Synthetic Studies Towards Natural Products Containing Bis-Spiroacetals

发布时间：2016年05月06日 来源：中国化学会

Margaret A. Brimble
Recipient of FACS Young Chemist Award 2001
Department of Chemistry, University of Auckland, Symonds St., Auckland, New Zealand

Our research is primarily concerned with the synthesis of biologically active natural products. Some of our current synthetic targets were described in the previous newsletter. In this lecture I have decided to focus on some of our synthetic work directed towards the synthesis of natural products that contain a bis-spiroacetal ring system which has interesting stereochemical issues for consideration. The pharmacological importance of spiroacetal containing compounds is evident from their widespread occurrence as metabolites from insects, microbes, plants, fungi and various marine organisms. The important biological activity of this class of compound has prompted a variety of methods for the synthesis of spiroacetals [1]. In contrast to their bicyclic analogues, the chemistry of tricyclic bis-spiroacetals, in which two acetal carbons are linked in a spiro fashion, has not been fully explored. 

POLYETHER ANTIBIOTICS EPI-17-DEOXY-(O-8)-SALINOMYCIN AND CP44,161

In 1973 the polyether antibiotic salinomycin 1, which contains the 1,6,8-trioxadispiro[4.1.5.3]pentadec-13-ene bis-spiroacetal unit, was isolated from the culture broth of Streptomyces albus [2] and was found to exhibit activity against mycobacteria and fungi and acted as a coccidiostat for poultry and as a growth promotant for ruminants. Using the same S. albus culture and a different medium, Westley et al. [3] established that epi-17-deoxy-(O-8)-salinomycin 2 was found to be present at much greater levels. Further examples of polyether antibiotics which contain the 1,6,8-trioxadispiro[4.1.5.3]pentadec-13-ene ring system include: narasin A (4-methylsalinomycin) 3 [4] from S. aureofaciens, noboritomycins A 4a and B 4b from S. noboritoensis [5], CP44,161 5 [6] from a Dactylsporangium species and the halogenated polyether antibiotic, antibiotic X-14766A 6 [7]. 

The stereochemistry of the bis-spiroacetal ring system in the above polyether antibiotics needs addressing. The four possible stereoisomers of the bis-spiroacetal ring system are depicted (Fig. 1). Diastereomer A depicts the stereochemistry adopted by salinomycin 1 and has three stabilizing anomeric effects but exhibits unfavourable 1,3-dipole-dipole interactions. 21-epi-salinomycin B has only one anomeric effect and is the thermodynamically least stable diastereomer. The 17-epi-diastereomer C exhibits three anomeric effects and although it exhibits unfavourable 1,3-diaxial interactions between the C17 oxygen atom and the C21 methylene it lacks the unfavourable 1,3-dipole-dipole interactions exhibited by diastereomer A and 17-epi-21-epi-diastereomer D. 

This qualitative analysis leads to the assumption that 17-epi-diastereomer C exhibits the most thermodynamically stable configuration. It transpires that in the cyclic structure that salinomycin 1 adopts, the repulsive 1,3-dipolar interactions in diastereomer A are compensated for by a hydrogen bond between the C9 and C20 hydroxy groups. Bis-spiroacetals whose structures preclude this remote hydrogen bond do not adopt the salinomycin configuration A. 

To date there have been three total syntheses of salinomycin 1 [8,9,10] which made use of a stereoselective aldol reaction to construct the C9-C10 bond (Scheme 1). Whilst the synthetic approaches to salinomycin 1 by Kishi, Yonemitsu and Kocienski, have focused on late assembly of the C ring after appending the D,E rings to the B ring, our synthetic efforts have focused on the construction of a tricyclic bis-spiroacetal core containing the B,C,D rings with the idea of appending the A and E rings at a later stage in the synthesis. 

Our initial work in this area focused on the synthesis of the 1,6,8-trioxadispiro[4.1.5.3]pentadec-13-ene ring system via oxidative cyclisation of an hydroxyspiroacetal to a bis-spiroacetal [11] (Scheme 2). Treatment of hydroxyspiroacetal 7 with iodobenzene diacetate in iodine under photolytic conditions afforded trans bis-spiroacetal 8 and cis bis-spiroacetal 9 in a 2.5:1 ratio. The stereochemistry of the major bis-spiroacetal 8 was the same as that present in epi-17-deoxysalinomycin 2 hence we embarked on a synthesis of epi-17-deoxysalinomycin 2 adopting the retrosynthesis outlined (Scheme 3). The key bis-spiroacetal 10 could be prepared by oxidative cyclisation of iodospiroacetal 11 since extensive model work [12] also established that an iodomethyl group was not only compatible with the key oxidative cyclisation step but was also thought to be readily converted to an aldehyde when elaboration of the E ring was required. 

