Determination of absolute structure of macroviracins by chemical synthesis

Article abstract of DOI:10.1021/ol034338k

(Matrix presented) The relative and absolute configurations of macroviracins have been established by the stereocontrolled synthesis of methyl ester 2a of the C₂₂ carboxylic acid, a constitutive fatty acid of macroviracin A (1), and their compa

Full text of DOI:10.1021/ol034338k

ORGANIC
LETTERS
2003
Vol. 5, No. 9
1555-1558
Determination of Absolute Structure of
Macroviracins by Chemical Synthesis
Shunya Takahashi,* Moritoshi Hosoya, Hiroyuki Koshino, and Tadashi Nakata
RIKEN (The Institute of Physical and Chemical Research),
Wako, Saitama 351-0198, Japan
shunyat@postman.riken.go.jp
Received February 26, 2003
ABSTRACT
The relative and absolute configurations of macroviracins have been established by the stereocontrolled synthesis of methyl ester 2a of the
22 carboxylic acid, a constitutive fatty acid of macroviracin A (1), and their comparison to a sample 2 derived from the natural product 1.
C
Macroviracins are a new class of macrocyclic compounds,
which were isolated from the mycelium extracts of Strept-
myces sp. BA-2836,1 and classified into eight types of
congeners (A-D) by the length (C22 or C24) of the constitu-
tive fatty acids. These natural products exhibit a powerful
antiviral activity against HSV-1 and VZV, and their potency
is reported to be 10 times that of acyclovir. Structurally,
macroviracins are related to sugar-fatty acid lactones such
as cycloviracins2 and fattiviracins,3 differing remarkably in
the size of the macrocyclic ring systems. Degradation studies
of macroviracins suggested the corresponding constitutive
acids to have the same relative stereochemistry. The absolute
configuration, however, remained unsolved. Furthermore, the
absolute configuration (D-form) of the sugar residue also has
not been established. In this paper, we describe the total
synthesis of methyl ester 2a of fatty acid constituted
macroviracin A (1) and determination of the absolute
configuration of macroviracins. Very recently, Fu¨rstner et
al. disclosed the total synthesis of cycloviracin B1, establish-
ing the absolute configuration.4
As methanolysis of 1 has been known to give the
corresponding methyl ester in degradation studies, the
benzoyl5 derivative 2 prepared therefrom was regarded as
an ideal synthetic target for this research. Our synthetic
strategy toward 2 is illustrated in Scheme 1. Thus, removal
of two sugar residues from 2 and depression of its oxidation
level leads to acyclic diol 3. This would be synthesized
through a coupling reaction of a nucleophile derived from 4
and terminal epoxide 5. As an alternate plan, we designed
two types of glucosides (6 and 7) as the left- and right-half
(1) Hyodo, T.; Tsuchiya, Y.; Sekine, A.; Amano, T. Jpn. Kokai Tokkyo
Koho Jpn. Pat. 11246587, Sept 14, 1999.
(2) Tsunakawa, M.; Komiyama, N.; Tenmyo, O.; Tomita, K.; Kawano,
K.; Kotake, C.; Konishi, M.; Oki, T. J. Antibiot. 1992, 45, 1467-1471.
Tsunakawa, M.; Kotake, C.; Yamasaki, T.; Moriyama, T.; Konishi, M.;
Oki, T. J. Antibiot. 1992, 45, 1472-1480.
(3) Uyeda, M.; Yokomizo, K.; Miyamoto, Y.; Habib, E.-S. E. J. Antibiot.
1998, 51, 823-828. Yokomizo, K.; Miyamoto, Y.; Nagao, K.; Kumagae,
E.; Habib, E.-S. E.; Suzuki, K.; Harada, S.; Uyede, M. J. Antibiot. 1998,
51, 1035-1039. Habib, E.-S. E.; Yokomizo, K.; Murata, K.; Uyeda, M. J.
Antibiot. 2000, 53 1420-1423.
(4) Fu¨rstner, A.; Albert, M.; Mlynarski, J.; Matheu, M. J. Am. Chem.
Soc. 2002, 124, 1168-1169. Fu¨rstner, A.; Mlynarski, J.; Albert, M. J. Am.
Chem. Soc. 2002, 124, 10274-10275.
(5) For a method for determination of the absolute configuration of
secondary alcohols by using benzoylated sugars, see: Trujillo, M.; Morales,
E. Q.; Va´zquez, J. T. J. Org. Chem. 1994, 59, 6637-6642.
10.1021/ol034338k CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/05/2003

