Download PDF
Synlett 2019; 30(16): 1855-1866
DOI: 10.1055/s-0037-1611899
account
© Georg Thieme Verlag Stuttgart · New York

Nitrosoalkenes: Underappreciated Reactive Intermediates for Formation of Carbon–Carbon Bonds

Steven M. Weinreb∗
Department of Chemistry, The Pennsylvania State University, University Park, PA 16802, USA   Email: smw@chem.psu.edu
› Author Affiliations
Further Information

Publication History

Received: 07 June 2019

Accepted after revision: 08 July 2019

Publication Date:
22 July 2019 (online)

 
 PDF Download Permissions and Reprints All articles of this category


This paper is dedicated to my longtime friend and colleague, Professor Philip DeShong, on the occasion of his 70th birthday.

Abstract

This Account describes studies carried out by my group during the past decade on both intra- and intermolecular conjugate additions of carbon nucleophiles to nitrosoalkenes. Using the Denmark protocol for the generation of nitrosoalkenes from α-chloro-O-silyloximes, a number of bridged and fused bicyclic ring systems can be prepared via the intramolecular version of this process. Intermolecular conjugate addition reactions of nitrosoalkenes with a wide variety of ester enolates as coupling partners can also be achieved efficiently using a similar procedure. Some stereochemical aspects of these nucleophilic additions have been studied with both acyclic and cyclic nitrosoalkenes. This methodology has been applied as key steps in synthetic approaches to some complex indole and Myrioneuron alkaloids.

1 Introduction

2 Conjugate Additions of Nitrosoalkenes

2.1 Background

2.2 Intramolecular Reactions

2.3 Intermolecular Reactions

2.4 Stereochemical Aspects

3 Applications to Natural Product Synthesis

3.1 Angustilodine and Related Alkaloids

3.2 Approach to Apparicine-Type Alkaloids

3.3 Myrioneurinol

4 Summary and Outlook


#

Key words

conjugate additions - alkaloid synthesis - total synthesis - enolonium ion equivalent - umpolung - indole alkaloids - Myrioneuron alkaloids

Biographical Sketch

Zoom Image

Steven M. Weinreb, Russell and Mildred Marker Professor of Natural Products Chemistry at Penn State, obtained an AB degree from Cornell University, and received his doctorate from the University of Rochester with Marshall Gates in 1967. He held NIH-sponsored postdoctoral fellowships at Columbia University during 1966–67 with Gilbert Stork and at MIT during 1967–70 with George Büchi. He began his independent scientific career at Fordham University and joined the faculty at Penn State in 1978. He was named Marker Professor in 1987. Dr. Weinreb was Head of the Department of Chemistry at Penn State from 1994 to 1998 and served as Interim Dean of the Eberly College of Science in 1998. He has served on a number of editorial boards and is currently the Executive Editor of Organic Reactions. Professor Weinreb’s research has focused on the total synthesis of natural products, heterocyclic chemistry, and the development of synthetic methods.

1

Introduction

The genesis of the work carried out by my laboratory in the area of nitrosoalkene chemistry can be traced back to two transformations that we discovered in the early 2000s. The first method involved the ring-closing metathesis (RCM) of vinyl chlorides to produce a wide variety of 5- to 7-membered carbocyclic and heterocyclic vinyl chloride products.[1] Thus, treatment of a chloro diene such as 1 with 10 mol% of the Grubbs II metathesis catalyst 2 in benzene at 65 °C afforded the chlorocyclohexene 3 in high yield (Scheme [1]). The second transformation of relevance was the observation that a chloroalkene could be converted regioselectively into the corresponding α-chloroketone under mild conditions using aqueous sodium hypochlorite in acetone/acetic acid at 0 °C.[2] For example, RCM product 3 could be transformed in excellent yield into α-chloroketone 4 using this procedure (Scheme [1]).

Zoom Image
Scheme 1 RCM of a chloroalkene and conversion of the product into an α-chloroketone

Nitrosoalkenes have been proposed as intermediates in organic chemistry for over a century, although relatively few have actually been isolated.[3] [4] These unstable, reactive species 7 are most commonly generated in situ from an α-chloro- or α-bromooxime such as 5 (easily prepared from the corresponding α-haloketone) by a dehydrohalogenation process mediated by a base (path a, Scheme [2]). A more modern and useful alternative method developed by the Denmark group[5] involves exposure of an α-halo-O-silyloxime 6 to a fluoride source to produce the nitrosoalkene 7 (path b). Nitrosoalkenes undergo a number of interesting transformations, including the ability to act as Michael acceptors with various nucleophiles, producing adducts such as 8.[4] Since our methodology described above provided a direct and regioselective access to a number of cyclic α-chloroketone derivatives, we decided to explore the feasibility of applying this chemistry in the realm of nitrosoalkenes, and were particularly interested in further investigating the scope of conjugate addition reactions of these intermediates for potential use in complex molecule synthesis.

