De novo formal synthesis of (-)-virginiamycin M2 via the asymmetric hydration of dienoates

Article abstract of DOI:10.1021/ol071145e

A de novo approach to the formal total synthesis of the macrolide natural product (-)-virginiamycin M₂ has been achieved via a convergent approach. The absolute and relative stereochemistry of the nonpeptide portion of (-)-virginiamycin M₂

Full text of DOI:10.1021/ol071145e

ORGANIC
LETTERS
2007
Vol. 9, No. 16
3105-3108
De Novo Formal Synthesis of
)-Virginiamycin M2 via the Asymmetric
(−
Hydration of Dienoates
Matthew S. Mortensen, Joshua M. Osbourn, and George A. O’Doherty*
Department of Chemistry, West Virginia UniVersity, Morgantown, West Virginia 26506
George.ODoherty@mail.wVu.edu
Received May 20, 2007
ABSTRACT
A de novo approach to the formal total synthesis of the macrolide natural product (−)-virginiamycin M2 has been achieved via a convergent
approach. The absolute and relative stereochemistry of the nonpeptide portion of (
asymmetric dihydroxylation reactions.
−)-virginiamycin M2 was introduced by two Sharpless
The problem associated with new microbes that develop
resistance mechanisms to antibiotics has fueled the never-
ending search for new antibacterial agents.1 In fact, infections
from methicillin-resistant S. aureus (MRSA) organism have
become increasingly common.2 More alarming is the dis-
covery of vancomycin-resistant S. aureus organisms.3 Cur-
rently, three antibiotics (linezolid, daptomycin, and quinu-
pristin-dalfopristin) have been approved to combat these
infections. The oldest of the three, quinupristin-dalfopristin,
is the admixture of two types of streptogramin antibiotics,
dalfopristin (type A) and quinupristin (type B) (Figure 1),
which derives its potency by the synergistic binding of two
weak ribosome binders.1
dehydroproline portion of virginiamycin M1 (Figure 2), which
enhances activity by providing improved solubility.1c As part
of a program aimed at developing new antibiotic structures,
we became interested in the synthesis of both virginiamycin
M2 as well as the synthesis of a library of novel virginiamycin
analogues. We envisioned replacing the sulfone group on
the proline with an aminoglycoside sugar, with the hope of
discovering new synergistic binding.4 As a prelude to these
medicinal chemistry studies/library synthesis, we decided to
investigate the feasibility by conducting a formal total
synthesis.
In addition to its potent antibiotic activity, the structural
novelty of virginiamycin M2 has also attracted the attention
The type A component dalfopristin is a semi-synthetic
compound prepared by a sulfinic acid addition across the
(1) (a) Walsh, C. T. Nature 2000, 406, 775-781. (b) Chang, S.; Sievert,
D. M.; Hageman, J. C.; Boulton, M. L.; Tenover, F. C.; Downes, F. P.;
Shah, S.; Rudrik, J. T.; Pupp, G. R.; Brown, W. J.; Cardo, D.; Fridkin, S.
K. New Engl. J. Med. 2003, 348, 1342-1349 and refs therein. (c) Manthous,
C. A.; Amaoteng-Adjepong, Y. Chest 2000, 118, 9-11. For clinical and
pharmacological background on streptomycin antibiotics, see: (d) Barriere,
J. C.; Berthaud, N.; Beyer, D.; Dutka-Malen, S.; Paris, J. M.; Desnottes, J.
F. Curr. Pharm. Des. 1998, 4, 155-180. (e) Manzella, J. P. Am. Fam.
Physician 2001, 64, 1863-1866. (f) Mukhtar, T. A.; Wright, G. D. Chem.
ReV. 2005, 105, 529-542. (g) Porse, B. T.; Garrett, R. A. J. Mol. Biol.
1999, 286, 375. For synthesis of new quinolone class antibiotic active against
VRSA, see: (h) de Souza, N. J.; Gupte, S. V.; Deshpande, P. K.; Desai, V.
N.; Bhawsar, S. B.; Yeole, R. D.; Shukla, M. C.; Strahilevitz, J.; Hooper,
D. C.; Bozdogan, B.; Applebaum, P. C.; Jacobs, M. R.; Shetty, N.; Patel,
M. V.; Jha, R.; Khorakiwala, H. F. J. Med. Chem. 2005, 48, 5232-5242.
(2) McCaughey, B. N.Y. Times Nov 14, 2006, p 27.
Figure 1. Quinupristin-dalfopristin.
(3) Schito, G. C. Clin. Microbiol. Infect. 2006, 12, 3-8.
10.1021/ol071145e CCC: $37.00
© 2007 American Chemical Society
Published on Web 07/03/2007

