 197.
66. Aebi, J. D.; Sutter, M. A.; Wasmuth, D.; Seebach, D. Annalen
1983, 2115.
75. Gemal, A. L.; Luche, J.-L. J. Am. Chem. Soc. 1981, 103, 5454.

M. A. Tius / Tetrahedron 58 ,2002) 4343±4367
4367
Biographical sketch
Marc Tius was born in 1953 in Izmir, on the Aegean coast of Turkey. He
moved to Greece when he was 5 years old. He attended Elementary school
for six years in Kavala, a town in Eastern Macedonia, and Gymnasium for
another six years in Thessaloniki, a larger city in northern Greece. In 1971
he enrolled as an undergraduate at Dartmouth College in Hanover, New
Hampshire, where he majored in Mathematics and Chemistry. His ®rst
research experience was in Professor Gordon Gribble's labs. In 1975 he
moved south from Hanover to Cambridge and started graduate studies at
Harvard, where he joined Professor E. J. Corey's group. For his thesis he
completed the synthesis of aphidicolin, working with Larry Blaszczak ®rst,
and then with Jagabandhu Das. After a brief postdoc in the Corey group, he
moved to Hawaii in August 1980, where he has been ever since. He
currently has a joint appointment in the Chemistry Department of the
University of Hawaii, and at the Cancer Research Center of Hawaii.