Zoom Image
Scheme 2 Formation of nitrosoalkenes from oxime derivatives and subsequent conjugate addition of nucleophiles

# 2

Conjugate Additions of Nitrosoalkenes

2.1

Background

As noted above, nitrosoalkenes have long been known to undergo rapid conjugate additions with a variety of both hetero- and carbon-nucleophiles to produce α-substituted oxime adducts like 8 in good yields (Scheme [2]).[4] However, when forming the nitrosoalkene species starting from an α-halooxime it has been commonplace to utilize at least two equivalents of a nucleophile, one of which acts as the base for the initial dehydrohalogenation step (path a). Such a procedure, however, is inefficient and wasteful when using high value nucleophiles. On the other hand, the Denmark method that relies on treatment of an O-silyl-α-halooxime 6 with fluoride to form intermediate 7 (path b) is potentially more efficient with regard to the nucleophile stoichiometry.[5] In fact, at the outset of our work, several scattered examples had appeared describing the production and trapping of nitrosoalkenes via this protocol using a nitrogen or oxygen hetero-nucleophile to afford the corresponding conjugate addition products.[6] Moreover, two publications were extant at that time that reported cases of the generation and intermolecular nitrosoalkene trapping with carbon nucleophiles utilizing α-halo-O-silyloximes.[7] It might be added that the nitrosoalkenes are acting as enolonium ion equivalents in this general type of transformation, thereby allowing an umpolung functionalization of the position adjacent to a carbonyl group.[8]

One fact that clearly stood out to us on perusal of the literature was that there seemed to be little in the way of systematic study of this type of conjugate addition reaction. In addition, this promising methodology appeared to be underappreciated by the synthetic community and had found little significant application in complex molecule/natural product synthesis.[9] Moreover, we noticed that no examples had been reported of intramolecular conjugate additions of nucleophiles to nitrosoalkenes. We therefore decided that our initial studies would focus on executing this type of transformation in the intramolecular sense, with the goal of forming various bridged and fused ring systems.


# 2.2

Intramolecular Reactions

Some important considerations needed to be addressed in devising a strategy to test the feasibility of effecting intramolecular nitrosoalkene conjugate additions. The plan was to utilize the two reactions shown in Scheme [1] to prepare suitable substrates that contain both an α-chlorooxime unit and a moiety that would act as a nucleophile. Since the large preponderance of carbon nucleophiles that had been used previously in intermolecular Michael-type additions to nitrosoalkenes are malonate-derived or related soft enolates,[4] we chose to initially incorporate such functionality into the substrates. In addition, due to the fact that nitrosoalkenes generally have very short lifetimes and are prone to polymerization, it seemed clear to us that the enolate nucleophile needed to be preformed and poised to rapidly intercept the reactive intermediate as it is generated. Consequently, it appeared that the Denmark procedure for forming nitrosoalkenes would be the method of choice.

Therefore, in our first experiments, easily prepared chloro diene substrate 9 was subjected to RCM promoted by the Grubbs II catalyst to afford chlorocyclohexene 10 (Scheme [3]).[10] This product was then transformed by our methodology into the α-chloroketone 11, produced as an inconsequential mixture of stereoisomers. Conversion of ketone 11 into the α-chloro-O-silyloxime 12 was cleanly affected using commercially available O-TBS-hydroxylamine, again affording a complex mixture of stereoisomers. After some experimentation, it was determined that the optimum procedure for the desired cyclization involved initial treatment of substrate 12 with sodium hexamethyldisilazide (or in some cases the corresponding potassium base) at –78 °C, followed by addition of TBAF and slowly warming the mixture to about 0 °C. These conditions produced the desired bridged product 14 in high yield as a single oxime (E)-geometric isomer, presumably via nitrosoalkene 13. Of significance is the fact that the strong amide base utilized seems to have no adverse effects on the α-chloro-O-silyloxime functionality in this case, nor in any of the other systems that we have investigated (vide infra). For further characterization purposes, oxime 14 was converted into bicyclic ketone 15 using the Dess–Martin periodinane.[11]

Zoom Image
Scheme 3 Intramolecular conjugate addition of a nitrosoalkene

This methodology has been extended to a number of other substrates, leading to a variety of bridged and fused ring systems. For instance, using the same experimental conditions indicated in Scheme [3], the examples shown in Scheme [4] employing malonate-type nucleophiles smoothly cyclized to give the expected products.[4a]

