Sulfamate Tethered Aza-Wacker Strategy for a Kasugamine Synthon

Abstract

We present our preparation of a kasugamine synthon, which proceeds in 14 steps from a literature epoxide. We expect that this kasugamine derivative can be used for the total syntheses of kasugamycin, minosaminomycin, and analogue antibiotics. A key step in the synthesis is our laboratory’s sulfamate-tethered aza-Wacker cyclization.

Graphical Abstract

Kasugamycin and minosaminomycin are structurally related aminoglycoside antibiotics (Figure 1). Kasugamycin was first isolated in 1965 from a bacterial strain (Streptomyces kasugiensis) found in a soil sample near the Kasuga shrine in Nara, Japan.^1–3 Shortly after isolation, empirical testing demonstrated that kasugamycin was active against a variety of bacteria and fungi.^{4, 5} While its antibacterial activity is much less than that of other members of the aminoglycoside family, its ability to control rice blast disease caused by the fungus Pyricularia oryzae was so remarkable that it continues to be used commercially as an agricultural fungicide (Kasumin).^{6, 7} Minosaminomycin is a structurally related antibiotic to kasugamycin and has demonstrable activity against a variety of pathogenic bacteria, including mycobacteria.^{8, 9}

Figure 1.

Kasugamycin and minosaminomycin are potent aminoglycoside antibiotics and are structurally related.

Our interest in kasugamycin and minosaminomycin stems from both fundamental and applied considerations. These antibiotics have been synthetically unexamined for close to fifty years.^{8, 10–16} Our laboratory has a deep interest in the development of new reactions which would enable the syntheses of complex antibiotics. Specifically, we envisioned that our laboratory’s sulfamate-tethered aza-Wacker cyclization could serve as a key step for the assembly of the kasugamine portion of these antibiotics.^17–24 Our interest in designing practical synthetic routes to antibiotics with new mechanisms of action^25–31 relative to approved agents is motivated by the global rise in resistance to antibacterial and antifungal agents currently in use.^32–34 It is our hope that de novo syntheses will allow for the rapid preparation of analogues^35–37 and will serve as a start of a drug-discovery effort to replenish worrisome antibacterial and antifungal pipelines. In this account, we describe our preparation of a kasugamine synthon, a common building block for both kasugamycin and minosaminomycin.

There are two general strategies for the syntheses of unusual monosaccharides.^38,49–51 One is the classic “heterocycle→heterocycle” approach; in such syntheses, chemists begin with a more readily available monosaccharide and then systematically manipulate it to the end target. A second, less explored approach is to commence with a linear precursor, build functional group complexity, and then effect a late-stage cyclization (“linear→cyclic”). The former approach generally has fewer total step counts; on balance, due to existing complexity, monosaccharides can be very challenging starting materials. For this reason, we prefer the latter approach, because of the greater flexibility it affords and chose kasugamine derivative A as our first target (Scheme 1).

Scheme 1.

Retrosynthetic analysis.

In accord with this plan, our retrosynthesis is depicted in Scheme 1. After a late stage hemiacetalization of heteroatom-rich linear precursor C, deoxygenation and azide reduction would furnish target A. C would be built systematically from epoxide F.^{39, 40} We envisioned our laboratory’s sulfamate-tethered aza-Wacker cyclization^{17, 21} would serve as a key step during the transformation of epoxide F into aldehyde C. The successful execution of this strategy would increase the prominence of both classical and tethered aza-Wacker chemistry for the stereoselective assembly of complex targets.^{19, 41}

Our forward synthesis (Scheme 2) commenced with morpholine amide 1, itself prepared in two steps from commercial D-erythrono-1,4-lactone in a procedure developed and disclosed by our laboratory.¹⁷ Morpholine amide 1 was converted into diol 2 by heating with NaBH₄ in EtOH. Treatment of diol 2 with benzoyl chloride in ice-cold CH₂Cl₂ allowed for selective esterification of the primary alcohol. The secondary alcohol of benzoate 3 was tosylated using a standard mixture of NEt₃, TsCl, and catalytic DMAP. Treatment of 4 with mixture of K₂CO₃ in MeOH allowed for the hydrolysis of the benzoate ester; the resulting alkoxide anion displaced the tosylate in an intramolecular S_N2 reaction to form epoxide 5. CuI mediated the regioselective opening of epoxide 5 by cis-1-propenyl-magnesium bromide. For this transformation, we have found that it is critical to freshly prepare cis-1-propenylmagnesium bromide; for reasons unknown, using commercial solutions of 1-propenyl-magnesium bromide did not allow for clean reactivity. Sulfamoylation of alcohol 6 proceeded smoothly with Johnson and Magolan’s protocol.⁴²

Scheme 2.

