Stereoselective preparation of (E)-configured 1,2-disubstituted propenes from two aldehydes by a two-carbon stitching strategy: Convergent synthesis of 18,21-diisopropyl-geldanamycin hydroquinone and its C7 epimer

Article abstract of DOI:10.1002/chem.201201600

The title compounds were synthesized in a longest sequence of 27 linear steps and with an overall yield of 2.9 and 3.9 %. In the course of the synthesis, two aldehydes representing carbon fragments C1-C7 (Eastern fragment) and C9-C21 (Western fragment) were prepared from D-mannitol, each of which incorporated a key stereogenic center at the respective secondary methyl ether group (C6, C12) from the chiral pool material. The assembly of the two aldehydes was achieved employing α-chloroethyl magnesium chloride as a two-carbon building block. The carbenoid reagent was generated from α-chloroethyl para-tolylsulfoxide by sulfoxide-magnesium exchange and it added smoothly to the highly sensitive aldehyde of the Eastern fragment (C1-C7). Upon oxidation, an α-chloroethyl ketone was generated, which underwent a clean and high-yielding reductive SmI₂-promoted addition to the other aldehyde fragment. Dehydration delivered the key double bond between C8 and C9 in an overall yield of 72 % over four steps. The method was shown to be generally applicable to the racemization-free conversion of several aldehydes into the respective α-chloroethyl ketone (11 examples, 64-95 %) and to the coupling protocol (5 examples, 66-90 %). The further course of the geldanamycin hydroquinone synthesis included a diastereoselective reduction at C7 and the implementation of the amino group at C20. Since deprotection of the two isopropyl protecting groups could not be achieved in significant yields, the structure of 18,21-diisopropyl-geldanamycin hydroquinone was proven by its independent synthesis from the natural product. Convergency in the synthesis of the geldanamycin skeleton has been achieved by employing a magnesium carbenoid as a two-carbon building block, connecting two aldehyde fragments in an efficient and high yielding manner. The preparation of α-chloroethyl ketones from aldehydes has been investigated (11 examples, 64-95 % yield) and the title compounds have been prepared from advanced precursor 1. Copyright

Full text of DOI:10.1002/chem.201201600

FULL PAPER
DOI: 10.1002/chem.201201600
Stereoselective Preparation of (E)-Configured 1,2-Disubstituted Propenes
from two Aldehydes by a Two-Carbon Stitching Strategy: Convergent
Synthesis of 18,21-Diisopropyl-Geldanamycin Hydroquinone
and Its C7 Epimer
Thomas Hampel,[a, c] Thomas Neubauer,[a] Thomas van Leeuwen,[b] and Thorsten Bach*[a]
Abstract: The title compounds were
synthesized in a longest sequence of 27
linear steps and with an overall yield of
2.9 and 3.9%. In the course of the syn-
thesis, two aldehydes representing
by sulfoxide–magnesium exchange and
it added smoothly to the highly sensi-
tive aldehyde of the Eastern fragment
mization-free conversion of several al-
dehydes into the respective a-chloro-
ethyl ketone (11 examples, 64–95%)
and to the coupling protocol (5 exam-
ples, 66–90%). The further course of
the geldanamycin hydroquinone syn-
thesis included a diastereoselective re-
duction at C7 and the implementation
of the amino group at C20. Since de-
protection of the two isopropyl protect-
ing groups could not be achieved in sig-
nificant yields, the structure of 18,21-
ꢀ
(C1 C7). Upon oxidation, an a-chloro-
ethyl ketone was generated, which un-
derwent a clean and high-yielding re-
ductive SmI2-promoted addition to the
other aldehyde fragment. Dehydration
delivered the key double bond between
C8 and C9 in an overall yield of 72%
over four steps. The method was shown
to be generally applicable to the race-
ꢀ
carbon fragments C1 C7 (Eastern frag-
ꢀ
ment) and C9 C21 (Western fragment)
were prepared from d-mannitol, each
of which incorporated a key stereogen-
ic center at the respective secondary
methyl ether group (C6, C12) from the
chiral pool material. The assembly of
the two aldehydes was achieved em-
ploying a-chloroethyl magnesium chlo-
ride as a two-carbon building block.
The carbenoid reagent was generated
from a-chloroethyl para-tolylsulfoxide
diisopropyl-geldanamycin
hydroqui-
none was proven by its independent
synthesis from the natural product.
ꢀ
Keywords: carbenoids · C C cou-
pling · halogenation · natural prod-
ucts · synthetic methods
Introduction
is structurally characterized by a 19-membered macrolactam
with six stereogenic centers within the ring system
(Figure 1). In contrast to most other ansamycin antibiotics
(e.g., herbimycins[6] and macbecin I[7]), geldanamycin (1)
bears a C17-methoxy group as an additional substituent at
The discovery that the heat shock protein 90 (Hsp90) is ef-
fectively blocked by the ansamycin antibiotic geldanamycin
(1) and that this inhibition has potential consequences for
the treatment of various cancer types[1,2] has sparked signifi-
cant synthetic interest in geldanamycin (1).[3,4] The com-
pound, which was first isolated in 1970 by De Boer et al.,[5]
ꢀ
its 2,6-disubstituted 1,4-benzoquinone ring (C16 C21).
The methoxy group lends sig-
nificant liver toxicity to the
compound, which is the reason
why not geldanamycin (1) itself
but C17 analogues[8] have been
put to advanced clinical trials.
The methoxy group also disa-
bles a straightforward protect-
[a] Dr. T. Hampel,+ Dipl.-Chem. T. Neubauer,+ Prof. Dr. T. Bach
Lehrstuhl fꢀr Organische Chemie 1
Technische Universitꢁt Mꢀnchen
Lichtenbergstrasse 4, 85747 Garching (Germany)
Fax : (+49)89-289-13315
E-mail: thorsten.bach@ch.tum.de
Figure 1. Structure and num-
bering of geldanamycin (1).
ing group strategy for the hy-
droquinone precursor of the
[b] T. van Leeuwen
University of Groningen
Nijenborgh 4
sensitive benzoquinone. While
several successful ansamycin total syntheses[9] have em-
ployed a 1,4-dimethoxybenzene as a protected form of the
benzoquinone,[10] it has turned out that a 1,2,4-trimethoxy-
benzene is not a suitable protecting group for the 2-me-
thoxy-1,4-benzoquinone unit of geldanamycin (1). Oxidative
deprotection (HNO3, HOAc) in the final step of the Andrus
synthesis[3a] led mainly to the formation of ortho-quinogelda-
namycin (50%) and only to minor quantities of the desired
9747 AG Groningen (Netherlands)
[c] Dr. T. Hampel+
New address:
Boehringer Ingelheim Pharma GmbH & Co. KG
Global Department Process Development Chemicals
Binger Strasse 173, 55216 Ingelheim (Germany)
[+] Both authors contributed equally to the project.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201201600.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

natural product (5%). Based on extensive precedence for
the successful deprotection of 1,4-diisopropoxybenzenes[11]
by BCl3 or AlCl3 we had in model studies towards simple
geldanamycin analogues employed isopropyl groups for pro-
tection of the hydroxy groups at C18 and C21.[4a,c,d] Their re-
moval was readily accomplished by treatment with BCl3 in
CH2Cl2. Following this analogy and encouraged by the fact,
that Qin and Panek had accomplished the transformation of
diisopropylether 2 into geldanamycin (1),[3b] compound 2
was considered a good starting point for further retrosyn-
thetic disconnections (Scheme 1). In an effort to make the
assembly of two independently synthesized geldanamycin
fragments (Western and Eastern part) as convergent as pos-
sible, our attention shifted from an earlier pursued ring clos-
ing metathesis strategy (bond formation between C4 and
C5) to a strategy, in which the olefinic double bond between
C8 and C9 was to be formed.
thetic efforts are discussed in this paper as are the deprotec-
tion attempts, which unfortunately did not lead to significant
amounts of geldanamycin (1) and its C7 epimer.
Results and Discussion
Synthesis of the Western and Eastern fragments and their
connection via reductive coupling: The synthesis of aldehyde
3 commenced with known homoallylic alcohol 5,[12b] which is
readily obtained in ten synthetic steps from 2,5-di-O-methyl-
d-mannitol[15] (29% overall yield). Formation of benzyl
ether 6 was accomplished with benzyl bromide and potassi-
um hydride in THF (Scheme 2). The subsequent oxidation
of the terminal double bond was best performed in a two-
step sequence. Oxidation with N-methylmorpholin-N-oxide
[16]
(NMO) in the presence of catalytic amounts of OsO4 de-
livered a diastereomeric mixture of the respective diols,
which were cleaved with sodium periodate in aqueous ace-
tone (ac).[16b,17] Aldehyde 3 was used immediately without
further purification for the subsequent reactions (see
below).
Scheme 1. Convergent retrosynthetic disconnection of geldanamycin pre-
cursor 2 into the Western part, aldehyde 3, and the Eastern part, alde-
hyde 4.