Scheme 1. Structures of Macroviracins and Retrosynthetic Analysis
1
model of 2, respectively. Prior to the synthesis of 2, syntheses
of 6 and 7 and comparison of their spectral data with those
of 2 were conducted.
(dd, J ) 16, 8.3 Hz, H-2b), 3.26 (s, OMe)}. The H NMR
spectra of the model compounds (6a and 6b) were compared
with that of 2. The methyl ester 2 exhibited δΗ 2.45 (dd, J
) 16, 6.8 Hz, H-2a), 2.88 (dd, J ) 16, 5.9 Hz, H-2b), and
3.59 (s, OMe), which resembled the data of 6a. These results
show that the stereochemistry of the C-3 position in 2 should
be R if the sugar moiety is a D-form.
We began with the synthesis of the left-half models (6a,b).
10-Undecenal 8 was subjected to Wittig reaction, and
subsequent DIBAL reduction afforded allyl alcohol 9
(Scheme 2). The alcohol 9 was oxidized by Sharpless
epoxidation to give epoxide 10. Red-Al reduction6 of 10 and
monosilylation provided a 1,3-diol derivative 11. This
underwent hydrogenation to afford alcohol 12. Glycosidation7
of 12 was best realized by using 2.5 equiv of thioglycoside
13 in the presence of 6.0 equiv of N-iodosuccinimide (NIS)
and a catalytic amount of triflic acid (TfOH) in CH2Cl2 to
give the corresponding â-glucoside 14. After de-silylation
of 14, the resulting alcohol was oxidized to give methyl ester
6a {[R]D +20.7° (c 0.35), δΗ 2.45 (dd, J ) 16, 6.8 Hz, H-2a),
2.89 (dd, J ) 16, 5.9 Hz, H-2b), 3.59 (s, OMe)}.8 On the
other hand, SN2 inversion of 119 via the corresponding
monochlate11 provided the epimer 15. Hydrogenation of 15
gave alcohol 16, which was glycosylated to give glucoside
17. This was transformed into the diastereomer 6b {[R]D
-2.0° (c 1.01), δΗ 2.34 (dd, J ) 16, 4.2 Hz, H-2a), 2.44
Synthesis of the right-half model 7 started with a trans-
formation of ester 1812 into unsaturated ester 19 via a 6-step
sequence including a one-carbon homologation and Wittig
reaction (Scheme 3). After DIBAL reduction of 19, the
resulting allyl alcohol was oxidized to give epoxide 20. 1,2-
Reduction6 of the epoxide followed by an oxirane formation
provided terminal epoxide 5a. Nucleophilic addition of Bu2-
CuLi to 5a proceeded cleanly, giving alcohol 21. A hydroxy
protection-deprotection sequence of 21 afforded 2R,8S-
benzoate 23 via silyl ether 22. Glycosidation of 23 provided
â-glucoside 7a {mp 99-100 °C, [R]D + 2.2° (c 0.41)}. On
the other hand, an SN2 inversion reaction of 239 afforded
2R,8R-isomer 24, which was glycosylated to give 7b {[R]D
-4.5° (c 0.40)}.
SN2 inversion of 22 relied on the Mitsunobu reaction.10
The benzoate 25 thus obtained was transformed into 2S,8S-
alcohol 26, and then glucoside 7c {mp 105-107 °C, [R]D
+13.6° (c 0.99)}. According to the method for preparation
of 24 from 23, 26 was changed to the epimer 27, which was
glycosylated to give 7d {[R]D +11.4° (c 0.73)}. In their 1H
NMR spectra, the signals corresponding to H-2 of 7b and
7d were observed at 5.00 ppm, while those of 7a and 7c
(6) Finan, J. M.; Kishi, Y. Tetrahedron Lett. 1982, 23, 2719-
2722.
(7) Veeneman, G. H.; van Leeuwen, S. H.; van Boom, J. H. Tetrahedron
Lett. 1990, 31, 1331-1334. Konradsson, P.; Udodong, U. E.; Fraser-Reid,
B. Tetrahedron Lett. 1990, 31, 4313-4316.
(8) Values of [R]D and δH were measured for the solution in CHCl3 and
CDCl3, respectively, at 23 ( 2°.
(9) Mitsunobu reaction10 of this compound failed.
(10) Mitsunobu, O. Synthesis 1981, 1-28.
(11) Shimizu, T.; Hiranuma, S.; Nakata, T. Tetrahedron Lett. 1996, 37,
6145-6148.
(12) Fukui, H.; Tsuchiya, Y.; Fujita, K.; Nakagawa, T.; Koshino, H.;
Nakatta, T. Bioorg. Med. Chem. Lett. 1997, 7, 2081-2086.
1556
Org. Lett., Vol. 5, No. 9, 2003

