Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Total syntheses of (+̲)-Methyl cantabrenonate, (+̲)-Methyl epoxycantabronate, and (+̲)-Crinipellin B Renaud, Johanne 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_spring_phd_renaud_johanne.pdf [ 8.91MB ]
Metadata
JSON: 831-1.0061795.json
JSON-LD: 831-1.0061795-ld.json
RDF/XML (Pretty): 831-1.0061795-rdf.xml
RDF/JSON: 831-1.0061795-rdf.json
Turtle: 831-1.0061795-turtle.txt
N-Triples: 831-1.0061795-rdf-ntriples.txt
Original Record: 831-1.0061795-source.json
Full Text
831-1.0061795-fulltext.txt
Citation
831-1.0061795.ris

Full Text

TOTAL SYNTHESES OF (±)-METHYL CANTABRENONATE, (±)-METHYL EPDXYCANTABRONATE, AND (±)-CRINIPELLIN B by JOHANNE RENAUD B. Sc., Universite de Sherbrooke, 1987  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMISTRY  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA March 1993 © Johanne Renaud, 1993  In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature)  Department of  .,.....s2c3,  The University of British Columbia Vancouver, Canada Date -TheLA_C^1 I  DE-6 (2/88)  1'  7 19 9 3 ,  ABSTRACT  The syntheses of two structurally related target compounds, (±)-methyl cantabrenonate (13) and (±)-methyl epoxycantabronate (14) are described in the first part of this thesis, while the preparation of the naturally occurring crinipellin B (15) is discussed in the second part of the thesis. Two methylenecyclopentane annulation sequences previously developed in our laboratories played key roles in the syntheses of (±)-13, (±)-14 and (±)-15. A new cyclopentenone annulation procedure was elaborated to assemble the last 5-membered ring of crinipellin B (15). The syntheses of (±)-13 and (±)-14 first involved the conversion of 3-methy1-2cyclopenten- 1-one (64) into the enone 62 via known chemistry. The enone 62 was transformed into the keto alkene 60 in one step utilizing a methylenecyclopentane annulation method developed previously by Piers and Karunaratne. Thus, conjugate addition of the reagent 18 to the enone 62 afforded the lithium enolate anion 106. This intermediate was allowed to undergo intramolecular alkylation, upon addition of HMPA to the solution containing 106 and warming to room temperature. The keto alkene 60 was converted into the ketone 59 via a sequence of steps which allowed establishment of the correct configuration of the methyl group at C-9. Three synthetic operations on the ketone 59 afforded the enone 120 which was successfully transformed into (±)-13 and (±)-14. The synthesis of (±)-crinipellin B (15) was accomplished in 22 steps from the commercially available 2-methyl-2-cyclopenten- 1-one. The starting material was efficiently converted into the bicyclic enone 194 which was subjected to a methylenecyclopentane annulation sequence regioisomeric to that described above. Thus, treatment of 194 with the reagent 209 in the presence of TMSBr provided the vinylgermane 219. A trimethylgermyliodine exchange on 219 afforded the vinyl iodide 220, which was allowed to cyclize under  conditions ((Ph3)4Pd, t-BuOK, t-BuOH, THF) developed previously by Piers and Marais. The keto alkene 191 was obtained in good overall yield from 194. Three synthetic operations on 191 yielded the ketone 224. A newly elaborated cyclopentenone annulation procedure allowed the conversion of 224 into the enedione 267. Alkylation of 224 with (Z)-3-bromo- 1 iodopropene (251) gave an intermediate vinyl iodide, which was allowed to undergo cyclization by treatment with n-BuLi in THF. The resultant allylic alcohol 249 underwent oxidative rearrangement to furnish the enedione 267. Two synthetic steps provided the enedione epoxide 188, which was converted into (±)-crinipellin B (15).  CO 2 Me  14  13  (CuCNLi)  CI  64  ^  60  62  ^  ^  iii  59  18  ^  ^  106  120  0  Me3 Ge  0  194  (CuCNLi)  ^  209  ^  219 R = Me3Ge 220 R = I  Br  224  0  267  iv  251  O  188  TABLE OF CONTENTS  Page ABSTRACT^  ii  TABLE OF CONTENTS^  v  LIST OF TABLES^  xi  LIST OF FIGURES^  xiii  LIST OF ABBREVIATIONS ^  xiv  ACKNOWLEDGEMENTS^  xviii  I.  1  II.  INTRODUCTION^  I.1 . GENERAL^  1  1.2. BACKGROUND AND PROPOSAL^  4  1.2.1. Previous Work^  5  1.2.2. The Problem^  10  DISCUSSION - METHYL CANTABRENONATE (13) AND  METHYL EPDXYCANTABRONATE (14) - ^  13  II.1. ISOLATION^  13  II.2. RETROSYNTHETIC ANALYSIS ^  14  v  11.3. TOWARDS THE SYNTHESES OF THE METHYL ESTER ^ DERIVATIVES OF CANTABRIC ACIDS  17  ^ 11.3.1. Preparation of the Ketol 63  21  11.3.2. Preparation of the Enone 62 via the Mesylate 97^  26  11.3.3. Preparation of the Keto Alkene 60^ 26 11.3.4. Preparation of the Ketone 59^  33  11.3.5. Preparation of the a,13-Unsaturated Keto Ester 57^38 11.3.6. Attempts to Prepare (±)-Methyl Cantabradienate (55) and (±)-Methyl Cantabrenonate (13)^38 11.3.7. Preparation of the Enone 120^  43  11.3.8. Completion of the Syntheses of (±)-Methyl Cantabrenonate (13) and (±)-Methyl Epoxycantabronate (14)  45  11.4. CONCLUSION^ III. DISCUSSION  -  CRINIPELLIN B (15)  -^  49 54  I11.1. ISOLATION^  54  111.2. THE PROBLEM^  55  111.3. PREVIOUS SYNTHETIC APPROACHES TOWARDS THE SYNTHESIS OF THE CRINIPELLINS^56 III.3.1. Approach by Mehta et al.^ 56 ^ 111.3.2. Approach by Curran and Schwartz 61  vi  111.4. RETROSYNTHETIC ANALYSIS^  67  111.5. TOWARDS THE SYNTHESIS OF (±)-CRINIPELLIN B (15)^  69  111.5.1. Preparation of the Enone 194^  69  111.5.2. Preparation of the Angularly Fused Triquinane 191^  72  I11.5.3. Preparation of the Ketone 224^  79  111.5.4. Attempts to Assemble the Last 5-Membered Ring of the Crinipellins^  85  111.5.5. Preparation of the Enedione 267 and of the Enone 231^  91  111.5.6. Preparation of the Enedione Epoxide 188^102 111.5.7. Attempts to Prepare (±)-Crinipellin A (43)^108 111.5.8. Model Studies on the a-Hydroxylation Reaction ^109 111.5.9. Completion of the Synthesis of (±)-Crinipellin B (15)^  111.6. CONCLUSION^  113 135  IV. GENERAL CONCLUSION ^  138  V. EXPERIMENTAL^  141  V.1. GENERAL^  141  V.1.1. Data Acquisition and Presentation^ 141 V.1.2. Solvents and Reagents ^  vii  143  V. EXPERIMENTAL (CONTINUED)  V.2. EXPERIMENTAL SECTION FOR THE SYNTHESES OF (±)-METHYL CANTABRENONATE (13) AND (±)-METHYL EPDXYCANTABRONATE (14)  146  Preparation of the Keto Acetal 94^  146  Preparation of the Bicyclic Keto Alcohol 63 and of the Bridged Bicyclic Keto Alcohol 96^  148  Preparation of the Enone 62 via the Keto Mesylate 97^151 Preparation of the Tricyclic Keto Alkene 60^153 Preparation of the Aldehydes 110 and 111 via the Diols 109^  157  Preparation of the Keto Dithioacetal 112^ 159 Preparation of the Ketone 59^  160  Preparation of the Enone 123^  161  Preparation of the Tertiary Alcohol 121^  163  Preparation of the Enone 120^  164  Preparation of the Tricyclic Ketone 126^  165  Preparation of the Keto Ester 129^  167  Preparation of the Phenylseleno Ketones 130 and 131^168 Preparation of (±)-Methyl Cantabrenonate (13)^170 Preparation of (±)-Methyl Epoxycantabronate (14)^171  Viii  V. EXPERIMENTAL (CONTINUED)  V.3. EXPERIMENTAL SECTION FOR THE SYNTHESIS OF (±)-CRINIPELLIN B (15)^ Preparation of the Silyl Enol Ether 200^  173 173  Preparation of 2-Bromomethyl-l-butene (201)^175 Preparation of the Keto Alkene 204^  176  Preparation of the Diketone 196^  179  Preparation of the Enone 194^  182  Preparation of the Keto Germane 219^  184  Preparation of the Keto Iodide 220^  186  Preparation of the Keto Alkene 191^  187  Preparation of the Alcohol 221^  189  Preparation of the Alkene 223^  194  Preparation of the Ketone 224^  195  Preparation of (Z)-3-Bromo-l-iodopropene (251)^197 Preparation of the Keto Iodide 250^  199  Preparation of the Allylic Alcohol 249^  201  Preparation of the Enedione 267 and of the Enone 231^203 Preparation of the Dione Epoxide 268^  209  Preparation of the Enedione Epoxide 188^  211  Preparation of the Ketol Epoxide 282^  214  ix  V.  EXPERIMENTAL (CONTINUED) V.3. EXPERIMENTAL SECTION FOR THE SYNTHESIS OF (±)-CRINIPELLIN B (15) (CONTINUED) Preparation of the Keto Epoxide 285^  218  Preparation of the a-Hydroxy Ketone 286^ 220 Preparation of the Triketone 281^  223  Preparation of the Diol 291^  225, 234  Preparation of (±)-Crinipellin B (15)^  228  Preparation of the Diketone Silyl Ether 292^ 232 Preparation of the Diketo Alcohol 290^  233  VI. REFERENCES AND NOTES^  235  VII. APPENDIX^  246  LIST OF TABLES  Page  Table 1H nmr Data (400 MHz, CDC13) for the Bridged Bicyclic Ketol 96: Decoupling Experiments.  25, 150  1H nmr Data (400 MHz, C6D6) for the Keto Alkene 60: Decoupling Experiments.  35, 155  3  1H nmr Data (400 MHz, CDC13) for the Ketone 126.  46, 166  4  1H nmr Data (400 MHz, CDC13) for the Keto Alkene 204.  178  5  1H nmr Data (400 MHz, CDC13) for the Diketone 196.  181  6  1H nmr Data (400 MHz, CDC13) for the Alcohol 221.  81, 192  7  1H nmr Data (400 MHz, CDC13) for the Alcohol 222.  83, 193  8  1H nmr Data (400 MHz, CDC13) for the Enedione 267.  205  9  1H nmr Data (400 MHz, CDC13) for the Enone 231: Decoupling Experiments.  207  1H nmr Data (400 MHz, C6D6) for the Enone 231: Decoupling Experiments.  208  1  2  10  11  12 13  14  11-1 nmr Data (400 MHz, CDC13) for the Enedione Epoxide 188.  213  1H nmr Data (400 MHz, CDC13) for the Ketol Epoxide 282: Decoupling and NOE Experiments.  116, 217  1H nmr Data (400 MHz, CDC13) for the a-Hydroxy Ketone 286.  120, 222  1H nmr Data (400 MHz, CDC13) for the Diol 291.  125, 227  xi  15^1H nmr Data (400 MHz, CDC13) for (±)-Crinipellin B (15).^130, 230 16^1H nmr (500 MHz) and 13 C nmr (125.8 MHz) Data for (±)Crinipellin B (15).^  xii  131, 231  LIST OF FIGURES  Figure^  Page  1^1H nmr Spectrum (400 MHz, C6D6) of the Keto Alkene 60. 34 2^1H nmr Spectrum (400 MHz, CDC13) of Synthetic (±)Methyl Cantabrenonate (13).^  50  3^1H nmr Spectrum (400 MHz, CDC13) of Synthetic (±)Methyl Epoxycantabronate (14).^  51  4  ^  1H nmr Spectrum (400 MHz, CDC13) of (-)-Methyl Epoxycantabronate (14) from San Feliciano et al.^52  5^1H nmr Spectrum (400 MHz, CDC13) of the Keto Alkene 191.^  80  6^1H nmr Spectrum (400 MHz, CDC13) of the Allylic Alcohol 249.^  98  7 8  ^  Stereoview of the Enedione Epoxide 188.^  ^  nnu. Spectrum (400 MHz, CDC13) of the Enedione Epoxide 188.^  106  1H  1H  107  nmr Spectrum (400 MHz, CDC13) of the Diol 291.^124  10^1H nmr Spectrum (400 MHz, CDC13) of Synthetic (±)Crinipellin B (15).^  129  LIST OF ABBREVIATIONS  A^angstrom Ac^Acetyl AIBN^2,2'-azobisisobutyronitrile Anal.^elemental analysis APT^attached proton test aq^aqueous atm^atmosphere br^broad i-Bu or But^isobutyl n-Bu^normal-butyl t-Bu or But^tertiary-butyl ° C^degree Celcius calcd^calculated cm^centimeter COSY^correlation spectroscopy C-x^carbon number x d^doublet 8^scale (nmr), dimensionless A^heat 2 D^two-dimensional DBN^1,5-diazabicyclo[4.3.0]non-5-ene DBU^1,8-diazabicyclo[5.4.0]undec-7-ene DDQ^2,3-dichloro-5,6-dicyano-1,4-benzoquinone  DIBAL^diisobutylaluminum hydride DME^1,2-dimethoxyethane DMF^N,N-dimethylformamide DMSO^dimethyl sulfoxide equiv^equivalent(s) Et^ethyl g^gram(s) glc^gas-liquid chromatography h^hour(s) HMBC^lE detected multiple bond heteronuclear multiple quantum coherence HMDSH^1,1,1,3,3,3-hexamethyldisilazane HMDSK^1,1,1,3,3,3-potassium hexamethyldisilazide HMPA^hexamethylphosphoramide HMQC^1H detected heteronuclear multiple Quantum coherence H-x^hydrogen number x Hz^hertz (s -1 ) it^infrared J^coupling constant  kg^kilogram(s) LDA^lithium diisopropylamide LHMDS^1,1,1,3,3,3-lithium hexamethyldisilazide m^multiplet M^molar MCPBA^3-chloroperoxybenzoic acid Me^methyl mg^milligram(s) MHz^megahertz  xv  min^minute(s) mL^milliliter(s) 111-^microliter(s) mmol^millimole(s) mol^mole(s) MOM^methoxymethyl mp^melting point Ms^methanesulfonyl mult.^multiplicity N^normal neg^negative NMO^4-methylmorpholine N-oxide nmr^nuclear magnetic resonance NOE^nuclear Overhauser gffect p^page PCC^pyridinium chlorochromate PDC^pyridinium dichromate Ph^phenyl ppm^parts per million Pr^propyl i-Pr^isopropyl psi^pounds per square inch pyr^pyridine q^quartet rt^room temperature s^singlet sat.^saturated xvi  t^triplet TBDMS^tertiary-butyldimethylsilyl Tf^trifluoromethanesulfonyl TfO^trifluoromethanesulfonate THE^tetrahydrofuran tic^thin layer chromatography TMS^trimethylsilyl p-Ts^para-toluenesulfonyl p-TsO^para-toluenesulfonate TPAP^tetra-n-propylammonium perruthenate  ACKNOWLEDGEMENTS  A number of people need to be thanked for their support and help at different periods of my stay at UBC. My first acknowledgements are undoubtedly granted to my research supervisor, Professor E. Piers, for his constant support, incredible patience and excellent guidance throughout this work. Working in Pr. Piers' laboratories has not only been a stimulating learning experience but has also been very enjoyable. I am very grateful to the various members of the Piers and Andersen's groups with whom I have experienced numerous interesting and fruitful chemistry discussions. These exchanges have often led to breakthroughs in my projects. Thanks to Montse, Miguel, Pierre, Richard, Betty-Anne, Fraser, Jacques, Renata, Philip, Tim, Christine, Yongxin, Guy, Romano, Livain, Chantal, Veljko, Jana, Dave, Judy and Francisco. Thanks are also due to Giri, Adri, Sandra, Francisco K.-B., Narly, Golnar and Jurgen. I am very indebted to Dr. Yongxin Han, Ms Renata Oballa, Dr. Christine Rogers, Dr. Jurgen Wagner and Pr. J. R. Scheffer for proof-reading my thesis. Special thanks are due to Ms R. Oballa for work-related help in a number of circumstances and to Jacques and Dr. Sherman for advice needed concerning computers. I am very grateful to the UBC staff and particularly to Mrs. M. Austria and Mrs. L. Darge from the nmr laboratory, to Mr. P. Borda who ran all the elemental analyses in this thesis and to the mass spectrometry staff. The secretaries of the Chemistry department and especially Mrs. T. Schreinders are thanked for their valuable help on various occasions during my studies at UBC. Je remercie egalement Mean pour son amide et sa comprehension qui ont facilite mon sejour a Vancouver. Finalement, je remercie chaleureusement ces deux étres qui ont suivi, de pros ou de loin, mon cheminement a travers la vie et qui m'ont toujours encouragee. Je leur dois cette curiosite, ce desir d'explorer et de comprendre ainsi que cette perseverance qui m'ont permis de poursuivre des etudes en science: grand merci a mes parents. I have greatly appreciated the financial independence associated with the postgraduate scholarships I received from NSERC (Natural Sciences and Engineering Research Council of Canada), Bio-Mega Inc. and les fonds FCAR (fonds pour la formation de chercheurs et l'aide la recherche).  xvii i  A mes parents et a Marie-France Avec affection  xix  I. INTRODUCTION 1.1. GENERAL The course of progress in many disciplines of science has often been influenced, sometimes dramatically, by the social concerns, needs, and beliefs during a particular era.  1,2  The curiosity and personal interest of the scientists who are involved in research projects, along with the occurrence of accidental discoveries, also affect the evolution of the different spheres of science. Over the years, the boundaries between various scientific fields have faded (although each one still retains its identity) and new areas of research have resulted from the fusion between domains of these disciplines. Chemistry has now incorporated information from fields such as mathematics, physics, biology, pharmacology and medicine. On the other hand, some of these disciplines such as biology, pharmacology, medicine and material sciences have greatly benefited from the development of chemistry. It is therefore tempting to say, as suggested by Seebach 2 in his reply to the declaration that chemistry as a discipline had lost its identity that, on the contrary, chemistry has become a central science. The field of chemistry embraces many related subdisciplines (such as organic, inorganic, organometallic, analytical, theoretical, physical and biological chemistry) which have aroused the interest of many scientists. One of these fascinating subdisciplines, organic chemistry, has evolved enormously over the past years due to the advent of new or improved tools and methods in analytical chemistry. 2 Chromatographic methods (flash chromatography, gas chromatography and HPLC to name only a few of them) and spectroscopic methods (nmr spectroscopy -which includes a wide variety of useful experiments- infrared spectroscopy and mass spectrometry) have greatly facilitated the separation and the analysis of mixtures of products. These methods have also allowed the structural determination or confirmation of an 1  increasingly large number of important compounds. X-ray structure analysis has proven to be an invaluable tool to scientists. The developments of the refined technologies mentioned above have resulted in, among other things, the syntheses of more complex, unusual products originating either from the imagination or from nature. Recently, the preparation of enantiomerically pure compounds (of consequence for industries involved in the syntheses of perfumes and drugs, for example) has also been an important goal for the community of organic chemists. The syntheses of target molecules possessing intriguing structural features have allowed researchers to discover and apply new synthetic methods (for example, milder ways of forming carbon-carbon bonds and of achieving functional group manipulations). In some cases, syntheses served to confirm original structural assignments of natural compounds and to verify the feasibility of suggested synthetic plans. Another important motivation to synthesize a natural product that possesses interesting biological activity and may be a promising drug for pharmaceutical companies concerns the need to obtain appreciable quantities of the desired compound when it is found in nature in small quantities and is difficult to isolate. Taxol (1), 3 a very promising antileukemic and antitumor drug, found in small amounts in the bark of the ecologically threatened yew tree Taxus brevifolia , is an example of such a case. 0  Ph^0 Ks)1 C■  NH^i OH  OH 1 H OCOCH2  OCOPh  1  The planning of a synthesis is an intellectual challenge that relies on the chemist's creativity, area of interest and, obviously, on the molecular complexity of the target molecule. The molecular complexity, as explained by Corey, 4 is expressed in different ways, such as the 2  size of the molecule, the elements, functional groups and stereocenters it includes, its cyclic connectivity, its reactivity, and its instability. The successful synthesis of a complex compound therefore requires the logical analysis of the problem being faced and the careful design of a synthetic pathway. 4 The establishment of a synthetic route necessitates disconnection of the target molecule into simpler units. This process, called retrosynthetic analysis, 4 eventually leads to commercially available starting materials. Convergences is a useful concept to bear in mind during a retrosynthetic analysis. This process, in which two or more components are combined together in a key step to form a new intermediate, is the most efficient way to assemble complicated products. A variety of approaches for the efficient construction of functionalized carbon frameworks have appeared in the literature recently. A short way to prepare functionalized products from simpler starting material consists of using reagents whose structure will be incorporated partly or in whole into a newly formed substance. These building units have been called conjunctive reagents 6 or multiple coupling reagents.? Reagents of this type that possess two reactive sites within the same molecule are named bifunctional conjunctive reagents. 6 b The centers can be deployed either simultaneously or sequentially. Examples of reagents that include either two nucleophilic or donor (d) sites (2), two electrophilic or acceptor (a) centers (3), or one donor (d) and one acceptor (a) site (4 and 5) are shown below. OLi OLi  0  ad MgBr  OMe  0  2^3  4  Li  CI  5  The coexistence of a nucleophilic and an electrophilic center within the same molecule could lead to self destruction of the reagent. Masking one of the reactive sites is one way to prohibit such a reaction. The reagent 4, in which the acceptor site is masked as an acetal 3  group, illustrates this case. Generation of one of the reactive centers under conditions that allow the temporary coexistence of the two sites is another alternative. For example, the reagent 5 can be produced from a suitable precursor, such as 4-chloro-2-trimethylstanny1-1butene (6), by transmetallation with methyllithium at low temperature (equation 1). 8 af Higher reaction temperatures result in self-annihilation of the reagent.  Li  CI  (1)  5 1.2. BACKGROUND AND PROPOSAL.  Previous work carried out in our laboratories has been directed, among other things, towards the design of bifunctional conjunctive reagents to assemble functionalized 5-membered rings. 8 Two complementary regioisomeric methylenecyclopentane annulation procedures have been developed. 8  b -f,9 Thus, an enone of general structure 7 can be transformed into either of  the two regioisomeric methylenecyclopentane annulation products 8 or 9 (Scheme 1). In the case of the annulated adduct 9, when R = H, the exocyclic double bond isomerizes to the endocyclic position, and the enone 10 results.  \\^\^If R=H ^ ^ ^ R' 8  7  9  10  Scheme 1  The first annulation method played a key role in the total syntheses of (±)-pentalenene (11), 8 c , e and (±)-A 9 ( 12)-capnellene (12), 81 .f accomplished by Piers and Karunaratne. The total syntheses of two structurally related methyl ester derivatives of natural products, (±)-methyl  4  H 11  12  cantabrenonate (13) and (±)-methyl epoxycantabronate (14), also utilized this annulation method. This work will be described throughout the first part of this thesis. The second methylenecyclopentane annulation procedure was employed in the synthesis of a new diterpene tetraquinane, crinipellin B (15). The preparation of this structurally and biologically fascinating substance will be covered in detail in the second part of the thesis.  CO 2 Me 13  o  o  14  1.2.1. Previous Work.  Recent reports from our laboratories have described the preparation of 4-chloro-2trimethylstannyl- 1-butene (6) 8 f,1 ° and have demonstrated its utility as a precursor of a number of valuable bifunctional conjunctive reagents. 8 The two-step synthesis of the vinyltin compound 6 from the alkynol 16 is outlined in Scheme 2. Reaction of 3-butyn-l-ol (16) with (trimethylstannyl)copper(I)•dimethyl sulfide, in the presence of methanol 8 U° afforded regioselectively the desired 4-hydroxy-2-trimethylstannyl- 1-butene (17). The precursor 4-chloro-2-trimethylstanny1-1-butene (6) was derived from the reaction of the corresponding alcohol 17 with triphenylphosphine in carbon tetrachloride, in the presence of triethylamine (Scheme 2).  5  .......,.OH  1) Me3SnCu•Me2S, THF, Me0H -78 °C, 2.5 h; 0 °C, 3 h.  Me 3 Sn  2) NH4CI-NH4OH, H2O  16^  i  rsOH  17 Ph3P CC14 Et3N, A  Me 3Sn^CI  Scheme 2  ^6  4-Chloro-2-trimethylstanny1-1-butene (6) could be converted into either of the cuprate reagents 18 or 19. Transmetallation of the vinyltin 6 with MeLi at -78 °C afforded 4-chloro-2lithio- 1 -butene (5). 8 b , f Treatment of the yellow solution containing the lithio derivative 5 with either solid CuCN or solid CuSPh (1 equiv) provided, after brief warming of the resultant reaction mixture, the corresponding cuprate reagents 18 or 19 8 b , f (Scheme 3). Me 3Sn^  -CI 6  MeLi THF^Li -78 °C  ^  CI  5  CuCN or CuSPh, -78 °C to -63 °C  ^  CI N.-  18 M = Cu(CN)Li 19 M = Cu(SPh)Li  Scheme 3  A variety of cyclic enones underwent reaction with the cuprate reagents 18 or 19 (in the presence of an additive such as BF3•0Et2 when necessary) to afford, after basic workup, the corresponding 1,4-adducts (Scheme 4). 8 b Treatment of the various chloro ketones thus obtained with a base such as KH allowed intramolecular alkylation to occur and yielded bicyclic keto alkenes. Alternatively, the 1,4-addition-cyclization sequence could be effected in a one-pot operation. After completion of the 1,4-addition to the enone, the resulting enolate  6  was allowed to cyclize directly, upon addition of HMPA to the reaction mixture and warming to room temperature. OLi  CI  CI  ^1.—  18 M = Cu(CN)Li Fr 19 M = Cu(SPh)Li  HMPA, warm to rt  KH, THE  Scheme 4 In the annulation method described above, reagents 18 and 19 act as synthetic equivalents to the 1-butene d 2 ,a4 -synthon 1 1 20. The enones are synthetic equivalents of the  21  20  synthon 21. The combination of these donor-acceptor synthons creates a functionalized adduct, whose basic structure is found in an array of terpenoid natural products. The use of this annulation procedure is illustrated by two examples.  7  As mentioned earlier, the 5-membered ring annulation sequence was an important key step in a short synthesis of (±)-pentalenene  8c,e  The sequence is shown in Scheme 5.  The synthesis of (±)-pentalenene (11) began with the known keto ketal 22 which was transformed into the enone 23 via a series of synthetic operations. Compound 23 was set to  °  0 =Cly '0 H  H 23  22 1) BrMg CuBrMe2S THF, -78 °C, 1.5 h 2) aq NII4C1-NH4OH  KH THF  -4.11(---  H 26  80%  H 25  83%  Scheme 5  undergo the methylenecyclopentane annulation sequence devised by Piers and Karunaratne. 8 b Copper(I)-catalyzed conjugate addition of the Grignard reagent 24 (formed by treatment of 5 with MgBr2) to the a,13-unsaturated ketone 23 furnished the chloro ketone 25 in good yield (83%). The bicyclic adduct 25 issued from the conjugate addition of 24 to 23 was cis-fused. This result is not surprising in view of the fact that cis-fused 5-membered rings are generally formed preferentially to trans-fused compounds since the latter types of products are very strained. Reaction of the ketone 25 with KH in THF afforded the keto alkene 26 in 80% yield. It is interesting to note that the methylenecyclopentane annulation procedure allowed the 8  conversion of the relatively simple enone 23 into the more complicated substance 26 which contained appropriately positioned functionalities that could be utilized for further transformations. In fact, subjection of 26 to a series of suitable functional group manipulations gave (±)-pentalenene (11). In another application of the methylenecyclopentane annulation sequence, Piers and Karunaratne have published 8 d , f a very elegant synthesis of the sesquiterpene (±)-A 9 ( 12 )capnellene (12) via the route outlined in Scheme 6. In this synthesis, the annulation procedure served to construct two of the three rings of the linearly fused triquinane 12. The 1) BrMg^  II  , ....,..,CI 24 ,  CuBr•Me2S, BF3.0Et2 THF, -78 °C, 1.5 h 3) aq NH4C1-NH4OH  CI ^Do-  KII THF  28  75%  80%  1) KH THF  31 (68% from 29)  CI  CuBr•Me2S THF, -78 °C, 1.5 h 2) aq NH4C1-NH4OH  30  1)LiA1H4, Et20, -78 °C 2) NaH, CS2, MeI, THF 3) n-Bu3SnH AIBN, PhMe, reflux  Scheme 6  9  29  central ring of the skeleton of 12 was provided by the starting material 2-methy1-2cyclopenten- 1 -one. Thus, copper(I)-catalyzed 1,4-addition of 24 to 2-methyl-2-cyclopenten1-one gave an epimeric mixture of the chloro ketones 27 in 80% yield. The adducts 27 underwent intramolecular alkylation upon treatment with KH in THF. This reaction furnished the keto alkene 28 in good yield. Appropriate transformation of the alkene 28 gave the enone 29 which was subjected to the now well-known annulation procedure. The tricyclic keto  alkene 31 was isolated in 68% yield. The assembly of the carbon skeleton of (±)-A 9 ( 12)capnellene (12) was thus achieved efficiently in a few steps by a reiterative utilization of the methylenecyclopentane annulation sequence. Completion of the synthesis of 12 was readily accomplished in three synthetic operations which involved removal of the extraneous carbonyl group.  1.2.2 The Problem.  The two natural products synthesized as summarized above are solely constituted of carbon and hydrogen atoms. However, a wide variety of sesquiterpenoid natural products contain oxygen atoms. It would be gratifying to utilize the annulation method in the preparation of structurally more complex targets. A series of oxygenated sesquiterpenes with interesting carbon frameworks, the cantabric acids 32, 33, 34 and 35, 12 are promising candi-  CO2H  CO2H  32  CO2H  33  dates for the successful application of the 5-membered ring annulation. However, before discussing in more detail the isolation and the preparation of these interesting natural products, a brief overview of different types of angularly fused triquinane containing substances is given. 10  The cantabric acids incorporate into their structures the tricyclo[6.3.0.0 1,5]undecane carbon skeleton 36. A series of angularly fused triquinane natural products, classified  83 36  according to their substitution patterns, possess this carbon framework. 13,14 Among these groups of natural products are the isocomane 37, the pentalenane 38, the silphinane 39 and the silphiperfolane 40 families. The simpler members of these families are hydrocarbons.  .>\S  37  -466.<>.6Th 38  39  40  However, many oxygenated derivatives are known. 13,14 The tricyclo[6.3.0.0 1,5]undecane skeleton is also embedded in complex natural products such as retigeranic acid (41), laurenene (42), 16 crinipellin A (43) 17 and crinipellin B (15). 17  o  OH  43  11  15  The first silphiperfolane-type sesquiterpenoids were discovered in 1980 by Bohlmann and Jakupovic. They described the isolation and the structural elucidation of the hydrocarbons 713H- and 7aH-silphiperfol-5-ene 44 and 45, 18 silphiperfol-6-ene (46), 18 a and the ketones 47 19 and 48. 20  Since then, an increasing number of silphiperfolane-type sesquiterpenoids have been isolated from different sources such as plants and marine organisms. The structural formulas of some natural products of this type recently found are shown in 49-54. 21  HO HO AngO  49  50  1985  51 1989  1988  HO  52  HO  53 1991  1990  12  54 1991  Since the discovery of the first triquinane sesquiterpenoids, a large number of ingenious synthetic approaches and total syntheses have been reported. 22 The isolation of the unusual silphiperfolane skeleton and its oxygenated derivatives has received the attention of various synthetic groups and many imaginative syntheses have appeared in the recent literature. 23 As a result of our interest in the preparation of these types of compounds, we undertook the syntheses of two sesquiterpenes possessing the silphiperfolane skeleton. The isolation 12 and the structural elucidation 12 of cantabrenonic acid (33) and epoxycantabronic acid (34) and of their methyl esters 13 and 14 are delineated in the next section. The description of our syntheses of the methyl ester derivatives 13 and 14 follows.  II. DISCUSSION - METHYL CANTABRENONATE (13) AND METHYL EPDXYCANTABRONATE (14) II.1. ISOLATION.  The cantabric acids 32-35 were isolated 12 from the aerial parts of a small plant, Artemisia cantabrica (Lainz), Lainz, endemic to the north of Spain. Extraction of 1.3 kg of  air-dried material with hexane for 20 hours yielded 35.7 g of crude extracts. After removal of the fat products, the remaining material was dissolved in diethyl ether and extracted with a 4% solution of aqueous NaOH. The acidic materials (3.8 g), obtained by acidification of the base extracts, were treated with CH2N2. After repeated chromatographic separations (on silica gel and on silver nitrate impregnated silica gel ) and crystallization of the appropriate compounds, the four ester derivatives 55 (30 mg), 13 (330 mg), 14 (86 mg) and 56 (139 mg) were obtained. The structure of the epoxide 14 was first solved on the basis of data derived from various 2D nmr experiments ( 1 H- 13 C heteronuclear, 1 H- 1 H homonuclear correlation spectra, 2D NOE spectra) and from infrared, UV, 1 H and 13 C nmr and mass spectra. The structures of  13  the three compounds 55, 13 and 56 were deduced by comparison of their spectral data with the information obtained for the keto epoxide 14.  CO2Me 55  „,'CO2Me  CO2Me  0^OH 14^56  13  Methyl cantabradienate (55), a pale yellow oil and methyl cantabrenolate (56), a viscous oil, are both unstable in atmospheric conditions. Methyl cantabrenonate (13) and methyl epoxycantabronate (14) are both stable, crystalline materials. 11.2. RETROSYNTHETIC ANALYSIS. The four natural products 32-35 are, from a structural viewpoint, aesthetically pleasing. Their synthesis represented a challenge because of the presence of (at least) four contiguous chiral centers, two of which are quaternary, and by the presence of oxygen functional groups. A particularly attractive way of synthesizing one, two, or all of these natural products would consist of devising a centralized approach which would lead to one or the other product by modification of the last (or last few) synthetic steps. A possible retrosynthetic pathway, which includes the methylenecyclopentane annulation discussed above, is outlined in Scheme 7. Since the natural products were characterized as their methyl ester derivatives, we envisaged the syntheses of the latter compounds, which could be transformed into the corresponding acids. Three of the target compounds could be linked together via simple synthetic transformations. Methyl cantabrenolate (56) could be derived from methyl cantabrenonate  14  ^  1=>  <=1  .u. CO2Me  ■  ^  CO2 Me  5 57^  58  II d^a  +  20  61  III  M  60  59  0 Li(CN)Cu  i=>  CI + 18  /— 0  ^  62  ^  r...0  )  or^  v,  ^\--0^MgBr^0^MgBr 67^  66  Scheme 7  15  63  ^  64  + HOd 65  (13) by stereoselective reduction of the ketone function. Methyl epoxycantabronate (14) could  also be synthesized from methyl cantabrenonate (13) via an epoxidation reaction. Methyl cantabrenonate (13) and methyl cantabradienate (55) could originate from the oc,13-unsaturated keto ester 57. Formation of the last carbon-carbon double bond of 55 could be accomplished with a variety of methylenation reagents. Addition of methyllithium to the ketone carbonyl of 57, followed by dehydration, should also lead to the desired product 55. On the other hand,  rearrangement of this 1,2-adduct, resulting from reaction of 57 with MeLi, with chromium reagents such as PCC or PDC would yield methyl cantabrenonate (13). The a,13-unsaturated keto ester 57 could be synthesized in a few steps from the ketone 59. Carbomethoxylation of 59 and oxidation of the resultant keto ester 58 would yield the  desired 57. Functional group manipulations involving the keto alkene 60 would provide the ketone 59. Examination of molecular models indicates that one face of the exocyclic double bond of 60 is much more hindered than the other. Simple hydrogenation of 60 would be predicted to give the ketone epimeric at C-9 to the desired substance 59. An alternative sequence of reactions would therefore have to be planned to obtain the correct stereochemistry at this center. The keto alkene 60 could be viewed as being derived by the methylenecyclopentane annulation method devised in our laboratories. Thus, disconnection of the alkene containing ring would afford synthons 61 and 20. These synthons correspond to the synthetic equivalents 18 (whose precursor is 4-chloro-2-trimethylstannyl-1-butene (6)) and 62. It should be noted that the stereochemistry of the newly formed center resulting from the 1,4addition to the enone 62 is crucial for the success of the synthesis. Failure to get the desired epimer as the major product would require the design of a new synthetic plan. However, related cases led us to believe that the desired material would be formed as the major or only isomer.  16  ^  Finally, the enone 62 could originate from commercially available 3-methy1-2cyclopenten- 1-one (64) via a 1,4-addition of a synthetic equivalent to the donor-acceptor synthon 65, followed by an intramolecular aldol condensation. 11.3. TOWARDS THE SYNTHESES OF THE METHYL ESTER DERIVATIVES OF CANTABRIC ACIDS.  The Grignard reagents 66 and 67 are suitable synthetic equivalents to the synthon 65. Both of these bifunctional reagents have been used to convert cyclic enones to bicyclic or tricyclic ketols or enones. Helquist and coworkers developed 24 this useful annulation method, which involves conjugate addition of the Grignard species 66 to the enone substrate, followed by a cyclization reaction. An example of this type of annulation sequence had been reported previously by Heathcock and Brattesani, 25 but this group did not pursue further investigation. Helquist and coworkers 24 prepared the Grignard reagent 66 from the corresponding bromide 68 and magnesium powder, obtained from the reduction of anhydrous magnesium dichloride with potassium metal (Rieke's procedure) (Scheme 8). Alternatively, the Grignard  0  > C  68  66  CBr  Mg  0  ,„,„„  LO 1  C^0D 0°0 69 Wurtz coupling product Scheme 8  17  MgBr  reagent could be produced by using a three-fold excess of freshly ground magnesium turnings and a concentrated solution of the bromo acetal. In these cases, only a small amount of the Wurtz coupling product 69 was formed. Three examples of conversion of cyclic enones into bicyclic substances by Helquist's method 24 are presented in Scheme 9. 2-Cyclopenten- 1 -one, 2-cyclohexen- 1-one and 2-cyclohepten- 1-one were allowed to undergo copper(I)-catalyzed conjugate addition with the Grignard reagent 66. The resultant keto acetals 70-72 were isolated in yields ranging from 77% to 87%. These keto acetals 70-72 cyclized upon exposure to acidic conditions. The enones 74 and 75 were isolated directly after workup of the reaction mixture and purification of the crude material thus obtained. However, in the case of the bicyclo[3.3.0]octane ketol 73, spontaneous dehydration did not occur. It seems that (3-hydroxy ketones of this type do not undergo elimination of water in the reaction conditions.  1)  Coy\,,MgBr o  66  CuBr•Me2S, THF, Me2S -78 °C, 10 h; warm up to 0 °C over 6 h; 0 °C, 2 h  HCI, H2O, THF, rt 72 h  0,  70 n 71 n 72 n  2) aq NH4C1-NH4OH  =1 =2 =3  Scheme 9  77% 85% 87%  74 n = 2 89% 75 n = 3 80%  ^  73 5 4 %  Ito and coworkers, in a synthesis of (±)-silphinene (79), utilized this annulation procedure to construct the last ring of the target molecule. 26 Thus, conjugate addition of 66 in the presence of CuI to the enone 76, followed by intramolecular aldol condensation, provided  18  the tricyclic intermediate 78 (Scheme 10). The ketol 78 was converted into (±)-silphinene (79) in six synthetic operations.  0  OH  o 1)C Y\/ MgBr  HC1, THF  0^66  CuI 2) Workup 76  H2O  77  78  70%  98%  Scheme 10  In their synthesis of (±)-modhephene (83), Oppolzer and Marazza built the bicyclic enone 82 using Helquist's annulation method. 27 The sequence is shown in Scheme 11.  C  oy\ ,M g Br O ^66  ^0.-  CuBr•Me2S, THF, Me2S -78 °C, 4 h; -30 °C, 15 h; rt, 20 min  0 X = H or X = O Me  81 50% 1) aq HCOOH, HCOONa,n, 15 h 2) 1 N aq NaOH, Et20, rt, 2.5 h 3) MsC1, Pyr, 0 °C, 15 min; rt, 30 min 4) DBN, Et2O, rt, 4 h  82 45%  Scheme 11  19  I  In 1976, Stowell found 28 that the Grignard reagent 67 derived from 2-(2-bromoethyl)1,3-dioxane (84) is more stable and easier to prepare than that derived from 2-(2-bromoethyl)1,3-dioxolane (68). Paquette and Leone-Bay showed 29 in their synthesis of (±)-silphinene (79) the usefulness of this reagent (Scheme 12).  Copper(I)-catalyzed 1,4-addition of the reagent 67 to 4,4-dimethy1-2-cyclopenten-1one (85) and subsequent cyclization afforded the ketol 86. As observed in previous cases with this type of (3 hydroxy ketones, dehydration of 86 does not proceed spontaneously. -  Therefore, the hydroxyl function was transformed into a good leaving group (OSO2CH3) and  1)  85  C °>.■ \,,, M g Br OH  0^67 CuBr•Me2S, THF, Me2S -78 °C, overnight;  1) CH3S02C1, Et 3 N, CH2C12, 0 °C  4 h, 0 °C; aq NH4C1 2) HC1, aq acetone 72 h  2) DBU, CH2C12, rt  1.:  79%  OH 1)  78  71%  87  86  C  1) MeLi, ether; H2O 2) PCC, CH2C12  C))/MgBr  0  0^67 CuBr•Me2S, THF, Me2S  76  88  68%  Scheme 12  elimination was effected by use of the base DBU. Reaction of 87 with MeLi and oxidative rearrangement of the resultant alcohol with PCC afforded the enone 76, which could be subjected to the now familiar cyclopentenone annulation sequence. It should be noted that the  20  only difference between Paquette's (1982) and Itii's (1983) strategies for the formation of the ketol 78 from the enone 76 relates to the use of two different Grignard reagents, 66 and 67. A last example (among the many others that exist) of elaboration of a bicyclic ketol is illustrated in Scheme 13. Koreeda and Mislankar, 30 in an approach towards the synthesis of the antitumor agent coriolin A (93), efficiently built the third ring of the natural product. After applying the annulation sequence, they obtained the ketol 91 in good yield. Dehydration was accomplished in a manner similar to those described previously.  C  °MgBr 0 6 7  CuBr•Me2S, THF, Me2S, -78 °C, 2 h;  aq HC1, THF, rt, 10 h  - 40 °C, 30 min; 0 °C, 5 min; aq NH4C1  89  90  92%  i o 92 Scheme 13 11.3.1. Preparation of the Ketol 63.  With the examples discussed above in mind, we were ready to attempt the annulation procedure using 3-methyl-2-cyclopenten- 1-one (64) as the starting material. Kuwajima et al. reported 31 an experimental protocol to achieve conjugate addition of Grignard reagents to enones in a relatively short reaction time and in high yield. This method was utilized to accomplish the desired transformation. Treatment of 3-methyl-2-cyclopenten- 1-one (64) with the Grignard reagent 67 in the presence of a catalytic amount of CuBr•Me2S, TMSC1 and 21  HMPA afforded, after an appropriate workup, the keto acetal 94 along with a small quantity of the diacetal 95 (See equation 2). The two compounds were very difficult to separate by flash  1) C 64  °>/\MgBr 0^67  CuBr•Me2S, THF, TMSC1, HMPA ^ •.-78 °C, 6 h; -45 °C, 2 h 2) CH3COOH 3) aq NH4C1-NH4OH  +  (c(2) o 0^/2 95  94  70%  chromatography. 