The Western part 3, which evolved from the above-men-
tioned retrosynthetic considerations, seemed easily available
from d-mannitol based on previous work.[12] The stereogenic
center at C12 was retained from the chiral pool material
upon oxidative cleavage and the stereogenic center at C14
was established by a directed hydrogenation. The stereogen-
ic centers at C10 and C11 were formed by an enantioselec-
tive crotylation reaction.[13] In the Eastern part, it was envis-
aged that the stereogenic center at C6 could also stem from
d-mannitol. Aldehyde 4 seemed a useful building block for
the Eastern part, to which different reagents could be
added, with which the desired bond formation could be ach-
ieved.
In this account we report on the synthesis of aldehyde 4
and our attempts to connect it to the Western part 3 via an
appropriate two-carbon building block. It was found that 1-
chloroethyl magnesium chloride nicely serves this purpose
and a general method was established, with which aldehydes
can be converted into the respective a-chloroethyl ketones.
As previously shown by Jamison et al.,[14] the latter com-
pounds are in turn capable to undergo a reductive SmI2-pro-
moted coupling reaction with aldehydes to provide after de-
hydration the respective E-configured a,b-unsaturated ke-
tones. When applied to compounds 3 and 4 a successful as-
sembly was possible to an intermediate which was further
converted into the title compounds upon reduction at C7
and macrolactamization. The individual steps of these syn-
Scheme 2. Benzylation of previously reported[12b] secondary homoallylic
alcohol 5 to the respective homoallylic benzyl ether 6, which served as
immediate precursor for aldehyde 3.
ꢀ
The propensity of the C4 C5 double bond for Z/E iso-
merization let us choose a synthetic route to aldehyde 4,
which avoided the use of acidic conditions for the deprotec-
tion of the adjacent secondary alcohol at C6 (Scheme 3).
Literature known lactone 7 was obtained in five synthetic
operations from d-mannitol (42% overall yield).[18] Its re-
duction to the respective lactol was achieved with diisobuty-
laluminumhydride (Dibal-H). Wittig reaction with a stabil-
ized ylide[19] delivered with perfect E selectivity the a,b,g,d-
unsaturated ester 8, in which the hydroxy group at C6 was
ideally displayed for the subsequent methylation step.
A strict temperature control was mandatory to avoid Z/E
ꢀ
isomerization of the C4 C5 double bond in the methylation
of secondary alcohol 8. At temperatures above ꢀ208C iso-
merization set in, which presumably is facilitated by conju-
gate addition to the dienoate. If the methylation was per-
formed at ꢀ208C, the isomerization was surpressesd and
methyl ether 9 was formed quantitatively. While the subse-
quent ester reduction to primary alcohol 10 and its protec-
tion as tert-butyldiphenylsilyl (TBDPS) ether 11 in the pres-
ence of imidazole (im) proceeded uneventfully, the chemo-
selective deprotection of the primary alcohol at C7 required
&
2
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Natural Products
FULL PAPER
Scheme 3. Conversion of d-mannitol-derived lactone 7 to primary alcohol
12, which served as immediate precursor for aldehyde 4.
some optimization. Eventually, it was found that camphor-
sulphonic acid (CSA) at 08C was well suited to achieve not
only the desired selectivity in the deprotection step but also
to avoid an undesired Z/E isomerization at C4/C5. Apart
from the desired Z,E-diene 12 (92%), only minor amounts
of the respective E,E-diene were isolated (3%). Aldehyde 4
was prepared from primary alcohol 12 by oxidation with
o-iodoxybenzoic acid (IBX).[20]
Scheme 4. Synthesis of a-bromoketone 15, its reductive coupling with al-
dehyde 3 and subsequent dehydration to a,b-unsaturated enone 17.
metrically, the yield was 52%, with two equivalents of the
a-bromoketone the yield increased to 76% and with three
[24]
equivalents to 81%. The use of freshly prepared SmI2 is
strongly recommended to achieve optimum yields. For the
subsequent dehydration step we recurred to the use of the
Martin sulfurane[25] reagent, which had been previously used
by Jamison et al. for the same purpose.[14] The desired enone
17 was obtained as pure E isomer in a yield of 79%. Al-
though the presented route is a valid solution for the retro-
synthetic disconnection presented earlier (Scheme 1), it was
not fully satisfying from a preparative point of view. In par-
ticular, the a-bromination was not reproducible on large
scale and often provided yields below 40%. Given that an
excess of a-bromoketone was required in the reductive cou-
pling this loss of material was not acceptable. Since a signifi-
cant effort had been made to optimize the route to aldehyde
4 in an overall yield of 28% from d-mannitol in eleven syn-
thetic steps, we considered it worthwhile to also optimize
the access to key intermediate 17.
Aldehyde 4 served as starting material for the preparation
of potential nucleophiles at carbon atom C8, which were
ꢀ
tested in potential C C bond forming reactions with alde-
hyde 3. In these reactions, the former aldehyde turned out
to be very capricous, partially due to its propensity for Z/E
isomerization, partially due to its acidity at carbon atom C6.
Even the normally straightforward Grignard addition to
form alcohol 13 (Scheme 4) proceeded only in a yield of
69%. Unfortunately, the readily accessible ketone 14, which
was prepared by oxidation with the Dess–Martin periodi-
nane (DMP),[21] did not prove to be significantly more
stable. All attempts to convert it into an enolate (Li, B, Ti)
or a silyl enol ether by deprotonation at C8 failed complete-
ly. The synthesis of appropriate olefination reagents from al-
dehyde 4 was extensively pursued but did not lead to the de-
sired success. Given the repeated failure encountered in
these attempts, it was gratifying to note that a-bromoketone
15, which was obtained by treatment of ethyl ketone 14 with
phenyltrimethyl ammonium tribromide (PTAB),[22] under-
went a reductive Reformatsky-type[23] coupling with alde-
hyde 3. As nicely precedented by the work of Jamison
et al.,[14] SmI2 was found to be an excellent reductant for
this step. It is noteworthy that a-bromoketone 15, which—in
contrast to the tert-alkyl substrates studied by Jamison
et al.[14c]—contains a sensitive stereogenic center at the
second a-position, reacted smoothly and afforded the de-
sired product 16 in up to 81% yield. The yield, which is
based on aldehyde 3 as the limiting reagent, varied depend-
ing on the equivalents of a-bromoketone. If used stoichio-
Preparation of a-chloroketones and their reductive cou-
pling: An improved route for the preparation of enone 17:
In light of the encountered difficulties preparing ketone 15
and being aware of the fact that a-chloroketones had been
reported to be an equally potential coupling partner in the
SmI2-promoted reactions,[14] we considered the possibility to
prepare an a-chloroketone related to 15 from aldehyde 4 by
direct addition of an a-chloro Grignard reagent and subse-
quent oxidation. Carbenoids of this type have been shown
to be accessible by sulfoxide-magnesium exchange
(Scheme 5), the driving force of which has been proposed to
be the increased stabilization of the negative charge at the
carbon atom by the halogen substituent.[26]
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

T. Bach et al.
Scheme 5. General strategy for the formation of magnesium carbenoids
B from the respective aromatic a-chlorosulfoxides A.
Literature precedence for the preparation of a-chloroeth-
yl magnesium chloride (Alkyl=Me) and its addition to car-
bonyl compounds was not found. We consequently under-
took a few preliminary experiments regarding the nature of
the para-substituent X in the aryl sulfoxide and of the group
R’ in the reagent R’MgCl. It was found that the best perfor-
mance was achieved with X=Me and R’=iPr. The benefi-
cial influence of X=Me on the exchange reaction became
clear by comparison with X=Cl or X=H. The formation of
aryl magnesium species was minimized and only the desired
carbenoid was obtained. The superiority of R’=iPr over
R=Et is due to the lower nucleophilicity of the R’MgCl re-
agent. In order to ensure quantitative conversion of the sulf-
oxide, the sulfoxide-magnesium reaction was performed at
ꢀ788C with a slight excess of R’MgCl. For the subsequent
carbonyl addition to an aldehyde (see below) the mixture
was warmed to ꢀ608C. If R’ was ethyl, formation of the re-
spective ethyl substituted secondary alcohols was observed,
while this side reaction could be avoided with iPrMgCl. A
final side reaction to be eliminated was due to overoxidation
of sulfide 18 when applying a known procedure.[27] Although
the desired sulfoxide 19 was formed in high yield (81%) by
using a mixture of AgNO3 and SO2Cl2 in acetonitrile for the
oxidation/a-chlorination, minor quantities of ethyl para-tol-
ylsulfone were formed, which could not be separated from
the reagent. A variant of this procedure[27] using potassium
nitrate (Scheme 6) delivered the desired product 19 in lower
yields but without any inseparable impurities. Sulfoxide 19
was obtained as a 9:1 mixture of diastereoisomers.
Scheme 7. Preparation of a-chloroketones 20 from a-chlorosulfoxide 19
by sulfoxide–magnesium exchange, aldehyde addition and subsequent ox-
idation.
aldehyde is chiral the respective a-chloroethyl ketone is
formed as a mixture of diastereoisomers, which is not rele-
vant for its use in the reductive coupling reaction (see
below). Still, the ratio of diastereoisomers (d.r.) was deter-
mined for these aldehydes (20e–g) and is also provided in
Scheme 7. A racemization at the a-stereogenic center was
not observed. Product 20e, for example, was obtained from
the respective enantiomerically pure aldehyde in >95% ee.
