115 Progress in the Chemistry of Organic Natural Products Progress in the Chemistry of Organic Natural Products

Home , Luis Federico Leloir, Norman Haworth

Progress in the Chemistry of Organic Natural Products

A. Douglas Kinghorn · Heinz Falk · Simon Gibbons · Yoshinori Asakawa · Ji-Kai Liu · Verena M. Dirsch Editors 115 Progress in the Chemistry of Organic Natural Products Progress in the Chemistry of Organic Natural Products

Series Editors A. Douglas Kinghorn , College of Pharmacy, The Ohio State University, Columbus, OH, USA Heinz Falk , Institute of Organic Chemistry, University Linz, Linz, Austria Simon Gibbons , School of Pharmacy, University of East Anglia, Norwich, UK Yoshinori Asakawa , Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Tokushima, Japan Ji-Kai Liu , School of Pharmaceutical Sciences, South-Central University for Nationalities, Wuhan, China Verena M. Dirsch , Department of Pharmacognosy, University of Vienna, Vienna, Wien, Austria

Advisory Editors Giovanni Appendino , Department of Pharmaceutical Sciences, University of Eastern Piedmont, Novara, Italy Roberto G. S. Berlinck , Instituto de Química de São Carlos, Universidade de São Paulo, São Carlos, Brazil Jun’ichi Kobayashi, Graduate School of Pharmaceutical Sciences, Hokkaido University, Sapporo, Japan Agnieszka Ludwiczuk , Department of Pharmacognosy, Medical University of Lublin, Lublin, Poland C. Benjamin Naman , Department of Marine Pharmacy, Ningbo University, Zhejiang, China Rachel Mata , Facultad de Química, Universidad Nacional Autónoma de México, Mexico City, Distrito Federal, Mexico Nicholas H. Oberlies , Department of Chemistry and Biochemistry, University of North Carolina, Greensboro, NC, USA Deniz Tasdemir , Marine Natural Products Chemistry, GEOMAR Helmholtz Centre for Ocean Research, Kiel, Schleswig-Holstein, Germany Dirk Trauner , Department of Chemistry, New York University, New York, NY, USA Alvaro Viljoen , Department of Pharmaceutical Sciences, Tshwane University of Technology, Pretoria, South Africa Yang Ye , State Key Laboratory of Drug Research and Natural Products Chemistry Department, Shanghai Institute of Materia Medical, Shanghai, China The volumes of this classic series, now referred to simply as “Zechmeister” after its founder, Laszlo Zechmeister, have appeared under the Springer Imprint ever since the series’ inauguration in 1938. It is therefore not really surprising to ﬁnd out that the list of contributing authors, who were awarded a Nobel Prize, is quite long: Kurt Alder, Derek H.R. Barton, George Wells Beadle, Dorothy Crowfoot-Hodgkin, Otto Diels, Hans von Euler-Chelpin, Paul Karrer, Luis Federico Leloir, Linus Pauling, Vladimir Prelog, with Walter Norman Haworth and Adolf F.J. Butenandt serving as members of the editorial board. The volumes contain contributions on various topics related to the origin, distribution, chemistry, synthesis, biochemistry, function or use of various classes of naturally occurring substances ranging from small molecules to biopolymers. Each contribution is written by a recognized authority in the ﬁeld and provides a comprehensive and up-to-date review of the topic in question. Addressed to biologists, technologists, and chemists alike, the series can be used by the expert as a source of information and literature citations and by the non-expert as a means of orientation in a rapidly developing discipline. All contributions are listed in PubMed.

More information about this series at http://www.springer.com/series/10169 A. Douglas Kinghorn • Heinz Falk • Simon Gibbons • Yoshinori Asakawa • Ji-Kai Liu • Verena M. Dirsch Editors

Progress in the Chemistry of Organic Natural Products

Volume 115

With contributions by

Bernd Schmidt Søren Brøgger Christensen Á Henrik Toft Simonsen Á Nikolai Engedal Á Poul Nissen Á Jesper Vuust Møller Á Samuel R. Denmeade Á John T. Isaacs Jiří Pospíšil Á Daniela Konrádová Á Miroslav Strnad Steven D. Shnyder Á Colin W. Wright 123 Editors A. Douglas Kinghorn Heinz Falk College of Pharmacy Institute of Organic Chemistry Ohio State University Johannes Kepler University Columbus, OH, USA Linz, Oberösterreich, Austria