The synthesis of the key cyclisation precursor, iodide 11, initially required the synthesis of optically active lactone 12, and acetylene 13 with the required S configuration at C-2. Lactone 12 was prepared [13] using methodology developed by Evans and Bartroli [14] for the synthesis of Prelog-Djerassi lactone. Acetylene 13 was prepared from (S)-(-)-lactonic acid which in turn was readily available by resolution of racemic lactonic acid using cinchonine. 

Generation of the acetylide derived from 13 followed by the addition of lactone 12 afforded methyl acetals 14 after in situ treatment with acidic methanol (Scheme 4). Semi-hydrogenation of the acetylene to a cis-alkene followed by treatment with a catalytic quantity of of pyridinium p-toluenesulfonate afforded a 1:1 mixture of spiroacetals 15. This thermodynamically controlled cyclisation affords the most stable configuration at the newly formed spirocentre due to maximum stabilisation by the anomeric effect with the two isomers of spiroacetal 15 differing only in the position that the side chain adopts. With spiroacetals 15a and 15b in hand, the neopentyl-like tosylates were converted to the iodides 11a and 11b (via intermedicay of an epoxide) in readiness for the key oxidative cyclisation. 

Finally the individual iodides 11a and 11b were treated with iodobenzene diacetate and iodine with irradiation from a tungsten filament lamp to afford a 1.7:1 mixture of the trans bis-spiroacetal 16a and the cis bis-spiroacetal 16b. It was pleasing to note that the major isomer had the same stereochemistry as the bis-spiroacetal ring in epi-17-deoxy-(O-8)-salinomycin 2. 

At this stage we had prepared bis-spiroacetal 16 which was a suitable advanced intermediate for the synthesis of epi-17-deoxy-(O-8)-salinomycin 2, however, further elaboration to append the E ring required conversion of the iodide to a hydroxy group. Unfortunately, conversion of the neopentyl-like iodide 16a into a hydroxy group required the use of potassium superoxide and 18-crown-6 which also deprotected the tert-butyldiphenylsilyl ether at the left hand end of the molecule (Scheme 5). This problem was later solved [15] by using an acetate group rather than an iodide in the cyclisation precursor. 

We next addressed the strategy for attachment of the E ring to the BCD fragment. Towards this end, our attention initially focused on a model system, namely, the conversion of the simpler aldehyde 17 and bromide 18 to bicyclic ether 19 (Scheme 6) [16]. Chelation controlled addition of the Grignard reagent derived from bromide 18 to aldehyde 17 afforded predominantly erythro alcohol 19. Treatment of alcohol 19 with iodine in acetonitrile affords predominantly iodoether 20 with the exact ratio of 20:21 depending on the temperature used. Individual treatment of each iodide with silver carbonate in wet acetone afforded in each case a single, yet different ring expanded product (20 gives 22 and 21 gives 23). 

Given that iodoether 21 affords pyran 23 which has the same stereochemistry as the E ring of epi-17-deoxy-(O-8)-salinomycin 2, it then remained to alter the stereochemical outcome of the iodoetherification such that the amount of the desired iodoether 21 was increased. Following Bartlett's rationalisation [17] for the synthesis of cis-2,5-disubstituted tetrahydrofurans, bulkier ether derivatives were used in the iodoetherification. However, the critical iodoetherification could not be induced to favour the iodoether 21 required for elaboration to the desired bicyclic ether 23. An alternative approach [18] based on an acid catalysed cyclisation of a hydroxyepoxide and ring expansion of the mesylate derived from the resultant tetrahydrofuranyl alcohol provided little improvement in stereoselectivity for the desired pyran 23. 

Given that aldehyde 17 had been successfully converted to a bicyclic ether, we then tried to extend our model work to the incorporation of an E ring fragment onto the model bis-spiroacetal aldehyde 24 (Scheme 7). Addition of the Grignard reagent derived from bromide 18 to aldehyde 24 afforded predominantly the erythro alcohol 25, however, it was at this stage that we discovered that the critical iodoetherification step was incompatible with the sensitive bis-spiroacetal ring system. 