Scheme 1. Retrosynthesis of (-)-Virginiamycin M2 (1)
Figure 2. Type A streptogramins.
of the synthetic community.5,6 To date, several total syntheses
of the type A streptogramin antibiotics have been completed,5
along with a formal total synthesis.6 While all of the previous
syntheses of the streptogramin class of antibiotics derived
their asymmetry from the chiral pool, enzymatic resolution,
and/or chiral auxiliaries,5 we were interested in a de novo
asymmetric approach that would use asymmetric catalysis
to install the stereocenters of the non-amino acid portion of
type A streptogramins. Herein we describe our successful
efforts to implement this strategy for the de novo formal
total synthesis of virginiamycin M2.
Retrosynthetically, virginiamycin M2 (1) has been derived
from the seco-macrolide 2,5b which in turn could be prepared
from the known oxazole 3, and triene 4. Previously,
Schlessinger had demonstrated the conversion of 4 and 3 to
virginiamycin M2 (1).5a In our strategy (Scheme 1), we
envisioned the triene 4 as being assembled from D-proline,
allylic amine 5, and δ-hydroxy ester 6. Finally, we planned
to install the asymmetry of these two fragments by the
application of a regioselective Sharpless asymmetric dihy-
droxylation of diene fragments 7 and 8.7,8 In particular, we
were interested in using our asymmetric hydration strategy
for the preparation of the syn-γ-substituted δ-hydroxyenoate
6.9
To access useful quantities of dienoate 8, an efficient five-
step approach was developed (Scheme 2). While the route
featured standard Wittig/Horner-Emmons olefination chem-
istry, key to practical nature of this approach was the
recognition that TFA can catalyze the stereoselective isomer-
ization of enal 11 to the more stable E-isomer (98:2), which
when treated with the stabilized Wittig reagent provided good
yields of 8 (72%).10 While we previously had demonstrated
the asymmetric hydration of 8 to the enantiomer of 6, because
of the pseudo-enantiomeric nature of the Sharpless reagents
(4) For conformational studies of the virginiamycin type antibiotics,
see: (a) Dang, J.; Metzger, R. P.; Brownlee, R. T. C.; Chai, A. N.; Separovic,
F. Eur. J. Biophys. 2005, 34, 383-388. (b) Dang, J.; Separovic, F.;
Brownlee, R. T. C.; Metzger, R. P. J. Org. Biomol. Chem. 2004, 2, 2919-
2924. (c) Dang, J.; Separovic, F.; Brownlee, R. T. C.; Metzger, R. P. Aust.
J. Chem. 2004, 57, 415-418.
Scheme 2. Asymmetric Hydration of Dienoate 8
(5) For the first total synthesis of virginiamycin M2, see: (a) Schlessinger,
R. H.; Li, Y.-J. J. Am. Chem. Soc. 1996, 118, 3301-3302. For the second,
see: (b) Breuilles, P.; Uguen, D. Tetrahedron Lett. 1998, 39, 3149-3152.
For the first total synthesis of madumycin II, see: (c) Tavares, F.; Lawson,
J. P.; Meyers A. I. J. Am. Chem. Soc. 1996, 118, 3303-3304. For the second,
see: (d) Ghosh, A. K.; Liu, W. J. Org. Chem. 1997, 62, 7908-7909.
(6) Brennan, C. J.; Campagne, J.-M. Tetrahedron Lett. 2001, 42, 5195-
5197.
(7) The regioselectivity of the asymmetric dihydroxylation of di- and
trienoates has been studied by Sharpless. See: (a) Berker, H.; Soler, M.
A.; Sharpless, K. B. Tetrahedron 1995, 51, 1345-1376. Our group: (b)
Zhang, Y.; O’Doherty, G. A. Tetrahedron 2005, 61, 6337-6351.
(8) For the use of this approach in synthesis, see: (a) Smith, A. B.; Walsh,
S. P.; Frohn, M.; Duffey, M. O. Org. Lett. 2005, 7, 139-142. (b) Ahmed,
Md. M.; Akhmedov, N.; Cui, H.; Friedrich, D.; O’Doherty, G. A.
Heterocycles 2006, 70, 223-233. (c) Ahmed, Md. M.; Cui, H.; O’Doherty,
G. A. J. Org. Chem. 2006, 71, 6686-6689. (d) Gao, D.; O’Doherty, G. A.
J. Org. Chem. 2005, 70, 9932-9939. (e) Ahmed, Md. M.; O’Doherty, G.
A. Tetrahedron Lett. 2005, 46, 4151-4155.
(9) (a) Ahmed, Md.; Mortensen, M. M. S.; O’Doherty, G. A. J. Org.
Chem. 2006, 71, 7741-7746. (b) Hunter, T. J.; O’Doherty, G. A. Org.
Lett. 2001, 3, 1049-1052. For its use in the synthesis of natural products,
see: (c) Li, M.; O’Doherty, G. A. Org. Lett. 2006, 8, 3987-3990. (d) Li,
M.; O’Doherty, G. A. Org. Lett. 2006, 8, 6087-6090.
3106
Org. Lett., Vol. 9, No. 16, 2007

