C-Furanosides: Synthesis and Stereochemistry

By Peter Goekjian, Arnaud Haudrechy, Boudjema Menhour and Claire Coiffier

Carbon analogs of carbohydrates, dubbed C-glycosides, have remained an important and interesting class of mimetics, be it in natural product synthesis, for pharmacological applications, as conformational probes, or for biological studies. C-Furanosides: Synthesis and Stereochemistry provides a much-needed overview of synthetic and stereochemical principles for C-furanosides: analogs of a 5-membered ring carbohydrate glycoside (furanoside), in which the anomeric oxygen has been replaced with a carbon.

While our understanding of conformational behavior and of stereoselective synthesis in 6-membered ring compounds is quite good, our ability to predict the conformation of 5-membered ring compounds, or to predict the stereochemical outcome of a given reaction, remains anecdotal. Through a comprehensive review of literature approaches to the different C-furanoside stereoisomers, as well as an interpretation of the outcome in terms of a reasonable number of stereochemical models, C-Furanosides: Synthesis and Stereochemistry enables the reader to determine the best approach to a particular C-glycoside compound, and also hopes to provide a certain level of rationalization and predictability for the synthesis of new systems.

• Provides a comprehensive review of the growing literature in C-furanosides
• Enables readers to choose the most convenient approach to access a defined target in natural products synthesis or pharmacology and make reasonable predictions for the stereochemical outcome in unpublished cases
• Explores the various rational models for stereochemical analysis of furanoside reactivity, with a clear distinction made between physical chemical mechanisms and stereochemical models

• Language: English
• Publisher: Academic Press
• Release date: Nov 30, 2017
• ISBN: 9780128037898

Peter Goekjian

Dr. Peter Goekjian has been Professor of Chemistry at Université Claude Bernard–Lyon 1, France, since September 2000. His research interests are carbohydrate chemistry, total synthesis of glycosidic natural products, targeted methodology, and the role of glycosylation in signal transduction and gene expression. He was involved in the early work on the use of C-glycosides as conformational probes in the late 1980s and early 1990s.

Book preview

C-Furanosides

Synthesis and Stereochemistry

Peter Goekjian

Arnaud Haudrechy

Boudjema Menhour

Claire Coiffier

Table of Contents

Cover image

Title page

Copyright

The Stereochemistry of C-Furanosides

Part A. C-Glycosides of Lyxose and Ribose: galacto-, altro- and allo- Configurations

Chapter A. Introduction

Chapter A.1. galacto-C-Furanosides (I, β-C-Lyxose)

Disconnections

Natural Occurrence

A.1.1 Disconnection A

A.1.2 Disconnection B

A.1.3 Disconnection C

A.1.4 Disconnection D

A.1.5 Miscellaneous

A.1.6 Conclusion

Chapter A.2. d- and l-altro-C-furanosides (II/ent-II, α-C-Lyxose, α-C-Ribose)

Disconnections

Natural Occurrence

A.2.1 Disconnection A

A.2.2 Disconnection B

A.2.3 Disconnection C

A.2.4 Disconnection D

A.2.5 Disconnection E

A.2.6 Miscellaneous

A.2.7 Conclusion

Chapter A.3. allo-C-Furanosides (VI, β-C-Ribose)

Disconnections

Natural Occurrence

A.3.1 Disconnection A

A.3.2 Disconnection B

A.3.3 Disconnection C

A.3.4 Disconnection D

A.3.5 Disconnection E

A.3.6 Miscellaneous

A.3.7 Conclusion

Chapter A.4. Lyxose and Ribose C-Glycosides: Other Results and Further Insight Into Stereochemistry

A.4.1 Disconnection A

A.4.2 Disconnection B

A.4.3 Disconnection C

A.4.4 Disconnection D

A.4.5 Miscellaneous

A.4.6 Conclusion

Part B. C-Glycosides of Arabinose and Xylose: gluco-, ido- and manno- Configurations

Chapter B. Introduction

Chapter B.1. Gluco-C-Furanosides (III/ent-III, β-C-Arabinose, β-C-Xylose)

Disconnections

Natural Occurrence

B.1.1 Disconnection A

B.1.2 Disconnection B

B.1.3 Disconnection C

B.1.4 Disconnection D

B.1.5 Miscellaneous

B.1.6 Conclusion

Chapter B.2. ido-C-Furanosides (V/ent-V, α-C-Xylose)

Disconnections

Natural Occurrences

B.2.1 Disconnection A

B.2.2 Disconnection B

B.2.3 Disconnection C

B.2.4 Disconnection D

B.2.5 Miscellaneous

B.2.6 Conclusion

Chapter B.3. manno-C-Furanosides (VII/ent-VII, α-C-Arabinose)

Disconnections

Natural Occurrence

B.3.1 Disconnection A

B.3.2 Disconnection B

B.3.3 Disconnection C

B.3.4 Disconnection D

B.3.5 Conclusion

Chapter B.4. Arabinose and Xylose C-Furanosides: Other Results and Further Insight Into Stereochemistry

Disconnections

B.4.1 Disconnection A

B.4.2 Disconnection B

B.4.3 Disconnection C

B.4.4 Equilibration Processes

B.4.5 Miscellaneous

B.4.6 Conclusion

Part C. Applications of C-Furanosides

Introduction

C.1 Preparation of C-Furanosides

C.2 Agro-Industry and Farming

C.3 Ligands for Asymmetric Catalysis

C.4 Perfumes, Aromas, and Cosmetics

C.5 Imaging and Diagnostics

C.6 Pharmaceutical Industry

C.7 Polymers

Part D. ¹H NMR Vicinal Coupling Constants of C-Furanosides

Introduction

D.1 ¹H NMR Data in galacto-C-Furanosides (I, β-C-Lyxose), Corresponding to Chapter A.1

