Molecular editing and assessment of the cytotoxic properties of iejimalide and progeny

Article abstract of DOI:10.1002/chem.201100180

Systematic variation of all substructures embedded into the framework of iejimalide B (2) led to a panel of synthetic analogues of this polyunsaturated macrolide, featuring structural modifications that are not accessible by derivatization of the natural lead compound itself. The assessment of the cytotoxicity of these compounds with the aid of a monolayer proliferation assay (12 human tumor cell lines in vitro) as well as a colony formation assay (24 human tumor xenografts ex vivo) revealed the exceptional potency of 2 and several of its synthetic congeners, with IC₅₀ values in the picomolar range for the most sensitive cell lines. Whereas structural modifications of the macrocycle or of the side chain generally lead to a decrease in activity, changes of the peptidic terminus are not only well accommodated but engender increased tumor selectivity as well. 2 and two of the most promising analogues were selected for in vivo studies in mice, in which their activity against human tumor xenografts of breast cancer (MAXF 401) and prostate cancer (PRXF PC-3M) was tested. In contrast to a previous report in the literature however, which had claimed in vivo activity for the parent iejimalides, these tests did not show a significant therapeutic effect against the chosen solid tumors upon intraperitoneal administration. Copyright

Full text of DOI:10.1002/chem.201100180

FULL PAPER
DOI: 10.1002/chem.201100180
Molecular Editing and Assessment of the Cytotoxic Properties of Iejimalide
and Progeny
Julien Gagnepain,[a] Emilie Moulin,[a] Cristina Nevado,[a] Mario Waser,[a] Armin Maier,[b]
Gerhard Kelter,[b] Heinz-Herbert Fiebig,[b] and Alois Fꢀrstner*[a]
Abstract: Systematic variation of all
substructures embedded into the
framework of iejimalide B (2) led to a
panel of synthetic analogues of this
polyunsaturated macrolide, featuring
structural modifications that are not ac-
cessible by derivatization of the natural
lead compound itself. The assessment
of the cytotoxicity of these compounds
with the aid of a monolayer prolifera-
tion assay (12 human tumor cell lines
in vitro) as well as a colony formation
assay (24 human tumor xenografts ex
vivo) revealed the exceptional potency
of 2 and several of its synthetic conge-
ners, with IC50 values in the picomolar
range for the most sensitive cell lines.
Whereas structural modifications of the
macrocycle or of the side chain gener-
ally lead to a decrease in activity,
changes of the peptidic terminus are
not only well accommodated but en-
gender increased tumor selectivity as
well. 2 and two of the most promising
analogues were selected for in vivo
studies in mice, in which their activity
against human tumor xenografts of
breast cancer (MAXF 401) and pros-
tate cancer (PRXF PC-3M) was tested.
In contrast to a previous report in the
literature however, which had claimed
in vivo activity for the parent iejima-
lides, these tests did not show a signifi-
cant therapeutic effect against the
chosen solid tumors upon intraperito-
neal administration.
Keywords: antitumor
catalysis metathesis
natural products · total synthesis
agents
·
·
·
Introduction
During a program on the synthesis and evaluation of novel
anticancer agents, our attention was caught by early reports
from the Kobayashi laboratory on the promising biological
activities of the iejimalides 1–4.[1–3] These polyene macro-
lides of marine origin were described as remarkably cytotox-
ic against a variety of murine and human cancer cell lines in
vitro.[2b,4] Even more significant is the fact that they were re-
ported to show in vivo activity in mice bearing intraperito-
neally inoculated P388 leukemia. Specifically, the life-span
of the animals was significantly increased (T/C=120–150%)
upon administration of these compounds (200 mgkgꢀ1) into
the peritoneal cavity on days 1 and 5 of treatment.[2b]
The very limited supply of material from the natural sour-
ces may explain why these promising initial findings were
not followed up by a more detailed preclinical evaluation;[5]
only a few simple carbamate derivatives were reported in
the literature, which displayed activity similar to that of the
natural leads.[6] The available biochemical information sug-
gests that the iejimalides primarily interfere with osteoclast
vacuolar H+-ATPases (V-ATPase).[7,8] Despite the ubiquity
of these vital proton pumps located in the plasma mem-
brane of all eukaryotic cells, V-ATPases constitute potential
targets for tumor therapy.[9,10] Their overexpression is re-
sponsible for the acidic extracellular microenvironment of
many tumors, which in turn is innately linked to cancer pro-
gression, metastasis and multidrug resistance.
Our group took up the challenge to provide meaningful
amounts of the iejimalides for further testing, focusing on 2
as the most potent member of the series. After gathering in-
telligence during an initial synthesis campaign,[11] the prepa-
ration of the individual building blocks was optimized and
their assembly streamlined to provide gram amounts of this
valuable natural product (Scheme 1).[12–14] Since the underly-
ing blueprint is inherently flexible, we also saw ourselves in
a position to explore the chemical space surrounding this
polyunsaturated lead compound. This endeavor followed
the logic of “diverted total synthesis”,[15] focusing on such
[a] Dr. J. Gagnepain, Dr. E. Moulin, Dr. C. Nevado, Dr. M. Waser,
Prof. A. Fꢀrstner
Max-Planck-Institut fꢀr Kohlenforschung
45470 Mꢀlheim/Ruhr (Germany)
Fax : (+49)208-306-2994
E-mail: fuerstner@kofo.mpg.de
[b] Dr. A. Maier, Dr. G. Kelter, Prof. H.-H. Fiebig
Oncotest GmbH, Am Flughafen 12–14, 79108 Freiburg (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201100180.
Chem. Eur. J. 2011, 17, 6973 – 6984
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6973

target structures that cannot be obtained by chemical ma-
nipulations of the natural product itself (at least not without
undue effort).[16] Outlined below we present the results of a
first round of molecular editing, which resulted in a set of
fully synthetic iejimalide-like compounds.[17,18] Some of them
clearly surpass the natural lead with regard to tumor selec-
tivity while rivaling it in terms of potency, as evident from
the comparative assessments in cell-based as well as in
tumor colony-based assays. In contrast to earlier claims in
the literature, however, our preliminary in vivo data did not
show a noticeable therapeutic effect in mice.
Results and Discussion
Design aspects: As summarized in Scheme 1, iejimalide B
was built-up in a convergent manner from six commercial
starting materials.[13] The optimized assembly process capi-
talizes on previously acquired knowledge about the molecu-
lar intricacy of 2,[11,12] yet remains flexible enough to allow
for “structural point mutations” in every part of the target
structure (Figure 1). Because no information concerning
structure/activity relationships (SAR) was available at the
outset of the project,[19] we were guided by the following
considerations.
Figure 1. Focal points for structural editing of the molecular framework
of iejimalide B (2).
Early on, we learnt that the advanced synthetic intermedi-
ate
5 of the original total synthesis campaign (see
Scheme 2), which carries an N-Boc group instead of the N-
formyl serine residue, retains appreciable bioactivity.[12] As
this preliminary result suggested that the amino acid termi-
nus of 2 might be a permissive site for structural variation,
considerable efforts were devoted to this region. In so
doing, we hoped to improve the activity/selectivity and in-
crease the robustness of the compounds and their synthetic
precursors (N-formyl serine derivatives are prone to racemi-
zation, oxazole formation and hydrolysis). Moreover, this
site may allow one to attach adequate molecular probes as
needed for further biochemical investigations (fluorescence
tags, biotin etc.).
Scheme 1. Optimized route for the gram-scale synthesis of iejimalide B
(2);[13] the numbers in the cycles indicate the step count, the longest
linear route is shown in bold.
The 1,4-dimethyl substituted 1,3-diene “spacer” region
(C24–C28) between the lipophilic macrolactone core and
Scheme 2. Original tactics for the late-stage variation of the N-acyl terminus, cf. ref. [12].
6974
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6973 – 6984

