Abstract
A concise route to a glycosylated allenic erythronolide was achieved. Key findings include the preparation of a desosamine sulfoxide donor and the use of the donor to glycosylate bulky acceptors. Additionally, the new reagent was used to prepare allene-containing macrocycles and to realize a four-step synthesis of macrolide 6 from bis[allene] 5. The longest linear sequence required to prepare 6 from commercial reagents was 15 steps.
Similar content being viewed by others
Introduction
The 14-membered macrolides have been especially important in the treatment of diseases due to the wide use of erythromycin antibiotics (Fig. 1) [1, 2]. These range from the parent natural product erythromycin A (1) to its clinically relevant semisynthetic derivatives, e.g. clarithromycin, azithromycin, telithromycin, roxithromycin, and solithromycin (not shown). The total chemical synthesis of erythromycins and their close aglycon analogs (1–4) has underscored many advances in the field [3,4,5,6,7,8,9,10,11,12,13,14,15,16,17]. Most of these syntheses relied on strategies that used six-membered ring matrices and/or methods of acyclic stereocontrol to fashion fragments with the required functionality and stereochemical relationships. Of particular interest to the theory of strategic analysis, the Danishefsky synthesis of 4 stands as the only route that controlled all of the stereochemical relationships of the macrolide from a single stereoprogenitor site [18]. Although Woodward and Martin achieved the only total synthesis of glycosylated erythronolides 1 and 2 and Kang and Andrade reported variants of telithromycin [19,20,21,22,23,24,25], only Myers—in a coup de main—has described the preparation of numerous glycosylated erythronolides [26].
We have described efforts to gain rapid access to this macrolide structure space by way of highly functionalized macrocyclic bis[allenes] [27,28,29]. This novel strategy converged on 5. The longest linear sequence to this key intermediate from commercial reagents is 11 steps. The high degree of unsaturation and conformational constraints imparted by two allenic moieties facilitated the macrocylization of stereochemical variants of the precursor— a serious challenge identified early on for this class of seco acids [3, 26]. The allenic functionality also enabled direct access (3–5 steps) to many functionalized 14-membered macrolides. These consisted entirely of aglycon structures, whereas the erythromycins (1 and 2) house specially functionalized glycosyl units. Consistent with structural studies of the erythromycins bound to the bacterial ribosome, all erythromycins that exhibit antibiotic activity are glycosylated with desosamine (e.g., 2) or an elaborated variant of this glycan [30, 31]. Here, we describe the synthesis of 6, a glycosylated allenic macrolide.
Results and discussion
We aimed to realize short routes to glycosylated macrocyclic scaffolds. We used thioglycoside donors 12 (see inset) and 9, Scheme 1. Thiopyrimidinyl donor forms of desosamine 12 have been the only donors used for erythronolide acceptors and follow the precedent set by Woodward and Toshima [3, 32]. Thiophenyl donor 9 offered other opportunities, including the option to explore the use of the corresponding sulfoxide [33]. We prepared 12 according to the known procedures and prepared 9 as outlined here beginning with desosamine. Conversion of desosamine hydrochloride to the carbonate (7→8) followed by installation of the thiophenyl group gave 9β as the major isomer. This material was isolated and taken on in an effort to avoid anticipated practical difficulties associated with manipulating isomeric mixtures of glycosides and sulfoxides. The direct treatment of thioglycoside 9β with meta-chloroperbenzoic acid gave the undesired N-oxide (not shown). However, the conversion of the amine to its trifluoroacetic acid salt prior to the addition of oxidant avoided this outcome and shunted the reaction to the desired sulfoxides (9β→10→11). Although the major isomer of 11 is crystalline and stable to long-term storage, this mixture of diastereomers was taken on and used together.
The incompatibility of allenes with typical glycosylation reaction conditions represents a major concern, as strong Lewis acidic salts used to activate glycosyl donors tend to react with allenes. With one exception, the reactions involving allenes and potential glycosides aimed to engage and, thereby transform, the allenic group [34]. Our interest was to effect glycosylation in the presence of a highly substituted allene and to retain that functionality for late-stage transformation. The sulfoxide donor represented an opportunity to achieve macrolide glycosylation in the presence of the allenic functionality by way of low-temperature generation of an oxocarbenium ion, which would be both expedient to us and would have thought-provoking implications. We first assessed sulfoxide 11 as a suitable desosamine donor in three simple reactions (Scheme 2a). Sterically encumbered secondary and tertiary alcohols, 2,4-dimethyl-3-pentanol, t-butanol, and 2-methyl-2-adamantanol, rapidly reacted in the presence of 11 under the agency of triflic anhydride to give the corresponding glycosides as single isomers and in yields ranging from 60 to 85%. Rather than optimizing the outcomes we moved from these proxy acceptors to macrolides.