32 Nevertheless, the pure keto acetal 94 could be isolated in 70% yield. This material exhibited all the expected spectral data including, in the it spectrum, a strong band at 1741 cm -1 for a carbonyl function characteristic of cyclopentanones. The 1 H nmr spectrum of 94 showed a tertiary methyl resonance at 8 1.04 and hydrogen signals associated with the  presence of the cyclic acetal (two broad ddd, 2H each, at 8 3.77 and 4.12, and a broad triplet integrating for 1 hydrogen at 8 4.52). The 1 H nmr spectrum of the diacetal 95 displayed signals specific to the acetal group at 8 3.76 (m, 4H), 4.10 (m, 4H) and 4.51 (t, 2H). The conjugate addition having succeeded, the keto acetal 94 was treated with dilute aqueous hydrochloric acid in THF (equation 3). Refluxing the mixture allowed the reaction to  +  94  (3)  ^ 63 ^96  ^  73%  23%  proceed in a reasonable length of time; after 18 hours, the reaction had gone to completion. Two bicyclic compounds resulted from the cyclization process: the fused bicyclic ketol 63 and  22  the bridged bicyclic ketol 96. Interestingly, each of the products was formed as a single isomer. The fused bicyclic ketol 63 displayed a broad band (3461 cm -1 ) in the it spectrum due to the hydroxyl group, along with a carbonyl absorption at 1735 cm -1 . The 1 H nmr spectrum of 63 displayed a doublet due to the angular hydrogen at 8 2.28, a hydroxyl signal, which exchanged upon treatment with D20, at 8 2.66 and a carbinol hydrogen signal at 8 4.52. The stereochemistry of the carbinolic center was ascertained by nuclear Overhauser enhancement (NOE) difference experiments. Saturation at 8 1.22 (tertiary Me) caused enhancement of the signal at 8 2.28 attributed to the angular hydrogen. Irradiation of the angular hydrogen signal at 8 2.28 increased the intensity of each of the signals at 8 1.22 (tertiary Me)  and 4.52 (CHOH). Saturation of the resonance at 8 4.52 (CHOH) caused enhancement of the angular hydrogen signal (8 2.28) and the hydroxyl signal (OH) at 2.66 ppm. These experiments confirmed the expected cis stereochemistry of the ring junction (it is well known that trans-fused bicyclo[3.3.0]octane are very strained and that cis-fused compounds are usually produced). These 1 H nmr spectral data also demonstrated the cis relationship between the vicinal angular and carbinol hydrogens. The formation of the product 63 as the only epimer can be rationalized as follows. In the isomer 63, the orientation of the hydroxyl group renders feasible the formation of a strong intramolecular hydrogen bond between the hydroxyl proton and the carbonyl oxygen. However, in the case of the other epimer, hydrogen bonding 23  cannot occur so easily. It is possible that the resultant difference in stability of the two isomers is large enough to cause the formation of only one of them since the conditions employed are presumably equilibrating. In fact, aldol condensations are known to be reversible. The bridged bicyclic ketol 96 was identified from the spectral data collected. The presence of the hydroxyl and carbonyl groups was confirmed by the it absorptions at 3413 and 1740 cm -1 . A 3-hydrogen singlet at 8 1.17 and a 1-hydrogen ddd at 8 3.84 ascertained the presence of a bridgehead methyl group and of a hydrogen adjacent to an oxygen (H-b). The value (11.5 Hz) of one of the vicinal coupling constants associated with the carbinol hydrogen showed that it is axially oriented and that the hydroxyl group therefore occupies the equatorial position. Decoupling experiments allowed the assignment of a few other signals of the 1 H nmr spectrum of 96 (See Table 1). Irradiation of the ddd at 3.84 (H-b) simplified the unresolved  Hh  96  multiplet at 8 2.55 due to H-a, along with the signals due to H-c and H-d. Irradiation of the unresolved multiplet at 8 2.55 (H-a) simplified the ddd with J = 3.5, 6, 11.5 Hz at 6 3.84 (Hb) into a dd with J = 6, 11.5 Hz and allowed the identification of H-h (ddd with J = 2.5, 6, 12 Hz at 8 1.86). H-g has a dihedral angle with H-a close to 90° and therefore does not show any coupling with H-a (J 0 Hz). Saturation of the ddd (J = 2.5, 6, 12 Hz) at 8 1.86 (H-h) changed the dd (J = 3.5, 12 Hz) at 8 1.54 (H-g) into a d, J = 3.5 Hz. The unresolved m at 2.55 (H-a) was also modified into a broad singlet.  24  Table 1: 1 H nmr Data (400 MHz, CDC13) for the Bridged Bicyclic Ketol 96: Decoupling Experiments.  Hh  96  Signals Being Observed  Signal Being Irradiated Assignment H-xa H-a  8 ppm (initial mult., J(Hz), H-x) to mult. after irradiation, J(Hz)b  1 H nmr (400 MHz) 8 ppm  (mult., J(Hz))  2.55 (unresolved m)  1.86 (ddd, J = 2.5, 6, 12, H-h) to dd, J = 2.5, 12. 3.84 (ddd, J = 3.5, 6, 11.5, H-b) to dd, J = 6, 11.5.  H-b  3.84 (ddd, J = 3.5, 6, 11.5)  1.24-1.37 (m, H-c or H-d) to sharpened m. 1.99-2.17 (m, 4H, includes H-c or H-d), part of the m is modified. 2.55 (unresolved m, H-a) to br d, J = 6.  H-c or 1.24-1.37 (m) H-d  1.56-1.62 (m, 2H), the m is modified. 1.99-2.17 (m, 4H, includes H-c or H-d), part of the m is modified. 3.84 (ddd, J = 3.5, 6, 11.5, H-b) to unresolved m.  H-h  1.86 (ddd, J = 2.5, 6, 12)  1.54 (dd, J = 3.5, 12, H-g) to d, J = 3.5. 2.55 (unresolved m, H-a) to br s.  H-g  _ 1.54 (dd, J = 3.5, 12)  a- Irradiated hydrogen. b- Only the hydrogens for which changes in their signals could be unambiguously seen are recorded.  25  11.3.2. Preparation of the Enone 62 via the Mesylate 97. In order to obtain the desired enone 62, the fused bicyclic ketol 63 was allowed to react with MeS02C1 in the presence of triethylamine (Scheme 14). The keto mesylate 97 thus obtained underwent elimination of the elements of MeS03H upon treatment with DBU. This straightforward succession of reactions afforded the enone 62 in 74% yield from the ketol 63. HO  H^0  H Ms() CH3S02C1^. .. Et3N, CH2C12, 0 °C^ 1....  DBU CH2C12, rt  ,  63  ^  97  ^  Scheme 14  Me 62  74% from 63  A pure sample of the keto mesylate 97 was characterized by spectroscopic methods, which indicated the conversion of the hydroxyl group to a methyl sulfonate moiety. The appearance in the 1 H nmr spectrum of a 3-hydrogen singlet at 8 2.99, attributed to MgS03, and of a ddd at 8 5.25, assigned to the hydrogen adjacent to the methanesulfonate moiety, also proved that the desired transformation had been accomplished. Evidence for the formation of the enone 62 was revealed by the presence, in the it spectrum, of an absorption for a conjugated carbonyl at 1714 cm -1 and a C=C absorption at 1635 cm -1 . The 1 H nmr spectrum showed a 1-hydrogen dd at 8 6.44, caused by the vinyl hydrogen. 11.3.3. Preparation of the Keto Alkene 60. We now had in hand the enone which would allow us to verify the validity of the synthetic route suggested. A few cases of conjugate addition made on similar enones were known and two of these cases have been encountered at the beginning of this section in 26  connection with a discussion of the syntheses of (±)-silphinene  (79).26,29  These two  examples are displayed in equation 4. Thus, conjugate addition of the reagents 66 and 67 to the enone 76 afforded exclusively the adducts 77 and 88, in which the side chain had been introduced stereoselectively, cis to the angular hydrogen (equation 4).  1)(6 y\MgBr 3i  0 66 or 67^vsCuX 2) Workup  76  ^  (4)  n = 1 77 n = 2 88  In a previous synthetic approach to (±)-silphinene, Paquette and Leone-Bay had shown 29 a that conjugate addition of the reagent 67 to the enone 98 produced one compound (99), albeit in low yield (equation 5). The stereochemistry of the newly introduced stereogenic  center was not determined.  1  (  )  0.,r7\ Mg Br  C.,0 67 ,.  ,,,,,  THE  98  ^  (5 )  99  41%  These examples constitute reasonable models for the conversion of the a,(3-unsaturated ketone 62 into a 1,4-addition adduct. Indeed, treatment of the enone 62 with the cuprate reagent 18, prepared as described previously, yielded one compound which was subsequently  27  determined to be the keto chloride 100 (equation 6). The frequency of the carbonyl absorption  CI 0  (6) 62  ^  100  (1735 cm -1 ) in the it spectrum of 100 indicated that the compound possessed a non conjugated ketone function. In the 1 H nmr spectrum (400 MHz, CDC13 as solvent) of 100, two signals accounting for one hydrogen each at 8 4.88 and 5.00 confirmed the presence of the two vinylic hydrogens of the side chain. The two-hydrogen multiplet at 8 3.63-3.75 (CH2C1) also showed that the 1,4-addition had proceeded. However, at this point, conclusive proof for the relative stereochemistry of 100 was not obtained. Therefore, this material was subjected to reaction conditions that would promote intramolecular alkylation. If treatment of the conjugate addition product with a base would lead to an angularly fused triquinane, it would provide a strong indication that the 1,4-addition had occurred on the R face as shown in 100. Cyclization would yield an all cis-fused tricyclic product (60) (Scheme 15). On the other hand, if the newly introduced chain was situated on the a face, as in the adduct 101, a highly strained triquinane 102, in which one of the two ring junctions is trans-fused, would result. Consequently, in the case of the a isomer 101, it is more likely that other types of compounds would be formed. The diene 103, obtained from elimination of the elements of HC1, or the tricyclic ketone 104, which should be differentiable from the tricyclic ketone 60 by 1 H nmr spectroscopy, represent plausible products (Scheme 15) .  28  p0  CI^H  60  100 Cle  H ^X^-0101  ^  103  102  ^  104  Scheme 15 The keto chloride 100 was subjected to the reaction conditions (KH, THY, room temperature) described previously. 8 b , f Since no major transformation had occurred after stirring the reaction mixture at room temperature for a few hours (glc and tic showed mainly starting material), the reaction mixture was heated (equation 7). After workup and purification, the 1 H  +  100  ^  105  29  ^  (7) 60  nmr spectrum of the material showed that two products, which were very difficult to separate by flash chromatography on silica gel, had formed (ratio: —2 : 1). Each substance exhibited by 1 H nmr spectroscopy (400 MHz, CDC13) a singlet for an angular methyl group (at 8 1.08 for the minor product and at 8 1.26 for the major product). The presence of a series of vinylic hydrogen signals indicated that the major product was one in which elimination of hydrogen chloride had occurred. This material was assigned structure 105. A dd, integrating for 1H at 8 6.36 with J = 11, 17.5 Hz, revealed that this hydrogen (H e) was coupled to two others in a cis and trans fashion. Two 1-hydrogen doublets, with J = 11 and 17.5 Hz, at 8 5.10 and  5.49 were assigned to Ha and Hb, respectively. Finally, two 1-hydrogen singlets at 5 5.03 and 5.08 were attributed to Hd and He . The minor product, which was subsequently shown unambiguously to be the tricyclic keto alkene 60, possessed two vinylic hydrogens which gave rise to two singlets (8 4.75 and 4.87) in the 1 H nmr spectrum. A broad doublet due to the angular hydrogen at 8 2.93 with J = 10 Hz was also visible.  In order to collect more information about the generation of the two products, the keto chloride 100 was subjected to a number of different reaction conditions. In these reaction conditions, the two substances previously described were formed in various ratios. When the keto chloride was treated with t-BuOK either in t-BuOH or in t-BuOH/THF at —25-30 °C, the tricyclic compound 60 was the major one formed. However, due to the difficulty in separating the two products, it was desirable to find conditions that would lead specifically to the angularly fused triquinane. Consequently, a "one-pot" procedure for the direct conversion of 62 into 60 was attempted.  The enone 62 was allowed to react with the lower order heterocuprate 18 in THE at -78 °C (Scheme 16). The resultant enolate 106 was set to undergo intramolecular alkylation. At this stage, the reaction mixture was treated with dry HMPA and then was warmed to room temperature. It was gratifying to find that the tricyclic keto alkene 60 could be isolated in 73% 30  yield along with 12% of the uncyclized keto chloride 100, identical by 1 H nmr spectroscopy with the compound previously described.  CI HMPA  Cu(CN)Li 18^1....  62  ^  106  ^  60  73%  Scheme 16  A number of factors were crucial for the success of the annulation procedure. Firstly, the sequence had to be accomplished in one operation. It was shown (vide supra) that the intramolecular alkylation was not clean if carried out on the 1,4-adduct 100. Since kinetic deprotonation would yield the undesired enolate anion 107, conditions had to be used to allow  106  ^  107  equilibration between the two possible enolates 106 and 107. However, the formation of 106 is disfavoured due to the strain associated with placing a double bond at the C-1/C-2  position of bicyclo[3.3.0]octanes. Indeed, it is known 33 that bicyclo[3.3.0]oct-l-enes are less stable than the 2-enes because of angle strain. It is thus understandable that, even under  31  equilibrating conditions, the concentration of 106 in solution would be very low and that, therefore, the intramolecular alkylation of 100 to give 60 is a sluggish process. Secondly, in order to get good yields of the keto alkene 60, all the reagents and substrates had to be carefully purified and distilled immediately prior to use. The MeLi solution utilized in the transmetallation process had to be taken from a recently opened bottle. The reactions carried out on scales smaller than 500 mg of the enone 62 worked better than those carried out on larger scales. The stereochemical outcome of the conversion of 62 into 60 also requires comment. It appears to be well established that stereoelectronic factors play a key role in regulating the stereochemistry of conjugate addition of cuprate reagents to cyclic a,13 unsaturated ketones. -  This implies that the transition states for such additions are product-like (i.e. enolate-like) in shape and that the developing bond at the ri carbon of the enone system is created, as nearly as possible, in a direction perpendicular to the plane of the forming enolate anion. Molecular models show that, of the two possible enolate anions (106 and 108) that could result from the  106  108  reaction of 62 with 18, only 106 can comfortably adopt a conformation such that the newly introduced side chain is attached to the ring system in an orientation (nearly) perpendicular to the plane of the adjacent enolate double bond. Consequently, the conjugate addition of the cuprate 18 takes place preferentially cis to the angular methyl group, even though this is the more hindered face of the enone system. 32  The spectroscopic data gathered were consistent with the formation of the angularly fused tricyclic ketone 60. The ketone and alkene functionalities of the product 60 were indicated by the absorptions at 3071, 1732 and 1656 cm -1 in the it spectrum. The 1 H nmr spectrum (in CDC13) of this compound was not overly informative since a lot of the signals overlapped. However, the hydrogen signals in the 1 H nmr spectrum taken in benzene were nicely resolved (Figure 1). Decoupling experiments (see Table 2), along with measurement of vicinal coupling constants, allowed the identification of most of the 1 H nmr signals. Irradiation of one of the vinyl hydrogens (H-15) at 8 4.75 modified three signals, which were easily attributed to the three allylic hydrogens H-1, H-10 and H-10'. The two hydrogens (H11 and H-11') vicinal to H-10 and H-10' were assigned after irradiation of H-10'. Similarly, H-2 and H-2' were identified easily since saturation of H-1 modified each of their signals. Irradiation of the ddd with J = 4, 9.5 and 13.5 Hz at 8 1.19 (H-5) changed each of the two ddd at 8 1.97, J = 9.5, 9.5, 18.5 Hz and 2.09, J = 4, 9.5, 18.5 Hz into a dd. The large geminal coupling constant (18.5 Hz) and the chemical shifts indicate that the latter two resonances are due to the hydrogens adjacent to the carbonyl group (H-6 and H-6'). The geminal partner of H-5 (H-5') is included in the 4 H signal at 1.25-1.40. These spectroscopic data strongly suggested that the desired angularly fused triquinane had been synthesized. The 1 H nmr spectrum (benzene) undoubtedly dismisses the possibility that the acquired material was compound 104. Moreover, the eventual acquisition of two target molecules showed the assignment to be correct. 11.3.4. Preparation of the Ketone 59.  The assembly of the tricyclic structure was achieved in a short sequence of steps. Completion of the synthesis of one of the target molecules (13, 14, 55 or 56) required transformation of the exocyclic alkene function of 60 into a methyl group with the correct stereochemistry.  33  Figure 1: The 1 H nmr Spectrum (400 MHz, C 6 D 6 ) of the Keto Alkene 60.  Table 2: 1 H nmr Data (400 MHz, C6D6) for the Keto Alkene 60a: Decoupling Experiments.  12 60  _ Signal Being Irradiated Assign1 H nmr (400 MHz) 8 ppm ment (mult., J (Hz)) H-x H-1  2.83 (dm, J for d = 9.5)  Signals Being Observed 8 ppm (initial mult., J (Hz), H-x) to mult. after irradiation, J (Hz)b 1.43-1.53 (m, H-2) to sharper m 1.77 (dddd, J = 9.5, 9.5, 9.5, 13.5, H-2') to ddd, J = 9.5, 9.5, 13.5. 4.75 (m, H-15) to sharper signal. 4.89 (m, H-15') to sharper signal.  1.97 (ddd, J = 9.5, 9.5, 18.5, H-6) to dd, J = H-5 1.19 (ddd, J = 4, 9.5, 13.5) ^9.5,^18.5. 2.09 (ddd, J = 4, 9.5, 18.5, H-6') to dd, J = ^9.5,^18.5. H-10'e  2.64 (dddm, J for ddd = 7.5, 1.25-1.40 (m, 4H, includes H-11); part of the m is modified. 7.5, 15) 1.86 (ddd, 1H, J = 7.5, 7.5, 13, H-11') to dd, J = 7.5, 13. 2.22 (dddm, 1H, J for ddd = 7.5, 7.5, 15, H10) to unresolved m. 4.75 (m, H-15) to sharper signal. 4.89 (m, H-15') to sharper signal.  H-15  4.75 (m)  2.22 (dddm, 1H, J for ddd = 7.5, 7.5, 15, H10) to sharper signal. 2.64 (dddm, 1H, J for ddd = 7.5, 7.5, 15, H10') to sharper signal. 2.83 (dm, J for d = 9.5, H-1) to sharper signal.  a- Silphiperfolane numbering used for consistency b- Only the hydrogens for which changes in their signals could be unambiguously seen are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-10' is more downfield than H-10).  35  ^  As mentioned before, simple hydrogenation of 60 would undoubtedly proceed stereoselectively in the wrong sense to give, at least primarily, the C-9 epimer of 59. Therefore, the reductive conversion of the keto alkene 60 into 59 was carried out via a reaction sequence in which the correct configuration at C-9 was established by equilibration of the two epimeric aldehydes 110 and 111. At the outset, there appeared to be little doubt that the aldehyde 110, possessing the CHO function in an exo orientation, would be thermodynamically more stable than the corresponding endo isomer 111. Indeed, the resultant expectation that the equilibrium between 110 and 111 would favor the former isomer turned out to be correct. Treatment of 60 with an excess of borane 34 in THF, followed by the usual oxidation step, gave a mixture of the diols 109 in 96% yield (Scheme 17). Direct Swern oxidation 35 0^0 II^II ,C ,^,C OH^ H "^0 H 0^ 1) BH3•THF,^1) DMSO, (C0C1)2, H ,,^H, HO^  THF, rt,^ CH2C12, -78 °C  +  2) Et3N, 0 °C 3) Me0Na, Me0H, rt  60  109  96%  Scheme 17  110 80% after two recyclings  111  of the diols afforded a mixture of the keto aldehydes 110 and 111. Equilibration of this material with sodium methoxide in methanol produced an 8 : 1 mixture of the two epimers, with predominance of the desired aldehyde 110. The two compounds could be separated by flash chromatography and the unwanted isomer 111 could be recycled. After two such recycling procedures, the keto aldehyde 110 was obtained in 80% yield. The success of the oxidation process was witnessed by the appearance of a carbonyl stretch at 1729 cm -1 and a C-H stretching band at 2713 cm -1 in the it spectrum of the aldehyde 110. A resonance at 8 9.58 in the 1 H nmr spectrum evidenced the presence of an aldehyde hydrogen. 36  Spectroscopic data also confirmed the formation of the keto aldehyde 111. The 1 H nmr spectrum showed the expected signal at 8 9.78 due to the hydrogen of the aldehyde group. Conversion of 110 into the ketone 59 was accomplished in a straightforward manner. Treatment of 110 with 1,2-ethanedithiol in the presence of boron trifluoride etherate 36 afforded the crystalline dithioacetal 112 (Scheme 18). The 1 H nmr spectrum of this material exhibited a 4-hydrogen multiplet at 8 3.14-3.28 and a 1-hydrogen doublet at 8 4.54 accounting for the presence of the dithioacetal group. Desulfurization of the dithioacetal 112 with Raney nicke1 37 produced, in 63% overall yield from 110, the required tricyclic ketone 59. The 1 H nmr spectrum of 59 showed a 3-hydrogen doublet at 8 0.97 (MeCH). 0^  I I ^C ,, C  S  L,H  is,^ " , ^II 1,2-ethanedithiol,^H,, BF3.0E12,  H H ,,,  S  -----1  ,  Raney nickel, Acetone, rt  CH2C12, 0 °C  110  ^  112  59 63%  Scheme 18  The desulfurization reaction had to be executed under mild conditions to avoid overreduction of the ketone function of 59. Reflux of a solution of 59 in ethanol with Raney nickel caused desulfurization along with reduction of the carbonyl group. When the desulfurization reaction was carried out at room temperature, the amounts of over-reduced product were decreased, but the reaction yields were still relatively low. Acetone proved to be a suitable solvent for this reaction. The quality of the Raney nickel also affected the yield of the desulfurization reaction. The amounts of side products could be minimized by utilizing a solution of freshly prepared Raney nickel. The reagent had to be aged for at least two days prior to use otherwise it was too 37  active and caused the formation of over-reduced products. It also had to be utilized within 30 days of its preparation. Under these conditions, the yields of the desulfurization process varied from 62-74%. 11.3.5. Preparation of the a,(3-Unsaturated Keto Ester 57. The ketone 59 was treated with KH in THF, and the resultant potassium enolate was  allowed to react with dimethyl carbonate. This sequence of reactions provided the keto ester 58 in 95% yield (Scheme 19). Compound 58 was converted into the a,(3-unsaturated keto Me,,,  Me,,,,  1) KH, THF, A H,,, 2) Me0CO2Me,  1) KH, THF^H,,, 2) PhSeC1  COOMe  A 3) dilute HC1  59  COOMe  3) H2O2, H20, CH2C12  58  57  95%  90%  Scheme 19  ester 57 according to the procedure of Reich and coworkers. 38 Selenenylation of the potassium enolate derived from 58 was accomplished with PhSeC1 (Scheme 19). Oxidation of the intermediate seleno ketone with H202 and selenoxide elimination gave 57 in 90% yield. The it spectrum of 57 revealed three absorptions at 1752, 1719 and 1622 cm -1 which attested that the desired conversion had succeeded. The 1 H nmr spectrum of 57 exhibited three singlets at 8 8.04 (vinylic hydrogen), 3.84 (COOMe) and 1.17 (angular Me) and a doublet (MeCH) with J = 6 Hz at 8 1.00. 11.3.6. Attempts to Prepare (±)-Methyl Cantabradienate (55) and (±)-Methyl Cantabrenonate (13).  In order to achieve the synthesis of the simplest of the four methyl ester derivatives, methyl cantabradienate (55), the ketone group of 57 had to be transformed into an exocyclic 38  ^  ^H  alkene function. A variety of methylenation reagents are known and could lead to the desired product. The simplest of the Wittig reagents, methylenetriphenylphosphorane (113), 39 has seen a great deal of use for the conversion of aldehydes and ketones into alkenes. Consequently, the synthetic conversion of 57 into 55 was attempted with this reagent (equation 8). However, the reaction did not yield any of the desired product. Instead, recovered starting material was isolated. The carbonyl group of 57 is very hindered and therefore, the reaction with the phosphorane 113 is very sluggish. M e „,,^O^  Me,,  H ^  Ph3P=CH2 113 a.COOMe ^ X  57  H  s•  COOMe^(8)  55  Thus, a different method for the one carbon homologation of the ketone function of 57 was tried. The a,[3-unsaturated keto ester 57 was treated with the highly electrophilic reagent "BrZnCH2TiC13" 114 which was prepared from zinc dust, CH2Br2 and TiC14 40 (equation 9). Me„,  Me„,^O^ ,^  CH2Br2, TiC14,^H,,, Zn, THF-CH2C12  Other COOMe +^ products ( 9)  COOMe ^Jo57  55  (traces)  A series of products (none of which seemed to be a major one) were formed as witnessed by glc and tic analyses of the crude reaction mixture. After chromatographic separation on silica gel, a very small amount of an impure methyl cantabradienate (55) was obtained. The 1 H nmr spectrum of this impure fraction exhibited three 1-hydrogen singlets at 6 6.68, 5.63 and 4.96. Two singlets (3 H each) at 6 3.77 (COOMe) and 1.04 (angular Me) along with a doublet 39  (CH3CH) at 8 0.99 were also present. These data compared very favorably with those reported by San Feliciano and coworkers for methyl cantabradienate (55). 12 However, the reaction conditions did not allow the isolation of pure 55 in good yield and another methylenation method was attempted. The ester 57 was allowed to react with 1 equiv of methyllithium at -78 °C (equation 10). Various products resulted from this reaction. One of them was ascribed structure 115 on the basis of 1 H nmr spectroscopic data. Three singlets, at 8 8.00 (vinylic hydrogen), 2.48 (NRC=0) and 1.16 (angular Me) and a doublet at 5 1.00 (MeCH) supported this assignment. It seems that the hindered nature of the carbonyl function coupled with the reactive oc,(3unsaturated keto ester moiety favor side reactions among which is the addition of MeLi to the ester group of 57. ^Me.,,,^O^ ^H  s,  „  Me.,,,  ^H , 1) MeLi, THF, -78 °C^ Other COOMe 2) sat. aq NH4C1 a.^+^(10) ^products -  57  ^  115  In order to decrease the reactivity of the intermediate 57, the ester group of 57 was hydrolized to the acid function. Thus, consecutive treatment of 57 with aqueous NaOH and aqueous HC1 provided the oc,(3-unsaturated acid 116 in good yield (equation 11). Reaction of  1) 10% aq NaOH, THF  COOMe 57  2) 10% aq HCI^BM.-  ^  COOH^(11) 116  >90%  40  the acid 116 with methyllithium at -78 °C, followed by acidic workup (Scheme 20), furnished a mixture of products which were difficult to identify unambiguously by 1 H nmr spectroscopy. Consequently, this crude material was allowed to react with diazomethane 41 (Warning: toxic and explosive 42 ). Chromatographic separation allowed the isolation of a few  impure compounds. Two substances were tentatively assigned structures 117 and 118, respectively, on the basis of 1 H nmr spectroscopic data. Me„,  1) MeLi, THF, H,, -78 °C /2) 10% aq HC1 3) CH2N2  Me„, H,,  HO^ Me  ,  HIg, COOMe +  COOMe +  ^11.-  other products  Me  118  117  ,  Me„,  COON  PCC,  CH2C12  116 Me„,  1) CH2131'2,^H,,, Zn, TiC14 )... 2) CH2N2  COOMe  COOMe Me  119  13 Scheme 20  The structure of 118 was proven by carrying out a chemical transformation and correlating the spectral data thus obtained with those of a known compound. The mixture containing 118 was treated with PCC in CH2C12 (Scheme 20). Oxidative rearrangement of the tertiary alcohol 118 occurred and methyl cantabrenonate (13), one of the target molecules, was isolated after chromatographic separation. In a last effort to synthesize methyl cantabradienate (55) through this route, the acid 116 was successively treated with the reagent derived from Zn dust, CH2Br2 and TiC14 and  41  with diazomethane (See above, Scheme 20). The reaction was extremely messy. One of the compounds, the keto ester 119 was tentatively identified via its 1 H nmr spectrum It is obvious from the experiments described above that the plan towards the syntheses of the target molecules 13, 14, 55 and 56 needeed to be revised. A modified route was attempted which involved, as intermediates, molecules containing an acetal moiety in place of an ester group (Scheme 21). This modification should alter the reactivity of the various compounds and prevent the types of side reactions encountered earlier. However, this pathway was eventually abandoned due to the low yields and difficulties associated with a few of the reactions and to the instability of some of the intermediates. 0  Me,,, 1) KH, THE H,, 2) HCO2Et,  DDQ, PhH CHO Collidine, rt  ,  CHO  A 3) dilute HC1  59 H  p-TsOH, PhH A  Me,,,  Y  1) MeLi, THE^H , -78 °C 2) sat. aq NH4C1 1PCC, CH2C12, rt  CHO^- - - - 41 - 13 ."  Scheme 21  42  A route which would allow the introduction of the C-14 methyl group before the methyl ester group seemed to be a promising alternative. A plausible retrosynthetic plan is illustrated in Scheme 22. Methyl cantabrenonate (13) and methyl epoxycantabronate (14) could originate from a common precursor, the enone 120. This enone 120 would be obtained by oxidative rearrangement of the tertiary alcohol 121, which would be readily prepared from the ketone 59. In practice, this synthetic scheme turned out to be viable and resulted in the syntheses of the two target compounds, 13 and 14. The description of the last part of these syntheses follows.  13 Scheme 22  11.3.7. Preparation of the Enone 120.  The ketone 59 was converted into the enone 123 via the procedure of Saegusa and coworkers (Scheme 23). 43 Successive treatment of 59 with LDA and trimethylsilyl chloride gave the silyl enol ether 122. 44 Exposure of 122 to Pd(OAc)2 in CH3CN afforded the enone 123 in 88% yield from 59. The 1 H nmr spectrum of 123 indicated the presence of two vinylic hydrogen signals at 8 6.08 and 7.30. Interestingly, this enone has also been prepared by Paquette and coworkers, 23 a and Kakiuchi and coworkers, 23 h although via different approaches. The spectral data (derived from infrared and 1 H nmr spectroscopies) obtained for 123 synthesized via our route compared favorably with those reported by Paquette and  coworkers.23a  43  Me,„,  •^  OSiMe3 Pd(OAc)2, 1) LDA, THF, H., CH3CN, rt -78 °C^ 2) Me3SiC1  59  123  122  88% from 59 1) MeLi, THF, -78 °C 2) sat. aq NH4C1  OH -.K^  PCC, CH2C12  121  96%  Scheme 23 Reaction of 123 with MeLi in THF provided a single, crystalline tertiary alcohol 121 of unassigned stereochemistry (Scheme 23). The ir spectrum (CHC13) of 121 showed the expected hydroxyl absorptions (3608 and 3536-3374 cm -1 ). The 1 H nmr spectrum displayed a 3-hydrogen singlet at 8 1.29 (tertiary methyl) and two vinylic hydrogen doublets at 8 5.39 and 5.54. Oxidation of 121 with pyridinium chlorochromate 45 in dichloromethane produced 120, a 14-carbon tricyclic enone possessing the required configuration at each of the four chiral centers. The ir spectrum of 120 revealed stretching bands associated with =C-H, carbonyl and C=C moieties at 3065, 1703 and 1616 cm -1 respectively. In the 1 H nmr spectrum, the lowest field signal (8 5.79) was attributed to the vinyl hydrogen, and a doublet at 8 2.08 was assigned to the vinylic methyl group. The enone 120 (in enantiomerically pure form) had been previously reported by San Feliciano and coworkers. 12 They prepared the ketone 120 by degradation of (-)-methyl cantabrenonate (13). Hydrolysis of ( ) 13 and subsequent -  -  decarboxylation of the resultant acid provided the norsesquiterpenoid ketone 120.  44  11.3.8. Completion of the Syntheses of (±)-Methyl Cantabrenonate (13) and (±)-Methyl Epoxycantabronate (14). Conversion of the intermediate 120 into (±)-methyl cantabrenonate (13) required a "simple" methoxycarbonylation of the former substance at C-6 (silphiperfolane numbering). Unfortunately, all attempts to effect this transformation directly met with failure. Treatment of 120 with different base-electrophile combinations (e.g. KH-Me0CO2Me, LDA-NCCO2Me 46 ) under a variety of conditions produced 125 as the only C-methoxycarbonylation product (Scheme 24). Thus, the dienolate 124 prefers to react with Me0CO2Me or NCCO2Me at the y carbon rather than at the a carbon. The reason(s) underlying this preference is (are) not immediately obvious. In any case, these results required that an alternative protocol be developed for the conversion of 120 into 13. The transformation of the enone 120 into (±)methyl cantabrenonate (13) is described below. COOMe  0 0^ OM^ ^ 124^ 125 120 M = K or Li  Scheme 24 Hydrogenation (H2, Pd-C, EtOAc) of 120 gave, in 98% yield, a single tricyclic ketone that was shown to possess structure 126 (equation 12). Molecular models indicate that, in  H2, Pd-C  EtOAc, rt, ),  (12)  22 psi  120 45  terms of steric hindrance, the two faces of the carbon-carbon double bond in 120 are similar. Although one might expect some preference for exo hydrogenation to produce 126, the highly stereoselective nature of this reaction was surprising. Since previous studies 22 in triquinane chemistry have shown that, in some instances, the stereochemical outcomes of reactions involving functionalized tricyclo[6.3.0.0 1,5]undecane systems are difficult to predict, it was of interest to determine the configuration of C-7 in 126. This was accomplished by means of 1 H nmr spectroscopy (Table 3). Table 3: 1 H nmr Data (400 MHz, CDCI3) for the Ketone 126a.  H-x  1H nmr (400 MHz) 8 ppm (mult., J (Hz))  COSY Correlationsb  NOE Correlationsb  H-1  1.89 (br t, J = 8)  H-2, H-2', H-9  Me-14, Me-15  H-2 H-2'c  Part of the m (2H) at 1.15-1.29 —1.33-1.40 (part of the m (4H) at 1.33-1.59)  H-1, H-2', H-3, H 3' H-1, H-2, H-3, H-3'  H-3'  1.94 (dd, J = 6, 12)  H-2, H-2', H-3, Me-12  H-6  1.77 (dd, J = 13, 18)  H-6', H-7  H-6'  2.29 (dd, J = 7, 18)  H-6, H-7  H-7  2.15-2.21 (m)  H-6, H-6', Me-14  Me-12  0.95 (s)  H-3'  Me-14  1.11 (d, J = 7)  H-7  Me-15  1.03 (d, J = 6.5)^_ H-9  Me-14  H-1, H-7 H-1  a- Silphiperfolane numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  46  A homonuclear shift correlation spectroscopy (COSY) experiment, along with appropriate decoupling experiments, allowed identification of the 1 H nmr signals due to H-1, Me-14, and Me-15 (Table 3). Comparison of the multiplicity pattern and chemical shift of H-1 with the data obtained for previous intermediates and compounds of related structures 12, 18-21,23 also confirmed the assignment of this hydrogen. This information set the stage for the key NOE difference experiments. As can be seen from the data presented in Table 3, irradiation of the signals at 8 1.89 (H-1) and 1.11 (Me-14) caused mutual en-  hancement of these two resonances. These results are possible only if Me-14 has the orientation shown in formula 126. Thus, in the hydrogenation of 120, hydrogen was delivered exclusively to the exo face of the carbon-carbon double bond. A recent report on the syntheses of (±)-5-oxosilphiperfol-6-ene (47) and (±)silphiperfol-6-ene (46) by Kakiuchi and coworkers has demonstrated 23 h that 1,4-addition of lithium dimethylcuprate to the enone 127 affords a single ketone 128 in 98% yield (Scheme 25). The configuration of the newly formed center of 128 was not determined. Comparison Me„ ^Me„, ,  H,  H  Me2CuLi, Et20,^' 0 °C to rt^1...Me 0  127  128  47  98%  Scheme 25  of the 1 H nmr spectroscopic data provided for the ketone 128 with those acquired for the intermediate 126 clearly reveals that the two compounds are epimeric at C-7. 128 displays, in its 1 H nmr spectrum (90 MHz, CDC13), methyl signals at 8 0.96 (d, 3H, J = 6.4 Hz), 0.98 (d, 3H, J = 6.2 Hz) and 1.01 (s, 3H). A series of multiplets from 8 1.1 to 2.6 accounts for the remaining 13 H. In the 1 H nmr of 126, the signal at 5 0.95 is a singlet attributed to the angular methyl group. The secondary methyl groups appear at 8 1.03 and 1.11. 47  Consequently, it may be concluded that the conjugate addition of lithium dimethylcuprate to the enone 127 proceeds from the exo face of the carbon-carbon double bond, to yield, stereoselectively, compound 128. Methoxycarbonylation of 126 via the procedure of Mander and Sethi 46 provided the keto ester 129 (Scheme 26). Sequential treatment of 129 with LDA and benzeneselenenyl bromide 38 gave a mixture of the epimeric phenylseleno ketones 130 and 131 (-1 : 1), which were readily separable by flash chromatography. Not unexpectedly, oxidation  38  of 130 did  not provide acceptable yields of (±)-methyl cantabrenonate (13). However, treatment of 130 with PhSeNa in THF-HMPA produced, efficiently, the keto ester 129 and thus, 130 could be recycled. In this manner, after two recycling operations, the overall yield of 131 from the keto ester 129 was 76%. Me„,  Me„, ^Me„, Me H^1) LDA, THF, H -78 °C ,  COOMe  2) HMPA 0 NCCO2Me 83%  126  129  Me 0 ...COOMe  COOMe  -78 °C  2) PhSeBr  0  SePh  /  130  PhSeNa, THFHMPA  Me„,  Me  1) LDA, THF, H  Me,,,  Me„,,  Ozone,  H202, Hi' NaOH, Me0H H2O  COOMe  14  13  85%  77%  ,  0  Me  H,  SePh  CH2C12, -78 °C  'COOMe  131  0  76% after  recycling  Scheme 26  Oxidation of the intermediate 131 with ozone 38 afforded crystalline (±)-methyl cantabrenonate (13) (Scheme 26). The it spectrum of 13 showed absorptions at 3022, 1737, 1708 and 1615 cm -1 which confirmed the presence of the ketone, ester and carbon-carbon 48  double bond moieties. The 1 H nmr spectrum exhibited signals for the methyl ester at 8 3.83 and for the vinyl methyl group at 5 2.36 (Figure 2). The synthetic product displayed 1 H nmr and 13 C nmr spectra identical with those derived from esterified natural cantabrenonic acid. 12,47 Treatment of 13 with alkaline hydrogen peroxide provided a single product, (±)-methyl epoxycantabronate (14) (Figure 3). The highly stereoselective nature of this reaction shows once again that reagents prefer to approach the carbon-carbon double bond of silphiperfol-6-en5-ones from the exo face of the planar enone system. The synthetic product 14 exhibited 1 H nmr and 13 C nmr spectra identical with those derived from esterified natural epoxycantabronic acid (Figure 4).1247 11.4. CONCLUSION.  The work described in this section of the thesis establishes a new approach to the construction of silphiperfolane sesquiterpenoids and culminated in the total syntheses of two highly oxygenated members of this family of natural products. The key step of the overall synthetic sequence was the completely stereoselective conversion of the bicyclic enone 62 into the tricyclic keto alkene 60. This methylenecyclopentane annulation was readily accomplished via a one-pot process involving reaction of 62 with the novel bifunctional cuprate reagent 18. A summary of the syntheses of (±)-methyl cantabrenonate (13) and (±)-methyl epoxycantabronate (14) is displayed in Scheme 27.  49  Figure 2: The 1 11 nmr Spectrum (400 MHz, CDC1 3 ) of Synthetic (±)-Methyl Cantabrenonate (13).  Figure 3: The 1 H nmr Spectrum (400 MHz, CDC1 3 ) of Synthetic (±)-Methyl Epoxycantabronate (14).  Figure 4: The 1 H nmr Spectrum (400 MHz, CDC1 3 ) of (-)-Methyl Epoxycantabronate (14) from San Feliciano et al.  c, d  a  94  64  120  0 ^  121  ^  I  f  g, h  i, j  k  60  62  63  59  ^  110  11, m  P CO2 Me —0-  0  0 129 R = CO2Me  0 130 R = CO2Me R' = SePh 131 R = SePh R' = CO2Me  (±)-14  0  (a) Reagent 67 (1.3 equiv), CuBr•Me2S (0.08 equiv), Me3SiC1 (2 equiv), HMPA (2 equiv), THF, -78 °C, 6 h; - 45 °C, 2 h; rt, 15 min; add HOAc and aq NH4C1-NH4OH (pH 8-9), 70%; (b) 10% aq HC1, THF, reflux, 18 h, 73%; (c) MeS02C1, Et3N, CH2C12, 0 °C, 1 h; (d) DBU, CH2C12, rt, 1.5 h, 74% from 63; (e) reagent 18 (1.0 equiv), THF, -78 °C, 1.5 h; - 48 °C, 2.5 h; add HMPA and warm to rt over 1.5 h, then stir for 0.75 h, 73 %; (f) BH3 (2 equiv), THF, rt, 2 h; NaOH, H202, H2O-Me0H; (C0C1)2, DMSO, CH2C12, -78 °C, 1 h; Et3N, 0 °C, 1 h; Me0Na, Me0H, rt, 77%; (g) HSCH2CH2SH, BF3.OEt2, CH2C12, 0 °C, 1 h, 85%; (h) Raney nickel (W4), acetone, rt, 1.5 h, 74%; (i) LDA, THF, -78 °C; Me3SiC1, -78 °C to rt, 1 h; Pd(OAc)2, CH3CN, rt, 3.8 h, 88%; (j) MeLi, THF, -78 °C, 1.5 h, 96%; (k) PCC, CH2C12, rt, 1 h, 88%; (1) H2 (22 psi), Pd-C, EtOAc, rt, 4 h, 98%; (m) LDA, THF, -78 °C; HMPA, Me02CCN, -78 °C, 1.5 h, 83%; (n) LDA, THF, -78 °C; PhSeBr, 76%; (o) 03, CH2C12, -78 °C, 77%; (p) H202, NaOH, Me0H, rt, 14 min, 85%.  Scheme 27 53  ^  III. DISCUSSION - CRINIPELLIN B (15) III.1. ISOLATION. The second section of this thesis describes the synthesis of the tetraquinane diterpenoid (±)-crinipellin B (15). This highly oxygenated compound is part of a small family of structurally unprecedented natural products that share the 12-isopropy1-4,8,11-trimethyltetracyclo[6.6.0.0 1,11 .0 3,7 ]tetradecane skeleton 132. The other members of this family are crinipellin A (43), O-acetylcrinipellin A (133), tetrahydrocrinipellin A (134) and dihydrocrinipellin B (135). 0-Acetylcrinipellin A (133), an antibiotic which is active against Grampositive bacteria, was the first of these natural products to be isolated. The antibacterial metabolite 133 was found in the culture broth of the fungus Crinipellis stipitaria 7612.48 This fungus grows on the dead and living parts of grasses.  HO  OAc  OH  0  O  133  43 15  134 13  14  12  19  ^ 8  17 20 10^HO 16 ==9^0  OH  OH  132  15  135  Later on, various strains of the fungus Crinipellis stipitaria were examined. In addition to O-acetylcrinipellin A (133), two new biologically active substances, 43 and 15, were discovered. 17 Two inactive compounds, 134 and 135, were also isolated. 17 The three 54  crinipellin natural products 43, 15 and 133 that exhibit antibiotic properties have in common an a-methylene ketone moiety. In fact, this structural feature has been associated with the biological activity of various compounds. 49 The chemical structure of crinipellin A (43) was elucidated mainly by nnmr spectroscopic methods. 1 H and 13 C nmr data, 1 H- 1 H-correlation 2D nmr experiments, selective decoupling experiments, NOE difference measurements, and correlation of the nmr spectra of 43 with those of complicatic acid (136) and hypnophilin (137) led to the conclusion that the  COOH  Me  'Me  'Me 137  structural formula 43 correctly represents the constitution and relative stereochemistry of crinipellin A (43). The structure of crinipellin B (15) was derived in a similar manner. Moreover, the structure of 15 was unambiguously proven by X-ray analysis.  111.2. THE PROBLEM.  