Modular total synthesis of archazolid A and B

Article abstract of DOI:10.1021/jo901565n

(Chemical Equation Presented) A modular total synthesis of the potent V-ATPase inhibitors archazolid A and B is reported. The convergent preparation was accomplished by late-stage diversification of joint intermediates. Key synthetic steps involve asymmetric boron-mediated aldol reactions, two consecutive Still-Gennari olefinations to set the characteristic (Z,Z)-diene system, a Brown crotyboration, and a diastereo-selective aldol condensation of highly elaborate intermediates. For macrocyclization, both an HWE reaction and a Heck coupling were successfully employed to close the 24-membered macrolactone. During the synthetic campaign, a generally useful protocol for an E-selective Heck reaction of nonactivated alkenes and a method for the direct nucleophilic displacement of the Abiko-Masamune auxiliary with sterically hindered nucleophiles were developed. The expedient and flexible strategy will enable further SAR studies of the archazolids and more detailed evaluations of target-inhibitor interactions. 2009 American Chemical Society.

Full text of DOI:10.1021/jo901565n

pubs.acs.org/joc
Modular Total Synthesis of Archazolid A and B
Dirk Menche,*,†,‡ Jorma Hassfeld,‡,§ Jun Li,‡, Kerstin Mayer,† and Sven Rudolph‡,^
†
€
Institut fu€r Organische Chemie, Ruprecht-Karls-Universitat Heidelberg, Im Neuenheimer Feld 270,
D-69120 Heidelberg, and ‡Helmholtz-Zentrum fu€r Infektionsforschung, Medizinische Chemie,
Inhoffenstrasse 7, D-38124 Braunschweig, Germany. §Present address: Bayer Schering Pharma AG,
Berlin. Present address: Institut fu€r Pharmazeutische Wissenschaften, ETH Zu€rich. ^Present address:
MOLISA GmbH, Magdeburg.
dirk.menche@oci.uni-heidelberg.de
Received July 22, 2009
A modular total synthesis of the potent V-ATPase inhibitors archazolid A and B is reported. The
convergent preparation was accomplished by late-stage diversification of joint intermediates. Key
synthetic steps involve asymmetric boron-mediated aldol reactions, two consecutive Still-Gennari
olefinations to set the characteristic (Z,Z)-diene system, a Brown crotyboration, and a diastereo-
selective aldol condensation of highly elaborate intermediates. For macrocyclization, both an HWE
reaction and a Heck coupling were successfully employed to close the 24-membered macrolactone.
During the synthetic campaign, a generally useful protocol for an E-selective Heck reaction of
nonactivated alkenes and a method for the direct nucleophilic displacement of the Abi-
ko-Masamune auxiliary with sterically hindered nucleophiles were developed. The expedient and
flexible strategy will enable further SAR studies of the archazolids and more detailed evaluations of
target-inhibitor interactions.
Introduction
molecular level, they constitute powerful and selective inhi-
bitors of vacuolar type ATPases (V-ATPases),3 heteromulti-
meric, proton translocating proteins that are localized in a
multitude of eukaryotic membranes, where they energize
many different transport processes.4 As a malfunction of
these enzymes is associated with various diseases such as
cancer5 and osteoporosis, the development and molecular
understanding of novel inhibitors present important research
goals.4
Myxobacterial fermentation broths of Archangium ge-
phyra and Cystobacter violaceus are the natural sources of
the archazolids (Figure 1), structurally complex polyketide
macrolactones, which were originally reported by the group
of Hofle and Reichenbach.1,2 They constitute highly potent
€
antiproliferative agents which inhibit the growth of various
cancer cell lines in subnanomolar concentrations.1 On a
*To whom correspondence should be addressed. Phone: þ49 6221
546207. Fax: þ þ49 6221 544205.
(3) Huss, M.; Sasse, F.; Kunze, B.; Jansen, R.; Steinmetz, H.; Ingenhorst,
G.; Zeeck, A.; Wieczorek, H. BMC Biochem. 2005, 6, 13.
(4) For reviews, see: (a) Huss, M.; Wieczorek, H. J. Exp. Biol. 2001, 204,
2597. (b) Beyenbach, K. W.; Wieczorek, H. J. Exp. Biol. 2006, 209, 577.
(c) Huss, M.; Wieczorek, H. J. Exp. Biol. 2009, 212, 341.
€
(1) (a) Hofle, G.; Reichenbach, H.; Sasse, F.; Steinmetz, H. German Patent DE
€
41 42 951 C1, 1993. (b) Sasse, F.; Steinmetz, H.; Hofle, G.; Reichenbach, H.
J. Antibiot. 2003, 56, 520.
(2) For a recent review on polyketides from myxobacteria, see: Menche,
D. Nat. Prod. Rep. 2008, 25, 905.
(5) Lebreton, S.; Jaunbergs, J.; Roth, M. G.; Ferguson, D. A.;
De Brabander, J. K. Bioorg. Med. Chem. Lett. 2008, 18, 5879.
7220 J. Org. Chem. 2009, 74, 7220–7229
Published on Web 09/09/2009
DOI: 10.1021/jo901565n
r
2009 American Chemical Society

Menche et al.
JOCFeatured Article
SCHEME 1
FIGURE 1. Natural archazolids: polyketide macrolactones of
myxobacterial origin.
The unique three-dimensional architecture of the archa-
zolids is characterized by a 24-membered polyunsaturated
macrolactone ring together with a thiazole side chain termi-
nating in a methyl carbamate. In total, they contain eight
stereogenic centers. Archazolids A (1) and B (2), the most
potent metabolites, differ only in the methylation pattern at
C-2, while the glucosides 36 and 47 are less active by a factor
of ∼1000. Originally reported as planar structures, their
absolute and relative stereochemistry has been determined
in our group by a combination of J-based configurational
analysis with molecular modeling and chemical derivatiza-
tion.8 Recently, this assignment was independently validated
by an innovative method based on residual dipolar cou-
plings.9 SAR data suggest the hydroxyl at C-7 to be part of
the pharmacophoric region. In contrast, removal of the
carbamate leads only to slight loss of activity, suggesting
that the full biological potential of the archazolids can be
retained while simplifying the structural complexity.10
The important biological properties of these macrolide
antibiotics together with their natural scarcity and unique
and challenging structure render these metabolites attractive
synthetic targets to support further biological evaluation but
also to enable more elaborate structure-activity relationship
(SAR) studies. Recently, the first total synthesis of archazolid A
was accomplished in our group,11 which also unambiguously
confirmed our stereochemical assignment. Shortly thereafter,
an alternative route was reported to access archazolid B by
the Trauner group.12,13 Furthermore, an innovative syn-
thetic approach to the characteristic trisubstituted Z-alkenes
of the archazolids has been disclosed.14 Herein, we report in
full detail our synthetic strategies toward these potent
macrolide antibiotics, which culminated in a joint total
synthesis of archazolids A and B.
Results and Discussion
As shown in Scheme 1, our original retrosynthetic route
was based on three main building blocks of similar comple-
xity: a northwestern, an eastern, and a southern fragment
(5, 6, and 7) in combination with phosphonates 8 and 9. The
delicate 9Z,11Z,13E triene was planned to originate from a
Horner-Wadsworth-Emmons reaction (HWE reaction)
between phosphonate 5 and dienone 6, while a more challen-
ging Heck reaction between 5 and 7 was envisioned to set the
18E,20E-diene. The 2,3-double bond, in turn, should be
generated by another phosphonate coupling. In principle,
all three reactions should be suitable for macrocyclization as
an alternative for a more conventional Yamaguchi reaction
for ring closure. Importantly, deliberate choice of phospho-
nates 8 or 9 should enable access to both archazolid A and B
by late-stage diversification. Notably, this convergent and
modular approach is flexible and should be readily amenable
to analogue synthesis.