Scheme 2 also summarizes our initial macrocycle glycosylation findings. Allenic diol 16 was combined with triflic anhydride and then sulfoxide donor 11 at a low temperature. The reaction mixture rapidly gave glycosylated products 17 and 18 (site selectivity = 1:1) as single stereoisomers (Scheme 2b). In contrast, donor 9β in the presence of silver hexafluorophosphate effected slow glycosylation, gave low yields of products with similar selectivity (site selectivity = 1:1), and was accompanied by substantial decomposition of 16 (Scheme 2c). Toshima conditions (pyrimidinyl phenylthio glycoside 12 and AgOTf) [32] gave extensive decomposition of 16 and no evidence of desired product formation (not shown). The combination of 16 with silver triflate in the absence of glycosyl donor gave rise to a similar decomposition profile. As expected, silver salts appear to undermine the desired glycosylation and engage in allenic reaction pathways. Nevertheless, we were encouraged by some of the findings and focused on the use of sulfoxide donor 11.
Scheme 3 shows the final stage of the investigation. Our earlier report described the preparation of 5, its conversion to 19 by allene osmylation, and its subsequent reduction to 16 (not shown) [29]. The facile O-trimethylsilylation of 19 via the agency of silyl chloride and catalytic 4-dimethylaminopyridine followed by sodium borohydride reduction gave the desired product as a single isomer (19→20→21; Scheme 3a). The structure of 21 was confirmed by the removal of the silyl group (toluene sulfonic acid in MeOH), and this product was shown to be identical with 16. The corresponding O-methyl ether 23 was difficult to install. Conventional mild methylating conditions, including those with methyl iodide, dimethyl sulfate, and Meerwein’s salt [35], failed to alkylate 19. Excess methyl triflate and 2,6-di-tert-butyl-4-methylpyridine in hot dichloroethane was required and gave multiple minor products along with the desired methyl ether 22. Glycosyl acceptor 23 was prepared as a single isomer by the reduction of 22 and was obtained in a moderate overall yield from 19. The silyl ether (21) resisted glycosylation. The methyl ether 23 reacted and gave 6 in modest yield (30%). The primary concern in the glycosylation was the stability of the allene to the cationic donor. The secondary concern was the stability of the allene to the phenylsulfenyl triflate generated under the reaction conditions [36]. Gratifyingly, the use of 4-allyl-1,2-dimethoxybenzene significantly improved the reaction outcome and furnished glycosylated macrolide 6 in 63% yield as a single isomer (Scheme 3b).
Our findings complement exciting recent developments in synthetic macrolide antibiotics and expand on our previously described methods to gain direct access to diverse structural motifs [19,20,21,22,23,24,25,26,27,28,29]. We have reduced to practice many new chemical transformations based on allene epoxidation, osmylation, and other chemistries [37]. These unexplored structures and reactivity profiles motivated the strategy embodied in macrocyclic bis[allene] 5 [27,28,29]. The approach has enabled high-tempo seco acid synthesis, facile macrocyclization, selective or simultaneous manipulation of the allenyl groups in a macrolide context, and glycosylation of the macrolide scaffold.
In summary, among the most important advances of this study is the preparation of novel desosamine donors 9 and 11 and the use of these donors to glycosylate bulky and macrolide acceptors. The sulfoxide reagent 11 represents a distinct alternative to Woodward–Toshima glycosylation of the erythromycins and other macrolides. Pleasingly, the reactive species derived from glycosyl donor 11 was generated in the presence of a highly substituted allene without complication. The resultant route to glycosylated allenic erythronolide 6 is short. The sequence required four steps beginning from 5 and corresponds to fifteen steps from commercial reagents (longest linear sequence). Needless to say, compound 6 represents an opportunity to diversify—post glycosylation—into the high-value antibiotic structure space of erythromycin.
References
Rossiter SE, Fletcher MH, Wuest WM. Natural products as platforms to overcome antibiotic resistance. Chem Rev. 2017;117:12415–74.