An alternative approach was also hampered by the presence of the bis-spiroacetal (Scheme 8). Epoxidation of alcohol 25 afforded a 1:1 mixture of epoxides 26a and 26b which then underwent acid catalysed cyclisation to alcohols 27a and 27b respectively. Subsequent mesylation and attempted silver assisted ring expansion, however, resulted in destruction of the bis-spiroacetal. 

Having committed to the strategy of appending the tetrahydropyran E ring fragment to a BCD bis-spiroacetal it was disappointing to find that the proposed ring expansion was not feasible in the presence of the bis-spiroacetal. Our attention therefore turned to the synthesis of antibiotic CP44,161 5 which has a substituted tetrahydrofuran as the E ring and thereby avoiding the undesirable ring expansion step [19]. 

Antibiotic CP44,161 has not been synthesised to date and has the same bis-spiroacetal stereochemistry as salinomycin 1. Aside from the aromatic A ring and the five membered E ring in CP44,161 5, the main differences between the bis-spiroacetal moieties of these two molecules are the presence of an additional methyl group and an ethyl rather than a methyl group in the D ring of CP44,161 5. The retrosynthesis adopted for antibiotic CP44,161 (Scheme 9) also uses an aldol disconnection to afford an aromatic left hand portion and the right hand fragment 28 which is further disconnected to the bis-spiroacetal aldehyde 29 and (E)-alkene 30. Finally use of an oxidative cyclisation to construct bis-spiroacetal 29 affords the same lactone 12 as that used earlier and acetylene 31 which has the requisite ethyl and methyl groups at C2 and C4 respectively. 

The proposed synthesis of acetylene 31 (Scheme 10) commenced with the alkylation of propanoyloxazolidinone 32 with allyl iodide 33 [20] to form alkene 34. Based on the Sharpless mnemonic, asymmetric dihydroxylation [21] of the terminal olefin using potassium osmate and (DHQ)2PHAL, was expected to afford diol 35. However, in the final stages of this work, X-ray diffraction studies subsequently revealed that lactone 37, produced by cyclisation of the unexpected diol 36, was in fact the major product. Lactone 37 contains the incorrect stereochemistry at C4 to that required for the formation of acetylene 31. It is unclear why the facial selectivity of dihydroquinine ligands did not follow the Sharpless mnemonic, however, the presence of the chiral oxazolidinone moiety may have been a contributing factor. 

Lactone 37 was converted to acetylene 40 which was then used to produce a tetracyclic fragment resembling the B,C,D and E rings of antibiotic CP44,161 5 (Schemes 11, 12). The synthesis of acetylene 40 was completed by reduction of lactone 37 with lithium borohydride to afford triol 38 which, after protection of the 1,2-diol as an acetonide, was oxidised at the remaining primary alcohol to afford aldehyde 39. Grignard reaction of aldehyde 39 with propargylmagnesium bromide resulted in the formation of an alcohol which, after protection as a silyl ether afforded acetylene 40 as a 1:1 mixture of diastereomers.

With acetylene 40 and lactone 12 in hand, assembly of the bis-spiroacetal core was effected based on the earlier epi-17-deoxy-(O-8)-salinomycin 2 work (Scheme 4). Addition of the lithium acetylide derived from acetylene 40 to lactone 12 followed by treatment with acidic methanol afforded methyl acetals 41. After protection of the primary hydroxyl group as an acetate, partial hydrogenation to a cis-olefin followed by acid catalysed cyclisation resulted in a 1:1 mixture of spiroacetals 42a and 42b.

Spiroacetals 42a and 42b were treated with iodobenzene diacetate and iodine to afford a 3.3:1 mixture of tricyclic bis-spiroacetals 43a and 43b. The preference for cis bis-spiroacetal 43a in this cyclisation reaction can be attributed to the presence of the additional methyl group in the D ring, which causes unfavourable steric interactions upon formation of the minor trans bis-spiroacetal 43b. Trans Bis-spiroacetal 43b therefore undergoes rapid epimerisation at the allylic spirocentre to cis bis-spiroacetal 43c. The presence of the additional methyl group exhibited a marked effect on the stereochemical outcome of the oxidative cyclisation in that the oxidative cyclisation of spiroacetal 11 which lacks this methyl group provided the trans isomer as the major product (Scheme 4).

