Date University This is to certify that the thesis entitled I. A NEW HOST-SPECIFIC TOXIN FROM HELMINTHOSPORIUM CARBONUM II. A STUDY OF ACYCLIC STEREOSELECTION V_I__A A CHELATION CONTROLLED [2,3] SIGMATROPIC REARRANGEMENT presented by Benjamin A. Horenstein has been accepted towards fulfillment of the requirements for Master of Science degree in Chemistry Major professor 11/18/85 0-7 639 MS U i: an Affirmative Action/Equal Opportunity Institution IV1£3I~1 RETURNING MATERIALS: Place in book drop to LlanARIEs remove this checkout from ”- your record. FINES W111 be charged if book is returned after the date stamped below. I. A NEW HOST-SPECIFIC TOXIN FROM EEMHWZRMWORHIIGMHMWUV II. A STUDY OF ACYCLIC STEREOSELECTION VIA A CHELATION-CONTROLLED [2,3] SIGMATROPIC REAERANGEMENT By Benjamin A. Horenmtein A DISSERTATION Submitted to Michigan State University in partial fulfillment for the degree of MASTER OF SCIENCE Department of Chemistry 1985 70 57/5 y ABSTRACT I. A NB" HOST-SPECIFIC TOXIN FROM ”MIMI” aunxwuv By Benjamin A. Horenstein Previously characterized HC-toxin, elaborated by Helsintbaspon’us Cerbanu. cyclo-(Pro-Ala-Ala-Aoe). (Aoe = 2-amino-8-oxo-9,10-epoxy decanoic acid), is toxic to susceptible genotype of maize. It has been demonstrated that the epoxy moiety of the Ace residue is essential for toxicity; however, the actual mode of action of HO-Toxin is not yet understood. In an effort to understand the physiochemical behavior of EC-toxin, the culture filtrate of Helmz‘ntbospon’um cerbamn was examined for homologous forms of the toxin. This thesis describes the isolation and structural elucidation of two HC-toxin analogs, HC-toxin II and BC-toxin III. cyclo-(Pro-Ala-Gly-Aoe) and cycle-(3- trans-Bypro-Ala-Ala-Aoe), respectively. Purification of the toxins was achieved by column chromatography on silica gel, followed by reverse-phase HPLC and, finally, normal-phase HPLC. A variety of techniques established the aforementioned structures of HC-toxin II and III, including amino acid analysis, 81- and FAB-mass spectroscopy. 13-NMR (standard, 008?, NOE-difference), and 1°C-NHR (fully decoupled, DEPT). ABSTRACT II. ACYCLIC STEREOSELECTION VIA A CHELATION-CONTROLLED [2,3] SIGMATROPIC REARRANGEMENT By Benjamin A. Horenstein The [2,3] Wittig rearrangement of substrates bearing an unstabilized carbanionic center proceeds selectively to yield Z-olefins, given alkyl substitution at the 2-position of the migrating allylic moiety. In the absence of such substituents, the rearrangement occurs in a stereorandom fashion. This thesis describes the synthesis and [2,3] rearrangement of substrates lacking Z-alkyl substitution of the migrating allylic fragment, which are capable of internal chelation of the carbanionic center with an alkoxy group adjacent to the breaking C-O bond. We have observed olefin selectivity of 3.5:1 (Z- to E-), a value higher than that obtained for similarly substituted substrates incapable of internal chelation; a model accounting for these results is discussed. nsnnmumal This work is dedicated to Professor 8. P. Tanis. Professor Tanis’ enthusiasm, helpfulness and dedication have been appreciated and will be missed. ii I would like to thank Ms. Tonya Acre for her fine work in preparation of this manuscript. Thanks to Mr. Richard (Red-Shoes) Olsen for his tremendous help in acquisition of many mass spectra. Also, thanks to Mr. Brad Ackermann for his interest, helpful discussions and work on FAB-MS of the Hc-toxins. In addition to a rewarding academic experience under the guidance of Professor Tanis, I will always appreciate the good times enjoyed by our group during pursuit of Mexican food, dry-ice ballistics and the B.S.R.I. LIST OF EQUATIONS . LIST OF FIGURES . . LIST OF TABLES. . . HC-TOXINS Introduction . . Results and Discussion Experimental . . List of References [2,3] WITTIG EEAREANGEMENT Introduction . . Results and Discussion . Experimental . . List of References iv Page vi ix 26 31 33 48 65 84 [2,3] Iittig W Equation 1 Equation 2 Equation 3 Preparation of Protected Lactaldehydes . . . . . . . . . . Alternative synthesis of compound m. 0 O O O O O O O O O O O O O 0 Preparation of 1% via Zn(BH¢)s Reduction of Enone 34 . . . . . Page 49 55 64 ln%hldms Figure 1 Figure 2 Figure 3 Figure 4 Figure 5a Figure 5b Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 IJBTCIPIHIIIE General Diogenetio Pathway for the Common Amino Acids. . . . . . . . . . Dioassay Protocol . . . . . . . . . . . Isolation Flow Chart. . . . . . . . . . FAB-MS of RC-Toxins . . . . . . . . . . 250 MR: 1E-MMR of RC-Toxin II . . . . 260 MHz 1E-NMR of HC-Toxin III. . . . COSY Spectrum of RC-Toxin II. . . . . . COSY Spectrum of EC-Toxin III . . . . . Possible sequences for EC-Toxins II and III 0 O O O 0 O I O O O O O O O O O O O RC-Toxin I Solution Conformation. . . . Definitions of Peptide Dihedral Angles. [2,3] littig W Figure Figure Figure Figure Figure l Macrolide and Polyether Natural Products. . . . . . . . . . . . . Acyclic Analog of Bafilomycin . . . . Aldol Condensations . . . . . . . Formation/Functionalization of a 0:0 Bond. 0 O O O O O O O O O O O O 0 [3,3] Sigmatropic Rearrangements. . . . vi Page l4 l5 17 18 20 23 24 33 33 35 36 38 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 6 7a 7b 7c 10 ll 12 13 l4 15 16 17 18 19 [2,3] Sigmatropic Rearrangements. Mechanistic Investigation of [2,3] Wittig Rearrangement by Rautenstrauch . Mechanistic Investigation of [2,3] Wittig Rearrangement by Baldwin . . . Mechanistic Investigation of [2,3] Wittig Rearrangement by Still . . . . Stabilized [2,3] Wittig Rearrangements. Unstabilized [2,3] Wittig Rearrange- ment. . . . . . . . . . . . . . . . . Low z-olefin Selectivity in Unstabilized [2,3] Variant. . . . . . . Stills’ Model for the Unstabilized [2,3] Mittig Rearrangement. . . . Chelation-Controlled, Unstabilized [2,3] Variant - MODEL . . . . . . . . Synthetic Versatility of Chelation- controlled [2,3] Rearrangement Products. . . . . . . . . . . . . Allylic Alcohol Precursor to Chelation- controlled [2,3] Hittig Rearrangement Substrates. . . . . . . . . . . . . . Vinyl Organometallic Additions to Protected Lactaldehydes as a Route to Allylic Alcohols. . . . . . . . . . Route to E- and Z- propenyl-allylic Alcohols. . . . . . . . . . . . . Zn(DH4)2 Reduction of Ynone 26. . . . . Preparation of Tributylstannylmethyl Ether m. 0 O O O O O O I O O O O O O 0 [2,3] Rearrangement of 26 . . . . . vii 2212 39 39 40 42 42 42 44 44 45 46 48 49 52 54 54 54 Figure Figure Figure Figure Figure Figure Figure 20 21 22 23 24 25 26 Alternative Explanation for Z-olefin Selectivity in [2,3] Rearrangement of 26 by Stills’ Model . . . . . . . . . Preparation and [2,3] rearrangement of Tributylstannyl Methyl Ether 31 . . . . The Effect of Rearrangement Substrate Olefin Substitution Upon Product Olefin Diastereoselection in Terms of the Chelate and Non-chelate (Still model) . [2,3] Rearrangement of Allyl Tributylstannyl Methyl Ethers Raving Z- and E- Substitution . . . . . . .-. . . [2,3] Rearrangement of Allyl Tributylstannyl Methyl Ether 34 . . . . Manipulation of [2,3] Rearrangement Product Chirality Through Variation of Precursor Olefin Geometry vs. Variation of Breaking C-O Bond Stereochemistry. . Control of C-0 Bond Stereochemistry in [2,3] Rearrangement Substrates. . . . . viii Page 56 57 59 60 61 63 63 Page lfidkmdns Table 1 Hehjntbosporjm Fungi, Hosts, and TOXin. O O O O O O O O O O O O O O O O O 3 Table 2 13C-NMR Data for EC-Toxins II and III. . 12 Table 3 1R-NMR Data for RC-Toxins II and III . . 16 Table 4 _ NOE-Difference Experiments - RC-Toxins II and III. 0 O O O O O O O O O O O O O O O 21 Table 5 , Dihedral Angles for RC-Toxins I-III. . 25 [2,3] littig W Table 1 Vinyl Additions to Protected « Lactaldehydes. . . . . . . . . . . . . . 50 Table 2 Propynyl Additions to DOM-Protected Lactaldehyde . . . . . . . . . . . . . . 52 ix BDJDIHIIHMEDUEHHGI Host-selective toxins can be defined as those which originate from microorganisms that are restricted to colonization of certain susceptible plant strains and are toxic only to hosts of the microorganism. There are a number of reasons for interest in host- selective toxins. Many important crop plants have been attacked by host-selective, toxin-producing microorganisms, including maize, sugarcane, oats, sorghum, pear, citrus, apple, and tomato; significant crop losses have occurred. In all‘ cases, however, resistant host genotypes exist. Indeed, host-selective ltoxins have been used to screen new hybrids for resistance to fungal attack.1 Host-selective toxins, in contrast to nonspecific toxins, are required by the parent microorganism for successful colonization of the host. This allows study of disease induction at the molecular level. Ultimately, through a much greater understanding of structure-function relationships and the biOchemiCal role played by host-selective toxins, one may be able to design synthetic host-selective toxins applicable as ecologically sound herbicides.1 H. carbonm is a pathogenic fungus, toxic to susceptible genotypes of maize. Examination of culture filtrates of E. carbonm resulted in the isolation2 and structure determination3'5 of l RC-toxin (cyclo(Ala-Ala-Aoe- Pro); Aoe = 2-amino-8-oxo-9,lO-epoxy decanoic acid). RC- toxin illicited the same host-selective toxicity to maize as did H. carbonum. Related species of Helmintbospon’um also elaborate host-specific toxins, though crops other than maize are affected. The Helsintbospors fungi, toxins elaborated and their hosts are presented in Table l. The wide structural diversity displayed by the toxins suggests that the reactive site(s) in plant cells varies from toxin to toxin. Interestingly, other microorganisms elaborate bioactive tetrapeptidic metabolites containing Aoe and an imino acid (Pro or pipicolic acid). These include Cyl-Z, 7, a non- specific plant toxin from Cylindroclsdjm scoparz'um Morgan and chlamydocin, 8, an anti-tumor agent from Dibeterospore cbluydospon's. Despite their structural likeness to RC- toxin, these metabolites display markedly different bioactivity, illustrating that changes in the composition of the cyclic tetrapeptides can have an influence upon their biological activity. The pathways of amino acid biosynthesis are complex, organism-dependent pathways, though certain generalizations may be made. Carbohydrates are the carbon source for amino table 1 W FUNGI,HOSTS,and Toxms FUNGUS HOST TOXIN NAME STRUCTURE’ W maize HC- TOXIN I 1, R1-Mo,R1-H HC- TOXIN ll 5, RnRz- H HC- TOXIN I ll 5.91'M032'0H mum: oats HV - TOXIN 2 NWT maize HDT- TOXIN 3 him sugarcane Hs- TOXIN 4a- a- 5-0-(8- alactofuranosyI)-B- c galacto uranoslde '0 " O N 0 chlamydocin p" 3 HM N" . O (CH ) Emanuele o¢f>j‘*NH‘<>>IhAh Gw4--Aa| .Ns‘--Pn: Hm¢—-Jme Ala—+Pro Aoe—Inn Fri—'1!