Compounds 15, 43, 133, 134 and 135 are presently the only known diterpenoids to possess a tetraquinane carbon skeleton. Each of these natural products contains (at least) 8 contiguous stereogenic centers, 3 of which are quaternary chiral atoms (C-1, C-8, C-11, crinipellin numbering). The level of complexity of the molecules is further increased by the proximity of a number of oxygen functionalities in rings A and C (epoxide, enone moiety, a-ketol), some of which are labile under acidic or basic conditions. Crinipellin B (15), crinipellin A (43) and 0-acetylcrinipellin A (133) represent valuable target molecules for syntheses because of their unique structural features and their interesting biological properties.  55  They have attracted the attention of a few groups, including our own. We decided to attempt the preparation of these challenging natural products.  111.3. PREVIOUS SYNTHETIC APPROACHES TOWARDS THE SYNTHESIS OF THE CRINIPELLINS.  One approach towards the construction of the functionalized tetraquinane carbon framework of the crinipellins had been published 50,51 by Mehta and coworkers prior to the beginning of our work on the crinipellins. During the course of our synthesis of (±)-crinipellin B (15), one other report on a synthetic approach to the crinipellin diterpenoids was  published. 52 These approaches are outlined below. 111.3.1. Approach by Mehta et al. 50,51  Mehta and coworkers assembled the linearly fused triquinane 142 in a short and ingenious sequence of steps from cyclopentadiene (138) and benzoquinone (139) (Scheme 28). 50,51 Their synthesis started with a Diels-Alder addition of 139 to 138 to furnish the endo adduct 140 in high yield. The two alkene functionalities of 140 are appropriately positioned to undergo intramolecular photochemical cycloaddition. Thus, irradiation of 140 with a 450 W Hanovia medium pressure mercury vapor lamp efficiently provided pentacyclo[5 .4.0.0 2,6 .0 3,10 .0 5,9 ]undecane-8,11-clione (141), after crystallization of the crude material. The dione 141 is known to be reluctant to undergo thermal rearrangement. However, sublimation of the latter substance under reduced pressure (1 Ton) at high temperature (560 °C) afforded the linearly fused triquinane 142 in 96% yield. It is interesting to note that the only "reagents" needed to accomplish the transformation of 138 and 139 into 142 are heat and light. The dienone 142 was converted into the enone ketal 144 in 3 steps (Scheme 28). Heat-promoted isomerization of one of the double bonds of 142 smoothly produced 143. The  56  Me0H, -70 °C to rt )1.-  +  hu, EtOAc or PhH  o 138  140 93%  139  H H^0  141  92% 560 °C, 1 Ton  1) H2, Pd/C,  Diphenyl ether, reflux -NE  EtOAc, 1 atm  2) F 1 -  HO OH  HH 0^0  0^0  p-TsOH, PhH, reflux  143  142  80%  96%  1) Me2C=CH(CH2)2MgBr 145 CuBr•Me2S, THF, Me2S -78 °C, 2 h 2) MeI, HMPA, 16 h, 87%  1) LHMDS, THF, -78 °C 2) PhSeC1, THF,^I._ -78 °C to 25 °C, 10 h 3) 30% aq H202, THF, 25 °C 148  77% from 147  H \__/^L.,)  O Me r 0^  1) LHMDS, THF, -78 °C 2) PhSeC1, THF, -78 °C to 25 °C, 10 h 3) 30% aq H202, THF, 25 °C  1) Me2CuLi, THF-Me2Sether, 20 °C 2) aq NH4C1N1140H -ii^ 0 0  H  Men \._ j^..-,  150  149  74%  90% Scheme 28 57  70% HC104, EtOAc, reflux, 82%^  .. Me in +^...Mei7  H Me  ^ 0 OH Me, %.., ‘__/^  3:1 151  150^  ...Me17  152  H2, Pd/C (10%), EtOAc -..4^  e  0 153  80%  Scheme 28 (continued)  less substituted double bond of 143 was hydrogenated under controlled conditions, and one of the ketone moieties was selectively protected as an ethylene ketal function, as shown in 144. A solution of the enone 144 in THF-Me2S at -78 °C was treated with the Grignard reagent 145, in the presence of CuBr•Me2S. The resultant reaction mixture contained two different  enolate anions obtained, respectively, from the conjugate addition of 145 to the a and the 13 face of the enone 144. The reaction mixture containing those two intermediates was treated with methyl iodide. The enolate anion in which the newly introduced side chain had the 13 orientation underwent methylation in a cis fashion, as expected, to furnish the cis, anti, cis product 147. However, the other enolate anion did not react with methyl iodide and led, after work-up, to the formation of the cis, syn, cis compound 146. The two adducts 146 and 147 were produced in a 1 : 1.3 ratio, respectively. Transformation of the ketone 147 into the key intermediate 150 was accomplished in a straightforward manner. Conjugate addition of Me2CuLi to the enone 148, derived from the ketone 147 via Reich's procedure, 38 afforded 149. The enone moiety of 150 was regenerated using, once again, Reich's procedure. 38  Acquisition of compound 150 set the stage for the key reactions of the synthesis. Treatment of the enone 150 with catalytic amounts of perchloric acid furnished a 3 : 1 mixture 58  of the endocyclic and exocyclic olefins 151 and 152 (Scheme 28). Hydrogenation of the mixture of olefins produced a single crystalline dione 153. Unfortunately, the relative configuration of the newly introduced stereogenic center was opposite to that found in the crinipellins, as proven by an X-ray crystallographic structure determination. In the hydrogenation reaction, the hydrogen approaches the exo face of the olefin functions of 151 and 152 to give 153. It was initially hoped that the neighbouring exo-methyl group (Me-17, crinipellin numbering) would influence the stereoselectivity of the hydrogenation process and provide at least some compound resulting from hydrogenation from the endo face of the alkene. The highly stereoselective nature of this reaction does seem surprising. However, it is well-known that the stereoselectivity of reactions involving 5-membered rings is sometimes difficult to predict and that unusual selectivities can be achieved. 22,23 Transformation of 151 and 152 into 153 is an example of such a case. As an alternative method for the construction of the final five-membered ring, Mehta and coworkers investigated a route (Scheme 29) which employed a radical cyclization  BF3.OEt2,  MCPBA, Na2CO3, I, CHC13, 0 °C, 45 min  PhH, 0 °C to rt, 3 h  OR  Bu3SnH, AIBN, PhH, reflux, IIMe i7 .....r 2.5 h 0  H  m  eo  153  65%  1) NaBH4, Me0H, -15 °C to 5 °C, 45 min, 70% 2) MeC6H40C(S)C1 Pyr-CH2C12, rt, 2 h, 62%  -.1^  0 H Me  o  156 R= —C—O II S  Scheme 29  59  reaction. 50,51 The enone 150 was subjected to a chemoselective epoxidation with MCPBA, and the resultant epoxide 154 was allowed to rearrange upon exposure to BF3.0Et2. Concomittant hydrolysis of the acetal group of 154 occurred during the rearrangement process, and the trione 155 was obtained. Fortunately, it was possible to achieve a chemoselective reduction of one of the carbonyl groups of 155 with sodium borohydride. The intermediate alcohol was converted into the derivative 156 by reaction with MeC6H40C(S)C1. Tin hydride promoted radical cyclization of 156 furnished the undesired tetraquinane 153, which was found to be identical with the same substance synthesized previously (see above). Despite the fact that Mehta et al. had (twice) obtained the wrong configuration at the C-12 stereogenic center (crinipellin numbering), they further explored the pathway towards the crinipellins (Scheme 30). 50,51 The enedione 160 was prepared from 153 in a straight-  1) NaBH4, Me0H, 0-10 °C, 2 h 2) MsCI, Pyr, 0-25 °C, 3 h  H me 0 OMs  80 °C, 2 h, 96% H Me  0  158  157  1 + :  70%  1.2  t-BuOOH, PDC, Celite, PhH, 10-  NaHCO3, 30% H202  25 °C, 6 h  aq THF, 025 °C, 0.5 h 160  43% LiN(SiMe3)2, THF, MeI, -78 °C, 2 h  162  60%  Scheme 30  60  forward sequence of reactions. Compound 160 could be converted into the epoxide 161 by reaction with H202 under basic conditions. Alternatively, 160 could be regio- and stereoselectively alkylated with MeI to give 162. 111.3.2. Approach by Curran and Schwartz. 52  Curran and Schwartz have reported52 model studies and variations of an approach towards the synthesis of crinipellin A (43) utilizing a remarkable tandem radical cyclization that they developed. The synthesis leading to a number of cyclization precursors is displayed in Scheme 31. 1,3-Cyclopentanedione 163 was converted into the symmetrical diketone 164  by palladium-catalyzed diallylation. Double-methylenation of the ketone functions of 164 was accomplished with the reagent prepared from TiC14/CH2Br2JZn in THF. Acid-promoted isomerization of the resultant product 165 provided 166 in 41% yield from 164. The low yield associated with the conversion of 165 into 166 might be caused by a polymerization reaction of 166 in the presence of acid. 52 The tetraene 166 is, otherwise, reported to be a stable and easily isolable compound. The presence of the two C-5 substituents on the ring prevents DielsAlder dimerization and prohibits 1,5-hydrogen shifts that would cause migration of the double bonds within the ring. 52 The side chains of 166 were modified by achieving appropriate functional group manipulations. The two less hindered alkene groups of 166 were hydroborated chemoselectively with the bulky dicyclohexylborane reagent. Oxidative workup yielded the diol 167. This diol could not be selectively monoprotected (1 equiv NaH, THF; 1 equiv TBDMSC1). However, it was possible to obtain a statistical mixture of three products (167, 168 and 169) upon treatment of the diol 167 with one equivalent of TBDMSC1 in the presence of imidazole. The unreacted diol 167 and the bis(silyl ether) 169 could be recycled, and the alcohol 168 was eventually obtained in 83% yield. Swern oxidation of 168 and Wittig olefination of the  61  0 i,...../.0Ac 0^ 0 TiC14, Zn,  aq HI,  DBU, Pd (0)^u^CH2Br2  166 41% (from 164)  165  164 84%  163  PhH  OR' 1) (C6H11)2BH, 0 °C 2) NaOH, H2 02, 0 ° C to rt, 66%  RO--\___ 1)Swern oxidation 2) Ph3P=C(CH3)2 170 66% R = TBDMS  167 R=R'=H 168 R = TBDMS, R' = H 169 R = R' = TBDMS  3) TBDMSC1, imidazole, CH2C12, rt 83% after recycling  1) n-Bu4NF, THF, rt, 94% 2) MsC1, Et3N, ether, 0 °C, 100% 3) Nal, acetone, rt, 74%  +  171 X = OH 172 X =OMs 173 X=I  174  175 5  1  Bu3SnH, AIBN, PhH, 80 °C, 65%  t +  Exo transition state  Scheme 31  62  Endo transition state  resultant aldehyde gave 170. The protecting group of 170 was removed with tetra-nbutylammonium fluoride, and the intermediate alcohol 171 was allowed to react with MsCl. Subsequent displacement of the mesylate group of 172 with Nal furnished the iodide 173. The key intermediate 173 was set to undergo the radical-mediated tandem cyclization. Thus, subjection of 173 to cyclization conditions (Bu3SnH, AIBN, PhH, 80 °C) provided two angularly fused triquinanes, 174 and 175 in a 1 : 5 ratio (Scheme 31). 52 The major isomer was shown to possess the endo configuration of the isopropyl group, opposite to that found in the crinipellins. In the cyclization process, the initial species generated from 173 is probably a primary alkyl radical that adds to one of the alkene moieties of the ring of 173 in a 5-exo fashion to give a cis-fused bicyclic allyl radical. Two transition states can be postulated for ring closure of this allyl radical, the "exo" and the "endo" transition states, that would lead to the triquinanes 174 and 175, respectively. Apparently, the orientation of the isopropylidene group as in the "endo" transition state is sterically favored compared to the orientation shown for the "exo" transition state. 52 Consequently, the tricyclic alkene 175 is formed preferentially in comparison with its epimer 174. At the outset of the work, it seemed likely that nonbonded interactions encountered in the "endo" transition state would favour cyclization via the "exo" transition state. However, experimental results contradicted those predictions. Although this sequence does not provide the desired triquinane 174 as the major product, it is interesting to note that the tandem radical cyclization afforded products in which two new quaternary centers, contiguous to the initial quaternary carbon, were formed easily. A parallel approach that would allow the introduction of an appropriately positioned functional group was studied by Curran and Schwartz. 52 An attractive way of generating an alkyl radical by indirect use of a C-H bond had been developed and was utilized to transform 176 into the diastereoisomers 178 (Scheme 32). The cyclization precursor 176 was  prepared by treatment of the alcohol 171 with (2-bromophenyl)chlorodimethylsilane in the presence of imidazole. Subjection of this precursor to n-Bu3SnH/AIBN/PhH/80 °C presum63  ably generated an aryl radical, which subsequently underwent a 1,5-hydrogen atom transfer. The resultant radical species 177 could then undergo cyclization to afford a mixture of diastereoisomers as shown in 178. After deprotection of the resultant triquinanes 178, and oxidation of the alcohols 179, two ketones, epimeric at C-12, were isolated. The isomer 181 formed preferentially had, once again, the undesired configuration at the C-12 stereogenic center (crinipellin numbering). As might have been expected, 180 and 181 were formed in a 1 : 4.5 ratio, similar to the ratio observed in the cyclization of the model compound 173.  sime 2 a, CH 2 C 1 2 ^I.Imidazole, 0 °C to rt 171  176  70% n-Bu3SnH, AIBN, PhH, -.I 80 °C OSi(CH 3 ) 2 Ph 178  177  n-Bu4NF, THF, 64% rt  PDC, DMF, rt, 80% HO  179  +  ^lb-  0  0 180 1  Scheme 32  64  181  4.5  In a variation of the tandem radical cyclization approach, the "reverse" cyclization was attempted (Scheme 33). The functional groups of the two side chains of the substrate were modified so that the isopropyl-containing ring would be formed first upon cyclization of an alkyl radical to the existent cyclopentane ring. The third 5-membered ring of the triquinane  OH  Me 2ButSiO  6 steps 168  ^  182 n-Bu3Sal, AIBN, C6D6, 80 °C  +  184  1  ^  183  ^  10  Scheme 33 products would result from the cyclization of the bicyclic allyl radical to the alkyne-containing side chain. The secondary bromide 182, synthesized in 6 steps from the alcohol 168, was converted to the alkenes 183 and 184 upon treatment with n-Bu3SnH in the presence of AIBN. The two tricyclic compounds 183 and 184 were formed in a 1 : 10 ratio with the undesired triquinane 183 predominating. A more direct way to assemble a functionalized triquinane skeleton, employing an acyl radical in the cyclization step, was investigated. 52 The alcohol 171 was converted to the ester 185 by a two-step sequence involving oxidation to the acid and subsequent esterification with diazomethane (Scheme 34). The ester 185 was treated with dimethylaluminum 65  methaneselenolate to furnish the methyl seleno ester 186 in 87% yield. Subjection of 186 to cyclization conditions (n-Bu3SnH, AIBN, PhH, 80 °C) yielded three products. Two of them were the tricyclic ketones 180 and 181 formed in the a 1 : 5 ratio. The third compound was the bicyclic ketone 187, resulting from reduction of the intermediate ally' radical. This  s____.  HO —\  SeCH  OMe  0  OK_ 1)Swern^•, 2) Ag2O  (CH3)2A1(SeCH3) 11.PhCH3, CH2C 1 2, 0 to 25 °C  3) CH2N2 185 66%  171  3  186 87%  n-Bu3SnH, AIBN, PhH, 80 °C, 62% +  +  0  0  180  187 2  1  Scheme 34  sequence produced a valuable intermediate 180 for the synthesis of crinipellin A (43). The alkene function of ring C of 180 would serve as an entry to the needed ketol moiety found in crinipellin A (43). The ketone group would allow further elaboration of the ring A of the crinipellins via known methods. However, due to the fact that the radical-mediated tandem cyclization had afforded the desired triquinane as the minor product of the reaction, this approach does not provide a useful pathway towards crinipellin A (43).  66  111.4. RETROSYNTHETIC ANALYSIS.  Our retrosynthetic plan towards the syntheses of crinipellin A (43) or B (15) contrasts with the two other approaches discussed above. This plan is pictured in Scheme 35. Crinipellin A (43) could, in principle, result from stereoselective a-hydroxylation of the intermediate 188 under conditions that would preserve the integrity of the newly introduced stereocenter. Crinipellin B (15) could be indirectly obtained from 188 via an a-hydroxylation-isomerization sequence. Alternatively, oxidation of the a-hydroxylated product(s) to a triketone intermediate, followed by chemo- and stereoselective reduction of the appropriate ketone function could lead to 15. The heavily functionalized tetraquinane 188 could arise from the enone 189 via a series of reactions, namely epoxidation, a-methylenation, deprotection and oxidation. In theory, this series of reactions could be accomplished in various orders. The sequence of steps should be planned in a way that would take into account the reactivity of the different functional groups that might be present at any given stage of the sequence. The a,13-unsaturated ketone 189 could be derived from the ketone 190 via a fivemembered ring annulation process. A number of such annulation processes, which would place the enone moiety in the desired position, are known and could be attempted. The ketone 190 could be prepared from the keto alkene 191. Stereoselective reduction of the ketone  function of 191, protection of the resulting alcohol and oxidative cleavage of the exocyclic double bond would yield 190. The angularly fused triquinane 191 could result from the theoretical combination of the synthons 192 and 193. These synthons correspond to the enone 194 and an appropriate reagent such as 195. In the reagent 195, X represents an atom or group of atoms that would activate, as such or after appropriate transformation, the carbon center to which it is attached,  67  0  OP  188  189  d a+  0 193  192  II I  III  191  Cu(CN)Li +  0  195  194  196  197  198  III  III (Br R R' ,  LiO  199  1  <=  201 R, R' =  202 R, R' = (0042)2  or SiMe  /=—K  Me3SiOktii& ^  200^203  Scheme 35 68  CH2  I  and allow reaction of this acceptor center with a donor site. The enone 194 could be formed upon intramolecular aldol condensation of the diketone 196. Disconnection of 196 into simpler units, as shown in Scheme 35, could afford the two synthons 197 and 198. A synthetic equivalent to 197 could be the lithium enolate 199 (whose possible precursor would be the trimethylsilyl enol ether 200). The synthon 198 has a reactivity pattern opposite to the "normal" reactivity mode associated with a carbon alpha to a ketone function. The center alpha to a carbonyl is normally a donor site. A synthetic equivalent to the synthon 198 in which the carbonyl group is masked is thus required. The reagents 201, 202 and 203 could all be used as electrophilic equivalents to the synthon 198. Alkylation of 199 with 201, 202 or 203 would afford compounds that can be transformed into 196 after functional group manipulations. Finally, the intermediate enol ether 200 could originate from 2-methy1-2cyclopenten-l-one. 111.5. TOWARDS THE SYNTHESIS OF (±)-CRINIPELLIN B (15). 111.5.1 Preparation of the Enone 194.  The method of Kuwajima and coworkers 31 was used to prepare the enol ether 200. Thus, treatment of 2-methyl-2-cyclopenten- 1-one with isopropylmagnesium bromide, in the presence of CuBr•Me2S, HMPA and TMSC1, provided 200 (equation 13). This substance 1) Me2CHMgBr, CuBr•Me2S, HMPA, TMSC1, TIT, -78 °C  (13)  ^4.-  2) Et3N  200  94%  was obtained in excellent yield, after (quick) flash chromatography of the crude product and distillation of the material thus obtained. In the 1 H nmr spectrum of 200, a trimethylsilyl signal at 8 0.17, two doublets at 8 0.70 and 0.88 (isopropyl methyl groups) and a signal due to  69  the vinylic methyl at 8 1.47, confirmed that the reaction had succeeded. The enol ether 200 was used within a week of its preparation. The pure compound could be stored under inert atmosphere (argon) in a freezer (-11 °C) without appreciable hydrolysis. The enol ether 200 could be transformed into the diketone 196 via two routes. Generation of the lithium enolate 199 was performed by treatment of 200 with MeLi at 0 °C. 53 The enolate anion 199 was alkylated with (E)-1-iodo-2-(trimethylsilyl)-2-butene (203) 54 (route 2, Scheme 36). The ketone 205 thus obtained was allowed to react with MCPBA, and the resultant epoxide 206, upon treatment with sulfuric acid, afforded the desired diketone 196. Conversion of 200 to 196 could also be accomplished via route 1. In practice, this  ( MeLi, sw THF, 0 °C RO^ Lid R = SiMe3  200  CH2C12 2) Me2S 3) aq HClTHF  Route 1  ^  199  ^  204 76%  196  93% 16  from 204  .....^MCPBA, NaHCO3  CH2C12 0^ Me 3 Si Route 2  205  H2SO4, Me0H  0 Me3Si  0  206 Scheme 36  70  pathway was somewhat more convenient and was subsequently followed. Reaction of 2bromomethyl-l-butene (201) with the lithium enolate 199 in the presence of catalytic amounts of (Ph3P)4Pd 55 gave, stereoselectively, the monoalkylated product 204, along with minor amounts of dialkylated material. Compound 204 resulted, as expected, from approach of the reagent on the less hindered face of the enolate 199, opposite to the isopropyl group. (An Xray crystallographic structure determination of a more advanced intermediate confirmed the cis relationship of the methyl and the isopropyl groups). The ir spectrum of 204 exhibited absorptions at 3084, 1739 and 1641 cm -1 for the alkene and ketone moieties. A series of signals in the 1 H nmr spectrum showed the presence of the alkene-containing side chain. Among these signals were found a triplet for a methyl group (Me-16, crinipellin numbering) at 5 0.96 and resonances for vinylic hydrogens at 5 4.66 and 4.83. A methyl singlet at 5 0.90 showed that the alkylation reaction had proceeded regioselectively in the desired sense. The various 1 H nmr signals were assigned to their respective hydrogens by COSY and decoupling experiments. The results of these experiments are reported in Table 4 in the experimental section of the thesis. A solution of the keto alkene 204 in Me0H-CH2C12 was treated with a stream of ozone (Scheme 36). After reductive workup with Me2S, 56 the 1 H nmr spectrum of the crude  product showed MeO signals, which indicated the presence of a ketal function. Acid hydrolysis of the crude material produced 196 in 93% overall yield. Two ketone stretching bands were observed at 1741 and 1714 cm -1 in the ir spectrum of 196. The various hydrogen signals in the 1 H nmr spectrum could be identified from COSY experiments. These data are listed in Table 5 in the experimental section. Base-promoted cyclization of the dione 196 with Me0Na in Me0H afforded cleanly the enone 194 in excellent yield (equation 14). This enone could be stored for a few days under inert atmosphere at low temperature (freezer). However, to exclude the possibility of decomposition, 194 was not kept for extended periods of time. Absorptions at 1708 and 1669 71  cm -1 in the it spectrum of 194 were assigned to the ketone and alkene functions. In the 1 H nmr spectrum, a resonance at 8 1.63 was attributed to the vinylic methyl group.  Me0Na, Me0H  (14)  reflux  196  ^  194 97%  111.5.2. Preparation of the Angularly Fused Triquinane 191.  The enone 194 could now be transformed into the key intermediate 191. A methylenecyclopentane annulation procedure that would produce the desired compound with the exocyclic double bond appropriately positioned had been elaborated in our laboratories 9 prior to the beginning of the synthesis of the crinipellins. In order to provide background for this method, a brief discussion of previous work performed in our laboratories will be presented. The bifunctional reagent 209 proved to be a suitable synthetic equivalent to the synthon 193 and was subsequently used for this work. The cuprate reagent 209 was conveniently  prepared from 4-chloro-2-trimethylstanny1-1-butene (6) (Scheme 37). 9 Transmetallation of 6 with methyllithium at -78 °C, followed by trapping of the intermediate anion with  trimethylgermyl bromide, gave 4-chloro-2-trimethylgermyl-1-butene (207). Reaction of 207 with sodium iodide in refluxing acetone (Finkelstein reaction) furnished the iodide 208. 4Iodo-2-trimethylgermy1-1-butene (208) was allowed to react with 2 equiv of t-BuLi at -98 °C to yield the corresponding organolithium reagent. The bright yellow solution containing the intermediate organolithium species was warmed to -78 °C and solid CuCN was added. Brief  72  Nal, CI acetone,^Me3GeI  Me 3 SnCI 1) MeLi, -78 °C^Me 3 Ge 2) Me3GeBr  reflux  6^  207  ^  Cu(CN)Li  Me3Ge  208  1) t-BuLi (2 eq) TIM, -98 °C to -78 °C 2) CuCN, -78 °C to -35 °C  a^d  209  193^  Scheme 37  warming of the resulting suspension to -35 °C gave a homogeneous, pale yellow or tan solution, indicating the formation of the lower order heterocuprate 209. It was essential to carry the lithium-iodine exchange on 4-iodo-2-trimethylgermy1-1-butene (208) rather than on the corresponding trimethylstannyl compound. 57 Attempts to achieve the metal-halogen exchange on 4-iodo-2-trimethylstannyl-1-butene did not provide efficiently the corresponding organolithium reagent. It has been suggested 57 that side reactions occur, caused by the presence of the trimethylstannyl group. The germanium-carbon bond is stronger than the tincarbon bond and, therefore, the lithium-iodine exchange can be performed on 208 without affecting the vinylgermane group. A variety of enones were converted to methylenecyclopentane annulation products via the new annulation procedure as shown in Scheme 38. 9 This annulation sequence is described in detail below through the transformation of the bicyclic enone 210 to the angularly fused triquinane 213. 1,4-Addition of the cuprate reagent 209 to the tetrasubstituted enone 210 in the presence of trimethylsilyl chloride afforded, after workup of the reaction mixture,  the pair of diastereoisomers 211 in 63% yield. Conjugate addition of 209 to 210 took place  73  Cu(CN)Li  1) Me3Ge  20 9^  (Ph3P)4Pd ...H ^ t-BuOK, t-BuOH, THF  Me3SiC1, THF, -78 °C to -63 °C 2) NH4C1, H2O 3)12, CH2C12  210^  ^  211 R = Me3Ge 63% ^213 65% 212 R I^70%  1) Me3Ge Me3SiC1, THF, -78 °C to -63 °C 2) NH4C1, H2O  12, CH2C12  215a 215b 215c  214a 214b 214c  73% 74% 35%  (Ph3P)4Pd t-BuOK, t-BuOH, THF  R2 218a 59% 218b 58% 218c 74%  216a 97% 216b 87% 216c 83%  217a 217b 217c 2  214a-218a R =R =R3 = H 1 3 214b-218b R =R2 = Me, R = H 2 1 3 214c-218c R =R = Me, R = H 1  Scheme 38  74  from the same side as the angular hydrogen to yield the cis-fused bicyclo[3.3.0]octan-3-one 211. As already pointed out previously, cis-fused 5-membered rings are generally formed  preferentially to trans-fused compounds since the latter types of products are very strained. Treatment of the vinylgermane 211 with iodine 58 in CH2C12 allowed germane-iodine exchange to proceed and yielded 212. It is interesting to note that this exchange typically occurs over a period of 12 to 24 hours. In contrast, the tin-iodine exchange of a number of vinyltin compounds is a fast process and is, in many instances, complete within minutes.  57  This difference in rate of reaction might be correlated with the strength of the carbongermanium bond compared with the strength of the carbon-tin bond. After many investigations, 57 reaction conditions were found that allowed the conversion of the vinyl iodide 212 into the tricyclic keto alkene 213. The cyclization reaction occurred upon addition of a solution of t-BuOK in t-BuOH-THF to a solution of (Ph3P)4Pd (catalytic amount) and of the vinyl iodide 212 in THF. It was essential to maintain a low concentration of base in the reaction media since higher concentrations favored elimination of the elements of HI from 212. Slow addition of the solution of t-BuOK over a period of —3 hours (using a syringe pump) decreased the amount of elimination reaction. In this manner, the annulated product 213 was obtained in 65% yield from the vinyl iodide 212. The cyclic enones 214a-c were also subjected to the methylenecyclopentane annulation sequence (Scheme 38). 9,57 The conjugate additions of 209 to the cc,13-unsaturated ketones 214a-b were efficient processes. 1,4-Addition of 209 to 214b proceeded from the side  opposite to the neighbouring methyl group, for steric reasons. Conjugate addition of the cuprate reagent 209 to the more hindered enone isophorone (214c) provided 215c in only 35% yield, probably due to steric hindrance. In fact, 214c is known for its reluctance to undergo 1,4-addition. The vinylgermanes 215a-c were treated with iodine in dichloromethane to furnish the vinyl iodides 216a-c. Cyclization of 216a-c could be accomplished under the conditions described above for the conversion of 212 to 213. The intermediates 217a-c 75  initially resulted from the cyclization step. However, in these cases, isomerization of the alkene function from the exocyclic position to the endocyclic position is possible and occurs under the basic reaction conditions. The enones 218a-c were therefore isolated as the final products in yields ranging from 58 to 74%. Since the two enones 210 and 194 are structurally somewhat similar, it seemed possible that 194 would undergo the methylenecyclopentane annulation sequence and afford  210  194  the desired tricyclic keto alkene 191. However, the enone 194 is considerably more sterically crowded than 210 because of the presence of the angular methyl and isopropyl groups. In order to produce the required cis-fused bicyclo[3.3.0]octan-3-one 219, the cuprate reagent 209 would have to add to the enone moiety of 194 from the same side as the methyl and  isopropyl groups. The resultant steric interactions involving these groups and the incoming reagent 209 would be expected to disfavour the conjugate addition reaction of 209 to 194. The influence of these steric effects in the 1,4-addition process was verified by allowing the a,(3-unsaturated ketone 194 to react with the cuprate reagent 209 in the presence of trimethylsilyl chloride (THF, -78 °C to - 48°C) (equation 15). The vinylgermane 219 was 1) Me3Ge "Me  0  194  Cu(CN)Li  209 ^ Me3 SiC1, THF, -78 °C to - 48 °C 2) NH4C1-NH4OH, H2O  ^  0.-  O  219 34%  76  obtained, but the yield was low (34%). Despite the low yield obtained, this result was very encouraging since it confirmed that the required transformation could be acccomplished. Unfortunately, further attempts to improve (or even reproduce) the yield of this reaction were unsuccessful. Invariably, starting material was recovered along with varying amounts of the desired product (0 to —25%). Conjugate addition of 209 to 194 was a slow process at -78 °C. Higher reaction temperatures allowed the 1,4-addition to occur to a certain extent, but prolonged reaction times did not seem to improve the transformation of 194 into 219. It is possible that the reagent 209 is unstable under these reaction conditions. Decomposition of the organocuprate 209 would obviously lead to low conversions of starting material into product. Other additives such as BF3•0Et2 59 and BF3•0Et2-TMSC1 60 were utilized in the conjugate addition, but without substantial improvements. Eventually, it was discovered that replacement of the additive TMSC1 by TMSBr had a remarkable effect on the reaction. Thus, treatment of the enone 194 with 209 in the presence of TMSBr afforded the adducts 219, epimeric at C-8, in a gratifying 83% yield (equation 16). In contrast, conjugate addition of 209 to the sterically less hindered enone 210 (in the presence of trimethylsilyl chloride) had 1) Me3Ge  Cu(CN)Li 209  0 194  Me3SiBr, THF, -78 °C to - 48 °C 2) NII4C1-NH4OH, H2O  ^  O  219 83%  provided 211 less efficiently (in only 63% yield). It is of interest to note that the yields of the conjugate addition reaction of 209 to 194 were typically found to be —90% after chromatographic purification of the crude material and removal of traces of solvent (vacuum pump). This epimeric mixture of the vinylgermanes 219 was usually pure enough before distillation for direct transformation into the vinyl iodides 220. The 1 H nmr spectrum of 219 77  revealed two signals for the l'v3Ge group at 8 0.19 (major vinylic germane) and 0.21 (minor epimer) and two sets of vinylic hydrogens at 6 5.14 and 5.47 for the major vinylic germane, and at 6 5.18 and 5.52 for the minor epimer. The keto trimethylgermanes 219 were converted to the iodides 220 quantitatively upon treatment with iodine in dichloromethane at room temperature for 16 hours (equation 17). The  e  0  12, CH2 C12  ....-^  ^  219  (17)  0  ^  220  97-100%  vinylic hydrogen signals at 6 5.63 and 5.96 (major epimer) and at 6 5.68 and 6.03 (minor isomer) in the 1 H nmr spectrum confirmed the formation of 220 . The iodides 220 were subjected to the palladium(0)-catalyzed cyclization conditions described above (equation 18).  ..-Me  t-BuOK, t-BuOH THF, (Ph3P)4Pd ^O.-  I 0  220  ^  191  83%  Gratifyingly, the angularly fused triquinane 191 was isolated as a white solid (mp 22.0-22.5 °C) in 83% yield after chromatographic purification of the crude material and distillation of the oil thus obtained. The conversion of 220 into 191 is an extremely efficient process. Comparatively, cyclization of 212 and 216a-c provided the annulation products 213 and 218a-c in 65%, 59%, 58% and 74% yields, respectively. The it spectrum of 191 showed 78  three absorptions at 3080, 1737 and 1651 cm -1 associated with the carbonyl and alkene functionalities. The 1 1-1 nmr spectrum displayed resonances at 8 1.01 and 1.18 due to two angular methyl groups and signals at 8 4.86 and 4.95 arising from the two alkene protons (Figure 5). The overall conversion of 194 into 191 was highly stereoselective and produced  a functionalized tricycle in which the three contiguous quaternary stereogenic centers required for the eventual synthesis of (±)-crinipellin B (15) had been installed cleanly and efficiently. 111.5.3. Preparation of the Ketone 224.  The ketone function of 191 was reduced with lithium(diisobutyl)(n-butyl)aluminum hydride 61 in ether (equation 19). The reduction step was highly stereoselective and afforded the alcohol 221 in 93% yield. The minor epimer 222 was isolated in —2% yield. The it spectra of 221 and 222 each showed a broad absorption for an alcohol function (at 3319 and at 3510 cm -1 respectively). The stereochemistry at the carbinolic center of each product was determined by NOE difference measurements. The various 1 H nmr signals were assigned by COSY and irradiation experiments. These results are compiled in Tables 6 and 7.  1) n-Bu(i-Bu)2A1(H)Li  0  (19)  ether, -78 °C 2) Na2SO4.10H20; aq NaOH  E H9 '  Me OH 7 * 16  191  221  93%  222 —2%  In order to prove the configuration of the carbinolic center of the major product 221, it was especially important to assign specific 1 H nmr signals to the quaternary methyl groups and to the hydrogen (H-12) vicinal to the isopropyl hydrogen (Table 6). The COSY spectrum  79  2.4  ^  .I ^"^'  7.0^6 0  2.2^2.0^1 . 6  PPM  ^ ►  1.6^1. LI  ^ ^ ^ 0 4_0 5.0^2.0^1.0 PPM  Figure 5: The 1 H nmr Spectrum (400 MHz, CDC1 3 ) of the Keto Alkene 191.  Table 6: 1 H nmr Data (400 MHz, CDC13) for the Alcohol 221a.  Assignment H-x  COSY Correlationsb  H-2 H-2'e (cc) H-3 (cc)  1H nmr (400 MHz) 8 ppm (mult., J (Hz), # of H) 1H, part of the m (4H) at 1.13-1.44 2.01 (ddd, J = 4, 7, 12.5, 1H) 2.16-2.28 (m, 1H)  H-3'  2.28-2.38 (m, 1H)  H-7*d H-7*' H-9e  4.86 (br s, 1H) 5.10 (m, 1H) 3.82 (ddd, J = 6, 9, 9, 1H)  OH9e H-10 H-10' H-12 H-13  1.66 (d, J = 9, 1H) 1.06 (dd, J = 9, 13, 1H) 2.12 (dd, J = 6, 13, 1H) 1H, - 1.13-1.24, part of the m (4H) at 1.13-1.44. 1H, part of the m (4H) at 1.13-1.44.  H-13'  1H, part of the m (2H) at 1.69-1.86.  H-14 H-14' Me-16 Me-17  1H, part of the m (4H) at 1.13-1.44. 1H, part of the m (2H) at 1.69-1.86. 1.10 (s, 3H) 0.93 (s, 3H)  H-18 Me-19 Me-20  H-12, Me-19, Me-20 1.50-1.63 (m, 1H) H-18 0.88 (d, J = 6.5, 3H) 0.96 (d, J = 6.5, 3H)^_ H-18  H-2', H-3, H-3' H-2, H-3, H-3' H-2, H-2', H-3', H7*, H-7*' H-2, H-2', H-3, H7*, H-7*' H-3, H-3', H-7*' H-3, H-3', H-7* H-10, H-10'  NOE Correlationsb  H-7*, H-10', H-12, Me-16  H-9, H-10' H-9, H-10 H-13, H-13', H-18 H-12, H-13', H-14, H-14' H-12, H-13, H-14, H-14' H-13, H-13', H-14' H-13, H-13', H-14  H-7*, H-9, H-2', H-3, H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2). d- * indicates a hydrogen on a carbon that will not be found later on in crinipellin B (15). e- Coupling between H-9 and OH9 was not always observed in the 1 H nmr spectrum of 221. It depended on the preparation of the sample. In the COSY spectrum, no coupling was observed between H-9 and OH9.  81  allowed the identification of the isopropyl hydrogen (H-18, m at 8 1.50-1.63) through the correlation of its signal with the two isopropyl methyl resonances (two d at 8 0.88 and 0.96). In the COSY spectrum, a third 1 H nmr signal showed a correlation with the 11-18 multiplet, and was thus attributed to H-12 (m at —1.13-1.24, part of the m at 8 1.13-1.44) (Table 6). Homonuclear decoupling experiments led to the same conclusion (see experimental section). The NOE difference experiments allowed the assignment of Me-16 and Me-17 to their respective singlets. Saturation of the signal at 8 1.10 caused enhancement of the broad singlet (H-7*) at 8 4.86. This resonance was therefore attributed to Me-16. Irradiation at 8 1.10 also caused enhancement of the carbinol signal (11-9, 8 3.82), which strongly suggested that the hydroxyl group had the a orientation. Unequivocal confirmation of the relative configuration of 221 at C-9 was obtained upon irradiation of H-9. Increases in the intensity of the signals due to H-7*, H-10', Me-16 and 11-12 were observed. Enhancement of the H-12 multiplet can occur only when the carbinol hydrogen (11-9) is oriented as shown in 221.  221  ^  221  The information acquired (Table 7) for the alcohol 222 was consistent with the proposed structure. In NOE difference experiments, irradiation of each quaternary methyl group caused a small enhancement of the H-9 signal. Saturation of the resonance at 8 4.00 (H9) caused a small increase in the intensities of the singlets due to Me-16 and Me-17. Molecular  82  Table 7: 1 H nmr Data (400 MHz, CDC13) for the Alcohol 222a. 19 3 04 14 2^... 17 23  10  7  Assignment H-x  16  OH  222  1H nmr (400 MHz) 8 ppm (mult., J (Hz), # of H)  COSY Correlationsb  NOE Correlationsb  H-2  part of the m (6H) at 1.18-1.59, 11-1  H-2', H-3  H-2'  part of the m (511) at 1.74-1.93, 1H  H-2, H-3  H-3  2.36-2.43 (m, 2H)  H-2, H-2', H-7*, H-7*'  H_7*c  4.77 (br dd, J = 1.5, 2.2 Hz, 1H)  11-3  H-7*'d  4.82 (ddd, J = 1, 2, 2, 111)  H-3  H-9  4.00 (dd, J = 7.5, 7.5, 1H)  H-10, OH9  OH9  Part of the m (6H) at 1.18-1.59, 1H  H-9  H-10  part of the m (5H) at 1.74-1.93, 2H  Me-16  1.07 (s, 3H)  H-7*, H-9  Me-17  0.91 (s, 3H)  H-9  H-18  —1.47-1.59 (m, 1H), part of the m (6H) Me-19, Me-20 at 1.18-1.59  Me-19  0.88 (d, J = 6.5, 3H)  H-18  Me-20  0.92 (d, J = 6, 3H)  H-18  H-7*, H-3, Me-16, Me-17  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- * indicates a hydrogen on a carbon that will not be found later on in crinipellin B (15). d- H' indicates the hydrogen of a pair which is more downfield (H-7' is more downfield than H-7).  83  222 models demonstrate that, of the two possible compounds 221 or 222, only 222 can account for these observed enhancements. The hindered alcohol function of 221 was protected as a silyl ether by reaction with TBDMSOTf 62 in the presence of triethylamine in CH2C12 to yield 223 (Scheme 39).  Me  'while^1) 0s04, PYr,  Me2Bu tSiOTf,  2) aq NaHS03  rt  ^0.-  Et3N, CH2C12 E  E  OH^ 221^  3) Pb(OAc)4, THF, 0 °C  OSiMe2But^  223^ 98%^  0^"OSiMe2But  224 93%  Scheme 39 Resonances at 8 0.03, 0.05 and 0.90 in the 1 H nmr spectrum witnessed the presence of the t-BuMe2Si group. The oxidative cleavage of the exocyclic olefin function of 223 was initially a difficult process. Ozonolysis gave the desired ketone 224 in low and irreproducible yields. Various reaction conditions utilizing catalytic 0s04 oxidation of the olefm and in situ cleavage of the intermediate diol(s) with sodium metaperiodate were also attempted but did not provide 224 in good yield and in acceptable lengths of time. The alkene was thus converted to an  intermediate diol by reaction with 0s04 in pyridine. 63 The diol thus obtained was cleaved with 84  lead tetraacetate to afford, in 93% overall yield, the ketone 224 (Scheme 39). An absorption at 1736 cm -1 for the ketone function was observed in the it spectrum of 224. 111.5.4. Attempts to Assemble the Last 5-Membered Ring of the Crinipellins.  In theory, construction of the fourth required 5-membered ring could be performed by use of various cyclopentenone annulation methods. An interesting annulation procedure developed by Piers and Abeysekera, 64 which involved an intramolecular Wadsworth-Emmons reaction, 65 was initially attempted. Alkylation of the ketone 224 with dimethyl 3-bromo-2ethoxypropenylphosphonate (225) afforded two products, along with recovered starting material (unoptimized reaction conditions, Scheme 40). The two products were assigned structures 226 and 227 respectively, based on the data derived from their 1 H nmr spectra. The alkylation reaction of 224 on carbon was believed to occur from the less hindered exo face of the molecule to yield the phosphonate 227. Direct verification for the stereochemical outcome of this reaction was not obtained; however, it was eventually shown that alkylation of 224 with a different reagent (vide infra) provided the epimer with the desired configuration at  C-3. Hydrolysis of the enol ether moiety of 227 gave, after purification of the crude material by flash chromatography, a mixture of two products, as seen by 1 H nmr spectroscopy. It was concluded that the major compound was the diketo phosphonate 228. The minor product was thought to be either the diketo phosphonate 229, resulting from epimerization of 228 at C-3, or the enol 230. Since this minor compound was not obtained in pure form, it was difficult to identify unambiguously. In order to explore the feasibility of the annulation sequence, the mixture of products was subjected to cyclization conditions. Treatment of the mixture containing 228 with a variety of bases under different reaction conditions 64,65 (Scheme 40, 1 to 5) either decomposed the starting material or left it unreacted. Clearly, the hindered carbonyl moiety is reluctant to undergo reaction with the phosphonate anion under the conditions employed. It was thus obvious that another cyclopentenone annulation method was required. 85  1  ..... 