Total synthesis of paecilomycine A

Article abstract of DOI:10.1002/anie.200605058

(Chemical Equation Presented) Piecing together paecilomycine: The first total synthesis of paecilomycine A, a terpenoid that enhances nerve growth factor levels in certain human cells, has been achieved. Two highly stereoselective key steps in this concis

Full text of DOI:10.1002/anie.200605058

Angewandte
Chemie
DOI: 10.1002/anie.200605058
Natural Product Synthesis
Total Synthesis of Paecilomycine A**
Sun-Joon Min and Samuel J. Danishefsky*
Our group is interested in developing, for mechanistic and
efficacy evaluations, small-molecule natural product
We considered paecilomycine A (1), particularly in the
juxtapositioning of carbon atoms C3, C12, and C15, as
reflecting a b-aldol derivative masked as the hemiacetal of a
C3 ketone, with a formal primary alcohol at C15. At the
planning level, this line of thought led us back to hypothetical
structure 9 (Scheme 1). As we anticipated potential vulner-
(
SMNP)-derived agents that are of value in the treatment of
[
1]
neurological diseases. Not being a de novo SMNP discovery
group, we rely on total synthesis to gain access to the SMNP
motif. Molecular “editing” of the parent compound through
diverted total synthesis, as opposed to conducting “post-
fact” chemistry on the SMNP itself, allows a much
broader survey of molecular space to be carried out.
One of our targets in the neurodegenerative disorder
[
2]
area is non-peptidyl agents, which can manifest the
equivalent of neurotrophic activity. For instance, we
[
3]
described the total synthesis of scabronine G and the
corroboration of its claimed neurotrophic activity
through enhancement of the expression levels of nerve
growth factor (NGF). Some promising analogues have
been synthesized as a result of that effort.
Recently, we took note of the isolation and structure
assignments of several paecilomyces tenuipes terpe-
[
4]
noids from Isoria japonica. The report described that
compound 1 (paecilomycine A), at 10 nm, is capable of
fostering neurite outgrowth in PC12 cells, apparently by
enhancing the expression levels of neurotrophic factors
in previously conditioned human astrocytoma cells.
Also noted was the claim that paecilomycine A (1) is
substantially more active than scabronine G in enhanc-
ing NGF levels. The emergence of an interesting conceptual
framework to assemble 1 in the laboratory engendered a total
synthesis program. We describe herein an inaugural total
synthesis of racemic 1 during which we gained some
potentially valuable insights pertinent to the Diels–Alder
Scheme 1. Synthetic strategy toward paecilomycine A (1).
abilities in the formal b-aldol sector within an obvious
intermediate 10, the decision was made to install the hydroxy
group at C12 fairly late in the program (see 9!10). We
entertained another retrosynthetic simplification, that is, that
9 might be derived from a thus-far unspecified alkylation
course (8!9) in which an important stereochemical deter-
minant is installed at C5. It was further conjectured that, in
[
5]
reaction owing to several early setbacks in our efforts.
[
6]
principle, an intramolecular Pauson–Khand reaction con-
ducted on substrate 7 might lead to the desired intermediate
[
*] Dr. S.-J. Min, Prof. S. J. Danishefsky
The Department of Chemistry
Columbia University
Havemeyer Hall, New York, NY 10027 (USA)
Fax: (+1)212-772-8691
8. We were not unmindful of the fact that the projected
Pauson–Khand reaction would be creating a 6:5 rather than
the more traditional 5:5 fusion motif and that it must occur
vicinally to the neopentylic C6 center. Indulging this line of
thought still further, we envisioned 7 as arising by allylation of
an alcohol precursor following appropriate stereochemical
management of the functional groups already in place in
hypothetical 4.