(6) Menche, D.; Hassfeld, J.; Steinmetz, H.; Huss, M.; Wieczorek, H.;
Sasse, F. Eur. J. Org. Chem. 2007, 1196.
(7) Menche, D.; Hassfeld, J.; Steinmetz, H; Huss, M.; Wieczorek, H.;
Sasse, S. J. Antibiot. 2007, 60, 328.
Synthesis of building block 6 was initiated by olefination
of ketone 11, readily available from commercial alcohol 10
ꢀ
(8) Hassfeld, J.; Fares, C.; Steinmetz, H.; Carlomagno, T.; Menche, D.
Org. Lett. 2006, 8, 4751.
ꢀ
(9) Fares, C.; Hassfeld, J.; Menche, D.; Carlomagno, T. Angew. Chem.,
Int. Ed. 2008, 47, 3722.
(10) Menche, D.; Hassfeld, J.; Sasse, F.; Huss, M.; Wieczorek, H. Bioorg.
Med. Chem. Lett. 2007, 17, 1732.
(11) Menche, D.; Hassfeld, J.; Li, J.; Rudolph, S. J. Am. Chem. Soc. 2007,
129, 6100.
(12) Roethle, P. A.; Chen, I. T.; Trauner, D. J. Am. Chem. Soc. 2007, 129,
8960.
(13) For a review on archazolid syntheses, see: von Zezschwitz, P. Nachr.
Chem. 2008, 56, 535.
(14) Huang, Z. H.; Negishi, E. J. Am. Chem. Soc. 2007, 129, 14788.
J. Org. Chem. Vol. 74, No. 19, 2009 7221

Menche et al.
JOCFeatured Article
SCHEME 2
SCHEME 3
(Scheme 2). Initial attempts using conventional Wittig ylide
12, however, soon revealed this reagent to be too unreactive
for the desired C2-homologation. Even at elevated tempera-
tures and extended reaction times, only small degrees of
conversion could be observed, regardless of the presence of a
hydroxyl-protective group. In contrast, use of Horner-
Wadsworth-Emmons reagent 13 was more efficient. Eva-
luation of reaction parameters (see the table in Scheme 2)
revealed the methyl reagent to be more effective as compared
to the ethyl analogue 14 and use of potassium as counterion
to be superior to the use of sodium. Optimal conditions
involved a slight excess of KHMDS and phosphonate 13
(1.25 equiv each). Running the reaction in THF at room
temperature overnight resulted in the formation of desired
enoate 15 with preparatively useful yields (62%), albeit only
moderate E/Z selectivity (2:1). Nevertheless, this route was
selected to prepare multigram quantities of 15 due to the
robustness, low cost of starting materials, and good scala-
bility of the process.
The undesired Z-isomer could be readily removed by
chromatography after reduction to the corresponding allylic
alcohol (Scheme 3). Subsequent allylic oxidation (MnO2)
gave diastereomerically pure enal 16 in essentially quantita-
tive yield. Our successful strategy for converting 16 to
building block 6 was based on a boron-mediated Paterson15
anti-aldol reaction with lactate-derived ketone 17 to give the
desired aldol product 18 with excellent selectivity and yield
(99%, dr > 20:1). After conversion of 18 in three steps to
aldehyde 19 (TBSOTf, LiBH4, and NaIO4, 85% yield), two
consecutive Still-Gennari olefinations16 using phosphonate
20 with aldehyde 19 and subsequently with enal 22, derived
from ester 21, proceeded with excellent selectivity to give the
desired (Z,Z)-dienone 23 in high yield and Z-selectivity of
the newly generated double bonds (88%/87%, dr > 20:1).
This route proved to be superior in terms of overall chemical
yield, scalability, and cost to a likewise tested alternative
sequence of first introducing the required anti-propionate by
a Brown crotylation of aldehyde 16, esterification of the
derived secondary alcohol with methacrylic acid, and ring-
closing metathesis reaction of the derived ester (not shown).17
Overall, several grams of building block 23 could be readily
obtained from 15 in 10 steps and 56% overall yield. Reduc-
tion and allylic oxidation proceeded smoothly to give 6 in a
straightforward fashion.
For the synthesis of side-chain fragment 7, natural L-leu-
cine (24) was chosen as a cheap starting material, already
containing the characteristic aliphatic isopentyl residue of
the archazolids with the desired configuration. Following a
known procedure,18 nucleophilic substitution of the derived
diazo acid (NaNO2/H2SO4) gave the corresponding R-hy-
droxy acid 25 with overall retention of configuration (68%).
The corresponding amide 26 was best prepared by first
activation with thionyl chloride, conversion to the respective
methyl ester, and subsequent treatment with ammonia in
methanol (68%). TBS protection of hydroxy amide 26 as
silyl ether 28 proceeded smoothly under conventional con-
ditions in high yields (92%). Alternatively, amide 28 may
also be obtained from 25 by a shorter route involving
protecting the hydroxyl as TBS ether, IBC activation of the
derived acid as the mixed anhydride, and in situ substitution
with NH3. However, this sequence proved less scalable,
proceeded with slightly lower yields, and consequently, was
not further pursued. Subsequently, amide 28 was treated
with the Lawesson reagent19 to give thioamide 29 in high
yields (93%), which was then condensed with ethyl bromo-
pyruvate to form thiazole 30. At the stage of the derived free
alcohol 31, however, Mosher ester analysis revealed variable
degrees of isomerization, which presumably occurs during
thiazole formation, in agreement with previous findings.20,21
Loss of stereochemical purity at an earlier stage of the
synthesis can be excluded as the optical properties of 28
prepared by the two routes discussed above were identical.
At this stage, the enantiomeric purity of 31 could be readily
(18) (a) Barrow, R. A.; Hemscheidt, T.; Liang, J.; Paik, S.; Moore, R. E;
Tius, M. A. J. Am. Chem. Soc. 1995, 117, 2479. (b) Bitan, G.; Gilon, C.
Tetrahedron 1995, 51, 10513.
(15) For a review on asymmetric boron-mediated aldol reactions, see:
Cowden, C. J.; Paterson, I. Org. React. 1997, 51, 1.
(19) Thomsen, I.; Clausen, K.; Scheibye, S.; Lawesson, S.-O. Organic
Syntheses; Wiley: New York, 1990; Vol. 7, p 372.
(16) Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4405.
(17) For a similar sequence, see: Cossy, J.; Bauer, D.; Bellosta, V.
Tetrahedron 2002, 58, 5909.
(20) Riedrich, M.; Harkal, S.; Arndt, H.-D. Angew. Chem., Int. Ed. 2007,
46, 2701.
(21) Schmidt, U.; Gleich, P.; Griesser, H.; Utz, R. Synthesis 1986, 12, 992.
7222 J. Org. Chem. Vol. 74, No. 19, 2009

Menche et al.
JOCFeatured Article
SCHEME 4
SCHEME 5
Brown’s methodology25 gave the desired building block 7 in
high stereoselectivity (dr > 20:1) and preparatively useful
yields (65%). Overall, the route toward 7was readily adaptable
for a large-scale approach, and several grams of the fragment
could be obtained.