Pal S. A journey across the sequential development of macrolides and ketolides related to erythromycin. Tetrahedron. 2006;62:3171–200.
Woodward RB, Logusch E, Nambiar KP, Sakan K, Ward DE, Au-Yeung BW, et al. Asymmetric total synthesis of erythromycin. 3. Total synthesis of erythromycin. J Am Chem Soc. 1981;103:3215–7.
Martin SF, Hida T, Kym PR, Loft M, Hodgson A. The asymmetric synthesis of erythromycin B. J Am Chem Soc. 1997;119:3193–4.
Hergenrother PJ, Hodgson A, Judd AS, Lee WC, Martin SF. An abiotic strategy for the enantioselective synthesis of erythromycin B. Angew Chem Int Ed Engl. 2003;42:3278–81.
Breton P, Hergenrother PJ, Hida T, Hodgson A, Judd AS, Kraynack E, et al. Total synthesis of erythromycin B. Tetrahedron. 2007;63:5709–29.
Corey EJ, Hopkins PB, Kim S, Yoo SE, Nambiar KP, Falck JR. Total synthesis of erythromycins. 5. Total synthesis of erythronolide A. J Am Chem Soc. 1979;101:7131–4.
Nakata M, Arai M, Tomooka K, Ohsawa N, Kinoshita M. Total synthesis of erythronolide A. Bull Chem Soc Jpn. 1989;62:2618–35.
Muri D, Lohse‐Fraefel N, Carreira EM. Total synthesis of erythronolide a by MgII‐mediated cycloadditions of nitrile oxides. Angew Chem Int Ed. 2005;117:4104–6.
Muri D, Carreira EM. Stereoselective synthesis of erythronolide A via nitrile oxide cycloadditions and related studies. J Org Chem. 2009;74:8695–712.
Hikota M, Tone H, Horita K, Yonemitsu O. Stereoselective synthesis of erythronolide a by extremely efficient lactonization based on conformational adjustment and high activation of seco-acid1. Tetrahedron. 1990;46:4613–28.
Hikota M, Tone H, Horita K, Yonemitsu O. Chiral synthesis of polyketide-derived natural products. 27. Stereoselective synthesis of erythronolide A via an extremely efficient macrolactonization by the modified Yamaguchi method. J Org Chem. 1990;55:7–9.
Sviridov AF, Borodkin VS, Ermolenko MS, Yashunsky DV, Kochetkov NK. Stereocontrolled synthesis of erythronolides A and B in A (C5-C9) + (C3-C4) + (C1-C2) + (C11-C13) sequence from 1, 6-anhydro-β-D-glucopyranose (levoglucosan). Part 2. Tetrahedron. 1991;47:2317–36.
Corey EJ, Kim S, Yoo SE, Nicolaou KC, Melvin Jr LS, Brunelle DJ, et al. Total synthesis of erythromycins. 4. Total synthesis of erythronolide B. J Am Chem Soc. 1978;100:4620–2.
Sviridov AF, Ermolenko MS, Yashunsky DV, Borodkin VS, Kochetkov NK. Total synthesis of erythronolide B. 2. Skeleton assembly in (C5-C9) + (C3-C4) + (C1-C2) + (C11-C13) sequence. Tetrahedron Lett. 1987;28:3839–42.
Mulzer J, Kirstein HM, Buschmann J, Lehmann C, Luger P. Total synthesis of 9-dihydroerythronolide B derivatives and of erythronolide B. J Am Chem Soc. 1991;113:910–2.
Chandra B, Fu D, Nelson SG. Catalytic asymmetric synthesis of complex polypropionates: lewis base catalyzed aldol equivalents in the synthesis of erythronolide B. Angew Chem Int Ed. 2010;122:2645–8.
Myles DC, Danishefsky SJ, Schulte G. Development of a fully synthetic stereoselective route to 6-deoxyerythronolide B by reiterative applications of the lewis acid catalyzed diene aldehyde cyclocondensation reaction: a remarkable instance of diastereofacial selectivity. J Org Chem. 1990;55:1636–48.
Kim HC, Kang SH. Total synthesis of azithromycin. Angew Chem Int Ed. 2009;48:1827–9.
Velvadapu V, Paul T, Wagh B, Klepacki D, Guvench O, Mackerell JA, et al. Desmethyl macrolide analogues to address antibiotic resistance: total synthesis and biological evaluation of 4, 8, 10-tridesmethyl telithromycin. ACS Med Chem Lett. 2011;2:68–72.
