Alkyne haloallylation [with Pd(II)] as a core strategy for macrocycle synthesis: A total synthesis of (-)-haterumalide NA/(-)-oocydin A

Article abstract of DOI:10.1021/ja051749i

The rarely used haloallylation reaction, first described by Kaneda and Teranishi in 1974, employs a Pd(II) catalyst to join an alkyne with an allylic halide to produce a 1-halo-1,4-diene subunit. It is shown here that functionalized and tertiary allylic c

Full text of DOI:10.1021/ja051749i

Published on Web 04/22/2005
Alkyne Haloallylation [with Pd(II)] as a Core Strategy for Macrocycle
Synthesis: A Total Synthesis of (-)-Haterumalide NA/(-)-Oocydin A
Thomas R. Hoye* and Jizhou Wang
Department of Chemistry, UniVersity of Minnesota, Minneapolis, Minnesota 55455
Received March 18, 2005; E-mail: hoye@chem.umn.edu
Two reports in 19991 described the results of parallel studies,
Scheme 1
each of which led to the identification of the structurally unique,
phytopathogenic, and cytotoxic macrolide 1. The compound isolated
from the Okinawan sponge, Ircinia sp., was named haterumalide
NA, and the one from the bacterium, Serratia marcescens, oocydin
A. Haterumalide NA/oocydin A (1) was subsequently identified in
a soil bacterium, Serratia plymuthica.2 It is intriguing that a natural
product with this degree of structural complexity is present in such
seemingly disparate organisms.
Studies in the Kigoshi3 and Snider4 laboratories, each culminating
in the synthesis of the ent-methyl ester of 1, led to a revision of
the initially assigned relative configuration as well as to the
assignment of absolute configuration of haterumalide NA/oocydin
A. The 14-membered macrolactone in 1 bridges C11 and C13 on
the embedded THF ring in a trans fashion, thereby conferring steric
strain. This architecture seemingly places greater demands on
macrocyclization reactions used to close the large ring.5
A series of reports in the 1970s from the Kaneda/Teranishi labor-
atory described PdX2-catalyzed reactions of terminal alkynes (7)
with allylic chlorides and bromides (8) to produce crossed 1:1 ad-
Scheme 2
ducts 11 (Scheme 2).6 The predominant products result from net
cis-addition of the halogen atom and allylic moiety in 8 across the
alkyne with the regiochemistries implied by 11 (and the presumed
intermediates 9 and 10). Even though substantial excesses of the
allylic halide partner 8 have usually been employed (to overcome
competitive alkyne oligomer formation), we hypothesized that an
intramolecular, endocyclic version of this infrequently used reaction7
could be commandeered to close the large ring in 1 by simultaneous
creation of the C7-C8 trisubstituted Z-alkene and the C6-C7 bond.
The appropriate precursor with which to address this issue took
the form of 2 (Scheme 1). We have not found examples where
tertiary allylic halides have been used as reaction partners in the
Kaneda haloallylation reaction (and only two where secondary6,7a
were used).
Scheme 3
The synthesis of esters 2a and 2b evolved as summarized
retrosynthetically in Scheme 1. The absolute and relative stereo-
chemical properties of the 3-hydroxytetrahydrofuran 3a were
established in 5, which was constructed in an efficient fashion by
the method of Roush and Micalizio.8,9
Coupling reactions for joining two complex and valuable
fragments must be efficient at a nearly equimolar ratio of the two
reaction partners. This requirement pertains to both inter- and intra-
molecular versions. With this need as a guiding principle, we exam-
ined the intermolecular couplings shown in Scheme 3. The alkyne
substrate for these studies, 13, is the TIPS ether of 3a; three tertiary
allylic chlorides, 12a-c, were examined. The results were similar
regardless of whether a 5-fold (12a,b) or a 1.5-fold excess (12c)
of the chloride (relative to alkyne 13) was used. Slow addition of
the alkyne6,7f,g was clearly advantageous. Only Z-∆7,8-chloroalkene
geometry was observed (NOE H7-H9). Separable mixtures of Z-
and E-∆4,5-alkenes (∼2:1; NOE Me4-H5) were generated. Isolated
yields were >50% even though some loss due to acetonide clea-
a PdCl2(PhCN)2 (∼20 mol %), THF, RT, 12:13 molar ratio 5:1 (for 12a,b)
or 1.5:1 (for 12c); alkyne added last by syringe pump over 10-60 min;
conversion complete upon addition; [13]final ) 0.01 M. bPercent yield; 14Z +
14E ) 50-67%; 14Z:14E 1.8-2.5:1. After MPLC purification SiO2; varying
amounts of a 1,2-diol byproduct arising from acetonide cleavage were observed
(TLC for 14b and 14c) and isolated (∼10%, for 14a).