As shown in Scheme 6, our synthesis of the northwestern
fragment 5 was based on known vinylic iodide 39. A reported
route from propargylic alcohol 37 by zirconation-methy-
lation-iodination,26 however, gave only low yields in our
hands (22%). Also a carbocupration-iodination sequence
remedied by a two-step procedure, involving first Dess-
Martin oxidation22 and stereoselective reduction (er>20:1)
of the derived ketone 32 using the CBS method (Scheme 4).23
As shown in Scheme 5, enantiopure alcohol 31 was homo-
logated to carbamate 35 by CDI-treatment and trapping the
intermediate imidazole carboxylate with methylamine. At this
stage, also an alternative strategy as compared to the multistep
approach discussed so far (Scheme 4) was evaluated. Relying
on a method of Hoppe,24 this involved use of chiral amines for
stereoselective deprotonation of 34, readily available from 33
using our previously established sequence. Thiazole 33 was
readily available by condensation of hydroxyacetonitrile, hy-
drogen sulfide, and ethyl bromopyuvate (see the Supporting
Information). Accordingly, carbamate 34 was first deproto-
nated at -40 ꢀC with s-BuLi in the presence of (-)-sparteine,
and subsequently, the reaction mixture was treated with iso-
butyl bromide. However, only complex product mixtures were
obtained, with no formation of trace amounts of desired 35,
even after modification of reaction parameters. Presumably,
the carbamate was not sufficiently sterically hindered to over-
come direct carbonyl addition of BuLi. Accordingly, this
concept was not further pursued. For completing the synthesis
of building block 36, ester 35 was first directly reduced to
aldehyde 36 with DIBALH at low temperatures in CH2Cl2,
which proceeded with consistently good yields (75%) also on
large scale-up, depending on the use of K-Na tartrate solution
during workup. Subsequent asymmetric crotylation using
involving MeMgBr/CuBr SMe2 and I2 (not shown) failed to
3
give the desired product.27 In contrast, a much more prac-
tical synthesis was based on diethyl 2-methylmalonate (38),
following a procedure by Baker and Castro.28 In the course
of this study, the original two-step reaction sequence was
optimized. This involved first treatment of 38 with iodoform
and sodium hydride in Et2O at reflux to give intermediate 40,
which was, however, not isolated but directly converted into
acid 41 under basic conditions. This optimized procedure
resulted in higher yields and was readily amenable to pro-
duction of multigram quantities of acid 41 (40 g). Reduction
to alcohol 39 (LiAlH4) and subsequent allylic oxidation
(MnO2) to aldehyde 42 proceeded smoothly. For the requi-
red anti-propionate homologation, an Abiko-Masamune al-
dol reaction29 with norephedrin-derived ethyl ester 44 to give
46 proved the method of choice in terms chemical yield and
stereoselectivity (96%, dr>20:1), as compared to a likewise
tested alternative of using Evans methodology30 (43 to 45).
(25) Brown, H. C.; Bhat, K. S.; Randad, R. S. J. Org. Chem. 1989, 54,
1570.
(26) White, J. D.; Blakemore, P. R.; Green, N. J.; Hauser, E. B.;
Holoboski, M. A.; Keown, L. E.; Nylund Kolz, C. S.; Phillips, B. W.
J. Org. Chem. 2002, 67, 7750.
(27) For an example, see: Layton, M. E.; Morales, C. A.; Shair, M. D.
J. Am. Chem. Soc. 2002, 124, 773.
(28) Baker, R.; Castro, J. L. J. Chem. Soc., Perkin Trans. 1 1990, 47–65.
(29) Inoue, T.; Liu, J.-F.; Buske, D. C.; Abiko, A. J. Org. Chem. 2002, 67,
5250.
(30) Evans, D. A.; Tedrow, J. S.; Shaw, J. T.; Downey, C. W. J. Am.
Chem. Soc. 2002, 124, 392.
(22) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155.
(23) Corey, E. J.; Helal, C. J. Angew. Chem., Int. Ed. 1998, 37, 1986.
(24) Hoppe, D.; Hintze, F.; Tebben, P. Angew. Chem., Int. Ed. 1990, 29,
1422.
J. Org. Chem. Vol. 74, No. 19, 2009 7223

Menche et al.
JOCFeatured Article
SCHEME 6
SCHEME 7
SCHEME 8
A major drawback of this Abiko-Masamune aldol route
was, however, the subsequent difficulties to cleave the nor-
ephedrin-derived auxiliary of 47 with nucleophiles other
than hydride. Apparently this presents a general limitation
of the method, which so far may only be circumvented by
resorting to a derived auxiliary.31 As shown in Scheme 7,
direct treatment with metalated phosphonate 48 resulted in
extensive decomposition, giving eliminated derivative 49 as
the main product (38%), and only traces of the desired
phosphonate 5 (9%) were found. Therefore, at this stage
we had to resort to a four-step sequence, relying on first
removing the auxiliary reductively with LiAlH4, Dess-
Martin oxidation of the resulting primary alcohol, addition
of lithiated phosphonate 48 to the resulting aldehyde, fol-
lowed by another Dess-Martin oxidation, affording the
desired phosphonate 5 in acceptable yields (44%) over four
steps.
To improve this result, efforts were directed to enable a
direct displacement of the Abiko-Masamune auxiliary. As
shown in Scheme 8, our concept to promote such a direct
substitution was based on ester activation by suitable addi-
tives. As schematically shown in 50, they are expected to
chelate the characteristic hydroxy ester, leading to an activa-
tion of the ester moiety and thereby promoting nucleophilic
attack. According to this concept, a variety of Lewis acids
were evaluated in the substitution of 46 with metalated 48
using mild reaction conditions (-78 to -20 ꢀC, THF). Best
results were obtained with i-PrMgCl among those screened.
Furthermore, it was found that the choice of base (KHMDS)
was critical in order to avoid hydrodehalogenation to 51.
Under optimized conditions, the desired phosphonate could
be obtained in high yields (80%) in a straightforward fash-
ion. Notably, the chiral auxiliary may be recovered after the
reaction by chromatography, which adds to the efficiency of
this process. Methylation of phosphonate 52 to give building
block 5 proceeded smoothly. Notably, this direct procedure
was superior as compared to the stepwise displacement
discussed above (Scheme 7), in terms of steps and overall
yields.
With efficient approaches to the three building blocks
established, our strategy for fragment union was initiated
by attempted HWE reactions of phosphonate 5 with alde-
hyde 6 (Scheme 9). However, to our disappointment, this
coupling proved extremely challenging, leading to extensive
decomposition. Even under very mild conditions according
to the Masamune-Roush protocol (LiCl, DBU, CH3CN, rt),32
no conversion to the desired trienone 54 was observed and
mainly isomerization of aldehyde 6 to 53 was found.
Accordingly, the original strategy had to be revised, and
an alternative approach was pursued. This involved union of
the two northern fragments by an aldol condensation.
Accordingly, Abiko-Masamune aldol product 46 was
transformed to the required methyl ketone 55. Initially, a
five-step procedure was used involving methylation of Abi-
ko-Masamune aldol product 46, reductive cleavage of the
chiral auxiliary (LiAlH4), DMP oxidation of the derived
alcohol, addition of MeMgBr, and DMP oxidation (83%,
five steps). For subsequent coupling of methyl ketone 55 with
aldehyde 6, lithium- and boron-mediated aldol reactions
(31) (a) Fanjul, S.; Hulme, A. N.; White, J. W. Org. Lett. 2006, 8, 4219.
(b) Fanjul, S.; Hulme, A. N. J. Org. Chem. 2008, 73, 9788.
(32) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.; Masamune,
S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
7224 J. Org. Chem. Vol. 74, No. 19, 2009

Menche et al.
JOCFeatured Article
SCHEME 9
SCHEME 10
SCHEME 11
were evaluated, with the latter giving better results in terms
of yield and scalability. Optimum conditions involved use of
NEt3 and (c-Hex)2BCl giving aldol product 56 in high yields.
For elimination, a two-step procedure proved optimal,
which involved first acylating the free hydroxyl group with
Ac2O/DMAP and subsequent treatment of the acetate with
DBU, enabling access to 54 in excellent 95% yield over three
steps.