Velvadapu V, Paul T, Wagh B, Glassford I, DeBrosse C, Andrade RB. Total synthesis of (−)-4,8,10-tridesmethyl telithromycin. J Org Chem. 2011;76:7516–27.
Velvadapu V, Glassford I, Lee M, Paul T, DeBrosse C, Klepacki D, et al. Desmethyl macrolides: synthesis and evaluation of 4, 10-didesmethyl telithromycin. ACS Med Chem Lett. 2012;3:211–5.
Wagh B, Paul T, Glassford I, DeBrosse C, Klepacki D, Small MC, et al. Desmethyl macrolides: synthesis and evaluation of 4, 8-didesmethyl telithromycin. ACS Med Chem Lett. 2012;3:1013–8.
Glassford I, Lee M, Wagh B, Velvadapu V, Paul T, Sandelin G, et al. Desmethyl macrolides: synthesis and evaluation of 4-desmethyl telithromycin. ACS Med Chem Lett. 2014;5:1021–6.
Andrade RB. Total synthesis of desmethyl macrolide antibiotics. Synlett. 2015;26:2199–215.
Seiple IB, Zhang Z, Jakubec P, Langlois-Mercier A, Wright PM, Hog DT, et al. A platform for the discovery of new macrolide antibiotics. Nature. 2016;533:338–45.
Ghosh P, Zhang Y, Emge TJ, Williams LJ. Modeling a macrocyclic bis [spirodiepoxide] strategy to erythronolide A. Org Lett. 2009;11:4402–5.
Liu K, Kim H, Ghosh P, Akhmedov NG, Williams LJ. Direct entry to erythronolides via a cyclic bis [allene]. J Am Chem Soc. 2011;133:14968–71.
Yu L, Wang H, Akhmedov NG, Sowa C, Liu K, Kim H, et al. Direct entry to 4, 10-didesmethyl (9S)-dihydroerythronolide A via catalytic allene osmylation. Org Lett. 2016;18:2868–71.
Bulkley D, Innis CA, Blaha G, Steitz TA. Revisiting the structures of several antibiotics bound to the bacterial ribosome. Proc Natl Acad Sci USA. 2010;107:17158–63.
Tu D, Blaha G, Moore PB, Steitz TA. Structures of MLSBK antibiotics bound to mutated large ribosomal subunits provide a structural explanation for resistance. Cell . 2005;121:257–70.
Toshima K, Nozaki Y, Mukaiyama S, Tamai T, Nakata M, Tatsuta K, et al. Application of highly stereocontrolled glycosidations employing 2, 6-anhydro-2-thio sugars to the syntheses of erythromycin A and olivomycin A trisaccharide. J Am Chem Soc. 1995;117:3717–27.
Kahne D, Walker S, Cheng Y, Van Engen D. Glycosylation of unreactive substrates. J Am Chem Soc. 1989;111:6881–2.
Haydl AM, Breit B. Atom-eonomical dimerization strategy by the rhodium-catalyzed addition of carboxylic acids to allenes: protecting-group-free synthesis of clavosolide A and late-stage modification. Angew Chem Int Ed. 2015;127:15750–4.
Meerwein H, Hinz G, Hofmann P, Kroning E, Pfeil E. Über Tertiäre Oxoniumsalze, I. J Prakt Chem. 1937;147:257–85.
Gildersleeve J, Smith A, Sakurai K, Raghavan S, Kahne D. Scavenging byproducts in the sulfoxide glycosylation reaction: application to the synthesis of ciclamycin 0. J Am Chem Soc. 1999;121:6176–82.
Sharma R, Manpadi M, Zhang Y, Kim H, Akhmedov NG, Williams LJ. Spirodiepoxide-based cascades: direct access to diverse motifs. Org Lett. 2011;13:3352–5.
Acknowledgements
We gratefully acknowledge the partial financial support from the NIH (GM078145)
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare no conflict of interest.
Additional information
Dedication: This manuscript is dedicated with respect and admiration to Professor S J Danishefsky in recognition of his creative force, caring mentorship, and tremendous contributions to chemistry.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
About this article
Cite this article
Yu, L., Wang, H., Akhmedov, N. et al. Glycosylation of an allenic erythronolide. J Antibiot 72, 432–436 (2019). https://doi.org/10.1038/s41429-019-0156-1
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41429-019-0156-1