Simon Gibbons Yoshinori Asakawa School of Pharmacy Faculty of Pharmaceutical Sciences University of East Anglia Tokushima Bunri University Norwich, UK Tokushima, Japan

Ji-Kai Liu Verena M. Dirsch School of Pharmaceutical Sciences Department of Pharmacognosy South Central University for Nationaliti University of Vienna Wuhan, China Wien, Austria

ISSN 2191-7043 ISSN 2192-4309 (electronic) Progress in the Chemistry of Organic Natural Products ISBN 978-3-030-64852-7 ISBN 978-3-030-64853-4 (eBook) https://doi.org/10.1007/978-3-030-64853-4

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021, corrected publication 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Preface

It is a great pleasure to announce that with effect from the present Volume 115, Univ. Prof. Dr. Verena M. Dirsch, University Professor and Head, Department of Pharmacognosy, University of Vienna, Vienna, Austria has become a new Series Editor of “Progress in the Chemistry of Organic Natural Products”. Prof. Dirsch has published widely, and has worked in particular on bioassay method development and mechanism-of-action studies on naturally occurring molecules in medicinal plants. She has been a member of the Editorial Advisory Board of this book series since 2008 (Volume 92). Prof. Dirsch was responsible for organizing the seven chapters from internationally renowned authors of Volume 110 (“Cheminformatics in Natural Product Research”; 2019), with the purpose of this volume being to show how “big data” from natural product collections and databases can be mined by computerized methods for molecular target identiﬁcation. In addition to being an active scientist, Prof. Dirsch has had a number of additional senior administrative roles at the University of Vienna, including Vice-Dean of the Faculty of Life Sciences (2008–2014) and Deputy Speaker of the Center of Pharmaceutical Sciences (2014–present). Prof. Dirsch is warmly welcomed into her new role for our book series. Professor Jun’ichi Kobayashi, of the Graduate School of Pharmaceutical Sciences, Hokkaido University, Sapporo, Japan has stepped down from his past position as Series Editor for this book series, and will become a member of the Editorial Advisory Board. The other Series Editors and Springer are very grateful for his many contributions to the book series for Volumes 92–114 (2008–2021). Prof. Kobayashi has had an illustrious academic career in the natural products area, and has published prodigiously on the structure elucidation of bioactive molecules from all of marine organisms, terrestrial microbes, and higher plants. He has also been a research supervisor to over 100 graduate students in the past. In Volume 115, there are four chapters altogether that feature a varied group of natural products. In the ﬁrst of these, Prof. Dr. Bernd Schmidt, of the University of Potsdam, Germany, has written on the role that total synthesis may play in revising the structures proposed for decanolides, which are ten-membered lactones found primarily in fungi, frogs, and termites. Prof. Søren Brøgger Christensen, of the

v vi Preface

University of Copenhagen, Denmark, and colleagues describe the development of the intriguing plant-derived sesquiterpene lactone, thapsigargin, a potent inhi- bitor of the enzyme, SERCA (sarco-endoplasmic Ca2+ ATPase), with potential as a lead compound to treat cancer. The third chapter was written by a team headed by Prof. Miroslav Strnad, of the Czech Academy of Sciences, Institute of Experimental Botany, and Palacký University, Olomouc, Czech Republic. This covers the potential of various plant phenolic compounds for treating the tropical and sub-tropical infectious disease, leishmaniasis. In the ﬁnal chapter, Dr. Steven Shnyder and Prof. Colin Wright, of the University of Bradford, U.K. describe recent advances on the plant alkaloid, cryptolepine, which is of particular interest as a lead for the treatment of malaria, trypanosomiasis, and cancer.