In order to shorten the synthetic sequence to methyl
ketone 55, also a more direct access from Masamune aldol
product 46 was evaluated by direct displacement of the
auxiliary with Weinreb amine 57. Following our successful
concept for Lewis acid activation as established above, an
analogous approach was studied also for this purpose.33 As
shown in Scheme 10, this involved again Lewis acid activa-
tion to give intermediate 58 which was then treated with
metalated Weinreb amine. As shown in the table in Scheme 10,
best results were again obtained with i-PrMgCl, giving the
desired product 59 with preparatively useful yields (72%).
Subsequent methylation and substitution with MeMgBr of
derived ether 59 proceeded smoothly to enable access to the
desired methyl ketone 55 in a straightforward fashion (74%,
two steps).
After successful construction of the northern carbon
skeleton of the archazolids, efforts were then directed to
stereoselective reducing the C-15 ketone to alcohol 60 with
the desired R-configuration. Accordingly, a variety of redu-
cing agents were evaluated for this purpose, as shown in
Scheme 11. However, it soon became apparent that this
transformation was hampered by a strong substrate bias
toward the here-undesired (S)-isomer, presumably by coor-
dinative effects of the 15-OMe group. Using achiral reducing
agents (e.g., K-Selectride, LAH), this unnatural epimer was
obtained with high selectivity (9 to 10:1), which could also
not be overturned by addition of Lewis acids (e.g., LiI,
CeCl3).34 Finally, selective access to the desired (R)-isomer
with good stereoselectivity (dr = 7:1) was accomplished by
using the CBS method.23 At this stage, yields remained
however only low (3-9%), mainly due to isomerization or
decomposition of the labile triene substrate. Therefore, at
this stage the synthetic campaign was pursued with ketone 54
and the prospect of reducing the ketone at a later step of the
synthesis (vide infra).
As a prelude to forge the northern and southern ring
fragment of the archazolids by means of a Heck reaction,
first model studies were carried out with the southern frag-
ments 7 and 61 and simplified building blocks for the north-
ern part, i.e., 55 and 62, in order to evaluate substituent
effects and establish suitable reaction conditions. For this
purpose, part of the material of homoallylic alcohol 7 was
transformed into TES ether 61. As shown in the table in
Scheme 12, a free hydroxyl group on the iodide coupling
partner was beneficial in all cases (entries 2/3, 5/6, and 8/9).
This may be explicable by additional coordinative effects, in
agreement with previous results.35 However, to our surprise,
(33) Direct replacement of Abiko-Masamune aldol product 47 with the
Weinreb amid resulted in only low yields of the product, giving mainly
elimination.
(35) For an example of the beneficial effect of a free hydroxyl, see, e.g.:
Hosoya, T.; Takashiro, E.; Matsumoto, T.; Suzuki, K. J. Am. Chem. Soc.
1994, 116, 1004.
(34) Gemal, A. L.; Luche, J.-L. J. Am. Chem. Soc. 1981, 103, 5434.
J. Org. Chem. Vol. 74, No. 19, 2009 7225

Menche et al.
JOCFeatured Article
SCHEME 12
SCHEME 13
Accordingly, the desired diene 66 was obtained in acceptable
yields (55%) and selectivities (6:1).38 The minor, here unwanted
20-Z-isomer was readily removed after the reaction by HPLC
separation on normal phase. Subsequent esterification with
phosphonate 8 (BOP, DMAP, 93%) and PMB-deprotection to
the respective primary alcohol (DDQ) proceeded smoothly
(94%). In contrast, oxidation to aldehyde 67 proved trouble-
some, leading to isomerization of the trienone system (TPAP
oxidation) or elimination along the (7,8)- and/or (16,17)-bond
(TEMPO oxidation) using a number of conditions. It was only
after resorting to the Swern protocol that the desired transforma-
tion could be effected (88%).
essentially no E/Z selectivities were obtained by following
either a conventional protocol [entries 1-3: Pd(OAc)2, NEt3,
DMF/H2O, rt to 80 ꢀC]36 or Jeffrey conditions37 [entries
4-6: Pd(OAc)2, TBACl, NaHCO3, DMF, rt]. After consid-
erable experimentation, it was finally realized that this unfa-
vorable result may be remedied by modification of reaction
conditions (entries 7-9), giving diene 63 with preparatively
acceptable yield (61%) and diastereoselectivity (6:1) (entry 7).
Optimal conditions involved use of a more stable catalyst
[PdCl2(PPh3)2] and the presence of water as cosolvent (DMF/
CH3CN/H2O=5:5:1). Importantly, use of waters resulted in
drastically reduced reaction times (several hours to a few
minutes) and consistently good selectivities. These observa-
tions suggest that isomerization occurs after initial coupling
by a hydropalladation-reductive elimination sequence. This
assumption was further corroborated by minor degrees of
isomerization observed also along the labile northern triene
system.
Overriding this distinct propensity for elimination along
the (7,8)- and (16,17)-anti-configured bonds also constituted
a major challenge in the subsequent macrocyclization of 67
to 68. Even under very mild Masamune-Roush conditions
(DBU, LiCl)32 mainly elimination was observed. Decom-
position products were also obtained with K2CO3 or
KHMDS/18-C-6. Use of Ba(OH)2 resulted in no conversion
(reisolation of starting material). Finally, the desired cycliza-
tion to 68 could be effected by a serendipitously found
With these optimized conditions in hand, union of the
authentic coupling partners 54 and 7 proceeded smoothly, with-
out the necessity of further adapting the protocol (Scheme 13).
˚
combination of NaH as base and 4 A molecular sieves, still
present from drying the starting material. This resulted in
ring closure with preparatively acceptable yields (44%), with
(36) (a) Heck, R. F. J. Am. Chem. Soc. 1974, 96, 1133. (b) Heck, R. F. Org.
React. 1982, 27, 345.
(37) (a) Jeffrey, T. Chem. Commun. 1984, 1287. (b) Jeffrey, T. Tetrahedron
Lett. 1985, 26, 2267. (c) Jeffrey, T. Synthesis 1987, 70.
(38) In addition, certain degrees of isomerization along the 11Z-double
bond were observed. The unwanted 11E isomer was removed after cycliza-
tion to 68.
7226 J. Org. Chem. Vol. 74, No. 19, 2009

Menche et al.
JOCFeatured Article
SCHEME 14
polyketide macrolactone antibiotics archazolid A and B has
been accomplised. Key steps for their syntheses involve an
aldol condensation to construct the (Z,Z,E)-triene system,
an E-selective Heck reaction for the (22,23)-diene moiety,
and an HWE coupling. Two different methods, an HWE
reaction and a Heck coupling, were successfully utilized to
forge the macrocyclic core of these polyketides. Notably,
the joint synthesis of archazolid A and B was enabled by
late-stage diversification of joint late-stage intermediates,
demonstrating an expedient and flexible endgame strategy.
Furthermore, in the course of this campaign, efficient pro-
tocols for an E-selective Heck reaction of nonactivated
alkenes, a direct nucleophilic displacement of the Abiko-
Masamune auxiliary by Lewis acid activation, and a mild
and highly reactive HWE protocol were developed. With the
established robust and flexible synthetic approach to the
archazolids in hand, efforts can now be directed for further
SAR studies and more detailed target-inhibitor evaluations.