Cytotoxic Properties of Iejimalide and Progeny
FULL PAPER
the highly polar peptidic terminus of 2 was addressed with
the aim of modulating the stiffness of the side chain and, as
a consequence, the binding affinity to the still unknown re-
ceptor. Since the formation of the hindered ester bond at
C23 with the allylic alcohol flanking this region had also
been far from trivial,[11a] it seemed worth editing this struc-
tural motif out on chemical grounds.
to a small initial set of iejimalide analogues differing from
the natural product in their peptidic termini (Scheme 2).[12]
Since the Teoc-derivative 6 itself, however, is somewhat un-
stable even in a freezer (ꢁ48C) and the released amine 7 is
exceptionally sensitive, this route was not deemed satisfacto-
ry in preparative terms and allowed only for the synthesis of
small amounts of the desired products after purification by
HPLC.[12]
These shortcomings were addressed in two different ways.
First, it was found that a phthalimide moiety instead of the
Teoc group entails much higher stability; yet this group can
be selectively cleaved in the presence of the macrolactone
with the aid of methylamine in EtOH[20] under microwave ir-
radiation (Scheme 3). As the preparation of compound 23
requires no more than 11 steps in the longest linear se-
quence and uses only building blocks that are available on
large scale, this route represents the shortest entry into a
fully functional and stable iejimalide scaffold. The inherent
lability of the primary amine 7 derived from 23, however,
remains unaddressed. This compound must be processed
without delay as it decomposes within short periods of time
in solution as well as in the solid state.[12] Whereas its use in
larger-scale applications was therefore not preferred, we
pursued the phthalimide deprotection/amine coupling tactics
ꢀ
The skipped 1,4-diene motif at C2 C6 next to the lactone
is by far the most labile substructure of the iejimalides, since
migration of the double bonds in either direction is favored
by conjugation.[11–13] This bias makes the preparation of the
acid perimeter troublesome and renders the doubly allylic
chiral center at C4 highly racemization prone; both prob-
ꢀ
lems might be solved by formal saturation of either the C2
ꢀ
C3 or the C5 C6 olefin. Moreover, the comparison of the
analogue bearing a saturated ester with the bioactivity of 2
should also tell if the enoate unit, which constitutes a poten-
tial Michael acceptor for biological nucleophiles, is an essen-
tial part of the pharmacophore.
The NMR data of the iejimalides show that the methyl
ether residing at C17 is directed toward the inside of the
macrocycle.[2,12,18] As this particular conformation may pre-
vent the ether oxygen from serving as a hydrogen-bond ac-
ceptor upon binding of 2 to the receptor protein, we were
interested to see whether an an-
alogue devoid of the C17 MeO
substituent remains biologically
competent.
Molecular dynamics calcula-
tions had suggested that the
ꢀ
C18 C21 diene stands essential-
ly perpendicular to the plane of
the
24-membered
ring.[18]
Therefore it was planned to
preserve this conspicuous struc-
tural element throughout this
campaign. Rather, we intended
to probe if the rigidity imposed
onto the macrocycle by the
polyunsaturated
“southern”
sector is mandatory for high po-
tency by preparing an analogue
ꢀ
devoid of the C11 C14 diene.
Variation of the N-formyl
serine terminus: Our first at-
tempts to perform late-stage
variations of the peptide termi-
nus of iejimalide had failed be-
cause compound 5 decomposed
upon attempted removal of the
N-Boc group even under buf-
fered conditions (TMSOTf, 2,6-
Scheme 3. Short route to an iejimalide-type macrocycle and late stage modification by phthalimide cleavage/
amide formation. a) LiBHEt3, THF, ꢀ788C ! RT, quant.;[21a] b) MnO2, CH2Cl2, 75%;[21b] c) Pd
ACHTUNGTRENNUNG(OAc)2
(5 mol%), PPh3 (5 mol%), Et2Zn, THF, ꢀ788C ! ꢀ208C, 74%; d) TBAF, THF, 08C, 77%; e) (i) Bu3Sn-
SnBu3, BuLi, THF, ꢀ788C; (ii) CuCN, ꢀ788C; (iii) 15, THF, 258C, 72% (E/Z ꢂ16:1); f) I2, Et2O, 0 8C ! RT,
lutidine,
CH2Cl2,
08C).[11]
90%; g) [PdACTHUNTRGENNUG(PPh3)4] (4 mol%), CuTC, [Ph2PO2]AHCUTNGTREN[NUGN NBu4], DMF, 75%; h) EDC·HCl, 4-pyrrolidinylpyridine
(10 mol%), CH2Cl2, 08C ! RT, 75 %; i) 22 (10 mol%); toluene, 508C, 77%; j) MeNH2, EtOH, microwave,
608C; k) 25, EDC·HCl, HOBt, DMF, 70% (over two steps); EDC=1-ethyl-3-(3-dimethylaminopropyl)carbo-
diimide; CuTC=copper thiophene-2-carboxylate; HOBt=1-hydroxy-1H-benzotriazole, TBAF=tetra-n-butyl-
Cleavage of the corresponding
Teoc group with TBAF was
more successful, paving the way ammonium fluoride.
Chem. Eur. J. 2011, 17, 6973 – 6984
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6975

A. Fꢀrstner et al.
Table 1. Yields [%] obtained in the route to the iejimalide analogues de-
picted in Scheme 4.
for the synthesis of the biotin labeled product 24 (11 mg in
the single largest batch).
R1C(O)NHCH(R2)COOH
Steps a,b[a]
Step c
Step d
Step e
A much more scalable route to iejimalide analogues bear-
ing different peptidic residues started with the readily avail-
able building block 26 (Scheme 4), for which a multigram
synthesis had already been worked out.[13] Treatment with
Et4NF in MeCN[22] furnished the corresponding amine,
which was reacted with the desired amino acid derivative
under standard peptide coupling conditions. The subsequent
Stille reaction with stannane 18 under the neutral conditions
developed in this laboratory,[23,24] esterification of the result-
ing products with the acid sector 20,[13] followed by ring-clos-
ing metathesis (RCM)[25] were carried out according to the
protocols developed for the large scale synthesis of iejima-
lide B itself.[13] Although not fully optimized for all individu-
al compounds, this sequence was invariably high yielding
N-formyl glycine
N-formyl l-alanine
N-formyl l-phenylalanine
N-formyl l-methionine
N-formyl l-valine
98
96
86
92
99
98
61
88
73
71
73
74
61
71
75
82
70
79
75
74
84
46[b]
74
N-benzoyl O-TBS-l-serine
86
[a] Over both steps. [b] With 1 equiv of 22; the product was isolated as
the corresponding sulfoxide, see main text.
apolar macrocycle and the hydrophilic N-formyl serine resi-
due in 2 might allow the affinity of the compounds to the
biological receptor to be optimized. In our initial foray, only
a subtle change was envisaged by preparing the C27-des-
methyl analogue 40. Its preparation was straightforward and
strictly followed the optimized procedures developed for the
synthesis of iejimalide B itself (Scheme 5).[13]
(see Table 1), furnishing the targeted analogues 8,[12] [12] and
9
29a–e in good overall yields; several hundred milligrams of
products were prepared by this route. Only the RCM of the
methionine derivative 28d was difficult, in line with the
notion that divalent sulfur may interfere with the ruthenium
catalyst.[26] Increasing the loading of the second-generation
Grubbs complex 22,[27] however, provided a viable solution.
Interestingly, however, the resulting product underwent
rapid air oxidation and was obtained in form of the corre-
sponding sulfoxide 29d.
Modifications of the side chain: As outlined above, structur-
al modifications of the rigid 1,3-diene spacer between the
Scheme 5. Preparation of the C27-desmethyl analogue 40. a) Pd
ACHTUNGTRENNUNG(OAc)2
(2.5 mol%), P(o-tolyl)3 (5 mol%), Et3N, 1008C, 87%; b) trifluoroacetic
AHCTUNGTRENNUNG
acid, CH2Cl2; c) 4-nitrophenyl-2-trimethylsilylethyl carbonate, Et3N,
CH2Cl2, 76% (over two steps); d) Dibal-H, CH2Cl2, ꢀ788C, 92%; e)
MnO2, CH2Cl2; f) 13, PdACHTUNRGTNEUNG(OAc)2 (5 mol%), PPh3 (5 mol%), Et2Zn, THF,
Scheme 4. Scalable route to iejimalide analogues with different peptidic
ꢀ788C ! ꢀ208C, 50% (over two steps); g) TBAF, THF, 08C, 82%; h)
(Bu3Sn)2, BuLi, THF, CuCN, ꢀ788C ! RT, then 34, 92%; i) I2, CH2Cl2,
08C ! RT, 95 %; j) Et4NF, MeCN, 408C; k) O-TBS-N-formyl-l-serine,
EDC·HCl, HOBt, N-methylmorpholine, CH2Cl2, 90% (over two steps);
termini. a) Et4NF, MeCN, 408C; b) R1C(O)NHCH(R2)COOH,
EDC·HCl, HOBt, N-methylmorpholine, CH2Cl2; c) 18, [Pd
ACHTUNGTRNE(NUNG PPh3)4]
(4 mol%), CuTC, [Ph2PO2][NBu4], DMF; d) EDC·HCl, 4-pyrrolidinyl-
AHCTUNGTRENNUNG
pyridine (10 mol%), CH2Cl2, 08C ! RT, then 20; e) 22 (10 mol%); tolu-
ene, 508C; f) TBAF, THF, 08C, 99%. For the yields, see Table 1; Teoc=
2-(trimethylsilyl)ethoxycarbonyl.
l) 18, [PdACTHNGUTERNNU(G PPh3)4] (4 mol%), [Ph2PO2]ACHTUNGTRENN[NUG NBu4], CuTC, DMF, 92%, m) 20,
EDC·HCl, 4-pyrrolidinylpyridine (10 mol%), CH2Cl2, 08C ! RT, 83 %;
n) 22 (10 mol%), toluene, 508C, 63%; o) TBAF, THF, 08C, 97%.
6976
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6973 – 6984