Columbus, OH, USA Douglas Kinghorn Linz, Austria Heinz Falk Norwich, UK Simon Gibbons Tokushima, Japan Yoshinori Asakawa Wuhan, China Ji-Kai Liu Contents

The Role of Total Synthesis in Structure Revision and Elucidation of Decanolides (Nonanolides) ...... 1 Bernd Schmidt From Plant to Patient: Thapsigargin, a Tool for Understanding Natural Product Chemistry, Total Syntheses, Biosynthesis, Taxonomy, ATPases, Cell Death, and Drug Development ...... 59 Søren Brøgger Christensen, Henrik Toft Simonsen, Nikolai Engedal, Poul Nissen, Jesper Vuust Møller, Samuel R. Denmeade, and John T. Isaacs Antileishmanial Activity of Lignans, Neolignans, and Other Plant Phenols ...... 115 Jiří Pospíšil, Daniela Konrádová, and Miroslav Strnad Recent Advances in the Chemistry and Pharmacology of Cryptolepine ...... 177 Steven D. Shnyder and Colin W. Wright Correction to: Antileishmanial Activity of Lignans, Neolignans, and Other Plant Phenols ...... C1 Jiří Pospíšil, Daniela Konrádová, and Miroslav Strnad

vii The Role of Total Synthesis in Structure Revision and Elucidation of Decanolides (Nonanolides)

Bernd Schmidt

Contents

Abbreviations ...... 1 1 Introduction ...... 3 2 Recurring Methods in Decanolide Synthesis ...... 7 2.1 EsterificationandMacrolactonizationReactions ...... 9 2.2 Ring-ClosingOlefinMetathesis ...... 11 2.3 OlefinCross-Metathesis ...... 16 3 Examples of the Synthesis-Assisted Structure Elucidation of Ten-Membered Lactone Natural Products ...... 18 3.1 Fusanolides, Modiolides or Curvulalic Acid? ...... 18 3.2 Stagonolides and Curvulide A ...... 23 3.3 Seimatopolides ...... 34 3.4 Pheromones from Madagascan Mantellid Frogs ...... 41 3.5 A Potential Pheromone from the Queen of the Termite Silvestritermes minutus ...... 46 4 Conclusions...... 49 References...... 50

Abbreviations

Ac Acetyl BINAP 2,2-Bis(diphenylphosphino)-1,1-binaphthyl CD Circular dichroism CM Cross-metathesis COSY Correlation spectroscopy

B. Schmidt (B) Institut für Chemie, Universität Potsdam, Karl-Liebknecht-Strasse 24–25, 14476 Potsdam-Golm, Germany e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 1 A. D. Kinghorn, H. Falk, S. Gibbons, Y. Asakawa, J.-K. Liu, V. M. Dirsch (eds.), Progress in the Chemistry of Organic Natural Products, Vol. 115, https://doi.org/10.1007/978-3-030-64853-4_1 2 B. Schmidt

CSA Camphor sulfonic acid DCC Dicyclohexyl carbodiimide DDQ 2,3-Dichloro-5,6-dicyano-1,4-benzoquinone DET Diethyl tartrate DMAP 4-N,N-Dimethylamino pyridine ECM Exciton chirality method EDC 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide ee Enantiomeric excess EI Electron ionization EOM Ethoxymethyl ESI Electrospray ionization GC-FTIR Gas chromatography—Fourier transform infrared spectroscopy (-coupling) GC-MS Gas chromatography—mass spectrometry (-coupling) GCxGC-TOFMS Two-dimensional gas chromatography/time-of-flight mass spectrometry (-coupling) HMBC Heteronuclear multiple bond correlation HRMS High-resolution mass spectrometry HSQC Heteronuclear single quantum coherence Ipc Isopinocampheyl IR Infrared (-spectroscopy) LAH Lithium aluminum hydride m-CPBA meta-Chloroperbenzoic acid MNBA 2-Methyl-6-nitrobenzoic anhydride MOM Methoxymethyl NHC N-Heterocyclic carbene NHK Nozaki–Hiyama–Kishi (-coupling) NMR Nuclear magnetic resonance (-spectroscopy) 1D-NMR One-dimensional nuclear magnetic resonance (-spectroscopy) 2D-NMR Two-dimensional nuclear magnetic resonance (-spectroscopy) NOE Nuclear Overhauser effect NOESY Nuclear Overhauser enhancement and exchange spectroscopy PMB p-Methoxybenzyl PPAR Peroxisome proliferator-activated receptors RCM Ring-closing metathesis ROESY Rotating frame Overhauser effect spectroscopy salen 2,2-Ethylenebis(nitrilomethylidene)diphenol (-ligand) TBAF Tetrabutylammonium fluoride TBDPS tert-Butyldiphenylsilyl TBS tert-Butyldimethylsilyl TEMPO 2,2,6,6-Tetramethylpiperidine-1-oxyl TES Triethylsilyl OTf Trifluoromethanesulfonate Tos p-Tolylsulfonate The Role of Total Synthesis in Structure … 3