- Aoe¢—-— Gly Prod—— Gly Glyc—-—Aoe HC JOIN Ill Ah--*-Aos AT-—-“’Tml° Ahr-d-wao .Ms<————Hnwo AhFF———AOI pm¢.___au. toxin I has been sequenced by FAB-CAD-MS,12 GO-MS of derivatized partial hydrosylates,3, 2-D TLC of derivatized partial hydrosylates,5 and 18-NMR N08 difference spectroscopy."1° We chose to employ the latter technique both for its operational simplicity and the greater amount of information it offered. In this approach, irradiation of a nucleus on a given residue may result in a N08 of a neighboring nucleus. If 21 the two nuclei reside on different amino acids, their proximity and, therefore, a partial sequence is established. As applied to HC-toxins II and III, this methodology yielded the data presented in Table 4 from which overlapping partial sequences were obtained. From this data, structures 5 and 6 for HC-toxins II and III, respectively, are unambiguously established. uue4 NOE DIFFEBCE EXPEIIBHS HC-TOXNS I no I H W16) H MED xnoe PARTIAL SEWBJCE —_ Hc-roxm I Gly M1 N5 (I IS” may NOW 537 O 9% Gone Ne M1 Pro a 13% Pro-Na Pro 8 W.“ As. a 3x Aoe-Pro Pro 8 am“ 117: Pro adowiidd An I: . 5x Aoe-Pro m ' upfidd st FC-TOXN I Nfifli biana nx b$mwvum Aoe M-I “I: 0 98 Ala-Au Nhfli “In ux unfit 14W 5m“ A00 0' 4% Aos- t-aoHym I'3‘HYWO 0 awe“ 21X bit-Mn ‘domfleid Au s 13% Aoe-t-a-Hypro 1‘ 3'W ‘ded 1“ 22 0 0 NH CH3 NH HN NH HN NH 0 0 0 “W’s—<41 0 N “W’s—m 0 0 HC-TOXIN l ' HC-TOXIN l I l 5 0 CH3 NH c1137)“ yo HN NH 0 O N (C.‘1z)f,_Gly) observed have caused a change in their toxicity.9 Possible reasons include changes in receptor affinity, membrane permeability, conformation, or various combinations of the three. Due to the current lack of information concerning the mode of action for the HC-toxins, the first two possibilities are difficult to investigate. Conformation, however, is subject to investigation; and, indeed, HC-toxin I has had its solution . 23 (00013) conformation examined.‘»13'1° Figure 9 presents the suggested conformation for HC-toxin I, determined by 18- and lac-nun analyses. The conformation shown contains a his 7 turn, characterized by the presence of two sets of 1-3 figure 9 HC-TOXIN l SOLUTION CONFORMATION (CDCl3) hydrogen-bonded amino acid residues and is further supported by a FTIR1° study which identified hydrogen-bonded and non- hydrogen bonded, NH in a 2:1 ratio. The conformation of a peptide can be described by the appropriate combination of bond dihedral angles, as defined in Figure 10. For HC-toxin I, these angles were calculated from measurement of Jun-In constants and a series of N08 experiments."18 Although presently incomplete, the existing NMR data for HC-toxins II and III may be compared to the data employed in the conformational analysis of HC-toxin I. From a preliminary analysis, we conclude that HC-toxins I, II and III have 24 figure 10 25 rather similar conformations in 0001:. Thus, , angles for HC-toxin I were derived from a Harplus-type2° curve based upon the value of Jun-all. The Jun-an values for corresponding residues in HC-toxins II and III yield virtually identical values of p; Jun-an and 9 for all three HC-toxins are presented"1 in Table 5. As for HC-toxin I, the amide bonds in HC-toxins II and III are transoid, as the observation of NH to sH (or Pr05 to Aoes) N08’s is only consistent with transoid amide bonds. For HC-toxin I, 9 angles were determined from a combination of J's-II and extensive N08 data. Presently, there are insufficient data to calculate 5 angles for Hc-toxins II and III; however, the aforementioned data support a general similarity in the conformation of all three HC-toxins. 5595 9 DIHEDRAL ANGLES FOR HC- TOXINS l-lll HC- TOXIN I l I I I l AMINOACID (Jump?) (MI-I") (JR-«1114) Ala, (10.3. -1100) (995,4 20°) (10.53,-120°) Ana, (9.5, 50°) ( 980. 60°) cry (14.03, 70°) A05 (10.5,-120°) (10.853120') (10.40,-120°) 26 flDfflnHNSlflflilnlflfllL Reagent grade solvents were employed for column and thin-layer chromatography and were used as received. HPLC- quality hexanes were used as received for HPLC separations. Flash chromatography was performed with 230-400 nm silica gel (Merck) by the method of Stillzz; conditions are reported as follows: (column outer diameter in mm; solvent system; grams of silica gel employed; fraction sizes collected). TLC was performed on 0.25 mm silica gel plastic-backed plates (Macherery-Nagel, D-5160) which were developed with 4:3 acetone-methylene chloride. HC-toxins were visualized on TLC plates as blue spots when sprayed with 0.058 bromcresol green in acetone, followed by heating. HPLC was performed on a Varian 5000 instrument, equipped with a UV detector operated at 215 nm. Normal-phase HPLC separations utilized a 0.4 x. 25 cm "batman partisil 10 column eluted isochratically with 96:4 hexane-ethanol at a 4 ml/min flow rate. Reverse-phase separations employed a 0.78 x 30 cm Water u-Bondapac -C18 column eluted with the following gradient: 78 aqueous ethanol to 20% aqueous 27 ethanol over 30 minutes, 20% aqueous ethanol for 15 minutes, followed by a recycle to 7% aqueous ethanol over 20 minutes. 18- and 13C-NMR spectra were recorded in CDC13 at 250 and 68.9 MHz, respectively, on a Bruker WM-250 spectrometer. All chemical shifts are reported in ppm relative to internal CHCla. l3C-NMR spectra were acquired with broadband noise decoupling; COSY23 spectra were obtained as 1K x 512W data matrices using a I/Z-T-I/4-AQ pulse sequence. Data was multiplied by a gaussian function prior to transformation. DEPTZ‘ subspectra were obtained using the following pulse sequence: 1H(I/2-A-I-A~O-A-BB), 13C(----I/2---I—--AQ); editing was performed using the procedure described by Bendall and Pegg.25 NOE experiments employed a 2-second decoupler irradiation, followed by a 1/2 observation pulse and aquisition of the F10. FID’s were multiplied by an exponential function before transformation such that a 2H2- line broadening was introduced. Substraction of a transformed off-resonance control spectrum from the transformed on resonance spectrum produced the NOE difference spectrum from which NOE’s were determined.25 The ratio of the area of the enhanced peak to the area of the irradiated peak (normalized to equal unity) determined the X _NOE. FAB-MS were obtained10 on a Varian Mat CHS-DF double- focusing mass spectrometer. EI/MS were obtained?7 on a Finnigan 4000 mass spectrometer equipped with an INCOS 4201 data system. 28 The initial stages of the isolation and purification of HC-toxins and bioassay were performed by Jack B. Rasmussena, from initial culture of Helmjntbospon’um carbonum (race 1) to Sephadex LH-20 chromatography (refer to Figure 3). Variable quantities of crude HC-toxins were obtained as a function of culture size, age, viability, etc. In a typical isolation, 2-5 ml of the Sephadex eluate was dissolved in methylene chloride (50 m1), leaving behind a small amount of dark-colored insoluble material. The methylene chloride solution was washed with distilled water (3 x 30 ml), brine (1 x 50 ml) and dried (Na2804). Concentration in vacuo yielded an orange-brown gum which tended to foam at reduced pressures. This material (2.24g) was purified by flash chromatography22 (60mm, 2-step gradient: 3L-l:l:l hexane- acetone-methylene chloride, 4L - 4:3 acetone-methylene chloride; 300g; 1 x 3L, 1 x 300 mL, 1 x 700 mL, 1 x 3L). After concentration in vacuo, fraction 1 yielded 1.45 g of a pale yellow foam. Comparison of this material with an authentic sample of HC-toxin 1 (TLC: 1 spot, R! 0.64) showed that it consisted primarily of HC-toxin l; the 1H-NMR spectrum of this material compared favorably with that of HC-toxin I. Fraction 4 yielded 0.350g of an orange-brown oil which illcited host-specific toxicity (TLC: streak R: 0.55-0.20). This material was purified by flash chromatography (40 mm, 8:12:3 acetone-methylene chloride— ethanol, 100g, 50 ml initial fraction followed by 20 ml 29 fractions); fractions 8 and 9 were combined and concentrated in vacuo yielding 181 mg of a yellow oil which was active by bioassay (TLC: 2 poorly resolved spots, Rf 0.45-0.25). The oil was then dissolved in 820 and passed through a Waters C19 Sep-Pak flushed with 40* aqueous ethanol. Reverse-phase HPLC of 35.0 mg of this material separated it into 3 major components, retention times: 25, 35 and 47 minutes, designated RP-l, RP-Z and RP-3, respectively; 11.0 mg of RP-l was obtained. Normal-phase HPLC of RP-l (11.0 mg) separated it into two components, with retention times of 25 and 33 minutes, which displayed host-specific toxicity; these were designated HC-toxin II (4.0 mg) and HC- toxin III (2.4 mg) in order of the elution. HC-toxin II. cyelo[Alg:6117(Z;§gino-§-oxo-9.10-epoxy- decanoy1)-ProlL;§ ‘ EI-MS (70eV): 422 (M‘, 5.0%), 170 (C981602N’, 76.6%), 70 (C4HaN’, base). FAB-MS (glycerol): 423 (M+l', 48X), 226 (M-196, 10%), 170 (ConeOzN’, 13X), 70 (C4HeN’, base). High Resolution FAB-MS (CsI,g1ycerol) for C20H31N400, calculated 423.22436, found 423.22558; +2.88 ppm. 1H- and 13C-NMR Tables 3 and 2. HC-toxin ' III, cyclolA1a-A1a-(2-amino-8-oxg;§.lO-epoxy- decanoyll-trggg-3-Hyprol..