Stereoselective synthesis of the phorboxazole A macrolide by ring-closing metathesis

Article abstract of DOI:10.1021/ol0619922

Described is a regio- and stereoselective ring-closing metathesis (RCM) to form the C2-C3 alkene of the macrolide-containing domain of phorboxazole A. This work demonstrates a dramatic effect of reaction solvent on RCM product (E/Z)-selectivity. This proc

Full text of DOI:10.1021/ol0619922

ORGANIC
LETTERS
2006
Vol. 8, No. 23
5223-5226
Stereoselective Synthesis of the
Phorboxazole A Macrolide by
Ring-Closing Metathesis
Bo Wang and Craig J. Forsyth*
Department of Chemistry, Institute of Technology, UniVersity of Minnesota,
207 Pleasant Street Southeast, Minneapolis, Minnesota 55455
forsyth@chem.umn.edu
Received August 10, 2006
ABSTRACT
Described is a regio- and stereoselective ring-closing metathesis (RCM) to form the C2−C3 alkene of the macrolide-containing domain of
phorboxazole A. This work demonstrates a dramatic effect of reaction solvent on RCM product (E/Z)-selectivity. This process offers an alternative
assembly of the macrolide-containing domain of phorboxazole A, one of the most potent anticancer agents known.
Phorboxazoles A and B (Scheme 1) are potent cytostatic
natural products originally isolated from the Indian Ocean
sponge Phorbas sp. collected in 1993 off Western Australia.1a
More recently, phorboxazole A has also been isolated from
a 1996 collection of the Western Australian Indian Ocean
sponge Raspailia sp.2 This suggests that these natural
products may ultimately be biosynthesized by a currently
unidentified symbiotic marine microorganism. Despite these
natural sources, laboratory synthesis remains the most reliable
route to access the phorboxazoles.
The phorboxazoles exhibit S-phase cytostatic activity
against a broad spectrum of human cancer cell lines.1
Synthetic phorboxazole A3 was also reported to induce cancer
cell apoptosis.4 Moreover, the phorboxazoles display unique
modes of action with respect to currently viable cancer
therapeutics.1,4,5 Several total syntheses have yielded these
natural products3,6 and their structural variants.7 SAR studies
indicated that elements of both the macrolide and side chain
domains of phorboxazole A are necessary for potent cyto-
static activity.7a,c The macrolide has been assembled using
(4) (a) Uckun, F. M.; Narla, R. K.; Navara, C.; Forsyth, C. Clin. Cancer
Res. 1999, 5, 3729s. (b) Uckun, F. M.; Narla, R. M.; Forsyth, C. J.; Cink,
R. D.; Lee, C. S.; Ahmed, F. U.S. Patent 6797721, 2004, WO 01/36394,
2001, Priority: US 99-16545.2 19991115.
(5) Forsyth, C. J.; Ying, L.; Chen, J.; La Clair, J. J. J. Am. Chem. Soc.
2006, 128, 3858.
(6) (a) Evans, D. A.; Fitch, D. M.; Smith, T. E.; Cee, V. J. J. Am. Chem.
Soc. 2000, 122, 10033. (b) Smith, A. B., III; Minbiole, K. P.; Verhoest, P.
R.; Schelhaas, M. J. Am. Chem. Soc. 2001, 123, 10942. (c) Pattenden, G.;
Gonzalez, M. A.; Little, P. B.; Millan, D. S.; Plowright, A. T.; Tornos, J.
A.; Ye, T. Org. Biomol. Chem. 2003, 1, 4173. (d) Williams, D. R.; Kiryanov,
A. A.; Emde, U.; Clark, M. P.; Berliner, M. A.; Reeves, J. T. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 12058. (e) Smith, A. B., III; Razler, T. M.;
Ciavarri, J. P.; Hirose, T.; Ishikawa, T. Org. Lett. 2005, 7, 4399. (f) Li, D.;
Zhang, D.; Sun, C.; Zhang, J.; Yang, L.; Chen, J.; Liu, B.; Su, C.; Zhou,
W.; Lin, G. Chem.-Eur. J. 2006, 12, 1185.
(7) (a) Uckun, F. M.; Forsyth, C. J. Bioorg. Med. Chem. Lett. 2001, 11,
1181. (b) Hansen, T. M.; Engler, M. M.; Forsyth, C. J. Bioorg. Med. Chem.
Lett. 2003, 13, 2127. (c) Smith, A. B., III; Razler, T. M.; Pettit, G. R.;
Chapuis, J.-C. Org. Lett. 2005, 7, 4403. (d) Chen, J.; Ying, L.; Hansen, T.
M.; Engler, M. M.; Lee, C. S.; La Clair, J. J.; Forsyth, C. J. Bioorg. Med.
Chem. Lett. 2006, 16, 901.
(1) (a) Searle, P. A.; Molinski, T. F. J. Am. Chem. Soc. 1995, 117, 8126.
(b) Searle, P. A.; Molinski, T. F.; Brzeinski, L. J.; Leahy, J. W. J. Am.
Chem. Soc. 1996, 118, 9422. (c) Molinski, T. F. Tetrahedron Lett. 1996,
37, 7879.
(2) Capon, R. J.; Skene, C.; Liu, E. H.; Lacey, E.; Gill, J. H.; Heilan,
K.; Friedel, T. Nat. Prod. Res. 2004, 18, 305.
(3) Forsyth, C. J.; Ahmed, F.; Cink, R. D.; Lee, C. S. J. Am. Chem. Soc.
1998, 120, 5597.
10.1021/ol0619922 CCC: $33.50
© 2006 American Chemical Society
Published on Web 10/20/2006