0 II^,.. Me P  ---  i i I  Me  1) LDA, THE  (MeO)2  2)  0(okie)2 II , 2^Br^CH-13/1 0^OSiMe2But Et0 225  I  CH Et()  224  0^OSiMe2But  226 11% +  Recovered starting material (41%)^+  EtO  224 CH^0^OSiMe2But  P.^227  0^- (0Me)2  42% aq HC1, acetone  0 II^„. Me _1:'  (Me0) 2-^CH  + (Me O) 2  OH^0^OSiMe2But  P  II 0  0^OSiMe2But  > 90% yield  230  228 has the 13 side chain 229 has the a side chain Reaction conditions 1-5  Reaction conditions for the cyclization of 228 —Do.- 231 1) NaH, DME, 65 °C 2) K2CO3, 18-Crown-6, PhH, 60 °C 3) K2CO 3 , 18-Crown-6, THE 4) Cs2CO3, THF-DMF 5) Cs2CO3, CH3CN  OSiMe2But  231 Scheme 40  86  An annulation sequence that involved an aldol condensation was explored. 66 Treatment of the ketone 224 with LDA, followed by trapping of the resultant enolate anion with 2bromomethyl- 1-propene afforded 232 in 51% yield, along with recovered starting material (45%) (unoptimized conditions) (Scheme 41). The 1 H nmr spectrum of 232 showed  4c,r  ..... 1.--3:• Me  1) LDA, THF, -78 °C  2  0^OSiMe But 224  224 +  2) Br  45%  0 °C to rt over 6 h, rt overnight  2  0^OSiMe But 232 51% 1) 03/ Me0HCH2C12 2) Me2S  Me0Na/Me0H or t-BuOH, t-BuOK, THF  2  OSiMe But  0^OSiMe2But  231  233  95%  Scheme 41  resonances for a vinylic methyl group at 8 1.70 and for the alkene hydrogens at 8 4.68 and 4.74. Ozonolysis of the alkene group of 232, followed by reductive workup with Me2S, gave the diketone 233. The 1 H nmr spectrum of 233 revealed a signal for a methyl adjacent to a ketone function at 8 2.16. Base-promoted intramolecular aldol condensation of 233 could not be accomplished, either with Me0Na in Me0H or t-BuOK in t-BuOH-THF. In the former case, the diketone 233 and its epimer 234 (not shown), were obtained. Under the latter conditions, a new compound formed, whose spectroscopic data ( 1 H nmr, 13 C nmr, mass and 87  it spectra) did not correspond to the expected data for the desired enone 231. The newly formed compound could not be identified unambiguously. The aldol condensation is a reversible reaction. It is possible that the equilibrium between the enolate anion formed upon deprotonation of 233 at C-6 (crinipellin numbering) and the tetracyclic keto alkoxide resulting from cyclization lies far in the direction of the enolate anion. The cyclization is, no doubt, disfavored by the hindered nature of the carbonyl group of 224. A slightly longer alternative pathway to transform the ketone 190 into the enone 189 via the alkene 235 is illustrated in Scheme 42. A reagent corresponding to the donoracceptor synthon 236 is required to achieve the desired conversion. The mesylate 237 67 and the iodide 238 67 in which the donor center is masked as a trimethylsilyl group are suitable synthetic equivalents to 236.  mi  me  OP^  189^  HO  OP  235 +  SiMe3  237 x = oms 238 X=II  236  Scheme 42  Trost and coworkers have published 68 a few cases of this type of annulation sequence. One example is illustrated in Scheme 43. The sodium enolate of the keto sulfone 239 was  88  O CI:AAPSO 2 Fh  Me3Si  O  NaH, Nal DMF, 55 °C  n-Bu4NF SO 2 Ph THF, 55 °C  Me3Si^OSO2CH3  23 9  237  240 77%  Scheme 43  alkylated with the reagent 237. Alkylation proceeded from the less hindered exo (convex) face of the substrate to provide 240. Intramolecular fluoride-promoted addition 69 of the allylsilane moiety of 240 to the ketone function of 240 proceeded in excellent yield to afford 241. This cyclization sequence seemed to be promising and was attempted in our approach to the synthesis of the crinipellins. The cyclization precursor 244 was prepared from the alcohol 221 (Scheme 44). The alcohol function of 221 was protected as a MOM ether since a silyl ether protecting group would not survive the fluoride-mediated cyclization step. Treatment of 221 with MOMC1 and diisopropylethyl amine afforded compound 242 which was converted to the ketone 243 by oxidative cleavage (0s04, Pb(OAc)4). The ketone 243 was allowed to react sequentially with a solution of base (LDA or HMDSK) and with the reagent 238." In each case, two substances, 244 and 245, were isolated in low yield, along with recovered starting material 243. The ratios of 244 to 245 varied depending on the reaction conditions used. The  reaction between the enolate anion of 243 and the alkylating agent 238 seems to be a slow process, allowing side reactions to occur. In fact, Trost and Curran had noticed that alkylation reactions that employed reagent 238 were slower than those which utilized 2-iodomethy1-1propene, and attributed this observation mainly to steric effects. 71 The sluggishness of the alkylation process coupled with the difficulty to obtain pure products constitute obstacles to this annulation sequence. Nevertheless, the cyclization step was attempted since a small amount of one of the two isomers, thought to be compound 245 with the undesired configuration at C-3, 89  could be obtained pure. Subjection of 245 to cyclization conditions (n-Bu4NF, THF, 4 A molecular sieves) led to the isolation of three compounds which were characterized by 1 H nmr  P  47  z_  ..... 1.,^..... I-, 1) 0s04, Pyr, n Me^ 2) aq NaHS03  Me MOMC1, CH2C12 OH  ).-(Me2CH)2NEt, 20 h rt  .. Me  3) Pb(OAc)4, THF, 0 °C O OMOM  221^  242  243  97%  96% 1) LDA or HMDSK, THF 2) I  SiMe3 238  1  ...  +  O OMOM^ o Me3Si^245^Me3Si^  OMOM  244  n-Bu4NF, THF, molecular sieves  li  ..... 4, S. Me +  o  +  OMOM  246 Scheme 44  and it spectroscopies. One of the compounds, isolated in minute amount, was tentatively assigned structure 248. The two other alkenes were identified as the desilylated products, 246 and 247. Because of the difficulties encountered, this pathway was also abandoned.  90  111.5.5. Preparation of the Enedione 267 and of the Enone 231.  In view of the failure of known cyclopentenone annulations to effect the conversion of 224 into 231, it appeared that a new annulation method needed to be developed. It seemed  possible to synthesize the enone 231 by oxidative rearrangement, with chromium(VI) reagents, of the allylic alcohol 249 (Scheme 45). The tetracyclic compound 249 could be  => Hö OSiMe 2 But^OSiMe2But^6^I 0^OSiMe 2 But 231 ^249^ 250  .u.  /=\__ I  Br  +  1  ..... --.Me ,  0^OSiMe2But  251  224  Scheme 45  obtained from the intramolecular addition of the vinylic anion generated from 250 to the carbonyl group of 250. A lithium-iodine exchange reaction, triggered by n-BuLi or t-BuLi for example, would serve to unmask the latent donor site of the side chain (at C-6) of 250 and produce a vinylic anion. This vinylic anion, in close vinicity to the carbonyl acceptor center, should react with the latter function to afford the allylic alcohol 249. Previous studies from our laboratories 72 indicated that the lithium-halogen exchange should be faster than the intermolecular addition of the alkyllithium species to the carbonyl group. Alkylation of the 91  enolate derived from 224 with a reagent such as the bromide 251 should yield the substance 250 with the side chain at C-3 having the 13 orientation. In the alkylation reaction of 224, the  incoming reagent would be expected to approach the enolate from the exo (convex) face of the molecule and yield the product 250. The work that inspired the elaboration of the new 5membered ring annulation procedure is described below. Piers and Marais have developed 72 a valuable method that allows the formation of bicyclic systems containing an allylic angular hydroxyl group. This method could provide a valuable entry into syntheses of a number of natural products. The annulation procedure was used to convert the ketones 253a-d, synthesized from the keto ketal 252 by known procedures, 72 into bicyclic ketols as shown in Scheme 46. Treatment of the ketone 253a with n-butyllithium (-2.5 equivalents) in THE at -78 °C gave a 1:1 mixture of the cis- and trans-fused substituted bicyclo[4.3.0]nonanols 254a and 255a in 69% yield. On the other  A  A  A n-BuLi  --IN.- --II.-  +  ^OH  ^OH  252 253 ^254^255  p -.:  a n= 1 R=H^1 b n= 1 R=Me^> 99 c n=2 R=H^1 d n=2 R=Me^> 99  A= e  0  1 <1 15 <1  1 n-BuLi (2.5 equiv)  Me 2 N-N 256  ^  257  ^  258 76%  Scheme 46  92  69% 83% 79% 74%  hand, subjection of compound 253b to the cyclization conditions furnished exclusively the cisfused allylic alcohol 254b. The iodide 253c, which contains a side chain longer by one carbon than those of 254a-b afforded, upon cyclization, the substituted bicyclo[4.4.0]decanol compounds 254c and 255c with the trans-fused product predominating. Finally, reaction of the iodide 253d with n-BuLi provided the cis-fused product 254d. It is of interest to note that the precursors 253b and 253d, in which R represents a methyl group, lead exclusively to the cis-fused bicyclic alcohols 254b and 254d. These examples demonstrate that it is feasible to achieve an intramolecular addition of an appropriate anion to a ketone moiety. 73 The case of another related annulation procedure is also shown in Scheme 46. 72 The dimethylhydrazone 256 was alkylated with (Z)-1-chloro-3-tri-n-butylstanny1-2-butene, and the ketone function of 257 was regenerated by hydrolysis of the hydrazone group. A second alkylation reaction with  methyl iodide, followed by tin-iodine exchange provided the vinyl iodide 257. Subjection of 257 to n-BuLi in THE at -78 °C allowed the cyclization reaction to occur in 76% yield. The  resultant allylic alcohol 258 was cis-fused and possessed an endocyclic double bond. The proposed pathway to construct the last ring of crinipellin B (15) required that 250 undergo cyclization to give 249. Although the iodide 257 experienced smooth cyclization to provide 258, it was not assured that annulation of 250 would be successful. The presence of the relatively acidic hydrogen at C-3 could favour side reactions. The reagent n-BuLi, which should normally effect the lithium-iodine exchange, could also act as a base and remove a proton at the C-3 center of 250. After aqueous workup, the vinyl iodide 265, resulting from kinetic protonation of the anion derived from 250, would be formed. Intramolecular protonation of the vinylic anion generated from the lithium-iodine exchange of 250 could also occur. After metal-halogen exchange of 250, the resultant vinylic anion might abstract the proton at C-3 faster than it could react intramolecularly with the carbonyl group of 250 thus preventing the formation of the tetracyclic alcohol. However, further work demonstrated that the transformation of 250 into 249 could be performed.  93  The conversion of the tricyclic ketone 224 into a tetraquinane product required, as the first step, an alkylation reaction with the reagent 251. The bromide 251 was prepared from the alcohol 263 (Scheme 47). A few syntheses of the alcohol precursor 263 had been published. 74 In practice, it was somewhat more convenient to obtain 263 by reduction of the known methyl (Z)-3-iodopropenoate (262). 75 Initially, methyl (Z)-3-iodopropenoate (262) was synthesized from propiolic acid (260) by the procedure of Moss and coworkers (route 1, Scheme 47). 75 a The acid 260 was allowed to react with HI in the presence of CuI to furnish  H—CEC—COOH 260  HI, CuI ^ /==\ 8-10 °C^I^COOH Route 1^2 61 Me0H H2SO4, A, 24 h  H  —  Nal, AcOHDIBAL, THF ^10.– /--=\ C EC — COOMe ^0. 70 °C, 12 h^I^COOMe -78 °C to 0 °C; workup 2 6 4^Route 2  ^2 6 2^  ir=\--OH 263  Scheme 47 (Z)-3-iodopropenoic acid (261). Acid-promoted esterification of 261 with Me0H provided methyl (Z)-3-iodopropenoate (262). Later on, during the course of the synthesis of (±)-crinipellin B (15), a shorter way of gaining access to 262 was published. 75 d - e Thus, treatment of methyl propiolate (264) with Nal in AcOH at 70 °C directly provided 262 in good yield. Moss and coworkers reported that the reduction of the methyl ester group of 262 to a hydroxyl moiety could be accomplished with LiA1H4. In our hands, this procedure did not afford the alcohol 263 in satisfactory yields. The reduction of 262 was therefore carried out with DIBAL in THF to give, after workup and distillation of the crude material obtained, the alcohol 263. The bromide 251, a strong lachrymator, was obtained upon reaction of 263  94  with Ph3PBr2 in CH2C12 (equation 20). 76 The bromide 251 was inclined to decompose upon I 1—\--OH 263  Ph3P, Br2, CH2Cl2 ^No-  (20)  1— /—=\---Br  251  heating or exposure to light. Solutions of the reagent 251 in a solvent at room temperature started to turn pink within about one hour. However, after completion of the reaction (263 to 251) and appropriate workup and purification procedures, the alkylating reagent 251 could be stored for a few months without serious decomposition in a freezer (-11 °C) under an inert atmosphere (argon) over a piece of copper wire. The required annulation sequence could now be undertaken. The ketone 224 was treated with LDA in THF at -78 °C, and the resultant enolate anion was allowed to react with (Z)-3-bromo-l-iodopropene 251. After a few trials, two alkylated compounds, 250 and  265,  were isolated in 76% and 3% yields, respectively (equation 21). The major product 250 displayed, by it spectroscopy, two absorptions at 3072 and 1610 cm -1 associated with the double bond of the side chain. The 1 H nmr spectrum exhibited signals at 8 6.18-6.27 for the ..... L.^  .....4,  Me^ 1) LDA, THF  ..Me^ ,1.. Me ^ + (  z  2) I r--A—Br 0^OSiMe2But 251  ..... 4—  0^OSiMe2But  224  ^  (21)  1 0^OSiMe2But  ^265 250 76%^  3%  vinylic hydrogens of the side chain. The major product 250 was shown to be the desired epimer upon conversion into a more advanced intermediate in the synthesis of (±)-crinipellin B (15). The structure of this intermediate was confirmed by X-ray crystallography. Thus, the alkylation of 224 with 251 had proceeded as expected, in the desired sense, from the less  95  sterically hindered exo face of the substrate. The minor isomer 265 showed, in its ir spectrum, bands at 3064 and 1607 cm -1 and, in its 1 H nmr spectrum, a 2-hydrogen multiplet at 8 6.21-6.32. These signals confirmed that an alkene-containing appendage was embodied in the isomer 265. The spectroscopic data gathered for the minor product 265 were consistent with the proposed structural formula. Strangely, the two keto iodides 250 and 265 were invariably accompanied by unreacted starting material 224 (20%). It is possible that the enolate anion derived from 224 acted, in a competitive reaction, as a strong base and caused elimination of the elements of HI from the reagent 251 (equation 22). The starting material  4c  .... 9 Me  .....  1  THF 0^OSiMe2But  , C_ Br  „.Me^ H  LDA,  OLi  /=-  224 +^(22)  OSiMe2 But 251^  Br  266  224 224 and the acetylene 266 would result from such a side reaction. Nevertheless, the  alkylation reaction afforded the keto iodide 250 in good yield, provided that two precautions were taken. Specifically, the substrate 224, which had been recrystallized from acetonitrile, had to be distilled under reduced pressure prior to use. The bromide 251 was filtered through flame-dried basic alumina and freshly distilled immediately prior to use in the alkylation reaction. A pivotal stage in the synthesis of (±)-crinipellin B (15) had been reached, where the decisive cyclization of the keto iodide 250 could be attempted. Treatment of a solution of 250 in THF at -78 °C with a solution of n-BuLi in hexanes afforded one major product. It was gratifying to find that the allylic alcohol 249 had formed cleanly in 93% yield (equation 23). The ir spectrum of 249 showed bands at 3496, 3051 and 1620 cm -1 for the alcohol and alkene  96  n-BuLi, THE  \^-_ I 0^aSiMe2But  -78 °C  (23) HO  OSiMe 2 But  249  250  93%  functions. The 1 H nmr spectrum (Figure 6) displayed a sharp singlet at 8 5.29, which exchanged with D20, for the hydroxyl group. Moreover, the elemental analysis and the mass spectrum agreed with the molecular formula C25H4402Si. The orientation of the fourth ring in 249 was determined by the stereochemistry of the side chain of 250. The alkene-containing  group of 250 was oriented (3 and, therefore, attack of the carbonyl group occurred from the (3 face to yield the cis-fused tetraquinane 249. Consequently, the angular hydroxyl group of 249 had the a orientation and was in close proximity with the protected secondary hydroxyl  group. An X-ray structure analysis of a more advanced intermediate confirmed the expected configuration of the C-3 center. Completion of the annulation sequence required oxidative rearrangement of the allylic alcohol moiety of 249 with a Cr(VI) reagent. A number of these chromium reagents have been used in oxidation of allylic tertiary alcohols (for example PCC 29a,45b,c pp c77 and Cr03 ,  reagents 78 ). The desired conversion of 249 into the enone 231 was attempted under various reaction conditions using these reagents. The reactions involving PCC were the most successful (Scheme 48). Treatment of 249 with a large excess of PCC (5 or 10 equiv) in CH2C12 (with or without additives such as NaOAc, 4 A molecular sieves and/or Celite) gave, depending on the reaction conditions, two or more products in varying ratios. These products were assigned structures 267, 231, 268 and 269. For example, treatment of the intermediate 249 with PCC in the absence of any additive or in the presence of Celite furnished the  97  Figure 6: The 1 H nmr Spectrum (400 MHz, CDCI 3 ) of the Allylic Alcohol 249.  e PCC, CH2C12 ^via, b, c or d HO  249  OSiMe 2 Bu t  267 a) no additive b) NaOAc, 4 A molecular sieves c) NaOAc, 4 A molecular sieves, Celite d) Celite  231 +  e  <  .... M e  + 0  OSiMe 2 But  0 268  269  Scheme 48  enedione 267 as the major product along with the enone 231. The epoxide 268 79 was also produced in small quantity in the experiment that employed Celite as an additive (condition d, Scheme 48). On the other hand, when the tertiary alcohol 249 was allowed to react with  PCC in the presence of excess NaOAc (5-20 equiv, conditions b and c), the major product isolated was the enone 231. The enedione 267 and the epoxide 269 79 were also obtained under these reaction conditions. The relative amounts of 231, 267 and 269 formed in the latter oxidative rearrangement reactions seemed to be dependent on, among other factors, the quantity of NaOAc used. Interestingly, attempts to convert the enone 231 into the epoxide 269 with H202 in the presence of NaHCO3 in a mixture of H2O-CH2C12 at room temperature failed. After workup of the reaction mixture, starting material was recovered. The epoxidation reaction should normally proceed from the bottom (a) face of the enone system of 231 to yield the cis-fused intermediate 269. However, in this case, the a face of the 0-carbon of the enone system in 231 is very sterically hindered by the OSiMe2But group at C-9 and, thus, the epoxidation  99  reaction does not occur. The protected alcohol group at C-9 of 231 is probably responsible for the failure of the epoxidation reaction. In fact, it was shown later (vide infra) that the enedione 267 which possesses a less hindering ketone group at C-9 undergoes epoxidation smoothly  and efficiently. It was felt that the enedione 267 represented a valuable intermediate for the synthesis of (±)-crinipellin B (15), especially since it was not possible to prepare the epoxide 269 directly from 231. The fact that it is possible to achieve in one step three synthetic operations (oxidative rearrangement, deprotection and oxidation of the secondary alcohol) also rendered the conversion of 249 into 267 very attractive. Moreover, Mehta et al. had reported 5 ° ,51 that it is possible to introduce chemoselectively a substituent on the enedione 160 at C-4 (crinipellin numbering). This selectivity would allow the installation of the exocyclic alkene moiety found in the crinipellins. Since compound 160 differs from 267 only with respect to the orientation of the isopropyl group at C-12 in ring D (crinipellin numbering), it can be expected  160  ^ 0 ^  o 267  that the reactivity mode of 267 would be very similar to that of 160. Therefore, alkylation of the bis-enolate generated from 267 should also occur chemoselectively at C-4. Consequently, efforts were invested to optimize the transformation of 249 into 267. Conditions were eventually found (PCC (5 equiv), Celite, CH2C12, 3.5 h) that allowed the isolation of 267, 231 and 268 in 53%, 11% and —6% yield, respectively. This result was very satisfactory as the desired enedione 267 was acquired in an acceptable yield in a single operation instead of three (if each step had proceeded in 80% yield, the enedione would have been obtained in 51% yield!). Moreover, the two products 231 and 268 could potentially be 100  useful. The epoxide 268 could become an intermediate in the synthesis of (±)-crinipellin B (15). The enone 231 could probably be converted into 267 after appropriate functional group  manipulations. A few features of the oxidation reaction need to be discussed. Decreasing the quantity of oxidizing agent resulted in a very sluggish reaction rate. An increase in the amount of reagent (greater than 5 equiv) did not seem to enhance the formation of 267, and introduced complications in the isolation step. Lengthening the reaction time did not provide 267 in better yield. After —60 h, the ratio of the isolated compounds 267, 231 and 268 was similar to that obtained after 3.5 h, but 267 was contaminated with a small amount of an impurity. When 231 was subjected to the reaction conditions described above, some 267 was formed but the  conversion was not significant. The pathway(s) for the transformation of 249 into 267 is (are) not immediately obvious. The various intermediates were characterized by spectroscopic methods. The enedione 267 showed by infrared spectroscopy two carbonyl stretching bands at 1739 and 1708 cm -1  and alkene absorptions at 1617 and 1607 cm -1 . The 1 H nmr spectrum of 267 exhibited a doublet at 5 5.89 with J = 2 Hz for H-6. The various 1 H nmr signals were assigned by COSY experiments. The results of these experiments are given in Table 8 in the experimental section. The 13 C nmr spectrum also witnessed the formation of the enone 267. Four carbon signals at 5 124.5 (C=CH), 190.4 (C=CH), 209.3 (C=0) and 215.4 L=O) indicated the presence of the enone moiety and of an additional carbonyl function. The enone 231 displayed, in its infrared spectrum, bands at 1707 and 1619 cm -1 for the ketone and alkene functions. Its 13 C nmr spectrum exhibited diagnostic signals at 5 126.3 (C=CH), 197.7 (C=CH) and 210.7 (C=0) for the enone moiety. The 1 H nmr spectrum of 231 was consistent with the proposed structure. Detailed 1 H nmr data (400 MHz, CDC13 and C6D6), derived from decoupling experiments, are compiled in Tables 9 and 10 in the experimental section. The  101  epoxide 268 was identical with the product obtained from the reaction of the enedione 267 with hydrogen peroxide (vide infra). 111.5.6. Preparation of the Enedione Epoxide 188.  Completion of the synthesis of one of the crinipellins (crinipellin A (43), for example) required, in theory, the execution of only a few more synthetic operations on 267 (amethylenation, oc-hydroxylation and epoxidation). The order in which these different steps should be performed was debatable. It was decided that the a-hydroxylation reaction had to be accomplished last, in view of the lability of the a-ketol moiety and its incompatibility with some of the reaction conditions to be used. It seemed possible to carry out a chemoselective a-methylenation on 267 (based on related work by Mehta et al. 5 ° ,51 ) and to epoxidize selectively the more strained double bond of the resultant dienone 270. A second option consisted  267  ^ o^ o o^ ^ 270  268  of first performing an epoxidation reaction on 267, and then executing an alkylation reaction on the resulting epoxide 268. Both sequences seemed viable. The stability of the keto epoxide under the reaction conditions necessary to realize the a-methylenation was not known. Therefore, the first alternative was favored and was employed in the exploratory work. The chemoselective introduction of the exocyclic methylene group at the C-4 position (crinipellin numbering) of 267 could be performed in a variety of ways. Many synthetic routes to a-methylene carbonyl compounds have been developed over the years because of the utility of such substances as intermediates in syntheses and because a large number of natural products, many of which are biologically active, possess an a-methylene lactone or an a102  methylene ketone moiety as a dominant structural feature. One expedient procedure to achieve such a transformation involves the highly electrophilic reagent dimethyl(methylene)ammonium iodide" (273, Eschenmoser's salt). This method was chosen to accomplish the desired transformation. 81 In order to study the behavior of the enone 267, a solution of this compound in THE at -78 °C was treated with a large excess of LDA (equation 24). The bis-enolate anion which is  0  267  . .. <1) LDA (excess), THF, -78 °C u Me ^lb.2) CH2=NMe2I (excess) 2 7 3 0 3) Mel, Me0H 4) NaHCO3, EtOAc, H2O  ^ 0 o^(24) ^ 270^ 271  43%  8%  presumably formed under these conditions was allowed to react with a large excess of Eschenmoser's salt 273. After an appropriate workup procedure, the crude reaction mixture was successively treated with methyl iodide in Me0H, and with NaHCO3 in a mixture of water and EtOAc. Chromatographic separation of the crude material furnished two major products, 270 and 271, in 43% and 8% yield respectively. The dienedione 270 showed by 1 H nmr spectroscopy (400 MHz, C6D6) three 1-hydrogen signals at 8 4.90, 5.99 and 6.10 for the vinylic hydrogens and two doublets with J = 17.5 Hz at 8 1.80 and 2.45 attributed to H-10 and H-10' (crinipellin numbering). The 1 H nmr spectrum (400 MHz, C6D6) of compound 271 displayed 5 signals associated with the presence of the vinylic hydrogens at 8 4.63, 4.85, 5.70, 6.01 and 6.06. Treatment of the product 270 with hydrogen peroxide provided the cis-fused intermediate 188 in yields ranging from 35-56% (equation 25). The workup and purification procedures affected the yields of the reaction. For example, incomplete removal of H202 led to  103  30% aq H202 ^I.-  (25)  THF, NaHCO3  270  ^ 0 ^  0 188 35-56%  the formation of side products. Furthermore, chromatographic purification of the crude material on silica gel caused decomposition of the enedione epoxide 188. The poor and irreproducible yields of this epoxidation reaction prompted us to study the second pathway in which the order of the epoxidation and a-methylenation steps were reversed. This sequence was ultimately adopted. The enedione 267 was transformed into the diketo epoxide 268 by reaction with hydrogen peroxide in the presence of NaHCO3 in aqueous THF (equation 26). The cis-fused  I  Il<  "Me  30% aq H202^< 1.-^ ....Me^(26) THF, NaHCO3  ^ o^  267  0 268 86%  dione epoxide 268 was produced in preference to the corresponding trans-fused substance because the latter material is highly strained. The product 268 was stable once it had been obtained as a pure substance. However, it was prone to decompose in the workup step if the crude material was heated over —40 °C. Chromatography of the crude reaction mixture on iatrobeads allowed the isolation of 268, as a white solid, in 86% yield. This substance exhibited, by 1 H nmr spectroscopy, a singlet at 5 3.24 characteristic of the hydrogen (H-6) of the epoxide function.  104  Efforts to achieve efficiently the conversion of 268 into 188 were at first not satisfying. Compound 268 was successively treated with LDA and dimethyl(methylene)ammonium iodide 273. Consecutive reactions of the crude material with Mel in Me0H, and then with NaHCO3 in aqueous THF produced 188 in —30% yield along with some recovered starting material (equation 27). Fortunately, after many trials, the procedure could be modified  <  1) LDA (excess), THF, -78 °C  ....Me  ....Me +^268^(27) 40%  DP-  2) CH2=NMe2I (excess) 2 7 3^0 3) MeI, Me0H 4) NaHCO3, EtOAc,  O  268  O  188  H2O  28%  to our advantage. Thus, compound 268 was allowed to react with lithium 1,1,1,3,3,3-hexamethyldisilazide in THF and the resultant solution was treated with 273 (equation 28). Two chromatographic purifications (on iatrobeads) of the crude product mixture afforded the desired  <  ....Me  0 268  0  1) (Me3Si)2NLi, THF, -78 °C 2) CH2=NMe2I (excess) 2 7 3 3) Flash chromatography  ...Me  0 188  ^+ 268^ (28) 7%  0  78%  enone 188 in 78% yield. Some unreacted starting material 268 was also recovered (-7% ). The species that resulted after alkylation was presumably the amino dione 272 which  Me 2 N  O  272  105  underwent an elimination reaction during the purification process. The use of iatrobeads as the solid phase in the flash chromatography was essential since 188 is unstable on silica gel. The intermediate 188 displayed all the expected spectral characteristics including, in the infrared spectrum, bands at 3104, 3058 and 1641 cm -1 associated with the alkene function and, in the 1H  nmr spectrum (Figure 8), two signals for the vinylic hydrogens at 8 5.45 and 6.12. The  various 1 H nmr signals were identified by COSY experiments. These data are listed in Table 11 in the experimental section of the thesis. Furthermore, the structure of 188 was proven by  an X-ray crystallographic analysis (Appendix 1) (Figure 7). This analysis showed that all the stereocenters of the carbon skeleton of the crinipellins had been installed stereoselectively in the desired sense and that a-methylenation had occurred exclusively at the C-4 center of 268.  Figure 7: Stereoview of the Enedione Epoxide 188  106  Figure 8: The 1 11 nmr Spectrum (400 MHz, CDCI 3 ) of the Enedione Epoxide 188.  111.5.7. Attempts to Prepare (±)-Crinipellin A (43).  An exciting stage of the synthesis had been reached. A last synthetic operation which would involve stereoselective a-hydroxylation of 188 needed to be performed in order to produce (±)-crinipellin A (43). A number of reagents have been developed recently to accomplish this type of transformation. A few of these methods allow direct enolate oxidation. The procedure by Davis and coworkers, 82 which involves the use of 2-(phenylsulfony1)-3phenyloxaziridine (274), 83 was chosen to attempt the last step of the synthesis. Examination of molecular models indicated that the two faces of the enolate resulting from 188 were similar and that a high stereoselectivity might be difficult to achieve. Nevertheless, it seemed possible to obtain at least some of the desired natural product. The intermediate 188 was allowed to react with HMDSK 84 in THF at -78 °C (equation 29). The resultant orange reaction mixture was treated with the oxidant 2-(phenylsulfony1)-3-  "Me  1) HMDSK, THF, -78 °C  ....m e  2) PhS02N - CHPh \ 0/ 274  (29)  OH  O  O  188  43  phenyloxaziridine (274). Tic analysis of the resultant solution revealed that all the starting material had been consumed. However, it was not possible to isolate any hydroxylated material from the crude reaction mixture. In order to test the stability of 188 under basic conditions, this compound was again subjected to deprotonation with HMDSK (equation 30). This time, the resulting solution was quenched with aqueous NaHCO3. The 1 H nmr spectrum of the crude material revealed that at least one new product had formed. Some enone 188 was also present in the mixture. However the recovery of the product 188 was low, and therefore  108  1) HMDSK, THF, -78 °C  (30)  2) aq NaHCO3  188  ^ 0 ^  0 188  attempts to achieve the hydroxylation reaction in one step were not pursued further. Since the enedione epoxide 188 could not be converted directly into 43, efforts were directed towards the preparation of the trimethylsilyl ether 275, which could, in theory, serve as a precursor for the preparation of (±)-crinipellin A (43). Thus, 188 was allowed to react with LDA and TMSC1 in THF (equation 31). 85 Unfortunately, no enol ether 275 was isolated. It was felt that sensitive functionalities in ring A of the intermediate 188 were the cause of the difficulties encountered. However, before envisaging important modifications in the last steps of the synthesis, we decided to gain more insight into the a-hydroxylation reaction and to study the behavior of the resultant a-hydroxy carbonyl compound(s).  ',... LDA, TMSC1, THF, -78 °C  (31)  ^Ow-  OSiMe 3  0  188  275  111.5.8. Model Studies on the a-Hydroxylation Reaction.  A substance that embeds part of the carbon skeleton of 188 would serve as a good model compound for the study of the a-hydroxylation reaction. The keto alkene 191 is a much simpler intermediate than 188 since it does not incorporate the A ring of the crinipellins and all the sensitive functionalities. However, it embodies the three rings that will most  109  188  0  influence the stereochemical outcome of the hydroxylation reaction. The tricyclic compound 191 would thus serve as a suitable model for the required investigation. Consequently, the  triquinane 191 was treated successively with HMDSK (1.5 equiv) and 2-(phenylsulfony1)-3phenyloxaziridine (274, 1.5 equiv), (equation 32). 82 The process was, surprisingly, very  L  Me  0  -  1) HMDSK, THF, -78 °C ,„.  —OH  2) PhSO2N — CHPh \ / 0 274  7* 16 0  191  Me"  (32)  Hio  276 > 90%  selective and afforded one major product (> 90% yield) whose structure was shown to be 276 by spectroscopic methods. The ketol 276 displayed, by 1 H nmr spectroscopy (400 MHz, C6D6), four singlets, integrating for 1 hydrogen each, at 5 2.38 (OH), 3.86 (H-10), 5.05 (H7*) and 5.31 (H-7*'). Two singlets at 8 0.78 and 1.05 were attributed to the two angular methyl groups. The two isopropyl methyl doublets appeared at 5 0.80 and 1.00. In NOE  H 7*  276  1 10  difference experiments, saturation of the signal at 5 3.86 (H-10) enhanced the resonances at 5 1.05 (Me-16) and 2.38 (OH) while irradiation of the singlet at 1.05 (Me-16) caused enhancement of the singlets at 5 3.86 (H-10) and 5.31 (H-7*'). The mutual increase in intensity of the carbinol signal (H-10) and the Me-16 resonance is only possible if the OH group has the a orientation as shown in 276. Therefore, the oxaziridine 274 approached the potassium enolate derived from 191 almost exclusively from the same side as the neighbouring methyl group (Me-17) to afford the ketol 276 with the configuration at C-10 opposite to that found in crinipellin A (43). The configuration at the C-10 center would need to be somehow inverted to provide the required stereochemistry. Initial trials to accomplish this inversion provided either a mixture of ketols (MeONa, Me0H) or recovered starting material. Alternatively, attempts could be made to reverse the selectivity of the a-hydroxylation process. However, due to time constraints and to the availability of relatively limited amounts of the keto alkene 191, it seemed preferable to examine other options. We were intrigued by the possibility of reducing chemo- and stereoselectively one of the two carbonyl functions of the 1,2-diketone 277 that would result from oxidation of the ketol 276 (Scheme 49). The a-ketol 276 was therefore subjected to mild oxidation  , 18L ....  , . .. 1) n  -Bu(i- 13102A1(H)Li, Ether^ ,...Meil n 0 ^ 2) aq NH4C1 (2 drops) ^0 3) filtration on 0^ 16 H9 OH florisil  Me TPAP, NMO  ...OH CH2C12, 0 276  molecular sieves  277 80% crude  278 > 90%  Scheme 49  conditions (tetra-n-propylammonium perruthenate, N-methylmorpholine N-oxide (NMO), 4 A molecular sieves, CH2C12). 86 After workup of the reaction mixture, the dione 277, a bright  111  yellow-orange solid, was obtained in 80% yield. This material showed by infrared spectroscopy, a strong band at 1742 cm -1 for a carbonyl stretch and an absorption at 1652 cm -1 attributed to the alkene function. The dione 277 was allowed to undergo reduction with a bulky reducing reagent. A cold (-78 °C) solution of 277 in ether was treated with a solution of n-Bu(i-Bu)2A1(H)Li (1.2 equiv) in ether (Scheme 49). As the addition proceeded, the yellow solution containing the diketone 277 faded and eventually became colourless. A ketol, whose 1 H nmr spectroscopic data differed from those displayed by 276, was isolated in excellent yield. This substance was shown to possess structure 278. Interestingly, when the reduction of 277 was attempted with 3 equivalents of the reducing reagent, the same product 278 was obtained. The various signals of the 1 H nmr spectrum (400 MHz, C6D6) of 278 were assigned by COSY and NOE difference experiments (the detailed data from these experiments have not been included in the thesis). Among these signals were found two singlets at 5 0.87 (Me-17) and 1.19 (Me-16), three multiplets at 8 1.08-1.15 (H-12), 1.34-1.44 (H-18) and 2.07-2.16 (H-3'), two doublets at 8 2.61 (OH) and 4.39 (H-9) and two singlets at 5 4.93 (H-7*) and 5.32 (H-7*'). Irradiation at 5 5.32 (H-7*') enhanced the doublet at 8 2.61 due to the OH proton, the singlet associated with Me-16 at 8 1.19 and the singlet attributed to H-7* at 5 4.93. Saturation at 5 1.19 (Me-16)  L RA  H12  „ ivie i7  o '''OH Me 16 H ..)1  H k..../H 7* To jiLal  278  112  9  induced enhancements of the signals at 4.39 (H-9) and at 5.32 (H-7*'). Irradiation of the doublet at 8 4.39 (H-9) caused an increase in the intensities of the signals at 8 1.08-1.15 (H12), 1.19 (Me-16) and 2.61 (OH). This last experiment proved that the reduction of one of the carbonyl groups of 277 occurred selectively from the 13 face. Since the newly formed ketol was different from 276, it therefore had to possess the structure 278. The stereoselective reduction of the dione 277 was a very encouraging result because it could provide a way to gain access to (±)-crinipellin B (15). 111.5.9. Completion of the Synthesis of (±)-Crinipellin B (15).  Knowing that it was possible to accomplish an a-hydroxylation reaction in the absence of the ring A of the crinipellins, it became evident that the problems associated with the conversion of 188 into an a-ketol were caused by the presence of the functional groups of ring A. Masking the reactive enone moiety of 188 in the form of a protected alcohol as shown in 279 should allow the a-hydroxylation to proceed (Scheme 50). Transformation of 188 into 279 would require protection of the alcohol function of the intermediate resulting from the  selective reduction of the less hindered C-5 carbonyl group of the diketone 188. It should be possible to hydroxylate stereoselectively the ketone 279 to yield the ketol 280. The results of the model study discussed above indicated that the ketol prepared from 279 by Davis procedure 82 should have the configuration at C-10 as shown in 280. The ketol 280 could be transformed into the triketone 281 by functional group manipulations. The last step of the synthesis would require the chemo- and stereoselective reduction of the C-9 carbonyl group of the intermediate 281. Based on the information derived from the model study, little doubt was left concerning the selectivity of the reduction of the dione system. However reduction could also occur at the C-5 center because the carbonyl function of the enone moiety is much less sterically hindered than either of the carbonyl groups of the diketone functionality. We hoped that the presence of the carbonyl function at C-10 would enhance the reactivity of the carbonyl at C-9 and favor reduction at this center to furnish (±)-crinipellin B (15). 113  188  PO  PO ^ 0 ^  279  ^ 0 ^  0  280  -4-  15  ^ OH ^  0  281 Scheme 50  The enone 188 was converted chemo- and stereoselectively to the alcohol 282 (NaBH4, Me0H-THF) in 80% yield (equation 33). The ketol epoxide 282 was accompanied  15^ ,....^ NaBH4 ,...Mein +^mime .,..me^ Me0H-THF -78 °C to HO"' HO''‘ "I..,^  0  .1-..  - 63 °C^H 5  H6  16  0^  OH  282 ^283  188  80%  (33)  by small amounts of side products, one of which was determined to be the product of bisreduction 283. The diol 283 showed by 1 H nmr spectroscopy two carbinol signals at 5 3.97 and 4.57. The configuration of the C-5 and C-9 centers was not proven, but was probably as shown in 283. Compound 282 exhibited in its it spectrum a broad signal at 3474 cm -1 for an alcohol group and a ketone absorption at 1737 cm -1 which indicated that only one of the  114  carbonyl moieties had been reduced. The 1 H nmr spectrum of 282 displayed a broad doublet (OH5) at 8 1.70 (J = 11 Hz) which exchanged upon treatment with D20 and a 1-hydrogen doublet with J = 1.5 Hz at 8 3.52 for H-6. The carbinol signal (H-5) was a broad doublet (J = 11 Hz) at 8 4.52 which, upon treatment with D20, became a broad singlet. The two vinylic hydrogen signals (H-15 and H-15') at 8 5.17 and 5.30 witnessed the replacement of the enone moiety by a non-conjugated alkene function. In decoupling experiments, irradiation of the carbinol signal (H-5) at 8 4.52 collapsed the doublet (H-6) at 8 3.52 into a broad singlet. Undoubtedly, the reduction took place at the C-5 center. A few other signals in the 1 H nmr spectrum of 282 were identified by decoupling and NOE difference experiments. The results of these experiments are listed in Table 12. The configuration of the new stereocenter of 282 was ascertained by NOE difference experiments (Table 12). Saturation at 8 4.52 (H-5) induced enhancement of the signals attributed to H-6 at 8 3.52 and H-15' at 8 5.30. Irradiation of the signal associated with H-6 at  282  ^  282  8 3.52 caused an increase in the intensities of the signals at 5 4.52 (H-5) and 0.97 (Me-16). These experiments established the cis relationship between H-5 and H-6 and allowed the  115  Table 12: 1 H nmr Data (400 MHz, CDC13) for the Ketol Epoxide 282a: Decoupling and NOE Experiments.  282  Signals Being Observed  Signal Being Irradiated Assignment H-x  1H nmr (400 MHz) 5 ppm (mult., J (Hz))  H-2 H-2'e (oc)  1.25 (dd, J = 13, 14) 2.40 (dd, J = 7.5, 14)  H-3  2.82 (ddm, J = 7.5, 13)  H-5  4.52 (br d, J = 11)  OH5  1.70 (br d, J = 11 Hz)  H-6  3.52 (d, J = 1.5)  H-10 (a)  2.30 (d, J = 17.5)  H-10' (03) 2.66 (d, J = 17.5) Me-16 Me-17  0.97 (s) 1.12 (s)  H-18  part of the m at 1.53-1.63  Me-19  0.85 (d, J = 6.5)  Me-20  0.95 (d, J = 6.5)  5 ppm (initial mult., J (Hz), H-x) to mult. after irradiation. J(Hz). 