§_ EI-MS (70eV): 452 (M*, 3.93), 170 (CoH1602N*, base), 86 (C4830N’, 55.48). FAB-MS (glycerol): 453 (M+l*, 19X), 156 30 (M-196, 12%), 170 (C9H1602N’, 30X), 86 (CqHaON’, base). High Resolution FAB-MS (CsI, glycerol) for 021H33N402, calculated 453.23493, found 453.23221; —6.0 ppm. 1H- and 13C-NMR Tables 3 and 2. 10. ll. 12. 13. 31 “11.1.1810!" Scheffer, R. P.; Livingston, R. S. Science 190, 22% 17. Pringle, R. 0. Plant Physio]. mo, 46‘, 45. Walton, J. 0.; Earle, B. ,0.; Gibson, 0. W. Biochem. Iquvl.ka.¢klm.IMB2,.RWL 785. Walton, J. 0.; Rich, 0. R.; Kawai, M. Dim. Biophys. lb. 00-. 1-, 111, 398. Pope, M. R.; Ciuffetti, L. M.; Knoche, R. W.; McCrery, 0.; 0aly, J. M.; Dunkle, L. 0. Biochem. m, a, 3502. Liesch, J. M.; Sweeley, C. C.; Staffeld, G. 0.; Anderson, M. 8.; Weber, 0. J.; Scheffer, R. P. IbhnflkmkwalSBZ, 3% l” Ciuffetti, L. M.; Pope, M. R.; Dunkle, L. 0.; 0aly, J. M.; Knoche, R. W. BIocsbem. 1003, 22, 3507. Bioassay protocol: Scheffer, R. P.; Rasmussen, J. 0.; Tanis, S. P.; Rorenstein, B. A. finial. Plant Patb., submitted for publication. Biological characterization is reported in Reference 8. Ackermann, 0.; Watson, J. T., M.S.U. Regional Mass Spectroscopy Facility, Department of Biochemistry, Michigan State University, unpublished results. CAD-FAB-MS of cyclic tetrapeptides: Tomer, I. 0.; Crow, P. W.; Cross, M. L.; Kopple, K. 0. Anal. 0M. 1000, 56‘, 880. Gross, M. L.; McCrery, 0.; Crow, P.; Tomer, K. 0.; Pope, M. R.; Ciuffetti, L. M.; lnoche, R. W.; 0aly, J. M.; Dunkle, L. 0. mtrabedron Lett. 1032, 23, 5381. Analysis performed by Professor 0. T. A. Lamport and Mr. P. B. Muldoon III, MSU—DOB Plant Research Laboratory, Michigan State University, East Lansing, Michigan 48824. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 20. 27. 28. 29. 32 We thank Dr. Ashok Shenji, E. I. DuPont de Wemours and Company (Inc.), Central Research and Development Department, DuPont Experimental Station, Wilmington, Delaware 19898 for authentic samples of cis- and trans- 3-hydroxyproline. Assignments were aided by a comparison of 13C-NMR data for l, 5 and 6; and DEPT experiments. For example, see: Graden, 0. W.; Lynn, 0. G. J. a. Chem. Soc. m, 106, 1119; and references cited therein. The structure of 5 has recently been reported by another group: Knoche, E. W.; Kim, S.-0.; Dunkle, L. 0.; McCrery, 0. A.; Tomer, E. E. hum Lett. 1”, 26, 969. Mascagni, P.; Pope, M. Gibbons, W. A.; Ciuffetti, L. M.; Enoche, R. W. Diodes. £10m. Res. Cm. 1003, 113, 10. Pope, M.; Mascagni, P.; Gibbons, W. A.; Ciuffetti, L. M.; Knoche, E. W. J. Am. aha. Soc. m, 106, 3863. Bystrov, V. 8.; ”Progress in WMR Spectroscopy", 1976, 10, 41; Pergamon Press, Great Britain. Values for 1 are those reported in Reference 4. Still, w. c.; Kahn, M.; Mitra, A. J. J. om. Chem. 1978, 43, 2923. The standard Bruker COSY.AU program was employed. Bruker DEPT.AU was employed. Bendall, M. R.; Pegg, 0. T. J. m. Res. um, 53, 272. For an example, see: Jones, C. R.; Sikakana, C. T.; Rehir, S.; Euo, M.-C.; Gibbons, W. A. Bionbrl. J. 1978, 24, 815. Analysis performed by Mr. Richard Olsen, Department of Chemistry, Michigan State University. Rirota, A.; Suzuki, A.; Suzuki, R.; Tamura, S. Agr. 5’01. C“. m, 37, 6430 Closse, A.; Rugenin, R. Heir. Chim. Acta um, 57, 60. 33 [2,3] MC. The macrolide and polyether classes of natural products often display potent antibiotic activity. This, and the structural complexity of these highly oxygenated bioactive molecules makes them of tremendous interest to synthetic organic chemists. A few examples of the structural diversity these molecules display are shown in Figure l. flours 1 BAFILO MYCIN 1, BREVITOJCNB 1 figure 2 as on 9“ ”WWW“ E a o\ ACYCLIC ANALOG OF BAFILOMYCIN 34 Some structural simplification can be achieved by opening the cyclic array and considering the assemblage of acyclic stereocenters thus produced as the ultimate synthetic target. Consider, for example, the acyclic precursor of bafilomycin (Figure 2). The C14‘C19 segment of 1 contains five consecutive centers of asymmetry, with both alkyl and oxygen substituents in a variety of stereochemical relationships. Clearly, any efficient synthesis of this molecule must incorporate convergent and flexible techniques of acyclic stereoselection. For an approach to the synthesis of arrays of acyclic stereocenters to be considered general, the relative and absolute stereochemistry in the target must be accessible in a straightforward manner.*'-° Of paramount importance to the rational control of acyclic stereochemistry in the relative or absolute sense is an understanding of the reaction process and various steric and electronic factors responsible for stereo-discrimination in the transition state.2‘5 With such a model in hand, the synthetic chemist has often been able to rationalize and/or predict the outcome of various functional group transformations and carbon-carbon, bond-forming reactions. One of the most successful techniques of acyclic stereoselection is the "modern” aldol condensation. Exemplifying this is the work of Evans°, Mukaiyama’, and Reathcock‘, whose aldol chemistry has achieved high levels 35 of diastereo- and enantioselection (Figure 3). The success of the aldol reaction depends upon two factors: first, enolates of high geometric purity must be obtained and, second, the homogeneous enolate should produce a preponderance of one aldol diastereomer in the condensation. Since it is generally accepted that the majority of ”standard” aldol condensations proceed via the Zimmerman- Traxler‘ chair-like transition state, in which substituents adopt orientations which result in minimum steric interaction, one can generally predict the outcome of the figure 3 O OH 1) 958011 3 arr-C.H.CHO _ W (\ JK/ ”momma , mo o 1uflee o . 1) 80,001: 0 O" . awaowmo- M00 smut; :8“ /, 1)uuvnrfl _ 2) i'caH'FHO * t-Bu 97:3 «moojrso o 0 OH \\v/fl\\// ”&“”‘a““m' ammo i P" . 96:4 mom 36 process. The absolute configuration in the aldol process has also been controlled, gig chiral enolates, and the course of the reaction with chiral aldehydes is relatively well understood. One drawback to the aldol chemistry thus far reported is difficulty in routinely producing the threo- or anti-stereochemistry with the same levels of stereoselection observed for the erythro-syn manifold. Another technique for the. control of relative and absolute acyclic stereochemistry focuses on the development of a carbon-carbon double bond with proximal resident chirality (Figure 4). In this approach, a fragment nguro J ”W I)” OH "I” : “CHO 4- P408 O (30.51 3%. [Mcflzocfltph —-- a : m. EIZ- 1 ‘1 ' and MONENsm I Ho 5"; mom 0, . 0 ages, 2:,0 *31N_F§—_.$O. . .2 O Mmmwn b mason... -< HO O __..CH. ”0) ZNIHQOMs 5 0H H6 40% OLIVO$E OH 0...... WV mums“ 0 R0 ——-—muj/'\r\/ \ OH mum w R-HyHNIJGOmnduhunpuua 37 containing the resident chirality is typically coupled with an achiral ylid; the resulting unsaturated center may then be operated upon as demonstrated in Figure 4. Here, success is contingent on obtaining olefins of high geometric purity, a condition which may not always be realized. Sigmatropic rearrangements are also attractive routes to highly functionalized molecules as these intramolecular processes are often highly stereospecific. A number of variants of the [3,3] sigmatropic rearrangement have met with considerable success in this regard. Figure 5 provides examples of some Claisen and Cope [3,3] rearrangements. Note that in these reactions, the starting asymmetric center is lost, but one or even two centers and double bonds of known configuration are obtained with high transmission of chirality. This self-immolitive process is generally thought to proceed through a six-membered, chair-like transition state10 wherein substituents assume the most stable orientation. The [2,3] sigmatropic rearrangement has also been employed in natural products synthesis. A number of variants are shown in Figure 6, including the allyl sulfoxide-sulfenate rearrangement, allylic amine oxide rearrangement, the- ”Buchi”11 rearrangement and the extensively employed Wittig rearrangement. The stereoselectivity observed in these rearrangements is thought to derive via a cyclic 5—center transition statelz, 38 figure 5 [J.;] names CC rfrn 4 a ____. CfiCrm m" O . on 3" °)\\ _, "lb ‘ MN , .uvfism 1) LDATFF. ‘7. N CO‘H o \Jk CO. ““- \\/¢; :Lomsa A I/ . L’ nuns,“ o o, co,» \Jko nwamuwm. «- c "' t 00:14. 7310 m a 000M930. 4 a: mo -—m-————-. mu: co,cu-, OM. PO 5 o i (H. 110 l = 0 one m m W 50$. ”:4 figure 6 \-° 1. ‘ ONW A—"‘" o’ \” vk'H \ CH3 W 2 1)LDA.Msl.-50' / "(SW 2)(Mo0),P. mow ‘ W OH W \_/O:J\ : 2. dmsthyl 0MP m‘ (CH,)1N E ' - mon- A fl / O W figure 73 OH AA [23' /><‘\Ph ”A0 —""" 5 £1 dssodsfion \ 1mm... ”A0 + ,s‘t‘fl. association 1 recombination OH [2.31 i PnAo ...