1 Introduction

Decanolides, sometimes referred to as nonanolides, are naturally occurring ten- membered lactones. The first compound with this general structure was isolated from jasmine oil in 1942 [1], and later named jasmine ketolactone (1). The ten- membered lactone structure could not be assigned until 1964; this assignment was based on IR and NMR spectroscopic investigations, but elucidation of the absolute and relative configuration required a four-step conversion into the known (–)-methyl jasmonate (2), and comparison with the previously published specific rotation of 2 [2] (Scheme 1). In their 1996 landmark review on ten-membered lactone natural products [3], Dräger et al. named the diplodialides as the first examples of decanolides, probably because these metabolites are, like the majority of decanolides known to date (and in contrast to jasmine ketolactone (1)), pentaketides and originate from a microbial source, the plant pathogenic fungus Diplodia pinea [4]. Diplodia pinea is the causal agent of the diplodia tip blight, a fungal disease that affects pines and other conifers when exposed to stressed conditions (Plates 1 and 2). The first three decanolides from this source were discovered in 1975 by Ishida and Wada and named diplodialides A–C [4], and a fourth metabolite was isolated in the following year in very small quantities and named diplodialide D [5]. The strategy for structure elucidation of the diplodialides in the 1970s was very similar to that employed for jasmine ketolactone (1) a decade earlier: the ten-membered lactone structure and the (E)-configuration of the endocyclic C=C double-bond were established through a combination of IR and NMR spectroscopy. However, in contrast to the work on jasmine ketolactone (1) ten years before, high-resolution mass spectrometry (HRMS) had entered the scene and provided valuable information for eluci- dating the constitution. Assignment of absolute configurations to the stereocenters at

Scheme 1 Historical elucidation of absolute and relative conﬁguration of (–)-jasmine ketolactone (1) 4 B. Schmidt

Plate 1 Browning of needles of a pine tree caused by diplodia tip blight. Photograph courtesy of Pﬂanzenschutzamt Berlin. Copyright Pﬂanzenschutzamt Berlin. Reproduced with permission

Plate 2 Black fruiting structures of Diplodia pinea. Photograph courtesy of Pﬂanzenschutzamt Berlin. Copyright Pﬂanzenschutzamt Berlin. Reproduced with permission The Role of Total Synthesis in Structure … 5

C-3 and C-9 still required synthetic degradation to known compounds and comparison with their published specific rotations, occasionally after derivatization [6]. By oxidation of the secondary alcohol of diplodialide B (4) and hydrogenation of the C-4=C-5 double bond, Ishida and Wada were able to prove that the absolute configuration at C-9 is identical for diplodialides A–C. The (R)-configuration assigned to 3–5 was proven by reductive ozonolysis, which gave (R)-hexane-1,5-diol (7), a compound that could be compared to its known enantiomer. The configuration of diplodialides B (4) and C (5) at C-3 was elucidated by conversion of 4 to the known p-nitrobenzoate 9 of dimethyl malate in five steps: acetylation of 4 furnished the acetate 8, which underwent oxidative ozonolysis, basic cleavage of the acetate and the lactone, methylation with diazomethane and eventually derivatization as a para-nitrobenzoate 9 (Scheme 2). In their original publications, Wada and Ishida presented the structures of diplodialides A–D (3–6) as shown in Scheme 2, without assigning any absolute or relative configurations for diplodialide D (6)[5, 6]. Insufficient amounts of material prevented a further investigation at that time. Although it appeared highly likely that the configurations at C-3 and C-9 should be the same as for diplodialides B (4) and C (5), this was not established unambiguously until 2018, when the first enantioselective total synthesis of diplodialide D (6) was published by Ramanujan and Kumar [7]. These researchers started from (S)-glycidol tosylate (10), an enantiomerically pure building block with a reliably assigned absolute configuration. Epoxide 10 is available via Sharpless epoxidation of allyl alcohol [8] and was initially converted to compounds 11 and 12 in two and five steps, respectively. Both electrophilic coupling partners