Table 1. HWE Conversions of 3 to (2Z)-1 and (2E)-14a
Scheme 1. Phorboxazoles and Intermediates
temp
(°C)
time
(h)
yield
(%)
entry
3, R
Et(b),3
1/14
1
2
3
4
25
25
1.5
>5
5
69
∼1:20
4:1
4:1
Et(c),3,13
CF3CH2
CF3CH2
∼50
(c),3,13
-40 to -5
77
(c),6b,11
25
3
93
4:1
i
a Reagents applied: (b) Pr2NEt, LiCl, CH2Cl2;12 (c) K2CO3, toluene,
18-crown-6.8 The structure of 14 is given in Scheme 2.
C30 for macrolide construction.3 These intermediates were
first combined via de novo formation of the C16-C18
oxazole moiety. Subsequent elaboration of the C24 hydroxyl
and the C3 terminus provided C24 phosphonoacetates
bearing a C3 aldehyde (Scheme 1). Intramolecular Horner-
Wadsworth-Emmons reactions under a variety of conditions
generated the macrolide in variable (2E)/(2Z) ratios (Table
1). It was initially found that the (2E)-acrylate 14 (Scheme
2) could be formed in high yield and geometrical selectivity
under thermodynamically controlled12 HWE conditions
(Table 1, entry 1).3 In contrast, an ca. 4:1 ratio of (2Z)-1 to
(2E)-14 macrolides was obtained using intramolecular Still-
Gennari reactions (entries 2-4).3,8,11,13 Although the key
coupling processes of oxazole and acrylate formation have
been satisfactorily optimized, the extensive functional group
an intramolecular Still-Gennari reaction8 to form predomi-
nantly the (2Z)-alkene in all of the published total syntheses
of phorboxazole A and in the most recently published
assembly of phorboxazole B.3,6b-f The original synthesis of
phorboxazole B exploited the semireduction of a 2,3-alkyne
to provide the (2Z)-alkene in higher geometrical selectivity.6a
Described here is a remarkably regio- and stereoselective
ring-closing metathesis9 (RCM) to form the C2 alkene of
the macrolide-containing domain (1)3 of phorboxazole A.
This work demonstrates a dramatic effect of reaction solvent
on RCM product (E/Z)-geometry and can provide the (2Z)-
alkene almost exclusively.
Scheme 2. Synthesis and RCM of 4
The stereochemistry and conformation of the natural
products’ (2Z)-macrolide were originally determined on the
basis of NMR spectroscopy1a and subsequently corroborated
by X-ray crystallography of synthetic material.3 Alteration
of phorboxazole A’s (2Z)-acrylate moiety and macrolide
conformation by alkene saturation to give 2,3-dihydrophor-
boxazole A10 resulted in a dramatic loss of cytotoxic activity.4
The unnatural (2E)-isomer11 has more recently been reported
to have 30-80-fold diminished in vitro cancer cell line
growth inhibitory activity.7c Hence, the question of whether
our convergent strategy3 could be adapted to improve the
synthetic efficiency of macrolide formation via an RCM
process (Scheme 1) was coupled with the issue of C2 alkene
geometrical selectivity.
The original approach to the phorboxazole architecture
utilized synthetic fragments representing C3-C17 and C18-
(8) Still, W. C.; Gennari, C. Tetrahedron Lett. 1983, 24, 4495.
(9) (a) Fu¨rstner, A. Angew. Chem., Int. Ed. 2000, 39, 3013. (b) Trnka,
T. M.; Grubbs, R. H. Acc. Chem. Res. 2001, 34, 18. (c) Hoveyda, A. H.;
Schrock, R. R. Org. Synth. Highlights 2003, 210.
(10) Lee, C. S. Ph.D. Thesis, University of Minnesota, 1999.
(11) Hansen, T. M. Ph.D. Thesis, University of Minnesota, 2002.
5224
Org. Lett., Vol. 8, No. 23, 2006

