PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Org Lett. Author manuscript; available in PMC 2010 November 5.
Published in final edited form as:
PMCID: PMC2783931
NIHMSID: NIHMS148850

Brønsted Acid-Promoted Glycosylations of Disaccharide Glycal Substructures of the Saccharomicins

Abstract

An external file that holds a picture, illustration, etc.
Object name is nihms148850u1.jpg

An acid-promoted glycosylation and alkynol cycloisomerization sequence provided direct access to the 2-deoxytrisaccharide corresponding to the fucose-saccharosamine-digitoxose substructure of saccharomicin B. In the course of this work, the absolute stereochemistry of the repeating fucose-saccharosamine disaccharide of saccharomicins was also confirmed.

The heptadecasaccharide antibiotics saccharomicins A and B were first isolated from the actinobacteria Saccharothrix espanaensis (Figure 1), and were found to have activity against a wide array of Gram-positive and Gram-negative bacteria. Cross resistance was not observed with many other antibiotics, including vancomycin, piperacillin, or ciprofloxacin.1

Figure 1
Saccharomicins A and B

The saccharomicins possess several structural features of interest as synthetic targets. In addition to the 2-deoxysugar digitoxose (dig, Figure 1), the rare pyranosides saccharosamine (sac) and 4-epi-vancosamine (eva) are 2,6-dideoxysugars bearing 3-amino and 3-methyl substituents. Based on our earlier work that established the absolute stereochemistry of the terminal d-fucose (fuc-1) attached to the aglycon,2 as well as a synthesis of saccharosamine glycal via tungsten-catalyzed alkynol cycloisomerization,3 we now report the synthesis of a protected form of the fucose-saccharosamine disaccharide (3) present in several sectors of saccharomicins, specifically fuc-3/sac-2, fuc-5/sac-4, fuc-8/sac-7, and fuc-12/sac-11. In addition, we demonstrate the viability of Brønsted acid-promoted glycosylations to form 2-deoxyglycosides corresponding to the trisaccharides fuc-8/sac-7/rha-6, and the fuc-12/sac-11/rha-10 of saccharomicin A as well as fuc-12/sac-11/dig-10 of saccharomicin B.

Although the stereochemistry of fucose-1 in saccharomicins was established as the d-isomer,2 we chose to prepare fragments and develop methods based on the antipodal structures, due to the ready availability of l-fucose. Thus our synthesis began with the resolution of racemic beta-lactam 5 by glycosylation with the l-fucose-derived trichloroacetimidate 4 (Scheme 1).4 The success of this transformation was particularly sensitive to temperature, so that temperatures below −45 °C favored orthoester formation, whereas temperatures above 0 °C produced a greater proportion of the alpha-glycoside anomers. After acetate ester removal and selective protection of the cis-diols as acetonide esters, the diastereomers 6 and 7 were chromatographically separated.5

Scheme 1
Preparation of fucose-saccharosamine disaccharide glycals 11-12 and 14-15

The acidic protons of the fucosyl C2-hydroxyl and the terminal alkyne of compound 6 were protected to provide compound 8 bearing TBS ether and TMS-alkyne. From beta-lactam 8, sequential addition of methyl and hydride nucleophiles3 afforded the secondary alcohol 9, although stereoselective hydride addition was achieved only with the chiral oxazaborolidine reagent.6,7 Two protocols for protective group manipulations were explored in order to diminish basicity of the nitrogen substituent for the alkynol cycloisomerization step. Originally, we first removed the para-methoxyphenyl (PMP) substituent from 9 and formed the acetamide, followed by desilylations with tetrabutylammonium difluorotriphenylsilane and acetic acid.8 Subsequently, we observed that desilylations proceeded more cleanly from compound 9, followed by oxidative removal of the PMP substituent and N-acylation.

Tungsten-catalyzed alkynol cycloisomerization of compound 10 proceeded efficiently to afford the disaccharide glycal 11, and O-acetylation of the fucosyl 2-hydroxyl gave compound 12. With an eye to facile late-stage deprotection of the amino substituent, we also prepared substrate 13 bearing an N-allyloxycarbonyl (Alloc) protective group. In this case the alkynol cycloisomerization reaction was noticeably slower, and a relatively high loading of tungsten carbonyl was required for complete conversion to the disaccharide glycal 14. The N-Alloc alkene may have coordinated with tungsten carbonyl,9 as N-benzyloxycarbonyl and N-butoxycarbonyl-protected substrates are considerably more reactive in alkynol cycloisomerizations.3,10

As the methyl glycoside of a peracetylated derivative of the fucose-saccharosamine disaccharide (3) had been reported as a degradation product arising from acidic methanolysis of saccharomicins, we sought to prepare the antipode of this compound in order to conclusively establish the absolute stereochemistry of the natural product-derived material.1a Acidic methanolysis of disaccharide glycal 11 was accompanied by acetonide removal, and acetylation of the hydroxyl groups of fucose provided a mixture of ent-3 and the beta-anomer, 16. After chromatographic separation of the anomers, the minor alpha-anomer ent-3 matched the spectroscopic data provided for the antipode of this structure, thus confirming that the fucose and saccharosamine sugars in the repeating unit of the natural product saccharomicins both possessed d-stereochemistry.11 Furthermore, the major beta-anomer of 16 was also characterized by X-ray crystallography, thus unambiguously establishing the structure of our synthetic material.12,13