In studying the patterns of 6, one can hardly fail to take
note of its cyclohexenol moiety in the form of an allylic
alcohol corresponding to carbon atoms C9, C10, and C11 of
target structure 1 (with the ethereal oxygen serving as
position 1). At this stage in the planning process, the
possibility of a Diels–Alder strategy to reach 6 virtually
imposed itself. Needless to say, there would be an issue of
stereochemistry to be transacted in ensuring the required
relationship of the allylic ether and the quaternary center
E-mail: s-danishefsky@ski.mskcc.org
Prof. S. J. Danishefsky
The Laboratory for Bioorganic Chemistry
Sloan-Kettering Institute for Cancer Research
1
275 York Avenue, New York, NY 10021 (USA)
**] This work was supported by the National Institutes of Health
HL25848). We thank Prof. Yoshiteru Oshima (Tohoku University,
[
(
Japan) for kindly providing the sample of paecilomycine A. We also
thank Dr. Louis J. Toldaro (Hunter College, New York) and Daniella
Buccella (Columbia University) for crystal structure analysis, and
Rebecca Wilson for assistance with the preparation of the manu-
script.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 2199 –2202
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2199

Communications
(
corresponding to the eventual C6 center).
These thoughts are presented in a forward
direction in Scheme 1. While the gestalt implied
in Scheme 1 seemed appealingly concise, some
significant difficulties arose in the more
demanding world of experimental reality
whose solution, in the end, enhanced the
teaching dimension of the experience.
Focusing on the generic potential Pauson–
Khand substrate 7, our initial bias was that the
alkynyl group would be fashioned from an
aldehyde. This avenue of thought led back to
Scheme 2. Reagents and conditions: a) 1) Toluene, 1458C, 48 h; 2) TBAF, THF,
10 min, 88% (13a/14a=2:1) or benzene, 808C, 22 h, 87% (13b/14b=6:1); b) allyl
bromide, NaH, TBAI, HMPA, THF, 08C to 238C, 85%; c) DIBAL, CH Cl , 08C, 100%
6. If, however, 6 were to arise from a Diels–
Alder reaction of diene 2 and 3, an obvious
problem presents itself. Thus, the aldehyde
function in 6 is stereochemically situated such
as to require an exo-selective cycloaddition
reaction from an E-configured generic 2.
Given the difficulties in finding highly activat-
ing Diels–Alder groups which are also exo-
selective, it was decided to derive the protected
hydroxymethyl function at C6 from what was
2
2
(
from 13a); d) NaClO , NaH PO ·H O, tBuOH, H O, 2-methyl-2-butene, 99%; e) allyl
2 2 4 2 2
bromide, NaH, TBAI, HMPA, THF, 61% (+8% diastereomer); f) DIBAL, CH Cl , 08C,
2
2
1
00% (from 13b); g) TBSCl, imid, CH Cl , 100%; h) DDQ, CH Cl , 84%; i) DMSO,
2
2
2
2
(
COCl) , Et N, CH Cl , ꢀ788C to 08C, 90%; j) 17, K CO , MeOH, 85%. TMS=tri-
2
3
2
2
2
3
methylsilyl; PMB=p-methoxybenzyl; TBAF=tetra-n-butylammonium fluoride; TBAI=
tetra-n-butylammonium iodide; HMPA=hexamethylphosphoramide; DIBAL=diiso-
butylaluminum hydride; TBS=tert-butyldimethylsilyl; imid=imidazole; DDQ=2,3-
dichloro-5,6-dicyanobenzoquinone; DMSO=dimethyl sulfoxide.
the endo-selective functional group in 3. Conversely, the
aldehyde functional group identified for eventual conversion
into the acetylene group required for the Pauson–Khand
reaction would have actually been derived from the exo-
directing function in 3. Thus, the emerging plan envisioned a
(14b) products. The former product was converted into
[
12]
alcohol 15 as shown (Scheme 2).
Thus, perhaps not
surprisingly, the more reactive aldehydo dienophile exhibited
greater endo:exo selectivity owing to the activating aldehyde
function (see generic structure 3, Scheme 1).