Scheme 2 Historical structure elucidation of diplodialides A–C 6 B. Schmidt were successively coupled with lithiated 1,3-dithiane (13) in a Corey–Seebach reaction to furnish 14 [9]. In three steps the ω-hydroxycarboxylic acid 15 was obtained, which was cyclized to the ten-membered lactone 16 through a Yamaguchi macrolactonization [10, 11] and oxidative cleavage of the protecting group. From lactone 16, diplodialide C (5) was accessible via reductive cleavage of the dithiane, and diplodialide D (6) via hydrolytic cleavage of the dithiane (Scheme 3). Analytical data obtained for synthetic diplodialides C and D, including specific rotations, matched those previously reported by Wada and Ishida for the compounds isolated from the natural source [5, 6]. This eventually corroborated the assumed, but not proven, absolute configurations of the stereocenters and completed the structure elucidation of diplodialide D (6) after four decades. Over the twenty years following the isolation and structure elucidation of the diplodialides, ca. 50 decanolides were isolated from natural sources. The status of decanolide research in the mid-1990s, including a survey of compounds isolated between 1975 and 1995, results from biosynthesis studies, and investigations into their biological properties and an overview of strategies for their total synthesis have been summarized in the first review dedicated exclusively to the topic of decanolides

Scheme 3 Enantioselective total syntheses of diplodialides C (5)andD(6) and conﬁrmation of assigned absolute conﬁgurations The Role of Total Synthesis in Structure … 7

[3]. In this review, Dräger et al. noted that more than 80% of the decanolides known at that time originate from microbial sources (mostly fungi), where 15% were isolated from animals and just one decanolide was isolated from a plant, that is, jasmine ketolactone (1). In 2012, Sun et al. published a second review dedicated to the same topics that specifically covered the period after publication of the first review by Dräger et al. until mid-2011 [12]. Sun et al. noted the isolation and description of 63 new decanolides in the chemical literature during this time. Fungi remained the main source of decanolides. Other reviews were not intended to be comprehensive but covered specific sub-topics, for example, synthesis studies of specific decanolides [13]. Since the 1970s, instrumental analytical techniques have seen tremendous progress, which has changed the strategies used for structure elucidation of natural products fundamentally. Synthetic interconversion and degradation of novel metabolites to known compounds are used rarely nowadays to assign structures; in particular, because very sensitive high-field NMR spectrometers are routinely available, and elaborate 1D- and 2D-NMR experiments can be performed in a short time. In addition, the substantial expansion of ionization techniques available for mass spectrometry facilitates the measurement of high-resolution mass spectra and the reliable determination of a molecular formula, double-bond equivalents and ring closures. In light of this technological progress, it appears somewhat surprising that until the present time the structural assignment of newly isolated decanolides has remained incomplete occasionally and has turned out to be in part erroneous. This can—as in Wada’s and Ishida’s case of diplodialide D forty years ago—often be explained by the very small amounts of new metabolites obtained by isolation from the natural source, which makes purification and hence interpretation of the NMR spectra often very difficult. In such cases, the total synthesis of natural products from enantiopure starting materials with reliably assigned absolute configurations (either obtained ex- chiral pool or corroborated otherwise) can assist in structure elucidation. This may be done by proving an assumed structure, by completing structural information (which is mostly the case when configurations of one or more stereocenters remain elusive by NMR spectroscopy), or by revising erroneously assigned structures. Herein, the contribution of total synthesis to the structure elucidation of naturally occurring decanolides is discussed by representative case studies. Results from isolation studies and investigations into the biosynthesis and bioactivities of certain compounds will be mentioned briefly, if appropriate, but they are not the primary focus of this chapter.