Experimental Section
Acid 41. A solution of diethyl methylmalonate (38, 33.2 g,
0.190 mol, 1 equiv) in Et2O (60 mL) was added to NaH
(55-65% in mineral oil, 9.21 g, 0.230 mol, 1.2 equiv) in Et2O
(120 mL) during 40 min under vigorous stirring, and the
resulting mixture was refluxed for 3 h. After being cooled to
rt, CHI3 (75.0 g, 0.190 mol, 1 equiv) was added during 30 min,
and the mixture was refluxed for 32 h. At 0 ꢀC, 10% aqueous
HCl (100 mL) was added and the mixture stirred for 20 min. The
organic layer was separated, and the aqueous layer was extrac-
ted with Et2O (3 ꢀ 30 mL). The combined organic layers were
dried over MgSO4 and concentrated in vacuo. The residue
was dissolved in EtOH/H2O (3:1, 520 mL), and KOH (28.1 g,
0.500 mol) was added. The mixture was refluxed for 24 h. After
being cooled to rt, the mixture was concentrated in vacuo. The
residue was redissolved in 10% aqueous K2CO3 (300 mL), and
the precipitated CHI3 was removed by filtration. The filtrate was
washed with CH2Cl2 (2 ꢀ 50 mL), acidified with 12 M HCl
(pH<1, 130 mL) and extracted with CH2Cl2 (8 ꢀ 50 mL). The
combined organic layers were dried over MgSO4 and concen-
trated in vacuo. The residue was purified by column chromato-
graphy on silica gel (hexanes/EtOAc = 10:1, 0.5% AcOH) to
give the carboxylic acid 41 (31.1 g, 0.147 mol, 77%, over two
steps) as a white solid: Rf =0.25 (hexanes/EtOAc=10:1, 0.5%
AcOH); mp 58 ꢀC; 1H NMR (300 MHz, CDCl3) δ=2.05 (s, 3H),
8.03 (s, 1H), 12.08 (s, 1H); 13C NMR (75 MHz, CDCl3) δ=19.8,
102.0, 139.0, 169.2; HRMS calculated for C4H4IO2 [M - H]þ
210.9256, found 210.9254. The spectroscopic data were identical
to those previously reported.25
Direct Nucleophilic Substitution the Abiko-Masamune Aux-
iliary: Phosphonate 52. To a cooled solution (-78 ꢀC) of dimethyl
methylphosphonate (48, 156 mL, 2.93 mmol, 11.0 equiv) in THF
(1 mL) was added KHMDS (0.5 M in toluene, 2.66 mL, 1.33 mmol,
10.0 equiv), and the resulting suspension was stirred at -20 ꢀC for
2 h. To a cooled solution (-78 ꢀC) of the ester 46 (92.1 mg,
0.136 mmol, 1.0 equiv) in THF (1 mL) was added i-PrMgCl (2.0 M
in Et2O, 204 μL, 0.409 mmol, 3.0 equiv). After 20 min, the above
mixture was added via cannula. The reaction mixture was warmed
to -20 ꢀC for ca. 1.5 h and stirred at -20 ꢀC for 5 h. Saturated aq
NH4Cl (6 mL) and H2O (6 mL) were added, the organic layer was
separated, and the aqueous layer thoroughly extracted with EtOAc
(4 ꢀ 6 mL). The combined organic layers were dried over MgSO4
and concentrated in vacuo. Silica gel chromatography (hexanes/
EtOAc=1:2) afforded the phosphonate 52 (40.9 mg, 0.109 μmol,
80%) as a colorless oil and the chiral Masamune alcohol auxi-
liary as white solid (50.6 mg, 0.120 mmol, 88%): Rf = 0.15
regard to the lability and overall complexity of the substrate.
Importantly, addition of molecular sieves was essential, in
order to avoid elimination processes, which may be due to
the slightly acidic nature of this additive. Completion of the
total synthesis involved stereoselective reduction of the
ketone on the basis of our previous results with (S)-oxaza-
borolidine-assisted borane, which proceeded again with high
diastereoselectivity and this time also acceptable yields
(73%), presumably by macrocyclic stabilization of the labile
triene system by conformational restraints. Final deprotec-
tion of the secondary alcohol at C-7 with HF/pyridine
completed the total synthesis of archazolid A in a straight-
forward fashion. The spectroscopic data and optical rotation
were in agreement with those of the authentic natural
products.1
For total synthesis of archazolid B (2), a modified end-
game strategy was applied in order to demonstrate the
modularity of our synthetic approach (Scheme 14). This
involved first an intermolecular HWE coupling of phospho-
nate 69 with aldehyde 70. Both starting materials were
readily available from 7 and 54 by esterification with 9 and
PMB deprotection/Swern oxidation, respectively. Again
HWE coupling was efficiently carried out with NaH as base.
Subsequently, a Heck macrocyclization of 71 was effected by
a protocol of Ziegler.39 Completion of the total synthesis of
archazolid B proceeded uneventfully by stereoselective CBS
reduction and HF deprotection, following our previously
established protocols (41%, three steps). All data of syn-
thetic archazolid B were identical with those of the authentic
natural product.1,8
Conclusions
In conclusion, based on a modular synthetic strategy, an
expedient and joint total synthesis of the potent antiproliferative
(39) Ziegler, F. E.; Chakraborty, U. R.; Weisenfeld, R. B. Tetrahedron
1981, 37, 4035.
J. Org. Chem. Vol. 74, No. 19, 2009 7227

Menche et al.
JOCFeatured Article
(hexanes/EtOAc = 1:2); [R]20D = -34.9 (c = 0.97, CHCl3); H
NMR (300 MHz, CDCl3) δ=0.93 (d, J=7.0 Hz, 3H), 1.82 (s, 3H),
3.06 (dq, J=9.2, 7.0 Hz, 1H), 3.17 (dd, J=18.7, 13.8 Hz, 1H), 3.25
(dd, J=18.7, 13.8 Hz, 1H), 3.76 (s, 3H), 3.80 (s, 3H), 4.25 (d, J=9.2
Hz, 1H), 6.30 (s, 1H); 13C NMR (75 MHz, CDCl3) δ=13.7, 18.7,
41.2, 42.9, 50.2, 53.2, 53.3, 79.2, 80.9, 147.4, 205.4, 205.5; HRMS
calculated for C10H18IO5PNa [MþNa]þ 398.9834, found 398.9836.
Direct Nucleophilic Substitution the Abiko-Masamune Aux-
iliary: Weinreb Amide 59. To a solution of ester 46 (231 mg, 0.342
mmol, 1.0 equiv) in THF (1 mL) was added i-PrMgCl (∼2 M in
THF, 0.17 mL, 0.34 mmol, 1.0 equiv); after 10 min, a suspension
of magnesium chloride methoxy(methyl)amide complex, which
was prepared by addition of i-PrMgCl (∼2 M in THF, 3.42 mL,
6.84 mmol, 20 equiv) to a suspension of N,O-dimethylhydrox-
ylamine hydrochloride (334 mg, 3.42 mmol, 10 equiv) in THF
(3 mL) at -20 ꢀC, was added. The reaction mixture was stirred at
-20 ꢀC for 2 h and warmed to -10 ꢀC (ca. 1 h). The reaction was
quenched by addition of satd aq NH4Cl (5 mL). The poduct was
extracted into EtOAc (3 ꢀ 20 mL), and the combined organic
layers were dried over MgSO4 and concentrated in vacuo.
Flash chromatography (hexanes/EtOAc = 2:1 to 1:1) gave
amide (77.0 mg, 0.246 mmol, 72%) as a white solid: Rf = 0.15
(hexanes/EtOAc = 2:1); [R]20D = -13.2 (c = 1.94, CHCl3); mp
82-83 ꢀC; 1H NMR (400 MHz, CDCl3) δ =1.09 (d, J=7.1 Hz,
3H), 1.81 (d, J=1.0 Hz, 3H), 3.15 (m, 1H), 3.17 (s, 3H), 3.58 (d,
J = 6.1 Hz, 1H), 3.69 (s, 3H), 4.26 (t, J = 6.4 Hz, 1H), 6.30
(s, 1H); 13C NMR (100 MHz, CDCl3) δ=15.1, 20.1, 32.2, 38.2,
61.8, 79.0, 80.1, 148.1, 176.3; HRMS calcd for C9H16INO3Na
[M þNa]þ 336.0073, found 336.0075.
