Concise and enantioselective synthesis of the aminocyclitol core of hygromycin A

Article abstract of DOI:10.1021/ol0473750

(Chemical Equation Presented) Stereoselective aminohydroxylation and dihydroxylation using osmium(VIII) oxidants enabled the short and efficient synthesis of the aminocyclitol core of hygromycin A. In addition to allowing the selective introduction of the

Full text of DOI:10.1021/ol0473750

ORGANIC
LETTERS
2005
Vol. 7, No. 7
1275-1277
Concise and Enantioselective Synthesis
of the Aminocyclitol Core of
Hygromycin A
Timothy J. Donohoe,*,† Peter D. Johnson,† Richard J. Pye,† and Martine Keenan‡
Chemistry Research Laboratory, Department of Chemistry, UniVersity of Oxford,
Mansfield Road, Oxford OX1 3TA, UK, and Lilly Research Centre, Eli Lilly and
Company, Ltd., Erl Wood Manor, Sunninghill Road, Windlesham,
Surrey, GU20 6PH, UK
timothy.donohoe@chem.ox.ac.uk
Received December 21, 2004
ABSTRACT
Stereoselective aminohydroxylation and dihydroxylation using osmium(VIII) oxidants enabled the short and efficient synthesis of the aminocyclitol
core of hygromycin A. In addition to allowing the selective introduction of the heteroatoms N and O, the use of osmium (via an osmate ester)
as a protecting group for a 1,2-glycol is also reported. This tactic allowed efficient differentiation of otherwise equivalent hydroxyl groups and
allowed us to complete the synthesis in short order (14 steps) and excellent overall yield (12%).
Hygromycin A 1 is an antibiotic first isolated from the
fermentation broth of Streptomyces hygroscopicus in 1953
(Figure 1).1 It was identified as having a broad spectrum of
activity against both Gram-positive and Gram-negative
bacteria. Its mode of action is peptidyl transferase inhibition,
sharing a binding site on the ribosome with chloramphenicol.2
More recent studies on 1 have shown that it also has
hemagglutination inactivation activity, plus high antirepone-
mal activity.3 This has led to renewed interest in hygromycin
A as a treatment for mucohemorrhagic diseases, the control
of which is of economic importance to farmers.
A series of semisynthetic modifications to 1 were made
by a group from Pfizer in the mid-1990s and revealed that
the aminocyclitol was critical for the activity of the com-
pound while the furanoside unit was not.4
† University of Oxford.
‡ Eli Lilly and Company, Ltd.
(1) Pittenberger, R. C.; Wolfe, R. N.; Hohen, M. M.; Marks, P. N.; Daily,
W. A.; McGuire, M. Antibiot. Chemother. 1953, 3, 1268. Mann, R. L.;
Gale, R. M.; VanAbeele, F. R. Antibiot. Chemother. 1953, 3, 1279. Sumiki,
Y.; Nakamura, G.; Kawasaki, M.; Yamashita, S.; Anazi, K.; Isono, K.;
Serizawa, Y.; Tomiyama, Y.; Suzuki, S. J. Antibiot. 1955, 8, 170. Isono,
K.; Yamashita, S.; Tomiyama, Y.; Suzuki, S. J. Antibiot. 1957, 10, 21.
Wakisaka, Y.; Koizumi, K.; Nishimoto, Y.; Kobayashi, M.; Tsuji, N. J.
Antibiot. 1980, 33, 695.
(2) Guerrer, M. C.; Modolell, J. Eur. J. Biochem. 1980, 107, 409.
(3) Omura, S.; Hakagawa, A.; Fujimoto, T.; Saito, K.; Otoguro, K. J.
Antibiot. 1987, 40, 1619. Nakagawa, A.; Fujimoto, T.; Omura, S.; Walsh,
J. C.; Stoish, R. L.; George, B. J. Antibiot. 1987, 40, 1627.
Figure 1. Retrosynthesis of the aminocyclitol core of hygromycin
A.
10.1021/ol0473750 CCC: $30.25
© 2005 American Chemical Society
Published on Web 03/04/2005