Cytotoxic Properties of Iejimalide and Progeny
FULL PAPER
A much more profound molecular surgery was realized in
compound 51, which contains a fully saturated spacer
(Scheme 6). This product derived from the commercial sub-
strate 41 by Teoc protection, Dibal-H reduction of the ester
in 42 and reaction of the resulting aldehyde 43 with proparg-
yl mesylate 44 under the conditions described by Mar-
shall.[28] The further elaboration of product 45 (d.r. >10:1)
thus formed required temporary protection of the secondary
alcohol, hydrozirconation/iodination[29] of the triple bond in
46, cleavage of the pivaloate ester, and exchange of the
Teoc residue for the protected N-formyl serine group. Since
the resulting product 48 had been prepared before the Stille
cross-coupling tactics was fully optimized for the large scale
synthesis of 2,[13] the assembly process followed the Suzuki
route originally employed in this laboratory using the
known boronate 49 as the coupling partner.[12,30] All steps
worked satisfactorily, providing product 51 in a respectable
overall yield.
ally observed in the Marshall reaction of aliphatic alde-
hydes,[28] however, the reaction of the benzaldehyde deriva-
tive 54 with mesylate 13 under standard conditions gave
syn- and anti-55 in a 1:1 ratio, which were separated by flash
chromatography. The elaboration of anti-55 to the desired
iejimalide analogue 59 followed the established tactics and
proceeded in good overall yield.
In a complementary setting, the 1,3-diene spacer of the
side chain of 2 was replaced by a phenyl ring as shown in
Scheme 7. In contrast to the high diastereoselectivity gener-
Scheme 7. Iejimalide analogue containing an aromatic rather than
a
diene spacer. a) 4-Nitrophenyl-2-trimethylsilylethyl carbonate, Et3N,
CH2Cl2, 86%; b) (i) MeLi, THF, ꢀ788C; (ii) tBuLi, ꢀ788C; (iii) DMF,
ꢀ788C ! RT, 95 %; c) 13, Pd
ACHTNUGRTEN(NNUG OAc)2 (5 mol%), PPh3 (5 mol%), Et2Zn,
THF, ꢀ78 ! ꢀ208C, 80%; d) TBAF, THF, 08C, 82%; e) pivaloyl chlo-
ride, pyridine, DMAP cat., 62%; f) Cp2Zr(H)Cl, THF, then I2, 67%; g)
LiHBEt3, THF, 08C, 81%; h) TBAF, THF; i) O-TBS-N-formyl-l-serine,
EDC·HCl, HOBt, N-methylmorpholine, CH2Cl2, 54% (over two steps);
j) 49, [(dppf)PdCl2], (15 mol%), Ba(OH)2·8H2O, DMF, 95%; k) 20,
DCC, 4-pyrrolidinylpyridine, CH2Cl2, 08C ! RT, 52 %; l) 22 (20 mol%),
CH2Cl2, RT, 70%; m) TBAF, THF, 08C, 80%.
Conceptually more interesting are the modifications pres-
ent in compound 61 (Scheme 8). As outlined in more detail
in a previous paper,[18] we recognized a striking structural
and functional relationship between iejimalide B (2) and
archazolid A (60), a macrolactone isolated from a terrestrial
myxobacterium.[31] Both natural products incorporate seven
double bonds into their 24-membered rings. The lactone
connects a substituted enoate to a secondary alcohol at C23,
which in turn is flanked by a methyl group in a 1,2-anti rela-
tionship; these chiral centers are S,S-configured in 2 and 60
Scheme 6. Preparation of an iejimalide analogue with a flexible spacer. a)
4-Nitrophenyl-2-trimethylsilylethyl carbonate, Et3N, CH2Cl2, 96%; b)
Dibal-H, CH2Cl2, ꢀ788C, 72%; c) 44, Pd
ACHTUNGTREN(UNGN OAc)2 (5 mol%), PPh3
(5 mol%), Et2Zn, THF, ꢀ788C ! ꢀ208C, 55%; d) pivaloyl chloride, pyr-
idine, DMAP cat., 84%; e) Cp2Zr(H)Cl, THF, then I2, 69%; f) LiHBEt3,
THF, 08C, 90%; g) TBAF, THF; h) O-TBS-N-formyl-l-serine,
EDC·HCl, HOBt, N-methylmorpholine, CH2Cl2, 74% (over two steps);
i) 49, (dppf)PdCl2, (15 mol%), Ba(OH)2·8H2O, DMF, 68%; j) 20, DCC,
4-pyrrolidinylpyridine, CH2Cl2, 08C
CH2Cl2, RT, 55%; l) TBAF, THF, 08C, 66%; DCC=dicyclohexylcarbo-
diimide; DMAP=4-dimethylaminopyridine.
!
RT, 74 %; k) 22 (10 mol%),
ꢀ
alike. Moreover, a C18 C21 diene with an adjacent methyl
Chem. Eur. J. 2011, 17, 6973 – 6984
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6977