____.. \ P“ 40 analogous to that of [3,3] rearrangements. Early mechanistic studies of the [2,3] Wittig rearrangement examined the competition of the concerted vs. non-concerted pathways in the rearrangement. Rautenstrauch13 examined 7,7-dimethyl allyl benzyl ether 4 in a crossover experiment, shown in Figure 7a. Upon lithiation, he observed an 8:1 mixture of 5 and 6 at -80°C, whereas at 23°C at 6:1 mixture of 5 and 6 was observed. Additionally, rearrangement of 7 s,sbdimethyl allyl benzyl ether at -25° afforded a 1:1.4 mixture of 5 and 6. Two conclusions may be drawn from these experiments: the concerted pathway appears to dominate over the non-concerted at lower temperatures, and the nature of the migrating allyl fragment has some influence on the pathway chosen. BaldwinH performed a similar experiment, outlined in Figure 7b. Rearrangement of deuterium labeled 8 at 0°C flmxo7b l\ . P“ OHHD 0 o n-BuU 000 " 14¢ sxoec-z a “.0 2 2 \\ -~H ‘ ' Pb L\ ”' OHH D O n-BuU ol’K ~DXDOC-2 w s” H 41 resulted in the production of 9 as a mixture of cis- and trans-isomers. 1R-NMR indicated a vinylic deuterium content of 14158. At -80°C, the vinylic deuterium count fell to nearly 03, indicating the predominance of the [2,3] pathway at low temperature, in agreement with the findings of Rautenstrauch. In another variant of the Wittig rearrangement, Still reported15 that the rearrangement of allyl stannyl methyl ethers derived from s and p bicyclo dec-5—ene-7-ol formed [1,2] (non-concerted) and [2,3] Wittig products in low yield. This contrasts with the rearrangement of 3- cyclohexenyl stannyl methyl ether (Figure 7c) which provided the product shown in sax yield. Still suggests that steric hindrance at the migration terminus (p-carbon) and perhaps conformational problems can promote the non-concerted pathway, relative to the concerted [2,3] process. There are two general classifications one may impose. upon the Wittig [2,3] rearrangement. The first consists of [2,3] substrates which contain an anion-stabilizing group; Midland1° and, in particular, Wakail'“18 have examined this variant extensively (Figure 8). The second type of rearrangement substrate generates an unstabilized carbanion, as exemplified by the work of Still15 (Figure 9). Inspection of the rearrangement products of the anion stabilized substrates listed in Figure 8 establishes that these rearrangements typically proceed with good to 42 figure 7c (2.3] . OH n-BuLi. -7a \ .Q OCFhSnBu, (oz/B Isomors) [1.21 [uwmwwm OH mime. low yield "“““"'= C ......... OCH,Sn9u, ‘ on 93: we figure 0 1 ,, 1",...“ YU ..... W Wfiw Ind-mat: s1 suds-m 10 on Z-dsh 0% 26:: cums W mm (:1) MM =/—° "9 "-. '6’ OH OH moms WWO m wow-nuance figure 9 I m M m 31 .0 “'5 s 95 fl 2 5.“ “Old 43 selectivity for the E-olefin and can often achieve diastereoselectivities of better than 10:1. There is, however, a drawback to the stabilized Wittig, that is, one must carry along the anion-stabilizing function regardless of its appropriateness to the target structure. Therefore, additional steps may be required to convert this moiety to the desired functionality. In this regard, the unstabilized variant might be the method of choice; as after rearrangement, one is left with a rather versatile primary alcohol for subsequent elaboration. In contrast to the stabilized rearrangement, the unstabilized Wittig rearrangement often yields predominantly Z-olefin. This selectivity is a function of olefin substitution, as the Z- selectivity drOps to 60:4015 when the vinyl carbon adjacent to the breaking C-O bond is unsubstituted (Figure 10). To rationalize these data, Still has suggested the model depicted in Figure 11. Still assumes an early transition state for this exothermic reaction, with a dihedral angle of ca. 30° between the breaking C-O bond and the vinyl moiety. The olefin facial-selectivity, and, thus. the ratio of E- to Z-product alkenes, is determined by A1.2 interactions. In the example depicted in Figure 11, the interactions between Me and 'Bu result in rearrangement through the alternative rotamer, producing predominantly the Z-olefin. When the vinyl substituent is removed, the olefin facial-selectivity drops dramatically, resulting in low Z/E ratios (60:40). The requirement of an alkyl substituent at 0-2 of the 44 figure 1 0 (”MW OH -—" It! - 00:10 omgwmm OH C7H1s fmm911 ——. m o 4. a;\ at; re... AArOH ‘ . / We" substrate allylic alcohol limits the utility of the method; however, if one could routinely select the face of the olefin which is attacked, then the unstabilized Wittig rearrange-ant would become an attractive technique for acyclic stereoselection. One possible method for directing olefin facial- selectivity is by organization of the transition state through chelation1° (Figure 12). If we assume that the 45 figure 12 olefin 03:0 to enolate olefin ondo to Chelate o P\° Ll,,”/ J ' I“2 n3 a, [2.3] [2.3] v J PO 92 intermediate alkoxy carbanion is internally coordinated30, providing a six-membered chair-like chelate, than the sense of asymmetry at C-7 and the resulting product olefin geometry will be dictated by the face of the double bond which is attacked. The placement of the vinyl group exo to the chelated ring system, leading to an B-olefin, is appealing for steric reasons; however, stereoelectronic considerations should render such a conformation unproductive. Alternatively, placement of the double bond in the sterically more-congested position, under the ring 46 plane, leading to a. Z-olefin, is attractive as the bond angles and overlap closely approximate those assumed in the transition state proposed by Still. The substrates obtained upon rearrangement should be highly versatile synthetic intermediates, as illustrated in Figure 13. Many regio- and stereochemical patterns of substitution may be produced from such a substrate using a variety of stereoselective methods. These acyclic pieces are then new. 13 Ungmgo Ru “8 NR. _3_UL_‘__ War. Po 31 '5: PO PO W1 law an M0" MON '0 P0 on Hfihhflv W0" L. W0" cum YYVOH P0 P0 . 3 P0 0H 10.0. n c l 47 ideally suited for rapid assembly into desired target molecules; for example, the primary hydroxyl group obtained from the rearrangement can serve as an electrophile in an aldol reaction, or could serve as the hydroxyl terminus in a macrolactonization. 48 [2,3] “SUITS m DISCIBSIM To examine the chelation-controlled [2,3] Wittig rearrangement, allylic alcohols of the type shown in Figure 14 are needed. Based on the high diastereomeric ratios 119qu OH R] NR2 PO . R3 p-mnwomm a... @cuzocuz- n- @411,- noted21 in the literature for chelation controlled organometallic additions to abalkoxy carbonyl compounds, a possible route to the desired allylic alcohols involved vinyl organometallic additions to readily available protected lactaldehydes, as depicted in Figure 15. The two-step transformation of protected ethyl lactates to the desired protected lactaldehydes is shown in Equation 1. With a variety of protected lactaldehydes available, we next investigated the organometallic additions. figure 15 O OH OH \(‘k /\H W + : / H _____ YV P0 P0 P0 dsdsnonpoma (a) My“ (b) 10 P-BOM 1 3 {:3 P-BOM 1‘, ME!“ 14 a12- ME“ 1394” 15ggpqh equauon 1 0 0 LM Sm OEt —"’ You _"H u H P0 P0 P0 13, P-BOM ‘1 P-BOM Lo, P-BOM ll. P-MEM 2_o_, P-MEM 3.1.: MEN ‘3. Mn 2.1. P-Bn 13 9'5" To the best of our knowledge, the only report of highly diastereoselective additions of vinyl organometallic reagents to e—alkoxy aldehydes which has appeared in the literature is that of MacDonald, et. a1.22 Therefore, we surveyed a variety of reaction conditions to ascertain the influence of solvent, protecting group and metals present and/or added on product diastereoselectivity. A typical experiment consisted of side-by-side runs which examined a given solvent/protected aldehyde combination, with or without zinc halide added to vinyl magnesium bromide. The reactions were performed by adding the aldehyde solution to either vinyl magnesium bromide solution (control) or to a solution (2:1 mol) of the vinyl 50 Grignard and the desired zinc halide at -78°C, followed by warming to room temperature. Table 1 presents the results of these studies. Based on the general observation‘ that Grignard reagents react with abalkoxy aldehydes to favor the umo1 O on 9H Y“ zon- W * W PO p0 P0 cnehuon (a) non-chelation ( b) 1am, azb P solvent ZnCI,(+I—) y1010(crude) ‘H—NMR 13110 GO ratio Bonzyl THF - 77 x 1.6 :1 1.6 :1 BenzyI THF + 97 x 6.0 :1 6.8 :1 MEM THF - 79 x 1.8: 1 - MEM THF + 53 x 7.5: 1 - 80M 71-11= - 7a Rpm“) 1.8 :1 - 80M THF + 83 x 9.3 :1 22 :1 80M 51,0 + 88 X 3.4 :1 4.1 :1 80M 011,0, - 86 x 1.3 :1 - 80M cmcn, + 91 x 8.4 : 1 3.4 : 1 80M c1-1,c1, ZnBrz 79 x - 12.7 :1 product of chelation control23 as the major diastereomer, we assigned product stereochemistry as chelation or non- chelation by comparison of the product distribution in the presence 'of zinc halide with that obtained from the corresponding Grignard reaction without zinc halide. It is noteworthy that in all cases the addition of zinc halide increased the disstereoselectivity of the reaction relative_ 51 to the control. A maximum diastereoselectivity of 2 9.3:1 was observed for the BOM/THF/ZnClz system. This selectivity is appropriate for synthetic use; however, it is still far below the 49:1 obtained by MacDonald?2 in vinyl copper additions to a,p-dia1koxya1dehydes. With diastereoselective additions of simple vinyl organometallics established, it was of interest to see if the same strategy might be extended to vinyl homologs such as cis- or trans-propenyl organometallics. These reagents, if successfully added, would yield valuable substrates for examination of the [2,3] rearrangement sequence. By simply reversing the geometry of the organometallic reagent, the sense of asymmetry at the methyl-bearing center off the rearranged product could be changed. However, it is difficult to obtain configurationally pure E- or Z- propenyl organometallic species. To circumvent this problem, the plan shown in Figure 16 was developed. Given the availability of E- or Z-propenyl allylic alcohols from precursor propargylic alcohols, we envisioned preparing 22 and 23 by a chelation-controlled addition of a propynyl organometallic reagent to a protected lactaldehyde, followed by a partial hydrogenation. or a hydride reduction. This protocol was employed by Hart24 to prepare [3,3] sigmatropic rearrangement substrates and by Midland?s to prepare {2,3} rearrangement substrates. Table 2 lists the data from three experiments. The ratios observed for the two Zn mediated reactions are relatively low but demonstrate a reversal of 52 figure 16 0 9H 011 +11 ‘ + 80" 30" non 2E. .zm OH OH OH OH L 2" Newman 80" SUN muoz OH 9H Y‘o “$.31 \’\ + w 3°"° BOND CH1 BDMO (fig 25!: m ImnwwwwMzmm yum!) £52. 