The result was further substantiated by the fact that consec-
utive products derived from the respective a-chloroethyl
ketone were also obtained racemization-free (see below). It
should be mentioned, however, that steric hindrance at the
aldehyde has an influence on the reaction rate. The reactivi-
ty of the a-chloroethyl Grignard reagent is not sufficient to
attack an aldehyde with an adjacent quaternary carbon
center. Reduction to the alcohol was observed instead.
Some sulfur-containing substrates (e.g. thiophenes) are not
compatible with a-chloroethyl magnesium chloride, possibly
due to the formation of a sulfur ylide by reaction of the
sulfur atom with the carbenoid.
In a less extensive study the conversion of a-chloroethyl
ketones into the respective a,b-unsaturated ketones was in-
vestigated (Scheme 8). To this end, an a-chloroketone (see
Scheme 7) and the aldehyde substrate were dissolved in
THF and a freshly prepared solution of SmI2 in THF was
slowly added (see Experimental Section). Products were ob-
tained as mixtures of diastereoisomers after chromatograph-
ic purification and were taken directly into the dehydration
reaction with the Martin sulfurane reagent. The reaction
Scheme 6. Preparation of a-chlorosulfoxide 19 by chlorination of ethyl-
para-methylphenylsulfide (18).
Under optimized conditions several aldehydes were sub-
jected to an addition of the a-chloroethyl magnesium chlo-
ride reagent and the resulting alcohols were further oxidized
to the respective a-chloroethyl ketones (Scheme 7). For the
oxidation, the Dess–Martin periodinane (DMP) turned out
to be the best reagent. The functional group tolerance of the
reaction sequence Grignard addition/oxidation was high.
Specifically, the compatibility with protected hydroxy
groups (20a, 20c, 20e, 20g), with protected amines (20b, f),
with alkenes (20c), with alkynes (20d) and various aromatic
substitutents (20g–k) could be demonstrated. If the starting
&
4
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Natural Products
FULL PAPER
was strictly E-selective for most substrates and products
21a–d were obtained in high yield. For product 21c the re-
action sequence was proved to be racemization-free (>95%
ee). Only in one instance (product 21e), a mixture of E/Z
isomers was obtained. An explanation for the high E selec-
tivity has been provided earlier by Jamison et al.[14c]
was employed, according to which the pre-cooled SmI2 solu-
tion was added to the mixture of a-chloroketone 20l and al-
dehyde 3 in THF. The product 16 obtained with this proce-
dure was dehydrated with Martin sulfurane and delivered
the desired E-configured enone 17 in diastereomerically
pure form (71% yield from 20l). Racemization at C7 or at
any other stereogenic center was not observed.
Synthesis of 18,21-diisopropyl-geldanamycin hydroquinone
and its C7 epimer, deprotection attempts: The reduction of
enone 17 at C7 was performed under a variety of conditions.
Although it was anticipated that Felkin–Anh control[28]
would deliver the required 7S isomer, the fact that—irre-
spective of the reaction conditions—a single diastereoisomer
was consistently observed made us suspicious. Indeed, the
non-chelating reducing agents L-Selectride (94%) and K-Se-
lectride (91%) delivered the same product as the Luche re-
[29]
agent NaBH4/CeCl3 (89%) or ZnACHTNUTRGNEUNG(BH4)2 (91%). In partic-
ular, the latter reagent is known to exert a high degree of
chelation control,[30] which would mean in our case forma-
tion of the 7R isomer. Since the preparation of novel iso-
mers of geldanamycin was in the focus of our studies, we
looked for a reducing agent, which would make both the 7S
isomer and the 7R isomer of product 22 available. We found
the reduction with borane in the presence of 3,3-diphenyl-1-
methylhexahydropyrrolo-[1,2-c,1,3,2]oxazaborole[31] a useful
way to achieve this goal (Scheme 10). The use of the S
isomer of the oxazaborolidine delivered both products in
a 1:1 ratio (77% yield), whereas the R isomer favored for-
mation of the previously encoutered diastereoisomer (65%
yield). Since the separation of the diastereoisomers was
facile we continued the synthesis with both compounds.[32] It
turned out at the end of the synthetic sequence (see below)
that the diastereoisomer observed with achiral reducing
agents was the 7R isomer 22a, whereas the only way to
obtain meaningful quantities of the 7S isomer 22b was the
oxazaborolidine-mediated reduction. In this respect it was
fortitious that the 7R isomer could be readily oxidized with
the Dess–Martin periodinane to the starting enone and used
for a second reduction reaction. If this recycling step was in-
Scheme 8. Stereoselective preparation of a,b-unsaturated enones 21 from
a-chloroketones 20 by reductive coupling with an aldehyde and subse-
quent dehyration.
Although our studies had revealed that a-chloroethyl
magnesium chloride is well suitable as two-carbon building
block for the connection of two aldehydes, it remained open
whether the reaction sequence would be equally well appli-
cable to the assembly of aldehydes 3 and 4. A few modifica-
tions were implemented to achieve optimal yields in the
conversion to aldol product 16 (Scheme 9).
Scheme 9. Synthesis of a-chloroketone 20l and its reductive coupling
with aldehyde 3 to aldol product 16 (see Scheme 4).
The addition of the carbenoid was performed at ꢀ788C
and the reaction was also quenched at this temperature.
Upon work-up the intermediate products were not purified
but immediately oxidized with Dess–Martin periodinane in
the presence of NaHCO3. The reaction sequence yielded
after chromatographic purification a-chloroethyl ketone 20l
in 80% yield, which was also immediately used in the next
step. For the reductive coupling, an inverse addition mode
Scheme 10. Diastereoselective reduction of the keto group at C7 to the
undesired alcohol 22a and—employing a chiral reducing agent—to the
desired alcohol 22b.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

T. Bach et al.
cluded, secondary alcohol 22b could be obtained in a syn-
thetically acceptable yield of 52%. Mechanistically, we did
not investigate the reason for the high propensity towards
a formal chelation-control reduction of enone 17 any fur-
ther. A somewhat related enone has been reported to be re-
duced in a similar fashion but no explanation for the diaste-
reoselectivity was provided.[33] Preferences for the reduction
of a-methyl substituted a,b-unsaturated enones with oxaza-
borolidines can be found in the literature.[34] Following the
suggested model for this kind of reduction,[35] formation of
the 7S isomer is indeed expected with the (S)-oxazaboroli-
dine. Apparently, however, the reagent-controlled stereose-
lectivity is heavily mismatched[36] with the substrate-induced
diastereoselectivity.
Protection of the alcohol group at C7 was required to be
orthogonal to the existing TBDPS protection at C1. Since
precedence for the selective removal of the latter protecting
group at a primary alcohol in the presence of a secondary
TBS ether existed,[37] a silyl protection at C7 was performed
with TBSCl/imidazole (im) in DMF (Scheme 11). The regio-
selective nitration of the electron-rich arene in the resulting
products 23 was conducted under conditions,[38] which we
had successfully employed in previous syntheses of simple
geldanamycin analogues.[4a,c,d] The nitro group served as pro-
tected amino group and avoided any protecting group ma-
nipulations at this position while adjusting the oxidation
state at carbon atom C1. Introduction of the nitro group
prior to the reductive coupling is feasible and a Cbz-protect-
ed aniline can be generated from the respective nitro ben-
zene. It was found, however, that the Cbz-protected amino
group is incompatible with the dehydration conditions. The
chemoselective deprotection of the primary silyl ether was
readily achieved for diastereoisomer 24a with HF·py in
THF (97% yield). Unfortunately, the other diasteroisomer
24b was less well behaved. Apart from the desired primary
alcohol 24b (60%) significant amounts of the fully desilylat-
ed product (39%) were isolated. Luckily, the resulting diol
could be nicely converted into the desired product by disily-
lation with TBSCl/im in DMF (85%) and chemoselective
deprotection at C1 (84%; see Supporting Information).
Since the overall yield remained high, also in the 7S series,
we did not make an effort to change the protective group
strategy in particular because the selectivity in the earlier re-
quired silyl deprotection primary TBS ether vs. primary
TBDPS ether had been carefully optimized (see Scheme 4).
The primary alcohol at C1 was oxidized to the respective al-
dehyde, which in turn was further oxidized to the carboxylic
acid using Pinnickꢃs conditions.[39]
At the stage of the carboxylic acid it was attempted to
reduce the nitro group to the amino group according to the
Lalancette protocol (NaBH4, S),[40] which we had employed
earlier for the preparation of related anilines.[4a,c,d] It was
found, however, that the chemoselective reduction was not
possible and concomitant reduction of the carboxylic acid to
the allylic alcohol was observed. To avoid the reduction,
which was presumably due to the formation of borane in the
reaction mixture, the carboxylic acid was converted into the
methyl ester 25 by treatment with trimethylsilyl diazome-
thane.[41] The three-step sequence from the alcohol to the
methyl ester proceeded in high yields for both diastereoiso-
mers (25a: 93%; 25b: 91%). The reduction of the nitro
group could be subsequently performed cleanly and in high
yields. The resulting anilines 26 already displayed the nucle-
ophilic reaction partner for the required macrolactamiza-
tion, which could be accomplished upon liberating the car-
boxylic acid by saponification. The reagent of choice for the
macrolactamization was bis(2-oxo-3-oxazolidinyl)phosphinic
acid chloride (BOPCl)[42] due to his high functional group
tolerance and its reported suitability for related ring closure
reactions.[3a,10a,c,d,f] The 19-membered macrolactams 27 were
formed in good yields over two steps. While all NMR spec-
tra before the ring closure had been well resolved, macrolac-
Scheme 11. Elaboration of the aldol condensation products 22 into the advanced macrocycles 29 and 2. The reaction sequence was performed both for
the 7R (22a) and the 7S epimer (22b). The first yield (top) refers to the former series 7R (23a–28a), the second yield (bottom) to the latter series 7S
(23b–28b).