1.25 (dd, J = 13, 14, H-2) to d, J = 13 2.82 (ddm, J = 7.5, 13, H-3) to br d, J = 13 1.25 (dd, J = 13, 14, H-2) to d, J = 14 2.40 (dd, J = 7.5, 14, H-2') to d, J = 14 3.52 (d, J = 1.5, H-6) to s 1.70 (br d, J = 11 Hz, OH5) to s  NOE Correiationsb  H-6, H-15'  H-5, Me-16 2.66 (d, J = 17.5, H-10') to s 2.30 (d, J = 17.5, H-10) to s  0.85 (d, J = 6.5, Me-19) to s 0.95 (d, J = 6.5, Me-20) to s H-18, part of the m at 1.53-1.63 to sharpened m H-18, part of the m at 1.53-1.63 _ to sharpened m  H-6 H-3, H-2', H-10, H-18  a- Crinipellin numbering used for consistency. b- Only those NOE correlations that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  116  assignment of the two quaternary methyl groups and the two alkene hydrogens with their respective signals in the 1 H nmr spectrum. The data obtained from the irradiation of Me-16 and Me-17 are compiled in Table 12. The reduction reaction of 188 occurred on the less sterically encumbered ketone function and from the less sterically hindered face of the carbonyl group, opposite to the epoxide ring. The stereochemical outcome of the reduction of 188 could have been somewhat anticipated based on examination of molecular models and by analogy with the work accomplished by Matsumoto and coworkers in the course of their synthesis of hirsutic acid (284). 87 Matsumoto et al. performed the reduction of the linearly fused triquinane 136 with high stereoselectivity. They showed 87 that treatment of 136 with NaBH4 in EtOH at 0 °C for 15 min provided (±)-hirsutic acid (284) in 80% yield (Scheme 51). Reduction proceeded from the less hindered side of the carbonyl group, opposite to the epoxide and methyl moieties, to give the compound 284 that possessed the hydroxyl group with the a orientation. The two substances 136 and 188 have in common a ring (ring A) that includes a-methylene ketone and epoxide moieties. The A ring is cis-fused in each case to another 5-membered ring (ring B). The third ring of 136 and the remaining two rings of 188 should have a negligible effect on the outcome of the reduction process since they are well removed from the center undergoing  COOH^NaBH4 "Me^Et0H, 0 °C 15 min  HO H  136  NaBH4  ImMei7  Me0H-TEIF -78 °C^HO"' H5  0 188  H6  16 0  282 Scheme 51  117  80%  the reaction. The main difference between 136 and 188 therefore consists of the presence of the angular methyl group at C-3 in the former substance, and of a angular hydrogen at C-3 in the latter compound. Thus, the "bottom" face of 136 is more hindered than the "bottom face" of 188. Nevertheless, the reduction of complicatic acid (136) with NaBH4 to provide hirsutic acid (284) can still serve as a reliable model reaction for the conversion of 188 into 282. The stereoselective transformation of 136 into 284 provides further evidence for the assignment of the configuration at C-5 of 282. The ketol epoxide 282 was converted into the silyl ether 285 upon reaction with TBDMSOTf and Et3N in CH2C12 (equation 34). In the 1 H nmr spectrum of 285, two singlets  '....15^H^',...  Me2BuI Si0Tf  Et3 N, CH2C12,  "Me^(34)  ^N.-^  -78 °C^Me2ButSiO''‘%  o  o 282  285 96%  at 8 0.14 (6H) and at 8 0.92 (9H) witnessed the presence of the protecting group. The intermediate 285 was subjected to the cc-hydroxylation reaction conditions. Successive treatment of 285 with HMDSK 84 (1.5 equiv) and 2-(phenylsulfonyl)-3-phenyloxaziridine (274, 1.5 equiv) 82 furnished the ketol 286 (equation 35). The infrared spectrum of 286 ^ .... 1) HMDSK, THF, -78 °C  ...Me Me 2 ButSiOss°  0^-78 °C  ...Mei7  ^O.-  2) PhS02N — CHPh \ 0/ 274  Me 2 ButSiO•  '1.0H 0^Hio 16 0  286 ^(35)  285  68%  118  displayed a broad band at 3487 cm -1 . The 1 H nmr spectrum exhibited a doublet with J = 3.5 Hz at 5 2.49 (OH) which exchanged upon treatment with D20. Another doublet with J = 3.5 Hz at 5 4.06 was attributed to H-10. This signal became a singlet after D20 exchange. In order to prove the stereochemistry of the carbinol center (C-10) of 286, it was especially important to identify the 1 H nmr signals associated with H-12, Me-16 and Me-17. It has been demonstrated previously (in proving the structures of the products 221, 222, 276 and 278) that these signals were useful in the assignment of the configurations of the C-9 and C-10 centers (crinipellin numbering). The various signals of the 1 H nmr spectrum of 286 were assigned by COSY and NOE difference experiments (Table 13). NOE difference experiments served to confirm the configuration of the carbinol center (at C-10). Irradiation of  ....  4.__  ,^  H18  kiMe i7  Me 2 ButSi0.—  IV r a H6  e1611 0  ,11 12 Af--''OH Rio )'  286^  286  286  Table 13: 1 H nmr Data (400 MHz, CDCI3) for the a-Hydroxy Ketone 286a. 15  13  19 18  5  2)  Me 2 ButSiO'  286  Assignment H-x H-2 H-2'c (a) H-3 H-5 H-6 H-10 OH H-12 H-13 H-13' H-14 H-14' H-15 H-15' Me-16 Me-17 H-18 Me-19 Me-20  1H nmr (400 MHz) 8 ppm (mult., J (Hz)) 1.09 (dd, J = 12.5, 14) 2.42 (dd, J = 8, 14) 3.03 (dd, J = 8, 12.5) 4.66 (ddd, J = 2, 2, 3.5) 3.30 (d, J = 2) 4.06 (d, J = 3.5) 2.49 (d, J = 3.5) —1.27-1.39 (m), part of the m (2H) at 1.27-1.51 —1.40-1.51 (m), part of the m (2H) at 1.27-1.51 —1.84-1.94 (m), part of the m (2H) at 1.77-1.94 Part of the m (2H) at 1.55-1.71 Part of the m (2H) at 1.77-1.94 5.09 (m) 5.15 (m) 1.03 (s)  COSY Correlationsb H-2', H-3 H-2, H-3 H-2, H-2', H-5, H-15, H-15' H-3, H-6, H-15, H-15' H-5, H-15 OH H-10 H-13, H-13', H-18 H-12, H-13', H-14, H-14'  NOE Correlationsb  H-15  H-5, Me-16 OH, H-12 H-10, (OH neg)  H-12, H-13, H-14, H-14' H-13, H-13', H-14' H-13, H-13', H-14 H-3, H-5, H-6, H-15' H-3, H-5, H-15  H-3, H-15' H-5, H-15 H-6, H-10, H-14'  0.98 (s) —1.60-1.71 (m), part of the m H-12, Me-19, Me-20 (2H) at 1.55-1.71 H-18 0.89 (d, J = 6.5) 1.00 (d, J = 6.5) H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  120  the doublet at 8 4.06 (H-10) enhanced the signals due to H-12 and to the hydroxyl proton. Saturation of the multiplet at 8 —1.27-1.39 (H-12) increased the intensity of the doublet at 8 4.06 (H-10). Small enhancement of the H-10 signal was noticed upon irradiation of Me-16 at 8 1.03. These data indicated that hydroxylation of the potassium enolate derived from 285 took place from the same side as the neighbouring methyl group Me-17. Hydroxylation of 285 from the bottom face was, in fact, expected in view of the results obtained in the case of  the model compound 276. The stereochemistry at C-10 is opposite to that found in crinipellin A (43). The ketol 286 was transformed into the triketone 281 in two steps (equation 36). The silyl ether function of 286 was cleaved with tetra-n-butylammonium fluoride. This process afforded a mixture of ketols as indicated by the 1 H nmr spectrum of the crude reaction mixture. The sensitive ketol group of 286 had epimerized (or isomerized) during the deprotection reaction. The two hydroxyl groups of the intermediate diols were oxidized with the periodinane 287. 88 A suspension of Dess-Martin reagent 287 in dichloromethane was slowly added to a solution of the intermediate diols and pyridine in dichloromethane at room temperature (equation 36). After workup of the reaction mixture and purification of the crude  '....  1) n-Bu4NF, THE 2)  4..  .... Me^(36)  ^).-  -  Me 2 Bu tSiO'%*^'''OH 0^287 0 ^ Pyr, CH2C12 286 Z = I(0A0)3  0 0  281  44%  material by flash chromatography on iatrobeads, the triketone 281, a bright yellow-orange solid, was isolated in 44% yield from 286 as the only product. The yield obtained for the conversion of 286 into 281 was disappointing even though three functional group transformations had been accomplished in two steps. It is possible that the conversion of 286 121  into the triketone 281 affords side products such as the anhydride 288. Since workup involved the use of aqueous base, this substance would have been converted into the carboxylate salt of the diacid 289 and, therefore, would not have been isolated. The trione  .....  „0  4  Me  COOH COOH  0 289  288  281 displayed one carbonyl absorption at 1732 cm -1 in its it spectrum. The 1 H nmr spectrum  of 281 showed two vinylic hydrogen signals whose chemical shifts (5.43 and 6.11) indicated that the enone moiety had been regenerated. The  13 C  nmr spectrum of 281 revealed the  existence of three carbonyl groups at 8 195.4,207.7 and 208.7. The chemoselective reduction of the C-9 carbonyl group in 281 was then attempted. A solution of the triketone 281 in a mixture of THE-ether was treated with one equivalent of a solution of n-Bu(i-Bu)2A1(H)Li (equation 37). As the addition proceeded, the yellow solution  4  ,..  1) n-Bu(i-B1)2A1(H)Li (1 equiv)  ^ unMe 1 0^THF-Ether, -78 °C  11.-  0^2) NH4CI  o  o  281  290  did not undergo any colour change. This observation seemed to indicate that reduction had occurred at the less hindered C-5 carbonyl group. The 1 H nmr spectrum of the major product proved this assumption to be correct and displayed all the characteristics expected for structure 290. The 1 H nmr spectrum of 290 displayed two vinyl signals at 8 5.13 and 5.30 which  showed that the olefin moiety was no longer conjugated with the ketone group and that 122  therefore this carbonyl group had been reduced to an alcohol function. A broad doublet with J = 10.5 Hz at 8 4.50 revealed the presence of the C-5 carbinol hydrogen. The configuration of 290 at C-5 was not proven but was assumed to be as shown based on subsequent results (vide  infra). As a last resort, we decided to effect the double reduction at the C-5 and C-9 carbonyl  centers of 281 and to attempt the chemoselective oxidation of the resultant diol. The oxidation of the secondary allylic alcohol function should be a faster process than the oxidation of the secondary alcohol. Reaction of the triketone 281 with n-Bu(i-Bu)2A1(H)Li (excess) furnished a major product (41% yield), that was subsequently shown to be the diol 291, along with a mixture of other diols (--30%) (equation 38). The diol 291 revealed, by it spectroscopy, a broad alcohol absorption at 3446 cm -1 and a ketone stretching band at 1730 cm -1 . The 1 H nmr spectrum displayed two carbinol resonances at 8 4.50 and 4.68. These signals along with the other resonances of the 1 H nmr spectrum of 291 (Figure 9) were assigned by COSY and NOE  meMe  1) n-Bu(i-Bu)2A1(H)Li (excess)  THF-Ether, -78 °C 2) NH4Cl 0^ O  281  291  41%  difference experiments (Table 14). The structural formula of 291 was proven by NOE difference experiments and by its successful transformation into (±)-crinipellin B (15). The connectivities between the various hydrogens of the A and B rings of the tetraquinane 291 were established by a 2D homonuclear correlation experiment (Table 14). The two alkene singlets (H-15 and H-15') of 291 showed correlations with one of the carbinolic signals (H-5 at 8 4.50) and the angular hydrogen H-3 (8 2.48). The two hydrogens  123  S. 0^4. 0^3. 0^2.0^ 1.0 PPM  Figure 9: The 1 H nmr Spectrum (400 MHz, CDC1 3  ) of the Diol 291.  Table 14: 1 H nmr Data (400 MHz, CDC13) for the Diol 291a. 15  13  14  12 11  me  5  HO%S  ,  10 16  19  '4 a 17  OH  291  Assignment H-x  iii nmr (400 MHz) 8 ppm (mult., J (Hz))  COSY Correlationsb  H-2 H-2'c H-3 H-5 OH5 H-6 H-9  1.16 (dd, J = 13.5, 13.5) 2.10 (dd, J = 7, 13.5) 2.48 (br dd, J = 7, 13.5) 4.50 (br d, J = 10.5) 1.70 (br d, J = 10.5) 3.39 (d, J = 2) 4.68 (d, J = 6)  H-2', H-3 H-2, H-3 H-2, H-2', H-15, H-15' OH5, H-6, H-15, H-15' H-5 H-5, H-15 OH9  OH9 H-12d  3.02 (d, J = 6) Part of the m (2H) at 1.66-1.78  H-9 H-13, H-13', H-18  H-13 H-13'd H-14 H-14'd H-15 H-15' Me-16  Part of the m (3H) at 1.43-1.65 1.95-2.04 (m) Part of the m (3H) at 1.43-1.65 2.24 (ddd, J = 8.5, 8.5, 14) 5.09 (br s) 5.26 (br s) 1.30 (s)  Me-17  1.06 (s)  H-18 Me-19 Me-20  Part of the m (3H) at 1.43-1.65 0.79 (d, J = 6.5) 0.87 (d, J = 6.5)  NOE Correiationsb  H-5, Me-16 OH9, H-12, Me-16 H-9, (OH9 neg)  H-12, H-13, H-14, H-14' H-13, H-13', H-14 H-3, H-5, H-6, H-15' H-3, H-5, H-15  H-6, H-9 H-14' H-2' (a), H-3, H-18  H-12, Me-19, Me-20 H-18 H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2). d- This hydrogen has been assigned by comparison with the 1 H nmr spectrum of crinipellin B (15).  125  adjacent to H-3 (H-2 and H-2') were assigned through their correlation with H-3. The signals attributed to H-12, H-13, H-13', H-14 and H-14' were identified by comparison with the 1 H nmr spectrum of crinipellin B (15). In NOE difference experiments, irradiation of the methyl singlet at 6 1.06 caused enhancements of the signals associated with H-2' (a), H-3 and H-18. This angular methyl group was attributed to Me-17. Saturation of the other angular methyl singlet at 6 1.30 (Me-16) increased the intensities of the signals due to H-6, H-9 and  HO  ....  H5  291  ^  291  H-14'. Irradiation of the doublet at 6 4.68 (H-9) amplified the signals associated with H-12, Me-16 and OH9. A last experiment demonstrated that the H-9 doublet also increased in intensity upon saturation of the H-12 resonance. Undoubtedly, the reduction of 281 proceeded from the f3 face of the dione system and was correctly assumed to have occurred on the C-9 carbonyl group (vide infra). The configuration at C-5 was believed to be as shown in 291, based on previous results (see above, transformation of 188 into 282). Indeed, a NOE  difference experiment indicated the cis relationship of H-5 and H-6. Irradiation of the signal at 6 3.39 (H-6) caused enhancements of the broad doublet at 6 4.50 (H-5) and of the singlet due to Me-16 at 6 1.30. Another proof of the structure of 291 was obtained by comparing the 1 H nmr spectrum (CD3OD) of 291 with the 1 H nmr spectrum (CD3OD) of the naturally occurring  126  HO• HO^ OH^ 135^  „, 291  OH  (-)-dihydrocrinipellin B (135), 17 which possesses the configuration at C-5 opposite to that in 291. 89 The two spectra were clearly different. The 1 H nmr spectrum of 291 displayed two doublets with J = 6.5 Hz at 8 0.78 and 0.88 (Me-19, Me-20), two singlets at 8 1.01 and 1.28 (angular methyl groups), a multiplet integrating for one hydrogen at 8 2.24-2.35 and a dd (J = 6.5, 12.5 Hz, 1H) at 8 2.40. The signal associated with H-6 was hidden under the  methanol signal at 8 3.30. The presence of the two carbinol hydrogens was revealed by the singlets at 8 4.52 and 4.76. Finally, two multiplets due to the vinylic hydrogens were found at 8 5.03 and 5.12. The 1 H nmr spectrum of 135 exhibited signals associated with Me-19, Me20, Me-17 and Me-16 at 8 0.83, 0.93, 1.07 and 1.35, respectively. A multiplet integrating for 2 hydrogens was present at 8 2.37. Five signals attributed to H-6, H-5, H-9, H-15 and H-15' were found at 8 3.26, 4.30, 4.80, 5.13 and 5.22 respectively. We had thus achieved the synthesis of (±)-5-epi-dihydrocrinipellin B (291). In order to complete the synthesis of (±)-crinipellin B (15), the alcohol function at the C-5 center of the diol 291 needed to be oxidized chemoselectively. The oxidizing agent had to be chosen judiciously among the wide number of oxidants that are known. The reagent derived from the pyridine-sulfur trioxide complex and DMSO9° is known for its mildness and was used to accomplish the delicate transformation of 291 into 15 (equation 39). A solution of the diol 291 in CH2C12-DMSO was allowed to react with Pyr•S03 in the presence of triethylamine. After an appropriate workup of the reaction mixture and separation of the crude material by flash chromatography on iatrobeads, three substances were isolated. Fortunately, the major product was (±)-crinipellin B (15) (49% yield). The overoxidized substance 281  127  was produced in 9% yield. This product was identical with the compound obtained previously (conversion of 286 into 281). Starting material 291 was also recovered in 12 % yield.  (39) +  OH  0  291  281  12%  9%  (±)-Crinipellin B (15) was characterized by a number of spectroscopic methods. Infrared spectroscopy showed absorptions for the hydroxyl and ketone groups at 3482 (broad) and 1730 cm -1 . The 1 H nmr spectrum of 15 (Figure 10) displayed two doublets at 5 2.93 (OH) and at 4.75 (H-9) and two signals at 5 5.37 and 6.08 for the alkene hydrogens (H-15 and H-15') of the enone moiety. The 13 C nmr spectrum indicated the presence of two carbonyl functions at 8 196.8 and 217.5. An APT experiment allowed the differentiation of the signals due to quaternary carbons and to methylene (CH2) carbons from those associated with methine (CH) and methyl (CH3) carbons (See Table 16). Most of the signals of the 1 H nmr and 13 C nmr spectra were assigned through the use of 1 H, 1 H-homonuclear correlation and 1 H, 13 Cheteronuclear correlation 2D nmr spectra (COSY and HMQC experiments respectively; see Tables 15 and 16). A HMBC experiment provided evidence that the hydroxyl group was  128  Figure 10: The 1 H nmr Spectrum (400 MHz, CDC1 3 ) of Synthetic (±)-Crinipellin B (15).  Table 15: 1 H nmr Data (400 MHz, CDC13) for (±)-Crinipellin B (15)a.  Assignment H-x  1H nmr (400 MHz) 8 ppm (mult., J (Hz))  COSY Correlationsb  H-3 H-6 H-9  1.24 (dd, J = 13, 14) Part of the m (2H) at 2.20-2.30 2.71 (ddm, J= 7, 13) 3.31 (s) 4.75 (d, J = 6.5)  H-2', H-3 H-2, H-3 H-2, H-2', H-15, H-15' H-15 OH  OH H-12  2.93 (d, J = 6.5) 1.76 (ddd, J= 7, 10, 11.5)  H-9 H-13, H-13', H-18  H-13 H-13'  Part of the m (3H) at 1.45-1.68 1.97-2.06 (m)  H-14 H-14' H-15 H-15' Me-16  Part of the m (3H) at 1.45-1.68 Part of the m (2H) at 2.20-2.30 5.37 (br s) 6.08 (d, J = 1.5) 1.33 (s)  Me-17  1.11 (s)  H-18 Me-19 Me-20  Part of the m (3H) at 1.45-1.68 0.81 (d, J = 6.5) 0.88 (d, J = 6.5)  H-2 H-2'c (a)  NOE Correia_^tionsb  OH, H-12, Me-16 H-9, (OH neg)  H-12, H-13, H-14, H-14'  H-3, H-6, H-15' H-3, H-15 H-6, H-9 H-14' H-2' (a), H-3, H-18 Me-19, Me-20 H-18 H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  130  Table 16: 1 H nmr (500 MHz) and 13 C nmr (125.8 MHz) Data for (±)Crinipellin B (15)a. ^ 15  14  13  12  19 18 17 23  1141  nu  10 16 =^°  15  E N T  R  C-x  13 C nmr spectrum (125.8 MHz) 8 ppm, APTb  OH  IIMQCcoi  1H-13C HMBCc4  8 ppm (assignment)  Correlations  1H nmr Correlations (500 MHz)  Long-range H-x  Y a  2  38.9  b  3  42.5 (-ye)  c  4  145.1  H-2, H-6, H-15'  d  5  196.8  H-6, H-15, H-15'  e  6  55.7 (-ye)  3.31 (H-6)  f  9  79.8 (-ye)  4.75 (H-9)  g  10  217.5  h i  12 13  60.7 (-ye) 30.1  j  14  33.9  k  15  122.8  1  16  21.4 (-ye)  1.33 (Me-16)  H-2 (4 bonds), H-9  m  17  10.3 (-ye)  1.11 (Me-17)  H-2 (4 bonds)  n  18  29.9 (-ye)  Part of the m (3H) at 1.45-1.68 (H-18)  o  19  21.4 (-ye)  0.81 (Me-19)  1.24 (H-2) Part of the m (2H) at 2.20-2.30 (H-2') 2.71 (H-3)  H-2, H-2'e, Me-16 (4 bonds)  Me-16 H-9, Me-17  1.76 (H-12) Part of the m (3H) at 1.45-1.68 (H-13) 1.97-2.06 (H-13') Part of the m (3H) at 1.45-1.68 (H-14) Part of the m (2H) at 2.20-2.30 (H-14') 5.37 (H-15) 6.08 (H-15')  Me-17, Me-19, Me-20 H-2  Me-20  Me-19 0.88 (Me-20) p 20 22.7 (-ye) a- The quaternary carbon signals in the 13 C nmr spectrum of 15 have not been included in the table. b- The results of the APT experiment are given in parentheses eve for CH and CH3 carbon signals). c-The table reads from left to right. The assignment and the chemical shifts of the 13 C nmr spectrum are listed in the first and second columns, respectively. The third column shows the 1 H nmr signal(s) which correlate(s) with the carbon of the first two columns, as obtained from the HMQC experiment (1 bond correlation). The last column lists the hydrogen(s) which correlate(s) with the 13 C nmr signal of the first two columns as obtained from HMBC experiments (2, 3 and 4 bonds correlation (s)). d- Only those HMQC and HMBC data that could be unambiguously assigned are recorded. e- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  131  situated on the C-9 carbon (conversion of 281 into 291). The stereochemistry at C-9 was confirmed by NOE difference experiments. The results of these experiments are discussed below. Some signals of the 1 H nmr spectrum of 15 were readily identified because of their distinctive chemical shifts and coupling patterns (H-3, H-6, H-9, OH9, H-15, H-15', Me-19 and Me-20). A few resonances were assigned through their connectivity with known signals in the COSY spectrum (H-2, H-2' and H-18). The signal associated with H-12 could not be identified directly through its correlation with H-18 since the H-18 resonance is embedded in a 3-hydrogen multiplet. It was necessary to use the HMQC experiment to determine the position of the H-12 signal in the 1 H nmr spectrum (vide infra). After assignment of H-12 to its corresponding resonance, the remaining signals (H-13, H-13', H-14 and H-14') were easily identified from the COSY spectrum. The data derived from the 1 H, 13 C-correlation 2D nmr spectrum (HMQC experiment, Table 16) allowed the identification of the signal attributed to H-12 (a ddd at 8 1.76) in the 1 H  nmr spectrum of 15. The 2D spectrum established the link between the methine (CH) carbon signals from the 13 C nmr spectrum and their respective hydrogen resonances (CH) in the 1 H nmr spectrum (see entries b, e, f and n, Table 16). All the C/H pairs had been identified but one which was attributed to the C-12/H-12 pair (entry h, Table 16). NOE difference experiments were carried out to confirm the configuration of the C-9 center and to assign Me-16 and Me-17 to their respective 1 H nmr singlets (Table 15). Irradiation of the resonances due to H-9 and H-12 at 8 4.75 and 1.76 caused mutual enhancements of these two signals and proved that the hydroxyl group had the a orientation as in 15. The oxidation of 291 into 15 therefore preserved the integrity of the C-9 carbinol center. Saturation of the singlet at 8 1.33 (Me-16) caused increase in the intensities of the signals due to H-6, H-9 and H-14'. Upon irradiation of the other angular methyl group (Me17), amplification of the multiplets attributed to H-2', H-3 and H-18 was observed. 132  15  The position (at the C-9 or C-10 center) of the hydroxyl group was confirmed by a HMBC experiment (see Table 16). The 2D long-range 1 H, 13 C-heteronuclear correlation nmr spectrum showed correlations between C-9 and CH3-16 (3 bonds) and C-10 and C1_13-17 (3 bonds). These data provided an independent proof that the hydroxyl group was situated on C-9 and that, therefore, reduction had occurred as predicted on the C-9 carbonyl group. Moreover, the 1 H nmr spectrum of our synthetic (±)-crinipellin B (15) was found to be identical with that of natural crinipellin B. 91 Even though we had successfully completed the total synthesis of (±)-crinipellin B (15), we decided to try to improve the efficiency of the conversion of 286 into 291. The  ketol 286 was thus oxidized with tetra-n-propylammonium perruthenate (TPAP) in the presence of N-methylmorpholine N-oxide (NMO) and 4 A molecular sieves in CH2C12 (Scheme 52). After flash chromatography of the crude material on iatrobeads, the diketone 292, a bright yellow solid, was obtained as a relatively pure product (tic analysis of the  material showed the presence of one compound). However, if the compound 292 was left under high vacuum overnight at room temperature, it decomposed partially. Tlc analysis of  133  this material revealed that more polar substances had formed. Attempts to recrystallize the diketone 292 also led to decomposition of the material.  ....m e  TPAP, NMO  CH2C12 '''C)H^4 A molecular sieves Me2ButSiCr*.  Me 2 ButSie  0  O 292 74%  286  n-Bu4NF, THE  1)n-Bu(i-Bu)2A1(H)Li (excess)  -•[^  THF-Ether, -78 °C 2) NH4C1  0 290 68% Scheme 52  The unstable dione 292 was allowed to react with tetra-n-butylammonium fluoride to afford, in 50% yield from 286, the alcohol 290, which also showed signs of instability. The alcohol 290 could be purified by chromatography. However, tic analysis of the purified substance revealed the presence of small amounts of polar product(s). The it spectrum of 290 exhibited an hydroxyl absorption at 3475 cm -1 and two ketone stretching bands at 1751 and 1742 cm -1 . The diketo alcohol 290 was treated with an excess of n-Bu(i-Bu)2A1(H)Li to furnish the diol 291 in 53% yield. The intermediate 291 thus obtained was identical with the same compound prepared from the reduction of the triketone 281. This synthetic pathway was  134  slightly more efficient than the previous one. However, since two intermediates were unstable, this alternative pathway was, in practice, less convenient. 111.6. CONCLUSION  The synthesis of (±)-crinipellin B (15) has been successfully accomplished in 22 steps from the simple starting material 2-methyl-2-cyclopenten-l-one. The various stereocenters of (±)-crinipellin B (15) were installed cleanly and efficiently and, in most instances, with high stereoselectivities. Two new annulation methods developed in our laboratories played important roles in the assembly of the required tetraquinane carbon skeleton. The sequence elaborated by Piers and Marais allowed the efficient conversion of the bicyclic enone 194 into the triquinane 191. A new cyclopentenone annulation procedure was developed during the course of the synthesis to construct the fourth five-membered ring of 15 since known methods failed to accomplish the desired transformation of 224 into 267. The last part of the synthesis involved functionalization of the A and D rings of crinipellin B (15). The entire synthesis of (±)-crinipellin B (15) is outlined in Scheme 53.  135  b, c  a Me 3 SiO  ,,,,,  R 204R = CH2 196R =0  200  0 194 e, fi  h, i, j  g  R^OR'  0  221 R= CH2, R' = H 223 R = CH2, R' = TBDMS 224 R = 0, R' = TBDMS  219 R = Me3Ge 220 R = I  191  m  1  0  0^-OTBDMS 250  249  267  (a) i-PrMgBr, CuBr•Me2S, Me3SiC1, HMPA, THF, -78 °C, 4 h; Et3N (94%); (b) MeLi, THF, 0 °C; 2-bromomethyl-1-butene, (Ph3P)4Pd, THF, 2 h, -20 °C; 0 °C, 5 h (76%); (c) 03, Me0H-CH2C12, -78 ° C; Me2S, -78 °C to rt; concentrate mixture, add 10% HC1-H20 and THF, stir at rt for 18 h (93%); (d) Me0Na, Me0H, reflux, 15 h (97%); (e) reagent 209, Me3SiBr, THF, -78 °C, 8 h; - 48 °C, 2 h (83%); (f) 12, CH2C12, rt (98%); (g) (Ph3P)4Pd (19 mol %), t-BuOK, t-BuOH, THF, rt (84%); (h) n-Bu(i-Bu)2A1(H)Li, Ether, -78 °C, 2 h; 0 °C, 1 h (93%); (i) t-BuMe2SiOSO2CF3, Et3N, CH2C12, -78 °C, 75 min; 0 °C, 20 min (98%); (j) 0s04, C5H5N, rt, 23 h; NaHSO3, H2O, 1 h; Pb(OAc)4, THF, rt, 30 min (93%); (k) i-Pr2NLi, THF, -78 °C; (Z)-3-bromo-l-iodopropene, rt, 7.5 h (76%); (1) n-BuLi (2.5 equiv), THF, -78 °C, 110 min (93%); (m) C5H5N•Cr03•HC1, CH2C12, Celite, rt, 3.5 h (51 %). Scheme 53  136  267  n, o  P, q „,  (5  0 268 R = H2 188 R = CH2  282 R = H 285 R = TBDMS  s, t  v HO'‘s 291  (n) H202, NaHCO3, H2O-THF (1:2), rt, 55 min (84%); (o) (Me3Si)2NLi, THF, -78 °C; (H2C=NMe2)+I , -78 °C, 70 min; -70 °C, 18 min; flash chromatography (iatrobeads) (78%); -  (p) NaBH4, Me0H-THF, -78 °C, 85 min; -63 °C, 15 min (80%); (q) t-BuMe2SiOSO2CF3, Et3N, CH2C12, -78 °C, 2 h; 0 °C, 110 min (88%); (r) (Me3Si)2NK, THF, -78 °C; 2-(phenylsulfonyl)-3-phenyloxaziridine, -78 °C, 45 min (68%); (s) n-Bu4NF, THF, rt, 75 min; (t) DessMartin periodinane reagent (4 equiv), C5H5N (2 equiv), CH2C12, rt; Na2S2O3, NaHCO3, H2O (44% from 286); (u) n-Bu(i-Bu)2A1(H)Li (4.6 equiv), Et20-THF, -78 °C, 30 min (41%); (v) C5H5N•503, DMSO, Et3N, CH2C12, rt, 9.5 h (49%).  Scheme 53 (continued).  137  IV. GENERAL CONCLUSION. In summary, this thesis describes the syntheses of three target molecules whose carbon skeleton is formed of fused five-membered rings. The first two target compounds, methyl cantabrenonate (13) and methyl epoxycantabronate (14), are methyl ester derivatives of the naturally occurring cantabrenonic acid (33) and epoxycantabronic acid (34). Since the two acids 33 and 34 were characterized as their methyl esters 13 and 14, it seemed appropriate to synthesize these latter substances for comparison purposes. The third target compound synthesized was (±)-crinipellin B (15), one of the five related tetraquinane natural products found in nature.  CO2Me 13  s'CO2Me 14  0  15  OH  The preparation of (±)-methyl cantabrenonate (13), (±)-methyl epoxycantabronate (14) and (±)-crinipellin B (15) demonstrated the synthetic applicability of two complementary methylenecyclopentane annulation procedures developed in our laboratories and led to, in the synthesis of (±)-crinipellin B (15), the discovery of a new synthetic method. The methylenecyclopentane annulation sequence that was required for the conversion of the bicyclic enone 62 into the angularly fused tricyclic keto alkene 60 (syntheses of (±)-13 and (±)-14) involved the  CI d^a 20  62  60  138  Li(CN)Cu^CI 18^20  reagent 18 derived from the precursor 4-chloro-2-trimethylstannyl-1-butene (6). The cuprate reagent 18 acted as a synthetic equivalent to the synthon 20. In the annulation procedure that allowed the transformation of 194 into 191 (synthesis of (±)-crinipellin B(15)), the conjunctive reagent 209, obtained from 4-iodo-2-trimethylgermyl-1-butene (208), acted as a  Me3Ge  208  193  Me3Ge  o  Cu(CN)Li 209  194  a^d 193  191  synthetic equivalent to the donor-acceptor synthon 193. These two regioisomeric methylenecyclopentane annulations can be summarized as illustrated below.  8  ^  7  ^  9  In each synthesis, a number of obstacles related to structural features and inherent reactivity of the various intermediates involved were encountered. For example, in the synthesis of (±)-crinipellin B (15), it was not possible to introduce directly the a-hydroxyl group on the intermediate 188. Consequently, the initial plans towards the syntheses of the  0 188  139  target molecules had to be revised to overcome these problems. In a few instances during the course of the syntheses, conversion of one intermediate to another was achieved with high and unpredictable stereoselectivity. The transformation of the enone 120 into the ketone 126 is one such example. Spectroscopic methods (particularly NOE difference experiments) served to determine the configuration of the newly created center(s) and, thereby, to gain more insights into the stereochemical outcome of reactions involving systems formed of 5-membered rings.  H2, Pd-C  EtOAc, rt,^3,... 22 psi Me 0  126  Although a wide variety of naturally occurring substances that embed polyquinanes in their skeleton have been discovered since the early 1970's, these compounds continue to fascinate the community of organic chemists. Each year, new substances containing fused 5membered rings continue to be isolated and characterized. Nature is likely to provide more polyquinane-containing natural products that will be of interest to synthetic organic chemists.  140  V. EXPERIMENTAL SECTION. V.1. GENERAL. V.1.1. Data Acquisition and Presentation.  Proton nuclear magnetic resonance ( 1 H nmr) spectra were recorded on either a Bruker model WH-400 or AMX-500 spectrometer using deuteriochloroform (CDC13) as the solvent, unless otherwise noted. Signal positions (8 values) are given in parts per million and were measured relative to the signals for tetramethylsilane (TMS) for the first part of the thesis (synthesis of (±)-methyl cantabrenonate (13) and (±)-methyl epoxycantabronate (14)) or chloroform (8 7.26) for the second part of the thesis (synthesis of (±)-crinipellin B (15)), unless otherwise noted. Coupling constants (J values) are given in Hertz (Hz). The multiplicity, number of hydrogens, coupling constants, and assignments (when known) are given in parentheses. Abbreviations used are: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad. When a hydrogen was observed to be coupled with the same coupling constant to two, three, four or five hydrogens which are chemically and magnetically nonequivalent, the designation dd, ddd, dddd and dddt are used instead of using t, q, quintet and sextet, respectively. For compounds exhibiting AB and ABX type spin systems, the quoted values for chemical shifts and coupling constants are measured as if they were first order systems, although these values only approximate the real values. 92 In the 1 H nmr spectra, H-x and H-x' have been used to designate hydrogens on the same carbon, with H-x' being the hydrogen at lower field. Moreover, a star (*) has been used to indicate a hydrogen on a carbon that will not be found later on in crinipellin B (15). Decoupling experiments refer to 1 H- 1 H spin decoupling experiments. Carbon nuclear magnetic resonance ( 13 C nmr) spectra were recorded on a Varian model XL-300 spectrometer at 75.3 MHz or on a Bruker model AMX-500 (125.8 MHz) spectrometer, using deuteriochloroform as the solvent, unless otherwise noted. Signal 141  positions are given in parts per million (8) from TMS, measured relative to the chloroform signal at 8 77.0. 93 Signals with negative intensity in the attached proton test (APT) are so indicated in brackets (-ve) following the chemical shift. Infrared (ir) spectra were recorded on liquid films between sodium chloride plates or on potassium bromide pellets using either a Perkin-Elmer model 1710 or 1600 Fourier Transform Spectrophotometers with internal calibration. Low and high resolution electron impact mass spectra were recorded on a Kratos MS50 mass spectrometer (70 eV). Desorption chemical ionization mass spectra (DCIMS) were recorded with a Delsi Nermag R-10-10 C mass spectrometer. Elemental analyses were performed on a CARLO ERBA CHN elemental analyzer, model 1106, by the UBC Microanalytical Laboratory. Gas-liquid chromatography (glc) analyses were performed on Hewlett-Packard model 5880A or 5890 capillary gas chromatographs, both using a flame ionization detector and a fused silica column, either —20 m x 0.21 mm coated with cross-linked SE-54 or —25 m x 0.20 mm coated with cross-linked 5% phenyl-methyl silicone. Thin layer chromatography (tic) was carried out on commercial aluminium backed silica gel 60 plates (E. Merck, type 5554, 0.2 mm) Reverse phase tic was performed on commercially available, glass backed plates (Watman, type KCI8/KCI8F). Visualization was accomplished with either ultraviolet light (254 nm), a solution of phosphomolybdic acid (PMA) in EtOH (20% w/v, Aldrich), a 5% aqueous solution of ammonium molybdate in 10% aqueous sulfuric acid (w/v) or iodine stain. Flash chromatography 32 was performed using 230-400 mesh silica gel (E. Merck, Silica Gel 60). Chromatography of sensitive compounds was done using iatrobeads 6RS-8060 from Iatron Laboratories, Inc. 11-4, Higashi-Kanda 1-Chome,  142  Chiyoda-Ku, Tokyo, 101 Japan. (TELEX: 02656098 Iatron J; FAX: 03-3865-1610; PHONE: 03-3862-1761). Melting points were measured on a Fisher-Johns melting point apparatus and are uncorrected. Distillation temperatures refer to air-bath temperatures of Kugelrohr distillations and are uncorrected. All compounds that were subjected to high resolution mass spectrometry and elemental analysis were homogeneous by tic analyses and >95% pure by glc analyses. Unless otherwise stated, all reactions were carried out under an atmosphere of dry argon using glassware that had been thoroughly flame and/or oven (-140 °C) dried. The glass syringes, the needles and the Teflon® cannula for handling anhydrous solvent and reagents were oven dried while the plastic syringes were flushed with a stream of dry argon prior to use. Concentration, evaporation or removal of the solvent under reduced pressure (water aspirator) refer to solvent removal via a Btichi rotary evaporator at —15 Tom Cold temperatures were maintained by use of the following baths: 0 °C, ice/water; -10 °C, ice/acetone; -20 °C, -30 °C, and -48 °C, aqueous calcium chloride/CO2 ( 27, 35 and 47 g CaC12/100 ml H2O, respectively); 94 -60 °C, chloroform/CO2; -78 °C, acetone/CO2; -98 °C, Me011/liquid nitrogen.  V.1.2. Solvents and Reagents  Solvents and reagents were purified and dried using known procedures. 95 Petroleum ether refers to a hydrocarbon mixture with by 35-60 °C. Ether refers to diethyl ether. Dry 143  ether and THF were distilled from sodium benzophenone ketyl. Carbon tetrachloride  96  was  refluxed and then distilled from phosphorus pentoxide. Acetonitrile, benzene, 9 6 dichloromethane, diisopropylamine, DMSO, 1,1,1,3,3,3-hexamethyldisilazane, HMPA 97 (WARNING: carcinogenic), methanesulfonyl chloride, pyridine, 98 triethylamine,  trimethylsilyl chloride (TMSC1) and trimethylsilyl bromide (TMSBr) were refluxed over and then distilled from calcium hydride. Magnesium was added to Me0H and, after refluxing the mixture, the Me0H was distilled from the resulting solution of magnesium methoxide. 2Methy1-2-propanol was dried over activated 4A molecular sieves. Solutions of methyllithium (LiBr complex in ether), butyllithium (in hexanes) and tertbutyllithium (in pentane) were obtained from Aldrich Chemical Co., Inc. and were standard-  ized using either the procedure of Kofron and Baclawski 99 or that of Suffert. 100 Solutions of borane (BH3•THF) were standardized according to the procedure of Brown. 101 Lithium diisopropylamide (i-Pr2NLi) and lithium 1,1,1,3,3,3-hexamethyldisilazide solutions were prepared by the addition of a solution of methyllithium (1.0 equiv) in ether or butyllithium (1.0 equiv) in hexanes to a solution of diisopropylamine (1.05 to 1.1 equiv) or HMDSH (1.05 to 1.1 equiv) in THF at -78 °C. The resulting colourless or faintly yellow solution was stirred at 0 °C for ten minutes before use. Potassium 1,1,1,3,3,3-hexamethyldisilazide solutions were prepared from HMDSH and KH according to the procedure of B rown  .  84  Lithium (diisobutyl)(n-butyl)aluminum hydride was obtained by adding a solution of nBuli (1.0 equiv) in hexanes to a solution of diisobutylaluminum hydride (1.05 equiv) in hexanes at -78 °C. 61 The resultant white solid was diluted with ether and the solution was stirred at -78 °C for 15 min and at 0 °C for 10 min before use. Copper(I) bromide-dimethyl sulfide complex was prepared by the method described by Wuts. 102 Lead tetraacetate was recrystallized from acetic acid. N, N-dimethyl(methylene)144  ammonium iodide was recrystallized 80 from tetramethylene sulfone and the yellowish solid obtained after filtration under inert atmosphere (argon) was rinsed with dry ether and acetone (from a freshly opened bottle, HPLC grade) to provide a white solid that was dried under reduced pressure (vacuum pump). Chloroforml° 3 and deuteriochloroform 103 were dried by filtration through a short column of flame dried basic alumina. Aqueous ammonium chloride solution (pH 8-9) was prepared by addition of 50 mL of aqueous ammonium hydroxide (28%) to — 1L of saturated aqueous ammonium chloride.  145  V.2. EXPERIMENTAL SECTION FOR THE SYNTHESES OF (±)-METHYL CANTABRENONATE (13) and (±)-METHYL EPDXYCANTABRONATE (14).  Preparation of the Keto Acetal 94.  94  To a stirred suspension of magnesium turnings (404 mg, 16 6 mmol) in dry THF (5 mL) containing a small amount of iodine (1 or 2 crystals) were added a few drops of a solution of 2-(2-bromoethyl)-1,3-dioxane (84, 2.65 g, 13.59 mmol) in dry THF (5 mL). The browncoloured mixture was warmed to initiate the formation of the Grignard reagent 67, and then the remainder of the THF solution of the bromide was added (dropwise) at a rate such that reflux of the reaction mixture was maintained. After the addition was complete, the mixture was refluxed for an additional 30 min, was cooled to room temperature, was diluted with dry THF (10 mL), and then was cooled to -78 °C. Solid CuBr•Me2S (171 mg, 0.83 mmol) was added and the mixture was stirred at -78 °C for 1 h. Dry HMPA 97 (3.6 mL, 20 8 mmol) was added and, after 10 min, a solution of 3-methyl-2-cyclopenten- 1-one (64, 1.0 g, 10 4 mmol) and Me3SiC1 (2.6 mL, 20.8 mmol) in dry THF (10 mL) was added dropwise. The reaction mixture was stirred at -78 °C for 6 h, at -45 °C for 2 h, and then was warmed to room temperature over a period of 15 min. After addition of HOAc (2.4 mL, 41 4 mmol), the mixture was stirred at room temperature for 20 min. Et20 (10 mL) and aq NH4C1-NH4OH (pH 8-9) (an amount sufficient to make the mixture basic) were added, the mixture was opened to the atmosphere and was stirred vigorously until the aqueous phase became deep blue. The phases were separated and the aqueous layer was extracted with Et20 (3 x 100 mL). The 146  combined organic extracts were washed with brine (70 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (350 g of silica gel, 7 : 3 petroleum etherEtOAc) of the remaining material gave, upon concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), 1.54 g (70%) of the keto acetal 94, as a colourless oil. Ir (neat): 1741, 1462, 1406, 1380, 1241, 1148, 1081, 1047, 1001, 936, 881 cm -1 . 1H  nmr (400 MHz) 5: 1.04 (s, 3H, tertiary Me), 1.36 (dm, 1H, J for d = 13.5 Hz), 1.49-  1.88 (m, 6H), 2.00-2.15 (m, 3H), 2.26-2.33 (m, 2H), 3.77 (br ddd, 2H, J = 2, 11.5, 11.5 Hz, axial OCI-1_2), 4.12 (br ddd, 2H, J = 1.5, 5.0, 11.5 Hz, equatorial OCLI2), 4.52 (br t, 1H, J = 5 Hz, OCH). Anal. calcd for C12H2003: C 67.89, H 9.50; found: 67.71, H 9.60. Exact Mass calcd: 212.1412; found: 212.1417.  The initial fractions of the above flash chromatography contained 0.57 g of an oil that, on the basis of analyses by glc and 1 H nmr spectroscopy, consisted of a mixture of the keto acetal 94 and the diacetal 95, in a ratio of about 85 : 15, respectively. Since these two substances were difficult to separate, this material was discarded.  147  Preparation of the Bicyclic Keto Alcohol 63 and of the Bridged Bicyclic Keto Alcohol 96. 8 2.28 54.52 41'.^1 1- ."--.1'  HOI  Me  8 1.22^,..,..  96  63  A solution of the keto acetal 94 (4.26 g, 20 1 mmol) in a mixture of THE and 10% hydrochloric acid (2 : 1, 200 mL) was refluxed for 18 h. The dark mixture was cooled, carefully neutralized with sat. aq NaHCO3, and diluted with EtOAc (150 mL). The phases were separated, and the aqueous layer was extracted with EtOAc (3 x 150 mL). The combined organic extracts were washed with brine (100 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (300 g of silica gel, 3 : 1 : 1 petroleum ether-EtOAc-CH2C12) of the crude product gave, upon concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), two products. The initially eluted substance was the desired keto alcohol 63 (2.26 g, 73%), a thick oil. Ir (neat): 3461 (broad), 1735, 1456, 1409, 1149 cm -1 . 