A. Fꢀrstner et al.
ether substituent is displayed in both series and even the 9S-
OMe group with its neighboring E alkene of the iejimalides
seems to correspond closely to the 7S-OH group of archazo-
lid A located in the exact same microenvironment. Al-
though the side chains per se differ in structural detail, they
consist of a polar head group (N-formyl serine versus N-
methyl carbamate) and a rigid hydrophobic spacer branch-
ing off the macrocycle at C23 in both families. Like iejima-
lide, archazolid A is a selective V-ATPase inhibitor and
potent cyctotoxic agent.[31] Inspired by these structural and
functional analogies, we targeted compound 61, which hy-
bridizes the core of iejimalide B with the characteristic tail
of the archazolid. Its preparation was described in detail in
a previous publication.[18]
Scheme 9. Iejimalide analogue devoid of the Michael acceptor unit and
the 1,4-diene motif next to the lactone linkage. a) tBuPh2SiLi, THF,
CuCN, 08C; b) ICl, CH2Cl2, 08C, 90% (over both steps); c) [Pd
ACHTUNGTRENNUNG(PPh3)4]
(0.4 mol%), [Ph2PO2][NBu4], CuTC, DMF, 66%; d) aq. LiOH, THF/
AHCTUNGTRENNUNG
MeOH, 75%; e) EDC·HCl, 4-pyrrolidinylpyridine (10 mol%), CH2Cl2,
08C ! RT, then 66, 54%; f) 22 (10 mol%), toluene, 508C, 85%; g)
TBAF, THF, 08C, 93%.
Scheme 8. Comparison of iejimalide B and archazolid A; synthetic chi-
mera 5, which combines the iejimalide core and the archazolid side
chain.
Similarly straightforward was the preparation of analogue
ꢀ
74, in which the C5 C6 bond of the skipped diene present
in the iejimalides is formally removed (Scheme 10). Starting
from the known aldehyde 69, a sequence of Masamune–
Roush olefination,[33] oxidation of the allylic alcohol 70,
Brown allylation[34] and final O-methylation with Meerwein
salt gave the required fragment 71. The derived acid 72 was
attached to alcohol 66[13] with the aid of DCC to furnish the
polyunsaturated ester 73, which was cyclized as usual on ex-
posure to catalytic amounts of the Grubbs second genera-
tion catalyst 22.[27] Deprotection of the TBS-ether protecting
the serine OH group delivered analogue 74.
Modifications of the acid perimeter: Deep-seated structural
modifications were planned for the region flanking the lac-
tone linkage. The skipped 1,4-diene of 2 is a highly fragile
and potentially reactive site, and the chiral center at C4 is
labile in all synthetic precursors prior to ring closure. As a
consequence we planned to excise the methyl branches at
C2 as well as C4 and formally replace the enoate by a satu-
rated ester group. This structural variant should also reveal
whether or not the iejimalides serve as Michael acceptors
for biological nucleophiles.
To this end, alkyne 62 was reacted with the “lower-order”
silyl cuprate derived from PhMe2SiLi and CuCN,[32] and the
resulting alkenylsilane was subjected to iodine-for-silicon ex-
change (Scheme 9). Iodide 63 thus formed was cross-cou-
pled with stannane 64 under the conditions previously opti-
mized in our laboratory.[23] Saponification of the methyl
ester followed by attachment of the liberated acid 65 to the
southern sector 66[13] furnished product 67, which underwent
a highly productive RCM reaction to close the macrocyclic
scaffold. Treatment of the resulting product with TBAF
then gave the targeted analogue 68 in high yield.
Modifications of the southern sector: Following the ration-
ale outlined above, it seemed interesting to see if the biolog-
ical activity of the iejimalide analogue 81 devoid of the C17
MeO substituent remains uncompromised. 4-Pentyn-1-ol
(75) served as the point of departure, which was oxidized
with pyridine·SO3 complex[35] and the resulting aldehyde was
subjected to a Z-selective Still–Gennari olefination mediat-
ed by KHMDS in the presence of sub-stoichiometric
amounts of [18]crown-6 (Scheme 11).[36,37] Reduction of the
ester group in 76 with Dibal-H followed by hydroboration
of the terminal alkyne in the resulting product with excess
6978
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6973 – 6984

Cytotoxic Properties of Iejimalide and Progeny
FULL PAPER
Scheme 10. Alternative design to remove the skipped diene characteristic
for the acid perimeter of iejimalide. a) (EtO)2P(O)CH(Me)COOEt,
LiCl, iPr2NEt, MeCN, 85%; b) MnO2, CH2Cl2, 92%; c) (ꢀ)-(Ipc)2BOMe,
allylmagnesium bromide, Et2O, ꢀ788C; d) Me3OBF4, proton sponge,
CH2Cl2, 27% (over two steps, unoptimized); e) aq. LiOH, THF/MeOH,
97%; f) DCC, 4-pyrrolidinylpyridine, CH2Cl2, then 66, 89%; g) 22
(20 mol%), CH2Cl2, 66%; h) TBAF, THF, 08C, 65%; Ipc=isopinocam-
pheyl.
Scheme 11. 17-Desmethoxy analogue of iejimalide B. a) SO3·Pyridine,
Et3N, DMSO, CH2Cl2; b) (F3CCH2O)2P(O)CH(Me)COOMe, KHMDS,
[18]crown-6, toluene, ꢀ788C, 66% (over two steps); c) Dibal-H, CH2Cl2,
ꢀ788C; d) pinacol borane, 9-BBN (10 mol%), THF, 79%; e) Dess–
Martin periodinane, CH2Cl2, 78%; f) Ph3PCH3Br, BuLi, THF, ꢀ788C !
RT, 60 %; g) 27 (R1 =H, R2 =CH2OTBS),[13] (dppf)PdCl2, (15 mol%),
Ba(OH)2·8H2O, DMF, 59%; h) 20, EDC·HCl, 4-pyrrolidinylpyridine
(10 mol%), CH2Cl2, 08C ! RT, then 79, 86%; i) 22 (15 mol%), CH2Cl2,
75%; g) TBAF, THF, 08C, 95%.
pinacol borane (and catalytic amounts of 9-H-9-BBN dimer
to entrain the reaction)[38] delivered compound 77 in good
yield. This product could be oxidized with Dess–Martin per-
iodinane[39] to the corresponding aldehyde without interven-
ꢀ
tion of the oxidation-prone C B bond. A subsequent Wittig
reaction provided product 78, which was cross-coupled with
alkenyl iodide 27 (R1 =H, R2 =CH2OTBS)[13] by a Suzuki
reaction promoted by Ba(OH)2·8 H2O as the optimal
base.[12] Attachment of the common acid sector 20[13] to the
resulting compound 79, ring closing metathesis of polyene
80, and final cleavage of the TBS-ether on the serine moiety
gave the desired product 81 in respectable overall yield.
A final structural variant addressed the C11-C14 diene
motif of the natural leads (Scheme 12). To impart gradually
higher conformational flexibility on the macrocycle, one of
the double bond in this part of the molecule was replaced
by a -CH2CH2- chain. To this end, a Noyori-transfer hydro-
genation[40] of ketone 82[41] followed by O-methylation gave
product 83 in excellent optically purity (98% ee). If proto-
desilylated at this stage, the resulting ene-yne could not be
engaged in a selective hydroboration at the alkyne site;
rather, a complex mixture was obtained under a variety of
conditions (cf. Experimental Section). Therefore, we em-
barked on a short detour as shown in Scheme 12. Specifical-
ly, the double bond of 83 was oxidatively cleaved and the re-
suting crude aldehyde reduced with NaBH4; subsequent de-
silylation gave alcohol 84, which was subjected to hydrobo-
ration with pinacol borane in the presence of 9-H-9-BBN
rectly be subjected to a Wittig reaction to re-install the ter-
minal alkene required for the projected ring closure by
RCM.
The Suzuki cross-coupling of boronate 86 thus formed
with alkenyl iodide 27 (R1 =H, R2 =CH2OTBS)[13] followed
by esterification of the resulting product 87 with acid 20
gave polyene 88. When subjected to catalytic amounts of
the ruthenium complex 22[27] under high dilution conditions,
an inseparable 2:1 mixture of E/Z-89 was obtained, which
was deprotected as usual with the aid of TBAF. The ob-
served low stereoselectivity in this particular case is in
accord with the results reported for many other RCM-based
macrocyclizations in the literature[25,42–44] but stands in
marked contrast to all other metathesis reactions performed
during this investigation with substrates incorporating a tri-
substituted C13-C14 alkene unit;[12,13,18] as outlined above,
all RCM reactions of these substrates were distinguished by
virtually complete E selectivity. This striking preference
must hence be ascribed to the directing influence of the
methyl branch next to the site of ring closure that is missing
in the case of substrate 88, as well as to the limited confor-
mational flexibility imposed on the resulting polyunsatura-
ted ring, which does not adopt an additional Z-configured
olefin if all seven endocyclic double bonds are present.
(10 mol%).[38] As the oxidation of the OH group left the C
ꢀ
B bond of 85 untouched, the resulting aldehyde could di-
Chem. Eur. J. 2011, 17, 6973 – 6984
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6979