Zoom Image
Scheme 4 Additional intramolecular nitrosoalkene cyclizations of malonate-type enolates

We have also examined the use of some other carbon nucleophiles in intramolecular nitrosoalkene cyclizations. Substrates 16, where the carbanion-stabilizing substituents X and Y could be varied, were chosen for these studies (Scheme [5]). The cyclizations to produce bicyclo[2.2.1]heptanes 17 were effected under experimental conditions very similar to those used for the transformations shown in Schemes 3 and 4. Both malonate-, 1,3-diketone- and bis-phenylsulfonyl-stabilized carbanions performed nicely in these cyclizations, affording bridged adducts 18, 19 and 20, respectively, in good yields, all having the (E)-oxime configuration.

Zoom Image
Scheme 5 Intramolecular nitrosoalkene cyclizations of a variety of soft nucleophiles

Interestingly, the differentially substituted system 16 where X = NO2 and Y = CO2Et afforded a single stereoisomeric product 21, tentatively assigned the configuration shown by NMR analysis. Similarly, cyclizations were conducted leading to bridged systems 22–24, each as mixtures of stereoisomers with ratios as indicated. At this stage, we do not have a good rationale as to why we are seeing such wide variability in stereoselectivity for the products produced in these reactions.

One other nucleophile that was tested was an enolate derived from a simple ester. In this case, substrate 25 bearing an ethyl ester moiety was first deprotonated with potassium hexamethyldisilazide, followed by the addition of TBAF, affording a mixture of cyclization products 26 and 27 in high total yield (Scheme [6]). Ester 26 was obtained as a mixture of (E)- and (Z)-oxime geometric isomers, whereas ester epimer 27 was found to be a single oxime isomer (configuration not determined).

Zoom Image
Scheme 6 Intramolecular nitrosoalkene cyclization with a simple ester enolate

Although we have not invested much time exploring applications of hetero-nucleophiles in this process, a few systems have been investigated that involve conjugate additions of sulfonamide anions to nitrosoalkenes.[10b] [12] Interestingly, it was found in this variation of the methodology that an added base is not necessary, as TBAF is capable of sulfonamide deprotonation as well as being effective in generating the nitrosoalkene from the α-chloro-O-silyloxime functionality. Thus, when substrate 28 was treated with 2 equivalents of TBAF in acetonitrile at 0 °C, piperidine derivative 29 was produced in good yield (~2:1 oxime geometric isomers) (Scheme [7]). Similarly, cyclization of substrate 30 afforded the bridged azabicyclo[2.2.2]octane system 31 in a moderate 34% (unoptimized) yield, again as a mixture of oxime E/Z isomers. Finally, the isomeric sulfonamide precursor 32 led to the azabicyclo[3.2.1]octane system 33 in good yield (~5:1 E/Z mixture) under the same experimental conditions.

Zoom Image
Scheme 7 Intramolecular nitrosoalkene cyclizations with sulfonamide nucleophiles

# 2.3

Intermolecular Reactions

In addition to the research described in the previous section on intermolecular nitrosoalkene conjugate additions, we began to examine some analogous intramolecular reactions with the plan of eventually applying this methodology in the area of natural product total synthesis. In anticipation of these future applications (vide infra), we were particularly interested in testing the feasibility of executing intermolecular conjugate additions of nitrosoalkenes formed by the Denmark strategy, with a particular focus on using ester enolates as the nucleophiles. As previously noted,[7] at the time we began our studies in this area, relatively little had been documented with regard to using the Denmark protocol in such conjugate addition processes. Moreover, since additions of carbon nucleophiles to nitrosoalkenes derived from cyclic ketones[13] as well as aldehydes[7a] had received scant attention relative to those formed from acyclic ketones, we opted to concentrate on exploring reactions involving these sorts of systems in order to probe the scope of the methodology.

A general experimental procedure to effect this transformation was therefore developed and tested with a variety of substrate combinations (Scheme [8]).[14] Thus, an ester derivative 34 (1.2 equiv) can first be converted into the corresponding enolate with potassium hexamethyldisilazide in THF at –78 °C. Subsequent addition of an α-chloro-O-silyloxime 35 (1.0 equiv) to the enolate solution is then followed by the slow addition of TBAF in THF (1.2 equiv). The mixture is then slowly warmed to 0 °C, and after about two hours the reaction is worked up to yield the alkylation product 36. A number of representative examples of this process, including the isolated yields of oxime products, are listed in Scheme [8]. As can be seen, it is even possible to generate two contiguous quaternary carbon centers by this process when using aldehyde-derived nitrosoalkenes.