1:0 . ‘1 1:1.7 1:1 03 1.05:1 2:1 as 1.1 :1 radosrsweswfimsvwsgedGCmdNMflm 53 the diastereoselectivity observed for the propynyl lithium addition. This seems reasonable after consideration of the occasionally observed preference23 for organolithium reagents to give non-chelation products and on the preference for zinc organometallics to give chelation addition products. Support for these stereochemical assignments is drawn from the Zn(BHe)2 reduction of ynone 25, Figure 17, which yields alcohol film as the major product. The indicated stereochemistry of alcohol 24mis predicted by the work of Oishi.3°'a° Alcohol Zlm is identical (00,13 NMR) to the maJor isomer obtained in the propynyl lithium addition, supporting the stereochemistry predicted by the non-chelation model. These data indicate that the stereochemistry of the diastereomeric product from the zinc-mediated addition is that shown in Table 2, in accord with a chelation-controlled addition. As a result of the relatively low diastereoselectivity observed, this route to rearrangement substrates was not pursued further.37 The allylic alcohols obtained from the BOM-lactaldehyde vinyl additions (ZnCla, TBF) were then converted to the corresponding methyl stannyl ethers in preparation for the [2,3] rearrangement (Figure 18). Alkylation of the alkoxides 13m,b (KB, TBF, 18-0-6) with ICBzSnBus resulted in a 623 isolated yield of &,b. Rearrangement of 28a.b (Figure 19) was realized after treatment with two equivalents of n-BuLi (hexane, -78°0, 1 hr.) followed by warming to room temperature and quenching with saturated 54 figure 18 OH 105311311, .KH_ 'nrnoce ' BOHO flours 19 OCH28nBus zoqsu1 \\ ~75- hex-1e 21D figure 17 O OH 21113144): 25 29.1 OCHzSnBu3 80110 OH 30110 1 23.5 21. OH OH 25.1! 55 NHcCl (aq.). After workup and silica gel chromatography, a 57% yield of diastereomeric homoallylic alcohols 27a and 215 was obtained in 3.5:1 ratio (IR-NMR). The major isomer was assigned as z- on the basis of a 1B-NMR shift reagent study (Bu(FOD)9-d27) in which measurement of the vicinal J values for olefinic hydrogens established 3J coupling of 10.6Hs for the major isomer (Z-) and 16.3Bz for the minor (B-) isomer. Confirmation of Z-Z'Im as the major isomer was obtained by synthesis of this material by an alternate and unequivocable route,2° shown in Equation 2. The 3.5:1 Z/E ratio observed ammmnz OH //' cmjm 1MHumvan_ ‘// ‘_——“’ I has snide some 2 M 30110 30110 23 2h OH in this rearrangement is significant since the 2/8 selectivity in the absence of an alkyl substituent on the internal vinylic position of the stannyl methyl ether should have been about 60:40 according to Still’s model (refer to Figure 10). Although our observation of enhanced z-selectivity is consistent with the chelation model, one might argue that the branching adjacent to the breaking c—o bond (Figure 20) might contribute to the observed Z-selectivity; this factor 56 figure 20 OCH23nBu, Po BmJ P0 P0 V o ‘::::- ‘fi\ 0 1 " / We” WVOH was not present in Stills’ substrates. To test this possibility, a substrate having the desired branching yet incapable of chelation, was needed. Allylic alcohol 30 obtained by addition of vinyl magnesium bromide to isobutyraldehyde was selected for this purpose. Conversion of allylic alcohol 30 to. its methyl stannyl ether (ICHzSnBua, KB, l8-C-6, THF), 31, (87%), and subsequent rearrangement (hexane, 2 equivalents nBuLi, -78°C, 2 hrs.) gave a 2.1:1 ratio of z- and E-homoallylic alcohols (compound 32, Figure 21). This ratio is slightly higher than that predicted by the Still model, indicating that 57 nguro 21 0101,10-C-6 \\ semen, OH OCHzSnBu3 :19 31 2——"N rm W011: “MOM -78, hex-w 2.2 branching of the alkyl group adjacent to the breaking C-O bond has an effect, albeit small, upon rearrangement stereoselectivity.25 Consequently, it appears that the primary factor responsible for the higher Z-selectivity in this work is, not the branching factor, but the presence of an alkoxy function, capable of chelation. Chelation control, as an important factor in [2:3] rearrangement stereoselectivity, can be further examined by the rearrangement of substrates having a variety of alkoxy groups such as benzyl, MEM and methyl. The varying ability of these functions to coordinate to the Li counter cation would be expected to effect the degree of diastereoselection. lAdditionally, the rearrangement could be performed in a variety of solvents and/or exogenous lithium chelating agents could be added. In an attempt to 58 define solvent effects on the stereochemical outcome of the rearrangement, 26m was treated with 2 equivalents of n-BuLi (-78°--)0) in methylene chloride and TBF. In methylene chloride, a nearly quantitative recovery of starting material was realized; suggesting that transmetallation does not readily occur under standard conditions in methylene chloride. Bowever, rearrangement in TBF (2.1 sq. nBuLi; - 78°-->RT; 823) provided a 1.5:1 ratio of alcohols 27a and 27m. The lower diastereoselectivity observed for the rearrangement of 28a in THF vs. rearrangement in hexane is consistent with a chelation model since, in the highly coordinating THF solvent, it is possible that THF-Li’ association could reduce the efficiency of chelation, leading to lower olefin diastereofacial selection. In hexane, the lack of solvent -Li* coordination could facilitate formation of the chelated transition state for rearrangement. Temperature studies should also be conducted in order to determine whether the selectivity is enhanced at low temperature. Olefin substitution is another means by which rearrangement diastereoselectivity may be modified. The effect of alkyl substitution at the internal position of the olefin has been examined by Still in a non-chelated variant of the [2,3] rearrangement. The same steric factors are expected to apply in the rearrangement of a suitably substituted chelatable substrate, as shown in Figure 22. In such a case, steric interactions would dictate that the 59 figure 22 P chelate model P l l ,o ,0 LI' Ll' ' *-_—.:—...— ; ’fi _$\\Cl -—K\WD ti [2.3] ‘ x I! OH \ p0 P0 H OH lam PC) C) non-chelate( Still) model olefin swing under the ring to minimize unfavorable A1.2 interactions, with rearrangement leading to a predominance of Z-olefin. Note that, whether chelation occurs or not, the same major product is expected. Substitution of the olefin terminus in the Z-position should lead to chelate A in Figure 23. However, it is expected that the olefinic 60 moiety would be strongly discouraged from residing under the ring due to severe steric interactions. This would inhibit rearrangement from the chelate, perhaps leading to rearrangement m a non-chelated intermediate in which we expect low diastereoselectivity.:3° E-33 (Figure 23) should rearrange via chelate B since the alkyl moiety at the olefin 119m 23 OCHISnBu, R P0 1 3111.1 Mea L150 PO 8 PO \.,0 —-—-—--' o “—-——-—-—-- o R \ \ 1 1m ml [2.3] R -— OH PO R PO 1 l P OCH230803 Ll \ W -——~B““ W . —'-'—+ WW P0 R P0 R ‘“ chemo B 61 terminus resides in a favorable steric environment, away from the chelated-ring. We expect that the product diastereoselectivity should be comparable to that observed for the rearrangement of substrate 28a. Alternatively, the stereochemistry of the breaking C-O bond could be inverted which would allow rearrangement of substrates having one less axial substituent (C33) in the transition state (Figure 24). This remains to be examined experimentally. figure 24 gowns». W a PO lit-NU 62 The rearrangement of substrates having substitution at the olefin terminus (Figure 25, R£#82) results in the transmission of chirality from the resident C—O bond to the new c-c bond. The sense of chirality installed at this center is dependent on two variables, the stereochemistry of the breaking C-O bond and the olefin geometry. The result of altering one of these variables while holding the other constant is presented in Figure 25. Therefore, one can manipulate the sense of asymmetry at the newly formed C-C bond either by variation of olefin geometry or 0-0 bond stereochemistry. We have suggested that manipulation of asymmetry, based upon reversal of the geometry of a vinyl organometallic, is impractical on steric grounds and from a consideration of organometallic precursor availability. However, a manipulation of the C-0 stereochemistry for the control of chirality in the rearrangement appears to‘ be a superior approach25 (Figure 26). The precursor to substrate 2% has been prepared in an unoptimized diastereomeric ratio of 7:1 as shown in Bountion 3. Treatment of 84 with ICH28nBu3 should yield an». rearrangement of which will provide insight as to the effect of alternate C-O stereochemistry upon product olefin diastereoselectivity. . This chelation variant of the [2,3] Wittig rearrangment is clearly effective for the rearrangement of substrates lacking alkyl substitution at the internal position of the 63 figure 25 OCH; 3. OCH; R; E [2 3| R J—L W0" We. a P0 P0 OCH, R. 2.3] [23] 9°": R: «L—. '5. OH ¢—- 2 3: po R. R: R‘ PO PO figure 26 O \ 1) R—-—|-1 O Y\ 2) unmom : WI! PO 3) [0) \‘).:/¢\/s‘H mm M\ OH WR WN: Po PO 64 «wmwn3 OH 11:: ' / OH o BOMO M" ”w" = 21L”. bzs .