“
role reversal” in progressing from the opening Diels–Alder
Seeking still greater endo:exo selectivity and a higher
[
13]
phase toward intermediate 6 required for the all-critical
level of convergence, we examined compound 19a
as a
Pauson–Khand strategy. We anticipated that, in essence, the
potential dienophile. In the event, an approximately 10:1 ratio
of endo (20a) to exo (21a) products was obtained (Scheme 3).
Remarkably, the resident ethynyl function had conferred far
higher endo selectivity upon the ester group than was the case
with dienophile 12a. Once again, the cycloadduct could
readily be converged, as shown, with previously encountered
intermediates. Finally, in this regard, even higher endo:exo
stereoselectivity and ultimate synthetic convergence followed
[
7]
umpolung-like reversal could be accomplished by redox-
type adjustment of functional groups to reach 6. The
acetylene function in 7 would have been derived from the
aldehyde group of 6 which, following the role reversal, would
have been derived from the exo-directing function of 3.
[
8]
We first studied the Diels–Alder reaction of 11 and
[
9]
1
2a, each a known compound, under strictly thermal
[
14]
activation (Scheme 2). When the cycloaddition reaction was
conducted in toluene at approximately 1408C, and then
followed by deprotection, an 88% yield of a 2:1 mixture of
from the reaction of aldehyde 19b with diene 11, leading to
[
15]
20b with greater than 20:1 selectivity. The improvement in
endo:exo selectivity in the presence of the acetylenic group
provided an unexpected dividend. Note also that the con-
ditions for effective non-catalyzed cycloadditions with 11
1
3a and 14a was obtained. It was decided for the moment to
use this quick and processible route—albeit of uninspiring
selectivity—to provide material for exploration. We were
hopeful that if the overall route would be shown to have
merit, means would be found to improve upon the poor
Diels–Alder endo/exo ratio. O-alkylation of 13a proceeded
smoothly as shown to afford the allyl ether (Scheme 2).
Reduction of the ester group allowed for access to 15, which
was protected as a silyl ether. Deprotection of the PMB group
and oxidation of the resultant alcohol to aldehyde was
accomplished to give compound 16. The latter compound
proved to be a competent substrate for a Gilbert ethynyla-
[
10]
tion to afford the milestone compound 18, on which the
critical Pauson–Khand reaction could be undertaken.
Before we disclose the route by which paecilomycine A
Scheme 3. Reagents and conditions: a) 11, CH Cl , 238C, 3 h, 86%
2
2
(
endo/exo>10:1) or 11, CH Cl , 08C, 0.5 h, 58% (endo/exo>20:1);
2 2
(1) was reached from compound 18, we describe other routes
b) LiAlH , THF, 08C; c) K CO , MeOH, (85%, two steps from 20a);
4
2
3
that are considerably more stereoselective in reaching this
d) NaBH , MeOH, 08C; e) K CO , MeOH, (72% two steps from 20b);
4
2
3
key intermediate. In this vein, cycloaddition of diene 11 with
f) TBSCl, imid, CH Cl , 89%; g) allyl bromide, NaH, DMF, 70%.
2
2
[
11]
aldehyde 12b afforded a 6:1 ratio of endo (13b) to exo
DMF=N,N-dimethylformamide.
2
200 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2199 –2202

Angewandte
Chemie
were simplified from 1458C in the case of 12a to 08C in the
case of 19b.
With excellent routes to 18 well in hand, it was subjected
to intramolecular Pauson–Khand reaction at 1008C under the
[
16]
conditions shown (Scheme 4). A single stereoisomer, 23,
was obtained in 37% yield. While the yield of the Pauson–
Khand reaction was disappointingly modest, the reaction was
scalable and the product could be readily purified and
advanced to reach paecelomycine A (see below).