D.2 ¹H NMR Data in D- and L-altro-C-Furanosides (II/ent-II, α-C-Lyxose, α-C-Ribose), Corresponding to Chapter A.2

D.3 Comparison Between ¹H NMR Data in galacto-C-Furanosides (I, β-C-Lyxose), Corresponding to Chapter A.1, and in D- and L-altro-C-Furanosides (II/ent-II, α-C-Lyxose, α-C-Ribose), Corresponding to Chapter A.2

D.4 ¹H NMR Data in allo-C-Furanosides (VI, β-C-Ribose), Corresponding to Chapter A.3

D.5 Comparison Between ¹H NMR Data in D- and L-altro-C-Furanosides (II/ent-II, α-C-Lyxose, α-C-Ribose), Corresponding to Chapter A.2, and in allo-C-Furanosides (VI, β-C-Ribose), Corresponding to Chapter A.3

D.6 ¹H NMR Data in D- and L-gluco-C-Furanosides (III/ent-III, β-C-Arabinose, β-C-Xylose), Corresponding to Chapter B.1

D.7 ¹H NMR Data in D- and L-ido-C-Furanosides (V/ent-V, α-C-D-Xylose, α-C-L-Xylose), Corresponding to Chapter B.2

D.8 Comparison Between ¹H NMR Data in D- and L-gluco-C-Furanosides (III/ent-III, β-C-Arabinose, β-C-Xylose), Corresponding to Chapter B.1, and in D- and L-ido-C-Furanosides (V/ent-V, α-C-D-Xylose, α-C-L-Xylose), Corresponding to Chapter B.2

D.9 ¹H NMR Data in D- and L-manno-C-Furanosides (VII/ent-VII, α-C-Arabinose), Corresponding to Chapter B.3

D.10 Comparison Between ¹H NMR Data in galacto-C-Furanosides (I, β-C-Lyxose), Corresponding to Chapter A.1, and in D- and L-gluco-C-Furanosides (III/ent-III, β-C-Arabinose, β-C-Xylose), Corresponding to Chapter B.1

D.11 Comparison Between ¹H NMR Data in D- and L-altro-C-Furanosides (II/ent-II, α-C-Lyxose, α-C-Ribose), Corresponding to Chapter A.2, and in D- and L-ido-C-Furanosides (V/ent-V, α-C-D-Xylose, α-C-L-Xylose), Corresponding to Chapter B.2

D.12 Conclusion and Summary of Expected Coupling Constants

Index

Copyright

Academic Press is an imprint of Elsevier

125 London Wall, London EC2Y 5AS, United Kingdom

525 B Street, Suite 1800, San Diego, CA 92101-4495, United States

50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom

Copyright © 2018 Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

ISBN: 978-0-12-803739-3

For information on all Academic Press publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: John Fedor

Acquisition Editor: Emily M. McCloskey

Editorial Project Manager: Amy M. Clark

Production Project Manager: Bharatwaj Varatharajan

Cover Designer: Victoria Pearson

Typeset by TNQ Books and Journals

The Stereochemistry of C-Furanosides

The stereochemical analysis of C-furanosides is discussed below, covering the various diastereomers of these compounds and their nomenclature, their symmetries and pseudosymmetries, and finally the stereochemical models for asymmetric induction commonly used to rationalize the stereochemical outcome of the syntheses.

1. C-Furanosides: Stereochemistry and Nomenclature

The various syntheses of C-furanosides will be detailed in subsequent chapters, so we would like to present here a quick overview of their stereochemistry and the nomenclature that we will use throughout this book. We chose to simplify this initial discussion to establish the stereochemical relationships between these cyclic sugar analogs and the 2,5-anhydrohexose series, as this will form the basis for the organization of the book. However, C-furanosides usually belong to more complex structures, and symmetry considerations (a plane of symmetry in this case) can no longer be considered, only a pseudosymmetry. The stereochemical nomenclature in individual cases may therefore differ.

Beginning with the case I (Fig. 1), where all the groups appear on the same face, the compound can be seen from one face as a D-sugar (I-a), or from the other face by rotation, as an L-sugar (I-b). At the same time, a mirror plane reflection of I-a gives its expected enantiomer ent-I. There is a clear identity between the templates ent-I, I-a, and I-b, due to the Cs symmetry of the meso- sugar. The stereochemistry of compound I in anhydrohexose nomenclature is thus galacto-.

Figure 1  Analysis of D - galacto - series.

To generate the diastereomers of I, the configuration at C-5 (anhydrohexose numbering) will be held fixed. A single inversion of the configuration at the C-2 carbon will generate stereoisomer II. Similar rotation of II-a gives II-b, while reflection will provide ent-II. The compound is chiral. The C-2 epimer of galacto- as drawn in II-a would be D-talo-, yet II-b would be regarded as D-altro- (Fig. 2). The carbohydrate nomenclature convention is that alphabetical order takes precedence, so that compound II is D-altro-, and its enantiomer ent-II is L-altro- (2).

Figure 2  Analysis of D - altro - series.

Inversion of the configuration at the C-3 carbon (anhydrohexose numbering) generates stereoisomer III. Similar rotation of III-a gives III-b, while reflection will provide ent-III. The compound is also chiral. The C-3 epimer of galacto- as drawn in III-a would be D-gulo- or L-gluco- as in III-b (Fig. 3). Again alphabetical order takes precedence, so that compound III is L-gluco-, and its enantiomer ent-III is D-gluco-.

Figure 3  Analysis of L - gluco - series.

Inversion of the configuration at C-4 similarly, generates IV-a, IV-b, and ent-IV. One clearly notes the identity between IV-b and ent-III and between ent-IV and III-b. IV and ent-IV are therefore D-gluco- and L-gluco-, respectively (Fig. 4).