The major bis-spiroacetal 43a isolated from this oxidative cyclisation has the 17-epi-21-epi-salinomycin stereochemistry (Fig. 1) whereas the minor cis isomer 43c has the correct bis-spiroacetal stereochemistry for salinomycin 1 and CP44,161 5. In view of the fact that in the previous total syntheses of salinomycin 1 the correct stereochemistry for the bis-spiroacetal ring system was obtained via a thermodynamically controlled cyclisation after the whole carbon skeleton of the natural product was assembled, it was decided to pursue appendage of the E ring to the major isomer of the BCD fragment 43a. This approach was justified in that it is well established that long range hydrogen bonding in the final molecule can dramatically alter the position of the bis-spiroacetal equilibrium. 

With the bis-spiroacetal ring assembled, hydrolysis of the major cis bis-spiroacetal acetate 43a followed by oxidation using tetrapropylammonium perruthenate afforded aldehyde 44. Applying methodology established in synthetic approaches to the D and E rings of epi-17-deoxy-(O-8)-salinomycin 2, the union of bromide 30 and aldehyde 44 using a Barbier reaction resulted in the successful synthesis of alcohol 45. Treatment of alcohol 45 with dimethyl dioxirane resulted in a 1:1 mixture of epoxides 46 and 47 which, after treatment with a catalytic quantity of pyridinium p-toluenesulfonate, underwent cyclisation to afford polyethers 48 and 49 which were separated by HPLC. 

In conclusion, polyethers 48 and 49 were synthesised from aldehyde 44 and bromide 30. Noteworthy features of the synthetic strategy adopted include the oxidative cyclisation of a bicyclic hydroxyspiroacetal to a bis-spiroacetal which provides cis bis-spiroacetal aldehyde 44 preferentially; the addition of a Grignard reagent derived from bis-homoallylic bromide 30 to a neopentyl-like aldehyde 44; and acid catalysed cyclisation of a -hydroxyepoxide to a disubstituted tetrahydrofuran in the presence of a sensitive bis-spiroacetal. 

The synthetic work outlined herein provides a framework on which to synthesise the B,C,D and E rings of antibiotic CP44,161 5 after synthesising acetylene 31 from lactone 50 (via the correct diol 35) which then provides access to aldehyde 51 (Scheme 13). 

SHELLFISH TOXINS SPIROLIDES B AND D.

During chemical investigations of polar bioactive molecules from microalgae and shellfish, Wright et al. [22] isolated two lipid-soluble macrocycles, spirolides B 52 and D 53, from the digestive glands of both mussels (Mytilus edulis) and scallops (Placopecten magellanicus). These macrocycles contain a novel spiro-linked tricyclic ether ring system and an unusual seven-membered spiro-linked cyclic iminium moiety. The spirolides cause potent and characteristic symptoms in the mouse bioassay and their toxicological properties are under investigation. They were also found to be weak activators of type L calcium channels. 

In the same year Uemura et al. [23] isolated pinnatoxins A 54 and D 55 from the shellfish Pinna muricato and proposed a biosynthetic pathway for construction of the G ring. The pinnatoxins are also calcium channel activators and are the principle toxins responsible for outbreaks of Pinna shellfish intoxication in China and Japan. The spirolides contain a 6,5,5-bis-spiroacetal system whereas the pinnatoxins contain a 6,5,6-system. The cyclic imine is common to both the spirolides and the pinnatoxins. 

To date there is no synthesis of the spirolides, however Kishi et al. [24] have recently reported a total synthesis of pinnatoxin A 54 in which the bis-spiroacetal moiety was assembled by thermodynamically controlled cyclization of a dihydroxyketone precursor in which the tertiary alcohol in the B ring was constructed using a Sharpless asymmetric dihydroxylation. A similar strategy was used by Hirama et al. [25] for the synthesis of the bis-spiroacetal moiety of pinnatoxin A 54 whereas Murai et al. [26] introduced the tertiary alcohol by stereoselective methylation of a ketone after assembly of the bis-spiroacetal system. 

Kishi's total synthesis of pinnatoxin A 54 has confirmed that the absolute stereochemistry of pinnatoxin A 54 is in fact the antipode of the structure drawn, whereas the relative and absolute stereochemistry of the spirolides has yet to be determined. It is therefore important that any synthetic strategy directed towards the spirolides is flexible in its approach so that the synthesis of synthetic fragments will aid the stereochemical assignment of the natural products. We have therefore embarked on the synthesis of the novel bis-spiroacetal moiety of the spirolides using an approach in which the relative and absolute stereochemistry of the substituents on the B and D rings can be varied. 