Scheme 1
Scheme 2
of intermediate 8 was conveniently measured by HPLC).
However, our initial attempts to introduce the amino alcohol
functionality onto C-2/3 using the tethered aminohydroxy-
lation of carbamate 10 failed and returned staring material
in each case (protection of the free hydroxyl group did not
improve the situation). Analysis of a model of 9 and 10 led
us to postulate that the pseudoaxial nature of the carbamate
within 10 was responsible for the failure to oxidize the alkene
unit.10
Therefore, the conformation of the TA precursor was
altered by formation of a temporary bridge between the two
carbocyclic rings using an ether oxygen link (Scheme 2).
The ethereal linkage within 11 acts as a “drawstring”,
imposing greater strain throughout the molecule while pulling
the alcohol closer to a (more desirable) pseudoequatorial
position.11 If the model predicted above is correct, this may
permit a TA reaction to occur.
Compound 9 was reacted with NBS, forming bromoether
11 in 93% yield; after carbamate formation (12), it was
possible to try the key oxidation again. Pleasingly, carbamate
12 reacted smoothly under modified TA conditions12 furnish-
ing 13 in 67% yield (with 16% recovered starting material).
As expected from previous work, this reaction showed
complete regio- and stereoselectivity.8
To date, there has been only one total synthesis of
hygromycin A, completed by Ogawa and co-workers in
1989.5 However, the aminocyclitol unit of the natural product
has also been prepared by Trost (en route to a synthesis of
C-2 epihygromycin A),6 and some formal synthetic work has
been published by Arjona.7
A short and efficient synthesis of the key inositol subunit
2 was envisaged that would test some directed oxidation
methodology developed within the group (Figure 1). Our
unique approach would rely upon two key stereoselective
reactions: (i) the facially selective dihydroxylation of
intermediate 3 and (ii) the regio- and stereoselective tethered
aminohydroxylation (TA) reaction of allylic carbamate 4.8
Thus, the commercially available diketone 5 was converted
into masked cyclohexadiene 7 in two steps, 84% yield, and
98% ee (of known absolute configuration) following the
protocol of Ogasawara (Scheme 1).9
Inversion of the alcohol 7 under Mitsunobu conditions and
ester hydrolysis furnished trans-diol 9 in good yield (the ee
(4) Hecker, S. J.; Minich, M. L.; Werner, K. M. Bioorg. Med. Chem.
Lett. 1992, 2, 533. Hecker, S. J.; Lilley, S. C.; Minich, M. L.; Werner, K.
M. Bioorg. Med. Chem. Lett. 1992, 2, 1015. Hecker, S. J.; Lilley, S. C.;
Werner, K. M. Bioorg. Med. Chem. Lett. 1992, 2, 1043. Hecker, S. J.;
Cooper, C. B.; Blair, K. T.; Lilley, S. C.; Minich, M. L.; Werner, K. M.
Bioorg. Med. Chem. Lett. 1993, 3, 289. James, B. H.; Elliott, N. C.; Jefson,
M. R.; Koss, D. A.; Schicho, D. L. J. Org. Chem. 1994, 59, 1224. Cooper,
C. B.; Blair, K. T.; Jones, C. S.; Minich, M. L. Bioorg. Med. Chem. Lett.
1997, 7, 1747.
(9) Takano, S.; Higashi, Y.; Kamikubo, T.; Moriya, M.; Ogasawara, K.
Synthesis 1993, 948. Konno, H.; Ogasawara, K. Synthesis 1999, 1135.
Marchand, A. P.; LaRoe, W. D.; Sharma, G. V. M.; Suri, S. C.; Reddy, S.
D. J. Org. Chem. 1986, 51, 1622. Hiroya, K.; Kurihara, Y.; Ogasawara, K.
Angew. Chem., Int. Ed. Engl. 1995, 34, 2287.
(5) Chida, N.; Ohtsuka, M.; Nakazawa, K.; Ogawa, S. J. Chem. Soc.,
Chem. Commun. 1989, 436. Chida, N.; Ohtsuka, M.; Nakazawa, K.; Ogawa,
S. J. Org. Chem. 1991, 56, 2976.
(6) Trost, B. M.; Dudash, J.; Dirat, O. Chem. Eur. J. 2002, 8, 259. Trost,
B. M.; Dudash, J.; Hembre, E. J. Chem. Eur. J. 2001, 7, 1619.
(7) Arjona, O.; deDios, A.; Plumet, J.; Saez, B. J. Org. Chem. 1995, 60,
4932. Arjona, O.; deDios, A.; Plumet, J.; Saez, B. Tetrahedron Lett. 1995,
36, 1319.
(10) Subsequent studies with allylic carbamates axially disposed on a
six-membered ring have supported this hypothesis: it is presumed that the
reaction fails because orbitals on the oxidant cannot overlap correctly with
the π-system of the alkene. The fact that high (syn) stereoselectivity is
observed during the TA reaction of chiral acyclic allylic carbamates is also
consistent with a strict stereoelectronic constraint on the reaction. Donohoe,
T. J.; Johnson, P. D.; Pye, R. J.; Keenan, M. Org. Lett. 2004, 6, 2583.
(11) Takano, S.; Inomata, K.; Ogasawara, K. Chem. Lett. 1989, 359.
Honzumi, M.; Hiroya, K.; Taniguchi, T.; Ogasawara, K. Chem. Commun.
1999, 1985.
(8) Donohoe, T. J.; Johnson, P. D.; Cowley, A.; Keenan, M. J. Am. Chem.
Soc. 2002, 124, 12934.
1276
Org. Lett., Vol. 7, No. 7, 2005