Figure 4  Analysis of D - gluco - series.

Inversion of the configuration at C-2 and C-3 generates V-a, V-b, and by mirror plane reflection ent-V. The molecule is chiral and has C2 symmetry. Both V-a and V-b are read as D-ido-, and ent-V is L-ido- (Fig. 5).

Figure 5  Analysis of D - ido - series.

Inversion of the configuration at C-3 and C-4 generates VI-a, VI-b, and by mirror plane reflection ent-VI. The molecule is achiral, as VI-b is clearly identical to ent-VI. VI is therefore meso- and has the stereochemistry allo- (Fig. 6).

Figure 6  Analysis of D - allo - series.

Inversion of the configuration at C-2 and C-4 generates VII-a, VII-b, and by mirror plane reflection ent-VII. The molecule is chiral and again has C2 symmetry. VII has the stereochemistry D-manno-, and ent-VII is L-manno- (Fig. 7).

Figure 7  Analysis of D - manno - series.

Finally, inversion of the configuration at C-2, C-3, and C-4 generates VIII-a, VIII-b, and by mirror plane reflection ent-VIII. One clearly notes the identity between VIII-b and II-a (and of course VIII-a and II-b). VIII and ent-VIII are therefore D-altro- and L-altro-, respectively (Fig. 8).

Figure 8  Analysis of D - altro - series (2) .

As there are two stereochemistries that are redundant, there are no talo- or gulo- configurations in the simple systems and there are therefore six diastereomeric series. Of these, two are achiral (allo- and galacto-), although individual examples will of course be chiral. As the chemistry in one enantiomeric series is necessarily applicable to the other enantiomeric series (with the significant difference of the access to the starting material), both enantiomers will be treated together.

2. Relationships Between Hexose-Stereochemical Nomenclature and the C-Glycosides of Furanose-Type Frameworks

The structures under discussion can be considered as C-glycosides of furanoside sugars. Thus, for example, galacto- structure I-a can be regarded as the anomeric β-C-glycoside of D-lyxose, and as such can be tied to structure D-altro-II-a, the corresponding anomeric α-C-glycoside (Fig. 9), the sphere corresponding to the substituent added to the anomeric position. Stereoselective addition to D-lyxose can therefore lead to either galacto- or D-altro- compounds I-a and II-a.

Figure 9  Analysis of α- and β- D - lyxo - C -glycosides.

Structure I-a can also be accessed by adding the anomeric carbon substituent to the other side. The galacto- configuration I-a can therefore also be considered as a β-C-glycoside of L-lyxose and therefore be tied to the α-C-glycosides ent-VIII (=ent-II) with the stereochemistry L-altro- (Fig. 10).

Figure 10  Analysis of α- and β- L - lyxo - C -glycosides.

The D-altro- configuration II-b can also be considered as an α-C-glycoside of D-ribose and therefore be tied to the β-C-glycosides VI-a with the stereochemistry D-allo- (Fig. 11). In the enantiomeric series, the L-altro- and L-allo- structures can thus be considered as α- and β-C-glycosides of L-ribose.

Figure 11  Analysis of α- and β- D - ribo - C -glycosides.

There is therefore an underlying stereochemical relationship between the C-glycosides of lyxose and ribose, leading to C-glycosides of the galacto-, altro-, and allo- configurations. These will be described in Part A. Unfortunately, in addition to these less-than-intuitive stereochemical relationships, this dual perspective on the structure of the compounds described in this review leads to the use of two incompatible numbering systems, namely the anhydrohexose numbering (which corresponds coincidentally to the tetrahydrofuran numbering, i.e., the anomeric position is position 2, although position 1 is the exocyclic carbon in the former case and the ring oxygen in the latter case) and the pentose C-glycoside numbering, in which the anomeric position is numbered as 1. We will generally favor the anhydrohexose numbering system, although in some cases the pentose numbering will be difficult to circumvent. Of course, there is also the numbering of the target structure, so our numbering will generally differ from that of the original authors.

In much the same way, the L-gluco- structure III-a can be regarded as the β-C-glycoside of D-xylose and is therefore tied to the corresponding α-C-glycoside, namely the D-ido- structure V-a (Fig. 12). In the enantiomeric series, the D-gluco- structure is therefore tied to L-ido- as α- and β-C-glycosides of L-xylose.

Figure 12  Analysis of α- and β- D - xylo - C -glycosides.

If the L-gluco- structure III-a is drawn with the anomeric position on the other side, it is now tied to ent-VII (L-manno-) as α- and β-C-glycosides of L-arabinose, respectively (Fig. 13). In the enantiomeric series, the D-gluco-structure is therefore tied to D-manno- as α- and β-C-glycosides of D-arabinose.

Figure 13  Analysis of α- and β- L - arabino - C -glycosides.

The C-glycosides of xylose and arabinose leading to C-glycosides of the gluco-, ido-, and manno- configurations will therefore be described in Part B.

The different stereochemical relationships thus obtained between the configurations, symmetries, and C-glycosides are summarized in Table 1.

Table 1

Stereochemical Relationships of C-Furanosides

In cases where the two anhydrohexose names differ, the first one is given priority according to the IUPAC rules. In cases with Cs symmetry, the absolute stereochemistry (D- or L-) is given by the relative priority of the substituents.