1H); 13C NMR (100 MHz, CDCl3, data for major 11Z,20E-
diastereomer) δ= -4.8, -4.3, 10.9, 14.3, 15.7, 16.7, 17.3, 18.2,
20.0, 22.2, 23.1, 24.6, 24.7, 25.9, 27.7, 39.7, 40.7, 43.9, 44.5, 46.4,
55.3, 56.3, 69.1, 72.2, 72.6, 73.0, 74.5, 90.0, 113.8, 114.5, 126.7,
127.4, 129.0, 129.3, 130.2, 130.7, 137.7, 131.9, 132.6, 133.6,
134.8, 135.8, 138.9, 141.1, 156.0, 158.2, 159.2, 171.1, 203.4;
HRMS calcd for C53H82N2O8NaSiS 957.5459, found 957.5483.
HWE Macrocyclization: Macrocycle 68. Aldehyde 67 (1.1 mg,
1.1 μmol, 1 equiv) was dissolved in absolute THF (1.5 mL) under
1
˚
argon and treated with molecular sieves 4 A (20 mg). This
solution was cooled to 0 ꢀC before NaH (60% in mineral oil,
29 μg, 1.2 μmol, 1.1 equiv) was added. After the addition, the
resulting suspension was stirred for 1 h at 0 ꢀC and then warmed
to ambient temperature and stirred for a further 4 h. Then the
reaction was stopped by addition of 2 mL of pH 7 buffer and
then diluted with Et2O (10 mL). The organic layer was separated
and the aqueous phase reextracted twice with Et2O (10 mL).
After the layers were combined, the organic layer was washed
with brine, dried (MgSO4), filtered, and separated from the
solvent under reduced pressure. The residue was then purified
by column chromatography (EtOAc/hexanes = 1:2) to obtain
0.41 mg (0.48 μmol) of the diastereomerically pure cyclization
product 68 (44%): Rf = 0.61 (EtOAc/hexanes = 1:2); [R]20
=
D
-12.5 (c=0.04, CHCl3); 1H NMR (600 MHz, CD3OD) δ=0.03
(s, 3H), 0.06 (s, 3H), 0.91 (s, 9H), 0.94 (d, J=6.8 Hz, 3H), 0.95
(d, J=6.8 Hz, 3H), 1.02 (d, J=6.8 Hz, 3H), 1.03 (d, J=6.8 Hz,
3H), 1.12 (d, J=6.8 Hz, 3H), 1.72 (s, 3H), 1.75 (s, 3H), 1.79 (m,
2H), 1.85 (m, 2H), 1.83 (s, 3H), 1.91 (s, 3H), 1.94 (s, 3H), 2.40
(ddd-like, J=17.0 Hz, J=13.2 Hz, J=6.4 Hz, 1H), 2.74 (s, 3H),
3.03 (d, J=7.9 Hz, 2H), 3.05-3.10 (m, 1H), 3.15 (s, 3H), 3.17 (d,
J=6.8 Hz, 1H), 3.72 (d, J=6.8 Hz, 1H), 4.21 (dd, J=9.1 Hz,
J=6.0 Hz, 1H), 5.22 (m, 1H), 5.23 (m, 1H), 5.63 (dd, J=15.1 Hz,
J=8.7 Hz, 1H), 5.90 (d, J=11.0 Hz, 1H), 5.95 (d, J=5.3 Hz,
1H), 6.04 (dd, J=9.1 Hz, J=4.5 Hz, 1H), 6.17 (dd, J=15.1 Hz,
J=10.6 Hz, 1H), 6.22 (d, J=16.2 Hz, 1H), 6.30 (s, 1H), 7.00 (t,
J = 7.6 Hz), 7.15 (s, 1H), 7.55 (d, J = 16.2 Hz, 1H); 13C NMR
(100 MHz, CD3OD) δ=-4.5, -4.1, 12.5, 14.0, 15.1, 15.9, 18.1,
18.3, 19.0, 19.4, 22.4, 23.3, 24.7, 25.8, 26.4, 27.5, 30.7, 39.2, 42.8,
43.1, 50.3, 57.2, 73.3, 74.0, 77.5, 88.5, 116.6, 127.9, 128.4, 128.5,
129.3, 129.5, 133.6, 133.9, 134.3, 135.0, 135.7, 139.9, 140.2,
142.2, 142.9, 149.2, 158.2, 168.2, 200.1; HRMS calcd for
C48H74N2O7NaSSi 873.4884, found 873.4880.
E-Selective Heck Reaction of Nonactivated Alkenes: Diene 66.
To a solution of the vinyl iodide 54 (74.5 mg, 99.3 μmol, 1.0
equiv) and alkene 7 (46.0 mg, 147 μmol, 1.5 equiv) in DMF was
˚
added 100 mg of molecular sieves 5 A under argon atmosphere.
This mixture was stirred for 15 min at rt. Then PdCl2[PPh3]2
(35.0 mg, 50 μmol, 0.5 equiv), NaHCO3 (29.0 mg, 345 μmol,
3.5 equiv), and TBACl (13.8 mg, 50 μmol, 0.5 equiv) were added.
After the mixture was warmed to 80 ꢀC, NEt3 (30 μL, 197 μmol,
2.0 equiv) was added, and the mixture was stirred at this
temperature for further 10 min followed by an addition of a
5:1 mixture of CH3CN and H2O (2.4 mL). The solution was then
allowed to cool to rt and was stirred for further 20 min before it
was diluted by addition of Et2O (5 mL), and the organic layer
was separated. The aqueous layer was reextracted by Et2O (3 ꢀ
15 mL), and the combined organic layers were dried (MgSO4)
and evaporated under reduced pressure. Purification of the residue
by column chromatography on silica gel (EtOAc/hexanes =
1:2), and subsuent purification by preparative HPLC (hexane/
iPrOH=95:5) gave the desired 20E-configurated compound 66
(43.6 mg, 46.0 μmol) as an unseparable mixture together with
the 11E(20E)-isomer (6.1 mg, 6.5 μmol) in a ratio of 6:1 (11Z/
11E = 6:1) as a yellow oil (combined yield 55%), as well as the
11Z,20Z-isomer (7.3 mg, 7.8 μmol, likewise ratio: 6:1), which
was separated by HPLC purification (overall yield of all iso-
mers: 61%). The desired product 66 was characterized as a
mixture together with its 11E isomer (11Z:11E=6:1): Rf=0.37
(EtOAc/hexanes = 1:2); [R]20D = þ11.6 (c = 1.02, MeOH, dr
Archazolid A. A solution of compound 68 (1.50 mg, 1.76 μmol)
in THF (0.3 mL) under argon atmosphere was cooled to 0 ꢀC.