One obvious possible pathway to advance from 23 to 1
was frustrated by our inability to achieve under any con-
ditions conjugate nucleophilic methylation by using cuprate
reagents. Momentary encouragement arose when it was found
Scheme 5. Reagents and conditions: a) NaBH
, CeCl ·7H O, MeOH,
3 2
4
8
6% (8% a-alcohol); b) nBuLi, Et O, ꢀ208C and then Et Zn, CH I ,
2
2
2 2
benzene, 53% (11% 26 + 4% bis-cyclopropane); c) TPAP, NMO,
4
-Š M.S., CH Cl , 93%; d) Li, NH , THF, ꢀ338C, 65% (75% brsm).
[
17]
2
2
3
that the corresponding Nagata conjugate cyanation could
TPAP=tetra-n-propylammonium perruthenate; NMO=N-methylmor-
pholine N-oxide; brsm=based on recovered starting material.
be accomplished in 63% yield (see compound 24). However,
[
18]
X-ray crystallographic analysis, conducted on its derived
(Scheme 6). Ketone 28 was converted into
enone 29 by dehydrogenation of the silyl enol
[
22]
ether according to the protocol of Itoh et al.
Nucleophilic epoxidation of the resulting enone
using hydrogen peroxide and sodium hydrox-
Scheme 4. Reagents and conditions: a) [Co (CO) ], 4-Š M.S., toluene, 238C, 2 h, then
2
8
[
23]
ide occurred syn to the angular methyl group
see epoxyketone 30, Scheme 6). The stereo-
1
008C, 24 h, 37%; b) Et AlCN, toluene, 63%; c) NaBH , MeOH, 80%.
2 4
(
alcohol 25 revealed that the cyano group was in the a-
position, that is, syn to the emerging hydroxymethyl group at
the incipient C6 center and the hydrogen atom at C12 (see
compound 25, Scheme 4). While the structure of the Pauson–
Khand product 23 had in effect been validated, the stereo-
chemical outcome of the somewhat surprising Nagata reac-
tion seemed to be non-remediable.
To solve the problem of introducing a b-methyl group at
C5, we took advantage of an important lesson to be gleaned
[
19]
from the work of Corey and Virgil a decade earlier. The
logic of their method exploits local stereochemical biases to
reduce a cyclic enone to a particular allylic alcohol. The
alcohol then directs stereospecific cyclopropanation in the syn
sense. In the case at hand, it was applied as follows: The
Pauson–Khand enone 23 was reduced under Luche condi-
Scheme 6. Reagents and conditions: a) 1) LiTMP, TMSCl, THF, ꢀ78!
8C; 2) Pd(OAc) , CH CN, 68% (71% brsm); b) aq. H O , 6n NaOH,
0
2
3
2
2
MeOH, 08C, 83% (89% brsm); c) Li, NH , THF, ꢀ338C, 63% (24%
3
29); d) HF/CH CN (5:95), 86%. TMP=2,2,6,6-tetramethylpiperidine.
3
[
20]
tions
to provide 26, which underwent cyclopropanation
[
21]
with the in situ generated Furukawa reagent, prepared by
addition of diiodomethane to diethylzinc. Under the precisely
defined conditions shown (Scheme 5), the desired cyclopro-
pane 27 was obtained along with a small quantity of the
starting alcohol 26 and a bis-cyclopropane. Oxidation of the
hydroxy group in 27, followed by dissolving metal reduction
of the resultant activated cyclopropane, afforded the cyclo-
pentanone 28 in good yield. The stereochemistry of the
angular methyl group was elucidated at this stage by a
NOESY experiment, in which we observed NOE enhance-
ment signals between the C14-methyl hydrogen atom and five
neighboring hydrogen atoms. The crystal structure eventually
obtained on paecilomycine A confirmed the soundness of this
analysis (see below).