Scheme 2a
Scheme 3a
a Reagents and conditions: (a) (EtO)2P(O)CH2CO2Et, NaH, THF,
0 °C; (b) DIBAL, CH2Cl2, -78 °C, 98% (2 steps); (c) (-)-DET,
Ti(Oi-Pr)4, t-BuO2H, MS4A, CH2Cl2, -23 °C, 82%; (d) Red-Al,
THF, 0 °C to rt; (e) NaIO4, CH2Cl2-acetone-H2O, 76% (2 steps);
(f) TBDPSCl, imidazole, DMF, 0 °C, 99%; (g) H2, Pd/C, EtOAc,
rt, 99%; (h) 13, NIS, TfOH, MS4A, CH2Cl2, -15 °C, 66% for 14,
84% for 17; (i) TBAF, THF, 0 °C; (j) Jones reagent, acetone, 0
°C; (k) CH2N2, ether, 0 °C, 68% for 6a, 81% for 6b; (l) McCl,
2,6-lutidine, CH2Cl2, 0 °C; (m) CsOAc, 18-crown-6, toluene, 80
°C; (n) K2CO3, MeOH, rt, 79% (3 steps).
a Reagents and conditions: (a) NaBH4, NiCl2‚6H2O, MeOH, 0
°C; (b) LAH, THF, 0 °C, 83% (2 steps); (c) p-TsCl, pyridine, 0
°C; (d) NaCN, DMSO, rt, 86% (2 steps); (e) DIBAL, CH2Cl2, -78
°C; (f) (EtO)2P(O)CH2CO2Et, NaH, THF, 0 °C, 99%; (g) (+)-DET,
Ti(Oi-Pr)4, t-BuO2H, MS4A, CH2Cl2, -23 °C, 89%; (h) DIBAL,
CH2Cl2, 0 °C, 80%; (i) p-TsCl, Bu2SnO, Et3N, CH2Cl2, rt, 82%;
(j) K2CO3, MeOH, rt, 97%; (k) n-BuLi, CuI, Et2O, -60 to -40
°C, 97%; (l) TBDPSCl, imidazole, DMF, rt, 90%; (m) PPTS, EtOH,
40 °C, 78%; (n) BzCl, pyridine, 0 °C, 99%; (o) TBAF, THF, 40
°C, 99%; (p) 13, NIS, TfOH, MS4A, CH2Cl2, -15 °C, 70% for
7a, 74% for 7b, 61% for 7c, 73% for 7d; (q) McCl, 2,6-lutidine,
CH2Cl2, 0 °C; (r) CsOAc, 18-crown-6, toluene, 80 °C; (s) concd
HCl, MeOH, 0 °C to rt, 73% (3 steps) for 24, 63% (3 steps) for
27; (t) BzOH, DEAD, Ph3P, THF, 0 °C; (u) TBAF, THF, 0 °C to
rt, 79% (2 steps).
were observed at 5.09 ppm, quite similar to that of 2 {δH
5.08}. In respect to the C-2 (or 21 for 2) proton, clear
discrimination between 7a and 7c was very difficult.
However, there was similarity between 7a and 2 in the
splitting pattern of signals derived from the phenyl groups;
see Supporting Information.
On the basis of these results, we estimated that 2 should
have the stereochemistry of 3R, 15S, and 21R if 2 possesses
a D-glucose residue. To confirm the estimation, 3R,15S,21R-
methyl ester 2a was synthesized as shown in Scheme 4. The
hydroxyl group in 11 was initially protected as the corre-
sponding THP ether to give 28. Hydroboration of 28
followed by iodination provided iodide 4a. Generation of
alkyllithium derived from primary iodide 4a (t-BuLi, -78
°C to rt) in ether-hexane13 followed by treatment with CuI
provided an organocopper reagent, which reacted with
(13) Bailey, W. F.; Punzalan, E. R. J. Org. Chem. 1990, 55, 5404-
5406.
Org. Lett., Vol. 5, No. 9, 2003
1557