Opening sequence of reactions.

As a first step to evaluating the sulfamate-tethered aza-Wacker cyclization with 7, we tested our standard protocol of Pd(OAc)₂/Cu(OAc)₂ in hot CH₃CN and a ~1 atm O₂ atmosphere. We were pleased that this delivered desired oxathiazinane 8 in a 36% yield, an excellent starting point for further optimization (Scheme 3, Entry 1). Switching to Pd₂(dba)₃ at a higher loading increased the yield to 50% (Scheme 3, Entry 2). Using PdCl₂(nbd) and Pd(CH₃CN)₂Cl₂ in place of Pd₂(dba)₃ did not help the reaction (Scheme 3, Entries 3–5). We have previously observed a dramatic improvement in sulfamate-tethered aza-Wacker cyclizations when mono-protected amino acid ligands and 4Å molecular sieves are used. We hypothesize, in accord with the Yu laboratory’s observations,^43–45 that mono-protected amino acids (MPAA) ligands help promote the Pd (II) – Pd (0) catalytic cycles operative during aza-Wacker transformations. Indeed, in line with our predictions, the yield of oxathiazinane 8 increased to 63% upon the addition of MPAA ligand Fmoc-Gly-OH and 4Å molecular sieves (Scheme 3, Entry 6); furthermore, starting material was fully consumed. Despite the presence of molecular sieves, it was critical to use anhydrous DMSO for the best reaction yield. X-ray structural analysis of 8 (CCDC: 2289376) allowed us to unambiguously confirm its identity and absolute stereochemistry. We note that this product formed as a single diastereomer (within the limits of ¹H NMR analysis). Test reactions were conducted on a 30 mg scale, but our optimized reaction conditions could be applied to batches as large as 1.7 g.

Scheme 3.

Optimization of the key sulfamate-tethered aza-Wacker cyclization.

Moving forward, a Cbz group was appended to the oxathiazinane nitrogen for a subsequent ring-opening with NaN₃ (Scheme 4).^{46, 47} The alkene of 10 was cleaved by dihydroxylation followed by treatment with NaIO₄. Aldehyde 12 was unstable to purification and was immediately treated with TsOH to form hemiacetal 13 as a mixture of anomers. This anomeric mixture was bis-acetylated with Ac₂O/pyridine, and the anomeric acetate was exchanged with thiophenol using BF₃•OEt₂. For 15, we have assigned the configuration at the anomeric carbon based on the small coupling constant of the anomeric proton (~1.5 Hz) and by analogy to the crystal structure of a later intermediate (vide infra). The acetate of 15 was removed using K₂CO₃ in MeOH.

Scheme 4.

Towards the endgame.

Initially, it was our plan to tosylate the alcohol of 16 and to displace it with an appropriate hydride source (Scheme 5A). Accordingly, tosylate 17 was prepared by stirring 16 with NEt₃ and TsCl. However, despite multiple trials, including with NaBH₄, LiAlH₄, and NaBH₃CN, we were unable to cleanly obtain 18, likely because of the presence of an azide functionality. We thus sought to develop a protocol which would allow for the clean formation of the methyl group and the primary amine of kasugamine in a single step. Knowing that there was precedent for the Pd-catalyzed hydrogenolysis of azides and alkyl iodides,⁴⁸ we prepared iodide 19 using a Finkelstein reaction (Scheme 5B). Indeed, stirring 19 with NEt₃ and Pd-C under an atmosphere of H₂ gas cleanly delivered kasugamine derivative 20. Crystal structure analysis of 20 (CCDC 2291427) allowed us to unambiguously confirm its identity and absolute stereochemistry.

Scheme 5.

(A) Displacement of the tosylate by hydride does not give expected product under multiple conditions. (B) Upon treatment with Pd/C and H₂ gas, deiodination and azide reduction occur in one step to furnish the desired target.

In summary, we have completed a concise synthesis of a kasugamine synthon, which we envision can be used in preparations of kasugamycin, minosaminomycin, and related analogues. A highlight of the synthesis is our sulfamate-tethered aza-Wacker cyclization. The march towards total syntheses of these antibiotics continues in our laboratory.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.