Zoom Image
Scheme 8 Intermolecular conjugate additions of nitrosoalkenes with ester enolates via the Denmark protocol (bold line indicates bond formed)

Some notable observations were made during these studies. Interestingly, with a few of the ester/α-chloro-O-silyloxime substrate combinations, variability in yield was observed depending on the particular metal enolate used (i.e., Li, Na or K).[14] However, in most cases the yield of the conjugate addition product was not affected significantly by the type of ester enolate employed. Surprisingly, it was observed that the enolates of both 1,3-diketones and simple ketones did not add to nitrosoalkenes generated under these conditions. We cannot rationalize this failure since there are published examples of such Michael reactions of nitrosoalkenes, although these species were produced by direct base-promoted dehydrohalogenation of α-halooximes (Scheme [2], path a).[4] [15]

Moreover, with some nitrosoalkene systems, it was observed that tautomerization competes with the desired conjugate addition of a nucleophile. Thus, with the α-methyl-substituted substrate 37, only the α,β-unsaturated oxime 39 was obtained in all attempted enolate additions (Scheme [9]). This by-product is presumably formed via base-promoted tautomerization of the intermediate nitrosoalkene 38.

Zoom Image
Scheme 9 Tautomerization of a highly substituted nitrosoalkene

One other aspect of intermolecular nitrosoalkene conjugate additions which we have investigated involves the utilization of aryl cuprate reagents as nucleophiles.[16] Using this methodology, one can easily convert readily available α-chloroaldoximes 40 into a variety of α-aryl nitriles 43 using aryl lithiocyanocuprate reagents (2 equiv) in a one-pot procedure (Scheme [10]). This transformation relies on the initial formation of a transient, reactive nitrosoalkene 41, which undergoes a conjugate addition with the organocuprate to form an α-arylaldoxime 42. Without work-up or isolation, this intermediate oxime can be dehydrated in situ with DCC (dicyclohexylcarbodiimide)[17] to produce the α-arylated nitrile in good overall yield. Some selected examples of the nitriles formed by this method are shown in Scheme [10].

Zoom Image
Scheme 10 Conversion of α-chloroaldoximes into α-aryl nitriles (bold line indicates bond formed)

This transformation was also applied to the exocyclic nitrosoalkene 45 in order to probe the stereochemistry of the process (Scheme [11]). Thus, α-chloroaldoxime 44 was treated with two equivalents of phenyl lithiocyanocuprate, followed by the usual in situ oxime dehydration with DCC, to yield arylated nitrile 46 as a single stereoisomer, resulting from equatorial conjugate addition to intermediate nitrosoalkene 45.

Zoom Image
Scheme 11 Stereoselective arylation of an exocyclic nitrosoalkene

# 2.4

Stereochemical Aspects

During the course of our studies on the utilization of nitrosoalkenes in synthesis, we decided to explore some stereochemical issues relevant to this chemistry which had not previously been addressed. Our initial studies in this area were focused on probing the stereoselectivity of conjugate additions of acyclic aldehyde-derived nitrosoalkenes, bearing a γ-stereogenic center, with carbon nucleophiles.

Thus, α-chloro-O-silyloxime substrates 47 and 48, which could be easily prepared, were converted into the corresponding nitrosoalkenes via the Denmark protocol in the presence of a malonate nucleophile (Scheme [12]).[18] In these experiments, diethyl 2-allylmalonate (2 equiv) was first deprotonated with potassium hexamethyldisilazide in THF at low temperature, followed by addition of the α-chloro-O-silyloxime 47 or 48 (1 equiv). TBAF (2 equiv) in THF was then added at –78 °C and the mixture was warmed to 0 °C to generate the nitrosoalkene 49. We were pleased to find that with both substrates the malonate enolate addition was totally stereoselective, producing exclusively the anti-diastereomeric adducts 50 and 51. Some additional experiments with related oxime substrates and a few other malonate-type enolates, as well as the use of a sulfonamide nucleophile, led to the anti conjugate addition products in all cases.