vage10 was always observed. Excess 12 could be recovered and
did not isomerize to its allylomeric primary chloride under the
reaction conditions.6
Encouraged by these results, we examined the alkynyl allylic
chlorides 15, 2a, and 2b as substrates for intramolecular macro-
cyclization (Scheme 4). Each of these allylic chlorides smoothly
underwent cyclization when exposed to PdCl2(PhCN)2. The yield
from the PMB-containing substrates 15 and 2b was excellent,
9
6950
J. AM. CHEM. SOC. 2005, 127, 6950-6951
10.1021/ja051749i CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Scheme 4
Acknowledgment. This investigation was supported by a grant
awarded by the DHHS (GM-65597). We thank Professors G.
Strobel (sample of 1), H. Kigoshi (NMR data), and B. Snider (NMR
and experimental data) for their very helpful assistance.
Supporting Information Available: Spectroscopic data for com-
1
pounds 1-5, 13-21, 23, and 24; PDFs of H NMR spectra for 1, 2,
17-21, and 24; and representative procedures for inter- and intramo-
lecular chloroallylation (56 pages, print/PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Takada, N.; Sato, H.; Suenaga, K.; Arimoto, H.; Yamada, K.; Ueda,
K.; Uemura, D. Tetrahedron Lett. 1999, 40, 6309-6312. (b) Strobel, G.;
Li, J.; Sugawara, F.; Koshino, H.; Harper, J.; Hess, W. M. Microbiology
1999, 145, 3557-3564. (c) A related ester, haterumalide B, was reported
contemporaneously: Ueda, K.; Hu, Y. Tetrahedron Lett. 1999, 40, 6305-
6308.
(2) Thaning, C.; Welch, C. J.; Borowicz, J. J.; Hedman, R.; Gerhardson, B.
Soil Biol. Biochem. 2001, 33, 1817-1826.
(3) Kigoshi, H.; Kita, M.; Ogawa, S.; Itoh, M.; Uemura, D. Org. Lett. 2003,
5, 957-960.
(4) (a) Gu, Y.; Snider, B. B. Org Lett. 2003, 5, 4385-4388. (b) Gu, Y. Ph.D.
Thesis, Brandeis University, 2004.
a PdCl2(PhCN)2 (∼20-30 mol %), THF, RT, slow addition over 2 h; [enyne
(5) (a) In the Kigoshi synthesis,3 the macrocycle was closed by a Reformatsky-
like reaction in 9% yield. In the Snider synthesis,4 Yamaguchi macro-
lactonization was accompanied by competitive dimer and trimer formation.
Our studies5c of a relay ring closing metathesis (RRCM) process5b can
also be interpreted as evidence for strain in the bridged bicycle of 1.
Collectively, these results reinforce that the macrocycle in 1 is a chal-
lenging one to close in a substrate with an intact THF ring. (b) Hoye, T.
R.; Jeffrey, C. S.; Tennakoon, M. A.; Wang, J.; Zhao, H. J. Am. Chem.
Soc. 2004, 126, 10210-10211. (c) Relay substrates i gave only the
products of truncation ii upon treatment with the Grubbs metathesis
initiator G2 (i.e., the metathesis relay, but not the subsequent ring closing
event occurred). By contrast and under the same conditions, each of the
“relaxed” substrates iii closed efficiently to the 14-membered lactones
iv, suggesting that the ring-closing event was faster than (bimolecular)
truncation.9
substrate]final ) ∼0.3 mM. bPlus the diol from acetonide cleavage.10
Scheme 5 a
a Reagents and conditions: (a) NaBH4, CeCl3‚7H2O, EtOH, -30 °C (90%);
(b) Ac2O, Et3N, DMAP, DCM, 0 °C (99%); (c) DDQ, DCM, H2O, 0 °C (94%);
(d) DMP, DCM, 0 °C; (e) 21 + 22 in DMSO, rt, then add CrCl2/NiCl2 (0.30 wt
% NiCl2) in drybox, 5-10 h (two steps: 50% for R3 ) PMB; 10:1 ratio of
24:15-epi-2412); (f) TFA, Et3SiH, DCM, 0 °C, 1 h.