1H  nmr (400 MHz) 8: 1.22 (s, 3H, angular Me), 1.57-1.65 (m, 1H), 1.68-2.09 (m, 5H), 2.28  (d, 1H, J = 8 Hz, angular H), 2.40 (t, 2H, J = 8 Hz, Ca2C=--0), 2.66 (br s, 1H, exchanges with D20, OH), 4.52 (ddd, 1H, J = 5.5, 5.5, 8 Hz, CHOH). In decoupling experiments, irradiation of the ddd (CHOH) at 8 4.52 collapsed the d (angular H) at 2.28 into a s and simplified the multiplets (hydrogens a to CHOH) at —1.68-1.78 and 148  --2.00-2.09 (part of the m (5H) at 1.68-2.09). Saturation at 8 2.28 (angular H) collapsed the ddd (J = 5.5, 5.5, 8 Hz) at 4.52 into a dd (J = 5.5, 5.5 Hz). In nuclear Overhauser enhancement (NOE) difference experiments, irradiation at 8 1.22 (Me) caused enhancement of the signal at 2.28 (angular H); irradiation at 5 2.28 increased the intensity of the signals at 1.22 (Me) and 4.52 (CHOH); saturation of the resonance at 8 4.52 caused enhancement of the signals at 2.66 (OH) and 2.28 (angular H). Exact Mass calcd for C9H14O2: 154.0993; found: 154.0992. The bridged bicyclic keto alcohol 96 (0.73 g, 23%) was the material eluted second and was isolated as an oil. It exhibited it (neat): 3413, 1740, 1460, 1403, 1349, 1238, 1180, 1068 cm -1 . 1H  nmr (400 MHz) 8: 1.17 (s, 3H, angular Me), 1.24-1.37 (m, 1H, H-c or H-d), 1.54 (dd,  1H, J = 3.5, 12 Hz, H-g), 1.56-1.62 (m, 2H), 1.86 (ddd, 1H, J = 2.5, 6, 12 Hz, H-h), 1.99-2.17 (m, 4H, including OH signal, a br s that exchanges with D20), 2.55 (unresolved m, 1H, H-a), 3.84 (ddd, 1H, J = 3.5, 6, 11.5 Hz, H-b). Detailed 1 H nmr data, derived from decoupling experiments, are given in Table 1.  149  Table 1: 1 H nmr Data (400 MHz, CDC13) for the Bridged Bicyclic Ketol 96: Decoupling Experiments.  H  Signal Being Irradiated Assignment H-xa H-a  1H  Signals Being Observed  nmr (400 MHz) 8 ppm (mult., J(Hz))  2.55 (unresolved m)  8 ppm (initial mult., J(Hz), H-x) to mult. after irradiation, J(Hz)b 1.86 (ddd, J = 2.5, 6, 12, H-h) to dd, J = 2.5, 12. 3.84 (ddd, J = 3.5, 6, 11.5, H-b) to dd, J = 6, 11.5.  H-b  3.84 (ddd, J = 3.5, 6, 11.5)  1.24-1.37 (m, H-c or H-d) to sharpened m. 1.99-2.17 (m, 4H, includes H-c or H-d), part of the m is modified. 2.55 (unresolved m, H-a) to br d, J = 6.  H-c or 1.24-1.37 (m) H-d  1.56-1.62 (m, 2H), the m is modified. 1.99-2.17 (m, 4H, includes H-c or H-d), part of the m is modified. 3.84 (ddd, J = 3.5, 6, 11.5, H-b) to unresolved m.  H-h  1.86 (ddd, J = 2.5, 6, 12)  1.54 (dd, J = 3.5, 12, H-g) to d, J = 3.5. 2.55 (unresolved m, H-a) to br s.  H-g  _ 1.54 (dd, J = 3.5, 12)  a- Irradiated hydrogen. b- Only the hydrogens for which changes in their signals could be unambiguously seen are recorded.  150  Preparation of the Enone 62 via the Keto Mesylate 97.  97  62  To a cold (0 °C), stirred solution of the keto alcohol 63 (4.94 g, 32 0 mmol) in dry CH2C12 (300 mL) were added successively freshly distilled, dry Et3N (13.4 mL, 96.2 mmol) and MsC1 (5 mL, 64.2 mmol). After the mixture had been stirred at 0 °C for 1 h, sat. aq NH4C1 (100 mL) was added and the layers were separated. The organic phase was washed with 10% aqueous HC1 (100 mL), sat. aq NaHCO3 (100 mL), brine (100 mL), and then was dried over anhydrous magnesium sulfate and concentrated. The crude keto mesylate 97, a yellowish solid, was used directly in the next step. Flash chromatography (silica gel, 3 : 1 : 1 petroleum ether-EtOAc-CH2C12) of a small amount of the crude keto mesylate 97 derived from another experiment, followed by three recrystallizations of the derived solid from Et20, provided material that exhibited mp 77-78.5 °C. Ir (KBr): 1745, 1351, 1183, 1171, 976, 943, 893 cm - I. 1H  nmr (400 MHz) 8: 1.23 (s, 3H, angular Me), 1.72-1.88 (m, 2H), 1.97-2.28 (m, 4H),  2.34-2.53 (m, 3H), 2.99 (s, 3H, Me -S 03), 5.25 (ddd, 1H, J = 3, 5, 7.5 Hz, CH - OS02). Exact Mass calcd for C9H1302 (M+-MeS02): 153.0915; found: 153.0911.  The crude keto mesylate 97 obtained from the reaction described above was dissolved in dry CH2C12 (250 mL) and 9.6 mL of 1,8-diazabicyclo[5.4.0]undec-7-ene (64.1 mmol) were added. The mixture was stirred at room temperature for 1.5 h, was washed with sat. aq 151  N114C1 (75 mL) and brine (75 mL), and then was dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (150 g of silica gel, 4 : 1 petroleum ether-ether) of the remaining material, followed by distillation (air-bath temperature 90-95 °C/--15 Torr) of the acquired liquid, afforded 3.24 g (74% from the keto alcohol 63) of the enone 62, a colourless oil. 1r (neat): 1714, 1635, 1448, 1221, 1094 cm -1 . 1H  nmr (400 MHz) 8: 1.21 (s, 3H, angular Me), 1.74 (ddd, 1H, J = 8.5, 12, 12 Hz), 1.89-  2.08 (m, 3H), 2.52 (ddd, 1H, J = 1, 8.5, 18 Hz), 2.60-2.72 (m, 2H), 2.92 (dddd, 1H, J = 2, 6.5, 11, 18 Hz), 6.44 (dd, 1H, J = 2, 3.5 Hz, vinyl hydrogen). Anal. calcd for C911120: C 79.37, H 8.88; found: C 78.98, H 8.89. Exact Mass calcd: 136.0888; found: 136.0896.  152  Preparation of the Tricyclic Keto Alkene 60.  60^  100  To a cold (-78 °C), stirred solution of freshly distilled 4-chloro-2-trimethylstanny1-1butene (6, 921 mg, 3.64 mmol) in dry THF (20 mL) was added a solution of MeLi in ether (3.4 mL, 4.13 mmol). After the yellow solution had been stirred at -78 °C for 30 min, solid CuCN 104 (264 mg, 2.95 mmol) was added in one portion. The slurry was warmed briefly (58 min) to -50 °C and the resultant pale yellow solution of the cuprate reagent 18 was recooled to -78 °C. A solution of the freshly distilled enone 62 (402 mg, 2.95 mmol) in dry THF (10 mL) was added slowly. The bright yellow mixture was stirred at -78 °C for 1.5 h and at -48 °C for 2.5 h. Dry HMPA 97 (1 mL, 5.9 mmol) was added, the solution was allowed to warm to room temperature over a period of 1.5 h, and then was stirred for an additional 45 min. The solution was poured into a mixture of Et20 (50 mL) and aq NH4C1-NH4OH (pH 89) (40 mL) and the resultant mixture was opened to the atmosphere and was stirred vigorously until the aqueous layer was blue. The phases were separated and the aqueous layer was extracted with Et20 (2 x 50 mL). The combined organic extracts were washed with brine (60 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (50 g of silica gel, 9 : 1 petroleum ether-ether) of the residual material, followed by distillation of the liquids thus obtained, gave 407 mg (73%) of the less polar tricyclic keto alkene 60 and 79 mg (12%) of the more polar keto chloride 100, as colourless oils. Ketone 60 (air-bath temperature 75-84 °C/0.4 Ton).  153  Ir (neat): 3071, 1732, 1656, 1457, 1162, 880 cm -1 . 1 H nmr (400 MHz) 6: 1.08 (s, 3H, Me-12), 1.49-1.86 (m, 7H), 2.01-2.15 (m, 1H), 2.23-  2.46 (m, 3H), 2.52-2.62 (m, 1H), 2.93 (br d, 1H, J = 10 Hz, H-1), 4.75 (br s, 1H, H-15), 4.87 (t, 1H, J = 1 Hz, H-15'). 1 H nmr (400 MHz, C6D6)  6: 0.74 (s, 3H, Me-12), 1.19 (ddd, 1H, J = 4, 9.5, 13.5 Hz, H-  5), 1.25-1.40 (m, 4H), 1.43-1.53 (m, 1H, H-2), 1.77 (dddd, 1H, J = 9.5, 9.5, 9.5, 13.5 Hz, H-2'), 1.86 (ddd, 1H, J = 7 .5, 7.5, 13 Hz, H-11'), 1.97 (ddd, 1H, J = 9.5, 9.5, 18.5 Hz, H-6), 2.09 (ddd, 1H, J = 4, 9.5, 18.5 Hz, H-6'), 2.22 (dddm, 1H, J for ddd = 7.5, 7.5, 15 Hz, H-10), 2.64 (dddm, 1H, J for ddd = 7.5, 7.5, 15 Hz, H-10'), 2.83 (dm, 1H, J for d = 9.5 Hz, H-1), 4.75 (m, 1H, H-15), 4.89 (m, 1H, H-15'). Detailed 1 H nmr data, derived from decoupling experiments, are given in Table 2.  Anal. calcd for C1314180: C 82.06, H 9.53; found: C 82.15, H 9.60. Exact Mass calcd: 190.1358; found: 190.1355.  154  Table 2: 1 H nmr Data (400 MHz, C6D6) for the Keto Alkene 60a: Decoupling Experiments.  12  60  Assignment H-x H-1  Signal Being Irradiated 1 H nmr (400 MHz) 8 ppm (mult., J (Hz))  Signals Being Observed 5 ppm (initial mult., J (Hz), H-x) to mult. after irradiation, J (Hz)b  2.83 (dm, J for d = 9.5)  1.43-1.53 (m, H-2) to sharper m 1.77 (dddd, J = 9.5, 9.5, 9.5, 13.5, H-2') to ddd, J = 9.5, 9.5, 13.5. 4.75 (m, H-15) to sharper signal. 4.89 (m, H-15') to sharper signal.  1.97 (ddd, J = 9.5, 9.5, 18.5, H-6) to dd, J = 1.19 (ddd, J = 4, 9.5, 13.5) H-5 ^9.5,^18.5. 2.09 (ddd, J = 4, 9.5, 18.5, H-6') to dd, J = ^9.5,^18.5. H_ io'c  2.64 (dddm, J for ddd = 7.5, 1.25-1.40 (m, 4H, includes H-11); part of the m is modified. 7.5, 15) 1.86 (ddd, 1H, J = 7.5, 7.5, 13, H-11') to dd,  J = 7.5, 13.  2.22 (dddm, 1H, J for ddd = 7.5, 7.5, 15, H10) to unresolved m. 4.75 (m, H-15) to sharper signal. 4.89 (m, H-15') to sharper signal. H-15  4.75 (m)  2.22 (dddm, 1H, J for ddd ---- 7.5, 7.5, 15, H10) to sharper signal. 2.64 (dddm, 1H, J for ddd = 7.5, 7.5, 15, H10') to sharper signal.  _ 2.83 (dm, J for d = 9.5, H-1) to sharper signal. a- Silphiperfolane numbering used for consistency b- Only the hydrogens for which changes in their signals could be unambiguously seen are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-10' is more downfield than H-10). 155  Keto chloride 100 (air-bath temperature 124-127 °C/0.5 Torr). Ir (neat): 3088, 1735, 1645, 1456, 896 cm -1 . 1 H nmr (400 MHz, CDC13)  8: 1.24 (s, 3H, angular Me), 1.58-1.96 (m, 6H), 2.08 (br d, 1H,  J = 6 Hz), 2.31-2.67 (m, 5H), 3.63-3.75 (m, 2H, Ca2C1), 4.88 (br s, 1H, vinylic  hydrogen), 5.00 (br s, 1H, vinylic hydrogen). 1 H nmr (400 MHz, C6D6)  8: 0.86 (s, 3H, angular Me), 1.10-1.41 (m, 4H), 1.43-1.56 (m,  2H), 1.77 (d, 1H, J = 5 Hz), 1.92-2.09 (m, 2H), 2.32-2.51 (m, 3H), 3.40-3.51 (m, 2H, Ca 2C1), 4.71 (br s, 1H, vinylic H), 4.87 (br s, 1H, vinylic H). Exact Mass calcd for C13H19C10: 226.1125; 226.1133.  156  Preparation of the Aldehydes 110 and 111 via the Diols 109.  0  OH  110  111  109  To a stirred solution of the keto alkene 60 (186.5 mg, 0.98 mmol) in dry THF (9.8 mL) was added a solution of BH3 in THF (0.78 M, 2.5 mL, 1.96 mmol) and the resultant mixture was stirred at room temperature for 2 h. The solution was cooled to 0 °C and Me0H (1.6 mL), 3 N aqueous NaOH (1.2 mL), and 30% aq 1-1202 (1.2 mL) were added sequentially. The mixture was refluxed for 1.5 h, was cooled to room temperature, and then was diluted with EtOAc (10 mL) and saturated with NaCl. The phases were separated and the aqueous layer was extracted with EtOAc (2 x 10 mL). The combined organic extracts were washed with brine (8 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (10 g of silica gel, 7 : 3 EtOAc-petroleum ether) provided 198 mg (96%) of a mixture of diols 109 as a white solid. This material was used directly for the next reaction. To a cold (-78 °C), stirred solution of DMSO (250 pL, 3.52 mmol) in dry CH2C12 (5 mL) was added oxalyl chloride (277 ptL, 3.17 mmol) and the resulting mixture was stirred for 15 min. A solution of the mixture of diols 109 (185 mg, 0.88 mmol) in dry CH2C12 (4 mL) was added slowly and the solution was stirred at -78 °C for 1 h. Dry Et3N (980 pL, 7.04 mmol) was added, the mixture was warmed to 0 °C, and then was stirred at this temperature for 1 h. Water (5 mL) was added and the layers were separated. The organic phase was washed with sat. aq NH4C1 (5 mL) and brine (5 mL), and then was dried over anhydrous magnesium sulfate and concentrated. The residual material, a mixture of the keto aldehydes 110 and 111, was dissolved in dry Me0H (4 mL) and a solution of Me0Na in Me0H (-0.15 M, 3 mL, 157  0,45 mmol) was added. The reaction mixture was stirred at room temperature for 2 h and then was diluted with EtOAc (20 mL). The mixture was washed sequentially with 10% aq HC1 (4 mL), water (4 mL), and brine (4 mL), and then was dried over anhydrous magnesium sulfate and concentrated. Analysis of the residual material by 1 H nmr spectroscopy showed that it consisted of a mixture of 110 and 111, in a ratio of -8 : 1. Flash chromatography (40 g of silica gel, 7 : 2 : 1 petroleum ether-CH2C12-Et20) of this mixture gave, in the earlier fractions, pure keto aldehyde 110, while the later fractions contained a mixture of 110 and 111. The latter material was recycled through the epimerization-flash chromatography sequence and the fractions containing a mixture of 110 and 111 were recycled again. These procedures provided a total of 145 mg (80%) of the desired keto aldehyde 110, as a colourless oil. Ir (neat): 2713, 1729, 1457, 1161, 1085 cm -1 . 1H  nmr (400 MHz) 8: 1.09 (s, 3H, angular Me), 1.45-1.85 (m, 7H), 1.90-2.08 (m, 2H),  2.11-2.22 (m, 1H), 2.26-2.48 (m, 3H), 2.73 (ddd, 1H, J = 3, 6, 9.5 Hz), 9.58 (d, 1H, J = 3 Hz, CHO). Anal. calcd for C13H1802: C 75.69, H 8.79; found: C 75.49, H 8.81. Exact Mass calcd: 206.1307; found: 206.1308.  The keto aldehyde 111, contaminated with a very small amount of the epimer 110, was obtained from the last fractions of the flash chromatographies described above. Ir (neat): 2726, 1729, 1458, 1162 cm -1 . 1H  nmr (400 MHz) 8: 1.09 (s, 3H, angular Me), 1.35-1.46 (m, 1H), 1.57-2.01 (m, 9H), 2.29  (ddd, 1H, J = 7, 9, 18 Hz), 2.50 (ddd, 1H, J = 7, 9, 18 Hz), 2.78-2.95 (m, 2H), 9.78 (br s, 1H, CHO).  158  Exact Mass calcd for C13H1802: 206.1307; found: 206.1307.  Preparation of the Keto Dithioacetal 112.  112  To a cold (0 °C), stirred solution of the keto aldehyde 110 (615 mg, 2.98 mmol) in dry CH2C12 (30 mL) were added sequentially 1,2-ethanedithiol (275 pL, 3.28 mmol) and BF3•Et20 (183 pL, 1.49 mmol). After the solution had been stirred at 0 °C for 1 h, it was washed with 10% aq NaOH (10 mL) and brine (10 mL) and then was dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (65 g of silica gel, 92 : 8 petroleum ether-EtOAc) of the crude product, followed by recrystallization of the acquired solid from Et20, afforded 712 mg (85%) of the dithioacetal 112, a crystalline substance that exhibited mp 93.5-95 °C. Ir (KBr): 1726, 1449, 1279, 1163 cm -1 . 1H  nmr (400 MHz) 5: 1.03 (s, 3H, angular Me), 1.39-1.47 (m, 1H), 1.55-2.10 (m, 10H),  2.21-2.33 (m, 2H), 2.41 (ddd, 1H, J = 4.5, 9.5, 19 Hz), 3.14-3.28 (m, 4H, CLI2SCHSCH2), 4.54 (d, 1H, J = 8 Hz, CH2SCHSCH2). Anal. calcd for C15H220S2: C 63.78, H 7.85, S 22.70; found: C 63.58, H 7.75, S 22.60. Exact Mass calcd: 282.1112; found: 282.1111.  159  Preparation of the Ketone 59.  59  The Raney nickel (W-4) 105 employed in this experiment was freshly prepared and stored under EtOH. It was aged for at least 2 days prior to use, but was used within 30 days of its preparation. A suspension of the Raney nickel (-200 mg) in EtOH (-1.2 mL) was allowed to settle and the EtOH was decanted. The residual material was washed with acetone (3 x 2 mL) and then was covered with acetone (2 mL). This black slurry was added to a solution of the dithioacetal 112 (42 mg, 0.15 mmol) in acetone (2 mL) and the resulting heterogeneous mixture was stirred vigorously 106 at room temperature for 1.5 h. The mixture was filtered through Celite (2 g, elution with Et20) and the filtrate was concentrated. Distillation (42-45 °C/0.15 Torr) of the residual liquid gave 21 mg (74%) of the ketone 59, as a colourless oil. (Depending on the quality of the Raney nickel used, it was sometimes necessary to purify, by flash chromatography, the crude material obtained, before distillation. In these cases, the solvent system was 95 : 5 petroleum ether-ether. The yields for this reaction varied from 62-74%.) Ir (neat): 1731, 1456, 1413, 1377, 1273, 1184, 1097 cm -1 . 1H  nmr (400 MHz, C6D6) 8: 0.78 (s, 3H, angular Me), 0.97 (d, 3H, J = 6.5 Hz, MeCH),  1.09-1.26 (m, 3H), 1.29-1.42 (m, 4H), 1.55-1.71 (m, 3H), 1.78-1.87 (m, 2H), 1.99 (ddd, 1H, J = 9.5, 9.5, 18.5 Hz), 2.10 (ddd, 1H, J = 4, 9.5, 18.5 Hz). Exact Mass calcd for C13H200: 192.1514; found: 192.1505.  160  Preparation of the Enone 123 via the Enol Si13 1 Ether 122. ,  OSiMe 3  122  123  To a cold (-78 °C), stirred solution of freshly prepared LDA (0.3 M, 1.4 mL, 0.43 mmol) in THF was added a solution of the ketone 59 (63 mg, 0.33 mmol) in dry THF (3.3 mL). After the solution had been stirred at -78 °C for 1 h, freshly distilled Me3SiC1 (57 1.1.L, 0.45 mmol) was added and the resulting mixture was warmed to room temperature and stirred for an additional 1 h. The mixture was concentrated under reduced pressure and the remaining material was triturated with pentane. The resultant mixture was filtered and the filtrate was concentrated. A repeat of the trituration-concentration sequence provided the crude enol silyl ether 122. To a suspension of Pd(OAc)2 (79.5 mg, 0.35 mmol) in dry MeCN (800 gL) was added a solution of the crude enol silyl ether 122 in dry MeCN (1 3 mL). After the mixture had been stirred at room temperature for 3.8 h, it was filtered through Florisil (4 g, elution with Et20). Concentration of the filtrate, followed by flash chromatography (5 g of silica gel, 95 : 5 to 9 : 1 pentane-Et20) of the residual material, gave 55 mg (88%) of the enone 123 as a white solid. Recrystallization (pentane, -22 °C) provided material that displayed mp 38.539 °C. 1r (KBr): 1696, 1585, 1454, 1448, 1376, 1344, 1305, 1272, 1195, 1139, 839, 695 cm -1 . 1H  nmr (400 MHz) 5: 0.99 (d, 3H, J = 6 Hz, MeCH), 1.12 (s, 3H, angular Me), 1.34-1.76  (m, 7H), 1.77-1.89 (m, 2H), 1.91-1.99 (m, 1H), 6.08 (d, 1H, J = 5.5 Hz, vinylic hydrogen), 7.30 (d, 1H, J = 5.5 Hz, vinylic hydrogen). 161  Anal. calcd for C13H180: C 82.06, H 9.53; found: C 82.10, H 9.54. Exact Mass calcd: 190.1358; found: 190.1354.  Preparation of the Tertiary Alcohol 121.  OH  121  To a cold (-78 °C), stirred solution of the enone 123 (55.3 mg, 0.29 mmol) in dry THE (2.9 mL) was added a solution of MeLi in Et20 (263 .tL, 0.38 mmol). After the reaction mixture had been stirred at -78 °C for 1.5 h, it was poured into a mixture of sat. aq NH4C1 (5 mL) and Et20 (10 mL) The phases were separated and the aqueous layer was extracted with Et20 (2 x 10 mL). The combined organic extracts were washed with brine (10 mL), dried over anhydrous magnesium sulfate and the solvent was removed. Flash chromatography (5 g of silica gel, 4 : 1 pentane-Et20) of the crude material produced 57.5 mg (96%) of the tertiary allylic alcohol 121 as a white solid. Recrystallization (pentane, -20 °C) gave white crystals, mp 44.5-45.5 °C. Ir (CHC13): 3608, 3536-3374, 3040, 3010, 1456, 1373, 1353, 1120 cm -1 . 1 H nmr (400 MHz) 8: 1.00 (s, 3H,  Me), 1.00 (d, 3H, J = 6.5 Hz, MeCH), 1.19-1.29 (m,  1H), 1.29 (s, 3H, Me), 1.38-1.59 (m, 7H, includes the signal (1.52, s) due to the OH, which exchanges with D20), 1.60-1.71 (m, 1H), 1.86 (ddd, 1H, J = 5.5, 5.5, 13.5 Hz), 2.25 (br t, 1H, J = 7 Hz), 5.39 (d, 1H, J = 6 Hz, vinylic hydrogen), 5.54 (d, 1H, J = 6 Hz, vinylic hydrogen). Exact Mass calcd for C14H220: 206.1671; found: 206.1675.  163  Preparation of the Enone 120.  120  0  To a stirred solution of the tertiary alcohol 121 (63 mg, 0.31 mmol) in dry CH2C12 (3.1 mL) was added, in one portion, solid PCC 107 (99 mg, 0.46 mmol). The mixture was stirred at room temperature for 1 h, was diluted with Et20, and then was filtered through Florisil (4 g, elution with Et20). The filtrate was concentrated. Flash chromatography (5 g of silica gel, 85 : 15 pentane-Et20) of the crude product afforded 55 mg (88%) of the enone 120, as a colourless oil.  1r (neat): 3065, 1703, 1616, 1458, 1377, 1293, 1140, 857 cm -1 . 1H  nmr (400 MHz) 8: 1.01 (d, 3H, J = 6.5 Hz, MeCH), 1.02 (s, 3H, angular Me), 1.23-  1.36 (m, 2H), 1.39-1.50 (m, 2H), 1.54-1.85 (m, 4H), 1.91 (dddd, 1H, J = 3, 6, 6, 12 Hz), 2.02 (dd, 1H, J = 6, 12 Hz), 2.08 (d, 3H, J = 1 Hz, vinylic Me), 5.79 (br s, 1H, vinyl hydrogen). Exact Mass calcd for C14H200: 204.1514; found: 204.1512.  164  Preparation of the Tricyclic Ketone 126.  Me 0 126 To a solution of the enone 120 (40 mg, 0.20 mmol) in EtOAc (4 mL) were added 12 mg of 10% Pd-C. The resultant mixture was stirred under an atmosphere of hydrogen (22 psi) for 4 h and then was filtered through Celite (2 g, elution with Et20). Concentration of the filtrate gave 39.8 mg (98%) of the tricyclic ketone 126, as a colourless oil. Ir (neat): 1738, 1461, 1378, 1246, 1179, 1132, 1056, 976, 700 cm -1 . 1 H nmr (400 MHz)  8: 0.95 (s, 3H, Me-12), 1.03 (d, 3H, J = 6.5 Hz, Me-15), 1.11 (d, 3H,  J = 7 Hz, Me-14), 1.15-1.29 (m, 2H), 1.33-1.59 (m, 4H), 1.61-1.68 (m, 1H), 1.68-1.77 (m, 1H), 1.77 (dd, 1H, J = 13, 18 Hz, H-6), 1.89 (br t, 1H, J = 8 Hz, H-1), 1.94 (dd, 1H, J = 6, 12 Hz, H-3'), 2.15-2.21 (m, 1H, H-7), 2.29 (dd, 1H, J = 7, 18 Hz, H-6'). In decoupling experiments, irrradiation of the d (Me-15) at 8 1.03 modified part of the m at 1.33-1.59. Saturation of the d at 8 1.11 (Me-14) simplified the m at 2.15-2.21 into a dd, J = 7, 13 Hz. Detailed 1 H nmr data, including those derived from COSY and NOE experiments, are given in  Table 3. Exact Mass calcd for C14H220: 206.1670; found: 206.1677.  165  Table 3: 1 H nmr Data (400 MHz, CDC13) for the Ketone 126a.  Me 0 12  126  H-x  COSY Correlationsb  1H nmr (400 MHz) 5 ppm (mult., J (Hz))  H-1  1.89 (br t, J = 8)  H-2, H-2', H-9  H-2  Part of the m (2H) at 1.15-1.29  H-1, H-2', H-3, H 3'  H-2'c  —1.33-1.40 (part of the m (4H) at  H-1, H-2, H-3, H-3'  NOE Correlationsb Me-14, Me-15  1.33-1.59) H-3'  1.94 (dd, J= 6, 12)  H-2, H-2', H-3, Me-12  H-6  1.77 (dd, J = 13, 18)  H-6', H-7  H-6'  2.29 (dd, J = 7, 18)  H-6, H-7  H-7  2.15-2.21 (m)  H-6, H-6', Me-14  Me-12  0.95 (s)  H-3'  Me-14  1.11 (d, J = 7)  H-7  Me-15  1.03 (d, J = 6.5)  H-9^_ H-1  Me-14  H-1, H-7  a- Silphiperfolane numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  166  Preparation of the Keto Ester 129.  COOMe  o 129  To a cold (-78 °C), stirred solution of LDA (0.32 M, 1.1 mL, 0.35 mmol) in THF was added a solution of the ketone 126 (28.5 mg, 0.14 mmol) in dry THF (1.4 mL) and the resultant mixture was stirred at -78 °C for 50 min. Dry HMPA 97 (60 lit, 0.35 mmol) was added and, after 5 min, the solution was treated with 49 p.1_, (0.63 mmol) of methyl cyanoformate. 108 After the reaction mixture had been stirred at -78 °C for 90 min, it was poured into a mixture of sat. aq NaHCO3 (5 mL) and Et20 (10 mL). The phases were separated and the aqueous layer was extracted with Et20 (2 x 10 mL) The combined organic extracts were washed with brine (5 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography on silica gel (3 g, 9 : 1 pentane-Et20) of the crude material produced 30 mg (83%) of the keto ester 129, as a colourless oil. Ir (neat): 1753, 1728, 1460, 1436, 1369, 1336, 1284, 1204, 1164, 1005 cm-1 . 1H  nmr (400 MHz) 8: 1.02 (s, 3H, Me-12), 1.03 (d, 3H, J = 6.5 Hz, MeCH), 1.12 (d, 3H,  J = 6.5 Hz, MeCH), 1.15-1.31 (m, 2H), 1.35-1.54 (m, 3H), 1.57-1.64 (m, 1H), 1.65-1.80 (m, 2H), 1.86 (br t, 1H, J = 8 Hz), 1.98 (dd, 1H, J = 6, 13 Hz), 2.60 (dq, 1H, J = 13, 6.5 Hz, H-7), 2.77 (d, 1H, J = 13 Hz, H-6), 3.76 (s, 3H, COOMe). Exact Mass calcd for C16H2403: 264.1725; found: 264.1727.  167  Preparation of the Phenylseleno Ketones 130 and 131.  SePh  COOMe^  130  -,^ Se Ph^ 0 ^  . COOMe 131  0  To a cold (-78 °C), stirred solution of LDA (0.31 M, 640 pit, 0.2 mmol) in dry THF was added a solution of the keto ester 129 (21.6 mg, 0.082 mmol) in dry THF (820 gL). After the mixture had been stirred at -78 °C for 40 min, a freshly prepared solution of PhSeBri°9 (0.25 mmol) in dry THF (400 ;IL) was added and stirring was continued for 25 min. The solution was poured into a mixture of sat. aq NaHCO3 (3 mL) and Et20 (5 mL) and the phases were separated. The aqueous phase was extracted with Et20 (2 x 5 mL) and the combined organic extracts were washed with brine (5 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography on silica gel (5 g, 9 : 1 to 7 : 3 pentaneEt20) of the residual material gave 16.6 mg (48%) of the less polar phenylseleno keto ester 131 and 16.4 mg (47%) of the more polar epimer 130, as yellowish and slightly impure oils. The unwanted a-phenylseleno ketone 130 could be recycled as follows. A cold (-78 °C), stirred solution of 130 (16.4 mg, 0.04 mmol) in dry THF (900 p.L) was treated with a freshly prepared (PhSeHl° 9 , NaH) solution of PhSeNa (0.06 mmol) in dry THF (170 1.1L) containing 0.06 mmol of dry HMPA, 97 and the mixture was stirred at -78 °C for 15 min. The mixture was diluted with Et20 (5 mL), washed with brine (3 mL), dried over anhydrous magnesium sulfate and concentrated. Column chromatography (4 g of silica gel, 9 : 1 pentane-Et20) of the residual material gave 9 mg of the I3-keto ester 129, which was subjected to the LDA-PhSeBr-flash chromatography sequence as described above. Two such recycling procedures provided an additional 9.4 mg of the desired a-phenylseleno ketone 131. Thus, the total yield of 131 from the keto ester 129 was 76%. 168  The phenylseleno keto ester 131 exhibited in its 1 H nmr spectrum (400 MHz) the following methyl signals: 8 0.68 (s, 3H, angular Me), 0.97 (d, 3H, J = 6 Hz, MaCH), 1.22 (d, 3H, J = 7 Hz, MeCH), 3.61 (s, 3H, COOMe). The methyl signals in the 1 H nmr spectrum (400 MHz) of the a-phenylseleno ketone 130 appeared at 8 1.00 (s, 3H, angular Me), 1.06 (d, 3H, J = 7 Hz, aqCH), 1.34 (d, 3H, J = 6 Hz, MeCH) and 3.59 (s, 3H, COOMe).  169  Preparation of (±)-Methyl Cantabrenonate (13).  CO2Me  13  A cold (-78 °C) solution of the a-phenylseleno ketone 131 (29.6 mg, 0.07 mmol) in dry CH2C12 (1 4 mL) was treated with a stream of ozone. The excess ozone was removed with a stream of argon, the mixture was allowed to warm to room temperature, and then was concentrated. Column chromatography (1 g of silica gel, 3 : 2 hexane-Et20) of the residual material gave 14.2 mg (77%) of (±)-methyl cantabrenonate (13) as a white solid. Recrystallization (hexane) provided material with mp 81-83 °C. 1r (CHC13): 3022, 2954, 2869, 1737, 1708, 1615, 1456, 1436, 1241, 1063, 994, 974 cm -1 . 1H  nmr (400 MHz) 8: 1.03 (d, 3H, J = 6 Hz, Me-15), 1.06 (s, 3H, Me-12), 1.20-1.55 (m,  4H), 1.60-1.70 (m, 1H), 1.74-1.91 (m, 3H), 1.97 (dddd, 1H, J = 2.5, 6, 6, 12 Hz), 2.082.15 (m, 1H), 2.36 (s, 3H, Me-14), 3.83 (s, 3H, OMe). 13 C  nmr (75.3 MHz) 5: 15.4 (C-14), 18.2 (C-15), 20.8 (C-12), 26.0 (C-2), 29.1 (C-11),  35.3 (C-10), 37.0 (C-3), 39.8 (C-9), 51.7 (OMe), 59.2 (C-1), 59.4 (C-4), 68.1 (C-8), 128.8 (C-6), 164.1 (C-13), 189.9 (C-7), 208.6 (C-5). Anal. calcd for C16H2203: C 73.25, H 8.45; found: C 73.28, H 8.54. Exact Mass calcd: 262.1569; found: 262.1560.  The synthetic product exhibited 1 H nmr and 13 C nmr spectra identical with those derived from esterified natural cantabrenonic acid. 47  170  Preparation of (±)-Methyl Epoxycantabronate (14).  14  0  To a solution of (±)-methyl cantabrenonate (13, 10 mg, 0.038 mmol) in Me0H (1.9 mL) were added sequentially 30% aq H202 (22 [IL, 0.215 mmol) and aq NaOH (1 N, 38 ;IL, 0.038 mmol). The mixture was stirred at room temperature for 14 min, was diluted with Et20 (10 mL), was washed with sat. aq NH4C1 (3 mL) and brine (3 mL), and then was dried over anhydrous magnesium sulfate and concentrated. Column chromatography (1 g of silica gel, 85 : 15 pentane-Et20) of the remaining material afforded 9 mg (85%) of (±)-methyl epoxycantabronate (14) a white solid that could be recrystallized from hexane at -20 °C to yield white crystals that exhibited mp 98-99.5 °C. Ir (KBr): 2951, 2862, 1752, 1439, 1401, 1379, 1295, 1234, 1069, 973, 777, 751 cm -1 . 1H  nmr (400 MHz) 6: 1.04 (d, 3H, J = 6.5 Hz, Me-15), 1.09 (s, 3H, Me-12), 1.21-1.52  (m, 5H), 1.53 (s, 3H, Me-14), 1.72 (ddd, 1H, J = 6.5, 11.5, 14 Hz, H-11), 1.85-1.97 (m, 2H, H-1, H-10'), 2.02 (ddd, 1H, J = 2, 7, 14 Hz, H-11'), 2.14-2.21 (m, 1H, H-3), 3.85 (s, 3H, Off. 13 C  nmr (75.3 MHz) 8:12.6 (C-14), 18.9 (C-15), 21.5 (C-12), 27.3 (C-2), 29.3 (C-11),  36.2 (C-10), 37.4 (C-3), 41.4 (C-9), 52.8 (OMe), 56.0 (C-1), 58.7 (C-4), 62.8 (C-8), 67.4 (C-6), 75.6 (C-7), 164.6 (C-13), 210.5 (C-5). Anal. calcd for C16H2204: C 69.04, H 7.97; found: 69.20, H 8.09. Exact Mass calcd: 278.1518; found: 278.1521.  171  The synthetic product exhibited 1 H nmr and 13 C nmr spectra identical with those derived from the esterification of natural epoxycantabronic acid.°  172  V.3. EXPERIMENTAL SECTION FOR THE SYNTHESIS OF (±)CRINIPELLIN B (15).  Preparation of the Say' Enol Ether 200.  To a cold (-78 °C), stirred solution of isopropylmagnesium bromide (49.5 mL of a 2 M solution in THF, 99 mmol) in dry THF (200 mL) was added copper(I) bromide-dimethyl sulfide complex (994 mg, 4.8 mmol, 7 mol %) and the resultant mixture was stirred for 15 min. Freshly distilled, dry HMPA 97 (23 mL, 132 mmol) was added and, after 20 min, a solution of 2-methyl-2-cyclopenten-1 -one (6.35 g, 66 mmol) (commercially available) and Me3SiC1 (16.8 mL, 132 mmol) in dry THE (20 mL) was added dropwise over 25 min. As the addition proceeded, the greyish to brownish reaction mixture turned bright yellow, then faded and became pale yellow. Stirring was continued for 4 h at -78 °C. Dry triethylamine (19.3 mL, 138.8 mmol) and, after 10 min, pentane (400 mL) were added. The reaction mixture was poured into water (200 mL) and the layers were quickly separated while each phase was still cold. The aqueous layer was extracted with pentane (200 mL) and the combined organic extracts were washed with water (3 x 200 mL), brine (200 mL), dried over anhydrous magnesium sulfate and concentrated. The crude silyl enol ether 200 was quickly purified by flash chromatography on silica gel (65 g, pentane) and the liquid thus obtained was distilled (air-bath temperature 107-115 °C/13 Ton) to provide 13.21 g (94%) of the silyl enol ether 200 as a colourless oil.  173  Ir (neat): 1690, 1467, 1382, 1328, 1253, 1087, 945, 911, 882, 843, 757 cm -1 . 1H  nmr (400 MHz) 5: 0.17 (s, 9H, Si Mel), 0.70, 0.88 (d, d, 3H each, J = 7 Hz in each  case, CHMe2), 1.44-1.59 (m, 4H, includes the signal (1.47, ddd, J = 1, 2, 2 Hz) due to the vinylic methyl), 1.69-1.79 (m, 1H), 1.82-1.91 (m, 1H, CHMe2), 2.16-2.24 (m, 2H), 2.41-2.48 (unresolved m, 1H, allylic methine). In decoupling experiments, irradiation at 8 1.82-1.91 (CHMe2) collapsed the two doublets (CHMe2) at 5 0.70 and 0.88 into two singlets, and the unresolved multiplet at 2.41-2.48 (allylic methine) became a broad triplet. Exact Mass calcd for C12H240Si: 212.1596; found: 212.1605.  174  Preparation of 2-Bromomethyl-l-butene (201). Br  201 2-Bromomethyl-l-butene (201) 11 ° was derived from the reaction of 2-hydroxymethyl1-butene 111 with triphenylphosphine and bromine in CH2C12. 760 2-Hydroxymethyl-l-butene was obtained by reduction (LiA1H4, ether) of the corresponding aldehyde, which was, in turn, prepared from butyraldehyde, an aq sol. of formaldehyde (37%) (1.2 equiv) and dimethylammonium chloride (1.2 equiv), as described in the literature. 111 To a cold (0 °C), stirred solution of triphenylphosphine (79.5 g, 0.303 mol) in dry CH2C12 (650 mL) was slowly added a solution of bromine (48.3 g, 0.303 mol) in dry CH2C12 (35 mL). A few crystals of triphenylphosphine were added to the mixture until it became colourless. The reaction mixture was stirred for 15 min and a solution of the 2hydroxymethyl- 1-butene (24.85 g, 0.289 mol) in dry CH2C12 (35 mL) was added dropwise. The solution was warmed to room temperature and stirred for 7 h. Petroleum ether was added (-500 mL) and a white precipitate formed. After refrigeration of the mixture for 10 h, it was filtered through Florisil (300 g, using petroleum ether as eluant) and the solvent was slowly removed by distillation at atmospheric pressure. Distillation (110-127 °C/760 Torr, literaturel 10 95-100 °C) of the brown liquid thus obtained afforded 30.94 g (72%) (purest fraction) of 2-bromomethyl-l-butene (201) as a colourless oil, a strong irritant and a lachrymator. Ir (neat): 3083, 1643, 1458, 1438, 1210, 908, 724 cm -1 . 1H  nmr (400 MHz) 8: 1.07 (t, 3H, J = 7.5 Hz, Cli3CH2), 2.23 (qm, 2H, J for  q = 7.5 Hz, CH3CH2), 3.97 (s, 2H, CH2Br), 4.95 (m, 1H), 5.14 (m, 1H).  175  Preparation of the Keto Alkene 204. 19 13 14  0  18  ^  20  '''' 17  10 1  9*  204 To a cold (0 °C) solution of the trimethylsilyl enol ether 200 (13.12 g, 61 8 mmol) in dry THF (240 mL) was added a solution of MeLi in Et20 (43 mL, 64 9 mmol). Stirring at this temperature was continued for 50 min The reaction mixture was cooled to -20 °C and a solution of 2-bromomethyl- 1 -butene (201) (14.64 g, 98.2 mmol) and tetrakis(triphenylphosphine)palladium (3.88 g, 3 4 mmol) in dry THF (30 mL) was added dropwise, over 15 min, via a cannula. The resultant mixture was stirred for 2 h at -20 °C and 5 h at 0 °C. The solution was poured into a mixture of sat. aq NH4C1 (200 mL) and ether (300 mL). The phases were separated, and the aqueous layer was extracted with ether (2 x 300 mL). The combined organic extracts were washed with brine (250 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatographies (1St: 350 g of silica gel; 2nd: 100 g; 3rd : -,,, iu g; 4th: 40 g, 98 : 2 to 96 : 4 petroleum ether-ether) of the crude product gave, in the earlier fractions, a mixture of compounds containing triphenylphosphine and small amounts of bisalkylated products and, in the later fractions, the desired alkene 204. Distillation (air-bath temperature 103-111 °C/15 Ton) of the liquid thus obtained provided 9.808 g (76%) of the keto alkene 204 as a colourless oil. Ir (neat): 3084, 1739, 1641, 1461, 1408, 1387, 1371, 1079, 894 cm -1 . 1H  nmr (400 MHz) 6: 0.90 (s, 3H, Me-17), 0.92 (d, 3H, J = 7 Hz, Me-19), 0.96 (t, 3H,  J = 7 Hz, Me-16), 1.00 (d, 3H, J = 7 Hz, Me-20), 1.35-1.49 (m, 1H, H-13), 1.59-1.72  176  (m, 1H, H-18), 1.72-1.88 (m, 3H, includes a br q (1.76, 2H, J = 7 Hz, H-8) and H-12), 2.01-2.18 (m, 3H, includes a d (2.15, 1H, J = 14 Hz, H-10), H-13' and H-14), 2.33 (br dd, 1H, J = 8, 18 Hz, H-14'), 2.65 (d, 1H, J = 14 Hz, H-10'), 4.66 (br s, 1H, H-9*), 4.83 (br d, 1H, J = 1.4 Hz, H-9*'). In decoupling experiments, irradiation at 8 2.65 (H-10') collapsed the doublet at 2.15 (H-10) into a singlet. Irradiation at 8 1.59-1.72 (H-18) collapsed the two doublets at 0.92 and 1.00 (Me-19 and Me-20) into singlets and simplified the multiplet (H-12) at — 1.80-1.88. Detailed 1 H nmr data, including those derived from COSY experiments, are given in Table 4. 13 C nmr (75.3 MHz) 8: 11.9 (-ve), 19.0 (-ve), 21.1 (-ve), 22.3 (-ve), 23.4, 29.3 (-ve), 29.4,  37.3, 44.1, 47.0 (-ye), 51.7, 112.7 (C=CH2), 147.8 (C=CH2), 223.7 (C=0). Anal. calcd for C14H240: C 80.71, H 11.61; found: C 80.55, H 11.60. Exact Mass calcd: 208.1828; found: 208.1821.  The earlier fractions contained two products of bisalkylation in addition to triphenylphosphine. By 1 H nmr spectroscopy, these products of bisalkylation showed two sets of vinyl hydrogen signals: 8 4.67 (br s, 1H), 4.73 (br s, 1H), 4.76 (br s, 1H), 4.84 (br s, 1H) and 4.68 (br s, 1H), 4.70 (br s, 1H), 4.79 (br s, 1H), 4.86 (br d, 1H, J = 1.5 Hz).  177  Table 4: 1 H nmr Data (400 MHz, CDC13) for the Keto Alkene 204a. 19 13 14^1. %18^23  0 Me  16^9*  204  Assignment H-x  COSY Correlationsb H-x  1H nmr (400 MHz) 6 ppm (mult., J(Hz), # of H)  H-8  1.76 (br q, J = 7, 2H)  H-9*', Me-16  H-9*e  4.66 (br s, 1H)  H-9*', H-10, H-10'  H-9*'d  4.83 (br d, J= 1.4, 1H)  11-8, H-9*, H-10'  H-10  2.15 (d, J = 14, 111)  H-9*, H-10'  H-10'  2.65 (d, J= 14, 111)  H-9*, H-9*', H-10  H-12  111, part of the m (311) at 1.72-1.88  H-13, H-13', H-14'e, H-18  H-13  1.35-1.49 (m, 1H)  H-12, H-13', H-14, H-14'  H-13'  1H, part of the m (3H) at 2.01-2.18  H-12, H-13, H-14, H-14'  H-14  111, part of the m (3H) at 2.01-2.18  H-13, H-13', H-14'  H-14'  2.33 (br dd, J= 8, 18, 1H)  H-12e, H-13, H-13', H-14  Me-16  0.96 (t, J = 7, 3H)  11-8  Me-17  0.90 (s, 3H)  H-18  1.59-1.72 (m, 1H)  H-12, Me-19, Me-20  Me-19  0.92 (d, J = 7, 311)  H-18  Me-20  1.00 (d, J = 7, 311)  H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations that could be unambiguously assigned are recorded. c- * indicates a hydrogen on a carbon that will not be found later on in crinipellin B (15). d- H' indicates the hydrogen of a pair which is more downfield (H-9' is more downfield than H-9). e- W-coupling.  178  Preparation of the Diketone 196. 19 13  A cold (-78 °C) solution of the keto alkene 204 (3.062 g, 14.7 mmol), in 4 : 1 CH2C12-Me0H (150 mL), was treated with a stream of ozone until a purple tint appeared. The solution was then allowed to stand for 15 min. The excess ozone was removed with a stream of argon, and Me2S (10.8 mL, 147 mmol) was added to the cold solution. The reaction mixture was warmed to room temperature and was then stirred for 2.5 h. The solution was concentrated under reduced pressure and the crude material was treated with a mixture of 10% HC1-H20 (50 mL) in THE (100 mL) for 18 h at room temperature. The reaction mixture was neutralized by careful addition of solid Na2CO3 and then was diluted with ether (60 mL). The phases were separated and the aqueous layer was extracted with ether (2 x 60 mL). The combined organic extracts were washed with brine (50 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (80 g of silica gel, 85 : 15 petroleum etherether) of the residual material, followed by distillation (air-bath temperature 101-110 °C/0.2 Torr) of the liquid thus obtained, afforded 2.861 g (93%) of the diketone 196 as a colourless oil. Ir (neat): 1741, 1714, 1460, 1407, 1113, 1083 cm -1 . 1H  nmr (400 MHz) 5: 0.86 (s, 3H, Me-17), 0.93 (d, 3H, J = 6 Hz, Me-19), 0.94 (d, 3H,  J = 6 Hz, Me-20), 1.00 (t, 3H, J = 7 Hz, Me-16), 1.39 (dddd, 1H, J = 8.5, 12.5, 12.5,  12.5 Hz, H-13), 1.53-1.66 (m, 1H, H-18), 1.94 (ddd, 1H, J = 6.5, 10, 12.5 Hz, H-12),  179  2.15 (dddd, 1H, J = 1, 6.5, 9.5, 12.5 Hz, H-13'), 2.21-2.45 (m, 3H, H-8, H-14), 2.56 (ddd, 111, J = 9.5, 12.5, 18.5 Hz, H-14'), 2.87 (d, 1H, J = 18.5 Hz, H-10), 2.93 (d, 1H, J = 18.5 Hz, H-10').  Detailed 1 H nmr data, including those derived from COSY experiments, are given in Table 5. Anal. calcd for C13H2202: C 74.24, H 10.54; found: C 74.04, H 10.46. Exact Mass calcd: 210.1620; found: 210.1623.  180  Table 5: 1 H nmr Data (400 MHz, CDC13) for the Diketone 196a. 19 13 14^018^al  Assignment H-x  1H nmr (400 MHz) 8 ppm (mult., J(Hz), # of H)  COSY Correlationsb H-x H-10/H-10'c, Me-16  H-8  2H, part of the m (3H) at 2.21-2.45.  H-10/H-10'c  2.87, 2.93 (d, d, J = 18.5 in each case) H-8, Me-17  H-12 H-13  1.94 (ddd, J = 6.5, 10, 12.5, 1H) 1.39 (dddd, J = 8.5, 12.5, 12.5, 12.5, 1H)  H-13, H-13', H-18 H-12, H-13', H-14, H-14'  H-13 'd  2.15 (dddd, J = 1, 6.5, 9.5, 12.5, 1H)  H-12, H-13, H-14, H-14'  H-14  1H, part of the m (3H) at 2.21-2.45.  H-13, H-13', H-14'  H-14'  2.56 (ddd, J = 9.5, 12.5, 18.5, 1H)  H-13, H-13', H-14  Me-16  1.00 (t, J = 7, 3H)  H-8  Me-17  0.86 (s, 3H)  H-10/H-10'c  H-18  1.53-1.66 (m, 1H)  H-12, Me-19, Me-20  Me-19  0.93 (d, J = 6, 3H)  H-18  Me-20  0.94 (d, J = 6, 3H)^_ H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations that could be unambiguously assigned are recorded. c- Since H-10 and H-10' appear very close to each other in the nmr spectrum, it is not possible to determine if only one, or the two of them, correlate to the indicated hydrogen(s). d- H' indicates the hydrogen of a pair which is more downfield (H-13' is more downfield than H-13).  181  Preparation of the Enone 194.  Me16  O  194  To a solution of the diketone 196 (2.861 g, 13.6 mmol) in Me0H (30 mL) was added a 2 M solution of sodium methoxide in Me0H (8.8 mL, 17.7 mmol). The resultant mixture was refluxed for 15 h, cooled to room temperature, diluted with ether (60 mL) and washed with sat. aq NH4C1 (25 mL). The aqueous layer was extracted with ether (3 x 60 mL) and the combined organic extracts were washed with brine (50 mL), dried over anhydrous magnesium sulfate (with stirring), and concentrated. Flash chromatography (90 g of silica gel, 85 : 15 petroleum ether-ether) of the residual material, followed by distillation (air-bath temperature 7683 °C/0.2 Ton) of the liquid thus obtained gave 2.537 g (97%) of the enone 194 as a colourless oil. Jr (neat): 1708, 1669, 1460, 1379, 1330, 1060, 1037 cm -1 . 1H  nmr (400 MHz) 8: 0.90 (d, 3H, J = 5 Hz, Me-19), 0.92 (d, 3H, J = 5 Hz, Me-20),  0.97 (s, 3H, Me-17), 1.15 (ddd, 1H, J = 8, 11, 11 Hz, H-12), 1.49-1.61 (m, 1H, H-18), 1.63 (dd, 3H, J = 0.8, 1.4 Hz, Me-16), 1.72 (dddd, 1H, J = 7, 11, 11, 13 Hz, H-13), 2.10 (dddd, 1H, J = 3.5, 8, 8, 13 Hz, H-13'), 2.25 (d, 1H, J = 17 Hz, H-10), 2.33 (d, 1H, J = 17 Hz, H-10'), 2.40-2.59 (m, 2H, H-14, H-14'). In decoupling experiments, irradiation of the two doublets at 8 0.90 and 0.92 collapsed the multiplet at 1.49-1.61 (H-18) into a broad doublet with J = 11 Hz. Irradiation of the ddd at 8 1.15 (H-12) sharpened the multiplet at 1.49-1.61 (H-18) and modified the signals at 1.72 182  (H-13) and 2.10 (H-13'). Irradiation of the multiplet at 8 1.49-1.61 (H-18) collapsed the ddd at 1.15 (H-12) into a broad triplet (unresolved dd) and the two doublets at 0.90 and 0.92 (Me19 and Me-20) into singlets. Anal. calcd for C13H200: C 81.20, H 10.48; found: C 81.14, H 10.39. Exact Mass calcd: 192.1514; found: 192.1523.  183  Preparation of the Keto Germane 219.  O  219  To a cold (-97 °C), rapidly stirred solution of freshly distilled 4-iodo-2trimethylgermyl- 1-butene (208) (893 mg, 2.99 mmol) in dry THF (30 mL) was quickly added a solution of tert-butyllithium in pentane (3.8 mL, 5.68 mmol). The resultant bright yellow and cloudy solution was stirred at -97 °C for 10 min and was then warmed to -78 °C. Copper(I) cyanide 104 (294.6 mg, 3.29 mmol) was added in one portion and the suspension became pale yellow after —5 min. Warming to -30 °C for 3 min provided a homogeneous solution, either tan or pale yellow in colour. The solution of the vinylgermane cuprate 209 was recooled immediately to -78 °C to avoid decomposition. At -78 °C, the solution became heterogeneous again. A mixture of the freshly distilled enone 194 (363.4 mg, 1.89 mmol) and TMSBr (1.5 mL, 11.3 mmol) in THF (4 mL) was added dropwise. As the addition proceeded, the reaction mixture turned bright yellow, then bright orange and eventually brownish and homogeneous. Stirring was continued for 8 h at -78 °C and at -48 °C for 2 h. The solution was poured into water (20 mL) and the resultant mixture was stirred for 15 min. Ether (30 mL) and aq NH4C1-NH4OH (pH 8-9) (20 mL) were added, and the mixture was stirred vigorously (open to the atmosphere) until the aqueous layer was blue (overnight). The phases were separated and the aqueous layer was extracted with ether (3 x 20 mL). The combined organic extracts were washed with brine (20 mL), dried over anhydrous magnesium sulfate and the solvent was removed. The resulting crude product was purified by flash chromatography (35 g of silica gel, 95 : 5 petroleum ether-ether) and the oil thus obtained was distilled (air-bath temperature 132-139 °C/0.1 Torr) to afford 571.7 mg (83%) of the keto 184  germane 219, a mixture of epimers, as a colourless oil. The ratio of those epimers varied somewhat from experiment to experiment but was found to be, in this case, 4 : 1. Ir (neat): 3052, 1738, 1472, 1411, 1378, 1236, 1183, 916, 825, 760, 599 cm -1 . 