Zoom Image
Scheme 12 Stereochemistry of conjugate additions of nucleophiles to aldehyde-derived nitrosoalkenes

In order to rationalize the anti-stereochemistry produced in these conjugate additions, we first considered the structure of the intermediate nitrosoalkene. Since aldehyde-derived nitrosoalkenes such as 49 have never been isolated, the double bond configuration of these species has not been established. We speculate, however, that these reactions might well occur via the (E)-geometric isomer. If this is indeed the case, the results of these nucleophilic additions can be nicely rationalized based upon a Felkin–Ahn-type transition state (Figure [1]). In such a model, it is expected that the large γ-phenyl group would be perpendicular to the olefinic double bond of the nitrosoalkene and the methyl/methoxy substituent would be in the position shown in the figure. Subsequent Burgi–Dunitz attack of the malonate nucleophile on this reactive conformation would produce the observed anti products 50 and 51.

Zoom Image
Figure 1 Felkin–Ahn model for attack of nucleophiles on (E)-nitrosoalkenes 49 (R = Me, OMe)

In addition to the studies on acyclic systems described above, we have also investigated the diastereoselectivity of a number of conjugate additions to endocyclic nitrosocyclohexenes.[19] In some comparative experiments, nitrosocyclohexene 53 was generated via the two methods (paths a and b) shown previously in Scheme [2]. In general, however, the differences in product yields and stereoselectivity observed using malonate-derived nucleophiles were relatively small, and in all cases the trans products 54 predominated over the cis isomers 55 (Scheme [13]).

Zoom Image
Scheme 13 Stereochemical studies on conjugate additions of malonates to a nitrosocyclohexene

A few specific examples utilizing the Denmark protocol for nitrosoalkene generation from the O-silyloxime 52 (cf. Scheme [2], path b) are outlined in Scheme [13]. Some variation in the diastereoselectivities was observed, however, depending on the particular malonate enolate employed. Thus, with the malonate and 2-ethyl-malonate enolates, only the trans isomers 54 were observed. Using the 2-methyl- and 2-allyl-malonate enolates, mixtures of trans 54 and cis 55 products were obtained, with the former isomer predominating. In all cases only the (E)-oxime geometric isomers were formed. These major trans isomers likely result from the preferred axial attack on the half-chair conformation of intermediate nitrosocyclohexene 53. It should also be added that similarly high levels of the products of axial attack were obtained using alkyl and aryl cuprates as nucleophiles, as well as with hetero-nucleophiles such as thiophenoxide and a sulfonamide.[19]


#
# 3

Applications to Natural Product Synthesis

Throughout the course of my 50-year independent career, I have felt that whenever we developed new synthetic methodology, we had a responsibility to probe its applicability and practicality in solving real-world problems and it seemed apropos that we do this with nitrosoalkene conjugate additions. To us, this ‘testing’ has meant applying the method as a key step in an approach to the total synthesis of a complex natural product, most commonly one containing nitrogen. Armed with what we had learned in the studies outlined above, and having a reasonable degree of confidence that we knew enough about the scope, as well as some of the limitations, of this nitrosoalkene-based Michael chemistry, we turned to designing some applications of the methodology. Discussed below are approaches to two different families of monoterpene indole alkaloids, albeit by strategies which are conceptually related. Also described is a total synthesis of a complex Myrioneuron alkaloid where a nitrosoalkene conjugate addition step was actually not originally planned, but rather was applied (almost in desperation) after more traditional approaches failed.

3.1

Angustilodine and Related Alkaloids

In 2004 Kam and Choo reported the isolation and characterization of angustilodine (56), a new skeletal type of monoterpene indole alkaloid, from the leaves of the Malayan plant Alstonia angustiloba (Figure [2]).[20a] This structurally unique metabolite is comprised of an indole moiety appended to a cis-fused 2-azadecalin ring system, bridged by an oxepane ether. In 2008, the Morita group investigated metabolites from the same plant and isolated the N-demethyl analog of angustilodine, alstilobanine E (57), along with alstilobanine A (58), which lacks the bridged oxepane ring present in compounds 56 and 57.[20b]

Zoom Image
Figure 2 Structures of the angustilodine-type alkaloids

We were intrigued by these natural products due to their very interesting and unusual structures, and believed they would make good targets for an initial application of the nitrosoalkene conjugate addition methodology we had been exploring. Our plan was to utilize an intramolecular version of the nitrosoalkene chemistry to first construct the C-15,16 bond of the alkaloids, and subsequently apply an intramolecular variant of the Romo acylammonium enolate methodology for forming β-lactones[21] to generate the C-19,20 bond, along with the attendant functionality and stereochemistry.[22]

The original strategy we had in mind was to affect the key nitrosoalkene conjugate addition step by applying our previously developed methodology shown above in Scheme [8]. However, some preliminary experiments aimed at generating a suitable nitrosoalkene via the Denmark procedure in the presence of one equivalent of an enolate derived from different indole-2-acetate derivatives proved disappointing.[23] It therefore became necessary to use the ‘classical’ procedure for nitrosoalkene generation starting from an α-chlorooxime (Scheme [2], path a), but we were loath to waste excess nucleophile to form the reactive intermediate. Fortunately, we have been able to devise a modification of this methodology which allows utilization of only one equivalent of the valuable enolate nucleophile.[22] [23]

Thus, it was discovered that easily prepared indole ester derivative 59 could be deprotonated with lithium hexamethyldisilazide to form a dianion 61 (Scheme [14]).