&
6
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Natural Products
FULL PAPER
tams 27 and their consecutive products showed relatively
1
broad peaks in the H NMR spectra. Spectra sharpened at
low temperature but several sets of signals became observa-
ble, which rather complicated than simplified the analysis.
High-temperature NMR studies seemed at this point not
warranted because we were afraid of product loss due to de-
composition and re-isolation problems. Since the individual
protons could be clearly assigned based on the recorded
spectra the synthetic efforts were continued. The reason for
the line broadening is presumably the existence of diastereo-
meric conformations due to the restricted movement of the
ansa chain around the central highly substituted benzene
ring.[4c,43] The relative ratio of these diastereoisomers is not
neccessarily equal and depends on the ring closure method
(see below). Upon deprotection of the silyl ether at C7, sec-
ondary alcohols 28 were generated, which were further con-
verted in the respective carbamates, employing trichloroace-
tyl isocyanate as the reagent.[43] The intermediary formed N-
trichloroacetyl carbamate was hydrolyzed with potassium
carbonate in methanol. An alternative protocol employing
sodium cyanate and trifluoroacetic acid (TFA)[44] was not
successful in this instance. No product was formed and un-
reacted starting material was recovered. It was expected
that one of the two carbamates should be identical in its
1H NMR data to the data of compound 2 earlier reported
by Qin and Panek.[3b] Unfortunately, this was not the case,
which is likely due to the fact that they had used another
protocol for the ring closure thus generating a product mix-
ture of different diastereomeric composition arising from
the previously mentioned planar chirality.
Figure 2. Structure of products 30a, 31, 32a, and 33a obtained by at-
tempted deprotection reactions.
ꢀ788C (83%). Deprotection attempts to geldanamycin and
its C7 epimer were performed with substrates 2, 29, and 30
(Figure 2). The Lewis acids AlCl3, BCl3 and TiCl4, which are
most frequently used for deisopropylation reactions, were
tested under a variety of conditions. In particular with
AlCl3, it was noted that the water content of the solvent is
very critical for its reactivity. In this respect it is likely that
previously used deprotection conditions (AlCl3, anisole)[3b]
failed for the complete deprotection of compounds 2 and 29
because we were unable to reproduce the precise reaction
conditions.
Gratifyingly, we could secure the constitution and configu-
ration of the above-mentioned carbamates, by deprotection
of the benzyl group at the C11 alcohol with TiCl4 at ꢀ788C
(Scheme 12). Authentic material was prepared via reduc-
The most important observations we made in the course
of our deprotection attempts were that TiCl4 led at higher
temperature in CH2Cl2 (ꢀ58C) to a partial deprotection of
the diisopropylether and the monoisopropylated hydroqui-
nones 31 were obtained for both diastereoisomers. The con-
stitution assignment is somewhat tentative but it is likely
that deprotection occurred at position C21. If these com-
pounds were further treated with AlCl3 in CH2Cl2 (ꢀ788C
! RT) traces of geldanamycin (1) and its C7-epimer could
be detected by HPLC analysis. The constitution of the mate-
rial was secured by HRMS data and the retention times let
us clearly distinguish between the natural product and its C7
epimer. Attempted oxidation of compound 31a gave a com-
plex product mixture with ceric ammonium nitrate (CAN)
in acetonitrile on one hand and resulted on the other hand
with 2,3-dichloro-5,6-dicyanobenzoquinone (DDQ) in
a clean conversion to a product, to which we assign the
structure of an ortho-azaquinone (32a). The latter observa-
tion supports the above-mentioned selectivity in the deiso-
propylation step. The use of BCl3, which had proven to be
a very reliable deprotecting reagent in model studies,[4a,c,d]
was disappointing. It induced already at ꢀ208C a demethox-
ylation at C6 and the respective chloride was detected. Re-
activity at this position was also observed when AlCl3 was
employed in acetonitrile as the solvent. A cyclic carbonate
was formed from precursor 31b, to which we assign struc-
ture 33b. As already mentioned above, the behavior of
ACHTUNGTRENNUNG
tion[8a] from the natural product, which is commercially
available. The isopropylation of the hydroquinone was not
optimized and the yield in this step remained relatively low.
Triple alkylation was observed as a side reaction occurring
presumably at the lactam nitrogen atom. Compounds 30b
obtained by derivatization of the natural product and by the
synthetic sequence shown in Schemes 10 and 11 were identi-
cal in all spectral data and in their chromatographic (HPLC-
MS) behavior. High temperature NMR spectra could be re-
corded in [D6]DMSO, which led to satisfactory spectra.
The C7 epimer of compound 30b, compound 30a, could
also be prepared by selective debenzylation with TiCl4 at
Scheme 12. Preparation of the potential relay compound 30b from syn-
thetically available benzyl ether 2 and from natural (+)-geldanamycin
(1).
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

T. Bach et al.
based on COSY, NOESY-2D, HMBC and HMQC experiments. 1H or
13C NMR signals are usually assigned using significant short sections of
the molecular formular. Optical rotations were measured using a Perkin–
Elmer 241 MC Polarimeter.
AlCl3 heavily depended on the solvent quality. For the
above-mentioned conversion of 31 into geldanamycin and
its C7-epimer CH2Cl2 was taken from a conventional solvent
purification system. If CH2Cl2 was further dried by treat-
ment with highly reactive alumina under absolute exclusion
of air and if the reaction was repeated under otherwise iden-
tical conditions, there was no significant conversion of the
starting material even at room temperature. Eventually, the
attempts to achieve a complete deprotection of compounds
2, 29, and 30 were discontinued.
Representative procedure for the synthesis of a-chloroketones 20
(Scheme 7) 2-Chloro-1-(1-tosyl-1H-indol-3-yl)propan-1-one (20k): A
solution of i PrMgCl (1.05 mL, 2.1 mmol, 2m in THF) was added at
ꢀ788C to a solution of 1-(1-chloroethyl)sulfinyl-4-methylbenzene (19)[27]
(400 mg, 1.8 mmol) in THF (5 mL) upon which the color turned slightly
yellow. The reaction mixture was stirred for 30 min at ꢀ788C, after which
a solution of 1-tosyl-1H-indole-3-carbaldehyde (144 mg, 0.48 mmol) in
THF (3 mL) was added dropwise and the reaction mixture was warmed
to ꢀ608C in one hour. The reaction was hydrolyzed by the addition of sa-
turated aq. NH4Cl (5 mL). Subsequently, CH2Cl2 (10 mL) and water
(5 mL) were added and the aqueous layer was extracted with CH2Cl2 (3ꢄ
5 mL). The combined organic layers were dried over Na2SO4 and the sol-
vent was evaporated in vacuo.
Conclusion
Despite the fact that a complete deprotection of the title
compounds 30 could not be achieved, the synthetic route to-
wards these compounds demonstrates the major benefit of
the chosen retrosynthetic disconnection at C8/C9, that is, its
high convergency. Two aldehyde building blocks, 3 and 4, of
comparable complexity have been assembled employing
a new strategy for the conversion of an aldehyde into an a-
chloroethyl ketone. This method has been shown to be gen-
erally applicable not only for the formation of these ketones
but also for the formation of enones with another aldehyde
in a SmI2-promoted reductive coupling and a subsequent de-
hydration. If access to modified geldanamycin derivatives is
desired by de novo synthesis the chosen synthetic route
opens an efficient pathway for their preparation. A modified
protection group strategy might be advisable if the arene
part of the modified derivative should still contain the 2,6-
disubstituted 3-methoxy-1,4-benzoquinone ring of geldana-
mycin.