We will discuss in the following chapters how the different research groups have accessed these C-furanoside configurations. Insofar as the reader may be interested in a particular target configuration, the book will provide six subchapters on the stereocontrolled access to each of the six diastereomeric structures. Enantiomeric series will of course be treated together, as chemistry developed in one series is necessarily relevant to the enantiomer. However, there are a number of stereochemically unselective C-glycosylation approaches, or syntheses where the stereochemistry is unspecified, which may be of interest to the reader, be it as an efficient means of obtaining a sample of a compound (a short unselective synthesis may be preferable to a more involved selective one if the reader simply wants a sample), as an appreciation and an overview of the stereochemical and chemical aspects of the domain, or as a means of finding the closest structural precedent for the target structure. Reflecting this dual organization, the review will be organized in two parts, the C-glycosides of lyxose and ribose, leading to the galacto-, altro-, and allo- configurations, and the C-glycosides of arabinose and xylose, leading to the manno-, gluco-, and ido- configurations. Each part is subdivided into three chapters covering reactions providing access to each of the configurations, and a chapter detailing results where the stereochemistry is nonselective or unspecified, which will be accompanied by a more in-depth discussion of selectivity. We have arbitrarily chosen 3:1 as a criterion to define a selective versus nonselective approach, keeping clearly in mind that a short synthesis with a 2.5:1 selectivity may nonetheless be of significant interest. The reader is therefore strongly encouraged to read Part A or B as a unit, as the answer they are looking for may lie in other chapters.

An additional caveat should be provided to the reader. Stereochemical assignments in the furanose series are far less straightforward than in the pyranose series, and a number of structures were originally misassigned. Some of these were later corrected by the same authors or by others. In these cases, we tried to point out the reassignment in the text. In other cases, there are internal inconsistencies within the paper (for example, between the text and the schemes, or between the mechanisms provided and the structures). In other cases, analysis of the data in light of later results might lead one to question the initial assignment. However, while we may point these out, it would be inappropriate for us to modify the conclusions of the initial authors, and these reports are thus presented as is.

3. Stereochemical Models for the Control of the Configuration at C-1

The use of stereochemical transition state models can be of a great utility in helping the reader make sense of the large amount of empirical data. We will therefore make liberal use of such models in describing the experimental results described in this book. However, it is critically important that the reader keep in mind that we are providing models, which are intended to be simple and predictive and not physical organic mechanisms. The number of examples for which physical organic studies have been performed is vanishingly small. The model that we may chose to rationalize the observed selectivity may differ from the rationale or model provided by the authors. There is nothing contradictory in this fact. The models provided in this book are the current authors' interpretations in light of the body of experimental results within the chapter, in such a way as to provide a framework for the reader. Under no circumstances should they be considered as experimental facts or as the conclusions of the cited authors (3).

3.1. Stereospecific Approach

The conceptually simplest approach to the control of C-furanoside configuration is the use of a stereospecific reaction. A stereospecific reaction is defined by Eliel as a transformation of an existing stereochemistry, by such a reaction that to obtain the other stereochemical result; it is necessary to change the mechanism (4). There is no additional stereochemical information in the product than in the starting material. The prototypical example is an SN2 reaction (Fig. 14). Thus the L-gluco- compound III-a could be obtained from a D-ido- precursor by SN2 inversion at C-2, or from L-manno- by inversion at C-5. It can also be obtained from D- or L-galacto- precursor by double inversion at C-2 or C-5. In reality, many other stereospecific approaches should be envisioned, and the relationships are far more complex. A detailed discussion is provided within each chapter.

Figure 14  Stereospecific approaches by S N 2 reactions.

Another useful stereospecific reaction is the addition to alkenes, which is both stereospecific and stereoselective. In the example presented in Fig. 15, the relative stereochemistry of the C-3 and C-4 centers is a transformation of the Z-stereochemistry of the alkene. However, the relative stereochemistry between C-2 and C-3 must be controlled by stereoselective induction. Nonetheless, this permits the stereochemistry at the C-4 center to be tied to the stereocontrol at the C-3 center (anhydrohexose numbering).

Figure 15  Stereospecific dihydroxylation approaches.

3.2. Stereoselective Reactions at the Anomeric Position Under Kinetic Control

Alternatively, the stereochemistry at the anomeric position can be controlled by the use of stereoselective additions, for example, to an sp²-hybridized carbon (oxonium ion or radical). Eliel defines a stereoselective reaction as one in which new stereochemical information is created in the product, and both stereochemical results are possible via the same mechanism, typically to different faces of a prochiral center, at different ends of a meso- precursor, or to one of two rapidly interconverting species. We can differentiate between internal stereocontrol, in which the stereoselectivity is controlled by centers within the target molecule, in which case we are constrained in our choice of this stereochemistry, and external stereocontrol, in which the stereoselectivity is controlled by centers outside the target molecule, such as a chiral auxiliary, chiral catalyst, reagent, or medium, in which case we are free to chose either stereochemistry. We will focus in this Section 3.2 on the different kinetic models for internal stereoinduction.

If the reaction is irreversible, the selectivity will be under kinetic control, and will be determined by the relative energy of the transition states of the rate-determining step. While the different transition state models for cyclic stereoinduction in the six-membered ring are well established, such as the axial addition to oxocarbenium ions, diaxial opening of epoxides, axial versus equatorial addition to exocyclic double bonds, the situation in five-membered rings is considerably less clearcut. Much progress has been made in recent years, however, and some of the models are presented below.

3.2.1. Bicyclic Induction

As we will see in the subsequent chapters, the most widely exploited stereochemical model involves the effect of bicyclic stereocontrol. If the five-membered ring is cis- fused to another ring, this will create a fairly rigid bicyclic framework having a concave and a convex face. Attack of the nucleophile will typically occur on the convex face. This applies, for example, to ribose and lyxose 2,3-O-acetonides and to cyclized forms of the C-3/C-5 positions in the lyxose and xylose series (Fig. 16).

Figure 16  Bicyclic stereoinduction.

