Macrolactonization reaction

By use of chelating tri- and tetramines 59 or 60 Kharasch additions can be performed in 65-75% yield using 0.3-10 mol% FeCl2 as the catalyst (Fig. 13) [116, 117]. The same catalysts are very efficient to promote otherwise difficult ATRC of co-alkenyl trichloroacetates 58 providing five-membered lactone 61 by 5-exo cycli-zation in 55% yield and eight- to ten-membered lactones 62 by 8-10-endo cyclizations in 34-50% yield, respectively. Even oo-allyl oligo(ethyleneoxy) trichloroacetates underwent radical macrolactonization reactions in 56-60% yield with 10 mol% of catalyst. 

The high reactivity of this polymeric-related DCC 5 is particularly noticeable in the macrolactonization reaction of 12, 13, 15 and 16-carbon to-hydroxy acids 6 (Scheme 7.2). Here, the combination of carbodiimide 5 and DMAP gave rise to the corresponding macrolactones 7 in identical yields to when DCC was used-the yield being even higher for the 13-membered lactone 7a (only a 32% yield was obtained using DCC) [16]. 

Scheme 7.2 Macrolactonization reaction using polymeric carbodiimide 5.

Diethyl azodicarboxylate (Et02C-N=N-C02Et, DEAD) is a key reagent in the Mitsunobu reaction (sec. 2.7.A.ii) and has also been used for macrolactonization. Reaction of 230 with DEAD gave 15% of 231 in White s synthesis of the antibiotic vermiculine. ... 

A variety of specialized reagents have been developed for macrolactonization reactions. Two of the more important are the Corey-Nicolaou reagent, 2,2 -dipyridyl disulfide (232)10 jjjg Mukaiyama reagent, which is 2-chloro-l-methylpyridinium iodide (233). A number of related reagents have been developed, including imidazole disulfide 234 [2,2 -dithio-(4-ferf-butyl-l-isopropylimidazole)]l 2 and imidazole 235 [N-(trimethylsilyl)imidazole].l 3 All of these reagents are effective for the cyclization of co-hydroxy acids, as shown in Table 6.1. For comparison, available cyclization results with DEAD and 234 are included. In this table, a hydroxy acid is converted to a lactone (219). 

White et aJ. [711 reported macrolactonization reaction that proceeds via allylic C - H oxidation of linear to-alkenoic acids to furnish 14- to 19-membered alkyl and aryl macrolides, promoted by 10mol% Pd(OAc)2/phenyl bissulfoxide/benzoquinone system (Scheme 6.8). 

Chemoselective carbonylation of a vinyl iodide 34 with alcohol containing a vinyl bromide moiety 35 has been successfully employed for the solid-phase synthesis of a macrosphelide precursor 36 [43]. After the 4-methoxyphenylmethyl (MPM) group was removed, the palladium-catalyzed carbonylative macrolactonization of the vinyl bromide 37 achieved the synthesis of the macrosphelide-supported derivative 38 (Scheme 9.13). The combinatorial synthesis of a 122-member macrosphelide library has been performed by the three-component strategy based on the palladium-catalyzed chemoselective carbonylation/macrolactonization reaction. 

Then, a macrolactonization reaction (Scheme 2.5) was reported by White in 2006 [10]. The strategy provided medicinal chemists with a useful tool to prepare macrolactonization of m-alkenoic acids via Pd-catalyzed C-H activation... 

Our strategy for the synthesis of (+)-dactylolide (2.217) is outlined in Scheme 2.69. We envisioned that the 20-membered macrolactone in 2.332 could be constructed by intramolecular iV-heterocyclic carbene (NHC)-catalyzed oxidative macrolactonization of co-hydroxy aldehyde 2.333. Intramolecular NHC-catalyzed oxidative esterification reactions have been recognized as an attractive tool and rapidly growing area in the synthetic community. Indeed, several examples of these reactions have recently been reported [208-216], which clearly provide a new opportunity for the development of catalytic acyl transfer agents in macrolactonization reactions of co-hydroxy aldehydes in the presence of oxidants. The substrate for the macrolactonization reaction would be derived firom the cyanohydrin alkylation of 2,6-dr-tetrahydropyran enal 2.335 with dienyl chloride 2.334. 2,6 -di-tetrahydropyran enal would in turn be constructed by employing the 1,6-oxa conjugate addition reaction of co-hydroxy 2,4-dienal 2.336. Despite the... 

Since NHCs have not yet been exploited as acyl transfer agents in macrolactonization reactions, our report, therefore, constitutes the first example of the NHC-catalyzed oxidative macrolactonization of co-hydroxy aldehydes. Because of significant benefits of the reaction, including the catalytic nature and mild reaction conditions of the reaction, the NHC-catalyzed oxidative macrolactonization reaction would provide a significant advance in the field of macrolactonization. 