The crude product, obtained from the previous step, was dissolved in
CH2Cl2 (5 mL). Dess–Martin periodinane[21] (600 mg, 1.4 mmol) was
added and the reaction mixture was stirred for three hours. Diethyl ether
(10 mL), saturated aq. NaHCO3 (10 mL) and saturated aq. Na2S2O3
(10 mL) were added to the reaction mixture. This biphasic system was
stirred for 10 min before the aqueous layer was extracted with CH2Cl2
(3ꢄ10 mL). The combined organic layers were dried over Na2SO4, fil-
tered and the solvent was evaporated in vacuo. Upon purification by
flash chromatography (pentane/diethyl ether 3:1), 20k (166 mg,
0.456 mmol, 95% over two steps) was isolated as a white amorphous
solid. Rf =0.35 (P/Et2O 3:1) [UV]; 1H NMR (360 MHz, CDCl3): d=1.80
(d, J=6.7 Hz, 3H, CClCH3), 2.39 (s, 3H, CarCH3), 5.07 (q, J=6.7 Hz,
1H, CHCl), 7.31 (virt. d, J=8.4 Hz, 2H, CarH), 7.35–7.45 (m, 2H, C5H +
C6H), 7.86 (virt. d, J=8.4 Hz, 2H, CarH), 7.96 (dd, J=6.7, 1.9 Hz, 1H,
C7H), 8.35 (dd, J=6.7, 1.9 Hz, 1H, C4H), 8.42 ppm (s, 1H, C2H);
13C NMR (91 MHz, CDCl3): d=20.2 (q, CClCH3), 21.7 (q, CarCH3), 54.6
(d, CCl), 113.1 (C7), 117.8 (s, C3), 123.1 (d, C5), 125.1 (d, C4), 126.0 (d,
C6), 127.2 (d, CarH), 127.9 (C3a), 130.3 (d, CarH), 132.6 (d, C2), 134.3 (s,
CarS), 134.8 (s, C7a), 146.1 (s, CarCH3), 189.7 ppm (s, C=O); IR (ATR): n˜
=1651 (s, C=O), 1533 (s, Ar), 1381 (s), 1169 (s, Ar), 1140 (s), 975 (s)
cmꢀ1; MS (EI): m/z (%): 361.04 (12) [M+], 298.05 (100) [MꢀCH3CH2Cl+
], 212.06 (24), 155.00 (66) [C7H7O2S+], 91.05 (76) [C7H7+]; HRMS (EI):
m/z: calcd for C18H16NO3ClS: 361.0534, found: 361.0525 [M+].
Representative procedure for the synthesis of a,b-unsaturated ketones 21
(Scheme 8): (E)-1,8-Bis(4-methoxybenzyloxy)-5,7,7-trimethyloct-5-en-
4-one (21a): To a solution of SmI2 in THF (0.1m, 5 mL, 0.5 mmol) at
ꢀ788C, a solution of 2-chloro-6-(4-methoxybenzyloxy)hexan-3-one (20a)
(27.1 mg, 0.1 mmol) and 3-(4-methoxybenzyloxy)-2,2-dimethylpropanal
(24.5 mg, 0.11 mmol) in THF (5 mL) was added dropwise, by slow addi-
tion of the individual drops via the walls of the flask. The blue reaction
mixture was stirred for 1 h at ꢀ788C, before bubbling air through the re-
action mixture. Saturated aq. NaHCO3 (10 mL), Na2S2O3 (10 mL) and
CH2Cl2 (20 mL) were added. The layers were separated and the aqueous
layer was extracted with CH2Cl2 (3ꢄ10 mL). The combined organic
layers were dried over Na2SO4 and filtered. Purification by flash chroma-
tography (pentane/diethyl ether 1:1) afforded an a-hydroxy ketone
(40.5 mg, 0.092 mmol, 92%), which was directly submitted to the next
step.
Experimental Section
General methods: All reactions involving water-sensitive chemicals were
carried out in flame-dried glassware with magnetic stirring under argon.
Tetrahydrofuran (THF), diethyl ether (Et2O) and dichloromethane
(CH2Cl2) were purified using a SPS-800 solvent purification system (M.
Braun). All other chemicals were either commercially available or pre-
pared according to the cited references. TLC was performed on silica
coated glass plates (silica gel 60 F254) with detection by UV (254 nm),
KMnO4 (0.5% in water) or ceric ammonium molybdate (CAM) with
subsequent heating. Flash chromatography was performed on silica gel
60 (Merck, 230–400 mesh) with the indicated eluent. Common solvents
for chromatography [pentane, ethyl acetate, diethyl ether (Et2O), di-
chloromethane, methanol (MeOH)] were distilled prior to use. HPLC
analyses was performed using the stationary phase hydrosphere C18
YMC (analytical: 125ꢄ2.1 mm, 3 mm) employing acetonitrile/water with
0.1% AcOH (10:90 ! 90:10, 30 min) as eluents and UV detection at
358C. IR: JASCO IR-4100 (ATR). GCMS: Agilent 6890 instrument with
Agilent N7873 mass detection, carrier gas: helium (method: 608C, 3 min;
60 ! 3008C, 16 min; 3008C, 5 min). MS/HRMS: Finnigan MAT 8200
(EI)/Finnigan MSD 5973 (HR-EI)/Finnigan LCQ classic (ESI)/Thermo
Finnigan LTQ FT (HRMS-ESI)/Thermo Sientific LTQ Orbitrap XL
To a solution of this a-hydroxy ketone (0.05 mmol, 22.0 mg) in CH2Cl2
(5 mL) was added via
a cannula a
solution of Martin sulfurane[25]
(100 mg, 0.15 mmol) in CH2Cl2 (5 mL). This mixture was stirred for one
hour, before hydrolyzing it by addition of sat. aq. NaHCO3 (5 mL). The
aqueous layer was extracted with CH2Cl2 (3ꢄ5 mL) and the combined
organic layers were washed with aqueous NaOH (1m, 4ꢄ5 mL), before
drying over Na2SO4. After filtration the solvent was evaporated in vacuo.
Purification by flash chromatography (pentane/diethyl ether 2:1) yielded
the desired 21a (18.3 mg, 0.0413 mmol, 83% over two steps) as a colorless
oil. Rf =0.37 (P/Et2O 2:1) [UV]; 1H NMR (360 MHz, CDCl3): d=1.21 [s,
(HRMS-ESI). H and 13C NMR: Bruker AV-250, Bruker AV-360, Bruker
1
AV-500, Bruker AV-500cr. Chemical shifts are reported relative to sol-
ACTHNUTRGNEUNG
vent signals [CDCl3 d(1H)=7.26 ppm, d(13C)=77.16 ppm]. The symbol *
6H,
CACHTGNUTERNN(UG CH3)2], 1.88 (d, J=1.3 Hz, 3H, =CCH3), 1.90–1.97 (m, 2H,
denotes an interconvertible assignment. The multiplicities of the
OCH2CH2), 2.77 (t, J=7.3 Hz, 2H, CH2C=O), 3.29 (s, 2H, CCH2O), 3.49
(t, J=6.2 Hz, 2H, OCH2CH2), 3.81 (s, 6H, OCH3), 4.44 (s, 2H, ArCH2),
13C NMR signal were determined by DEPT experiments, assignments are
&
8
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Natural Products
FULL PAPER
4.49 (s, 2H, ArCH2), 6.69 (q, J=1.3 Hz, 1H, =CH), 6.85–6.93 (m, 4H,
ArH), 7.23–7.31 ppm (m, 4H, ArH); 13C NMR (91 MHz, CDCl3): d=
The product was directly taken into the next step. Rf =0.10 (P/Et2O 4:1)
[UV, CAM].
12.5 (q, =CCH3), 24.8 [q, C
ACHTUNGTRENNUNG(CH3)2], 37.7 (t, CH2C=O), 55.3 (q, OCH3), 69.3 (t, OCH2CH2), 72.4 (t,
ACHTNUGTNERUN(GN CH3)2], 24.9 (t, CH2CH2CH2), 34.0 [s, C-
The diastereomeric mixture of b-hydroxy ketone 16 (314 mg, 0.34 mmol)
was dissolved in dichloromethane (25 mL) and cooled to 08C before
Martin sulfurane[25] (451 mg, 0.67 mmol) was added. After one hour at
08C saturated aq. NaHCO3 (10 mL) was added and the reaction was
warmed to room temperature. The aqueous layer was extracted with di-
chloromethane (3ꢄ10 mL) and the combined organic layers were washed
with brine (10 mL), followed by drying over Na2SO4 and concentrating in
vacuo. The crude product was purified by flash chromatography (pen-
tane/diethyl ether 6:1) to yield the desired a,b-unsaturated ketone 17
(294 mg, 0.27 mmol, 79%) as a colorless oil. Rf =0.74 (P/Et2O 2:1) [UV,
CAM]; [a]D20 = +72.2 (c=0.56, CDCl3); 1H NMR (500 MHz, CDCl3): d=
0.75 (d, 3J=6.6 Hz, 3H, C-14-CH3), 1.08 (d, 3J=6.6 Hz, 3H, C-10-CH3),
ArCH2), 72.9 (t, ArCH2), 79.1 (t, CCH2O), 113.7 (d, CarH), 113.8 (d,
CarH), 129.1 (d, CarH), 129.2 (d, CarH), 130.5 (s, CarCH2), 130.6 (s,
CarCH2), 137.0 (s, C=CH), 148.0 (d, =CH), 159.1 (s, CarOCH3), 159.1 (s,
CarOCH3), 202.6 ppm (s, C=O); IR (ATR): n˜ =1511 (s, Ar), 1244 (s,
OCH3), 1090 (s), 1032 (s), 818 (s) cmꢀ1; MS (EI): m/z (%): 183.11 (74),
121.04 (100) [C8H9O+]; HRMS (EI): m/z: calcd for C27H36O5: 440.2557;
found: 440.2550 [M+].