- 7:1 BOMO BOMO 0" mu: 35 m BOMO olefin. The effects of chelation, olefin substitution pattern, and C-0 stereochemistry need to be further examined in order to define the scope and utility of this technique. 65 UL3]IMUIBDIUHIL All reactions were conducted under Ar, with rigorous exclusion of moisture from glassware and reagents unless otherwise noted. Hexane, ether, and THE were dried by distillation from Na/benzophenone ketyl. Methylene chloride, triethylamine, and diisopropyl ethylamine were dried by distillation from calcium hydride. Dimethyl sulfoxide was dried by distillation'at reduced pressure from P205. Zinc chloride was dried by fusion under Ar, then was crushed and stored in a dry box. Zinc bromide was dried in an abderhalen (xylene) for 48 hours using P205 dessicant, then stored in a dry box. Ethyl-L—(+)-lactate (Aldrich,. 983) was used as received. n-Butyl lithium was purchased as a solution in hexane from Aldrich Chemical 00., Milwaukee, WI and was titrated by the method31 of Watson and Eastham. Vinyl magnesium bromide stock solution was prepared from 20 mesh Mg° and vinyl bromide in THF; the Grignard solution was titrated by the method of Watson and Eastham. Vinyl addition product diastereomeric ratios were measured by Capillary GC (carbowax column) or 250 MHz 1H-NMR. 66 1H-NMR spectra were recorded at 60, 80, or 250 MHz, as indicated, on Varian T-60, Varian FT-80 or Bruker WM-250 spectrometers, respectively. Broadband decoupled 13C-NMH spectra were recorded at 68.9 MHz on a Bruker WM-250 spectrometer. Spectra were recorded in 00013 unless noted otherwise. Chemical shifts are reported in ppm relative to a Meqsi internal standard unless noted otherwise. Coupling constants are reported in Hz; data are reported as follows: chemical shift (multiplicity, coupling constant, integration), multiplicities are defined as s=sing1et, d=doublet, t=trip1et, q=quartet, m=mu1tip1et, br=broad. IR spectra were recorded on a Perkin Elmer Model) 599 Spectrometer. Mass spectra were obtained on a Finnigan 4000 mass spectrometer equipped with an INCOS 4201 data system. Spectra were obtained in either electron impact (HI) or chemical ionization (CI) mode; for CI, methane was employed as ionizing gas. Gas chromatography was performed on a Hewlett-Packard 5880 Level 3 capillary gas chromatograph equipped with an FID detector. Methyl silicone (12.5m, He carrier) or carbowax (50m, Hz carrier) columns were employed. Ethyl-Z-(bengyloxygethyl)-(§)-lactate, 16 To 54.6g (0.35 mol) of benzyloxymethyl chloride37 in 200 m1 of methylene chloride, cooled to 0°C in an ice-water bath, was added 61.0 ml (0.35 mol) of N-ethyl diisopropyl 67 amine over a five-minute period. After stirring for 5 minutes, 28.3 ml (0.25 mol) of ethyl-(s)-lactate was added to the reaction vessel over ten minutes. The reaction mixture was warmed to room temperature and stirred for 12 hours. The resulting yellow solution was cast into CHzclz (100 ml), washed with 0.1 N HCl, saturated aq. NaHCOa, brine (100 ml each), and dried (NazSOc). Concentration in teams provided 49.06g of a yellow oil, which, after distillation, yielded 30.16g (51%) of 16, BPo.s 113-117°C. 1H-NMR (60 MHz): 6 = 1.1-1.5 (m, 6H), 3.95-4.4 (m, 3H), 4.62 (s, 2H), 4.82 (s, 2H), 7.34 (br s, 5H). CI-MS: 239 (M’l, IX), 131 (M-107, 86%), 91 (C7H7’, base). IR (neat): 3065 (m), 2985 (s), 1740 (s), 1210 (s), 1175 (s), 1115 (s) cm‘l. Ethyl-(gs):(mgthoxyethoxyggthy1)-1actate, 17 According to the procedure for synthesis of 16, treatment of 35.25g (0.3 mol) of ethyl-(s)-lactate, 65 ml (0.373 mol) of N-ethyl-diisopropyl amine and 46.50g (0.373 mol) of methoxyethoxymethyl chloride32 provided, after distillation, 34.78g (56*) of 17, BPo.ss .79-90°C. 1H-NMR (60 MHz): 6 = 1.3 (t, J=7Hz, 3H), 1.43 (d, J=6Hz, 3H), 3.4-3.85 (m, 4H), 4.20 (q, J=7Hz, 2H), 4.24 (q, J=6Hz, 1H), 4.8 (s, 2H). CI-MS: 207 (M+l, 0.3%), 131 (M-75, 71%), 59 (M-l47, base). IR (neat): 2990 (s), 2940 (s), 2890 (s), 1745 (s), 1180 (s), 1030 (s), 1100 (s) cm'l. » a a“. 68 gZS)-§methoxyethoxyggthyl)-propanediol,‘gg To a suspension of 2.02g (0.053 mol) of lithium aluminum hydride in 500 m1 of anhydrous ether was added 17.56g (0.085 mol) of 17 in 50 ml of ether at such a rate so as to maintain a gentle reflux. After the addition was complete, the mixture was heated under reflux for 7 hours. The reaction mixture was then cooled in an ice bath, and with rapid mechanical stirring, was cautiously quenched with 20% aqueous sodium hydroxide (ca. 80 ml). The resulting heterogeneous solution was filtered through a pad of celite; the filter pad was rinsed with ether. The biphase filtrate was transfered to a separatory funnel and the aqueous layer drawn off. The organic layer was washed with H20 (2 x 200 ml), brine (100 m1) and dried (Na2804). Concentration in vacua yielded 12.58g of a pale green oil. Distillation provided 9.18g (66*) of 20, BPo.9 73-80°C. 1H-NMH (60 MHz): 6 1.04 (d, J=6Hz, 3H), 2.89 (s, 1H), 3.24- 3.84 (m, 7H), 4.79 (s, 2H). CI-MS: 165 (M+l, 113), 89 (M- 75, base). IR (neat): 3450 (s), 2970 (s), 2930 (s), 1450 (m), 1115 (s), 1040 (s) cm'l. g28)-(benzyloxymethyl)-propanediol, 19 According to the procedure employed for 20, 55.7g (0.234 mol) of 16 was treated with 5.54g (0.146 mol) of lithium aluminum hydride. In a modified workup, the A1 salts collected by filtration were dissolved in a large 69 volume of dilute sodium hydroxide and extracted with ether (4 x 50 ml). The resulting ether extracts were combined with the organic phase obtained from the filtrate of the quenched reaction mixture and were washed with brine (200 m1) and dried (NazSOe). Concentration in vacuo yielded 47.19g of a clear colorless oil. Distillation afforded 40.15g (88%) of 19, BPl.5 122-12400. 1H-NMH (60 MHz): 6 = 1.2 (d, J=6Hz, 3H), 2.7 (s, 1H), 3.5 (m, 2H), 3.83 (m, 1H), 4.64 (s, 2H), 4.83 (s, 2H), 7.32 (br s, 5H). CI-MS: 197 (M+1, 23), 179 (M-l7, 88), 91 (C7H7*, base). IR (neat): 3475 (m), 3100 (w), 3070 (w), 2980 (s), 1040 (s), 1105 (s) cm‘l. ZS - benz lox math 1 - ro anal 10 The procedure described33 by Wuts and Bigelow was employed. 5.0 ml (0.055 mol) of oxalylchloride in 100 m1 of dry methylene chloride was cooled to -60°C (CHCla/COz). 8.5 ml (0.110 mol) of dry DMSO in 25 m1 of methylene chloride was slowly added to the stirring chilled solution, keeping the solution temperature below -50°. After the addition was complete, stirring was continued for 20 minutes. Then 3.92g « (0.020 mol) of 19 in 25 ml of methylene chloride was added to the solution over 10 minutes, resulting in a cloudy white solution. After stirring for 15 minutes, 35 ml of triethylamine (0.250 mol) was added, keeping the temperature below -40°C. The flask was warmed to room temperature and the reaction mixture cast into 2:1 H20/methylene chloride 70 (300 ml). The aqueous phase was extracted with methylene chloride (2 x 50 ml); the combined organic phases were washed with 1% H01 (50 m1), 5% NazCOa (50 ml), brine (50 m1) and dried (Na2804). After concentration in vacuo, Kugelrohr distillation (100°C, 0.02mmHg) afforded 3.50g (98%) of 10. 1H-NMH (60 MHz): 6 = 1.3 (d, J=6Hz, 3H), 4.1 (dq, J=2,6Hz, 1H), 4.66 (s, 2H), 4.84 (s, 2H), 7.36 (br s, 5H), 9.6 (d, J=2Hz, 1H). EI-MS (70eV): 165 (M-29, 43), 107 (C7H70’, 42x). 91 (c137*, base). IR (neat): 3035 (m), 2395 (s), 2720 (w), 1735 (s), 1105 (s), 1040 (s) cm‘l. g282-(methoxyethoxymethyl)-propana1, 11 Using the procedure employed for the preparation of 10, 5.13g (0.031 mol) of 20 provided 3.30g (65%) of 11, BPo.oos 40-45°C. 1H-NMR (60 MHz): 6 = 1.33 (d, J=6Hz, 3H), 3.4 (s, 3H), 4.45- 4.8 (m, 4H), 4.15 (dq, J=2,6Hz, 1H), 4.82 (s, 2H), 9.62 (d, J=2Hz, 1H). CI-MS: 163 (M+1, IX), 89 (M-73, base). IR> (neat): 2980 (s), 2935 (s), 2725 (w), 1735 (s), 1100 (s), 1035 (s) cm‘l. ZS -benz lox ro anal 12 Using the procedure employed for preparation of 10, 5.20g (0.031 mol) of 21" provided 4.72g (92*) of 12 after distillation (Kugelrohr, 100°o;os). 1H-NMH (60 MHz): 6 = 1.34 (d, J=6Hz, 3H), 3.9 (dq, J=2,6Hz, 1H), 4.62 (s, 2H), 7.34 (br s, 5H), 9.62 (d, J=2Hz, 1H). 71 EI—MS (70eV): 164 (M’. 0.8%), 153 (M“1. 5*). 135 (M-29. 22%), 91 (CvHv’, base). IR (neat): 3035 (m), 2980 (s), 2940 (s), 2718 (m), 1735 (s), 1160 (s) cm'l. General Procedure for Vinyl Organometallig:Aggitiong:to Protected Lactaldehzdes The reactions were performed at ca. 1 mmol scale. One equivalent of anhydrous zinc halide was introduced to a round bottom flask in a dry box and 10 m1 of dry solvent (ether, THF, or CHzclz) was added to both the zinc halide flask and a control flask. The flasks were cooled to -78° in a dry ice/acetone bath. With stirring, 2 equivalents of vinyl magnesium bromide were added to the zinc halide- containing flask, while one equivalent of vinyl Grignard was introduced into the control flask. The resulting solutions were stirred at -78°C for 15-30 minutes, after which time, a slow addition of 1 equivalent of protected lactaldehyde in ca. 1 ml of reaction solvent was made. Stirring was continued at -78°C for 30 minutes, followed by slow warming to room temperature over ca. 60 minutes. The reactions were cast into 50 m1 of 1:1 ether/saturated NHqu. The aqueous phase was extracted three times with Et20. The combined organic layers were washed with saturated ,NaHCOa, saturated brine (50 ml each) and dried (NazSOe). Concentration .in vacuo yielded diastereomeric allylic alcohols which were analyzed with no further purification. 