Zoom Image
Scheme 14 Intermolecular nitrosoalkene conjugate addition of an indole ester enolate (TMSE = β-trimethylsilylethyl)

Addition of one equivalent of α-chlorooxime 60 [24] to dianion 61 leads to nitrosoalkene 64 along with what appears to be an equilibrium mixture of monoanions 62 and 63. Much to our delight, conjugate addition with the transient nitrosoalkene 64 occurs only at the carbon adjacent to the ester, leading to a 1.2:1 mixture of epimeric C-16 esters 65 in 99% combined yield. Interestingly, no products resulting from N-alkylation of the indole, or alkylation at the C-3 position of the indole were detected. This was a very fortuitous result, since this high regioselectivity could not have been predicted confidently in advance and we know of no good precedent for this transformation.

To continue the synthesis, the mixture of epimeric products 65 was transformed in an eight-step sequence into keto acid 66 (Scheme [15]). This intermediate was then subjected to the conditions developed by Romo for β-lactone construction.[21] Thus, treatment of keto acid 66 with 4-pyrrolidinopyridine (PPY), 2-bromo-N-propylpyridinium triflate and diisopropylethylamine in methylene chloride, along with a small amount of acetic acid to suppress C-16 ester epimerization, led to formation of the desired cis-azadecalin 67 in 94% isolated yield after reductive removal of the Cbz group. This pivotal compound, which has the requisite stereochemistry at C-15,19,20, could then be converted into all three of the alkaloids 56–58.

Zoom Image
Scheme 15 Romo cyclization and completion of the syntheses of the angustilodine-type alkaloids

# 3.2

Approach to Apparicine-Type Alkaloids

(–)-Apparicine (68), a tetracycle isolated from Aspidosperma dasycarpon (Figure [3]), is a member of a small subclass of monoterpene indole alkaloids characterized by the loss of one of the two carbons of the biosynthetic precursor, tryptamine.[25] [26]

Zoom Image
Figure 3 Structures of apparicine and the vallesamines

This alkaloid has been shown to have both antimicrobial and antiviral activity.[27] A number of other monoterpenoid indole alkaloids structurally related to apparicine were subsequently discovered, including the vallesamines (69 and 70), which were isolated from Vallesia dichotoma.[28]

Based on what had learnt during our work on the angustilodine alkaloids,[22] we decided to investigate a potentially direct route towards construction of the tetracyclic ring system of the apparicine-type alkaloids.[29] Thus, known 3-formyl indole ester 71 [30] was first treated with 2.5 equivalents of lithium hexamethyldisilazide to generate dianion 72 (Scheme [16]). Subsequent addition of one equivalent of α-chloroketoxime 60 [24] to this dianion then led to the desired coupled product 75 as a ~1:1 mixture of C-15,16 diastereomers, and as a single oxime isomer in nearly quantitative yield. As was the case with the related system described above for the angustilodine studies (cf. Scheme [14]),[22] this transformation likely proceeds by way of an initial dehydrochlorination of the α-chlorooxime 60 by the dianion 72 to generate the nitrosoalkene 64, along with an intermediate monoanion that is, once again, probably an equilibrium mixture of 73 and 74. As previously observed, the conjugate addition to nitrosoalkene 64 occurs exclusively via the ester enolate form 74 to afford the observed product 75.

Zoom Image
Scheme 16 Ester enolate nitrosoalkene conjugate addition

Compound 75 could then be transformed into O-silyloxime amine 76 in two steps (Scheme [17]). In order to complete the synthesis of an apparicine-type ring system, simply treating amino aldehyde 76 with sodium borohydride in methanol at 0 °C led to the tetracycle 78 in high yield. This transformation probably occurs by initial reduction of aldehyde 76 to the corresponding alcohol, which then eliminates in situ to form the azafulvene intermediate 77, followed by nucleophilic addition of the piperidine nitrogen to form the observed product 78.[31] We expect that tetracycle 78, which has functional handles in place at C-16 and C-20 for further elaboration, should be a versatile intermediate for construction of a variety of alkaloids of the apparicine class.