1H  nmr (400 MHz) (selected methyl and vinyl hydrogen signals) 8: 0.19 (s, 9H, ne3Ge,  major epimer), 0.21 (s, 9H, auGe, minor epimer), 0.86 (d, 3H, J = 6.5 Hz, CHMe2, minor), 0.88 (d, 3H, J = 6 Hz, major), 0.95 (d, 3H, J = 6 Hz, major), 1.01 (s, 3H, angular Me, minor), 1.05 (d, 3H, J = 7 Hz, major), 1.07 (s, 3H, angular Me, major), 1.08 (d, 3H,  J = 7.5 Hz, minor) and 5.14 (1H, m, major), 5.18 (1H, m, minor), 5.47 (1H, m, major), 5.52 (1H, m, minor).  Anal. calcd for C20H36GeO: C 65.80, H 9.94; found: C 65.98, H 10.04. Exact Mass calcd: 366.1978; found: 366.1972.  185  Preparation of the Keto Iodide 220.  O  220 To a solution of the keto germane 219 (514.0 mg, 1.41 mmol) in dry CH2C12 (28 mL) was added in one portion solid iodine (536.2 mg, 2.11 mmol) and the resultant dark red mixture was stirred at room temperature for 16 h. The solution was diluted with CH2C12 (50 mL), washed successively with sat. aq Na2S2O3 (45 mL) and brine (30 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography on silica gel (26 g, 95 : 5 petroleum ether-ether) provided 525.3 mg (quantitative yield) of the keto iodide 220, a mixture of epimers, as a white solid. The ratio of the epimers varied from experiment to experiment but was found, in the present case, to be 2 : 1. Ir (KBr): 1730, 1615, 1469, 1407, 1376, 1184, 1162, 1123, 901 cm -1 . 1H  nmr (400 MHz) (selected methyl and vinyl hydrogen signals) 8: 0.86 (d, 3H,  J = 6.5 Hz, minor), 0.88 (d, 3H, J = 6 Hz, major), 0.95 (d, 3H, J = 6 Hz, major), 1.05 (d, 3H, J = 7 Hz, major), 1.07 (s, 3H, angular Me, major), 5.63 (d, 1H, J = 1.5 Hz, major), 5.68 (d, 1H, J = 1.5 Hz, minor), 5.96 (dd, 1H, J = 1.5, 2.5 Hz, major), 6.03 (dd, 1H, J = 1.5, 3 Hz, minor). Anal. calcd for C17H27I0: C 54.55, H 7.27, I 33.90; found: C 54.73, H 7.32, I 33.71. Exact Mass calcd: 374.1109; found: 374.1117.  186  Preparation of the Keto Alkene 191. 19  To a stirred solution of tetrakis(triphenylphosphine)palladium (535 mg, 0.46 mmol) in dry THF (15 mL) at room temperature was added a solution of the keto iodide 220 (805.2 mg, 2.15 mmol) in dry THF (8 mL). The reaction mixture was stirred for 5 min and a solution of potassium tert-butoxide in a 4 : 1 mixture of dry THF and dry tert-butyl alcohol (0.21 M, 11.8 mL, 2.47 mmol) was added via a syringe pump over a period of 4 h. After the solution had been stirred for an additional hour, it was diluted with ether (50 mL) and washed with sat. aq NH4C1 (40 mL). The aqueous layer was extracted with ether (2 x 50 mL) and the combined organic extracts were washed with brine (50 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (90 g of silica gel, 97 : 3 petroleum etherether), followed by distillation (air-bath temperature 89-97 °C/0.2 Torr) of the liquid obtained, gave 443.2 mg (84%) of the keto alkene 191 as a white solid, mp 22.0-22.5 °C. Ir (neat): 3080, 1737, 1651, 1456, 1412, 1369, 1254, 1219, 1184, 891, 558 cm -1 . 1H  nmr (400 MHz) 8: 0.86 (d, 3H, J = 6.5 Hz, Me-19), 0.93 (d, 3H, J = 6.5 Hz, Me-  20), 1.01 (s, 3H, angular Me), 1.13 (ddd, 1H, J = 9, 9, 10 Hz), 1.18 (s, 3H, angular Me), 1.37-1.50 (m, 1H), 1.51-1.64 (m, 3H), 1.73-1.98 (m, 3H), 2.21-2.37 (m, 3H, includes a d (2.25, J = 17.5 Hz, H-10)), 2.43 (d, 1H, J = 17.5 Hz, H-10'), 4.86 (dd, 1H, J = 2, 2 Hz, H-7*), 4.95 (dd, 1H, J = 2, 2 Hz, H-7*').  187  13 C nmr (125.8 MHz) 8: 16.0, 22.3, 22.4, 22.8, 29.3, 30.8, 31.8, 33.4, 36.2, 47.9, 49.4,  55.8, 61.5, 66.0, 106.2 (C=CH2), 156.2 (C=CH2), 219.1 (L=0). Anal. calcd for C17H260: C 82.87, H 10.64; found: C 82.95, H 10.75. Exact Mass calcd: 246.1984; found: 246.1988.  188  Preparation of the Alcohol 221. 19 13  13 14  14  20  2  2  16 OH^  7*^7*  221^  18  16 OH  222  To a cold (-78 °C), stirred solution of the keto alkene 191 (1.332 g, 5.41 mmol) in dry ether (30 mL) was added a 0.4 M solution of lithium (diisobutyl)(n-butyl)aluminum hydride in ether (17 mL, 6.97 mmol) . The solution was stirred at this temperature for 2 h and at 0 °C for 1 h. Finely ground, solid Na2SO4•10 H2O (871.6 mg, 2.71 mmol) was added to the solution and the mixture was stirred for 15 min. Aqueous NaOH (1 N, 30 mL) was slowly poured into the reaction vessel and stirring was continued for another 10-15 min. The mixture was diluted with ether (40 mL) and the phases were separated. The aqueous layer was extracted with ether (2 x 75 mL). The combined organic extracts were washed with sat. aq NH4C1 (75 mL), brine (60 mL), dried over anhydrous magnesium sulfate and the solvent was removed. Flash chromatography (80 g of silica gel, 96 : 4 to 92 : 8 petroleum ether-ether) of the crude material gave, upon concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), two products. The initially eluted substance was the desired alcohol 221 (1.257 g, 94%), a white solid that could be recrystallized from CH3CN at -11 °C (1.244 g from 2 crops, 93%), mp 53.0-53.5 °C. Ir (KBr): 3319 (broad), 3096, 1648, 1474, 1367, 1270, 1156, 1098, 1030, 979, 893 cm-1.  189  1H  nmr (400 MHz) 8: 0.88 (d, 3H, J = 6.5 Hz, Me-19), 0.93 (s, 3H, Me-17), 0.96 (d, 3H,  J = 6.5 Hz, Me-20), 1.06 (dd, 1H, J = 9, 13 Hz, H-10), 1.10 (s, 3H, Me-16), 1.13-1.44  (m, 4H, H-2, H-12, H-13, H-14), 1.50-1.63 (m, 1H, H-18), 1.66 (d, 1H, J = 9 Hz, OH signal that exchanges with D20), 1.69-1.86 (m, 2H, H-13', H-14'), 2.01 (ddd, 1H, J = 4, 7, 12.5 Hz, H-2'), 2.12 (dd, 1H, J = 6, 13 Hz, H-10'), 2.16-2.28 (m, 1H, H-3), 2.282.38 (m, 1H, H-3'), 3.82 (ddd, 1H, J = 6, 9, 9 Hz, H-9; upon exchange with D20, this signal becomes a dd with J = 6, 9 Hz), 4.86 (br s, 1H, H-7*), 5.10 (m, 1H, H-7*'). In decoupling experiments, irradiation at 8 3.82 (H-9) changed each of the dd at 1.06 (H-10) and 2.12 (H-10') into a d (J = 13 Hz each); irradiation of the signal at 8 1.50-1.63 (H-18) collapsed the two d (Me-19 and Me-20) at 0.88 and 0.96 into singlets and simplified the m (H12) at —1.13-1.24 (part of the multiplet at 1.13-1.44). Detailed 1 H nmr data, including those derived from COSY and NOE experiments, are given in  Table 6. 13 C  (75.3 MHz) 8: 17.9 (-ve), 22.9 (-ve), 23.0 (-ve), 23.5 (-ve), 29.0 (-ve), 29.6, 34.7,  35.5, 35.8, 47.8, 49.9, 57.6, 58.7 (-ve), 65.5, 79.2 (-ve, CHOH), 107.7 (C=CH2), 158.5 (C=CH2)Anal. calcd for C17H280: C 82.20, H 11.36; found: C 82.24, H 11.40. Exact Mass calcd: 248.2140; found: 248.2141.  The more polar compound, the alcohol 222 was isolated in — 2% yield (31.2 mg, slightly impure). The minor alcohol was combined with other samples of the same product obtained from different experiments, distilled (air-bath temperature 130-140 °C/4 Ton) and recrystallized from CH3CN at -11 °C. The white crystals exhibited mp 52.5-54 °C. Ir (KBr): 3510 (broad), 3071, 1646, 1470, 1452, 1087, 1049, 1026, 887 cm -1 .  190  1H  nmr (400 MHz) 8: 0.88 (d, 3H, J = 6.5 Hz, Me-19), 0.91 (s, 3H, Me-17), 0.92 (d, 3H,  J = 6 Hz, Me-20), 1.07 (s, 3H, Me-16), 1.18-1.59 (m, 6H, includes a br s (OH signal that  exchanges upon treatment with D20)), 1.74-1.93 (m, 5H), 2.36-2.43 (m, 2H), 4.00 (dd, 1H, J = 7 .5, 7.5 Hz, H-9), 4.77 (br dd, 1H, J = 1.5, 2 Hz, H-7*), 4.82 (ddd, 1H, J = 1, 2,  2 Hz, H-7*'). Detailed 1 H nmr data, including those derived from COSY and NOE experiments, are given in Table 7. 13 C  (125.8 MHz) 8: 17.2, 19.3, 22.3, 22.9, 29.0, 31.1, 33.6, 34.4, 35.6, 49.1, 49.4, 57.9,  61.0, 66.9, 80.4 (CHOH), 103.7 (C=CH2), 161.2 (C=CH2). Anal. calcd for C17H280: C 82.20, H 11.36; found: C 82.14, H 11.22. Exact Mass calcd: 248.2140; found: 248.2140.  191  Table 6: 1 H nmr Data (400 MHz, CDC13) for the Alcohol 221a. 19  Assignment H-x H-2 H-2'e (a) H-3 (a) H-3' H-7*d H-7*' H-9e  COSY Correlationsb 1H nmr (400 MHz) S ppm (mult., J (Hz), # of H) 1H, part of the m (4H) at 1.13-1.44 H-2', H-3, H-3' H-2, H-3, H-3' 2.01 (ddd, J = 4, 7, 12.5, 1H) H-2, H-2', H-3', H2.16-2.28 (m, 1H) 7*, H-7*' H-2, H-2', H-3, H2.28-2.38 (m, 1H) 7*, H-7*' H-3, H-3', H-7*' 4.86 (br s, 1H) H-3, H-3', H-7* 5.10 (m, 1H) H-10, H-10' J = 6, 9, 9, 1H) 3.82 (ddd,  H-13  1.66 (d, J = 9, 1H) 1.06 (dd, J = 9, 13, 1H) 2.12 (dd, J = 6, 13, 1H) 1H, - 1.13-1.24, part of the m (4H) at 1.13-1.44. 1H, part of the m (4H) at 1.13-1.44.  H-13'  1H, part of the m (2H) at 1.69-1.86.  H-14 H-14' Me-16 Me-17  1H, part of the m (4H) at 1.13-1.44. 1H, part of the m (2H) at 1.69-1.86. 1.10 (s, 3H) 0.93 (s, 3H)  H-18 Me-19 Me-20  1.50-1.63 (m, 1H) 0.88 (d, J = 6.5, 3H) 0.96 (d, J = 6.5, 3H)  OH9e H-10 H-10' H-12  NOE Correlationsb  H-7*, H-10', H-12, Me-16  H-9, H-10' H-9, H-10 H-13, H-13', H-18 H-12, H-13', H-14, H-14' H-12, H-13, H-14, H-14' H-13, H-13', H-14' H-13, H-13', H-14  H-12, Me-19, Me-20 H-18 H-18  H-7*, H-9, H-2', H-3, H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2). d- * indicates a hydrogen on a carbon that will not be found later on in crinipellin B (15). e- Coupling between H-9 and OH9 was not always observed in the 1 H nmr spectrum of 221. It depended on the preparation of the sample. In the COSY spectrum, no coupling was observed between H-9 and OH9.  192  Table 7: 1 H nmr Data (400 MHz, CDC13) for the Alcohol 222a.  Assignment H-x  1H nmr (400 MHz) 8 ppm (mult., J (Hz), # of H)  COSY Correlation sb  H-2  part of the m (6H) at 1.18-1.59, 111  H-2', H-3  H-2' H-3  part of the m (5H) at 1.74-1.93, 111 2.36-2.43 (m, 211)  H-2, H-3 H-2, 11-2', H-7*, H-7*'  H-7*c  4.77 (br dd, J = 1.5, 2.2 Hz, 1H)  H-3  H-7*'d H-9  4.82 (ddd, J = 1, 2, 2, 1H) 4.00 (dd, J = 7.5, 7.5, 1H)  H-3 H-10, OH9  0149  Part of the m (6H) at 1.18-1.59, 1H  H-9  H-10  part of the m (511) at 1.74-1.93, 2H  Me-16  1.07 (s, 311)  Me-17 H-18  0.91 (s, 3H) —1.47-1.59 (m, 1H), part of the m (611) Me-19, Me-20 at 1.18-1.59  Me-19  0.88 (d, J = 6.5, 311)  H-18  Me-20  0.92 (d, J = 6, 3H)  H-18  NOE Correlationsb  H-7*, H-3, Me-16, Me-17  H-7*, 11-9 H-9  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- * indicates a hydrogen on a carbon that will not be found later on in crinipellin B (15). d- H' indicates the hydrogen of a pair which is more downfield (H-7' is more downfield than H-7).  193  Preparation of the Alkene 223.  7*^  OSiMe2But  223  To a cold (-78 °C) solution of the alcohol 221 (376.3 mg, 1.51 mmol) in CH2C12 (15 mL) were added successively dry Et3N (275 mL, 1.97 mmol), and TBDMSOTf (400 gL, 1.74 mmol). Stirring was continued at this temperature for 75 min and at 0 °C for 20 min The reaction mixture was diluted with CH2C12 (10 mL) and poured into sat. aq NaHCO3 (10 mL). The layers were separated and the organic phase was washed with brine (10 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (31 g of silica gel, petroleum ether) of the crude material, followed by distillation (air-bath temperature 128-142 °C/0.2 Ton) provided 535.7 mg (98%) of a white solid. Recrystallization (2 : 10 EtOAcCH3CN, -11 °C) of a small sample afforded white crystals that exhibited mp 45.0-45.5 °C. Ir (KBr): 3096, 1782, 1648, 1471, 1255, 1164, 1111, 1059, 900, 875, 835, 771 cm -1 . 1 H nmr (400 MHz)  8: 0.03, 0.05 (s, s, 3H each, SiMe2But), 0.88 (d, 3H, J = 6.5 Hz, Me-  19), 0.90 (s, 9H, SiMe2But), 0.91 (s, 3H, angular Me), 0.94 (d, 3H, J = 6.5 Hz, Me-20), 1.01-1.43 (m, 8H, includes a s at 5 1.03 (3H, angular Me)), 1.51-1.72 (m, 3H), 1.86 (dd, 1H, J = 6, 12.5 Hz), 2.02 (ddd, 1H, J = 3.5, 7.5, 13 Hz), 2.20-2.40 (m, 2H, H-3), 3.85 (dd, 1H, J = 6, 10 Hz, H-9), 4.93 (dd, 1H, J = 2, 4 Hz, H-7*), 5.05 (dd, 1H, J = 2, 4 Hz, H-7*'). Anal. calcd for C23H42OSi: C 76.17, H 11.67; found: C 75.95, H 11.80. Exact Mass calcd: 362.3004; found: 362.3007. 194  Preparation of the Ketone 224. 19 20  0^öSiMe2But  224  To a  solution of the alkene 223 (507.5 mg, 1.40 mmol) in pyridine (14 mL) was added  in one portion 0504 112 (396.7 mg, 1.56 mmol) and the resultant brown mixture was stirred at room temperature for 23 h. Solid NaHSO3 (728.4 mg, 7 mmol), pyridine 98 (3 mL) and water (16 mL) were added and the reaction mixture was stirred for 1 h. The mixture was diluted with CH2C12 (30 mL) and the phases were separated. The aqueous layer was extracted with CH2C12 (2 x 20 mL) and the combined organic extracts (which were colourless or pale yellow) were combined, washed with 10% hydrochloric acid (20 mL), sat. aq NaHCO3 (20 mL), dried over anhydrous magnesium sulfate and the solvent was removed. If necessary, the crude material was put under reduced pressure (vacuum pump) in order to remove most of the leftover pyridine. Flash chromatography of the remaining material on silica gel (66 g, 60 : 40 petroleum ether-ether) gave an intermediate diol. This material was used directly in the next step. To a cold (0 °C) solution of the diol (obtained as described) above in THE (35 mL) was added, in one portion, solid Pb(OAc)4 113 (859.7, 1.94 mmol) and the resultant white suspension was stirred for 30 min. Ethylene glycol (625 .tL, 11 2 mmol) was added and stirring was continued for another 10 min. The mixture was diluted with ether (100 mL) and washed with 1 N aqueous NaOH (30 mL), brine (20 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography on silica gel (28 g, 95 : 5 petroleum ether-  195  ether) provided 483 mg (95%) of the ketone 224 as a white solid. Recrystallization (2 crops from CH3CN, 475.6 mg, 93%) gave white crystals, mp 47-48.5 °C. Ir (KBr): 1736, 1470, 1375, 1253, 1170, 1076, 1041, 995, 927, 893, 839, 778 cm -1 . 1 H nmr (400 MHz) 8: -0.01, 0.00 (s, s, 3H each, SiMe7,But), 0.84 (s, 9H, SiMe?But), 0.89  (d, 3H, J = 6.5 Hz, Me-19), 0.92 (d, 3H, J = 6.5 Hz, Me-20), 1.05 (s, 3H, angular Me), 1.10 (s, 3H, angular Me), 1.20 (ddd, 1H, J = 6, 10, 12 Hz), 1.29-1.60 (m, 5H), 1.72-1.91 (m, 2H), 2.09-2.20 (m, 2H), 2.26 (ddd, 1H, J = 3.5, 8, 18 Hz), 2.46 (ddd, 1H, J = 8, 11.5, 11.5 Hz), 4.04 (dd, 1H, J = 2.5, 5 Hz, H-9). 13 C nmr (75.3 MHz) 8: -5.3, -4.8, 17.8, 18.9, 19.3, 22.3, 22.9, 25.7, 29.4, 31.5, 31.6,  33.8, 40.1, 50.8, 54.9, 60.8, 63.4, 65.7, 86.0 (CHOH), 221.4 (C=0).  Anal. calcd for C22H40O2Si: C 72.47, H 11.06; found: C 72.37, H 11.05. Exact Mass calcd for C22H39O2Si (M+-1): 363.2719; found: 363.2715.  196  Preparation of (Z)-3-Bromo-l-iodopropene (251).  /=\__ /=\___ /=\ Br OH^I I I^COOMe^ 262^263^251 To a cold (-78 °C), stirred solution of freshly distilled methyl (Z)-3-iodopropenoate (262) 114 (7.729 g, 36 5 mmol) in dry THE (360 mL) was added a 1 M solution of DIBAL  (91 mL, 91 mmol) in hexanes. The resultant solution was stirred for 140 min at -78 °C and 45 min at 0 °C. Finely ground, solid Na2SO4•10 H2O (29.4 g, 91 2 mmol) was added and the mixture was diluted with EtOAc (300 mL). Stirring was continued for 100 min and the gelatinous mixture was filtered through Celite (35 g) using a 150 mL 60 M fitted glass filter and EtOAc as the eluant. 115 The solvent was removed; the water-bath temperature was maintained below 40 °C, otherwise the liquid would turn yellow or pink. Distillation (air-bath temperature 56-80 °C/0.2 Ton) of the liquid thus obtained afforded (Z)-3-iodo-2-propen- 1 -ol (263) in essentially quantitative yield. This compound was used directly for the next step.  To a cold (-30 °C), stirred solution of triphenylphosphine (10.53 g, 40 2 mmol) in dry CH2C12 (300 mL) was added a solution of bromine (4.42 g, 40 2 mmol) in CH2C12 (35 mL). The solution turned pale yellow. A few crystals of triphenylphosphine were added until the resulting solution turned colourless. Stirring at -30 °C to -25 °C was continued for 15 min. A solution of the freshly distilled (Z)-3-iodo-2-propen-1-ol (obtained as described above) in CH2C12 (30 mL) was added to the mixture, and the cold bath was removed. Stirring was continued for 1 h. Most of the solvent was removed under reduced pressure. During this process, the water bath was maintained under 30 °C. Petroleum ether was slowly added to the residual material until a small amount of a white precipitate appeared. The mixture was filtered through Florisil (74 g, elution with petroleum ether) and the eluate was concentrated. Traces of solvent were removed by putting the resulting yellowish to pinkish liquid under reduced pressure (-20 Ton) for 5-10 min at 45-50 °C. The compound was then distilled under reduced 197  pressure (air-bath temperature 25-45 °C/0.2 Torr). The iodide 251 (7.417 g, 82% for the two steps) was obtained as a colourless (or sometimes slightly pink) oil. It could be stored for a few months in a freezer (-11 °C), under inert atmosphere (argon), over a piece of copper wire, without serious decomposition. It is to be noted that this compound is a lachrymator. Ir (neat): 3065, 1649, 1602, 1432, 1300, 1203, 726, 635 cm -1 . 1H  nmr (400 MHz) 8: 3.99 (d, 2H, J = 7 Hz), 6.44-6.54 (m, 2H, vinylic hydrogens).  13 C  nmr (125.8 MHz) 5: 32.6, 88.3, 136.2.  Exact Mass calcd for C3H4BrI: 245.8544 and 247.8524; found: 245.8547 (19.91%) and 247.8525 (21.25%).  198  Preparation of the Keto Iodide 250.  H 5(4 6I  ,II  0^aSiMe2But  I  265  To a cold (-78 °C), stirred solution of freshly distilled, dry diisopropylamine (2661.IL, 1 9 mmol) in dry THF (9 mL) was added a solution of MeLi in ether (1.1 mL, 1.77 mmol). The resultant mixture was stirred at -78 °C for 15 min, at 0 °C for 5 min and then was recooled to -78 °C. A solution of the ketone 224 (crystallized from CH3CN and distilled (138-146 °C/0.2 Torr), 496.4 mg, 1.36 mmol) in dry THF (2.5 mL) was added via a cannula. The resultant yellow reaction mixture was stirred at this temperature for 135 min. A solution of (Z)-3-bromo- 1 -iodo-propene (251) (filtered through flame-dried basic alumina and freshly distilled, 1.286 g, 5.21 mmol) in THF was added. The pinkish solution was warmed to room temperature and stirred for 7.5 h. Over this period of time, the solution became pale yellow, then darkened from orange to deep orange or brownish. The reaction mixture was poured into sat. aq NH4C1 (15 mL) and diluted with ether (15 mL). The phases were separated and the aqueous layer was extracted with ether (2 x 20 mL) The combined organic extracts were washed with brine (15 mL), dried over anhydrous magnesium sulfate and the solvent was removed. The crude material was immediately purified by flash chromatography on silica gel (83 g, 98 : 2 to 96 : 4 petroleum ether-ether) to afford, upon concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), two products along with recovered starting material. The initially eluted compound was the desired keto iodide 250 (545.5 mg, 76%), a colourless oil. 199  Ir (neat): 3072, 1739, 1610, 1472, 1367, 1282, 1255, 1116, 888, 837, 776 cm -1 . 1H  nmr (400 MHz) 8: -0.01, 0.03 (s, s, 3H each, SiMe7But), 0.85 (s, 9H, SiMeaut), 0.88,  0.97 (d, d, 3H each, J = 6.5 Hz in each case, Me-19, Me-20), 1.03 (s, 3H, angular Me), 1.08 (s, 311, angular Me), 1.15-1.26 (m, 211), 1.28-1.48 (m, 3H), 1.52-1.65 (m, 1H), 1.721.83 (m, 2H), 2.08-2.24 (m, 2H, includes a dd at 2.19 with J = 6.5, 13.5 Hz), 2.34-2.48 (m, 3H), 4.01 (dd, 1H, J = 6.5, 8.5 Hz, H-9), 6.18-6.27 (m, 2H, H-5, H-6). Anal. calcd for C25H43IO2Si: C 56.59, H 8.17, I 23.92; found: C 56.29, H 8.06, I 24.01. Exact Mass calcd for C25H42IO2Si (W-1): 529.2001; found: 529.1994. The second eluted compound consisted of another alkylated product 265 (19.9 mg, 3%) (the epimer? of 250), a slightly impure solid. It was combined with other samples of the same compound from different experiments and recrystallized from CH3CN (rt to -11 °C) to afford white crystals mp 71.5-73.0 °C. Jr (KBr): 3064, 1729, 1607, 1473, 1374, 1292, 1257, 1157, 1043, 1003, 928, 887, 836, 809, 780, 665 cm -1 . 1H  nmr (400 MHz) 8: 0.02, 0.03 (s, s, 3H each, SiMe2But), 0.87 (s, 911, SiMe2But), 0.89  (d, 3H, J = 7 Hz, Me-19), 0.90 (d, 3H, J = 7 Hz, Me-20), 1.07 (s, 3H, angular Me), 1.10 (s, 3H, angular Me), 1.17 (ddd, 1H, J = 5.5, 10, 12 Hz), 1.32 (ddd, 111, J = 1.5, 8, 12.5 Hz), 1.38-1.58 (m, 2H), 1.63 (dd, 1H, J = 1.5, 14 Hz), 1.68 (dd, 111, J = 6.5, 10.5 Hz), 1.77 (ddd, 1H, J = 7 .5, 10, 12.5 Hz), 1.83-1.93 (m, 111), 2.07-2.31 (m, 4H), 2.53-2.62 (m, 1H), 4.11 (dd, 1H, J = 1.5, 5 Hz, 11-9), 6.21-6.32 (m, 21-1, vinylic hydrogens). Anal. calcd for C25H43IO2Si: C 56.59, H 8.17, I 23.92; found: C 56.29, H 8.11. Exact Mass calcd for C25H42IO2Si (M -F-1): 529.2001; found: 529.1994. 200  The last eluted compound was the recovered starting material (98 mg, 20%).  Preparation of the Allylic Alcohol 249. 19 17 20  OSiMe 2 But  249  To a cold (-78 °C) solution of the keto iodide 250 (511.7 mg, 0.964 mmol, kept under reduced pressure (vacuum pump) overnight) in dry THE (9 6 mL) was added a 1.36 M solution of n-BuLi in hexanes (1.8 mL, 2.45 mmol). The resultant colourless solution was stirred at this temperature for 110 min and then was poured into a mixture of ether (20 mL) and sat. aq NaHCO3 (15 mL) The phases were separated and the aqueous layer was extracted with ether (2 x 20 mL). The combined organic extracts were washed with brine (10 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography on silica gel (70 g, 97 : 3 to 96 : 4 petroleum ether-ether) of the crude material gave, after concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), the allylic alcohol 249 as a white solid (361.2 mg, 93%). Recrystallization of a small sample from CH3CN at  -11 °C provided white crystals (mp 61.5-63.0 °C). (Special care had to be taken with this compound since it is very acid sensitive. It decomposed in CDC13 that had not been passed through basic alumina). Jr (KBr): 3496, 3051, 1620 (weak), 1471, 1449, 1371, 1361, 1349, 1261, 1082, 1064, 1042, 1028, 1008, 885, 834, 781, 749, 558 cm -1 . 1H  nnu. (400 MHz) 8: 0.09, 0.12 (s, s, 3H each, SiMe2But), 0.88 (d, 3H, J = 6.5 Hz, Me-  19), 0.91 (s, 9H, SiMe2But), 0.96 (s, 3H, angular Me), 0.96 (d, 3H, J = 6.5 Hz, Me-20), 201  1.07 (s, 3H, angular Me), 0.93-1.24 (m, 4H), 1.49-1.70 (m, 3H), 1.88 (br dd, 1H, J = 1.5, 17 Hz, H-4), 1.96 (dd, 1H, J = 10, 12.5 Hz, H-10), 2.01 (dd, 1H, J = 7, 12.5 Hz, H10'), 2.28-2.44 (m, 2H), 2.66 (dddd, 1H, J = 2, 2, 7.5, 17 Hz, H-4'), 4.01 (dd, 1H, J = 7, 10 Hz, H-9), 5.29 (s, 1H, OH; exchanges with D20), 5.63-5.72 (m, 2H, vinyl  hydrogens). In decoupling experiments, irradiation of the m (vinyl hydrogens) at S 5.63-5.72 collapsed the dddd at 2.66 (H-4') into a dd (J =7 .5, 17 Hz) and the dd (H-4) at 1.88 into a d (J = 17 Hz). Irradiation at 8 4.01 (H-9) simplified the dd (H-10 and H-10') at 1.96 and 2.01  into two doublets with J = 12.5 Hz. Irradiation of the dddd (H-4') at 2.66 simplified the signal (vinyl hydrogens) at 5.63-5.72 and collapsed the dd (H-4) at 1.88 into a broad s. 13 C nmr (75.3 MHz) 8116: -5.1, -4.5 (both -ve, SiMe2But), 17.8, 18.2 (-ve), 21.0 (-ve),  23.1 (-ve), 25.7 (-ve), 28.8 (-ve), 29.5, 36.1, 38.6, 43.9, 47.7, 49.3, 49.8 (-ve), 56.5, 59.9 (-ve), 66.2, 82.6 (-ve, LHOH), 100.5 (SOH), 129.8 (-ve, CLI=CH), 136.0 (-ve, CH=C11). Anal. calcd for C25H44O2Si: C 74.20, H 10.96; found: C 74.40, H 10.90. Exact Mass calcd: 404.3110; found: 404.3113.  202  Preparation of the Enedione 267 and of the Enone 231. 13  <  19  17 20 16 E  267  ^  OSiMe 2 But  231  To a suspension of flame-dried Celite (3.46 g) and PCC 107 (3.69 g, 17 1 mmol) in dry CH2C12 (24 mL) was added a solution of the tertiary alcohol 249 (1.385 g, 3.42 mmol) in dry CH2C12 (10 mL). The mixture was vigorously stirred for 3.5 h at room temperature and then was diluted with ether (10 mL) The resultant mixture was sonicated for a few minutes and filtered through Florisil (42 g; elution first with ether and then with AcOEt to ensure that all the enedione 267 had been eluted). The solvent was removed from the eluate and the dark residual material was purified by flash chromatography (96 g of silica gel, 80 : 10 : 10 to 60 : 20 : 20 petroleum ether-ether-CH2C12). Concentration of the appropriate fractions and removal of traces of solvent (vacuum pump) gave three major products. The first eluted substance consisted of the enone 231 (153.4 mg, 11%), a white solid that could be recrystallized from pentane to afford white crystals (mp 86.0-87.0 °C). The second set of fractions contained the epoxide 268 (-6%) identical with the product obtained from the reaction of the enedione 267 with hydrogen peroxide. The last eluted compound was the enedione 267 (519.4 mg, 53%), a white solid that could be recrystallized from ether (465.4 mg, 3 crops; last crop from 80 : 20 pentane-ether, 31.4 mg, 51%) (mp 130-132 °C). The enedione 267 exhibited it (KBr): 1739, 1708, 1617, 1607, 1454, 1421, 1382, 1252, 1219, 1164, 868, 826, 666 cm -1 . 1H  nmr (400 MHz) 5: 0.88, 0.97 (d, d, 3H each, J = 6.5 Hz in each case, Me-19, Me-20),  1.12 (ddd, 1H, J = 8.5, 8.5, 11 Hz), 1.15 (s, 3H, angular Me), 1.22 (dd, 1H, J = 13.5, 203  13.5 Hz, H-2 or H-4), 1.34 (s, 3H, angular Me), 1.39-1.50 (m, 1H), 1.56-1.68 (m, 2H, includes H-18), 1.75-1.94 (m, 2H), 2.01 (dd, 1H, J = 3, 18 Hz, H-4 or H-2), 2.31 (d, 1H,  J = 17.5 Hz, H-10), 2.54-2.66 (m, 2H, H-2', H-4'), 2.70 (d, 1H, J = 17.5 Hz, H-10'), 2.81-2.92 (m, 1H, H-3), 5.89 (d, 1H, J = 2 Hz, H-6). Detailed 1 H nmr data, including those derived from COSY experiments, are given in Table 8. 13 C nmr (75.3 MHz) 8: 16.2 (-ye), 18.0 (-ye), 22.6 (-ve), 22.7 (-ye), 28.4, 30.7 (-ye), 33.0,  40.6, 42.5 (-ye), 43.2, 49.6, 52.0, 56.2 (-ye), 61.4, 68.9, 124.5 (-ye, C=CH), 190.4 (C=CH), 209.3 (C=0), 215.4 (C=0).  Anal. calcd for C19H2602: C 79.68, H 9.15; found: C 79.61, H 9.22. Exact Mass calcd: 286.1933; found: 286.1929.  204  Table 8: 1 H nmr Data (400 MHz, CDC13) for the Enedione 267a. 13 H^14^12^19 4^1114 5^1" 17 a) 10 16  0  267  Assignment H-x  1H nmr (400 MHz) S ppm (mult., J(Hz))  COSY Correlationsb H-x  H-2 (or H-4)  1.22 (dd, J= 13.5, 13.5)  H-2' (or H-4'), 11-3  H-2'e  Part of the m (2H) at 2.54-2.66  Seed  H-3  2.81-2.92 (m)  11-2, H-2'd, H-4, 11-4'd  H-4 (or H-2)  2.01 (dd, J= 3, 18)  H-4' (or H-2'), H-3  H-4'  Part of the m (2H) at 2.54-2.66.  Seed  H-6  5.89 (d, J = 2)  H-10  2.31 (d, J = 17.5)  H-10'  H-10'  2.70 (d, J = 17.5)  H-10  H-18  Part of the m (2H) at 1.56-1.68  Me-19, Me-20  Me-19  0.88 (d, J = 6.5)  H-18  Me-20  0.97 (d, J = 6.5)  H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-13' is more downfield than H-13). d- Since H-2' and H-4' appear very close to each other in the 1 H nmr spectrum, it is not possible to determine which one of them correlates to the indicated hydrogen(s).  205  The enone 231 exhibited it (KBr): 1707, 1619, 1463, 1258, 1107, 1064, 1042, 892, 852, 838, 773, 666 cm -1 . 1H  nmr (400 MHz, CDC13) 8: 0.00, 0.04 (s, s, 3H each, SiMe2But), 0.84 (s, 9H, SiMe7But),  0.90 (d, 3H, J = 6.5 Hz, Me-19), 0.98 (s, 3H, angular Me), 1.00 (d, 3H, J = 6.5 Hz, Me20), 1.02 (dd, 1H, J = 13, 13 Hz, H-2), 1.17-1.41 (m, 7H, includes a s at 1.19 (3H, angular Me) and a dd at 1.24 (1H, J = 10.5, 13 Hz, H-10)), 1.54-1.67 (m, 1H, H-18), 1.70-1.89 (m, 2H), 1.95 (dd, 1H, J = 3, 17.5 Hz, H-4), 2.17 (dd, 1H, J = 6.5, 13.0 Hz, H-10'), 2.35 (dd, 1H, J = 7, 13 Hz, H-2'), 2.57 (dd, 1H, J = 6, 17.5 Hz, H-4'), 2.812.91 (m, 1H, H-3), 4.10 (dd, 1H, J = 6.5, 10.5 Hz, H-9), 5.81 (d, 1H, J = 2 Hz, H-6). Detailed 1 H nmr data, derived from decoupling experiments, are given in Table 9. 13 C  nmr (75.3 MHz) 8116 -4.9, -4.4 (both -ve, SiMe2But), 17.3 (-ve), 18.0, 22.7 (-ve), 23.1 :  (-ve), 25.8 (-ve), 29.0, 29.6 (-ye), 33.7, 41.4, 43.3, 45.1 (-ye), 47.1, 51.7, 57.0 (-ve), 67.4, 79.7 (-ye), 126.3 (-ye, C=CH), 197.7 (C=CH), 210.7 (f=0). 1H  nmr (400 MHz, C6D6) 8: 0.00, 0.02 (s, s, 3H each, SiMe9But), 0.60 (dd, 1H, J = 13,  13 Hz, H-2), 0.72 (s, 3H, angular Me), 0.81 (d, 3H, J = 6.5 Hz, Me-19), 0.92 (s, 9H, SiMe2But), 0.93 (d, 3H, J = 6.5 Hz, Me-20), 0.98 (s, 3H, angular Me), 1.04-1.16 (m, 3H), 1.19 (dd, 1H, J = 10.5, 13 Hz, H-10), 1.38-1.48 (m, 1H, H-18), 1.51-1.66 (m, 2H), 1.72 (dd, 1H, J = 3, 17.5 Hz, H-4), 1.91 (dd, 1H, J = 7, 13 Hz, H-2'), 2.10 (dd, 1H, J = 6.5, 13 Hz, H-10'), 2.34 (dd, 1H, J = 6, 17.5 Hz, H-4'), 2.44-2.54 (m, 1H, H-3), 4.00 (dd, 1H, J = 6.5, 10.5 Hz, H-9), 5.97 (d, 1H, J = 2 Hz, H-6). Detailed 1 H nmr data, derived from decoupling experiments, are given in Table 10. Anal. calcd for C25H42O2Si: C 74.57, H 10.51; found: C 74.66, H 10.60. Exact Mass calcd: 402.2954; found: 402.2954.  206  Table 9: 1 H nmr Data (400 MHz, CDC13) for the Enone 231a: Decoupling Experiments.  16 =  OSiMe2 But  231  Assignment H-x  Signal Being Irradiated 1H nmr (400 MHz) 8 ppm (mult., J (Hz))  H-2 H-2'b  1.02 (dd, J = 13, 13) 2.35 (dd, J = 7, 13)  H-3  2.81-2.91 (m)  H-4  1.95 (dd, J = 3, 17.5)  H-4'  2.57 (dd, J = 6, 17.5)  H-6 H-9  5.81 (d, J = 2) 4.10 (dd, J = 6.5, 10.5)  H-10 H-10'  1.24 (dd, J = 10.5, 13) 2.17 (dd, J = 6.5, 13)  1.54-1.67 (m) H-18 Me-19 0.90 (d, J = 6.5) Me-20^_1.00 (d, J = 6.5)  Signals Being Observed 8 ppm (initial mult., J (Hz), H-x) :^Mult. after irradiation, J (Hz) 1.02 (dd, J = 13, 13, H-2) 2.81-2.91 (m, H-3) 1.02 (dd, J = 13, 13, H-2) 1.95 (dd, J = 3, 17.5, H-4) 2.35 (dd, J = 7, 13, H-2') 2.57 (dd, J = 6, 17.5, H-4') 5.81 (d, J = 2, H-6) 2.57 (dd, J = 6, 17.5, H-4') 2.81-2.91 (m, H-3) 1.95 (dd, J = 3, 17.5, H-4) 2.81-2.91 (m, H-3)  dc . sharpened m ' dc d, J = 17.5 . d, J = 13 d, J = 17.5  1.24 (dd, J= 10.5, 13, H-10) 2.17 (dd, J = 6.5, 13, H-10')  ' d, J = 13 • d, J = 13  1.24 (dd, J = 10.5, 13, H-10) 4.10 (dd, J = 6.5, 10.5, H-9)  . d, J = 10.5 ' d, J = 10.5  1.54-1.67 (m, H-18) 1.54-1.67 (m, H-18)  a- Crinipellin numbering used for consistency. b- H' indicates the hydrogen which is most downfield (H-2' is more downfield than H-2). c- The J could not be obtained since the other part of the d is hidden under a peak. d- The decoupling is incomplete; therefore the J could not be obtained.  207  sharpened m d, J = 6 dd sharpened m  sharpened m sharpened m  Table 10: 1 H nmr Data (400 MHz, C6D6) for the Enone 231a: Decoupling Experiments.  16 = OSiMe 2 But 231  Assignment H-x H-2  Signal Being Irradiated tH nmr (400 MHz) 8 ppm (mult., J (Hz)) 0.60 (dd, J = 13, 13)  H-2'e  1.91 (dd, J = 7, 13)  H-3  2.44-2.54 (m)  H-4  1.72 (dd, J = 3, 17.5)  H-4'  2.34 (dd, J = 6, 17.5)  H-6 H-9  5.97 (d, J = 2) 4.00 (dd, J = 6.5, 10.5)  H-10 H-10'  1.19 (dd, J = 10.5, 13) 2.10 (dd, J = 6.5, 13)  H-18  1.38-1.48 (m)  Me-19  0.81 (d, J = 6.5)  Signals Being Observed 8 ppm (initial mull., J (Hz), H-x) NIult. after irradiation, J(Hz) 1.91 (dd, J = 7, 13, H-2') 2.44-2.54 (m, H-3) 0.60 (dd, J = 13, 13, H-2) 2.44-2.54 (m, H-3) 0.60 (dd, J = 13, 13, H-2) 1.72 (dd, J = 3, 17.5, H-4) 1.91 (dd, J = 7, 13, H-2') 2.34 (dd, J = 6, 17.5, H-4') 5.97 (d, J = 2, H-6) 2.34 (dd, J = 6, 17.5, H-4') 2.44-2.54 (m, H-3) 1.72 (dd, J = 3, 17.5, H-4) 2.44-2.54 (m, H-3)  sharpened m d, J = 13 sharpened m d, J = 13 d, J = 17.5 d, J = 13 d, J = 17.5 s d, J = 6 sharpened m br s sharpened m  1.19 (dd, J = 10.5, 13, H-10) 2.10 (dd, J = 6.5, 13, H-10')  d, J = 13 d, J = 13  db  d, J = 10.5 1.19 (dd, J = 10.5, 13, H-10) d, J = 10.5 4.00 (dd, J = 6.5, 10.5, H-9) 0.81, 0.93 (d, d, each has J = 6.5, s Me-19, Me-20) 1.38-1.48 (m, H-18)  sharpened m  a- Crinipellin numbering used for consistency. b- The decoupling is incomplete; therefore the J could not be obtained. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  208  Preparation of the Dione Epoxide 268. 19 20  268  0  To a cold (0 °C) solution of the enedione 267 (300.1 mg, 1.05 mmol) in 2 : 1 THFH20 (15.9 mL) were added solid NaHCO3 (880 mg, 10.5 mmol) and 30% aq H202 (1.6 mL) The reaction mixture was warmed to room temperature and stirred for 55 min. The solution was diluted with CH2C12 (20 mL) and water (10 mL) and the layers were separated. The aqueous layer was extracted with CH2C12 (3 x 10 mL). The combined organic extracts were washed with a 10% aqueous solution of NaHSO3 (10 mL), brine (10 mL), dried over anhydrous magnesium sulfate and concentrated. During the concentration process, the waterbath temperature was kept below 30 °C. The crude material was immediately purified by flash chromatography on iatrobeads (7.5 g, 60 : 20 : 20 pentane-ether-CH2C12) to provide 271.9 mg (86%) of the epoxide 268 as a whitish solid that could be recrystallized (2 crops from EtOAc at -11 °C, 1 crop from ether at -11 °C; 267.3 mg, 84%) to yield white crystals (mp 141.5-144 °C). Ir (KBr): 1734, 1454, 1418, 1368, 1256, 1182, 976, 883 cm -1 . 1H  nmr (400 MHz, CDC13) 8: 0.85 (d, 3H, J = 6.5 Hz, Me-19), 0.95 (d, 3H, J = 6.5 Hz,  Me-20), 0.97 (s, 3H, angular Me), 1.09 (ddd, 1H, J = 8.5, 8.5, 11 Hz), 1.11 (s, 3H, angular Me), 1.19 (dd, 1H, J = 15, 16 Hz), 1.42 (dddd, 1H, J = 6, 11, 11, 13 Hz), 1.531.66 (m, 2H), 1.73-1.83 (m, 1H), 1.89 (ddd, 1H, J = 6, 9.5, 14.5 Hz), 2.03 (d, 1H,  J = 18.5 Hz), 2.30 (d, 1H, J = 17.5 Hz, H-10), 2.39-2.56 (m, 3H), 2.71 (d, 1H, J = 17.5 Hz, H-10'), 3.24 (s, 1H, H-6).  209  1 H nmr (400 MHz, C6D6) 8: 0.57 (dd, 1H, J = 13.5, 13.5 Hz), 0.60 (s, 3H, angular Me),  0.66 (d, 3H, J = 6.5 Hz, Me-19), 0.70 (s, 3H, angular Me), 0.74 (d, 3H, J = 6.5 Hz, Me20), 0.85-1.13 (m, 3H), 1.27 (dddd, 1H, J = 6.5, 6.5, 8.5, 13 Hz), 1.33-1.49 (m, 2H), 1.61 (d, 1H, J = 19 Hz), 1.75 (dd, 1H, J =7 , 13.5 Hz), 1.84-1.94 (m, 2H, includes a dd at 8 1.89 (1H, J = 0.4, 17 Hz)), 2.07 (dd, 1H, J = 7.5, 19 Hz), 2.48 (d, 1H, J = 17 Hz), 3.02 (s, 1H, H-6).  Anal. calcd for C19H2603: C 75.46, H 8.67; found: C 75.28, H 8.74. Exact Mass calcd: 302.1882; found: 302.1886.  210  Preparation of the Enedione Epoxide 188. 19 18  17 20  188  To a cold (-78 °C) solution of lithium 1, 1, 1, 3, 3, 3-hexamethyldisilazide (4 mL of 0.15 M, 0.60 mmol) in dry THF was added a solution of the dione epoxide 268 (60.9 mg, 0.201 mmol, dried overnight under reduced pressure (vacuum pump) at room temperature) in dry THF (2 mL). The resultant colourless solution was stirred at -78 °C for 18 min. Solid N,N-dimethyl(methylene)ammonium iodide (273, 150.6 mg, 0.814 mmol, recrystallized and dried at 80 °C under reduced pressure (vacuum pump) for 2 h) was added to the solution in one portion. Stirring at -78 °C was continued for 70 min and at -70 °C for 18 min. The reaction mixture was poured into a flask containing sat. aq NaHCO3 (10 mL) and EtOAc (10 mL). The phases were separated and the aqueous layer was extracted with EtOAc (4 x 10 mL). The combined organic extracts were washed with brine (10 mL), dried over anhydrous magnesium sulfate and concentrated. Filtration of the crude material through iatrobeads (2.7 g, elution first with EtOAc and then with Me0H), followed by flash chromatography on iatrobeads (1st flash: 7.5 g, 80 : 20 : 20 pentane-ether-CH2C12, 2nd flash: 2.7 g, 80 : 20 : 20 pentane-ether-CH2C12) afforded, after concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), 49.6 mg (78%) of the enedione epoxide 188 as a white solid. Recrystallization from EtOAc yielded white crystals (mp 157.5-158.5 °C). X-ray crystallographic analysis confirmed the structure of this solid. The experimental details of this analysis are listed in the appendix.  211  The second eluted substance was the recovered starting material, the dione epoxide 268 (4.4 mg, 7%). The epoxide 188 exhibited it (KBr): 3104, 3058, 1728, 1641, 1425, 1267, 1140, 962, 935, 876 cm -1 . 1 H nmr (400 MHz) 5: 0.86, 0.96 (d, 3H each,  J = 6.5 Hz, Me-19, Me-20), 1.00 (s, 3H,  angular Me), 1.09 (ddd, 1H, J = 8.5, 8.5, 11.5 Hz, H-12), 1.16 (s, 3H, angular Me), 1.33 (dd, 1H, J = 13, 14.5 Hz, H-2), 1.43 (dddd, 1H, J = 6, 11.5, 11.5, 13 Hz, H-13), 1.541.66 (m, 2H, includes H-14 and H-18), 1.74-1.84 (m, 1H, H-13'), 1.90 (ddd, 1H, J = 6, 9.5, 14 Hz, H-14'), 2.35 (d, 1H, J = 17.5, H-10), 2.57 (dd, 1H, J = 7.5, 14.5 Hz, H-2'), 2.72 (d, 1H, J = 17.5 Hz, H-10'), 3.05 (ddm, 1H, J = 7.5, 13 Hz, H-3), 3.45 (s, 1H, H-6), 5.45 (m, 1H, H-15), 6.12 (d, 1H, J = 1.5 Hz, H-15'). Detailed 1 H nmr data, including those derived from COSY experiments, are given in Table  1 1. 13 C nmr (75.3 MHz) 5: 15.9 (-ve), 16.1 (-ye), 22.6 (-ye), 22.7 (-ye), 28.2, 30.7 (-ye), 32.6,  38.3, 41.7 (-ve), 49.8, 52.1, 53.9, 56.0 (-ve), 57.9 (-ve), 64.4, 77.9, 123.1 (C=CH2), 145.9 (C=CH2), 196.6 (C=0), 216.2 (C=O).  Anal. calcd for C201-12603: C 76.40, H 8.33; found: C 76.49, H 8.25. Exact Mass calcd: 314.1882; found: 314.1881.  212  Table 11: 1 H nmr Data (400 MHz, CDCl3) for the Enedione Epoxide 188a. 19  1H nmr (400 MHz) 5 ppm (mult., J (Hz))  Hydrogen H-x  COSY Correlationsb  H-2  1.33 (dd, J = 13, 14.5)  H-2', H-3  H-2'c  2.57 (dd, J = 7.5, 14.5)  H-2, H-3  H-3  3.05 (ddm, J = 7.5, 13)  H-2, H-2', H-15, H-15'  H-6  3.45 (s)  H-15  H-10  2.35 (d, J = 17.5)  H-10'  H-10'  2.72 (d, J = 17.5)  angular Me (5 1.16), H-10  H-12  1.09 (ddd, J = 8.5, 8.5, 11.5)  H-13, H-13', H-18  H-13  1.43 (dddd, J = 6, 11.5, 11.5, 13)  H-12, H-13', H-14, H-14'  H-13'  1.74-1.84 (m)  H-12, H-13, H-14, H-14'  H-14  Part of the m (2H) at 1.54-1.66  H-13, H-13', H-14'  H-14'  1.90 (ddd, J = 6, 9.5, 14)  H-13, H-13', H-14  H-15  5.45 (m)  H-3, H-6, H-15'  H-15'  6.12 (d, J = 1.5)  H-3, H-15  angular Me  1.16 (s)  H-10'  H-18  Part of the m (2H) at 1.54-1.66  H-12, Me-19, Me-20  Me-19  0.86 (d, J = 6.5)  H-18  Me-20  0.96 (d, J = 6.5)  H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  213  Preparation of the Ketol Epoxide 282. 15  19  19  ,i,' 17^20  20  8  HO"  He  10 16  0^  282^  283  OH  To a cold (-78 °C) solution of enedione epoxide 188 (242.8 mg, 0.77 mmol) in a 7.5 : 2 mixture of dry Me0H and dry THE (9.5 mL) was added a cold (-78 °C) 0.33 M solution of NaBH4 in dry Me0H (2.8 mL, 0.92 mmol). The resultant white suspension was stirred at -78 °C for 85 min and at -63 °C for 15 min. The reaction mixture was poured into brine (5 mL) and the resultant mixture was diluted with EtOAc (5 mL). The mixture was stirred for 10 min and most of the organic solvents were removed under reduced pressure. EtOAc (10 mL) was added to the residual mixture and the layers were separated. The aqueous layer was extracted with EtOAc (3 x 10 mL) and the combined organic extracts were washed with brine (10 mL), dried over anhydrous magnesium sulfate and the solvent was removed. Flash chromatography (silica gel, 7.1 g, 70 : 30 to 65 : 35 to 60 : 40 pentane-EtOAc ) of the crude material afforded, upon concentration of the appropriate fractions and removal of traces of solvent (vacuum pump) three products. The first eluted compound was the desired ketol epoxide 282 (195.4 mg, 80%), a white solid that could be recrystallized from etherpentane at -11 °C to yield white crystals, mp 146-147.5 °C. The next eluted substance was a white solid (slightly impure) that consisted largely of the diol 283 (13.4 mg). This compound was identified by its 1 H nmr and it spectra as described below. The third eluted fraction consisted of a mixture of 2 compounds (26.2 mg), the 1 H nmr spectrum of which showed two methoxy signals and no signals due to olefinic hydrogens.  214  The ketol epoxide 282 exhibited ir (KBr): 3474, 1737, 1667, 1461, 1426, 1375, 1300, 1251, 1173, 1108, 1044, 1020, 919, 895 cm -1 . 1H  nmr (400 MHz) 8: 0.85, 0.95 (d, 3H each, J = 6.5 Hz, Me-19, Me-20), 0.97 (s, 3H,  Me-16), 1.07 (ddd, 1H, J = 8.5, 8.5, 11 Hz), 1.12 (s, 3H, Me-17), 1.25 (dd, 1H, J = 13, 14 Hz, H-2), 1.41 (dddd, 1H, J = 6, 11.5, 11.5, 13 Hz), 1.53-1.63 (m, 2H, H-18 is one of the H's), 1.67-1.82 (m, 2H, includes a br d at 1.70, J = 11 Hz; exchanges upon treatment with D20), 1.88 (ddd, 1H, J = 6, 9.5, 14 Hz), 2.30 (br d, 1H, J = 17.5 Hz, H-10), 2.40 (dd, 1H, J = 7.5, 14 Hz, H-2'), 2.66 (d, 1H, J = 17.5 Hz, H-10'), 2.82 (ddm, 1H, J = 7.5, 13 Hz, H-3), 3.52 (br d, 1H, J = 1.5 Hz, H-6), 4.52 (br d, 1H, J = 11 Hz; upon exchange with D20, this signal becomes a br s, H-5), 5.17 (m, 1H, H-15), 5.30 (dd, 1H, J = 1.5, 1.5 Hz, H-15') Detailed 1 H nmr data, including those derived from NOE and irradiation experiments, are given in Table 12. 13 C nmr (75.3 MHz) 8 117 : 15.7 (-ve), 15.9 (-ve), 22.5 (-ve), 22.6 (-ve), 28.2, 30.7 (-ve),  32.6, 37.7, 45.1 (-ve), 49.9, 52.2, 53.9, 55.9 (-ve), 60.8 (-ve), 64.1, 72.8 (-ve), 114.9 (C=QH2), 152.1 (C=CH2), 217.4 (C=O). Anal. calcd for C20112803: C 75.91, H 8.92; found: C 75.83, H 9.03. Exact Mass calcd: 316.2038; found: 316.2037. The diol 283 exhibited ir (KBr): 3413, 1675, 1469, 1454, 1092, 1058, 1028, 895 cm -1 . 1H  nmr (400 MHz) 8: 0.87 (d, 3H, J = 6.5 Hz, Me-19), 0.99 (d, 3H, J = 6.5 Hz, Me-  20), 1.03, 1.04 (s, s, 3H each, two angular Me groups), 1.10-1.36 (m, 4H, also includes a dd at 1.16 (1H, J = 13, 13 Hz)), 1.46 (dd, 1H, J = 11.5, 13 Hz), 1.54-1.85 (m, 3H), 1.92 (unresolved d, 1H, OH; this signal exchanges upon treatment with D20), 2.21 (dd, 1H, J = 7.5, 13 Hz), 2.32 (dd, 1H, J = 6.5, 13 Hz), 2.97 (ddm, 1H, J = 7.5, 13 Hz, H-3), 215  3.15 (unresolved d, 1H, OH; this signal exchanges upon treatment with D20), 3.50 (d, 1H, J = 2.5 Hz, H-6), 3.97 (unresolved ddd, 1H, H-9; upon treatment with D20, this signal  becomes a dd with J = 6.5, 11.5 Hz), 4.57 (br s, 1H, H-5; upon exchange with D20, this signal sharpens), 5.14 (br s, 1H, H-15), 5.28 (dd, 1H, J = 2, 2 Hz, H-15').  216  Table 12: 1 H nmr Data (400 MHz, CDC13) for the Ketol Epoxide 282a: Decoupling and NOE Experiments. 15  19 14^ 13 12^ 1111 18  5  11111  7  ^  2o  He „^10 16  282  Signals Being Observed  Signal Being Irradiated Assignment H-x  1H nmr (400 MHz) 8 ppm (mult., J (Hz))  8 ppm (initial mult., J (Hz), H-x) to mult. after irradiation. J(Hz).  H-2 1.25 (dd, J = 13, 14) H_2 c ( c ) 2.40 (dd, J = 7.5, 14)  NOE Correiationsb  1.25 (dd, J = 13, 14, H-2) to d, J = 13 2.82 (ddm, J = 7.5, 13, H-3) to br d, J = 13 2.82 (ddm, J =7 .5, 13) 1.25 (dd, J = 13,^14, H-2) to d, H-3 J= 14 2.40 (dd, J = 7.5, 14, H-2') to d, J= 14 H-6, H-15' 3.52 (d, J = 1.5, H-6) to s 4.52 (br d, J = 11) H-5 1.70 (br d, J = 11 Hz, OH5) to s 1.70 (br d, J = 11 Hz) OH5 H-5, Me-16 3.52 (d, J = 1.5) H-6 H-10 (a) 2.30 (d, J = 17.5) 2.66 (d, J = 17.5, H-10') to s H-10' ((3) 2.66 (d, J = 17.5) 2.30 (d, J = 17.5, H-10) to s H-6 Me-16 0.97 (s) H-3, H-2', 1.12 (s) Me-17 H-10, H-18 H-18 part of the m at 1.53-1.63 0.85 (d, J = 6.5, Me-19) to s 0.95 (d, J = 6.5, Me-20) to s H-18, part of the m at 1.53-1.63 to 0.85 (d, J = 6.5) Me-19 sharpened m H-18, part of the m at 1.53-1.63 to Me-20 0.95 (d, J = 6.5) sharpened m ,  a- Crinipellin numbering used for consistency. b- Only those NOE correlations that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  217  Preparation of the Keto Epoxide 285.  