ACHTUNGTRENNUNG(4S,5Z,7E)-9-(tert-Butyldiphenylsilyloxy)-2-chloro-4-methoxy-8-meth-
ylnona-5,7-dien-3-one (20l): To a solution of 1-(1-chloroethylsulfinyl)-4-
methylbenzene (19) (554 mg, 2.71 mmol) in THF (9.5 mL) at ꢀ788C was
added a solution of i PrMgCl (2.75 mL, 2.75 mmol, 2m in THF) dropwise
and the reaction mixture was stirred for 30 min. Subsequently, aldehyde 4
(300 mg, 0.69 mmol) in THF (7.0 mL) was added and the solution was
stirred for two hours at ꢀ788C. After hydrolysis with saturated aq.
NH4Cl (5 mL) at ꢀ788C, the mixture was allowed to reach room temper-
ature, before the aqueous layer was extracted with diethyl ether (3ꢄ
10 mL). The combined arganic layers were washed with brine (10 mL),
dried over Na2SO4 and concentrated in vacuo. The yellow oil was directly
submitted to the next step. Rf =0.10 (P/Et2O 4:1) [UV, CAM].
1.09 [s, 9H, Si(Ph)2CACHTUNRGTNEN(UG CH3)3], 1.24–1.33 [m, 13H, OCHAHCUTNGTREN(NUGN CH3)2, H-13],
1.68 (s, 3H, C-2-CH3), 1.72–1.78 (m, 1H, H-13), 1.81 (d, 4J=1.0 Hz, 3H,
C-8-CH3), 2.00–2.07 (m, 1H, H-14), 2.49 (dd, 2J=12.4, 3J=8.7 Hz, 1H,
H-15), 2.58 (dd, 2J=12.4, 3J=5.9 Hz, 1H, H-15), 2.70–2.80 (m, 1H, H-
10), 3.19–3.24 (m, 1H, H-12), 3.29 (s, 3H, C-6-OCH3), 3.30 (s, 3H, C-12-
OCH3), 3.44 (dd, 3J=7.9, 3J=2.6 Hz, 1H, H-11), 3.80 (s, 3H, CarOCH3),
4.14 (s, 2H, H-1), 4.34–4.44 [m, 2H, CHACTHNUTRGNEUNG
(CH3)2], 4.48 (d, 2J=11.4 Hz,
1H, OCH2Ph), 4.77 (d, 2J=11.4 Hz, 1H, OCH2Ph), 5.12 (d, 3J=9.3 Hz,
1H, H-6), 5.38 (virt. t, 3J ffi 9.9 Hz, 1H, H-5), 6.47 (d, 3J=8.9 Hz, 1H,
CarH), 6.53 (virt. t, 3J ffi 11.3 Hz, 1H, H-4), 6.62 (d, 3J=11.8 Hz, 1H, H-
3), 6.65–6.67 (m, 1H, H-9), 6.68 (d, 3J=8.9 Hz, 1H, CarH), 7.29–7.32 (m,
The crude a-chlorohydrine was dissolved in dichloromethane (7.0 mL)
before adding solid NaHCO3 (85.5 mg, 1.02 mmol) and Dess–Martin peri-
odinane[21] (439 mg, 1.06 mmol). After three hours at room temperature,
diethyl ether (20 mL), saturated aq. NaHCO3 (10 mL) and saturated aq.
Na2S2O3 (10 mL) were added. After stirring the slurry for ten minutes at
room temperature the aqueous layer was extracted with diethyl ether
(2ꢄ10 mL). The combined arganic layers were washed with saturated aq.
Na2S2O3 (10 mL) and saturated aq. NaHCO3 (10 mL) before drying with
Na2SO4 and concentration in vacuo. The crude product was purified
through a short pad of silica (pentane/diethyl ether 8:1) to yield a mixture
of diastereoisomers (d.r. 78:22) of the desired a-chloroketone 20l
(259 mg, 0.55 mmol, 80%) as a colorless oil. Analytic data are provided
for the major diastereoisomers: Rf =0.49 (P/Et2O 10:1) [UV/CAM];
[a]2D0 = +33.1 (c=0.5, CDCl3); 1H NMR (500 MHz, CDCl3)): d=1.09 [s,
5H, OCH2Ph), 7.35–7.45 [m, 6H, Si(Ph)2C
4H, Si(Ph)2C
(CH3)3]; 13C NMR (90.6 MHz, CDCl3): d=12.3 (q, C-8-
CH3), 14.0 (q, C-2-CH3), 16.6 (q, C-10-CH3), 19.5 (q, C-14-CH3), 19.5 [s,
Si(Ph)2C(CH3)3], 22.4 [q, OCH(CH3)2)], 22.5 [q, OCH(CH3)(CH3)], 22.5
[q, OCH(CH3)(CH3)], 27.0 [q, SiC(Ph)2A(CH3)3], 30.4 (d, C-14), 32.3 (t, C-
15), 36.4 (d, C-10), 37.6 (t, C-13), 56.5 (q, C-6-CH3), 57.5 (q, C-12-CH3),
60.5 (q, CarOCH3), 68.2 (t, C-1), 70.0 [d, OCH(CH3)2], 71.8 [d, OCH-
(CH3)2], 74.2 (t, OCH2Ph), 80.2 (d, C-6), 82.1 (d, C-12), 82.7 (d, C-11),
107.6 (d, C-4’), 114.9 (d, C-5’), 117.4 (d, C-3), 124.5 (d, C-5), 125.4 (s, C-
1’), 127.6 (d, OCH2Ph), 127.9 [d, Si(Ph)2C(CH3)3], 128.0 (d, OCH2Ph),
128.4 (d, OCH2Ph), 129.8 (d, C-4), 129.9 [d, Si(Ph)2C(CH3)3], 133.4 [s,
(CH3)3], 135.6 [d,
(CH3)3], 139.0 (s, OCH2Ph), 141.6 (s,
ACHTUGNTRENUN(GN CH3)3], 7.67–7.69 ppm [m,
AHCTUNGTRENNUNG
A
E
N
ACHTUNGTRENNUNG
A
R
CHTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
Si(Ph)2C
Si(Ph)2C
C
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
9H, Si(Ph)2C
3.37 (s, 3H, COCH3), 4.17 (s, 2H, H-9), 4.77 (q, 3J=6.9 Hz, 1H, H-2),
ACHTUNGTRENNUNG
(CH3)3], 1.57 (d, 3J=6.9 Hz, 3H, H-1), 1.72 (s, 3H, CCH3),
C-2), 144.4 (s, C-3’), 146.1 (d, C-9), 150.3 (s, C-2’), 151.2 (s, C-6’),
198.7 ppm (s, C-7); IR (ATR): n˜ =2971 (w, CH), 2929 (w, CH), 2856 (w,
CH), 1663 (m, C=O), 1589 (w), 1476 (m, C=C), 1428 (m, C=C), 1371
(m), 1248 (s), 1107 (CO), 701 cmꢀ1 (s); MS (ESI): m/z: 941 [M+Na+],
887 [MꢀCH3O+]; HRMS (ESI): m/z: calcd for C57H78O8NaSi: 941.5358,
found: 941.5357 [M+Na+].
3
3
5.00 (d, J=9.2 Hz, 1H, H-4), 5.29 (virt. t, J ffi 9.5 Hz, 1H, H-5), 6.65 (d,
3J=12.5 Hz, 1H, H-7), 6.86–6.74 (m, 1H, H-6), 7.37–7.43 [m, 6H,
Si(Ph)2C
13C NMR (90.6 MHz, CDCl3): d=14.0 (q, CCH3), 19.5 [s, Si(Ph)2C-
(CH3)3], 20.5 (q, C-1), 27.0 [q, SiC(CH3)3], 54.5 (d, C-2), 57.3 (q,
COCH3), 67.9 (t, C-9), 79.9 (d, C-4), 116.8 (d, C-7), 122.5 (d, C-5), 127.9
[d, Si(Ph)2C(CH3)3], 129.9 [d, Si(Ph)2C(CH3)3], 131.3 (d, C-6), 133.4 [s,
Si(Ph)2C(CH3)3], 133.5 [s, Si(Ph)2C(CH3)3], 135.6 [d, Si(Ph)2C(CH3)3],
135.6 [d, Si(Ph)2C(CH3)3], 142.2 (s, C-8), 202.5 ppm (s, C-3); IR (ATR):
ACHTUNGTRENNUNG AHCTUNGTREN(NUGN CH3)3];
(CH3)3], 7.68 ppm [d, 3J=6.3 Hz, 4H, Si(Ph)2C
U
ACHTUNGTRENNUNG
N
ACHTUNGTRENNUNG
G
A
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
Acknowledgements
n˜ =2960 (w, CH), 2929 (w, CH), 2856 (w, CH), 1732 (w, C=O), 1428 (m,
C=C), 1105 (s, CO), 822 (m), 739 (m), 700 (s) cmꢀ1; MS (ESI): m/z: 435
[MꢀCl+], 181 [C26H33O3Si+]; HRMS (ESI): m/z: calcd for C27H35O3Si:
435.2350, found: 435.2353 [MꢀCl+].