manipulations required for the deployment of the original
C3-C17 and C18-C30 fragments for these couplings have
prompted the exploration of alternative strategies.
Scheme 3. Partitioning of 4 between 1 and 14 via RCM
To explore an RCM-based macrolide assembly for phor-
boxazole A, our established C3-C17 bispyran 514 and C18-
C30 acid 63 building blocks (Scheme 2) could be employed
with only minor functional group modifications. First, the
C3 PMB ether of 5 was selectively cleaved with DDQ15 and
the resultant primary alcohol 7 was oxidized with the Dess-
Martin periodinane reagent16 to yield the C3 aldehyde. Wittig
olefination provided terminal alkene 8. Exposure of the latent
C16,17 vicinal amino alcohol from N-Boc-N,O-acetonide 8
was accomplished in one step by treatment with TFA in
CH2Cl2 to give 9. The secondary amine of 9 was joined with
C18 carboxylic acid 63 via EDCI-mediated amide formation17
to yield the C17 hydroxy, C16 amide 10. Optimized Dess-
Martin oxidation11,16 of the C17 alcohol gave aldehyde 11,
poised for dehydrative oxazole formation.18 Inspired by
variations19 of Crimmin’s cyclodehydration protocol,18
R-amido aldehyde 11 was subjected to a single manipulation
using halogen electrophile-activated Ph3P and Hu¨nig’s base
in CH2Cl2 to generate oxazole 12 in 80% yield from 10.11
This procedure provides a uniquely efficient conversion of
an R-N-acyl serine derivative into a 2,4-disubstituted oxazole.
Selective cleavage of the C24 TES ether of 12 was
accomplished with TBAF to yield 13. The hindered C24
hydroxyl group of 13 was acylated using acryloyl chloride
and Hu¨nig’s base to form the C24 acrylic acid ester 4.
It was expected that cross-metathesis of acrylate 4 at high
dilution would favor intramolecular C2-C3 double bond
formation due, in part, to the steric hindrance of the C7 and
the C27-C28 alkenes and conjugation of the C19-C20
alkene. RCM studies began with addition of the second-
generation Grubbs catalyst (G2, Scheme 2)9b,20 to a dilute
solution of 4 in toluene at reflux. Macrolides (2Z)-1 and (2E)-
14 were obtained in a 1:1.7 ratio and in 70% combined yield
(Table 2, entry 1). Repeating the RCM of 4 in toluene at 70
Table 2. Solvent and Temperature Effects on RCM of 4a
temp
(°C)
time
(h)
catalyst
(mol %)
yield
(%)
entry
solvent
1/14
1:1.7
1:1.2
>10:1
1
2
3
4
5
6
toluene
toluene
hexanes
pentane
CH2Cl2
CH2Cl2
110
70
70
36
40
70
0.4
1
1
3
16
3
15
25
25
25
25
25
70
∼90
65
<10
0
nd
nd
nd
nd
a Results shown represent ca. three replicate experiments for each solvent
entry. nd ) not determined.
°C for 1 h at higher catalyst loading gave an improved yield
with diminished E-selectivity (Table 2, entry 2). Attempts
to perform the RCM of 4 in CH2Cl2 at reflux gave no
detectable RCM products.
The failure of 4 to undergo cyclization in CH2Cl2 using
the G2 catalyst may reflect an important influence of solvent
Org. Lett., Vol. 8, No. 23, 2006
5225