Scheme 4a
a Reagents and conditions: (a) DHP, PPTS, CH2Cl2, rt, 99%; (b) BH3‚MeS, THF, 0 °C and then 3N NaOH, H2O2, 0 °C rt, 87%; (c) I2,
Ph3P, imidazole, benzene, 10 °C, 92%; (d) t-BuLi, CuI, 5a, hexane-Et2O, -78 °C; (e) PPTS, EtOH, 50 °C, 85%; (f) BzCl, pyridine-
CH2Cl2, -30 °C, 30%; (g) 13, NIS, TfOH, MS4A, CH2Cl2, -15 °C, 49%; (h) TBAF, THF, 0 °C, 82%; (i) Jones reagent, acetone, 0 °C;
(j) CH2N2, ether, 0 °C, 86% (2 steps).
epoxide 5a to give a coupled product 29 in 78% yield. Upon
treatment with PPTS, 29 afforded triol 30. Benzoylation14
of 30 with 1.05 equiv of BzCl in CH2Cl2-pyridine at -30
°C provided the desired benzoate 3a as a major product along
with several dibenzoate derivatives. Glycosidation of 3a was
performed by using 5.0 equiv of 13 in the presence of NIS
(12 equiv) and TfOH (cat.) to afford â-glycoside 31 in 49%
yield. Finally, removal of the silyl group of 31, oxidation,
and methylation gave the 3R,15S,21R-carboxylic acid methyl
ester 2a {[R]D24 +9.2° (c 0.60, CHCl3)} in good yield. The
1H, 13C NMR, and [R]D data were identical with those of 2
{[R]D24 +9.1° (c 0.22, CHCl3)}, showing macroviracin A to
have the structure 1a as depicted in Scheme 4.
In summary, the absolute configuration of macroiviracins
was unambiguously established by chemical synthesis of C22-
fatty acid methyl ester 2a. Now, the total synthesis of 1a is
underway.
Acknowledgment. We are grateful to Mr. T. Hyodo
(Kaken Pharm. Co.) for providing us natural macroviracins.
We also express our thanks to Ms. K. Harata at RIKEN for
mass spectral measurements.
1
Supporting Information Available: H NMR spectra for
compounds 6a,b, 7a-d, 2, and 2a; 13C NMR spectra for 2
and 2a. This material is available free of charge via the
Internet at http://pubs.acs.org.
(14) From the results of several model experiments, we expected that
benzoylation at the C-2 position would proceed regioselectively.
OL034338K
1558
Org. Lett., Vol. 5, No. 9, 2003