72 Table 1 reports yields and ratios for product l3a,b, 14a,b, and 15a,b as a function of various reaction parameters. Sample Procedure for a ZnClg-mediated Preparation of 13a,b 0.38g (2.79 mmol) of ZnClz) was introduced to a flask in a dry box. 10 ml of dry THF was added to the flask and was cooled to -78° (C02/acetone). 6.2 ml (5.6 mmol) of vinyl magnesium bromide was added to the flask; the resulting solution was stirred for 30 minutes, after which time 0.500g (2.6 mmol) of 10 in 3 m1 THF was slowly added. Stirring was continued for 30 minutes at -78°C, after which time, the reaction mixture was gradually warmed (over 1 hour) to room temperature and was cast into 50 ml of 1:1 ether-saturated NHqu. The aqueous layer was extracted with ether (3 x 25 ml); the combined organic phases were washed with saturated NaHCOa (40 ml), brine (50 mL), and dried (Na2804). Concentration in vacuo yielded 0.480g (843) of a yellow oil. (4S,3S)- and (4S,3H)-4-(ggnzyloxymethyl)-3-hydroxy-l+pentene 13242 1&3: 1H-NMR (250 MHz): 6 = 1.2 (d, J=6Hz, 3H), 2.57 (br s, 1H), 3.66 (quintet, J=6Hz, 1H), 3.97 (br t, J=6Hz, 1H), 4.65 (ABq, UAH=17HZ, J=12.5Hz, 2H), 4.85 (m, 2H), 5.23 (br d, J=10.4Hz, 1H), 5.38 (br d, J=17Hz, 1H), 5.79-5.96 (m, 1H), 7.23-7.4 (m, 5H). 73 1 : 1.18 (d, J=6Hz, 3H), 2.57 (br s, 1H), 3.85 (dq, J=3.2,6Hz, 1H). 4.15 (m, 1H), 4.63 (ABq, vxa=l7Hz, J=12.5Hz, 2H), 4.85 (m, 2H), 5.19 (br d, J=10.2Hz, 1H), 5.28 (br d, J=16.3Hz, 1H), 5.78-5.92 (m, 1H), 7.25-7.4 (m, 5H). CI-MS: 223 (M+1, 1.4%), 205 (M-17, 5.4%), 91 (C7H7’, base). IR (neat): 3450 (s), 3095 (m), 3082 (m). 2990 (s), 2895 (s), 1100 (s), 990 (s) cm‘l. (43,38)-gang4(48,3R)-4-benzyloxy-3-h1droxy-1:pggtene._y§hh_ lggz 1H-NMR (250 MHz, CsDe, relative to Me2C0): 6 = 1.12 (d, J=6Hz, 3H), 2.66 (br s, 1H), 3.35 (quintet, J=6Hz, 1H), 4.07 (br t, J=6Hz, In), 4.40 (ABq, VA0=26H2, J=11.9Hz, 2H), 5.20 (br d, J=10Hz, 1H), 5.45 (br d, J=16.7Hz, 1H), 5.86-6.00 (m, 1H), 7.16-7.40 (m, 5H). lgb: 1.18 (d, J=6Hz, 3H), 2.66 (br s, 1H), 3.45 (m, 1H), 4.07 (br t, J=6Hz, 1H), 4.42 (ABq, vA0=26Hz, J=11.9Hz, 2H), 5.20 (br d, J=10Hz, 1H), 5.45 (br d, J=16.7Hz, 1H), 5.86-6.00 (m, 1H), 7.16-7.40 (m, 5H). CI-MS: 193 (M+1, 11.5%), 91 (C7H7‘, base). IR (neat): 3440 (s), 3090 (w), 3030 (m), 2980 (s), 1185 (s), 990 (s) cm'l. 74 (3S,4S)- and (3R,4S)-3-hYdroxy-4-(gethoxyethoxygethyl)~1- pentene 14s,!z 1&5: 1H-NMR (250 MHz, CsHs): 6 = 1.06 (d, J=6.2Hz, 3H), 3.11 (s, 3H), 3.28 (t, J=6Hz, 2H), 3.45-3.65 (m, 5H), 4.0 (br t, J=6Hz, 1H), 4.58 (s, 2H), 5.11 (br d, J=9.2Hz, 1H), 5.4 (hr d, J=16.6Hz, 1H), 5.76-5.91 (m, 1H). 159: 1.12 (d, J=6.2Hz, 3H), 3.09 (s, 3H), 3.25 (t, J=6Hz, 2H), 3.51-3.71 (m, 5H), 4.22 (br m, 1H), 4.56 (s, 2H), 5.13 (br d, J=9.2Hz, 1H), 5.42 (br d, J=16.6Hz, 1H), 5.82-5.97 (m, 1H). CI-MS: 191 (M+l, 3.4%), 89 (M-lOl, base). IR (CHCla): 3470 (s), 3080 (w), 3030 (w), 2980 (s), 1130 (s), 1110 (s), 1040 (s) cm‘l. (58,4R)- and (5S,4S)-5-(benzyloxymethyl)-4-hydroxy-2-hexyne, 24a,b via lithium methylagetylidg Approximately 5 m1 of propyne was condensed into a flask at -78°C, then diluted with 4 m1 of THF. To this solution was added 0.67 ml (1.46 mmol) of 2.18N n-BuLi over ca. 1 minute; the resulting cloudy white solution was stirred for 30 minutes at -78°. To the lithium acetylide- mixture was added 0.120g (0.73 mmol) of 10 in 4.0 m1 of THF over 3 minutes. The reaction was monitored by capillary GC (methyl-silicone column) and after 3.5 hours at -78°C, only 7% conversion was observed. The reaction mixture was then warmed slowly to room temperature (3.5 hours), then let stir an additional hour. The mixture was cast into 50 ml of 1:1 75 ether/saturated NHqu solution; the organic layer was separated and washed with brine (25 ml) then dried (Na2804). Concentration in vacuo yielded 99.0 mg of a crude yellow oil which contained a 67% yield of 24a,b as shown by GC analysis. gag: 1H-NMR (250 MHz): 6 = 1.28 (d, J=7Hz, 3H), 1.82 (d, J=2.8Hz, 3H), 2.80 (br d, J=4.9Hz, 1H), 3.77 (quintet, J=7Hz, 1H), 4.2 (m, 1H), 4.65 (ABq, VAB=14.8H2, J=11.5Hz, 2H), 4.79—4.9 (0, 2H). 7.27-7.36 (m, 53). Q: 1.25 (d, J=7.0Hz, 3H), 1.85 (d, J=7.8Hz, 3H), 3.25 (br d, J=8.8Hz, 1H), 3.82 (m, 1H), 4.31 (m, 1H), 4.68 (ABq, vxa=53Hz, J=ll.5Hz, 2H), 4.79-4.9 (m, 2H), 7.27-7.36 (m, 5H). CI-MS: 235 (M+l, 0.6%), 217 (M-l7, 1.3%), 91 (C7H7*, base). IR (CHCla): 3440 (m), 3070 (m), 3040 (m), 2980 (m), 2230 (w), 1040 (s), cm‘l. M filzl gnCngi-ng-Cg; In a procedure analogous to that employed for the Zn mediated vinyl additions, 2 mmol of CH3CEC-Li (5 m1 of a 0.40 M THF solution) was added to 0.270g (2 mmol) of ZnClz in 5 m1 of THF at 0° over a two-minute period. Stirring was continued for 30 minutes, after which time 0.164g (1 mmol) of 10 in one ml of THF was added, all in one portion. Stirring. was maintained for five hours at 0°C, then continued at room temperature overnight. The reaction mixture was cast into ether/saturated aqueous NHqu (50 ml, 1:1). The organic layer was separated and washed with brine 76 (50 ml) and dried (Na2504). Concentration in vacuo yielded 0.189g (93%) of a yellow oil, shown to consist (GC/IH-NMR) of 241) and 24a in a ca. 1.65:1 ratio. 24a,b via 2:1 CH3-CsC-Li/ZnClg The procedure and conditions employed were identical to those of the preceding reaction. Addition of 0.164g (1 mmol) of 10 to a' 4 mmol:2 mmol mixture of lithium methylacetylide and zinc chloride yielded 0.176g (86%) of a yellow oil shown to consist (GC/lH-NMR) of 24b and 24a in a 1.1:1 ratio. 24a,b via Znflflglz reduction of ynone 25 To 30.1mg (0.15 mmol) of 25 in 4 m1 of dry ether, cooled to -78°C, was added 2.5 ml of 0.18M Zn(HH4)235 in ether (0.45 mmol) over five minutes. After stirring for one hour at -78°, the reaction mixture was warmed to room temperature then cast into 1:1 ether/saturated NH4CI (50 ml), the aqueous layer was separated and extracted with ether (3 x 25 ml). The organic phases were combined and washed with saturated aq. NaHCO: (50 m1), brine (50 ml), and dried (Na2804). Concentration in vacuo yielded 31 mg of a cloudy white oil. Chromatography (10 mm, 2:1 ether/hexane, 10g) yielded 13.8 mg (45%) of 24a and diastereomer 24b in a 6.3:1 ratio (lH-NMR). 77 ,gZS2-benzyloxymethyl-hex-4-xn-3-one, 25 According to the procedure for 10, Swern oxidation36 of 24a,b (0.147g, 0.72 mmol) yielded 0.134g (92%) of 25 as a pale yellow oil. 1H-NMR (60 MHz): 6 = 1.4 (d, J=7Hz, 3H), 2.0 (s, 3H), 4.29 (q, J=7Hz, 1H), 4.65 (s, 2H), 4.8 (s, 2H), 7.3 (br s, 5H). EI-MS (70eV): 165 (M-67, 0.15%), 107 (C7H70*, 28%), 91 (C7H7’, base). IR (neat): 3090 (m), 3060 (m), 2980 (s), 2218 (s), 1675 (s), 1040 (s) cm‘l. (48,38)-4-benzyloxymethyl-pent-l-en-3- (tribgtylstangylmethyl)-gthgr.g§§;_ To 0.250g (6.25 mmol) of HR (23.6% in oil), washed 3 times with pentane, then suspended in 25 ml of dry THF was added 1.llg (5.0 mmol) of 13a (which contained 4% 13b by GC analysis) in 10 ml of THF over 1 hour. The stirring was continued for 1 hour at room temperature, then 0.55g (2.08 mmol) of lB-crown-S was added, followed by a slow addition of 2.37g (5.5 mmol) of iodomethyltributylstannane in 10 m1 THF. Stirring was continued at room temperature for 4.0 hours; the reaction mixture was then carefully decanted into 300 ml of 20:19:1 saturated NHqu/hexane/ether. The organic phase was separated, washed with saturated brine (100 ml) and dried (Na2804). Concentration in vacuo provided a crude product which was purified by f1ash37 chromatography (15:1 78 hexane/ether), yielding 1.35g (52%) of 26a as a colorless oil. 1H-NMR (60 MHz): 6 = 0.8-1.7 (m, 30H), 3.35-3.9 (m, 4H), 4.60 (s, 2H), 4.75 (s, 2H), 5.5-5.8 (m, 3H), 7.28 (br s, 5H). CI-MS: 469 (M-57, 3%, for 12°Sn), 467 (M-57, 2.5%, for 118Sn), 91 (07H7*, base). IR (neat): 3035 (w), 3018 (w), 2980 (s), 1645 (m), 1585 (m), 1105 (s), 695 (m) cm‘l. E- and Z-(5§)-bengy1oxygethyl-hgx-3-en-l-o1.;Zflhh_ To allyl stannyl methyl ether 26m (0.262g, 0.5 mmol) in 15 m1 of dry hexane, cooled to -78° (C02/acetone) was added 0.4 ml (1.0 mmol) of n-HuLi (2.5M/hexane) over a two-minute period. Stirring was continued for 1 hour at -78°, after which time the reaction mixture was slowly warmed to room temperature and quenched by casting into saturated NH4Cl/ether (100 ml, 50:50). The organic phase was separated and washed with saturated NaHCOa (50 ml), brine (50 m1), and dried (Na2804). Concentration in vacuo yielded 0.240g of a biphase oil. Purification by flash chromatography (1:1 hexane/ether) yielded 67.5 mg (57%) of a 3.5:1 Z/H mixture of 27. 1H-NMR (250 MHz): 6 = 1.20 (d, J=6.2Hz, 3H), 1.85 (br s, 10), 1.93—z.2 (m, 2H), 3.40 (br t, J=7Hz, 20), 4.55 (ABq, vas=24Hz, J=5.9Hz, 2H), 5.3-5.46 (m, 2H), 7.0-7.35 (m, 5H). CI-MS: 107 (C7H70’, 10%), 99 (M-137, base), 91 (C?H7*, 61%). IR (neat): 3410 (s), 3060 (w), 3020 (w), 2980 (s), 2940 (s), 79 1100 (s) cm‘l. 13C-NMR (CsDe): 6 = Z-27a: 21.7, 31.5, 61.9, 67.4, 69.3, 91.7, 129.3, 133.8, 138.7; E-Zfln 21.7, 36.0, 61.9, 67.4, 73.2, 91.7, 129.5, 134.4, 138.7 (some aromatic signals obscured by solvent). (I)-3-benzyloxymethyl-l-butyne, 28 Following the procedure described for preparation of 16, 3.50g (0.050 mol) of (i)-l-butyne-3-ol was treated with 10.92g (0.070 mol) of benzyloxymethyl chloride yielding 6.91g (77%) of 28, BPo.2s 84-9100. 1H-NMR (60 MHz): 6 = 1.4 (d, J=7Hz, 3H), 2.4 (d, J=2Hz, 1H), 4.60 (dq, J=2,7Hz, 1H), 4.82 (ABq, vaa=20Hz, J=14Hz, 2H), 5.22 (s, 2H), 7.3 (br s, 5H). gt)-5-benzyloxymethyl-hex-S-yn-l-ol 29 1.90g (0.010 mol) of 28 in 20 ml of dry THF was cooled to -78°. 