Interested in facilitating access to the macrolactone, we also explored the macrolactonization by employing the Shiina s procedure [191-193] (Scheme 2.81). To this end, the seco-acid 2.359 of the aldehyde 2.333 was independently prepared and subjected to Shiina s MNBA macrolactonization reaction conditions. The reaction smoothly proceeded to provide the macrolactone 2.332 in 81 % yield. At this point, we hypothesized that the result was most likely due to the gem-disubstituent effect induced by the TBS-protected cyanohydrin functionality. As mentioned, in the previous syntheses of (+)-dactylolide, several attempts for the macrolactonization of the structurally related substrates without the cyanohydrin group either failed to afford the corresponding macrolactone or resulted in low to modest yields (21 8 %) under various macrolactonization conditions, suggesting that the TBS-protected cyanohydrin in our substrate appears to have a noticeable effect in the NHC-catalyzed oxidative macrolactonization. 

Use of 2,4,6-trichlorobenzoyl chloride, Et3N, and DMAP, known as the Yamaguchi method,128 is frequently used to effect macrolactonization. The reaction is believed to involve formation of the mixed anhydride with the aroyl chloride, which then forms an acyl pyridinium ion on reaction with DMAP.129... 

Intramolecular lactonization can also be carried out with DCCI and DMAP. As with most other macrolactonizations, the reactions must be carried out in rather dilute solution to promote the intramolecular cyclization in competition with inter-molecular reaction, which leads to dimers or higher oligomers. A study with 15-hydroxypentadecanoic acid demonstrated that a proton source is beneficial under these conditions and found the hydrochloride of DMAP to be convenient.130... 

K. C. Nicolaou s group at Scripps Research Institute developed two synthetic routes to epothilone A. One of the syntheses involves closure of the lactone ring as a late step. Three major fragments were synthesized. The bond connection at C(6)-C(7) was made by an aldol reaction. The C(12)-C(13) bond was formed by a Wittig reaction and later epoxidized. The ring was closed by macrolactonization. 

This synthesis is shown in Scheme 13.59. Two enantiomerically pure starting materials were brought together by a Wittig reaction in Step C. The aldol addition in Step D was diastereoselective for the anti configuration, but gave a 1 1 mixture with the 6S, 1R-diastereomer. The stereoisomers were separated after Step E-2. The macrolactonization (Step E-4) was accomplished by a mixed anhydride (see Section 3.4.1). The final epoxidation was done using 3-methyl-3-trifluoromethyl dioxirane. 

A Mitsunobu process simultaneously coupled the enyne acid fragment 4 to /J-lactam 10 and inverted the CIO stereochemistry to the required (S)-configured ester 11 in 93% yield. A deprotection provided alcohol 12, the key /J-lactam-based macrolactonization substrate, which, under conditions similar to those reported by Palomo for intermolecular alcoholysis of /J-lactams (Ojima et al, 1992, 1993 Palomo et al, 1995), provided the desired core macrocycle 13 of PatA 13 (Hesse, 1991 Manhas et al, 1988 Wasserman, 1987). Subsequent Lindlar hydrogenation gave the required E, Z-dienoate. A Stille reaction and final deprotection cleanly provided (-)-PatA that was identical in all respects to the natural product (Romo etal, 1998 Rzasaef al, 1998). This first total synthesis confirmed the relative and absolute configuration of the natural product and paved the way for synthesis of derivatives for probing the mode of action of this natural product. 

The synthesis of the disaccharide subunit 85 of tricolorin A, a cytotoxic resin glycoside isolated from lpomoea tricolor, provides a unique opportunity to compare the efficiency of an RCM-based macro cyclization reaction with that of a more conventional macrolactonization strategy. Furthermore, this specific target molecule challenges the compatibility of the catalysts with various functional groups. 

Nicolaou et al. were the first to report the successful use of RCM to prepare the 16-membered macrolactone nucleus of the epothilones and present a strategy for their total synthesis based on this reaction. The approach involved formation of the C12,C13 olefin and is outlined in Scheme 2 [12,13]. 

In parallel investigations, Danishefsky and coworkers accomplished the preparation of the 16-membered lactone of a model epothilone system via an alternative C9,C10 disconnection [14] (Scheme 4). In this case, coupling of epoxy-alcohol 17 with acids 18a and 18b afforded trienes 19a and 19b respectively. RCM of 19a under the influence of ruthenium initiator 3 produced dienes 20a as a 1 1 mixture of Z -isomers. Under identical conditions, cyclization of 19b produced a single product 20b (tentatively assigned as the Z-isomer). The variable stereoselectivity observed in these reactions was inconsequential since the olefinic functionality could be reduced to afford the corresponding saturated macrolactones. Schrock s molybdenum initiator 1 promoted the cyclization of 19a and 19b with similar efficacy [14]. 

Aldol reaction of keto-acid 21 with aldehyde 10 and esterification of the resulting acids with alcohol 22 led rapidly to cyclization precursor 23 and its 6S,7R-diastereomer (not shown). RCM using ruthenium initiator 3 (0.1 equiv) in dichloromethane (0.0015 M) at 25 °C afforded macrolactones 24a and 24b in a 1.2 1 ratio. Deprotection and epoxidation of the desired macrolactone, 24a, afforded epothilone A (4) via 25a (epothilone C) (Scheme 5). Varying a number of reaction parameters, such as solvent, temperature and concentration, failed to improve significantly the Z-selectivity of the RCM. However, in the context of the epothilone project, the formation of the E-isomer 24b could actually be viewed as beneficial since it allowed preparation of the epothilone A analog 26 for biological evaluation. 

---