The advantages of this reagent are (i) that the complex is
very reactive; (ii) hydrogen bonding control with OsO4/
TMEDA is well established and so the correct stereoselec-
tivity should ensue; (iii) the initial product of osmylation is
a stable osmate ester13 that could conceivably be used as a
protected form of the glycol.
Scheme 3
The results in Scheme 3 show that this key dihydroxylation
of 16 worked as planned and furnished osmate ester 17 as
the sole product in quantitative yield. Moreover, it was
possible to exploit the stability of osmate ester 17 and use it
as a glycol protecting group: compound 17 was easily
benzylated to give 18 (which could be isolated if required),
and the reaction was quenched with acidic methanol to
release the osmium from the glycol and form 19. This two-
step procedure (16f19) solved all the problems of poor
reactivity and stereo- and regioselectivity that had previously
arisen.
Finally, all that remained was formation of the methylene
acetal 205 (the stereochemistry of this fully protected
aminocyclitol was confirmed by X-ray crystallography14) and
then one-pot deprotection of the benzyl groups and oxazo-
lidinone under conditions reported by Trost.6 The data (1H/
13C NMR and specific rotation) for the final compound (-)-2
was a good match with that reported in the literature.
To conclude, a new synthesis of the key inositol portion
of hygromycin A is reported that proceeds in 14 steps and
12% overall yield; this represents the highest yielding route
reported for this compound. The key reactions in this
sequence were the tethered aminohydroxylation reaction and
directed dihydroxylation, both of which were completely
stereoselective. Moreover, the novel idea of using an osmate
ester as a protecting group for a 1,2-glycol is one that may
find further use in organic synthesis.
Next, the OH and NH groups were benzyl protected (14),
the ether bridge removed (Zn, AcOH), and a retro Diels-
Alder reaction performed on 15 to produce alkene 16 in good
overall yield.
To complete the synthesis, a diastereoselective dihydroxy-
lation reaction of 16 was required, followed immediately by
a regioselective protection of the two new hydroxyl groups
as a methylene acetal (Scheme 3).
Protection of the C-4 hydroxyl group of 16 prior to
dihydroxylation was required in order to allow regioselective
acetal formation afterward. However, such protection led to
a series of compounds that underwent a nonstereoselective
dihydroxylation reaction. Moreover, the oxidation was also
low yielding and sluggish (a similar problem had beset
Ogawa’s original synthesis).5
As a solution, it was decided to use the hydroxyl group at
C-4 to perform a directed dihydroxylation using the mixture
of OsO4/TMEDA (a reagent we developed some years ago).13
Acknowledgment. We thank the EPSRC and Eli Lilly
for funding this project. Pfizer, Novartis, AstraZeneca, and
Merck are thanked for generous unrestricted funding.
Supporting Information Available: Experimental pro-
cedures and analytical and spectroscopic characterization of
all starting materials and products. This material is available
free of charge via the Internet at http://pubs.acs.org.
OL0473750
(13) Donohoe, T. J.; Moore, P. R.; Waring, M. J.; Newcombe, N. J.
Tetrahedron Lett. 1997, 38, 5027. Donohoe, T. J.; Blades, K.; Moore, P.
R.; Waring, M. J.; Winter, J. G.; Helliwell, M.; Newcombe, N. J.; Stemp,
G. J. Org. Chem. 2002, 67, 7946.
(12) Modified conditions involving the use of sodium citrate as a ligand
for osmium increase the yield of the TA and also improve the workup. For
the original report on the use of citrate in the dihydroxylation reaction,
see: Dupau, P.; Epple, R.; Thomas, A. A.; Fokin, V. V.; Sharpless, K. B.
AdV. Synth. Catal. 2002, 344, 421.
(14) Crystallographic data for compound 20 has been deposited with the
Cambridge Crystallographic Data Centre as supplementary publication
CCDC 249567. Copies of the data can be obtained free of charge on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK (fax:
(+44)1223-336-033; E-mail: deposit@ccdc.cam.ac.uk).
Org. Lett., Vol. 7, No. 7, 2005
1277