chemical outcome of this reaction was tentatively assigned on
the basis of the assumption that a cis ring juncture would be
preferred in a [4.3.0]bicyclic ring system. Many attempts to
open the epoxide ring in 30 failed due to the susceptibility of
its derived b-hydroxyketone 31. Fortunately, dissolving metal
reduction of 30 did provide 31 along with a small amount of
enone 29. Cleavage of the silyl protecting group was also
problematic. Thus, attempted deprotection of 31 under basic
or acidic conditions using TBAF, trifluoroacetic acid, pyridi-
[
24]
nium para-toluenesulfonate, AcOH, and HF/pyridine,
resulted in the formation of the intramolecular Michael
product 32. In the end, the TBS protecting group was
successfully removed through the use of hydrofluoric acid in
[
25]
Completion of the synthesis of paecilomycine A required
introduction of a tertiary alcohol at C12 and removal of the
silyl protecting group, and ring closure to the hemiacetal
acetonitrile, whereupon the resulting hydroxyketone spon-
taneously cyclized to furnish (ꢁ )-paecilomycine A (1;
Figure 1) in 86% yield. The spectral data obtained with our
Angew. Chem. Int. Ed. 2007, 46, 2199 –2202
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2201

Communications
[
6]For a recent review of the Pauson–Khand reaction, see: a) K. M.
Brummond, J. L. Kent, Tetrahedron 2000, 56, 3263. For a
relevant analogy, see: b) T. F. Jamison, S. Shambayati, W. E.
Crowe, S. L. Schreiber, J. Am. Chem. Soc. 1997, 119, 4353.
7]D. Seebach, Angew. Chem. 1979, 91, 259; Angew. Chem. Int. Ed.
Engl. 1979, 18, 239.
[
[
[
8]E. W. Colvin, I. G. Thom, Tetrahedron 1986, 42, 3137.
9]Y. Okamoto, S. Habagami, T. Nakano, Jpn. Kokai Tokkyo Koho,
JP 2004027207, 2004. However, we prepared 12a from ethyl a-
(hydroxymethyl)acrylate (H.-S. Byun, K. C. Reddy, R. Bittman,
Tetrahedron Lett. 1994, 35, 1371) by protection with the PMB
group (PMB-trichloroacetimidate, camphorsulfonic acid,
CH Cl , 97%).
2
2
[
10]a) S. Ohira, Synth. Commun. 1989, 19, 561; b) S. Müller, B.
Liepold, G. J. Roth, H. J. Bestmann, Synlett 1996, 521; c) D.
Seyferth, R. S. Marmor, P. Hibert, J. Org. Chem. 1971, 36, 1379;
d) J. C. Gilbert, U. Weerasooriya, J. Org. Chem. 1979, 44, 4997.
11]We easily prepared the aldehyde 12b from reduction of 12a
Figure 1. ORTEP representation of paecilomycine A (1) with ellipsoids
shown at the 20% probability level.
[
2 2
(DIBAL, CH Cl , 87%), followed by oxidation (Dess–Martin
periodinane, CH Cl , 93%).
2
2
synthetic material were congruent with those obtained from a
sample kindly provided by Professor Yoshiteru Oshima
[
12]In contrast to the case of 13a, attempted deprotection of 13b
leads to further equilibration with 14b through a retro-aldol
pathway. We dealt with this complication by inserting an
oxidation step prior to deprotection in the manner shown
(
Tohoku University, Japan). The entire set of structured
arguments was further validated by a crystallographic inves-
[
18]
(Scheme 2).
tigation of fully synthetic 1.
[
[
13]C. Spino, J. Crawford, J. Bishop, J. Org. Chem. 1995, 60, 844.
14]C. Thongsornkleeb, R. L. Danheiser, J. Org. Chem. 2005, 70,
In summary, the goal of a total synthesis of paecilomyci-
ne A has been accomplished in a highly stereoselective
manner. While the yield of the Pauson–Khand reaction in
the case of substrate 18 was not unexpectedly modest, the
reaction served such a powerful enabling role that, for the
moment, its uninspiring yield is survivable. Also capable of
remediation, in principle, is that the route leads to racemic 1.
These two chemical matters are being addressed even as the
primary emphasis of the project now shifts to the evaluation
of the utility, the mechanistic basis, and the potential
applicability of the neurotrophic-like action of paecilomy-
cine A.