Zoom Image
Scheme 17 Reductive closure of piperidine aldehyde 76 to tetracycle 78

# 3.3

Myrioneurinol

The Myrioneuron alkaloids are a small family of plant secondary metabolites which generally incorporate a cis-decahydroquinoline component in their structure, and also contain a 1,3-oxazine and/or 1,3-diazine moiety.[32] A new, structurally complex tetracyclic Myrioneuron alkaloid, (+)-myrioneurinol (79), was isolated from the leaves of the Vietnamese plant Myrioneuron nutans in 2007 (Figure [4]).[33] (+)-Myrioneurinol shows weak inhibitory activity against KB cells, and has moderate antimalarial activity against Plasmodium falciparum. In view of the interesting structure and biological profile of myrioneurinol, we became interested in this metabolite as a target for total synthesis.[34] As discussed below, our original strategy for accessing myrioneurinol did not include a nitrosoalkene conjugate addition. However, we ran into an intractable problem with one of the key steps that had been envisioned, and nitrosoalkene methodology proved to provide a good solution.

Zoom Image
Figure 4 Structure and conformation of (+)-myrioneurinol

The first pivotal step in our approach to myrioneurinol was to form the spirocyclic A/D ring system of the alkaloid via a stereoselective intramolecular Michael reaction. After some experimentation, it was found that readily prepared lactam α,β-unsaturated ester substrate 80, upon treatment with titanium tetrachloride/triethylamine in methylene chloride from 0 °C to room temperature, led to a single stereoisomeric spirocycle 82 in excellent yield (Scheme [18]). We postulate that product 82 is probably formed via a chelated titanium enolate intermediate like 81.

Zoom Image
Scheme 18 Stereoselective intramolecular Michael reaction of a lactam titanium enolate

The next stage of the synthesis was to annulate the B-ring onto spirocycle 82 by effecting a stereoselective alkylation at C-7. However, much to our dismay, despite considerable effort we were unable to generate an enolate from ester 82 or any other related derivatives. Similarly, although ester 82 could be converted into aldehyde 83 and subsequently used to prepare enamine 84, we failed in all attempts to carry out a Michael reaction or any other type of alkylation of this enamine (Scheme [19]).

Zoom Image
Scheme 19 Attempted alkylation at C-7

As a result of this inexplicable problem in effecting the required C-7 alkylation by standard methods, we turned to the possibility of using a nitrosoalkene in an umpolung-type alkylation. Spirocyclic ester 82 was therefore converted in a few steps into α-chloro-O-silylaldoxime 85 (Scheme [20]). We were pleased to find that using the Denmark method for nitrosoalkene generation, it was possible to affect an alkylation of intermediate 85 with dimethyl malonate to produce the desired adduct 87 as a single diastereomer at C-7 (mixture of oxime geometric isomers) in excellent yield.

Zoom Image
Scheme 20 Alkylation at C-7 via a diastereoselective nitrosoalkene conjugate addition

To rationalize this outcome, it is assumed that the transformation proceeds via conjugate addition of the malonate enolate on the least hindered face of a nitrosoalkene conformer 86, producing the observed configuration at C-7. However, we are presently unable to satisfactorily explain why this particular conformation would be the reactive one. In addition, as was discussed above (cf. Scheme [12]), one can only speculate that an (E)-nitrosoalkene is involved here since the geometry of such an aldehyde-derived species is presently unknown.

With alkylation product 87 in hand, it was possible to complete the synthesis via the sequence shown in Scheme [21]. Formation of the B-ring of the alkaloid could be executed by a key step involving a stereoselective N-sulfonyliminium ion/allyl silane cyclization. Thus, conversion of compound 87 into N-sulfonyllactam allyl silane 88 could be affected in a few straightforward operations. Partial reduction of this lactam with diisobutylaluminum hydride to aminal 89, followed by FeCl3-promoted formation of N-sulfonyliminium ion 90, and subsequent cyclization with the allyl silane, presumably via the chair-like conformation shown, yielded tricycle 91 as a single stereoisomer. This compound could then be converted into racemic myrioneurinol (79) in a few additional steps.

Zoom Image
Scheme 21 Completion of the synthesis of (±)-myrioneurinol

#
# 4

Summary and Outlook

Although we were certainly not the first to investigate conjugate additions of carbon nucleophiles to nitrosoalkenes,[4] it is hoped that our work outlined here will lead to an increased awareness by the synthetic community of the scope and potential of this methodology. In particular, the demonstration that the Denmark procedure for nitrosoalkene formation can be used efficiently in conjugate additions of carbon nucleophiles might lead to utilization of this chemistry in the future for complex molecule synthesis. Moreover, the intramolecular version of this process may provide convenient access to some interesting bridged and fused carbocycles and heterocycles.