2 Recurring Methods in Decanolide Synthesis

In their major review from 1996, Dräger et al. classiﬁed synthesis approaches to decanolides as follows: “(i) oxidative fragmentation of annulated decanolides, (ii) ring closure by C–C-bond formation, (iii) ring closure by macrolactonization” [3]. 8 B. Schmidt

The methods for ring closure by C–C-bond formation used at that time were Pd- catalyzed allylic substitutions, SmI2-mediated Reformatsky reactions or intramolecular radical additions to C=C double-bonds. These methods, as well as the oxidative fragmentation of annulated decanolides, no longer play a significant role in the total synthesis of ten-membered lactones due to the development of efficient olefin metathesis catalysts starting in the mid-1990s. Comparison of the review on decanolides in general by Sun et al. from 2012 [12] and with Ishigami’s review on the synthesis of selected ten-membered lactone natural products from 2009 [13] reveals how olefin metathesis has revolutionized the total synthesis of decanolides. However, macrolactonization remains a method of constant importance over the decades; in particular, Yamaguchi’s method has been, and is, used routinely in decanolide synthesis [14]. As a notable example for the efficient use of an oxidative ring expansion strategy [15] in contemporary decanolide total synthesis, the cephalosporolides should be mentioned [16–18]. Many contemporary total syntheses of decanolide natural products proceed via either of these two general strategies: (i) an ω-enoic acid 17 and an ω-alkenol 18 are esterified (most commonly using Steglich’s esterification or a variant [19]) and then cyclized by ring-closing olefin metathesis (RCM) or (ii) an ω-enoic acid 17 and an ω-alkenol 18 are connected through cross-metathesis (CM) and the resulting seco- acid 20 is then cyclized using Yamaguchi macrolactonization [14] or an alternative mixed-anhydride method (Scheme 4). An alternative way to connect two advanced fragments to an unsaturated seco- acid 20, which is subsequently cyclized via macrolactonization, involves carbonyl olefination methods. The Julia–Kocienski olefination, for example, often shows very high levels of (E)-stereoselectivity [20, 21]. This method has been used as a key step in the synthesis of the plant metabolites cytospolides C and D [22]. In this section a brief overview of methods often used in the synthesis of naturally occurring decanolides is given.

Scheme 4 General strategies for contemporary decanolide total synthesis The Role of Total Synthesis in Structure … 9

2.1 Esteriﬁcation and Macrolactonization Reactions

Steglich’s reaction [19] is the most commonly used esterification method in decanolide total synthesis. The method relies on the use of dicyclohexylcarbodiimide (DCC) in combination with a catalytic amount of N,N-dimethylaminopyridine (DMAP) in a moderately polar aprotic solvent, such as dichloromethane. A typical example is the esterification step in Prisinzano’s synthesis of aspinolide A ((5R)-25) and 5-epi-aspinolide A ((5S)-25)[23]. Sometimes 1-ethyl-3-(3- dimethylaminopropyl)carbodiimide (EDCI), which is normally employed as its hydrochloride (EDCI•HCl), is used advantageously [24]. In contrast to DCC, the use of EDCI allows aqueous workup at pH 4–6 to remove the urea co-product formed during the esterification. An example is the coupling of carboxylic acid 26 and alcohol 27 to ester 28, a precursor of the ten-membered lactone pinolide [25]. In both examples, a carboxylic acid and an alcohol can be used in equimolar or near equimolar amounts (Scheme 5). The carbodiimides DCC and EDCI•HCl were originally introduced as peptide coupling reagents. An overview of these and related reagents is given in a study on the assessment of their thermal stability [26]. Although Yamaguchi’s method has occasionally been used for difficult intermolecular esterifications instead of Steglich’s conditions, this method shows its strength best in its intramolecular version, the lactonization. Yamaguchi’s reaction relies on the formation of an activated mixed anhydride of the carboxylic acid to be coupled with a nucleophile and 2,4,6-trichlorobenzoyl chloride [14]. In contrast

Scheme 5 Examples of Steglich esteriﬁcations 10 B. Schmidt to intermolecular esteriﬁcations, high- or pseudo-high dilution conditions have to be applied for medium-sized macrolactones [27]. For example, in a synthesis of stagonolide B (31), the hydroxycarboxylic acid 29 underwent lactonization to 30 under Yamaguchi conditions when a solution of the mixed anhydride (formed in situ from 29 and 2,4,6-trichlorobenzoylchloride) was added to a highly diluted solution of DMAP in toluene via syringe pump. The effective concentration was 0.5 mM in this case [28]. Occasionally, Yamaguchi’s macrolactonization conditions have been reported to produce only dimers or polymers, or the reagent reacts with other nucle- ophilic groups in the lactonization precursor. All of these problems occurred in a total synthesis of cytospolide D (35) when Yamaguchi’s conditions were tested for the lactonization of 32 [29, 30], but could be overcome by using Shiina’s reagent, 2- methyl-6-nitrobenzoic anhydride (MNBA), instead [31]. Like Yamaguchi’s reagent MNBA also reacts with the carboxylic acid to a mixed anhydride. In this particular case the elimination product 34 was observed as a side product, which presumably is formed through intermediate formation of a β-lactone [29] (Scheme 6). In a follow-