2
364.
15]The NMR spectra showed some indication of a minor product
< 5%) in the reaction of 19b and 11. This product has not yet
been identified as the exo compound.
[
(
[16]a) L. PØrez-Serrano, L. Casarrubios, G. Dominguez, J. PØrez-
Castells, Org. Lett. 1999, 1, 1187; b) M. Ishizaki, K. Iwahara, K.
Kyoumura, O. Hoshino, Synlett 1999, 587; c) J. Castro, A.
Moyano, M. A. Pericàs, A. Riera, A. E. Greene, Tetrahedron:
Asymmetry 1994, 5, 307.
17]a) W. Nagata, M. Yoshioka, S. Hirai, J. Am. Chem. Soc. 1972, 94,
4635; b) W. Nagata, Org. React. 1977, 25, 255; c) L. E. Overman,
D. J. Ricca, V. D. Tran, J. Am. Chem. Soc. 1997, 119, 12031.
18]CCDC 630741 ( 25) and 630742 (1) contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[
[
Received: December 14, 2006
Published online: February 15, 2007
[
19]a) E. J. Corey, S. C. Virgil, J. Am. Chem. Soc. 1990, 112, 6429. See
also: b) S. D. Guile, J. E. Saxton, M. Thornton-Pett, J. Chem.
Soc. Perkin Trans. 1 1992, 1763.
20]J. L. Luche, J. Am. Chem. Soc. 1978, 100, 2226.
21]a) J. Furukawa, N. Kawabata, J. Nishimura, Tetrahedron Lett.
Keywords: Diels–Alder reaction · natural products ·
Pauson–Khand reaction · total synthesis
.
[
[
[
[
1]R. M. Wilson, S. J. Danishefsky, Acc. Chem. Res. 2006, 39, 539.
2]a) D. Dawbarn, S. J. Allen, Neuropathol. Appl. Neurobiol. 2003,
1966, 7, 3353; b) J. Furukawa, N. Kawabata, J. Nishimura,
Tetrahedron 1968, 24, 53.
29, 211; b) D. Kirik, B. Georgievska, A. Bjorklund, Nat. Neuro-
[
[
22]Y. Itoh, T. Hirao, T. Saegusa, J. Org. Chem. 1978, 43, 1011.
23]M. Miyashita, T. Suzuki, A. Yoshikoshi, J. Am. Chem. Soc. 1989,
sci. 2004, 7, 105.
[
[
[
3]S. P. Waters, Y. Tian, Y.-M. Li, S. J. Danishefsky, J. Am. Chem.
111, 3728.
Soc. 2005, 127, 13514.
[
24]a) C. Prakash, S. Saleh, I. A. Blair, Tetrahedron Lett. 1989, 30, 19;
b) M. J. Robins, V. Samano, M. D. Johnson, J. Org. Chem. 1990,
4]H. Kikuchi, Y. Miyagawa, Y. Sahashi, S. Inatomi, A. Haganuma,
N. Nakahata, Y. Oshima, Tetrahedron Lett. 2004, 45, 6225.
5]a) J. G. Martin, R. K. Hill, Chem. Rev. 1961, 61, 537; b) F.
Fringuelli, A. Taticchi, Organic Reactions: Dienes in the Diels–
Alder Reaction, Wiley, New York, 1990; c) Y. Kobuke, T. Fueno,
J. Furukawa, J. Am. Chem. Soc. 1970, 92, 6548; d) T. Inukai, T.
Kojima, J. Org. Chem. 1966, 31, 2032; e) J. Sauer, H. Wiest, A.
Mielert, Chem. Ber. 1964, 97, 3183.
55, 410; c) E. Carreira, J. Du Bois, J. Am. Chem. Soc. 1995, 117,
8106.
[
25]R. F. Newton, D. P. Reynolds, M. A. W. Finch, D. R. Kelly, S. M.
Roberts, Tetrahedron Lett. 1979, 20, 3981.
2
202 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2199 –2202