polarity, not just temperature. To explore this hypothesis, 4
was subjected to additional RCM conditions with varying
solvent and temperature (Table 2). Repeating the reaction
of 4 in CH2Cl2 using a sealed flask at 70 °C led to little
conversion (analyzed by TLC), even after several hours and
with up to 1 equiv of catalyst. Attempted RCM of 4 in
n-pentane at a temperature and catalyst loading similar to
those used initially with CH2Cl2 yielded only minor amounts
(TLC) of macrolides after 3 h. Prolonged reaction time in
n-pentane resulted in a colored reaction solution, an observa-
tion that is consistent with catalyst degradation. In contrast,
a remarkable result was obtained when the RCM reaction
of 4 was performed in hexanes at 70 °C. A combined yield
of macrolide formation comparable to that obtained in
toluene at 110 °C was achieved after 1 h (45 min catalyst
addition plus an additional 15 min). However, the RCM of
4 in hexanes at 70 °C was highly Z-selective [>10:1, (2Z)-
unlikely to reengage in cross-metathesis (Scheme 3), undergo
equilibration under the reaction conditions, or lead to
selective degradation of one of the geometrical isomers.
Hence, the (2E/Z)-selectivities observed in the RCM products
reflect initial kinetic preferences that vary among the
conditions used.
The basis of the dramatic differences observed in kinetic
selectivities between RCM of 4 in toluene and hexanes is
not understood but may reflect the partitioning of 4 among
cis- and trans-metallocyclobutanes. Although a solvent-
dependent, substrate conformational preference may be
operative, the effect of solvent polarity on differentially
stabilizing lateral (cf. 19, Scheme 3) vs axial (15) alkene
coordination with Ru in the G2 catalyst is expected to be
large.21 Specifically, toluene would better stabilize the more
polar lateral alkene coordination (19, cis-chlorides) and
subsequent metallocyclobutane-forming transition state (cf.
20) than would hexanes. The much less polar axial coordina-
tion (cf. 15, trans-chlorides) and derived transition state (16)
would benefit less from the more polar solvent. Whether a
solvent-dependent correlation exists between lateral (20) vs
axial (16) transition states and cis- (cf. 17, 21) vs trans- (cf.
22, 18) metallocyclobutane generation with the G2 catalyst
and substrate 4 remains to be determined.22 Further optimiza-
tion of this Z-selective RCM may involve lowering the
catalyst loading, performing the reaction at higher concentra-
tion, and surveying alternative catalysts and solvents.
1
1/(2E)-14] as determined by H NMR spectroscopy. Only
trace amounts of 4 remained in the hexane reaction mixture
at 1 h. A prolonged reaction time in hexanes did not improve
the yield or change the distribution of macrolide geometrical
isomers but led to darkening of the reaction mixture.
The ratios of (2E/Z)-macrolides 1 and 14 did not change
measurably (1H NMR spectroscopy) when the isolated
mixtures were resubjected to refluxing toluene in the presence
(25 min) or absence (3 h) of the G2 catalyst. Similarly,
prolonged treatment of a 1.7:1.0 ratio of 14 to 1, respectivley,
to the G2 catalyst in hexanes at 70 °C did not appreciably
change the ratio of 14 to 1. These results indicate that once
the metallocyclobutane intermediates collapse to generate
macrolides 1 and 14 the newly generated C2 alkenes are
Acknowledgment. This research was supported by the
NIH (R01 CA099950). We thank L. Ying and Y. Lu
(University of Minnesota) for the provision of 5 and 6.
(12) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.;
Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
(13) Cink, R. D. Ph.D. Thesis, University of Minnesota, 1998.
(14) Lu, Y. M.S. Thesis, University of Minnesota, 2003.
(15) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu, O.
Tetrahedron 1986, 42, 3021.
Supporting Information Available: Experimental pro-
cedures and characterization data for 1, 4, 7, 8, 10, 12, and
14. This material is available free of charge via the Internet
at http://pubs.acs.org.
(16) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
(17) Sheehan, J. C.; Cruickshank, P. A.; Boshart, G. L. J. Org. Chem.
1961, 26, 2525.
(18) Crimmin, M. J.; O’Hanlon, P. J.; Rogers, N. H.; Sime, F. M.;
Walker, G. J. Chem. Soc., Perkin Trans. 1 1989, 2059.
(19) Wipf, P.; Lim, S. J. Am. Chem. Soc. 1995, 117, 558.
(20) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1,
953.
OL0619922
(21) (a) Benitez, D.; Goddard, W. A., III. J. Am. Chem. Soc. 2005, 127,
12218. (b) Correa, A.; Cavallo, L. J. Am. Chem. Soc. 2006, 128, 13352-
13353.
(22) Funk, T. W.; Berlin, J. M.; Grubbs, R. H. J. Am. Chem. Soc. 2006,
128, 1845.
5226
Org. Lett., Vol. 8, No. 23, 2006