The key disconnection in our retrosynthesis of spirolide B 52 and D 53 (Scheme 14) makes use of a Ni(II)/Cr(II)-mediated Kishi-Nozaki coupling to form the C9-C10 bond of the macrocyclic ring. A similar disconnection was successfully used by Kishi et al [24] in their related synthesis of pinnatoxin A. A subsituted allyl side chain is appended to C23 of the 6,5,5-bis-spiroacetal ring system via Lewis acid mediated addition of allylstannane 56 to bis-spiroacetal 57 that bears an acetate group at the anomeric position. Precedent for this step has already been established by this research group on a related bicyclic spiroacetal system [27]. 

The C27-C28 bond can be constructed by addition of allyltrimethylsilane to an iminium ion generated from lactam acetal 59 followed by conversion of the resultant allylated product to the key allylstannane 56. The protected hydroxymethyl group at C8 provides the functionality for introduction of the vinyl iodide at C9 via the intermediacy of an acetylene. It is then proposed to assemble the unusual bicyclic spiro aminoacetal 58 by intramolecular Diels-Alder addition of unsaturated aminoacetal 59 to diene 60.

Bis-spiroacetal acetate 57 is available from hydration of unsaturated bis-spiroacetal 61 which in turn is synthesized from alkene 62 via epoxidation followed by base induced rearrangement of the resultant epoxide. This procedure elegantly introduces an hydroxyl group at C19 in preparation for introduction of the necessary tertiary alcohol onto the B ring. Precedent for this latter step has also been demonstrated by this research group on a simpler bicyclic spiroacetal ring system [28]. 

Having established that the two key reactions required for assembly of the allyl substituted bis-spiroacetal fragment of the spirolides, were feasible (albeit using bicyclic spiroacetal model systems) our focus then turned the synthesis of the key bis-spiroacetal intermediate 62 making use of two iterative oxidative cyclizations to construct the novel 6,5,5-bis-spiroacetal ring system (Scheme 15). Addition of aldehyde 63 to the Grignard reagent 64 affords alcohol 65 which can then undergo oxidative cyclization to a bicyclic spiroacetal 66. Deprotection of the triethylsilyl ether to an alcohol then allows a second oxidative cyclization to the bis-spiroacetal 62 to take place. 

Aldehyde 63 provides the substituents required for the D ring of the 6,5,5-bis-spiroacetal ring system which are introduced in a stereocontrolled manner by the addition of Brown's chiral crotyl borane [29] to a protected 3-hydroxyl-propanal. The synthesis of the bis-spiroacetal moiety is therefore sufficiently flexible such that bis-spiroacetals of varying relative and absolute stereochemistry can be synthesised by the appropriate choice of chiral crotyl borane used. 

With the successful synthesis of bis-spiroacetal 62 in hand current work is addressing the stereochemistry present in the bis-spiroacetal ring system of the synthetic material compared to the naturally occurring spirolides. Work on the synthesis of the spiroimine portion of the spirolides has also been initiated. 

ACKNOWLEDGEMENTS

The author wishes to express her gratitude to the research workers, whose names appear in the references, for the invaluable contributions they have made to the research described herein. Financial support from the Australian Research Council, The University of Sydney and the University of Auckland is also gratefully acknowledged. 