Then 2 μL of a 1 M solution of 2-methyl (S)-2-methyl-CBS-
oxazaborolidine in toluene (0.55 mg, 2.00 μmol, 1.1 equiv) and
BH3 SMe2 (1.3 mg, 17.6 μmol, 10 equiv) was slowly added to this
3
solution. After the mixture was stirred for 1 h at this temperature,
the reaction was quenched carefully by addition of ethanol (0.5 mL)
and warmed to rt. Additional H2O (5 mL) and EtOAc (5 mL) were
added, the organic layer was separated, and the aqueous phase was
extracted with EtOAc (2ꢀ 5 mL). The combined organic layer was
dried with MgSO4 and filtered, and the solvent was evaporated
under reduced pressure. After purification by column chromatog-
raphy on silica gel (EtOAc/hexanes = 1:4 to EtOAc/hexanes =
1:2), 1.1 mg (1.3 μmol, 73%) of the alcohol was obtained as a
colorless oil: Rf=0.44(hexanes/EtOAc=2:1);[R]20D=-74.3 (c=
1.4 mg/mL, CHCl3; 1H NMR (CD3OD, 600 MHz) δ = 0.03
(s, 3H), 0.06 (s, 3H), 0.78 (d, J=7.3 Hz, 3H), 0.88 (d, J=6.8 Hz,
3H), 0.91 (s, 9H), 1.01 (d, J= 6.80 Hz, 3H), 1.03 (d, J =6.4 Hz,
3H), 1.12 (d, J=6.8 Hz, 3H), 1.61 (d, J=1.1 Hz, 3H) 1.72 (d, J=
1.1 Hz, 3H), 1.75 (ddq, J=8.7Hz, J=4.2Hz, J=6.8 Hz, 1H), 1.78
(s, 3H), 1.80 (m, 2H), 1.84 (dd, J=8.3 Hz, J=4.9 Hz, 1H), 1.91 (s,
3H), 1.93 (d, J=1.1 Hz, 3H), 2.30 (ddq, J=9.8 Hz, J=8.6 Hz, J=
6.8 Hz, 1H), 2.74 (s, 3H), 2.95 (dd, J=15.1 Hz, J=6.8 Hz, 1H),
3.00 (dd, J=14.4 Hz, J=9.1 Hz, 1H), 3.06 (ddq, J=7.6 Hz, J=
4.5 Hz, J=6.8 Hz, 1H), 3.13 (s, 3H), 3.29 (d, J=8.7 Hz, 1H), 4.08
(dd, J=8.9, J=8.9 Hz, 1H), 4.20 (dd, J=7.2 Hz, J=4.2 Hz, 1H),
1
(11E:Z) = 6:1); H NMR (300 MHz, CD3OD, data for major
diastereomer) δ=-0.01 (s, 3H), 0.02 (s, 3H), 0.86 (d, J=6.6 Hz,
3H), 0.89 (s, 9H), 0.94 (d, J=6.6 Hz, 3H), 1.01 (d, J=6.0 Hz,
3H), 1.02 (d, J=6.0 Hz, 3H), 1.09 (d, J= 6.8 Hz, 3H), 1.66 (s,
3H), 1.68 (s, 3H), 1.84 (s, 3H), 1.98 (s, 3H), 2.30 (t, J =6.9 Hz,
2H), 2.39 (m, 1H), 2.75 (s, 3H), 2.86 (m, 1H), 3.05 (s, 3H), 3.08
(m, 1H), 3.57 (t, J=6.7 Hz, 2H), 3.67 (d, J=9.8 Hz, 1H), 3.82 (s,
3H), 4.20 (dd, J=9.0 Hz, J=5.6 Hz, 1H), 4.46 (s, 2H), 4.71 (d,
J=5.5 Hz, 1H), 5.15 (d, J=8.9 Hz, 1H), 5.28 (d, J=10.2 Hz,
1H), 5.76 (dd, J=15.1 Hz, J=8.3 Hz, 1H), 6.03 (m, 2H), 6.25
(m, 1H), 6.31 (s, 1H), 6.38 (d, J=15.6 Hz, 1H), 6.92 (d, J=8.7 Hz,
2H), 7.29 (d, J=8.7 Hz, 2H), 7.32 (s, 1H), 7.61 (d, J=15.6 Hz,
7228 J. Org. Chem. Vol. 74, No. 19, 2009

Menche et al.
JOCFeatured Article
5.16 (dt, J=10.1 Hz, J=1.2 Hz, 1 H), 5.19 (d, J=9.1 Hz, 1 H),
5.55 (dd, J=15.1, J=7.6 Hz, 1H), 5.80 (dd, J=15.9 Hz, J=7.9
Hz, 1H), 5.80 (s, 1H), 5.83 (d, J=9.4 Hz, 1H), 5.96 (d, J=4.5 Hz,
1H), 6.03 (dd, J=9.1 Hz, J=4.9 Hz, 1H), 6.06 (dd, J=14.7 Hz,
J=10.6 Hz, 1H), 6.38 (d, J=15.9 Hz, 1H), 6.87 (ddd, J=8.3 Hz,
J=6.8 Hz, J=1.5 Hz, 1H), 7.20 (s, 1H); 13C NMR (CDCl3, 100
MHz) δ = -4.71, -4.20, 11.4, 12.6, 13.8, 14.4, 16.6, 16.9, 17.7,
19.83, 22.2, 23.0, 24.4, 24.7, 25.9, 27.6, 39.8, 40.8, 41.1, 42.0, 44.4,
56.1, 72.1, 73. 8, 76.2, 76.3, 90.5, 115.1, 126.3, 128.6, 128.8, 129.3,
130.5, 130.6, 131.6, 132.3, 132.3, 133.0, 133.2, 134.6, 140.8, 141.1,
155.2, 156.0, 168.3, 174.3; HRMS calcd for C48H76N2O7NaSSi
875.5040, found 875.5035.
Heck product (1.80 mg, 2.20 μmol, 60%), which was directly used
in the next step due to its base sensitivity. A solution of this Heck
product (1.80 mg, 2.20 μmol) in THF (0.3 mL) under argon
atmosphere was treated at 0 ꢀC with 3 μL of a 1 M solution of
2-methyl (S)-2-methyl-CBS-oxazaborolidine in toluene (0.83 mg,
3.00 μmol) and BH3 SMe2 (2.0 mg, 27.0 μmol). After the mixture
3
was stirred for 1 h at this temperature, ethanol (1 mL) was added
carefully, the reaction mixture was warmed to rt, H2O (5 mL) and
EtOAc (5 mL) were added, the organic layer was separated,
and the aqueous phase was extracted with EtOAc (2 ꢀ 5 mL).