Having previously established a viable acid-catalyzed glycosylation of a disaccharide glycal in our synthesis of digitoxin,14 we explored the glycosylation of glycal 15 with the d-rhamnose acceptor 17. In initial experiments, the yield of trisaccharide 18 was diminished by competitive hydration of the glycal due to adventitious water. However, excellent yields of 18 were obtained when the glycosylation was conducted in the presence of molecular sieves (Scheme 3). The optimized transformation required excess camphorsulfonic acid due to the mildly basic nature of the molecular sieves. Glycosylation occurred with complete stereoselectivity trans- to the C3-acetamido substituent.15

Scheme 3
Acid-promoted glycosylations of disaccharide 15 with rhamnose acceptor 17

Building on the concept of acyclic alkynyl alcohols as glycosyl acceptors,14,16 we subsequently explored the glycosylation of disaccharide 15 with alkynyl alcohol 19 as the precursor to digitoxose. In this case, the glycoside 20 was generated in excellent yield and with high stereoselectivity for the beta-anomer. After removal of the ester protective groups, the alkynyl alcohol substrate 21 underwent cycloisomerization to the trisaccharide glycal 22, although this transformation required stoichiometric tungsten carbonyl due to the N-Alloc substituent.

In summary, this work has confirmed the structure of the repeating fucose-saccharosamine disaccharide of saccharomicins, and has provided insights into the viability of Brønsted acid-promoted glycosylations with this disaccharide to provide 2-deoxyglycoside trisaccharide substructures observed in the saccharomicins.

Scheme 2
Synthesis of disaccharides ent-3 and 16
Scheme 4
Acid-promoted glycosylation of disaccharide 15 with alkynyl alcohol 19, and cycloisomerization to digitoxose-terminated trisaccharide 22

Supplementary Material

1_si_001

2_si_002

3_si_003

4_si_004

Acknowledgments

This work was initially supported by the National Institutes of Health, grant CA-59703.

Footnotes

Supporting Information Available. Experimental procedures and characterization and spectral data for new compounds are available free of charge via the Internet at http://pubs.acs.org.

References

1. (a) Kong F, Zhao N, Siegel MM, Janota K, Ashcroft JS, Koehn FE, Borders DB, Carter GT. J Am Chem Soc. 1998;120:13301. (b) Singh MP, Petersen PJ, Weiss WJ, Kong F, Greenstein M. Antimicrob Agents Chemother. 2000;44:2154. [PubMed]
2. Pletcher JM, McDonald FE. Org Lett. 2005;7:4749. [PubMed]
3. Cutchins WW, McDonald FE. Org Lett. 2002;4:749. [PubMed]
4. (a) Schmidt RR, Wegmann B, Jung KH. Liebigs Ann Chem. 1991:121. (b) Schmidt RR, Kinzy W. Adv Carbohydr Chem Biochem. 1994;50:21. [PubMed]
5. The identity of glycosides 6 and 7 were confirmed by X-ray crystallography (Supporting Information).
6. (a) Corey EJ, Bakshi RK. Tetrahedron Lett. 1990;31:611. (b) Corey EJ, Helal CJ. Angew Chem Int Ed. 1998;37:1986.
7. A variety of substrate-controlled reduction methods gave poor stereoselectivity; see Supporting Information for a summary of these results. The (R)-diastereomer was produced in approximately 5% yield in the oxazaborolidine reduction, and was recycled by IBX oxidation to the methyl ketone.
8. (a) Evans CM, Kirby AJ. J Chem Soc, Perkin Trans II. 1984:1269. (b) Hecker SJ, Heathcock CH. J Am Chem Soc. 1986;108:4586.(c)
Desilylation with tetrabutylammonium fluoride in the presence of acetic acid gave byproducts consistent with hydration of the alkyne, possibly from 5-exo-cyclization of the alkynyl alcohol and hydrolysis of the exocyclic enol ether, whereas the desilylation with tetrabutylammonium difluorotriphenylsilane afforded good yields of alkynyl alcohol 10.
9. Barluenga J, Diéguez A, Rodríguez F, Fananás FJ, Sordo T, Campomanes P. Chem Eur J. 2005;11:5735. [PubMed] (b) Katz TJ. Angew Chem Int Ed. 2005;44:3010. [PubMed] (c) Fuchibe K, Iwasawa N. Chem Eur J. 2003;9:905. [PubMed]
10. Davidson MH, McDonald FE. Org Lett. 2004;6:1601. [PubMed]
11. We thank Drs. Fangming Kong and Guy T. Carter (Wyeth Research) for providing 1H and 13C NMR spectra of compound 3 (Supporting Information).
12. We thank Drs. Rui Cao and Kenneth I. Hardcastle (Emory University) for solving the crystal structure of compound 16 (Supporting Information).
13. The l-fucosyl-d-saccharosamine diastereomer of 16 has also been prepared by a similar protocol beginning with compound 7 (Supporting Information).
14. McDonald FE, Reddy KS. Angew Chem Int Ed. 2001;40:3653. [PubMed]
15. These results are consistent with anchimeric assistance from the N-carbonyl substituent at C3. For examples of glycosylations trans- to C3-ester substituents, see:(a) Tsai TYR, Jin H, Wiesner K. Can J Chem. 1984;62:1403. (b) Komarova BS, Tsvetkov YE, Knivel YA, Zähringer U, Pier GB, Nifantiev NE. Tetrahedron Lett. 2006;47:3583. (c) Chiba S, Kitamura M, Narasaka K. J Am Chem Soc. 2006;128:6931. [PubMed]
16. McDonald FE, Wu M. Org Lett. 2002;4:3979. [PubMed]