(DHQ/DHQD), the regiochemistry of this reaction was an
open question. In practice, we found that the monomeric
4-methyl-2-quinolylether linked DHQD ligand provided the
best balance in terms of regio- and stereochemistry.11 Thus,
when 8 was dihydroxylated with the OsO4/DHQD-4-MEQ
reagent, good yields of diol 12 as a single regioisomer were
isolated (75%, 90% ee), which was cyclized into carbonate
13 in good overall yield (90%). Exposure of carbonate 13
to the palladium(0)-catalyzed reduction conditions (HCO2H/
Et3N) provided δ-hydroxyenoate 6 in good yield (98%).
With δ-hydroxyenoate 6 in hand, we next investigated the
synthesis of amine 5 (Scheme 3). Unfortunately, our initial
10:1, E/Z). The enoate 18 was reduced with DibalH to give
allylic alcohol 19, which in turn was converted into allylic
nitrile 20 by a two-step protocol (PPh3/I2 then NaCN in AN;
50%). Base-promoted elimination (K2CO3/MeOH, 81%) of
20 stereoselectively gave the E,E-diene 21 (98:2, EE:EZ).
The secondary allylic alcohol in 21 was protected as a
TBDPS-ether (TBDPSCl/imid, 89%), and the nitrile was
cleanly reduced with AlH3 to give amine 5 in good yield
(70%).12
We next investigated the synthesis of the first fragment
coupling product ester 25 (Scheme 4). While 24 could be
Scheme 4. Synthesis of Proline Ester 25
Scheme 3. De Novo Synthesis of Allylic Amine 5
prepared in one step from a Troc-protected D-proline, for
this model system, we decided to use the already available
Boc-protected D-proline 22. Coupling proline 22 and alcohol
6 with DCC/DMAP followed by deprotection of the Boc
group (TFA) gave good yields of amine 23 (88%, two steps).
The secondary amine 23 was protected with TrocCl/Pyr
(71%) to give ester 24, which was selectively hydrolyzed
(NaOH, 56%) to give the desired carboxylic acid 25.13
With the final two fragments 5 and 25 in hand, we
investigated their coupling to form our desired target
molecule 4 (Scheme 5). After screening several coupling
plan to use a regioselective asymmetric dihydroxylation of
dienoate 7 was thwarted by our inability to prepare 7 by
Horner-Emmons olefination of 15. Instead, of 7, only
Michael addition products were observed. Thus, we decided
to preform the dihydroxylation a step earlier on enone 15,
which could be easily prepared in three steps (63%) from
1,3-propane diol (14). Exposure of enone 15 to the typical
Sharpless asymmetric dihydroxylation procedures gave a
good yield (71%) of diol 16 and in high enantiopurity (90%
ee). The diol product was protected as an acetonide (2,2-
DMP/5% CSA, 72%), and the ketone product 17 underwent
Horner-Emmons olefination to form enoate 18 (84%;
Scheme 5. Completion of the Formal Synthesis of
(-)-Virginiamycin M2
procedures, we found the DCC/DMAP to give the best yields
and to be operationally the simplest. Thus, exposing a 1:1
(10) We have found this acid-catalyzed isomerization of R,â-unsaturated
aldehydes to be quite general and very useful in our diene syntheses. See:
Varelis, P.; Johnson, B. Aust. J. Chem. 1997, 50, 43-52.
(11) In addition to the electronic effect, we have found that subtle steric
effects can lead to the formation of regioisomers in the Sharpless asymmetric
dihydroxylation reaction of dienoates. See refs 7 and 9a.
(12) The chemoselective AlH3 reduction of dienylnitriles like 21 is
precedented; see ref 5a.
(13) We surmised that this lower than expected yield was due to the
water solubility of 25; no elimination to dienoate 8 was observed.
Org. Lett., Vol. 9, No. 16, 2007
3107

mixture of 5 and 25 to a CH2Cl2 solution of DCC/DMAP
gave 82% yield of amide 4, which was physically (optical
rotation) and spectroscopically (1H NMR, 13C NMR, IR, and
MS) identical to the material previously reported by
Schlessinger.14
of mixed aminoglycoside/virginiamycin M2 analogues is
ongoing, and these results will be reported in due course.
Acknowledgment. We are grateful to the NIH (GM63150)
and NSF (CHE-0415469) for the support of our research
program, and NSF-EPSCoR (0314742) for a 600 MHz NMR
and an LTQ-FT mass spectrometer at WVU.
In conclusion, a short formal de novo asymmetric synthesis
of virginiamycin M2 has been developed. This highly enantio-
and diastereocontrolled route illustrates the utility of our
dienoate asymmetric hydration strategy for natural product
synthesis. Further application of this approach to the synthesis
Supporting Information Available: Complete experi-
mental procedures and spectral data for all new compounds
can be found in the Supporting Information. This material
is available free of charge via the Internet at http://pubs.acs.org.
(14) Unfortunately, a typographical error occurs in the data reported by
Schlessinger such that the 11 signals below 24 ppm are missing from their
reported 13C NMR spectral data. The remaining 25 signals above 24 ppm
did match the reported data; see ref 5a.
OL071145E
3108
Org. Lett., Vol. 9, No. 16, 2007