4.6 ml (10 mmol) of 2.18M n-HuLi/hexane was added over 5 minutes; stirring was continued for 15 minutes, after which time, 1.33 ml BFa-OEtz38 was slowly added, stirring was continued for'lO minutes. Ca. 40 m1 of ethylene oxide was introduced to the reaction vessel over a 15 minute. period. The mixture was stirred at -78° for 45 minutes, then warmed to room temperature. Removal of excess ethylene oxide ‘was facilitated by partial evacuation of the reaction vessel with aspirator suction for ca. 10 minutes. The mixture was subsequently cast into 1N NHqu, and filtered through a pad of celite. The celite pad was rinsed (100 m1 80 Ethyl acetate); and the organic phase was separated and washed with saturated HaHCOs (50 m1), saturated brine (50 ml) and dried (HazSOo). Concentration in vacuo yielded 11.69g of a milky fluid. Flash chromatography (2 times), (6:4 hexane/ethyl acetate) of a portion of this material (3.35g) yielded 104 mg of a clear oil shown by CC analysis to consist of 80% 29 (12% yield). 1H-NMR (60 MHz): 6 = 1.42 (d, J=7Hz, 3H), 2.43 (dt, J=2,7Hz, 20), 2.64 (br s, 1H), 3.61 (t, 3:752. 23), 4.3-4.7 (m, 1H), 14.6 (s, 2H), 4.85 (ABq, vus=18Hz, J=9Hs, 2H), 7.3 (br s, 5H). Impurity Peaksz‘: 6 1.8-1.6 (m, 1.5H), 3.6-3.35 (3H). RI-MS (70eV): 234 (M’, 0.1%), 91 (0787*, 37%), 40 (M- 194, base). IR (neat): 3430 (s), 2980 (m), 2940 (s), 2960 (s), 2238 (w), 1100 (s), 1035 (s) cm‘l. Z- :t -benz lox meth l-hex-3-en-l-ol 27m In a procedure similar to that employed by Cram and Allinger39, 104 mg (0.43 mmol) of 29*was semi-hydrogenated (2 mg, 5%, Pd-Ba804, 2 mg quinoline, 10 m1 methanol, 1.5 atm.H2, 45 minutes). Workup was effected by filtration through a pad of celite. concentration of the filtrate in vacuo, and redissolution of the crude product in ether (20 ml). This ethereal solution was washed (2 x 20 ml) with saturated HHeCl solution and dried (NazSOo). Concentration In vscuo yielded a yellow oil. which was purified by flash chromatography (65:35 ether/hexane) yielding 43 mg (41%) of 81 l-Z-27m, spectroscopically identical to Z-27a obtained via [2,3] rearrangement of 26m. 4-methyl-pen-l-ggj3-ol 30 To 0.020 mol of vinyl magnesium bromide in 25 ml of THF at 0°C was added 0.72g (0.010 mol) of isobutyraldehyde in 5 ml of THF over 5 minutes. Stirring was continued for 1 hour at the end of which time the mixture was cast into 1:1 ether/saturated WHeCl (50 ml). The aqueous layer was extracted with ether (2 x 25 m1) and the combined organic phases washed with brine (50 ml) and dried (Hassoc). After concentration In vecuo, Kugelrohr distillation (120°, 760mmHg) afforded 0.433; (43:) of 30. 1H-HMR (60 MHz): 6 = 0.90 (d, J=6Hz, 6H), 1.75 (s, 1H), 1.3- 1.8 (m, 1H), 3.84 (m, 1H), 5.0-5.4 (m, 2H), 5.6-6.2 (m, 1H). RI-MS (70eV): 100 (M’, 7.1%), 83 (M-17, base), 82 (m-18, 54%). IR (neat): 3380 (s), 3080 (m), 2965 (s), 2880 (s), 1470 (m), 1385 (m), 990 (s) cm‘l. Following the procedure for 26s, 1.72g (17.2 mmol) of 30 was treated with 8.14g (18.9 mmol) of iodomethyltributylstannane, yielding 7.23g of a pale green oil.. Purification by flash chromatography (hexane, 2.15g of crude employed) afforded 2.08g (87%) of 31 as a colorless oil. 82 1H-NMR (60 MHz): 6 = 0.8-2.0 (m, 34H), 3.1 (br t, J=6Hz, 1H), 3.62 (ABq, vxa=24Hz, J=10Hz, 2H), 5.0-6.0 (m, 3H). EI-MS (70eV): 347 (M-57, 5.1% for 12°Sn), 345 (M-57, 3.8%, 118Sn), 177 (M-267, base). IR (neat): 3077 (w), 2960 (s), 2920 (s), 1585 (s), 1050 (s), 992 (m) cm'l. Z-ggnd §-5:!ethyl-hex-en-l-ol §g_ Following the procedure for 27a,b, 0.449g (1.1 mmol) of 31 was treated with 0.82 ml (2.2 mmol) of nBuLi (2.5M/hexane) to provide a pale yellow oil after workup. Flash chromatography (pentane-ether, 2:1) afforded 15.1 mg (12%) of E- and Z-32 in a 1:2 ratio as shown by a 1H-NMR shift reagent study. 1H-NMR (250 MHz, Cst): 6 = 0.93 (major isomer, d, J=6.1Hz, 6H), 0.90 (minor isomer, d, J=6.1Hz, 6H), 2.0-2.15 (m, 2H), 2.42-2.56 (m, 1H), 3.30-3.42 (m, 2H), 5.12-5.41 (m, 2H). CI-MS: 114 (M’, 2.2%), 96 (M-18, 63%), 81 (M-33, base). IR (neat): 3350 (s), 3008 (m), 2960 (s), 2935 (s), 1465 (s), 1380 (m), 1045 (s) cm'l. Attempted WittigTRearrangement of 26m in CH2012 Following the procedure employed for rearrangement of 26a in hexane, treatment of'26m with 2 equivalents of n-HuLi in CH2C12 failed to provide a rearrangement product with 94% recovery of 26a as realized by TLC and NMR data. 83 (gS)-benzzloxymethzl)-pegt-1-en-3-one 34 Follwing the procedure employed for Swern oxidation of 19, 2.02g (9.1 mmol) of 13 was converted to the desired enone 34 (0.30g, 15%) after isolation by flash chromatography (hexane-ethyl acetate, 10:1). 1H-HMR (80 MHz, Cube): 6 = 1.15 (d, J=8Hz, 3H), 4.12 (q, J=8Hz, 1H), 4.4 (s, 2H), 4.5 (s, 2H), 5.1-5.3 (m, 1H), 6.6- 6.2 (m, 2H), 7.1 (br s, 5H). CI-MS: 221 (M+l, 0.84%), 113 (M-107, 39%), 91 (C?H1*, base). IR (neat): 3100 (m), 3035 (m), 2980 (m), 2940 (m), 1700 (s), 1107 (s), 1030 (s) cm'l. 4S - and 48 38 -4-benz lox meth -3-h drox - - tene Ass To 0.134g (0.61 mmol) of 34 in 15 ml of dry ether, cooled to -22° (002, 0014), was added slowly 1.2 m1 of ethereal 0.17M Zn(HH4)2. After the addition was complete, stirring was continued for 25 minutes, then the reaction mixture was cast into 75 m1 of 1:1 ether-H20. The aqueous layer was drawn off and acidified to pH 3 and extracted (2 x 25 ml) with ether. The ether fractions were combined, washed with brine (50 m1) and dried (Mg804). Concentration in vacuo yielded 0.133g (98%) of 13 and 13m in a 7:1 ratio (GC, carbowax column). lH-NMR data compared identically with that of 13¢ obtained from vinyl addition to 10. 10. ll. 12. 13. 14. 15. 16. 84 [2.3] 1.151 or .- a) Bartlett, P. A. Tetrahedron 19m, 36‘, 2. b) see: "Asymmetric Organic Reactions"; Morrison, J. D.; Ed.; Academic Press, NY, 1984. Masamune, S.; Choy, W.; Peterson, J. S.; Sita, L. R. Aug. Chem. Int. Ed. 19%, 24, l . Eliel, E. L.; Allinger, N. L.; Angyal, S. J.; Morrison, G. A., "Conformational Analysis”, Am. Chem. Soc., 1965. Cram, D. J.; Kopecky, K. R J. Am. Chem. Soc. lfi9, 81, 2748. Anh, N. T.; Eisenstein, 0. jbjd. 1973, Q, 6146. For example, see: Evans, D. A.; Bartroli, J.; Shih, T. L. ibid. 181, 103, 2127. Early use of the Boron enolate. See: Mukaiyama, T.; Inoue, T. Bull. Chem. Soc. Jpn. 1950, 53, 174; and references cited therein. Heathcock, C. R.; Pirrung, M. C.; Youn8._S. 0.; Hagen, J. P.; Jarvi, E. T.; Badertscher, U.; Mfirki, H.-P.; Montgomery, 8. H. J. Am. Chem. Soc. 1&4, 106‘, 8161. Zimmerman, H.; Traxler, M. J. ibid. 157, 79, 1920. Shea, K. J.; Phillips, R. B. ibid. 1978, 100, 654. Blichi, C.; Cushman, M.; Wheat, H. ibz’d. 1974, w, 5563. Hoffman, R. W. Aug. Chem. Int. Ed. 1979, 18, 563. Rautenstrauch J. Chem. Soc. Chem. Comm. 1970, 4. Baldwin, J. E.; Patrick, J. E. J. Am. Chem. Soc. 1971, .93, 3556. Still, W. C.; Mitra, A. jbz'd. 1978, 100, 1927. Midland, M. M.; Tsai, D. J.-S. ibid. 195, 107, 3915; and references cited therein. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 85 Nakai, T.; Azuma, K.-I.; Mikami, K. Tetrahedron 1984, 40, 2303. Nakai, T.; Kishi, N.; Kimura, Y.; Mikami, K. J. Org. Chem. 1&3, 48, 279; and references cited therein. Tanis, S. P., Research Proposal, 1983. McGarvey, G. J.; Kimura, M. J. Org. Chem. 1982, 47, 5422. Reetz, M. T. Angew. Chem. Int. Ed. 1984, 23, 556; and references cited therein. MacDonald, T. L.; Mead, K. .J. Org. Chem. 1%, 50, 422. Still, W. C.; Schneider, J. A. Tetr. (Lett. 198), 21, 1035. Hart, D. J.; Hutchinson, D. K.; Chillous, S. E. J. Org. Chem. 1%, 47, 5418. Midland, M. M.; Kwon, Y. C. Tetrahedron Lett. 1”, 26, 5013. a) Oishi, T.; Tanaka, T.; Nakata, T. Tetrahedron Lett. 1&3, 26‘, 2653. b) Takahashi, T.; Miyazawa, M.; Tsuji, J. jhjd. 1%, 26', 5139. At the time of this writing, Takahashi (Reference 26b) reported low diastereoselectivity for alkynyl additions to a-alkoxyaldehydes; a variety of protecting groups, metal addends, and temperatures were examined. The BFa'OEtz catalyzed addition of 28 to ethylene oxide resulted in a contaminant which was quite difficult to remove from product 29 and semi-hydrogenation product (t)-27a. Unpublished work, Professor 8. P. Tanis. Recent work by Midland (Reference 25) suggests that rearrangement may yield E-olefin as the major isomer. Watson, S. L.; Eastham, J. F. J. Organomet. Chem. 1%, 9, 165. Prepared analogously to the procedure described for MEMCl; Corey, E. J.; Gras, J.-L.; Ulrich, P. Tetrahedron Lett. 1976, 809. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 86 Wuts, P. G. M.; Bigelow, S. S. J. Org. Chem. 1983, 48, 3489. 21 was prepared by Professor 8. P. Tanis. Prepared by the method of: Gensler, W. J.; Johnson, P.; Sloan, A. 0. B. J. Am. Chem. Soc. 130, 82, 6074. Swern, D.; Huang, S.-L.; Mancuso, A. J. J. Org. Chem. 1978, 43, 2480. Still, W. C.; Kahn, M.; Mitra, A. J. jbjd. 1978, 43, 2923. Reaction performed by the .method of: Yamaguchi, M.; Hirao, I. Tetrahedron Lett. 1%, 24, 391. Cram, D. J.; Allinger, N. L. J. Am._Chem. Soc. 156, 78, 2518. Kishi, Y.; Akasaka, K.; Fukuyama, T.; Schmid, G. 11nd. 1979, 101, 259. Roush, W. R.; Brown, R. J. J. Org. Chem. 1982, 47, 1373. Overman, L. E.; McCready, R. J. Tetrahedron Lett. 1982, 23, 2355. Vittorelli, P.; Hansen, H.-J.; Schmid, H. Helv. Chim. Acta 1975, 58, 1293. Sucrow, W. Angew. Chem. Int. Ed. 1%8, 7, 629. Ireland, R. E.; Mueller, R. H.; Willard, A. K. J. Am. Chem. Soc. 1976, .98, 2868. Evans, D. A.;-Nelson, J. V. ibid. 1900, 102, 774. Overman, L. E.; Mendelson, L. T. ibid. 1981, 103, 5579. Moriwaki, M.; Yamamoto, Y.; Oda, J.; Inouye, Y. J. Org. Chem.. 1976, 41, 1385. Evans, D. A.; Andrews, G. C.; Fujimoto, T. T.; Wells, D. Tetrahedron Lett. 1973, 1385. Chan, K.-K.; Saucy, G. J.~Org. Chem. 1977, 42, 3828. A TY "'TITJ'ITrLMILflMLu)flinnjfgt'iflmmiflfigflmfi"