3.2.2. 1,3-Induction

A general model of cyclic induction in addition to nonbicyclic five-membered ring oxonium ions has recently been proposed by Woerpel et al. conformation in the intermediate in which the oxygen occupies the pseudoaxial position. Thus, by the same argument as above, inside attack cis- to the oxygen via a ²T3 conformation is favored over attack trans- to the oxygen by a ⁵E transition state (Fig. 17). This model explains the weak effect of the 4-hydroxymethyl substituent. Furthermore, the authors showed that this model can be applied to the six-membered ring series: in the absence of a hydroxymethyl substituent, the C-4-alkoxy group directs the incoming nucleophile to the trans- position, where thus both the C-4-alkoxy and the nucleophile are axial in the transition state (5d).

Figure 17  Woerpel model.

3.2.3. 1,2-Induction

The kinetic induction of the position 2 seems to differ significantly between the oxygen and carbon nucleophiles. Indeed, anchiomeric assistance by acyl groups in the position 2 is perhaps the most reliable induction mode in O-glycosides, yet does not play a significant role when allyltrimethylsilane is used as a nucleophile in six-membered ring series (6). One could rationalize this by a later transition state for C-glycosylation. In five-membered rings, we will see that the results are somewhat mixed, as anchiomeric assistance in C-2 seems to be observed in some cases and not in others. An alternative variant of the anchiomeric participation is internal delivery of the reagent, either through covalent or noncovalent association. In this case, a 1,2-cis- relationship will be obtained (Fig. 18).

Figure 18  1,2-Induction models.

In the case of nonparticipating groups in C-2, there is again a difference between oxygen and carbon nucleophiles. Boons et al. proposed that in the arabinose series, the O-2 substituent prefers a pseudoaxial, Felkin–Ahn-like orientation in the transition state of the O-glycosylation (7), whereas Woerpel et al. showed that in the carbon series, the directing effect is in the opposite direction, in a 85:15 ratio in favor of a 1,2-cis- selectivity, thus increasing the α-selectivity in the ribose series and reducing the corresponding selectivity in the arabinose series (5c). The benzyloxy group prefers an equatorial arrangement in the ²T3 twist transition state emanating from inside attack on the ³E intermediate oxonium ion (Fig. 17). Again, this is consistent with a later transition state, although Boons' result is also consistent with an enhanced Woerpel selectivity from the 3-alkoxy group.

By definition, there is always a transition state leading to one stereochemical result, and a transition state leading to the opposite result. It may be that several models predict a given result, and it is not necessarily obvious, which is true, or when one model versus the other should dominate. We will thus talk of the bicyclic induction product versus the Woerpel product, or the product of stereospecific substitution versus the product of 1,2-induction. It is our hope that as we provide an overview of the literature results, some trends will emerge on which will be predictive.

3.3. Stereoselective Reactions Under Thermodynamic Control

If the reaction is reversible, or an equilibrium is established between the different stereoisomers by another mechanism, then the most stable stereochemical product will be observed. Unfortunately, conformational analysis in the five-membered ring series is more complex due to the lack of symmetry of the different conformers, and it is not immediately obvious which stereoisomer is more stable (8). A discussion of the thermodynamic properties of each configuration is provided within the individual chapters.

References

1. For previous reviews of C-glycosides, please see for example:(a) Levy D.E, Tang C. The Chemistry of C-Glycosides. In: Tetrahedron Organic Chemistry Series. Vol. 13. Oxford: Elsevier; 1995(b) Postema M.H.D. C-Glycoside Synthesis. Boca Raton: CRC Press Inc.; 1995

2. Rule 2-Carb-2-1-3, IUPAC Nomenclature of Carbohydrates. Pure Appl. Chem. 1996;68:1919–2008.

3. It is useful in this context to differentiate clearly between three different things that we may call mechanisms. The distinction lies above all in the purpose for which it is intended by the authors, and therefore on what standards it should be judged. A mechanistic rationale is a plausible explanation for a single observed result. It should be judged on whether it is reasonable, and needs to be neither predictive nor true, but simply be based on reasonable precedent. A mechanistic model is a mnemonic mechanistic proposal intended to predict the stereochemical outcome of a given class of reactions. It must be simple and predictive but need not be true. Its value, and therefore that on which it must be judged, is on its ability to predict the outcome. This is what is provided here. It is more esthetically pleasing if they are true, but this not really germane to the discussion. For example, Lewis structures are clearly not true but are nonetheless one of the most useful and essential predictive systems in all of chemistry. The third possibility is a physical organic mechanism, which must be true, and conclusive evidence must be provided for each and every aspect of the proposal. It need be neither simple (it rarely is) nor predictive. A physical organic mechanism is only rigorously true in the specific case and conditions under investigation, and as soon as it is used in another system (prediction), then it is by definition no longer true. There is an essential interplay between physical organic mechanisms (understanding), mechanistic rationales (analysis), and mechanistic models (prediction and manipulation). It is important to clearly define which is being presented, so that it will be interpreted appropriately by the reader. It is difficult to attribute correctly such an established distinction, other than to say that it is not ours. Although our own source is Edwin Vedejs, it most likely originated in the early discussions of Hammett, Hammond, Ingold, Breslow, Wiberg, or Isaacs.