A. Fꢀrstner et al.
Table 2. Antitumor activity of iejimalide B (2) as determined with a
panel of 12 human tumor cell lines.
Histotype
Cell line
IC50 [nm]
IC70 [nm]
colon
stomach
lung, non small cell
HT-29
0.014
0.77
0.54
0.36
0.66
>144
2.89
117
0.11
85
1.0
0.94
14
>144
10
>144
>144
1.4
10
>144
GXF 251L
LXFL 529L
LXFA 629L
MAXF 401NL
MEXF 462NL
OVXF 899L
PAXF 1657L
22Rv1
breast
melanoma
ovary
pancreas
prostate
mesothelioma
kidney
7.2
0.71
5.8
PXF 1752L
RXF 486L
UXF 1138L
uterus
5.8
IC50) and the total number of cell lines investigated for a
given compound (Table 3).
These in vitro data reveal the truly impressive cytotoxicity
of iejimalide B (2) and several of its congeners, reaching
IC50 values well below 1 nm for several cell lines; the highest
Table 3. Antitumor potency and selectivity of iejimalides and analogues
as determined in a panel of 12 human tumor cell lines.
Compound
Mean IC50
[nm]
Mean IC70
[nm]
selective/
total[a]
Selectivity
[%]
iejimalide B
2.3
12
6/12
50
(2)
iejimalide A 14
46
4/12
33
(1)
29a
29b
29d
9
29c
40
68
6
51
81
23
59
61
90
2.0
2.0
6.0
7.0
13
16
5.0
46
5.0
6.0
16
22
31
41
61
322
733
730
846
2902
5161
11902
37
3/12
2/12
5/12
6/12
3/12
5/12
4/12
3/12
4/12
4/12
3/12
5/12
3/12
3/12
4/12
2/12
25
17
42
50
25
42
33
25
33
33
25
42
25
25
33
17
Scheme 12. Iejimalide analogue devoid of the trisubstituted alkene at
C13-C14. a) 91 (6 mol%), iPrOH, 97% (98% ee); b) BuLi, THF, then
MeI, DMSO, ꢀ788C ! RT, quant.; c) OsO4 (10 mol%), NaIO4, 2,6-luti-
dine, 1,4-dioxane/water; d) NaBH4, MeOH, 70% (over two steps); e)
TBAF, THF, 08C, 85%; f) pinacol borane, 9-H-9-BBN (10 mol%), THF;
g) Dess–Martin periodinane, CH2Cl2, 61% (over two steps); h)
Ph3PCH3Br, BuLi, THF, ꢀ788C, 67%; i) 27 (R1 =H, R2 =CH2OTBS),[13]
[(dppf)PdCl2], (15 mol%), Ba(OH)2·8H2O, DMF, 65%; j) 20, EDC·HCl,
4-pyrrolidinylpyridine (10 mol%), CH2Cl2, 08C ! RT, then 87, 58%; k)
22 (15 mol%), CH2Cl2, 64% (E/Z = 2:1); g) TBAF, THF, 08C, 73%.
87
118
272
608
2017
5544
17
doxorubicin
5-fluorouracil 707
1386
Biological activity of iejimalides and analogues: In a first
round of our investigations, the ability of iejimalide A and B
and the assortment of fully synthetic congeners to inhibit
the proliferation of 12 human tumor cell lines was deter-
mined using a propidium iodide-based monolayer prolifera-
tion assay. The assay measures the number of viable cells
after four days of incubation and allows the IC50 and IC70
values to be calculated by non-linear regression.[45] The
known antitumor agents doxorubicin and 5-fluorouracil
were used as reference compounds.
As a representative example, the experimental data ob-
tained for 2 are shown in Table 2. From such data sets, geo-
metric mean IC values were calculated and the compounds
were ranked according to their overall potency (Table 3).
Anti-tumor selectivity was assessed by forming the quotient
of the number of sensitive cell lines (IC50 <1/2 of the mean
[a] Number of cell lines, the individual IC50 value of which is <1= of the
mean IC50 for the total panel of 12 tested cell lines.
2
activity was found against the colon cancer cell line HT-29,
with an IC50 value of only 14pm! 2 is about six-fold more
potent than its naturally occurring relative iejimalide A (1)
in this assay, which proves that the methyl branch next to
the lactone linkage is critical. Importantly, several synthetic
analogues retain IC50 and even IC70 values in the single digit
nanomolar range and exhibit remarkable tumor selectivity
as well. Variations of the peptidic terminus are well accom-
modated, with the glycine and alanine containing siblings
29a and 29b being highly potent. Only the phthalimide de-
rivative 23 is markedly less active, even though its mean
6980
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6973 – 6984

