Studies toward the Synthesis of Oximidines I and II

Article abstract of DOI:10.1021/ol036007d

(Equation presented) The synthesis of compound 1, a precursor for the synthesis of the oximidine II core structure 2, is described. An undesired C8-C9 isomerization occurred during the intramolecular Castro-Stephens reaction leading to macrocyle 3. The th

Full text of DOI:10.1021/ol036007d

ORGANIC
LETTERS
2003
Vol. 5, No. 26
5019-5022
Studies toward the Synthesis of
Oximidines I and II
,†
,§
Torsten Haack,§ Serdar Kurtkaya,† James P. Snyder,* and Gunda I. Georg*
Department of Medicinal Chemistry, UniVersity of Kansas, 1251 Wescoe Hall DriVe,
Lawrence, Kansas 66045-7582, and Department of Chemistry, Emory UniVersity,
1515 Dickey DriVe, Atlanta, Georgia 30322
georg@ku.edu
Received October 14, 2003
ABSTRACT
The synthesis of compound 1, a precursor for the synthesis of the oximidine II core structure 2, is described. An undesired C8−C9 isomerization
occurred during the intramolecular Castro−Stephens reaction leading to macrocyle 3. The thermodynamic driving force for this unexpected
isomerization was established by DFT and MP2 calculations.
The oximidines I and II1 (4 and 5, Figure 1) belong to a
family of recently discovered macrolides, which contain a
mon biological targets of these natural products have been
identified as mammalian vacuolar-type (H+)-ATPases.3 The
possibility of forming structural analogues to gain more
information about the mechanism of action has inspired
researchers to develop total syntheses toward these com-
pounds. The first total synthesis of the oximidines I and II,
isolated from Pseudomonas sp. in 1999,1 was reported in
2003.4 Further synthetic efforts have been limited to model
studies for the formation and the installation of the enamide
side chain5 and to the preparation of simplified macrocyclic
core structures.6-8 We are now reporting on our approach
toward a synthesis of the fully functionalized oximidines I
and II.
Assuming that it might be possible to transform the C12-
C13 double bond in oximidine II (5) into the corresponding
Figure 1. Oximidines I (4) and II (5).
(2) For a review see: Yet, L. Chem. ReV. 2003, 103, 4283-4306.
(3) Boyd, M. R.; Farina, C.; Belfiore, P.; Gagliardi, S.; Kim, J. W.;
Hayakawa, Y.; Beutler, J. A.; McKee, T. C.; Bowman, B. J.; Bowman, E.
J. J. Pharmacol. Exp. Ther. 2001, 297, 114-120.
salicylic acid moiety and an unusual enamide side chain.
This family of natural products also includes the salicyli-
halamides, the apicularens, and the lobatamides.2 The com-
(4) Wang, R.; Porco, J. A., Jr. J. Am. Chem. Soc. 2003, 125, 6040-
6041.
§ University of Kansas.
(5) (a) Shen, R.; Porco, J. A., Jr. Org. Lett. 2000, 2, 1333-1336. (b)
Raw, S. A.; Taylor, R. J. K. Tetrahedron Lett. 2000, 41, 10357-10361.
(c) Snider, B. B.; Song, F. Org. Lett. 2000, 2, 407-408. (d) Kuramochi,
K.; Watanabe, H.; Kitahara, T. Synlett 2000, 397-399.
† Emory University.
(1) Kim, J. W.; Shin-ya, K.; Furihata, K.; Hayakawa, Y.; Seto, H. J.
Org. Chem. 1999, 64, 153-155.
10.1021/ol036007d CCC: $25.00 © 2003 American Chemical Society
Published on Web 12/04/2003