4. Eliel E.L, Wilen S.H, Mander L.N. Stereochemistry of Organic Compounds. New York: John Wiley & Sons; 1994.

5. (a) Larsen C.H, Ridgway B.H, Shaw J.T, Woerpel K.A. A Stereoelectronic Model to Explain the Highly Stereoselective Reactions of Nucleophiles with Five-Membered-Ring Oxocarbenium Ions. J. Am. Chem. Soc. 1999;121:12208–12209(b) Smith D.M, Woerpel K.A. Using Stereoelectronic Effects to Explain Selective Reactions of 4-Substituted Five-Membered Ring Oxocarbenium Ions. Org. Lett. 2004;6:2063–2066(c) Larsen C.H, Ridgway B.H, Shaw J.T, Smith D.M, Woerpel K.A. Stereoselective C-Glycosylation Reactions of Ribose Derivatives: Electronic Effects of Five-Membered Ring Oxocarbenium Ions. J. Am. Chem. Soc. 2005;127:10879–10884(d) Smith D.M, Woerpel K.A. Electrostatic Interactions in Cations and Their Importance in Biology and Chemistry. Org. Biomol. Chem. 2006;4:1195–1201(e) Lucero C.G, Woerpel K.A. Stereoselective C-Glycosylation Reactions of Pyranoses: The Conformational Preference and Reactions of the Mannosyl cation. J. Org. Chem. 2006;71:2641–2647(f) Yang M.T, Woerpel K.A. The Effect of Electrostatic Interactions on Conformational Equilibria of Multiply Substituted Tetrahydropyran Oxocarbenium Ions. J. Org. Chem. 2009;74:545–553.

6. See for example. Uchiyama T, Vassilev V.P, Kajimoto T, Wong W, Lin C.-C, Huang H, Wong C.-H. Design and Synthesis of Sialyl Lewis X Mimetics. J. Am. Chem. Soc. 1995;117:5395–5396.

7. Zhu X, Kawatkar S, Rao Y, Boons G.-J. Practical Approach for the Stereoselective Introduction of α-Arabinofuranosides. J. Am. Chem. Soc. 2006;128:11948–11957.

8. (a) Coiffier C, Barberot C, Nuzillard J.-M, Goekjian P.G, Henon E, Haudrechy A. Ring Dihedral Principal Component Analysis of Furanose Conformation. Carbohydr. Chem. 2014;40:378–400(b) Ohrui H, Jones G.H, Moffatt J.G, Maddox M.L, Christensen A.T, Byram S.K.C-Glycosyl Nucleosides. V. Some Unexpected Observations on the Relative Stabilities of Compounds Containing Fused Five-Membered Rings with Epimerizable Substituents. J. Am. Chem. Soc. 1975;97:4602–4613.

Part A

C-Glycosides of Lyxose and Ribose: galacto-, altro- and allo- Configurations

Outline

Chapter A. Introduction

Chapter A.1. galacto-C-Furanosides (I, β-C-Lyxose)

Chapter A.2. d- and l-altro-C-furanosides (II/ent-II, α-C-Lyxose, α-C-Ribose)

Chapter A.3. allo-C-Furanosides (VI, β-C-Ribose)

Chapter A.4. Lyxose and Ribose C-Glycosides: Other Results and Further Insight Into Stereochemistry

Chapter A

Introduction

Abstract

This chapter introduces the stereochemical relationships between the C-furanosides of ribose and lyxose, as well as some general points necessary to understand Chapters A1–A4.

Keywords

allo- configuration; altro- configuration; galacto- configuration; Lyxose; Ribose; talo- configuration

Chapter Outline

References

As discussed in the general introduction, the β-C-glycosides of both D- and L-lyxose correspond to the galacto- configuration I-a, each one providing access to a different side chain on the tetrahydrofuran (Fig. A.1). In the illustrations below, the sphere corresponds to the structure added as the glycoside, which in many synthetic applications can correspond to either side chain of the target structure. Individual compounds will be either D-galacto- or L-galacto- depending on the nature of the substituents at each position.

Figure A.1  Access to galacto - C -furanosides.

Access to galacto-structures by a C-glycosylation-type process on a lyxose precursor can also give the α-C-glycoside corresponding to an altro-C-furanoside (II-a and ent-II). The C-glycosides of α-D-lyxose and α-D-ribose both correspond to the D-altro- configuration II-a and the C-glycosides of α-L-lyxose and α-L-ribose to the L-altro- configuration ent-II (=ent-VIII, Fig. A.2).

Figure A.2  Access to altro - C -furanosides.

A C-glycosylation on a ribose precursor can also give the β-isomer VI-a (Fig. A.3). Both β-D-ribose and β-L-ribose C-glycosides correspond to the allo- configuration VI-a, again each one providing access to a different side chain on the tetrahydrofuran.

Figure A.3  Access to allo - C -furanosides.

Because of these underlying relationships, we have chosen to group all structures of the galacto-, altro-, and allo- configurations, belonging to lyxose and ribose C-glycoside series in Part A. Stereoselective C-glycosylation reactions can give rise to mixtures in varying ratios, and in some cases the stereochemistry of the product is not defined in the original publication. We will therefore divide the chemistry into four chapters: the first three chapters will be devoted to stereoselective approaches to the galacto-, altro-, and allo- configurations, respectively; in Chapter A.4, we will cover poorly selective cases (arbitrarily defined as less than 3:1) and those reports in which stereoselectivity is undefined. The fourth chapter will provide a useful overview of the chemistry in this area and an in-depth discussion of the factors influencing stereochemical outcomes. The C-glycosides of arabinose and xylose, leading to the manno-, gluco-, and ido- series, will be covered in Part B.

A few clarifications of the data presented in the subsequent chapters may be warranted:

• In many cases, ratios of products are explicitly reported and are therefore provided in this chapter; to make comparisons easier, all selectivities were converted to a ratio relative to 1 versus other expressions that may appear in the publications (diastereomeric excesses, percentages, etc.). Any errors in the conversions are our own. We have made no attempt to evaluate the precision of the reported ratios, as an estimate is sufficient for the purposes of the current work. If more precise ratios are needed (e.g., for physical chemical studies, etc.), the readers are encouraged to consult the original publications.

• In other cases, the authors specifically sought to determine a ratio but only observed one product to the limits of detection of the method (usually by NMR). In such cases, any differences in ratio should not be overinterpreted, as they typically reflect the author's perception of the precision of their method. If a limiting ratio was reported (e.g., >20:1 or >9:1), it is retranscribed here.