Scheme 6 Examples of the application of Yamaguchi and Shiina macrolactonization The Role of Total Synthesis in Structure … 11 up study on the synthesis of cytospolide D analogs, it was discovered that small differences in the structures of the precursors can be crucial for the success of a macrolactonization. If, for instance, the C-3 hydroxy group of 32 is protected as a TBS ether and/or if the C-2 epimer of 32 is used, the lactonization fails completely, even under optimized Shiina conditions [30]. Contemporary methods for ester coupling reactions, their mechanisms and representative applications in the synthesis of complex natural products have been reviewed [27].

2.2 Ring-Closing Oleﬁn Metathesis

Ring-closing oleﬁn metathesis (RCM) has emerged as one of the most versatile C–C-bond forming reactions used in the synthesis of carbo- and heterocyclic target molecules [32, 33], and was triggered by the discovery of stable and deﬁned homoge- neous precatalysts based on Ru- [34, 35] and Mo-carbenes [36]. The general mechanism of an RCM reaction leading to an unsubstituted ten-membered lactone is shown in Scheme 7. It proceeds through a series of (2+2)-cycloaddition/cycloreversion reac-

Scheme 7 General mechanism of an RCM reaction 12 B. Schmidt tions. The catalytically active species, a metal alkylidene complex, reacts with one of the two double-bonds to a metallacyclobutane, which undergoes a cycloreversion with liberation of ethene. The new metal alkylidene reacts intramolecularly with the second C=C double-bond of the starting material to a bicyclic metallacyclobutane, which eventually undergoes a cycloreversion to regenerate the catalyst and furnish the product. Ring-closing metathesis reactions are entropy-driven and rely on the formation of a volatile co-product, which in most cases is ethene. In principle, all steps (and the overall reaction) are reversible, but removal of ethene shifts the equilibrium to the formation of the product [37]. Due to their robustness, low sensitivity toward air and moisture and high functional group tolerance in particular, the Ru-based precatalysts introduced by Grubbs and co-workers have found many applications in the total synthesis of natural products [38–42]. First-generation catalysts (A) are less active (which can be an advantage in some cases, as will be discussed below) but are cheaper and more conveniently accessible. Second-generation catalysts (B)have one N-heterocyclic carbene (NHC) ligand. They are more active and therefore lower catalyst loadings are required. These catalysts are also suitable for the generation of triple- and even tetra-substituted double bonds, but they are more expensive and their synthesis requires additional steps. The two most commonly used metathesis catalysts are A1 [34] and B1 [35]. While the formation of five- and six-membered rings through RCM proceeds without difficulties in most cases at substrate concentrations >0.1 M and catalyst loadings <2 mol%, medium-sized rings normally require higher dilution and increased catalyst loadings of ca. 10 mol%. Whenever an RCM reaction is slow, as in the case of medium-sized rings, catalyst decomposition can successfully compete with the olefin metathesis reaction. This problem makes it necessary to use increased amounts of catalyst from the outset or to add the catalyst in portions over a longer period of time. In RCM reactions giving five- to nine-membered rings the configuration of the newly formed double-bond is always cis due to steric reasons. However, for ten-membered and larger rings the RCM reactions can yield either (E)- or (Z)- configured cycloalkenes, or mixtures of both diastereomers. This limitation was observed initially in the first synthesis of a ten-membered ring via RCM ever docu- mented in the literature. In 1997, Fürstner and Müller reported on the synthesis of racemic jasmine ketolactone ((rac)-1, cf. Scheme 1)[43], which is obtained by RCM of diene 36 in the presence of 10 mol% of catalyst A2 (an ancestor of the first-generation Grubbs’ catalyst A1 with comparable activity) [44]. The (Z)- and (E)-isomers were obtained in the ratio of 2.5:1 in toluene at 70°C. A small, but detrimental effect on the selectivity was observed when dichloromethane at 40°C was used as a solvent; under these conditions a similar yield was obtained, but the (Z):(E) ratio decreased to 1.4:1 (Scheme 8). Controlling the double-bond configuration in RCM macro-cyclization reactions became an important issue in subsequent years. Kalesse et al. discovered in the course of a study directed at the synthesis of epothilones [45] that diene 37 cyclizes preferably to decanolide (Z)-38, but that the minor isomer (E)-38, after chromatographic isolation, can be converted to (Z)-38 by exposing it to the original RCM conditions. This observation suggests that (Z)-38 is the thermodynamically preferred product, The Role of Total Synthesis in Structure … 13