regardless of whether the more relaxed, 15-membered (16) or the
14-membered macrolactone (17b) was being generated. Again, only
the Z-∆7,8-alkene was observed, but the ∆4,5-alkene was produced
as a separable mixture of alkene isomers. This result raises the
intriguing possibility that these intramolecular haloallylations are
stereospecificsthat each of the epimeric allylic chlorides engenders
a single C4-C5 alkene geometry. This hypothesis will be explored.
To complete the synthesis of 1, Luche reduction (Scheme 5) of
the C3-ketone in the Z-isomer of 17b provided 18 as a single alcohol
having the 3R configuration (MTPA). This extremely high prefer-
ence for the natural configuration is in contrast to that observed
for reduction of the conformationally relaxed, larger ketolactone
16. This substrate gave a ca. 2:1 mixture of epimeric alcohols under
similar conditions. Acetylation of 18 (to 19) and PMB removal (to
20) were uneventful. Careful Dess-Martin oxidation at 0 °C pro-
vided aldehyde 21. The Cr(II)/Ni0-mediated Nozaki-Hiyama-Kishi
(NHK) reaction11 gave the methyl ester 233,4 and, ultimately, the
PMB ester 24.12 Consistent with others’ experience,3,4 ester 23 could
not be successfully hydrolyzed under basic nor essentially neutral
(Me3SnOH, DCE13) conditions. However, the PMB ester in 24 was
cleanly cleaved (TFA, Et3SiH)14 to the acid 1, completing this first
synthesis of haterumalide NA/oocydin A (1). Key to our success
were the Pd(II)-mediated chloroallylative cyclization of 2b to close
the strained macrocycle in 17b and the choice of the acid-labile
PMB ester in 24.
(6) Kaneda, K.; Uchiyama, T.; Fujiwara, Y.; Imanaka, T.; Teranishi, S. J.
Org. Chem. 1979, 44, 55-63 and references therein to earlier communica-
tions.
(7) (a) Ma, S.; Lu, X. J. Org. Chem. 1991, 56, 5120-5125. (b) Ma, S.; Lu,
X. J. Org. Chem. 1993, 58, 1245-1250. (c) Ma, S.; Zhu, G.; Lu, X. J.
Org. Chem. 1993, 58, 3692-3696. (d) Ba¨ckvall, J.-E.; Nilsson, Y. I. M.;
Gatti, R. G. P. Organometallics 1995, 14, 4242-4246. (e) Xie, X.; Lu,
X.; Liu, Y.; Xu, W. J. Org. Chem. 2001, 66, 6545-6550. (f) Thadani, A.
N.; Rawal, V. H. Org. Lett. 2002, 4, 4317-4320. (g) Thadani, A. N.;
Rawal, V. H. Org. Lett. 2002, 4, 4321-4323.
(8) (a) Micalizio, G. C.; Roush, W. R. Org. Lett. 2001, 3, 1949-1952. (b)
(9) Details of the conversion of 5 into 3a and 3b8b as well as of the syntheses
of (()-4 and the three cyclization substrates 2a, 2b, and 15 (Scheme 4)
are described: Wang, J. Ph.D. Thesis, University of Minnesota, 2005.
(10) Lipshutz, B. H.; Pollart, D.; Monforte, J.; Kotsuki, H. Tetrahedron Lett.
1985, 26, 705-708.
(11) Fu¨rstner, A. Chem. ReV. 1999, 99, 991-1045.
(12) The sense of the diastereoselectivity of this NHK addition was supported
by Mosher ester analysis of the R- and S-MTPA esters arising from
treatment of 15-epi-24, the minor diastereomer, with S- and R-MTPA-Cl,
respectively.
(13) Nicolaou, K. C.; Estrada, A. A.; Zak, M.; Lee, S. H.; Safina, B. S. Angew.
Chem., Int. Ed. 2005, 44, 1378-1382.
(14) Pearson, D. A.; Blanchette, M.; Baker, M. L.; Guindon, C. A. Tetrahedron
Lett. 1989, 30, 2739-2742.
JA051749I
9
J. AM. CHEM. SOC. VOL. 127, NO. 19, 2005 6951