• Some papers stated that they only observed one isomer, without providing a limiting ratio, in which case we noted only α observed. These results leave considerable ambiguity, as it depends on how the observation was made, or may simply mean that the authors did not identify the other isomer. We can assume that the authors showed due diligence, even if we do not have a ratio; but it is important for the interpretation of the data to know that >9:1 should be considered as a higher standard of selectivity than only α observed.

• In yet other cases, the authors simply report a yield of the desired isomer, in which event we state only α reported. It is important to note that we have no information on the actual ratio of stereoisomers. If the authors report only one product in 60% yield, the ratio could, in fact, be anywhere from 1.5:1 to >99:1.

• Finally, if a ratio is provided for a stereospecific reaction, this should be regarded as the ratio of the desired reaction to a particular side reaction; this is different from a selectivity, in which both products are of the same reaction.

A note on the nomenclature system used here: The precursors used in the syntheses of C-furanosides can be derived from higher sugars; for example, an α-C-glycosylation of a D-mannofuranose may lead to a D-glycero-D-talo-C-furanoside using the formal IUPAC-approved nomenclature system. However, this nomenclature is rather unwieldy, and in order to focus on the ring stereochemistry, we will refer to the D-mannofuranoside as a D-lyxo-precursor and to the product as a D-altro-C-furanoside. Similarly, α and β, as used here, will correspond to the altro-C-furanoside: α is 2,5-trans and β is 2,5-cis. Again, in the example above, α and β should formally be relative to the C-5 center of mannose, not to the ring. The nomenclature used in this book may, therefore, differ from the correct IUPAC nomenclature used in the publication.

Another important note concerns the interpretation of the observed selectivity. If the explanation appears before the reference, then it is that of the original authors. If the model is provided after the reference, then it is our own interpretation in an effort to provide the reader with a few guiding predictive models to make sense of the data (1). It is important, however, that the reader not attribute these interpretations to the original authors.

Finally, we are aware that some readers may only read individual chapters, and have therefore attempted to make them relatively self-sufficient. There is, therefore a certain amount of repetition that will undoubtedly strike those who read the entire book. We apologize to these more engaged readers.

The altro-, galacto-, and allo-C-furanosides, can be considered as C-glycosides of ribose and lyxose; of course, the stereochemical analysis is more complex. In fact, a full stereochemical analysis of these structures, using Quiral (2), shows the richness of the stereochemical relationships between the sugars. A detailed analysis of the D-altro-stereoisomer II-a using Quiral is shown in Figs. A.4–A.8; NA shows the nonanomeric moiety of the carbohydrate and A shows the anomeric position. Similar analyses will be provided for the remaining stereoisomers in the corresponding chapter.

The Quiral program is designed to identify underlying stereochemical relationships between a target structure and potential carbohydrate precursors. It shows all stereochemical possibilities; although cheap sugars are typically more attractive as starting materials, there are a variety of contexts where other possibilities are relevant.

Figure A.4  Identification of the stereotetrads D -altrose and D -talose.

Figure A.5  Stereotriads D -lyxose and D -ribose corresponding to C -glycosylation at the anomeric position.

Figure A.6  Stereotriads D -arabinose and L -ribose corresponding to C -glycosylation at the nonanomeric position.

Figure A.7  Stereochemical transformations leading to L -ribose and D -lyxose.

Figure A.8  Stereochemical transformations leading to D -altrose and D -talose.

The Quiral program applied to the D-altro-C-furanoside II-a mapped out several carbohydrate skeletons, including the stereotetrads of D-altrose and D-talose (Fig. A.4) and the stereotriads of D-lyxose and D-ribose (Fig. A.5), as expected based on the stereochemical analysis in the general introduction. In addition to the α-glycosides at the anomeric position of D-lyxose and D-ribose, the D-altro-stereoisomer is also accessible as the α-C-glycoside of L-ribose or of D-arabinose performed on the oxidized nonanomeric (C-5) position (Fig. A.6). Such head-to-tail interconversions are particularly relevant to target-oriented synthesis, as in most cases both the anomeric and nonanomeric positions must be modified, and it is therefore not more difficult to first functionalize the anomeric position before oxidizing the C-5 position to allow for C-glycosylation.

While D-ribose and D-arabinose are viable starting materials, L-ribose and D-lyxose are less likely precursors for the D-altro configuration. A simple click on the synthesis button gives a view of all the single transformations (level 1) leading to each of these sugars (Fig. A.7). In addition to the head-to-tail (exch) transformation of D-ribose and D-arabinose, one notes that L-ribose corresponds to inversion at the C-2 position of L-arabinose or to double inversion at C-3 and C-4 positions of D-arabinose; D-lyxose corresponds to inversion at the C-2 position of D-xylose, inversion at the C-3 position of D-arabinose, double inversion at C-2 and C-3 positions from D-ribose, to the C-1/C-5 fragment of D-mannose, and to the C-2/C-6 fragment of D-galactose. D-mannofuranosides are the most commonly used glycoside donors in the D-lyxose series.