(2E,4Z,6S,8E,10S,11R,12S,14R)-11-(Benzyloxy)-1-(tert-butyldiphenyl-
silyloxy)-15-(3’,6’-diisopropoxy-2’-methoxyphenyl)-6,12-dimethoxy-
2,8,10,14-tetramethyl-pentadeca-2,4,8-trien-7-one (17): A solution of a-
This project was supported by the Deutsche Forschungsgemeinschaft (Ba
1372-8). Dr. Daniela Rosenbeiger is acknowledged for preliminary stud-
ies towards the synthesis and reductive coupling of compound 15. We
thank Olaf Ackermann for excellent technical assistance throughout the
project and Burghard Cordes for recording the HPLC/MS spectra. T.H.
is the recipient of a Novartis graduate fellowship in organic chemistry.
T.N. wishes to thank the Studienstiftung des Deutschen Volkes (PhD
scholarship) and the TUM graduate school for support.
chloroketone 20l (51.8 mg, 0.11 mmol) and aldehyde
3 (35 mg,
0.07 mmol) in THF (4.0 mL) was cooled to ꢀ788C before dropwise
adding SmI2 (80.8 mg, 0.2 mL, 0.20 mmol, 0.1m solution in THF) within
15 min. To the reaction mixture was added another portion of a-chloro-
ketone 20l (51.8 mg, 0.11 mmol) in THF (2.0 mL) followed by SmI2
(80.8 mg, 0.20 mL, 0.20 mmol, 0.1m solution in THF) within 15 min.
After one hour at ꢀ788C saturated aq. NaHCO3 (10 mL) was added,
before warming up to room temperature. The aqueous layer was extract-
ed with diethyl ether (3ꢄ2 mL). The combined organic layers were
washed with brine (2 mL), dried with Na2SO4, filtered, and concentrated
in vacuo. The crude product was purifed by flash chromatography (pen-
tane/diethyl ether 4:1 ! 1:1) to yield a mixture of diastereomers of the
desired b-hydroxy ketone 16 (63.0 mg, 67.0 mmol, 90%) as a colorless oil.
[1] a) L. Whitesell, E. G. Mimnaugh, B. De Costa, C. E. Myers, L. M.
Neckers, Proc. Natl. Acad. Sci. USA 1994, 91, 8324–8328; b) C. E.
Stebbins, A. A. Russo, C. Schneider, N. Rosen, F. U. Hartl, N. P.
Pavletich, Cell 1997, 89, 239–250; c) S. M. Roe, C. Prodromou, R.
OꢃBrien, J. E. Ladbury, P. W. Piper, L. H. Pearl, J. Med. Chem. 1999,
42, 260–266.
[2] Selected reviews: a) Y. Fukuyo, C. R. Hunt, N. Horikoshi, Cancer
Lett. 2010, 290, 24–35; b) Y. L. Janin, Drug Discovery Today 2010,
15, 342–353; c) A. G. Ortiz, J. M. Salcedo, Clin. Transl. Oncol. 2010,
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

T. Bach et al.
12, 166–173; d) S. Sharp, P. Workman, Adv. Cancer Res. 2006, 95,
323–348; e) E. McDonald, P. Workman, K. Jones, Curr. Top. Med.
Chem. 2006, 6, 1091–1107; f) L. Whitesell, S. L. Lindquist, Nat. Rev.
Cancer 2005, 5, 761–772; g) L. Neckers, T. W. Schulte, E. Mim-
naugh, Invest. New Drugs 1999, 17, 361–373.
[14] a) R. M. Moslin, T. F. Jamison, J. Am. Chem. Soc. 2006, 128, 15106–
15107; b) R. M. Moslin, T. F. Jamison, J. Org. Chem. 2007, 72, 9736–
9745; c) B. A. Sparling, R. M. Moslin, T. F. Jamison, Org. Lett. 2008,
10, 1291–1294.
[15] C. R. Schmid, J. D. Bryant, M. Dowlatzedah, J. L. Phillips, D. E.
Prather, R. D. Schantz, N. L. Sear, C. S. Vianco, J. Org. Chem. 1991,
56, 4056–4058.
[16] a) V. VanRheenen, R. C. Kelly, D. Y. Cha, Tetrahedron Lett. 1976,
17, 1973–1976; b) B. M. Trost, H. C. Shen, J. P. Surivet, J. Am.
Chem. Soc. 2004, 126, 12565–12579.
[3] Completed total sytheses: a) M. B. Andrus, E. L. Meredith, B. L.
Simmons, B. B. V. Soma Sekhar, E. J. Hicken, Org. Lett. 2002, 4,
3549–3552; M. B. Andrus, E. L. Meredith, E. J. Hicken, B. L. Sim-
mons, R. R. Glancey, W. Ma, J. Org. Chem. 2003, 68, 8162–8169;
b) H.-L. Qin, J. S. Panek, Org. Lett. 2008, 10, 2477–2479.
[4] De novo syntheses of analogues: a) T. Bach, A. Lemarchand, Synlett
2002, 1302–1304; b) C. J. Davis, C. J. Moody, Synlett 2002, 1874–
1876; c) A. Lemarchand, T. Bach, Tetrahedron 2004, 60, 9659–9673;
d) A. Lemarchand, T. Bach, Synthesis 2005, 1977–1990; e) C. S. P.
McErlean, N. Proisy, C. J. Davis, N. A. Boland, S. Y. Sharp, K.
Boxall, A. M. Z. Slawin, P. Workman, C. J. Moody, Org. Biomol.
Chem. 2007, 5, 531–546; f) I. E. Wrona, A. Gozman, T. Taldone, G.
Chiosis, J. S. Panek, J. Org. Chem. 2010, 75, 2820–2835.
[17] L. Malaprade, Bull. Soc. Chim. Fr. 1934, 883.
[18] a) A. K. Ghosh, S. Leshchenko, M. Noetzel, J. Org. Chem. 2004, 69,
7822–7829; b) S. Takano, A. Kurotaki, M. Takahashi, K. Ogasawara,
Synthesis 1986, 403–406.
[19] a) O. Isler, H. Gutmann, M. Montavon, R. Ruegg, G. Ryser, P.
Zeller, Helv. Chim. Acta 1957, 40, 1242–1249; b) H. O. House, G. H.
Rasmusson, J. Org. Chem. 1961, 26, 4278–4281; c) Y. Elemes, C. S.
Foote, J. Am. Chem. Soc. 1992, 114, 6044–6050.
[5] a) C. De Boer, P. A. Meulman, R. J. Wnuk, D. H. Peterson, J. Anti-
biot. 1970, 23, 442–447; b) K. Sasaki, K. L. Rinehardt, G. Slomp,
M. F. Grostic, E. C. Olsen, J. Am. Chem. Soc. 1970, 92, 7491–7593.
[6] a) S. Omura, Y. Iwai, Y. Takahashi, N. Sadakane, A. Nakagawa, H.
Oiwa, Y. Hasegawa, T. Ikai, J. Antibiot. 1979, 32, 255–261; b) A. Na-
kagawa, N. Sadakane, S. Omura, Nippon Kagaku Kaishi 1981, 892–
894; c) Y. Iwai, A. Nakagawa, N. Sadakane, S. Omura, H. Oiwa, S.
Matsumoto, M. Tagahashi, T. Ikai, Y. Ochiai, J. Antibiot. 1980, 33,
1114–1119.
[20] a) C. Hartmann, V. Meyer, Chem. Ber. 1893, 26, 1727–1732; b) M.
Frigerio, M. Santagostino, S. Sputore, J. Org. Chem. 1999, 64, 4537–
4538; c) J. D. More, N. S. Finney, Org. Lett. 2002, 4, 3001–3003;
d) review: A. Duschek, S. F. Kirsch, Angew. Chem. 2011, 123, 1562–
1590; Angew. Chem. Int. Ed. 2011, 50, 1524–1552.
[21] a) B. D. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155–4156;
b) R. E. Ireland, L. Lin, J. Org. Chem. 1993, 58, 2899.
[22] a) H.-J. Bissinger, H. Detert, H. Meier, Liebigs Ann. Chem. 1988,
221–224; b) H.-Y. Lꢀ, Synlett 2009, 330–331.
[7] M. Muroi, M. Izawa, Y. Kosai, M. Asai, J. Antibiot. 1980, 33, 205–
212.
[23] Reviews: a) R. Ocampo, W. R. Dolbier, Tetrahedron 2004, 60, 9325–
9373; b) A. Fꢀrstner, Synthesis 1989, 571–590.
[8] a) R. C. Schnur, M. L. Corman, R. J. Gallaschun, B. A. Cooper,
M. F. Dee, J. L. Doty, M. L. Muzzi, J. D. Moyer, C. I. DiOrio, E. G.
Barbacci, P. E. Miller, A. T. OꢃBrien, M. J. Morin, B. A. Foster,
V. A. Pollack, D. M. Savage, D. E. Sloan, L. R. Pustilnik, M. P.
Moyer, J. Med. Chem. 1995, 38, 3806–3812I; b) T. W. Schulte, L.