epoxide leading to oximidine I (4), we focused our efforts
on a total synthesis of oximidine II (5). The main challenge
in a total synthesis of 5 is the formation of the strained 12-
membered macrocycle containing nine sp2 carbon centers.
Our retrosynthetic disconnections are set at the ester bond,
at the enamide between carbons C17 and C18, and between
carbons C9 and C10 leading to fragments 6, 7, and 8
(Scheme 1).
process by using an esterification and a Takai-olefination.
A 1:4 cis:trans mixture of 13 was obtained, which was
separable by silica gel chromatography.
We chose a chiral-pool strategy for the preparation of the
aliphatic building block. Compound 8 can be obtained from
protected 2-deoxysugar 14, which is a derivative of L-xylose
(Scheme 3).
Scheme 3
Scheme 1
Utilizing a modified literature procedure11,12 L-xylose was
transformed into known alcohol 15 in 70% yield over four
steps. The alcohol functionality was protected as pivaloate
ester and removal of the acetonide in 70% acetic acid at room
temperature provided diol 16. A three-step sequence involv-
ing TBS-protection of the primary alcohol in 16, MOM
protection of the secondary alcohol followed by removal of
the silyl group, provided the primary alcohol 17 in 75%
Since the enamide side chain is presumably chemically
unstable, we aimed first at the preparation of the macrocycle
from fragments 7 and 8. The subsequent installation of the
side chain with amide 6 requires an appropriate elongation
of the protected aldehyde in 8. For the preparation of the
1,2,3-trisubstituted aromatic fragment 13 we selected the
synthetic route shown in Scheme 2. The Diels-Alder
reaction9,10 between alkyne 9 and diene 10 forms the desired
salicylate 11 in 80% yield.
Scheme 4
Scheme 2
overall yield. Dess-Martin oxidation of the primary alcohol
(Scheme 5) furnished the unstable corresponding aldehyde,
which was immediately subjected to a Corey-Fuchs reaction.
The resulting dibromoalkene was stereoselectively debro-
minated to form cis-vinyl bromide 18 in 71% overall yield.
Removal of the pivaloate with DIBAL-H, Sonogashira
(9) Fresko, J. N.; Morrow, G. W.; Swenton, J. S. J. Org. Chem. 1985,
50, 805-810.
(10) For other recent Diels-Alder approaches toward related natural
products, see: (a) Yang, Z.-Q.; Danishefsky, S. J. J. Am. Chem. Soc. 2003,
125, 10636-10637. (b) Graetz, B. R.; Rychnovsky, S. D. Org. Lett. 2003,
5, 3357-3360.
(11) Kim, J. Y.; Nam Shin, J. E.; Chun, K. H. Bull. Korean Chem. Soc.
1996, 17, 478-480.
(12) Wong, M. Y. H.; Gray, G. R. J. Am. Chem. Soc. 1978, 100, 3548-
3553.
Acidic hydrolysis9 of 11 yielded isobenzofuranone 12,
which was transformed to the trans iodide 13 by a two-step
(6) Scheufler, F.; Maier, M. E. Synlett 2001, 1221-1224.
(7) Coleman, R. S.; Garg, R. Org. Lett. 2001, 3, 3487-3490.
(8) Harvey, J. E.; Raw, S. A.; Taylor, R. J. K. Tetrahedron Lett. 2003,
44, 7209-7212.
5020
Org. Lett., Vol. 5, No. 26, 2003

1 in low yield. Compound 1 is a fully functionalized
precursor for an intramolecular Castro-Stephens reaction.7
Reaction of 1 under Castro-Stephens conditions yielded
macrocycle 3 that arises from the desired intramolecular sp-
sp2 coupling. Generally, this reaction is stereospecific with
respect to the precursor vinyliodide,13 and in the present case
a trans C8-C9 double bond was expected. However, NMR
Scheme 5
3
analysis yielded a coupling constant of J ) 12 Hz for the
double bond, a difference of 4 Hz by comparison with the
results from Coleman and Garg,7 who reported a value of 3J
) 16 Hz for trans-configured C8-C9 centers in several
model systems.
To put these values in perspective, we calculated Z and E
vicinal 3J(H,H) values across C8-C9 for the global minimum
energy conformations (see below) of model systems 21 and
22 (Scheme 7). The coupling constants were obtained by
computing the four contributing terms according to Ramsey’s
non-relativistic approach14 in the context of the B3LYP/6-
311G(d,p) method.15
elongation, and subsequent deprotection of the alkyne under
mild nucleophilic conditions gave the desired aliphatic
building block 19.
With the two building blocks 13 and 19 in hand, our next
aim was the formation of the macrocyclic core structure of
oximidine II. Our strategy was the saponification of 13,
esterification of resulting 20 with 19, and a macrocyclization
under Castro-Stephens conditions.7 Ester hydrolysis of 13
was achieved with TFA/CH2Cl2 to give acid 20 in 77% yield
(Scheme 6). The subsequent esterification with 19 was
difficult because of the low reactivity of the acid moiety due
to steric and electronic reasons. However, following the
Yamaguchi procedure we were able to form the desired ester
Scheme 7
Scheme 6
3
Summing the Ramsey contributions furnishes J(H,H)
values of 14.2 and 20.4 Hz, respectively (Table 1). Since
3
the predicted J(H,H) values are slightly overestimated by
comparison to the experimental values, we scaled the
3
predictions by employing the Z and E J values of acrylo-
nitrile as standards (∆3J ) -2.4 and -4.5 Hz, respectively;
Table 1). In this way, the values for 21 and 22 are reduced
to 11.8 and 15.9 Hz, respectively, in complete accord with
the experimental values of 12 and 16 Hz. To test whether
the aromatic methoxy substituent in 3 might lower the E-3J
from 16 to 12 Hz, truncated forms of 22 and 3 (27 and 28,
respectively; see Supporting Information) were subjected to
coupling constant analysis (Table 1). The methoxy perturba-
3
tion has no influence on the vicinal J value.
We conclude that compound 3 is the Z,Z isomer. A strong
NOE between the protons at C8 and C9 supports this
deduction. Presumably, isomerization took place from the
(13) (a) Kabbara, J.; Hoffmann, C.; Schinzer, D. Synthesis 1995, 299-
302. (b) von der Ohe, F.; Bruckner, R. New J. Chem. 2000, 24, 659-669.
(c) Rawat, D. S.; Zaleski, J. M. Synth. Commun. 2002, 32, 1489-1494.
(14) Ramsey, N. F. Phys. ReV. 1953, 91, 303-307.
(15) For definitions of the various terms and the method of calculation,
see: Barone V.; Peralta, J. E.; Contreras, R. H.; Snyder, J. P. J. Phys. Chem.
A 2002, 106, 5607-5612.
Org. Lett., Vol. 5, No. 26, 2003
5021