In addition to the C-glycosylation process starting from a pentofuranose sugar, an alternative approach is from a hexose precursor by reaction between the C-2 and C-5 hydroxyl groups. D-altrose and D-talose themselves are not likely precursors, as this would require substitution of the C-2 or C-5 position with retention of configuration; yet a stereoselective approach by selective reduction of the corresponding ketose sugars, D-tagatose and D-psicose, can be envisioned. Once again, a simple click on the synthesis button provides a view of the single stereochemical transformations leading to D-altrose and D-talose (Fig. A.8). In the case of D-altrose, in addition to the D-arabinose and D-ribose precursors listed above (addRed and addNonRed), and the relationship to D-talose (exch) noted in the general introduction, the D-altro-compound II-a could be obtained by SN2 inversion from the D-allo- configuration at the C-2 position, from the D-manno- configuration at the C-3 position, from the D-ido- configuration at the C-4 position and from the L-galacto- configuration at the C-5 position (Fig. A.9). The most interesting approach is by cyclization with C-2 inversion of D-allose, while inversions at the C-3 and C-4 positions can be performed on the corresponding C-furanosides. Double inversions correspond an epoxide formation followed by a regioselective opening, and, for example, the D-altro- configuration II-a can be obtained by a double inversion at C-2 and C-3 positions from D-glucose, by 5-endo-trig opening of the D-allo-epoxide. Similar analysis of the D-talo- configuration (Fig. A.8) provides the expected relationships to D-lyxose (addRed), L-ribose (addNonRed), and D-altrose (exch), but also a highly relevant approach by inversion of C-2 of D-galactose (Fig. A.9). The Quiral analysis is designed to find carbohydrate starting materials, so even this extensive analysis is far from exhaustive, as many noncarbohydrate strategies are also available. We will be able to see within individual chapters which of these strategies have been realized and which may represent future opportunities.

Figure A.9  Stereospecific transformations leading to II - a .

Similar analyses will be provided within each chapter. The following three chapters will cover access to galacto- configuration (β-C-lyxose, Chapter A.1), the altro- configuration (α-C-lyxose and α-C-ribose, Chapter A.2), and the allo- configuration (β-C-ribose, Chapter A.3). Chapter A.4 will cover the less selective/unspecified approaches to galacto-, altro-, and allo-C-furanosides and will provide an in-depth discussion of the different factors that control the stereochemistry and regiochemistry in the synthesis of C-furanosides. Both enantiomeric series are treated in each chapter, since all chemistry developed in one enantiomer is necessarily applicable to the other, providing that the alternative starting material is available. In studies where both selective and unselective results were reported, these results will be presented in separate chapters, but both will be discussed in Chapter A.4. We encourage the reader with a specific structure in mind to read Part A in its entirety, as a close structural analogy to the desired target may be more relevant than the observed selectivity in a more distantly related case. Indeed, one may generally start from the assumption that a given selectivity can be improved or even reversed. A poorly selective result may therefore be a useful starting point for developing a selective approach to the opposite stereoisomer than the one targeted by the authors of the initial report. Furthermore, in some cases, the selectivity is simply unassigned, but may be quite good.

References

1. As discussed in the general introduction, the authors wish to make a firm distinction between a stereochemical model, which must be simple and predictive, and a physical organic mechanism, for which conclusive evidence must be brought to bear on each aspect of the mechanistic proposal. The models presented here are intended to help the reader analyze and predict the stereochemical outcome of the many reactions reviewed. Few if any physical organic mechanistic studies have been performed, and therefore the stereochemical discussions in this chapter should not be considered as true. The interpretations are the authors' own, and may differ from the original publication's author's interpretation or may be contradicted by later mechanistic studies.

2. (a) Nuzillard J.-M, Banchet A, Haudrechy A. Application of the Quiral Program to the Challenge of Myoinositol Synthesis. J. Chem. Inf. Model. 2007;47:1979–1985(b) Nuzillard J.-M, Haudrechy A. Quiral: A Computer Program for the Synthesis of Chiral Molecules from Sugars. Tetrahedron Lett. 2007;48:2311–2313.

Chapter A.1

galacto-C-Furanosides (I, β-C-Lyxose)

Abstract

This chapter reviews the synthesis of C-furanosides of the galacto- configuration series. Approaches include nucleophilic and radical additions, reductions at the anomeric carbon, hydrogenation of exo-glycals, stereospecific and stereoselective cyclization reactions, stereochemical inversions at C-2 or C-3, and cycloaddition reactions. The principal stereocontrol elements are bicyclic induction, internal delivery, and thermodynamic control.

Keywords

galacto-C-furanoside; Halichondrins; Hydride reductions; Hydrogen reductions; Kirromycin; Lyxose; Stereoselective reactions; Stereospecific cyclizations

Chapter Outline

Disconnections

Natural Occurrence

A.1.1 Disconnection A

A.1.1.1 Coupling Between an Electrophilic Anomeric Carbon and a Nucleophilic Carbon Donor

A.1.2 Disconnection B

A.1.2.1 Reduction of an Anomeric Hemiketal With a Hydride Donor

A.1.2.2 Radical or Transition Metal Catalyzed Reduction of an Anomeric Heteroatomic Moiety

A.1.2.3 Hydrogenation of an Exocyclic Enol Ether (exo-Glycal) at the Anomeric Position

A.1.2.4 Hydroboration of an Exocyclic Enol Ether at the Anomeric Position (exo-Glycal)

A.1.2.5 Radical Addition to an Exocyclic Enol Ether at the Anomeric Position (exo-Glycal)

A.1.3 Disconnection C

A.1.3.1 Acid-Catalyzed Cyclization of 1,4-Diols

A.1.3.2 Intramolecular Reaction Between an Alcohol or Ether and a Sulfonate Leaving Group

A.1.3.3 Cyclization of a 1,4-Diol by a Mitsunobu-Type Reaction

A.1.3.4 Intramolecular Reaction Between an Alcohol and an Epoxide

A.1.3.5 Intramolecular Addition of an Alcohol to an Unsaturated Moiety

A.1.4 Disconnection D

A.1.4.1 Reduction of a Ketone

A.1.4.2 Stereospecific Inversion

A.1.5 Miscellaneous

A.1.5.1 Diels–Alder Approach

A.1.6 Conclusion

References

Disconnections

This chapter will be divided into four parts, each one dealing with a different retrosynthetic possibility. The discussion will focus only on the key