Cytotoxic Properties of Iejimalide and Progeny
FULL PAPER
IC50 and IC70 are still below 1 mm. Formal exchange of the
N-formyl group on the serine terminus for an N-benzoyl
substituent, as materialized in compound 9, is also possible
without compromising the activity.
saturated ester unit for the pharmacophore of 2. In accord-
ance with the cell-based screen, the phthalimide derivative
23 is up to three orders of magnitude less potent than 2 or
the best synthetic analogues, although its tumor selectivity is
remarkable.
Interestingly, product 68 with a saturated rather than an
a,b-unsaturated ester linkage retains appreciable potency;
however, this compound was not amongst the best ana-
logues in the colony formation assay described below. All
structural variations of the spacer between the macrolide
core and the peptidic terminus entailed a significant loss of
activity. Even the deletion of only one methyl group on this
substructure causes a drop in cytotoxicity of one order of
magnitude (compare 2 with 40), although an IC50 of 16 nm
recorded for 40 is still noteworthy. The compounds with a
fully saturated side chain (51) or with a phenyl group in the
linker (59) are less potent. Likewise, the iejimalide/achazolid
chimera 61 performed rather poorly (mean IC50 ꢁ 2 mm).
Although NMR data and computational studies showed
that the 17-OMe substituent of 2 points toward the interior
of the macrocyclic ring, its deletion is detrimental as evident
from the results obtained with compound 81. The least
active of all synthetic congeners in this in vitro assay was
Importantly, the colony formation-based assessment exer-
cise showed that deletion of the hydroxymethyl group of the
serine moiety of 2 or exchange for more lipophilic substitu-
ents, as materialized in compounds 8 and 29a–d, increased
tumor selectivity to a noticeable extent (cf. the number of
sensitive tumors color-coded in green in Figure 2). Replace-
ment of the N-formyl group in 2 by an N-benzoyl substitu-
ent, as displayed by compound 9, engenders the same im-
provement. Likewise, the C27-desmethyl derivative 40 was
found significantly more selective than the parent compound
2. In this particular case, the individual IC50 values for as
many as 14 of the 24 tested tumors were less than half the
average IC50 for the entire panel.
Based on these results, iejimalide B (2), the N-formyl ala-
nine analogue 29b as well as the N-benzoyl serine derivative
9 were selected for a first round of in vivo testing in mice,
carrying subcutaneously implanted tumor xenografts. These
particular compounds were chosen as the arguably best
compromise between activity, selectivity and accessibility.
They were tested for therapeutic effects against xenografts
of breast cancer (MAXF 401) and prostate cancer (PRXF
PC-3M) on the basis of the high sensitivity which these
tumors had shown in the cell-based as well as the colony-
based screenings. Dose-finding experiments led to no or
only little body weight loss of healthy animals at exposures
to doses of 0.6 mgkgꢀ1 dꢀ1 and 2 mgkgꢀ1 dꢀ1, although com-
pound 29b was found toxic at a dose of 2.4 mgkgꢀ1 dꢀ1.
Given the fact that the Kobayashi group had reported in
vivo activity of the iejimalides against the P388 mouse leu-
kemia cell line at doses of 200 mgkgꢀ1 dꢀ1 (T/C 150% for 4,
120% for 2),[2b] it came as a surprise that we found little if
any therapeutic effects, even though up to six-fold higher
doses were given. The reasons for these strikingly different
outcomes remain to be elucidated. However, it should be
pointed out that the antitumor effects observed by the Ko-
bayashi group were obtained in an in vivo system, where
P388 mouse leukemia cells were inoculated intraperitoneally
and the iejimalides were then also also administered intra-
peritoneally starting on day one after inoculation. Thus, the
previously observed in vivo antitumor effect (increase in life
span) could be due to a direct interaction of the tumor and
the administered compounds in the peritoneal cavity of the
animal; this set-up may actually correspond more closely to
the monolayer proliferation and colony formation assays de-
scribed above, which are also based on direct contacting. By
contrast, the present in vivo studies rely on xenografts being
subcutaneously transplanted in the flanks of immunodefi-
cient nude mice, followed by intraperitoneal administration
of the compounds to result in systemic exposure. This ap-
proach mimics the clinical situation to a better extent and
takes relevant effects like bioavailability, distribution, and
metabolism of the administered compounds into account. In
ꢀ
product 90, in which the trisubstituted alkene group at C13
C14 of the natural lead compound was replaced by a
-CH2CH2- chain. These results advocate the notion that the
integrity and rigidity of the macrocyclic edifice is indispensi-
ble for high cytotoxicity.
The inhibition of clonogenicity of tumor cells by the 10 so
far most active and/or selective compounds was further eval-
uated using a colony formation assay. In this assay the abili-
ty of the compounds to inhibit anchorage-independent
growth and ex vivo colony formation of xenograft-derived
tumor cells in semi-solid medium was investigated. Because
of the predictive value of the assay, candidate tumors can be
selected for subsequent in vivo tests.[45] To this end, tumor
cell suspensions from 24 human tumor xenografts of 13 dif-
ferent histotypes were prepared from freshly explanted solid
tumor xenografts grown in serial passage on immune defi-
cient nude mice. Cells were plated out in soft agar and incu-
bated with different concentrations of the respective iejima-
lide derivative. Inhibition of colony formation was taken as
the measure for antitumor potency.
Figure 2 summarizes the results in form of a selectivity
heatmap. Overall, the colony formation assay confirmed the
significant potency of iejimalide B and congeners. All ana-
logues differing from 2 in the amino acid residue retain a re-
markable level of cytotoxicity; in particular, the N-formyl-
glycine and the N-formylalanine containing derivatives 29a
and 29b rival or even surpass the natural product in terms
of activity, with IC50 values as low as 1nm for several sensi-
tive tumors. Analogue 40 lacking the C27-methyl substituent
on the diene-spacer between the core and the peptide resi-
due is only a factor of 2–3 less active (rather than a factor of
10 in the cell-based assay) than the most potent compounds,
whereas the activity of 68 comprising the saturated ester
linkage was rather low. As a result, no simple answer can be
given at this stage as to the importance of an intact a,b-un-
Chem. Eur. J. 2011, 17, 6973 – 6984
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6981

A. Fꢀrstner et al.
Figure 2. Antitumor potency and tumor selectivity for iejimalide B and analogues as determined by a colony formation assay using 24 human tumor xeno-
grafts. IC50 values are shown in nm concentrations, the tumor selectivity is color coded as shown in the insert.
any case, ongoing work in our laboratories has to address
possible metabolic and/or hydrolytic instabilities of the natu-
rally occurring iejimalides and the synthetic analogues pre-
pared so far by adequate structural modifications.
Acknowledgements
Generous financial support by the MPG, the Chemical Genomics Center
(CGC Initiative of the MPG), the Fonds der Chemischen Industrie, the
Alexander von Humboldt Foundation (fellowships to E.M. and C.N.), the
Association pour la Recherche sur le Cancer (fellowship to E.M.), and
the Austrian Fonds zur Fçrderung der Wissenschaftlichen Forschung (fel-
lowship to M.W.) is gratefully acknowledged. We sincerely thank the
NMR, MS and chromatography departments of our Institute for their ex-
cellent support during the entire project.
Conclusion
Whereas marine organisms range amongst the most promis-
ing sources of novel cytotoxic metabolites, they hardly ever
allow for the isolation of sufficient quantities of material for
in-depth biological and preclinical evaluations. A customary
role of chemical synthesis is to fill this gap; the iejimalide
case reported in this and the accompanying paper corrobo-
rates this notion. At the same time, it shows that contempo-
rary total synthesis is well positioned to provide analogues
with deep seated structural mutations and improved proper-
ties, which traditional derivatization of a given lead could
not provide. Yet, this investigation also reminds us that it
may take more than one round of molecular editing before
even an exceptionally potent cytotoxic agent may become a
clinically viable therapeutic candidate.
[1] a) J. Kobayashi, J. Cheng, T. Ohta, H. Nakamura, S. Nozoe, Y.
Hirata, Y. Ohizumi, T. Sasaki, J. Org. Chem. 1988, 53, 6147–6150;
b) Y. Kikuchi, M. Ishibashi, T. Sasaki, J. Kobayashi, Tetrahedron
Lett. 1991, 32, 797–798.
[2] a) M. Tsuda, K. Nozawa, K. Shimbo, H. Ishiyama, E. Fukushi, J. Ka-
wabata, J. Kobayashi, Tetrahedron Lett. 2003, 44, 1395–1399; b) K.
Nozawa, M. Tsuda, H. Ishiyama, T. Sasaki, T. Tsuruo, J. Kobayashi,
Bioorg. Med. Chem. 2006, 14, 1063–1067.
[3] A recent paper suggested that these compounds might actually be
produced by symbiotic cyanobacteria rather than the tunicates
themselves, see: T. L. Simmons, R. C. Coates, B. R. Clark, N.
Engene, D. Gonzalez, E. Esquenazi, P. C. Dorrestein, W. H. Ger-
wick, Proc. Natl. Acad. Sci. USA 2008, 105, 4587–4594.
[4] A second set of activity data is available from the NCI homepage
(http://www.dtp.nci.nih.gov/docs/dtp_search.html). (last accessed
April, 2011.
[5] An exception is a preliminary report from the Walther Cancer Re-
search Center, which claimed stunning morphological changes upon
injection of iejimalides into tumor xenografts (http://www.nd.edu/
science/documents/waltherbrochure.pdf). To the best of our knowl-
Experimental Section
The entire experimental section, including compound characterization
data, can be found in the Supporting Information.
6982
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6973 – 6984