To a cold (-78 °C) solution of the ketol epoxide 282 (197.5 mg, 0.624 mmol) in dry CH2C12 (6.5 mL) were added successively dry Et3N (2081AL, 1.49 mmol) and TBDMSOTf (288 111,, 1.25 mmol). The reaction mixture was stirred at -78 °C for 2 h and at 0 °C for 110 min. The solution was poured into sat. aq NaHCO3 (7 mL) and the resultant mixture was diluted with CH2C12 (20 mL). The layers were separated and the organic phase was washed with brine (7 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (6.2 g of silica gel, 97 : 3 to 92 : 8 pentane-EtOAc) of the crude material gave, upon concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), the keto epoxide 285 (258.2 mg, 96%) as a white solid. Recrystallization (ether, -11 °C, 3 crops, 236.5 mg, 88%) produced white crystals that exhibited mp 187.0-188.5 °C. Ir (KBr): 1730, 1674, 1473, 1413, 1372, 1251, 1119, 1023, 877, 837, 778 cm -1 . 1H  nmr (400 MHz) 8: 0.14 (s, 6H, SiMe7,But), 0.85 (d, 3H, J = 6.5 Hz, Me-19), 0.92 (s,  9H, SiMe2But), 0.95 (d, 3H, J = 6.5 Hz, Me-20), 0.97 (s, 3H, Me-16), 1.07 (ddd, 1H, J .8.5, 8.5, 11 Hz), 1.12 (s, 3H, Me-17), 1.26 (dd, 1H, J = 13.5, 13.5 Hz, H-2), 1.40 (dddd, 1H, J = 6, 11, 11, 13 Hz), 1.51-1.63 (m, 2H, includes H-18), 1.71-1.82 (m, 1H), 1.89 (ddd, 1H, J = 6, 9.5, 14 Hz), 2.30 (d, 1H, J = 17.5 Hz, H-10), 2.39 (dd, 1H, J = 7.5, 13.5 Hz, H-2'), 2.64 (d, 1H, J = 17.5 Hz, H-10'), 2.85 (br dd, 1H, J = 7.5, 13.5 Hz, H-3), 3.35 (d, 1H, J =2 Hz, H-6), 4.65 (ddd, 1H, J = 2, 2, 3.5 Hz, H-5), 5.13 (m, 1H, H-15), 5.17 (dd, 1H, J = 1, 1 Hz, H-15').  218  Anal. calcd for C26H42O3Si: C 72.51, H 9.83; found: C 72.79, H 10.00. Exact Mass calcd: 430.2903; found: 430.2901.  219  Preparation of the a-Hydroxy Ketone 286. 15  ^  13  286 To a cold (-78 °C) 0.32 M solution of HMDSK (2.5 mL, 0.814 mmol) in dry THE was added a solution of the keto epoxide 285 (233.7 mg, 0.543 mmol, dried overnight under reduced pressure (vacuum pump) at room temperature) in dry THE (4 mL). The colourless solution was stirred at -78 °C for 35 min 2-(Phenylsulfonyl)-3-phenyloxaziridine (274) (215.2 mg, 0.824 mmol) in dry THE (1 4 mL) was added to the solution which turned yellow, then orange. Stirring was continued for 45 min. The solution was poured into sat. aq NaHCO3 (7 mL) and the mixture was diluted with CH2C12 (15 mL). The phases were separated and the aqueous layer was extracted with CH2C12 (3 x 5 mL). The combined organic extracts were washed with brine (5 mL), dried over anhydrous magnesium sulfate and concentrated. The crude material was immediately purified by flash chromatography over iatrobeads (1St: 13.2 g, 96 : 4 to 90 : 10 benzene-EtOAc; 2nd: 7.4 g, same solvent mixtures) to afford 164.3 mg (68%) of the a-hydroxy ketone 286 as a white solid, along with 28.7 mg of the same compound contaminated with a UV active substance that was very difficult to separate. Recrystallization (ether-pentane) yielded white crystals, mp 190.5-192 °C (while melting, the compound turned yellow). Ir (KBr): 3487 (broad), 1738, 1672, 1474, 1372, 1254, 1125, 901, 880, 840, 778 cm -1 . 1H  nmr (400 MHz) 5: 0.14, 0.15 (s, 3H each, SiMe2But), 0.89 (d, 3H, J = 6.5 Hz, Me-  19), 0.93 (s, 9H, SiMe2But), 0.98 (s, 3H, Me-17), 1.00 (d, 3H, J = 6.5 Hz, Me-20), 1.03 (s, 3H, Me-16), 1.09 (dd, 1H, J = 12.5, 14 Hz, H-2), 1.27-1.51 (m, 2H, H-12, H-13), 220  1.55-1.71 (m, 2H, includes the m for H-18 at 1.60-1.71, and H-14), 1.77-1.94 (m, 2H, H13', H-14'), 2.42 (dd, 1H, J = 8, 14 Hz, H-2'), 2.49 (d, 1H, J = 3.5 Hz, OH; exchanges with D20), 3.03 (br dd, 1H, J = 8, 12.5 Hz, H-3), 3.30 (d, 1H, J = 2 Hz, H-6), 4.06 (d, 1H, J = 3.5 Hz, H-10; upon treatment with D20, this signal becomes a s), 4.66 (ddd, 1H, J = 2, 2, 3.5 Hz, H-5), 5.09 (m, 1H, H-15), 5.15 (m, 1H, H-15').  Detailed 1 H nmr data, including those derived from COSY and NOE experiments, are given in  Table 13. 13 C  nmr (75.3 MHz) 8: -4.6 , -4.3 (both -ve, SiMe2But), 10.8 (-ve), 17.5 (-ve), 18.2, 22.1  (-ve), 23.1 (-ve), 25.8 (-ve), 28.8, 29.3 (-ve), 35.9, 39.2, 44.8 (-ve), 50.8, 51.8, 57.2 (-ve), 59.0 (-ve), 62.1, 73.5 (-ve), 75.5, 80.9 (-ve), 113.9 (C=CH2), 150.7 (C=CH2), 217.7 (C=O). Anal. calcd for C26H42O4Si: C 69.91, H 9.48; found: C 69.67, H 9.47. Exact Mass calcd: 446.2853; found: 446.2844.  221  Table 13: 1 H nmr Data (400 MHz, CDC13) for the a-Hydroxy Ketone 286a.  286  Assignment H-x  1H nmr (400 MHz) 8 ppm (mult., J (Hz))  COSY Correlationsb  H-2 H-2'e (a) H-3  1.09 (dd, J = 12.5, 14) 2.42 (dd, J = 8, 14) 3.03 (dd, J= 8, 12.5)  H-2', H-3 H-2, H-3 H-2,^H-2',^H-5,^H-15, H-15' H-3, H-6, H-15, H-15' H-5, H-15 OH H-10 H-13, H-13', H-18  H-5 H-6 H-10 OH H-12 H-13 H-13' H-14 H-14' H-15 H-15' Me-16 Me-17 H-18 Me-19 Me-20  4.66 (ddd, J= 2, 2, 3.5) 3.30 (d, J= 2) 4.06 (d, J = 3.5) 2.49 (d, J = 3.5) —1.27-1.39 (m), part of the m (2H) at 1.27-1.51 —1.40-1.51 (m), part of the m (2H) at 1.27-1.51 —1.84-1.94 (m), part of the m (2H) at 1.77-1.94 Part of the m (2H) at 1.55-1.71 Part of the m (2H) at 1.77-1.94 5.09 (m) 5.15 (m) 1.03 (s)  H-12, H-13', H-14, H-14'  NOE Correlationsb  H-15  H-5, Me-16 OH, H-12 H-10, (OH neg)  H-12, H-13, H-14, H-14' H-13, H-13', H-14' H-13, H-13', H-14 H-3, H-5, H-6, H-15' H-3, H-5, H-15  H-3, H-15' H-5, H-15 H-6, H-10, H-14'  0.98 (s) 1.60-1.71 (m), part of the m H-12, Me-19, Me-20 (2H) at 1.55-1.71 H-18 0.89 (d, J = 6.5) H-18 1.00 (d, J = 6.5)  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  222  Preparation of the Triketone 281.  0  To a stirred solution of the a-hydroxy ketone 286 (125 mg, 0.28 mmol) in dry THF (5.6 mL) at room temperature was added a 1 M solution of tetra-n-butylammonium fluoride (560 p.L, 0.56 mmol) in THF. The resultant yellow solution was stirred for 75 min, poured into sat. aq NH4C1 (5 mL), diluted with EtOAc (7 mL) and the phases were separated. The aqueous phase was extracted with EtOAc (3 x 7 mL) and the combined organic extracts were washed with brine (5 mL), dried over anhydrous magnesium sulfate and the solvent was removed. Purification of the remaining material by flash chromatography on iatrobeads (3.1 g, 70 : 30 EtOAc-pentane) afforded a mixture of diols that was used directly in the next step. To a solution of the diol mixture and pyridine (45 ilL, 0.56 mmol) in dry CH2C12 (2.8 mL) at room temperature was added over a period of 35 min, a suspension of the periodinane 287 (421 mg, 1.12 mmol) in CH2C12 (6.5 mL) The resultant yellow-orange suspension was  stirred for another 20 min. Solid sodium thiosulfate (310 mg, 1.96 mmol) and sat. aq NaHCO3 (5 mL) were added and stirring was continued for 10 min. The mixture was diluted with EtOAc (15 mL) and the phases were separated. The aqueous phase was extracted with EtOAc (2 x 10 mL) and the combined organic extracts were washed with brine (10 mL), dried over anhydrous magnesium sulfate and concentrated. Flash chromatography on iatrobeads (1 st 2 8 g CH2C12; 2nd: 2.8 g, 75 : 10 : 15 pentane-CH2C12-EtOAc) of the crude material :  .  ,  gave 40.8 mg (44% from 286) of the triketone 281, a bright yellow-orange solid.  223  Recrystallization (CH2C12-pentane) yielded yellow-orange needles that exhibited mp 188-189.5 °C. Ir (KBr): 1732, 1641, 1476, 1455, 1383, 1262, 1247, 1135, 982, 960 cm -1 . 1H  nmr (400 MHz) 8: 0.72, 0.86 (d, d, 3H each, J = 6 Hz in each case, Me-19, Me-20),  1.21 (s, 3H, angular Me), 1.25 (s, 3H, angular Me), 1.40 (dd, 1H, J = 13, 14.5 Hz, H-2), 1.58-1.70 (m, 3H), 1.76-1.86 (m, 1H), 1.97-2.18 (m, 2H), 2.47 (dd, 1H, J = 7, 14.5 Hz, H-2'), 2.73 (ddm, 1H, J = 7, 13 Hz, H-3), 3.48 (br s, 1H, H-6), 5.43 (br s, 1H, H-15), 6.11 (d, 1H, J = 1.5 Hz, H-15'). 13 C  nmr (75.3 MHz) 8 117 : 9.0 (-ye), 15.1 (-ye), 21.1 (-ye), 22.6 (-ye), 29.1, 31.1 (-ye),  32.9, 37.6, 41.8 (-ye), 49.1, 54.6 (-ye), 57.3 (-ye), 59.9, 63.5, 123.8 (C=H2), 144.7 (C=CH2), 195.4 (C=0), 207.7 (C=0), 208.7 (=0).  Exact Mass calcd for C20112404: 328.1674; found: 328.1682.  224  Preparation of the Diol 291. 13  15  19 18 17  '  20  To a cold (-78 °C), stirred (yellow) solution of the triketone 281 (24.1 mg, 0.073 mmol) in a 3 : 1 mixture of dry THE-ether (2.9 mL) was added a 0.14 M solution of lithium (diisobutyl)(n-butyl)aluminum hydride in ether (2.4 mL, 0.34 mmol). The resultant colourless solution was stirred for 30 min, and two drops of sat. aq NH4C1 were added. After 5 min, the solution was warmed to room temperature, stirred for another 8 min, dried over anhydrous magnesium sulfate and filtered through iatrobeads (1 g, EtOAc as eluant). The solvent was removed under reduced pressure, while the water-bath was maintained below 40 °C. Flash chromatography (1 g of silica gel, 60 : 40 to 40 : 60 pentane-EtOAc) of the residual material provided, after concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), a mixture of diols (7.2 mg) in the first eluted fractions and the desired diol 291 (10 mg, 41%) in the later fractions. The diol could be recrystallized from ether-pentane to  yield white crystals. 118 Ir (KBr): 3446 (broad), 1730, 1670, 1453, 1101, 997, 978 cm -1 . 1H  nmr (400 MHz) 8: 0.79, 0.87 (d, d, 3H each, J = 6.5 Hz in each case, Me-19, Me-20),  1.06 (s, 3H, Me-17), 1.16 (dd, 1H, J = 13.5, 13.5 Hz, H-2), 1.30 (s, 3H, Me-16), 1.431.65 (m, 3H, H-13, H-14, H-18), 1.66-1.78 (m, 2H, includes the OH5 signal at 1.70 (d, J = 10.5 Hz) that exchanges upon treatment with D20 and H-12), 1.95-2.04 (m, 1H,  H-13'), 2.10 (dd, 1H, J =7 , 13.5 Hz, H-2'), 2.24 (ddd, 1H, J = 8.5, 8.5, 14 Hz, H-14'), 2.48 (br dd, 1H, J = 7, 13.5 Hz, H-3), 3.02 (d, 1H, J = 6 Hz, OH9, exchanges 225  upon treatment with D20), 3.39 (d, 1H, J = 2 Hz, H-6), 4.50 (br d, 1H, J = 10.5 Hz, H-5, upon exchange with D20, the signal becomes a s), 4.68 (d, 1H, J = 6 Hz, H-9 119 ), 5.09 (br s, 1H, H-15), 5.26 (br s, 1H, H-15'). Detailed 1 H nmr data, including those derived from COSY and NOE experiments, are given in Table 14. Anal. calcd for C20H2804: C 72.26, H 8.49; found: C 71.88, H 8.48. Exact Mass calcd: 332.1987; found: 332.1991.  226  Table 14: 1 H nmr Data (400 MHz, CDCI3) for the Diol 291a.  Assignment H-x  1H nmr (400 MHz) 8 ppm (mult., J (Hz))  COSY Correlationsb  H-2 H-2'c H-3 H-5 OH5 H-6 H-9  1.16 (dd, J = 13.5, 13.5) 2.10 (dd, J = 7, 13.5) 2.48 (br dd, J= 7, 13.5) 4.50 (br d, J= 10.5) 1.70 (br d, J = 10.5) 3.39 (d, J= 2) 4.68 (d, J = 6)  H-2', H-3 H-2, H-3 H-2, H-2', H-15, H-15' OH5, H-6, H-15, H-15' H-5 H-5, H-15 OH9  OH9 H-12d  3.02 (d, J = 6) Part of the m (2H) at 1.66-1.78  H-9 H-13, H-13', H-18  H-13 H-13 'd H-14 H-14'd H-15 H-15' Me-16  Part of the m (3H) at 1.43-1.65 1.95-2.04 (m) Part of the m (3H) at 1.43-1.65 2.24 (ddd, J= 8.5, 8.5, 14) 5.09 (br s) 5.26 (br s) 1.30 (s)  Me-17  1.06 (s)  H-18 Me-19 Me-20  Part of the m (3H) at 1.43-1.65 0.79 (d, J = 6.5) 0.87 (d, J = 6.5)  NOE Conelationsb  H-5, Me-16 OH9, H-12, Me-16 H-9, (OH9 neg)  H-12, H-13, H-14, H-14' H-13, H-13', H-14 H-3, H-5, H-6, H-15' H-3, H-5, H-15  H-6, H-9 H-14' H-2' (a), H-3, H-18  H-12, Me-19, Me-20 H-18 H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2). d- This hydrogen has been assigned by comparison with the 1 H nmr spectrum of crinipellin B (15).  227  Preparation of (±)-Crinipellin B (15). 15  13  19 18 17 20  To a solution of the diol 291 (9.7 mg, 0.029 mmol) in 2 : 1 dry CH2C12-DMSO (600 pL) were added dry triethylamine (53 tL, 0.38 mmol) and sulfur trioxide-pyridine complex (43.2 mg, 0.27 mmol). The resultant brown-yellow solution was stirred at room temperature for 9.5 h. It was diluted with CH2C12 (10 mL) and the resultant mixture was washed with water (3 mL) and brine (3 mL). The organic extracts were dried over anhydrous magnesium sulfate, and the solvent was removed under reduced pressure. Flash chromatography of the residual material on iatrobeads (3.6 g, 80 : 20 to 70 : 30 pentane-EtOAc to EtOAc) gave, after concentration of the appropriate fractions and removal of traces of solvent (vacuum pump) three substances. The least polar material was the triketone 281 (0.9 mg, 9%) while the most polar compound was the starting material, the diol 291 (1.2 mg, 12%). The middle fractions contained the (highly) desired (±)-crinipellin B (15) (4.7 mg, 49%), a white solid that could be recrystallized from ether-hexane to yield white crystals (mp 153.5-155 °C; the solid turned yellow while melting). Ir (KBr): 3482 (broad), 1730, 1642, 1473, 1380, 1256, 1104, 1011 cm -1 . 1 H nmr (400 MHz) 5: 0.81, 0.88 (d, d, 3H each, J = 6.5 Hz in each case, Me-19, Me-20),  1.11 (s, 3H, Me-17), 1.24 (dd, 1H, J 13, 14 Hz, H-2), 1.33 (s, 3H, Me-16), 1.45-1.68 (m, 3H, H-13, H-14, H-18), 1.76 (ddd, 1H, J =7 , 10, 11.5 Hz, H-12), 1.97-2.06 (m, 1H, H-13'), 2.20-2.30 (m, 2H, H-2', H-14'), 2.71 (ddm, 1H, J = 7, 13 Hz, H-3), 2.93 (d, 1H, J = 6.5 Hz, ow20) 3.31 (s, 1H, H-6), 4.75 (d, 1H, J = 6.5 Hz, H-9120), 5.37 (br s, ,  228  1H, H-15), 6.08 (d, 1H, J = 1.5 Hz, H-15'). The 1 H nmr spectrum of our synthetic (±)crinipellin B (15) was found to be identical with that of natural (-)-crinipellin B. 91 13 C  nmr (75.3 MHz) 8 117 : 10.3 (-ve, C-17), 21.4 (-ve, C-16, C-19), 22.7 (-ve, C-20), 29.9  (-ve, C-18), 30.1 (C-13), 33.9 (C-14), 38.9 (C-2), 42.5 (-ve, C-3), 43.4, 55.7 (-ve, C-6), 57.5, 60.7 (-ve, C-12), 63.6, 77.6, 79.8 (-ve, C-9), 122.8 (C-15), 145.1 (C-4), 196.8 (C-5), 217.5 (C-10). Detailed 1 H nmr and 13 C nmr data, including those derived from COSY, NOE, HMQC and HMBC experiments, are given in Tables 15 and 16. Exact Mass calcd for C20142604: 330.1831; found: 330.1827.  229  Table 15: 1 H nmr Data (400 MHz, CDC13) for (±)-Crinipellin B (15)a. 15  13  19 18  15  Assignment H-x  1H nmr (400 MHz) 8 ppm (mult., J (Hz))  COSY Correlationsb  H-3 H-6 H-9  1.24 (dd, J = 13, 14) Part of the m (2H) at 2.20-2.30 2.71 (ddm, J = 7, 13) 3.31 (s) 4.75 (d, J = 6.5)  H-2', H-3 H-2, H-3 H-2, H-2', H-15, H-15' H-15 OH  OH H-12  2.93 (d, J = 6.5) 1.76 (ddd, J = 7,10, 11.5)  H-9 H-13, H-13', H-18  H-13 H-13'  Part of the m (3H) at 1.45-1.68 1.97-2.06 (m)  H-14 H-14' H-15 H-15' Me-16  Part of the m (3H) at 1.45-1.68 Part of the m (2H) at 2.20-2.30 5.37 (br s) 6.08 (d, J = 1.5) 1.33 (s)  Me-17  1.11 (s)  H-18 Me-19 Me-20  Part of the m (3H) at 1.45-1.68 0.81 (d, J = 6.5) 0.88 (d, J = 6.5)  H-2 H-2'c (a)  NOE Correiationsb  OH, H-12, Me-16 H-9, (OH neg)  H-12, H-13, H-14, H-14'  H-3, H-6, H-15' H-3, H-15  H-6, H-9 H-14' H-2' (a), H-3, H-18  Me-19, Me-20 H-18 H-18  a- Crinipellin numbering used for consistency. b- Only those COSY correlations and NOE data that could be unambiguously assigned are recorded. c- H' indicates the hydrogen of a pair which is more downfield (H-2' is more downfield than H-2).  230  Table 16: 1 H nmr (500 MHz, CDC13) and 13 C nmr (125.8 MHz, CDC13) Data for (±)-Crinipellin B (15). 15  Carbon  numbera  13c nmr  2  spectrum (125.8 MHz) 8 ppm 38.9  3  42.5  4 5 6 9 10 12 13  145.1 196.8 55.7 79.8 217.5 60.7 30.1  14  33.9  15  122.8  16 17 18 19 20  21.4 10.3 29.9 21.4 22.7  13  19  umQcb,c 1H nmr Correlations (500 MHz) 5 ppm (assignment) 1.24 (H-2) Part of the m (2H) at 2.20-2.30 (H-2') 2.71 (H-3)  3.31 (H-6) 4.75 (H-9) 1.76 (H-12) Part of the m (3H) at 1.45-1.68 (H-13) 1.97-2.06 (H-13') Part of the m (3H) at 1.45-1.68 (H-14) Part of the m (2H) at 2.20-2.30 (H-14') 5.37 (H-15) 6.08 (H-15') 1.33 (Me-16) 1.11 (Me-17) Part of the m (3H) at 1.45-1.68 (H-18) 0.81 (Me-19) 0.88 (Me-20)  1H-13C HMBCb,c Long-range Correlations Hx -  H-2, H-2'd, Me-16 (4 bonds) H-2, H-6, H-15' H-6, H-15, H-15' Me-16 H-9, Me-17 Me-17, Me-19, Me-20 H-2  H-2 (4 bonds), H-9 H-2 (4 bonds) Me-20 Me-19  a- The quaternary carbon signals in the 13 C nmr spectrum of 15 have not been included in the table. b-The table reads from left to right. The assignment and the chemical shifts of the 13 C nmr spectrum are listed in the first and second columns, respectively. The third column shows the 1 H nmr signal(s) which correlate(s) with the carbon of the first two columns, as obtained from the HMQC experiment (1 bond correlation). The last column lists the hydrogen(s) which correlate(s) with the 13 C nmr signal of the first two columns as obtained from HMBC experiments (2, 3 and 4 bonds correlation (s)). c- Only those HMQC and HMBC data that could be unambiguously assigned are recorded. d- H' indicates the hydrogen of a pair which is more downfield (11-2' is more downfield than H-2).  231  Preparation of the Diketone Silyl Ether 292. 15  To a suspension of 4-methylmorpholine N-oxide (22.7 mg, 0.193 mmol) and 4 A molecular sieves in dry CH2C12 (1 mL), stirred for 20 min, was added a solution of the alcohol 286 (57.6 mg, 0.129 mmol) in dry CH2C12 (5.5 mL). The mixture was stirred at room  temperature for 15 min and solid tetra-n-propylammonium perruthenate (22.6 mg, 0.064 mmol) was added in one portion. The resultant black-green reaction mixture was stirred for 40 min. The solution was diluted with 85 : 15 pentane-EtOAc and filtered through iatrobeads (1.3 g, 85 : 15 pentane-EtOAc as eluant). The eluate was concentrated and the residual material was purified by flash chromatography (2.8 g of iatrobeads, 90 : 10 pentane-ether) to afford 42.5 mg (74%) of the diketone 292, a bright yellow solid. Attempts to recrystallize the solid failed since the diketone decomposed upon handling. Ir (KBr): 1742, 1474, 1374, 1255, 1125, 876, 841, 778 cm -1 . 1H  nmr (400 MHz) 5: 0.12 (s, 6H, SiMe2But), 0.70 (d, 3H, J = 5.5 Hz, Me-19), 0.84 (d,  3H, J = 5.5 Hz, Me-20), 0.91 (s, 9H, SiMe2But), 1.18 (s, 3H, angular Me), 1.20 (s, 3H, angular Me), 1.32 (dd, 1H, J = 13.5, 13.5 Hz, H-2), 1.50-1.67 (m, 3H), 1.72-1.82 (m, 1H), 1.95-2.16 (m, 2H), 2.29 (dd, 1H, J = 7, 13.5 Hz, H-2'), 2.50 (br dd, 1H, J = 7, 13.5 Hz, H-3), 3.38 (s, 1H, H-6), 4.62 (s, 1H, H-5), 5.07 (s, 1H, H-15), 5.15 (s, 1H, H-15'). Exact Mass calcd for C26H40O4Si: 444.2695; found: 444.2690.  232  Preparation of the Diketo Alcohol 290. 15  19 17 20  290 To a solution of the diketone 292 (30 mg, 0.067 mmol) in dry THF (1 mL) at room temperature was added a 1 M solution of tetra-n-butylammonium fluoride in THF (81 gL, 0.081 mmol). The yellow solution became brownish after the addition. Stirring was continued for 15 min. Brine (5 mL) was added and the mixture was diluted with CH2C12 (10 mL). The phases were separated and the aqueous layer was extracted with CH2C12 (10 mL) and EtOAc (10 mL). The combined organic phases were dried over anhydrous magnesium sulfate and concentrated. Flash chromatography (1.1 g of iatrobeads, 40 : 40 : 20 pentaneEtOAc-CH2C12) provided 15.2 mg (68%) of the diketo alcohol 290 as a yellow solid. Since this compound also showed signs of instability, it was not recrystallized. Ir (KBr): 3475, 1751, 1742, 1667, 1466, 1380, 1115, 921 cm -1 . 1H  nmr (400 MHz) 8: 0.71, 0.85 (d, d, 3H each, J = 6 Hz in each case, Me-19, Me-20),  1.18 (s, 3H, angular Me), 1.21 (s, 3H, angular Me), 1.31 (m, 1H), 1.50-1.69 (m, 4H, includes OH signal that exchanges upon treatment with D20), 1.72-1.82 (m, 1H), 1.94-2.15 (m, 2H), 2.30 (dd, 1H, J = 7, 14 Hz), 2.49 (br dd, 1H, J = 7, 13 Hz), 3.55 (s, 1H), 4.49 (br s, 1H, upon D20 exchange, this signal sharpens), 5.12 (br s, 1H), 5.29 (br s, 1H). Exact Mass calcd for C20H26O4: 330.1831; found: 330.1833.  233  Preparation of the Diol 291.  To a cold (-78 °C), stirred solution of the diketone 290 (15.2 mg, 0.046 mmol) in 3 : 1 ether-THF (2 mL) was added a 0.14 M solution of lithium (diisobutyl)(nbutyl)aluminum hydride in ether (755 pi, 0.106 mmol). The yellow solution turned colourless as the addition proceeded. It was stirred at -78 °C for 30 min, and a few drops of sat. aq NH4C1 were added to destroy the excess reducing agent. The mixture was warmed to room temperature and stirred for 10 min. The solution was dried over anhydrous magnesium sulfate and filtered through iatrobeads (1.4 g, elution first with EtOAc and then with 1 :1 Et0Ac-Me0H). The eluate was concentrated. Flash chromatography (1 g of silica gel, 60 : 40 to 40 : 60 pentane -EtOAc) of the crude material gave, after concentration of the appropriate fractions and removal of traces of solvent (vacuum pump), 3.7 mg of a mixture of diols and 8.1 mg (53%) of the more polar diol 291, identical with the same compound obtained from the reduction of the triketone 281. The 1 H nmr spectra of the two diols were indistinguishable.  234  REFERENCES AND NOTES 1. 2. 3.  For a discussion about the evolution and the future of chemistry see: Whitesides, G. M.  Angew. Chem. Int. Ed. Engl. 29, 1209 (1990). Also see 2.  Seebach, D. Angew. Chem. Mt. Ed. Engl. 29, 1320 (1990) and pertinent citations therein. Wani, M. C.;Taylor, H. L.; Wall, M. E.; Coggon, P.; McPhail, A. T. J. Am. Chem.  Soc. 93, 2325 (1971).  4.  Corey, E. J.; Cheng, X.-M. The Logic of Chemical Synthesis; Wiley: New York, 1989; pp 1-16.  5.  (a) Velluz, L.; Valls, J.; Mathieu, J. Angew. Chem. Mt. Ed. Engl. 6, 778 (1967). (b) For a discussion about analysis of molecular complexity and convergence, see: Bertz, S. H. J. Am. Chem. Soc. 104, 5801 (1982).  6.  (a) "The term conjunctive reagent is introduced to focus on those reagents which are simple building blocks that are incorporated in whole or in part into a more complex system and to differentiate them from reagents that operate on but are not normally incorporated into a substrate. Thus, methyl vinyl ketone would be a conjunctive reagent and chromic acid would be a simple reagent." See Trost, B. M. Acc. Chem. Res. 11, 453 (1978). (b) Trost, B. M.; Brandi, A. J. Org . Chem. 49, 4811 (1984) and references given therein.  7.  Seebach, D.; Knochel, P. Hely. Chim. Acta 67, 261 (1984).  8.  (a) Piers, E.; Karunaratne, V. J. Org . Chem. 48, 1774 (1983). (b) Piers, E.; Karunaratne, V. J. Chem. Soc., Chem. Commun. 935 (1983). (c) Piers, E.; Karunaratne, V. J. Chem. Soc., Chem. Commun. 959 (1984). (d) Piers, E.; Karunaratne, V. Can. J. Chem. 62, 629 (1984). (e) Piers, E.; Karunaratne, V. Can. J. Chem. 67, 160 (1989). (f) Piers, E.; Karunaratne, V. Tetrahedron 45, 1089 (1989).  9.  Piers, E.; Marais, P. C. J. Org . Chem. 55, 3454 (1990) and appropriate citations therein.  10.  For small scale reactions, see: (a) Piers, E.; Chong, J. M. J. Chem. Soc., Chem. Commun. 934 (1983).  235  (b) Piers, E.; Chong, J. M. Can. J. Chem. 66, 1425 (1988). For large scale reactions see 8f. 11. 12. 13.  Seebach, D. Angew. Chem. Mt. Ed. Engl. 18, 239 (1979). San Feliciano, A.; Miguel Del Corral, J. M.; Caballero, E.; Alvarez, A.; Medarde, M.  J. Nat. Prod. 49, 845 (1986).  Some of the recent references on the topic are listed below: Fraga, B. M. Nat. Prod. Rep. 1, 105 (1984); 2, 147 (1985); 3, 273 (1986); 4, 473 (1987); 5, 497 (1988); 7, 61 (1990); 7, 515 (1990) and pertinent citations therein.  14.  Some of the recent references on the topic are listed below: Faulkner, D. J. Nat. Prod. Rep. 1, 251 (1984); 1, 551 (1984); 3, 1 (1986); 4, 539 (1987); 5, 613 (1988); 7, 269 (1990); 8, 97 (1991); 9, 323 (1992) and pertinent citations therein.  15. 16. 17. 18.  Kaneda, M.; Takahashi, R.; Iitaka, Y.; Shibata, S. Tetrahedron Lett. 4609 (1972). Corbett, R. E.; Couldwell, C. M.; Lauren, D. R.; Weavers, R. T. J. Chem. Soc.,  Perkin Trans. 1 1791 (1979).  Anke, T.; Heim, J.; Knoch, F.; Mocek, U.; Steffan, B.; Steglich, W. Angew. Chem.  Mt. Ed. Engl. 24, 709 (1985).  (a) Bohlmann, F.; Jakupovic, J. Phytochemistry 19, 259 (1980). (b) Bohlmann, F.;Wallmeyer, M.; Jakupovic, J.; Ziesche, J. Phytochemistry 22, 1645 (1983).  19.  Bohlmann, F.; Suding, H.; Cuatrecasas, J.; Robinson, H.; King, R. M.; Phytochemistry 19, 2399 (1980).  20.  Bohlmann, F.; Zdero, C.; Bohlmann, R.; King, R. M.; Robinson, H. Phytochemistry 19, 579 (1980).  21.^(a) Groweiss, A.; Fenical, W.; Cun-heng, H.; Clardy, J.; Zhongde, W.; Zhongnian, Y.; Kanghou, L. Tetrahedron Lett. 26, 2379 (1985). (b) Jakupovic, J.; Zdero, C.; Paredes, L.; Bohlmann, F. Phytochemistry 27, 2881 (1988). (c) Coll, J.C.; Wright, A. D. Aust. J. Chem. 42, 1591 (1989). (d) Wright, A. D.; Coll, J. C. J. Nat. Prod. 53, 845 (1990). (e) Weyerstahl, P.; Marschall-Weyerstahl, H.; SchrOder, M.; Brendel, J.; Kaul, V. K.  Phytochemistry 30, 3349 (1991). The two main compounds 53 and 54 are shown.  A few other constituents possessing the silphiperfolane skeleton and isolated in trace amounts have not been included in the list. 236  ^ ^  22.^Some references taken from the numerous papers and reviews published to date are listed below: (a) Paquette, L. A. Top. Curr. Chem. 79, 41 (1979); 119, 1 (1984). (b) Ramaiah, M. Synthesis 529 (1984). (c) Paquette, L. A.; Doherty, A. M. Polyquinane Chemistry: Synthesis and Reactions (Reactivity and Structure: Concepts in Organic Chemistry, Vol. 26); Springer-Verlag: Berlin, 1987. (d) Hudlicky, T.; Sinai-Zingde, G.; Natchus, M. G.; Ranu, B. C.; Papadopolous, P. Tetrahedron 43, 5685 (1987). (e) Hudlicky, T.; Price, J. D. Chem. Rev. 89, 1467 (1989). (f) Hudlicky, T.; Fleming, A.; Radesca, L. J . Am. Chem. Soc. 111, 6691 (1989). 23.^See, for examples, as leading references: (a) Paquette, L. A.; Roberts, R. A.; Drtina, G. J. J. Am. Chem. Soc. 106, 6690 (1984). (b) Wender, P. A.; Singh, S. K. Tetrahedron Lett. 26, 5987 (1985). (c) Curran, D. P.; Kuo, S.-C. J. Am. Chem. Soc. 108, 1106 (1986). (d) Curran, D. P.; Kuo, S.-C. Tetrahedron 43, 5653 (1987). (e) Meyers, A. I.; Lefker, B. A. Tetrahedron 43, 5663 (1987). (f) Iwata, C.; Takemoto, Y.; Doi, M.; Imanishi, T. J. Org . Chem. 53, 1623 (1988). (g) Demuth, M.; Hinsken, W. Hely. Chim. Acta 71, 569 (1988). (h) Kakiuchi, K.; Ue, M.; Tsukahara, H.; Shimizu, T.; Miyao, T.; Tobe, Y.; Odaira, Y.; Yasuda, M.; Shima, K. J. Am. Chem. Soc. 111, 3707 (1989). (i) Brendel, J.; Weyerstahl, P. Tetrahedron Lett. 30, 2371 (1989). (j) Dickson, J. K., Jr.; Fraser-Reid, B. J. Chem. Soc., Chem. Commun. 1440 (1990). (k) Wender, P. A.; deLong, M. A. Tetrahedron Lett. 31, 5429 (1990). (1) Weyerstahl, P.; Brendel, J. Liebigs Ann. Chem. 7, 669 (1992). 24.^(a) Marfat, A.; Helquist, P. Tetrahedron Lett. 4217 (1978) and pertinent citations therein. (b) Bal, S. W.; Marfat, A.; Helquist, P. J. Org. Chem. 47, 5045 (1982) and pertinent citations therein. 237  25.  Brattesani, D. N.; Heathcock, C. H. J. Org. Chem. 40, 2165 (1975).  26.  Tsunoda, T.; Kodama, M.; Ito, S. Tetrahedron Lett. 24, 83 (1983).  27.  Oppolzer, W.; Marazza, F. HeIv. Chim. Acta 64, 1575 (1981).  28.  (a) Stowell, J. C.; J. Org. Chem. 41, 560 (1976). (b) Stowell, J. C. Chem. Rev. 84, 409 (1984).  29.^(a) Paquette, L. A.; Leone-Bay, A. J. Am. Chem. Soc. 105, 7352 (1983) and appropriate references therein. (b) Paquette, L. A.; Leone-Bay, A. J. Org . Chem. 47, 4173 (1982). 30.^Koreeda, M.; Mislankar, S. G. J. Am. Chem. Soc. 105, 7203 (1983). 31.  Horiguchi, Y.; Matsuzawa, S.; Nakamura, E.; Kuwajima, I. Tetrahedron Lett. 27, 4025, 4029 (1986).  32.  Still, W. C.; Kahn, M.; Mitra, A. J. Org . Chem. 43, 2923 (1978).  33.  Agosta, W. C.; Wolff, S. J. Org. Chem. 40, 1699 (1975).  34.  For leading references concerning the hydroboration of alkenes see: (a) Brown, H. C. Boranes in Organic Chemistry; Cornell Univ. Press: Ithaca, 1972; pp 255-297 and appropriate citations therein. (b) Brown, H. C. Organic Synthesis via Boranes; Wiley: New York, 1975 and appropriate citations therein. (c) Brown, H. C.; Zweifel, G. J. Am. Chem. Soc. 82, 4708 (1960).  35.^(a) Omura, K.; Swern, D. Tetrahedron 34, 1651 (1978). (b) Mancuso, A. J.; Huang, S.-L.; Swern, D. J. Org . Chem. 43, 2480 (1978). (c) For a recent review on "oxidation of alcohols by activated dimethyl sulfoxide" see: Tidwell, T. T. Synthesis 857 (1990). 36.^Hatch, R. P.; Shringarpure, J.; Weinreb, S. M. J. Org . Chem. 43, 4172 (1978) and appropriate citations therein. 37. 38.  For the preparation of the Raney nickel catalyst used, see: Pavlic, A. A.; Adkins, H. J.  Am. Chem. Soc. 48, 1471 (1946).  Reich, H. J.; Renga, J. M.; Reich, I. L. J. Am. Chem. Soc. 97, 5434 (1975) and pertinent citations therein.  39.^(a) Wittig, G.; Schoellkopf, U. Org. Syn. 40, 66 (1960). 238  (b) Greenwald, R.; Chaykovsky, M.; Corey, E. J. J. Org . Chem. 28, 1128 (1963). 40.^To consult the original references, see: (a) Takai, Y.; Hotta, Y.; Oshima, K.; Nozaki, H. Tetrahedron Lett. 2417 (1978). (b) Takai, Y.; Hotta, Y.; Oshima, K.; Nozaki, H. Bull. Chem. Soc. Jpn. 53, 1698 (1980). For modifications to the procedure, see: (c) Lombardo, L. Tetrahedron Lett. 23, 4293 (1982). (d) Lombardo, L. Org . Synth. 65, 81 (1987). (e) Hart, T. W.; Comte, M.-T. Tetrahedron Lett. 26, 2713 (1985). 41.^Diazomethane was prepared from N-methyl-N-nitroso-p-toluenesulfonamide (Diazald ®). See Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis ; Wiley: New York, 1967; Vol. 1, pp 191-195 and pertinent citations therein. 42.  Diazomethane should be made and used in a well-ventilated fumehood. Special glassware is needed for the preparation of this reagent. Handling of the reagent behind a safety screen is also recommended.  43.  Ito, Y.; Hirao, T.; Saegusa, T. J. Org . Chem. 43, 1011 (1978).  44.  House, H. O.; Czuba, L. J.; Gall, M.; Olmstead, H. D. J. Org . Chem. 34, 2324 (1969).  45.  (a) Corey, E. J.; Suggs, J. W. Tetrahedron Lett. 2647 (1975). (b) Dauben, W. G.; Michno, D. M. J. Org. Chem. 42, 682 (1977) and appropriate citations therein. (c) Sundararaman, P.; Herz, W. J. Org . Chem. 42, 813 (1977).  46.  Mander, L. N.; Sethi, S. P. Tetrahedron Lett. 24, 5425 (1983).  47.  We are grateful to Professor San Feliciano for copies of the 1 H nmr and 13 C nmr spectra of (-)-methyl cantabrenonate (13) and for a sample of (-)-methyl epoxycantabronate (14).  48.  Kupka, J.; Anke, T.; Oberwinkler, F.; Schramm, G.; Steglich, W. J. Antibiot. 32, 130 (1979).  49.  (a) Fujita, E.; Nagao, Y. Bioorg. Chem. 6, 287 (1977). (b) Anke, T.; Watson, W. H.; Giannetti, B. M.; Steglich, W. J. Antibiot. 34 1271 (1981).  239  50.  Mehta, G.; Rao, K. S.; Reddy, M. S. Tetrahedron Lett. 29, 5025 (1988) and pertinent references therein.  51.  Mehta, G.; Rao, K. S.; Reddy, M. S. J. Chem. Soc., Perkin Trans. 1 693 (1991) and pertinent references therein.  52.  Schwartz, C. E.; Curran, D. P. J. Am. Chem. Soc. 112, 9272 (1990).  53.  (a) Stork, G.; Hudrlik, P. F. J. Am. Chem. Soc. 90, 4462, 4464 (1968). (b) House, H. O.; Gall, M.; Olmstead, H. D. J. Org . Chem. 36, 2361 (1971). (c) d'Angelo, J. Tetrahedron 32, 2979 (1976).  54.  Paquette, L. A.; Galemno Jr., R. A.; Caille, J.-C.; Valpey, R. S. J. Org . Chem. 51, 686 (1986) and pertinent references therein.  55.  (a) Negishi, E.; Luo, F.-T. J. Org . Chem. 48, 2427 (1983). (b) Negishi, E.; John, R. A. J. Org . Chem. 48, 4098 (1983). (c) Welch, S. C.; Assercq, J.-M.; Loh, J.-P.; Glase, S. A. J. Org . Chem. 52, 1440 (1987).  56.  Pappas, J. J.; Keaveney, W. P. Tetrahedron Lett. 4273 (1966).  57.  See Marais, P. Ph. D. thesis, 1990, University of British Columbia, Vancouver, B.C., Canada.  58.  Oda, H.; Morizawa, Y.; Oshima, K.; Nozaki, H. Tetrahedron Lett. 25, 3221 (1984).  59.  (a) Yamamoto, Y. Angew. Chem. Int. Ed. Engl. 25, 947 (1986). (b) Lipshutz, B. H.; Ellsworth, E. L.; Siahaan, T. J. Am. Chem. Soc. 111, 1351 (1989).  60.  (a) Wu, T.-C.; Xiong, H.; Rieke, R. D. J. Org . Chem. 55, 5045 (1990). (b) Piers, E.; Roberge, J. Y. Tetrahedron Lett. 32, 5219 (1991).  61.  Kovacs, G.; Galambos, G.; Juvancz, Z. Synthesis 171 (1977).  62.  (a) Corey, E. J.; Cho, H.; Rucker, C.; Hua, D. H. Tetrahedron Lett. 22, 3455 (1981). (b) Mander, L. N.; Sethi, S. P. Tetrahedron Lett. 25, 5953 (1984).  63.  (a) Baran, J. S. J. Org . Chem. 25, 257 (1960). (b) For a review on cis hydroxylation with 0s04, see: SchrOder, M. Chem. Rev. 80, 187 (1980). 240  ^ ^  64.^(a) Piers, E.; Abeysekera, B.; Scheffer, J. R. Tetrahedron Lett. 3279 (1979). (b) Piers, E.; Abeysekera, B. Can. J. Chem. 60, 1114 (1982) and pertinent references therein. 65.^(a) The potential use of (3,e-diketo phosphonates for the formation of cyclopentenones had been investigated by Heathcock and coworkers; see: Clark, R. D.; Kozar, L. G.; Heathcock, C. H. Synth. Commun. 5, 1 (1975). Some references in which the intramolecular Wadsworth-Emmons reaction is employed are listed below: (b) Aristoff, P. A. J. Org . Chem. 46, 1954 (1981). (c) Aristoff, P. A. Synth. Commun. 13, 145 (1983). (d) Begley, M. J.; Cooper, K.; Pattenden, G. Tetrahedron Lett. 22, 257 (1981). (e) Begley, M. J.; Cooper, K.; Pattenden, G. Tetrahedron 37, 4503 (1981). (f) Dauben, W. G.; Walker, D. M. Tetrahedron Lett. 23, 711 (1982). (g) Liu, H.-J.; Llinas-Brunet, M. Can. J. Chem. 66, 528 (1988). 66.^See for example: (a) Miyano, M.; Dorn, C. R. J. Org . Chem. 37, 268 (1972). (b) Brown, E.; Ragault, M. Tetrahedron Lett. 1927 (1973). (c) Evans, D. A.; Sims, C. L.; Andrews, G. C. J. Am. Chem. Soc. 99, 5453 (1977). 67.^The preparation of the mesylate 237 from the corresponding alcohol is given in reference 68a. The iodide 238 is obtained from the mesylate 237 as seen in reference 71. For the description of the synthesis of the alcohol see: Trost, B. M.; Chan, D. M.T.; Nanninga T. N. Org . Synth. 62, 58 (1984). 68.  (a) Trost, B. M.; Vincent, J. E. J. Am. Chem. Soc. 102, 5680 (1980). (b) Trost, B. M.; Curran, D. P. J. Am. Chem. Soc. 103, 7380 (1981).  69.  An intramolecular example of a reaction of an allylsilane and a carbonyl group had been reported previously by Anderson and Sarkar. See Sarkar, T. K.; Anderson, N. H. Tetrahedron Lett. 3513 (1978).  70.  Detailed procedures for alkylation of various substrates with Trost's reagent are given in a paper by Majetich and coworkers. These procedures have inspired this work. See Majetich, G.; Desmond, R. W., Jr.; Soria, J. J. J. Org. Chem. 51, 1753 (1986).  71.^Trost, B. M.; Curran, D. P. Tetrahedron Lett. 22, 5023 (1981).  241  72.  Piers, E.; Marais, P. C. Tetrahedron Lett. 29, 4053 (1988) and appropriate citations therein.  73.  Examples of cyclization of this type involving the use of cuprate reagents were previously reported by Corey and coworkers. See: (a) Corey, E. J.; Kuwajima, I. J. Am. Chem. Soc. 92, 395 (1970). (b) Corey, E. J.; Narisada, M.; Hiraoka, T.; Ellison, R. A. J. Am. Chem. Soc. 92, 396 (1970). Since the beginning of this work, a few reports have appeared on the use of intramolecular addition of carbon side chains to ketone functions to form new rings. See for example: (c) Mori, M.; Watanabe, N.; Kaneta, N.; Shibasaki, M. Chem. Lett. 1615 (1991). (d) Barbero, A.; Cuadrado, P.; Gonzalez, A. M.; Pulido, F. J.; Rubio, R.; Fleming, I.  Tetrahedron Lett. 33, 5841 (1992).  74.^(a) Cowell, A.; Stille, J. K. J. Am. Chem. Soc. 102, 4193 (1980) and pertinent citations therein. (b) Feldman, K. S. Tetrahedron Lett. 23, 3031 (1982). (c) Jung, M. E.; Light, L. A. Tetrahedron Lett. 23, 3851 (1982). (d) Sato, Y.; Honda, T. Tetrahedron Lett. 33, 2593 (1992). 75.^Some of the references published concerning the syntheses of methyl (Z)-3-iodopropenoate and other (Z)-3-halopropenoates are listed below. See: (a) Moss, R. A.; Wilk, B.; Krogh-Jespersen, K.; Westbrook, J. D. J. Am. Chem.  Soc. 111, 6729 (1989).  (b) Ma, S.; Lu, X. J. Chem. Soc., Chem. Commun. 1643 (1990). (c) Ma, S.; Lu, X. Tetrahedron Lett. 31, 7653 (1990). (d) Marek, I.; Alexakis, A.; Normant, J.-F. Tetrahedron Lett. 32, 5329 (1991) and appropriate citations therein. (e) Ma, S.; Lu, X.; Li, Z. J. Org . Chem. 57, 709 (1992). 76.  (a) Wiley, G. A.; Hershkowitz, R. L.; Rein, B. M.; Chung, B. C. J. Am. Chem. Soc. 86, 964 (1964). (b) Schaefer, J. P.; Higgins, J. J. Org . Chem. 32, 1607 (1967).  77.  (a) Majetich, G.; Condon, S.; Hull, K.; Ahmad, S. Tetrahedron Lett. 30, 1033 (1989) and appropriate citations therein. 242  (b) Liotta, D.; Brown, D.; Hoekstra, W.; Monahan III, R. Tetrahedron Lett. 28, 1069 (1987) and pertinent references therein. 78.^See for example a, b, c, d and pertinent citations therein: (a) Glotter, E.; Greenfield, S.; Lavie, D. J. Chem. Soc. C 1646 (1968). (b) Biichi, G.; Egger, B. J. Org. Chem. 36, 2021 (1971). (c) McCurry, P. M. Jr.; Singh, R. K. J. Org . Chem. 39, 2317 (1974). (d) Glotter, E.; Rabinsohn, Y.; Ozari, Y. J. Chem. Soc., Perkin Trans. I 2104 (1975). 79.^The formation of products containing an epoxide in the rearrangement step is not surprising. A number of such cases have been reported. See 29a, 45b-c and citations therein. 80. 81.  Schreiber, J.; Maag, H.; Hashimoto, N.; Eschenmoser, A. Angew. Chem. Int. Ed.  Engl. 10, 330 (1971)  See for example: (a) Danishefsky, S.; Schuda, P. F.; Kitahara, T.; Etheredge, S. J. J. Am. Chem. Soc. 99, 6066 (1977) and citations therein. (b) Roberts, J. L.; Borromeo, P. S.; Poulter, C. D. Tetrahedron Lett. 1621 (1977).  82.  Davis, F. A.; Vishwakarma, L. C.; Billmers, J. M.; Finn, J. J. Org. Chem. 49, 3241 (1984).  83.  (a) Davis, F. A.; Lamendola, J., Jr.; Nadir, U.; Kluger, E. W.; Sedergran, T. C.; Panunto, T. W.; Billmers, R.; Jenkins, R., Jr.; Turchi, I. J.; Watson, W. H.; Chen, J. S.; Kimura, M. J. Am. Chem. Soc. 102, 2000 (1980). (b) Davis, F. A.; Stringer, O. D. J. Org . Chem. 47, 1774 (1982).  84.  Brown, C. A. J. Org . Chem. 39, 3913 (1974).  85.  Corey, E. J.; Gross, A. W. Tetrahedron Lett. 25, 495 (1984).  86.  (a) Griffith, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D. J. Chem. Soc.,  Chem. Commun. 1625 (1987).  (b) Griffith, W. P.; Ley, S. V. Aldrichim. Acta 23, 13 (1990). 87.  Hashimoto, H.; Tsuzuki, K.; Sakan, F.; Shirahama, H.; Matsumoto, T. Tetrahedron  Lett. 3745 (1974).  88.^Dess, D. B.; Martin, J. C. J. Org . Chem. 48, 4155 (1983).  243  89.  We are grateful to Professor W. Steglich for sending us a copy of the 1 H nmr spectrum of natural (-)-dihydrocrinipellin B.  90.  (a) Parikh, J. R.; Doering, W. von E. J. Am. Chem. Soc. 89, 5505 (1967). (b) Nicolaou, K. C.; Duggan, M. E.; Hwang, C.-K. J. Am. Chem. Soc. 111, 6666, 6676, 6682 (1989).  91.  We are grateful to Professor W. Steglich for sending us a copy of the 1 H nmr spectrum of natural (-)-crinipellin B.  92.  Cooper, J. W. Spectroscopic Techniques for Organic Chemist; Wiley: New York, 1980; pp 94-96.  93.  Silverstein, R. M.; Bassler, G. C.; Morrill, T. C.; Spectrometric Identification of Organic Compounds; Fourth Ed.; Wiley: New York, 1981.  94.  Bryan, W. P.; Byrne, R. H. J. Chem. Ed. 47, 361 (1970).  95.  Perrin, D. D.; Armarego, W. L. F.; Perrin, D. R. Purification of Laboratory Chemicals; Pergamon: Oxford, 1980.  96.  CCLt and benzene are cancer suspect agents. They should be handled with care, in a fumehood.  97.  HMPA is highly toxic and is a cancer suspect agent. It should be handled with care, in a well-ventilated fumehood. Gloves should be worn when handling this substance. Solutions containing HMPA should be sent for disposal.  98.  Pyridine is a toxic chemical. Solutions containing this substance should be sent for disposal.  99.  Kofron, W. G.; Baclawski, L. M. J. Org. Chem. 41, 1879 (1976).  100.  Suffert, J. J. Org . Chem. 54, 509 (1989).  101.  See reference 34b, pp 241-246.  102.  Wuts, P. G. M. Synth. Commun. 11, 139 (1981).  103.  This substance has been listed as a carcinogen and should be handled with care.  104.  CuCN is poisonous. A dust mask and two pairs of disposable gloves should be worn when handling this compound. Solutions containing this compound or derivatives of it should be sent for disposal.  105. Raney nickel is a cancer suspect agent and should be handled with great care. Gloves should be worn when handling this substance. Raney nickel is also potentially pyrophoric as a dry solid. It should therefore be kept in solution. During filtration, this solid should never be allowed to "dry" in the filtration funnel.  244  106.  Raney nickel is paramagnetic.  107. PCC is a cancer suspect agent. Gloves should be worn when handling this substance. 108.  Methyl cyanoformate is highly toxic and should be handled in a fumehood. Solutions containing this compound should be sent for disposal.  109. Phenylselenenyl bromide is highly toxic. Special care should be taken when handling this reagent. Solutions containing this compound should be sent for disposal. 110.  Takano, S.; Hirama, M.; Araki, T.; Ogasawara, K. J. Am. Chem. Soc. 98, 7084 (1976).  111.  Green, M. B.; Hickinbottom, W. J. J. Chem. Soc. 3262 (1957).  112.  Osmium tetroxide is highly toxic. It produces poisonous vapors. This substance is an irritant that can cause damages to the eyes, the respiratory system and the skin. Special care should be taken when opening containers of 0s04 and handling this product; two pairs of gloves should be worn and the reagent should be handled in the fumehood at all times. Solutions containing 0s04 should be sent for disposal.  113.  Lead tetraacetate is a cancer suspect agent. Special care should be taken when handling this reagent; two pairs of gloves should be worn and the compound should be handled in the fumehood at all times  114. For the preparation of methyl (Z)-3-iodopropenoate (262), see references 75d and 75e. 115.  The use of a non-aqueous workup procedure for the preparation of (Z)-3-iodo-2propen-l-ol (263) was suggested and first used by Dr. C. Rogers in our laboratories. This modification resulted in a substantial increase of the yield for the conversion of 262 into 263.  116. Apparently, carbon signals are overlapping. 117.  Two carbon signals are overlapping.  118.  The melting point was difficult to determine; at —180°C, the solid turned pale yellow and progressively darkened from orange to red. At —219-225°C, a thick reddish liquid formed. The temperature at which the orange-red solid turned into the thick liquid was not exactly reproducible.  119. After D20 exchange, the signal should be a s; however, the multiplicity of this signal could not be ascertained since the s is hidden under the DOH peak, after exchange. 120. In most nmr spectra, the H-9 and OH signals appeared as two d. However, in the sample sent for D20 exchange, the signals were s; the OH signal was part of the m (4H) at 1.45-1.68 and exchanged upon treatment with D20.  245  Appendix 1: X-Ray Crystallographic Data for the Enedione Epoxide 188.  188  0  Formula^  C201-12603  Crystal System^  Monoclinic  Space Group^  P2 1/n  Lattice Parameters^  a (A) = 7.244 (2) b (A) = 20.628 (3) c (A) = 12.176 (2) 0 (°) = 106.42 (2) V(A 3 ) = 1745.0 (7) 4  Z value^  Number of reflections used in refinement^2625 R^  0.038  Rw^0.040  246  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061795/manifest

Comment

Related Items