predicted to be more stable by 11-16 kcal/mol, highlighting
the thermodynamic basis for isomerization.
3
Table 1. Ramsey Contributions and J(H,H)total (Hz) for
C8-C9 with the DFT B3LYP/6-311G(d,p) Model
Reduction of the C12-C13 double bond (i.e. 23 and 24)
is predicted to lower the energy difference, but to still favor
the Z-form 23 by 7-11 kcal/mol. On the other hand, the
calculations suggest that reduction of the acetylene to a cis
double bond (i.e. 25 and 26)7 reverses the stability order to
favor 26-E,Z,Z by 2-5 kcal/mol. These results imply that
the presence of the acetylene unit at C10-C11 is decisive
for determining the relative stability of the C8-C9 double
bond isomers; namely, the presence of the acetylene unit in
the 12-membered ring stabilizes the C8-C9 Z-alkene. That
this carries over to 3 was shown by substituent supplementa-
tion of the MMFF global minima of 21 and 22 followed by
MMFF optimization to give 3 in a twisted conformation and
the corresponding trans isomer. The former is favored by
15.3 kcal/mol (∆Erel, MMFF).
DSO
PSO
FC
SD
total
21
22
Z-Aa,b
E-Aa,b
27
-0.002
-2.74
-0.76
-3.30
-2.84
-2.81
-0.57
3.02
0.072
4.77
3.05
4.39
14.78
19.72
14.89
20.56
19.64
19.67
-0.052
0.39
-0.043
0.33
0.40
0.40
14.2
20.4
14.2c
22.4c
20.2
21.7
28
a 3J(H,H) for acrylonitrile Z-A and E-A are 11.8 and 17.9 Hz,
respectively: (i) http://chemistry.utah.edu/faculty/vogler/LectureNotes331/
CH331Chapter13.pdf; (ii) Lambert, J. B.; Shurvell, H. F.; Lightner, D. A.;
Cooks, R. G. In Organic Structural Spectroscopy; Prentice Hall: Englewood
Cliffs, NJ, 1998; p 72. b Geometry optimized with the B3LYP/6-311G(d,p)
c
model. 3J scaling factors from Z and E acrylonitrile are -2.4 and -4.5
Hz, respectively.
E to a Z double bond under the reaction conditions. The
reasonableness of these assertions was evaluated by perform-
ing exhaustive conformational analyses on model structures
21-Z,Z and 22-E,Z as well as on the corresponding dihydro
analogues 23-26. Table 2 summarizes the results and
includes both molecular mechanics and quantum chemical
relative energies for the C8-C9 configurational isomers. For
21 and 22 with three unsaturated centers, the Z,Z form is
In summary, we have developed a convenient synthesis
of an aromatic and an aliphatic building block of the fully
functionalized macrocycle of oximidine II. We have also
combined these building blocks to the desired core structure,
albeit with the Z-configured C8-C9 double bond. In
addition, DFT and MP2 calculations were used to determine
the thermodynamic driving force for this isomerization.
Further investigations toward the use of the building blocks
for the synthesis of the oximidines are currently in progress.
Table 2. Relative Energies of Annelated and Unsaturated
Acknowledgment. We are grateful for financial support
from the NIH (CA 84173). This work was also supported
by a fellowship (T.H.) within the Postdoc-Program of the
German Academic Exchange Service (DAAD). S.K. and
J.P.S. are grateful to Prof. Dennis Liotta (Emory University)
for encouragement and support.
12-Membered-Ring Lactones (kcal/mol)a
∆E(rel), kcal/mol
MMFFb
DFTc
MP2d
21
22
23
24
25
26
0.0
11.4
0.0
10.9
4.8
0.0
14.4
0.0
7.5
2.6
0.0
15.7
0.0
9.0
2.4
Supporting Information Available: Experimental pro-
cedures including physical data for all synthesized com-
0.0
0.0
0.0
3
pounds; Ramsey term contributions to J(H,H) and compu-
a Only MMFF global minima derived by Monte Carlo conformer
searching are compared. b Molecular mechanics/MMFF energies. c B3LYP/
6-311G(d,p)//MMFF. d MP2/6-311G(d,p)//MMFF (see Supporting Informa-
tion).
tational details. This material is available free of charge via
the Internet at http://pubs.acs.org.
OL036007D
5022
Org. Lett., Vol. 5, No. 26, 2003