Cytotoxic Properties of Iejimalide and Progeny
FULL PAPER
edge, however, these results have not been published in form of an
original research report.
[6] D. Schweitzer, J. Zhu, G. Jarori, J. Tanaka, T. Higa, V. J. Davisson, P.
Helquist, Bioorg. Med. Chem. 2007, 15, 3208–3216.
[23] A. Fꢀrstner, J.-A. Funel, M. Tremblay, L. C. Bouchez, C. Nevado,
M. Waser, J. Ackerstaff, C. C. Stimson, Chem. Commun. 2008,
2873–2875.
[24] For applications, see: a) A. Fꢀrstner, J. Ackerstaff, Chem. Commun.
2008, 2870–2872; b) A. Fꢀrstner, L. C. Bouchez, J.-A. Funel, V. Lie-
pins, F.-H. Porꢂe, R. Gilmour, F. Beaufils, D. Laurich, M. Tamiya,
Angew. Chem. 2007, 119, 9425–9430; Angew. Chem. Int. Ed. 2007,
46, 9265–9270; c) A. Fꢀrstner, L. C. Bouchez, L. Morency, J.-A.
Funel, V. Liepins, F.-H. Porꢂe, R. Gilmour, D. Laurich, F. Beaufils,
M. Tamiya, Chem. Eur. J. 2009, 15, 3983–4010; d) A. Francais, A.
Leyva, G. Etxebarria-Jardi, S. V. Ley, Org. Lett. 2010, 12, 340–343;
e) R. S. Paley, K. E. Berry, J. M. Liu, T. T. Sanan, J. Org. Chem.
2009, 74, 1611–1620; f) I. Paterson, S. B. J. Kan, L. J. Gibson, Org.
Lett. 2010, 12, 3724–3727.
[25] a) T. M. Trnka, R. H. Grubbs, Acc. Chem. Res. 2001, 34, 18–29;
b) A. Fꢀrstner, Angew. Chem. 2000, 112, 3140–3172; Angew. Chem.
Int. Ed. 2000, 39, 3012–3043; c) K. C. Nicolaou, P. G. Bulger, D.
Sarlah, Angew. Chem. 2005, 117, 4564–4601; Angew. Chem. Int. Ed.
2005, 44, 4490–4527; d) A. H. Hoveyda, A. R. Zhugralin, Nature
2007, 450, 243–251; e) A. Gradillas, J. Pꢂrez-Castells, Angew. Chem.
2006, 118, 6232–6247; Angew. Chem. Int. Ed. 2006, 45, 6086–6101;
f) A. Deiters, S. F. Martin, Chem. Rev. 2004, 104, 2199–2238.
[26] a) A. Fꢀrstner, G. Seidel, N. Kindler, Tetrahedron 1999, 55, 8215–
8230; b) A. Fꢀrstner, O. R. Thiel, C. W. Lehmann, Organometallics
2002, 21, 331–335.
[7] a) S. Kazami, M. Muroi, M. Kawatani, T. Kubota, T. Usui, J. Ko-
bayashi, H. Osada, Biosci. Biotechnol. Biochem. 2006, 70, 1364–
1370; b) M. Muroi, S. Kazami, K. Noda, H. Kondo, H. Takayama,
M. Kawatani, T. Usui, H. Osada, Chem. Biol. 2010, 17, 460–470.
[8] a) W.-L. W. Wang, P. McHenry, R. Jeffrey, D. Schweitzer, P. Hel-
quist, M. Tenniswood, J. Cell. Biochem. 2008, 105, 998–1007; b) P.
McHenry, W.-L. W. Wang, E. Devitt, N. Kluesner, V. J. Davisson, E.
McKee, D. Schweitzer, P. Helquist, M. Tenniswood, J. Cell. Biochem.
2010, 109, 634–642.
[9] For V-ATPases as possible anti-cancer targets, see: a) M. Pꢂrez-
Sayꢃns, J. M. Somoza-Martꢄn, F. Barras-Angueira, J. M. Gꢃndara
Rey, A. Garcꢄa-Garcꢄa, Cancer Treat. Rev. 2009, 35, 707–713;
b) E. J. Bowman, B. J. Bowman, J. Bioenerg. Biomembr. 2005, 37,
431–435.
[10] For the total synthesis of other ATPase inhibitors from our group,
see: a) A. Fꢀrstner, T. Dierkes, O. R. Thiel, G. Blanda, Chem. Eur.
J. 2001, 7, 5286–5298; b) A. Fꢀrstner, O. R. Thiel, G. Blanda, Org.
Lett. 2000, 2, 3731–3734; c) A. Fꢀrstner, M. Bindl, L. Jean, Angew.
Chem. 2007, 119, 9435–9438; Angew. Chem. Int. Ed. 2007, 46, 9275–
9278; d) M. Bindl, L. Jean, J. Herrmann, R. Mꢀller, A. Fꢀrstner,
Chem. Eur. J. 2009, 15, 12310–12319.
´
[11] a) A. Fꢀrstner, C. Aꢅssa, C. Chevrier, F. Teply, C. Nevado, M. Trem-
[27] M. Scholl, S. Ding, C. W. Lee, R. H. Grubbs, Org. Lett. 1999, 1, 953–
956.
[28] J. A. Marshall, J. Org. Chem. 2007, 72, 8153–8166.
blay, Angew. Chem. 2006, 118, 5964–5969; Angew. Chem. Int. Ed.
2006, 45, 5832–5837; b) A. Fꢀrstner, C. Nevado, M. Tremblay, C.
´
Chevrier, F. Teply, C. Aꢅssa, M. Waser, Angew. Chem. 2006, 118,
[29] D. W. Hart, J. Schwartz, J. Am. Chem. Soc. 1974, 96, 8115–8116.
[30] For a general overview, see: A. Suzuki, J. Organomet. Chem. 1999,
576, 147–168.
5969–5974; Angew. Chem. Int. Ed. 2006, 45, 5837–5842.
[12] A. Fꢀrstner, C. Nevado, M. Waser, M. Tremblay, C. Chevrier, F.
´
Teply, C. Aꢅssa, E. Moulin, O. Mꢀller, J. Am. Chem. Soc. 2007, 129,
[31] a) F. Sasse, H. Steinmetz, G. Hçfle, H. Reichenbach, J. Antibiot.
2003, 56, 520–525; b) J. Hassfeld, C. Farꢆs, H. Steinmetz, T. Carlo-
magno, D. Menche, Org. Lett. 2006, 8, 4751–4754.
9150–9161.
[13] See the accompanying paper in this issue: J. Gagnepain, E. Moulin,
A. Fꢀrstner, Chem. Eur. J. 2011, 17, 6964–6972 (preceding paper).
[14] For a second total synthesis of 2, see: D. Schweitzer, J. J. Kane, D.
Strand, P. McHenry, M. Tenniswood, P. Helquist, Org. Lett. 2007, 9,
4619–4622.
[15] a) R. M. Wilson, S. J. Danishefsky, J. Org. Chem. 2006, 71, 8329–
8351; b) R. M. Wilson, S. J. Danishefsky, Angew. Chem. 2010, 122,
6168–6193; Angew. Chem. Int. Ed. 2010, 49, 6032–6056; c) A. M.
Szpilman, E. M. Carreira, Angew. Chem. 2010, 122, 9786–9823;
Angew. Chem. Int. Ed. 2010, 49, 9592–9628.
[16] For earlier programs in this laboratory following the logic of “divert-
ed total synthesis”, see: a) A. Fꢀrstner, D. Kirk, M. D. B. Fenster, C.
Aꢅssa, D. De Souza, O. Mꢀller, Proc. Natl. Acad. Sci. USA 2005,
102, 8103–8108; b) A. Fꢀrstner, D. De Souza, L. Turet, M. D. B.
Fenster, L. Parra-Rapado, C. Wirtz, R. Mynott, C. W. Lehmann,
Chem. Eur. J. 2007, 13, 115–134; c) A. Fꢀrstner, D. Kirk, M. D. B.
Fenster, C. Aꢅssa, D. De Souza, C. Nevado, T. Tuttle, W. Thiel, O.
Mꢀller, Chem. Eur. J. 2007, 13, 135–149; d) A. Fꢀrstner, E. Kattnig,
G. Kelter, H.-H. Fiebig, Chem. Eur. J. 2009, 15, 4030–4043; e) A.
Fꢀrstner, Isr. J. Chem. 2011, 51, 329–345.
[32] S. Sharma, R. A. Martin, Heteroat. Chem. 1994, 5, 429–436.
[33] M. A. Blanchette, W. Choy, J. T. Davis, A. P. Essenfeld, S. Masa-
mune, W. R. Roush, T. Sakai, Tetrahedron Lett. 1984, 25, 2183–2186.
[34] a) U. S. Racherla, H. C. Brown, J. Org. Chem. 1991, 56, 401–404;
b) U. S. Racherla, Y. Liao, H. C. Brown, J. Org. Chem. 1992, 57,
6614–6617; c) for large scale applications from this laboratory, see:
A. Fꢀrstner, M. D. B. Fenster, B. Fasching, C. Godbout, K. Radkow-
ski, Angew. Chem. 2006, 118, 5632–5636; Angew. Chem. Int. Ed.
2006, 45, 5506–5510; d) C. Aꢅssa, R. Riveiros, J. Ragot, A. Fꢀrstner,
J. Am. Chem. Soc. 2003, 125, 15512–15520.
[35] C. Mukai, I. Nomura, S. Kitagaki, J. Org. Chem. 2003, 68, 1376–
1385. It should be noted that the use of Dess–Martin periodinane as
well as Swern conditions led to complex mixtures.
[36] W. C. Still, C. Gennari, Tetrahedron Lett. 1983, 24, 4405–4408.
[37] S. J. Mickel, G. H. Sedelmeier, D. Niederer, F. Schuerch, M. Seger,
K. Schreiner, R. Daeffler, A. Osmani, D. Bixel, O. Loiseleur, J.
Cercus, H. Stettler, K. Schaer, R. Gamboni, A. Bach, G.-P. Chen, W.
Chen, P. Geng, G. T. Lee, E. Loeser, J. McKenna, F. R. Kinder, Jr.,
ˇ
K. Konigsberger, K. Prasad, T. M. Ramsey, N. Reel, O. Repic, L.
[17] For a preliminary report on three prototype analogues, see ref. [12].
[18] For an iejimalide/archazolid chimera, see: E. Moulin, C. Nevado, J.
Gagnepain, G. Kelter, H.-H. Fiebig, A. Fꢀrstner, Tetrahedron 2010,
66, 6421–6428.
Rogers, W.-C. Shieh, R.-M. Wang, L. Waykole, S. Xue, G. Florence,
I. Paterson, Org. Process Res. Dev. 2004, 8, 113–121.
[38] a) A. Arase, M. Hoshi, A. Mijin, K. Nishi, Synth. Commun. 1995,
25, 1957–1962; b) K. C. Nicolaou, K. C. Fylaktakidou, H. Monen-
schein, Y. Li, B. Weyershausen, H. Mitchell, H. Wei, P. Guntupalli,
D. Hepworth, K. Sugita, J. Am. Chem. Soc. 2003, 125, 15433–15442.
[39] a) D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155–4156;
b) S. D. Meyer, S. L. Schreiber, J. Org. Chem. 1994, 59, 7549–7552.
[40] a) K. Matsumura, S. Hashiguchi, T. Ikariya, R. Noyori, J. Am.
Chem. Soc. 1997, 119, 8738–8739; b) K.-J. Haack, S. Hashiguchi, A.
Fujii, T. Ikariya, R. Noyori, Angew. Chem. 1997, 109, 297–300;
Angew. Chem. Int. Ed. Engl. 1997, 36, 285–288; c) R. Noyori, S. Ha-
shiguchi, Acc. Chem. Res. 1997, 30, 97–102; d) for an application
from this laboratory see: A. Fꢀrstner, P. Hannen, Chem. Eur. J.
2006, 12, 3006–3019.
[19] The only available information was that 2 (R1 =Me) carrying a
methyl branch next to the ester linkage is more cytotoxic than 1
(R1 =H) with a proton at this site, see refs. [1,2].
[20] F. Tavares, J. P. Lawson, A. I. Meyers, J. Am. Chem. Soc. 1996, 118,
3303–3304.
[21] a) Under these conditions, reduction of the ester to the alcohol is ac-
companied by reduction of one of the carbonyls of the phthalimide
to the corresponding hemiaminal; b) MnO2 concomitantly oxidizes
the allylic alcohol and the partially reduced phthalimide.
[22] L. A. Carpino, J.-H. Tsao, H. Ringsdorf, E. Fell, G. Hettrich, J.
Chem. Soc. Chem. Commun. 1978, 358–359.
Chem. Eur. J. 2011, 17, 6973 – 6984
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6983