The combined organic layer was dried with MgSO4 and filtered,
and the solvent was evaporated under reduced pressure. A solution
of the crude product in THF (0.2 mL) was treated at 0 ꢀC with
pyridine (0.6 mL, 7.5 mmol) and HF (0.4 mL, 70% in pyridine,
15.4 mmol), and the resulting mixture was stirred at rt for 4. After
addition of pH 7 buffer (10 mL) and dilution with CH2Cl2 (5 mL),
the organic layer was separated and the aqueous phase extracted
twice with CH2Cl2 (2 ꢀ 5 mL). The combined organic layer was
washed with saturated, aqueous NaHCO3 solution (10 mL) and
brine (10 mL), dried (MgSO4), and filtered, and the solvent was
removed in vacuo. The crude product was purified by HPLC
(Nucleosil 100-7 C18, 250/21; CH3CN/H2O = 70:30 to CH3CN/
H2O = 85:15) to give 1.0 mg (1.49 μmol, 41%, three steps) of
archazolid B (2) as a colorless oil: Rf = 0.22 (hexanes/EtOAc =
2:1); [R]22D = -63.8 (c = 64, MeOH) [lit.12 [R]25D = -61.9 (c =
0.52, MeOH)]; 1H NMR (CD3OD, 600 MHz) δ=0.72 (d, J=7.1
Hz, 3H), 0.81 (d, J=6.6 Hz, 3H), 1.01 (d, J=6.6 Hz, 3H), 1.02 (d,
J=6.1 Hz, 3H), 1.11 (d, J=6.8 Hz, 3H), 1.67 (d, J=1.2 Hz, 3H),
1.74 (d, J=1.0 Hz, 3H), 1.78 (m, 1H), 1.79 (m, 1H), 1.79 (s, 3H),
1.92 (m, 2H), 1.93 (d, J = 1.1 Hz, 3H), 2.30 (ddq, J = 9.5, J =
9.5 Hz, J= 7.0 Hz, 1H), 2.75 (s, 3H), 2.91 (dd, J= 14.7 Hz, J=
6.6 Hz, 1H), 2.96 (dd, J=14.7 Hz, J=8.6 Hz, 1H), 3.09 (ddq, J=
6.6 Hz, J=4.6Hz, J=6.8 Hz, 1H), 3.19 (s, 3H), 3.47 (d, J=9.2Hz,
1H), 4.03 (dd, J=9.2 Hz, J=9.2 Hz, 1H), 4.38 (dd, J=5.6 Hz, J=
3.2 Hz, 1H), 5.18 (dd, J=9.2Hz,J=1.0Hz,1H),5.28(d,J=9.7Hz,
1H), 5.75 (dd, J=15.3 Hz, J=6.6 Hz, 1H), 5.77 (dd, J=16.2 Hz,
J=5.6 Hz, 1H), 5.79 (s, 1H), 5.90 (d, J=10.2 Hz, 1H), 5.90 (d, J=
4.1 Hz, 1H), 5.92 (d, J= 15.7 Hz, 1H), 6.03 (dd, J=8.9 Hz, J=
4.8 Hz, 1H), 6.28 (ddd, J=15.3 Hz, J=10.7 Hz, J=1.0 Hz, 1H),
6.60 (d, J=15.8 Hz, 1H), 6.94 (ddd, J=15.5 Hz, J=8.4 Hz, J=
6.4 Hz, 1H), 7.31 (s, 1H); 13C NMR (CD3OD, 100 MHz) δ=11.6,
11.9, 16.2, 16.3, 17.1, 19.2, 21.6, 22.7, 24.1, 25.1, 26.8, 40.5, 40.7,
43.2, 43.3, 45.3, 55.7, 72.6, 73.0, 74.0, 76.5, 89.3, 116.7, 122.6,
126.5, 128.6, 129.3, 130.0, 130.0, 132.2, 132.6, 132.7, 133.8,
134.7, 134.8, 135.5, 148.3, 154.9, 157.5, 165.9, 173.5; HRMS
calcd for C41H60N2O7S 724.4121, found 724.4119. All data were
identical to those previously reported for archazolid B from
A. gephyra.1,8
The alcohol as prepared above (4.0 mg, 4.7 μmol) was
dissolved in THF (0.2 mL). The resulting solution was cooled
to 0 ꢀC, and then pyridine (0.6 mL, 7.5 mmol) and HF (0.4 mL
(70% in pyridine, 15.4 mmol) were added. The resulting mixture
was warmed to rt, stirred at this temperature for 4 h, and
quenched by addition of pH7 buffer (10 mL) afterward. The
mixture was diluted with CH2Cl2 (5 mL), the organic layer was
separated, and the aqueous phase was extracted twice with
CH2Cl2 (2 ꢀ 5 mL). The combined organic layer was washed
with saturated, aqueous NaHCO3 solution (10 mL) and brine
(10 mL), dried (MgSO4), and filtered, and the solvent was
removed in vacuo. The crude product was purified by HPLC
(Nucleosil 100-7 C18, 250/21; CH3CN/H2O=70:30 to CH3CN/
H2O=85:15) to give 2.8 mg (3.8 μmol, 80%) of archazolid A (1)
as a colorless amorphous solid: Rf = 0.25 (hexanes/EtOAc =
2:1); [R]20D = -47.3 (c = 1.2 mg/mL, MeOH) [lit.4 [R]22
=
D
-64.0 (c = 12.8 mg/mL, MeOH)]; 1H NMR (CD3OD, 600
MHz) δ=0.74 (d, J=7.0 Hz, 3H), 0.84 (d, J=7.0 Hz, 3H), 1.01
(d, J=6.0 Hz, 3H), 1.02 (d, J=6.0 Hz, 3H), 1.15 (d, J=7.0 Hz,
3H), 1.64 (d, J = 1.2 Hz, 3H), 1.73 (d, J = 1.2 Hz, 3H), 1.77
(m, 1H), 1.80 (ddq, J=9.0 Hz, J=3.0 Hz, J=7.0 Hz, 1H), 1.80
(s, 3H), 1.91 (d, J=1.2 Hz, 3H), 1.92 (m, 2H), 1.93 (d, J=1.2 Hz,
3H), 2.30 (ddq, J=9.8, J=9.5 Hz, J=7.0 Hz, 1H), 2.75 (s, 3H),
2.91 (dd, J=15.0 Hz, J=7.5 Hz, 1H), 3.03 (dd, J=15.0 Hz, J=
7.5 Hz, 1H), 3.10 (ddq, J=7.5 Hz, J=4.0 Hz, J=7.0 Hz, 1H),
3.16 (s, 3H), 3.40 (d, J=9.0 Hz, 1H), 4.03 (dd, J=9.4Hz,J=9.3Hz,
1H), 4.31 (dd, J=6.4 Hz, J=3.2 Hz, 1H), 5.21 (dd, J=9.5 Hz,
J=1.2 Hz, 1H), 5.27 (d, J=9.6 Hz, 1H), 5.63 (dd, J=15.2 Hz,
J=7.0 Hz, 1H), 5.79 (dd, J=15.5 Hz, J=6.0 Hz, 1H), 5.81 (s,
1H), 5.87 (dd, J=10.9 Hz, J=1.2 Hz, 1H), 5.97 (d, J=4.1 Hz,
1H), 6.04 (dd, J=9.1 Hz, J=4.5 Hz, 1H), 6.17 (dd, J=15.5 Hz,
J = 10.9 Hz, 1H), 6.56 (dd, J = 15.5 Hz, J = 0.9 Hz, 1H), 6.82
(ddd, J=7.9 Hz, J=7.5 Hz, J=1.3 Hz, 1H), 7.21 (s, 1H); 13
C
NMR (CD3OD, 100 MHz) δ=12.6, 12.6, 13.0, 16.8, 17.6, 17.7,
19.9, 22.4, 23.4, 24.7, 25.8, 27.5, 40.6, 41.8, 42.0, 44.5, 46.0, 56.2,
73.3, 73.6, 75.5, 77.6, 89.8, 116.7, 127.6, 129.3, 129.7, 130.8,
130.4, 130.7, 134.8, 132.7, 133.6, 133.6, 135.1, 135.7, 136.9,
142.1, 156.1, 158.2, 168.3, 173.9; HRMS calcd for C42H62N2O7S
738.4277, found 738.4274. All data were identical to those
previously reported for archazolid A from A. gephyra.1,8
Acknowledgment. This work was supported by the Volks-
wagenstiftung, the Fonds der Chemischen Industrie, the
Wild-Stiftung, and the HZI. We thank Antje Ritter, Tatjana
Archazolid B. A solution of PdCl2(MeCN)2 (1.50 mg, 5.78 μmol)
in MeCN (0.5 mL) was treated with iodide 71 (3.50 mg, 3.63 μmol)
in MeCN (0.5 mL), NEt3 (5 μL, 35.5 μmol), and formic acid
(0.1 mL of a premixed solution of 4 μL formic acid in 1 mL MeCN,
corresponds to 0.4 mL formic acid, 10.9 μmol) and stirred at rt for 2
h. Filtration through a pad of Celite, evaporation of the solvent in
vacuo, and purification by HPLC (Nucleosil 100-7 C18, 250/21;
hexane/MTBE=15:1 to 6:3 gradient elution) gave the corresponding
€
Arnold, Henning Stockmann, and Nicole Horstmann for
technical support.
Supporting Information Available: Full experimental details
1
and copies of the H and 13C NMR spectra for all new com-
pounds. This material is available free of charge via the Internet
at http://pubs.acs.org.
J. Org. Chem. Vol. 74, No. 19, 2009 7229