What’s next? Probably nothing by me since I am planning to retire soon. However, there are some obvious gaps in this area which need to be filled. For example, methodology for effecting enantioselective reactions of nitrosoalkenes is in short supply. One nice report has appeared on catalytic enantioselective conjugate additions of heteronucleophiles (i.e., aryl thiols) to a number of nitrosoalkenes generated from α-chlorooximes, giving products with good enantiomeric ratios.[35] However, there appear to be no examples published to date of enantioselective conjugate additions of carbon nucleophiles. One would expect that such methodology could be exceptionally valuable. It might be added that it is well known that nitrosoalkenes can also act as heterodienes in [4+2] cycloadditions, and a report has appeared that describes a catalytic enantioselective version of this process.[36]

Finally, although a number of methods other than those shown in Scheme [2] exist for generating nitrosoalkenes,[4] most are rather obscure, and appear to have little synthetic value. In order to move this area forward even further, it would be useful if some new practical methods can be devised for producing a variety of structural types of nitrosoalkenes. As was mentioned above, we have had some difficulties in additions of certain kinds of carbon nucleophiles to nitrosoalkenes formed via the Denmark procedure, and these problems might be alleviated if other usable methods of generation become available. Any new methods for producing nitrosoalkenes will, of course, have to take into account the fact that these are highly reactive intermediates that are prone to polymerization, as well as other side reactions such as tautomerization (cf. Scheme [9]).[37]


#
#

Acknowledgment

S.M.W. thanks the numerous former coworkers whose efforts have made our contributions in the area of nitrosoalkenes possible. Their names are indicated in the publications cited in this Account.



   Permissions and Reprints All articles of this category
Zoom Image
Zoom Image
Scheme 1 RCM of a chloroalkene and conversion of the product into an α-chloroketone
Zoom Image
Scheme 2 Formation of nitrosoalkenes from oxime derivatives and subsequent conjugate addition of nucleophiles
Zoom Image
Scheme 3 Intramolecular conjugate addition of a nitrosoalkene
Zoom Image
Scheme 4 Additional intramolecular nitrosoalkene cyclizations of malonate-type enolates
Zoom Image
Scheme 5 Intramolecular nitrosoalkene cyclizations of a variety of soft nucleophiles
Zoom Image
Scheme 6 Intramolecular nitrosoalkene cyclization with a simple ester enolate
Zoom Image
Scheme 7 Intramolecular nitrosoalkene cyclizations with sulfonamide nucleophiles
Zoom Image
Scheme 8 Intermolecular conjugate additions of nitrosoalkenes with ester enolates via the Denmark protocol (bold line indicates bond formed)
Zoom Image
Scheme 9 Tautomerization of a highly substituted nitrosoalkene
Zoom Image
Scheme 10 Conversion of α-chloroaldoximes into α-aryl nitriles (bold line indicates bond formed)
Zoom Image
Scheme 11 Stereoselective arylation of an exocyclic nitrosoalkene
Zoom Image
Scheme 12 Stereochemistry of conjugate additions of nucleophiles to aldehyde-derived nitrosoalkenes
Zoom Image
Figure 1 Felkin–Ahn model for attack of nucleophiles on (E)-nitrosoalkenes 49 (R = Me, OMe)
Zoom Image
Scheme 13 Stereochemical studies on conjugate additions of malonates to a nitrosocyclohexene
Zoom Image
Figure 2 Structures of the angustilodine-type alkaloids
Zoom Image
Scheme 14 Intermolecular nitrosoalkene conjugate addition of an indole ester enolate (TMSE = β-trimethylsilylethyl)
Zoom Image
Scheme 15 Romo cyclization and completion of the syntheses of the angustilodine-type alkaloids
Zoom Image
Figure 3 Structures of apparicine and the vallesamines
Zoom Image
Scheme 16 Ester enolate nitrosoalkene conjugate addition
Zoom Image
Scheme 17 Reductive closure of piperidine aldehyde 76 to tetracycle 78
Zoom Image
Figure 4 Structure and conformation of (+)-myrioneurinol
Zoom Image
Scheme 18 Stereoselective intramolecular Michael reaction of a lactam titanium enolate
Zoom Image
Scheme 19 Attempted alkylation at C-7
Zoom Image
Scheme 20 Alkylation at C-7 via a diastereoselective nitrosoalkene conjugate addition
Zoom Image
Scheme 21 Completion of the synthesis of (±)-myrioneurinol