A. Fꢀrstner et al.
[41] J. Ackroyd, M. Karpf, A. S. Dreiding, Helv. Chim. Acta 1985, 68,
338–344.
[42] a) A. Fꢀrstner, K. Langemann, J. Org. Chem. 1996, 61, 3942–3943;
b) A. Fꢀrstner, K. Langemann, Synthesis 1997, 792–803.
[44] For E/Z- control in the formation of medium-sized rings, see: a) A.
Fꢀrstner, K. Radkowski, C. Wirtz, R. Goddard, C. W. Lehmann, R.
Mynott, J. Am. Chem. Soc. 2002, 124, 7061–7069; b) A. Fꢀrstner, T.
Nagano, C. Mꢀller, G. Seidel, O. Mꢀller, Chem. Eur. J. 2007, 13,
1452–1462; c) A. Fꢀrstner, M. Schlede, Adv. Synth. Catal. 2002, 344,
657–665.
[45] For a description of the assays, see: a) H.-H. Fiebig, D. P. Berger,
W. A. Dengler, E. Wallbrecher, B. R. Winterhalter, in Immunodefi-
cient Mice in Oncology (Eds.: D. P. Berger, H. H. Fiebig), Contrib.
Oncol., Vol. 42, Karger, Basel, 1992, pp. 321–351; b) T. Roth, A. M.
Burger, W. Dengler, H. Willmann, H.-H. Fiebig, in Relevance of
Tumor Models for Anticancer Drug Development (Eds.: H. H.
Fiebig, A. M. Burger), Contrib. Oncol., Vol. 54, Karger, Basel, 1999,
pp. 145–156; c) W. A. Dengler, J. Schulte, D. P. Berger, R. Mertels-
mann, H. H. Fiebig, Anti-Cancer Drugs 1995, 6, 522–532; d) H. H.
Fiebig, A. Maier, A. M. Burger, Eur. J. Cancer 2004, 40, 802–820.
[43] Because of the lack of control, our group has usually combined
RCM macrocyclizations with a subsequent reduction, oxidation or
isomerization of the newly formed double bond. For representative
examples, see: a) A. Fꢀrstner, K. Langemann, J. Am. Chem. Soc.
1997, 119, 9130–9136; b) A. Fꢀrstner, T. Mꢀller, J. Am. Chem. Soc.
1999, 121, 7814–7821; c) A. Fꢀrstner, J. Grabowski, C. W. Lehmann,
J. Org. Chem. 1999, 64, 8275–8280; d) A. Fꢀrstner, F. Jeanjean, P.
Razon, Angew. Chem. 2002, 114, 2203–2206; Angew. Chem. Int. Ed.
2002, 41, 2097–2101; e) B. Scheiper, F. Glorius, A. Leitner, A. Fꢀrst-
ner, Proc. Natl. Acad. Sci. USA 2004, 101, 11960–11965; f) A. Fꢀrst-
ner, T. Nagano, J. Am. Chem. Soc. 2007, 129, 1906–1907; g) T.
ˇ
Nagano, J. Pospꢄsil, G. Chollet, S. Schulthoff, V. Hickmann, E.
Moulin, J. Herrmann, R. Mꢀller, A. Fꢀrstner, Chem. Eur. J. 2009,
15, 9697–9706.
Received: January 18, 2011
Published online: May 9, 2011
6984
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 6973 – 6984