Scheme 8 Synthesis of jasmine ketolactone via RCM which was corroborated by theoretical calculations, whereas (E)-38 is the kinetic product that can react to the (Z)-isomer via ring-opening/ring-closing metathesis [46] (Scheme 9). The development of the significantly more active second-generation catalysts B allowed the control of the double-bond configuration to a certain extent. In the course of their syntheses of herbarumins I and II, Fürstner and co-workers found that the first-generation catalyst A3 converts the acetonide-protected precursor 39 selectively to (E)-40, which was deprotected to furnish herbarumin I (41). Under otherwise identical conditions the same precursor 39 was cyclized to (Z)-40 when the second-generation catalyst B2 was used [47]. The different stereoselectivities can be understood when thermodynamic and kinetic control are taken into account: in this case, the diastereomer (Z)-40 is thermodynamically more stable than (E)-40,but the first-generation catalyst A3 is not sufficiently active to promote a ring opening metathesis of the kinetic product (E)-40. The more active catalyst B2, however, can initiate an equilibration of (E)- and (Z)-40 via ring-opening/ring-closing metathesis (Scheme 10). A similar observation was made during a synthesis of pinolide through the RCM of diene 28 (cf. Scheme 5). With the less active first-generation catalyst A1,a fully protected pinolide was selectively obtained in the required (E)-configuration, whereas the second-generation catalyst B1 gavea2:1mixtureof(E)- and (Z)-isomers [48]. These examples illustrate that in unsaturated ten-membered rings the relative thermodynamic stabilities of (E)- and (Z)-isomers are not only governed by the

Scheme 9 Thermodynamic versus kinetic control of (E)/(Z)-selectivity 14 B. Schmidt

Scheme 10 Control of (E)/(Z)-selectivity through choice of catalyst configuration of the endocyclic double-bond but also by the number, steric demand and relative configurations of substituents. As a conclusion, it can be stated that in those cases where the target structure is the thermodynamically less stable geomet- rical isomer, first-generation Ru-alkylidenes will be the catalysts of choice because the RCM step should be kinetically controlled. Another option for tuning the yield and stereoselectivity of the RCM step is a deliberate design of the precursor. By choosing more or less sterically demanding or conformationally restraining protecting groups, or by readjusting the configuration of certain substituents after the ring closure, it is sometimes possible to facil- itate the RCM step and favor the desired double-bond configuration. An example is the synthesis of ent-pinolidoxin ((ent)-45)[49], which led—together with other syntheses published simultaneously [47]—to a revision of the originally assigned absolute configuration [50]. The acetonide-protected RCM precursor 42 undergoes highly (Z)-selective RCM with the second-generation catalyst B1, whereas the diol 44 reacts in the presence of the same catalyst preferably to the (E)-configured product, (ent)-pinolidoxin ((ent)-45), unfortunately only in a 2:1 ratio. The (E):(Z) ratio can be increased to 4.9:1 if the sorbate is removed and the RCM is performed with the triol (Scheme 11). From the example of pinolidoxin outlined above, it should not be concluded that removing the hydroxy protecting groups prior to the RCM step will always solve stereoselectivity and reactivity problems; sometimes the opposite is the case. Nonenolide ((E)-48), a fungal metabolite with notable antimalarial activity, was first synthesized using an RCM approach. In the course of this synthesis it was discovered that protection of both secondary alcohols in the RCM precursor is mandatory to obtain the (E)-configuration required for the target structure [51, 52]. Thus, the bis- PMB protected diene 46a was converted to 47 with an (E):(Z) ratio of 9:1, whereas the