Neckers, Cancer Chemother. Pharmacol. 1998, 42, 273–279; c) Hol-
stein, D. Robertson, F. Di Stefano, P. Workman, P. A. Clarke, Cancer
Res. 2001, 61, 4003–4009; d) J. R. Sydor, E. Normant, C. S. Pien,
J. R. Porter, J. Ge, L. Grenier, R. H. Pak, J. A. Ali, M. S. Dambski,
J. Hudak, J. Patterson, C. Penders, M. Pink, M. A. Read, J. Sang, C.
Woodward, Y. Zhang, D. S. Grayzel, J. Wright, J. A. Barrett, V. J.
Palombella, J. Adams, J. K. Tong, Proc. Natl. Acad. Sci. USA 2006,
103, 17408–17413.
[24] T. Imamoto, M. Ono, Chem. Lett. 1987, 16, 501–502.
[25] R. J. Arhart, J. Martin, J. Am. Chem. Soc. 1972, 94, 5003–5010.
[26] a) T. Satoh, K. Takano, Tetrahedron 1996, 52, 2349–2358; b) R. W.
Hoffmann, Chem. Soc. Rev. 2003, 32, 225–230; c) T. Satoh, Chem.
Soc. Rev. 2007, 36, 1561–1572.
[27] Y. H. Kim, H. H. Shin, Y. J. Park, Synthesis 1993, 209–210.
[28] Review: A. Mengel, O. Reiser, Chem. Rev. 1999, 99, 1191–1224.
[29] J.-L. Luche, L. Rodriguez-Hahn, P. Crabbe, J. Chem. Soc. Chem.
Commun. 1978, 601–602.
[30] a) T. Oishi, T. Nakata, Acc. Chem. Res. 1984, 17, 338–344; b) D. A.
Evans, P. Nagorny, K. J. McRae, D. J. Reynolds, L.-S. Sonntag, F.
Vounatsos, R. Xu, Angew. Chem. 2007, 119, 543–546; Angew.
Chem. Int. Ed. 2007, 46, 537–540.
[9] Review: I. E. Wrona, V. Agouridas, J. S. Panek, Compt. Rend.
Chimie 2008, 11, 1483–1522.
[31] a) A. Hirao, S. Itsuno, S. Nakahama, N. Yamazaki, J. Chem. Soc.
Chem. Commun. 1981, 315–317; b) E. J. Corey, R. K. Bakshi, S. Shi-
bata, J. Am. Chem. Soc. 1987, 109, 5551–5553; c) S. F. Sabes, R. A.
Urbanek, C. J. Forsyth, J. Am. Chem. Soc. 1998, 120, 2534–2542.
[32] Formation of the Mosher ester (J. A. Dale, H. S. Mosher, J. Am.
Chem. Soc. 1973, 95, 512–519) from alcohol 22a was not successful.
[33] J. D. Frein, R. E. Taylor, D. L. Sackett, Org. Lett. 2009, 11, 3186–
3189.
[34] a) I. E. Wrona, J. T. Lowe, T. J. Turbyville, T. R. Johnson, J. Beignet,
J. A. Beutler, J. S. Panek, J. Org. Chem. 2009, 74, 1897–1916; b) X.
Liao, Y. Wu, J. K. De Brabander, Angew. Chem. 2003, 115, 1686–
1690; Angew. Chem. Int. Ed. 2003, 42, 1648–1652.
[10] Examples: a) R. Baker, J. L. Castro, J. Chem. Soc. Chem. Commun.
1989, 378–381; b) M. Nakata, T. Osumi, A. Ueno, T. Kimura, T.
Tamai, K. Tatsuta, Bull. Chem. Soc. Jpn. 1992, 65, 2974–2991;
c) D. A. Evans, S. J. Miller, M. D. Ennis, P. L. Ornstein, J. Org.
Chem. 1992, 57, 1067–1069; d) J. S. Panek, F. Xu, A. C. Rondꢅn, J.
Am. Chem. Soc. 1998, 120, 4113–4122; e) K. D. Carter, J. S. Panek,
Org. Lett. 2004, 6, 55–57; f) S. Canova, V. Bellosta, A. Bigot, P.
Mailliet, S. Mignani, J. Cossy, Org. Lett. 2007, 9, 145–148; g) K. Be-
lardi, G. C. Micalizio, Angew. Chem. 2008, 120, 4069–4072; Angew.
Chem. Int. Ed. 2008, 47, 4005–4008.
[11] Examples: a) J. P. Gillepsie, L. G. Amoros, F. R. Stermitz, J. Org.
Chem. 1974, 39, 3239–3241; b) T. Sala, M. V. Sargent, J. Chem. Soc.
Perkin Trans. 1 1979, 2593–2598; c) J. Zhu, R. Beugelmans, S. Bour-
det, J. Chastenet, G. Roussi, J. Org. Chem. 1995, 60, 6389–6396;
d) M. G. Banwell, B. L. Flynn, S. G. Stewart, J. Org. Chem. 1998, 63,
9139–9144; e) S. Boisnard, A.-C. Carbonnelle, J. Zhu, Org. Lett.
2001, 3, 2061–2064; f) K. Hasse, A. C. Willis, M. G. Banwell, Eur. J.
Org. Chem. 2011, 88–99; g) B. L. Flynn, M. Banwell, Heterocycles
2012, 84, 1141–1170.
[12] a) T. Horneff, E. Herdtweck, S. Randoll, T. Bach, Bioorg. Med.
Chem. 2006, 14, 6223–6234; b) T. Horneff, T. Bach, Synlett 2008,
2969–2972.
[13] W. R. Roush, K. Ando, D. B. Powers, A. D. Palkowitz, R. L. Halter-
man, J. Am. Chem. Soc. 1990, 112, 6339–6348.
[35] E. J. Corey, C. J. Helal, Angew. Chem. 1998, 110, 2092–2118;
Angew. Chem. Int. Ed. 1998, 37, 1986–2012.
[36] For a seminal review on double stereodifferentiation, see: S. Masa-
mune, W. Choy, J. S. Petersen, L. R. Sita, Angew. Chem. 1985, 97, 1–
31; Angew. Chem. Int. Ed. Engl. 1985, 24, 1–30.
[37] a) A. N. Cuzzupe, C. A. Hutton, M. J. Lilly, R. K. Mann, K. J.
McRae, S. C. Zammit, M. A. Rizzacasa, J. Org. Chem. 2001, 66,
2382–2393; b) S. Kirsch, T. Bach, Angew. Chem. 2003, 115, 4833–
4835; Angew. Chem. Int. Ed. 2003, 42, 4685–4687; c) P. Ruiz, J.
Murga, M. Carda, J. A. Marco, J. Org. Chem. 2005, 70, 713–716;
d) S. F. Kirsch, T. Bach, Chem. Eur. J. 2005, 11, 7007–7023.
[38] D. L. Boger, R. M. Garbaccio, J. Org. Chem. 1999, 64, 8350–8362.
[39] a) B. S. Bal, W. E. Childers Jr., H. W. Pinnick, Tetrahedron 1981, 37,
2091–2096; b) L. S. M. Wong, M. S. Sherburn, Org. Lett. 2003, 5,
&
10
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Natural Products
FULL PAPER
3603–3606; c) short review: A. Raach, O. Reiser, J. Prakt. Chem.
2000, 342, 605–608.
[40] a) J. M. Lalancette, A. Freche, J. R. Brindle, M. Laberte, Synthesis
1972, 527–532; b) J. M. Lalancette, A. Freche, R. Monteux, Can. J.
Chem. 1968, 46, 2754–2757.
[43] For further examples, see: a) S. Grimme, J. Harren, A. Sobanski, F.
Vçgtle, Eur. J. Org. Chem. 1998, 1491–1509; b) Modern Cyclophane
Chemistry (Eds.: R. Gleiter, H. Hopf), Wiley-VCH, 2004, Wein-
heim.
ˇ
´
[44] P. Kocovsky, Tetrahedron Lett. 1986, 27, 5521–5524.
[45] B. Loev, M. F. Kormendy, J. Org. Chem. 1963, 28, 3421–3426.
[41] A. Presser, A. Hꢀfner, Monatsh. Chem. 2004, 135, 1015–1022.
[42] J. Diago-Meseguer, A. L. Palomo-Coll, J. R. Fernandez-Lizarbe, A.
Zugaza-Bilbao, Synthesis 1980, 547–551.
Received: May 8, 2012
Published online: && &&, 0000
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!

T. Bach et al.
Convergency in the synthesis of the
Total Synthesis
geldanamycin skeleton has been ach-
ieved by employing a magnesium car-
benoid as two-carbon building block,
connecting two aldehyde fragments in
an efficient and high yielding manner.
The preparation of a-chloroethyl
ketones from aldehydes has been
investigated (11 examples, 64–95%
yield) and the title compounds have
been prepared from advanced precur-
sor 1.
T. Hampel, T. Neubauer,
T. van Leeuwen, T. Bach* . . &&&& — &&&&
Stereoselective Preparation of (E)-
Configured 1,2-Disubstituted Propenes
from two Aldehydes by a Two-Carbon
Stitching Strategy: Convergent
Synthesis of 18,21-Diisopropyl-Gelda-
namycin Hydroquinoneand Its C7
Epimer
&
12
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!