REFERENCES AND NOTES

[1] For reviews see: (a) A. F. Kluge, Heterocycles, 24, 1699 (1986); (b) F. Perron and K.F. Albizati, Chem. Rev., 89, 1617 (1989); (c) T. L. B. Boivin, Tetrahedron, 43, 3309 (1987); (d) M. F. Jacobs and W. B. Kitching, Current Organic Chemistry, 2, 395 (1998). 
[2] H. Kinashi, N. Otake, H. Yonehara, S. Sato and Y. Saito, Tetrahedron Lett., 4955 (1973). 
[3] J. W. Westley, J. F. Blount, R. H. Evans, Jr. and C. Liu, J. Antibiot., 30, 610 (1977). 
[4] D. H. Berg and R. L. Hamill, J. Antibiot., 30, 610 (1978).
[5] J. L. Occolowitz, D. H. Berg, M. Debono and R. L. Hamill, Biomed. Mass Spectrometry, 3, 272 (1976). 
[6] J. Tone, R. Shibakawa, H. Maeda, K. Inoue, S. Ishiguro, W.P. Cullen, J. B. Routien, L. R. Chappel, C. E. Moppett, M. T. Jefferson and W.D. Celmer, Abstract 171, 18th ICACC Meeting, Atlanta, Georgia, 1978. 
[7] J. W. Westley, R. H. Evans, L. H. Sello, N. Troupe, C. Liu, J. F. Blount, R. G. Pitcher, T. H. Williams and P. A. Miller, J. Antibiot., 34, 139 (1981). 
[8] Y. Kishi, S. Hatakeyama and M. D. Lewis, Frontiers of Chemistry Plenary Keynote Lecture, 28th IUPAC Congress (1981).
[9] (a) K. Horita, Y. Oikawa, O. Yonemitsu, Chem. Pharm. Bull., 37, 1698 (1989); (b) K. Horita, S. Nagato, Y. Oikawa and O. Yonemitsu, Chem. Pharm. Bull., 37, 1705 (1989); (c) K. Horita, Y. Oikawa, S. Nagato and O. Yonemitsu, Chem. Pharm. Bull., 37, 1717 (1989); (d) K. Horita, S. Nagato, Y. Oikawa and O. Yonemitsu, Chem. Pharm. Bull., 37, 1726 (1989).
[10] (a) R. C. Brown and P. J. Kocienski, Synlett, 417 (1994); (b) R. C. Brown and P. J. Kocienski, J. Chem. Soc., Perkin Trans. 1, 9 (1998).
[11] R. Baker and M. A. Brimble, J. Chem. Soc. Perkin Trans. I, 125 (1988). 
[12] R. Baker, G. M. Williams and M. A. Brimble, J. Chem. Soc. Perkin Trans. I, 2221 (1991). 
[13] M. A. Brimble, Aust. J. Chem., 43, 1035 (1990). 
[14] D. A. Evans and J. Bartroli, Tetrahedron Lett., 23, 807 (1982). 
[15] P. R. Allen, M. A. Brimble and F. A. Fares, J. Chem. Soc. Perkin Trans. I, 2403 (1998).
[16] M. A. Brimble and M. K. Edmonds, Tetrahedron, 51, 9995 (1995). 
[17] S. D. Rychnovsky and P. A. Bartlett, J. Amer. Chem. Soc., 103, 3963 (1981).
[18] M. A. Brimble and H. Prabaharan, Tetrahedron, 54, 2113 (1998). 
[19] P. A. Allen, M. A. Brimble and H. P. Prabaharan, J. Chem. Soc. Perkin Trans. I, 379 (2001). 
[20] R. M. Magid, O. S. Fruchey, W. L. Johnson and T. G. Allen, J. Org. Chem., 44, 359 (1979). 
[21] H. C. Kolb, M. S. Van Nieuwenhze and K. B. Sharpless, Chem. Rev. 94, 2483 (1994).
[22] T. Hu, J. M. Curtis, Y. Oshima, M. A. Quillam, J. A. Walter, W. Watson-Wright and J. L. C. Wright, J. Chem. Soc., Chem. Commun., 2159 (1995).
[23] (a) D. Uemura, T. Chuo, T. Haino, A. Nagatsu, S. Fukuzawa, S. Zheng and H. Chen, J. Amer. Chem. Soc., 117, 1155 (1995); (b) T. Chuo, O. Kamo and D. Uemura, Tetrahedron Lett., 37, 4023 (1996); (c) T. Chuo, T. Haino, M. Kuramoto and D. Uemura, Tetrahedron Lett., 37, 4027 (1996).
[24] J. A. McCauley, K. Nagasawa, P. A. Lander, S. G. Mischke, M. A. Semones and Y. Kishi, J. Amer. Chem. Soc., 120, 7647 (1998).
[25] T. Noda, A. Ishiwata, S. Uemura, S. Sakamoto and M. Hirama, Synlett., 298 (1998).
[26] (a) T. Sugimoto, J. Ishihara and A. Murai, Tetrahedron Lett., 38, 7379 (1997); (b) J. Ishihara, T. Sugimoto and A. Murai, Synlett., 603 (1998).
[27] M. A. Brimble, F. A. Fares and P. Turner, J. Chem. Soc. Perkin Trans. I, 677 (1998).
[28] M. A. Brimble, R. H. Furneaux, A. D. Johnston, J. Org. Chem., 63, 471 (1998).
[29] H. C. Brown and K. S. Bhat, J. Am. Chem. Soc., 108, 293 (1986).